Byzantine Navy 500-1204

THE AGE OF THE DROMWN
THE
MEDIEVAL MEDITERRANEAN
PEOPLES, ECONOMIES AND CULTURES, 400-1500
EDITORS
Hugh Kennedy (St. Andrews)

Paul Magdalino (St. Andrews)
David Abulafia (Cambridge)
Benjamin Arbel (Tel Aviv)
Mark Meyerson (Toronto)
Larry J. Simon (Western Michigan University)
VOLUME 62
THE AGE OF THE DROMWN
The Byzantine Navy ca 500-1204
BY
JOHN H. PRYOR
AND
ELIZABETH M. JEFFREYS
WITH AN APPENDIX
TRANSLATED FROM THE ARABIC OF
MU\AMMAD IBN MANKALI
BY
AHMAD SHBOUL
BRILL
LEIDEN • BOSTON
2006
On the cover: ‘Dromon using Greek Fire’ from John Skylitzes’ Matritensis Græcus (formerly
Codex Græcus Matritensis Ioannis Skyllitzes, 5-3 N-2). © Biblioteca Nacional, Madrid, vitr. 26-2,
fol. 34v.
Despite our efforts we have not been able to trace all rights holders to some copyrighted material.
The publisher welcomes communications from copyrights holders, so that the appropriate acknowledgements
can be made in future editions, and to settle other permission matters.
This book is printed on acid-free paper.
Library of Congress Cataloging-in-Publication Data
A C.I.P. record for this book is available from the Library of Congress.
ISSN 0928–5520
ISBN-13: 978-90-04-15197-0
ISBN-10: 90-04-15197-4
© Copyright 2006 by Koninklijke Brill NV, Leiden, The Netherlands

Koninklijke Brill NV incorporates the imprints Brill Academic Publishers,
Martinus Nijhoff Publishers and VSP.
All rights reserved. No part of this publication may be reproduced, translated, stored in
a retrieval system, or transmitted in any form or by any means, electronic,
mechanical, photocopying, recording or otherwise, without prior written
permission from the publisher.
Authorization to photocopy items for internal or personal

use is granted by Brill provided that
the appropriate fees are paid directly to The Copyright
Clearance Center, 222 Rosewood Drive, Suite 910
Danvers, MA 01923, USA.
Fees are subject to change.
printed in the netherlands

To my children, Shefali and Sean
John Pryor
To my daughter, Katharine
Elizabeth Jeffreys
CONTENTS
Preface and appreciations………………………………...…….. xi
List of figures, maps, and tables ……………………………...... xiii
Note on orthography, translation, citation, and dating ………… xix
Note on metrology and hours of daylight ……………………… xxiii
Selective gazeteer of historical place names …………………... xxv
Selective glossary of Greek, Latin, and Arabic terminology ….. xxxix
Glossary of English nautical terminology ……………………... xlix
Note on citation of Greek and Latin glossaries ………………... lix
Abbreviations ………………………………………………….. lxiii
Maps …………………………………………………………… lxv
Introduction …………………………………..………………... 1
Chapter One: the operational context ………………………….. 7

The first period, ca 400-560: the Germanic assault and
imperial recovery …………………………………….. 7
The second period, ca 560-750: the Muslim assault and
imperial recovery …………………………………….. 19
The third period, ca 750-875: equilibrium of chaos …….. 34
The fourth period, ca 875-1025: Byzantine ascendancy ... 50
The fifth period, ca 1025-1204: the triumph of the Latin
West ………………………………………………….. 76
Chapter Two: the origins of the Dromon …………………….... 123

First mentions …………………………………….……... 123
Deck and oarage system ………………………………… 128
The ram and the spur ……………………………………. 134
Square sails and lateen sails …………………………….. 153
viii CONTENTS
Chapter Three: from the sixth to ninth centuries ………………. 163
Chapter Four: the Dromon in the age of the Macedonian

emperors ……………………………………………………. 175
(a) The sources ………………………………………. 175
(b) Terminology and ship types ……………………… 188
(c) Hull ………………………………………………. 192
(d) Prow ……………………………………………… 203
(e) Poop ……………………………………………… 215
(f) Deck and castles ………………………………….. 227
(g) Masts, yards, and sails …………………………… 238
(h) Rigging …………………………………………… 252
(i) Crews …………………………………………….. 254
(j) Oarage system and dimensions …………………... 276
(k) Horse transports ………………………………….. 304
(l) Performance capabilities, water supplies, and
logistics ………………………………………... 333
(m) Armaments ……………………………………….. 378
(n) Tactics, strategy, and techniques ………………… 382
Chapter Five: The demise of the Dromon ……………………... 407
Chapter Six: The triumph of the Galea ………………………... 423
Conclusion ……………………………………………………... 445

Postscript ………………………………………………... 453
Appendix One: Syrianos Magistros, Naumacivai Surianou'

Magivstrou, edition and translation ………………………… 455
Appendix Two: Leo VI, Naumacika; Levonto" Basilevw", edition

and translation ………………………………………………. 483
Appendix Three: Anonymous, Naumacika; suntacqevnta para;

Basileivou patrikivou kai; parakoimoumevnou, edition and
translation …………………………………………………... 521
Appendix Four: fleets, armaments, and equipment for dromons,

pamphyloi, and ousiaka chelandia according to the invento-
ries for the expeditions to Crete of 911 and 949 in the De
cerimoniis aulae Byzantinae attributed to Constantine VII
Porphyrogenne2tos …………………………………………... 547
CONTENTS ix
Appendix Five: Nike2phoros Ouranos, Peri; qalassomaciva",

edition and translation ……………………………………… 571
Appendix Six: Greek Fire ……………………………………... 607
Appendix Seven: The galleys of the manuscript, Madrid,

Biblioteca National, vitr. 26-2, of John Skylitze2s’ Synopsis
historio2n and its dating and art styles ………………………. 633
Appendix Eight: Muh5ammad ibn Mankalı3, [a] Al-adilla al-

rasmiyya fı3 ’l-ta‘a2bı3 al-h5arbiyya [b] Al-ah5ka2m al-mulu2kiyya
wa ’l-d5a wa2bit al-na2mu2siyya, translation …………...………. 645
Bibliography and abbreviations …………….....………………. 667
Indices …………………………………………………………. 711

[a] General Index …………………………………………… 711
[b] Technical Index …………………………………………. 738
Acknowledgments …………………………………………....... 751

PREFACE AND APPRECIATIONS
We have shared a happy collaboration in this book. Our friendship

even survived the translation of the texts for Appendices One to Five,
where we were concerned to achieve a delicate balance between
producing a readable English translation for Byzantinists and the
general public and preserving the literal meaning of technical
language as closely as possible for maritime and military historians.
John Pryor has been responsible for the historical research and for the
writing of the text. Elizabeth Jeffreys has been responsible for the
editing of the texts in Appendices One to Five, for the translation of
Greek texts, for matters philological, and for the interpretation of the
milieux of the Byzantine sources. We are indebted to Ahmad Shboul
for his collaboration in Appendix Eight.
An earlier version of some parts of the book was published by John
Pryor as “From dromo2n to galea: Mediterranean bireme galleys AD
500-1300”, in J. Morrison, ed., The age of the galley: Mediterranean
oared vessels since pre-classical times (London, Conway Maritime
Press, 1995), 101-16. A much shorter version of the first four parts of
Chapter One was published by John Pryor as “The Mediterranean
breaks up: 500-1000”, in D. Abulafia, ed., The Mediterranean in
History (London, Thames & Hudson, 2003), 155-82, and parts of
Chapter Four were also used by John Pryor in “Types of ships”,
“Byzantium and the Sea”, and “Stadiodromikovn”. We acknowledge
the permission of the various publishers to reuse material here.
We are grateful to friends and colleagues who kindly read and
commented on preliminary drafts of this study or who made valuable
contributions in other ways: Professor Lionel Casson, Professor John
Dotson, Professor Michael McCormick, Dr John O. Ward, and Mr
Nigel Wilson. Professor Reinhold Mueller traced down at our request
the medal forged by Alvise Meneghetti and attributed to a Doge Pietro
Candiano. Associate Professor Dexter Hoyos was generous in
assistance with some tortuous passages in Latin. Mr Maxwell Walkley
was our consultant for Old French and Professors Margaret Clunies
Ross and Geraldine Barnes for Old Norse. Mr John Coates displayed
endless patience and good will in replies to numerous questions from
an academic (even if a nautically knowledgable one) to a practised
seaman and naval architect, and Ms Ann Hyland was our consultant
on horses. We are particularly indebted to Professor Michael Carter
xii PREFACE
for his extensive assistance with matters Arabic in the last years of
production of the book and to Professor John Haldon for the unfailing
generosity of his collaboration over many years. It is not possible to
acknowledge individual contributors everywhere, but they will
recognize where we are indebted to them.
We are also grateful to David Frendo for allowing us access to a
draft of his translation of the On the capture of Thessalonike2 of John
Kaminiate2s and also to Ann Moffat for her translation of chapters
II.44, 45 of the De cerimoniis of Constantine Porphyrogenne2tos. We
are also indebted to John Haldon for allowing us access to the
manuscript of his commentary on chapters II, 44 & 45 of the De
cerimoniis, before it was published as “Theory and practice”, and for a
pre-publication copy of his “‘Greek Fire’ revisited”.
John Pryor acknowledges that his interest in the Byzantine treatises
on naval warfare and in the drovmwn was originally aroused by a copy
of a translation of, and commentary on, the first few paragraphs of the
Naumacika; Levo nto" Basilevw" of Leo VI by the late R. H. Dolley,
which was made available to him by Mr Brian Dolley, his brother’s
executor. He is grateful to him for his generosity. R. H. Dolley had
intended to complete a translation of, and commentary on, the whole
of the Naumacika; Levonto" Basilevw", but never did so. We also
acknowledge the unfailingly helpful assistance of the staff of the
Bodleian Library, Oxford, in particular of Ms Christine Mason, and of
Fisher Library, University of Sydney, in particular of Rod Dyson,
Bruce Isaacs, and Megan O’Brien of Inter Library Loans and of Terry
O’Brien (no relation), the Library’s wizard at finding “lost” books.
Finally we would like to acknowledge the patience, and also the
assistance, of Michael Jeffreys, who suffered endless phone calls to
his wife from John Pryor, whom he came to know as “dromons
personified”, and of Gail Pryor, who suffered the absence of her
husband “away at sea” for years on end. Research for this book, the
original draft of which was a 12,000-word article, began in 1987.
John Pryor
Sydney
Elizabeth Jeffreys
Oxford
January 2006
LIST OF FIGURES, MAPS, AND TABLES
[a] Figures
Frontispiece: Reconstruction of a bireme dromon of the era of

the Macedonian emperors
Figure 1: Liburnae in the Vatican Vergil (Rome, Biblioteca
Apostolica Vaticana, MS. Vat. Lat. 3225, fol.
43v), late fourth century .…................................ 136
Figure 2: Liburnae on Trajan’s column, ca 114 C.E. …......... 137
Figure 3: Galley on a mosaic from the baths at Themetra,
near Hadrumetum, Tunisia, ca 200-220 C.E. …. 138
Figure 4: Dromons in the Roman Vergil (Rome, Biblioteca
Apostolica Vaticana, MS. Vat. Lat. 3867, fol.
77r), late fifth century ……………………….... 139
Figure 5: Catalan galley on a painted beam from a church
near Teruel (Barcelona, Museu nacional d’Art
de Catalunya, Ref. 15839), late thirteenth
century ………………………………………… 140
Figure 6: Dromon in the Ilias Ambrosiana (Milan,
Biblioteca Ambrosiana, Cod. Ambros. F. 205
Inf., min. VIII), early sixth century ………….... 141
Figure 7: Dromons in the Ilias Ambrosiana (Milan,
Biblioteca Ambrosiana, Cod. Ambros. F. 205
Inf., min. XXXVIII), early sixth century ……... 141
Figure 8: Two-masted, lateen-rigged dromons in a
manuscript of the Sacra Parallela attributed to
St John of Damascus (Paris, Bibliothèque
Nationale, MS. Gr. 923, fol. 207r), third quarter
of the ninth century ………………..….………. 142
Figure 9: Dromons rolling over Rho2s ships with their spurs
in the Synopsis historio2n of John Skylitze2s
(Madrid, Biblioteca Nacional, vitr. 26-2, fol.
130r), ca 1160 ………..……………………….. 144
Figure 10: The Athlit ram, third-second centuries B.C.E. …... 146
Figure 11: Caulkers at work extracting old pitch from the hull
of a ship in a manuscript of the De materia
medica of Dioskoride2s (N.Y., Pierpont Morgan
Library, Cod. 652, fol. 240r), tenth century …... 151
xiv LIST OF FIGURES, MAPS, AND TABLES
Figure 12: Square sail on a galley in a manuscript of the

Bible commissioned by abbot Vivian of St
Martin of Tours (Paris, Bibliothèque Nationale,
MS. Lat. 1, fol. 3v), ca 850 ………………...…. 155
Figure 13: Lateen sails on dromons in the Ilias Ambrosiana
(Milan, Cod. Ambros. F. 205 Inf., min.
XXVII), early sixth century, in the 1835 edition
by Angelo Mai …………………………..……. 156
Figure 14: Lateen? sails on dromons in the Ilias Ambrosiana
(Milan, Cod. Ambros. F. 205 Inf., min.
XXVII), early sixth century ………………..…. 156
Figure 15: Lateen-rigged ship in a manuscript of the Sermons
of St Gregory of Nazianzos (Paris, Bibliothèque
Nationale, MS. Gr. 510, fol. 367v), ca 879-82 158
Figure 16: Lateen-rigged ship in a manuscript of the Psalms,
the Khludov Psalter (Moscow, Historical
Museum, MS. 129 D, fol. 88r), ca 843-7 ……... 159
Figure 17: Two-masted, lateen-rigged ship in a painting from
Kellia, Egypt, ca 600-630 ……..…...…………. 160
Figure 18: Galley on a lustre-ware bowl from al-Fayyu2m,
Egypt (Cairo, Museum of Islamic Art, Inv. No.
7900), tenth century, probably Fa2t6i mid period 165
Figure 19: Midships section of a dromon of the era of the
Macedonian emperors ……………………….... 193
Figure 20: Longitudinal section of a dromon of the era of the
Macedonian emperors ……………………….... 205
Figure 21: Through-hull rudder mounts on a galley
representing the constellation Argo in a manu-
script of the Aratea attributed to Germanicus
Caesar (Boulogne-sur-Mer, Bibliothèque Muni-
cipale, MS. 188, fol. 78), eleventh century …… 223
Figure 22: Chelandium on a medal forged by Alvise
Meneghetti (1691-1768) attributed to a Venetian
Doge Pietro Candiano, as published by
Augustin Jal …………………...………………. 237
Figure 23: Ship borne siege engines in a treatise on
poliorcetics attributed to He2ro2n of Byzantium
(Rome, Biblioteca Apostolica Vaticana, MS.
Vat. Gr. 1605, fol. 40r), eleventh century …….. 243
Figure 24: Mosaic of a galley with a lowered mast from a
sepulchre at Hadrumetum, Tunisia, third
LIST OF FIGURES, MAPS, AND TABLES xv
century ……………………………………….... 249

Figure 25: Section of Figure 32 to demonstrate the effect of
overloading by ca 9.35 tonnes ………………... 263
Figure 26: Dromons in the Kyne2g etika of Pseudo-Oppian
(Venice, Biblioteca Marciana, MS. Gr. 479
[coll. 881], fol. 23r), eleventh century ………... 273
Figure 27: Midships oars of a dromon of the era of the
Macedonian emperors, drawn at a right angle to
the centre-line of the ship ……………………... 277
Figure 28: The oarage system of a bireme dromon of the era
of the Macedonian emperors, I: Interscalmia … 289
of the Macedonian emperors, II: straight-hulled
lower midships section with thalamian oars of
Olympias Mark II and scaled-down version ….. 293
of the Macedonian emperors, III: straight-
hulled midships section with full sized oars
above and below ……………………......…….. 295
of the Macedonian emperors, IV: straight-sided
midships section with oversized and full sized
oars above and below …………………………. 297
of the Macedonian emperors, V: midships
section with oversized and full sized oars above
and below and flared upper hull ………………. 301
of the Macedonian emperors, VI: midships
section with oversized and full sized oars above
and below, flared upper hull, and two oarsmen
for oversized oars above deck ……………….... 303
Figure 34: The fleet of Thomas the Slav advancing on
Abydos and carrying horses aboard a galley in
the Synopsis historio2n of John Skylitze2s
(Madrid, Biblioteca Nacional, vitr. 26-2, fol.
31v), ca. 1160, based on an eleventh-century
original …………………………………….….. 309
Figure 35: Mosaic of a galley from the Dermech district of
Roman Carthage, early fourth century .............. 310
Figure 36: Mosaic of a galley from Piazza Armerina, Sicily,
xvi LIST OF FIGURES, MAPS, AND TABLES
early fourth century …………………….……... 311

Figure 37: Horses unloaded from ports at the sterns of galleys
in a manuscript of Les livres des histoires du
commencement du monde (Paris, Bibliothèque
National, MS. Fr. 301, fol. 58v), fourteenth
century …………………………………...……. 313
Figure 38: Horse transports of the era of the Macedonian
emperors, I: standard bireme dromon with a
fifteen-hand horse ……………….……………. 321
Figure 39: Horse transports of the era of the Macedonian
emperors, II: modified dromon/chelandion with
a fifteen-hand horse ……………………..…….. 323
Figure 40: Bireme dromon of the era of the Macedonian
emperors heeling under sail to ten degrees ….... 337
Figure 41: Curve of sustainable speed against time for a
trie2re2s ………………………….…………….... 339
Figure 42: The oars of a dromon of the era of the Macedonian
emperors, drawn in the middle of the return
stroke at 67˚ to the centre line ……………….... 353
Figure 43: The cista Ficoronica: Jason and the Argonauts
watering at the spring of the Bebrycians …….... 361
Figure 44: Stowage of barrels or amphorae aboard a bireme
dromon of the era of the Macedonian emperors 369
Figure 45: Anchoring, mooring, and beaching galleys …........ 375
Figure 46: The harbour of Kastron Samos, adapted from
Great Britain, Admiralty, Hydrographic Office,
Chart No 1568 of 1967 ………………………... 377
Figure 47: Dromon in a manuscript of the Sermons of St
Gregory of Nazianzos (Mount Athos,
Pantelee2mon, Cod. 6, fol. 138r), twelfth century 409
Figure 48: Graffito of a katergon? From the monastery of the
Blatado2n at Thessalonike2, post 1355 ………..... 419
Figure 49: Graffito of a katergon? From Hagia Sophia,
Trebizond, probably fourteenth century …….... 421
Figure 50: Galleys in the Annales Ianuenses of Genoa (Paris,
Bibliothèque Nationale, MS. Suppl. Lat. 773),
ca 1160-1200 …………………………...……... 425
Figure 51: Bireme Muslim galley in the Synopsis historio2n of
John Skylitze2s (Madrid, Biblioteca Nacional,
vitr. 26-2, fol. 111v), ca 1160 ……………….... 427
Figure 52: Bireme galleys in the Synopsis historio2n of John
LIST OF FIGURES, MAPS, AND TABLES xvii
Skylitze2s (Madrid, Biblioteca Nacional, vitr.

26-2, fol. 145r), ca 1160 ……………………..... 427
Figure 53: Bireme galleys in the Synopsis historio2n of John
Skylitze2s (Madrid, Biblioteca Nacional, vitr.
26-2, fol. 146v), ca 1160 …................................ 428
Figure 54: Sicilian galley in a manuscript of the De rebus
Siculis carmen of Peter of Eboli (Berne,
Burgerbibliothek, MS. 120, fol. 119r), early
thirteenth century ……………………….…….. 429
Figure 55: The alla sensile bireme oarage system …………... 437
Figure 56: Galleys of Charles I of Anjou, ca 1269-1284, and
the bireme alla sensile oarage system ................ 441
Figure 57: Dromon using Greek Fire in the Synopsis historio2n
of John Skylitze2s (Madrid, Biblioteca Nacional,
vitr. 26-2, fol. 34v), ca 1160 ………………….. 613
Figure 58: Wolfenbüttel, Herzog August Bibliothek, Cod.
Guelf., 96 Gud. lat., fol. 157r-v ………….….... 615
Figure 59: Soldier using a hand-held flame thrower in a
treatise on poliorcetics attributed to He2ro2n of
Byzantium (Rome, Biblioteca Apostolica
Vaticana, MS. Vat. Gr. 1605, fol. 36r), eleventh
century ………………………………….……... 619
Figure 60: The Sotiel Coronada Roman force pump from
Valverde, Huelva, Spain, probably dated to the
first century C.E., Madrid, Museo Arqueo-
logico Nacional …………….…………………. 625
Figure 61: “Greek” or liquid fire sipho2n built by Colin Hewes
and Andrew Lacey under the direction of John
Haldon …………………………………………….. 629
[b] Maps
Map 1: The Atlantic and western Mediterranean:

antiquity and the Early Middle Ages ………... lxv
Map 2: The Atlantic and western Mediterranean: the
High Middle Ages …………………………… lxvi
Map 3: The central and eastern Mediterranean …………. lxvii
Map 4: Southern Italy and Sicily ……………………….. lxviii
Map 5: The Ionian Sea ………………………………….. lxix
Map 6: The Adriatic Sea ………………………………... lxx
xviii LIST OF FIGURES, MAPS, AND TABLES
Map 7: The southern Aegean and Lycia ………………... lxxi

Map 8: The northern Aegean and the Balkans ………….. lxxii
Map 9: The Dardanelles to the Black Sea ………………. lxxiii
Map 10: Byzantine Asia Minor …………………………... lxxiv
Map 11: Greater Syria ………………………………….… lxxv
Map 12: Egypt and Palestine …………………………….. lxxvi
Map 13: Greater Iraq ……………………………………... lxxvii
[c] Tables
Table 1: Rulers of the first period, ca 400-560 ……………. 10

Table 2: Rulers of the second period, ca 560-750 ………… 20
Table 3: Rulers of the third period, ca 750-875 …………… 36
Table 4: Rulers of the fourth period, ca 875-1025 ………… 52
Table 5: Rulers of the fifth period, ca 1025-1204 …………. 78
Table 6: Specifications for conjectured anchoring systems
in the De cerimoniis …………………………... 211
Table 7: Some reported voyages of ancient and medieval
galleys and galley fleets ………………………. 344
Table 8: Artists of the manuscript Madrid, Biblioteca
Nacional, Vitr. 26-2 of the Synopsis historio2n
of John Skylitze2s according to (1) Grabar and
Manoussacas and (2) Tsamakda ……………… 634
Table 9: The galleys of the manuscript Madrid, Biblioteca
Nacional, Vitr. 26-2 of the Synopsis historio2n
of John Skylitze2s classified by artistic style ….. 636
Table 10: Comparison of styles of depictions of galleys in
the manuscript Madrid, Biblioteca Nacional,
Vitr. 26-2 of the Synopsis historio2n of John
Skylitze2s with the general styles and artists as
identified by (1) Grabar and Manoussacas and
(2) Tsamakda …………………………………. 640
Table 11: Galley Group artists correlated to (1) Grabar and
Manoussacas and (2) Tsamakda ……………… 642
NOTE ON ORTHOGRAPHY, TRANSLATION, CITATION, AND
DATING
Orthography
This work is intended for the use of not only Byzantinists who can
read Greek but also maritime historians and medievalists of all
persuasions who can not. Therefore, we have tried to be as
accomodating as possible, frequently giving terms not only in Greek,
but also in transliteration and in translation where we have judged it
useful to readers to do so. We have also created a Glossary of
selective Arabic, Greek, and Latin technical terms. Usages have
usually been given in the text in Roman transliteration or English
translation except where specific reference is made to a particular
Greek text. All Arabic citations have been made in transliteration.
Terms included in the Glossary have been given in italics in the text.
There is also a separate Glossary of English nautical terminology
for the assistance of those unfamiliar with matters maritime.
In transliterating Greek we have distinguished h from e by adding a
makron to “e” for the h, as in “e2”. Similarly, we have distinguished
the Greek “w” from “o” by adding a makron to “o” for the w as in “o22”.
The only exception we have made to this rule is the word drovmwn
itself. It would have been pedantic beyond words to have used
“dromo2n” and “dromo2nes” on hundreds of occasions. Except where
the use of the word itself is at issue, we have simply used “dromon”
and “dromons”. In transliterating Arabic and Turkish we have
followed a modified version of the Encyclopedia of Islam system,
only replacing “dj” by “j”, “k5” by “q”, and omitting the underlining of
digraphs; thus, Aghlabid rather than Aghlabid, Shah rather than Shah,
etc.
To assist readers lacking a strong historical background we have
created a selective Gazeteer of historical place names which can not
be found in commonly available English-language atlases. Place
names included in the Gazeteer have been given in italics in the text.
In the Gazeteer, and throughout, Greek names of people, places,
institutions, etc. have been standardized to the usage of The Oxford
dictionary of Byzantium or, failing that, Smith’s Dictionary of Greek
and Roman geography, or the British Admiralty Mediterranean Pilot.
We have preserved the Greek orthography in the transliteration of
xx NOTE ON ORTHOGRAPHY
proper nouns except where there is a common English equivalent, thus

Constantine rather than Ko2nstantinos, Constantinople rather than
Ko2nstantinoupolis, Leo rather than Leo2n , etc. However, Nike2phoros,
Lekape2nos, etc. On occasions when the extent of a common English
equivalent is ambiguous we have had to make a choice, thus
Thessalonike2 rather than Thesssaloniki or Salonika, etc.
Arabic names and titles have been standardized to the
transliterations of the Encyclopedia of Islam as modified above.
Translation
In translating Greek, Latin, and Arabic texts we have have kept in

mind throughout our intended readership; namely, not only
Byzantinists but also maritime historians and others who cannot read
these languages but who are concerned with the close technical
meaning of the texts. Therefore we have made our translations as
literally close to the Greek, Latin, and Arabic texts as it is possible to
do without making the English incomprehensible.
Square brackets [ ] in translations indicate our additions to texts
where the corresponding words in English do not exist in the texts but
need to be added to make the English comprehensible.
Citation of Primary Sources
Because not all readers will have access to the same editions we have
used, we have followed the principle that in citation of primary
sources we have given any text subdivisions, for example, book,
chapter, verse, etc. first. We have then added the page numbers of the
editions we have used in parentheses.
On the one hand, Greek and Latin texts have been cited according
to the best editions known to us. Where translations known to us into
European languages exist, we have cited the most convenient of these
in the Bibliography for the information of readers; however, we have
not cited them in the notes.
On the other hand, for Arabic texts we have followed the principle
that because standard editions in Arabic are in many cases very
difficult to obtain, even in major libraries, and because few readers of
this book will be able to read Arabic, we have used translations. In
many cases this has meant using a variety of translations, frequently
NOTE ON ORTHOGRAPHY xxi
obscure, of selections from, and parts of, texts. The only exceptions to
this rule have been where a point has needed to be made in the text
from a part of an Arabic text which has no translation known to us or
where a word with a technical meaning is at issue.
Dating
All dates refer to the “Christian” (or “Common”) Era (C.E.) or “Anno
Domini” (A.D.) unless otherwise specified. “B.C.E.” is used for the
pre-Christian era, “A.H.” for the years of the Hijrah, the Muslim
calendar dating from 16 July 622, and “A.M.” for Annus Mundi, the
Byzantine sytem of dating from the Creation, reckoned to B.C.E.
5,508.
When citing the regnal dates of Byzantine emperors, we have used
their entire reigns, irrespective of whether or not they were only co-
emperors for part of that time.
NOTE ON METROLOGY AND HOURS OF DAYLIGHT
[a] Metrology 1
In terms of the practical limitations of medieval shipbuilding and also

of medieval mensuration, measurements or trigonometrical
calculations taken to tenths or even hundredths of a centimetre, or
other equivalents, would have been totally impracticable of course.
We have used such equivalents as derived from the nineteenth-century
metrological manuals compiled by Martini, Doursther, and others
after the metrification of Europe only as a base on which to build
more realistic medieval measures. Similarly, we have rounded out the
results of often complex mathematical applications to sensible figures.
Modern
British Imperial Metric

ton tonne
.984 1 1.016
mile kilometre
.621 1 1.609
Byzantium
Length
pou'" (pous) = 16 daktyloi = 31.23 cms
ph'cu" (pe2chys) = 24 daktyloi [= 1.33 = 46.8 cms
podes]
mivlion (milion) = 4,200 or 4,500 podes = 1.31 or 1.40 kms
Weight
Livtra (litra) = approx. 0.32 kgs
Liquids
mevdimno"// movdio" = 40 litrai = 17.1 litres
(basilikos modios)
------------------------------
1
Sources: Doursther, Dictionnaire universel; Martini, Metrologia; Oxford
dictionary of Byzantium.
xxiv NOTE ON METROLOGY
Genoa
Liquids
quartarolo = 39.75 litres
barilio = 2 quartaroli = 79.5 litres
mezzaruola = 2 barili = 159 litres
Marseilles
Length
palmus = 0.252 metres
Liquids
millayrola/milhairocla = 63.5 litres
Naples
Liquids
barile = 43.625 litres
Paris
Liquids
quarta = 2 pinte = 1.86 litres
[b] Hours of Daylight
For hours of daylight we have utilized the United States Naval

Observatory, Complete sun and moon data for one day, @
http:/aa.usno.navy.mil/data/docs/RS_OneDay.html. This enables one
to enter any town in the world for any day of any year back to 1800 to
obtain hours of daylight and moonlight for that day and place. It gives
the hours of the beginning of daylight, sunrise, the midday transit,
sunset, and the end of twilight. We have used the year 1800
throughout because it is the earliest for which the Observatory has
computed data, because it eliminates the effects of modern climate
change, and because the climate at that time is considered by
historians of climate to have been similar to that of the Middle Ages.
Even if the climate was not quite the same, the differences would have
been negligable and would not affect any calculations made here.
SELECTIVE GAZETEER OF HISTORICAL PLACE NAMES1
Abbreviations
Ar. Arabic lit. literally

Byz. Gr. Byzantine Greek med. Lat. medieval Latin
cl. Lat. classical Latin prov. province
et var. et variae (and variants) q.v. quod vide, which see
Gr. Greek R. Roman
Abydos “Abudo", et var. (cl. & Byz. Gr.), Abydus, et var.

(cl. & med. Lat.): town and port on the
Dardanelles, near Çanakkale, Turkey, ruined.
Achelo2on Acelw'/on (Byz. Gr.): battlefield near a river or
fortress of the name, W coast of the Black Sea.
Actium “Aktion (cl. Greek), Actium (cl. & med. Lat.): naval
battle off the promontory of the name, S
entrance to the Gulf of Amvrakia, Greece.
Aegates Aegates insulae (cl. Lat.), possibly from Gr.
Aijgavde": Egadi islands, W Sicily.
Africa Africa (cl. Lat.), ’Afrikhv (cl. Gr.): prov. of R.
North Africa, N Tunisia.
Aigaion Pelagos AiJgaio'n Pevlago" (Byz. Gr.): lit. “Aegean Sea”,
Byzantine naval thema†, headquarters Lesbos.
Aigina Ai[gina (cl. & Byz. Gr.), Aegina, et var. (cl. & med.
Lat.): island, Aiyina, Greece
Al-Andalus Al-Andalus (Ar.), Andalus/Hispania/Iberia (med.
Lat.): Muslim Spain, connected hypothetically to
the Arabic for the Vandals, al-Andalı3sh.
Alarcos Alarcos (med. Spanish), Al-Arak (Ar.): battlefield
near Santa Maria de Alarcos, prov. Calatrava la
Vieja, Spain.
Al-Farama2’ Phlouvsion (Gr.), Pelusium (cl. Lat.), Pelusium, et
------------------------------
1
The gazeteer is a guide only to those historical place names which may not be
known commonly and which cannot be found in the commonly-available English
language atlas: J. C. Bartholomew, The world atlas, any ed. (Edinburgh). The style of
the Wisconsin History of the Crusades [Setton, HC] has been followed except that for
Turkish and Arabic names a modified version of The Encyclopedia of Islam [EI.2] has
been followed and for Byzantine names The Oxford Dictionary of Byzantium [ ].
Graesse, Orbis Latinus, has also been used. All are acknowledged throughout.
xxvi SELECTIVE GAZETEER
var. (med. Lat.), al-Farama2’ (Ar.): town, al-

Farama2’, Egypt.
Al-Fayyu2m (Ar.), Phiom (Coptic): fertile agricultural region of
Middle Egypt, capital Arsinoe (q.v.).
Al-Fust6a2t6 Al-Fust6at2 6 (Ar.), Babylon/Babilonia, et var. (med.
Lat.): Muslim town north of Babylon (q.v.), later
Qas5r al-Sham‘, now Mis5r, Old Cairo, Egypt.
Al-Ku2fa Al-Ku2fa (Ar.), Cufa/Kufa (med. Lat.): town founded
on the middle Euphrates in 638 C.E., near al-
Najaf, Iraq, ruined.
Al-Mahdiyya Al-Mahdiyya (Ar.), Madia (med. Lat.): town and
port founded by ‘Ubayd Alla2h al-Mahdı3 in 912-
13 C.E., Mahdia, Tunisia.
Al-Qayrawa2n Al-Qayrawa2n (Ar.): Aghlabid capital, founded by
Mu‘a2wiya ibn H4udayj in 665 C.E., Kairouan,
Tunisia.
Al-Rawd5a Al-Rawd5a (Ar.): island in the Nile at al-Fust6a2t6
(q.v.).
Ampurias Emporivai/Empovrion (cl. Gr.), Emporiae/Emporium
(cl. Lat.), Emporitanus, et var. (med. Lat.):
medieval county, from the ancient city of the
same name, Gulf of Rosas, Catalonia, Spain.
Anatolikon Anatolikovn: original thema† of Asia Minor,
central Anatolia south of Opsikion (q.v.).
Antioch on Antiovceia ejpi; Kravgw/ (cl. Gr.), Antiochea super
Cragus Cragum, et var. (med. Lat.): fortress city on the
Gulf of Antalya, Turkey, lost.
Aphrodite2 Afrodivth" povli"/Afrodivtw (cl. & Byz. Gr.),
Aphrodito/Aphroditopolis, et var. (cl. & med.
Lat.): city of al-Fayyu2m (q.v.), Kom Ishgaw,
near Atfih, Egypt.
Aquileia Aquileia (cl. Lat.), Akulhiva (Byz. Gr.), Aquilegia,
et var. (med. Lat.): town, Aquileia, Italy.
Aquitania Aquitania Secunda (cl. Lat.): prov. of R. Gaul on
Secunda the Bay of Biscay, N of the Garonne, France.
Arados ÔH “Arado", et var. (cl. Gr.), Aradus (cl. & med.
Lat.), Arwa2d/Ruwa2d (Ar.): island off Lebanon
near Tortosa/Tartus, Ruad, Syria.
Arginousai Arginou'sai (cl. Gr.), Arginousae (cl. Lat.):
promontory or small islands on the mainland
opposite Mityle2ne2, Turkey.
SELECTIVE GAZETEER xxvii
Armeniakon Armeniakovn: original thema† of Asia Minor,

eastern Anatolia from Cappadocia to the
Euphrates, reduced in the 9th century to the
coast of the Black Sea, bounded by Paphlagonia
(q.v.) to the W, Kolo2neia to the E, and
Charsianon to the S, Turkey.
Arsinoe Arsinovh (cl. Gr.), Crocodilopolis (cl. Lat.),
Arsinoe/Crocodilopolis (med. Lat.), Madı3nat al-
Fayyu2m (Ar.): capital of the area of Middle
Egypt known as al-Fayyu2m (q.v.), Madı3nat al-
Fayyu2m, Egypt.
Arsuf Apollwniva (cl. Gr.), Apollonia (cl. Lat.),
Apollonia/Arsur/Sozusa, et var. (med. Lat.):
Arsuf, Israel.
As5ı3la Zh'li", et var. (cl. Gr.), Zilia (cl. Lat.), Arzilia, et
var. (med. Lat.), As5ı3la (Ar.): town and port on
the Atlantic coast, As(z)ilah, Morocco.
Astakos, Gulf of Astakov" (cl. & Byz. Gr.), Astacus (cl. Lat.),
Astacenus sinus (med. Lat.): Gulf, Izmit Körfezi,
Turkey.
Avlona Aulon (cl. Lat.), Aujl wvn (Byz. Gr.), Aulona/Valona,
et var. (med. Lat.): town, Vlorë, Albania.
Ayas Agai (cl. Lat.), Ayacium (med. Lat.), A›ya2s (Ar.):
Ayas, prov. Yumurtalı£k, Turkey.
Babylon Babylon (cl. Lat.), Babulwvn (cl. Gr.): military camp
and fortress, N of al-Fust6a2t6 (q.v.), Egypt, ruined.
Ba2dis Parietina? (cl. Lat.), Ba2dis (Ar.), Bardias, et var.,
(med. Lat.): town at the mouth of the
Ta2lemba2d es river, territory of Naku2r (q.v.),
Morocco (to Spain), ruined.
Barqa Bavrkh (Gr.), Barce (cl. Lat.), Barca, et var. (med.
Lat.), Barqa (Ar.): town and prov., al-Marj,
Libya.
Bija2ya Saldae (cl. Lat.), Savldai (cl. Gr.), Bogia (med.
Lat.), Bija2ya (Ar.): town and port, Béjaïa,
Algeria.
Bilbays Phelbès (Copt.), Bilbays (Ar.), Belbeis (med. Lat.):
town, Bilbays, Egypt.
Bithynia Biquniva (Gr.), Bithynia (cl. Lat.): prov. of NW Asia
Minor.
Bona Hippo Regius (cl. Lat.), ÔIppw;n Basilikov" (cl. Gr.),
xxviii SELECTIVE GAZETEER
Bu2na, et var. (Ar.), Bona, et var. (med. Lat.):

town and port, al-‘Anna2b a, Algeria.
Boulgarophygon Boulgarovfugon (Byz. Gr.): battlefield, Baba Eski,
near Edirne, Turkey.
Bouthro2ton Bouqrwtovn (cl. Gr.), Buthrotum (cl. & med. Lat.):
coastal town, Butrint, Albania.
Busta Gallorum Boustagallwvrwn (Byz. Gr.): battlefield near
Gualdo, Umbria, Italy.
Byzacena Valeria Byzacena (cl. Lat.), Bussavti" cwvra (Byz.
Gr.): prov. of R. North Africa created by
Diocletian, southern Tunisia.
Byzantion Buzavntion (cl. Gr.), Byzantium (cl. Lat.): Greek city
on the site of the later Constantinople.
Caesarea Kaisavreia (cl. & Byz. Gr.), Caesareia (cl. Lat.),
Caesarea, et var. (med. Lat.): town, Kayseri,
Turkey.
Calatrava Qal‘at Raba2h5 (Ar.), Callatriva (med. Lat.): town
and castle, Calatrava la Vieja, prov. Ciudad
Real, Spain, ruined.
Camarina Kamavrina/Kamari'na (cl. Gr.), Camarina (cl. Lat.):
Greek city on the S coast of Sicily, 30 kilometres
E of Gela, ruined.
Cannae Kavnnai (cl. Gr.), Cannae (cl. Lat.): town on the S
bank of the river Ofanto, Apulia, Italy, 10
kilometres inland.
Capitanata Capitanata (med. Lat.): northern Apulia, prov. of
Foggia, Italy. From katepano2†, because the Byz.
katepanate repopulated it in the eleventh
century.
Cartenna Kartevnnai (cl. Gr.), Cartenna (cl. Lat.),
Cartenna/Teneza, et var. (med. Lat.): city and
port, Ténès, Algeria.
Carthage Karchdoniva (Gr.), Carthago (cl. Lat.), Carthago, et
var. (med. Lat.), Qart6aj2 anna (Ar.): Roman
capital of Africa (q.v.), on the site of the Carth-
aginian capital, superseded by Muslim Tunis.
Castronuovo Castellum novum (med. Lat.): Muslim fortress in
central Sicily NE of Prizzi.
Cerami Ceramum (med. Lat.): town, Cerami, district of
Enna, Sicily.
Cetara Cetara (med. Lat.): unknown site, not far from
SELECTIVE GAZETEER xxix
Salerno.
Chalke2do2n Calkhdwvn (cl. & Byz. Gr.), Chalcedon (cl. Lat.),
Calcedonia, et var. (med. Lat): town opposite
Constantinople across the Bosporos, Kadıköy,
Turkey.
Chandax Al-Khandaq (Ar.), Cavndax (Byz. Gr.), Candia/
Chandax, et var. (med. Lat.): port, Iraklion,
Crete.
Chelidonia, Hiera acra (cl. Lat.): from the Celidovniai,
Cape Chelidoniae insulae (cl. Gr. & med. Lat.), two
islets off its tip, Cape Gelidonya, Turkey.
Cherso2n Cerswvn (Byz. Gr.), Cherso/Chersonium (med.
Lat.): town, Kherson, Crimea, Ukraine.
Christiana Cristiana;, Ta;, (Byz. Gr.): islet SW of The2ra,
Greece.
Chrysopolis Crusovpoli" (cl. & Byz. Gr.), Chrysopolis (cl.
Lat.), Chrysopolis/Scutarium (med. Lat.):
harbour on E shore of Bosporos upstream from
Constantinople, Üsküdar, Turkey.
Cilicia Kilikiva (cl. & med. Gr.), Cilicia (cl. & med. Lat.):
prov. of R. SE Asia Minor between the Gulf of
Antalya and the Gulf of Alexandretta, Turkey.
Civitate Teanum (cl. Lat.), Teavnon (cl. Gr.), Civitas (med.
Lat.): town on Fortore river, Capitanata (q.v.),
Apulia, Italy.
Classe Classis, et var. (med. Lat.), Klavsai (Byz. Gr.): port
of Ravenna, Italy.
Covadonga Cova Sancte Marie/cova-domenica (med. Lat.):
rock or outcrop in the Picos de Europa Peña
Vieja, E of Cangas de Onis, prov. Asturias,
Spain
Dacia Dacia (cl. Lat.), Dakiva (cl. & Byz. Gr.), Dacia, et
var. (med. Lat.): R. prov. ca 101-270/75 C.E., N
of the Lower Danube, central Romania.
Darna Davrni" (cl. Gr.), Darnis et var. (med. Lat.), Darna
(Ar.): town and port, Darna/Derna, Libya.
De2me2trias Dhmhtriav" (cl. & Byz. Gr.), Demetrias (cl. & med.
Lat.): port on the Gulf of Pagasaí, SE of Volos,
Greece, ruined.
Develtos Debeltov"/Dhbeltov" (Byz. Gr.), Deultum/Develtus,
et var. (med. Lat.): city and fortress controlling
xxx SELECTIVE GAZETEER
the road on the W coast of the Black Sea, 20

kilometres S of Burgas, Debelt, Bulgaria.
Dorostolon Dorovstolon (Byz. Gr.), Durostorum/Silistra/ Distra
et var. (med. Lat.): city and fortress on the
Danube, Silistra, Bulgaria.
Drako2n Dravkwn (Byz. Gr.): geographical hapax legomenon
mentioned by Prokopios, a river somewhere near
Mt Vesuvius.
Duklja Diovkleia (Byz. Gr.), Dioclea/Zenta, et var. (med.
Lat.): stronghold at the junction of the Zeta and
Morava rivers, name extended to the coastal
plain north of the thema† of Dyrrachion (q.v.),
Serbian heartland, Serbia.
Dyrrachion Epidamnus/Dyrrachium (cl. Lat.), Durravcion (Byz.
Gr.), Dy(i,o,u)rrachium/Epidamnus, et var.
(med. Lat.): town and port, Durrës, Albania.
Eknomos “Eknomo" (cl. Gr.), Ecnomus (cl. Lat.): site near the
mouth of the river Salso, S coast of Sicily.
Elvira Ilbı3ra (Ar.), Eleberis/Iliberris, et var. (med. Lat.):
town on part of the site of Granada, Spain,
ruined.
Epidaurus Epivdauro" (cl. Gr.), Epidaurus (cl. Lat.): town in
Argolis, E coast of Peloponne2sos, Greece,
ruined.
Epidauros Epivdauro" (cl. Gr.), Epidaurus (cl. Lat.): town in
Illyricum, abandoned for Ragusa (q.v.), Croatia,
ruined.
Epiros “Hpeiro" (cl. & Byz. Gr.), Epiros/Epirus (cl. &
med. Lat.): prov. of NW Greece bounded to the
N by the Gulf of Vlorës and to the S by the Gulf
of Amvrakia, Greece & Albania.
Euripos Eu[ripo" (cl. Gr.), Euripus (cl. & med. Lat.): 9th-
century synonym for Chalkis, Greco-Byz. town
on Evvoia at the straits separating the island
from the mainland, Greece.
Fraxinetum Fraxinetum/Fraxinum (med. Lat.): Muslim corsair
nest, La Garde-Freinet, Dép. Var, France.
Galata Ta; Galavtou/ Galata'" (Byz. Gr.), Calata/Galatas/
Sycae, et var. (med. Lat.): originally Sykai,
fortress and suburb opposite Constantinople on
the NE of the Golden Horn.
SELECTIVE GAZETEER xxxi
Galatia Galativa (cl. & Byz. Gr.), Galatia (cl. & med. Lat.):
N-central Anatolian plateau E from the river
Sakarya to the Kizil Irmak and S from
Paphlagonia (q.v.) to the Salt Lake, Tuz Gölu,
Turkey.
Grado Gravdon (Byz. Gr.), Gradum/Nova Aquileia (cl. &
med. Lat.): late R. fortress and city on an island
south of Aquileia (q.v.), Grado, Italy.
Hadrumetum Hadrumetum et var. (cl. Lat.), ÔAdravmhto", et var.
(Byz. Gr.): capital of Byzacena (q.v.), town and
port, Su2sa, Tunisia.
Harmathous ÔArmaqou'" (cl. Gr.), Harmathus (cl. Lat.): town on
the N coast of the Gulf of Edremit E of Cape
Baba, Turkey.
H4at6t6ı3n, Horns of H4at6t6ı3n (Ar.), Carnehatin (med. French), Mares-
calcia (med. Lat.): village and hill around 4
kilometres SW of Tiberias, Israel.
H4aydara2n Al-H4aydara2n (Ar.): battlefield, probably between
30-60 kilometres S-SW of al-Qayrawa2n (q.v.),
Tunisia.
Heliopolis of ÔHliouvpoli", (cl. & Byz. Gr.), Heliopolis Aegypti
Egypt (cl. & med. Lat.): city in Egypt, 12 miles N of
Babylon (q.v.), ruined.
Heliopolis of ÔHliouvpoli", (cl. & Byz. Gr.), Heliopolis Syriae (cl.
Syria Lat.), Heliopolis Libanesia, et var. (med. Lat.),
Ba‘labakk (Ar.): city, Baalbek, Lebanon.
Hellas ÔEllav" (Byz. Gr.): thema†, E-central Greece, capital
Thebes (q.v.), Greece.
He2rakleia Pevrinqo" (cl. Gr.), ÔHravkleia (Byz. Gr.), Perinthus
(cl. Lat.), Heraclea (med. Lat.): town and port,
Marmara Ereg°li, Turkey.
He2rakleia ÔHravkleia Povntikh (cl. & Byz. Gr.) Heraclea
Pontike2 Pontica (cl. Lat.), Heraclea/Pontaracia, et var.
(med. Lat.): town and port, Black Sea, Ereg°li,
Turkey.
Hierax ÔIevrako" (Byz. Gr.), Hierax/Zarax (med. Lat.): port
N of Monemvasia on the E coast of the
Peloponne2sos, Greece.
Hunayn Hunayn (Ar.): port, near Beni Saf, Algeria.
Ibe2ria Ibhriva (Byz. Gr.): NE thema† in Asia Minor
created from Armenian lands, Turkey.
xxxii SELECTIVE GAZETEER
Ifrı3qiya Ifrı3qiya (Ar.): Arabic name for the R.-Byz. prov. of

Africa (q.v.).
Ivrea Iporegia, et var. (med. Lat.); town and marquisate,
prov. Turin, Italy.
Kardia Kardiva (cl. & med. Gr.), Cardia (cl. & med. Lat.);
town at the head of the Gulf of Saros, Turkey,
lost.
Kasion Kavsion/ Kavsio" o[ro" (cl. Gr.), Casius mons (cl. Lat.):
small summit south of Lacus Sirbonis (Sabkhat al
Bardawil), Egypt.
Katasyrtai Katasuvrtai (Byz. Gr.): battlefield in Thrace not far
from Constantinople, unidentified.
Kellia Kelliva (Byz. Gr.): monastic complex near the W
edge of the Nile delta, Egypt.
Ke2poi Kh'poi (Byz. Gr.): “the Gardens”, aple2kton† at the
mouth of the Maeander (q.v.) river, Turkey.
Khura2sa2n Khura2sa2n (Ar.), Corasania, et var. (med. Lat.):
prov. of NE Iran, W Afghanistan, and S
Turkmenistan.
Kibyrrhaio2tai Kiburraiw'tai (Byz. Gr.): [“those of Kibuvrra”,
from Kibuvra the Greater, classical city of the
name in Phrygia]; originally part of the fleet of
the Karabisianoi but erected into the first Byz.
naval thema†, around the Gulf of Antalya,
Turkey.
Kleidion Kleidivon (Byz. Gr.), Cimbalongus (med. Lat.):
battlefield in a mountain pass near the Struma
valley, Kljuc°, Bulgaria.
Korkyraioi Korkurai'oi (cl. Gr.) Corcyraei (cl. Lat.):
inhabitants of Kovrkura/Corcyra, Corinthian
colony, Corfu, Greece.
Kyzikos Kuvziko" (cl. & Byz. Gr.), Cyzicus (cl. & med. Lat.):
city and port on the S coast of the Sea of
Marmara, Balkız, Turkey.
Lamos Lavmo" (cl. & Byz. Gr.), Lamus (cl. & med. Lat.):
river in Cilicia (q.v.), boundary between Muslim
and Byz. territory in the 9th-10th centuries.
Lampsakos Lavmyako" (cl. & Byz. Gr.), Lampsacus (cl. & med.
Lat.): city on the E shore of the Dardanelles
opposite Gallipoli, Turkey, lost.
Las Hormigas Formigueras (med. Lat.): group of islets off the
SELECTIVE GAZETEER xxxiii
coast of Catalonia, NE of Palamos, Spain.

Las Navas de Las Navas de Tolosa (med. Spanish), Al-‘Iqa2b
Tolosa (Ar.): battlefield 9 kilometres NW of Las Navas
de Tolosa, prov. Ciudad Real, Spain.
Laureate2 Laureavth (Byz Gr.): geographical hapax
legomenon mentioned by Prokopios, somewhere
on the coast of Dalmatia.
Lebounion, Mt Lebouvnion (Byz. Gr.): hill or mountain near the
mouth of the Maritsa river, Turkey.
Lilybaion Liluvbaion (cl. Gr.), Lilybaeum (cl. Lat.), Marsa
‘Alı3 (Ar.), Marsala (med. Lat.): Carthaginian
settlement, Roman town, on Cape Bœo, Marsala,
Sicily.
Longobardia Loggibardiva/ Lagoubardiva (Byz. Gr.), Longobar-
dia thema (med. Lat.): Byz. thema†, Apulia and
NE Basilicata, Italy.
Lycia Lukia (cl. & Byz. Gr.), Lycia/Lucia (cl. & med.
Lat.): R. prov. of SW Asia Minor.
Maeander Maivandro" (cl. & Byz. Gr.), Maeander/ Maeandrus
(cl. & med. Lat.): river Mendere, Turkey.
Mamora Banasa (cl. & med. Lat.), Bavnassa (cl. Gr.), Al-
Mamora (Ar.): ancient town on the river Sebou,
Morocco, lost.
Mams, the Al-Mams (Ar.): watercourse about 50 kilometres W
of al-Qayrawa2n (q.v.), Hanshı3r Dwı3mı3s, Tunisia.
Mantzikert Mantzikievrt (Gr.), Manazkert (Arm.) Mala2zgird
(Ar.), Malazgirt (Turkish): fortress city N of
Lake Van, Malazgirt, Turkey.
Mauritania Mauritania Caesariensis (cl. Lat.): prov. of R.
Caesariensis North Africa, N-W Algeria.
Mauritania Mauritania Sitifensis (cl. Lat.): prov. of R. North
Sitifensis Africa, N-central Algeria.
Mesopotamia Mesopotamiva (Byz. Gr.): thema† of S-E Asia
Minor, Turkey/Syria/Iraq.
Morea Moreva (Byz. Gr.), Morea (med. Lat.): the
Peloponne2sos, Greece.
Mouikouron Moui>kou'ron (Byz Gr.): geographical hapax
legomenon mentioned by Prokopios, near
Salo2nes (q.v.), Croatia.
Mylae Mulaiv (cl. Gr.) Mylae (cl. & med. Lat.): town and
port, Milazzo, Sicily.
xxxiv SELECTIVE GAZETEER
Myra Muvra (cl. & Byz. Gr.), Myra (cl. & med. Lat.):
town and port on the river Andracus, about 3.5
kilometres inland, Dembre, Turkey.
Myriokephalon Muriokevf alon (Byz. Gr.): battlefield E of Homa in
the Maeander (q.v.) valley, Turkey.
Nahr al-Ibra2hı3m “Adwni" (cl. Gr.), Adonis (cl. & med. Lat.): river, S
of Jubail, Lebanon.
Naku2r Naku2r (Ar.), Nocor/Necur (med. Lat.): city and
state around al-H4usayma/Alhucemas Bay,
Morocco (to Spain).
Narbonensis Narbonensis Prima (cl. Lat.): prov. of R. Gaul,
Prima Languedoc.
Nicaea Nivkaia (cl. & Byz. Gr.), Nicaea/Nicea (cl. & med.
Lat.): city, Iznik, Turkey.
Nikome2deia Nikomhvdeia (cl. & Byz. Gr.), Nicomedia (cl. &
med. Lat.): city, Izmit, Turkey.
Nikopolis Nikovpoli" (cl. Gr.), Nicopolis (cl. & med. Lat.):
city on the Gulf of Amvrakia, Greece, ruined.
Thema† of Nikopolis, capital Naupaktos.
Nineveh Nivno"/Nineuvh (cl. Gr.), Ninus (cl. Lat.), Ninive, et
var. (med. Lat.): capital of ancient Assyria on
the Tigris river opposite Mosul, Iraq.
Novem Populi Novem Populi (cl. Lat.): prov. of R. Gaul, Gascony.
Numidia Numidia (cl. Lat.): prov. of R. North Africa, N-E
Algeria.
Ophryneion Ofruvneion (cl. Gr.), Ophrynium (cl. Lat.): town
near lake Pteleos on S coast of the Dardanelles.
Opsikion Oyivkion (Byz. Gr.): original thema† of Asia
Minor, NW Asia Minor N of Anatolikon (q.v.),
capital Ankara, Turkey.
Ostia Ostia (cl. Lat.), Hostia, et var. (med. Lat.): city and
port of Rome at the mouth of the Tiber river,
synonym for Portus, a new port built by
Claudius some 3 kilometres N of Ostia.
Outremer Ultra mare (med. Lat.), Outremer (med. French):
the Crusader states in the Levant.
Oxyrhynchus Oxuvrugco" (cl. Gr.), Pemje (Coptic), al-Bahnasa2
(Ar.): town in upper Egypt, Bahnasa, ruined.
Pamphylia Pamfuliva (cl. & Byz. Gr.), Pamphylia (cl. & med.
Lat.): prov. of S Asia Minor around the Gulf of
Antalya, Turkey.
SELECTIVE GAZETEER xxxv
Pannonia Pannoniva (cl. & Byz. Gr.), Pannonia (cl. Lat.),

Pannonia/Hungaria (med. Lat.): R. territories S
& W of the middle Danube, E Austria and W
Hungary.
Paphlagonia Paflagoniva (cl. & Byz. Gr.), Paphlagonia (cl. &
med. Lat.): coastal region of N Asia Minor,
thema†, capital Çankırı, Turkey.
Patara Pavtara (cl. Gr.), Patara/Patera (cl. & med. Lat.):
town and port E of river Etshen, Turkey, ruined.
Pechina Bajja2na (Ar.): Muslim town in the hills behind
Almeria, Spain.
Pe2g e2 Phghv (Byz. Gr.): church, monastery, and palace
outside the walls of Constantinople.
Peukia, Ta Peukiva, Ta; (Byz. Gr.): “the Pines”, perhaps
Pefkhia near Ophryneion (q.v.), Turkey.
Phoinikous Foinikou'" (cl. & Byz. Gr.), Phoenicus (cl. Lat.),
Finica/Phoenix (med. Lat.): port near Finike,
Turkey, ruined.
Phygela Fuvgela (cl. & Byz. Gr.), Figila, et var. (med. Lat.):
town and port, Kuscadasi, Turkey.
Picenum Picenum (cl. Lat.), Pikhnoiv, et var. (cl. Gr.): region
around Ancona, NE Italy.
Portus See Ostia.
Praeneste Praivnesto"/Praivneste (cl. Gr.), Praeneste (cl. &
med. Lat.): town, Palestrina, Italy.
Preslav the Presqlabivtza (Byz. Gr.): town in the Danube
Little delta, now abandonned, near Tulcea, Bulgaria.
Proikonne2sos Proikovnnhso" (cl. & med. Gr.), Proconnesus (cl. &
med. Lat.): Marmara island, Turkey.
Pylai Puvlai (Byz. Gr.), Yalaqa2ba2d (Ar.): town, Yalova,
Gulf of I‹zmit, Turkey.
Pylos Puvlo" (cl. Gr.), Pylus (cl. Lat.), Pylus/Ionchium/
Navarinum (med. Lat.): town and gulf (Pilos) on
the W coast of the Peloponne2sos, Greece.
Ragusa ÔRagouvsion (Byz. Gr.), Ragusium, et var. (med.
Lat.): port and city founded from Epidauros
(q.v.), capital of a thema†, Dubrovnik, Croatia.
Rhaidestos Bisavnqh (cl. Gr.), ÔRaidestov" (Byz. Gr.), Bisanthe
(cl. Lat.), Rhaedestus/Bisanthe, et var. (med.
Lat.): town and port, Tekirdag°, Turkey.
Rhoiteion To; ÔRoivteion (cl. Gr.), Rhoeteum (cl. Lat.); town and
xxxvi SELECTIVE GAZETEER
promontory at the entrance to the Dardanelles,

near It-ghelmes, Turkey.
Romania Romania (med. Lat.), ÔRwmaniva (Byz. Gr.):
synonym for the Byzantine Empire, applied in
the Latin West to all territories that belonged, or
had belonged, to the Empire.
Ru2m Al-Ru2m (Ar.), from Byz. Gr. Rwmai'oi, Romans.
Used in Arabic mainly for the lands and the
people of the Byzantine Empire but also for
those of the Christian West.
Saepinum Saepinum (cl. Lat.), Saivpinon (cl. Gr.): town on the
E slopes of the Monte Matese, abandoned in
favour of Sepino, Italy.
Sagrajas Sacralias (med. Lat.), Zallaque (med. Spanish), Al-
Zalla2qa (Ar.): battlefield on the river Guerrero, 8
kilometres NE of Badajoz, Spain.
Salo2nes Salw'na (cl. Gr.), Savlwne" (Byz. Gr.), Salona/
Salonae (cl. & med. Lat.): town and port in
Illyricum, abandoned for Split in the 7th century,
revived as Solin, Croatia.
Samosata Samovsata (cl. & Byz. Gr.), Samosata (cl. & med.
Lat.), Sumaysa2t6 (Ar.): town, Samsat, Turkey.
Se2lymbria Shlumbriav, et var. (cl. & Byz. Gr.), Selymbria, et
var. (cl. & med. Lat.): port, Silivri, Turkey.
Senogallia Senagavllika (cl. Gr.), Senogavllia (Byz. Gr.),
Senogallia, et var. (cl. & med. Lat.): town N of
Ancona, Senigallia, Italy.
Septimania Septimania (med. Lat.): name of the Mediterranean
coastal region held by Visigoths after the battle
of Vouillé, from the establishment there of the
Roman Seventh Legion by Augustus.
Sigeion Sivgeion (cl. Gr.), Sigeum (cl. Lat.): promontory and
town on the S coast of the mouth of the
Dardanelles, ruined.
Sophon Boavnh livmnh (cl. Gr.), Sunonensis lacus (cl. Lat.),
Sophon lacus (med. Lat.); lake Sapança, Turkey.
Stenon, the To; Stenovn (Byz. Gr.); “the Straits”, the Bosporos.
Stilo, Punta di Cocinthus (cl. Lat.), Cocintum promontorium (med.
Lat.), Sth'lai (Byz. Gr.): cape, Punta di Stilo,
Calabria, Italy.
Strobilos Strovbilo" (Byz. Gr.) Strovilus/Strobilus et var.
SELECTIVE GAZETEER xxxvii
(med. Lat.): fortress and port on the SW coast of

Asia Minor, Aspat or Çifut Kalesi, Turkey.
St Symeon Sancti Symeonis Portus, et var. (med. Lat.), al-
Suwaydı3ya (Ar.): town and port, Samandag°,
Turkey.
Sybota Suvbota (cl. Gr.), Sybota (cl. Lat.): island in the
Straits of Corfu, Syvota, Greece.
T4abarqa Qavbraka (cl. Gr.), Thabraca (cl. Lat.), T4abarqa
(Ar.): town and port, T4abarqa, Tunisia.
Ta2hart Ta2hart (Ar.): town founded by ‘Abd Alla2h ibn
Rustam about 15 kilometres from ancient
Tingartia, W. Algeria.
Tahu2da Thabudeus (Lat.), Tahu2da (Ar.): battlefield S of
Biskra, Algeria.
Tenedos Tevnedo" (cl. & Byz. Gr.), Tenedos/Tenedus (cl. &
med. Lat.): island, Aegean Sea, Bozca Ada,
Turkey.
Thebes Qh'bai (cl. & Byz. Gr.), Thebae (cl. & med. Lat.):
chief city of Boeotia, Thivai, Greece.
Themetra Themetra (cl. Lat.): Chott Maria, Tunisia.
Thermopylae Qermopuvlai (cl. & Byz. Gr.), Thermopylae (cl. &
med. Lat.): pass leading from Thessaly to Locris,
fortifications restored by Justinian I, Greece.
Thrake2sio2n Qrakhsivwn (Byz. Gr.): thema† in W Asia Minor,
W of Anatolikon, S of Opsikion, and N of
Kibyrrhaio2tai, capital Cho2nai, Honaz, Turkey.
Thugga Thugga (cl. Lat.): Douga, Tunisia.
Tingitania Tingitania/Mauritania Tingitana (cl. Lat.), Maure-
tania Tingitana (med. Lat.): prov. of R. North
Africa, N Morocco.
Tinnis Qinnhsov" (cl. Gr.), Tenessus (cl. Lat.), Tanis, et var.
(med. Lat.), Tinnı3s (Ar.): town on an island in
Lake Manzala, Egypt, ruined.
Tmutorakan Ta; Mavtraca (Byz. Gr.), Matrega/Tmutaracha, et
var. (med. Lat.): town and region on N coast of
the Black Sea E of the Straits of Kerch, region of
Krasnodar, Ukraine.
Trajan’s Gates Succorum angustiae (late Lat.), ÔAgivou Basileivou
kleivsoura/Basilikhv puvlh (Byz. Gr.), Traiani
porta/Sancti Basilii porta (med. Lat.): pass on
the Via militaris between Ikhtiman and
xxxviii SELECTIVE GAZETEER
Pazardc°ik, Vratnik, Bulgaria.

Tricamaron Trikamavron (Byz Gr.): battlefield, geographical
hapax legomenon mentioned by Prokopios, in
Africa (q.v.).
Tripolitana Trivpoli" (cl. & Byz. Gr.), Tripolitana (cl. & med.
Lat.): prov. of R. North Africa, W Libya.
Tziliapert Tziliavpert (Byz. Gr.): unkown, possibly a
corruption of To; Pai?perte, perhaps Üçyol
(Turkish), Gjuljabert (Georgian), on the
Turkish/Georgian border.
Utica ÔH Ituvkh (Gr.), Utica (cl. & med. Lat.): Phoenician
settlement, later Roman colony, in Africa (q.v.),
near the mouth of the river Merjeda, Gulf of
Tunis, NW of Carthage (q.v.), ruined.
Versinikia Bersinikiva (Byz. Gr.): battlefield N of Edirne, near
Malamirovo, Bulgaria.
Volubilis Oujoloubiliv" (cl. Gr.), Volubilis (cl. & med. Lat.):
Roman colony and town on the river Sebou in
Mauritania Tingitana, Morocco, ruined.
Zenta (Zeta) and Zevnta and Stavmno" (Byz. Gr.): region usually
Stamnos identified with Duklja (q.v.), Serbian toparchia†,
Serbia.
Zichia Ziciva/Zhkciva (Byz. Gr.), Zechia/Zichia et var.
(med Lat.): N-E Black Sea region, separated
from Tmutorakan (q.v.) by the river Kuban.
SELECTIVE GLOSSARY OF GREEK, LATIN, AND ARABIC
TERMINOLOGY
Abbreviations and cross references
Ar. Arabic p. cl. post classical

Byz. Byzantine pl. plural
cl. classical q.v. quod vide, which see
esp. especially R. Roman
Gr. Greek sing. singular
Lat. Latin W. Western
lit. literally * See Gazeteer
‘abd (Ar.) [pl. ‘abı3d]: lit. “slave”; in Ifrı3qiya* and

Egypt, black soldiers, originally bought as
slaves.
‘ahd (Ar.): covenant; esp. Muslim covenant of
peace with non-Muslim power.
akation/akatenarion ajkavtion/ajk atenavrion [pl. ajkavtia/ajkatenav-
ria]: diminutive derivative synonyms for
akatos (q.v.).
akatos a[kato" (cl. & Byz. Gr.): light merchant galley.
akrite2s ajkrivth" (Byz. Gr.) [pl. ajkrivtai/akritai]: from
akra, “edges, border regions”; troops
stationed in, or inhabitants of, frontier
territories.
Al-H4ajj (Ar.): pilgrimage to Mecca, obligatory at least
once in life for Muslims able to accomplish
it, one of the five “pillars” of Islam.
amı3r (Ar.) [pl. umara2’]: lit. “commander”, usually in
the Middle Ages, a military commander of
various rank, also a title.
amı3r al-juyu2sh (Ar.): commander of the armies.
Amı3r al-Muslimı3n (Ar.): “Commander of the Muslims”, title of
Almoravid rulers.
aple2kton a[plhkton (Byz. Gr.) [pl. a[plhkta/aple2kta]: lit.
“fortified camp”, staging area, rendezvous
for forces, magazine where resources were
stockpiled.
xl GREEK, LATIN, AND ARABIC TERMINOLOGY
Arithmos Ariqmov": in Byz. Gr. the regiment of imperial

guards also known as the Bivgla/Vigla.
ase2kre2tis aJshkrh'ti" (Byz. Gr.) [sing. & pl.]: upper
echelon of imperial secretaries in the 6th-
12th centuries.
bourgesioi bourgevsioi (Byz. Gr.): legal order or category
created by Manuel I Komne2nos for
foreigners permanently domiciled in the
Empire.
but6sa (Ar.) [pl. but6sa2t]: used in the 12th-13th
centuries for large merchant sailing ships of
the type referred to in contemporary W.
sources as naves (q.v.).
Caesar Kai'sar: in Byz. Gr. a dignity applied to junior
emperors, often emperors’ sons but also
others on occasions.
cattus/gatus (med. Lat.) [pl. catti/gati]: type of galley
referred to commonly in W. sources of the
late 11th and early 12th centuries, from the
Ar. qit6‘a (q.v.).
chrysobull crusovboullon (Byz. Gr.): solemn docucument
bearing the imperial gold bulla, seal.
comes see kome2s.
comes Africae late imperial provincial administrator of the
province of Africa.
da2‘ı3 (Ar.): lit. “caller/summoner”, missionary pro-
pagandist; esp. spreading Isma2‘ı3lism.
die2re2s dihvrh" [pl. dihvrei"/die2reis]: in cl. Gr. a “two”
or bireme, rowing two files of oars each
side; in Byz. literature a literary affectation
for a warship but, occasionally, with appar-
ent reference to an actual number of oar
banks.
domestikos to2n domevstiko" tw'n scolw'n (Byz. Gr.): command-
scholo2n er of the regiment of the scholae, the imper-
ial guard.
doux douvx (Byz. Gr.) [pl: doukavde"/doukades]: esp.
military commanders and provincial
governors from the 11th century.
droungarios drouggavrio" (Byz. Gr.) [pl. drouggavrioi/
droungarioi]: high military rank in the 7th-
GREEK, LATIN, AND ARABIC TERMINOLOGY xli
8th centuries, esp. naval commands.

droungarios tou ploi- drouggavrio" tou' ploi?mou/ drouggavrio" tw'n
mou/droungarios ploi?mw'n (Byz. Gr.): droungarios of the
to2n ploimo2n ship(s), of the imperial fleet at Constan-
tinople.
Eidikon Eijdikovn (Byz. Gr.): etymology disputed,
imperial treasury and warehouse.
eparchos of the city e[parco" th'" povlew" (Byz. Gr.): governor of
Constantinople, responsible for law and
order, justice, and commerce.
exarchos e[xarco" (Byz. Gr.) [pl. e[xarcoi/exarchoi): lit.
“leader”, governor of certain provinces in
the 6th-8th centuries, holding both civil and
military power, esp. of Ravenna and
Carthage*, hence ejxarciva/exarchia, “exar-
chate”.
geniza genı3za (Hebrew): storehouse, depository for
documents bearing name of God, in parti-
cular that attached to the synagogue in al-
Fust6a2t6*.
Ghassa2nı3 (Ar.): from the Banu2 Ghassa2n, south Arabian
tribe which settled around Damascus in the
3rd and 4th centuries C.E.
ghura2b (Ar.) [pl. aghriba]: used for oared galleys of
indeterminate types, often interchangeably
with shı3nı3 (q.v.).
golafrus/garabus (med. Lat.) [pl. golafri/garabi]: name for W.
galleys of some kind in the late 11th and
early 12th centuries, from the Ar. ghura2b
(q.v.)
gumbaria (med. Lat.) [pl. gumbariae]: Latinization of the
Byz. Gr. koumb(p)arion (q.v.) for a war
galley.
h5a2jib (Ar.): lit. “chamberlain”, superintendant of a
place, head of government, chief minister;
esp. in al-Andalus*.
h5arra2qa (Ar.) [pl. h5arra2qa2t]: used in a variety of
contexts for light vessels of various kinds,
often river boats and pleasure craft, but also
for warships, fire ships in particular.
hippago2gos iJppagwgov" (cl. & Byz. Gr.) [pl. iJppagwgoi;/
xlii GREEK, LATIN, AND ARABIC TERMINOLOGY
hippagogoi]: horse transport ship.

ima2m (Ar.): religious leader, in particular the ima2ms
of the Shı3‘a (q.v.).
jiha2d (Ar.): lit. “striving”, in particular, here, holy
war to extend the boundaries of the da2r al-
Isla2m, the world ruled by Muslims.
jizya (Ar.): poll tax levied on non-Muslim men
under Muslim rule and recognized
religiously as dhimmı3, peoples of the book.
jund (Ar.): esp. in al-Andalus, a regiment attached
to a ku2ra (q.v.).
Karabisianoi Karabisiavnoi (Byz. Gr.): lit. “those of the
kavraboi” (q.v.), first regular fleet of the
Byzantine Empire.
karabion/karabos karavbion/ kavrabo" (Byz. Gr.) [pl. karavbia(oi)/
karabia(oi)]: word for a war galley which
first appeared in Muslim Egypt but then
spread in the Byz. Empire.
kate2na kath'na (Byz. Gr.) [pl. kath'nai]: possibly the
same as akation/akatenarion (q.v.), etymo-
logy otherwise unknown, possibly Ar. qit6‘a,
transport vessel; used by Theophane2s the
Confessor for Muslim transport ships.
katepano2 katepavnw (Byz. Gr.) [sing. & pl.]: lit. “the one
above”, used in the 11th century, esp., for
governors of major provinces, esp. Italy,
Mesopotamia*, and Bulgaria.
Kha2rijı3] (Ar.) [pl. Khawa2rij]: lit. “seceders”, hearkened
back to a supposed purity of the age of
Caliph ‘Umar ibn al-Khat6t6ab2 , believing that
succession to the Caliphate should be
decided by God alone.
kleisourarche2s kleisouravrch" (Byz. Gr.) [pl. kleisouravrcai/
kleisourarchai]: commander of a kleisoura,
territory smaller than, sometimes part of, a
thema (q.v.).
kome2s comes (Lat.), kovmh" (Byz. Gr.) [pl. kovmhte"/
kome2tes): late imperial title for officials of
various ranks with assignments, sometimes
military, often linked to the imperial family;
various functionaries in later times, also
GREEK, LATIN, AND ARABIC TERMINOLOGY xliii
commander of a squadron of a fleet.

koubikoularios koubikoulavrio" (Byz. Gr.): general term for a
palace eunuch waiting upon the emperor;
appointed to many commands, including
military ones.
koumb(p)arion koumb(p)avrion (Byz. Gr.) [pl. koumb(p)avria/
koumb(p)aria]: term for a Muslim galley,
perhaps from ghura2b (q.v.).
ku2ra (Ar.): landed district or dependent province,
from Gr. cwvra, place.
logothete2s logoqevth": lit. “auditor”, in Byz. Gr. a generic
term for a high official, head of a
department of the imperial administration.
logothete2s tou logoqevth" tou' drovvmou (Byz. Gr.): head of the
dromou department of the drovvmo"/dromos, originally
the public post but by the eighth century
concerned with foreign affairs.
logothete2s to2n agelo2n logoqevth" tw'n ajgelw'n (Byz. Gr.): head of the
department of the ajgevlai, herds, of the
imperial horses and mules.
magister militum (p. cl. Lat.): commander in chief of armies in
the later R. empire.
Magnaura Magnauvra (Byz. Gr.): ceremonial hall on the
edge of the imperial palace in Constantin-
ople.
Mahdı3 Al-Mahdı3 (Ar.): lit. “the guided one”. Among
the Shı3‘a (q.v.), the hidden ima2m (q.v.),
descendent of ‘Alı3 ibn Abı3 T4al2 ib.
maks (Ar.) [pl. muku2s]: non-Qur’a2nic tax.
malik (Ar.) [pl. mulu2k]: ruler, loosely, “king”.
mamlaka (Ar.) [pl. mama2lik]: realm, loosely, “king-
dom”.
mamlu2k (Ar.) [pl. mama2lı3k]: “one owned”, a slave.
Slaves imported from non-Muslim regions
and raised to be soldiers loyal to their
owner/commander.
Mardaites Mardai?tai (Byz. Gr.): a people, originally
from the Tauros, who became famous as
warriors in Byz. service, esp., here, in
various fleets of the naval themata (q.v.).
markab (Ar.) [pl. mara2kib]: generic term for ships in
xliv GREEK, LATIN, AND ARABIC TERMINOLOGY
general.
mawa2lı3 (Ar.): from mawla2, “client”: non-Arab converts
to Islam adopted into Arab tribes as clients.
megas domestikos mevga" domevstiko" (Byz. Gr.): supreme military
commander below the emperor, esp. in
11th-12th centuries
megas doux mevga" douvx (Byz. Gr.): admiral of the imperial
fleet from the late 11th century.
mone2re2s monhvrh" [pl. monhvrei"/mone2reis]: in cl. Gr. a
“one” or monoreme, rowing one file of oars
each side; in Byz. literature a literary
affectation for a small warship, perhaps also
actually used in practice.
muja2hidu2n (Ar.): those who wage jiha2d, both holy war and
also other forms of “striving for God”.
mulu2k al-t6a wa2’if (Ar.): t6a2’ifa, “party” “kings”; a generic, the
rulers themselves used various titles.
myoparo2n muopavrwn (cl. Gr.) [pl. muopavrwne"/myoparo2-
nes]: light pirate ship.
navis [pl. naves]: in med. Lat. a generic for ships in
general but having the specific sense of
round-hulled, lateen-rigged, sailing ships in
the 12th-14th centuries.
nipsistiarios niyistiavrio" (Byz. Gr.): originally an official
whose function was to hand the emperor a
basin in which to wash his hands, became a
title.
Oghuz Ghuzz (Ar.), Ou\zoi (Byz. Gr.): the Oghuz
Turks, confederations of Turkish peoples
from S of the Aral Sea.
ostiarios ojstiavrio" (Byz. Gr.), ostiarius (Lat.):
doorkeeper, originally an office whose
function was to introduce dignitaries to the
emperor, became a title.
parakoimo2menos parakoimwvmeno" (Byz. Gr.): lit. “sleeping at
the side [of the emperor]”, high
chamberlain, guardian of the emperor’s
bedchamber, head of the civilian
government in the tenth century.
paria (med. Spanish), from med. Lat. pariare, to pay
a fine; tribute paid by one state to another in
GREEK, LATIN, AND ARABIC TERMINOLOGY xlv
return for protection or non-aggression.

patrikios patrivkio" (Byz. Gr.), patricius (Lat.): high
ranking dignity, not an office, granted to
important generals and governors.
pente2konteros penthkovntero" [pl. penthkovnteroi]: in cl. Gr., a
fifty-oared galley; in Byz. literature a
literary affectation.
phortago2goi fortagwgoiv (Byz. Gr.): lit. “load carriers”,
transport ships.
phorte2goi forthgoiv (Byz. Gr.): lit. “load carriers”,
transport ships.
ploion ploi'on (cl. & Byz. Gr.) [pl. ploi'a/ploia]:
generic for ships in general.
ploia makra ploi'a makrav: literary synonym in Byz. Gr. for
“long ships”, war galleys.
praefectus praivf ekto" (Byz. Gr.), praefectus (Lat.): title
conferred on local rulers.
praefectus Aegypti title of governors of Egypt under the early R.
empire.
praetorian prefect praefectus praetorio (Lat.), e[parco" tw'n
praitwrivwn/eparchos to2n praito2rio2n (cl. &
Byz. Gr.): originally commander of the
imperial bodyguard, but a high-ranking civil
functionary from the 4th century, deputy to
the emperor.
proedros provedro": in Byz. Gr. a dignity rather than an
office, ranking with the office of proedros
[president] of the Senate.
pro2tase2kre2tis prwtashkrh'ti" (Byz. Gr.): head of the
secretaries, ajshkrh'ti"/ase2kre2tis (q.v.) of the
imperial chancery.
pro2tosebastos prwtosevbasto" (Byz. Gr.): lit. “first sebastos”,
high title conferred on close relatives of the
emperor, esp. under the Komne2noi
emperors.
pro2tospatharios prwtospaqavrio" (Byz. Gr.): lit. “first sword-
bearer”, first of the spatharioi, dignity
awarded to military commanders, governors
of themata (q.v.), foreign princes, and
others. Varied in significance over the
centuries.
xlvi GREEK, LATIN, AND ARABIC TERMINOLOGY
pro2tospatharios te2s prwtospaqavrio" th'" fiavl h" (Byz. Gr.):

phiale2s pro2tospatharios of the “basin”: judge of the
oarsmen of the imperial ships, probably
from the “basin” of the small harbour at
Boukoleon, S of the imperial palace and the
Hippodrome on the Sea of Marmara.
qa2’id (Ar.): military commander of various rank, in
al-Andalus frequently commander in chief,
admiral of a fleet.
qa2rib (Ar.) [pl. qawa2rib]: sometimes a small boat or
ship’s boat but also a generic for sailing
ships in general.
qit6‘a (Ar.) [pl. aqt6a2‘]: word used esp. for war
galleys but also for transport ships, both
galleys and sailing ships.
Rho2s ÔRw'"//ÔRwv" (Byz. Gr.): Scandinavians settled in
the Ukraine, especially along the Dnepr
river and around Kiev.
riba2t6 (Ar.) [pl. riba2t6a2t]: fortress/“monastery” on the
frontiers of the Muslim world where
military and religious duties were
performed.
safı3na (Ar.) [pl. sufun]: common word for ships in
general, esp. transport ships.
sage2na saghvna (Byz. Gr.) [pl. saghvnai/sage2nai]:
etymology unknown, possibly from saghvnh,
a fishing net. Used for small ships of
Muslim or Dalmatian pirates but by the
Strate2gikon of Maurice for a Byzantine
galley of some kind.
saio sagio, et var. (p. cl. & med. Lat.): royal retainer
in the Ostrogothic kingdom, agent of the
king.
sa(k)toura Sa(k)touvra (Byz. Gr.) [pl. sa(k)touvra/sa(k)-
tourai]: term for Cretan ships used in the
Vita Basilii of the Thephane2s continuatus,
probably from Ar. shakhtu2r, some kind of
small ship.
sandanum (med. Lat.): from Byz. Gr. celavndion/
chelandion, W. transport galley in the 12th-
13th centuries.
GREEK, LATIN, AND ARABIC TERMINOLOGY xlvii
sekreton sevkreton (Byz. Gr.): a department of the

imperial administration.
shalandı3 (Ar.) [pl. shalandiyya2t]: term used frequently
in Muslim sources for Byz. galleys, then
emulated by the Muslims, esp. in Egypt;
from Byz. Gr. chelandion.
sharı3‘a (Ar.): from the root shara‘a, “prescribe”, the
revealed law of Islam as found in the
Qur’a2n, the sunna (q.v.), and the schools of
law.
Shı3‘a (Ar.): from “shı3‘at ‘Alı3”, “party of ‘Alı3”,
Muslims believing (among other things)
that the Caliphate should have descended
through the line of Muh5ammad’s son-in-
law, ‘Alı3 ibn Abı3 T4a2lib.
shı3nı3 (Ar.) [pl. shawa2nı3]: common generic term for a
war galley.
skevophoron skeuofovron (Byz. Gr.) [pl. skeuofovra/skevo-
phora]: lit. “carrying vessel”, supply ship.
skevos, barytera skeu'o", baruvtera (Byz. Gr.) [pl. skeuvh,
baruvtera/skeve2, barytera: (lit.) “vessel of
burden”, cargo ship.
spatharios spaqavrio" (Byz. Gr.): lit. “sword-bearer”, a
dignity and title rather than an office.
strate2gos strathgov" (Byz. Gr.) [pl. strathgoi;/strate-2
goi]: cl. Gr. for a military commander, by
the 8th century applied to governors of
themata (q.v.), holding both civil and
military authority.
sult6a2n (Ar.): power or authority, secular title of ruler,
alongside Caliph.
Sunnı3 Ahl al-sunna wa ’l-ijma2‘ (Ar.), “people of the
sunna and the consensus”: Muslims
adhering to the sunna, the orthodox
example of the Prophet Muh5ammad, to one
of the four schools of law, and not
attributing any importance to the descend-
ants of ‘Alı3 ibn Abı3 T4a2lib.
tarida (med. Lat.) [pl. taride]: W. transport galley,
especially for horses, of the 12th and 13th
centuries, from the Ar. t6arı3da.
xlviii GREEK, LATIN, AND ARABIC TERMINOLOGY
t6arı3da (Ar.) [pl. t6ara2’id]: transport galley, esp. for

horses.
tetre2re2s tetrhvrh": in cl. Gr. a “four” or quadrireme,
rowing four files of oars each side.
thema Qevma (Byz. Gr.) [pl. qevmata/themata]: territory
and army administered by a strate2gos (q.v.),
also a squadron of a fleet.
toparche2s topavrch": in Byz. Gr. a term for military and
civil authorities, by the 10th century used
for independent, neighbouring rulers.
topote2re2te2s topothrhthv": in Byz. Gr. deputy governor,
lieutenant commander, esp. “port admiral”?
tourmarchai to2n tourmavrcai tw'n ploi?mwn (Byz. Gr.): lieutenant
ploimo2n commanders of the fleet of the Stenon*,
under the droungarios to2n ploimo2n (q.v.).
tourmarche2s tourmavrch" (Byz. Gr.) [pl. tourmavrcai/tour-
marchai): military commander of a tourma,
regiment of a thema (q.v.), also a naval
commander.
triachonte2re2s triaconthvrh": in cl. Gr. a thirty-oared galley;
in Byz. literature an affectation.
trie2re2s trihvrh" [pl. trihvrei"/trie2reis]: in cl. Gr. a
“three” or trireme, rowing three files of oars
each side; in Byz. literature an affectation
for a warship.
Vestiarion basilikon Bestiavrion basilikovn (Byz. Gr.): lit. “imperial
wardrobe”, warehouse, arsenal for military
forces.
GLOSSARY OF ENGLISH NAUTICAL TERMINOLOGY1
abeam: across a ship, at right angles to the centre line (q.v.)

from stern (q.v.) to bow (q.v.).
aft: towards the stern (q.v.), in the after part of a ship.
ahead: of the wind, from ahead, blowing onto the bows
(q.v.).
amidships: in the middle or “waist” of a ship.
astern: of the wind, from astern, blowing onto the stern
(q.v.).
back water: to put a galley into reverse by pulling (q.v.) the oars
in reverse, from stern (q.v.) to bow (q.v.)
backstay: rope bracing a mast against forces pushing it forward
(q.v.) and running from the masthead (q.v.) to
some point aft (q.v.).
bank, of used here for any arrangement of oarsmen on the
oarsmen: same horizontal level.
beam: (1) a transverse timber serving to brace the hull (q.v.) of a
ship apart. See also through beams.
beam: (2) transverse width of a ship at the widest part amiships
(q.v.).
beat to to sail to windward by a series of alternate tacks
windward: (q.v.).
bench: bench or seat on which an oarsman sat. Cf. thwart.
bench used here for the same position abeam (q.v.) of any
position: number of benches (q.v.) in a galley (q.v.),
irrespective of vertical arrangement.
bend (to): to attach a rope to something by means of a “bend”, a
“knot” in general usage. Esp., to bend a sail to a
yard.
bilges: where the floor timbers (q.v.) turn upwards and the
angle of the frames (q.v.) moves from more than
45˚ to less than 45˚ to the vertical.
bilge keel: small keel parallel to the central keel (q.v.) of a ship
at or just below the turn of the bilge (q.v.).
bilge water: water collecting in the bilges (q.v.) from rain, seas
washing over the decks, or seepage through the
------------------------------
1
In this Glossary we have used extensively, Kemp, Ships and the sea.
l ENGLISH NAUTICAL TERMINOLOGY
seams (q.v.) between the strakes (q.v.).

bilge pump: one of many different varieties of pumps for
extracting bilge water (q.v.).
bireme: used here for a galley (q.v.) rowing two oars from
each bench (q.v.) position, irrespective of any
vertical arrangement of the benches or of the
number of oarsmen on the same bench.
block: a shell or casing in which sheaves (q.v.) or pulleys
(q.v.) are fitted, over which ropes run. Blocks
consist of a shell, one or more sheaves, the
swallow (q.v.) or hole in which the sheaves are set,
and pins or spindles on which the sheaves turn.
blockmast: section of a masthead (q.v.) containing a block (q.v.)
for raising and lowering the yard (q.v.) by a
halyard (q.v.).
bow(s): the forward (q.v.) end of a ship.
brails: ropes sewn into the foot of a square sail and running
through fairleads (q.v.) to its head at the yard (q.v.)
by which the sail can be taken up to reduce its
plane area in heavy weather.
brow(s): sloping ramps or gangways leading up or down from
one deck to another or from a deck into a hold for
the movement of animals from above to below or
vice versa.
bulwark: low wall of frames (q.v.) and planking (q.v.) above
the deck to prevent water coming inboard and
seamen being washed overboard.
bung (and a hole in the floor (q.v.) of a ship through which bilge
bung hole): water (q.v.) can be drained off when the ship is out
of water. When afloat, the bung hole is stopped up
with a wooden plug or bung.
buoy: a float used to mark the position of a submarine
object.
camber of a curvature of the deck downwards from the centre line
deck: (q.v.) towards the hull (q.v.), both to strengthen the
deck and also to promote the draining off of water
shipped inboard.
carling: fore-and-aft (q.v.) timber inserted between frames
(q.v.).
cathead: heavy timber fitted with sheaves (q.v.), over which
the anchor cables (or cat tackles on large ships)
ENGLISH NAUTICAL TERMINOLOGY li
run, projecting out over both bow quarters of ships

for raising and letting-go the anchors.
caulking: any material (flax, oakum, tow, pitch, etc.) used to
coat the outside hull (q.v.) of a ship or to fill the
seams (q.v.) between strakes (q.v.) to seal them
against water penetration.
centre line: line of a ship down the centre from the sternpost
(q.v.) to the stempost (q.v.).
cowl: a ventilating shaft with bell-shaped top above deck
which can be turned to catch the breeze from any
direction. If turned away from the wind, it can be
used as an exhaust.
crab (to catch when an oarsman loses control of an an oar when
a): lifting the blade from the water at the end of the
stroke and it remains caught in the water.
crutches: crescent-shaped holders set on posts on the deck
down the centreline (q.v.) of a ship to take masts
and yards (q.v.) when lowered.
cutwater: section of the stempost (q.v.) at the waterline (q.v.).
deadweight the difference between the tonnage displacement of a
tonnage: ship fully fitted out and the maximum tonnage
displacement she can carry with an additional full
cargo or military equipage.
fairlead: any ring or other device used to lead a rope in a
desired direction.
file, of used here for a row of oarsmen from stern (q.v.) to
oarsmen: bow (q.v.).
flared: of a hull (q.v.), curved outwards; of a stempost (q.v.),
raked (q.v.) forward (q.v.).
floor: the flatest section of the hull (q.v.) and frames (q.v.)
of a ship, running from the keel (q.v.) to the curve
of the bilges (q.v.).
floor timbers: the lowest futtocks (q.v.) of the frames (q.v.) of a
ship, fitted to the keel (q.v.).
fluke: of an anchor, triangular spade at the end of each arm,
which digs into the sea bed.
fore-and-aft: in a line from stern to bow.
forecastle: used here for any superstructure built over the
stempost (q.v.) of a ship at the bows (q.v.).
foremast: mast at the bows (q.v.).
forestay: rope bracing a mast against forces pushing it aft and
lii ENGLISH NAUTICAL TERMINOLOGY
running from the masthead (q.v.) to some point

forward (q.v.).
forward: towards the bows (q.v.), in the fore part of a ship.
frame: timber or rib (q.v.) of a ship running from the keel
(q.v.) to the gunwale (q.v.). On all large wooden
ships frames were composed of several timbers
known as futtocks (q.v.).
freeboard: distance from the waterline (q.v.) to the deck.
furled: of a sail, rolled up and secured to the yard (q.v.).
futtock: any sectional timber comprising part of a frame (q.v.)
or rib (q.v.).
galley: used here throughout as a generic for oared ships of
every kind, except where the specific type of
galeva/galea is discussed.
garboard first planking (q.v.) or strake (q.v.) of the hull (q.v.)
strakes: rabbeted (rebated) to the keel (q.v.).
gearing ratio the ratio of an oar’s length from the mid-point of the
of an oar: oarsman’s hands on the handle to the thole to that
from the thole to the centre of water pressure on
the blade.
gripe (to): to come up into the wind when beating to windward.
grommet: of an oar, ring of twisted leather or cordage used to
attach an oar to its thole pin (q.v.).
gunwale: the term is a modern one associated with the guns of
broadside sailing ships. However, it has become
generalized in usage for stringers (q.v.) which
cover the heads of the frames (q.v.) of small boats.
We have used it in relation to dromons for
stringers which covered the heads of the bulwarks
(q.v.).
half-deck: small deck at the bow (q.v.) and/or stern (q.v.) of an
otherwise open boat.
halyard: generic term for a rope or series of ropes and tackles,
such as ties (q.v.) and tie tackles (q.v.), by which a
yard (q.v.) is raised and lowered.
hatch: opening in a deck through which access to the hold
(q.v.) is provided.
heel of a mast: lower end of a mast, usually squared off to fit into the
mast step (q.v.).
heel (to): (of a ship) to lean over to one side.
hog (to): when the bow and stern of a ship droop when the
ENGLISH NAUTICAL TERMINOLOGY liii
midship section is on the crest of a wave and the

weight of the bow and stern causes the hull to flex,
it is said to hog. Cf. sag.
hold: in general terms, the space inside the hull (q.v.) below
deck.
hull: the “skin” of a ship, consisting of frames (q.v.) and
planking (q.v.).
keel: the lowest timber (or composite of timbers) of a ship,
extending from sternpost (q.v.) to stempost (q.v.).
The whole ship is built upwards from the keel.
keelson: stringer (q.v.) laid over the floor timbers (q.v.) and
keel (q.v.) to provide fore-and-aft rigidity and to
lock the floor timbers to the keel.
knot: one nautical mile (6,080 feet), or 1.15 English statute
miles, per hour.
lee: lee shore: a shore or coast onto which the prevailing
winds blow directly, downwind from any ship off
it and thus dangerous.
leech: the trailing or after edge of any fore-and-aft sail
(q.v.). The outer edges of any square sail (q.v.).
leeway: the distance a ship slides to leeward from her true
course as the wind, or tide or current, forces her to
slide sideways through the water.
lifts: ropes running from mastheads (q.v.) to the ends of
yards (q.v.) on ships with square sails (q.v.) to
control their angle to the horizontal.
limber (holes): holes cut in the underside of floor timbers (q.v.) close
to the keel (q.v.) to allow bilge water (q.v.) to flow
to a sump (q.v.) at the lowest part of the keel.
long boat: a ship’s main auxiliary boat, usually towed behind.
luff: the leading or forward edge of any fore-and-aft (q.v.)
sail.
luff up (to): to turn the bows (q.v.) up in to the wind.
marine: general term used here for any soldiers who fought at
sea. There is no reference here to the type of
specialized soldiers trained for service at sea in the
English and other navies from the 17th century.
Such specialized marines did not exist in the
Middle Ages. Some oarsmen also doubled as
marines.
mast step: framework of timber built on the keel (q.v.) and
liv ENGLISH NAUTICAL TERMINOLOGY
keelson (q.v.), forming a receptacle to take the heel

(q.v.) of a mast.
masthead: uppermost section of a mast, in which one or more
blocks (q.v.) could be set to raise and lower the
yard (q.v.).
midships a mast located amidships (q.v.) in the waist of a ship.
mast:
mizzen mast: a third mast, usually small, right aft at the stern (q.v.).
monoreme: a galley (q.v.) rowing one oar from each bench (q.v.)
position.
moor (to): to anchor a ship with two or more anchors to
minimize the swing of the ship with wind or tide.
mortise and method of joining planking (q.v.) or strakes (q.v.)
tenon together edge to edge by chiselling rectangular
joints: holes (mortises) in both planks and inserting a
piece of wood (tenon) into each, hammering the
strakes together, and holding the tenon in place
with treenails (q.v.) inserted into holes bored
through the strakes and tenons.
oar ports: holes cut in the planking (q.v.) of the hull (q.v.)
through which oars are rowed.
oar thongs: grommets (q.v.), rings of twisted leather or rope used
to hold oars to thole pins (q.v.).
parrell: a hoop of multiple ropes rove through circular balls
(trucks) and flat spreaders (ribs) to hold a yard
(q.v.) against a mast.
part deck: narrow deck extending only part way from the hull
(q.v.) towards the centre line (q.v.) of a ship.
pitch (to): ships pitch when waves lift their bows and then travel
towards the sterns, thus lifting those and causing
the bows to pitch down.
planking: planks making up the strakes (q.v.) of the hull (q.v.)
of a ship which form its “skin”.
point of a one-eighth division of a quarter of a circle: 11
sailing: degrees, 15 minutes (11.25˚).
poop: general term for the stern (q.v.) of a ship, including
the sterncastle (q.v.) and sternpost (q.v.).
port: facing from stern (q.v.) to bow (q.v.), the left-hand
side of the ship.
ports: generally speaking, holes cut in the hull of a ship,
especially for loading and entry purposes.
ENGLISH NAUTICAL TERMINOLOGY lv
prow: general term for the bows of a ship, including the

forecastle (q.v.) and stempost (q.v.).
pull (to): to row an oar, to “pull” it through the stroke. Hence, a
“pull” under oars.
pulley: see sheave.
quarters: the two parts of the stern (q.v.) either side of the
centreline.
raked: angled.
ram, waterline projection at the bow of a ship at the waterline, made
of wood sheathed in bronze, used on classical war
galleys to fracture the hull (q.v.) of enemy vessels.
rib: see frame.
roll (to): a ship rolls from side to side as she passes over waves
unless they are dead ahead or astern.
rudders, sometimes called “steering oars”, rudders on long
quarter: shafts mounted on both of the stern quarters of
ancient and medieval Mediterranean ships.
sag (to): when the bow and stern of a ship rise when the
midship section is in the trough of two waves and
its weight causes the hull to flex, it is said to sag.
Cf. hog.
sail, lateen: triangular, fore-and-aft (q.v.) sail, whose luff (q.v.) is
set on a yard (q.v.). The forward end of the yard is
hauled down so that it is oblique to the mast.
sail, square: rectangular sail suspended from a yard (q.v.) set
square to the mast, abeam (q.v.) the ship.
scuppers: drainage holes in bulwarks (q.v.) allowing water on
the deck to drain away over the side.
seam: gap between planking (q.v.) or strakes (q.v.).
shaft of an the cylindrical section of an oar between the blade
oar: and the loom (handle).
sheave: circular wheel that revolves in a block (q.v.) and
which is grooved to take the rope that runs over it.
shell con- method of construction of the hull (q.v.) of a ship in
struction: which the planking (q.v.) is set up first, before the
frames (q.v.) are added. Cf. skeleton.
shores: also known as “legs”: timbers used to hold a ship
upright when aground.
skeleton con- method of construction of the hull (q.v.) of a ship by
struction: building a skeleton of keel (q.v.), stempost and
sternpost (q.v.), frames (q.v.) and stringers (q.v.)
lvi ENGLISH NAUTICAL TERMINOLOGY
first, and then attaching the planking (q.v.) to the

frames. Cf. shell.
sleeve of an leather seal attached to the shaft (q.v.) of an oar
oar: outboard of the hull (q.v.), and to the inside of the
hull around the oar ports (q.v.) to seal the ports
against shipping water inboard.
spur: long wooden beam, perhaps sometimes sheathed in
iron, attached to the stempost (q.v.) and suspended
by a chain or coupling from its head.
stand-and-sit rowing stroke in which oarsmen rise to a semi erect
stroke: position at the commencement of the stroke and
then, bracing one foot against a footrest, pull the
oar through the stroke by falling back on to the
bench (q.v.).
starboard: facing from stern (q.v.) to bow (q.v.), the right-hand
side of a ship.
stempost: the fore-most timber of the frames (q.v.) of a hull
(q.v.) of a ship at the bow (q.v.), set on the keel
(q.v.) at its foot.
stern: the after end of a ship.
sterncastle: used here for any superstructure built over the
sternpost (q.v.) of a ship at the stern (q.v.). Cf.
poop.
sternpost: the aftermost timber of the frames (q.v.) of a hull
(q.v.) of a ship at the stern (q.v.), set on the keel
(q.v.) at its foot.
stocks: also known as “keel blocks”, the line of blocks on
which a keel (q.v.) is laid down when a ship is
being built.
strakes: name given to each line of planking (q.v.) of a hull
(q.v.). A strake may be composed of several planks
joined together.
stringer: any timber running fore-and-aft (q.v.) in the skeleton
of a ship; especially those locking frames (q.v.)
and deck beams (q.v.) in place. Wales (q.v.) and
keelsons (q.v.) are stringers.
sump: a box set at the lowest part of the keel (q.v.) into
which bilge water (q.v.) flowed through limber
holes (q.v.) and from which it was pumped out by
a bilge pump (q.v.).
swallow: hole in the casing of a block (q.v.) in which the
ENGLISH NAUTICAL TERMINOLOGY lvii
sheave (q.v.) is set.

tack (to): to move the bows (q.v.) of a ship across the direction
of the wind when sailing into it so as to bring the
wind onto the opposite side of the ship.
tackle: a combination of two or more blocks (q.v.) and ropes
used to increase mechanical advantage when
hauling any weight.
thole: pin set in a gunwale (q.v.) or other timber to which an
oar is held by a grommet (q.v.).
through beams (q.v.) which projected through the hulls of
beams: ancient and medieval ships.
thwart: bench (q.v.) or seat on which an oarsman sat. Cf.
bench.
tie: rope attached to a yard (q.v.) and rove (passed)
through a block (q.v.) in the masthead (q.v.) and
connected to a tie tackle (q.v.), by which the entire
complex of the halyard (q.v.) is taken in or let out
to raise or lower the yard (q.v.) respectively.
tie tackles: tackles (q.v.) used to haul on the ties (q.v.) attached to
the yards (q.v.), by which the halyards (q.v.) are
taken in or let out.
tiller: handle by which a rudder (q.v.) is manœuvred.
treenails: wooden pins or dowels inserted into bored holes to
hold any two timbers together.
trireme: used here for a galley (q.v.) rowing three oars from
each bench (q.v.) position, irrespective of any
vertical arrangement of the benches or of the
number of oarsmen on the same bench.
wale: a heavy stringer (q.v.) fastened to the outside of a hull
(q.v.), esp. at points where protection from
abrasion or collision is required.
waterline: level of flotation of a ship.
weather helm: when the tiller (q.v.) of a rudder (q.v.) has to be held
to windward, to the weather, in order to prevent
the ship griping (q.v.).
weight in hand the downward force on the mid-handle needed to
of an oar raise the oar from the water and to balance it at the
pivot at the thole.
windlass: a horizontal cylinder fitted with bars to turn it, around
which an anchor cable can be wound.
windsail: a cloth funnel able to have its mouth rotated into the
lviii ENGLISH NAUTICAL TERMINOLOGY
wind so that fresh air is deflected below deck by

the spout led through a hatch (q.v.) or deck
opening of some kind.
yard: spar from which a sail is set; i.e., hung.
NOTE ON CITATION OF GREEK AND LATIN GLOSSARIES
At several points in the text, we have cited editions of some of the

extant manuscripts of Greek and Latin glossaries, word lists with
explanations of the meanings of words, which contain material
relevant to our study. Some of these were bilingual, Greek-Latin or
Latin-Greek. Some were simply Latin. In all cases, the glossaries had
complex transmission and manuscript histories and, in all instances
cited, the manuscripts referred to post-date the original compilation of
the glossaries, in some cases by many centuries.
In order to avoid tedious repetition in the text, we have given the
details of the glossaries and manuscripts here. Cross references to the
discussion here have been added to the notes at the appropriate places.
The manuscripts are discussed here not in the order in which they
appear in the text but rather in an order which best facilitated
discussion of transmission processes.
We are perfectly well aware that compilers of glossaries constantly
altered, re-arranged, and modified what they had before them.
Sometimes they worked ab initio from manuscripts or from marginal
glosses on manuscripts but more usually they worked from older
glossaries, producing “collected glossaries”. On the one hand, it is
clear that the scribes of the manuscripts of the glossaries as we have
them frequently had no idea what the words that they were glossing
had originally meant. They sometimes produced weird and wonderful
explanations, sometimes based on false etymologies. But they
themselves may not have been responsible for the loss of
understanding. That could have occurred anywhere in the process of
transmission to them. Alternatively, the meanings of words may have
changed, as they frequently do. On the other hand, even when they do
appear to have understood the antique meanings of words that they
glossed, and in those cases we have had to assume for want of
argument that they really did understand, even that may not
necessarily have been the case. They may have been simply copying
something they did not understand themselves, or they may have been
just guessing. It is rarely possible to know the date and provenance of
a gloss, whether it was that of the manuscript in which it survives, or
that of the original compilation of the glossary, or that of some point
in the transmission process between. We have used the glossaries with
lx GREEK AND LATIN GLOSSARIES
extreme caution, being fully conscious of their notorious unreliability

and the difficulty of interpretation of these sources.1
1: Rome, Biblioteca Apostolica Vaticana, MS. Vat. Lat. 3321 (=
Lowe, CLA, vol. 1, no. 15).
Written on parchment in Uncial majuscule letters, this manuscript
is generally dated to the mid eighth century with a provenance in
central Italy. Folios 2r-163r were edited in Goetz, Glossarii Latini,
vol. 4, 1-198, as Glossae Codicis Vaticani 3321.
The glossary was a copy of an earlier one, also probably produced
in Italy, which was the common ancestor of both that of this
manuscript and also that of the later tenth-century manuscript Monte
Cassino, Biblioteca dell’ Abbazia, MS. Cass. 439, which Goetz noted
in his critical apparatus.
Both glossaries contain compilations of two separate earlier
glossaries, that known as the Abolita (from its first lemma), which was
produced in Spain in the late seventh century, and that known as the
Abstrusa (again from its first lemma), which was probably produced
in France, perhaps as early as the sixth century. In Goetz’s edition, the
Abolita entries are contained within square brackets.
2: Paris, Bibliothèque Nationale, MS. Lat. 7651.
Produced in the ninth century in France, possibly at Laon, and
probably compiled from an earlier Uncial exemplar, this manuscript,
whose Latin text is in Caroline minuscule, contains the oldest extant
copy of the Latin-Greek glosses attributed to Philoxenos. It was
intended for Greeks attempting to read Latin. Folios 1r-218r were
edited in Goetz, Glossarii Latini, vol. 2, pp. 1-212 as Glossae Latino-
Graece.
Who the compiler of the Philoxenos was is unkown, but he was not
the consul of 525 C.E. He was probably a monk in an Italian, Greek-
literate monastery. He had access to a copy of the Ars grammatica of
Flavius Sosipater Charisius (fl. ca 375), so the earliest possible dating
would be to the fifth century. However, a sixth-century dating is
widely accepted.
3: St Gall, Stiftsbibliothek, MS. 912 (= Lowe, CLA, vol. 7, no.
967a).
------------------------------
1
The sources that we have used for this note on the glossaries include CLA;
Dionisotti, “Greek grammars and dictionaries”; Goetz, Glossarii Latini; Kaster,
Guardians of language; Law, Grammar and grammarians; Lindsay, Early mediaeval
Latin glossaries.
GREEK AND LATIN GLOSSARIES lxi
Written on parchment as a palimpsest over six other texts dating

from the fourth to eighth centuries, in rude Uncial majuscule, this
manuscript was once thought to have been produced at St Gall;
however, it has been shown to have been written in North Italy in the
eighth century, at approximately the same time as Vat. Lat. 3321. The
glossary on folios 2v-160v was edited in Goetz, Glossarii Latini, vol.
4, pp. 199-298 as Glossae Codicis Sangallensis 912.
The sources of the St Gall glossary included a composite Abstrusa-
Abolita glossary as well as a Philoxenos glossary, St Isidore’s
Etymologiae, and some bilingual glossary probably produced in an
Italian monastery where Greek was spoken or studied, such as
Vivarium or Bobbio.
4: London, British Library, MS. Harley 5792 (= Lowe, CLA, vol.
2, no. 203).
A parchment Uncial manuscript probably produced in Italy,
possibly Byzantine Italy, possibly as early as the seventh century or
alternatively as late as ca 800, this soon found its way to Merovingian
France, as is shown by some annotations on it in Merovingian
minuscule. Folios 1v-240v were edited in Goetz, Glossarii Latini, vol.
2, pp. 213-483 as Glossae Graeco-Latinae.
The glossary was a copy of the Greek-Latin glosses attributed to a
certain Cyrillus and may well have been copied from a papyrus
exemplar. The Cyril glosses have been tentatively dated to the sixth
century; however, who Cyrillus was is unknown. He was not the fifth-
century Patriarch of Alexandria.
The sources of the Cyril glosses included a Latin-Greek glossary
similar to the pseudo-Philoxenos and then turned back to front, and a
Latin grammar composed for Greeks.
5: Munich, Bayerische Staatsbibliothek, Cod. Monac. Lat. 13002.
In a dated manuscript produced in 1158 at the German monastery
of Prüfening in Bavaria, this glossary is known as the Hermeneumata
Monacensia. Folios 209r-218r were edited in Goetz, Glossarii Latini,
vol. 3, pp. 117-220.
The glossary was a much-removed copy of glosses from a third-
century Greco-Latin schoolbook attributed, falsely, to Dositheus, the
author of a Latin grammar for Greeks composed in the early third
century which achieved a wide circulation.
6: Metz, Bibliothèque Publique, Cod. Metensis 500.
Folios 9r-24v and 136r-160v of this eleventh-century manuscript
contain a late tenth-century copy of a glossary known as the Glossae
Aynardi from their inscription attributing them to a certain Aynardus
lxii GREEK AND LATIN GLOSSARIES
in the year 969. Excerpts from the glosses were edited in Goetz,
Glossarii Latini, vol. 5, pp. 615-625.
This was a unique glossary, sui generis, with no known links to any
other glossary. The author was an unknown grammarian associated
with Toul, at the tomb of St Evre of which he dedicated the glossary,
according to his own preface. He knew Origen and St Ambrose, the
De compendiosa doctrina of Nonius Marcellus (fl. ca 280) on the
literature of Republican Rome, the grammarian Servius (Rome, late
fourth to early fifth century), and Horace, Vergil, and Juvenal.
7: Erfurt, Wissenschaftliche Allgemeinbibliothek, Amplon. Fol.
42.
Folios 1-14v contain the so-called Amplonianum Primum, the First
Amplonian Glossary. The Second Amplonian Glossary follows on
folios 14-34 of the manuscript. The First Amplonian Glossary dates
from the ninth century and was probably produced in Germany. It was
edited in Goetz, Glossarii Latini, vol. 5, pp. 337-401. It was akin to
the manuscript Épinal MS. 7, also of the ninth century, which Goetz
included in his critical apparatus.
The glossary included material from the Ars de nomine et verbo of
the grammarian Phocas (Rome, fifth century), Hermeneumata
materials, glosses of the Antiochene grammarian Rufinus (mid fifth -
early sixth century) on Eusebius, Orosius, St Jerome, De viris
illustribus, the Vulgate Bible, and the Abstrusa and Abolita glossaries.
8: Rome, Biblioteca Apostolica Vaticana, MS. Vat. Lat. 6925.
Folios 67r-78v contain the so-called Hermeneumata Vaticana, like
the Hermeneumata Monacensia a much-removed copy of glosses
from the Greco-Latin schoolbook attributed to Dositheus. The
manuscript is dated to the tenth century. The glossary was edited in
Goetz, Glossarii Latini, vol. 3, pp. 421-438.
9: Leiden, Bibliothek der Rijksuniversiteit, MS. BPL 67F (=
Lowe, CLA, vol. 10, no. 1575).
In Caroline minuscule, this manuscript was written somewhere in
North-East France in the age of Charlemagne and signed on folio
158v by a certain GAwsQ MA Rws (Gao2sthmaro2s). Folios 142v-147r
were edited in Goetz, Glossarii Latini, vol. 5, pp. 637-651 as the
Glossae Nonii.
Compiled by Agellus and Marcellus, the Glossae Nonii contained
glosses derived from the De compendiosa doctrina of Nonius
Marcellus. It was probably compiled from marginal notes in a
manuscript of Nonius, one of which may have been taken to Tours by
Alcuin.
ABBREVIATIONS
AnalBoll Analecta Bollandiana

BA-S Amari, M., trans., Biblioteca Arabo-Sicula.
Versione italiana, 2 vols and Appendice
(1880-81, 1889; rpt, Turin and Rome, 1982).
BGA De Goeje, M. J., ed., Bibliotheca geographorum
Arabicorum, 3rd ed. (Leiden, 1967).
BMGS Byzantine and Modern Greek Studies
BZ Byzantinische Zeitschrift
CCContMed Corpus Christianorum, Continuatio medievalis,
173- vols (Turnhout, 1971-2000-).
CCSL Corpus Christianorum, Series Latina, 176 vols
(Turnhout, 1954-65).
CFHB Corpus fontium historiae Byzantinae (Berlin et
al., 1967-).
CI Corpus iuris civilis. Vol. II: Codex Iustinianus,
ed. P. Krueger (1877; rpt, Berlin, 1970).
CLA Lowe, E. A., ed., Codices Latini antiquiores, 11
vols (Oxford, 1934-71).
CSHB Corpus Scriptorum Historiae Byzantinae, 34
tomes (Bonn, 1828-53), plus tome 34, vol. 3
(Bonn, 1897).
EHR English Historical Review
FstI Fonti per la Storia d’Italia
IJNA International journal of nautical archaeology
and underwater exploration
JHS Journal of Hellenic Studies
JMH Journal of medieval history
JRS Journal of Roman Studies
MGHAA Monumenta Germaniae historica, Auctores
antiquissimi, 15 vols (1877-1919; rpt, Berlin,
1961).
MGHEp Monumenta Germaniae historica, Epistolae, 8
vols (1891-1939; rpt, Berlin, 1957-75).
MGHScriptRerGerm Monumenta Germaniae historica, Scriptores
rerum germanicarum in usum scholarum ex
monumentis Germaniae historicis, nova
series, vols 1-7 (1922-30; rpt, Berlin, 1964);
lxiv ABBREVIATIONS
vols 8-13 (Berlin, 1955-67); vol. 6 recusi

(1985; rpt, Hanover, 1950); vol. 41
separatim editi (Hanover and Leipzig, 1915).
MGHScriptRerLang Monumenta Germaniae Historica, Scriptores
rerum Langobardicarum et Italicarum saec.
VI-IX (1878; rpt, Hanover, 1964).
MGHScriptRerMer Monumenta Germaniae historica, Scriptores
rerum Merovingicarum, 7 vols (Hanover and
Leipzig, 1885-1919/20; vols 1-4 rpt,
Hanover, 1969-77).
MGHSS Monumenta Germaniae historica, Scriptores, 30
vols (1826-1934; rpt, Stuttgart and N.Y.,
1963-64).
MM Mariner’s Mirror
PG Patrologiae cursus completus, Series Graeca,
ed. J.-P. Migne, 161 vols (Paris, 1857-66).
PL Patrologiae cursus completus, Series Latina, ed.
J.-P. Migne, 217 vols (Paris, 1878-90).
PO Patrologia Orientalis, 47 vols [212 nos] (Paris
and Turnhout, 1903-98).
REB Revue des études byzantines
RHCHOcc Recueil des historiens des Croisades, Historiens
occidentaux, 5 vols (1844-95; rpt, N.Y.,
1967).
RHCHOr Recueil des historiens des Croisades, Historiens
orientaux, 5 vols (1872-1906; rpt, N.Y.,
1967).
RISS Rerum italicarum scriptores, ed. L. A. Muratori,
25 vols (Milan, 1723-51).
RISSNS Rerum italicarum scriptores, nova series, 34
tomes (Bologna, 1900-35).
ROL Revue de l’Orient Latin
Rolls Series Great Britain, Public Record Office, Rerum
Britannicarum medii aevi scriptores; or,
Chronicles and memorials of Great Britain
and Ireland during the Middle Ages, 99
tomes (London, 1857-96).
RSBN Rivista di studi bizantini e neoellenici
TM Travaux et Mémoires
MAPS lxv
Map 1: The Atlantic and western Mediterranean: antiquity and the Early
Middle Ages
lxvi MAPS
Map 2: The Atlantic and western Mediterranean: the High Middle Ages
MAPS lxvii
Map 3: The central and eastern Mediterranean

lxviii MAPS
Map 4: Southern Italy and Sicily

MAPS lxix
Map 5: The Ionian Sea

lxx MAPS
Map 6: The Adriatic Sea

MAPS lxxi
Map 7: The southern Aegean and Lycia

lxxii MAPS
Map 8: The northern Aegean and the Balkans

MAPS lxxiii
Map 9: The Dardanelles to the Black Sea

lxxiv MAPS
Map 10: Byzantine Asia Minor

MAPS lxxv
Map 11: Greater Syria

lxxvi MAPS
Map 12: Egypt and Palestine

MAPS lxxvii
Map 13: Greater Iraq

INTRODUCTION
There are few images more representative of the Mediterranean Sea in

the Early Middle Ages than that of the famous Byzantine war galley
known as the dromo2n. At sea, the succession of the dromon to the
Roman bireme liburna and its predecessors, especially the Greek
trie2re2s, has been presented in the conventional historiography of the
maritime history of the Mediterranean as marking a transition from
Rome to Byzantium. Similarly, the succession of the Western galea to
the dromon in the late eleventh and twelfth centuries has been
presented as marking a transition from the Early Middle Ages to the
High Middle Ages in so far as the maritime history of the
Mediterranean is concerned.
Behind this conventional presentation lie two intellectual
assumptions which have underpinned the historiography. The first is
that specific ship types, known by different names, existed in different
chronological periods, or in different civilizations, and that these had
distinctive construction features which either can be ascertained or, if
they cannot be ascertained, would be able to be ascertained if
sufficient evidence were available. The second assumption is that
when the writers of ancient and medieval texts used terms such as
trie2re2s, liburna, dromo2n, or galea, they actually intended to refer to
such specific ship types because these names were applied to the ship
types by their contemporaries. Therefore, if a new name began to be
used in the texts from a certain period, this reflects the fact that a new
type of ship appeared in that period. Conversely, if a name faded from
use in the texts in a certain period, then this indicates that the type of
ship to which it referred had disappeared. It has been assumed that
there were definite relationships between the words and the physical
objects to which they referred, relationships which were both stable
over long periods of time and also consistent in usage from place to
place and person to person at any one time.
This study was begun by John Pryor as an attempt to research the
construction characteristics of the Byzantine dromon in the age of the
Macedonian emperors in this conventional way. When it was
commenced, there was an implicit acceptance of the assumptions of
the conventional historiography. However, in the course of our
research we have been led to question them. As a result, we have also
been led to question the very basis of any attempt to know what “the”
2 INTRODUCTION
Byzantine dromon actually was. In certain periods Byzantines

certainly referred to galleys by the term dromo2n , and also by
chelandion and other terms, but did they always really intend that
their use of these terms should actually designate specific galley types
with distinctive design characteristics?
On the one hand, maritime historians know well that throughout
history gradual evolution has almost invariably been the norm in so
far as ship design is concerned. There has very rarely been any sudden
technological innovation which has produced a distinctive new ship
type overnight. Even submarines and aircraft carriers were developed
gradually as new features were experimented with. Ship types have
never remained static and fixed in design over time. They have always
evolved slowly as generation after generation has progressively
refined them and adapted them to changing circumstances. The
evolutionary norm has been that eventually changes have become so
marked that the ships have become distinctive new types which can be
distinguished from their progenitors. Sometimes a previous name or
term for a ship type has been taken into a new technological context;
for example, the medieval Italian galeone for a small galley eventually
became galleon for sailing ships of the sixteenth century. Sometimes a
term for a ship type has been replaced by another term; for example,
the Scandinavian knörr, which evolved in England into the Anglo-
Norman buss. This being the case, we are led to consider whether “a”
distinctive Byzantine warship, known as a dromo2n, ever actually
existed at any time or whether, in fact, different forms of galleys over
many centuries were referred to by Byzantines and others by the name
dromo2n? There is no reason per se why the same term used in, say,
the sixth century and the tenth, should not have been used with
reference to quite different ship types. There is no reason, per se, why
the same name should not have continued in use even if the
construction features of the ships had changed dramatically.
On the other hand, when we examine texts which use terms such as
dromo2n for ships, the reality for us lies in the texts and terms
themselves. In most cases, we cannot see beyond the terms and cannot
know whether two authors using the same term, even in the same time
period, really had the same type of ship in mind. The same would true
of the use of terminology in different geographical regions. Was a
ship referred to as a chelandion in Byzantine South Italy in the tenth
century really the same as that which was referred to by the same
name in Constantinople? Futhermore, in most cases we cannot even
know whether authors really even intended to refer to any specific
INTRODUCTION 3
ship type by their use of such terms. Indeed, in many cases, collateral
evidence suggests that their use of them was no more specific than is
that of “yacht” in our own time: a term which began with a specific
reference to a seventeenth-century Dutch ship but which has since
been applied to almost any kind of sailing pleasure craft. The popular
use of “battleship” is another case in point. The word is correctly used
for first-rate capital ships of the modern era of iron ships but is
frequently used in popular literature with many other references.
Nelson’s Victory, for example, is often referred to as a “battleship”;
whereas, she was properly a “first-rate ship of the line”. Only if we
had texts which empirically described the construction or operation of
galleys referred to as dromo2nes at any particular time could we be
confident that we were being informed about actual ships in
contemporary use, but even then only for that time and place and for
those texts.
To this general problem of the use of technical and technological
terminology in texts, we need to add another consideration especially
prominent in Byzantine literary texts. As is well known, in most
periods most educated Byzantine authors aped the style and
vocabulary of classical Greece. Their models were, for example,
Homer, Herodotos, and Thucydides. Moreover, Byzantine literati
learned their classical Greek by reading and memorizing these and
other authors. As a result, classical vocabulary and expressions
continually recurred to them when searching for ways in which to
express what they wished to say. When writing, they might, on the
one hand, attempt to display their education to their intended readers
by deliberately quoting or paraphrasing snippets from classical
authors. On the other hand, such snippets might find their way to their
pens quite unconsciously simply as a product of their education
because a word or phrase or clause remembered from their education
sprang to their minds as a way to say something. A similar problem
occurs in Western medieval Latin texts when authors used short
passages of scripture to express something. Often, one can not know
whether the quotation was deliberate or simply a product of their
education, during which much of the Bible had been memorized.
Consequently, when we find Byzantine authors using technical
vocabulary derived from a distant past, such as trie2re2s for a “three” or
three-banked galley, or triakonte2re2s for a “thirty” [oared galley], or
pente2konteros for a “fifty” [oared galley], we can never be sure that
they intended to convey to their readers that fact that the ships in
question had the technological characteristics to which the terms had
4 INTRODUCTION
originally referred. They may simply have been using a word for a
ship which was known to be classical, and therefore approved,
without any intent at all to link it in their readers’ minds with the
technological characteristics of the ships of their own day referred to.
The latter may or may not have had three banks or thirty or fifty oars,
etc. There is simply no way of knowing from the texts per se.
Leaving aside the question of subconscious utilization of classical
terminology, there is no doubt that educated Byzantines did also
deliberately and consciously ransack classical texts for their own
purposes. Unfortunately for modern maritime historians, this was the
case with the “Naval warfare, commissioned by Basil, the patrician
and parakoimo2menos”, a treatise compiled by an anonymous author
for the parakoimo2menos Basil Lekape2nos, in 958-59, and which is the
only surviving text which purports to describe the contemporary
construction of dromo2nes and chelandia.1 In the past, this treatise has
been accepted as a virtual “shipwright’s manual” by maritime
historians; although, its derivative nature has been recognized by
literary historians.2 It will be shown to have been little more than an
exercise in classicizing philology, and therefore to be of only limited
use for study of the construction of actual tenth-century dromons.
Since we have been led to question seriously the underlying
assumptions for empirical study of the construction of the ships, we
have then approached the reality of “the” Byzantine dromon from
alternative perspectives. On the one hand, from the sixth to the twelfth
centuries, Byzantines and others certainly referred to some kinds of
war galleys by the name dromo2n . On the other hand, real war galleys
certainly existed. But, what did contemporaries intend their
terminology to signify and what can we know of the physical objects
to which they referred? Beyond that, with what degree of confidence
can we use their texts to research the construction characteristics of
the galleys and the ways in which they may have evolved over time?
Our primary objective has become an attempt to elucidate the
meanings of terminology as used by contemporaries and how such
meanings may have varied from time to time or from author to author.
------------------------------
1
Naumachika syntachthenta para Basileiou patrikiou kai parakoimoumenou.
Hereafter, this text is referred to as the Anonymous and its author as “the
Anonymous”. We have edited the text from a microfilm of the manuscript and
translated it here in Appendix Three because the text published as Naumacika;
Suntacqevnta para; Basileivo u patrikivo u kai; para-koimoumevnou, in Dain Naumachica,
pp. 57-68 has been found to be completely unsatisfactory. On the text and its
classicizing terminology, see below pp. 183-6.
2
Hunger, Hochsprachliche profane Litteratur, vol. 2, p. 334.
INTRODUCTION 5
As a result, large sections of this study have evolved into an

etymological and philological hunt for linguistic chimerae. We are
well aware, of course, that the hunt was always doomed to only partial
success at the best. The passage of the centuries and the disappearance
of so many sources has made the recovery of the meaning of
terminology in the past possible only in part. Some readers may
consider that in some places we have pushed the search for
understanding of the meaning of terminology to excessive lengths, or
that we have presented the evidence at excessive length. In response,
we point out that the search has been successful in some places in
elucidating the meaning of some terms whose meaning has been
completely forgotten; for example, perovnh (perone2) for the “spur” of
a galley, kalu(m)bomavto" (kaly(m)boma2tos) for a water tank or
possibly something to do with a bilge pump, and trocanth'r
(trochante2r) for a part of a rudder to which the rudder tackles were
attached. These are merely three examples and there are many points
at which we consider that the results have justified the hunt. Those
who find the presentation of the evidence tedious can simply skip to
the conclusions; however, there will be some readers who will want
the evidence for the conclusions properly presented.
In retrospect, now that the research has been done and the book has
been written, there will no doubt be some who will consider that we
have made much ado about nothing. What else would one expect but
that words, even technical and technological terms, varied in meaning
from time to time, place to place, and author to author? What else
would one expect but that Byzantine galleys of the tenth century were
not the same as those of the sixth century? What else would one
expect but that Byzantine authors wrote classicizing philological
treatises rather than shipwrights’ manuals? We would respond that
these have not been the assumptions of the traditional historiography
of maritime history, that there has been an assumption that something
called “the dromon” did exist and remained the same for centuries,
and that texts referred to actual ships. We came to a full appreciation
of the extent of the methodological difficulties and to our questioning
of the assumptions of the maritime historiography only slowly.
We have been mindful of all of these considerations throughout
and have attempted to avoid referring to dromons as though they were
a single reality. Only in Chapter Four, which deals with the
construction, equipment, and armaments of tenth-century Byzantine
war galleys during the era of the Macedonian emperors, have we used
the word dromon as an intellectual shorthand, as though it did
6 INTRODUCTION
represent a single reality. To have avoided it there would have

involved endless and tedious circumlocutions. Moreover, since most
of the texts under discussion in Chapter IV emanated from
Constantinople over a comparatively short period between ca 900 and
960, there is some justification for considering that fairly well-known
and standardized types of galleys may have been referred to as
dromo2nes in those texts.
CHAPTER ONE
THE OPERATIONAL CONTEXT1
The first period, ca 400-560: the Germanic assault and imperial

recovery
It has been claimed frequently that from Actium in 43 B.C.E. to the

battle of the Dardanelles between forces of Constantine and Licinius
in 324 C.E. the Mediterranean was a Roman lake and that therefore
the Roman imperial navy retained throughout only the skeletal forces
needed for state communications and occasional suppression of
piracy. When the Crimean Goths crossed the Black Sea into the Sea of
Marmara in 267, the naval forces sent against them were inadequate.2
In reality, however, the Empire had always maintained significant
naval forces, both in the Mediterranean and in the North, and by the
fourth century authorities had the capability to put considerable naval
forces to sea. At the Dardanelles, Constantine’s forces supposedly
engaged Licinius’ 350 trie2reis with 200 triacontoroi.3 Constantius
gathered a large fleet in the East in 352 for his assault on the usurper
Magnentius in the West. Theodosius I sent Valentian II to Italy with a
squadron in 388 when Maximus seized power there and Maximus
himself gathered a fleet in the Adriatic to intercept them. In 398 the
magister militum of the West, Stilicho, sent Mascezel, brother of the
rebel comes Africae Gildo, with a fleet against his brother.
------------------------------
1
This chapter could obviously have been another book and the sources that could
be adduced in documentation of it are so numerous that they would have expanded the
bibliography unrealistically. Only the most pertinent have been adduced here. By and
large, only earlier, more contemporary sources have been cited and later, more
derivative ones have been omitted; even though it is appreciated that contemporaneity
is not always the best index of reliability. We have made occasional exceptions to this
rule where there are good reasons to do so, particularly in the cases of some
information supplied uniquely by the great Muslim historians Ibn al-Athı3r and Ibn
Khaldu2n, that of the indispensable historian of Maghribin and Andalusi affairs Ibn
‘Idha2rı3, and the Egyptian historian Al-Maqrı3zı3, but generally we have adhered to it.
References to modern secondary literature have been kept to an essential minimum.
2
Zo2simos, Historia nova, A.34-5 (pp. 24-5).
3
Zo2simos, Historia nova, B.22 (pp. 78-9). Zo2simos’ use of the classical words
trie2reis and triacontoroi for three-banked triremes and thirty-oared galleys
respectively was not technical. The language really meant no more than that Licinius’
galleys, collected from Egypt, Phoenicia, Africa, and elsewhere, were larger than
those of the fleet Constantine had built at Thessalonike2.
8 CHAPTER ONE
Stilicho also sent naval forces against the Visigoths in the Balkans
and when the Gothic magister militum Gainas tried to cross the
Dardanelles in 399 on hastily assembled ships or rafts, his forces were
massacred by Roman liburnae under the command of Fravitta. In 410
the Western emperor Honorius I was besieged in Ravenna by Attalus,
a usurping emperor created by Alaric the Visigoth, and prepared to
flee by sea but was saved by the arrival of six regiments of 4,000 men
from the East, suggesting considerable capability to transport troops
by sea. When Heraclianus, the comes Africae, rebelled against
Honorius in 413 and crossed to Italy, Orosius reported that he had
3,700 ships, a gross exaggeration no doubt but nevertheless indicating
that considerable naval forces could be gathered in Africa. In 417 the
magister militum Constantius penned the Visigoths in Narbonne,
cutting off supplies by sea and forcing them to evacuate and cross the
Pyrenees into Spain. He, also, must have had considerable naval
forces. And, finally, in 425 Theodosios II sent forces against the rebel
John under the magister militum Ardabourios and his son Aspar which
stormed Salo2nes and then made a sea-borne expedition to Aquileia.4
In 429 a confederation of Siling and Asding Vandals and Alans
under the Asding king Gaiseric took ship from Cartagena to
Tingitania, possibly by invitation of Boniface, the comes Africae.
From Tingitania Gaiseric pushed east into the provinces of
Mauritania Caesariensis, Mauritania Sitifensis, and Numidia. A
combined expedition of Western and Eastern forces under Boniface
and Aspar, the magister militum in Constantinople, failed to dislodge
him in 431 and by 435 the Western emperor Valentinian III was
forced to cede possession of the two Mauretaniae and Numidia,
retaining only Carthage and the province of Africa for the Empire.
But in 439 Gaiseric finally took Carthage, making it the capital of the
Vandal kingdom. An expedition sent against him reached Sicily in
441 but was recalled because of attacks by Attila the Hun in Thrace. A
treaty ceded Africa, Byzacena, Tripolitana, and eastern Numidia to the
Vandals in 442 while returning the Mauretaniae and western Numidia
to the Empire, at least in theory.5
------------------------------
4
Claudian, De bello Gildonico, ll. 417, 489-91, 515-26 (pp. 69, 71-3); idem, De
quarto consulatu Honorii, ll. 459-65; idem, De consulatu Stilichonis, I.170-74;
Eunapios, History (Blockley), Frag. 64.1 (p. 94); Julian [emperor], Orations, I.40
(vol. 1, p. 104); Olympio2doros, Books of history, fr. 43.2 (p. 208); Orosius, Historiae
adversum paganos, VII.42.13 (p. 298), 43.1 (p. 299); Zo2simos, Historia nova, D.45.3-
46.1, E.11.3-4 (pp. 203, 228-9). See also Manfroni, Marina italiana. I, pp. 3-4; Reddé,
Mare nostrum, pp. 605-47.
5
Gregory of Tours, Historiae, II.2 (fasc. 1, pp. 39-40); Hydatius, Chronicle,
THE OPERATIONAL CONTEXT 9
The Vandals appear not to have established a navy in the sense of a

dedicated battle fleet; however, from Africa they immediately
launched raids on Sicily using ships captured in Carthage and others
they built. Under Gaiseric a fleet sailed to Ostia in 455 and
systematically pillaged Rome for fourteen days. A fleet of 60 ships
sent to Corsica probably in 456 was defeated by the magister militum
of the West, Ricimer. The Vandals raided Campania in 458 and 461-3,
occupied the Balearics some time after 455, Corsica some time after
456, raided Sardinia and occupied it temporarily some time after 455
and permanently from 482/3, raided Sicily and Italy annually and
occupied Sicily some time after 468, and sacked Nikopolis in Epiros
and Zakynthos around 474.6
Theodosios II sent 1,100 ships, according to Theophane2s the
Confessor, against Gaiseric in 448 but the fleet only reached Sicily.
Then, the Western emperor Majorian attempted to gather a fleet in
southern Spain around Cartagena in 460 to mount an attack but the
Vandals destroyed it. Not until 465-8 did the Eastern Empire attempt
an offensive against the Vandals when Leo I mounted a major three-
pronged attack. A comes Marcellinus expelled the Vandals from
Sicily in 465 and from Sardinia probably in 466. At the same time the
magister militum He2rakleios of Edessa landed at Tripoli and marched
on Carthage by land. However, the entire operation ended in disaster
when the magister militum, the emperor’s brother-in-law Flavius
Basiliskos, anchored his fleet in an exposed position off Cape Bon and
it was destroyed by Vandal fire ships. Leo’s successor Zeno realized
that he had to come to terms and and negotiated peace in 474.7
------------------------------
§§80, 107 (pp. 90, 94); Jordanes, Getica, §167 (pp. 101-2); idem, Romana, §330 (p.
42); Marcellinus Comes, Chronicon, Annus 439 (p. 17); Priskos, History, fr. 9.4 (p.
240); Prokopios, History of the wars, III.iii.23 - III.iv.15 (vol. 2, pp. 28-38); Prosper,
Epitoma chronicon, §§1295, 1321, 1332, 1339, 1347 (pp. 472-7); Theophane2s,
Chronographia, A.M. 5931, 5943 (pp. 95, 104); Victor Vitensis, Historia, I.i.1-2,
I.iv.12-13 (pp. 2-4). See also Courtois, Vandales.
6
Hydatius, Chronicle, §§160, 169 (pp. 104, 109); John Malalas, Chronographia,
IDV[14].26 (pp. 287-8); Jordanes, Getica, §235 (p. 118); Malchos, Byzantiaka, fr. 5 (p.
410); Marcellinus Comes, Chronicon, Annus 455 (p. 22); Priskos, History, frr. 30.1 &
3, 31.1, 38.1-2, 39.1 (pp. 330-4, 340, 342); Prokopios, History of the wars, III.v.1-5,
III.v.22-25, III.xxii.16-18, IV.xiv.40 (vol. 2, pp. 46-8, 52-54, 188, 338); Prosper,
Epitoma chronicon, §1375 (p. 484); Sidonius Apollinaris, Carmina, V.388-92 (p.
197); Theophane2s, Chronographia, A.M. 5947 (pp. 108-9); Victor of Tunnuna,
Chronica, Annus 455 (p. 186); Victor Vitensis, Historia, I.iv.13, I.viii.24, III.v.20 (pp.
4, 7, 45). See also Courtois, Vandales; Reddé, Mare nostrum, pp. 648-50.
7
Hydatius, Chronicle, §§195, 223 (pp. 112, 116); John Malalas, Chronographia,
IDV[14].44 (p. 296); Malchos, Byzantiaka, fr. 17 (pp. 424-6); Marius of Avenches,
Chronica, Annus 460 (p. 232); Priskos, History, frr. 36.1-2; 53.1, 3-5 (pp. 338, 360-
68); Prokopios, History of the wars, III.vi.7-24, III.vii.26-7 (vol. 2, pp. 56-62, 70-72);
10 CHAPTER ONE
Table 1: Rulers of the first period, ca 400-560
Eastern Empire Western Empire
Emperors magistri militum Emperors magistri militum
Arkadios (395-408) Gainas (399-400) Honorius (393- Arbogast (388-94)

Fravitta (400-401) 423) Stilicho (395-408)
Constantius (411-
21)
Theodosios II (408- Aspar (431-71) Constantius III Castinus (423-5)
50) Flavius (421) Felix (425-30)
Basiliskos (468-74) John (423-5) Aetius (430-32)
Valentinian III Boniface (432)
(425-55) Aetius (433-54)
Marcian (450-57) PetroniusMaximus
(455)
Avitus (455-7)
Leo I (457-74) Zeno (473-4) Majorian (457-61) Ricimer (457-72)
Liberius
Severus (461-5)
Anthemius (467-
72)
Anicius Olybrius
(472)
Glycerius (473)
Julius Nepos (473-
75)
Leo II (474)
Zeno (474-91) Theodoric Strabo Romulus Orestes (475-6)
(474) Augustulus Odovacar (476-)
Theodoric the (475-6)
Amal (476-8) _____
Anastasios I (491-
518)
Justin I (518-27) Vitalian (518-20)
Justinian I (527-65) Belisarios (529-48)
Justin II (565-78)
From the death of Gaiseric, the Vandals ceased to be a threat and were
eventually overthrown by Justinian I’s general Belisarios in 533-4.
Just how massive an undertaking the sea-borne invasion of Africa from
------------------------------
Theophane2s, Chronographia, A.M. 5941-2, 5961, 5963 (pp. 101-2, 115-17).
(Table 1 continued)
Ostrogoths Visigoths Vandals
Alaric (395-410) Godegisel († 406)
Athaulf (410-15) Gunderic (406-28)

Valia (415-18) Gaiseric (428-77)
Theodoric I (419-51)
Thorismund (451-3)
Theodoric II (453-66)
Euric (466-84)
Odovacar (476-93) Alaric II (484-507) Huneric (477-84)

Gunthamund (484-96)
Thrasamund (496-523)
Theodoric the Great (493- Theodoric the Great (507-

526) 26)
Amalaric (507-31)
Athalaric (526-34) Hilderic (523-30)
Gelimer (530-34)
____________
Theodahad (534-36) Theudis (531-48)
Witigis (536-40) Theudegesil (548-9)
Hildibad (540-41) Agila (549-54)
Eraric (541) Athanagild (554-67)
Totila (541-53)
Teia (553)
_________
Leovigild (568-86)
Constantinople was is revealed by Prokopios’s report of the advice of

the praetorian prefect, John of Cappadocia, to Justinian to think twice
about the undertaking. After the defeat and capture of the Vandal king
Gelimer, Belisarios’s lieutenants subsequently recovered Sardinia,
Corsica, the Mauretaniae, Tingitania, the Balearics, and Lilybaion in
12 CHAPTER ONE
Sicily. Belisarios himself recovered the whole of Sicily in 535 and,

following Berber uprisings and civil wars, by 546 Africa had been
pacified.8
During the reign of Gaiseric from 439 to 474 the Vandals had
broken the Romanized homogeneity of the Mediterranean for the first
time. Even though they established a polity based on grain production
integrated into Mediterranean maritime commercial networks rather
than the corsair kingdom which their state used to be characterized
as,9 nevertheless the establishment of their kingdom marked a first
stage in the breakup of the Mediterranean.
When the Visigothic king Alaric I moved into Italy he attempted to
cross to Africa in 409-10. Then after entering Spain, the Visigoths
under Valia also tried to cross from Gibraltar in 415 but did not have
the necessary ships. Having moved back across the Pyrenees in 418 to
settle in the provinces of Novem Populi and Aquitania Secunda, they
did not return to Spain until 468, occupying most of the peninsula by
473 except for Galicia, which remained in the hands of the Suevi.
Then, in 507 at Vouillé, near Poitiers, the Frankish army of Clovis
annihilated that of Alaric II and the Visigoths were pushed back into
Spain; although, they long retained territory along the Mediterranean
coast of Narbonensis Prima. In 511 the Ostrogoth Theodoric took
over the Visigothic kingship and ruled in Spain and Toulouse through
a governor, Theudis, on behalf of Alaric II’s son Amalaric. On
Amalaric’s death Theudis seized the crown but from this time on
Visigothic emigration into Spain stepped up. In 551 a noble by the
name of Athanagild revolted against King Agila and called in support
from the Emperor Justinian, whose troops occupied south-east Spain
from Cartagena to Malaga and some distance inland by 555.
Athanagild became king in Toledo, his successor Leovigild finally
eliminating the kingdom of the Suevi in 585, but the Byzantines were
not finally expelled from their Spanish province until after 621.10
Throughout the history of their kingdom, the Visigoths appear to
have had few, if any, naval forces, their only known naval expedition
------------------------------
8
Prokopios, History of the wars, III.x - IV.iv, III.x.7-17, IV.v, IV.viii.9-25,
IV.x.1-xiii.45, IV.xiv.1-2, IV.xiv.7-xxviii.52, (vol. 2, pp. 90-246, 92-6, 246-54, 272-8,
284-326, 326, 328-458); Theophane2s, Chronographia, A.M. 6026 (pp. 186-216).
9
Courtois, Vandales, pp. 205-14; Hodges and Whitehouse, Mohammed, pp. 26-
30.
10
Gregory of Tours, Historiae, II.37, IV.8 (fasc. 1, pp. 87-8, 140); Isidore of
Seville, Historia, Aerae CCCCXLVII, DXCII, DCLVIIII (pp. 275, 286, 292);
Jordanes, Getica, §173 (p. 103); Orosius, Historiae adversum paganos, VII.43.10-12
(p. 300); Zo2simos, Historia nova, ı.7 (p. 288). See also Thompson, Goths in Spain;
García Moreno, “Byzantium’s Spanish province”.
occurring in 547. At some time before 534 they managed to cross the
Straits of Gibraltar and take Ceuta from the Vandals. Belisarios’s
forces expelled them in 534 but in 547 Theudis again crossed the
Straits in an unsuccessful attempt to regain it. No attempt against the
Balearics, which were also recovered for the Empire by Belisarios, is
known to have been made. Although Isidore of Seville claimed that
from the reign of Sisebut the Visigoths acquired eminence at sea, no
evidence supports this.11
After the overthrow of the Empire in the West by Odovacer, and
then under the Ostrogoth Theodoric the Great, some naval forces were
maintained in the Adriatic; however, they appear to have been
minimal. According to Malchos, when Theodoric captured
Dyrrachion imperial authorities were so alarmed by the prospect of
his acquiring naval forces that he was ordered to advance no further
and to seize no ships. According to the Fasti Vindobonenses priores
and the Köbenhaven continuations of Prosper, Theodoric gathered
dromones at Rimini before besieging Odovacer in Ravenna. Agnellus
of Ravenna recorded that Odovacer fled from Ravenna before
Theodoric in 491 “cum dromonibus”.12
In 508 Anastasios I sent 100 war galleys of this new kind known as
dromones to ravage the coasts of Italy. His relations with Theodoric
were hostile but the precise purpose of the expedition is obscure;
possibly it was to dissuade the Ostrogoth from intervening in the
Languedoc after Vouillé. Whatever the case, Theodoric appears to
have had no naval forces with which to mount any opposition at sea.13
Only late in his reign did he begin to consider naval forces.
Cassiodorus drafted four letters on his behalf between 523 and 13
June 526 referring to them: the first two addressed to the praetorian
prefect Abundantius and the others to the Count of the Patrimony
------------------------------
11
CI, I.27.2,§2; Isidore of Seville, Historia, Aerae DLXVIIII, DCLVIIII
(recapitulatio) (pp. 284, 294-5); Prokopios, History of the wars, IV.v.6-9 (vol. 2, p.
248).
12
Agnellus, Liber pontificalis, §39 (p. 303); Malchos, Byzantiaka, fr. 20 (p. 442).
Consularia Italica, pp. 318-19. These last are anonymous manuscripts. The Fasti
Vindobonenses are in MS. 3416 (antea hist. Lat. 56 sive hist. prof. 452) of the old
Imperial Library of Vienna, now the Österreichisches Staatsbibliotek, written in 1480,
and in the eleventh-century manuscript, St Gall, MS. 878. The Köbenhaven
continuations of Prosper are in what is now the Kongelike Bibliotek, Köbenhaven,
quondam MS., N o 454, of the Danish Royal Library, probably twelfth century.
13
Marcellinus Comes, Chronicon, Annus 508 (pp. 34-5): “Romanus comes
domesticorum et Rusticus comes scholariorum cum centum armatis navibus
totidemque dromonibus octo milia militum armatorum secum ferentibus ad
devastanda Italiae litora processerunt ...”. See also Manfroni, Marina italiana. I, pp.
5-12.
14 CHAPTER ONE
Vvilia and to the saio Aliulfus. In part they do not ring true. Heavy
rhetorical flourishes reek of redrafting when Cassiodorus later
compiled his Variae, the source in which they survive. Moreover, in
the first, Theodoric supposedly ordered construction of 1,000 dromons
for carriage of public grain supplies as well as defence against hostile
ships. On the one hand, it is extremely doubtful that the Ostrogothic
kingdom would have been capable of building and maintaining 1,000
dromons in any case. From where could it have obtained the 50,000
oarsmen at least needed, as well as officers and marines? Later in the
letter Theodoric discussed recruiting slaves for the purpose! On the
other hand, no one would ever have built war galleys such as dromons
to transport grain. That would have been the most inefficient means
possible of doing so. It is true that in the second letter to Abundantius
Theodoric congratulated him on having completed the task in a very
short period of time and said that the fleet was to rendezvous at
Ravenna on 13 June 526; however, that does not necessarily mean that
1,000 dromons were constructed. The second letter to Abundantius
suggests that Theodoric intended to use the fleet against either the
Byzantines or the Vandals, or both. Theodoric’s break with the
Vandals may have gone back to 508 when the Vandal fleet failed to
prevent the imperial fleet ravaging Italy, or to 510/11 when the
Visigothic claimant Gesalec had found refuge in Carthage, or to the
imprisonment of Theodoric’s sister Amalafrida, the widow of
Thrasamund, by his successor Hilderic and her death in 523.14
Whatever the case, his plans came to nothing because of his death and
a decade later the Ostrogoths appear to have had few naval forces with
which to oppose the imperial invasion of Italy.
The Gothic War opened in 535 with a two-pronged amphibious
assault on the outposts of the Ostrogothic kingdom. Belisarios was
sent with a fleet and and army to coccupy Sicily and in the following
year crossed to Calabria. Justinian also sent Ko2nstantianos, the
commander of the imperial grooms, to Dyrrachion in the following
year to gather forces to expel the Goths from Salo2nes. He sailed with a
fleet to Epidauros and then to Salo2nes. In the first major Gothic
deployment of naval forces, Witigis sent an army by sea, supposedly
with many ploia makra, to recover Salo2nes but they were scattered by
------------------------------
14
Cassiodorus, Variae, V.16 (p. 195): “..., deo nobis inspirante decreuimus mille
interim dromones fabricandos assumere, qui et frumenta publica possint conuehere et
aduersis nauibus, si necesse fuerit, obuiare.”; V.17 (p. 196 “..., Non habet quod nobis
Graecus imputet aut Afer insultet.”. Cf. also V.17-20 (pp. 196-9). See also Wolfram,
Goths.
Ko2nstantianos’s fleet.15 The first phase of the war resolved itself into a
Gothic defence of the heartlands of their kingdom while every year
Justinian sent new forces to Italy by sea. By the winter of 537-8
Belisarios’s forces commanded the sea and those of Witigis besieging
Rome were starving. In March 538 he was forced to raise the siege
and retire to Ravenna.16
In spring 539 Belisarios moved towards Osimo, guarding the
approaches to Ravenna. Rimini had been occupied by one of
Belisarios’s lieutenants and was under siege. He left 1,000 men
encamped outside Osimo by the shore, sent a fleet with an army to
Rimini while another advanced up the coast, and made a sweep to the
west himself. The sudden appearance of the fleet over the horizon
precipitated a Gothic flight from Rimini back to Ravenna. With
command of the Po and the Adriatic, Belisarios besieged Ravenna late
in the year and Witigis was starved into submission by forces smaller
than his own. Belisarios entered Ravenna unopposed in May 540 and
at the same time a grain fleet entered its port, Classe, to supply the
city.17
After becoming king in 541, Totila perceived the need for naval
forces to counter those the Byzantines had thrown against Italy since
the beginning of the war. In 542, after Totila had defeated Byzantine
forces in the North and had broken through to the South to besiege
Naples, Justinian sent out a fleet under the praetorian prefect
Maximinos. A strate2gos, De2me2trios, sent to Sicily with another fleet,
but who sailed to Rome instead, was attacked by Totila and destroyed
by many dromons when he brought his fleet to Naples, the first clear
mention of Gothic use of dromons. Maximinos went to Syracuse,
stayed there through the summer, and then in the autumn was
persuaded to send his fleet to Naples. Caught by a storm, it was driven
ashore near the Gothic camp and mostly destroyed.18 Totila’s
------------------------------
15
Anonymous addition to Marcellinus Comes, Chronicon, Annus 535 (p. 46);
Prokopios, History of the wars, V.v.1-7, V.v.12-19, V.vii.26-37, V.viii.1-7, V.xvi.5-
17 (vol. 3, pp. 42-8, 64-8, 68-70, 158-62). See also Manfroni, Marina italiana. I, pp.
12-21.
16
Anonymous addition to Marcellinus Comes, Chronicon, Anni 536-8 (pp. 46-8);
Prokopios, History of the wars, V.xxiv.18-21, VI.v.1, VI.vi.2, VI.vii.1, VI.vii.16-18
(vol. 3, pp. 342, 326-4, 336, 346, 352).
17
Prokopios, History of the wars, VI.xvi.18-24, VI.xvii.21, VI.xxviii.6-7, VI.xxix.31
(vol. 4, pp. 8-10, 16, 114, 132).
18
Prokopios, History of the wars, VII.vi.10-17, VII.vi.24-5 (vol. 4, pp. 200-202, 204):
“Toutivla" de; to;n pavnta lovgon ajmfi; tw'/ stovlw/ touvtw/ ajkouvsa" drovmwna" me;n pollou;"
a[rista plevonta" ejn paraskeuh'/ ei\cen, ...”, VII.vii.1-7 (vol. 4, pp. 204-8).
16 CHAPTER ONE
command of the sea induced Naples to surrender in spring 543.

In 544, with the Byzantine fortress at Otranto under siege,
Belisarios was appointed to command in Italy again and sent a relief
fleet from Salo2nes with a year’s supplies and a replacement garrison.
After the garrison had been replaced and the fleet had returned, he
sailed to Pula, and from there to Ravenna.19 In the following year he
sent to Justinian begging for a new army, money, arms, and horses.
Totila moved on Rome, which he cut off by sea by a light fleet
stationed in the Aeolian and other coastal islands. A fleet sent from
Sicily by Pope Vigilius to reprovision Rome was intercepted off Ostia
and captured but Belisarios at Dyrrachion was able to send an army to
Ostia under the command of Valentinos with instructions to join the
garrison in Portus. According to Prokopios, Totila attributed the
reconquest of Rome and Italy to being able to secure Sicily and supply
Rome from there.20
In 546, after receiving reinforcements at Dyrrachion, Belisarios
sailed to Rome while his lieutenant John, nephew of the usurper of
513-15, Vitalian, took the rest of the army to Otranto and then
marched on Rome, clearing out Gothic garrisons en route. Outside
Rome Totila bridged the Tiber to prevent boats reaching the city. At
Ostia Belisarios fortified 200 dromons with wooden parapets with
bow-slits and made other preparations to ascend the Tiber.21
The following year Justinian sent reinforcements to Otranto. Near
the winter solstice, general Valerian reached the Adriatic but did not
cross because he thought that there would not be provisions sufficient
for the men and horses because of the season. Belisarios in Rome
sailed for Sicily and then Taranto with 700 cavalry and 200 foot but
was forced by storm to put in at Crotone. He stayed there with the
infantry but sent the cavalry ahead to secure passes and supplies.22
This was the first occasion during the progress of the war in which
Prokopios was clear that the Byzantines were transporting cavalry and
horses around the coasts, although they almost certainly had been
doing so earlier.
------------------------------
19
Prokopios, History of the wars, VII.ix.22 - x.9, VII.x.13, VII.xi.1 (vol. 4, pp.
228-32, 232, 236).
20
Prokopios, History of the wars, VII.xii.1-3, VII.xiii.5-7, VII.xv.9-13, VII.xv.1,
VII.xvi.16-21 (vol. 4, pp. 248, 256, 278-80, 274-6, 286-8).
21
Anonymous addition to Marcellinus Comes, Chronicon, Anni 546-7 (p. 51);
Prokopios, History of the wars, VII.xviii.1-4, 8-10, VII.xix.5 (vol. 4, pp. 300-302,
304, 312).
22
Prokopios, History of the wars, VII.xxvii.1-4, VII.xxvii.13-17, xxviii.3-7 (vol. 4,
pp. 386-8, 390, 394).
In 548 Justianian sent 2,000 infantry by sea to Sicily and Belisarios

at Otranto gathered a large fleet and sailed to the relief of Rossano;
however, it was scattered by storm and after regrouping was detered
from landing by Totila’s cavalry lining the beaches. Retiring to
Crotone, they decided to land the men and horses and march overland
to Picenum. Again, they must have been carrying horses.23 But what is
clearly apparent here is that the imperial forces did not have the
capability to land cavalry ashore against opposition. This is a theme to
which we shall return.
Belisarios was recalled to Constantinople in 548, having spent 5
years in Italy but having been confined to coastal landings, as a result
of which Totila still controlled most of the peninsula. In the same year
one of Belisarios’s guardsmen, named Indoulph, deserted to Totila
and went to Dalmatia to a place called Mouikouron near Salo2nes and
then to Laureate2. The Byzantine commander at Salo2nes, Klaudianos,
attacked him with a fleet of dromons but was defeated, the crews
abandoning their ships in the harbour. After this Indoulph returned to
Italy, where Totila gave him and other commanders 47 ploia makra to
besiege Ancona. This led to an attempt to relieve the city in 551 and to
the decisive battle at sea off Senogallia.24
In 549 Totila readied 400 ploia makra as well as a fleet of large
sailing ships sent from the East which had been captured. In response
to a Gothic attack on Reggio, Justinian sent a fleet and army under the
patrikios Liberios to Sicily, later replacing him by Artabane2s and
making Germanos, sent out with another army, commander in chief in
Italy. Germanos died in the following year and Justinian appointed
Germanos’s son-in-law John, son of Vitalian, and his son Justinian to
lead to Italy the army which Germanos had gathered at Sofia. They
intended to winter at Salo2nes before going around the north of the
Adriatic because they had no ships. Meanwhile Liberios had sailed to
Sicily and forced entry to the besieged Syracuse. Artabane2s’ fleet
following him from Kefalle2nia was scattered by storm off Calabria
and driven to Malta. Totila and the Goths had meanwhile plundered
Sicily for grain and treasure and took it back to Italy on their ships.25
While Narse2s was marching to Salo2nes in 551 to take command of
------------------------------
23
Prokopios, History of the wars, VII.xxx.1, VII.xxx.9-14, VII.xxx.15 (vol. 4, pp.
406, 408-10, 410).
24
Prokopios, History of the wars, VII.xxxv.1-2, VII.xxxv.23-30, VIII.xxiii.1-3
(vol. 4, pp. 458, 464-6; vol. 5, pp. 286-8).
25
Prokopios, History of the wars, VII.xxxvii.5, VII.xxxix.6-10, VII.xl.10-19 (vol.
5, pp. 12-14, 28-30, 40-44).
18 CHAPTER ONE
the forces under John and Justinian before going on to Italy, Totila
manned 300 ploia makra and sent them to Corfu. They reached and
plundered it and the opposite mainland and then sailed along the coast
capturing many Roman ships, including some carrying provisions to
Narse2s. The imperial commander at Ravenna sent a message to John
at Salo2nes asking him to relieve Indoulph’s siege of Ancona and,
contrary to his orders, John manned 38 ploia makra and sailed from
Salo2nes. The Goths at Ancona sailed out to give battle and the two
fleets met off Senogallia in the only naval engagement of the war, the
Goths being defeated and only 11 ships under Indoulph escaping.
These were burned to prevent their falling into Byzantine hands and
this led to the abandonment of the siege of Ancona, the Goths
retreating to Osimo. Prokopios wrote that Senogallia broke the spirit
and weakened the power of Totila and the Goths. By now Artabane2s
in Sicily had reduced all of the Gothic fortresses in the island,
although Totila could still assemble a fleet to send to Corsica and
Sardinia to subjugate both islands. The Byzantine commander in
Libya, another John, sent a fleet against them but was defeated outside
Cagliari and retired to Carthage.26
The long war was brought to a conclusion by the victories of
Narse2s over Totila at Busta Gallorum near Gualdo Tadino in late June
or early July 552 and over his successor Teias at the “Milk Mountain”
beneath Mt Vesuvius on 30 October or November. In this last phase of
the war control of the sea proved critical. In spring 552 the garrison in
Crotone was under siege by the Goths and Justinian ordered the
garrison of Thermopylae in Greece to sail to its relief, which it did
successfully. In 552 when Narse2s moved from Salo2nes against the
Goths he marched around the head of the Adriatic, which must have
meant that he did not have enough ships to ship all his troops across
the Adriatic, although he did have some, as Prokopios suggested. In
552, when Narse2s was facing Teias across the river Drako2n near Mt
Vesuvius, the Goths at first controlled the sea and were able to bring
in provisions by ship. However, Narse2s captured their ships through
an act of treason on the part of the Goth in charge of their shipping
and this forced Teias into the battle of the “Milk Mountain”.27
In 561 the last Gothic garrisons in Verona and Brescia capitulated
------------------------------
26
Prokopios, History of the wars, VIII.xxii.17-32, VIII.xxiii.4-9, VIII.xxiii.10-38,
VIII.xxiii.42, VIII.xxiv.3, VIII.xxiv.31-6 (vol. 5, pp. 282-6, 288-90, 290-300, 302,
312-14).
27
Prokopios, History of the wars, VIII.xxv.24 - xxvi.2, VIII.xxvi.1-25, VIII.xxxv.
12-38 (vol. 5, pp. 324-6, 326-36, 410-18).
and the Empire again controlled Italy, all of the islands, and all
Mediterranean coasts except for the strip held by the Visigoths in
Spain and the Franks in the Languedoc and Provence. But neither of
these were bellicose at sea and the unity of the Mediterranean was
restored again until the invasion of Italy by the Lombards from 568.
The second period, ca 560-750: the Muslim assault and imperial

recovery
The Lombard invasion of Italy under Alboin in 568 was precipitated

by pressure on their Pannonia homeland from the Avars in the mid
sixth century. A nomadic Turkic people, the Avars first made contact
with the Empire in 558, conquering and eliminating the Kutrigurs and
Antai north of the Black Sea and the Gepids in Dacia. Once
established in Pannonia around the confluences of the Danube, Sava,
and Tisza rivers, inevitable frictions with the Empire eventually led to
a combined Avar and Persian siege of Constantinople in 626.
However, the Avars’ dugout canoes were destroyed by imperial
squadrons and the siege dissipated, after which they suffered defeats
at the hands of the Croats and Bulgars and their threat diminished.28
The Lombard invasion precipitated the flight of the Roman
populace of Aquileia to the islands of the lagoons and led to the
foundation of Venice. The northern inland cities quickly fell, leaving
in imperial hands only coastal strongholds which could be supplied
from the sea. Pavia fell after three years and within seven years most
of Italy had been occupied. In 571 they swept into southern Italy,
taking Benevento and establishing a southern duchy centred on that
city which would become a principality and which, together with its
twin at Spoleto, would dominate south Italy for hundreds of years.
Between 584 and 588 the Romans in Ravenna built a fleet and, with
the help of a disgruntled Lombard duke, drove the Lombards from its
port of Classe, thus establishing the exarchate of Ravenna. Imperial
presence in Italy became confined to the exarchate and a belt of
territory running south-west to Rome, together with most of Apulia
and Calabria. Over the next two centuries it waned progressively, Rome
being lost to the Papacy during the first half of the eighth century,
and by the opening of the ninth century was confined to southern
------------------------------
28
Chronicon Paschale, Annus 726 (pp. 715-26); George of Pisidia, Bellum
Avaricum; Theophane2s, Chronographia, A.M. 6117 (p. 316).
20 CHAPTER ONE
Table 2: Rulers of the second period, ca 560-750
Byzantine Empire The Muslims
Muh5ammad and the

Governors of al-Andalus
Caliphs
Justin II (565-78)
Tiberios II (578-82)
Maurice (582-602)
Pho2kas (602-10)
Herakleios I (610-41) The Prophet Muh5ammad
(to 632)
The Rightly Guided

Caliphs
Abu2 Bakr (632-4)

‘Umar ibn al-Khat6t6a2b
(634-44)
Constantine III (641)
Heraklo2nas (641)
Constans II (641-68) ‘Uthma2n ibn ‘Affa2n (644-
56)
‘Alı3 ibn Abı3 T4a2lib (656-
61)
The Umayyad Caliphs
Mu‘a2wiya I ibn Abı3-

Sufya2n (661-80)
Constantine IV (668-85) Yazı3d I (680-83)
Mu‘a2wiya II (683-4)
Marwa2n I ibn al-H4akam
(684-5)
Justinian II (685-95) ‘Abd al-Malik (685-705)
Leontios (695-8)
Tiberios III (698-705)
Justinian II (705-11) Al-Walı3d I (705-15)
Philippikos (711-13) ‘Abd al-‘Azı3z ibn Mu2sa2

(714-16]
(Table 2 continued)
Lombards Visigoths Merovingians on the

Mediterranean
(K) Kings
(K) Kings
(M) Arnulfing Mayors of
(B) Dukes of Benevento
the Palace
Alboin (K 568-72) Guntram (K 561-92)

Zotto (B 571-91)
Cleph (K 572-4)
Authari (K 584-80) Recared I (586-601) Childebert II (K 593-5)

Interregnum Leova II (601-3) Theodoric II (K 595-613)
Agilulf (K 590-616)
Arichis I (B 591-641)
Witterich (603-10)
Adaloald (K 616-26) Gundemar (610-12) Sigibert II (K 613)
Ariold (K 626-36) Sisebut (612-21) Clovis II (K 639-56)
Rothari (K 636-52 Recared II (621)
Swinthila (621-31)
Sisenand (631-6)
Chintila (636-40
Tulga (640-42)
Ayo I (B 641-2) Chindaswinth (642-52) Clothar III (K 656-61)

Radoald (B 642-6) Receswinth (653-72)
Grimoald I (B 646-62)
Rodoald (K 652)
Aribert I (K 652-61)
Godepert/
Perctarit (K 661-2)
Grimoald (K 662-71)
Perctarit (K 671-88) Wamba (672-80) Childeric I (K 673-75)

Romoald I (B 662-87) Erwig (680-87) Theodoric III (K 675-90)
Pepin II (M 680-714)
Grimoald II (B 687-92) Egica (687-701) Clovis III (K 690-4)

Gisulf I (B 692-706) Childebert III (K 694-711)
Cunipert (K 688-700)
Luitpert (K 700) Witiza (701-9)

Aribert II (K 700-12)
Romoald II (B 706-30)
Roderick (709-11)
Muslim conquest
Ansprand (K 712) Dagobert III (K 711-15)
22 CHAPTER ONE
(Table 2 continued)
Muh5ammad and the

Governors of al-Andalus
Caliphs
Anastasios II (713-15)
Theodosios III (715-17) Sulayma2n (715-17) Ayyu2b ibn H4abı3b al-
Lakhmi (716)
Al-H4u rr ibn ‘Abd al-
Rah5m a2n al Thaqafı3
(717-19)
Leo III (717-40) ‘Umar ibn ‘Abd al-‘Azı3z Al-Samh5 ibn Malik al-
(717-20) Khawla2nı3 (719-21)
Yazı3d II (720-24) ‘Abd al-Rah5m a2n al-
Hisha2m (724-43) Gha2fiqı3 (721)
‘Anbasa ibn Suh5aym al-
Kalbı3 (721-5)
‘Udhra ibn ‘Abd Alla2h al-
Fihrı3 (725-6)
Yah5ya2 ibn Sala2m a al-
Kalbı3 (726-8)
H4udhayfa ibn al-Ah4was5
al-Qaysı3 (728-9)
‘Uthma2n ibn Abı3 Nas5r al-
Khath‘amı3 (729)
Al-H4aytham ibn ‘Ubayd
al-Kila2b ı33 (729-30)
Muh5ammad ibn ‘Abd
Alla2h al-Ashja‘ı3 (730)
‘Abd al-Rah5m a2n ibn ‘Abd
Alla2h al-Gha2fiqı3 (730-
32)
‘Abd al-Malik ibn Qat6an
al-Fihrı3 (732-4, 740-
41)
‘Uqba ibn al-H4ajja2j al-
Salu2lı3 (734-40)
Constantine V (740-75) Al-Walı3d II (743-4) Tha‘laba ibn Sala2m a al-
Yazı3d III (744) ‘A›milı3 (742-3)
Ibra2hı3m (744) Abu2 al-Qat6t6a2r al-H4usa2m
Marwa2n II al-H4ima2r (744- (743-5)
50) Thawa2b a ibn Yazı3d (745-
6)
‘Abba2sid Caliphs Yu2suf ibn ‘Abd al-
Rah5m a2n al-Fihrı3 (746-
Al-Saffa2h5 (749-54) 56)
Al-Mans5u2r (754-75) _________
Umayyad amı3rs
(Table 2 continued)
Lombards Visigoths Merovingians on the

Mediterranean
(K) Kings
(K) Kings
(M) Arnulfing Mayors of
(B) Dukes of Benevento
the Palace
Liutprand (K 713-44) Charles Martel (M 714-41)
Audelaius (B 730-32) Chilperic II (K 719-20)

Gregory (B 732-9) Theodoric IV (K 721-37)
Godescalc (B 739-42)
Gisulf II (B 742-51) Carloman (M 741-7)

Hildebrand (K 744) Pepin III (M 741-51)
Ratchis (K 744-49) Childeric III (K 743-51)
Aistulf (K 749-56) _________
Desiderius (K 756-74)
Liutprand (B 751-58) Carolingian
Arichis II (B 758-87) kings
24 CHAPTER ONE
Apulia and Calabria.29 At some time prior to 725 the Lombard Duke
Farwald II of Spoleto took Classe from the Byzantines; however,
King Liutprand ordered him to restore it and the Lombards do not
appear to have been concerned to gain control of the Italian coasts
except when Rothari swept up the Byzantine outposts on the
Tyrrhenian from Luni to Provence around 642 and Aistulf finally took
Ravenna in 751.30 Although they certainly disrupted Italy, and
possibly had an encounter at sea with Byzantine forces in Sardinia,31
they never assumed a Mediterranean presence such as that of the
Vandals and they had little impact on the Sea as a whole. Its maritime
integrity remained in imperial hands.
In 627 He2rakleios brought the long Romano-Persian conflict to a
successful conclusion when he led the Byzantine armies into Persia
and won a decisive victory over Khusraw II near Nineveh, effectively
ending the Persian Empire. In Constantinople it would have appeared
that the world had been restored to rights.32 The East was secure, the
Empire again controlled the sea, and the Visigoths, Franks, and
Lombards were mostly confined to hinterlands and posed no threat.
The emperor might look forward to a long and peaceful reign.
However, it was not to be so for a bare nine years later the forces of
the newly emergent Muslim Caliphate annihilated the imperial armies
in Transjordan at the battle of the Yarmu2k in August or September of
636. The Muslims occupied Syria and Palestine and in 640-42 Egypt
also fell to them. The religious unity of the Mediterranean world was
broken. The assault of the Arian Vandals, Visigoths, and Ostrogoths
had been as nothing compared to that which the Muslims were about
to unleash.
While the Byzantines still had mastery at sea and could attack at
will, as in 645-6 when they reoccupied Alexandria and raised a revolt
in Egypt, watch towers and a signalling system were established along
the coasts. However, the governors of Syria and Egypt, Mu‘a2wiya ibn
Abı3 Sufya2n and ‘Abd Alla2h ibn Sa‘d ibn Abı3 Sarh5 respectively, began
to create naval forces, at first crewed by native Christians.
------------------------------
29
Paul the Deacon, Historia Langobardorum, II.6-8, II.10, II.14, II.15-17, II.32,
III.32-3 (pp. 75-7, 78, 81, 81-2, 90-91, 112). See also Delogu, Longobardi e Bizantini.
30
Fredegar, Continuations, §72 (p. 60); Paul the Deacon, Historia Langobar-
dorum, III.18-19, IV.45, VI.44 (pp. 101-2, 135, 180).
31
See Fiori, Cosentino hypatos.
32
The mood of the years is reflected by Theophane2s the Confessor, followed by
his despair following the battle of the Yarmu2k. See Theophane2s, Chronographia,
A.M. 6118-21 (pp. 317-32). See also Nike2phoros I, Historia syntomos, §§12-17 (pp.
54-65).
Construction began on al-Rawd4a island in the Nile opposite al-Fust6a2t6

and the fleet first went into action against Cyprus in 649, subjecting
the island to a covenantary status of ‘ahd, a covenant of peace which
Muslims could make with non-Muslim peoples living outside the
Muslim polity, under which the Cypriots were to remain neutral
between the Empire and the Caliphate. Crete, Rhodes, and Sicily were
raided between 652-4 and the fleet also returned to Cyprus in 653.33
There is also a possibility, although only a remote one, that
Mu‘a2wiya’s forces attacked Constantinople itself in 654.34
In 655 the first real hammer blow to the maritime integrity of the
Mediterranean fell. A Muslim fleet under ‘Abd Alla2h ibn Sa‘d
engaged the main Byzantine fleet commanded by Constans II off
Phoinikous in Lycia. The Byzantines were annihilated and the
emperor lucky to escape with his life. This so-called “Battle of the
Masts”, “Dha2t al-S4awa2rı3”, opened the central Mediterranean to
Muslim attack, even if the Muslims did suffer heavily in it and Cyprus
reverted to Byzantine rule under a truce concluded between Constans
and Mu‘a2wiya in 659. The Byzantine fleet of the Karabisianoi, based
on Samos, was probably created as a front line of defence shortly
thereafter.35
Sometime after 660 occurred the curious episode of Constans II
leaving Constantinople, moving to Italy, where he arrived in 663,
campaigning against the Lombards and advancing on Rome, but
eventually retiring to Syracuse in Sicily, where he settled, established
a thema, created an army and fleet, and was eventually murdered in
668. His son, Constantine IV, most probably led a fleet from
Constantinople to Sicily in 669 to avenge him.36 Constans’ motives for
------------------------------
33
Agapios, Kita2b al-‘Unwa2n, tome 8, fasc. 3,pp. 455, 480, 482; Al-Bala2dhurı3,
Kita2b Futu2h5 al-Bulda2n, part 2, ch. 13, part 5, ch. 2, part 7, ch. 1 (vol. 1, pp. 235-6,
347-8, 375); Al-Nuwayrı3, Niha2yat ‘al-Arab (Amari), p. 112; Al-T4abarı3, Ta’rı3kh (Yar-
Shater), A.H. 28 (vol. 15, pp. 25-32); Denys of Tell-Mah5ré, Chronique, p. 7; Paul the
Deacon, Historia Langobardorum, V.13 (p. 150); Pseudo al-Wa2qidı3, Futu2h5 al-S!a2m
wa-Mis5r, pp. 329-38; Theophane2s, Chronographia, A.M. 6140, 6145 (pp. 343-4,
345). See also Beihammer, “Zypern”; Bosworth, “Arab attacks on Rhodes”; Cheira,
Lutte, pp. 88-101; Vasiliev, Byzance et les Arabes. Tome I, pp. 61-3.
34
The argument is based on Sebeos, Armenian history, §50 (vol. 1, pp. 143-6) and
vol. 2, pp. 274-6. See O’Sullivan, “Arab attack”. In our opinion the evidence is not
convincing.
35
Agapios, Kita2b al-‘Unwa2n, tome 8, fasc. 3, pp. 483-4; Al-T4abarı3, Ta’rı3kh (Yar-
Shater), A.H. 31, 34 (vol. 15, pp. 74-7, 131); Sebeos, Armenian history, §45 (vol. 1,
pp. 111-12) and vol. 2, pp. 259-62; Theophane2s, Chronographia, A.M. 6146 (vol. 1,
pp. 345-6). Christides, “Dha2t as5-S4awa2rı3”; Stratos, “Naval engagement at Phoenix”.
36
Agapios, Kita2b al-‘Unwa2n, tome 8, fasc. 3, pp. 490-91; Gesta episcoporum
Neapolitanorum, pp. 417-8; Paul the Deacon, Historia Langobardorum, V.6-11 (pp.
146-50); Theophane2s, Chronographia, A.M. 6153-60 (pp. 348-52). On the
26 CHAPTER ONE
abandoning Constantinople may have had as much to do with the

precariousness of his domestic position there as with any desire to re-
establish Rome as the imperial capital or some notion that the Empire
could best be defended from the central Mediterranean.
When Mu‘a2wiya became Caliph in 661 and established the capital
of the Caliphate at Damascus, he began to use his naval capabilities to
good effect. Rhodes was again assaulted in 667 and was occupied by a
garrison and squadron in 673 until abandoned after Mu‘a2wiya’s death
in 680. In 669 the Egyptian fleet sailed to attack Byzantine Africa,
although it returned after wintering over in Sicily. Cyprus was
probably reoccupied around 670 in preparation for the forthcoming
great assault on Constantinople and Crete was also attacked in 672 as
part of the same assault. According to the so-called Chronicle of
Alfonso III, during the reign of Wamba a Muslim fleet of 270 ships
attacked Visigothic Spain but was destroyed and burned. However, it
is improbable that this expedition actually occurred because all
Umayyad forces were committed at the time to the assault on
Constantinople.37
The assault began most probably in 671-2 when two fleets entered
the Aegean and wintered at Izmir and in Cilicia and Lycia. It is
possible, although unlikely, that during this preliminary phase of the
assault a battle occurred off Lycia in which the Byzantines used a new
incendiary, “Greek Fire”, for the first time.38 In 672 the Muslims
moved into the Sea of Marmara and began the siege, which lasted for
seven years, although it was not maintained as a close blockade
------------------------------
Karabisianoi see also Antoniadis-Bibicou, “Stratège des Caravisiens”; idem, “Thème
des Caravisiens”; Cosentino, “Flotte byzantine”, pp. 4-7; Whittle, “Carabisiani”.
37
Al-Bala2dhurı3, Kita2b Futu2h5 al-Bulda2n, part 7, ch. 1 (vol. 1, pp. 375-6); Al-
Nuwayrı3, Niha2yat ‘al-Arab (Caussin), p. 402; Al-T4abarı3, Ta’rı3kh (Yar-Shater), A.H.
53 (vol. 18, p. 166); Chronicle of Alfonso III, pp. 6-7, 72, 110; Gesta episcoporum
Neapolitanorum, §31 (p. 419); Ibn ‘Idha2rı3, Al-baya 2n al-mughrib, vol. 1, pp. 10-13;
Sa2wı3ris, History of the Patriarchs (Evetts), I.15 (vol. V.1, p. 4).
38
Three Byzantine patrikioi were reported by Agapios and by the much later
Jacobite Patriarch of Antioch, Michael the Syrian, to have defeated a Muslim fleet,
the remnants of which were then destroyed by a new fire weapon developed by a
Greek according to Agapios or by a Syrian named Kallinikos according to Michael
the Syrian. See Agapios, Kita2b al-‘Unwa2n, tome 8, fasc. 3, p. 492; Michael the
Syrian, Chronique, XI.xiii (vol. 2, p. 455). However, the story is not confirmed by any
Greek or Muslim source and neither Agapios nor Michael the Syrian recounted any
further naval conflict between the Byzantines and Muslims until after the death of
Mu‘a2wiya, both passing over the subsequent assault on Constantinople in silence.
Their chronology is also very confused at this point. It seems most likely that they or
their sources confused the development and use of Greek Fire during the subsequent
defence of Constantinople with some report of a prelimary success over the
approaching Muslim fleets, which may or may not have actually taken place.
throughout. Muslim squadrons variously retired to Kyzikos, Crete, and

Rhodes to winter over, returning each spring. In the end Greek Fire
shot from weapons mounted on the prows of dromons annihilated the
Muslim fleets and forced the lifting of the siege. The remnants of the
Muslim armada were destroyed by storms during the retreat and
Mu‘a2wiya had to conclude a thirty-year truce and his son Ya2zid I had
to evacuate Cyprus and Rhodes.39
Until the end of the century the focus of conflict moved to Africa.
As early as 643 the governor of Egypt, ’Amr ibn al-’As, had launched
a tentative against Tripoli and then in 665 Mu‘a2wiya had sent an army
under Mu‘a2wiya ibn H4udayj al-Saku2nı3 to Ifrı3qiya. Although it
accomplished little, during the expedition a force under ‘Abd Alla2h
ibn al-Zubayr encountered and defeated a Byzantine amphibious force
commanded by a patrikios named Neqfur, Nike2phoros, at Hadru-
metum. Mu‘a2wiya ibn H4udayj also founded what was to become the
capital of Ifrı3qiya, al-Qayrawa2n. In 669 ‘Uqba ibn Na2fi‘ al-Fihrı3
returned, establishing an advance base at al-Qayrawa2n . From there he
raided into the interior against the Berber tribes. However, in 681 he
overreached himself with a long-range expedition across the entire
Maghrib to Tangier and down the Atlantic coast to near Aga2dı3r.
Byzantine naval forces cut his lines of communication and during his
return he was defeated and killed by a coalition of Berber tribes under
the convert Kusayla or Kası3la ibn Lamzan, the chief of the Awraba
Berbers, and Byzantine forces at Tahu2da near an old Roman fortress
called Thabudeus in Algeria. The Berbers and Byzantines then took
al-Qayrawa2n and forced the Muslims to evacuate back to Barqa on
the Egyptian frontier.40
They did not return to the attack until 688 when a new Caliph,
‘Abd al-Malik, sent an old lieutenant of ‘Uqba ibn Na2fi‘, Zuhayr ibn
Qays al-Balawı3, back to Ifrı3qiya. Zuhayr defeated and killed Kusayla
------------------------------
39
Al-T4abarı3, Ta’rı3kh (Yar-Shater), A.H. 54-9 (vol. 18, pp. 172-99); Theophane2s,
Chronographia, A.M. 6164-5 (pp. 353-4), 6169 (pp. 355-6). Sebeos almost certainly
confused this expedition with earlier plans for an assault on Constantinople which he
dated to 654. Sebeos, Armenian history, §50 (vol. 1, pp. 143-6) and vol. 2, pp. 274-6.
40
Agapios, Kita2b al-‘Unwa2n, tome 8, fasc. 3, p. 491; Akhba2r Majmu2‘a, p. 17; Al-
Bakrı3, Kita2b al-mughrib, pp. 32-5, 75-6, 150-51, 213-14; Al-Bala2dhurı3, Kita2b Futu2h 5
al-Bulda2n, part 5, ch. 5 (vol. 1, pp. 356-60, 362); Al-Nuwayrı3, Niha2yat ‘al-Arab (De
Slane), vol. 1, pp. 324-36; Al-Tija2nı3, Rih5la, ser. 4, 20, pp. 104-6, ser. 5, 1.1, pp. 138-
9; Al-Ya‘qu2bı3, Al-Bulda2n, pp. 208-9; Ibn ‘Abd al-H4akam, Afrique, pp. 56-75; Ibn al-
Athı3r, Al-Ka2mil (Fagnan), pp. 9-25; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 1, pp. 10-
19; Ibn Khaldu2n, ‘Ibar (De Slane), vol. 1, pp. 210-11, 311, 327-30; idem, ‘Ibar (Des
Vergers), pp. 5-22; Nike2phoros I, Historia syntomos, §34 (pp. 84-7); Theophane2s,
Chronographia, A.M. 6161 (p. 352). See also T4a2ha, Muslim conquest.
28 CHAPTER ONE
near the Mams, about 50 kilometres west of al-Qayrawa2n, but was

forced to retire by Byzantine amphibious forces which had reoccupied
Barqa in his rear. He was killed in battle near Darna in 690.41 Five
years later the conquest of Ifrı3qiya began in earnest. A huge army
under H4assa22n ibn al-Nu‘ma2n al-Ghassa2nı3 reduced the Byzantine
fortresses one by one and finally captured Carthage, the Byzantine
garrison evacuating by sea. A Berber uprising led by a mysterious
queen or soothsayer known as al-Kahı3na was accompanied by a
Byzantine amphibious assault on Carthage under a patrikios named
John. However, al-Kahı3na was defeated and, not having the forces to
resist the Muslims, John was forced to evacuate Carthage and retire to
Crete for reinforcements. He had never had the forces necessary to
resist the Muslims on his own and the days of Byzantine Africa were
over. On Crete John was murdered by mutineers who proclaimed
Apsimaros, the droungarios of the Kibyrrhaio2tai, emperor and who
then sailed on Constantinople.42
Since the harbour of Carthage had proved too vulnerable to attack
from the sea, H4assa2n commenced building a new Muslim capital and
fortress arsenal at Tunis by connecting an inland lake to the sea by a
canal through the coastal strip. The governor of Egypt, ‘Abd al-‘Azı3z,
sent 1,000 Coptic shipwrights to populate the new city and to
construct a fleet of 100 warships which, under a new governor, Mu2sa2
ibn Nus5ayr, from 704 began to open the way to the conquest of the
Maghrib by denying the Byzantines access to remaining outposts. The
fleet also began to raid across the western Mediterranean to Sicily,
Sardinia, and the Balearics.43
------------------------------
41
Al-Bakrı3, Kita2b al-mughrib, p. 22; Al-Bala2dhurı3, Kita2b Futu2h5 al-Bulda2n, part 5,
ch. 5 (vol. 1, p. 360); Al-Nuwayrı3, Niha2yat ‘al-Arab (De Slane), vol. 1, pp. 337-8; Ibn
‘Abd al-H4akam, Afrique, pp. 74-7, 80-85; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 1,
pp. 19-22; Ibn Khaldu2n, ‘Ibar (De Slane), vol. 1, pp. 211-12-; idem, ‘Ibar (Des
Vergers), p. 23.
42
Al-Bakrı3, Kita2b al-mughrib, pp. 81-4; Al-Bala2dhurı3, Kita2b Futu2h5 al-Bulda2n, part
5, ch. 5 (vol. 1, pp. 360-1); Al-Ma2liki, Muh@tas5ar Kita2b Riya2d5, pp. 295-9; Al-Nuwayri,
Niha2yat ‘al-Arab (De Slane), vol. 1, pp. 338-43; Al-Tija2nı3, Rih5la, ser. 4, 20, pp. 67,
119-22, 202-3; Ibn ‘Abd al-H4akam, Afrique, pp. 76-87; Ibn al-Athı3r, Al-Ka2m il
(Fagnan), pp. 25-35; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 1, pp. 23-31; Ibn
Khaldu2n, ‘Ibar (De Slane), vol. 1, pp. 213-15, vol. 3, pp. 192-4; idem, ‘Ibar (Des
Vergers), pp. 24-8; Nike2phoros I, Historia syntomos, §41 (pp. 98-101); Theophane2s,
Chronographia, A.M. 6190 (pp. 370-1).
43
Al-Bakrı3, Kita2b al-mughrib, pp. 83-5; Al-Bala2dhurı3, Kita2b Futu2h5 al-Bulda2n, part
5, ch. 6 (vol. 1, p. 362); Al-Nuwayrı3, Niha2yat ‘al-Arab (Caussin), pp. 402-4; idem,
Niha2yat ‘al-Arab (De Slane), vol. 1, pp. 343-4, 357-9; Al-Tija2nı3, Rih5la, ser. 4, 20, pp.
69-70; Ibn ‘Abd al-H4akam, Afrique, pp. 86-87; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol.
1, pp. 31-6; Ibn Khaldu2n, ‘Ibar (Des Vergers), pp. 28-30; idem, Muqqadimah, vol. 2,
p. 40; Pseudo ibn Qutaybah, Ah5a2d ı3th al-Ima2mah, pp. lxv-lxvii.
The circumstances leading to the Muslim invasion of Spain are

obscure. For some reason T4a2riq ibn Ziya2d , the governor of Tangier,
sent an exploratory force across the Straits under Abu2 Zar‘a T4arı3f ibn
Malik al-Mu‘a2firı3 in 710 on four ships provided for reasons of his own
by the governor of Ceuta, a certain Count Julian. The success of the
probe persuaded T4a2riq to lead a full-scale follow-up himself in the
following year, again in ships provided by Julian, landing at the foot
of what was to become known as Jabal T4a2riq, Tariq’s Mount or
Gibraltar, probably in April 711. The Muslim forces were not large
but the Visigothic king, Roderick, was away in the north-east
occupied in the Basque country. He marched south attempting to rally
all available Visigothic forces and the armies met somewhere around
the Guadalete river. The king disappeared in the battle and the
Visigothic kingdom then disintegrated in a welter of local
insurgencies by governors unable to resist T4a2riq’s forces. T4a2riq’s
successes prompted a flood of Muslim adventurers from the Maghrib
to cross to Spain and in Ramad5a2n A.H. 93 (June-July 712 C.E.) Mu2sa2
ibn Nus5ayr himself landed at Algeciras with a new army of 18,000
men. T4a2riq met his master outside Toledo and, after having resolved
initial hostilities, the two proceeded to pacify the peninsula, with the
exception of part of the Asturian mountains. Although Muslim rule
remained shaky for some time and the Caliph ‘Umar ibn ‘Abd al-
‘Azı3z reportedly considered abandoning the conquest in 718, the
Muslims consolidated their rule and would eventually push north
across the Pyrenees. Seville became the naval base of the new
province: al-Andalus.44
A Visigothic noble named Pelagius or Pelayo whose power base
lay in the Asturian mountains raised a revolt against Muslim rule and
was attacked by forces sent by the governor, ‘Anbasa ibn Suh5aym, but
won a victory near the rock of Covadonga, on which he was later
besieged by a punitive expedition under the governor ‘Uqba ibn al-
H4ajja2j. He defied the Muslims, reportedly until they abandoned the
------------------------------
44
Akhba2r Majmu2‘a, pp. 19-31; Al-Bakrı3, Kita2b al-mughrib, p. 204; Al-Bala2dhurı3,
Kita2b Futu2h5 al-Bulda2n, part 6, ch. 1 (vol. 1, pp. 365-6); Al-Marra2kushı3, Al-Mu‘jib, pp.
7-10; Al-Nuwayrı3, Niha2yat ‘al-Arab (De Slane), vol. 1, pp. 345-53; Al-Qurt6ubı3,
Kita2bu-l-iktifa2, pp. xliii-l; Al-T4abarı3, Ta’rı3kh (Yar-Shater), A.H. 92, 93 (vol. 23, pp.
182, 201); Fath5 al-Andalus, pp. 2-15; Ibn ‘Abd al-H4akam, Spain, pp. 18-25; Ibn al-
Athı3r, Al-Ka2m il (Fagnan), pp. 35-50, 53; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 1, pp.
36-7 & vol. 2, pp. 5-25, 34-5; Ibn Khaldu2n, ‘Ibar (De Slane), vol. 1, p. 215, vol. 2, p.
136; Pseudo ibn Qutayba, Ah5a2dı3th al-Ima 2ma, pp. lxx-lxxviii; Chronicle of Albelda,
§46 (col. 1136); Chronicle of Alfonso III, pp. 14-17, 75-6,114-5; Chronicle of 754,
§§68-73 (pp. 352-4); Historia pseudoIsidoriana, pp. 387-8. See also T4a2ha, Muslim
conquest, pp. 84-109; Collins, Arab conquest, esp. pp. 23-51.
30 CHAPTER ONE
siege when only 30 men and 10 women survived of the Christians.45

Thus was born the Kingdom of the Asturias, from which the Spanish
Reconquista would eventually develop.
The Muslim conquest was in some ways relatively peaceful. Being
vastly outnumbered, the Muslims simply had to accomodate the
Visigothic nobility: the most well-known example being the
agreement between ‘Abd al-‘Azı3z ibn Mu2sa2 and duke Theodamı3r,
dated 5 April 713. In return for submission he, his lords, and the
inhabitants of his seven towns were confirmed in their possessions
subject to payment of an annual tribute: the jizya. However, this was
not always the case. At Narbonne the turning of the town into a
Muslim military encampment was preceded by wholesale slaughter of
the men and enslavement of the women and children.46
Successive governors of al-Andalus both pressed against Christian
resistance in the northern mountains and also crossed the Pyrenees.
Al-H4urr ibn ‘Abd al-Rah5ma2n al-Thaqafı3, appointed governor in 716,
invaded the Languedoc. Al-Samh5 ibn Malik al-Khawla2nı3 marched
through Zaragoza to Narbonne but was killed in battle near Toulouse
in 721. The next governor but one, ‘Anbasa ibn Suh5aym al-Kalbı3,
captured Carcassonne and reached as far as Nîmes, Autun, and Sens,
30 kilometres from Paris, dying during the withdrawal in 726. The
most famous of all, ‘Abd al-Rah5man al-Gha2fiqı3, brought back vast
booty; although, his final expedition ended in his defeat and death at
the hands of Charles Martel at the battle of Tours in 732. ‘Uqba ibn al-
H4ajja2j al-Salu2lı3 resumed the campaigns. In 737 he sent an army by sea
to relieve Narbonne but it was defeated after landing by Charles
Martel. Gothic Septimania across to the Rhône became a Muslim
march centred on Narbonne, the Muslim presence being welcomed by
some indigenous aristocracy as a counter-balance to the pretension of
the Franks under Charles Martel. The Muslim assault on Gaul petered
out eventually only because of over-extended resources and of a
Berber revolt which broke out in the Maghrib in 739 and spilled over
into al-Andalus. Charles Martel’s victory at Tours was not nearly as
decisive as it has been made out to be traditionally.47
------------------------------
45
Akhba2r Majmu 2‘a, pp. 38-9; Chronicle of Albelda, §50 (coll. 1136-7); Chronicle
of Alfonso III, pp. 18-29, 80-83, 115-8. On the Pelayo legend and the contradictory
accounts of his revolt see Collins, Arab conquest, pp. 141-51.
46
Chronicle of Moissac, Annus 715 (p. 290).
47
Akhba2r Majmu2‘a, pp. 36, 39-54; Al-Marra2kushı3, Al-Mu‘jib, pp. 10-13;
Chronicle of Moissac, Anni 715, 725, 732, 734 (pp. 290-92); Chronicle of 741, §2 (p.
14); Chronicle of 754, §§80, 86, 90, 103-6, 109-10, 112-25 (pp. 356, 358, 359 361-2,
362-3, 363-5); Fath5 al-Andalus, pp. 28-9, 31, 34-47; Fredegar, Continuations, §§13,
The second assault on Constantinople by the Muslims caught the

Empire at a nadir in its fortunes following two politically tumultuous
decades. In 695 Justinian II had been overthrown by the strate2gos of
Hellas, Leontios. He in turn was overthrown by Apsimaros, Tiberios
III, in 698. In exile at Cherso2n Justinian II married the sister of the
Khan of the Khazars and in 705, with the help of the Bulgar Khan
Tervel, he regained the empire, executing both Leontios and Tiberios.
He was overthrown in his turn by a mutiny in the fleet he sent to
Cherso2n in 711 led by its strate2gos, Bardane2s, who became emperor
Philippikos. Following Tervel’s devastation of Thrace in reprisal for
the overthrow of Justinian II and Muslim attacks along the Asiatic
frontiers, the army of the thema of Opsikion revolted and raised the
pro2tase2kre2tis Artemios to the throne as Anastasios II. An expedition
into Galatia in 714 led by Maslama ibn ‘Abd al-Malik, a brother of
the Caliph al-Walı3d I, led Anastasios to sue for peace but al-Walı3d’s
preparations for a large-scale assault on Constantinople prompted him
to prepare for siege, readying the fleet, repairing the walls and
mounting engines on them, collecting provisions, and ordering those
who could not lay up sufficient for three years to leave the city.
However, a pre-emptive naval expedition sent to destroy the Muslims’
fleet and timber supplies in Lycia broke up in disarray in Rhodes and
the army of Opsikion again revolted, this time in favour of a tax
collector from Edremit who assumed the throne as Theodosios III.48
When the Muslim assault on the capital began to gather momentum in
717, he in turn was overthrown by the strate2gos of Anatolikon, who
became emperor as Leo III on 25 March 717.
The assault had begun in 715 with Maslama’s army moving into
Anatolia and wintering the first year in Cilicia and the next at Nicaea.
It marched on Constantinople in spring 717 but the fleet did not arrive
until 1 September. For the first time known in Byzantine history, the
entrance to the Golden Horn was closed by a chain, a{lusi" (alysis).
With many ships attacked and burned by Greek fire, starving and
freezing through a harsh winter, with Christian Egyptian crews of
------------------------------
20 (pp. 90-91, 93-5); Gesta abbatum Fontanellensium, cc. 9-12 (pp. 282-5); Gesta
episcoporum Neapolitanorum, §38 (p. 422); Ibn ‘Abd al-H4akam, Spain, pp. 33-8; Ibn
al-Athı3r, Al-Ka2m il (Fagnan), pp. 54-74, 84-9; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol.
1, pp. 50-66 & vol. 2, pp. 36-7, 39-57; Paul the Deacon, Historia Langobar-dorum,
VI.46, 54 (pp. 180-81, 183).
48
Agapios, Kita2b al-‘Unwa2n, tome 8, fasc. 3, pp. 499-501; Al-T4abarı3, Ta’rı3kh
(Yar-Shater), A.H. 96, 97 (vol. 24, pp. 28, 30); John Zo2naras, Epitome2 historio2n,
XIV.22-8 (vol. 3, pp. 323-33); Paul the Deacon, Historia Langobardorum, VI.36 (p.
177); Theophane2s, Chronographia, A.M. 6206-7 (pp. 383-5).
32 CHAPTER ONE
reinforcements which arrived the following spring deserting,

squadrons again destroyed by Greek Fire, and with forces around the
city attacked by Tervel, with whom Leo III made an alliance,
Maslama was ordered by the new Caliph, ‘Umar ibn ‘Abd al-‘Azı3z, to
abandon the siege in August 718. Most of the remaining fleet was
destroyed by storms during the retreat.49
Shortly thereafter the fleet of the Karabisianoi was dissolved and
replaced by two new naval commands: an imperial fleet, basilikon
plo2imon, at Constantinople and the fleet of the Kibyrrhaio2tai, now
become a naval thema, based at Antalya. The Karabisianoi either had
performed poorly during the siege or had supported the deposed
Anastasios II in a revolt against Leo in 719.50
In 726 Leo ordered the removal of an icon of Christ from the
Chalke2 entrance vestibule of the Great Palace in Constantinople, thus
precipitating the iconoclast dispute which would wrack and weaken
the Empire. In 727 the fleets of Hellas and the Cyclades islands
proclaimed a certain Kosmas as emperor and sailed on Constantinople
but were scattered by the imperial fleet using Greek Fire. When
imperial officials attempted to enforce iconoclasm in Italy revolts
flared and the Lombard king Liutprand seized Luni and perhaps
Corsica An expedition sent out under Mane2s , the strate2gos of the
Kibyrrhaio2tai, was wrecked in the Adriatic, probably in 732. In 735
Pope Gregory III and the Lombards drove the exarchos from Ravenna
but the Papacy soon fell out with the Lombards and in 742 Pope
Zachary and Venice returned it to the Empire. Disorder continued
until 787 when iconoclasm was condemned at the Second Council of
Nicaea. It resurfaced in the ninth century from 813 during the reigns
of Leo V, Michael II, and Theophilos until Empress Theodo2ra
restored iconophile orthodoxy in 843, but the second period lacked the
intensity of the first and was not nearly as destructive for the Empire.51
------------------------------
49
Agapios, Kita2b al-‘Unwa2n, tome 8, fasc. 3, pp. 500-502; Al-T4abarı3, Ta’rı3kh
(Yar-Shater), A.H. 98-9 (vol. 24, pp. 39-42, 74); Denys of Tell-Mah5ré, Chronique,
pp. 12-14; John Zo2naras, Epitome2 historio2n, XV.1, 2 (vol. 3, pp. 333-6, 337-8);
Nike2phoros I, Historia syntomos, §§49-57 (pp. 116-27); Theophane2s, Chronographia,
A.M. 6208-10 (pp. 386-99), esp. p. 396, l. 18.
50
On the Kibyrrhaio2tai, see Ahrweiler, Byzance et la mer, pp. 81-5, 131-5;
Cosentino, “Flotte byzantine”, 7-8; Savvides, “Secular prosopography”;
Yannopoulos, “Cibyrra”.
51
Genesios, Basileiai, D.2-3 (pp. 56-8); John Zo2naras, Epitome2 historio2n, XV.19
(vol. 3, pp. 381-2); Nike2phoros I, Historia syntomos, §§60, 72, 80-81 (pp. 128-31,
142-5, 152-5); Theophane2s, Chronographia, A.M. 6215-80 (pp. 401-63) passim, esp.
6215, 6217, 6218, 6224 (pp. 401-2, 404, 404-5, 410); Theophane2s continuatus, I.17;
II.8, 28; III.2, 10-14; IV.1-6 (pp. 28, 47, 83-4, 86-7, 99-106, 148-54).
The Umayyad assault petered out in a whimpering coda around

750, marked by the defeat of a large fleet off Cyprus in 747 by that of
the Kibyrrhaio2tai,52 by the overthrow of the Umayyads themselves by
the ‘Abba2sids in 750, by the last raid on Sicily from Ifrı3qiya for a half
a century in 752, and by the seizure of power in al-Andalus by the
refugee Umayyad ‘Abd al-Rah5ma2n in 756. For the next half century
the Empire would have virtually the only naval forces in the
Mediterranean. However, during the preceding half century the
struggle had turned the entire Mediterranean into something of a no-
man’s land.
Cyprus was recovered and garrisoned by the Muslims in 693 only
to be lost to the Byzantines the following year and recovered again in
695. The islanders clearly assisted the Empire continually in violation
of their covenant and the Muslims had to enforce their suzereinty
again in 713 and 725. In 743 al-Walı3d II deported many of them to
Syria. However his action was judged too severe by Muslim jurists
and Yazı3d III eventually allowed them to return. Crete was also raided
during the reign of al-Walı3d I, probably in 713.53 In 703 an Egyptian
fleet attacked Sicily at the request of the governor of Ifrı3qiya, Mu2sa2
ibn Nus5ayr. The island was attacked again in 704 by Mu2sa2’s son ‘Abd
Alla2h and from then on became subject to virtually incessant Muslim
raids for the next fifty years. Sardinia was attacked in 708 and 711 and
the Balearics were also raided in 708.54 In reverse, the Byzantines
attacked Egypt and captured the fleet commander Kha2lid ibn Kaysa2n
in 709. They attacked Latakia in 718 and in the following year
returned to Egypt, attacking Tinnis. There were further raids on Egypt
in 720, 725, and 736. Then in 739 a reported 360 Byzantine ships
attacked Damietta.55
On land, Al-T4abarı3’s accounts of the interminable Umayyad
campaigns against Byzantine frontiers and of Byzantine responses, a
------------------------------
52
Historia miscella, lib. XXII (col. 1095); Nike2phoros I, Historia syntomos, §68
(pp. 140-41); Theophane2s, Chronographia, A.M. 6238 (p. 424).
53
Agapios, Kita2b al-‘Unwa2n, tome 8, fasc. 3, p. 511; Al-Bala2dhurı3, Kita2b Futu2h 5
al-Bulda2n, part 2, ch. 13 (vol. 1, pp. 235-43); part 7, ch. 1 (vol. 1, p. 376); Al-T4abarı3,
Ta’rı3kh (Yar-Shater), A.H. 107, 125 (vol. 25, p. 25; vol. 26, pp. 119-20); Theophane2s,
Chronographia, A.M. 6234 (p. 417). On the status of Cyprus, see Jenkins, “Cyprus”;
Kyrris, “Cyprus”.
54
Gesta episcoporum Neapolitanorum, §36 (p. 422); Ibn Khaldu2n , ‘Ibar (Des
Vergers), pp. 33-4, 44; Ibn S4abba2t6, Dı3wa2n S4ilat al-simt6, pp. 348-50; Pseudo ibn
Qutaybah, Ah5a2dı3th al-Ima2ma, pp. lxvi-lxvii.
55
Al-Bala2dhurı3, Kita2b Futu 2h5 al-Bulda2n, part 2, ch. 9 (vol. 1, p. 204); Al-T4abarı3,
Ta’rı3kh (Yar-Shater), A.H. 90 (vol. 23, pp. 149-50). See also Brooks, “Relations
between the empire and Egypt”.
34 CHAPTER ONE
tale of raid and counter-raid, devastation and slaughter, capture and

enslavement, makes depressing reading.56
The third period, ca 750-875: equilibrium of chaos
In 750 the last Umayyad Caliph, Marwa2n II al-H4ima2r, was defeated at

the battle of the river Great Zab and eventually killed by the forces of
Abu2 ’l-‘Abba2s al-Saffa2h5, who had been proclaimed at al-Ku2fa in 749.
The ‘Abba2sids claimed descent from Muh5ammad’s uncle al-‘Abba2s
ibn ‘Abd al-Mut6t6alib and swept to power on a wave of support from
the disgruntled within the Muslim community: the Arab tribes of Iraq,
the Shı3‘a, and the mawa2lı3. The Shı3‘a had been persecuted, their
ima2ms, descendents of ‘Alı3’s second son H4usayn, forced underground
and hunted by Umayyad secret police. The mawa2lı3 had been treated
unequally by Arab Muslims and especially in Iraq and Persia had
become a resentful underclass.57 Following the disaster at the Great
Zab many Umayyads were hunted down and killed; however, one,
‘Abd al-Rah5ma2n ibn Mu‘a2wiya, later called al-Da2khil, “the Incomer”,
managed to flee via the Maghrib to al-Andalus, where he seized
power in 756.58 The establishment of an Umayyad amı3rate in al-
Andalus was the first rupture in the Muslim polity, although it was
only the beginning. In some respects it represented an inevitability
because the Caliphate was stretched so far geographically that the
contemporary technology of travel and communications was simply
incapable of holding it together. Even given the best of
circumstances, a message could take a year or more to reach
Morocco from Baghdad and return. Reflecting a shift in the centre of
gravity of the Muslim polity, the second ‘Abba2sid Caliph, al-Mans5u2r,
founded a new capital at Baghdad in 762.59 Henceforth the Caliphate
would focus towards the east and south rather than the Mediterranean.
The ‘Abba2sids quickly rose to great power, reaching their zenith
during the reign of Ha2ru2n al-Rashı3d.60 However, upon his death in 809
civil war broke out. Al-Ma’mu2n was victorious but although his reign
------------------------------
See al-T4abarı3, Ta’rı3kh (Yar-Shater), vols 23-5, passim under “Byzantines”.
56
Al-Mas‘u2dı3, Muru2j, vol. 6, pp. 35-76; Ibn ‘Idha2rı3, Al baya 2n al-mughrib, vol. 1,
57
pp. 67-70; Sa2w ı3rus, History of the Patriarchs, I.18 (vol. V.1, pp. 134-88).
58
Akhba2r Majmu 2‘a, pp. 55-109; Al-Marra2kushı3, Al-Mu‘jib, pp 13-15; Fath5 al-
Andalus, pp. 50-63; Ibn al-Athı3r, Al-Ka2mil (Fagnan), pp. 91-102; Ibn ‘Idha2rı3, Al-
baya2n al-mughrib, vol. 2, pp. 61-73.
59
Al-Ya‘qu2bı3, Al-Bulda2n, pp. 6-19.
60
See Kennedy, Early Abbasid Caliphate.
was a great one for Muslim cultural development, politically it was

one of disintegration, the great Persian province of Khura2sa2n being
delegated to the T4a2hirids from 821. His successor, al-Mu‘tas5ı3m, began
to surround himself with a personal corps of Turkish guards, whose
disorderly conduct forced him to evacuate Baghdad in 836 and build a
new capital at Samarra, which became the capital until 889. There the
guards imprisoned the Caliphs within their own city, making and
breaking Caliph after Caliph and not being brought under control
until the reign of al-Mu‘tamid. In the meantime Khura2sa2n remained
lost to the T4a2hirids, then to the S4affa2rids from 873, and to the
Sa2ma2nids from 900. Southern Iraq was thrown into turmoil by a
dangerous revolt of Negro slaves, the Zanj, from 869-83.
Disruption and weakness in the heartlands led to disintegration in
the West. In the Maghrib the first ‘Abba2sid governor sent to Ifrı3qiya in
761-2, Muh5ammad ibn al-Ash‘a2th ibn al-‘Uqba al-Khuza2‘ı3, fortified
al-Qayrawa2n and made the city the new capital of the province.
However, the ‘Abba2sids proved incapable of exercizing authority west
of Ifrı3qiya and the governors were fully occupied trying to maintain
authority over fractious Arab settlers and Berber tribes.61
Those Berber tribes who had been converted had been heavily
influenced by the Khawa2rij, who had survived in remote provinces as
populist groups opposed to central Caliphal authority, in particular in
the Maghrib the Iba2d5ı3yya, followers of the seventh-century Arab ‘Abd
Alla2h ibn Iba2d5. These Kha2rijı3 Iba2d5ı3 Berber tribes resisted the
authority of the Umayyad governors and their ‘Abba2sid successors.
From the third quarter of the eighth century the Maghrib
fragmented. ‘Abd Alla2h ibn Rustam, leader of the Iba2d5ı3 Zena2ta
Berbers, occupied al-Qayrawa2n temporarily and then fled to western
Algeria, where he founded an amı3rate centred on Ta2hart, becoming
ima2m of all the Iba2d5ı3yya in 777. Eventually the Rustamids would be
confronted by the Shı3‘a Idrı3sids to the west and the Sunnı3 Aghlabids
to the east and would forge an alliance with the Umayyads of al-
Andalus. Andalusi seamen from Almeria/Pechina established a colony
near ancient Cartenna on the coast north of Ta2hart in 875-6, bringing
the Rustamid state and Umayyad al-Andalus into economic relations.
The last Rustamid, Yaqz5a2n ibn Muh5ammad, would be overthrown by
the Keta2ma Berbers of the Fa2t6imid Abu2 ‘Abd Alla2h in 909.62
------------------------------
61
Al-Bakrı3, Kita2b al-mughrib, p. 57; Al-Nuwayrı3, Niha2yat ‘al-Arab (De Slane),
vol. 1, pp. 374-97; Ibn Khaldu2n, ‘Ibar (des Vergers), pp. 55-9.
62
Al-Bakrı3, Kita2b al-mughrib, pp. 128-9, 139-41; Al-Ya‘qu2bı3, Al-Bulda2n, pp. 216-
17, 224; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 1, pp. 283-7.
36 CHAPTER ONE
Table 3: Rulers of the third period, ca 750-875
The Byzantine
Empire The Muslims
(A) Aghlabids of
(U) Umayyads Tunisia
The ‘Abba2sid of al-Andalus (I) Idrı3sids of
Caliphs (C) Amı3rs of Crete Morocco
(S) Amı3rs of Sicily (R) Rustamids of
Algeria
Constantine V (740- Abu2 ’l-Abbas al- ‘Abd al-Rah5m a2n I

75) Saffa2h 5 (749-54) (U 756-88)
Al-Mans5u2r (754-
75)
Leo IV (775-80) Al-Mahdı3 (775-85) ‘Abd al-Rah5m a2n

ibn Rustam (R
777-84)
Constantine VI Al-Ha2dı3 (785-6) Hisha2m I ( U 788- ‘Abd al-Wahha2b

(780-97) Ha2ru2n al-Rashı3d 96) (R 784-823)
(786-809) Al-H4akam I (U Idrı3s I ( I 789-93)
796-822) Idrı3s II (I 793-828)
Staurakios (811)
Michael I (811-13) ‘Abd Alla2h I (A
812-17)
(Table 3 continued)
(B) Bulgaria The Carolingians

(V) Doges of The Iberian rulers The Lombards and their
Venice63 successors
(K) Kings of the

Franks
(K) Kings
(A) Asturias (E) Emperors
(B) Benevento
(Ar) Aragon (I) Italy
(C) Capua
(N) Navarre (S) Salerno (P) Provence
(B) Burgundy
(A) Aquitaine
Kormisos° (B 739- Pelayo (A ca Liutprand (K 712-

56) 718/22-37) 44)
Favila (A 737-9)
Alfonso I (A 739-
57)
Vinekh (B 756-ca Fruela (A 757-68) Hiltiprand (K 744) Pepin I (K 751-68)
761) Aurelio (A 768-74) Ratchis (K 744-9) Charlemagne (K
Telec (B ca, 761-4) Silo (A 774-83) Aistulf (K 749-56) 768-814, E
Sabin (B ca 764-7) Desiderius (K 757- 800-814)
Umar (B 767) 74) Louis I (A 781-
Toktu (B 767-ca Adelchis (K 759- 838)
769) 74)
Pagan (B ca 770) Gisulf II (B 742-
Telerig (B ca 770- 51)
77) Liutprand (B 751-
8)
Arichis II (B 758-
87)
Giovanni Galbaio
(V 775-804)
Kardam (B 777-ca
803)
Mauregato (A 783- Grimoald III (B Pepin (I 781-810)
8) 787-806) Bernard (I 812-17)
Vermudo I (A 788-
91)
Alfonso II (A 791-
842)
Krum (B ca 803-14) Aznar I Galíndez Grimoald IV (B

Obelerio degli (Ar ca 809-39) 806-17)
Antenori (V Iñigo Iñiguez
804-11) Arista (N ca
810-52)
Agnello Partecipa-
zio (V 811-27)
------------------------------
63
From 800.
38 CHAPTER ONE
(Table 3 continued)
The Byzantine
Empire The Muslims
(A) Aghlabids of
(U) Umayyads Tunisia
The ‘Abba2sid of al-Andalus (I) Idrı3sids of
Caliphs (C) Amı3rs of Crete Morocco
(S) Amı3rs of Sicily (R) Rustamids of
Algeria
Leo V (813-20) Al-Ma’mu2n (813- Ziya2dat Alla2h I (A

33) 817-38)
Michael II (820-29) ‘Abd al-Rah5m a2n II Abu2 Sa‘ı3d Aflah5 (R
(U 822-52) 823-72)
Abu2 H4afs5 ‘Umar Muh5ammad al-
ibn Shu‘ayb (C Muntas5ir (I
ca 824-55) 828-36)
Asad ibn al-Fura2t
(S 827-8)
Muh5ammad ibn
Abı3 ’l-Jawa2rı3 (S
828-9)
Theophilos (829-42) Al-Mu‘tas5im (833- Zuhayr ibn al- ‘Alı3 I (I 836-49)
42) Ghawth (S 829) Abu2 ‘Iqa2l al-
Al-Wa2thiq (842-7) As5bagh ibn Wakı3l Aghlab (A 838-
(S 829) 41)
‘Uthma2n ibn Muh5ammad I (A
Qurhub (S 829) 841-56)
Muh5ammad ibn
‘Abd Alla2h ibn
al-Aghlab (S
832-5)
Ibra2hı3m ibn ‘Abd
Alla2h (S 835-
51)
Michael III (842- Al-Mutawakkil Muh5ammad I (U Yah5ya2 I (I 849-?)
67) (847-61) 852-86) Yah5ya2 II (I ?)
Al-Muntas5ir (861- Al-‘Abba2s ibn al- ‘Alı3 I (I ?)
2) Fad5l (S 851-61) Yah5ya2 III (I ?)
Al-Musta‘ı3n (862- Shu‘ayb I ibn Ah5mad (A 856-63)
6) ‘Umar (C ca Ziya2dat Alla2h II (A
Al-Mu‘tazz (866-9) 855-?) 863)
Ah5mad ibn Ya‘qu2b Abu2 ’l-Ghara2nı3q
(S 861) Muh5ammad II
‘Abd Alla2h ibn al- (A 863-75)
‘Abba2s (S 861)
Khafa2ja ibn Sufya2n
(S 862-9)
(Table 3 continued)
(B) Bulgaria The Carolingians

(V) Doges of The Iberian rulers The Lombards and their
Venice64 successors
(K) Kings of the

Franks
(K) Kings
(A) Asturias (E) Emperors
(B) Benevento
(Ar) Aragon (I) Italy
(C) Capua
(N) Navarre (S) Salerno (P) Provence
(B) Burgundy
(A) Aquitaine
Omurtag (B 814-31) Sico (B 817-33) Louis I (E 813-40)

Pepin I (A 817-38)
Giustiniano
Partecipazio (V
827-9)
Giovanni Sicard (B 833-9) Lothar I (I 822-55,

Partecipazio (V Radelchis I (B 839- E 840-55)
829-36) 51) Pepin II (A 838-48)
Malamir (B 831-6) Sikenolf (B 839-
Presiam (B 836-52) 49, S 849-51)
Pietro Tradonico (V
836-64)
Boris I (B 852-89) Ramiro I (A 842- Radelgar (B 851-3) Louis II (I 840-75,

Orso Partecipazio I 50) Adelchis (B 853- E 855-75)
(V 864-81) Galindo I Aznár 78) Charles the Bald
(Ar ca 844-67) Peter (S 853-6) (A 848-66)
Ordoño I (A 850- Ademar (S 856-61) Lewis II (A. 866-
66) Guaifer (S 861-80) 79)
García Iñiguez (N
852-70)
Alfonso III (A 866-
910)
------------------------------
64
From 800.
40 CHAPTER ONE
As opposed to the Kha2rijı3 Rustamids, the Idrı3sids were Shı3‘a,

descended from a great grandson of ‘Alı3 ibn Abı3 T4a2lib’s son al-H4asan
who was involved in an ‘Alı3d uprising against the ‘Abba2sids in the
H4ija2z in 786 and who fled to the Maghrib and was recognized by
several chiefs of the Zena2ta Berbers, founding a new capital, Fez, near
the site of the old Roman town of Volubilis in 793. He and his
descendants were proclaimed as Caliphs. However, the Idrı3sid
Caliphate began to fragment during the reign of Muh5ammad al-
Muntas5ir as the towns were parcelled out among his many brothers. In
the tenth century Yah5ya2 IV was forced to recognize the suzereinty of
the Fa2t5imid ‘Ubayd Alla2h and in 921 Fez was occupied. Idrı3sid rule
survived in some outlying towns but their history is obscure. In 931
the Umayyads of al-Andalus initiated a forward defence policy in the
Maghrib against the Fa2t6imids by occupying Ceuta and the last Idrı3sids
were taken off to Cordoba in 974.65
From a Christian perspective, the most important Maghribin
splinter state was that of the Aghlabids, who were descended from
Ibra2hı3m ibn al-Aghlab, the son of a Khura2sa2nian Arab officer of the
‘Abba2sids who had moved to Egypt. He was appointed governor of
Ifrı3qiya in 800 by Ha2ru2n al-Rashı3d but in practice became independent
of Baghdad. The Aghlabids took to the sea with ferocity and
determination, the third, Ziya2dat Alla2h I, beginning the definitive
conquest of Sicily in 827. Aghlabid fleets harried south Italy, Corsica,
Sardinia, and even the Maritime Alps. Malta was occupied in 868 and
by 878 the conquest of Sicily was virtually complete. The island was
ruled by governors, at first Aghlabid and then later Fa2t6imid. The Shı3‘a
propaganda of the da2‘ı3 Abu2 ‘Abd Alla2h, precursor of the Fa2t6imid
Mahdı3, Ubayd Alla2h, stirred up the Keta2ma Berbers and a military
uprising drove the last Aghlabid, Ziya2d at Alla2h III, out into Egypt.66
There was also a small independent Sunnı3 dynasty centred on the
------------------------------
65
Al-Bakrı3, Kita2b al-mughrib, pp. 205, 231-59; Al-Ya‘qu2bı3, Al-Bulda2n, pp. 223-4;
Ibn abı3 Zar‘, Rawd5 al-Qirt6a2s, pp. 15-65, 103-30; Ibn ‘Idha2rı3, Al-baya2n al-mughrib,
vol. 1, pp. 96-9, 289-95, 303-11; Ibn Khaldu2n, Muqaddimah, Introduction (vol. 1, pp.
47-53); idem, ‘Ibar (De Slane), vol. 2, pp. 559-71; idem, ‘Ibar (Des Vergers), p. 89.
66
Al-Bakrı3, Kita2b al-mughrib, p. 99; Al-Bala2dhurı3, Kita2b Futu2h5 al-Bulda2n, part 4,
ch. 1 (vol. 1, pp. 369-70); Al-Ma2liki, Muh@tas5ar Kita2b Riya2d5, pp. 306-9; Al-Mas‘u2dı3,
Muru2j, vol. 1, p. 370 & vol. 8, p. 246; Al-Nuwayrı3, Niha2yat ‘al-Arab (Amari), pp.
113-24, 146-7; idem, Niha2yat ‘al-Arab (Caussin), pp. 404-16; idem, Niha2yat ‘al-Arab
(De Slane), vol. 1, pp. 397-447; Ibn al-Athı3r, Al-Ka2m il (Fagnan), pp. 157-60, 167-8,
173-6, 181-94, 207-8, 210-18, 225-30, 235-41, 247-62, 263-71; Ibn ‘Idha2rı3, Al-baya2n
al-mughrib, vol. 1, pp. 111-29 et passim ad p. 163; Ibn Khaldu2n, ‘Ibar (Des Vergers),
pp. 83-9. See also Pertusi, “Ordinamenti militari”, pp. 688-95; Talbi, L’émirat
Aghlabide.
town of Naku2r in Alhucemas Bay in Morocco whose foundation by

S5a2lih5 ibn Mans5u2r went back as far as 710. This small city state
remained independent of the Rustamids and Idrı3sids and was friendly
with the Umayyads of al-Andalus.67
Despite the breakup of the Caliphate, the period from the ‘Abba2sid
revolution to the accession of the Byzantine emperor Basil I in 867
witnessed the peak of Muslim eminence. Much later, Ibn Khaldu2n
penned the famous lines that: “... the Muslims gained control over the
whole Mediterranean. Their power and domination over it was vast.
The Christian nations could do nothing against the Muslim fleets,
anywhere in the Mediterranean. All the time the Muslims rode its
waves for conquest”.68 Even if this was in fact never more than
comparatively true at best, conquest of the islands and the presence of
fleets across the sea did neverthless give Muslim powers an
ascendancy. However, they did not have it all their own way and the
period was characterized more by inchoate thrust and counter-thrust.
Moreover, neither the Muslim nor Christian worlds were internally
untroubled. The various Muslim splinter states became at odds with
each other and with the Caliphate in Bagdhad. The Byzantines had to
reckon with the First Bulgarian Empire under its great Khans Krum
and Omurtag and the Italian peninsula was wracked by strife between
the Lombards and Franks.
By land, the ‘Abba2sids maintained the interminable raids and
counter-raids across the Tauros frontiers begun by the Umayyads.69
The eighth and ninth centuries were the age of the muja2hidu2n and the
akritai. Muslims built numerous riba2ta2t, from which muja2hidu2n
waged jiha2d. Their Byzantine counterparts were the border lords, the
akritai, of whom the most famous was the fictional Digene2s Akritas.
Encounters across the borders were reflected in the epic romance of
Dha2t al-Himmah and in many tales of later versions of the Thousand
and One Nights.
By sea the first major assault came in 790 when a Muslim fleet
sailing against Cyprus encountered that of the Kibyrrhaio2tai in the
Gulf of Antalya and the Byzantine strate2gos, Theophilos, was
defeated, captured, and later killed. In 806 Ha2ru2n al-Rashı3d
------------------------------
67
Al-Bakrı3, Kita2b al-mughrib, pp. 183-96; Al-Ya‘qu2bı3, Al-Bulda2n, pp. 222-3; Ibn
‘Idha2rı3, Al-baya2n al-mughrib, vol. 1, pp. 249-55; Ibn Khaldu2n, ‘Ibar (De Slane), vol.
2, pp. 137-43.
68
Ibn Khaldu2n, Muqqadimah, vol. 2, p. 41.
69
Al-T4abarı3, Ta’rı3kh (Yar-Shater), vols 29-30, passim under “Byzantines”;
Theophane2s, Chronographia, A.M. 6243-6305 (pp. 427-503), passim.
42 CHAPTER ONE
despatched a fleet to Cyprus which deported many Cypriots, once

again for suspicion of not remaining neutral. In the following year he
sent a fleet against Rhodes; however, it could not capture the fortress.
Crete was also assaulted some time during his Caliphate.70
In al-Andalus ‘Abd al-Rah5ma2n I was fully occupied consolidating
his own rule, though he also defended himself successfully against
Charlemagne’s expedition to Zaragoza in 778, which ended in the
famous disaster for the Franks in the Pyrenees pass of Roncesvalles.
His successors carried the attack to their Christian neighbours, not
only to the Kingdom of the Asturias, the Basque lands and the
incipient Christian states in Aragon and the Frankish March of
Barcelona, but also across the Pyrenees. In 793 Hisha2m I’s amı3r ‘Abd
al-Malik ibn Mughı3th destroyed Gerona and pushed on to Narbonne
and Frankish territory, defeating a force sent against him and returning
with great booty. In a two-year siege in 801-3 Louis of Aquitaine
captured Barcelona, thus establishing the Spanish March, but in 808
and 809 Umayyad columns threw back his troops from Tortosa and in
813 or 815 they defeated the Franks outside Barcelona. During the
reign of ‘Abd al-Rah5ma2n II, war broke out in the Spanish March in
827 between a rebel noble called Aizo and Count Bernard of
Barcelona. Aizo called in an army sent in 828 by ‘Abd al-Rah5ma2n
under the amı3r Abu2 Marwa2n ‘Ubayd Alla2h which engaged the
Frankish forces at Barcelona and devastated the lands north to Gerona.
Barcelona was retaken from the Franks in 852.71
At the turn of the eighth and ninth centuries, Al-H4akam I began to
turn his attention to the islands of the western Mediterranean,
attempting to extend his rule as far as Corsica and possibly Sardinia.
An Andalusi fleet raided the Balearics as early as 798 and it may have
been this same fleet which was defeated at Naples in the same year. In
806 the King of Italy, Pepin, sent a fleet from Italy to Corsica against
Andalusi Moors who had pillaged the island and in the following year,
since they had come to frequent the island, Charlemagne sent his
Count of the Stable, Burchard, to Corsica with a fleet which gained a
------------------------------
70
Al-Bala2dhurı3, Kita2b Futu2h5 al-Bulda2n, part VII, ch. 1 (vol. 1, p. 376); Al-T4abarı3,
Ta’rı3kh (Yar-Shater), A.H. 190 (vol. 30, p. 265); Theophane2s, Chronographia, A.M.
6282, 6298, 6300 (pp. 465, 482-3).
71
Annales Bertiniani, Annus 852 (p. 447); Annales regni Francorum, Annus 778,
801, 826-7 (pp. 50-53, 116, 170-73); Chronicle of Moissac, Annus 793 (p. 300); Ibn
al-Athı3r, Al-Ka2mil (Fagnan), pp. 102-7, 109-12, 118-22, 125-44, 150-55, 160-72, 174-
5, 177-80, 194-9, 200-201, 208-9, 211-12, 220-25; Ibn ‘Idha2rı3, Al-baya2n al-mughrib,
vol. 2, pp. 73-91, 101-2, 117, 119-20; Vita Hludowici, §§10, 13-16, 40-41 (pp. 611-
15, 629-30). The best account of the Umayyad amı3rate in al-Andalus, from our
perspective, is still that of Lévi-Provençal, L’Espagne musulmane, vols 1-2.
victory over them in some harbour, probably Bonifacio. In 810 they

again sent a fleet to Sardinia and Corsica and conquered the latter
almost entirely. However, in 813 Count Irmingar of Ampurias
intercepted them returning from Corsica off the Balearics and
captured 8 ships. In revenge they ravaged Civitavecchia and Nice and
attacked Sardinia but were repelled. Andalusi corsairs raided
Marseilles in 838, Arles in 842 and 850, and established a base in the
Camargue at the mouth of the Rhône some time before 869. In 849
‘Abd al-Rah5man II sent 300 mara2kib to the Balearics to reduce them
to the same covenantary status of ‘ahd viz-a-viz al-Andalus as Cyprus
had to the Caliphate, although they were not conquered and made a
ku2ra, dependent province, of Cordoba until 903.72
Development of major naval forces in al-Andalus was also
stimulated by Norse attacks. In 844 a Norse fleet of 80 mara2kib and
other smaller boats sailed up the estuary of the Tagus river and
assaulted Lisbon. Beaten off, they sailed south and sacked Seville.
Mauled by Muslim cavalry, they then re-embarked, attacked As5ı3la in
Morocco, and retired to winter in Aquitaine. They returned again in
859-60, when 62 mara2kib anchored off the mouth of the Guadalquivir
river. Deterred by Muslim forces, they went on to sack Algeciras and
part of the fleet assaulted Naku2r. The remainder ravaged the Balearics,
a detachment raided up the valley of the Ebro as far as Pamplona, and
they then went on to sack Arles, Nîmes, and Valence in Provence and
Luni in Liguria before retiring.73
In Italy the old order was changing. The establishment of the
Lombard kingdom and confining of imperial influence to the
exarchate of Ravenna left the Papacy in Rome isolated from
Constantinople. The last emperor to visit Rome was Constans II in
663 and the last Pope to visit Constantinople was Constantine in 711.
In an imperial vacuum the Papacy turned to the Frankish kingdom
under its Mayor of the Palace, Charles Martel, and then his son Pepin
III. In 751, in response to a set-up question from Pepin as to who
should bear the title of “king”, he who held real power or the titular
------------------------------
72
Annales Bertiniani, Anni 838, 842, 849, 850, 869 (pp. 432, 439, 444-5, 485);
Annales regni Francorum, Anni 798, 806, 807, 810, 812, 813 (pp. 104, 122, 124, 130,
137, 139); Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 2, p. 145; Neapolitanorum victoria
facta. The final conquest of Majorca is reported only by Ibn Khaldu2n, ‘Ibar (Bu2laq),
vol. 4, p. 164. See also Manfroni, Marina italiana. I, pp. 35-6, 40-41.
73
Al-Bakrı3, Kita 2b al-mughrib, pp. 184-5, 219-20; Annales Bertiniani, Anni 844,
859-60 (pp. 441, 453-4); Chronicle of Alfonso III, pp. 54-5, 64-5, 102, 125; Ibn
‘Idha2rı3, Al-baya2n al-mughrib, vol. 2, pp. 141-4, 157-8; Ibn Khaldu2n, ‘Ibar (De Slane),
vol. 2, p. 139.
44 CHAPTER ONE
Merovingian incumbent, Pope Zachary gave him the answer he

wanted and Pepin was crowned King. A fateful alliance was forged
between the Papacy and the Frankish Kingdom which eventually led
to the coronation of Pepin’s son Charlemagne as Emperor in Rome on
Christmas day 800 and to the creation of the medieval Western
Roman Empire.74
The immediate Papal concern was the Lombards. In 752 Aistulf
demanded tribute from Rome and control of the fortresses of Papal
territories. Appealed to by Stephen II, Pepin invaded Italy and
defeated Aistulf, who had to come to terms. But he broke the peace in
756 and Pepin invaded again. When Desiderius attacked Papal
territories in 772-3, Hadrian I appealed to Charlemagne, who invaded,
defeated the Lombard army, besieged and took Pavia, and had himself
crowned king in 774. The Lombard kingdom was succeeded by a
Frankish one which eventually became the Franco-Lombard Kingdom
of Italy when the Frankish Empire began to fragment.75 In response,
empress Eire2ne2, regent for her son Constantine VI and the power
behind the throne in Byzantium, sent an embassy to Charlemagne in
781 with a view to betrothing his daughter Rotrud to Constantine. An
agreement was reached; however, in 788 she broke the agreement and
despatched the refugee Lombard pretender, Desiderius, together with
a logothete2s named John, to Italy to counter Charlemagne but the
expedition was defeated.76
The Carolingian conquest left Italy with a Frankish kingdom in the
North and Papal territories in the centre. The South, under the
Lombard duke of Benevento, Arichis, became subject to the
Carolingians, technically at least. In 787 Arichis died and in return for
release of his heir, Grimoald, who was a hostage at Charlemagne’s
court, his widow Adalperga accepted Carolingian suzereinty. But it
remained ephemeral. By the early ninth century the Lombards, now
with a second capital established by Arichis at Salerno, were dominant
in the South.77 Their principalities would continue to dominate it until
------------------------------
74
Annales regni Francorum, Anni 749-50, 801 (pp. 8-11, 112-13). On early
medieval Italy and the Empire, see Gay, L’Italie méridionale.
75
Annales regni Francorum, Anni 750-56, 773-4 (pp. 8-15, 34-41); Chronicon
Salernitanum, §§2-9 (pp. 4-11); Fredegar, Continuations, §§36-9 (pp. 104-9); Leo
Marsicanus, Chronica, I.8, 12 (pp. 585-6, 589).
76
Theophane2s, Chronographia, A.M. 6274, 6281 (pp. 456, 463-4).
77
Annales regni Francorum, Anni 787, 812 (pp. 72-5, 137); Chronicon
Salernitanum, §§9-14 (pp. 11-20); Erchempert, Historia Langobardorum, §§2-6 (pp.
235-7); Leo Marsicanus, Chronica, I.12 (p. 589). On the Franks, Lombards,
Byzantines, and Muslims in Italy in the ninth and tenth centuries, see Kreutz, Before
the Normans; Loud, Robert Guiscard.
their conquest by the Normans in the eleventh century.

In the North, internal strife amongst the communities of the
Venetian lagoons eventually led Doge Obelerio degli Antenori to
make Venice’s submission to Charlemagne in 805. After the defeat of
a Byzantine expedition sent to bring Venice back to her allegiance in
809-10, Pepin I, king of Italy, attempted in 810 to incorporate her into
his own domains but was forced to withdraw by the Venetian defence
of the lagoons, having won only payment of an annual tribute. Venice
began the rise to power that would make her master of the Adriatic.78
Even after the ‘Abba2sid overthrow of the Umayyads, the
incumbent governor of Ifrı3qiya, ‘Abd al-Rah5ma2n ibn H4abı3b, sent a last
raid against Sicily and Sardinia in 752.79 After that, chaotic politics in
the Maghrib and the weakness of the ‘Abba2sid governors gave the
central Mediterranean islands a respite for half a century. But after
Ibra2hı3m I ibn al-Aghlab’s seizure of power in Ifrı3qiya, Aghlabid
squadrons began to raid across the central Mediterranean. In 805 they
raided the Peloponne2sos, and in 812 and 813 Corsica and Sardinia,
Lampedusa, Ponza, and Ischia. In 820 corsairs captured eight
merchant ships returning to Italy from Sardinia. In the following year
squadrons raided Sardinia, but were thrown back. Ziya2dat Alla2h I
began the conquest of Sicily in 827. Carolingian forces attempted to
return these compliments in kind and in 828 ships from Pisa and Luni
raided Bona in Algeria and Count Boniface, governor of Corsica,
together with his brother and some Tuscan counts, sailed to Ifrı3qiya
and raided between ancient Utica and Carthage.80
The first half of the ninth century was disastrous for the Byzantine
Empire. Byzantine-Bulgarian relations had been reasonably amicable
during the reigns of the Khans Telerig and Kardam, apart from
occasional skirmishes in which the Empire had mostly prevailed. But
in 807 hostilities flared with a new Khan of a very different ilk, Krum.
On 26 July 811 he trapped a Byzantine army in a mountain defile,
defeated and killed the Emperor Nike2phoros I and reportedly had his
skull made into a drinking cup. On 22 June 813 he routed another
------------------------------
78
Annales regni Francorum, Anni 806, 810 (pp. 120-21, 130); Constantine VII, De
administrando imperio, §28 (pp. 118-120); John the Deacon, Cronaca Veneziana, pp.
104-5. See also Manfroni, Marina italiana. I, pp. 37-9.
79
Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 1, p. 70; Nike2phoros I, Historia syntomos,
§78 (pp. 150-51).
80
Annales regni Francorum, Anni 812, 820, 828 (pp. 137, 153, 176); Constantine
VII, De administrando imperio, §49 (pp. 228-9); Al-Nuwayrı3, Niha2yat ‘al-Arab (De
Slane), vol. 1, p. 412; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 1, p. 120; Leo III,
Epistolae, 6-7 (pp. 96-9); Vita Hludowici, §42 (p. 632).
46 CHAPTER ONE
army under Michael I at the battle of Versinikia and advanced on

Constantinople. The capital was spared only by his premature death
and fortunately his son Omurtag concluded a thirty-year peace treaty
in 816.81
No sooner was this threat neutralized than the Empire was rocked
by the revolt in 820 of Thomas the Slav. This revolt, whose origins
and purposes are obscure, succeeded in winning over most of the
themata of Asia Minor, including the front-line maritime thema of the
Kibyrrhaio2tai. Thomas’s appeal has been attributed variously to anti-
Greek discontent amongst ethnic communities, reaction against
iconoclasm, and inchoate social discontent. Thomas was recognized
as emperor by the Caliph al-Ma’mu2n but, although able to besiege
Constantinople from late 821 to spring 823, his fleets were eventually
scattered by imperial squadrons using Greek Fire and his armies by
the Bulgarian Khan Omurtag, who came to the assistance of Michael
II in fulfillment of his treaty.82
Weakening of naval defences in the approaches to the Aegean by
the Kibyrrhaio2tai defection to Thomas may well have been what made
it possible for Andalusi corsairs under the leadership of Abu2 H4afs5
‘Umar ibn ‘Iflsa2 to land in Crete some time between 824 and 827.
Opposition was weak and they succeeded in consolidating their hold.83
Three expeditions sent out before the death of Michael II on 2 October
829 to retake the island were all unsuccessful: the first under
Pho2teinos, the strate2gos of the thema of Anatolikon; a second under
Krateros, the strate2gos of the Kibyrrhaio2tai; and a third under
O›oryphas, who was probably the droungarios tou ploimou. The last
did not even reach Crete. The Byzantines returned again in 843 when
the magistros and logothete2s tou dromou, Theoktistos, achieved brief
------------------------------
81
Genesios, Basileiai, A.1-2, 10 (pp. 3-4, 9); John Zo2naras, Epitome2 historio2n ,
XV.8, 12, 15, 17, 19 (vol. 3, pp. 353-5, 364, 372-4, 376-7, 380-81); Theophane2s,
Chronographia, A.M. 6265-7, 6283-4, 6288, 6303, 6304-5 (pp. 446-8, 467-8, 470,
489-92, 495-502). See also Runciman, First Bulgarian Empire; Fine, Early medieval
Balkans.
82
Genesios, Basileiai, B.2-8 (pp. 23-31); John Skylitze2s, Synopsis historio2n ,
Micah;l oJ Traulov".10-12 (pp. 36-9); John Zo2naras, Epitome2 historio2n, XV.22-3 (vol.
3, pp. 392-7); Theophane2s continuatus, II.9-20 (vol. 1, pp. 49-71).
83
Al-Bala2dhurı3, Kita2b Futu2h5 al-Bulda2n, part VII, ch. 1 (p. 376); Al-T4abarı3,
Ta’rı3kh (Yar-Shater), A.H. 210 (vol. 32, pp. 164-5); Constantine VII, De
administrando imperio, §22 (pp. 94-6); Genesios, Basileiai, B.10 (pp. 32-3); Ibn
Khaldu2n, ‘Ibar (De Slane), vol. 2, p. 544; John Skylitze2s, Synopsis historio2n, Micah;l
oJ Traulov".16 (pp. 42-3); John Zo2naras, Epitome2 historio2n, XV.24 (vol. 3, pp. 397-8),
XVI.7 (vol. 4, pp. 20-21); Sa2wı3ris, History of the Patriarchs (Evetts), I.19 (vol. X.5,
pp. 429-32); Theophane2s continuatus, II.21, 23 (pp. 73-8). See also Christides,
Conquest of Crete; Tsougarakis, Byzantine Crete.
success before returning to Constantinople, leaving his forces behind

on Crete to be slaughtered. Another attempt by the Caesar Bardas in
866 was prematurely terminated when he was murdered in the
presence of Emperor Michael III in camp at Ke2poi at the mouth of the
Maeander river by a rival, Basil the Macedonian, the future emperor
Basil I.84 After that there would be no further Byzantine expeditions
against Crete until the reign of Leo VI.
The loss of Crete altered fundamentally the strategic makeup of the
eastern Mediterranean. From a new fortress port at Chandax on the
north coast of the island, the Andalusi raided the Aegean for slaves
and booty, exercized some control over the southern Aegean, and
occupied some islands periodically: Aigina, Kos, Kythe2ra, and
Karpathos, for example. Some others were forced to pay tribute: for
example, Naxos. They almost certainly exercized influence over
Rhodes and Cyprus also, although they never attempted to occupy
them. Around 839 they inflicted a major defeat on a Byzantine fleet
off Thasos and around 860 they raided the Cyclades and the mainland,
penetrating through the Dardanelles as far as Proikonne2sos.85
Probably in response to their depredations, in the second half of the
century, alongside the Kibyrrhaio2tai, the northern Aegean islands
were erected into the maritime thema of Aigaion Pelagos and the
southern ones into that of Samos. How effective these measures were
is debatable, although the Byzantines did have some success. Around
840-42 Constantine Kontomyte2s, the strate2gos of Thrake2sio2n
destroyed a Cretan force ravaging the mainland. A large Muslim fleet
sailing on Constantinople in 842 was destroyed by storm off Cape
Chelidonia. Then in 852-3, having realized that ‘Abba2sid Egypt was
the power behind Crete, a Byzantine fleet, reportedly 100 mara2kib of
the shalandiyya2t type strong, attacked Damietta, sacking it, seizing
weapons destined for Crete, and destroying naval supplies. They
returned six years later to sack al-Farama2’; however, the Egyptians
replied in kind with raids on Byzantine coasts.86
------------------------------
84
Genesios, Basileiai, B.12-13, D.20-23 (pp. 34-5, 73-6); John Skylitze2s, Synopsis
historio2n, Micah;l oJ Traulov".16, 18 (pp. 43, 45), Micah;l oJ uiJo;" tou' Qeovfilou.22 (pp.
110-12), Basivleio" oJ Makedw'n.12 (p. 128); Theophane2s continuatus, II.22, 25-6,
IV.39, 41, V.17 (pp. 76-7, 79-81, 203, 204-6, 235-8). See also Makrypoulias,
“Byzantine expeditions”; Pertusi, “Ordinamenti militari”, pp. 695-700.
85
John Skylitze2s, Synopsis historio2n, Micah;l oJ uiJov" tou' Qeovf ilou.18 (p. 107);
John Zo2naras, Epitome2 historio 2n, XVIII.5 (vol. 4, pp. 15-16); Theophane2s
continuatus, III.39, IV.34 (pp. 137, 196). See also Christides, “Raids of the
Moslems”.
86
Al-T4abarı3, Ta’rı3kh (Yar-Shater), A.H. 238 (vol. 34, pp. 124-7); Brooks,
“Relations between the empire and Egypt”, p. 391 [referring to al-Kindı3, Governors
48 CHAPTER ONE
As destructive to the Byzantine position in the East as was the loss

of Crete, it was less important than that of Sicily. Following their
earlier raids from Sardinia to the Peloponne2sos, the Aghlabid conquest
of Sicily began in earnest in 827 when a Byzantine naval commander
in Sicily, the tourmarche2s Euphe2mios, revolted and offered Ziya2dat
Alla2h I suzereinty in return for recognition of himself as governor. An
expedition under Asad ibn al-Fura2t and Euphe2mios sailed for Sicily in
827. Landing at Mazara, it encountered stiff resistance and an assault
on Syracuse failed. A Veneto-Byzantine relief expedition sent by
Michael II may have prevented the fall of the city. However, Palermo
fell in 831, by which time the Muslims controlled most of the west of
the island. By 843 they had captured Messina and controlled its
crucial straits.87 But from then on their progress was slow. According
to Ibn al-Athı3r, a Byzantine fleet of 300 chelandia sent to relieve the
island after the fall of Enna in 858 failed; however, Syracuse would
not fall until 878 and the last strongholds not until 907.88
From Messina southern Italy lay exposed to Muslim incursions,
which had begun even before the fall of Messina. Brindisi and Taranto
were seized in 838 and 839 and a Venetian fleet of 60 bellicosae
naves sent to relieve Taranto at imperial request was defeated in 840.
In 841 Bari was captured by H4abla, a freedman of Abu2 ‘Iqa2l al-
Aghlab. From there Muslim forces raided north, sacking Ancona and
inflicting a major defeat on the Venetians in the Gulf of Kvarner in
842. Turning their attentions to Calabria and the west coast, they went
on to attack Rome itself and to pillage St Peter’s in 846. Although
------------------------------
and judges, p. 203]; George Hamartolos, Chronikon syntomon, IV.cclxviii.4 (col.
1033; Sa2wı3ris, History of the Patriarchs (Burmester), vol. 2, pt 1, pp. 13-14;
Theophane2s continuatus, III.39 (p. 137); Vita Theodorae, p. 11. See also Kubiak,
“Byzantine attack”; Levi della Vida, “Damietta raid”.
87
Chronicle of Cambridge, pp. 24-5, 50-53; Chronicon Salernitanum, §60 (p. 59-
60); Erchempert, Historia Langobardorum, §11 (p. 239); Al-Nuwayrı3, Niha2yat ‘al-
Arab (Amari), pp. 113-19; Ibn al-Athı3r, Al-Ka 2mil (Amari), p. 367; idem, Al-Ka2mil
(Fagnan), pp. 187-94, 216; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 1, pp. 128-9; Ibn
Khaldu2n, ‘Ibar (Des Vergers), pp. 103-12; John of Naples, Gesta, §54 (pp. 429-30);
John Skylitze2s, Synopsis historio2n , Micah;l oJ Traulov".20 (pp. 46-7); John the Deacon,
Cronaca Veneziana, p. 109; Leo Marsicanus, Chronica, I.21 (p. 596); Theophane2s
continuatus, II.27 (pp. 81-3); Vita di Sant’Elia, p. 7. The capture of Messina is
reported only by the later historians Ibn al-Athı3r and Ibn Khaldu2n. See Ibn al-Athı3r,
Al-Ka2mil (Fagnan), p. 216; Ibn Khaldu2n, ‘Ibar (Des Vergers), p. 118.
88
John Zo2naras, Epitome2 historio2n, XV.24 (vol. 3, pp. 399-400); Chronicle of
Cambridge, pp. 28-9; Al-Nuwayrı3, Niha2yat ‘al-Arab (Amari), pp. 119-24, 146-52;
idem, Niha2yat ‘al-Arab (Caussin), pp. 404-16; Ibn al-Athı3r, Al-Ka2mil (Fagnan), pp.
216-18, 225-9, 237-40, 253-4, 263-6; Ibn Khaldu2n, ‘Ibar (Des Vergers), pp. 116-23.
There is no corroborating evidence for the despatch of this fleet and from where Ibn
al-Athı3r derived this information about Sicily in the ninth century is unknown.
driven out by forces of Lothar I and eventually dispersed by the

Neapolitan fleet at Gaeta, during these years various Muslim bands
established themselves in strongholds all around the coasts. Bari
became the capital of an amı3rate which survived for thirty years.89
Southern Italy became polarized between a Muslim-controlled
swathe running from Bari to Taranto and the Lombard principalities
of Benevento and Salerno to the north and west. In response to
Muslim raids Lothar sent his son Louis II against Apulia in 848-9 and
he supposedly had some success, although a second expedition and
siege of Bari in 852 failed.90 Subsequent Muslim raids as far as
Benevento and into Campania induced Louis to intervene again in
866. In the following year an Aghlabid assault on Dubrovnik induced
Basil the Macedonian, become emperor Basil I in 867, to send the
droungarios tou ploimou, Nike2t as O›oryphas, with a fleet to relieve the
city. A new Venetian Doge, Urso Partecipazio I, sent a Venetian fleet
which defeated the Muslims off Taranto. An alliance was then
concluded between Louis II and the Empire and Nike2tas O›oryphas
sailed to Italy again; however, an allied siege of Bari in 869 failed and
not until 871 did Louis’s forces, now assisted by a Croatian fleet,
finally take the city. The third and last amı3r of Bari was taken off to
Benevento. An attempt by the Muslims of Taranto to reverse the
setback was driven back. Subsequently Louis’s behaviour alienated
the Beneventans, who imprisoned him and then sent him back north
under promise never to return. When he died in 875, the Byzantines
occupied Bari, which then became the capital of Byzantine Italy for
the next 200 years.91
------------------------------
89
Al-Bala2dhurı3, Kita2b Futu2h5 al-Bulda2n, part 6, ch. 1 (pp. 371-2); Annales
Bertiniani, Annus 846 (p. 442); Chronicon Casinense, §225 (p. 225); Chronicon
Salernitanum, §§72, 81 (pp. 70-1, 79); Erchempert, Historia Langobardorum, §§20,
29 (pp. 242, 245); John of Naples, Gesta, §60 (pp. 432-3); John the Deacon, Cronaca
Veneziana, pp. 113-15; Leo Marsicanus, Chronica, I.25, 27 (pp. 598-600); Liber
Pontificalis, 104 (Sergius II), §44, (vol. 2, p. 99). See also Manfroni, Marina italiana.
I, pp. 45-7, 49-53; Musca, Emı3rato di Bari.
90
Chronicon Salernitanum, §93 (pp. 93-4); Erchempert, Historia Langobardorum,
§§20, 29 (pp. 242, 245); John of Naples, Gesta, §61 (p. 433); Leo Marsicanus,
Chronica, I.29-30, 35 (pp. 601, 604).
91
Andrew of Bergamo, §§14-16 (pp. 227-9); Annales Bertiniani, Anni 869, 871
(pp. 485, 492); Chronicon Casinense, §7 (pp. 224-5); Chronicon Salernitanum, §§
103-9, 120 (pp. 104-22, 134); Constantine VII, De administrando imperio, §29 (pp.
126-34); Cronaca Capuana, pp. 123-4; Erchempert, Historia Langobardorum, §§29,
33-4, 38 (pp. 245, 247, 249); John of Naples, Gesta, §§64-5 (pp. 434-5); John
Skylitze2s, Synopsis historio2n, Basivleio" oJ Makedw'n.26 (pp. 146-7); John the Deacon,
Cronaca Veneziana, pp. 119-21; Leo Marsicanus, Chronica, I.35, I.36, (pp. 604, 605-
6); Theophane2s continuatus, V.53, 55-7 (pp. 289-90, 292-6). See also Manfroni,
Marina italiana. I, pp. 53-6.
50 CHAPTER ONE
The fourth period, ca 875-1025: Byzantine ascendancy
Basil I’s seizure of the throne in 867 through his murder of Michael
III marked the beginning of a period in which the balance of power in
the Mediterranean turned against the Muslims, even if it might not
have seemed so at the time. Muslim incursions continued in the east
against the Empire, in the central Mediterranean against remaining
Byzantine possessions in Sicily and also into the Lombard territories
on the west coast of Italy, and in the west into Provence. Nevertheless,
in retrospect it is clear that the Muslim offensive was running out of
vigour. The following century saw most Muslim expansion negated
and the Mediterranean frontiers pushed back to the south everywhere,
except in al-Andalus.
Political fragmentation of the Muslim polity continued apace. In
Egypt a Turkish soldier of fortune, Ah5mad ibn T4u2lu2n, who had been
sent to Egypt as deputy to the ‘Abba2sid governor, acquired the
governorship in 868 and extended his authority into Palestine, Syria,
and the Hija2z. Theoretically subordinate to the ‘Abba2s ids, in practice
the T4u2lu2nids were independent. A powerful fleet, the first Muslim
fleet about which more than skeletal details are known, projected
T4u2lu2nid influence throughout the Levant.92 Muslim Cilicia came
under their control from 878 to 882 and again from 892 to 897.
Although their inability to keep under control Qarmat6ı3 Shı3‘a sectarians
in Syria provoked the Caliph al-Muktafı3 to send to Egypt an army
which ended T4u2lu2nid rule in 905, re-establishment of direct ‘Abba2sid
authority was short-lived. Another Turkish soldier of fortune,
Muh5ammad ibn T4ughj al-Ikhshı3d, sent to Egypt as governor in 935,
defended his independence against the ‘Abba2sids and the H4amda2nids
in Syria, holding on to Damascus even though acknowledging
nominal ‘Abba2sid suzereinty. However, his two sons were mere
puppets and real power passed to a Nubian slave named Abu2 ’l-Misk
Ka2fu2r, who he appointed regent before he died and who became the
actual ruler on the death of the second son in 966. Only after his own
death in 968 was a weak grandson of Muh5ammad ibn T4ughj installed
as ruler, only to fall before the Fa2t6i mids the next year.
The latter were descended from ‘Ubayd Alla2h al-Mahdı3, a Shı3‘a
ima2m claiming descent from the seventh ima2m, Isma2‘ı3l, and hence
ultimately from ‘Alı3 ibn Abı3 T4a2lib and his wife: Muh5ammad’s
------------------------------
92
Sa2wı3ris, History of the Patriarchs (Burmester), vol. 2, pt 2, p. 110. See also
Fahmy, Muslim naval organization, pp. 42-5, 49-50.
daughter Fa2t6ima. Groundwork among the Keta2ma Berbers by a da2‘ı3, a

missionary propagandist, Abu22 ‘Abd Alla2h al-Shı3‘ı3, prepared his
move to Ifrı3qiya and he then overthrew the Aghlabids and Rustamids
and made the Idrı3sids tributary. A new capital and naval base, from
which Sicily was conquered from its Aghlabid governors and
operations were launched against the Byzantines, was built at al-
Mahdiyya. Early attempts to conquer Egypt from the ‘Abba2sids failed.
However the campaign of 919 did see one of the very few naval
engagements between two Muslim forces when the Caliph al-
Muqtadir sent a fleet from Tarsos to engage the Fa2t6imid fleet and won
a victory off Rosetta. In 969 the Fa2t6imid general Jawhar al-S4aqlabı3
conquered Egypt and paved the way for the fourth Fa2t6imid Caliph, al-
Mu‘izz, to move there in 973. He constructed a new capital al-Qa2hira,
“the Victorious”, Cairo, and from Egypt the Fa2t6imids extended their
authority over Palestine, Syria, and the Hija2z.93 As governor in Ifrı3qiya
he left the S4anha2ja Berber chief Yu2suf Buluggı3n ibn Zı3rı3, who overran
the Maghrib as far as Ceuta. However, his possessions proved too
unwieldy to hold together and under his grandson, Na2s5ir al-Dawla
Ba2dı3s, they were divided between the main branch of the family with
its capital at al-Qayrawa2n and the H4amma2did branch which ruled in
Algeria with a capital at Qal‘at Banı3 H4amma2d , near Ması3la. In theory
the Zı3rids remained subject to the Fa2t6imids in Egypt but in 1041 the
fourth Zı3rı3d, Sharaf al-Dawla al-Mu‘izz, rebelled and proclaimed the
‘Abba2sid Caliph. In response the Fa2t6i mid al-Mustans5ir unleashed
against the Maghrib the Bedouin tribes of the Banu2-Hila2l and Banu2-
Sulaym, who swept through the interior. The Banu2-Hila2l inflicted a
crushing defeat on the Zı3rı3ds at H4aydara2n in 1052, forcing them to
relocate to the coast at al-Mahdiyya. The H4amma2dids founded Bija2ya
on the coast as a refuge in 1067-68 and moved there in 1090-91.94
While the Muslim polity continued to fragment, the Byzantines had
problems of their own. The threat from Bulgaria had been neutralized
temporarily, with the incumbent Khan Boris I, who converted to
Christianity in 864, being quite amicable, although the menace would
------------------------------
93
Al-Bakrı3, Kita2b al-mughrib, pp. 65-8; Al-Maqrı3zı3, Al-Muqaffa2, pp. 76-8; Al-
Nuwayrı3, Niha2yat ‘al-Arab (Caussin), pp. 417-20; Al-Tija2nı3, Rih5la, ser. 5, 1.1,
pp.141-2, 357-63; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 1, pp. 163-237, 321-2, 332;
Ibn al-Athı3r, Al-Ka2m il (Fagnan), pp. 272-309, 313-20, 366-7, 370-74; Ibn-Khaldu2n,
Muqaddimah, Introduction (vol. 1, pp. 41-4); idem, ‘Ibar (De Slane), vol. 2, pp. 506-
51; idem, ‘Ibar (Des Vergers), pp. 144-56.
94
Al-Tija2nı3, Rih5la, ser. 4, 20, pp. 85-96, ser. 5, 1.1, pp. 369-73; Ibn al-Athı3r, Al-
Ka2mil (Fagnan), pp. 369-70, 374-9, 391-2, 394-406, 412-20, 448-50, 452-5, 456-60,
468-80; Ibn Khaldu2n, ‘Ibar (De Slane), vol. 2, pp. 9-26, 43-59.
52 CHAPTER ONE
Table 4: Rulers of the fourth period, ca 875-1025
Byzantine
The Muslims
Empire
(A) Aghlabids of
Tunisia
(F) Fa2t6imids
(I) Idrı3sids of
Morocco
(U) The Spanish
The ‘Abba2sid (Ik) Ikhshı3dids of
Umayyads
Caliphs Egypt
(C) Amı3rs of Crete
(B) Bu2y id amı3rs (R) Rustamids of
(S) Amı3rs of Sicily
Algeria
(T) T4u2lu2nids of
Egypt
(Z) Zı3rids of
Algeria
Basil I (867-86) Al-Muhtadı3 (869- Muh5ammad ibn Ah5mad ibn T4u2lu2n

70) Khafa2ja (S 869- (T 868-84)
Al-Mu‘tamid (870- 71) Abu2 Bakr ibn Aflah5
92) Ah5mad ibn Ya‘qu2b (R 872-?)
(S 871) Abu2 ’l-Yaqz5a2n
Al-H4u sayn (S 871) Muh5ammad (R
Abu2 ’l-‘Abba2s (S ?)
871-2) Ibra2hı3m II (A 875-
‘Abd Alla2h ibn 902)
Muh5ammad (S Khuma2rawayh ibn
872-4) Ah5mad (T 884-
Ah5mad ibn Ya‘qu2b 96)
(S 874-8)
Al-Aghlab ibn
Muh5ammad (S
878)
Al-H4u sayn ibn
Raba2h5 (S 878-
81)
‘Umar II ibn
Shu‘ayb (C ca
880-95)
Al-H4asan ibn al-
‘Abba2s (S 881-
2)
Muh5ammad ibn al-
Fad5l (S 882-5,
892)
Al-H4u sayn ibn
Ah5mad (S 885)
(Table 4 continued)
Bulgaria, Croatia, The Carolingians

Serbia, Venice, The Iberian rulers The Lombards and their
Rho2s of Kiev successors
(K) Kings of the

(B) Bulgaria (A) Aragon Franks
(C) Croatia (B) Barcelona (B) Benevento (E) Emperors
(K) Kiev (L) Asturias/León- (C) Capua (I) Italy
(S) Serbia Castile (S) Salerno (P) Provence
(V) Venice (N) Navarre (B) Burgundy
(A) Aquitaine
Ordoño I (L 850-
66)
Alfonso III (L 866-
910)
Branimir (C 879- Fortún Garcés (N Gaideris (B 878- Charles the Bald (E
92) 870-905) 81) 875-7)
Vlastimir (S mid Aznar II Galindo Radelchis II (B Charles the Fat (E
9th century) (A 867-93) 881-4) 881-7)
Giovanni Wifred I (B 878- Ayo II (B 884-91) Boso (B 878-87)
Partecipazio II 98) Guaimar I (S 880- Carloman (B & A
(V 881-7) 900) 879-84)
54 CHAPTER ONE
(Table 4 continued)
Byzantine
The Muslims
Empire
(A) Aghlabids
of Tunisia
(F) Fa2t6imids
(I) Idrı3sids of
(U) The Spanish
Umayyads Morocco
The ‘Abba2sid (Ik) Ikhshı3dids
(C) Amı3rs of
Caliphs of Egypt
Crete
(B) Bu2y id amı3rs (R) Rustamids
(S) Amı3rs of
of Algeria
Sicily
(T) T4u2lu2nids of
Egypt
(Z) Zı3rids of
Algeria
Sawa2da ibn
Muh5ammad (S
885-7)
Leo VI (886-912) Al-Mu‘tad5id (892- Al-Mundhir (U Abu2 H4a2tim Yu2suf
902) 886-8) (R 894-7)
Al-Muktafı3 (902-8) ‘Abd Alla2h (U 888- Jaysh (T 896)
Al-Muqtadir (908- 912) Ha2ru2n (T 896-905)
32) Abu2 ’l ‘Abba2s ibn Ya‘qu2b ibn Aflah5
‘Alı3 (S 887-90) (R 897-901)
Sawa2da (S 890-92) Abu2 H4a2tim Yu2suf
Ah5mad ibn ‘Umar (R 901-7)
(S 892-900) ‘Abd Alla2h II (A
Muh5ammad ibn 902-3)
Shu‘ayb (C ca Ziya2dat Alla2h III
895-910) (A 903-9)
‘Abd Alla2h ibn Yah5ya2 IV (I 905-
Ibra2hı3m (S 900- 22)
902) Shayba2n (T 905)
Ziya2dat Alla2h ibn Yaqz5a2n ibn Abı3 ’l-
‘Abd Alla2h (S Yaqz5a2n (R 907-
902-3) 9)
Muh5ammad ibn al- ‘Ubayd Alla2h al-
Sarqu2sı3 (S 903) Mahdı3 (F 909-
‘Alı3 ibn 34)
Muh5ammad (S
903, 909)
Ah5mad ibn al-H4u -
sayn (S 903-9)
Al-H4asan ibn
Ah5mad (S 910-
12)
Yu2suf ibn ‘Umar II
(C ca 910-15)
Alexander (912-13) ‘Abd al-Rah5m a2n
III (U 912-61)
(Table 4 continued)

(K) Kings of the

(A) Aquitaine
Pietro Candiano I Galindo II Azárez Ursus (B 891-2) Berengar I (I 888-

(V 887) (A 893-922) Guy IV of Spoleto 924, E 915-24)
Pietro Tribuno (V Wifred Borrell I (B (B 895-7) Louis III (I 900-5)
888-912) 898-912) Radelchis II (B Rudolf I (B 888-
Mutimir (S ?-891) Sancho I Garcés (N 897-900) 912)
Vladimir (B 889- 905-25) Atenolf I (C 887- Guy of Spoleto (I
93) García I (L910-14) 910, B 900- 889-95, E 891-)
Prvoslav (S 891-2) 910) Louis the Blind (P
Peter Gojnikovic) (S Guaimar II (S 900- 890-928, I 900-
892-917) 46) 928)
Mutimir (C 892- Atenolf II (CB Lambert (I 891-8,
910) 910-40) E 892-)
Symeon (B 893- Landolf I (CB 901- Arnulf (I. 894-6, E
927) 43 896)
Tomislav (C ca 910- Louis III (I 900-
28) 905, E 905)
Orso Partecipazio II Suñer (B 912-54) Rudolf II (B 912-

(V 912-32) 37, I 922-6)
56 CHAPTER ONE
(Table 4 continued)
(A) Aghlabids of
Tunisia
(F) Fa2t6imids
(I) Idrı3sids of
Morocco
(U) The Spanish (Ik) Ikhshı3dids of
The ‘Abba2sid
Umayyads
Caliphs Egypt
(B) Bu2y id amı3rs (C) Amı3rs of Crete (R) Rustamids of
(S) Amı3rs of Sicily Algeria
(T) T4u2lu2nids of
Egypt
(Z) Zı3rids of
Algeria
‘Alı3 ibn ‘Umar al-

Balawı3 (S 912-
13)
Constantine VII Ah5mad ibn Ziya2d at
(913-20) Alla2h (S 913-
16)
‘Alı3 ibn Yu2suf (C
ca 915-25)
Mu2sa2 ibn Ah5m ad
(S 916-17)
Sa2lim ibn Rashı3d
(S 917-37)
Ro2manos I (920-44) Al-Qa2hir (832-4) Ah5mad ibn ‘Umar Al-H4asan al-
Al-Ra2d5ı3 (934-40) II (C ca 925-40) Hajja2m (I 922-
Al-Muttaqı3 (940- Khalı3l ibn ‘Ish5a2q (S 6)
44) 937-41) Al-Qa2’im (F 934-
Shu‘ayb II ibn 6)
Ah5mad (C ca Muh5ammad ibn
940-43) T4ughj al-Ikhshı3d
Ibn al-Ku2fı3 (S 941- (Ik 935-46)
48)
‘Alı3 ibn Ah5mad (C
ca 943-9)
Constantine VII Al-Mustakfı3 (944- Al-H4asan ibn ‘Alı3 U›nu2ju2r (Ik 946-61)
(944-59) 6) al-Kalbı3 (S 948- Al-Mans5u2r (F 946-
Al-Mut6ı3‘ (946-74) 54) 53)
Mu‘izz al-Dawla Ah5mad ibn al- Al-Mu‘izz (F 953-
Ah5mad (B 945- H4asan (S 954- 75)
67) 69)
Ro2manos II (959- ‘Abd al-Azı3z ibn ‘Alı3 (Ik 961-6)

63) Shu‘ayb II (C
960-61)
(Table 4 continued)

(K) Kings of the

(A) Aquitaine
Igor (K ?-945)
Pavel Branovic) (S Ordoño II (L 914-

917-21) 24)
Fruela II (L 924-5)
Zaharije Prvoslav- Alfonso IV (L 925- Atenolf III (CB Hugh (I 926-47)

ljevic) (S 921-4) 30) 933-43) Lothar (I 931-50)
C!aslav Klonimiro- García I Sánchez Landolf II (CB Conrad (B 937-93
vic) (S ca 927/8- (N 925-71) 939-61
60) Ramiro I (L 930-
Peter (B 927-67) 50)
Pietro Candiano II Miro (B 940-66)/
(V 932-39) Borrell II (B
Pietro Partecipazio 940-92)
(V 939-42)
Pietro Candiano III
(V 942-59)
Svyatoslav (K Ordoño III (L 950- Gisulf I (S 946-77) Berengar II (I 950-
945—72) 55) Landolf III (CB 62)
Kresimir II (C 949- Borrell II (B 954- 959-68/9) Otto I (I, 951-73, E
69) 92) 962-)
Pietro Candiano IV Sancho I (L 955-
(V 959-76) 57, 960-67)
Ordoño IV (L 957-
60)
Pandolf I (CB 961-
81, S 977-81)
58 CHAPTER ONE
(Table 4 continued)
(A) Aghlabids of
Tunisia
(F) Fa2t6imids
(I) Idrı3sids of
(U) The Spanish
Umayyads Morocco
Caliphs Egypt
Algeria
(T) T4u2lu2nids of
Egypt
(Z) Zı3rids of
Algeria
Byzantine conquest
of Crete
Al-H4akam II (U
961-76)
Basil II (963)
Nike2phoros II (963- ‘Izz al-Dawla Ka2fu2r (Ik 966-9)
9) Bakhtiya2r (B Ah5mad (Ik 969)
967-78)
John I (969-76) Al-T4a2’i‘ (974-91) Ya‘ı3sh (S 969) Yu2suf Buluggı3n I
Abu2 ’l-Qa2sim (S ibn Zı3rı3 (Z 972-
970-82) 84)
Al-‘Azı3z (F 975-96)
Basil II (976-1025) Al-Qa2dir (991- Hisha2m II (U 976- Al-Mans5u2r ibn
1031) 1009) Buluggı3n (Z
‘Ad5ud al-Dawla Ja2bir (S 982-3) 984-96)
Fanna2 Khusraw Ja‘far ibn Muh5am- Na2s5ir al-Dawla
(B 978-83) mad (S 983-5) Ba2dı3s (Z 996-
S4ams5a2m al-Dawla ‘Abd Alla2h ibn 1016)
Marzuba2n (B Muh5ammad (S Al-H4a2k im (F 996-
983-7) 985-9) 1021)
Sharaf al-Dawla Yu2suf (S 989-98) Sharaf al-Dawla al-
Shı3rdhı3l (B Ja‘far (S 998-1019) Mu‘izz (Z 1016-
987-9) Muh5ammad II (U 62)
Baha2’ al-Dawla 1009) Al-Z4a2hir (F 1021-
Fı3ru2z (B 989- Hisha2m II (U 1010- 36)
1012) 13)
Sult6a2n al-Dawla (B Sulayma2n al-
1012-21) Musta‘ı3n (U
Musharrif al-Dawla 1013-18)
H4asan (B 1021- ‘Abd al-Rah5m an
4) IV al-Murtad5a2
(U 1018-23)
Ah5mad al-Akhal (S
1019-36)
(Table 4 continued)

(K) Kings of the

(A) Aquitaine
Landolf IV (CB
968/9-82)
Boris II (B 967-71) Ramiro III (L 967-

84)
Stjepan Drz°islav (C Sancho II Garcés Otto II (E 967-83)

969-97) (AN 970-94)
Yaropolk (K 972-8)
Samuel, Aaron, Vermudo II (L 982- Pandolf II (B 982- Rudolf III (B 993-

Moses, David 99) 1014) 1032)
(B 976-86) Ramon Borrell I (B Ladenolf (C 982- Otto III (E 996-
Pietro Orseolo I (V 992-1019) 93) 1002)
976-8) García II Sánchez Laidolf (C 993-9) Henry II (I 1004-
Vitale Candiano (V (AN 994-1000) Landolf V (B 987- 24, E 1014-)
978-9) Alfonso V (L 999- 1033) Conrad II (I 1024-
Tribuno Memmo 1027) Landolf V (C 39, E 1027-)
(V 979-91) Sancho III Garcés 1000-7)
Samuel (B 986- (AN 1000-35) Pandolf II (C
1014) Berenguer Ramon I 1007-22
Pietro Orseolo II (V (B 1019-35) Pandolf III (B
991-1008) 1101-60)
Vladimir I (K 978- Manso (S 981-3)
1015) John II (S 983-9)
Svetoslav (C 997- Guaimar III (S
1000 989-1027)
Otto Orseolo (V Pandolf III (C
1008-26) 1014-26)
Kresimir III (C Pandolf IV (C
1000-30) 1016-49)
60 CHAPTER ONE
(Table 4 continued)
(A) Aghlabids of
Tunisia
(F) Fa2t6imids
(I) Idrı3sids of
(U) The Spanish
Umayyads Morocco
Caliphs Egypt
Algeria
(T) T4u2lu2nids of
Egypt
(Z) Zı3rids of
Algeria
Abd al-Rah5m an V
al-Mustaz5hir (U
1023-4)
Muh3ammad III al-
Mustakfı3 (U
1024-7)
recur in vengeance under Tsar Symeon. But a new threat had emerged
in the Ukraine. Scandinavians settled along the Dnepr river around the
rapids and especially at Kiev, who became known to the Byzantines
as the ÔRw'", the Rho2s, launched the first of several attacks on
Constantinople in 860.95 Although beaten off after ravaging the
environs of Constantinople, the attack presaged a new and powerful
force which would affect the Empire for centuries, especially up to the
conversion of Prince Vladimir I of Kiev in 988 and the defeat of the
last Rho2s attack on Constantinople in 1043. On the Tauros frontier the
virtually independent frontier amı3rs of the ‘Abba2sids continued the
interminable warfare of annual raids. The river Lamos in Cilicia west
------------------------------
95
Photios, Homilies, III & IV (pp. 82-110); John Zo2naras, Epitome2 historio2n,
XVI.5 (vol. 4, p. 15); Russian primary chronicle, p. 60; Theophane2s continuatus,
IV.33 (p. 196). See also Vasiliev, Russian attack on Constantinople.
(Table 4 continued)

(K) Kings of the

(A) Aquitaine
Gabriel Radomir (B Pandolf V (C

1014-15) 1020-57)
John Vladislav (B Pandolf VI (C
1015-18) 1022-26)
Svjatopolk (K
1015-19)
Msistlav (K 1019-
36)
John Vladmir (S
late 10th C. -
1016)
Interregnum (S
1016-34
of Tarsos marked the border and from 805 to 946 its banks witnessed
a depressing series of prisoner exchanges and redemptions.96 At sea
the Cretans continued their corsair war, raiding Dalmatia in 872,
although more normally they confined themselves to the Aegean.
However, imperial squadrons began to get the better of them from the
870s. Around 873 Nike2tas O›oryphas engaged them off Kardia at the
head of the Gulf of Saros, destroying 20 skaphe2 with Greek Fire. Then
in 879 he inflicted a major defeat on them when he destroyed in the
Gulf of Corinth a squadron which had been raiding in the Ionian.97
------------------------------
96
Al-Mas‘u2dı3, Muru2j, vol. 8, pp. 224-5; idem, Al-Tanbı3h, pp. 356-61; Al-T4abarı3,
Ta’rı3kh (Yar-Shater), A.H. 189, 192, 231, 241, 245-6, 283, 292, 295 (vol. 30, pp. 257,
291; vol. 34, pp. 38-43, 137-41, 156, 168-70; vol. 38, pp. 32-3, 153, 185).
97
John Skylitze2s, Synopsis historio2n, Basivleio" oJ Makedw'n.30-31 (pp. 152-4);
John Zo2naras, Epitome2 historio2n , XVIII.9 (vol. 4, pp. 27-8, 31-2); John the Deacon,
Cronaca Veneziana, pp. 119-20; Theophane2s continuatus, V.60-61 (pp. 299-301).
62 CHAPTER ONE
The ‘Abba2sids also began to rebuild naval forces in Cilicia and by

860 squadrons based at Tarsos were sufficiently powerful to attack
Antalya.98 Perhaps Ah5mad ibn T4u2lu2n had been doing so also, since a
century later Constantine VII reported that around 875 he had been
preparing a vast fleet in Egypt and Syria to attack the Empire. 99
Possibly in response to this threat or to the build-up of Muslim forces
in Cilicia, Basil I is said to have sent an expedition to recover Cyprus
under a strate2gos called Alexios and to have made it into a thema. But
even if the expedition actually took place, the Byzantines held the
island only for a few years.100
Tarsos became such a threat to Byzantine territory that in 883 a
large army was sent against it under the domestikos to2n scholo2n Kesta
Styppeio2te2s; however, he was attacked at night while unprepared and
defeated and killed by Ya2zama2n al-Kha2dim. Ya2zama2n was amı3r of
Tarsos from 882 until his death in 891 and became renowned for the
raids of his naval squadrons. Shortly after 883 he led a raid by 30
koumparia against Euripos but was beaten off by the strate2gos of
Hellas, a certain Oiniate2s, who used Greek Fire against the Muslim
ships. He continued hostilities until his death, after which Tarsos was
held by governors of the T4ulu2nids of Egypt until 897 and then of the
‘Abba2sids, all of whom continued the annual raids into Byzantine
territory.101 In 898 a Tarsiote squadron under Ra2ghib, a client of al-
Muwaffaq, a brother of the Caliph al-Mu‘tamid, encountered a
Byzantine fleet, probably that of the Kibyrrhaio2tai, and defeated it,
capturing numerous ships and beheading 3,000 seamen. The victory
exposed the Aegean to the depredations of Leo of Tripoli, known to
the Muslims as Gula2m Zura2fa, a former Byzantine seaman from the
Kibyrrhaio2tai who had been taken prisoner and had then converted to
Islam. In 904 he led a devastating raid into the Aegean which
penetrated the Dardanelles and sacked Abydos, the droungarios tou
ploimou Eustathios Argyros declining battle. Leo then turned back
------------------------------
98
Al-T4abarı3, Ta’rı3kh (Yar-Shater), A.H. 246 (vol. 34, p. 167). On the fleets of
Tarsos, see Fahmy, Muslim naval organization, pp. 56-63.
99
John Skylitze2s, Synopsis historio2n, Basivleio" oJ Makedw'n.36 (pp. 157-8);
Theophane2s continuatus, V.68 (pp. 308-9).
100
Constantine VII, De thematibus, XV.40 (pp. 80-81). This story is corroborated
by no other sources, neither Greek nor Arabic, and there is considerable reason to
doubt its historicity. See Pertusi’s Introduction to De thematibus, pp. 26-7.
101
Al-Mas‘u2dı3, Muru2j, vol. 8, pp. 71-2; Al-T4abarı3, Ta’rı3kh (Yar-Shater), A.H. 269-
70, 272, 274-5, 278 (vol. 37, pp. 81-2, 143-4, 152, 155, 157, 175); John Skylitze2s,
Synopsis historio2n, Basivl eio" oJ Makedw'n.25, 29 (pp. 144-5, 151); John Zo2naras,
Epitome2 historio2n, XVI.9 (vol. 4, pp. 31-2); Theophane2s continuatus, V.50-51, 59
(pp. 286-8, 298-9).
and captured and sacked Thessalonike2 before retiring back to Tripoli

via Crete with his fleet laden with booty and prisoners. According to
al-T4abarı3, Leo killed 5,000 people in Thessalonike2, delivered to
freedom 4,000 Muslim captives, captured 60 ships and took thousands
of prisoners. Each man on the expedition received 1,000 gold dinars
from the proceeds of the booty.102 A letter written by Patriarch
Nicholas I Mystikos to the amı3r of Crete may have called for the
release of the prisoners taken at Thessalonike.103
In response to the depredations of these corsair amı3rs of Cilicia,
Leo VI sent a large fleet to the Levant in 910 under the patrikios and
logothete2s tou dromou Himerios, who had already won an important
victory over the Muslims on the “Day of the Apostle Thomas”,
probably in 905 or 906. Although there is great confusion in the
sources about Himerios’s expedition, complicated by the fact that
what purports to be a collection of inventories for an invasion of Crete
by it in 911 was inserted into the treatise De cerimoniis,104 the
expedition certainly began in 910 as an assault on Muslim naval
forces in the Levant. In response, the amı3r of Tarsos, Damianos,
another convert to Islam, ravaged Cyprus for four months and took
many prisoners, probably because the inhabitants had broken the
terms of their covenant by assisting Himerios.105 Himerios probably
assaulted Crete unsuccessfully the following year and his fleet was
annihilated north of Chios in October 912 by Leo of Tripoli and
Damianos. The expedition began successfully but ended disastrously
as a three-year effort to reduce Muslim capabilities in the Levant and
at the entrance to the Aegean. A second letter of Patriarch Nicholas I
Mystikos, dated to 913-14, pleaded with the Caliph al-Muqtadir, for
------------------------------
102
Al-T4abarı3, Ta’rı3kh (Yar-Shater), A.H. 285, 291 (vol. 38, pp. 73, 148); Al-
Mas‘u2dı3, Muru2j, vol. 2, pp. 318-9; John Kaminiates, De expugnatione Thessalonicae;
John Skylitze2s, Synopsis historio2n , Levwn oJ filovsofo".23 (pp. 182-3); Theophane2s
continuatus, VI.Basileiva Levo nto" aujtokravtoro".20-21 (pp. 366-8); Vita di Sant’Elia,
pp. 108-9.
103
Nicholas I, Letters, No 2 (pp. 12-17). There is doubt about the dating of this
letter and it is possible that rather than it being dated to 904-5 and calling for the
release of the prisoners of Thessalonike2 it should be dated to 913-14 and refer to
prisoners taken during the defeat of Himerios’s expedition of 910-12.
104
See below pp. 186-7.
105
Al-Mas‘u2dı3, Muru2j, vol. 8, pp. 281-2; John Skylitze2s, Synopsis historio2n, Levwn oJ
filovsofo".29, 33 (pp. 186, 191); John Zo2naras, Epitome2 historio2n, XVI.14 (vol. 4, pp.
44-5); Theophane2s continuatus, VI.Basileiva Levo nto" aujtokravtoro".26, 31 (pp. 371-
2, 376-7), VI.Basileiva Alexavndrou uiJou' Basivl eiou.5 (pp. 379-80). See also
Vasiliev/Canard, Byzance et les Arabes. Tome II, part 1, pp. 196-216. It is even
possible that there was no actual Cretan expedition at all in 911. See Haldon, “Theory
and practice”, pp. 241-2. See also Appendix Four [a].
64 CHAPTER ONE
the release of the Cypriote captives.106 Between 909 and 916 three
inscriptions raised at Antalya, the headquarters of the thema of the
Kibyrrhaio2tai, recorded the strengthening of the walls against Muslim
attack by the droungarios Stephen. He added a second, inner wall to
prevent Muslim ships using flying bridges from their mastheads to
overtop the walls.107 Not until the third decade did the tide really begin
to run out. Possibly the accession to the imperial throne in 920 of one
of the only two Byzantine admirals ever to do so, the droungarios tou
ploimou Ro2manos I Lekape2nos, was instrumental. Leo of Tripoli was
defeated by the patrikios John Rhade2nos off Lemnos in 923 and
probably killed. Damianos died besieging the Kibyrrhaio2tai fortress at
Strobilos in the following year, after which the threat to the Empire
from Muslim naval forces and corsairs in Syria, Cilicia, and Crete
began to dissipate.108
In Sicily the fall of Enna in 858 confined the Byzantines to the
coastal strip from Taormina to Syracuse and Cape Passero. Pressure
mounted on Syracuse by a new governor, Khafa2ja ibn Sufya2n, led
Basil I to send a new expedition to the island and the fleet, which may
well have been that of Nike2tas O›oryphas which had just relieved
Dubrovnik, reached Syracuse in 868 and was engaged at sea by
Muslim fire ships, harra2qa2t, under the command of Khafa2ja’s son
Muh5ammad, while Khafa2ja himself engaged the Byzantine forces by
land. The campaigns were apparently indecisive and the fleet may
then have retired back to the Adriatic in time to join Louis II in his
assault on Bari in the following year.109 Under the governorship of
Muh5ammad ibn Khafa2ja all of the islands around Sicily fell to the
Aghlabids. Malta fell on 29 August 870. From Sicily the amı3rs and
their Aghlabid masters harrassed Italy mercilessly. In 871 they
launched a massive attack on Salerno, which was beseiged for over a
year but eventually relieved by an army sent by Louis II. In 875 they
penetrated the Adriatic as far as the Gulf of Trieste and besieged
Grado unsuccessfully, razing Comacchio during their retreat. At the
same time they were raiding Campania and the west coast of Italy as
------------------------------
106
Nicholas I, Letters, No 1 (pp. 2-13). The letter is addressed to the amı3r of Crete
but it has been demonstrated beyond doubt that it was actually addressed to al-
Muqtadir.
107
See Trombley, “War, society and popular religion”, pp. 125-7.
108
John Skylitze2s, Synopsis historio2n, Kwnstanti'no" oJ uiJo;" Levo nto".7 (p. 202),
ÔRwmano;" oJ Lakaphnov".11 (p. 218); Theophane2s continuatus, VI.Basileiva
Kwnstantivnou uiJo u' Levo nto".9 (p. 388), Basileiva Rwmanou'.14 (p. 405).
109
Ibn al-Athı3r, Al-Ka 2mil (Fagnan), pp. 238-9; Ibn ‘Idha2rı3, Al-baya2n al-mughrib,
vol. 1, pp. 148-9; Ibn Khaldu2n, ‘Ibar (Des Vergers), pp. 124-5.
far north as Rome. Pope John VIII tried to organize a coalition against
them in 875 and in 877 asked the Byzantine strate2gos of Longobardia
at Bari, Gregory, to send 10 chelandia to defend Rome. In 880 he
asked Basil I himself to send a fleet.110
If the fall of Bari to Louis II and then to the Byzantines had
virtually eliminated the Muslim threat in Apulia, the situation was
different in Calabria and Campania on the west coast, where the ill-
defined political structures of the Lombard principalities and the three
quasi independent Byzantine duchies of Gaeta, Naples, and Amalfi
provided opportunities. In 880 the bishop-duke of Naples, Athanasius
II, allowed a Muslim band to settle at the foot of Mt Vesuvius and
others from Saepinum raided as far north as Spoleto. Later another
band settled at Cetara on the Gulf of Salerno. Naples and Salerno
combined between 881 and 883 to drive out these various nests but
they moved north and joined others on the Garigliano river near
Gaeta. In 884 the great abbey of Montecassino was sacked.111
The amı3rs of Sicily attempted to take Syracuse again in 869 and
873 and the city finally fell in 878. A relieving fleet under a patrikios
named Adrian was supposedly delayed for 50 days by contrary winds
at Hierax in the Peloponne2sos until too late.112 The remaining Greeks
held out around Taormina until it itself fell to the amı3r ‘Abd Alla2h ibn
Ibra2hı3m ibn Ah5mad in 902.113
In 880 the Aghlabid Ibra2hı3m II took his fleet into the Ionian Sea,
raiding Kefalle2nia and Zakynthos. Basil I responded by sending out a
fleet of 45 warships (trie2reis, die2reis, and ne2es tachynautousai) under
------------------------------
110
Andrew of Bergamo, Historia, §18 (p. 229); Chronicle of Cambridge, pp. 30-33,
60-61; Chronicon Salernitanum, §§111-8 (pp. 123-33); Erchempert, Historia
Langobardorum, §§35, 39 (pp. 247-8, 249); John the Deacon, Cronaca Veneziana, p.
121; John VIII, Registrum, Ep. 22, 31, 32, 47 (pp. 19-21, 29-30, 31-2, 45-6); Leo
Marsicanus, Chronica, I.40 (pp. 608-9).
111
Chronicon Salernitanum, §§126, 130, 136 (pp. 139-40, 142-3, 145);
Erchempert, Historia Langobardorum, §§44, 49, 51, 61 (pp. 251-4, 255-6, 259); Leo
Marsicanus, Chronica, I.43-4 (pp. 609-10); Liudprand of Cremona, Antapodosis,
II.44-5 (pp. 57-8).
112
Hierax was a small port with a sheltered harbour some 15 kilometres north of
Monemvasia. See Kalligas, Monemvasia, pp. 51-4.
113
Chronicle of Cambridge, pp. 32-3, 38-9, 60-61, 68-9; Genesios, Basileiai, D.33
(pp. 82-3); Ibn al-Athı3r, Al-Ka2mil (Fagnan), pp. 239-40, 244, 247-9, 253-4; Ibn
‘Idha2rı3, Al-baya2n al-mughrib, vol. 1, p. 152; John Skylitze2s, Synopsis historio2n,
Basivleio" oJ Makedw'n.37 (pp. 158-9), Levwn oJ filovsofo".21 (p. 181); Theodosios the
Monk, Espugnazione di Siracusa; Theophane2s continuatus, V.69-70, VI.Basileiva
Levonto" aujtokravtoro".18 (pp. 309-12, 365); Vita di Sant’Elia, pp. 74-5, 80-83. It
would be extremely improbable, meteorologically impossible in fact, for a fleet to be
prevented by contrary winds from rounding Cape Malea into the Ionian Sea for 50
continuous days.
66 CHAPTER ONE
the patrikios and droungarios tou ploimou Nasar. Although some of

his crews mutinied at Metho2ne22, Nasar destroyed the Aghlabid fleet off
western Greece by the extremely bold tactic of attacking at night.
Night battles at sea were rare because darkness made tactical
manœuvring impossible and outcomes unpredictable. Continuing
west, Nasar landed near Palermo, ravaged its district, captured many
Muslim ships, and won another victory off Punta di Stilo while
returning. These successes enabled the Byzantines to send a squadron
under a spatharios Gregory, a tourmarche2s Theophylaktos, and a
kome2s Diogene2s to Naples, where they won another important victory.
The fleets returned in triumph to Constantinople and after their
departure the Byzantine commander in Italy, the patrikios Leo
Apostype2s, finally succeeded in taking Taranto from the Muslims.114
Basil I followed up these successes by sending to Italy an army
under the domestikos to2n scholo2n, Nike2phoros Pho2kas, who rapidly
reduced many towns and fortresses with conduct so exemplary that
when Leo VI produced his famous Taktika, his manual of strategy and
tactics for war, around 905, he devoted a section to it. Acting upon the
success of Nike2phoros Pho2kas, Pope Stephen VI asked Basil to send
chelandia to protect the coasts of Rome and the Byzantines even
captured Benevento in 891, although they held it for only four years.
But at sea they had a major setback in October 888 while attempting
to invade Sicily when the fleet was defeated either in the Straits of
Messina or off Milazzo and the Muslims then sacked Reggio.115
In following decades the Empire was occupied by new threats from
the North. Symeon of Bularia attempted to make Bulgaria the equal of
the Empire and even to conquer it. Even though educated in
Constantinople, he proved to be an implacable and formidable enemy.
Opening hostilities in 893, in 896 he annihilated a Byzantine army at
Boulgarophygon, after which he agreed to a truce, which held, by and
large, for around 16 years. It is probable, even though there is some
doubt about it, that in 907 Oleg, the Rho2s prince of Novgorod and
perhaps of Kiev also, attacked Constantinople and forced Leo VI to
------------------------------
114
Chronicle of Cambridge, pp. 32-3, 62-3; Genesios, Basileiai, D.34 (pp. 83-5);
Ibn al-Athı3r, Al-Ka2m il (Fagnan), p. 257; John VIII, Registrum, Ep. 245, 263 (pp. 214,
233); John Skylitze2s, Synopsis historio2n, Basivl eio" oJ Makedw'n.32-4 (pp. 154-7);
John Zo2naras, Epitome2 historio2n, XVI.10 (vol. 4, p. 32); Theophane2s continuatus,
V.62-6 (pp. 302-6); Vita di Sant’Elia, pp. 36-9.
115
Chronicle of Cambridge, pp. 34-5, 64-5; Erchempert, Historia Langobardorum,
§81 (p. 264); Ibn al-Athı3r, Al-Ka2mil (Fagnan), p. 262; Ibn ‘Idha2rı3, Al-baya2n al-
mughrib, vol. 1, pp. 157-8; Leo Marsicanus, Chronica, I.49 (p. 615); Leo VI, Taktika
(PG), XV.38 (coll. 895-6); Theophane2s continuatus, V.71 (pp. 312-13); Vita di
Sant’Elia, pp. 56-7. On the letter of Stephen VI, see below p. 168 & n. 27.
grant the Rho2s rights of access. In 913 Symeon of Bulgaria again

marched on Constantinople and the administration of Emperor
Alexander, headed by the Patriarch Nicholas I Mystikos, was forced
to come to terms and to crown him. Hostilities were nevertheless
renewed and Symeon crushed Byzantine armies at Achelo2on and
Katasyrtai in 917. Bulgarian columns raided as far south as the Gulf
of Corinth, creating the conditions in which the droungarios tou
ploimou, Ro2manos Lekape2nos, was able to seize the throne. His
attempts to neutralize Symeon had only limited success and in 922
Symeon invaded again and won a victory near Pe2ge2 outside the walls
of Constantinople. Adrianople was temporarily lost. Negotiations for
peace failed in 924 and the Bulgarian threat was removed only by
their defeat at the hands of Prince Tomislav of Croatia in 826 and by
Symeon’s death in 927. His son Peter came to terms.116
A political entity amongst Croatian Slavs who had entered the
Balkans in the seventh century was recognized by the Papacy as early
as 879, during the rule of Prince Branimir. However, it matured only
during that of Tomislav, who was created King. South of Croatia and
along the coast in the province of Duklja north of the Byzantine thema
of Dyrrachion lived other groupings of Slavs who became known as
Serboi and whose first rulers included Vlastimir in the mid ninth
century and C!aslav Klonimirovic°, who was the real founder of an
independent Serbian polity. Around the mouth of the Neretva river
and north as far as the Cetina river and on islands off the coasts lived
the tribes of the Neretljani, known to the Byzantines as the Arentanoi
and to the Venetians as Narentan “pirates”, who seriously menaced
their shipping from the ninth century. Doge Pietro Tradonico sailed
against them with a fleet as early as 839 and Pietro Candiano I led
another expedition in 887 which ended in defeat and his own death in
battle. Sixty years later in 948 his grandson Pietro Candiano III
returned to the attack with a fleet of 33 gumbariae with limited
success. But their menace could not be eliminated and Venice
continued to pay protection money for safe passage of her ships along
------------------------------
John Skylitze2s, Synopsis historio2n, Levwn oJ filovsofo".12, 14 (pp. 175-8),
116
Kwnstanti'no" oJ uiJo;" Levo nto".3-6, 8-9 (pp. 200-205), ÔRwmano;" oJ Lakaphnov".3, 6, 10,
12, 16 (pp. 213-15, 218-22); John Zo2naras, Epitome2 historio2n , XVI.12, 16-18 (vol. 4,
pp. 40-41, 52-61); Theophane2s continuatus, VI.Basileiva Levonto" aujtoktravtoro".9-
10 (pp. 357-60), VI.Basileiva Kwnstantivnou uiJo u' Levonto".5-8, 10 (pp. 385-8, 389-
90), VI.Basileiva Rwmanou'.5, 8, 10, 13, 15-16, 20-23 (pp. 400-409, 411-15). See also
Runciman, First Bulgarian Empire; Fine, Early medieval Balkans.
The Rho2s attack of 907 is not reported in any Greek sources and its veracity has
therefore been questioned. It is reported only in the Russian primary chronicle, pp.
64-9.
68 CHAPTER ONE
the Dalmatian coast, which the Venetians had to use because the east
coast of Italy was a dangerous lee shore. Not until the year 1000 did
Doge Pietro Orseolo II finally subdue them in a victorious campaign
and consolidate Venetian hegemony in the Adriatic.117
In Italy the Aghlabids made a last major effort in June 901 when
Abu2 ’l-‘Abba2s , the son of the amı3r ‘Abd Alla2h II ibn Ibra2hı3m, seized
Reggio. If we can believe Ibn al-Athı3r, a Byzantine relief fleet under
Eustathios, the strate2gos of Calabria, was defeated and lost 30 ships
off Messina in 902. ‘Abd Alla2h then crossed into Calabria after taking
Taormina but died besieging Cosenza and his army melted away.
Nevertheless, the Muslim menace was still sufficient to persuade
Constantinople to order Eustathios to conclude a truce with Ah5mad
ibn Ziya2dat Alla2h, the amı3r of Sicily, sometime around 914, agreeing
to a humiliating annual tribute of 22,000 pieces of gold. Then in 915
combined forces of the Papacy, Spoleto, Gaeta, Camerino, Benevento,
and Salerno finally eliminated the corsair nest on the Garigliano, a
Byzantine fleet closing the mouth of the river. But even if the threat of
Muslim conquest had passed, Sicilian corsairs continued to harrass the
coasts of Italy for another 50 years and were joined by squadrons of
the new Fa2t5imid Caliphate operating from al-Mahdiyya in Ifrı3qiya.118
In the north-west of the Sea the second half of the ninth century
and the tenth was a period of political disintegration. Even when
someone continued to hold a title, the kings of Italy, Aquitaine,
Burgundy, and Provence became increasingly titular. Real authority
drifted to the margarves, counts, and dukes of Tuscany, Provence,
Burgundy, Gascony, Toulouse, and Aquitaine. At the same time the
Umayyad amı3rate of al-Andalus reached its zenith during the long
reign of ‘Abd al-Rah5ma2n III, who assumed the title of Caliph.
As early as 879 Muh5ammad I had built a fleet at Cordoba with
which to attack Galicia; however, it broke up at sea, the mara2kib no
doubt being unsuitable for Atlantic conditions. Early in his reign ‘Abd
al-Rah5ma2n III sent light flotillas cruising the Straits of Gibraltar to
prevent rebels against his rule receiving reinforcements and provisions
------------------------------
117
Constantine VII, De administrando imperio, §§29-36 (pp. 122-65); John
Skylitze2s, Synopsis historio2n, Basivleio" oJ Makedw'n.26 (pp. 145-7); Theophane2s
continuatus, V.52, 54 (pp. 288-9, 291-2); John the Deacon, Cronaca Veneziana, pp.
113, 128-9, 136, 155-60. See also Fine, Early medieval Balkans.
118
Chronicle of Cambridge, pp. 40-48, 70-82; Ibn al-Athı3r, Al-Ka2m il (Fagnan), pp.
263-6; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 1, p. 175; Ibn Khaldu2n, ‘Ibar (Des
Vergers), pp. 136-44; John Skylitze2s, Synopsis historio2n, Nikhfovra" oJ Fwka'".4 (p.
263); Kita2b al-‘uyu2n, A.H. 289 (p. 221); Leo Marsicanus, Chronica, I.52 (pp. 616-7);
Liudprand of Cremona, Antapodosis, II.51-4 (pp. 61-2); Translatio Sancti Severini, p.
457, n. 1; Vita di sant’Elia, pp. 62-3, 82-3.
from the Maghrib. When he perceived the danger posed by the new
Fa2t6imid fleet, he built up his own and this enabled him to take Melilla
in 927 and Ceuta in 931. In 953 an Umayyad amı3r, Ah5mad ibn Ya‘la2,
led the fleet on a raid to Galicia, returning in triumph with the bells
and crosses of Christian churches. The Fa2t6imid fleet sacked Almeria
in 955 and in response ‘Abd al-Rah5ma2n III sent the Umayyad fleet to
ravage Fa2t6imid territory. The fleet was active in the Maghrib again in
958. If we can believe Ibn Khaldu2n, during ‘Abd al-Rah5ma2n’s reign
the fleets of both the Umayyads and Fa2t6imids had grown to the
formidable numbers of 200 mara2kib each.119
In the tenth century, the most important Umayyad naval base was
Pechina in the hills behind Almeria, which was inhabited by an
admixture of Ghassa2nı3 Arabs originally from Syria and sailors of
Andalusi origin, some Muslim and some Christian. Prior to 884 they
formed a self-governing community at Pechina and succeeded in
resisting attempts by the Arab chiefs of Elvira to take them over. In
the mid tenth century they moved down to the port of Almeria. Under
‘Abd al-Rah5ma2n III the fleet of Pechina was the main Umayyad fleet
operating against the Fa2t6imids in the Maghrib.120
Around 890 a group of Andalusi corsairs landed in the bay of St
Tropez and fortified themselves on a hilltop at Fraxinetum. There they
established for almost a century a Muslim enclave from which they
raided as far west as Marseilles, as far north as Vienne, as far east as
Asti, and as far to the north-east as the abbey of St Gall in
Switzerland. Attempts to expel them in 931 and 942 failed and not
until 972 did the Counts of Provence and Turin succeed in doing so
with the help of a Byzantine fleet. Wrecks of tenth-century Muslim
ships found off the coast of Provence suggest that this enclave enjoyed
lively maritime communications with the main Muslim world.121
By the 920s Italy was divided between a Byzantine thema of
Longobardia, the Lombard principalities of Capua-Benevento and
Salerno, Papal territories around Rome to the west of the Apennines,
and the Kingdom of Italy in the North. However, although there
continued to be kings, real control of much of the North lay with the
------------------------------
119
Akhba2r Majmu2‘a, pp. 133-42; Al Bakrı3, Kita2b al-mughrib, pp. 179, 205; Ibn al-
Athı3r, Al-Ka2mil (Fagnan), p. 257; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 1, pp. 289-
95, vol. 2, pp. 170, 339, 363, 366, 369; Ibn Khaldu2n, Muqqadimah, vol. 2, p. 40. The
word translated by Rosenthal as “vessels” is actually mara2kib.
120
Al-Bakrı3, Kita2b al-mughrib, pp. 128-9, 158-9, 163.
121
Liudprand of Cremona, Antapodosis, I.2-4, II.43, IV.4, V.16-17 (pp. 5-6, 56-7,
104-5, 139); Ralph Glaber, Historiae, I.9 (pp. 20-23). See also Senac, Musulmans et
Sarrasins.
70 CHAPTER ONE
margraves of Tuscany, Ivrea, and Friuli, and the Dukes of Spoleto.

The Magyars had already raided into northern Italy in 899-900 and
904-5 but from 922 the whole of the peninsula would be seriously
disrupted by their raids, which recurred in 937, 940, and again in
either 947 or 949, and which reached as far south as Apulia and
Salerno.122 At the same time they raided into the Balkans, reaching
Constantinople in 934. Their assaults were halted only by their defeat
by the Western Emperor Otto I at the battle of the Lech in 955.
In Sicily rule by Fa2t5imid amı3rs replaced that of the Aghlabids from
al-H4asan ibn Ah5mad in 910. Although indigenous revolts flared from
time to time, from then until 948 the amı3rs would mainly be Fa2t6imid
appointees. Sicilian squadrons raided Calabria and the Basilicata in
925-6 and 929, in spite of the truce bought in 914.123 However, by this
time such Sicilian raids had become mere pin-pricks. More
threatening were the exploits of Fa2t5imid squadrons. In 925 the h5a2jib
Abu2 Ah5mad Ja‘far took the fleet to Apulia and sacked Bruzzano and
Oria, taking many Jewish prisoners back to Ifrı3qiya.124 Then, in 927,
and possibly again in 928 and 929, the Slavic amı3r S4a2bir sailed from
al-Mahdiyya with 44 mara2kib. Taranto was sacked, probably in 928.
In 935 they even sacked Genoa.125
Resistance to Fa2t6i mid attempts to impose their hegemony over the
powers of the Maghrib led to a struggle for influence between them
and ‘Abd al-Rah5ma2n III, the Umayyad seizure of Melilla in 927 and
Ceuta in 931 being part of it. When an Umayyad ship attacked and
captured a Fa2t6imid one off Ifrı3qiya in 955, it led to war and the amı3r
of Sicily, al-H4asan ibn ‘Alı3 al-Kalbı3, now back in Fa2t6imid service, was
ordered to use his fleet against al-Andalus. He attacked Almeria and
destroyed the Umayyad fleet there. In riposte an Umayyad fleet of 70
ships attacked Ifrı3qiya, sacking al-Kala, Su2sa, and T4abarqa. The
Fa2t6imid Caliph al-Mu‘izz then sent his general Jawhar al-S4aqlabı3 to
------------------------------
122
Annales Barenses, Annus 949 (p. 53); Annales Beneventani, Annus 922 (p. 175);
Leo Marsicanus, Chronica, I.55 (p. 619); Liudprand of Cremona, Antapodosis, II.7-
16, 42, 61-2 (pp. 41-5, 56, 64-5); Lupus Protospatharios, Annales, Annus 947 (p. 54);
Romuald of Salerno, Chronicon, pp. 165-6.
123
Ibn al-Athı3r, Al-Ka2m il (Fagnan), p. 320; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol.
1, p. 301.
124
Chronicle of Cambridge, pp. 42-3, 72-3; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol.
1, p. 271; Lupus Protospatharios, Annales, Annus 924 (p. 53).
125
Chronicle of Cambridge, pp. 42-3, 72-3; Al-Nuwayrı3, Niha2yat ‘al-Arab (Amari),
p. 128; idem, Niha2yat ‘al-Arab (Caussin), pp. 420-21; Ibn al-Athı3r, Al-Ka2m il
(Fagnan), p. 317; idem, Al-Ka2mil (Amari), pp. 411-12; Ibn ‘Idha2rı3, Al-baya2n al-
mughrib, vol. 1, pp. 274, 277, 279; Ibn Khaldu2n, ‘Ibar (De Slane), vol. 2, p. 529;
idem, ‘Ibar (Des Vergers), pp. 162-3; idem, Muqqadimah, vol. 2, p. 41; Kita2b al-
‘uyu2n, A.H. 315 (p. 223). See also Lev, “Fa2t6imid navy”.
the Maghrib in 958-9 and in a triumphal campaign he reimposed

Fa2t6imid rule to the shores of the Atlantic, with the exception of Ceuta
and Tangier, which the Umayyads continued to hold.126
In this context, when Constantine VII was planning the expedition
to recover Crete which would take place in 949, he sent an embassy to
Cordoba in 946 which was received with great pomp and ceremony.
Another followed in 949. The intent was no doubt to neutralize any
Fa2t6imid assistance to Crete and perhaps also to open the way to
reinforce Byzantine rule in southern Italy.127 The Cretan expedition
was, however, a disastrous fiasco, probably because of the
incompetence of the commander, the he2ge2tor naumachias
(“commander of the sea battle”), Constantine Gongyle2s .128
Probably in the following year Constantine VII sent a patrikios
named Malake2nos to Italy with a fleet under the command of
Makroio2anne2s . In response the amı3r of Sicily, ‘Alı3 ibn Abı3 ’l-H4usayn
al-Kalbı3, sought reinforcements from his Fa2t6imid master, al-Mans5u2r bi
’lla2h, landed in Calabria in 951, and inflicted a series of severe defeats
on the Byzantines. Returning in 952 he again defeated Malake2nos near
Gerace. Constantine VII appears to have then sent the ase2kre2tis John
Pilatos to conclude a truce with al-Kalbı3 and another embassy to al-
Mahdiyya to confirm it with his master.129 Following these reversals,
and because Duke John of Naples had been pursuing a policy of
accomodation with the Muslims of Sicily and the Lombards of Capua-
Benevento, Constantine sent another large expeditionary force to Italy
------------------------------
126
Akhba2r Majmu2‘a, pp. 133-42; Al-Bakrı3, Kita2b al-mughrib, pp. 179, 205; Ibn al-
Athı3r, Al-Ka2m il (Fagnan), pp. 358-9; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 1, pp.
289-95, vol. 2, pp. 170, 339, 363, 366, 369; Ibn Khaldu2n, ‘Ibar (Bu2laq), vol. 4, pp.
139-40; idem, ‘Ibar (De Slane), vol. 2, pp. 542-4.
127
Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 2, pp. 353, 357; Liudprand of Cremona,
Antapodosis, VI.4 (p. 337).
128
We say “probably” because the sources are extremely hostile to him.
Theophane2s continuatus has no details, perhaps because the expedition was such a
disaster that Theodore Daphnopate2s, the presumed author of the relevant section, did
not wish to include it in his account of the reign of Constantine VII. He2ge2tor
naumachias was an author’s description not his dignity or office. He was a palace
eunuch, one of the chamberlains, apparently without military experience. “Gongyle2s”,
“Turnip”, was a derogatory nickname, not his real name. See John Skylitze2s, Synopsis
historio2n, Kwnstantivno" pavlin aujtokravtwr.15 (pp. 245-6); John Zo2naras, Epitome 2
historio2n, XVI.22 (vol. 4, p. 70); Leo the Deacon, Historiae, I.bV (pp. 6-7);
Theophane2s continuatus, VI.Autokratoriva Kwnstantivnou.1-3 (vol. 1, pp. 436-8);
John Zo2naras, Epitome2 historio2n, XVI.22 (vol. 4, p. 70).
129
Chronicle of Cambridge, pp. 44-7, 74-7; Ibn al-Athı3r, Al-Ka 2mil (Fagnan), pp.
350-54; Ibn Khaldu2n, ‘Ibar (De Slane), vol. 2, pp. 540-41; idem, ‘Ibar (Des Vergers),
pp. 167-9; John Skylitze2s, Synopsis historio2n, Nikhfovra" oJ Fwka'".7, 8 (p. 266); Kita2b
al-‘uyu2n, p. 225; Lupus Protospatharios, Annales, Annus 951 (p. 54).
72 CHAPTER ONE
in 956 under the patrikios Marianos Argyros to bring Naples and the
Lombards back to their allegiances and to relieve Calabria and
Campania from Sicilian attentions. A fleet conveying an army from
Ifrı3qiya to Palermo under the command of al-H4asan ibn Alı3 al-Kalbı3
and his brother ‘Amma2r was wrecked by storm in 958 and following
the disaster al-Mu‘izz apparently agreed to a new truce with
Constantine VII which held until the Byzantine assault on Crete in
960. An Ifrı3qiyan fleet may have assaulted Naples at that time.130 The
Fa2t6imids and their now-independent Sicilian Kalbı3te amı3rs still posed
a threat to southern Italy but the worst had passed.
A third Rho2s attack on Constantinople in 941 was scattered by the
imperial fleet under the patrikios Theophane2s. In the following year
Emperor Ro2manos I Lekape2nos was able to respond to a request from
Hugh of Arles, the King of Italy, for assistance against the Muslims of
Fraxinetum by sending a squadron of chelandia which destroyed the
Muslim ships with Greek Fire. The Empire was slowly gathering
strength and by 944-5, according to al-Mas‘u2dı3, Rhodes had become a
Byzantine arsenal where warships were constructed which attacked
Egypt.131
In the East relations with the ‘Abba2sids and Ikhshı3dids had been
relatively calm during the first half of the century. However, Tarsos
still remained a threat that was not eliminated until the strate2gos of the
Kibyrrhaio2tai, Basil Hexamilite2s, won a famous victory over its fleet
off Lycia in 956, freeing the way for a new assault on Crete.132
Ro2manos II gave command to Nike2phoros Pho2kas, who successfully
completed the task between July 960 and March 961.133 Cretan
appeals for help to both the Fa2t6imid al-Mu‘izz and the Ikhshı3did
regent Abu2 ’l-Misk Ka2fu2r fell on deaf ears because the Fa2t6imid
proposals to Ka2fu2r for a combined operation were ignored since the
------------------------------
130
Chronicon Salernitanum, §161 (p. 168); Ex miraculis Sancti Agrippini; Ibn al-
Athı3r, Al-Ka2m il (Fagnan), pp. 353-6, 359; John Skylitze2s, Synopsis historio2n,
Nikhfovra" oJ Fwka'".8 (pp. 266-7); Kita2b al-‘uyu2n, A.H. 345 (p. 225); Theophane2s
continuatus, VI.Autokratoriva Kwnstantivnou.30-31 (pp. 453-5).
131
Al-Mas‘u2dı3, Muru2j, vol. 2, p. 423; Liudprand of Cremona, Antapodosis, V.9,
14-16 (pp. 134-9); Russian primary chronicle, p. 72; Theophane2s continuatus,
VI.Basileiva Rwmanou'.39 (pp. 423-6); John Zo2naras, Epitome2 historio2n, XVI.19 (vol.
4, p. 63).
132
Ibn al-Athı3r, Al-Ka2m il (Vasiliev/Canard), A.H. 345 (p. 162); Theophane2s
continuatus, VI.Autokratoriva Kwnstantivnou.29 (pp. 452-3).
133
Leo the Deacon, Historiae, I.gV-qV, II.ıV-hV (pp. 7-16, 24-29); John Skylitze2s ,
Synopsis historio 2n, Rwmano;" oJ Nevo ".4 (pp. 249-50); John Zo2naras, Epitome2 historio 2n,
XVI.23 (vol. 4, pp. 72-3); Pseudo Symeon magistros, Chronographia, pp. 758-60;
Theophane2s continuatus, VI.Basileiva Rwmanou' uJiou' Konstantivnou tou'
Porfurogenhvtou.7-12 (pp. 473-8). See also Farello, “Niceforo Foca”.
latter realized quite rightly that to cooperate would be to surrender the

political and moral advantage to the Fa2t6imids.
Nike2phoros Pho2k as pushed into Cilicia in 963. An attempt by
Ka2fu2r to reconstitute the Egyptian fleet to relieve Tarsos failed when
the squadron of 36 ships was mauled in a storm and the remnants
defeated off Cyprus in 965. Both Tarsos and Cyprus were recovered
for the Empire. Nike2phoros followed up these successes with a push
into Syria in 969 which returned to the Empire Antioch and all of
northern Syria west of Aleppo and north of Tripoli.134 His successor,
John I Tzimiske2s, continued the policy, capturing Beirut in 975 and
forcing Damascus to pay tribute, although his siege of Tripoli failed.
Under John’s successor Basil II the Empire reached its medieval
zenith. First, however, he had to face and defeat serious revolts by the
provincial aristocracy. In 976 one of John Tzimiske2s’ generals, Bardas
Skle2ros, revolted in Mesopotamia and marched on the capital. Initially
successful, he was defeated in 979 and took refuge amongst the
Muslims. He tried again in 987 but was taken captive by another rebel
general, Bardas Pho2kas, who was himself killed in battle on 13 April
989. Bardas Skle2ros continued the struggle but was eventually
reconciled to Basil II in October 989. Provincial fleets joined both of
these revolts but were scattered by the imperial fleet using Greek
fire.135
Basil II eventually became known to history as the
Boulgaroktonos, the “Bulgar slayer”. Svjatoslav of Kiev invaded
Bulgaria in 969 and deposed Tsar Boris II, intending to transfer his
own capital from Kiev to Preslav the Little, south of the Danube.
However, John Tzimiske2s’ army relieved Preslav and restored Boris,
forcing Svjatoslav to retire to Dorostolon. Then a Byzantine fleet
mounted the Danube and destroyed his ships, forcing him to
surrender. He was intercepted and killed by the Pechenegs at the
Dnepr rapids during his withdrawal. Bulgaria was annexed to the
Empire and Boris was taken to Constantinople.136 But a rebellion in
Bulgaria brought to the throne a new Tsar, Samuel, who trapped a
Byzantine army in the pass of Trajan’s Gates in 986 and inflicted a
------------------------------
134
John Skylitze2s, Synopsis historio2n, Nikhfovra" oJ Fwka'".11-13, 17 (pp. 268-70,
271-3).
135
John Skylitze2s, Synopsis historio2n, Basivl eio" kai; Konstanti'no".1-9, 16-18 (pp.
314-27, 334-8); John Zo2naras, Epitome2 historio2n, XVII.5-7 (vol. 4, pp. 107-110, 112-
17).
136
John Skylitze2s, Synopsis historio2n, Iwavnnh" oJ Tzimiskhv".5, 9-12, 14-18 (pp. 288,
295-302, 305-10); John Zo2naras, Epitome2 historio2n, XVII.2-3 (vol. 4, pp. 96-102);
Russian primary chronicle, pp. 87-90.
74 CHAPTER ONE
devastating defeat on it. Basil returned to the attack from 991 with
almost continuous campaigns designed to extinguish the Bulgarian
Khanate once and for all and at the battle of Kleidion in July 1014 he
finally triumphed. The sight of 14,000 Bulgarian captives sent home
blinded was said to have led to the Tsar’s death within two days.
Although some resistance continued, by 1018 Bulgaria had been
pacified and incorporated into the Empire.137
In the West the Empire was less successful. Following a Greek
uprising against the Muslims, Nike2phoros Pho2kas sent reinforcements
to Sicily and Taormina and Rametta were recovered temporarily in
963-5. However, an army and relief fleet sent out under the command
of a patrikios named Manuel were both defeated by the Fa2t6imids, the
fleet in the Straits of Messina, in 965,138 and soon after that a truce
was concluded because both had other concerns: the Fa2t6imids with a
long planned invasion of Egypt and the Byzantines with the Rho2s , the
Bulgarians, and the Western Emperor Otto I.
In February 962 Otto I came to Rome to be crowned. An interest in
southern Italy was aroused by the question of the Lombard
principalities of Capua-Benevento and Salerno, ruled at the time by
Pandolf I and Gisolf I respectively, over which both Western and
Eastern empires claimed suzereinty. He visited Benevento in 967 and
in 968 returned to both Benevento and Capua and then layed siege to
Byzantine Bari. Finding it impregnable he sent an embassy to
Constantinople headed by Bishop Liudprand of Cremona which
Nike2phoras Pho2k as dismissed contemptuously, giving rise to the
narrative of it by Liudprand, the most famous narrative of an embassy
in medieval history. The struggle was resumed in 969 but then
terminated by Nike2phoros’s assassination by John Tzimiske2s, who
proposed a marriage between his niece Theophano2 and Otto’s son
Otto (II). The wedding took place in St Peter’s on 14 April 972 and
Otto I withdrew from southern Italy.139
------------------------------
137
John Skylitze2s, Synopsis historio2n , Basivl eio" kai; Konstanti'no".12, 20, 23-7, 30,
35-6, 41, 43 (pp. 331, 339, 341-4, 346, 348-50, 357-9, 363-5); John Zo2naras, Epitome2
historio2n, XVII.6, 8, 9 (vol. 4, pp. 111-12, 118-19, 121-4). On Basil, see Stephenson,
Basil the Bulgar-slayer, esp. pp. 1-48.
138
Al-‘Ayni, ‘Iqd al-Juma2n; John Zo2naras, Epitome2 historio 2n, XVI.24 (vol. 4, pp.
78-9); Chronicle of Cambridge, pp. 46-7, 78; Al-Nuwayrı3, Niha2yat ‘al-Arab, (Amari),
pp. 130-34; idem, Niha2yat ‘al-Arab (Caussin), pp. 423-9; Ibn al-Athı3r, Al-Ka2m il
(Fagnan), pp. 362-6; Ibn Khaldu2n, ‘Ibar (De Slane), vol. 2, p. 545; idem, ‘Ibar (Des
Vergers), pp. 169-71; John Skylitze2s, Synopsis historio2n, Nikhfovr a" oJ Fwka'".3-4, 9
(pp. 261-2, 267); Leo the Deacon, Historiae, IV.zV-hV (pp. 64-8).
139
Chronicon Salernitanum, §§166-74 (pp. 170-7); Liudprand of Cremona, Relatio;
Romuald of Salerno, Chronicon, p. 167
From 965 south Italian waters were left to local Byzantine forces,
to those of the growing cities of Naples, Gaeta, and Amalfi, and of the
princes of Salerno, and above all to those of the Kalbı3te amı3rs of
Sicily descended from al-H4asan ibn ‘Alı3 al Kalbı3, now become
independent from the Fa2t6imids in practice although still their clients in
theory.
Even though there was no longer any hope of permanent Muslim
occupation of areas of southern Italy, Kalbı3te raids on the mainland
resumed from 975 and continued into the 980s. They contributed to
inducing Otto II to intervene. In 981 he marched into Apulia and
Calabria, provoking the amı3r, Abu2 ’l-Qa2sim, to cross the Straits. At
Punta di Stilo the armies met in a disastrous defeat for the German,
who, in a famous story narrated by Thietmar of Merseburg, escaped
only by swimming his horse out through the waves to take refuge on
one of two Byzantine chelandia which he had previously taken into
his service at Taranto as corsairs.140 Eventually he made his way back
north, his policies in ruins, leaving southern Italy to its own devices.
His son Otto III would trouble the Lombard principalities of Capua
and Benevento, but not the Byzantine and other territories further
south.141
After al-Mu‘izz moved from Ifrı3qiya to Egypt in 973 the Fa2t6imids
took no action against the Empire for some time. What naval forces
they had were initially occupied countering Syrian squadrons
supporting Qarmat6ı3 and other rebels in Syria and Palestine. Hostilities
with the Empire intensified only in the 990s with a series of
encounters, including a Byzantine raid on Alexandria in 993. Probably
as a response, in 996 the Caliph al-‘Azı3z began to construct a large
new fleet at Cairo. A fire which destroyed some of the ships provoked
suspicion of Christian merchants from Amalfi and a mob killed 100 of
them and looted local Christian churches. The fleet was reconstructed
and 24 ships were despatched to Tripoli but it was wrecked on the
Syrian coast. However, 20 ships could still be sent in 998 to assist in
putting down a rebellion in Tyre and they were capable of defeating a
Byzantine squadron assisting the rebels. In the following year a ten-
------------------------------
140
Chronicle of Cambridge, pp. 46, 48, 80, 82; Ibn al-Athı3r, Al-Ka 2mil (Fagnan), pp.
389-90; Ibn Khaldu2n, ‘Ibar (Des Vergers), pp. 172-5; John the Deacon, Cronaca
Veneziana, p. 145; Leo Marsicanus, Chronica, II.9 (p. 635); Romuald of Salerno,
Chronicon, pp. 168, 170-71; Thietmar of Merseburg, Chronicon, III.20-23 (pp. 122-
9).
141
Annales Beneventani, Anni 997-1001 (p. 177); Cronaca Capuana, pp. 133-4;
Leo Marsicanus, Chronica, II.10, 15, 24 (pp. 635-6, 638, 642-3); Ibn al-Athı3r, Al-
Ka2mil (Fagnan), pp. 379-80.
76 CHAPTER ONE
year truce was concluded.142

The impressive naval forces which ‘Abd al-Rah5ma2n III had created
in al-Andalus were maintained by his successors. When the Norse
returned to Lisbon with 48 “mara2kib” in 966, they were defeated off
Silves by the fleet of Seville. Nevertheless al-H4akam II ordered a fleet
of Norse style to be built at Cordoba in order to be able to deceive the
Vikings and thus to be able to close with them. Another assault in 971
was countered by bringing the Almeria squadrons around to Seville
and in the following year the fleet was sent to Ceuta and then to
Tangier, which it captured from the last Idrı3sids. Al-H4akam died in
976, succeeded by his young son Hisha2m II who was only eleven
years old, and power passed to the h5a2jib, or chief minister, Abu2 ‘Amı3r
Muh5ammad al-Mans5u2r, “the Victorious”, known to the Christians as
Almanzor. He assembled the fleet at Alcacer do Sal for his famous
campaign against Santiago de Compostela in 997 which brought the
bells of the cathedral to Cordoba. Having made the Caliphate the
terror of the Christian states to the north, he died on campaign in
1002. However, ironically, the very policies by which he had done so,
reliance on Slavic slave mama2lı3k and Berber and Christian
mercenaries, rather than the Muslim jund of the Arabic aristocracy of
al-Andalus, led to disintegration after his death. His son, ‘Abd al-
Malik, died in mysterious circumstances in 1008, after which the
Caliphate began to collapse.143
The fifth period, ca 1025-1204: the triumph of the Latin West
Liudprand of Cremona recorded that when he was in Constantinople

in 968 Nike2phoros Pho2kas had boasted that he alone had naval power
and that he would attack Otto I with his fleets, destroy his maritime
cities, and reduce to ashes those along the rivers.144 There would have
been much truth to the claim at the time. However, it appears that
from that point on, the very success of the Empire against its enemies
in the East, the pacification of Levantine, Aegean, and Black Sea
------------------------------
142
Yah5ya2 ibn Sa‘ı3d, Histoire, XXIII.3, pp. 447-9, 455.
143
Al-Marra2kushı3, Al-Mu‘jib, pp. 22-33; Ibn al-Athı3r, Al-Ka2m il (Fagnan), pp. 382-
5, 392-4, 400-401, 406-7 ; Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 2, pp. 394-5, 399,
405, 491-8.
144
Liudprand of Cremona, Relatio, §11 (p. 182): “Nec est in mari domino tuo
classium numerus. Navigantium fortitudo mihi soli inest, qui eum classibus aggrediar,
bello maritimas eius civitates demoliar et, quae fluminibus sunt vicina, redigam in
favillam.”.
waters, and the absence of any significant hostile naval forces, led to
neglect of its own naval forces. In the Escorial Taktikon, a list of titles
and offices compiled ca 971-5, the droungarios tou ploimou of the
imperial fleet in Constantinople appeared in the 133rd place after all
of the strate2goi of the themata and the strate2goi of the Kibyrrhaio2tai,
Samos, and Aigaion Pelagos appeared in 57th, 69th, and 70th places
respectively.145 This almost certainly reflected a decline in the
importance of naval forces by comparison to the preceding period.
Reflecting the lack of need for them, the Byzantine sources record no
naval expeditions except for the fleet sent by John Tzimiske2s to the
Danube against Svjatoslav of Kiev in 971 and that assembled by Basil
the parakoimo2menos for Basil II against the revolt of Bardas Skle2ros
in 989. Not until the last years of Basil II was a major new naval
expedition considered, this time an attempt to recover Sicily.146
This was precipitated by preceding events in south Italy. In 1009 an
Apulian noble named Melo had led a revolt against Byzantine rule
which lasted until suppressed by Basil Mesardonite2s, a new katepano 2
of Longobardia sent from Constantinople in 1010. Melo fled to the
Lombards. However, he returned in 1017 accompanied by Norman
mercenaries. Basil Mesardonite2s had died early in the same year but a
new katepano2, Basil Boio2anne22s, won a decisive victory near ancient
Cannae in October 1018, enabling the Byzantines to reconsolidate
their rule in northern Apulia and the Capitanata, and to exercize
overlordship over the Lombard principalities. It also suggested to
Basil II a possible reconquest of Sicily and a certain Oreste2s was sent
out with a fleet. The expedition failed and Oreste2s was eventually
replaced during the reign of Ro2manos III Argyros. Ro2manos had an
adventurous foreign policy and in 1033 sent an expedition to Egypt
which was a disastrous failure. However, the Empire’s last aggressive
foreign policy initiative in the Mediterranean for the century would
come in 1038 when the strate2gos autokrato2r George Maniake2s was
sent with a fleet to attempt a reconquest of Sicily. With the assistance
of a regiment of Varangians and Norman mercenaries he had initial
success and recovered the east of the island. However, accusations
were made against him at court and he was recalled in 1040 and the
------------------------------
145
Oikonomides, Listes de préséance, pp. 264-8.
John Skylitze2s, Synopsis historio2n, Iwavnnh" oJ Tzimiskhv".12 (pp. 300-301),
146
Basivleio" kai; Kwnstanti'no".8 (p. 324), ÔRwmanov" oJ Argurov".8 (pp. 383-4), 17 (p.
389); John Zo2naras, Epitome2 historio2n, XVII.9 (vol. 4, p. 124); Leo the Deacon,
Historiae, VIII.aV (p. 129), X.zV (pp. 169-70).
78 CHAPTER ONE
Table 5: Rulers of the fifth period, ca 1025-1204
(U) Spanish
(F) Fa2t6imids
The ‘Abba2sid Umayyads
(A) Ayyu2bids
Caliphs (S) Amı3rs of Sicily
(Z) Zangids
(B) Bu2y id amı3rs (Z) Zı3rids
(S) Salju2qids of
(S) Salju2qid sult6a 2ns (A) Almoravids
Ru2m (Al) Almohads
Constantine VIII Jala2l al-Dawla (B Hisha2m III al-

(1025-8) 1025-44) Mu‘tadd (U
1027-31)
End of dynasty
T4awa2’if kings
Ro2manos III (1028- Al-Qa2’im (1031-
34) 75)
Michael IV (1034- Al-Mustans5ir (F ‘Abd Alla2h ibn al-

41) 1036-94) Mu‘izz (S 1038-
40)
Al-H4asan al-S4am-
s5a2m al-Dawla (S
1040-44)
Civil war and
Norman invasion
Michael V (1041-2)
Zo2e & Theodo2ra
(1042)
Constantine IX ‘Ima2d al-Dı3n al-
(1042-55) Marzuba2n (B
1044-8)
Al-Malik al-Rah5ı3m
Khusraw Firu2z
(B 1048-55)
Salju2qid conquest
(Table 5 continued)
The Balkans, the The Western

Rho2s, Venice The Iberian rulers Italy Empire and France
(A) Aragon
(B) Bulgaria (B/C) Barcelona/
(B) Benevento (F) France
(C) Croatia Catalonia
(C) Capua Western Empire
(K) Kiev (L) Asturias/León
(S) Salerno (K) Kings
(S) Serbia147 (C) Castile)
(N) Normans (E) Emperors
(V) Venice (N) Navarre
(P Portugal)
Hugh Capet (F 987-

96)
Robert II (F 996-
1031)
Pietro Centranico Guaimar IV (S
(V 1026-32) 1027-52)
Stjepan I (C 1030- Vermudo III (L Landolf VI (B Henry I (F 1031-60)

58) 1028-37) 1038-77)
Domenico Pandolf VI (B
Flabanico (V 1054-74)
1032-43)
Yaroslav I (K 1036- Ramon Berenguer I Henry III (K 1039-
54) (B 1035-76) 56, E 1046-)
Stefan Vojislav (S Fernando I (L
1040-1043) 1035-65)
García III Sanchez
(N 1035-54)
Ramiro I (A 1035-
63)
Joint family rule (S Sancho IV Garcés William the Iron

1043 - ca 1046) (N 1054-76) Arm (N Ct
Domenico Contarini Apulia 1042-6)
(V 1043-71) Landolf VI (C
Michael (S 1046- 1047-58)
81/2) Gisulf II (S 1052-
Vyacheslav (K 77)
1054-7) Drogo (N Ct
Apulia 1046-
51)
Humphrey (N Ct
Apulia 1051-7)
------------------------------
147
Serbian rulers include those of Duklja and Ras°k a, some of whom overlapped.
80 CHAPTER ONE
(Table 5 continued)
(U) Spanish
(F) Fa2t6imids
(A) Ayyu2bids
(Z) Zangids
(S) Salju2qids of
Ru2m (Al) Almohads
Theodo2ra (1055-6) Rukn al-Dunya2 wa

’l Dı3n T4oghrïl I
(S 1055-63)
Michael VI (1056-
7)
Isaac I (1057-9)
Constantine X ‘Ad5ud al-Dawla Yu2suf ibn Ta2shufı3n

(1059-67) Alp Arslan (S (A 1061-1106)
1063-72) Tamı3m (Z 1062-
1108)
Michael VII (1067-

8)
Ro2manos IV (1068-
71)
Michael VII (1071- Al-Muqtafı3 (1075- Sulayma2n ibn
8) 94) Qutulmı£sh (S
Jala2l al-Dawla 1077-86)
Malik-Sha2h I (S
1072-92)
Nike2phoros III
(1078-81)
Alexios I (1081- Na2s5ir al-Dı3n Al-Musta‘lı3 bi ‘lla2h ‘Alı3 (A 1106-42)
1118) Mah5mu2d I (S (F 1094-1101) Yah5ya2 (Z 1108-
1092-4) Qı£lı£j Arslan I (S 1116)
Al-Mustaz5hir 1092-1107) ‘Alı3 (Z 1116-21)
(1094-1118) Al-A›mir (F. 1101-
Rukn al-Dı3n 1130)
Barkya2ru2q (S Malik-Sha2h (S
1094-1105) 1107-1116)
Malik Sha2h II ibn Rukn al-Dı3n
Barkya2ru2q (S Mas‘u22d I (S
1105) 1116-56)
(Table 5 continued)

(A) Aragon
(S) Serbia (C) Castile)
(P Portugal)
Pandolph IV (C
1049-57
Norman conquest
Henry IV (K 1056-
1106, E. 1084-)
Peter Kresimir IV Robert Guiscard
(C 1058-74) (D. Apulia
Igor (K 1057-60) 1057-85)
Svjatoslav II (K Sancho I Ramírez Philip I (F 1060-
1060-76) (A 1063-94) 1108)
Sancho II (C 1065-
72)
Alfonso VI (L
1065-1109)
Domenico Selvo Ramon Berenguer Roger I (N Great

(1071-84) II (B/C 1076- Count of Sicily
Demetrius Zvonimir 82) 1072-1101)
(C 1075-89/90) Berenguer Ramon
Izyaslav (K 1076-8) II (1076-97)
Vsevolod (K 1078-
93)
Constantine Bodin Pedro I (AN 1094- Roger Borsa (N D. Henry V (K 1106-
(S 1081/2-ca 1104) Apulia 1085- 25, E 1111-)
1101) Ramon Berenguer 1111) Louis VI (F 1108-
Vukan (S 1083/4-ca III (B/C 1097- William (N D. 37)
1122) 1131) Apulia 1111-
Vitale Falier (V Alfonso I (AN 27)
1084-96) 1104-34) Roger II (N Ct
Urraca (CL 1109- Sicily 1105, D.
Interregnum (C
26) Apulia 1128, K.
1090-93)
Sicily 1130-54)
Peter (C 1093-7)
Svjatopolk II (K
1093-1113)
82 CHAPTER ONE
(Table 5 continued)
(U) Spanish
(F) Fa2t6imids
(A) Ayyu2bids
(Z) Zangids
(S) Salju2qids of
Ru2m (Al) Almohads
Ghiya2th al-Dı3n
Muh5ammad I
(S 1105-18)
John II (1118-43) Al-Mustarshid Al-H4a2fiz5 (F 1131- Al-H4asan (Z 1121-

(1118-35) 49) 48)
Mughı3th al-Dı3n ‘Ima2d al-Dı3n Zangı3 Norman, then
Mah5mu2d II (S (Z 1127-46)
Almohad conquest
1118-31)
Ghiya2th al-Dı3n ‘Abd al-Mu’min (Al
Da2wu22d (S 1130-63)
1131-2) Ta2shufı3n (A 1143-
Rukn al-Dı3n 5)
Toghrı£l II (S
1132-4)
Ghiya2th al-Dı3n
Mas‘u2d (S
1134-52)
Al-Ra2shid (1135-6)
Al-Muqtafı3 (1136-
60)
Manuel I (1143-80) Al-Mustanjid Nu2r al-Dı3n (Z Ibra2hı3m (A 1146)

(1160-70) 1146-74) Ish5a2q (A 1146-7)
Al-Mustad5ı3’ (1170- Al-Z4a2fir (F 1149-
Almohad
80) 54)
conquest
Mu‘ı3n al-Dı3n Al-Fa2’iz (F 1154-
Malik-Sha2h III 60) Abu2 Ya‘qu2b Yu2suf
(S 1152-3) I (Al 1163-84)
(Table 5 continued)

(A) Aragon
(P Portugal)
Vitale Michiel I
1096-1102)
Michael &
Dobroslav (S ca
1101-2)
Ordelafo Falier (V
1102-18)
Koc°apar (S ca 1102-
3)
Vladimir (S ca
1103-8)
Oleg (K 1113-15)
Yaroslav II (K
1115-23)
Vladimir II (K Alfonso VII (CL Lothar I (K 1125-
1123-25) 1126-57) 37, E 1133-)
Juraj (S ca 1118) Afonso Henriques Conrad III (K 1138-
Grubes°a (S ca 1118- (P 1128-85) 52)
25) Ramon Berenguer Louis VII (F 1137-
Domenico Michiel IV (B/C 1131- 80)
(V 1118-30) 62)
Juraj (S ca 1125-7) Ramiro II (A 1134-
Mstislav-Harald (K 7)
1125-32) García IV Ramírez
Uros° I (S ca 1125- (N 1134-50)
45)
Gradinja (S ca
1127-46)
Pietro Polani (V
1130-48)
Yaropolk II (K
1132-9)
Yuri Dolgoruky (K
1139-57)
Uros° II (S ca 1145- Sancho VI (N William I (N 1154- Frederick I (K
62) 1150-94) 66) 1152-90, E
Radoslav (S ca Sancho III (C William II (N 1155-)
1146-?) 1157-8) 1166-89)
Domenico Morosini Fernando II (L
(V 1148-56) 1157-88)
Vitale Michiel II Alfonso VIII (C
(1156-72) 1158-1214)
84 CHAPTER ONE
(Table 5 continued)
(U) Spanish
(F) Fa2t6imids
(A) Ayyu2bids
(Z) Zangids
(S) Salju2qids of
Ru2m (Al) Almohads
Rukn al-Dı3n ‘Izz al-Dı3n Qı£lı3j

Muh5ammad II Arslan II (S
(S 1153-60) 1156-92)
Ghiya2th al-Dı3n Al-‘A›d5id (F 1160-
Sulayma2n Sha2h 71)
(S 1160-61) Ayyu2bid
Mu‘izz al-Dı3n Conquest
Arslan (S 1161-
76) Nu2r al-Dı3n Isma2‘ı3l
Rukn al-Dı3n (Z 1174-81)
T4oghrı£l III (S S4ala2h5 al-Dı3n (A
1176-94) 1169-93)
Extinction of
Salju2qid sult6a2n s by
Khwa2razm-Sha2hs
Alexios II (1180- Al-Na2s5ir (1180-
83) 1225)
Andronikos I (1183- Abu2-Yu2suf Ya‘qu2b
5) al-Mans5u2r (Al
1184-99)
Isaac II (1185-95) Al-Afd5al Nu2r al-
Dı3n ‘Alı3 (A#
1186-96)148
Ghiya2th al-Dı3n
Kay-Khusraw I
(S 1192-6)
Al-‘Azı3z ‘Ima2d al-
Dı3n (A* 1193-
8)149
Alexios III (1195- Al-‘A›dil I Sayf al- Muh5ammad al-
1203) Dı3n (A# 1196- Na2s5ir (Al 1199-
1218) 1214)
Rukn al-Dı3n
Sulayma2n II (S
1196-1204)
Al-Mans5u2r Na2s5ir
al-Dı3n (A*
1198-1200)
------------------------------
148
# = Line in Damascus
149
* = Line in Egypt
(Table 5 continued)

(A) Aragon
(P Portugal)
Desa (S ca 1162-6) Alfonso II (AB/C

Tihomir (S ca 1166- 1162-96)
7)
Joint rule (S ca
1168-71)
Stefan Nemanja (S
ca 1171-96)
Sebastiano Ziani (V
1172-8)
Orio Mastropiero
(V 1178-92)
Philip II (F 1180-
1223)
Peter & John Ase2n Alfonso IX (L Tancred of Lecce Henry VI ( K

(B 1186-96) 1188-1230) (N 1190-94) 1190-97, E
Enrico Dandolo (V Sancho VII (N William III (N 1191)
1192-1205) 1194-1234) 1194)
Hohenstaufen
conquest
Stefan the First- Pedro II (AB/C Philip of Swabia

Crowned (S 1196-1213) (K 1198-1208)
1196-1217) Otto IV (K 1198-
John & Kalojan 1218, E 1209-)
Ase2n (B 1196-7)
Kalojan Ase2n
(1197-1207)
86 CHAPTER ONE
(Table 5 continued)
(U) Spanish
(F) Fa2t6imids
(A) Ayyu2bids
(Z) Zangids
(S) Salju2qids of
Ru2m (Al) Almohads
Al-‘A›dil I Sayf al-

Dı3n (A*1200-
18)
Alexios IV (1203-4)
Alexios V (1204)
Latin Empire
of
Constantinople
last Byzantine attempt against Sicily collapsed after his departure. The
evidence for the extensive effort made for Maniake2s’s expedition to
Sicily does not accord with that of Michael Psellos for the degraded
state of Byzantine naval forces when the capital was attacked by the
Rho2s again in 1043. According to Psellos, to face the Rho2s attack only
a few derelict vessels could be found to be armed with Greek Fire to
oppose them. However, the success of the Byzantine fleet suggests
that Psellos exaggerated its weakness for literary purposes. According
to Kekaumenos, probably writing ca 1075-8, the Byantine fleet was
still the “glory of Romania” in his own day.150 And even in the late
1070s there were still some detachments around Thrace and in Asia
Minor who assisted in putting down the revolt of Nike2phoros
Bryennios and who welcomed Nike2phoros Botaneiate2s.151 That being
------------------------------
150
Chronica monasterii Casinensis, II.37 (pp. 236-40), II.66 (pp. 298-9); John
Skylitze2s, Synopsis historio2n, Basivleio" kai; Kwnstanti'no".34 (p. 348), 47 (p. 368),
ÔRwmanov" oJ Argurov".16-17 (pp. 388-9), Micahvl oJ Paflagwvn.9-20 (pp. 398-407); John
Zo2naras, Epitome2 historio2n , XVII.15, 22, 24 (vol. 4, pp. 139-42, 160-62, 167-9);
Kekaumenos, Strate2gikon (Wassiliewsky), Nouqethtiko;" pro;" Basileva, §22: “To;n
stovlon ajgwnivzou pavntote ajkmavzein kai; e[cein aujto;n ajnelliph': oJ ga;r stovlo" ejsti;n hj
dovxa th'" ÔRwmaniva".”; Leo Marsicanus, Chronica, II.37 (pp. 651-3), II.66 (pp. 675-6);
Lupus Protospatharios, Annales, Anni 1009-1019 (pp. 56-7); Michael Psellos,
Chronographia, Constantin IX, §76 (vol. 2, p. 1), §§90-95 (vol. 2, pp. 8-12); Romuald
of Salerno, Chronicon, pp. 174-5.
151
John Skylitze2s continuatus, pp. 175, 178; Michael Attaleiate2s, Historia, pp. 254,
268-9, 269-72; Nike2phoros Bryennios, Hyle historias, III.22-3 (pp. 249-51).
(Table 5 continued)

(A) Aragon
(P Portugal)
No rulers began their rules in this period
said, there can be no doubt that by the accession of Alexios I

Komne2nos, the Byzantine navy had virtually disappeared and
certainly could not project its forces far afield. Faced by the attack on
the Balkans in 1081 by the Normans under Robert Guiscard, Alexios
had no choice but to offer a new imperial chrysobull to Venice,
granting extensive economic privileges within the Empire to her
merchants in return for the assistance of her fleet against the
Normans.152
To some degree at least, the decline of Byzantine naval forces over
the century can be explained by a lack of enemies to be concerned
about. The Bulgarians had been eliminated and the Rho2s of Kiev had
been largely pacified by an alliance made with Vladimir I in 987 and
his marriage to Basil II’s sister Anna and conversion to Christianity in
the following year. The only outbreak of hostilities with the Rho2s was
the attack on Constantinople in 1043 by Vladimir’s son Jaroslav I,
probably as a result of disputes over trading rights. The defeat of his
forces led to a peace treaty in 1046 sealed by the marriage of
Constantine IX’s daughter to Jaroslav’s son Vsevolod.
In the South it appears that the Fa2t6imid navy went into decline from
the end of the tenth century, at least if the complete lack of mention in
the sources of any activity by it is a reliable index. During the reigns
of al-H4ak2 im and al-Z4a2hir, relations with Byzantium were relatively
peaceful, except for some clashes by land in Syria, and there were
long periods of truce. Then in 1062 civil war broke out between
Turkish and black ‘abı3d troops in Egypt which was ended only by the
appointment in 1073 of Badr al-Jama2lı3, the governor of Acre, as amı3r
al-juyu2sh, commander of the armies. Badr died in 1094 and was
------------------------------
152
Anna Komne2n e2, Alexiade, IV.i-iii, VI.v (vol. 1, pp. 145-50; vol. 2, pp. 50-55).
88 CHAPTER ONE
succeeded by his son al-Afd4al. During all this period Muslim sources
make no mention of any operations of Fa2t6imid fleets. Who the Muslim
corsairs who attacked Myra and pillaged the Cyclades in 1035, only to
be defeated by the fleet of the Kibyrrhaio2tai in the Cyclades in the
following year, were is unknown.153 After that Muslim incursions into
the Aegean ceased entirely. As a result the three great maritime
themata of the Kibyrrhaio2tai, Samos, and Aigaion Pelagos declined.
The fleet of the Kibyrrhaio2tai and its strate2gos are last mentioned
during the reign of Constantine IX. During the eleventh century small
squadrons based locally at Kefalle2nia, Abydos, Samos, Chios,
Naupaktos, and other places in support of terrestrial forces and against
corsairs became more important.
In al-Andalus the death of al-Mans5u2r’s son, ‘Abd al-Malik, in 1008
led to a series of short-lived Caliphates interspersed with rules by
members of the H4ammu2did family of Malaga. The Caliphate finally
collapsed in 1031 and was succeeded by local dynasties ruling in
various regions and cities and known as the mulu2k al-t6awa2’if, the taifa
or “Party” kings because the various rulers were descended from
either Arabs, or Berbers, or Slavic mama2lı3k. Muslim al-Andalus began
a slide into military impotence which would lead to domination of it
by the Christian rulers of the North and to progressive loss of territory
to them. This would culminate in 1085 with the fall of the Dhu2 ’l-
Nu2nid taifa mamlaka of Toledo to Alfonso VI of León/Castile. Many
of the taifa mulu2k were compelled to pay protection money, paria, to
the northern Christian monarchs and to freebooters such as Rodrigo
Díaz de Vivar, el Cid, to prevent being attacked by them.154
There was, however, one exception to the nature of the taifa
mama2lik. This was the mamlaka of Denia and the Balearics founded
by the renowned Muja2hid al-Muwaffaq, a Slavic mamlu2k of al-
Mans5u2r who became governor of Denia and then independent ruler
from ca 1009. In 1015-16 he attempted to conquer Sardinia but was
ejected by the combined fleets of Pisa and Genoa in 1017.155 At some
time after that he annexed the Balearic islands, which he ruled until
------------------------------
153
John Skylitzes, Synopsis historio 2n, Micahvl oJ Paflagwvn.6-8 (pp. 396-8); John
Zo2naras, Epitome2 historio2n, XVII.10, 14 (vol. 4, pp. 127, 139); Sa2wı3ris, History of the
Patriarchs (Khater & Burmester), vol. 2, part 3, pp. 314-16, 388-9.
154
Al-Marra2kushı3, Al-Mu‘jib, pp. 33-111; Ibn al-Athı3r, Al-Ka2mil (Fagnan), pp. 408-
12, 420-46, 480-81;
155
Bernardo Maragone, Annales Pisani, MXVI-MXVII (pp. 4-5); Ibn al-Athı3r, Al-
Ka2mil (Amari), Anno 406 (vol. 1, pp. 358-9); idem, Al-Ka2m il (Fagnan), p. 444; Ibn
Khaldu2n, Muqaddimah, III.32, VI.10 (vol. 2, pp. 41, 441). See also Codera,
“Mochéhid”.
his death in 1044/5, to be succeeded there by his son ‘Alı3 ibn Muja2hid
until 1076, when both Denia and the Balearics were occupied by al-
Muqtadir ibn Hu2d, the malik of Zaragoza. They were occupied by the
Almoravids in 1115 after the governor called for their help when
under attack by an invasion force of Genoa, Pisa and Barcelona.
During all this period the Balearics became a hub for Muslim corsairs
operating throughout the western Mediterranean.
Around 1050 the Byzantine Empire reached its greatest extent for
centuries and at sea faced no enemies. Armenia was finally absorbed
completely in 1045 and the Empire then stretched from Lake Urmia to
Syria, the Danube, and the Adriatic. Its maritime commerce
flourished. But as a consequence the maintenance of military and
naval forces was allowed to fall away. Services, both military and
those supporting the military, were commuted for cash payments,
which might or might not be spent on the military. In his advice to an
emperor, Kekaumenos reflected the growing malaise:
Strive to have commanders of the fleet who are above all giving and
receiving, for if the commanders of the fleet are greedy and accept gifts,
listen to what they will do. In particular, they allow the forces to be
excused, accepting from them money, not the amount they normally give
for this service of the fleet but a double amount, and the chelandion
becomes defective.156
This merely reflected corruption at an individual level. Much more

serious was the commutation of services for cash on a system-wide
basis and the consequent neglect of military forces. The army of
Ibe2ria/Mesopotamia was dissolved by Constantine IX around 1050,
leaving the eastern frontiers exposed. Constantine also initiated a
series of changes to military structures which led to progressive
debilitation of the traditional thema armies. The Empire became
increasingly reliant on foreign mercenaries such as Normans, Turks,
and Pechenegs. With the exception of Isaac I Komne2nos and Ro2manos
IV Diogene2s, the emperors of the period neglected the armies,
preferring to spend their revenues on other things, and the state to
which the armies declined is reflected in the descriptions of the rag-
------------------------------
156
Kekaumenos, Strate2gikon (Wassiliewsky), Nouqethtiko;" pro;" Basileva, §22 (p.
102): “ajgwnivzou de; e[cein kai; tou' stovlou a[rconta" ajnwtevrou" panto;" dwvrou kai;
lhvmmato", ei; gavr ei;sin leivxouroi kai; dwrolh'ptai oiv tou' stovlou a[r conte", a[kouson tiv
poiou'sin. ejn prwvtoi" me;n strateiva" ejw'sin e;xkouseuvesqai lambavnonte" ejx aujt w'n
nomivsmata oujc o{sa h[qelon dou'nai eij" th;n ejphvreian tou' stovl ou, ajll ejn diplh'/
posovthti, kai; givnetai celavndion ejllipev".”.
90 CHAPTER ONE
tag battalions which Ro2manos IV assembled to face the Turkish

menace in the East in 1068-71.157
The nature of Byzantine naval forces changed. The droungarios
tou ploimou in Constantinople in charge of a sekreton became
virtually the admiral in chief. The strate2goi of the provincial themata
in general lost importance and in the case of the naval themata, their
fleets disappeared. What naval forces remained were either sent out
from the centre by the droungarios tou ploimou or became dependent
upon provincial military commanders such as doukade2s and
katepano2.158
In the central Mediterranean the Zı3rı3ds were the successors to the
Aghlabids at sea. According to Ibn al-Athı3r, who is however
unsupported and there must be some doubt about it, in 1025 Sharaf al-
Dawla al-Mu‘izz sent a huge fleet of 400 ships to sea in response to
the Byzantine tentative against Sicily but it was destroyed by storm
near Pantelleria. The activities of Zı3rı3d corsairs provoked the Pisans to
attack Bona in 1034. Then in 1047 Sharaf al-Dawla sent his fleet
raiding into the Ionian.159
In Sicily in 1035 a rebel against the rule of the Kalbı3te amı3r Ah5mad
al-Akhal by the name of Abu2-H4afs called in the Zı3rı3ds and Sharaf al-
Dawla sent his son ‘Abd Alla2h ibn al-Mu‘izz with an expeditionary
force. Al-Akhal was defeated and killed in 1038 but ‘Abd-Alla2h was
in turn defeated and expelled by another Kalbı3te amı3r, al-H4asan al-
2 al-Dawla, in 1040. However, H4asan was deposed in 1044 and
S4ams5am
in any case ruled over part of the island only, which from the 1040s
came to be ruled by various amı3rs in various cities, of whom ‘Alı3 ibn
Ni‘ma ibn al-H4awwa2s in the region of Enna was the most powerful.
Then between 1053 and 1060 another amı3r by the name of
Muh5ammad ibn Ibra2hı3m ibn al-Thumna rose to power in Syracuse.
When Ibn al-Thumna lost support after unsuccessfully besieging Enna
and then turned to Roger of Hauteville for aid, the Muslims
themselves had sown the seeds of their own destruction.160
The advent of the Normans in southern Italy is a confused tale of
------------------------------
157
John Skylitze2s continuatus, pp. 124-5; John Zo2naras, Epitome2 historio2n,
XVIII.10, 11 (vol. 4, pp. 202-3, 207); Michael Attaleiate2s, Historia, p. 103.
158
See Ahrweiler, Byzance et la mer, pp. 155-71; Guilland, “Drongaire”, pp. 539-
42; Malamut, “Les insulaires”.
159
Ibn al-Athı3r, Al-Ka2mil (Fagnan), pp. 450-51, 455. See also Idris, Berberie
orientale, pp. 167-71.
160
Al-Nuwayrı3, Niha2yat ‘al-Arab (Amari), pp. 141-4; idem, Niha2yat ‘al-Arab
(Caussin), pp. 434-9; Amatus of Monte Cassino, L’ystoire de li Normant, V.viii (pp.
147-8); Geoffrey Malaterra, De rebus gestis, II.iii (p. 30); Ibn al-Athı3r, Al-Ka2mil
(Fagnan), pp. 498-504; Ibn Khaldu2n, ‘Ibar (Amari), pp. 200-203.
bloodshed and mayhem.161 Most probably the first of them to attract

attention were pilgrims returning from Jerusalem and either they or
others became mercenaries in the employ of the prince of Salerno and
they, and perhaps yet others, joined the second rebellion of Melo in
1017. After Cannae the survivors took service with the Lombards.
Byzantine rule in Apulia and the Capitanata remained secure until the
1040s but meanwhile Norman mercenaries found employment on all
sides in the internecine strife of the period and in 1030 Duke Sergius
IV of Naples made one of their leaders, Rainulf, count of Aversa, thus
giving them their first landed power base. His countship was
confirmed by the Western Emperor Conrad II when he visited the
South in 1038, at which time he also deprived Pandolf IV of his
principality of Capua and invested Guaimar IV of Salerno with it.
In 1041 a Milanese mercenary by the name of Arduin who had
once been in Byzantine service invaded Apulia with the aid of the
Normans of Aversa. Arduin soon disappeared, to be replaced by
Lombard allies and by Argyrus, the son of Melo, but by 1042 the
Normans and their allies controlled all of Apulia apart from Trani,
Taranto, and Bari and the peninsula south. In these circumstances,
George Maniake2s was restored to favour and sent to Italy. He
managed to restore the situation to some extent, in the process
inducing Argyrus to abandon his Norman friends, but when his enemy
Constantine IX came to the throne he attempted to replace Maniake2s
in Italy, inducing him to revolt and march on Constantinople, during
the course of which he lost his life. However, Argyrus’s defection
caused the Normans in Apulia to choose their own leader for the first
time. They chose William “the Iron Arm”, William of Hauteville, the
son of a poor knight of Normandy. In 1042 William and the other
Norman leaders carved up the lands they had acquired, William taking
Ascoli and his brother Drogo Venosa. The rise to power of the
Hautevilles had begun. William received the title of Count from
Guaimar IV of Salerno and married one of his nieces. The sucession
was disputed but after William’s death in 1045, Drogo did eventually
succeed him. In the next year or so there arrived from Normandy
Drogo’s younger half-brother Robert, who Drogo knew well as a
potential rival and did not welcome. He shunted him off to Calabria to
make a living as a freebooting brigand and it was at this time that he
acquired his sobriquet Guiscard, “the wily” because he lived off his
wits very successfully.
------------------------------
161
See also Kreutz, Before the Normans; Loud, Robert Guiscard.
92 CHAPTER ONE
In the winter of 1046-7 the Western Emperor Henry III came to

Italy, primarily to resolve the problem of three competing claimants to
the Papacy. However, he also took the opportunity to visit the South,
depriving Guaimar IV of Capua and restoring Pandolf IV, but also
confirming Rainulf II as Count of Aversa and Drogo as Count in his
own lands. By the early 1050s the Normans were rapidly becoming
masters of the South but their depredations had alienated all sections
of the native populace and in 1051 Drogo was murdered. Pope Leo
IX, by now ruling Benevento on behalf of the Western Empire, and
Argyrus, recently sent back to Italy as katepano2 from Constantinople,
began negotiations for an anti-Norman coalition; however, a third
Hauteville brother, Humphrey, who had succeeded Drogo as Count,
struck against Argyrus and defeated him in 1052. In the following
year, when the Pope in person led a substantial force into the South
intending to unite forces with Argyrus in Apulia, the Normans
intercepted him at Civitate, south of the Fortore river in the
Capitanata, and won an overwhelming victory, the Pope being
surrendered to them by the men of Civitate. He died in the following
year, followed by Constantine IX in 1055 and Henry III in 1056,
leaving the Normans free to do as they wished. On Humphrey’s death
in 1057 Robert Guiscard succeeded him. By 1060 all of Calabria had
fallen to the Normans and the last Byzantine garrisons were
evacuated. Capua had been taken in 1058, although Salerno would
survive until 1076. At Melfi, in August 1059, Robert Guiscard was
invested by Pope Nicholas II as Duke of Apulia and Calabria. All that
remained was to mop up the last Byzantine possessions in Apulia.
However, that would prove to be a very protracted process,
exacerbated by the diversion of Norman energies to Sicily from 1061.
Not until August 1068 was Robert Guiscard ready to commence the
siege of Bari. The city’s land walls could not be stormed and the
Normans were unable to prevent supplies being brought in by sea, one
squadron succeeding in breaking into the city in 1069. But early in
1071 a second was defeated by Robert’s younger brother Roger who
had brought a squadron around from Sicily. The city was compelled to
surrender on 16 April.162
The Normans launched a preliminary raid against Sicily in 1061
which succeeded in capturing Messina, delivering mastery of the
Straits of Messina to them and allowing their forces to be reinforced
------------------------------
162
Amatus of Monte Cassino, L’ystoire de li Normant, V.xxvii (pp. 159-64);
Geoffrey Malaterra, De rebus gestis, II.40, 43 (pp. 48-9, 50-51); Romuald of Salerno,
Chronicon, pp. 187-8; William of Apulia, Gesta, III, ll 111-46 (p. 268).
and supplied easily. Muslim naval forces from Palermo were forced to
withdraw. The Normans then moved inland against Enna and defeated
the army of Ibn al-H4awwa2s. However, the conquest bogged down for
many years as a result of a number of factors: inadequate Norman
numbers, disputes between Roger and Robert Guiscard and the latter’s
need to complete the conquest of Apulia, stiffening Muslim resistance
and difficulties in taking some heavily fortified positions, worsening
relations with the Greek population of the island, and the arrival of
Zı3rı3d reinforcements under the sons of the new amı3r Tamı3m, Ayyu22b
and ‘Alı33. Although they were defeated at Cerami in 1063 and ‘Alı3 was
killed, Ayyu22b remained on the island until 1069.163
In 1064 Roger, with the assistance of Robert Guiscard, besieged
Palermo for four months but the Normans had neither the numbers nor
the naval forces necessary for such an assault. However, after the
capture of Bari the circumstances became very different. In the
summer of 1071 Robert and Roger marshalled their forces and moved
against Palermo. For the first time the Normans had significant naval
forces and these engaged a Zı3rı3d relief force, drove them into the
harbour, and broke through its chain and set the surviving Muslim
ships on fire. The outer walls were penetrated and the city surrendered
on 10 January 1072.164
Nevertheless, it would be another 20 years before the last Muslim
fortress fell. Of immediate concern was Enna, which was too
powerfully fortified to be besieged. So an offensive fortress was built
on Monte Calascibetta, two kilometres to the north, to contain the
Muslims. A Zı3rı3d assault on Mazara in 1075 was only beaten off with
great difficulty, but successful assaults on Trapani in 1077,
Castronuovo in 1078, and Taormina in 1079, began to turn the tide
decisively. In August 1086 Syracuse fell, a Pisan, Genoese, and
Amalfitan expedition in 1087 against al-Mahdiyya neutralised Tamı3m,
and in 1087 Enna’s amı3r finally gave up the struggle. The last Muslim
fortress, Noto, surrendered in 1091.165
------------------------------
163
(Caussin), pp. 439-40; Amatus of Monte Cassino, L’ystoire de li Normant, V.ix-x (pp.
148-50); Geoffrey Malaterra, De rebus gestis, II.v-vi, viii-xii, xvii, xxxiii (pp. 31-4,
42-5); Ibn al-Athı3r, Al-Ka2m il (Fagnan), pp. 502-3. See also Idris, Berbérie orientale,
pp. 283-9; Loud, Age of Robert Guiscard, pp. 146-85; Waley, “Combined
operations”.
164
Amatus of Monte Cassino, L’ystoire de li Normant, V.xxvi (p. 159), VI.xiv, xvi
(pp. 178-80); Geoffrey Malaterra, De rebus gestis, II.xxxvi (pp. 46-7), II.xlv (pp. 52-
3); Lupus Protospatharios, Annales, 1072 (p. 60); William of Apulia, Gesta, III, ll.
225-339 (pp. 270-72).
165
Al-Nuwayrı3, Niha2yat ‘al-Arab (Caussin), p. 440; Geoffrey Malaterra, De rebus
94 CHAPTER ONE
Although the Empire had made some efforts to relieve Bari in

1068-71, there can be little doubt that in Constantinople the perils of
its last remaining posession in Italy would have appeared insignificant
compared to those posed by the irruption of the Turks into its eastern
provinces. In 1038 a Salju2qid chieftain of the Oghuz Turks, Rukn al-
Dunya2 wa ’l Dı3n T4oghrïl I, had proclaimed himself sult6a2n at Nishapur.
In 1055 he entered Baghdad, overthrew the Bu2yids, and was
recognized by the ‘Abba2sid Caliph al-Qa2’im. In 1063-4 a Salju2qid
chieftain by the name of Qutalmïsh rebelled unsuccessfully and was
killed; however, his son Sulayma2n ibn Qutalmïsh moved north into
the mountains south of the Caspian Sea and then into Anatolia and
Byzantine territory. The activities of the tribes under Sulayma2n led to
Ro2manos IV Diogene2s assembling the Byzantine armies and moving
against them. This provoked the new sult6a2n ‘Ad5ud al-Dawla Alp-
Arslan to move north to their defence and the two armies met at
Mantzikert in August 1071. The ensuing two-day battle was a
disastrous defeat for the Empire. Although treated with respect by
Alp-Arslan and released, Ro2manos was overthrown in Constantinople
by supporters of Michael VII, defeated in the civil war that ensued,
and blinded.166 In the chaos that descended on Anatolia, the Turkish
bands under Sulayma2n ibn Qutalmïsh moved west and by 1078 he had
captured Nicaea and made it his capital. Alexios I Komne2nos
recognized his boundaries in 1081. The Turks of Nicaea would figure
prominently in the circumstances leading to the First Crusade.
The rise of the Salju2qids was parallelled in the West by that of the
Almoravids. Yah5y a2 ibn Ibra2hı3m, the leader of a branch of the S4anha2ja
Berbers in the western Sahara, made the H4ajj in the early eleventh
century and on his way home engaged in Ifrı3qiya a famous preacher
called ‘Abd Alla2h ibn Ya2sı3n. ‘Abd Alla2h’s preaching was so austere
that Yah5ya2 ibn Ibra2hı3m’s own tribesmen drove him out but he found
refuge in a riba2t6 surrounded by water, probably on an island in either
the Niger or Senegal rivers, or perhaps on some promontory on the
Atlantic coast. His movement became known as the al-Mura2bit6u2n,
Hispanicized as Almoravids. After his death, the Almoravid Amı3r al-
Muslimı3n , Yu2suf ibn Tashufı3n, overran Morocco and founded
------------------------------
gestis, III.vii-ix, xi-xii, xvii-xviii (pp. 60-61, 62-4, 66-7), IV.i-ii, v-vi, xii-xv (pp. 85-
8, 92-3); Ibn al-Athı3r, Al-Ka2mil (Fagnan), p. 503.
166
Nike2phoros Bryennios, Hyle historias, I.13-17 (pp. 104-119); John Skylitze2s
continuatus, pp. 144-55; John Zo2naras, Epitome2 historio 2n, XVIII.14-15 (vol. 4, pp.
213-19); Michael Attaleiate2s, Historia, pp. 142-79; Sa2wı3ris, History of the Patriarchs
(Khater & Burmester), vol. 2, part 3, pp. 308-11.
Marrakesh as his capital in 1062. In 1086 he was invited by the taifa

mulu2k to come to their assistance and won a great victory over
Alfonso VI of León/Castile at Sagrajas. He retired back to Morocco
but, despairing of the squabbling taifa mama2lik, returned in 1090 and
by 1095 had suppressed all of them, with the exception of Zaragoza.167
The Almoravids were fundamentalist Berber warriors who initally
had little but contempt for the softness of the civilization of Moorish
Spain. They considered payment of paria to Christians as a breach of
the sharı3‘a, which forbad Muslims from being subject to non-
Muslims. They also denounced the imposition of non-Qur’a2nic taxes,
muku2s , by the taifa mulu2k, and the exercise of authority by Jews over
Muslims, as had occurred in the taifa mamlaka of Granada, as another
breach of the sharı3‘a. They were aided by the support of jurists of the
Malı3kı3 school and the common populace, who were wearied of the
corruption, luxury, and inability to resist the Christians of the taifa
ruling classes. Increased restrictions were imposed on Christians and
Jews and conditions for Christians became so bad that in 1125-6 large
numbers returned to the North to settle there with Alfonso I of Aragon
when he led a large raid into Muslim territory. Although the
Almoravid takeover stabilized the frontiers for some years, they were
unable to retake Toledo and Zaragoza was taken by Alfonso I of
Aragon in 1118.
According to Ibn Khaldu2n, in their heyday the Almoravids had
fleets totalling 100 ships and their hereditary admirals were the banu 2
Maymu2n, a family from Cadiz. Muh5ammad II al-Mu‘tamid, the last
‘Abba2did malik of Seville, had handed over his fleet to Yu2suf ibn
Tashufı3n in 1083 to blockade Ceuta and Yu2suf had built more ships of
his own. Although there is little record of the activities of the
Almoravid fleets, in 1122 ‘Ali ibn Yu2suf did send a fleet to Calabria
commanded by Abu2 ‘Abd Alla2h Muh5ammad ibn Maymu2n and he
captured Nicotera. In 1127 the Almoravid admiral sacked Patti and
Syracuse. The strength of the Almoravid fleets may have been
underestimated because they appear never to have engaged Christian
naval forces in open battle. Traditionally, the twelfth century has been
considered to mark the begining of the naval domination of the
Mediterranean by the Latin West and it is true that the naval forces of
Genoa and Pisa were beginning to flex their muscles in the Almoravid
------------------------------
Al-Bakrı3, Kita2b al-mughrib, pp. 311-21; Al-Marra2kushı3, Al-Mu‘jib, pp. 110-16,
167
136-8; Ibn abı3 Zar‘, Rawd5 al-Qirt6a2s, pp. 162-223; Ibn al-Athı3r, Al-Ka2mil (Fagnan),
pp. 462-8, 482-6, 491-8; Ibn Khaldu2n, ‘Ibar (De Slane), vol. 2, pp. 67-80. See also
Lagardère, Almoravides, pp. 45-141.
96 CHAPTER ONE
period. However, when the two cities allied with Count Ramon
Berenguer III of Barcelona to attack the Balearics in 1113-4 in
revenge for corsair raids and attacks on Christian coasts, they could
not hold onto them against ‘Ali ibn Yu2suf’s counter-attack in 1115.
According to a letter of 1116 appointing an unnamed person as
governor of the Balearics, the Almoravid invasion fleet had consisted
of 300 shawa2nı3 galleys under the command of a qa2’id by the name of
Ibn Ta2fratust. The Almoravid appointed Muh4ammad ibn ‘Alı3 ibn
Yu2suf al-Masu2fi ibn Gha22niya governor of the Balearics in 1126 and
his family would continue to rule them for generations, even after the
overthrow of the Almoravids by the Almohads.168
In time the Almoravids were captured by the culture and luxury of
Moorish Spain and their whole state began to lose cohesion. By the
mid twelfth century Portugal had gained Lisbon, Castile had pushed
south across the Guadiana river to Calatrava, and Aragon had secured
the frontier of the Ebro river. Only in this late period of their decline
do the Almoravids appear to have begun to cede naval supremacy in
the western Mediterranean. In 1136 a Genoese squadron raided the
H4amma2did capital of Bija2ya and in 1146 the Genoese consul Caffaro
assaulted Minorca. In the following year Count Ramon Berenguer IV
of Barcelona seized Almeria with Pisan and Genoese assistance, and
the following year Tortosa, although the Almohads recovered Almeria
in 1157. From the 1130s the Almoravids appear to have sought more
peaceful relations with some Christian powers. In 1136 two galeae
commanded by Abu2 ‘Abd Alla2h Muh5ammad ibn Maymu2n arrived in
Pisa to conclude a ten-year treaty of peace between the city and the
Almoravids and the amı3r of Tlemcen.169
The Almohads stemmed from a movement for religious reform
known as al-Muwah5h5idu2n, “the unitarians”, founded by Muh5ammad
ibn2 ‘Abd Alla2h ibn Tu2mart, a Mas5mu2da Berber from the Atlas
mountains. They brought fervour, piety, and reform of religious mores
to the Muslim world in the West, waging jiha2d against their
Almoravid and other opponents. After his death in 1130, his follower,
------------------------------
168
Al-Maqqarı3, Nafh5 al-t5ı3b, VII.iv (vol. 2, pp. 257-8); Al-Marra2kushı3, Al-Mu‘jib,
pp. 230-33; Bernardo Maragone, Annales Pisani, A.D. MCXIIII (p. 8); Gesta
triumphalia per Pisanos facta, pp. 90-94; Ibn abı3 Zar‘, Rawd5 al-Qirt6a2s, p. 204; Ibn al-
Athı3r, Al-Ka2mil (Fagnan), pp. 513-15; Ibn Khaldu2n, ‘Ibar (De Slane), vol. 2, pp. 77,
87-9; idem, Muqqadimah, vol. 2, p. 43. See also Bel, Benou Gha2nya; Lagardère,
Almoravides, pp. 205-7; idem, Djihad Andalou, pp. 39-41, 268-70.
169
Al-Marra2kushı3, Al-Mu‘jib, pp. 146-54, 179-82; Bernardo Maragone, Annales
Pisani, A.D. MCXXXIIII (p. 9); Caffaro, Annales Januenses, pp. 28, 33-5; Ibn abı3
Zar‘, Rawd5 al-Qirt6a2s, pp. 224-42; Ibn al-Athı3r, Al-Ka 2mil (Fagnan), pp. 562, 567, 582-
4; Ibn Khaldu2n, ‘Ibar (De Slane), vol. 2, pp. 80-86.
the Caliph ‘Abd al-Mu’min, continued his work bringing all of the
Maghrib under Almohad control by the time that he annexed the last
remaining Norman possessions in Ifrı3qiya in 1160. Even earlier, in
1146, he had been compelled to intervene in al-Andalus when the
Almoravid amı2rate began to break up. By 1148 he controlled the
south-west but the south-east remained in the hands of the malik of
Valencia, Abu2 ‘Abd Alla2h Muh5ammad ibn Sa‘d ibn Mardanı3sh,
known as the “Wolf king” to Christians. The Almohads closed in on
him and after he died in 1172 his family submitted to Abu2 Ya’qu2b
Yu2suf. In time the Almohads themselves would go into decline but in
their heyday, before the distastrous defeat of Muh5ammad al-Na2s5ir by a
Christian coalition at Las Navas de Tolosa in 1212, they brought
renewed vigour to al-Andalus. In particular, their crushing victory
over Alfonso VIII of León/Castile at Alarcos in 1195 turned back the
progress of the Christian Reconquista for over a decade.170
Even more so than the Almoravids, the Almohads made a
determined effort to achieve naval strength in the western
Mediterranean and especially to control the 400 kilometre channel
through the Straits of Gibraltar from Cadiz to Almeria. They were
never able to prevent ingress through the Straits of large Northern
Crusader fleets because the contemporary technology of naval warfare
and the maritime geography and meteorology of the channel made
that impossible; however, S4ala2h5 al-Dı3n considered that they could
have attempted to do so at least. In 1189 he sent an envoy to the
Almohad Caliph Abu2 Yu2suf Ya‘qu2b al-Mans5u2r requesting that he use
his fleet to prevent Crusader fleets reaching the East. But other
Christian shipping sailed in these waters only with Almohad
permission, as the treaties with them concluded by Pisa, Genoa, and
Sicily testify. No Christian powers wanted to antagonize the
Almohads at sea.171
The Almohads acquired the fleet of Seville when its admiral, ‘Alı3
ibn Ifisa2 ibn Maymu2n lent it to them for a siege of Ceuta in 1146. After
------------------------------
170
On the Almohads in general in this period see Al-Marra2kushı3, Al-Mu‘jib, pp.
154-281, esp. pp. 245-6 (Alarcos), 279-80 (Las Navas de Tolosa); Ibn abı3 Zar‘, Rawd 5
al-Qirt6a 2s, pp. 242-342, esp. pp. 309-22 (Alarcos), 339-42 (Las Navas de Tolosa); Ibn
al-Athı3r, Al-Ka2mil (Fagnan), pp. 526-66, 609-12 (Alarcos); Ibn Khaldu2n, ‘Ibar (De
Slane), vol. 2, pp. 161-227, esp. pp. 213-14 (Alarcos), 224-6 (Las Navas de Tolosa).
For a detailed account of Almohad history 1160-73, see Ibn Sa2h5ib, Al-Mann bi ’l-
ima2ma, pp. 40-234.
171
De Mas Latrie, Traités de paix, vol. 2, Documents, Pise III-IV (pp. 27-30),
Gênes II (p. 108), Deux-Siciles I (p. 152); Ibn Khaldu2n, ‘Ibar (De Slane), vol. 2, p.
215; idem, Muqqadimah, vol. 2, pp. 43-5. See also Gaudefroy-Demombynes, “Lettre
de Saladin”; Rosenberger, “Contrôle du détroit”, pp. 24-5.
98 CHAPTER ONE
the conquest of Ceuta and Tangier, ‘Abd al-Mu‘min named his son
Abu2 Sa‘ı3d governor of Ceuta and the Almohad fleet began to be built
up there and in Tangier. It was the fleet of Ceuta commanded by Abu2
Sa‘ı3d which was instrumental in the recovery of Almeria in 1157. He
also built a powerful new base on the Atlantic: the Riba2t6 al-Fath on
the southern bank of Wadı3 Abu2 Rak5ra2k5 opposite Salé, now Rabat. In
November 1158 he sent a circular from there ordering the coastal
tribes to construct ships and as early as 1160 the Almohad fleet was
able to beat off a Sicilian fleet sent to relieve the siege of al-Mahdiyya
by ‘Abd al-Mu‘min. According to Ibn al-Athı3r, the Almohad fleet
included 70 shawa2nı3, tara2’id and shalandiyya2t. According to Ibn abı3
Zar‘, in 1162 he had 400 ships built: 120 at Mamora upstream from
Rabat, 100 at Tangier, Ceuta, Ba2dis and other ports of the Rif, 100 in
the Maghrib at Oran and Hunayn, and 80 in al-Andalus. By the 1180s
the Almohad fleet was a force to be reckoned with by all. In 1179/80
the Caliph Abu2 Ya‘qu2b Yu2suf I sent the fleet under Gha2nim ibn
Muh5ammad ibn Mardanı3sh to blockade Lisbon, which he was trying
to recover. According to Ibn Khaldu2n, he returned with considerable
booty, although Ibn ‘Idha2rı3 recorded that he was defeated and
captured. In the following year a Muslim officer of Roger II of Sicily
by the name of Ah5mad al-S4iqillı3, who had fled to Ifrı3qiya when
William I succeeded to the throne and then to the court at Marrakesh
where he was made admiral of the fleet, reorganised it and gained a
victory over the Portuguese, capturing 40 ships according to Ibn
‘Idha2rı3; although, according to Ibn Khaldu2n the victory was gained by
the admiral of Seville, ‘Abd Alla2h ibn Ish5aq2 ibn Ja2mı3, and only 20
ships were captured. In 1184 Ah5mad al-S4iqillı3 was again sent with the
fleet to blockade Lisbon in conjunction with a land assault by Abu2
Ya‘qu2b Yu2suf, who died at Santarem. When Alı3 ibn Gha2niya of the
Balearics seized Bija2ya in 1185, the Almohad fleet under the
command of Ibn Jamı3 quickly drove him from it. The Balearic fleet of
the banu2 Gha2niya was by now no match for that of the Almohads. Ibn
Khaldu2n wrote that it reached “a size a nd quality never, to our
knowledge, attained before or since”. As long as the banu2 Gha2niya in
the Balearics continued the corsair war against Christians they were
tolerated. But their rebellious activities in the Maghrib provoked
punitive expeditions against the islands from Ceuta and Almeria. In
1203 an expedition consisting of 300 ships, of which 70 were
aghriba, 30 t6ara2’id, 50 mara2kib, and the rest qawa2rib and merchant
ships, was mobilized in Denia. The Balearics were occupied by the
172
Almohads.
Having eliminated the menace of the Neretljani in 1000 Venice
became if not yet the mistress of the Adriatic at least the dominant
power in it. Only two years later the Doge led the Venetian fleet to the
relief of Byzantine Bari, besieged by the Muslims, and defeated their
naval forces in a three-day battle outside the harbour. For many years
thereafter the Adriatic was peaceful; although, Doge Domenico
Contarini was led to recapture Zara and reimpose Venetian authority
along the Dalmatian coast in 1062 by increasing confusion caused by
pressure from Croats, Hungarians, and Byzantines. Thereafter the
Venetian fleet remained inactive until Alexios I Komne2nos was
compelled to seek Venetian help against Robert Guiscard in 1081.173
Guiscard was a man of enormous ambition. In 1074 he accepted a
proposal for a engagement between his daughter Olympias and
Constantine, the infant son of Michael VII Doukas, in return for
imperial titles carrying with them very substantial stipends, money of
which Guiscard had sore need. But the overthrow of Michael VII by
Nike2phoros III Botaneiate2s changed all of that. The marriage was
called off and Olympias, rebaptized Helena, became a virtual prisoner
in Constantinople. The stipends no doubt ceased to be paid. An
opportunity to invade the Empire beckoned. An imposter claiming to
be the deposed Michael VII was used as an excuse and in the spring of
1081 Robert’s forces crossed the Adriatic. His son Bohemond first
seized Corfu and in June the united forces laid siege to Dyrrachion.174
In April 1081 Botaneiate2s himself was overthrown by Alexios I
Komne2nos, who was married to Eire2ne2 Doukaina. Constantine and
Helena were rehabilitated. But by the time that news of that reached
the West the attack had gone in and was not to be called off. The siege
of Dyrrachion continued despite the best efforts of Alexios. His
Venetian allies lost their initial engagement at sea with the Normans,
------------------------------
172
Al-Baydhaq, Histoire des Almohades, pp. 176-7, 200-201; Al-Marra2kushı3, Al-
Mu‘jib, pp. 274-5; Al-Nuwayrı3, Niha2yat ‘al-Arab (Caussin), pp. 445-8; Al-Tija2nı3,
Rih5la, ser. 5, 1.1, pp. 398-9; Ibn abı3 Zar‘, Rawd5 al-Qirt6a2s, pp. 275, 284, 323, 386; Ibn
al-Athı3r, Al-Ka2mil (Fagnan), pp. 584-90, 603-8; Ibn Khaldu2n, ‘Ibar (De Slane), vol. 2,
pp. 180-81, 183-4, 190, 192, 202, 204, 208-9, 217-19; vol. 4, p. 63; idem,
Muqqadimah, vol. 2, pp. 43-5; Romuald of Salerno, Chronicon, pp. 241-2.
173
John the Deacon, Cronaca Veneziana, pp. 165-7;
174
Amatus of Monte Cassino, L’ystoire de li Normant, VII.xxvi (pp. 213-14);
Andrea Dandolo, Chronica, IX.viii (p. 216); Anna Komne2n e2, Alexiade, I.x.3, I.xii,
I.xiv, I.xvi (vol. 1, pp. 37, 42-7, 51-3, 56-61); Geoffrey Malaterra, De rebus gestis,
III.13 (p. 65); Lupus Protospatharios, Annales, 1076 (p. 60); Romuald of Salerno,
Chronicon, pp. 189, 191; William of Apulia, Gesta, IV, ll. 171-207 (p. 283). On the
Norman campaign in the Balkans see also Loud, Age of Robert Guiscard, pp. 209-23.
100 CHAPTER ONE
although they regrouped and successfully re-engaged, and Alexios’s

own relief force by land was overwhelmingly defeated in October.
The emperor was lucky to escape with his life and the city was
eventually betrayed in February 1082. However, events in Italy
compelled Robert to return home leaving Bohemond in command.
Although he fought hard for another eighteen months and had several
victories, he was eventually outmanœuvred by Alexios’s guerrilla
tactics, the latter’s suborning of some of his forces, inability to take
some strategic towns and fortresses, scarcity of supplies, and a
Venetian recovery of Dyrrachion. He was compelled to evacuate at
the end of 1083.175
After subduing his enemies in Italy, Robert returned to the Balkans
in 1084 to find both Corfu and Dyrrachion back in Byzantine hands.
He was initially defeated at sea by a combined Venetian and
Byzantine fleet although he turned the tables on them in a second
engagement. The Byzantine squadron in this allied fleet appears to
have been commanded by a Michael Maure2x, who had led squadrons
against Guiscard in Apulia in the 1060s, had formed his own private
force in Paphlagonia in the 1070s, and was reported by Nike2phoros
Bryennios to have had great experience of maritime affairs. His
squadron was probably his own private force since Alexios does not
appear to have attempted to reconstitute any Byzantine imperial
squadrons until some time later. Despite this success, however, the
Normans suffered cruelly over the winter and when Robert died at the
outset of an attack on Kefalle2nia in July 1085, the whole campaign
disintegrated.176
Alexios had survived the first crisis of his reign but the next ten
years would not be easy. In addition to a great many domestic
problems he had to face the pressure of the Pechenegs, who had
moved into the lower Balkans from ca 1078 and who reached the Sea
of Marmara by the end of the next decade. Constantinople was
besieged by them in the winter of 1090-91. However, Alexios made
an alliance with their enemies the Kumans, who were Qipc°aq Turks,
------------------------------
175
Andrea Dandolo, Chronica, IX.viii (p. 216); Anna Komne2ne2, Alexiade, II.ii.3,
IV.i-viii (vol. 1, pp. 68, 144-68), V.iv-vii, VI.i (vol. 2, pp. 17-32, 41-3); Geoffrey
Malaterra, De rebus gestis, III.26-8 (pp. 72-5); John Zo2naras, Epitome2 historio 2n,
XVIII.22 (vol. 4, pp. 237-9); Romuald of Salerno, Chronicon, pp. 192-6; William of
Apulia, Gesta, IV, ll. 272-505 (pp. 284-9), V, ll. 1-105 (pp. 290-92).
176
Andrea Dandolo, Chronica, IX.viii (p. 216); Anna Komne2ne2, Alexiade, VI.v-vi
(vol. 2, pp. 50-57); Romuald of Salerno, Chronicon, pp. 196-7; William of Apulia,
Gesta, V, ll. 143-336 (pp. 293-7). On Maure2x see Anna Komne2ne2, Alexiade, IV.iii.1
(vol. 1, pp. 148-9); Nike2phoros Bryennios, Hyle historias, II.26 (pp. 196-7); William
of Apulia, Gesta, V, ll. 98-102 (p. 292).
and on 29 April 1091 he and the Kumans annihilated the Pechenegs at

the battle of Mt Lebounion in Thrace near the mouth of the Maritsa
river. Their remnants were absorbed and became part of the
mercenary forces in the imperial armies.177
Of even more concern were the activities of the Salju2qids of Ru2m
who, under Qı£lı£j Arslan I, controlled almost all of Asia Minor by the
mid 1090s. An independent amı3r known to the Byzantines as Tzachas,
Turkish Çaka, constructed a fleet at Smyrna ca 1088-9 and began to
raid across the Aegean and to seize control of its islands. This was in
many ways an even more formidable threat than that of the Normans
and Pechenegs because it threatened the essential heartland of the
Empire: its islands and sea lanes in the Aegean. He entered into
negotiations with the Pechenegs but lost his allies after the battle of
Mt Lebounion. He was then engaged by a new Byzantine fleet under
Alexios’s brother-in-law John Doukas, for whom, in keeping with the
centralizing changes to Byzantine naval forces, a new title of megas
doux, had been created. Tzachas was forced back to his base at
Smyrna. Eventually his fleet was destroyed by Constantine
Dalasse2nos and he was betrayed to his Turkish rival Qı£lı£j Arslan I by
Alexios and killed.178
In the spring of 1096 bands of Western pilgrims began moving
from France and Germany towards Constantinople in what was to
become known as the First Crusade. According to Ekkehard of Aura,
Alexios had been in communication with Pope Urban II, requesting
military aid against the Turks. Then, according to two obscure
sources, envoys from Alexios had met Urban at the Council of
Piacenza in March 1095 and had requested military aid. An even more
obscure source, almost incoherent in its Latinity, recorded that Urban
had first uttered the call to Crusade at Piacenza, before the momentous
call was made after the Council of Clermont, on 27 November
1095.179 However, it is in fact extremely unlikely that Alexios sent
------------------------------
177
Anna Komne2ne2, Alexiade, VIII.i-vi (vol. 2, pp. 127-46); John Zo2naras, Epitome2
historio2n, XVIII.23 (vol. 4, pp. 241-2).
178
Anna Komne2ne2, Alexiade, VII.viii, VIII.iii.2, IX.i.3-9, IX.iii (vol. 2, pp. 110-16,
134, 158-62, 164-6); John Zo2naras, Epitome2 historio 2n, XVIII.22 (vol. 4, p. 239).
179
Bernold of St Blasien, Chronicon, 1095 (p. 462); Ekkehard of Aura,
Hierosolymita, c. 5 (pp. 14-15); Fragmentum historiae monasterii-novi Pictaviensis,
coll. 1219-20; Theodore Skoutario2te2s, Synopsis Chronike2 (Sathas), pp. 184-5.
Bernold of St Blasien was an obscure German monastic chronicler; however, he was
present at Piacenza. Theodore Skoutario2te2s was a thirteenth-century bishop of Kyzikos
and from where he obtained this information, which is reported by no surviving
earlier Byzantine historian, is unknown. Ekkehard of Aura was an important historian
of the Crusade but his report was not replicated by any other major Latin historian of
102 CHAPTER ONE
envoys to Italy to seek military assistance from the Pope. Envoys

arriving in Italy in early spring 1095 would have had to have been sent
in autumn 1094 but by then the Norman threat had been eliminated,
the remnants of the Pechenegs had been absorbed, Tzachas’s fleet had
been destroyed and he himself killed, the Kumans had been thrown
back by a decisive campaign in the summer of 1094, and Alexios had
begun to take the offensive against the Turks by constructing a canal
from Lake Sophon to the Gulf of Nikome2deia, to isolate much of
Bithynia from Turkish raids and thus to create a beachhead for an
attempt at reconquest of Asia Minor. It is more probable that the
confused reports of requests made at Piacenza for assistance reflected
earlier overtures to the West such as that addressed to Count Robert I,
“the Frisian”, of Flanders ca 1091 seeking assistance against the
Pechenegs,180 and it is at least possible that there had been some
earlier communication between Alexios and the Pope as Ekkehard
reported, even though no hard evidence for it survives.
The Pisan and Genoese expulsion of Muja2hid al-Muwaffaq, from
Sardinia was one of the first manifestations of the growing rise to
maritime power of these two great trading cities of the Tyrrhenian
Sea. Amalfi had preceded them in developing maritime commerce
with the Muslim world but they would overtake and eclipse her in the
later eleventh century, although there are only fragments of evidence
from the eleventh century to indicate the growing reach of their
maritime commerce. A letter from the Cairo geniza, written in
Alexandria and dated to ca 1060, recorded that: “There arrived ships
from Genoa and other Ru2m places, and three other ships are expected
from Spain”. This is, in fact, the earliest known reference to Genoese
voyaging to the Levant. For Pisa there is the intriguing archaeological
evidence of the Pisan bacini, glazed ceramic shallow basins of
Muslim origin inserted into the walls of churches, 628 of which have
survived. Two from the church of St Piero a Grado, which dates from
the early 11th century, have been identified as coming from the
Balearics. Others can be identified as coming from the Maghrib and
------------------------------
the Crusade. The Fragmentum historiae monasterii-novi Pictaviensis, the only text to
report that Urban actually preached the Crusade at Piacenza, was written by a certain
monk Martin. It ends immediately after its report on Piacenza and Clermont and was
presumably very contemporary to the events. However, as it survives the text is
extremely corrupt.
180
Hagenmeyer, Epistulae, No I (pp. 129-36). See also Joranson, “Spurious letter”.
As we have it, the letter from Alexios to Robert of Flanders is undoubtedly false but,
it may well have been based on an authentic original.
al-Andalus.181
Of the two Tyrrhenian cities of Pisa and Genoa, it was Pisa which
was initially the more aggressive at sea. In 1063 the Pisans proposed a
joint attack on Palermo to the Normans but Count Roger considered it
premature and did not take part. Nevertheless, the Pisan fleet attacked
the port, seizing its chain and six magne naves. In 1087 Pisa, Genoa,
and Amalfi attacked al-Mahdiyya in revenge for the raids of Tamı3m’s
corsairs on Christian shipping and the coasts. There were many
Christian prisoners in his prisons and his assistance to the Sicilian
amı3rs against the Normans had also increased Christian hostility to
him. This was an important expedition because for the first time it
embodied some elements of the ideology of Crusading. The
expedition was successful and Tamı3m was forced to free his prisoners,
pay tribute, grant access to al-Mahdiyya to Christian merchants, and
promise to restrain his corsairs.182
Even though it appears that from early in his reign Tamı3m had
some form of truce with Sicily, Zı3rı3d relations with the rest of the
western Mediterranean remained fraught. In 1105 a Christian fleet of
“shawa2nı3” and “mara2kib” attacked al-Mahdiyya but was unable to
bottle up the Zı3rı3d fleet and was defeated by it. On the death of Tamı3m
in 1108 George of Antioch, a young Christian who had come from the
East with his father early in Tamı3m’s reign and had become his
financial official but who feared the animosity of his son and
successor Yah5ya2, fled to Sicily. He would become a renowned
admiral of Roger II. Yah5y a2 devoted considerable attention to his fleet,
increasing the number of ships, multiplying corsair raids on Christian
coasts and, according to Ibn Khaldu2n, compelling Genoa, Sardinia,
and “the French” to pay tribute. According to Muslim sources, in A.H.
503 (31 July 1109 – 20 July 1110) Yah5ya2 sent 15 shawa2nı3 or aghriba
to raid Ru2m but they were defeated by a Christian squadron with the
loss of six ships; however, in the autumn of 1113 the fleet of al-
Mahdiyya returned from the lands of the Ru2m with many captives.183
------------------------------
181
Abulafia, “Pisan Bacini”; Goitein, Mediterranean society, p. 318.
182
Al-Nuwayrı3, Niha2yat ‘al-Arab (Amari), pp. 153-4; Al-Tija2nı3, Rih5la, ser. 5, 1.1,
pp. 374-6; Bernardo Maragone, Annales Pisani, MXXXV, MLXIII, MLXXXVIII (pp.
5, 6); Ibn al-Athı3r, Al-Ka 2mil (Fagnan), pp. 487-8; Ibn ‘Idha2rı3, Al-baya2n al-mughrib,
vol. 1, p. 301; Ibn Khaldu2n, ‘Ibar (De Slane), vol. 2, p. 24. The major source for the
Mahdia campaign is the Carmen in victoriam Pisanorum, edited by Cowdrey in his
“Mahdia campaign”. See also Idris, Berbérie orientale, pp. 286-91; Manfroni, Marina
italiana. I, pp. 84-104.
183
Al-Tija2nı3, Rih5la, ser. 5, 1.1, pp. 376-8; Ibn al-Athı3r, Al-Ka2mil (Fagnan), p. 519;
Ibn ‘Idha2rı3, Al-baya2n al-mughrib, vol. 1, pp. 452, 455, 456; Ibn al-Khat6ı3b, A‘ma2l al-
a‘lam, p. 458; Ibn Khaldu2n, ‘Ibar (De Slane), vol. 2, p. 25.
104 CHAPTER ONE
Relations between Ifrı3qiya and Sicily began to worsen in the reign

of Yah5ya2’s son ‘Alı3. The governor of Gabes, an Arab chieftain by the
name of Ra2fi‘ ibn Maggan ibn Ka2mil, attempted to make himself
independent and was besieged by Alı3’s forces in 1117. He called on
Roger II for help and the Sicilian sent a fleet of 24 “shawa2nı3”, which
was, however, defeated or at least forced to withdraw by the Zı3rid
fleet. Whether there was an actual engagement is unclear. The
Almoravid raid on Calabria by Abu2 ‘Abd Alla2h Muh5ammad ibn
Maymu2n, which captured Nicotera, had been instigated by the Zı3rı3ds
and in reprisal Roger II sent a fleet, supposedly of 300 “shawa2nı3”,
carrying 30,000 men and 1,000 cavalry commanded by George of
Antioch and another admiral Christodoulos against Ifrı3qiya in 1123.
Although hit by a storm off Marsala some ships reached Pantelleria
and then went on to al-Mahdiyya. However, the raid was beaten off,
the Normans lost two thirds of their ships, and their last forces left
ashore were massacred on the night of 7-8 August.184
The storm clouds were gathering for the last Zı3rı3d, al-H4asan. In
1127 Roger II’s forces occupied Malta and in 1135 the H4amma2did
amı3r of Bija2ya, Yah5y a2 ibn al-‘Azı3z, attacked al-Mahdiyya and in
extremis al-H4asan turned to Roger II for assistance. Also in 1135 the
Normans occupied Jerba and then in A.H. 536 (6 August 1141 – 27
July 1142) George of Antioch assaulted al-Mahdiyya with 25
“aghriba”, returning to attack Tripoli of Libya the following year. The
Zı3rids began to slide into dependency on the Kingdom of Sicily,
which acquired a kind of protectorate over al-Mahdiyya. Tripoli was
occupied in 1146 and then in 1148 the Sicilian fleet was launched
against al-Mahdiyya itself and al-H4asan fled without offering any
resistance. The Norman occupation of Ifrı3qiya had begun. Although
Norman rule rapidly began to disintegrate, it was not finally ended
until the Almohad ‘Abd al-Mu’min overran the last possession, al-
Mahdiyya, in January 1160.185
------------------------------
184
(Caussin), p. 441; Al-Tija2nı3, Rih5la, ser. 4, 20, pp. 148-50, 177, ser. 5, 1.1, pp. 143,
378-82; Ibn al-Athı3r, Al-Ka2m il (Fagnan), pp. 523-5, 547-9; Ibn ‘Idha2rı3, Al-baya 2n al-
mughrib, vol. 1, pp. 461-3; Ibn Khaldu2n, ‘Ibar (De Slane), vol. 2, pp. 25-7, 35-6. See
also Manfroni, Marina italiana. I, pp. 182-93.
185
Al-Marra2kushı3, Al-Mu‘jib, pp. 195-7; Al-Nuwayrı3, Niha2yat ‘al-Arab (Amari),
pp. 156-9; idem, Niha2yat ‘al-Arab (Caussin), pp. 441-8; Al-Tija2nı3, Rih5la, ser. 4, 20,
pp. 110-11, 133-6, 150-51, ser. 5, 1.1, pp. 384-401; Ibn abı3 Zar‘, Rawd5 al-Qirt6a2s, pp.
279-81; Ibn al-Athı3r, Al-Ka2mil (Fagnan), pp. 550-53, 555-67, 576-80, 584-90; Ibn
‘Idha2rı3, Al-baya2n al-mughrib, vol. 1, pp. 469-77; Ibn Khaldu2n, ‘Ibar (De Slane), vol.
2, pp. 27-9, 193-4; Romuald of Salerno, Chronicon, pp. 226-7. On the Norman
conquest of Africa, see also Abulafia, “Norman kingdom”; Idris, Berbérie orientale,
It is no coincidence that the Genoese annals of Caffaro di

Caschifellone began with the account of the small Genoese expedition
to the Levant for the First Crusade. Consisting of twelve galleys of a
new Western type, galeae, and a transport galley, a sandanum, a
Latinization of the Greek chelandion, the fleet left Genoa in July 1197
and reached St Symeon in late November, taking around four months
for the voyage. The larger Pisan and Venetian fleets did not even
attempt to make the voyage in a single passage in one season. They
left in the autumns of 1098 and 1 099 respectively, wintering in the
Ionian islands and on Rhodes and reaching the Levant in the autumn
of 1099 and the spring of 1100 respectively.186
Voyages such as these by war fleets, across the length of the
Mediterranean far from one’s own territory and logistical support, had
not been seen before. Not since Belisarios’s voyage from
Constantinople to Vandal Africa had there been anything like them
and in his case he had had Byzantine territory all the way to the Ionian
islands and then the hospitality of the Ostrogothic regent Amalasuntha
in Sicily. In all the age of naval warfare between the Muslim and
Christian worlds, fleets had always coasted their own territory before
making short passages to targets of attack. The Byzantines had never
attempted to reconquer Crete from Constantinople. They always
brought their forces overland to one of the aple2kta on the south-west
coast of Asia Minor before making a short crossing to Crete. In 960
Nike2phoros Pho2kas had marshalled his forces at Phygela for the final,
successful attack.
What was undertaken by the Genoese, Pisans, and Venetians
between 1096 and 1098 was unprecedented in the Mediterranean and
it is no wonder that they took between one and a half and three years
to prepare for their voyages. The fleets of these maritime republics
were ad hoc assemblies composed of privately-owned ships and it
would have taken time to gather them if scattered across the sea.
Moreover the logistical problems of provisioning, and especially of
watering, large numbers of men on galleys, would have created great
problems in an age when developed port facilities were few and far
between. Even Western galleys of the new galea type, which did not
------------------------------
pp. 338-93.
186
Albert of Aachen, Historia Hierosolymitana, VII.xviii-xx (pp. 519-20); Andrea
Dandolo, Chronica, lib. IX, c. X(L) (pp. 221-3); Bernardo Maragone, Annales Pisani,
A.D. MXCVIIII (p. 7); Caffaro, De liberatione, p. 102; Raymond of Aguilers,
Historia Francorum, c. V (p. 242); Translatio Sancti Nicolai, cc. I-V (pp. 254-57);
William of Tyre, Chronicon, 9.14-15 (vol. 63, pp. 438-41).
106 CHAPTER ONE
have a bank of oarsmen below deck as Byzantine dromons had had,

and which therefore could stow many more provisions and much more
water in their holds, could still only stay at sea for around five days
before they needed to water.187 Watering large galley fleets from
springs and wells must have been extremely time consuming.
In 1122-3 Venice launched a new Crusade but the Venetians still
wintered over in Corfu and then took over two months to make
Outremer from Crete in the spring of 1123. In their case, however, the
wintering and the slow rate of voyaging was probably necessitated by
the fact that they were the first to attempt to transport horses from the
West to Outremer. But already by then matters were evolving. As
early as 1100 a larger Genoese fleet of 26 galeae and either 4 or 6
sailing naves reached Latakia from Genoa between 1 August and the
onset of winter. A Genoese fleet of 60 galeae which participated in
the siege of Tripoli in 1109 and another of 22 which joined in that of
Beirut in 1110, as well as the Venetian fleet of 1109 which assisted at
the siege of Sidon in 1110, all appear to have made the voyage in a
single passage in one season. By the time of the Third Crusade and
that of Frederick II, voyages to the Levant were being made regularly
by large fleets, even carrying horses, in a matter of three to four
weeks.188
Between the Venetian Crusade of 1122-3 and the Third Crusade,
no actual Crusader fleets reached Outremer except for the Northern
fleet for the Second Crusade. However, according to Odo of Deuil,
when he was preparing his own expedition for the Second Crusade,
Louis VII of France wrote to Roger II of Sicily suggesting that the
French would need the support of his kingdom, to which Roger
responded with the offer of a fleet and logistical support. If this was
true and what was meant was that Sicily would provide transportation
to Outremer for the entire French army for the Second Crusade, Roger
II’s offer would have been a major innovation, well beyond even what
the Venetians had provided in 1122-3. It would suggest that Western
naval powers had made extraordinary progress in their ability to
project naval force far afield. The Sicilian experience in Ifrı3qiya
would no doubt have contributed. That the Sicilians perhaps did
already have such capabilities is also suggested by the fact that in
1147 Roger II was able to send his fleet under George of Antioch
against Corfu and the Ionian islands and into the Aegean, where
------------------------------
187
Consult Chapter Six below.
188
Andrea Dandolo, Chronica, lib. IX, cc. X(L)I, X(L)II (pp. 228, 233); Caffaro,
Annales Januenses, pp. 5, 14-15; Fulcher of Chartres, Historia Hierosolymitana,
III.xiv-xv (pp. 656-8); William of Tyre, Chronicon, 12.22 (vol. 63, p. 573).
Thebes and Corinth were plundered. Sicily certainly did have the
ability to send its naval forces far afield by 1154, when what appears
to have been a small squadron attacked Tinnis in Egypt, and again in
1174 when a fleet which attacked Alexandria was very large, being
said to have included 200 “shawa2nı3” as well as 36 “tara2’id”/taride,
horse transport galleys, six “sufun” transport ships carrying war
machines, and 40 “mara2kib” transports with provisions.189
Throughout the twelfth century, however, it was sailing naves
rather than galleys by which Western naval power was projected into
the Levant and which sustained the economic and human life-lines
which maintained the Crusader states. Fa2t6imid, and later Ayyu2bid,
galley forces operating out of Egypt proved ineffectual against them.
For so long as the Fa2t6imids of Egypt held some of the coastal towns
after the conquest of Jerusalem in 1099, shipping moving along the
Syro-Palestinian coast was subject to the attacks of their squadrons.
Moreover, Egyptian fleets were able to operate against the fledgling
Crusader Kingdom of Jerusalem and also to reinforce coastal towns
when under siege. We would argue that the Fa2t6imids in fact had little
grasp of the strategic situation and failed to really coordinate the
movement of their armies and squadrons; nevertheless, they did
attempt to make use of their naval forces. A squadron was present off
the coast during the battle of Ascalon on 12 August 1099 although it
did not become involved. Another participated in an assault on Jaffa
in 1102 which was unsuccessful because the army was defeated. In
the summer of 1103 a squadron participated in another campaign
against Jaffa but then distributed its grain amongst some of the coastal
towns when the campaign failed. Baldwin I’s siege of Sidon in 1108
failed because of the successful intervention of a Fa2t6imid squadron,
although another fleet sent to relieve Tripoli in 1109 arrived too late.
Nineteen Fa2t6imid ships entered Beirut in 1110 when it was under
siege but could not prevent its fall and an Egyptian relief fleet which
arrived too late to prevent the fall of Sidon in 1110 was intercepted by
------------------------------
Abu2 ’l-Fı3d a, Mukhtas5ar, A.H. 548 (p. 30); Abu2 Sha2m a, Kita2b al-rawd5atayn,
189
A.H. 570 (vol. 4, pp. 164-5); Andrea Dandolo, Chronica, lib. IX, c. X(L)III (p. 242);
Anonymous, Kita2b al-ilma2m , pp. 26-38; Bernardo Maragone, Annales Pisani, A.D.
MCLXXV [sic!] (p. 61); Continuatio Praemonstratensis, 1154 (p. 456); Ibn al-Athı3r,
Al-Ka2mil (Fagnan), pp. 568-9, 600-601; idem, Al-Ka2mil (RHCHOr), A.H. 548, 570
(vol. 1, pp. 491, 611-12); Ibn al-Qala2nisı3, Dhayl ta’rı3kh Dimashq, A.H. 549 (pp. 321-
2); Ibn Shadda2d, Al-Nawadir al-sult6a2niyya, p. 50; John Kinnamos, Historiae, III.4-5
(pp. 96-101); Al-Maqrı3zı3, Sulu2k (Broadhurst), A.H. 569 (pp. 48-50); idem, Sulu2k
(Blochet), A.H. 570 (vol. 8, p. 524); Nike2tas Cho2niate2s, Historia, Basivleiva Manouh;l
tou' Komnhnou' BV, (pp. 72-6); Odo of Deuil, De profectione Ludovici VII, bk III (pp.
58-9); Romuald of Salerno, Chronicon, p. 227; William of Tyre, Chronicon, 21.3
(vol. 63A, p. 963) and see also Eracles, XXI.3 (p. 1007).
108 CHAPTER ONE
Frankish ships and captured. Ibn al-Qala2nisı3 made the revealing

comment that by 28 November 1111 when the Franks lay siege to
Tyre, the inhabitants despaired of al-Afd5al being able to relieve them
by sea. Nevertheless, they successfully defied the siege, which was
lifted on 7 April 1112.190
After a decade of unsuccessful attempts to use their squadrons
against the coastlines of the incipient Crusader states, Fa2t6imid naval
resources were becoming stretched thin. The threat from their corsairs
against Western shipping arriving off the coasts of Outremer appears
to have dissipated. Nevertheless, the Egyptians were still able to send
a relief fleet into Tyre in 1113 and there also appears to have been an
attack on Jaffa in 1115 by the garrison of Ascalon, accompanied by an
Egyptian squadron which went on to Tyre. In 1118 another squadron
sailed to Ascalon to participate in a campaign but again went on to
Tyre. In 1122 a squadron again sailed to Tyre and replaced the
governor and in the following year the Fa2t6imids sent yet another army
and fleet to Palestine. However, on this occasion they were unlucky
because their expedition coincided with the arrival of the Venetian
Crusade off the coast. Doge Domenico Michiel destroyed the
Egyptian fleet off Ascalon and in the following year the siege of Tyre
could not be relieved by sea and the city fell on 7 or 8 July 1124.191
If we can believe Fulcher of Chartres, and William of Tyre
following him, in 1126 the Fa2t6imids sent 24 “galeae” north to raid the
coasts but it ran out of water near Beirut because of the loss of Tyre
and was severely mauled when it tried to take on water surreptitiously
from streams and springs. After that no more is heard of the Egyptian
fleet until 1151. In that year, probably as a response to mounting
pressure on Ascalon, a large fleet of 70 warships raided from Jaffa
------------------------------
190
Abu2 ’l-Mah5a2sin, Al-Nuju2m , pp. 488, 490, 492; Fulcher of Chartres, Historia
Hierosolymitana, II. xxxiii.1-2, xlii.2, xliv.5, (pp. 501-2, 535-6, 547-8); Ibn al-Athı3r,
Al-Ka2mil (RHCHOr), A.H. 496, 501, 503, 504, 505 (pp. 216, 257, 273-4, 279, 283-6);
Ibn Muyassar, Akhba2r Mis5r, A.H. 496 (p. 465); Ibn al-Qala2n isı3, Dhayl ta’rı3kh
Dimashq, A.H. 492, 496, 501-3, 505 (pp. 48, 58-9, 86-7, 89, 99-100, 119-25);
Saewulf, pp. 75-6; Sa2wı3ris, History of the Patriarchs (Khater & Burmester), vol. 2,
part 3, pp. 398-9; Sibt6 ibn al-Jawzı3, Mir’a2t al-zama2n, A.H. 496, 501-3, 505 (pp. 526,
535-7, 539, 543-5); William of Tyre, Chronicon, 11.3, 13 (vol. 63, pp. 498-500, 515-
16). See also Lev, State and society, pp. 107-13.
191
Abu2 ’l-Mah5a2sin, Al-Nuju2m , p. 492; Albert of Aachen, Historia Hierosolymitana,
XII.xvii (pp. 699-700); Fulcher of Chartres, Historia Hierosolymitana, II.liii.4-6,
III.ii.1, xvi.3, xx, xxvii-xxxiv.3 (pp. 585-6, 617-18, 661, 669-72, 695-735); Ibn al-
Athı3r, Al-Ka2mil (RHCHOr), A.H. 517-18 (pp. 354, 356-9); Ibn Muyassar, Akhba2r
Mis5r, A.H. 516-17 (pp. 468-9); Ibn al-Qala2nisı3, Dhayl ta’rı3kh Dimashq, A.H. 507,
518 (pp. 142-3, 170-72); Sibt6 ibn al-Jawzı3, Mir’a2t al-zama2n, A.H. 507, 517-18 (pp.
547, 563-5); William of Tyre, Chronicon, 11.24, 12.6, 21-3, 13.5-13 (vol. 63, pp. 531-
2, 552-3, 571-5, 591-602).
north to Tripoli. Then, when the Franks lay siege to Ascalon two years
later in 1153, the city was again provisioned and reinforced by the
men of a Fa2t6imid fleet, said by William of Tyre to have numbered 70
galeae plus supply naves. On this occasion there were a large number
of Frankish galeae and also naves participating in the siege and the
Egyptians had to fight their way in. But, since Ascalon had no
harbour, merely an open beach, the ships must have been abandoned
and then lost when the city surrendered on 22 August.192
Even the loss of the squadron at Ascalon did not put an end to the
activities of the Fa2t6imid navy. In 1155 squadrons raided coastal
shipping and temporarily seized the harbour at Tyre. In 1157 a raiding
fleet was said to have returned to Egypt with 700 captives. Another
small squadron of five galleys from Cairo did the same the next year,
1158, and both the Alexandria and Damietta squadrons were also
active in the same year. After that, however, no more is heard of the
activities of the Fa2t6imid fleet until its destruction by fire in its arsenal
at al-Fust6a2t6 in 1168 during a campaign against Egypt by Amalric of
Jerusalem. In the following year, when the Franks again invaded
Egypt in conjunction with a Byzantine fleet, there were apparently no
Fa2t6imid naval forces to oppose the Byzantines.193
In spite of Alexios Komne2nos’s efforts to rebuild something of a
Byzantine fleet, Byzantine naval forces had remained weak well into
the twelfth century. After the death of Tzachas the fleet reconstructed
by Alexios had been sent under the command of John Doukas to
suppress revolts in Crete and Cyprus. Eumathios Philokale2s was
appointed governor of Cyprus. By the time of the First Crusade,
squadrons of the fleet had sufficient capability to evacuate the
remnants of Peter the Hermit’s forces back to Constantinople and
Eumathios Philokale2s’ squadron on Cyprus played a significant role
off the Syrian coast during the First Crusade. His squadron was also
capable of transporting Robert of Flanders and Robert of Normandy,
presumably with their immediate entourages only, back from Latakia
------------------------------
192
Fulcher of Chartres, Historia Hierosolymitana, III.lvi (pp. 803-5); Ibn al-Athı3r,
Al-Ka2mil (RHCHOr), A.H. 548 (pp. 490-91); Ibn al-Qala2n isı3, Dhayl ta’rı3kh Dimashq,
A.H. 546, 548 (pp. 307-8, 316-17); William of Tyre, Chronicon, 13.20, 17.23-5, 27-
30 (vol. 63, pp. 611-12; vol. 63A, pp. 792-5, 797-805). Ibn al-Qala2nisı3 wrote that in
1153 some of the Ascalonites departed by sea; however, it is much more likely that
the Franks would have seized any ships still surviving and that the Ascalonites would
have evacuated to Egypt by land, as William of Tyre recorded.
193
Ibn al-Athı3r, Al-Ka2m il (RHCHOr), A.H. 564, 565 (pp. 555, 568-70); Ibn al-
Qala2nisı3, Dhayl ta’rı3kh Dimashq, A.H. 550, 553 (pp. 323-4, 346); William of Tyre,
Chronicon, 20.7, 14-17 (vol. 63A, pp. 919-20, 927-34).
110 CHAPTER ONE
to Constantinople in 1099.194
Anna Komne2ne2 also related the very improbable story of a major
encounter between the Byzantine fleet and the Pisans on their way
east for the First Crusade. To give credence to the story, when Alexios
heard of Pisan pillaging in the Ionian islands during their wintering
there in 1098-9, he constructed a new fleet armed with Greek Fire and
entrusted it to Tatikios, the general who he had appointed to guide the
First Crusade across Asia Minor and who had recently returned from
Antioch, and to a Latin mercenary called Landulf. Tatikios and
Landulf defeated the Pisan fleet off Lycia and its remnants, conceiving
the idea of pillaging Cyprus, were beaten off by Eumathios Philokale2s
and went on to Latakia. Supposedly the Byzantine fleet was then
destroyed by storm on its return.195 However, there is absolutely no
corroborating evidence for this story in any other source and the
whole account rings of implausibility. It is followed by an equally
improbable account of a similar attack on a Genoese fleet the
following year. Supposedly forseeing that this Genoese fleet would
cause trouble, Alexios sent his general Kantakouze2nos by land to
Lycia while Landulf took the fleet around. Landulf supposedly
intercepted the Genoese off Cape Malea but judged his forces
inadequate to engage them and retired. The Genoese eluded
Kantakouze2nos who then assaulted Latakia.196 Again the entire episode
is unsupported by any other evidence and reeks of an attempt to
explain Bohemond of Taranto’s return to the West and gathering of a
new Crusade against the Empire.
Bohemond, the mastermind of the victory of the Frst Crusade, who
had become Prince of Antioch and who had been captured and
imprisoned in August 1100 but released in May 1103, returned to the
West in the autumn of 1104 to raise an army to attack the Byzantines.
He succeeded in persuading Pope Paschal II to give to his proposed
expedition the standing of a Crusade and in the autumn of 1107
crossed to Avlona and besieged Dyrrachion. In the meantime Alexios
------------------------------
194
Albert of Aachen, Historia Hierosolymitana, VI.xlv, lvi-lx (pp. 493-4, 501-4);
Anna Komne2ne2, Alexiade, IX.ii, X.vi.5, XI.vii.4 (vol. 2, pp. 162-4, 212; vol. 3, p. 34);
Caffaro, De liberatione, p. 114; Ibn al-‘Adı3m, Zubdat al-h5alab, p. 578; Ibn Muyassar,
Akhba2r Mis5r, A.H. 546, 550, 551, 553 (pp. 470-72); Ralph of Caen, Gesta Tancredi,
c. lviii (p. 649).
195
Anna Komne2n e2, Alexiade, XI.x.1-8 (vol. 3, pp. 41-5).
196
Anna Komne2ne2, Alexiade, XI.xi (vol. 3, pp. 46-9). Note that these two episodes
fall into the later section of Anna’s account of the First Crusade identified by Howard-
Johnston as not having been completed by her husband Nike2phoros Bryennios but
rather compiled by her from poorly understood files. See Howard-Johnston, “Anna
Komnene”, pp. 291-2.
had made Thessalonike2 his base, ordered a fleet to be prepared in the

Cyclades and maritime cities in Asia Minor and the Balkans, and sent
it to Dyrrachion under a new megas doux, Isaac Kontostephanos.
Kontostephanos made an ill-advised and unsuccessful attempt against
Otranto and then was unable to intercept Bohemond’s fleet and
prevent it crossing to Avlona. Bohemond was eventually out-
manœuvred and forced to surrender, but not because of anything the
Byzantine fleet achieved. During the last decade of Alexios’s life
squadrons of the fleet were used occasionally; for example, to convey
Manuel Boutoumite2s on an embassy to Tripoli and Jerusalem in 1111.
Anna Komne2ne2 also has a garbled account of a combined Pisan,
Genoese, and Lombard naval assault which Alexios supposedly
warded off but there is no evidence that this event ever took place.
There is certainly evidence that Alexios did reconstruct Byzantine
naval forces to some degree during his reign, but none to suggest that
by his death they were anything like a major force to be reckoned
with.197
This was demonstrated clearly when in response to John II
Komne2nos’s refusal to renew the commercial privileges which his
father had granted to the Venetians, the Venetian fleet for the Crusade
of 1122-3 attacked Corfu with impunity on its way east and then on its
return in 1125 sacked Samos, Chios, Lesbos, and Andros. The
Byzantine fleet could offer no resistance and John was forced to
renew the privileges. In fact, if we can believe Nike2tas Cho2niate2s,
although he was a strong military emperor, John was actually
persuaded by his collector of revenues, John of Poutze2, to allow the
fleet to decay. John convinced him to divert revenues raised for the
fleet into general revenues and then to pay for ships only when they
were needed.198
Even as late as 1147 Byantine squadrons did not have the
------------------------------
197
Anna Komne2ne2, Alexiade, XII.i.6, iv.1-3, viii.1, viii.3-ix.3, XIV.ii.6-14 (vol. 3,
pp. 56, 64-5, 77, 78-83, 148-54). Our assessment is very different to that of
Ahrweiler, Byzance et la mer, pp. 189-97, which we regard as uncritical and
Byzantino-centric. That the Empire exercized a “thalassocratie … au début du XIIe
siècle sur la Méditerranée orientale”, is simply untrue.
198
Andrea Dandolo, Chronica, lib. IX, c. X(L)II (pp. 232, 233, 235); Historia
ducum Veneticorum, §2 (pp. 73-4); John Kinnamos, Historiae, VI.10 (p. 281);
Nike2tas Cho2niate2s, Historia, Basivleiva Manouh;l tou' Komnhnou' AV (pp. 54-5). The
story of John of Poutze2 is used by Nike2tas as an explanation for why “Now, [i.e., at
the time of Nike2tas’s writing this early part of his Historia in the 1190s] as a result of
this ill-advised policy or pennypinching, pirates rule the seas and the Roman maritime
provinces are harassed by pirate ships, and the enemy gloats.”. It is not necessarily to
be believed.
112 CHAPTER ONE
capability to oppose the Norman fleet which raided into the Aegean
that year. Not until the 1160s did Byzantine naval forces become
really formidable again as a consequence of the ambitious foreign
policies of Manuel I Komne2nos, which would see adventurous
expeditions sent to Egypt, Italy and into Anatolia. He would be
criticized by Nike2tas Cho2niate2s for the enormous expenditure
occasioned, although even Cho2niate2s recognized that there was good
reason behind his adventurism and his reign did witness the last great
flourish of Byzantine naval power.199
Almost immediately after his succession, Manuel sent an army
against Antioch to drive its prince, Raymond, back out of Cilicia. The
army was accompanied by a fleet under De2me2trios Branas which was
instrumental in the success of the expedition, which compelled
Raymond to go to Constantinople to make his peace.200 Manuel’s
second expedition against Antioch in 1158-9, which imposed
Byzantine suzereinty over the principality, was apparently
unaccompanied by a fleet, at least the sources make no mention of
one. Byzantine squadrons continued to operate in Levantine waters
out of Antalya and Cyprus, but it would not be until 1169 that another
large Byzantine fleet made its appearance off the coasts of Outremer.
Exactly how large various Byzantine squadrons were at the time of
the Second Crusade is a matter of conjecture. Certainly the Byzantines
transported the French and German armies across the Bosporos, but
that would not have needed imperial ships. Local craft were hired or
impressed, as had no doubt been the case during the First Crusade.
The same would have been true of the ships which were used by the
imperial government to supply Crusader forces. However, Manuel
certainly had sufficient ships to ferry Conrad III and his immediate
household from Constantinople to Acre in 1148, even while most of
the Byzantine naval forces were engaged at Corfu, and then back from
Acre to Thessalonike2 in the autumn after the failure of the Second
Crusade before Damascus.201
The Norman attack on the Ionian and Aegean in 1147 led to the
mobilization of extraordinary forces and to yet another request to
------------------------------
199
John Kinnamos, Historiae, III.2 (p. 92); Nike2tas Cho2niate2s, Historia, Basivleiva
Manouh;l tou' Komnhnou' BV, ZV (pp. 72-6, 203); Romuald of Salerno, Chronicon, p.
227.
200
John Kinnamos, Historiae, II.3 (pp. 33-5); Nike2tas Cho2niate2s, Historia,
Basivleiva Manouh;l tou' Komnhnou' AV (p. 52); William of Tyre, Chronicon, 18.23 (vol.
63A, pp. 844-5).
201
Eustathios of Thessalonike2, “Oratio ad Manuelem imperatorem [4]”, p. 107;
John Kinnamos, Historiae, II.19 (pp. 86-7).
Venice for assistance in return for privileges. Nike2t as Cho2niate2s’s

enumeration of the fleet sent to Corfu is redolent of literary
affectation. The figure of nearly 1,000 trie2reis, “fire-bearing”
pyrphoroi, pente2konteroi, myoparo2nes, horse transport hippago2goi,
transport phortago2goi, and light piratical epaktrokele2tes, is not to be
taken literally. This was just a collection of classical names for ships;
however, the fleet was clearly a very large one assembled from all
quarters. John Kinnamos numbered it at 500 trie2reis and 1,000
hippago2goi and supply ships.202 Under the megas doux Stephen
Kontostephanos the siege of Corfu town began in the autumn of 1148
and although it did not progress smoothly, the garrison eventually
surrendered in the summer of 1149.203 Sometime during the siege
Roger II may have sent a Sicilian squadron under George of Antioch
raiding into the Aegean, no doubt as a diversion. The fleet was said to
have reached Constantinople and to have demonstrated off the
imperial palace but to have been intercepted on its return by a
detachment of the imperial fleet from Corfu under a certain Chouroup
and heavily defeated somewhere near Cape Malea. Supposedly, Louis
VII of France, returning from the Crusade on a Sicilian ship, became
caught up in the battle. The story is supported by the Historia ducum
Veneticorum and Andrea Dandolo, who, however, ascribe the victory
to the Venetians.204
Even immediately after the ending of the siege of Corfu, Manuel
may have planned to cross over into Italy to attack Roger II in his own
territory. According to Kinnamos, Manuel crossed to Avlona and from
there ordered the megas domestikos, John Axouch, who had
succeeded the dead Stephen Kontostephanos in command of the fleet,
to Ancona, from there to harry Italy. Axouch apparently prevaricated
and stopped at the river Vijosë and the fleet was severely mauled by a
------------------------------
202
Andrea Dandolo, Chronica, lib. IX, c. X(L)III (p. 243); Historia ducum
Veneticorum, §§3-4 (p. 75); John Kinnamos, Historiae, III.2, 4 (pp. 92, 96-7);
Michael Rhetor, “Oratio ad Manuelem imperatorem”, p. 156; Nike2tas Cho2niate2s,
Historia, Basivleiva Manouh;l tou' Komnhnou' BV (pp. 76-7); Tafel and Thomas,
Urkunden, vol. 1, no. 51 (pp. 113-24).
203
Andrea Dandolo, Chronica, lib. IX, c. X(L)III (p. 243); Historia ducum
Veneticorum, §§3-4 (pp. 75-6); John Kinnamos, Historiae, III.4-5 (pp. 96-101);
Nike2tas Cho2niate2s, Historia, Basivleiva Manouh;l tou' Komnhnou' BV (pp. 77-88).
204
Andrea Dandolo, Chronica, lib. IX, c. X(L)III (p. 243); Continuatio
Praemonstratensis, 1149 (p. 454); Historia ducum Veneticorum, §§3-4 (p. 75); Ibn al-
Athı3r, Al-Ka2mil (RHCHOr), A.H. 544 (p. 477); John Kinnamos, Historiae, II.19, III.5
(pp. 87-8, 100-101).
114 CHAPTER ONE
storm before returning to Constantinople.205

Not until 1155 was Manuel ready to return to Italy. The imperial
fleet was entrusted to his uncle, Constantine Angelos, and ordered to
Monemvasia to await reinforcements from provincial squadrons.
Angelos, however, grappled with a Sicilian fleet before the provincial
reinforcements had arrived and was decisively defeated and taken off
to prison in Sicily. It is possible that this Sicilian fleet was one
returning from the reported attack on Tinnis in Egypt in the summer of
1154. The campaign that eventuated in Italy in 1155-6, which at first
was crowned with success with the capture of Bari but which ended in
total defeat outside Brindisi on 28 May 1156 was fought largely by
Byzantine gold and disaffected Normans and Italians. At the
beginning the Byzantine commanders had only ten ships and small
numbers of forces, although Manuel later sent reinforcements of Latin
and Turkish mercenaries and later still a fleet carrying another army
under the command of the megas doux Alexios Komne2nos Bryennios,
the son of Anna Komne2ne2. The naval forces, however, played no part
in the outcome. All that remained of the Italian adventure was the city
of Ancona, to which Alexios Axouch, son of John Axouch, was sent
out in 1158, and which remained in Byzantine hands. John Axouch
was instrumental in arranging a truce between the Empire and Sicily
in the spring of 1158. The raid on Constantinople by a Sicilian
squadron commanded by Stephen, the brother of Maio of Bari,
reported by both Nike2tas Cho2niate2s and Romuald of Salerno probably
took place in 1157 and contributed to the climate which induced
Manuel to seek peace with Sicily.206
Manuel sought to project Byzantine authority not only to the West
but also into Levantine waters. He had from the beginnings of his
reign established a kind of religious protectorate over the Orthodox
churches in the Holy land. Amalric of Jerusalem had invaded Egypt
several times and prospects for the conquest of the land of the Nile
------------------------------
205
John Kinnamos, Historiae, III.6, 9 (pp. 101-2, 113); Nike2tas Cho2niate2s ,
Historia, Basivleiva Manouh;l tou' Komnhnou' BV (p. 89); Theodore Prodromos,
Historical poems, no. 30 (pp. 349-60), esp. ll. 213-22, 402-11.
206
Abu2 ’l-Fı3d a, Mukhtas5ar, A.H. 548 (p. 30); Continuatio Praemonstratensis, 1154
(p. 456); Ibn al-Athı3r, Al-Ka2mil (RHCHOr), A.H. 548 (p. 491); Ibn al-Qala2nisı3, Dhayl
ta’rı3kh Dimashq, A.H. 549 (pp. 321-2); John Kinnamos, Historiae, III.12, 13, IV.1-
13, 14 (pp. 118-21, 134-68, 169-70); Nike2tas Cho2niate2s, Historia, Basivleiva Manouh;l
tou' Komnhnou' BV (pp. 91, 94-6, 96-8, 99); Romuald of Salerno, Chronicon, pp. 227,
239-41. There are so many similarities in the accounts of the Sicilian raids on
Constantinople in 1148/9 and 1157 that it is quite possible that the first never in fact
took place. The second must have because Maio of Bari did not become amı3r of amı3rs
in Sicily until 1154 and his brother Stephen was appointed amı3r shortly thereafter.
were not at all unrealistic. The native populace had been demilitarized
for centuries and the key to the country lay in possession of the three
key cities: Cairo, Damietta, and Alexandria. In retrospect all of the
Byzantine and Crusader attempts to conquer Egypt between 1163 and
1248 appear to have been futile; however, the prospects of success
were in fact quite real. In 1168 Manuel sent an embassy to Jerusalem
to propose an allied invasion of Egypt and Amalric responded by
sending a legation back to him to draw up the plans. A fleet under the
command of the megas doux Andronikos Kontostephanos, enum-
erated by William of Tyre at 150 longe naves rostrate known as galee,
60 naves maiores horse transports. and 10-12 naves maxime transports
known as dromones, and by Nike2tas Cho2niate2s at 200 ploia makra,
reached Acre in the summer of 1169. The assault on Damietta failed,
however, because of poor coordination and because the Byzantines
ran out of food. Manuel had provided provisions for three months
from August but the siege dragged on into December before being
abandoned. Reportedly, much of the fleet was lost in storms on the
way home.207 However, the destruction of fleets in storms on the way
home after unsuccesful expeditions was something of a literary topos
and subsequent events suggest that much of the fleet must have
returned. Accusations by John Kinnamos and Nike2tas Cho2niate2s that
the debacle was all due to the Franks may also be unwarranted
because when Amalric visited Constantinople in person in 1171 he
was warmly welcomed by the emperor.208
In that same year Manuel sent orders to officials throughout the
Empire to imprison and confiscate the property of every Venetian in
their jurisdictions on 12 March. The action against them was taken in
response to their overstepping the bounds of peaceful and law-abiding
bourgesioi when they sacked a newly established Genoese quarter in
Constantinople sometime after August 1170. Amazingly the secret
orders remained undiscovered by the Venetians and thousands of them
------------------------------
207
A.H. 565, 570 (pp. 149-53, 173); Ibn al-Athı3r, Al-Ka2mil (RHCHOr), A.H. 565 (pp.
568-70); Ibn Shadda2d, Al-Nawadir al-sult6a2niyya, pp. 45-6; John Kinnamos, Historiae,
VI.9 (pp. 278-80); John Tzetze2s, Epistulae, Epp. iV, igV (pp. 19, 21-4); Nike2tas
Cho2niate2s, Historia, Basivl eiva Manouh;l tou' Komnhnou' EV (pp. 159-68); William of
Tyre, Chronicon, 20.4, 13-17 (vol. 63A, pp. 915-17, 926-34). See also below pp. 415-
17.
208
Eustathios of Thessalonike2, “Oratio ad Manuelem imperatorem [2]”, pp. 39-40;
John Kinnamos, Historiae, VI.10 (p. 280); Nike2tas Cho2niate2s, Historia, Basivl eiva
Manouh;l tou' Komnhnou' EV (pp. 171-2); William of Tyre, Chronicon, 20.22-4 (vol.
63A, pp. 940-43).
116 CHAPTER ONE
were seized and their ships and property confiscated.209 In response, at

the insistence of the Venetian people, Doge Vitale Michiel II left
Venice in September 1171 with a Venetian fleet of 100 galeae and 20
transport naves in an attempt to induce the emperor to reverse his
actions. The fleet lay siege to Euripos but the Doge and the governor
came to an agreement to send a joint delegation to Constantinople to
urge Manuel to release his Venetian prisoners. It then crossed to
Chios, where they wintered, waiting for a reply. Manuel, however,
refused to see the envoys, although he sent his own to Chios inviting
them to send another delegation, which was done. Again Manuel
refused to see the envoys but sent his own inviting yet a third
delegation. Over the winter disease broke out in the fleet and in the
spring it moved to the tiny islet of Panaghia, and then to Lesbos and
Skyros in attempts to rid itself of the pestilence, after which it returned
to Venice and the Doge was murdered by the mob. Now, it may have
been that the Doge was indeed deceived by the emperor’s
prevarications and failed to take any military action because he sought
to achieve his objectives by negotiation. However, it is very curious
that whereas the Venetian sources make no mention of naval actvity
by the Byzantines against the fleet, John Kinnamos recorded that they
were opposed by Byzantine forces on Chios and that an imperial fleet
pursued them back across the Aegean to Cape Malea. Nike2tas
Cho2niate2s and Theodore Skoutariote2s following him stated that
Manuel sent Andronikos Kontostephanos to Chios with 150 trie2reis
and that he pursued the Venetians to Cape Malea. In fact the
Byzantine account makes more sense and also explains Michiel’s
failure to even attempt to enter the Dardanelles. The Venetians never
had the numbers to force the straits against the powerful Byzantine
fleet and Manuel knew that he had no need to see Venetian envoys.
Michiel may indeed have wished to negotiate, but negotiations always
proceed more smoothly when backed by some force. In this case there
was none and when disease had done its work the Byzantine fleet
simply drove the Venetians out of the Aegean. In 1171-2, it was not as
it had been in 1125. In the following year, 1173, Byzantine naval
forces were instrumental in enabling Ancona to endure a seven-month
siege by the forces of the Western Emperor Frederick I Barbarossa
------------------------------
209
Andrea Dandolo, Chronica, lib. IX, c. X(L)V (p. 250); Genoa, Codice
diplomatico, vol. 2, no. 53 (pp. 121-3); Historia ducum Veneticorum, §6 (p. 78); John
Kinnamos, Historiae, VI.10 (pp. 281-2); Nike2tas Cho2niate2s, Historia, Basivleiva
Manouh;l tou' Komnhnou' EV (pp. 172-3). See also Madden, Enrico Dandolo, pp. 52-7.
and his Venetian allies.210

That these parameters are correct is suggested by the fact that even
after the failed expedition of 1169 to Damietta, the Egyptians had sent
envoys to Constantinople to make a treaty of peace in exchange for
annual tribute. Whether a peace was actually concluded, as Nike2tas
Cho2niate2s wrote, or rejected, as John Kinnamos wrote, is unknown.
But in 1176, at the same time as he was preparing his attack on the
Salju2qids of Ru2m which culminated in his defeat at Myriokephalon,
Manuel sent the fleet back to Outremer. John Kinnamos numbered it
at 150 ships. It reached Acre, where William of Tyre numbered it at
70 galee plus other naves; however, negotiations over a campaign
against Egypt stalled and the fleet eventually returned home.211
When S4ala2h5 al-Dı3n seized the Egyptian throne in 1171, the country
was virtually defenceless at sea. No naval resistance appears to have
been mounted against either the allied attack on Damietta in 1169 or
the Sicilian attack on Alexandria in 1174. However, probably in
response to the threat from the Byzantine fleet in 1176 in 1177 S4ala2h5
al-Dı3n visited Alexandria and ordered the construction of new
warships. By 1179 he had 60 shawa2nı3 and 20 tara2’id. The extension
of Ayyu2bid rule across the Maghrib to Ifrı3qiya, al-Qayrawa2n being
taken from the Almohads in 1187, also opened up access to timber
supplies and Maghribin sailors. Maghribin sailors and marines are
reported aboard S4ala2h5 al-Dı3n’s ships. A squadron raided Acre on 14
October 1179, capturing and sinking several ships. By 1181 there
were 50 ships stationed at Damietta to protect the coasts and they beat
off a Frankish assault on Tinnis. Another squadron captured two large
sailing but6sa2t on their way in to Outremer in the same year, three
others were taken in 1182 and another in 1183. However a major land
and sea assault on Beirut in 1182 failed because the fleet was defeated
and driven back to Egypt by a Frankish squadron raised in Acre and
Tyre. Al-Maqrı3zı3 reported that 31 Egyptian shawa2nı3 and other
h5arra2qa2t left on a cruise in May 1184 but gave no other details.
Despite this activity, however, and William of Tyre’s gloomy
------------------------------
210
Eustathios of Thessalonike2, “Oratio ad Manuelem imperatorem [2]”, pp. 36-9;
idem, “Oratio ad Manuelem imperatorem [4]”, pp. 109-17; Historia ducum
Veneticorum, §7 (pp. 79-80); Andrea Dandolo, Chronica, lib. IX, c. X(L)V (pp. 251-
3); John Kinnamos, Historiae, VI.10, 12 (pp. 282-6, 288-9); Nike2tas Cho2niate2s,
Historia, Basivleiva Manouh;l tou' Komnhnou' EV, ZV (pp. 172-3, 201-2); Theodore
Skoutario2te2s, Synopsis Chronike2 (Sathas), p. 281. See also Abulafia, “Ancona”.
211
John Kinnamos, Historiae, VI.9, VII.3 (pp. 280, 300); Montfaucon,
Palaeographia Graeca, pp. 47-8; Nike2tas Cho2niate2s, Historia, Basivleiva Manouh;l tou'
Komnhnou' EV (p. 168); William of Tyre, Chronicon, 21.15(16)-17(18) (vol. 63A, pp.
981-6). See also Magdalino, Manuel I Komnenos, pp. 95-8.
118 CHAPTER ONE
assessment of the effects of the Ayyu2bid conquest of Egypt on access

by sea from the West to Outremer, it is clear that in fact S4ala2h5 al-
Dı3n’s uses of Egyptian naval forces against the maritime lifelines of
the Crusader states were little more than pin pricks. The real struggle
would come during the siege of Acre.212
On 4 July 1187 S4ala2h5 al-Dı3n all but destroyed the army of the
Kingdom of Jerusalem at the Horns of H4at6t6ı3n. Acre was surrendered
shamefully on the 10th of July and Jerusalem on 2 October, by which
time all that was left to the Franks was Antioch, Tyre, Tripoli,
Tortosa, and a few isolated fortresses. A squadron of the Egyptian
fleet under its Armenian admiral, the h5a2jib H4usa2m al-Dı3n Lu’lu al-
Mas‘u2dı3, moved up to Ascalon. However, an assault on Tyre was
beaten off when the Egyptian fleet was engaged by galeae from Tyre
and completely defeated. The king, Guy of Lusignan, was released
from captivity in the spring of 1188 and on 27 August 1189 began a
siege of Acre with the assistance of a Pisan fleet which had reached
Tyre in April. In fact, however, the Pisan fleet had not been the first to
reach the East after the disaster at H4at6t6ı3n. That honour had belonged to
the Sicilians. In 1185 the Normans had launched a new assault on the
Byzantine Empire, advancing overland on Thessalonike2 while a fleet
came round into the Aegean. After the defeat of the land army, the
fleet was attacked by the Byzantines in the Gulf of Astakos in
November 1185 and forced to withdraw through the Dardanelles to
Crete, where it wintered under agreement with Isaac Komne2nos, the
self-proclaimed emperor of the island. Some time during the stay in
Cyprus command passed from Count Tancred of Lecce to admiral
Margaritus of Brindisi, who defeated a Byzantine fleet of 70 ploia
makra sent against the island in the following spring. In the spring of
1188 William II of Sicily sent Margaritus with a fleet of 50 galeae to
the Levant and his forces helped to save Tripoli, Tyre, and Antioch,
and contributed much to enabling the Franks to survive through 1188
and the winter of 1188-9 before other Crusader forces reached the
East.213
------------------------------
212
Abu2 Sha2ma, Kita2b al-rawd5atayn, A.H. 575, 578, 583 (pp. 200, 203, 209-11,
223, 230, 235, 239-40, 342); Ibn Khaldu2n, ‘Ibar (De Slane), vol. 2, pp. 91-5; Al-
Maqrı3zı3, Sulu2k (Broadhurst), A.H. 570-72, 574, 577-80, 583 (pp. 52, 53, 56, 59, 64,
70, 71, 77, 87); Al-Tija2nı3, Rih5la, pp. 153-63; Ibn al-Athı3r, Al-Ka2m il (RHCHOr), A.H.
578, 579 (pp. 653, 660); William of Tyre, Chronicon, 20.10, 22.18(17)-19(18) (vol.
63A, pp. 923-5, 1032-6). See also Ehrenkreutz, “Place of Saladin”, pp. 100-110; Lev,
Saladin in Egypt, pp. 161-84; Odetallah, “S4ala2h al-Dı3n and the sea”.
213
A.H. 584 (pp. 362-3); Ambroise, L’estoire de la guerre sainte, ll. 2730-86 (coll. 73-
5); Continuation de Guillaume de Tyr, §§40-81, esp. §§64, 72-3, 75 (pp. 52-89, esp.
The struggle for Acre resolved itself into a complex series of

strategic and technological parameters with the garrison in the city
besieged by Crusader forces, which were too few to close off ingress
into the city at first, while they themselves were confronted by S4ala2h 5
al-Dı3n’s forces which arrived in September 1189. An initial fleet may
have broken into the port on 31 October and H4usa2m al-Dı3n Lu’lu
definitely broke in with 50 shawa2nı3 on 25 or 26 December 1189,
capturing some Christian ships in the process. Another fleet broke
through on 15 June 1190 and a large but6sa from Beirut made it
through the cordon of Crusader ships at the end of August. Three
more from Egypt entered the harbour on 17 September but of either 7
or 15 more which arrived around 31 December 1190 some, if not all,
were dashed on the rocks in heavy weather. S4ala2h5 al-Dı3n established
Haifa, which was protected from the Crusaders by his own
encampment, as a base where ships entering and leaving Acre were
loaded. He did attempt to use the ships in Acre harbour to attack the
Crusader ships shortly before Easter 1190 but the Egyptian shawa2nı3
were defeated. The Crusaders’ grip around the city gradually
tightened as more and more reinforcements arrived from the West,
and when S4ala2h5 al-Dı3n’s last throw of the dice to slip another but6sa
into the city failed because it was intercepted and sunk by the arriving
fleet of Richard Cœur de Lion, the die was cast. The city surrendered
on 12 July 1191. In fact Egyptian squadrons appear to have acquitted
themselves quite well during the siege, breaking through into the city
several times and also defeating a Crusader attempt to take the Tower
of the Flies, to which the chain of the harbour was attached and which
would have given the Crusaders free access to the inner harbour if
taken. However, as well as the Pisans, Venetian and Genoese fleets
also sailed to Acre for the Crusade and they were joined by huge
numbers of ships from elsewhere in the Mediterranean and from
Northern Europe. The Egyptians were simply outnumbered and even
------------------------------
pp. 82-3, 85-7); Eracles, XXIII.xl-XXIV.xiv, esp. XXIV.iii, vi-vii, xi, xiv (vol. 2, pp.
62-125, esp. 107-9, 113-15, 119-21, 124-5); Ernoul, cc. XX-XXII (pp. 236-47, 251,
257); Eustathios of Thessalonike2, Capture of Thessaloniki, §§54, 59, 138 (pp. 66, 74,
150); Ibn al-‘Adı3m, Zubdat al-h5alab (Blochet), p. 184; Ibn al-Athı3r, Al-Ka2mil
(RHCHOr), A.H. 583-5 (pp. 677-744, esp. 697, 708-9, 718, 720-21); Ibn Shadda2d,
Al-Nawadir al-sult6a 2niyya, pp. 72-9, 91, 96-8; ‘Ima2d al-Dı3n, Al-fath5 al-qussı3, A.H.
583-5 (pp. 13-168, esp. pp. 44, 68-72, 125-6); Itinerarium peregrinorum (Mayer), pp.
257-75, esp. 267-8, 271-2; Itinerarium peregrinorum (Stubbs), I.14, 25-6 (pp. 27-8,
59-62); Al-Maqrı3zı3, Sulu2k (Broadhurst), A.H. 583-5 (pp. 81-91); Nike2tas Cho2niate2s,
Historia, Basileiva Andronivkou tou' Komnhnou' AV , BV, Basvleiva Isaakivou tou' Aggelou'
AV (pp. 297-308, 317-20, 355-63, 369-70); Sa2wı3ris, History of the Patriarchs (Khater
& Burmester), vol. 3, part 2, pp. 119-41.
120 CHAPTER ONE
if some of the sailing supply ships may have managed to break out of
the port again, the galleys’ crews were committed to the defence of
the walls and the ships were trapped. When the city surrendered they
were taken by the Crusaders. That was one of the terms of
surrender.214
After the fall of Acre, little more is heard of the activities of the
Egyptian fleet, although some units of it certainly continued to exist.
The administration and financing of the fleet was reformed in 1191
and S4ala2h5 al-Dı3n gave command to his brother Sayf al-Dı3n al-‘Adil.
The arsenal at Cairo was still functioning in 1194, ships were sent
down the rivers to relieve Bilbays in 1195, Frankish but6sa2t were
attacked several times between 1196 and 1198, and an Egyptian amı3r
who had some castles near Sidon armed 20 galleys in 1203 and used
them to raid shipping off Cyprus and to re-provision his castles. In
1198 Sayf al-Dı3n al-‘Adil, by now sult6a2n in Damascus, made a truce
with the Franks for six years by land, but not by sea, suggesting that
Egyptian naval forces were still capable of hitting Western shipping.
Nevertheless, it is abundantly clear that Egyptian naval forces had
been severely mauled at Acre and that they would have been no match
for the massive battle and transport fleets gathered by Venice for the
projected Fourth Crusade assault on Egypt, which, of course, was
diverted by a series of circumstances to Constantinople.215
There can be little doubt that at the death of Manuel Komne2nos in
1180 Byzantine naval forces were still a power to be reckoned with in
the eastern Mediterranean. There is nothing to suggest that the fleet
which sailed to Outremer in 1176 did not return safely and an
------------------------------
214
Abu2 ’l-Fı3da, Mukhtas5a r, A.H. 585 (p. 62); Andrea Dandolo, Chronica, X.ii (p.
270); Continuation de Guillaume de Tyr, §§82-124, esp. §120 (pp. 89-127, esp. pp.
121-3); Ibn al-‘Adı3m , Zubdat al-h5a lab (Blochet), pp. 197-9; Ibn al-Athı3r, Al-Ka2mil
(RHCHOr), A.H. 585-7 (pp. 7-48, esp pp. 16, 21-2, 29, 32-3, 43); Ibn Shadda2d, Al-
Nawadir al-sult6a2niyya, pp. 98-161, esp. pp. 112-13, 123-4, 126-7, 141-2, 151; ‘Ima2d
al-Dı3n , Al-fath5 al-qussı3, A.H 585-7 (pp. 168-326, esp. pp. 170-71, 175, 190, 193, 197-
201, 210, 224-6, 246-9, 253-5, 273-4, 276, 279, 280-81, 289-92, 297-9, 320-21, 324);
Itinerarium peregrinorum (Mayer), pp. 304-57, esp. pp. 319-24, 327, 329, 345-6, 348;
Itinerarium peregrinorum (Stubbs), I.x-III.xviii, esp. I.x, xiv, xxv-vi, xxxiii-iv,
xxxvii, xxxix, lx, II.xlii (pp. 23-234, esp. pp. 23-5, 27-8, 59-62, 77-81, 86, 88, 114-15,
204-9); Al-Maqrı3zı3, Sulu2k (Broadhurst), A.H. 585, 586-7 (pp. 90, 91-3); Ottobono
Scriba, Annales Ianuenses, pp. 32-6; Sa2wı3ris, History of the Patriarchs (Khater &
Burmester), vol. 3, part 2, pp. 144-56; Tafel and Thomas, Urkunden, No 73 (vol. 1, pp.
204-6). See also Lev, Saladin in Egypt, pp. 172-5.
215
Abu2 Sha2m a, Kita 2b al-rawd5atayn, A.H. 590 (vol. 5, pp. 152-3); Eracles,
XXVIII.vii (vol. 2, p. 258); Ernoul, c. XXXII (pp. 354-5); Al-Maqrı3zı3, Sulu2k
(Broadhurst), A.H. 587, 590, 592, 600 (pp. 95, 107, 111, 123, 146); Sa2wı3ris, History
of the Patriarchs (Khater & Burmester), vol. 3, part 2, p. 171. See also Pryor,
“Venetian fleet”.
anonymous rhetor claimed that at Manuel’s death piracy had been

suppressed and maritime routes opened to commerce and navigation.
However, medieval technology for protecting ships from rot and ship-
worm was not particularly effective and neglect would lead quickly to
the material decay of fleets. War fleets also depended on skilled
seamen, who could not be produced overnight. What crews hardened
by years at the oars could achieve by comparison to raw recruits was
immeasurable. A few years of neglect could lead quickly to total
collapse. It is no wonder that it took the Komne2noi emperors almost
80 years to rebuild Byzantine fleets into naval forces capable of taking
on the best that the Mediterranean had to offer. However, the results
of all their efforts were lost in the next twenty.216
As early as 1182 the pro2tosebastos Alexios Komne2nos had to turn
to Latin mercenaries to man some of the trie2reis despatched to face
the revolt of Andronikos Komne2nos, Nike2tas Cho2niate2s commenting
that their quality was superior to that of Byzantine crews. However, in
1185 there were still sufficient Byzantine squadrons for Andronikos to
prepare 100 ploia makra to aid cities threatened by the Sicilian attack,
to block access to the Golden Horn, and for the fleet to later engage
the Sicilians in the Gulf of Astakos and force them to withdraw out
into the Aegean. After the defeat of the Norman expedition, an
alliance was reached between the Empire and Sicily under the terms
of which Sicily was required to furnish a fleet for the Empire if
requested. Together with a similar clause in Isaac II Angelos’s later
treaty with Venice, it suggests that a decline in Byzantine naval forces
was becoming felt keenly, a decline which became reflected in the
impunity with which Western fleets and corsairs began to operate in
the Aegean and the Sea of Marmara.217
According to Nike2tas Cho2niate2s, one of the main contributors to the
decay of Byzantine naval strength was the megas doux Michael
Stryphnos, who during the 1190s sold off spikes and anchors, ropes
and sails, emptying the arsenals of ploia makra to line his own
pockets. The Byzantines became incapable of controlling even their
own waters. In 1196 war broke out between Pisa and Venice in
Romania and a Venetian fleet penetrated the Dardanelles as far as
------------------------------
216
Anonymous Rhetor, Laudatio funebris Manuelis imperatoris, p. 195.
217
Annales Ceccanenses, 1185 (p. 287); George Tornikios, “Oratio ad Isaacium
Angelum imperatorem”, pp. 277-8; Nike2tas Cho2niate2s, Historia, Basileiva Alexivo u
tou' Komnhnou', Basileiva Andronivkou tou' Komnhnou' BV, Basivleiva Isaavkiou tou'
Aggelou' AV (pp. 246-50, 320, 362-3); Sergios Kolybas, “Oratio ad Isaacium Angelum
imperatorem”, pp. 289-90.
122 CHAPTER ONE
Abydos. Although ordered home by Doge Enrico Dandolo, its

commanders stayed put and its presence probably persuaded Alexios
III Angelos to come to terms with Venice. Then between 1197 and
1199 a Genoese squadron under the command of a corsair named
Gafforio attacked the Aegean coasts and islands and was overcome
only by another Western corsair in Byzantine service, John Steiriones
from Calabria.218 The operations of these Western corsairs and fleets
in the Aegean and Sea of Marmara indicate clearly the degree to
which the naval forces of the Empire laboriously built up by the
Komne2noi emperors were allowed to go to wrack and ruin in only a
few years. Nike2tas Cho2niate2s is not to be taken literally because he
was searching for an explanation of why civilization as he knew it had
been eclipsed, nevertheless he wrote that only 20 decaying, worm-
eaten, and unseaworthy little skiffs, he used the diminutive and
pejorative term skaphidia, could be found in the Golden Horn in 1203
to oppose the Venetian fleet of the Fourth Crusade. That this was
perhaps an exaggeration is suggested by the fact that the Venetians did
not attempt to attack the Byzantine ships ranged behind the chain
drawn across the mouth of the Golden Horn to defend it, but rather the
Crusaders chose to attack the Tower of Galata on the northern side of
the Golden Horn, where the chain came ashore, and thus to break the
chain that way. Nevertheless, it is clear from the way that the
Venetians quickly overcome all resistance in the Golden Horn once
the chain was broken, that what remained of Byzantine naval forces
were by 1203 no match for the Venetian battle fleet of 50 galeae. The
days of glory of the Byzantine navy were over forever, as were those
of the dromon.
------------------------------
218
Genoa, Codice diplomatico, vol. 3, No 40 (pp. 112-15); Michael Cho2niate2s, Ta
So2zomena, vol. 2, pp 105-7; Miklosich and Müller, Acta et diplomata, vol. 3, N o 8
(pp. 46-7); Nike2tas Cho2niate2s, Historia, Basileiva Alexivo u tou' Aggelou' AV, BV (pp.
481-3, 491, 540-44); Tafel and Thomas, Urkunden, N o 78 (vol. 1, pp. 216-25).
CHAPTER TWO
THE ORIGINS OF THE DROMON
First mentions
To the best of our knowledge, the earliest reference to a galley by a

form of the word dromo2n occurs in a fragment of a work of an
unknown Greco-Roman author, possibly the History of Eunapios of
Sardis (345/6-post 414 C.E.).1 This text, once attributed to Eunapios,
mentioned either “thirty-oared dromades in the form of liburnae”, or
“swift triakonte2reis in the form of liburnae”, or “dromades, thirty-
oared ships, in the form of liburnae”. We prefer the first reading but
all three are possible.2
It is possible that the earliest mentions of galleys explicitly called
dromo2nes are in charters from Ravenna dated to the late fifth century,3
------------------------------
1
Interpretation of the word drovmwno["] (dro?m?o2no[s?]) found totally without
context in a papyrus fragment of unknown provenance dated to 126 C.E. as referring
to a ship of a type known as dromo2n, as suggested “probably” in the 1996 revised
supplement of Liddell and Scott, Greek-English lexicon, is in fact highly improbable.
See Kiessling, Sammelbuch, §9855 (p. 125). See also below pp. 164-5 & nn. 8, 9.
2
In the anonymous lexicon compiled in the circle of the tenth-century Byzantine
encyclopedic movement and known as Souda, the following citation is given at
Lambda, §490: “Livberna : ei\do" ploivou. karavbia. phxavmeno" dromavda" triakonthvrei"
Libernivdwn tuvpw/”. See Souda, L.490 (vol. 3, p. 267).
This text was attributed to Eunapios by Boissonade. See Eunapios, History
(Boissonade), Eunapii fragmenta ex Suida, §44 (vol. 1, p. 525). The attribution was
accepted by the nineteenth-century editors of Eunapios such as Niebuhr. See,
Eunapios, History (Bekker), p. 115. However, it is entirely speculative. The Souda did
not specify the source of the quotation. The latest editor of Eunapios, Blockley, has
apparently rejected the attribution since he does not include the text. See Eunapios,
History (Blockley). All that may be said is that some late Greco-Roman source used
by the Souda had employed this clause.
The ambiguity in meaning is created by the fact that dromavda" and
triakonthvrei" are both accusative plurals and both may be either nouns or adjectives.
Since there is no context in which to understand the meaning of the clause, either
word may qualify the other or both may have been nouns used in apposition.
The attempt by Clover to attribute the text to Eunapios and a non-surviving
source of his is interesting but ultimately unconvincing. See Clover, “Count Gaïnas”,
pp. 65-8.
3
See Ahrweiler, Byzance et la mer, p. 411 and n. 1. Ahrweiler cites Serre,
Marines de guerre, vol. 3, p. 24. However, all attempts to trace a third volume of
admiral Serre’s work in all major libraries around the world have failed. It does not
appear to exist and this reference cannot be found in volumes one and two of Serre’s
work. A search of the Ravenna papyri in Tjäder has not found any charters containing
references to dromons as such; however, see n. 8 below. We cannot assert
124 CHAPTER TWO
but it is only from the sixth century that the sources which referred to
them by this name really proliferated rapidly; for example, the
chronicle of Marcellinus (ca 518),4 the Emperor Justinian I in a
rescript of 534,5 Cassiodorus in his Variae (537-8),6 and John Lydos
in his On the magistracies (ca 551-65).7 Dromonarii, crews manning
the dromons of the fleet, and a praepositus or commander of the
dromunarii, are attested to in rescripts of the Ostrogothic king
Theodoric the Great dated to 507-11 and in a charter from Ravenna
dated to 539. It is clear that by this time squadrons of dromons must
have been stationed at Ravenna and there may well have been others
elsewhere in northern Italy even earlier. A sixth-century epitaph from
the church of St Saturninus at Cagliari refers to a certain Gaudiosus
who was probably a dromonarius and who died aged around the age
of 24 on 17 July in a year which was the first indiction.8
------------------------------
categorically that this reference to dromons at Ravenna in the fifth century does not
exist, but our best efforts to verify it have failed.
4
Marcellinus Comes, Chronicon, Annus 508 (pp. 34-5): “Romanus comes
domesticorum et Rusticus comes scholariorum cum centum armatis navibus
totidemque dromonibus octo milia militum armatorum secum ferentibus ad
devastanda Italiae litora processerunt ...”. This section of the chronicle was written ca
518, even though Marcellinus continued it later up to 534. See Croke, Count
Marcellinus, pp. 20-35.
5
CI, I.27.2.§2: “Iubemus etiam, ut in traiectu, qui est contra Hispaniam, quod
Septem dicitur, quantos providerit tua magnitudo, de militibus una cum tribuno suo, ...
constituas, ... In quo traiectu etiam dromones, quantos provideris, ordinari facias.”.
6
Cassiodorus, Variae, V.16 (p. 195): “..., deo nobis inspirante decreuimus mille
interim dromones fabricandos assumere, qui et frumenta publica possint conuehere et
adversis nauibus, si necesse fuerit, obuiare.”. Cf. also V.17, 18, 20 (pp. 196-7, 197-8,
198-9).
These letters were drafted by Cassiodorus on behalf of Theodoric the Great
between 523 and 526. The first two were addressed to the praetorian prefect
Abundantius and the last two to the Count of the Patrimony Vvilia and to the saio
Aliulfus. On these letters and their historical context see above, pp. 13-14.
7
John Lydos, On powers, pt. II, §14 (p. 106): “..., ejkei'no prolevgwn w{" eijsin e[t i
kai; nu'n porqmivde" trei'" th'/ ajrch'/ pro;" ta;" ajntipovrqmou" diaperaiwvsei" ejk th'"
basilivdo" ejpi; ta;" geivtona" hjpeivrou". bavrka" aujtav", ajnti; tou' drovmwna", patrivw"
ejkavlesan oiJ palaiovteroi kai; kevlwka", oi|on tacinav", o{t i kevl er kat aujtou;" oJ tacu;"
levgetai, kai; sarkinariva", ajnti; tou' oJlkavda", o{ti savrkina kat aujtou" to; a[cqo"
kalei'tai.”. Note, however, that in part III, §43 (pp. 200-201), John Lydos described
the fleet sent against the Vandal king Gaiseric in Africa by Leo I and Anthemios in
468 C.E. as being composed of 10,000 liburnae (libuvr nai), completely impossible of
course.
8
A rescript of Theodoric reproduced in Cassiodorus’s Variae, was addressed to
the dromonarii of the river Po. See Cassiodorus, Variae, II.31 (p. 79). A second, also
dated to 507-11, mentioned 21 dromonarii from some unspecified location. See IV.15
(p. 152): “Illustris et magnifici viri comitis patrimonii suggestione comperimus
dromonarios viginti et unum de constituto numero mortis incommodo fuisse
subtractos.”. Tjäder, Nichtliterarischen lateinischen papyri, vol. 2, Pap. 30 (p. 58):
“...: Casanovam, iuris quond(am) Secund[i] [drom]onarii, ... [et] fundum Kalegaricus
iuris quond(am) Andreatis b(onae) m(emoriae), pra[epo]siti dromunariorum, ...”.
ORIGINS OF THE DROMON 125
In his Chronicle (expl. 563), John Malalas mentioned dromons four

times,9 and at book 9, §10, which was in a part of his chronicle
composed at Antioch in the early 530s, in discussing Marcus
Antonius’s preparation of the fleet to engage Octavian at Actium, he
wrote that: “..., he also built many dromo2n ships and war liburna ... ”.
Malalas appears to have identified galleys that he himself knew as
dromo2nes in his own time with those that he knew the Romans had
called liburnae in the first century B.C.E. and to have equated the two
terms for galleys in his own mind.10
There can be little doubt that the word dromo2n became used for
some war galleys, or perhaps rather for some specific type of war
galley, because these galleys were unusually fast, faster than the
standard Roman liburnae war galleys of the late Empire, which had
been developed by the Romans from the ships of the Illyrian people
known as the Liburni during the first century B.C.E.11 The word
dromo2n was derived from the Greek “drovmo"” (dromos), meaning a
“race”, and the root “drom-(avw)” (drom-ao2), meaning “run”.12 Writing
in the 550s, Prokopios of Caesarea, who accompanied Belisarios as
his secretary on the Byzantine expedition sent to Africa against the
Vandal king Gelimer in 533, was clear about this. In his History of the
Wars, he wrote that the dromons of this expedition were capable of
great speed.
And they also had ships of war [long ships] prepared as for sea-fighting,
------------------------------
On the dromonarius named Gaudiosus, see Cosentino, “Epitafio sardo”.
Indictions were 15-year taxation cycles instituted from 312 C.E. The first indiction
was the first year of any cycle.
9
John Malalas, Chronographia, QV [9].10 (p. 166): “... poihvsa" ploi'a dromwvnwn
pollw'n, kai; livburna polemikav ...”; IAV [11].3 (p. 205): “oJ de; basileu;" Trai>ano;" h]
movnon katevfqase tw'/ drovmwni ...”; IıV [16].16 (p. 331): “..., lambavnonta" ploi'a
dromwvnwn kai; stratiwvta". ... kai; ajganakthvsa" kat aujt w'n oJ basileu;" e[balen aujtou;"
e[xw tou' palativou, kai; keleuvsa" Marivnw/ tw'/ Suvr w/ labei'n tou;" drovmwna" ...”; IHV
[18].90 (p. 407): “oJ de; aujto;" basileu;" pevmya" Narsh'n to;n koubikoulavrion meta;
dromwvnwn ...”.
10
On Malalas, QV [9].10 (p. 166) see also John Malalas, Chronographia, trans.
Jeffreys et al., pp. xxiii and 116.
11
See Vegetius, Epitoma, IV.33 (p. 151).
Quite remarkably, in spite of the fact that they were the Empire’s major
warships for four centuries or more, perhaps even less is known about Roman
liburnae than about the dromons which succeeded them. The latest study is
Höckmann, “Liburnian”. See also Morrison, Greek and Roman oared warships, pp.
131, 165, 170-5, 253, 264, 316-7; Reddé, Mare nostrum, pp. 104-10.
12
It was once argued that the word was derived from the Gothic word droma,
meaning to go slowly. However, the etymology is extremely doubtful and it does not
accord with what we know of the ships in any case. See Jal, Archéologie navale, p.
230.
126 CHAPTER TWO
to the number of ninety-two, and they were single-banked ships covered

by decks [ojrofa;", orophas], in order that the men rowing them might if
possible not be exposed to the bolts of the enemy. Such ships are called
dromo2nes by those of the present time; for they are able to attain a great
speed. In these sailed two thousand men of Byzantium, who were all
rowers as well as fighting men; for there was not a single superfluous man
among them/in them.13
In addition to the testimony of Prokopios, there is also that of St

Isidore of Seville who wrote in his Etymologiae that: “The dromon is
so called from ‘running down’; for the Greeks call ‘running’
drovmon”.14
This reference to “speed” may have referred to any one or more of
a number of quite different things in the context of war galleys. On the
one hand, it may have referred to overall or general speed, to an
ability to outpace other galleys over long distances. On the other hand,
it may have referred rather to short-term sprint speed in battle.
Alternatively, it may have referred to manœuvrability, which would
also translate in practice into “speed” in battle.15 Whatever may have
been these alternative possibilities for the type of “speed” to which
Prokopios and St Isidore referred, and the possible technological
------------------------------
13
Prokopios, History of the wars, III.xi.15-16 (vol. 2, p. 104): “h\san de; aujtoi'" kai;
ploi'a makrav, wJ" ej" naumaciva n paraskeuasmevna, ejnenhvkonta duvo, monhvrh mevntoi kai;
ojrofa;" u{perqen e[conta, o{pw" oiJ tau'ta ejr evssonte" pro;" tw'n polemivwn h{kista
bavllointo. drovmwna" kalou'si ta; ploi'a tau'ta oiJ nu'n a[nqrwpoi: plei'n ga;r kata; tavco"
duvnantai mavlista. ejn touvtoi" dh; Buzavntioi discivlioi e[pleon, aujterevtai pavnte":
perivnew" ga;r h\n ejn touvtoi" oujdeiv".”. The last touvtoi" may refer to either the men or
the ships; although, most probably to the men. Thus the final clause probably meaned
that there was not a superflous man among the 2,000; although, it is possible that it
meaned that there was not a superflous man in the ships, which does not necessarily
amount to the same thing.
Cf. Theophane2s, Chronographia, A.M. 2026 (vol. 1, p. 189): “... ajrchgo;" de; ei|"
ejpi; tai'" nausi; Kalwvnumo" Alexandreuv". h\san de; kai; drovmwne" dia; naumaciva n
ejnenhvkonta.”.
14
Isidore of Seville, Etymologiae, XIX.i.14: “Longae naves sunt quas dromones
vocamus, dictae eo quod longiores sint ceteris: cuius contrarius musculus, curtum
navigium. Dromo autem a decurrendo dictus; cursum enim Graeci drovmon vocant.”.
Elsewhere we have pointed out that Isidore apparently knew very little about
Roman war galleys. See below pp. 128, 134-5. It is therefore quite possible that he
was merely writing philologically, drawing an explanation of the Latin word dromon
from his knowledge of Greek. Consequently, whether he can really be considered as
an independent witness to the meaning of the word, or more importantly to whether or
not any real ships of his own age which he knew as dromons were unusually fast, is
arguable.
15
The Latin mobilitas could also mean either “speed” or “manœuvrability” in the
context of a ship. See Pryor, “Rutilius Namatianus”, pp. 272-3.
reasons for them,16 there can be little doubt that it was because these
galleys which were becoming referred to as dromo2nes had these
particular qualities that the term became applied to them. It is perhaps
significant that when Prokopios discussed in an earlier part of his
History of the wars the expedition of Flavius Basiliskos to Africa in
468 he did not use the term dromo2nes for the ships of Flavius’s fleet,
but rather the conventional terms for ships, nau'" (naus) and ploi'on
(ploion).17
Meagre though it is, the evidence suggests that the early use of the
term dromo2n was philological rather than technological in its import.
As the apparent identification of liburnae and dromons by pseudo-
Eunapios and John Malalas indicates, these war galleys which were
becoming referred to as dromo2nes by the sixth century were almost
certainly the product of a gradual evolution of Roman liburnae over a
considerable period of time rather than of some dramatic and sudden
“invention” of a new design. Some liburnae may have become called
dromo2nes at some point in time simply because they were particularly
fast rather than because, as yet, they had significantly different design
characteristics. If this was the case, however, we are left with the
question of what it was that gave such liburnae this extra speed? Only
when evidence for significant design changes begins to occur can we
begin to be confident that technological evolution had finally
produced a ship type which was qualitatively different from its
predecessors and distinctive.
Evidence for changes in the design characteristics of war galleys
during the Late Roman Empire suggests that evolution in three key
areas eventually led to these new galley types becoming distinguished
from liburnae. And, since no other new term for war galleys was
coined and became widely used in the period, we may reasonably
connect the use of the term dromo2n to the evidence for these changes
in design. This evidence associates the term in the first case with
smaller galleys which had only 50 oarsmen but which were fully-
decked and were therefore distinguished from half-decked bireme and
trireme liburnae.18 Secondly, there is evidence for the replacement of
the classical waterline rams by abovewater spurs and for changes in
hull design and construction at the bow in particular consequent upon
that. Finally, there is evidence for the replacement of the traditional
square sails of antiquity by lateen sails. Not all of these changes may
------------------------------
16
See also below pp. 139, 143.
17
Prokopios, History of the wars, III.vi.5-27 (vol. 2, pp. 56-63).
18
See Höckmann, “Liburnian”, pp. 196-7.
128 CHAPTER TWO
have been necessary before the term dromo2n became applied to such
galleys. Nor is it necessarily the case that all galleys becoming known
as dromo2nes had the same design characteristics. For example,
continued use of square sails may well have co-existed side by side
with increased use of lateeen sails. Such changes obviously occurred
slowly and progressively over time and it will always be impossible to
know to what particular point of technological change the use of the
term dromo2n corresponded, if, indeed, there ever was such a single
point of correspondence. More probably there was not and even to put
the question in such terms is inappropriate to what were slow and
progressive evolutions in both technology and terminology.
Evidence for discontinuation of the use of the word liburna, or for
misunderstanding of what it had once meant, may also be revealing.
Although the late-Roman writer on military tactics Publius Vegetius
Renatus wrote that in his own day, which was the second quarter of
the fifth century, warships of the Empire were still known as liburnae,
and although the word libevrno" (libernos) for a ship is found as late
as the fifth and sixth centuries in the Oxyrhynchus papyri from
Egypt,19 by the 630s St Isidore of Seville no longer understood either
the meaning of the word or its etymology. He thought that it was
derived from “Libya” and that liburnae were merchant ships.20 It is
significant that use of the word was discontinued and knowledge of its
meaning was lost in the same chronological period in which use of the
word dromo2n began.
Deck and oarage system
As Prokopios presented them, early dromons were monoreme galleys

with a full deck to protect the oarsmen beneath it. In classical Greek
the word katastro2ma had been used for the lateral part-decks which
trie2reis and other polyremes had had along both sides.21 It was used in
conscious contradistinction to other words for the decks of ships such
as i[kria (ikria), especially a half-deck at the bow or stern of an
otherwise open boat, stevgh or stegov" (stege2 or stegos), lit. “roof”, and
the group of words saniv" (sanis), sanivdion (sanidion), and sanivdwma
------------------------------
19
See Vegetius, Epitoma, IV.33 (p. 151); Grenfell, Oxyrhynchus papyri, N os
2032.52 & 54 (p. 255) (sixth century) and 2042.11 (p. 264) (fifth century).
20
Isidore of Seville, Etymologiae, XIX.i.12: “Liburnae dictae a Libyis; naves enim
sunt negotiatorum.”.
21
See Morrison, et al., Athenian trireme, esp. pp. 158-61.
(sanido2ma), literally “planking”. Terminological usage was not

inflexible but it is clear that katastro2ma did have this specialized
meaning. When the Romans developed liburnae as their own pre-
eminent galleys in the period from ca 50 B.C.E., they are thought to
have modelled them on Hellenistic galleys and to have built them with
the same lateral part-decks; although, some Roman triremes had full
decks and some liburnae may also have been given full decks.22 But,
curiously, the Romans did not take the word katastro2ma into Latin,
and they used a variety of words for the decks of ships, none of which
seem to have had a meaning confined to the idea of a katastro2ma or
even to war galleys in particular: pons, forus, constratum, stega (from
stevgh), tecta (past participle, “decked”).23 If Roman liburnae really
did have lateral part-decks the linguistic evidence for them is
unknown. No Latin text known to us associates any particular word
with what were clearly lateral part-decks and the pictorial evidence
does not help since pictures of ships were invariably drawn from the
side rather than from a bird’s-eye view. No picture known to us shows
part-decks.
Irrespective of what types of decks liburnae of the Roman Empire
may have had, the salient innovation to which Prokopios appears to
have pointed was to give monoremes, many of which were certainly
undecked in prior centuries, a full deck. Prokopios was employing
literary, rather than technical, language and he used the word orophe2,
literally a “roof” or “ceiling”, rather than katastro2ma. But he probably
chose a word for a complete covering because these were full-decks
rather than part-decks. He made it clear that the purpose of the deck
was to provide greater protection for the oarsmen from incoming
missiles than the construction of traditional galleys gave and this
implies something different from katastro2mata; surely, full decks. By
the tenth century dromons were certainly fully decked.
Prokopios’s reference to dromons having full decks was probably
predated by around thirty years in the second letter of Theodoric the
Great to the praetorian prefect Abundantius, penned by Cassiodorus
------------------------------
22
See Höckmann, “Liburnian”, pp. 196-7; Casson, Ships and seamanship, pp. 123-
4, 141-6, 178-9; Morrison, Greek and Roman oared warships, pp. 170, 264. See
below pp. 231-2.
23
See, for example, Tacitus, Annals, II.6 (vol. 3, p. 392): “... multae pontibus
stratae super quas tormenta veherentur ...”; Lucan, Civil War, III.630 (p. 160): “Et,
postquam ruptis pelagus conpagibus hausit, / Ad summos repleta foros descendit in
undas.”; Caesar, Civil wars, I.56 (p. 78): “Dum haec Ilerdam geruntur, Massilienses
usi L. Domitii consilio naves longas expediunt numero XVII, quarum erant XI
tectae.”.
130 CHAPTER TWO
between 523 and 526. In this letter the king, or rather Cassiodorus,
congratulated Abundantius on having completed his task of
constructing a fleet of dromons in quick time, almost as speedily as
they were customarily sailed. He described the dromon as: “..., a
‘trireme’ conveyance carrying a great many oars but carefully
concealing the form of the men”.24 Where else could oarsmen have
been concealed but below deck?
Being monoremes, Prokopios’s dromons must have been smaller
than their Greco-Roman bireme and trireme predecessors. If we can
believe Zo2simos, monoremes had prevailed over biremes and triremes
in the victory of the fleet of Constantine I over that of Licinius at the
battle of the Dardanelles in 324. According to him, Constantine’s fleet
of small monoreme triakontoroi defeated that of Licinius which
supposedly included 160 trie2reis from Egypt.25 However, whether the
victory was really due to any superiority of monoremes over biremes
and triremes in the context of changing conditions of naval warfare is
debatable. Leadership and tactics may also have been important and
there is insufficient corroborating evidence from elsewhere to reach
any definite conclusion. And, whether the use of the word dromo2n
was confined to monoremes alone in the sixth century is also
arguable. By the tenth century there were certainly bireme galleys
which were also called dromo2nes and this may have been the case as
early as the late eighth century. It is possible that Prokopios was
referring to only one class of dromons and that the term had become
applied to galleys distinguished from liburnae because of their speed
and perhaps other design characteristics in addition to the deck,
irrespective of whether they were monoremes or not. The reference to
“speed” in the “racer” etymology of dromo2n suggests that this may
------------------------------
24
Cassiodorus, Variae, V.17 (p. 196): “Renuntias ilico completum, quod uix credi
poterat inchoatum, ut paene quanta uelocitate nauigari solet constructio nauium, tanta
sit celeritate completa. ... , trireme uehiculum remorum tantum numerum prodens, sed
hominum facies diligenter abscondens.”. The adjective trireme need not be read
literally. The phrase reeks of a rhetorical re-writing by Cassiodorus when later
compiling his Variae.
25
Zo2simos, Historia nova, 2.22 (pp. 78-9): “kai; triakovntoroi me;n eij" diakosiva "
kateskeuavsqhsan, nau'" de; fortivd e" sunhvcqhsan plevon h] discivliai, ... Likivnnio" de;
Kwnstanti'non ajkouvsa" ejn paraskeuai'" ei\nai, dievp empen ajggevl ou" kata; ta; e[qnh
ploi'a polemika; kai; dunavmei" pezav" te kai; iJppika;" eujtrepei'" poih'sai keleuvwn. kai;
su;n panti; tavcei trihvrei" ejxevp empovn oiJ Aijguvptioi me;n ojgdohvkonta, Foivnike" de; ta;"
i[sa", ...”; 2.23 (p. 80): “ajf ikomevnou de; tou' stovlou kata; to; prostacqevn, oiJ me;n
Kwstantivnou strathgoi; movnai" ojgdohvkonta triakontovroi" tai'" a[rista pleouvsai"
e[gnwsan naumacei'n oi|a tou' tovpou dia; th;n stenovthta plhvqei new'n oujk o[nto"
ejpithdeivou, ...”. A triakovntoro" (triakontoros) was a thirty-oared galley, very small,
and certainly a monoreme.
have been the case since small ships are never faster than larger ones
unless they have different design characteristics which cause this to be
so. Large yachts are always faster than smaller ones with the same
design because they can carry more sail per unit of wetted hull
surface. All other factors being equal, rowing fours will outpace eights
over short distances but will fall behind over the long haul.26 This
consideration adds weight to other evidence, especially that which can
be extrapolated from the replacement of the ram by the spur, which
suggests that dromons may have had new hull design characteristics,
and that use of the term may therefore not necessarily have been
confined to monoremes. When considered together with the general
observation that ship types evolve over time in any case, it would also
help to explain how biremes as well as monoremes could have been
referred to as dromo2nes by the tenth century.
How large were Belisarios’s dromons? Supposedly, Prokopios
wrote that the 92 dromons of the fleet were manned by 2,000 men,
although this figure was no doubt an approximation.27 The word used
for the crews, aujterevtai (auteretai), referred to men who were both
oarsmen and marines at the same time. Prokopios also used it with the
same sense in reference to the ships of the people he referred to as the
“Angili” of the island of “Brittia”,28 and it is reasonably well attested
with this meaning, particularly by Thucydides but also by Heliodo2ros
of Emesa of the second-fourth centuries, Longus of the late second to
early third centuries, and Philostratos the elder in the third century. It
was referenced by Hesychios in the fifth-sixth centuries and remained
known with this meaning into the Middle Byzantine period in the
Souda and by Pho2tios.29 There are two possibilities. Either the 2,000
------------------------------
26
At the Olympic Games, rowing over courses of 2,000 metres, gold-medal
winning eights habitually outperform coxed fours by between around 25 and 40
seconds, or about 7-12%.
27
We say “supposedly”, because the earliest extant manuscripts from which the
received edition has been compiled date from the fourteenth century and it is
impossible to know what violence may have been done to the figures in the
intervening eight centuries. See Prokopios, Opera omnia, vol. 1, pp, xxviii-liv. An
earlier, thirteenth-century manuscript has been discovered recently but is unpublished.
Given the fact that 2,000 oarsmen for 92 dromons appears to be a figure inexplicably
low, it is possible that in the manuscript transmission process between the sixth and
fourteenth centuries the figure was mistranscribed and corrupted at some point.
28
Prokopios, History of the wars, VIII.20.31 (vol. 5, p. 260): “perivnew" de; oujk h\n
ejn touvtw/ tw'/ stovlw/, ajll aujterevtai pavnte".”.
29
It does not necessarily imply that Prokopios’s evidence on this point is not to be
trusted, nor that he was not saying something about the real characteristics of the
dromons of Belisarios’s fleet; nevertheless, it seems highly probable that his choice of
language here was inspired by the passage of Thucydides referring to the composition
of the fleet of Philokte2te2s for the Trojan War. See Thucydides, Peloponnesian war,
132 CHAPTER TWO
auteretai were the oarsmen of the dromons and they fought as well as
rowed. Or, alternatively, these men were marines in addition to the
normal complements of oarsmen and they doubled as oarsmen when
necessary. On the one hand, the figure of approximately 22 oarsmen
per dromon which the first possibility gives is not so far removed
from the supposed 30 oarsmen of the triakontoroi of Constantine’s
fleet at the battle of the Dardanelles as to rule it out. Moreover, it did
become normal in the Middle Ages for galley oarsmen to also fight in
battle once the galleys became locked together. They were always
armed, at least lightly. On the other hand, it is very difficult to accept
that any serious warship could have only eleven or so pairs of oars.
Such a ship would have been a mere long boat.30 What could they
have been intended to be used for? Surely not to engage the Vandal
fleet. This being said, all 92 dromons need not necessarily have been
of the same size and any estimate of the size of the dromons on the
basis of the crews is therefore not be possible for that reason alone.
As we have seen, Cassiodorus wrote that dromons had a great
many oars. This alone suggests that they must have had more than a
mere eleven or so pairs of them. However, in addition to that, one
further consideration above all leads us to conclude that these 2,000
men were in reality marines in addition to the normal oarsmen: the
fact that the dromons were fully decked with the explicit purpose of
protecting the oarsmen from missiles. Obviously, there could be no
incoming missiles until after battle had been joined and, since
Prokopios wrote that the oarsmen continued rowing after that, they
therefore cannot have fought as marines. In any case, how could
oarsmen rowing below deck be called upon to then engage in battle as
marines? They would have had to scramble up on deck through
hatches, leaving the ships powerless in order to do so. Both the
leaving of the ships without power and also the time elapsed before
they could take up their arms and join battle would be unacceptable
from the point of view of both manœuvrability in battle and advantage
------------------------------
I.x.4 (vol. 1, p. 20): “aujterevtai de; o{t i h\san kai; mavcimoi pavnte", ejn tai'" Filokthvtou
nausi; dedhvlwken: toxovta" ga;r pavnta" pepoivhke tou;" proskwvpou". perivnew" de; oujk
eijko;" pollou;" xumplei'n ...”. See also II.18.4 & VI.91.4, where the word was also
used. The word was scholiated in the tenth-century Patmos manuscript at I.x.4 as:
“aujt erevtaiÚ oiJ aujtoi; ejrevtai kai; stratiw'tai”. See Hude, Scholia, p. 14. See also
Heliodo2ros, Aithiopika, BV.ii.2 (vol. 1, p. 49); Longus, Daphnis and Chloe, II.20 (p.
94); Philostratos, Eijkovne" A, I.12.(1) (p. 50); Hesychios, Lexicon (Schmidt), A.8385
(vol. 1, p. 325); Souda, A.4468 (vol. 1, p. 418); Pho2tios, Lexicon (Theodoridis),
A.3202 (p. 295).
30
In the thirteenth century the ships’ boats of large sailing ships rowed from 32 to
52 oars. See Pryor, “Naval architecture”, pp. 372-3.
in initial engagement. We conclude that the 2,000 auteretai were

marines supernumerary to the oarsmen and that this figure tells us
nothing about the size of the dromons of Belisarios’ fleet or of the
number of oarsmen aboard them.
What is possibly the earliest evidence for the size of dromons
occurs in the Histories of Theophylaktos Simokatte2s, which were
written in the late 620s. Theophylaktos referred twice to dromo2nes,
once to fast-sailing ships, tacunautouvsa" oJl kavda" (tachynautousas
olkadas), commonly known as dromo2nes, in use by the strate2gos
Priskos on the Danube against the Avars at Belgrade in 595 and once
to the Emperor Maurice fleeing Constantinople on a dromon in 602.31
However, in another section, he used almost the same term for “fast
sailing ships”, tacunautou'sai nh'e" (tachynautousai ne2es), for those
on which Maurice took ship at Se2lymbria for He2rakleia in 590, and he
then said that the one on which he was travelling was a pente2konteros;
that is, it supposedly had fifty oars.32 Assuming that Theophylaktos
was not merely aping Herodotos or Thucydides by using the classical
pente2konteros simply as an approved word for a war galley, he most
probably also meant that these “fast sailing” pente2kontoroi were of the
type commonly known as dromo2nes. Since, as we shall see, in the
tenth century the Theophane2s continuatus said that monoreme
dromons had fifty oars and the emperor Leo VI (886-912) that bireme
dromons had fifty oars in the lower bank,33 the evidence suggests that
the standard size of galleys to which the word dromo2n was applied
was fifty oars by the early seventh century at least, and most probably
in the age of Prokopios as well. In fact, a figure of 50-54 oars for any
two files of oarsmen, whether arranged in superimposed banks or in
some other way, remained the norm for standard galleys of the line of
------------------------------
31
Theophylaktos Simokatte2s, Historiae, 7.10.3 (p. 262): “oJ me;n ou\n Privsko" th'/
nhvvsw/ ta;" dunavmei" peribalwvn, tacunautouvsa" oJlkavda" parasthsavmeno", a}" drovmwna"
ei[wqen ojnomavzein to; plh'qo", ejpi; ta; Kwnstantivola paragivnetai.”; 8.9.7 (p. 300): “... oJ
Maurivkio" ... thvn te uJphrevtin oJlkavda parasthsavmeno" (drovmwna de; tauvthn eijwvqasi ta;
plhvqh ajpokalei'n) kai; crhvmata ej" aujth;n ejmbalw;n a{ma tw'/ gunaivw/ kai; toi'" paisi;n ejp
aujth'" ejpibaivnei, ...”. Cf. Theophane2s, Chronographia, A.M. 6090 (vol. 1, p. 276):
“tou'to maqw;n Privsko" paragivnetai eij" th;n nh'son tou' “Istrou, kai; paralabw;n
drovmwna" paragivnetai pro;" to;n Cagavnon eij" Kwnstantivola, qevlwn aujtw'/ oJmilh'sai.”;
6094 (vol. 1, p. 288): “oJ de; Maurivkio" mesouvsh" th'" nukto;" ajpodusavmeno" th;n
basivleion stolh;n kai; ijdiwtikh;n peribalovmeno" eij" drovmwnav te eijselqw;n meta; th'"
gunaiko;" kai; tw'n tevknwn aujtou' kai; Kwnstantivnou ajpodidravskei.”.
32
Theophylaktos Simokatte2s, Historiae, 6.1.1-2 (p. 220): “Toivnun tacunautou'sai
nh'e" parh'san th'" peri; aujta;" ajrtiva" kataskeuh'" mhde;n pro;" basivleion ejkpomph;n
ajpodevousai. ... to;n de; aujtokravtora a{ma th'/ peri; aujto;n penthkontovrw/ paradovxw/
swthriva/ ejn tw'/ Daonivw/ diaswqh'nai th;n nauagivan ejkklivnanta.”.
33
See also below pp. 190, 283-4.
134 CHAPTER TWO
battle until the end of the thirteenth century. Twenty five to 27 oar
benches for any one file of oarsmen became the norm, almost
certainly determined by the technological limitations imposed by
building ships as long and as narrow as galleys with such a flexible
material as wood.
The ram and the spur
Of the two most salient design characteristics which eventually

distinguished Byzantine galleys from their Roman predecessors, the
first was the replacement of the waterline ram, embolos, by an
abovewater spur. The last known use of the classical words embolos,
“a ram”, or emballein, “to ram”, in a context which might possibly
indicate that classical waterline rams were still in use occurred in
Prokopios’s History of the wars. In his account of the battle of
Senogallia in 551, Prokopios referred twice to the Byzantine fleet
ramming the Gothic ships.34 But Prokopios was not present at
Senogallia and he was writing in a classical tradition, using classical
Greek terminology for an imperial and educated audience. His use of
emballein cannot be taken as proof that Byzantine war galleys of the
period, which he referred to at Senogallia by the generic terms for
ships, ploi'a (ploia) and nh'e" (ne2es), rather than as dromo2nes, still had
waterline rams. Later Byzantine authors continued to use the words
embolos and emballein, for example the Anonymous author of the
treatise commissioned by the patrikios and parakoimo2menos Basil;35
however, it is clear from other evidence that by his age spurs rather
than rams were in use and that “ramming” meant simply engaging, not
even necessarily by the bow. At §6.2 of his treatise the Anonymous
used the noun embolas to express the ideas contained in the participle
ejpercomevnou" (eperchomenous) used by Leo VI for attacking, but
clearly in the sense of engaging side to side, at §28 of his Naumachika
Leontos Basileo2s.36
Isidore of Seville suggested that by his age, the late sixth and early
seventh centuries, the real function of the ram was no longer
understood in the West. He wrote that: “Ships with rams are so called
------------------------------
34
Prokopios, History of the wars, VIII.xxiii.31 and 34 (vol. 5, pp. 296-8).
35
Appendix Three, §§2.15, 6.2, 7.3. On the spurs of dromons in the tenth century,
see below pp. 203-10. At §2.15 the Anonymous used emboloi for cables used to
control the tillers, oi[ake" (oiakes), of the quarter rudders, revealing clearly that he did
not understand the meaning of the word.
36
See Appendix Three, §6.2 and Appendix Two [a], §28.
since they have at the bow copper rams [rostra] on account of rocks,
so that they [the ships] do not collide [with the rocks] and be
destroyed”.37 It is surprising that someone of Isidore’s erudition had
not learned the real function of the rostrum from the classical sources
that he was familiar with. Nevertheless, he seems to have merely
deduced by the application of logic that the purpose of the Roman
rostrum had been to act as a sort of fender at the bow against rocks,
thus indicating that he had no familiarity with its use as a ram in naval
warfare. The equation between the Greek embolos and the Latin
rostrum continued to be understood but there is no evidence that the
actual functions of the object to which these terms had referred in the
context of a ship remained known.38
The literary sources are inconclusive as to whether or not the
waterline ram had been replaced by the abovewater spur as early as
the sixth century. However, the pictorial evidence suggests that it had
been. The earliest surviving depiction of what may have been a
dromon is an illustration in the Roman Vergil manuscript of the
Aeneid in the Vatican Library, a late fifth-century manuscript whose
provenance was some metropolitan centre in the West.39 It is followed
shortly thereafter by miniatures of what must surely have been many
dromon galleys in the illustrations of the manuscript of the Iliad of
Homer in the Ambrosiana Library (Ilias Ambrosiana), which is dated
to the early sixth century with a provenance in Constantinople.40
These manuscripts show galleys with bow characteristics significantly
different from those of their Greco-Roman predecessors. They are
------------------------------
37
Isidore of Seville, Etymologiae, XIX.i.13: “Rostratae naves vocatae ab eo quod
in fronte rostra aerea habeant propter scopulos, ne feriantur et conlidantur.”.
38
See “Note on citations of Greek and Latin glossaries”, p. lxix above. See the
Greek-Latin Cyril glosses of London, British Library, MS. Harley 5792; Goetz,
Glossarii Latini, vol. 2, p. 376, l. 7: “Newnramfh htoi emboloi : rostra
singularenonhabet [sic]”. The gloss on ejmbavllw in the same manuscript shows that
the author was unaware of what the verb meant in the context of naval warfare. Ibid.,
vol. 2, p. 295, l. 45: “Emba2llw inmitto inicioconicio [sic]”. The same is true of the
Philoxenos Latin-Greek glosses of Paris, Bibliothèque Nationale, MS. Lat. 7651;
ibid., vol. 2, p. 114, l. 30: “Contorquet inmittit etiaculatur emba2llei [sic]” and p. 175,
l. 23: “Rostra embolai2 : rugchxelunia [sic]”. However, the equation of the word
embolos with rostrum remained known in the hermeneumata attributed to Dositheus
in the Hermeneumata Monacensia. See ibid., vol. 3, p. 205, l. 29: “embolon rostrum
nauis”.
39
Rome, Biblioteca Apostolica Vaticana, MS. Vat. Lat. 3867. See Rosenthal,
Vergilius Romanus, plate VIII (p. 52).
40
Milan, Biblioteca Ambrosiana, Cod. Ambros. F. 205 Inf. See Bandinelli,
Hellenistic-Byzantine miniatures, esp. figs 44 (p. 56) [= fig. 96 (pl. 9)], 63 (p.67) [=
fig. 190 (pl. 34)], 74 (p. 73) [= Min. XXXVIII (colour plate III)]. All that survives of
this manuscript are the illustrations.
136 CHAPTER TWO
quite different, for example, to the bows of the many galleys shown in
the Vatican Vergil manuscript of the late fourth c entury.41
Classical Greek trie2reis and other galleys had a straight stempost
with a pronounced waterline ram extending from it which had at least
Figure 1
Liburnae in the Vatican Vergil (Rome, Biblioteca Apostolica Vaticana,
MS. Vat. Lat. 3225, fol. 43v), late fourth century.
two, and perhaps three, horizontal fins and a central vertical post at
the impact zone.42 In the Hellenistic period the head of the stempost
became recurved towards the stern and the ram was now invariably
three-finned.43 The bows of galleys of the navy of the Roman
Republic had this same recurved stempost and three-finned ram and
were no doubt modelled on Hellenistic galleys.44 However, it appears
that during the first century C.E. the Romans abandoned the three-
finned ram and replaced it with a single-pointed one. Some of the
galleys were also given forecastles.45 The many galleys depicted on
Trajan’s column of ca 114 C.E., celebrating the emperor’s Dacian
------------------------------
41
Rome, Biblioteca Apostolica Vaticana, MS. Vat. Lat. 3225. See Stevenson,
Miniature decoration, pictures 17, 21, 25, 28, 29, 30, 39.
42
See Casson, Ships and seamanship, p. 85 and plates 81-2, 84, 88-90.
43
See Casson, Ships and seamanship, p. 117 and plates 107, 109, 110, 116.
44
See Casson, Ships and seamanship, p. 146 and plates 124, 125, 129, 130-32.
45
See Casson, Ships and seamanship, p. 146 and pll. 122-3, 127, 141; Stevenson,
Miniature decoration (as per n. 41 above).
victories, appear to show rams that also curve upwards from the keel
below water to terminate in a single point above water.46
It is possible that this may mean nothing more than that the
sculptors attempted to show the ram while at the same time showing
the water. However, even if this was the case with Trajan’s column,
there is too much evidence for similarly curved rams in some of the
late Roman mosaics to maintain the argument that all we are looking
Figure 2
Liburnae on Trajan’s column, ca 114 C. E.
at is artistic license.47 The mosaic of a galley from the baths at

Themetra, near Hadrumetum in Tunisia, dated to ca 200-220 C.E., is a
good example.48 If rams continued to have the same function that they
had had for centuries, that is to fracture the hulls of enemy ships at the
waterline, why were they apparently begun to be built from this time
------------------------------
46
Lepper and Frere, Trajan’s column, plates 26, 35, 58, 59, 61, 63.
47
See also Höckmann, “Liburnian”, pp. 200-202.
48
Foucher, Navires et barques, fig. 9 (p. 17). See also figs 2 (p. 7) and 12 (p. 21):
galleys on mosaics from the house of L’Oued Blibane, Hadrumetum, ca 190-210, and
the baths at Themetra, ca 200-220, respectively. See also the Dionysios mosaic from
Thugga in Dunbabin, Mosaics of Roman North Africa, pl. 16. One of the galleys in
the great hunting scene mosaic at Piazza Armerina, Sicily, also has a similarly curved
ram. See Figure 36.
138 CHAPTER TWO
commencing at the waterline at the stempost but then curving up to a

point well above the waterline?
The Roman Vergil illustration appears to show bow configurations
very similar to those of Trajan’s column and the Africa mosaics.
However, there is a critical difference. In the Roman Vergil what
appear to have been the rams are sustained from the stemposts by
couplings not seen on earlier depictions of Greco-Roman galleys. 49
This indicates that they were not, in fact, rams. No ram built as an
integral part of the keel and stempost structure needed to be sustained
Figure 3
Galley on a mosaic from the baths at Themetra near Hadrumetum,
Tunisia, ca 200-220 C.E.
in this way. The couplings must have indicated something new. They
cannot be considered to have been a mere artist’s aberration because
in the Latin West in the Middle Ages exactly the same type of
coupling was used to sustain the spurs of galleys.
They can be seen in a late thirteenth-century painting of a Catalan
galley from a church near Teruel in Spain,50 and are also specified in
the earliest surviving contracts for the construction of galleys, from
the Angevin court in the reign of Charles I of Anjou, King of Sicily
(ca 1269-84).51 The illustration of the Roman Vergil manuscript surely
shows the first known depiction of spurs and a new type of war galley.
Whether the same can be said of all the depictions of galleys with
------------------------------
49
Wright, Codicological notes, p. 82.
50
Foerster, “Warships of Aragón”, fig. 6 and p. 28.
51
See Pryor, “Galleys of Charles I of Anjou”, p. 40: “ Et de iugo prore usque ad
palmam habet palmincellum [palmentellum] palmorum XVI et medii usque ad ferrum
quod sustinet speronum [speronem].”.
similarly curved “rams” from the time of Trajan’s column onwards is

a moot point.
Figure 4
Dromons in the Roman Vergil (Rome, Biblioteca Apostolica Vaticana,
MS. Vat. Lat. 3867, fol. 77r), late fifth century.
The illustrations of the Ilias Ambrosiana depict for the first time
galleys with stemposts raked strongly forward and flared bows of a
type which ought to have decreased water resistance and increased
speed, but which would have made constructing ships with waterline
rams impossible. Significantly, no projection of any kind at or near the
waterline which might possibly be a waterline ram is shown in any of
these illustrations. Some Roman merchant galleys had also had such
140 CHAPTER TWO
raked and flared bows but they, of course, had no need for rams.52 In
most cases it is not possible to distinguish any more detail of the bows
of the Ilias Ambrosiana galleys since they are hidden either by other
galleys or by promontories of land. In a few cases the stemposts
(KIWTG
%CVCNCPICNNG[QPCRCKPVGFDGCOHTQOCEJWTEJPGCT6GTWGN
$CTEGNQPC
Figure 5
/WUGQFGN#TVGFG%CVCNWÅC4GHNCVGVJKTVGGPVJEGPVWT[
Catalan galley on a painted beam from a church near Teruel (Barcelona,
~/0#%n/WUGW0CEKQPCNF #TVFG%CVCNWP[C$CTEGNQPC
Museu nacional d’Art de Catalunya, Ref. 15839), late thirteenth century.
2JQVQITCRJGTU%CNXGTCU/½TKFC5CITKUV´
appear to be clean, having no forward projection of any kind.53

However, there are two illustrations where it is possible to make out a
narrow beam projecting forward from the stempost about half way
between the water and the gunwale and raked slightly upwards. These
also must surely have been spurs. These beams, or “spurs” as we
would argue that they were intended to represent, were shown clearly
in the first reproduction of the drawings of the manuscript, done in
1819 under the direction of Cardinal Angelo Mai.54
Although we would not wish to nail our colours to the mast on the
issue, it is certainly arguable that the earliest known “Byzantine”
illustrations of dromons occur in the manuscript Paris, Bibliothèque
Nationale, MS. Grec 923 of the Sacra Parallela attributed to St John
of Damascus (ca 675-753/4) at fol. 207r. It is now generally accepted
that the manuscript may be dated to the third quarter of the ninth
century, with a provenance in Constantinople.55 Although the artist
------------------------------
52
See Casson, Ships and seamanship, plates 138, 139.
53
Bandinelli, Hellenistic-Byzantine miniatures, figs 47 (p. 58) [= fig. 116 (pl. 15)],
63 (p. 67) [= fig. 190 (pl. 34)].
54
See Mai, Iliad, pl. VIII (= Figure 6 here), pl. XXXVIII (= Figure 7 here). The
1819 edition has no page numbers.
55
See Weitzmann, Sacra Parallela, fig. 203 (pl. LIII). The illustration is to Psalm
106 (107) verses 23-4: “They that go down to the sea in ships, working [rather than
“doing business”] in many waters; these [men] have seen the works of the Lord, and
his wonders in the deep.” [The Septuagint version of the Old Testament, with an
English translation, p. 765] quoted in abbreviated form from John Chrysostom,
Homily in Lazarum. There are other illustrations of ships with some of the same
characteristics, but nowhere near as complete, at fols 29v and 206v (fig. 206, pl.
LXIV and fig. 316, pl. LXXII). On the dating and provenance of the manuscript see
Figure 6
Dromon in the Ilias Ambrosiana (Milan, Biblioteca Ambrosiana, Cod.
Ambros. F. 205 Inf., min. VIII), early sixth century.
Figure 7
Dromons in the Ilias Ambrosiana (Milan, Biblioteca Ambrosiana, Cod.
Ambros. F. 205 Inf., min. XXXVIII), early sixth century.
------------------------------
Brubaker, Vision and meaning, p. 25.
142 CHAPTER TWO
has distorted the ships to make them fit the margins of the manuscript,
he has clearly depicted oars as well as two masts with lateen sails.
These are the only depictions of two-masted ships known to us in
European art between the second and twelfth centuries. Moreover, on
the stemposts, below the line of the oars, there are forward projections
of some kind similar to those of the Ilias Ambrosiana galleys. It is
difficult to imagine what they might have been intended to depict
other than spurs of dromons.
Figure 8
Two-masted, lateen-rigged dromons

in a manuscript of the Sacra Parallela
attributed to St John of Damascus
(Paris, Bibliothèque Nationale, MS.
Gr. 923, fol. 207r), third quarter of the
ninth century.
Cliché Bibliothèque nationale de France
The replacement of the waterline ram by the abovewater spur must

have been accompanied by changes in hull design. With a waterline
ram the ship itself was the weapon rather than the ram, just as, in the
High Middle Ages, the weapon was not the knight’s lance but rather
the entire combination of horse and rider. The lance was merely the
delivery system. In the case of the ram, the impetus of the entire
galley was funnelled into the impact zone of the ram. Therefore, the
keel, stempost, ram, and hull had to be constructed in such a way as to
absorb the shock of impact.56 This precluded raking and flaring of the
stempost and hull at the bows. Replacement of the ram by the spur
permitted war galleys to have stemposts and bows constructed in the
same raked and flared way as some merchant galleys had had. This
may have decreased water resistance at the bow somewhat, possibly
producing a concomitant increase in speed. However, water resistance
is an extremely complex matter,57 and suffice it to say here that such a
change in the structure of the bows ought not in itself to have
produced any marked increase in speed per se. The speed of trie2reis,
which did have rams of course, was legendary. It may have made the
ships somewhat more manœuvrable since they would have had a
shorter keel and therefore less length of keel to drag across the water
when turning but, against this, the reconstructed Greek trie2re2s
Olympias has proved to be highly manœuvrable. Neither of these
factors are likely to have been sufficiently significant in their own
right to explain the disappearance of the ram or the etymology of the
word dromo2n .
The new spur of Byzantine and medieval galleys was not designed
for “ramming”, in the traditional classical meaning of that term.58 It
was not meant to fracture the hull of an enemy ship in order to sink it.
Rather, it was designed to ride up and over the oars of an enemy ship,
smashing them and disabling its power source so that it would be
------------------------------
56
See Steffy, “Ram and bow timbers”, pp. 37-9.
57
Water resistance to a ship is largely comprised of, first, frictional resistance due
to the ship dragging water along with it and, secondly, the wave-making action of the
ship as it moves through the water and creates drag. The first is effectively
proportional to the wetted surface of the hull and to the speed of the ship to the power
of 1.85. In the case of a hull shaped like that of a galley and of the same length, the
second is not very significant below about 6 knots but rises to become about
equivalent to the frictional resistance at around 9 knots. Communication from John
Coates.
58
Contrary to the opinion of Dolley in “Warships”, p. 48, and also of other
scholars.
144 CHAPTER TWO
rendered helpless and vulnerable to attack by marines and archers.59

This function is revealed by the etymology of “calcar” one of the two
medieval Latin terms for such spurs, the other being speronus. In
Latin, “calcare” meant “to tread under foot, to ride over, to trample”.
In the twelfth-century Sicilian manuscript of John Skylitze2s’
Synopsis historio2n galleys intended to represent Byzantine war galleys
defeating Rho2s ships in the Bosporos in 941 are shown rolling the
Rho2s ships over and smashing their oars with their spurs and bows.
Figure 9
Dromons rolling over Rho2s ships with their spurs in the Synopsis
historio2n of John Skylitze2s (Madrid, Biblioteca Nacional, vitr 26-2, fol.
130r), ca 1160.
There has been much speculation about possible reasons for the
replacement of the ram by the spur. Some have thought that it may
have had something to do with the invention of “Greek Fire”, on
which see Appendix Six. However, it is clear that the development of
------------------------------
59
This is made clear in the chronicle composed by an anonymous chaplain of the
Templars in the Holy Land in 1191-2 known as the Itinerarium peregrinorum. See
Itinerarium peregrinorum (Mayer), p. 322: “Quod autem antiqui dixere liburnam,
moderni galeam media producta nominant, que longa, gracilis et parum eminens
lignum a prora prefixum habet, et vulgo calcar dicitur, quo rates hostium transfiguntur
percusse.”.
the spur predated the invention of Greek Fire by at least a century and
a half. More probably, the change was related to the evolution of hull
construction in late antiquity.
Maritime archaeologists have now produced clear evidence that
during late antiquity the classical technique of constructing the hulls
of ships shell first was changing. In the classical Mediterranean form
of shell construction,60 hulls were constructed from the keel outwards
by fitting the planks or strakes edge to edge and holding them together
with closely-spaced mortise and tenon joints pegged with treenails.
Frames were not inserted until hulls had been built up to a point where
they could usefully be placed in position. The finest surviving
archaeological example of this form of construction is the wreck of a
small sailing ship of the fourth century B.C.E. found off Kyrenia,
Cyprus. In this ship, the tenons were fitted tightly in the mortises and
were approximately 4.3 centimetres wide with gaps of only around 7.5
centimetres between them, around 11.8 centimetres from centre to
centre of adjacent tenons. They were also long; the mortises being cut
in to each plank to a depth of around 6.1 centimetres, the tenons being
around 12 centimetres long. In addition, the tenons were pegged fast
in the mortises by treenails after the planks had been hammered
home.61 Internal frames were added after the hull had already been
built up to a certain point. This technique produced light and strong,
but very inflexible hulls. Almost certainly the waterline ram had been
specifically designed for use against hulls constructed in this way.
The only classical waterline ram, embolos or rostrum, so far
recovered from the Mediterranean seabed is the ram found off Athlit,
Israel, which survives from what was in all probability a Hellenistic
tetre2re2s of the third-second centuries B.C.E.62 In a seminal study of
this ram Steffy has related its structural operation to the construction
of the hulls against which it was used and has argued persuasively that
its peculiar structure of the ram, with three horizontal fins and a
central vertical post at the impact zone, was specifically designed to
operate against hulls constructed from planks edge-joined by closely-
spaced mortise and tenon joints. It was not intended to penetrate the
hull. Rather, it was designed to deliver a blow to the moving hull of an
------------------------------
60
Shell construction as such was also used elsewhere with various plank-joining
techniques. For example, in Scandinavia and Northern Europe until the Late Middle
Ages shell construction was also used but with the clinker technique of joining the
planks to each other.
61
See Steffy, “Shell to skeleton”, pp. 1-2; Casson, Ships and seamanship, p. 214.
62
Murray, “Athlit ship”.
146 CHAPTER TWO
enemy ship which would shatter its waterline wale or at least cause it
to flex markedly, dislodging frames and and tearing loose the mortise
and tenon joinery of adjacent planks. This would probably cause the
planks to split down the middle.63 They would be sprung irrepairably,
resulting in flooding of the hull that could not be stopped by damage
control.
Figure 10
The Athlit Ram, third-second centuries B.C.E.
Collateral evidence to support this thesis that the classical ram was
specifically designed to operate against a particular type of hull
construction may be found in Julius Caesar’s comment that the rams
of his galleys were useless against the oak hulls of the ships of the
Veneti in the English Channel.64
The evidence of late antique wrecks shows that by around the
fourth century the mortise and tenon joinery was becoming looser and
less structurally important while the internal frames in the hull were
becoming more important. By the time of the fourth-century wreck
found at Yassı Ada islet in the Chuka Channel between Pserimo and
Turkey, the tenons had become less tightly fitting, wider (7-9
------------------------------
63
Steffy, “Ram and bow timbers”, pp. 37-8. Cf. Shaw, “Steering to ram”, p. 99,
the quotation from J. Haywood. See, for example, the split plank from the Grand
Congloué wreck in Casson, Ships and seamanship, plates 159-60.
64
Caesar, Gallic war, III.13 (p. 156): “Neque enim his [navibus] nostrae rostro
nocere poterant (tanta in eis erat firmitudo, ...”.
centimetres), but shorter, the mortises ranging between 5.0 and 5.5
centimetres, and were spaced some 24.3 centimetres apart. However,
they were still pegged in the mortises by treenails. In the wreck of ca
400 known as Port-Vendres A in Roussillon the mortises and tenons
were spaced at intervals of between 6 and 15 centimetres. In the wreck
of the fifth century known as Dramont E, found off the Ile d’Or,
Provence, the tenons were loose fitting in the mortises, irregularly
sized, and spaced between around 10 and 30 centimetres apart, but
still pegged in the mortises by treenails. The evolutionary process was
even more clearly apparent in the seventh-century Yassı Ada wreck,
in which the tenons were only around 3 centimetres wide, very loose
fitting, and strongly tapered at the ends in mortises up to 5 centimetres
wide but only around 3.5 centimetres deep, and varied in spacing
between around 35 and 90 centimetres apart. The wreck excavated
near Bozburun, Turkey, whose timbers were felled in 874 according
to dendo-chronological analysis, shows no signs of mortise and tenon
edge-joining of planks.65 By the eleventh century, in the Serçe Limani
wreck, mortise and tenon joining of planks had definitely disappeared
and skeleton construction over a framework of ribs and stringers had
replaced the classical shell construction technique. Other wrecks
which display little or no evidence of mortise and tenon plank joining,
and which were skeleton built, include the seventh-century Saint
Gervais B wreck, the tenth-century Agay wreck, the twelfth-century
Pelagos wreck, and the tenth-century Muslim ship at Plane in
Marseilles Bay.66 As more wrecks from the centuries spanning the
first millennia B.C.E. and C.E. are found and excavated in the future,
the precise details of this evolution in hull construction in the
Mediterranean will become more completely fleshed out. But, even
now, enough has been learned from nautical archaeology to confirm
the general parameters of the evolution. Slowly, over the centuries, the
entire conception of the building of hulls of ships changed.
It has also been suggested that a change from waterproofing hulls
by means of a coating of waterproof material covered with lead
sheathing to hold it in place, to doing so by a caulking of tow or
oakum driven into the seams between the strakes, may also have
------------------------------
65
Personal communication from Frederick M. Hocker to John Pryor. See also
Hocker, “1995 field season”; “1997 field season”; “Final campaign”.
66
See Bass, History of seafaring, pp. 138, 143; idem, Yassı Ada volume I, p. 55;
Parker, Ancient shipwecks, pp. 42, 306, 314, 330, 373, 454-5; Pryor, “Mediterranean
round ship”, pp. 65-7; Santamaria, “L’épave Dramont”, p. 144; Steffy, “Shell to
skeleton”.
148 CHAPTER TWO
accompanied the progressive evolution away from mortise and tenon

plank joining.67 Obviously, with closely-spaced mortise and tenon
edge joining of the planks, it was not possible to force caulking into
the seams. The tenons would have prevented much of the caulking
being driven in very far in any case and to have even tried to do so
would have loosened the tenons in their mortises, weakening the
whole structure and negating the entire rationale on which the
construction of the hull was based.68 This only became possible and
necessary when the number and structural integrity of the mortises
and tenons decreased. Since lead sheathing has not been reported
reliably on any wrecks dated to later than the end of the second
century C.E., this evolutionary process may have begun much earlier
than has hitherto been suspected. That being said, the archaeological
evidence is very ambiguous, confused by unprofessional excavations
and reports and the problem of knowing whether coatings of
waterproofing, or caulking between strakes, were applied at the time
of construction of ships or much later in attempts to prolong their
life.69
------------------------------
67
See Hocker, “Lead hull sheathing”. See also Casson, Ships and seamanship, pp.
214-16.
68
See Basch, “Note sur le calfatage”, p. 188.
69
The following does not pretend to be a comprehensive discussion of the issues.
It merely indicates the ambiguous parameters of the subject given the current state of
research. References are omitted because of their multiplicity.
On the one hand, the wrecks of Kyrenia (ca 310-300 B.C.E.), Marsala Punic
ship (ca 250-175 B.C.E.), Grand Congloué and Spargi (2nd century B.C.E.),
Anticythera A, Mahdia, Albenga, and Le Madrague de Giens, (1st century B.C.E.),
Ben-Afelí (near Almazora, Spain), Nemi, and Port-Vendres B (1st century C.E.),
Saint Gervais D (ca 50-150 C.E.), Grand Bassin C at Gruissan, Roussillon (ca 120
C.E.), and Procchio, Elba (ca 160-200 C.E.), and possibly a ship dated tentatively to
ca 200 C.E. excavated off Grado, all had lead or copper sheathing. However, the
sheathing was applied to the Kyrenia ship in her old age in an attempt to prolong her
life and, while the Marsala ship was lead-sheathed but was newly-built when she
sank, in many of the other cases it is impossible to know when the sheathing was
applied and for what purpose. Whether any coating of waterproofing between the hull
and the sheathing existed is often not specified.
On the other hand, no lead sheathing is reported for the wrecks of La Chrétienne
A and C (2nd century B.C.E.), Dramont A and Le Titan (1st century B.C.E.), Mateille
B at Gruissan (1st century C.E.), Torre Sgarrata (ca 180-205 C.E.), Monaco A (ca
200-250 C.E.), La Lique B, Provence (ca 300-325 C.E.), Yassı Ada, Dramont F, and
Port-Vendres A (ca 400 C.E.), or for any wrecks later than ca 200 C.E.
Lead plates have been found in some other later wrecks such as those at Ayia
Galini, Crete (ca 276-90 C.E.), Femmina Morta, Sicily (early 4th century C.E.), and
Isola delle Correnti (3rd-4th centuries C.E.?). However, it is unclear whether these
plates were to replace sheathing come loose or for patching. The hull of the Culip D
wreck at Cap de Creus, Spain (ca 70-80 C.E.) was patched with lead rather than
sheathed.
Under the year 718, Theophane2s the Confessor wrote that the
Muslim fleet retiring through the Aegean after the failed siege of
Constantinople in 716-18 was struck by a “fiery shower” which made
the sea boil up and that the ships were then sunk because their pitch
was gone.70 This is an obviously improbable story, but the point is that
such a melting of the pitch would effect only ships which depended on
intra-seam caulking. Ships with mortise and tenon joined planks and a
coat of pitch over the whole hull would not be sunk by its melting, at
least not straight away. He had died when Theophane2s was young but
Theophane2s’s father had held some kind of command in the islands of
the Aegean Sea and Theophane2s may have had some real knowledge
------------------------------
As for coatings of waterproofing or intra-seam driven caulking, the hull of the
Dramont A wreck had been coated on the inside with some protective material and the
La Chrétienne C wreck (ca 175-150 B.C.E.) had been coated inside and out with
resin. The Mateille B wreck had coats of pitch on the the hull both inside and outside.
The Grand Bassin C wreck had a layer of fabric and pitch between the hull and the
lead sheathing. The Monaco A wreck had pitch on the outside of the hull but the
Pomegues A wreck (3rd century C.E.) had it on the inside. The fourth-century Yassı
Ada wreck had pitch applied to the undersides of the frames before they were laid in
place but this was probably to stop water being trapped between the frames and
planks and thus rotting out both. The Dramont F wreck of ca 400 C.E. also had pitch
applied to the hull both inside and out. The seventh-century Yassı Ada wreck had a
coat of pitch applied over the whole of the inside hull after the frames had been put in
place and to the outside of the hull before the ship was launched. The Port Vendres A
wreck had caulking of tow driven between the planks but this may have been a repair
job late in the ship’s life. The Dramont F wreck (ca 420-25 C.E.) had coats of pitch
over both the inside and outside of the hull. The inside coating was done when the
ship was built but that on the outside only shortly before the ship sank.
Strangely enough, in the seventh-century Yassı Ada wreck, there is no mention
of intra-seam driven caulking; although, what appears to have been a caulking tool
was found. There was, however, caulking in the L’Anse Saint-Gervais B wreck (ca
600-625 C.E.) and in the ninth/tenth-century Bozburun wreck a fibrous material,
probably grass, was driven between the seams as caulking and the outside of the hull
was covered in a layer of pitch or resin. Similarly, in the wreck of the eleventh-
century Serçe Liman ship a complete set of caulking tools was found and the coating
of grass and pitch applied to the outside of the hull was also driven into the seams.
70
Theophane2s, Chronographia, A.M. 6210 (vol. 1, p. 399): “... oiJ de;
perileifqevnte" parhvrconto to; Aijgai'o n pevlago", kai; a[fnw ejph'lqen aujtoi'" qeomhniva
foberav: cavlaza ga;r puvrino" katelqou'sa ejp aujtou;" to; u{dwr th'" qalavssh" kaclavsai
pepoivhken, kai; th'" pivssh" luqeivsh", au[tandroi aiJ nau'" eij" buqo;n kathnevcqhsan:”.
In his note on this passage, Mango comments that “though doubtless
embellished”, the report of boiling waters in the Aegean may well have been
connected with unusual volcanic activity which culminated in the eruption of The2ra in
726, also reported by Theophane2s. See, Theophane2s, Chronographia, trans. Mango
and Scott, A.M. 6210 (pp. 550-551 & n. 9) and cf. A.M. 6218 (p. 559). Note that the
translation, “... and as the pitch of their keels dissolved, their ships sank ... ” is
incorrect. Keels were not mentioned in the text, which simply said that they sank
because the pitch was gone. Ships had pitch as caulking in the seams between strakes,
not on their keels. Or at least, if they did have pitch on their keels, loss of it would not
lead to them sinking, as it would if the pitch was caulking in the seams.
150 CHAPTER TWO
of ships and the sea and an unusual familiarity with Byzantine fleets. 71
His story seems to reflect an age in which ships depended on intra-
seam driven caulking for watertightness.
Although the first known use of a particular word for any practice
is only circumstantial evidence for the chronology of its emergence,
the first known usages of the words which became the medieval Greek
for a “caulker”, kalaphate2s, and “caulking”, kalaphatizein, occur in
Egyptian papyri dated to the 560s. By the eighth century, they were
common in the Aphrodite2 papyri.72 They appear in Byzantium itself in
the De cerimoniis attributed to Constantine VII. In inventories for
expeditions to Crete in 911 and 949, flax for, and the cost of,
kalaphate2seo2s, “caulking”, was included.73 The first known illustration
of caulkers at work on the hull of a ship is at folio 240r of the Pierpont
Morgan manuscript of the De materia medica of Dioskoride2s, which
was probably made for Constantine VII. The manuscript was a re-
working of the early sixth-century Vienna manuscript, which,
however, did not include this illustration, or indeed any human
figures. It appears that the process of caulking became known in the
Byzantine world between the ages of composition of these two
manuscripts.74 In Latin the word first appeared as calafata in the
Antapodosis of Liudprand of Cremona, written ca 958-62, where it
was a transliteration of the Greek term, but misunderstood as meaning
shipwrights,75 as though he encountered the word in Constantinople
but its meaning was new to him. These words were not known in
classical Greek and Latin and their appearance clearly reflected new
practices for waterproofing the hulls of ships, presumably by intra-
seam driven caulking, as early as the sixth century in Egypt at least.
Although all of the wrecks upon which research into the evolution
------------------------------
71
Vita Theophanis, III.5 (p. 4): “... tw'/ kata; savrka patri; ejn th'/ diepomevnh/ aujt w'/ tw'n
Aijgaiopelagitw'n ajrch'/ teleuthvsanti, ...”. By comparison to the authors of subsequent
Byzantine histories and chronicles such as Theophane2s continuatus, Genesios, George
Hamartolos, John Skylitze2s, Leo the Deacon, and Symeon Logothete2s, Theophane2s
does show a particular interest in, and knowledge of, matters maritime.
72
Bell, Greek papyri. IV, Nos 1391, 1410, 1433-36, 1446, 1514; idem, Greek
papyri. V, N o 1852 (p. 270); Rea, Oxyrhynchus papyri, No 3804.262 (p. 113); Turner,
Oxyrhynchus papyri, N o 2480.33 (p. 185). See also Kahane and Tietze, Lingua
Franca, §775 (pp. 513-14).
73
Appendix Four [a], §15; [b], §VI.14 [= Haldon, “Theory and practice”, pp. 211,
231; Constantine VII, De cerimoniis, II.44, 45 (vol. 1, pp. 659, 675)].
74
Dioskoride2s, De materia medica (Pierpont Morgan), fol. 240r; ibid. (Wellmann),
I.72 (vol. 1, p. 72); ibid. (Vienna), passim for lack of human figures.
75
See Liudprand of Cremona, Antapodosis, V.15 (p. 138): “Quod ut audivit, tou'"
kalafavta", tus calafatas, hoc est navium compositores, ad se venire praecepit.”.
Figure 11
Caulkers at work extracting old pitch from the hull of a ship in a
manuscript of the De materia medica of Dioskoride2s (N.Y., Pierpont
Morgan Library, Cod. 652, fol. 240r), tenth century.
of hull construction has been based were merchant ships, there is no

evidential reason to suppose that warships were not subject to similar
processes.76 Certainly, by the age of Leo VI it is clear that dromons
------------------------------
76
It is possible that, because of the value of lightness and strength in the hulls of
warships, and because governments may have been less susceptible to the socio-
economic constraints that contributed to propelling commercial shipping towards
skeletal construction, shell construction may have survived in warships longer than in
merchantmen. Against this should be balanced the consideration that governments
frequently found it necessary to build war fleets in considerable haste and that the
skeleton method of construction was much quicker, less labour intensive, less costly,
and required fewer carpentry skills than the shell method. These considerations ought
to have been extremely attractive in moments of crisis.
Since no underwater archaeological evidence for the hull construction of either
classical or early medieval galleys has yet been discovered, it is not possible to affirm
positively whether or not shell construction continued to be used in them after the
change was under way in merchant ships. The only wreck known to us of what was
probably a galley of the early Byzantine period is that investigated by Purpura at
Cefalù, Sicily. The evidence of its pottery dates it to the late sixth or early seventh
centuries and it appears to have been around 35 metres in length and perhaps up to 6
metres in beam, although that may have been a product of the hull collapsing
outwards as it disintegrated. However, so little of the ship has been reported that no
hypotheses as to its hull construction can be ventured. That it was of the type of the
152 CHAPTER TWO
were caulked with tow or oakum driven into the seams because the
emperor wrote in his Naumachika Leontos Basileo2s that ships should
carry extra floor timbers, planks, tow (stuppiva, styppia), pitch (pivssa,
pissa), and liquid pitch (uJgrovpisson, hygropisson). Given that the
specification for tow and pitch comes after the mention of extra
timbers and before a requirement that one oarsman should be a
naupe2gos, a shipwright, with the requisite tools, there can be no doubt
that what the emperor was referring to was caulking for the seams.77
We suggest that when hull construction changed so that the mortise
and tenon joinery of the planks became far less frequent and tight,
with the results that the planks were more flexible and not so
susceptible to splitting, that the waterproofing of the seams became
more dependent upon caulking, and that the frames became heavier
and more integral to the construction of the hull, the Greco-Roman
ram no longer worked in the way it had done in the past. Heavier and
more frequent frames would better sustain the hull against any impact
and any breach in it would be more localized and more easily sealed
from within by damage control. Without the structural weakening of
the planks down their centres caused by the frequent chiselling out of
the mortises on both sides, they would be far less susceptible to
splitting. Therefore, the ram was replaced by a different offensive
weapon, the spur, which was also designed to disable an enemy ship,
but in a completely different way.
These developments cannot be dated precisely. All that can be said
is that the evidence for disappearance of lead sheathing suggests that
some changes in hull construction were under way by the end of the
first century C.E., that the evidence of the fourth-century Yassı Ada
wreck shows clearly that they were considerably advanced by that
time, and that the evidence for spurs on the galleys of the Roman
Vergil and Ilias Ambrosiana manuscripts suggests that the processes
of change were so far advanced by the turn of the fifth and sixth
centuries that the waterline ram had been abandoned by that time.78
------------------------------
dromon, as suggested by Purpura, is entirely hypothetical. It could have been almost
any kind of galley, perhaps a merchant galley, judging from the pottery aboard it. See
Purpura, “Relitto bizantino di Cefalu”. The wreck has not been excavated and nothing
is known about the construction of its hull.
77
See Appendix Two [a], §5. Basch reached the same conclusion on the basis of
this text. See his “Note sur le calfatage”, p. 197.
78
Harris, “Bessarion on shipbuilding”, p. 292, has claimed that dromons were built
using the new skeletal construction techniques. However, this is sheer speculation and
none of the sources he cites, including Pryor, Geography, technology, and war, pp.
57-60, actually say this.
Square sails and lateen sails
The square sail of antiquity evolved gradually into the lateen sail of
the Middle Ages by an evolutionary process of setting the sails more
fore-and-aft than square and then tailoring the luff and leech.79
Whether the dromons of Belisarios’s fleet still had square sails as the
the trie2reis and liburnae of antiquity had had, or whether they already
had the lateen sails of medieval galleys, Prokopios did not make clear
since he referred to Belisarios’s command ships by the generic ne2es
rather than as dromons. However, he did write that: “The sails of the
three ships in which he [Belisarios] and his following were carried he
painted red from the upper corner for about a third of their length”.80
Because of its reference to an “upper corner”, this passage has been
widely considered to indicate that the sails of at least part of the
Byzantine fleet, perhaps including the dromons, were triangular,
presumably lateen. What is possibly the first direct literary reference
to ships with lateen sails occurred in the Life of St Caesarius of Arles
(ca 470-542) in a paragraph apparently written by a deacon named
Stephen between the death of the saint and 549. Stephen wrote that
sometime between 508 and 516 the Burgundian kings Gundobad
(474-516) and Sigismund (516-32) sent relief to famine-stricken Arles
in the form of: “..., three large ships, which they call latenae, full of
wheat ...”.81 It is difficult to imagine what else latenae could have
meant here other than lateen-rigged ships. That being said, this is in
fact the only use of such a word for a sail known to us from medieval
Latin and no equivalent is known from medieval Greek. It appears to
be a hapax legomenon. The word did not find currency for a sail in
either Greek or Latin in the Middle Ages and the origin of the modern
English word “lateen” is from the French “latine”, itself not known
before the sixteenth century.82
------------------------------
79
See Casson, “Origin of the lateen”, pp. 49-51; idem, Ships and seamanship, pp.
243-5, 273-8 and plates 180-182.
80
Prokopios, History of the wars, III.xiii.3 (vol. 2, p. 118): “triw'n new'n, ejn ai|"
aujtov" te kai; hJ qerapeiva e[plei, ta; iJstiva ejk gwniva" th'" a[nw kai; ej" trithmovrion mavlista
e[crise mivltw /, ...”.
81
Vita Caesarii Arelatensis, II.9 (p. 487): “..., antequam ipsa lux diei claresceret,
tres naves quas latenas vocant maiores plenas cum tritico direxerunt.”. On the
interpretation of latenae here see Jal, Glossaire nautique, p. 915.
82
The first European citation apparently occurs in the account of his travels
between 1435 and 1439 by Pero Tafur, probably composed in the 1450s. See Pero
Tafur, Andanças e viajes, pp. 75-6: “... despues el Adelantado me fizo dar un navío,
para yr á Babylonia, que llaman gerba, que son tan luengos como una grant galea é
todo fecho á cámaras de un cabo é de otro para aposentamiento, é llanos de carena,
154 CHAPTER TWO
Some pictorial evidence suggests that square sails survived in the

Mediterranean into the sixth century, or even beyond. The most well-
known example is the apparently square sail on a sailing ship in a
mosaic dated to ca 504-26 in the church of St Apollinare Nuovo,
Ravenna. However, the sail is depicted in reverse. The ship is sailing
backwards, indicating that the mosaicist had little familiarity either
with square sails or with ships in general. His testimony to the
survival of square sails is questionable.83 This is certainly true of the
artist of the galley with a square sail in the manuscript of the Bible
commissioned by abbot Vivian of St Martin of Tours, sometimes
known as the First Bible of Charles the Bald because it was presented
to him around 850. This manuscript’s illustrations were based on an
earlier, late antique manuscript, possibly even a Bible commissioned
by Pope Leo I (440-61).84 No really unchallengeable evidence for the
survival of the square sail in the Mediterranean is known to us from
later than the sixth century.85 From then on it disappeared until re-
introduced from Northern Europe in the fourteenth century.
The earlier Vatican Vergil manuscript definitely showed square
sails on all the galleys depicted in it under sail.86 However, the sails of
the Roman Vergil manuscript are more problematical. The galley in the
------------------------------
porque naden en poca agua; levan grandes cargos, traen la vela tan alta como una
carraca, salvo que es angosta é latina como de galea; ...”
Under “barca” in the manuscript, Florence, Biblioteca Riccardiana, Cod. 217,
which is an Arabic-Latin and Latin-Arabic dictionary compiled in Eastern Spain in
the thirteenth century, the Arabic word la2t6ana, pl. lawa2t6in, appears among a list of
Arabic words for ships considered equivalent to barca. See Schiaparelli, Vocabulista,
p. 267. However, whether the author of the manuscript really did intend this word to
refer to a lateen-rigged ship is highly doubtful. The word became common in
European languages only from the sixteenth century. See Kahane and Tietze, Lingua
Franca, §361 (p. 272).
83
See Bass, History of seafaring, ch. 6, pl. 18 (p. 154). The mosaic is discussed in
Martin, Art and archaeology, pp. 31-3. It was altered from the original in some ways
in 561 and suffered a poor restoration at the hands of Felice Kibel in 1855. See
Bovini, “Felice Kibel”, pp. 93-6; Bonino, Archeologia, p. 48. Martin believes that the
sail may in fact have been either lateen or square; however, it is quite clear from
Bovini’s reproduction in his fig. 7 of a drawing of the mosaic published by Ciampini
in 1699, which clearly shows the yard of the sail, that it represented a square sail.
Ciampini, Vetera monimenta, vol. 2, tav. XXVII (pp. 98-9) [non vidimus].
84
Paris, Bibliothèque Nationale, MS. Lat. 1, fol. 3v. See Dufrenne and Villain-
Gandossi, “Bateaux figurés”, pl. 14 (p. 254). On the illustrations of the manuscript see
Beckwith, Early medieval art, pp. 52-6 and n. 53.
85
The square sail on a ship in the tenth-century manuscript, whose provenance was
probably Constantinople, Paris, Bibliothèque Nationale, MS. suppl. Grec 247, of the
Theriaca and Alexipharmaca of the poet Nikander, fol. 12r, is a copy of a picture
from a late antique manuscript. See Omont, Miniatures, pp. 34-5 and plate LXVI.
86
See Stevenson, Miniature decoration, pictures 17, 21, 25, 30, 39.
background appears to have no mast and to have a square sail but it

may be a lateen sail with the slope of the yard towards the bow
reversed through artistic ignorance. That in the foreground has a
broken mast and the sail is attached directly to it.87 The artist either did
not understand sails or made no attempt to depict them accurately.
Both sails could be equally as well lateen or square.
Miniature XXVII of the Ilias Ambrosiana, in the print of the 1835
edition by Angelo Mai, shows two lateen-rigged galleys with the
direction of the sails reversed. They are sailing backwards.88 This is
the only miniature of the codex showing unfurled lateen sails. In all
Figure 12
Square sail on a galley in a manuscript of the Bible commissioned by
abbot Vivian of St Martin of Tours (Paris, Bibliothèque Nationale, MS.
Lat. 1, fol. 3v), ca 850.
------------------------------
87
The illustration is of Aeneid, I.84-101: storm at sea. A broken mast fits the
context.
88
Bandinelli, Hellenistic-Byzantine miniatures, fig. 63 (p. 67) [= fig. 190 (pl. 34)].
The original edition which contained the copperplate engraving from which fig. 63 is
reproduced, is Mai, Homerus et Virgilius.
156 CHAPTER TWO
Figure 13
Lateen sails on dromons in the Ilias Ambrosiana (Milan, Cod. Ambros. F.
205 Inf., min. XXVII), early sixth century, in the 1835 edition by Angelo
Mai.
others bar one the sails are furled and may be either lateen or square.
The exception is Miniature VIII, in which the sail is clearly square.86
Figure 14
Lateen? sails on dromons in the Ilias Ambrosiana (Milan, Cod. Ambros.
F. 205 Inf., min. XXVII), early sixth century.
------------------------------
86
Bandinelli, Hellenistic-Byzantine miniatures, fig. 44 (p. 56) and fig. 96 (Plate
9).
There is a problem, however. The original reproduction of Miniature

XXVII in the edition of 1819 shows the two galleys just as they are
today in the mutilated manuscript and, indeed, the torn corner of the
manuscript just as it is today. It is unclear whether the sails are lateen
or square. The engraver of 1819 was faithful to the manuscript and
drew only the foot of the sail of the right-hand galley. It looks as
though he also made it out to be a lateen sail but drew it billowing in
the correct way. This illustration of the manuscript has not
deteriorated further since 1819 and the completion of the reversed
lateen sails was the work of the engraver of the 1835 edition.87
Nevertheless, there is perhaps just enough of the foot of the sail of the
right-hand galley surviving in the manuscript to suggest that these
may have been lateen sails reversed as he thought. The sail certainly
appears to be cut differently to that of the galley in Miniature VIII. If
lateen sails were still somewhat unusual when the manuscript was
produced, the artist may have reversed the sails out of ignorance. At
the turn of the sixth century both rigs may well have been used in the
Mediterranean and galleys referred to as dromo2nes may equally well
have employed either rig at this time.
In conventional historiography, the first definite depictions of
lateen sails are usually said to be those of lateen-rigged ships in the
Paris, Bibliothèque Nationale, Ms. Grec 510, manuscript of the
Sermons of St Gregory of Nazianzos, fols 3r & 367v, which has been
dated to ca 879-82 with a provenance in Constantinople.88 However,
the little-known miniature of a sailing ship in the manuscript known as
the Khludov Psalter, Moscow, Historical Museum, MS. 129 D, fol.
88r, also clearly shows a lateen sail and this manuscript has been
arguably dated to 843-7, also with a provenance in Constantinople.89 If
------------------------------
87
The original 1819 reproduction may be found in both Mai, Iliad, pl. XXVII,
and in idem, Picturae, pl. 27 (p. 27).
88
See Weitzmann, Byzantine book illumination, ch. IV, fig. 1. On the dating and
provenance of the manuscript see now Brubaker, Vision and meaning, pp. 5-7. The
illustration on fol. 367v is of Orthodox Christians fleeing Arian persecution, to
illustrate Homily XXXIII: Pro;" Areianou;", kai; eij" eJautovn, probably the lines: “Tivna"
presbutevrou" ejnantivai fuvsei", u{dwr kai; pu'r, ejmerivsanto, purso;n a[ranta" xevnon ejp i;
qalavssh", kai; th/' nhi; sumflecqevnta" ejf h|" ajnhvcqhsan; ...”. See Gregory of
Nazianzos, Logoi, col. 220.
89
See Khludov Psalter, No 88 (no page numbers). The illustration is to Psalm 88
(89) [Septuagint], verse 9 (“You rule the power of the sea; and you calm the tumult of
its waves.”. See The Septuagint version of the Old Testament, with an English
translation, p. 751], prefiguring Mark, 4.37-39 (“And there arose a great storm of
wind, and the waves beat into the ship, so that it was now filling up. And he was in
the stern of the ship, asleep on a pillow: and they awoke him, and said to him, Master,
158 CHAPTER TWO
this is true, then the Khludov Psalter illustration predates those of the
Sermons of the Paris Gregory of Nazianzos by some thirty years.
Figure 15
Lateen-rigged ship in a manuscript of the Sermons of St Gregory of
Nazianzos (Paris, Bibliothèque Nationale, MS. Gr. 510, fol. 367v), ca
879-82.
The illustration of the dromons in the Paris manuscript of the Sacra

Parallela, which is contemporary with the Paris Gregory of Nazianzus
also shows lateen sails. [See Figure 8] There is, however, even earlier
evidence than this for lateen-rigged ships from Byzantine, or possibly
early-Muslim Egypt. A number of Egyptian graffiti and pictures of the
sixth to seventh centuries appear to show lateen-rigged ships, the most
definitive of which is a painting from the monastic complex at Kellia
------------------------------
do you not care that we are perishing? And he arose, and rebuked the wind, and said
to the sea, be quiet and still. And the wind ceased, and there was a great calm.”). See
Greek New Testament, pp. 153-4.
On the dating and provenance of the manuscript see Brubaker, Vision and
meaning, p. 25. The manuscript was overwritten for the most part in a dark minuscule
script in the twelfth century. Originally the illustrations were merely ink drawings but
at a later date they were coloured with thin, light colours. Many, and parts of many
others, were later repainted with heavy colours, especially ultramarine, no earlier than
the late fourteenth century.
in the Nile delta, about 60 kilometres south-east of Alexandria, which

has been dated to ca 600-630.90 In fact, as Casson has shown, there is
ample evidence to prove that the origins of the lateen sail in the
Mediterranean reached back to the pre-Christian era.91 Most probably,
all that occurred was that in the late-Roman, early-Byzantine period
this sail became adapted from its previous use on small craft and
merchant ships for the warships of the Empire.
Figure 16
Lateen-rigged ship in a manuscript of the Psalms, the Khludov Psalter
(Moscow, Historical Museum, MS. 129 D, fol. 88r), ca 843-7.
------------------------------
90
See Basch, “Navires et bateaux coptes”. Probable lateen sails can also be seen in
figures 22 (sepulchre at Anfouchy, Alexandria, 1st or 2nd centuries C.E.) and 23
(house at Kôm el-Dikka, Alexandria, probably late 6th century C.E.).
91
See Casson, “Origins of the lateen”, pp. 49-51; idem, Ships and seamanship,
pp. 243-5.
160 CHAPTER TWO
Figure 17
Two-masted, lateen-rigged ship in a painting from Kellia, Egypt,
ca 600-630.
Even the emperor Leo VI, an “arm-chair sailor” who had never
been to sea, appreciated that ship design was always a matter of
compromise between various objectives. He understood that it was
possible to design a galley which would be light and which would
have good speed and other qualities in battle. However, the same ship
would probably be swamped and sunk if caught at sea in heavy
weather and would be too slight to withstand enemy attack. As he
wrote of dromons of his age:
The construction of the dromons should be neither too heavy, or they will
be sluggish when under way, nor built too lightly, or they will be weak
and unsound and quickly broken up by the waves and the attacks of the
opposition. Let the dromon have suitable workmanship, so that it is not
too sluggish when under way and remains sturdy and unbroken when in a
gale or struck by the enemy.92
The theme of the emperor’s comments was surely true also of the
processes by which Roman liburnae evolved over the centuries to
become the galleys which he knew as dromons. Generation after
generation no doubt applied practical expertise to the design
characteristics of battle galleys as they knew them in order to improve
them. They made innovations which gave superior performance, they
adapted design features to changing conditions of naval warfare and
changing technology, and they made whatever compromises between
various performance desiderata were necessary to achieve the best
possible overall designs. Like all other ship types throughout history,
liburnae and dromons evolved continuously and the latter would
continue to do so from from the sixth century to the tenth century and
beyond.93
------------------------------
92
Appendix Two [a], §4. Cf. Appendix Five, §3. Interestingly, Ibn Mankalı3
chose to include the quite literal translation of these comments in his Al-ah5ka2m al-
mulu2kiyya. See Appendix Eight [b], p. 20.
93
Cf. Alexandres, ÔH qalassiva, pp. 57-60.
CHAPTER THREE
FROM THE SIXTH TO NINTH CENTURIES
Although dromons are mentioned in Byzantine sources between the

age of Prokopios and the late ninth century, no detailed descriptions of
them and no pictorial representations survive from this period.1
The Strate2gikon attributed to Maurice (ante 630) differentiated
dromons as warships from ploia, ships in general, sage2nai, and
barytera skeve2, vessels of burden, used for transportation purposes.2
The second, anonymous, collection of the miracles of St De2me2trios of
Thessalonike2, composed in the late seventh century, recorded in its
fourth miracle that an emperor sent a dromon to warn Thessalonike2 of
the flight from Constantinople of a certain king Perboundos of a
Slavic people referred to as Rynchinoi.3
In the West, in 590 the exarchos of Italy, Ro2manos, wrote to
Childebert II of the Franks, announcing the receipt of news that a
Frankish army had been sent to Italy against the Lombards, rehearsing
details of the ensuing campain in which the Lombard king Authari had
shut himself up in Pavia, and announcing plans, which were never put
into action, to besiege him there with combined Frankish and
Ravennese forces, including dromons, which would have had to sail
-----------------------------
1
The graffito of a lateen-rigged galley found on a piece of ceramic at Malaga has
sometimes been thought to have represented a dromon of the sixth century from the
period when Malaga and southern Spain were still under Byzantine control (until
621). See Höckmann, Antike Seefahrt, fig. 109 (p. 120); Viereck, Römische Flotte, p.
287; Alertz, “Naval architecture”, pp. 155-6. The projection at the bow shown on it
has been variously interpreted as either a sivfwn (sipho2n) for Greek fire or as a “pole-
like ram prow”.
This graffito now exists only in a reproduction in the Museo Naval in Madrid
because the original was stolen from the Museo Arqueologico in Malaga. However,
the original ceramic was excavated at Malaga in a level of the old city dated to the
fourteenth century. We owe this information to Larry Mott who spoke to the Director
of the Museo Arqueologico about the graffito.
The graffito dated from the High to late Middle Ages and clearly depicted a
medieval galley with a spur at the bow. It is interesting in its own right because it
appears to show a row of oar ports in the lower hull, but it was definitely not a
representation of a Byzantine dromon. It was most probably a representation of a late
medieval Muslim galley.
2
Maurice, Strate2g ikon, XI.4.88-9 (pp. 376-8): “Tou;" de; drovmwna" katasth'sai ejn
toi'" trevktoi" toi'" ajnagkaivo i"”. See also XI.4.138-9 (p. 380), XIIB.21.1-2, 21-2 (p.
468). On sage2nai see Woody, “Sagena piscatoris”.
3
Miracles of Saint Demetrius, §237 (vol. 1, pp. 199, 210).
164 CHAPTER THREE
up the Po river.4 Pope Gregory the Great twice referred to dromons in

letters to Innocentius the praetorian prefect of Africa and Smaragdus
the exarchos of Italy dated to 600 and 603 respectively.5
The chronicle known as the Parastaseis syntomoi chronikai, dated
to the early eighth century, referred for the first time to a dromo2n
basilikos, “imperial dromon”, which pursued and overtook certain
thieves.6 By this time it had apparently become customary for one or
more dromons of the imperial fleet in Constantinople to be specially
designated for the use of the emperor. The practice would continue
until dromons disappeared from the sources after the twelfth century.7
By the eighth century, Greek scribes in Muslim chanceries in Egypt
were using dromonarion, together with other terms for warships such
as akation/akatenarion and karabo2s/karabion, for war galleys in the
Egyptian fleet, although their Muslim rulers referred to the same ships
by different Arabic names.8 Akation/akatenarion was derived from the
Greek a[kato"/ajkavtion (akatos/akation) for a light merchant galley.
Presumably dromons were known in Egypt in the Byzantine era before
the Muslim conquest of the country in 639-42; however, neither that
term nor akation/akatenarion appear in the papyri record of the
Romano-Byzantine period.9 Karabo2s/karabion is even more
-----------------------------
4
Epistolae Austrasicae, No 40 (p. 146).
5
Gregory I, Epistulae, Bk. X, Ep. 16 (vol. 2, p. 845) and Bk. XIII, Ep. 34 (vol. 2,
p. 1037). We are indebted to Michael McCormick for this and many of the following
references to Western sources in the remainder of this chapter.
6
Parastaseis syntomoi chronikai, §43 (vol. 1, p. 50).
7
It must be acknowledged that the untitled treatise attributed to Constantine VII
Porphyrogenne2tos written between 948 and 952 and given the Latin title De
administrando imperio in the seventeenth century by its first editor, Joannes Meursius
the elder, said that there was no designated “imperial dromo2nion”, basilikovn
dromwvnion (basilikon dromo2nion) until the reign of Leo VI. The diminutive
dromo2nion had no particular significance. It was merely used as a derivative synonym
for dromo2n, a common practice in Byzantine Greek. De administrando imperio said
that until the reign of Basil I, emperors had used a “scarlet barge”, rJouvsion ajgravrion
(rousion agrarion), when they wished to make a progress by water and that Basil had
been the first to use dromo2nia for longer journeys. See Constantine VII, De
administrando imperio, §51 (p. 246). Either the compiler of the De administrando
imperio was unaware of the designation of a dromon for imperial use before the age
of Leo VI or else the author of the Parastaseis syntomoi chronikai did not intend
anything so specific by his use of the term. He may have meant to refer merely to a
dromon of the imperial fleet.
8
See Bell, Greek papyri. IV, Pap. 1337.3 (p. 7), 1348.3 (p. 21), 1369.4 (p. 44),
1376.6 (p. 50), 1387.6 (p. 61), 1390.2 (p. 62), 1391.4 (p. 63), 1408.5 (p. 79), 1410.2
(p. 79), 1435.10 & 95 (pp. 325, 329), 1442.22 & 135 (pp. 309, 315).
For the Arabic equivalents see Becker, “Arabische Papyri”, pp. 84, 88; idem,
“Papyrusstudien”, pp. 150-51. The Arabic terms were qa2dis, mawa2(‘)ı3n, and safı3na.
See also Pryor, “From dromo2n to galea”, p. 107.
9
Neither Dromo2n/dromonarion nor akation/akatenarion are cited in Johnson and
SIXTH TO NINTH CENTURIES 165
problematical. The term does not appear to have been Greek at all,
first appearing in papyri from Arsinoe and al-Fayyu2m dating from the
seventh to eighth centuries,10 and then in Aphrodite2 papyri of ca 709-
715/16, where some karaboi were qualified as die2reis, “twos”, that is
biremes, some as being kastellatoi, that is castellated in some way,
and one as being both a die2re2s and also castellated.11 By that time they
had clearly evolved into major units of the Egyptian fleet.
Figure 18
Galley on a lustre-ware bowl from al-Fayyu2m , Egypt (Cairo, Museum of
Islamic Art, Inv. No. 7900), tenth century, probably Fa2t6imid period.
------------------------------
West, Byzantine Egypt, pp. 139-40, nor in Merzagora, “Navigazione in Egitto”, both
of which are based on the papyri.
10
See Wessely, Studien, Nos 718 and 900.
11
See Bell, Greek papyri. IV, Pap. 1433.64, 129, 179, 227, 319 (pp. 287, 290, 292,
294, 297) for dihvrei" kavraboi; 1434.35, 1435 98 & 103, 1441.102, 1464 (pp. 310,
329, 330, 347, 424) for kavraboi kastella'toi; 1449.94 (p. 376) for kavraboi
kastella'toi dihvrei".
166 CHAPTER THREE
In 758 Pope Paul I wrote to the Frankish king Pepin I requesting

assistance to force the Lombard king Desiderius to return to the
Papacy some cities he had occupied. In it he referred to previous
negotations between Desiderius and George, the legate of Emperor
Constantine V in South Italy, for a combined assault on Otranto by
land and sea with the participation of a Sicilian Byzantine fleet of
dromons.12 Also in the eighth century, the North-Italian compiler of
the Latin glosses in the St Gall, Stiftsbibliothek, MS. 912 knew the
word as used by “Greeks” in the form “dulcones” and equated it to the
classical “trieres”; however, he was merely paraphrasing his source:
Isidore of Seville.13
Letters of the ninth-century Popes Nicholas I and John VIII also
referred to dromons, both Byzantine and Western, operating in the
Tyrrhenian Sea in the years ca 860-80. In 860 Nicholas I asked the
emperor Michael III to send his ambassadors back to Rome on
suitable dromons. In 872-3, in two letters to the praefecti of Amalfi
and the Carolingian empress Engelberga, John VIII referred to
“dromones nostros” and to his building dromons and other ships to
defend Rome. In 877 he asked the bishop of Benevento to request the
Byzantine strate2gos to arrange for ten dromons to be sent to Rome to
help defend it against the Muslims. In 878 he was also said to have
taken ship for France from Rome on three dromons which had come
from Naples to convey him there.14 In two letters of 879 he referred to
“dromonibus nostris” and congratulated the spatharios Gregory, the
tourmarche2s Theophylaktos, and kome2s Diogene2s on their victory
over the Muslims and asked them to come to Rome “cum aliquantis
dromonibus”. In the following year he wrote to the emperor Basil I
asking him to send dromons to defend the lands of St Peter.15
The word dromo2n was also used many times by Theophane2s the
Confessor; 16 however, a new term for a warship, chelandion, also
-----------------------------
12
Codex Carolinus, No 17 (p. 515): “Nam et hoc cum eodem Georgio imperiali
misso constituit, ut dromonorum Siciliae stolum in Otorantina civitate dirigatur, ut
tam Graeci quamque Langobardi ipsam opsidentes conprehendere valeant civitatem,
...”.
13
Glossae codicis Sangallensis 912, in Goetz, Glossarii Latini, vol. 4, p. 292, l.
29: “Trieres nauis magna quas greci dulcones uocant”. See “vi.. Note on citations of
Greek and Latin glossaries”, p. lxix above. Cf. Isidore of Seville, Etymologiae,
XIX.1.10: “Trieris navis magna, quam Graeci durconem vocant.”.
14
Auxilius, In defensionem, p. 63; Nicholas I, Epistolae, N o 82 (p. 439); John VIII,
Register, pp. 258-9.
15
John VIII, Fragmenta, N os 5 (p. 276), 11 (p. 279) and Registrum, N os 46 (pp. 44-
5), 217 (p. 194), 245 (p. 214), 259 (p. 229). Cf. above p. 66.
16
For example, Theophane2s, Chronographia, A.M. 6171 (vol. 1, p. 358).
appeared among surviving sources for the first time in Theophane2s in

the context of the imperial expedition sent to Cherso2n by Justinian II
in 711, which consisted of: “... all kinds of ships: dromo2nes, and
trie2reis, and 10,000 [modioi]-carrying vessels, and boats, and
chelandia”. Theophane2s also wrote that when Constantine V led a
fleet against the Bulgars in 774, he himself sailed with “the red
chelandia” (ta; ÔRouvsia celavndia).17 The word was certainly derived
from kevl h", kele2s, which in classical Greek had the sense of a
“courser”. It became applied to fast-sailing monoreme galleys as well
as to riding-horses. In fact, chelandion almost certainly originated as a
term for horse transports, although its use did not remain confined to
them.18 Almost contemporaneously, St Theodore of Stoudios
mentioned in his letters both a ship known as a chelandion and a crew
member of such a ship, to whom he referred by the term
chelandarios.19 So also Patriarch Nicholas I Mystikos referred to
chelandia being sent to the relief of Lampsakos in a letter addressed to
the then Caesar Ro2manos Lekape2nos in late 920.20 The chronicle
attributed to Symeon Logothete2s referred to an imperial chelandion
and an anonymous chronicle of the reign of Leo V also referred to an
unarmed chelandion sent to negotiate with the Bulgars.21 Dromo2n and
chelandion became sometimes interchangeable terms, although in
some sources the latter appears to have had the specific meaning of an
oared transport ship, especially for horses.22 Written around 995, the
text known as the Patria Ko2nstantinoupoleo2s referred to an imperial
dromon, dromonion basilikon, in use in the Golden Horn, supposedly
during the reign of Justin I, and then later to a fleet of chelandia with
which the droungarios of the Kibyrrhaio2tai Apsimaros seized the
-----------------------------
17
Theophane2s, Chronographia, A.M. 6203 (vol. 1, p. 377): “... pa'san nau'n
dromwvnwn te kai; trihrw'n kai; skafw'n muriagwgw'n kai; aJliavdwn kai; e{w" celandivwn,
...”. See also A.M. 6265 (vol. 1, p. 446).
18
The extraordinary attempt by Moutsos, “Greek CELANDION ”, to investigate the
etymology of celandivw(o)n, completely overlooks the obvious. He does not consider
the classical Greek kevlh", nor the origin of the ships as horse transports, although he
does realize that they were originally transports of some kind.
19
See Theodore of Stoudios, Epistulae, 108, l. 25 and 116, l.1 (vol. 2, pp. 226,
235).
20
See Nicholas I, Letters, 95, ll. 10-14 (p. 362): “Nu'n ou\n genevsqw hJ pa'sa fronti;"
kai; ejpimevleia, i{na qeou' sunergou'nto" mh; avrch;n lavbh/ to; toiou't on kakovn, ei[te di
ejntopivwn celandivwn, eja;n eijsivn, ei[te monerivwn ejnteu'qen ka]n duvo ajpostellomevnwn eij"
parafulakh;n tou' tovpou kai; swthrivan.”.
21
See Symeon Logothete2s, Chronographia, Life of Nike2phoros I, p. 202, ll. 22-3;
Syngraphe2 chronographia, p. 342, l. 18.
22
See Appendix Three, §2.16. See also Eickhoff, Seekrieg und Seepolitik, p. 136.
168 CHAPTER THREE
throne as Tiberios III in 698.23 By ca 852, according to the Cronaca

Veneziana attributed to John the Deacon, Venice had attempted to
build zalandriae, that is chelandia, for her own use.24 In a letter which
was probably written by Anastasius Bibliothecarius in 871, and which
purported to be from the Western emperor Louis II to Basil I
announcing the capture of Bari, Louis asked Basil to send a fleet to
prevent Muslim forces reaching Italy, especially Naples, from Sicily
and Africa because the “stratigus” Georgius had too few chelandia to
stop them.25 On the same day in 877, 17 April, on which John VIII had
asked the bishop of Benevento to convey to the Byzantine strate2gos
his request for 10 dromons to defend the coasts of Rome, he also
wrote directly to the strate2gos of Longobardia at Bari, Gregory,
asking him to send 10 chelandia to defend Rome. He used the two
words, chelandium and dromo, interchangeably.26 To a letter to the
emperor Basil I in 885-6, Pope Stephen VI appended an appeal asking
him to send some chelandia, equipped for a year, to defend the coast
of Rome against the Muslims from April to September and to place
them under the command of a reliable man who would have the task
at heart and would not abuse the local population by raiding and
pillaging on his own account.27
-----------------------------
23
See Patria Ko2nstantinoupoleo2s III, §65 (p. 187) and §207 (p. 280).
24
John the Deacon, Cronaca Veneziana, p. 115: “Illud etiam non est
pretermittendum quod antedicti duces [Doge Pietro Tradonico (836-64) and his son
Giovanni († 863)] ad sua tuenda loca eo tempore duas bellicosas naves tales perficere
studuerunt, quales numquam apud Veneciam antea fuit, que greca lingua zalandriae
dicuntur.” [written post 1009]. Cf. p. 145: “Imperator siquidem [Otto II], licet ingenti
difficultate. per medias barbarorum acies vix ad litus usque pervenit, ... ubi duae
Grecorum naves, quae lingua illorum zalandriae nuncupantur, non procul a terra
anchoris herebant; a quibus ipse cum duobus suis vernaculis susceptus, minime
agnitus est.” [describing Otto’s defeat by the Kalbı3te amı3r of Sicily, Abu2 ’l Qa22sim, in
982].
25
Louis II, Epistola, p. 394. This letter survives only as an insert in the
anonymous, tenth-century Chronicon Salernitanum. See Chronicon Salernitanum,
§107 (pp. 107-21, esp. p. 120, l. 9). The point is that the word chelandia was used,
and its meaning was clearly understood, either by Anastasius Bibliothecarius or by the
chancery of Louis II. In the chronicle itself the anonymous author used an
orthography which was probably closer to the way the word was pronounced in the
vernacular in South Italy towards the end of the tenth century: scelandria. Chronicon
Salernitanum, §107 (p. 107, l. 11). On the letter, its authenticity, and its milieu, see
Gay, L’Italie méridionale, pp. 84-99.
26
John VIII, Registrum, N o 47 (pp. 45-6).
27
The Greek version of this letter was edited by Grumel from the fourteenth-
century manuscript, Mt Sinai, monastery of St Catherine, MS. 482 (1117). The
original language of the letter is considered by Grumel to have been Latin; however,
no trace of a Latin version survives. If this was the case, then this was another case of
a ninth-century use of the word chelandia in Latin. See Grumel, “Lettre du Pape
Étienne V”, §61 (p. 147): “Parakalw' de; to; a{gion uJmw'n kravto" celavndia ejxwplismevna
Al-T5abarı3 recorded that in 852-3 a Byzantine fleet of 100 mara2kib

of the shalandiyya2t type, ships of the chelandia type, each carrying
between 50 and 100 men attacked Damietta.28 In his S4u2rat al-Ard5, Ibn
H4awqal used forms of both dromu2n and shalandi as generics,
indicating that the words were known commonly, and also the
adjectival form shalandiyya2t to qualify mara2kib ships.
29
Theophane2s the Confessor who, as noted, may have had some real
familiarity with matters naval, wrote that in 672-3, in response to the
Muslim assault on Constantinople, Constantine IV armed huge fire-
carrying die2reis with cauldrons for Greek Fire and also dromons
equipped with sipho2nes for Greek Fire. When describing the response
of Anastasios II in 713 to the projected Muslim assault on
Constantinople by Maslama ibn ‘Abd al-Malik, he wrote that the
emperor: “... began to build dromo2nes and fire-carrying die2reis and
huge trie2reis”. The Umayyad fleet of caliph Sulayma2n was described
as being composed of “huge ships, fighting kate2nai, and dromo2nes”.
Another Muslim fleet which arrived in spring 717 was described as
consisting of “four hundred grain-carrying kate2nai and dromo2nes”.
Finally, the new emperor Leo III prepared “fire-bearing sipho2nes and
put them aboard dromo2nes and die2reis, then dispatched them against
the two [Muslim] fleets”.30 Elsewhere Theophane2s referred to both
------------------------------
meta; tw'n creiw'n aujt w'n ejniausiaivwn ajpo; mhno;" Aprillivo u e{w" Septembrivo u
ajpostei'lai, o{pw" fulavttwsi th;n paraqavlassan hJmw'n ajpo; th'" tw'n Agarhnw'n pagavnwn
ejkporqhvsew": ...”.
This Greek version of the letter was later abridged by the compiler of the anti-
Pho2tian collection, a collection of materials pertaining to the eighth Ecumenical
Council held in Constantinople in 869-70, which restored Patriarch Ignatios and
exiled Pho2tios, and which was compiled by a partisan of Ignatios, probably Nike2tas
David Paphlagon, the author of the Life of Ignatios. In this version, the details of the
Pope’s request were altered slightly and the second request about the commander was
omitted, but this makes no difference to the point here. The anti-Pho2tian version of the
letter was edited from five manuscripts in Epistolae ad res Orientales spectantes, esp.
p. 374: “Parakalw' de; to; a{gion uJmw'n kravto" celavndion ejxoplivsai meta; kai; tw'n
creiw'n aujtw'n ajpo; mhno;" Aprillivo u e{w" Septembrivou kai; ajpostei'l ai, o{pw" fulavttwsi
ta; pro;" qavlassan hJmw'n ajpo; th'" tw'n Agarhnw'n ejkporqhvsew".”. The Latin versions of
the letter in Mansi, the Patrologia Latina, and elsewhere were Renaissance
translations, not the supposed Latin original. See Mansi, Concilia, coll. 419-26;
Stephen V, Epistolae, 1 (col. 789).
28
Al-T4abarı3, Ta’rı3kh, (de Goeje), vol. 3, pp. 1417-18; Al-T4abarı3, Ta’rı3kh (Yar-
Shater), A.H. 238 (vol. 34, pp. 124-7)
29
Ibn H4awqal, Kita2b S5u2rat al-Ard, pp. 151, 197-198.
30
Theophane2s, Chronographia, A.M. 6164 (vol. 1, p. 353): “... kai; aujto;" dihvrei"
eujmegevqei" kakkabopurfovrou" kai; drovmwna" sifwnofovrou" ...”; A.M. 6206 (p. 384):
“e[sthse de; ejp eivkta" kai; h[rxato ktivzein drovmwnav" te kai; dihvrei" ‹pursofovrou" kai;
megivsta" trihvrei"› ...”; A.M. 6209 (p. 395): “... e[cwn pammegevqei" nau'" kai; polemika;"
kathvna" kai; drovmwna" ...”; A.M. 6209 (p. 396): “..., e[cwn kathvna" sitofovrou" uV kai;
drovmwna".”; A.M. 6209 (p. 397): “... sivf wna" pursofovrou" kataskeuavsa" eij"
170 CHAPTER THREE
Byzantine and Muslim warships as either dromo2nes or chelandia,

amongst other words.31
The “letter” attributed to Pseudo John of Damascus and addressed
to the emperor Theophilos, which may be dated to any time between
ca 840 and 940, referred to a fleet of 120 dromo2nes sent by
Theodosios III to Italy in 716-17 under the command of Leo the
Isaurian, who then seized the throne as Leo III on his return to
Constantinople.32
Anastasius Bibliothecarius, the ninth-century cardinal priest of St
Marcellus, Antipope, and Papal archivist, in his translation of the
chapter of Theophane2s’ Chronographia describing the preparations
undertaken by Anastasios II against the projected Muslim attack on
Constantinople by Maslama ibn ‘Abd al-Malik, translated literally and
wrote that the emperor ordered the construction of “dromo2nes, and
fire-carrying dieres, and very large trieres, ...”. But, curiously, when
this text was later incorporated into his Historia miscella by the
otherwise unknown Lombard author Landolfus Sagax, he altered
Anastasius’ dieres igniferas to trieres igniferas.33
In the ninth century, in his Liber pontificalis ecclesiae Ravennatis
Agnellus of Ravenna mentioned dromons four times: first, in the
context of Odovacer fleeing from Ravenna before Theodoric the Great
in 491 “cum dromonibus”, and secondly, in the context of a certain
abbot John of the monastery of St John “ad Titum” being unable to
find a ship to carry him from Constantinople to Ravenna or Sicily
some time between 692 and 708, having searched out “all carabi and
celandria and dromones”.34 He certainly knew the word as that for
major units of Byzantine fleets. Writing of a supposed battle outside
Ravenna between iconodule Ravennese and an expedition sent by the
iconclast emperor Leo III sometime after ca 727, Agnellus wrote that
------------------------------
drovmwnav" te kai; dihvrei" touvtou" ejmbalw;n kata; tw'n duvo stovlwn ejxevpemyen.”.
31
Theophane2s, Chronographia, A.M. 6171 (vol. 1, p. 358), A.M. 6178 (p. 368),
A.M. 6210 (p. 398), A.M. 6235 (p. 420), A.M. 6238 (p. 424), A.M. 6254 (pp. 432-3),
A.M. 6261 (p. 444), A.M. 6265 (p. 446), A.M. 6266 (pp. 447-8), A.M. 6267 (p. 448),
A.M. 6289 (pp. 471-2), A.M. 6295 (p. 479).
32
See Munitz, Letter of the three Patriarchs, §11.b (p. 163): “... kai; katalabw;n ta;
ejkei'se ejn plwthrsivoi" drovmorsi [drovmwsi] to;n ajr iqmo;n rkV, ...”. “drovmorsi” as per the
MS. is meaningless. “drovmwsi” is a logical editorial emendation. See also below pp.
420-21.
33
Anastasius Bibliothecarius, Chronographia, p. 246: “Constituit autem
praepositos construentium naves, et coepit aedificare dromones et dieres igniferas et
maximas trieres, ...”. Cf. Historia miscella, lib. XX (col. 1071): “Constituit autem
praepositos construentium naves, et cœpit ædificare dromones, et trieres igniferas, et
maximas trieres, et ...”.
34
Agnellus, Liber pontificalis, §§39, 131 (pp. 303, 364).
the defeated Byzantines sought refuge on their dromons on the river

Badareno but were surrounded and killed by the Ravennese with their
35
light cymbae and carabi. These passages appear to have been well
informed. However, the fourth, referring to one of the annual wheat
transportations from Sicily to Ravenna, “..., exinde honeratis
dromonibus quinquaginta milia modiorum tritici, ...”, is obviously
problematical and therefore casts doubt upon the others.36 Noone
would ever have transported grain on galleys of any kind, except in
dire emergencies, because of their low capacity and high operating
costs. Sailing ships were the normal means of transporting grain.
Depending upon which modius Agnellus was referring to, the 50,000
modii of wheat would have ranged in volume from around 550 cubic
metres to around 850 cubic metres and in weight between around 440
and 680 tonnes. Given the low deadweight tonnage of medieval
galleys of all types, between 17 and 27 dromons would have been
needed.37 Although it is possible that Agnellus may have been familiar
with real galleys known as dromo2nes at Ravenna in his time, even
though Ravenna had not been part of the Empire since 751, it is much
more probable that he knew the word from literary sources and used it
as a generic for any major vessel, whether a warship or a transport.
The anonymous Greek author who wrote, sometime in the ninth or
tenth centuries, a treatise on the interpretation of dreams, an
oneirokritikon, under the Muslim name of Achmet the son of Seire2m,
also used the term dromo2n in a paragraph on the interpretation of
-----------------------------
35
Agnellus, Liber pontificalis, §153 (p. 377): “Videntes vero Pelasgi, cornu suum
esse cunfractum, coeperunt fugere infra dromonibis, putantes se liberare. Tunc
Melisenses, id est Ravenniani cives, circumdederunt eos cum cymbis et carabis, et
irruentes super Bizanteos, omnes interfecerunt et corpora eorum in Eridanum
praecipitaverunt.”. There is no other evidence to confirm this account of a supposed
iconodule uprising in Ravenna against Byzantine authority after the death of the
exarchos Paulos ca 727, nor for the defeat of a Byzantine expedition sent to return the
city to its allegiance.
36
Agnellus, Liber pontificalis, §110 (p. 350).
37
The various types of modioi used by Byzantines ranged in volume from
approximately eleven litres to approximately seventeen litres; although it is most
probable that Agnellus would have been referring to the “sea” (thalassios) or
“imperial” (basilikos) modios, of around 16-17 litres of wheat. Wheat weighs
approximately 80 kilogrammes per hectolitre, 800 kilogrammes per cubic metre. Even
in the thirteenth century, the deadweight tonnage of Western light galeae, that is the
weight of maximum cargo, was only around 40 tonnes at best, and of that some ten
tonnes would have to be allowed for the weight of the crews. See Pryor, “From
dromo2n to galea”, p. 114. Dromons were considerably smaller than Western galeae
and their deadweight tonnage is unlikely to have exceeded around 26 tonnes.
Allowing 25 tonnes for cargo would be very generous. See below p. 304. The grain
fleet would have had to have comprised a minimum of around 17 dromons up to a
maximum of around 27.
172 CHAPTER THREE
dreams concerning imperial ships.38 The account of the lives of Saints

David, Symeon, and George, bishops of Mityle2ne2, composed towards
the end of the ninth or in the early tenth century, related that Symeon
was provided with an imperial dromon by the empress Theodo2ra when
he was sent by her from Constantinople to take up the bishopric.39 In
his life of Patriarch Ignatios, Nike2tas David Paphlagon, also writing
around the same time, referred to dromons on several occasions.40
However, none of these sources contained any details about the
construction of the ships.
To say that the evidence for the dromon in these centuries is
extremely exiguous would be to understate the obvious. Nevertheless,
the few texts that do survive from this period illustrate one important
methodological issue. Theophane2s the Confessor and Anastasius
Bibliothecarius both used the terms dromo2n, die2re2s, and trie2re2s in
sequences connected by the Greek word kaiv (“and”) and the Latin et
(also “and”), respectively. Were they, first, using the classical terms to
distinguish monoreme galleys known as dromo2nes from bireme and
trireme warships still in use in Byzantine fleets? Or, secondly, were
they using them parenthetically as in “dromo2nes, both fire-carrying
die2reis and very large trie2reis, ...”, implying that both biremes and
triremes were already known as dromo2nes? Both kaiv and et could be
used in this way. Or, thirdly, were they simply using the classical
terms parenthetically to dromo2n as approved classical words for ships
in order to display their own classical erudition, but without any
intention to specify anything about the number of banks of oars on
dromons?
The answers to these questions affect the understanding of the
reality of the dromon at the time. When the word dromo2n was used,
did it refer to monoremes only and were polyremes distinguished from
dromons still called die2reis and trie2reis? But, did bireme and trireme
polyremes even exist at the time? If they did, were they already
known as dromo2nes in the vernacular? Was dromo2n a generic
vernacular term for all Byzantine war galleys by this time? Or, was
the use of the classical terminology totally without technical meaning?
We believe that in fact dromo2n had become the standard Byzantine
-----------------------------
38
See Achmet, Oneirocriticon, § rpV [180] (p. 141): “... to;n basiliko;n drovmwna ...
oJ drovmwn aujtou' ... o{ti nevo n drovmwna eijrgavsato, etc.”. The poor translation of
Oberhelman does not use the term dromo2n [§180 (pp. 177-8)].
39
See De Smedt, “Acta Graeca”, p. 253: “... kai; drovmwno" ejpibavnte" basilikou'
...”.
40
See Vita S. Ignatii, cols 540B, 544C.
term for a war galley long before the age of Theophane2s the Confessor
and that, given the deep attachment of both Byzantine and Latin
authors to displays of classical erudition, the classical terms die2reis
and trie2reis were used parenthetically to dromo2nes by Theophane2s the
Confessor and Anastasius Bibliothecarius without any intended
technical import, being used in this way simply because they were
approved classical words for war galleys. Although it is not possible
to actually prove this solely from grammatical analysis of the texts,
later evidence suggests that it was so. As we shall see, both monoreme
and bireme galleys could be called dromo2nes by the tenth century, but
there is no hard evidence that triremes even existed at that time and
there is some to suggest that the Byzantines had only monoremes and
biremes.41 Therefore, the use of the classical term trie2reis in the
sources of the tenth century and later can have been nothing more than
a classicizing affectation. Arguably, this was also the case in the
sources of the eighth and ninth centuries? It may have been the case
that some bireme galleys were also known as dromo2nes as early as the
fifth and sixth centuries if Prokopios and Theophylaktos Simokatte2s
were merely describing one class of dromon. However, the sources of
the eighth and ninth centuries cannot be relied upon for technical
details and therefore we cannot know what the evolution of the
dromon was until those of the tenth century provide new information.
All that we can be confident of is that the term dromo2n was
originally applied, primarily at least, to monoreme galleys of 50 oars,
that its use became more and more widespread not only in the Empire
but also in the Muslim world and the Latin West between the seventh
and ninth centuries, and that by the tenth century its primary reference
in Byzantium had changed to a bireme galley of 100 or more oars.
-----------------------------
41
See below pp. 276-304
CHAPTER FOUR
THE DROMON IN THE AGE OF THE MACEDONIAN

EMPERORS
(a) The sources
Only from the age of the Macedonian emperors between 886 and
1025 does information about the construction of galleys known by the
term dromo2n and associated names survive in any quantity and detail.
Three treatises on naval warfare dating from this period provide the
most detailed information about ships and naval warfare to survive
from anywhere in the Mediterranean world between antiquity and the
thirteenth century.
The first of these is the treatise Naumachika Leontos Basileo2s, The
naval warfare of the emperor Leo, which was in fact Constitution XIX
of the Taktika written by, or compiled under the auspices of, the
emperor Leo VI, which is dated to 905-6. Some time later, in the
compilation of the manuscript Milan, Biblioteca Ambrosiana, MS. B
119-sup. [Gr. 139] the text of the entire Taktika was included in the
manuscript at folios 186r-322r, with the exception of Constitution
XIX, which was excerpted, included at folios 323r-331v, and followed
by other materials on naval warfare. In his edition of the treatise, this
manuscript was referred to by Dain as MS. A and for convenience we
have retained this reference to it throughout.1 It was written in an
expert mid tenth-century hand and its compilation has been dated to
the years around 963 and has been associated strongly with Basil the
parakoimo2menos.2 It was produced in Constantinople for a client at
the highest levels of court society, most probably for Basil himself.
Leo VI’s Taktika in general, and the Naumachika in particular,
have been generally considered to have been practical in nature, even
if Leo’s use of classical texts on military tactics has been
acknowledged. It has even been suggested that the Taktika were
intended to be quasi-legal fighting instructions which the emperor’s
------------------------------
1
See Appendix Two [a].
2
See Dain, Naumachica, p. 11; idem, “Stratégistes”, p. 385. See also Bouras,
“Basil Lekapenos”; Cosentino, “Syrianos’s «Strategikon»”, pp. 245-7; Mazzuchi,
“Basilio Parakimomenos”, p. 293. Also discussed by Dennis and Gamillscheg in
Maurice, Strate2gikon.
176 CHAPTER FOUR
strate2goi would ignore at their peril.3 However, there are several

problems associated with Constitution XIX of Leo’s Taktika and we
have severe reservations about the practicality of some parts of it.
Leo VI was one of the few Byzantine emperors to that time who
had no practical experience as a strate2gos. He never took to the field
or to the sea.4 By his own admission, his knowledge of naval warfare
was derived from consultation with his strate2goi and from some
“ancient” tactical manuals. Having said at the start that he could find
nothing about naval warfare in the old tactical manuals, he later
referred his readers to “the book on ancient tactics and strategies” for
more information about weaponry used at sea. What this book was is
unknown, although it is probable that it was a composite manuscript
containing various ancient or early Byzantine treatises on warfare,
possibly, although improbably, having some sections on naval
warfare.5
We say improbably because our research has found only two
treatises dealing with naval warfare which survived to the era of Leo
VI and which he knew. To deal first with those which either may once
have existed but which had been lost by that time, or which appear to
have been unknown to Leo VI:
First: the Poliorke2tikon attributed to Aeneas the Tactician,
written around 357-56 B.C.E., had a section on naval warfare
appended to it, but all but the first few words of this have been lost
from the sole surviving medieval text in the tenth-century manuscript,
Florence, Biblioteca Medicea Laurenziana, MS. Laurentianus LV-4,
folios 153v-180v.6 Many other treatises in the same manuscript are
also incomplete, indicating that the losses had occurred before the
compilation of the manuscript(s) from which Laurentianus LV-4 was
itself compiled.
Second: there was a small section on naval warfare and besieging
cities from the sea at the end of Book VIII, the Poliorke2tika, of the
Mechanical construction, Me2chanike2 syntaxis, of Philo2n of
Byzantium, dated to the late third century B.C.E. This book survived
in three tenth- and eleventh-century manuscripts: Madrid, Escorial,
MS. Scorialensis Gr. Y-III-11 (late tenth century); Rome, Biblioteca
------------------------------
3
See Karlin-Hayter, “Military affairs”; Tougher, Leo VI, pp. 171 ff.; Magdalino,
“Non-juridical legislation”.
4
On this question see Tougher, “Imperial thought-world”.
5
See Appendix Two [a], §§1, 72. It is quite probable that this was in fact the
treatise of Syrianos Magistros. See pp. 178-81 below.
6
See Aeneas the Tactician, Poliorke2tikon, XL.8 (p. 198).
THE AGE OF THE MACEDONIAN EMPERORS 177
Apostolica Vaticana, MS. Vat. Gr. 1164 (early eleventh century); and
Paris, Bibliothèque Nationale, MS. Gr. 2442 and Barberianus II 97
(276) (early eleventh century).7 However, there is no suggestion in
Leo VI’s Naumachika that he knew this work.
Third: Asclepio2dotos, early first century B.C.E., briefly
mentioned naval forces in his Art of tactics, Techne2 taktike2, but there
is no discussion of naval warfare or tactics in the surviving text, which
in this case appears to be complete, at folios 132r-142v of the
Laurentianus LV-4 manuscript.8
Fourth: Ailian the Tactician, late first century C.E., the oldest text
of whose Theory of tactics, Taktike2 theo2ria, is also in the same tenth-
century manuscript Laurentianus LV-4 at folios 143r-153r, said that
he was going to discuss naval warfare in another work.9 However, if
he ever wrote this, it did not survive.10 Leo VI knew and used Ailian’s
Theory of tactics;11 however, there is no evidence that another work
on naval warfare, or even a now-lost section on naval warfare of the
Theory of tactics, had survived to the tenth century and was used by
the emperor.
Fifth: Athe2naios Me2chanikos, second century C.E., had a
section in his On machines, Peri me2chane2mato2n, now best preserved
at folios 18r-24v, 25r, and 32r-v of the tenth-century manuscript Paris,
Bibliothèque Nationale, MS. Suppl. Gr. 607, which dealt with the
construction of flying bridges from the mastheads of ships to the walls
of besieged towns.12 This was known in the tenth century to the author
of the Parangelmata poliorke2tika attributed to He2ro2n of Byzantium. 13
However, it was not known to Leo VI.
Sixth: in the late fourth century, Vegetius did have some
paragraphs on naval warfare in Book IV of his Epitoma rei militaris.14
Vegetius’s treatise survived in many manuscripts, the earliest of
which is dated to the seventh century, but as far as is known it was
never translated into Greek and there is no indication in Leo’s
------------------------------
7
Edited in Garlan, Poliorcetique grecque, pp. 291-327, esp. D.101-110 (pp. 326-
7).
See Asklepio2dotos, Techne2 taktike2, I.1 (p. 247).
8
See Ailian the Tactician, Taktike2 theo2ria, II.1 (p. 248): “kai; peri; me;n tw'n ejn tai'"
9
naumacivai" suntavxewn u{steron ejrou'men, ...”.

10
See Dain, Elien le tacticien, pp. 135-6.
11
R. Vari has shown throughout his edition that the sources that Leo did know
were the On the General, Strate2gikos, of Onasandros, the Taktike2 theo2ria of Ailian the
Tactician, and the Strate2gikon attributed to Maurice.
12
Published in Wescher, Poliorcetique, 3-40, esp. pp. 32-3.
13
See below p. 242 & n. 257.
14
See Vegetius, Epitoma, IV.31-46 (pp. 150-65).
178 CHAPTER FOUR
Naumachika that he was familiar with it.15

Seventh: the treatise On strategy, Peri strate2gike2s, once thought to
have been anonymous and composed in the age of Justinian I, now
found at folios 104-130v of the Laurentianus LV-4 manuscript and in
part at folios 8-21v of the manuscript Milan, Biblioteca Ambrosiana,
MS. B 119-sup. [Gr. 139], which is now argued convincingly to have
been part of the same treatise composed by Syrianos Magistros in the
ninth century which contained the Naumachiai Syrianou Magistrou,
signalled at §14 an intent to discuss naval warfare after land warfare
had been dealt with. This section on naval warfare was almost
certainly separated from the rest of the treatise before the tenth
century but survives in the Naumachiai Syrianou Magistrou of the
Ambrosiana manuscript at folios 333r-338v.16
Eighth: the anonymous author of the Precepts from strategical
practice, Hypotheseis ek to2n strate2giko2n praxeo2n , which consisted of
paraphrases of extracts from the Strate2ge2mata of Polyainos (second
century C.E.), in the same tenth-century manuscript Laurentianus LV-
4, folios 76v-103v, did have a chapter on naval warfare.17 However,
nothing in this chapter was used by Leo VI. He appears to have been
unaware of both it and of Polyainos, even though the treatise on
imperial military campaigns which was adapted from an earlier one by
the magistros Leo Katakylas and produced under the auspices of
Constantine VII for his son Ro2manos, the treatise known generally as
the Praecepta imperatori Romano bellum cogitanti ... observanda,
recommended that among the books which an emperor should take
with him on campaign were the works of Polyainos and Syrianos.18
None of these treatises were used by Leo VI and they are the sum
total of those known which contained any material on naval warfare.
In fact, the only earlier treatises that contained material on naval
warfare and which we have been able to establish that Leo knew and
------------------------------
15
No Greek translation is listed in Shrader, “Handlist”. According to a colophon
added to Book Four of many manuscripts of Class One, the Epitoma rei militaris was
revised at Constantinople in 450 by a Flavius Eutropius: “Fl[avius] Eutropius
emendavi sine exemplario Constantinopolim consulibus Valentiniano Augusto VII. et
Avieno.”. See Vegetius, Epitoma, pp. vi, xvii. However, from then on knowledge of
the text appears to have disappeared in Constantinople.
16
Syrianos Magistros, Peri strate2gike2s, §14.10-17 (p. 44). On the dating see also
Baldwin, “Peri Strate2gike2s”; Cosentino, “Syrianos’s «Strategikon»”, pp. 248-50;
Zuckerman, “Military compendium”.
17
See Anonymous, Hypotheseis, §57 (pp. 498-503).
18
Constantine VII, Praecepta, p. 467. Here and throughout we have used the new
edition by John Haldon: ”Osa dei' givnesqai tou' megavlou kai; uJyhlou' basilevw" tw'n
ÔRwmaivwn mevllonto" fossateu'sai. See Constantine VII, Three treatises, Text C (p.
106).
used are two fragments of earlier treatises later inserted in the same
manuscript, Milan, Biblioteca Abrosiana, MS. B 119-sup. [Gr. 139],
or some other version of them: that on crossing rivers, the Po2s dei
diapleein potamous ..., excerpted from the Strate2gikon of Maurice and
the Naumachiai Syrianou Magistrou of Syrianos Magistros.19 This can
hardly have been a coincidence.
The Strate2gikon attributed to Maurice contained a chapter on
crossing rivers in the face of the enemy, Po2s dei diapleein potamous,
which was part of a separate small treatise on infantry added to the
text as Book XII later.20 This treatise was transmitted in three
manuscript groups, of which the major surviving one is once again
Laurentianus LV-4, folios 3r-67v, where the treatise was attributed to
Urbikios. However, another version was incorporated into Milan,
Biblioteca Ambrosiana, MS. B 119-sup. [Gr. 139], at folios 6r-88v.
Subsequently, the compiler re-worked the chapter on crossing rivers at
Book XII.B.21 from the treatise and included this version in the
section on naval warfare at folios 331v-332v. The excerpt concludes
halfway down folio 332v, indicating that the manuscript still has the
whole of the text that interested the compiler. Leo VI knew and used
the chapter on crossing rivers but he also knew and used other parts of
the Strate2gikon.21 He must have had access to a complete text of it.
The chapters on naval warfare attributed to Syrianos Magistros, the
Naumachiai Syrianou Magistrou, contained in the Ambrosiana
manuscript at folios 333r-338v are unique to that manuscript. No
other text of them is known. However, they have now been identified
almost certainly as having originally been part of a treatise on strategy
composed by a certain Syrianos Magistros. The treatise is incomplete
at the beginning because a page has been lost from the manuscript
between what is now folio 332v and what is now folio 333r.22 Over
------------------------------
Both published by Dain from the manuscript under the rubrics Ek tou'
19
Maurikivo u pw'" dei' diaplevein tou;" potamou;" kai; ta;" diabavsei" aujtw'n poiei'sqai
ejcqrw'n ajntikaqistamevnwn (From Maurice, how you should sail across rivers and
make crossings when the enemy resist) and Naumacivai Surianou' Magivstrou (Naval
battles of Syrianos Magistros) in Naumachica, pp. 41-2 & 45-55, respectively. See
various notes to Appendix Two [a] which indicate sections where the treatise of Leo
VI was indebted to these earlier works.
20
See Maurice, Strate2gikon, XII.B.21 (pp. 468-73) and for convenience the
comments on the manuscripts in Dennis’s translation, pp. xxviii-xx.
21
See below p. 395 & n. 652.
22
The treatise now ends at the foot of fol. 338v, at the end of a paragraph but
without the normal explicit such as is found at the end of the Maurice text. There are
also stubs remaining from now missing folios after 338. The Syrianos text was
originally written on one quaternion of the manuscript, from which the outer leaves at
beginning and end have been lost.
180 CHAPTER FOUR
the text of the Po2s dei diapleein potamous from the Strate2gikon
attributed to Maurice, on folio 332v, can be discerned an impression
of a rubricated heading which was once on the facing page which is
now lost. Dain believed that he could decipher the words “Surianou'
Magivstrou Naumacivai”.23 The compiler had the treatise to its
beginning and obviously considered it to be important and therefore
gave it a major heading. A certain Syrianos appears at the beginning
of the Taktika of Nike2phoros Ouranos as one of his sources and the
same name is cited in the Praecepta imperatori Romano bellum
cogitanti ... observanda, attributed to Constantine VII.24 Leo VI knew
it but did he have access to the complete text or was what he had
already only the now separated section? Most probably he had the
whole since he also used the treatise Rhetorica militaris, which is also
contained in the Ambrosiana and Laurentian manuscripts, in
Constitution XVIII of his Taktika. In fact, the Naumachiai Syrianou
Magistrou, the Peri strate2gike2s, and the Rhetorica militaris, are now
all considered to have been part of the same treatise by Syrianos
Magistros. Nikephoros Ouranos also used the Peri strate2gike2s at
§§74-87, and the Naumachiai Syrianou Magistrou at §§119-21, of his
own Taktika, although he may conceivably have received them by an
independent tradition. Zuckerman considers that the treatise should be
dated to the sixth century; however, Cosentino, and we also, consider
that it should be dated to the ninth century. The Naumachiai Syrianou
Magistrou does not suit the sixth-century context but sits admirably in
that of Byzantine-Muslim naval conflict in the ninth century.25
When the compiler of the Ambrosiana manuscript came to the task
of assembling treatises on naval warfare, the only ones that predated
Constitution XIX of Leo VI’s Taktika and that he knew and included
were the Po2s dei diapleein potamous ... from Maurice and the sections
on naval warfare of Syrianos Magistros, significantly, the only
treatises on naval warfare also known to have been known to Leo VI.
The emperor extrapolated from some classical Greek historical
texts,26 and his language was studded throughout with classical
allusions and archaic, anachronistic terminology. Some of his advice
to his strate2goi also reads like that of an arm-chair sailor dreaming up
------------------------------
23
See Dain, Naumachica, pp. 43-4; idem, “Stratégistes”, p. 342.
24
Constantine VII, Praecepta, p. 467; idem, Three treatises, p. 107. See also Dain,
“Stratégistes”, p. 342.
25
See Cosentino, “Syrianos’s «Strategikon»”; Zuckerman, “Military compen-
dium”.
26
See Appendix Two [a], §45 and n. 32.
stratagems for naval warfare in front of a fire in the imperial palace.

For example, at §21, where Leo instructed his strate2goi to be aware of
the character and bravery of each soldier under his command. All very
well, but how could any commander ever acquire such knowledge
about each of thousands of soldiers or sailors? Similarly, at §74 the
emperor instructed them to ensure that the crews of the dromons of
their fleets should be no smaller than, and preferably larger than, those
of the enemy.27 This was almost certainly based on what he had read
in the treatise of Syrianos Magistros, and was all very well but, even
given the fact that the Byzantines employed spies and had intelligence
systems,28 as did other medieval military and naval powers, how could
any fleet commander preparing for battle know how large the crews of
the enemy’s ships would be? One might make a reasonable guess on
the basis of the known size of the enemy ships and the normal crews
of such ships, but no commander could ever know how his enemy
counterpart might crew his ships in the approach to battle.29 The
normal number of marines or oarsmen might well augmented by
supernumeraries, as the Byzantines themselves apparently did in the
case of the Cretan expeditions of 911 and 949.30 No doubt the
emperor’s advice was theoretically appealing, but it was very
impractical. There were a series of other stratagems suggested in
§§69-71 of Constitution XIX which also read very much as though
they were conceived by an arm-chair sailor.31
In the same Ambrosiana Library MS. B 119-sup. [Gr. 139] a short
collection of five paragraphs drawn from Constitution XX, §§196,
201, & 220 of the Taktika and from its Epilogue, §§44, 45, & 47 fin.
was included at folios 331r-v under the rubric From the Lord Leo, the
Emperor, Ek tou kyrou Leontos tou Basileo2s.32 The compiler of the
manuscript went to the trouble of culling the rest of the Taktika for
material relating to naval warfare not found in Constitution XIX.
Leo VI’s Constitution XIX was closely paraphrased by the
magistros Nike2phoros Ouranos as chapter 54, entitled Peri
------------------------------
27
See Appendix Two [a], §§21, 74.
28
See Christides, “Military intelligence”; Koutrakou, “Diplomacy and espionage”;
idem, “Spies of towns”.
29
The text of Syrianos Magistros merely amounted to general advice to have good
intelligence and this was taken up again later by Nike2phoros Ouranos. See Appendix
One, §9.8; Nike2phoros Ouranos, Ek to2n taktiko2n, §119.4 (8) (p. 94).
30
See below pp. 262-6, 370-71. However, note our remarks on the disruptive
effects this would have on the delicate gearing of oarage systems, pp. 262-3.
31
See below pp. 204-7, 387-406.
32
Published by Dain in Naumachica, pp. 37-8. See Appendix Two [b].
182 CHAPTER FOUR
thalassomachias by Dain, of his own Taktika, which was composed

around 1000-1011 during his tenure of the governorship of Antioch as
oJ kratw'n th'" Anatolh'" (ho krato2n te2s Anatole2s), “He who holds the
East”, the commander of Byzantine forces in the East.33 Nike2phoros
included two short passages found in the Ek tou kyrou Leontos tou
Basileo2s but not in the Naumachika Leontos Basileo2s,34 indicating that
he probably knew the Ambrosiana Library manuscript or one of its
ancestors; although, it is of course possible that he included them from
the manuscript of the entire Taktika of Leo VI to which he had access.
Nike2phoros’s text is a very useful check on that of Leo VI because his
paraphrase clarified some obscure points in that of his predecessor.
His language and syntax were more simple and down to earth and it is
clear that, as both an educated man and also a practised general, he
was able to interpret the emperor’s text for the use of soldiers, even if
on occasions he remained faithful to his imperial mentor’s impractical
advice.35 He frequently eliminated Leo VI’s classical allusions or
translated his archaic terminology into contemporary terms.
There are also problems associated with the surviving texts of
Nike2phoros’s treatise. All of the earliest manuscripts are unpublished
and the earliest of them all, Munich, Bayerische Staatsbibliothek,
Cod. Monac. Gr. 452, dates only from the fourteenth century. Dain
had wished to publish his edition from this manuscript but his
------------------------------
33
McGeer, “Tradition and reality”. See Appendix Five, esp. n. 1.
34
See Appendix Five, nn. 42, 44.
35
On several occasions Nike2phoros Ouranos reiterated the essence of what Leo VI
had written when, as an experienced general himself, he ought to have known that the
emperor’s advice was, as we consider, impractical. On Nike2phoros’s career, see Dain,
Nicéphore Ouranos, pp. 133-6; McGeer, “Tradition and reality”, pp. 130-31.
However, as far as is known, Nike2phoros’s experience in war was entirely land
based. His major achievements were against Tsar Samuel of Bulgaria and against the
Muslims on the land frontier in Syria and Armenia. There is no evidence that he had
any practical experience in naval warfare and this may help to explain why he
reiterated some of Leo VI’s impractical suggestions.
Almost the entire text of ch. 54 of Nike2phoros’s Taktika was a close paraphrase
of Leo VI’s Constitution XIX of his own Taktika. No doubt, considerations of having
his own Taktika accepted for “publication” in the highest circles of Byzantine court
society meant that Nike2phoros could not have cast aspersions on what a revered
former emperor was well known to have either written himself or to have been
responsible for the writing of. Nike2phoros was, after all, writing during the reign of
Basil II, the great grandson of Leo VI. His capacity to emend the text of Leo VI was
limited to paraphrase and clarification.
In fact, not only chapter 54 but also all of chapters 1-55 of Nike2phoros’s Taktika
were a paraphrase of Leo VI’s Taktika and chapters 56-62 were a paraphrase of the
Praecepta militaria of emperor Nike2phoros Pho2kas. Nike2phoros Ouranos did add
some original material of his own, but it was very limited. See McGeer, “Tradition
and reality”, esp. pp. 132-8; idem, Dragon’s teeth, esp. pp. 79-86.
transcripts were lost in the Second World War and he was compelled
to use those he had made from the sixteenth-century manuscript,
Florence, Biblioteca Medicea Laurenziana, MS. Laurentianus LVII-
31, which had been copied from Cod. Monac. Gr. 452 at Corfu by
Antonios Eparchos in 1564. We have edited our text from Cod.
Monac. Gr. 452. However, because even this manuscript postdates the
original time of composition by Nike2phoros Ouranos by over 300
years, not surprisingly we have have been able to identify several
points at which we believe that errors have surely crept in between
what should have been Nike2phoros’s original text and that of the
fourteenth-century manuscript.
In the manuscript Oxford, Bodleian Library, MS. Baroccianus
Graecus 131, dated to the first half of the fourteenth century, which
contains many other treatises, there was included another excerpt from
the Taktika of Nike2phoros Ouranos containing chapters 119-123,
which were largely concerned with naval warfare.36 However,
chapters 119 and 122 were simple paraphrases of Syrianos Magistros
and Maurice respectively and 120, 121, and 123 were a collection of
exempla from classical history. The excerpt has been of only limited
use to us.
At face value, the only treatise which described the actual
construction of dromo2nes and chelandia was the anonymous treatise
Naval warfare, commissioned by the patrikios and parakoimo2menos
Basil, executed by an anonymous client, “the Anonymous”, for the
patrikios and parakoimo2menos Basil Lekape2nos, an illegitimate son of
the emperor Ro2manos I who rose to great power from ca 947 to 959
under Constantine VII, and again under Nike2phoros II Pho2kas, John I
Tzimiske2s, and Basil II from 963 until his overthrow in 985. This
treatise was an unashamed attempt to impress an important patron by
parading a knowledge of classical Greek.37
The only surviving medieval manuscript of this text is contained at
folios 339r-342v of the same Ambrosiana Library, MS. B 119 sup.
[Gr. 139] manuscript which contains the Naumachika Leontos
Basileo2s of Leo VI. It is possible that the text of the Anonymous’s
treatise transcribed in this manuscript was the first part of the original
------------------------------
36
Folios 269r-272v. Published by Dain, Naumachica, pp. 89-104.
The Anonymous said that the construction of dromo2nes and chelandia used the
37
same types of timbers, even if two different names were used for the ships. See
Appendix Three, §2.16: “Au|tai me;n aiJ ojnomasiva i oijkei'a i celandivo u kai; drovmwno": ejk
tw'n ajutw'n ga;r nhi?wn xuvlwn ajmfotevrwn aiJ kataskeuai; givnontai, eij kai; peri; th;n
kaqovlou klh'sin dienhnovcasi: kai; to; me;n drovmwn wjnovmastai, to; de; celavndion.”. On
Basil the parakoimo2menos see Brokkaar, “Basil Lacapenus”.
184 CHAPTER FOUR
fair copy of the author’s holograph. As it survives the treatise is

incomplete because the manuscript ends in mid sentence at the bottom
of folio 342v.
In the poem with which the Anonymous opened his treatise, he
referred to Basil having defeated “Chambdan”, that is Sayf al-Dawla
‘Alı3 I, the H4amda2nid amı3r of Aleppo (945-67), and to Crete being at
the time still in Muslim hands.38 This enables us to date the
composition of the treatise very precisely. Under Constantine VII,
Basil was patrikios and parakoimo2menos, and paradynastevo2n te2s
synkle2tou, “co-president of the Senate”. He was the highest ranking
eunuch in the government, “Prime Minister” as he has been called. 39
He had participated with John Tzimiske2s, strate2gos of Anatolikon, in a
victorious campaign against Samosata, which was subject to Sayf al-
Dawla, in October/November 958.40 In November 959 Constantine
VII died and Basil was excluded from power by Constantine’s son
Ro2manos II. He returned to power only in 963, when he became
proedros and parakoimo2menos under Nike2phoros II Phokas. The final
campaign to recover Crete departed in July 960 and was brought to a
successful conclusion in March 961. Now, obviously the planning of
the expedition of 960-61 had taken some time and we suggest that the
Anonymous’s treatise was compiled for Basil between his return from
the Samosata campaign and his fall from grace: between November
958 and November 959. Because the Anonymous writes as though
Basil expected to participate in the Cretan campaign, he must have
completed his treatise before Basil’s fall from power. This is
confirmed by his addressing Basil as “the valiant attendant of our
valiant emperor”.41 This must have been written before the death of
Constantine VII. Moreover, since the treatise was addressed to him as
patrikios and parakoimo2menos rather than proedros and parakoi-
mo2menos, it cannot have been compiled after his return to power in
963. The redaction of the fine copy of the treatise in the form that we
have it may possibly have post-dated 963 but the original composition
of it can not have. More probably, the redaction of the entire surviving
manuscript was completed between November 958 and November
959.42
For several reasons, it is most probable that the Anonymous was a
------------------------------
38
See Appendix Three, opening poem.
39
Brokkaar, “Basil Lecapenus”, p. 213.
40
See Vasiliev/Canard, Byzance et les Arabes. Tome II, part 1, pp. 362-4.
41
See Appendix Three, Preface.4.
42
Cosentino, “Syrianos’s «Strategikon»”, pp. 244-5.
young member of Basil’s own household. First, he associated himself

with those who had taken pleasure in Basil’s achievements.43
Secondly, he referred to his own “youthful exuberance”.44 Thirdly, his
whole treatise reads very much as though it was a rather juvenile
exercise in book learning composed by someone whose education did
not sit lightly on him. For example, on occasions he employed
Platonic and Aristotelian terminology which reads awkwardly in such
a treatise, as though the author was using it pretentiously and self-
consciously.45 Also, when technical terms are used in the manuscript,
sometimes there are spaces left in the lines of writing before the
definitions of the terms, as though to emphasize the lexicographical
nature of the work. Fourthly, although the Anonymous claimed to be
using a variety of historical and strategical texts,46 we have been able
to identify positively only five texts used by him: the Naumachika
Leontos Basileo2s of Leo VI, the Onomasticon of Julius Pollux,
professor of rhetoric at Athens from ca 178 C.E., the Lexicon of
Hesychios of Alexandria (fifth-sixth centuries C.E.), a manuscript of
Thucydides which had scholia, and a manuscript of Homer’s Odyssey
which also had scholia.47 We have been able to find no evidence that
he had access to a manuscript of the Iliad, to any of the ancient
treatises on tactics, to the classical commentaries on Homer and
Homeric dictionaries by Aristarchos of Samothrace (ca. 217-145
B.C.E.), Apio2n (1st century C.E.), and Apollo2nios Sophista (ca 100
C.E.), or to any of the surviving encyclopedic and other treatises
compiled under the auspices of Constantine VII. His sources appear to
have been those of a rather limited private library.
The Anonymous derived most of his information about ships from
a manuscript of the Onomasticon, the “word book”, of Julius Pollux,
one of which had been copied in the ninth century in literary circles in
Constantinople and had then been owned and interpolated by bishop
Arethas of Caesarea (mid 9th century - 932/44).48 A manuscript of
Hesychios may also have been known to Arethas.49 The fate of
------------------------------
43
44
45
See Appendix Three, §§1.1, 1.2.
46
47
These sources are identified in the notes to Appendix Three.
48
First noted by Dain. See Naumachica, p. 58.
49
This manuscript of Pollux is now lost and of the surviving manuscripts, all of
which are incomplete, interpolated, and abridged, the earliest is dated to the tenth
century. See Pollux, Onomasticon (Bethe), vol. 1, pp. vi-vii. How much Pollux’s
original work may have been altered in the ninth century, or in earlier transitional
manuscripts, is impossible to say. On Arethas’s library, see Wilson, Scholars, pp.
186 CHAPTER FOUR
Arethas’s library after his death is unknown; however, it is arguable

that Basil the parakoimo2menos himself had acquired these
manuscripts from it and that the library that the Anonymous had
access to was Basil’s own. Pollux himself had used classical Greek
sources such as Plato, Thucydides, Herodotos, and Homer and the
section of the Onomasticon dealing with ships was never intended to
be a technical treatise. What Pollux had written about the construction
of ships reflected that of a trie2re2s of Themistoclean Athens more than
that of a Roman liburna of his own time.
Either the redactor of the Pollux manuscript, or even perhaps
Pollux himself, had misunderstood some of the classical Greek
terminology for ships and the Anonymous then misunderstood parts
of the Pollux manuscript. Far from describing the construction of a
tenth-century dromon, the text of the Anonymous as it survives
essentially describes a Greek trie2re2s but with numerous errors. It is
not a shipwright’s manual but rather an exercise in philology. That
being said, on the one hand, some of the characteristics ascribed to
tenth-century dromons and chelandia by it can be accepted because
they are confirmed by other sources and some of the terminology used
is so common to ships of all kinds that it can be presumed to be
relevant to dromons also. On the other hand, quite a few of the
characteristics ascribed to dromons by it can be rejected because they
are clearly anachronisms misunderstood. Many others lie in the realms
of probability or possibility.
To these sources can be added a series of works compiled either by
Constantine VII himself, or under either his auspices or those of Basil
the parakoimo2menos, including: the De administrando imperio;50 the
Praecepta imperatori Romano bellum cogitanti ... observanda,51 the
treatises produced as part of the encyclopedic Excerpta historica;52
and the anonymous collection of imperial biographies known as
Theophanes continuatus.53
However, the most important of the Porphyrogenne2tan treatises for
Byzantine naval affairs are some inventories, supposedly for the
unsuccessful expeditions to recover Crete in 910-12 led by the
patrikios and logothete2s tou dromou Himerios and in 949 led by
Constantine Gongyle2s , and for two other expeditions to Italy in 934
------------------------------
120-35.
50
Constantine VII, De administrando imperio.
51
Constantine VII, Three treatises, Text C.
52
Constantine VII, Excerpta historica.
53
Theophane2s continuatus.
and 935. These were inserted in the treatise De cerimoniis, On [court]

ceremonies, which is traditionally ascribed to Constantine VII but
which was actually compiled in its final form under the auspices of
Basil the parakoimo2menos during the reign of Nike2phoros II Pho2kas. 54
Attention has long been drawn to problems inherent in the De
cerimoniis compilation in general, and in these inventories in
particular. Why were the inventories compiled and from what
sources? As they survive, the various inventories represented different
chronological stages in the organization, and recording of that
organization, of these expeditions. Some appear to have anticipated
what ought to be done in the future. Others appear to have recorded
auditing after the event of what had been done. Still others can be read
as though they were standard “check lists” for such expeditions.55
Why were other bureaucratic sources which may be presumed to have
existed not included among the inventories? Why were the inventories
that were included inserted in a text to which they clearly did not
belong? Why are they ordered in the way that they are in the main
surviving manuscript: Leipzig, Univ./Urb., MS. 28 [Rep.i.17]? These
and many other questions remain unresolved. We have used the new
edition by Haldon, which incorporates some modifications to Reiske’s
text,56 but in Appendix Four have made our own translations of those
sections of the inventories for the expeditions of 911 and 949
concerned with the ships and their equipment and armaments.
Other treatises, such as the De expugnatione Thessalonicae of John
Kaminiate2s ,57 also date from the period and add some useful further
information. So also do scholia on various classical authors, most of
the earliest manuscripts of which are dated to the tenth or early
eleventh centuries. Leo the Deacon also referred to fire-bearing
------------------------------
54
Constantine VII, De cerimoniis, II.44 and 45 (vol. 1, pp. 651-78).
55
See Haldon, “Theory and practice”, pp. 236-9, 255, 267-8.
56
See Haldon, “Theory and practice”, pp. 202-35. See also Bury, “Ceremonial
book”, p. 221; S!evc°enko, “Constantine Porphyrogenitus”, p. 185 and n. 47;
Treadgold, “Army”; Featherstone, “Preliminary remarks”; idem, “Further remarks”.
57
John Kaminiates, De expugnatione Thessalonicae.
The authenticity of this text and its dating was questioned by Kazhdan, who
wished to redate the surviving text to the fifteenth century. However, its authenticity
has been defended by Christides and Khoury Odetallah. See Kazhdan, “Some
questions”; Christides, “Caminiates”; Khoury, “Leo Tripolites-Ghula2m Zura2fa”. See
also John Kaminiates, De expugnatione Thessalonicae, trans. Frendo, pp. xxxvii-xl;
Frendo, “St Demetrius”.
We add that the information about ships, shipping, and naval warfare found in
the treatise suits the context of the tenth century far better than that of the fifteenth.
That does not, of course, preclude the possibility that the surviving text may derive
from a later reworking of a tenth-century original.
188 CHAPTER FOUR
trie2reis which the “Romans” [of his day] called dromo2n es.58
(b) Terminology and Ship Types
In the Porphyrogenne2tan corpus, the terminology used for ship types

was very varied. On the one hand, in the De administrando imperio
the author used chelandion throughout as the word for a warship
except in §51, where he used dromwvnion (dromo2nion) for the imperial
galley.59 Dromo2n was similarly used in the De cerimoniis with
reference to the “imperial dromon” whose crew was given a donative
during the festival of the broumalia before Christmas.60 On the other
hand, in the Praecepta imperatori Romano bellum cogitanti ...
observanda, dromo2nion and dromo2n were used for warships. 61
Chelandion was not used at all in this text. In the various parts of the
so-called Theophane2s continuatus the usage was variable. Dromo2n is
found once in the first part compiled by the continuator of
Theophane2s the Confessor and four times in the third part, which is
similar to the chronicle of Symeon Logothete2s. It is also found once in
the fourth part, probably written by Theodore Daphnopate2s, with
reference to the invasion fleet sent to Crete under Nike2phoros Pho2kas
in 960.62 However, it is not found at all in the second part, the so-
called Vita Basilii, “Life of Basil I”, compiled under the auspices of
Constantine VII. Chelandion is found only once in any part of the
Theophane2s continuatus but in part four karabia was used for
------------------------------
58
Leo the Deacon, Historiae, I.3 (p. 7): “... kai; tacuplohvsa", purfovrou" te
trihvrei" pleivsta" ejpagovmeno" (drovmona" tauvta" ÔRwmai'oi kalou'si), ...”. See also
below pp. 308-9.
59
Constantine VII, De administrando imperio, §8 (p. 54): “Kai; basilikou'
ajpostellomevnou ejnteu'qen meta; celandivwn, ...”; §42 (p. 182): “... spaqarokandidavtou
Petrwna' meta; celandivwn basilikw'n plwi?mwn ajpevsteilen kai; celavndia tou' katepavnw
Paflagoniva".”. Cf. §29 (pp. 126-7) and §51 (pp. 246-7): “Peri; tou', tivni trovp w/
gevgonen to; basiliko;n dromwvnion, ...”.
60
Constantine VII, De cerimoniis, II.18 (vol. 1, p. 601).
61
Constantine VII. Three treatises, Text (C), ll. 321, 686, 827 (pp. 114, 138, 146).
62
Theophane2s continuatus, IV.44 (p. 208); VI.Basileiva Levonto" aujtokravtoro".9
(p. 358), VI.Basileiva Kwnstantivnou uiJo u' Levonto".11 (p. 391), VI.Basileiva
Rwmanou'.23 (p. 414), VI.Basileiva Rwmanou'.39 (pp. 423-4), VI.Basileiva Rwmanou'
uiJo u' Kwnstantivnou tou' porfurogennhvtou.10 (p. 475).
Theodore Daphnopate2s held high positions at court during the reign of Ro2manos
I Lekape2nos. He lost influence under Constantine VII but under Ro2manos II was
eparchos, or urban prefect, of Constantinople. On the question of whether he was the
author of the fourth part of the Theophane2s continuatus, see Theodore Daphnopate2s,
Correspondance, pp. 6-10.
warships, both Muslim and Byzantine.63

In a letter to Pope John XI (931-35) penned by Theodore
Daphnopate2s for Ro2manos I Lekape2nos in February 933, the Pope was
invited to send his sister to Constantinople on chelandia to be
provided by the Empire to marry one of the emperor’s sons, probably
Constantine, the youngest.64
In the inventory for the Cretan expedition of 911-12 chelandion
was used only for two ships supplied by the strate2gos of the thema of
the Kibyrrhaio2tai. The overwhelming majority of this fleet was said to
be composed of dromo2n es and pavmfuloi (pamphyloi).65 However, in
the inventory for the fleets sent to Italy in 934 and 935, the squadrons
were described respectively as being composed of 11 chelandia, and
11 chelandia plus seven karabia, which were transporting 415 Rho2s.
Here the transports appear to have been the karabia and the warships
to have been the chelandia.66 Finally, in the inventory for the Cretan
expedition of 949, the ships were referred to variously as pamphyloi,
oujsiaka; celavndia (ousiaka chelandia), celavndia pavmfula (chelandia
pamphyla), and dromo2nes.67
In Byzantine Italy, the Life of St Neilos of Rossano (ca 910-1004)
reveals that chelandia were constructed by the doux Nike2phoros
Magistros as part of a naval defence force in the tenth century.68
Liudprand of Cremona knew Byzantine warships from his first visit to
Constantinople in 949 and referred to those that used Greek fire as
chelandia. According to him, it was fifteen derelict old chelandia
which had been hastily repaired and equipped with Greek Fire which
scattered the Rho2s attack on Constantinople in 941.69 However, it was
------------------------------
63
Theophane2s continuatus, III.28 (p. 123); VI.Autokratoriva Kwnstantivnou.29 (p.
453); VI.Basileiva Rwmanou' uiJou' Kwnstantivnou tou' porfurogennhvtou.7, 10 (pp. 473,
475).
64
Theodore Daphnopate2s, Correspondance, letter 1 (pp. 40-41).
The letter was really meant for the Pope’s half brother Count Alberic of
Tusculum, the current master of Rome, who had imprisoned the Pope’s mother, the
infamous Senatrix Marozia, and who had himself sought his own marriage to one of
Ro2manos’s daughters. The letter invited the Pope to have his mother escort and
chaperone his sister. The proposal was a diplomatic ruse which both sides knew
would not be pursued, as indeed it was not.
65
Appendix Four [a], §§2-7, 13 [= Haldon, “Theory and practice”, pp. 203, 205,
209; Constantine VII, De cerimoniis, II.44 (vol. 1, pp. 652-4, 657)].
66
Haldon, “Theory and practice”, p. 213; Constantine VII, De cerimoniis, II.44
(vol. 1, p. 660).
67
Appendix Four [b], §I [= Haldon, “Theory and practice”, pp. 219, 221;
Constantine VII, De cerimoniis, II.45 (vol. 1, pp. 664-5)].
68
Vita S. Nili, coll. 105-8. See also Ahrweiler, Byzance et la mer, p. 116.
69
See Liudprand of Cremona, Antapodosis, V.9 (p. 135): “Quam ob rem Hugo rex
consilio accepto nuntios Constantinopolim dirigit rogans imperatorem Romano2n, ut
190 CHAPTER FOUR
probably in Byzantine Italy that Westerners in general became

familiar with the chelandion and emulated it as the chelandia/
scelandrium/salandria, etc.70
Muslim authors of the ninth and tenth centuries, including Ibn
H4awqal (died post 988 C.E.) and al-Muqaddası3 (al-Maqdisı3) (died 988
C.E.), also used varieties of the word for Byzantine war galleys.71
Smaller light galleys also existed. They were classed generically by
Leo VI as dromo2nes but specifically termed galevai (galeai) or
monhvrei" (mone2r eis) by him, and were said both by him and by the
Anonymous to be used especially for scouting purposes.72 Nicholas I
Mystikos also wrote in his letter of 920 to Ro2manos Lekape2nos that
at least one or two mone2r eis should be sent to the relief of Lampsakos
if chelandia were unavailable.73 The Vita Basilii of the Theophane2s
continuatus referred to ships of the Cretan corsairs as muopavrwne"
(myoparo2nes), a classical Greek word for a fast pirate ship, and
penthkovntoroi (pente2kontoroi), i.e. fifty-oared ships, and said that
they were commonly known as sa(k)touvrai (sa(k)tourai) and galeai.
In the first part, the continuator of Theophane2s also referred to ships
of the Cretan Muslims as galeai and in the fourth part the author
referred to the galeai of the Byzantine Cretan expedition of 960 as
being fast sailing and being used for scouting purposes.74 In the
inventory for the Cretan expedition of 911-12, it was said that the
katepano2 of the Mardaites of Antalya undertook to provide galeai for
------------------------------
naves sibi Greco cum igne transmittat, quas chelandia patrio sermone Greci
cognominant.” [written 958-62]. See also ibid., V.15 (p. 138). Cf. idem, Relatio, pp.
190, 192, 193.
70
Chronicon Salernitanum, §107 (p. 107): “Basilius imperator Grecorum ut
huiusmodi verba captasset, valde gavisus est, atque sine mora non pauca scelandria
misit, ...” and cf. §107 (p. 120) [written ca 974]; Thietmar of Merseburg, Chronicon,
III.23 [13] (pp. 126-127): “Sed ut in omnibus, lector carissime, certus efficiaris,
salandria quid sit vel cur ad has pervenerit horas, breviter intimabo. Haec est, ut
prefatus sum, navis mirae longitudinis et alacritatis, et utroque latere duos tenens
[habet] remorum ordines ac centum quinquaginta nautas” [written ca 1000-1018];
Anonymous chronicle of Bari, p. 152 (“Et chelandie incenderunt nave, que veniebat
de Calabria.”) and cf. p. 153 [written ca 1115 but based on much earlier sources].
71
Ibn H4awqal, S5u2rat al-Ard 5, pp. 197-8; al-Muqaddası3, Ah5san, p. 177.
72
Appendix Two [a], §10. Cf. §81; Appendix Three, §3.2. Cf. Appendix Five, §§9,
74.
73
See above p. 167 & n. 20.
74
Theophane2s continuatus, IV.34 (p. 196); V.60 (p. 299): “prosh'n de; aujtoi'"
ajnalovgw" kai; plh'qo" muoparwvnwn kai; penthkontovrwn, a}" saktouvra" kai; galeva "
ojnomavzein eijwvqasi pavmpolloi.”; VI.Basileiva Rwmanou' uiJou' Kwnstantivnou tou'
porfurogennhvtou.10 (p. 475): “e[mprosqen de; oJ suneto;" tacudrovmou" galeva"
ajposteivla" kataskoph'sai kai; krath'sai glw'ssan prosevtaxen.” and 11 (p. 477): “oiJ
de; ajposteivlante" galeva" eujq udrovmou" ...”.
Syria; however, no details about these were supplied.75 In that for the
Cretan expedition of 949, it was said that of the 15 galeai of Antalya,
six were left behind to protect the thema.76 Makrypoulias has argued
that when the galea first appeared in the sources in the tenth century,
the ship was especially associated with the Mardaites of Antalya,
Antioch on Cragus, and Karpathos, and that it was developed by them
as a corsair and scouting galley.77 Both galeai and also chelandia were
mentioned as being used by the Mardaites in a short text which may
have been a submission from the strate2gos of the Kibyrrhaio2tai to
Constantine VII on weather and seasonal navigation made at the
emperor’s request, a text now found among the final folios of a
manuscript of the late fourteenth or early fifteenth century which once
belonged to Cardinal Bessarion: Venice, Biblioteca Marciana, MS.
Gr. 335 [coll. 645].78 According to the tenth-century Life of St
Theodore of Kythe2ra, the tourmarche2s Melito2n was said to have been
sent to Crete around 920-21 by Ro2manos I Lekape2nos with four
chelandia, then glossed as die2reis. Later a chelandion put in to
Kythe2ra and found the saint dead.79
Pamphyloi were probably dromons or chelandia of a type
originally associated with the province of Pamphylia around the Gulf
of Antalya or alternatively with squadrons of the navy having picked
crews, since the two elements of the word, pa'" “all”, and fu'lon “tribe
or group”, might refer to a crew selected from all tribes or localities,
or all crews, as Leo VI implied.80 The province of Pamphylia was so-
called in ancient Greek because it was reputed to contain tribes or
peoples from all areas. In the inventory for the Cretan expedition of
------------------------------
75
Appendix Four [a], §13: “... , oJ de; katepavnw tw'n Mardai>tw'n eujtrepivsh/ galeva",
...” [= Haldon, “Theory and practice”, p. 209: Cf. Constantine VII, De cerimoniis,
II.44 (vol. 1, p. 657)].
76
Appendix Four [b], §I.12: “galevai th'" Attaliva" ieV. ejx aujt w'n kateleivfqh eij"
fuvlaxin tou' qevmato" galevai ıV.” [= Haldon, “Theory and practice”, pp. 219-21;
Constantine VII, De cerimoniis, II.45 (vol. 1, p. 665)].
77
See Makrypoulias, “Navy”, pp. 160-61.
78
The text was published by Lambros in “Triva keivmena sumbavllonta”, pp. 171,
173. On the dating to the age of Constantine VII and the attribution of the text to a
work on meteorology possibly compiled for him, see Dolley, “Meteorology”.
In the Praecepta imperatori Romano bellum cogitanti ... observanda, it was
recommended that an emperor take with him on campaign a treatise “which includes
information on fair weather and storms and squalls and rain and lightning and thunder
and the direction of winds” (our trans.). See Constantine VII, Three treatises, Text C
(p. 106). It is possible that the Lambros fragments were later incorporated in this
treatise, which is now apparently lost if it was ever in fact actually compiled.
79
See Vita S. Theodori, p. 287.
80
Appendix Two [a], §42. Cf. Ahrweiler, Byzance et la mer, pp. 415-17;
Alexandres, ÔH qalassiva duvnami", p. 74.
192 CHAPTER FOUR
911-12 pamphyloi were said to have had either 160 or 130 men, it not
being specified whether these were oarsmen or marines.81 The number
of men required rules out the possibility that they were sailing ships.
Such numbers can have been needed only for oared ships. However,
on the one hand, since no specific mention was made of marines for
them, in this fleet they were probably used as transports rather than in
battle. On the other hand, Leo VI suggested that pamphylos became
applied to picked crews selected for the personal dromons of
strate2goi; that is, if he was not just engaging in a piece of
etymological word play. Even if he was, the idea of picked crews for
fleet commanders’ own galleys was reiterated by Nike2phoros Ouranos
and the Arabic translator of Leo VI used by the fourteenth-century
Mamlu2k Egyptian official Muh5ammad ibn Mankalı3 in his Al-ahka2m
al-mulu2kyya wa ’l dawa2bit al-na2musiyya.82 Even if originally
transports, pamphyloi must soon have assumed more belligerent roles.
Their crews, possibly the descendants of the famous Mardaites settled
in Pamphylia by Justinian II, apparently acquired a great reputation.83
By the Macedonian age the term dromo2n had lost its specific
reference to a monoreme. It had become a generic for any war galley
which could take its place in the line of battle. The tenth-century
treatises are very clear that by then the standard Byzantine war galley
was a bireme; although, some small monoreme, and just possibly
some large trireme galleys, were also used. The term dromo2n could be
applied to all classes of galleys.
(c) Hull
According to the Anonymous, dromons had a druvocon (dryochon), a

keel (trovpi" [tropis]), tropivdia (tropidia), spei'ra (speira), and tropoiv
(tropoi).84 However, he was merely following Pollux here and that he
------------------------------
81
Appendix Four [a], §2: “pavmfuloi mV: ejx w|n oiJ me;n kV pavmfuloi ajna; ajndrw'n rxV,
oiJ de; e{t eroi kV ajna; ajndrw'n rlV, ...” [= Haldon, “Theory and practice”, p. 203;
Note that Treadgold assumes that these crews were all oarsmen. However, the
men of the pamphyloi are referred to simply as a[ndre" (andres), men, as opposed to
the a[ndre" kwphlavtai (andres ko2pe2latai), oarsmen, and polemistaiv (polemistai),
marines or “soldiers”, of the dromons. See Treadgold, “Army”, p. 110.
82
See Appendix Two [a], §42: “... kai; katasth'sai to;n th'" sh'" ejndoxovthto"
toiou'ton drovmwna, to;n dh; legovmenon pavmfulon.”. Cf. Appendix Five, §40; Appendix
Eight [b], p. 122.
83
See also Haldon, “Theory and practice”, p. 248, n. 45.
84
Appendix Three, §2: “Mevrh newv". §2.1: Druvocon, trovpi", tropivdia, spei'ra,
tropoiv: ...”. Cf. Pollux, Onomasticon (Bethe), I.85 (vol. 1, p. 27): “ [mevrh de; new;"]
druvocon, trovpi", [trovpide"], tropivdia, stei'ra [tropoiv].”.
Figure 19
Midships section of a dromon of the era of the Macedonian emperors.85
© John H. Pryor
really understood the meaning of all of these terms and that all of
them were actually structural elements of dromons of his age is
extremely doubtful.
In classical Greek a dryochon or dryochos was one of the stocks on
which a hull was built. But Pollux did not explain the term and the
Anonymous defined its meaning in terms which suggested a stringer.
He equated it to a koravkion (korakion), whatever he meant by that,
and defined it as: “every long continuous timber which fastens
together many other short timbers.”86 Such misunderstanding of
------------------------------
85
Dimensions based upon those calculated in Chapter Four (j): Oarage system and
dimensions.
86
Appendix Three, §2.1: “Kai; druvocon me;n su;n polloi'" a[lloi" noeivsqw te kai;
legevsqw to; kalouvmenon para; pa'si koravkion, o} sunevcei pavnta kai; sugkratei' kai; w|/
prosdevdentai kai; oiJo nei; ejpereivdontai ta; loipav. Koinw'" me;n ga;r wjnovmastai druvocon
a{pan xuvlon ejp ivmhke" sunevcon kai; oiJonei; proshlou'n e{t era braceva te kai; pollav.”.
What exactly the Anonymous meant by his equation of dryochos/dryochon with
whatever he meant by korakion is unclear. Both sentences here suggest the equation
of dryochos/dryochon with a “stringer”.
The Greek kovrax (korax) and its diminutive koravkion (korakion) were most
194 CHAPTER FOUR
dryochoi was common. The famous line of the Odyssey, XIX.574

which referred to the setting up of axes in Odysseus’ hall in a line like
the dryochoi of a ship under construction (“... i{stasc eJx eivh",
druovcou" w{", dwvdeka pavnta"”) was frequently misunderstood because
the meaning of dryochoi had been forgotten.87 Prokopios, who wrote
------------------------------
normally used in a maritime context for some sort of grappling iron or hook for
seizing on to an enemy ship. The word, in its connotation of “raven”, had been used
by the Romans in the Latin form corvus for the famous boarding bridge devised by
them to grapple Carthaginian ships during the First Punic War (264-41 B.C.E.).
However, there is no evidence for the use of the corvus as such at any time after the
Punic wars. Thereafter, the word seems to have been understood as meaning
something by which one could grapple on to, or join, any two ships together,
presumably a grappling iron for the most part. See Casson, Ships and seamanship, pp.
121-2; Morrison, Greek and Roman oared warships, p. 45. The Anonymous appears
to have understood the word in the sense of something that joined other things
together, with no more specificity than that.
Boarding bridges continued to be used in naval warfare and were known in
Greek as ejp ibavqrai, but these were not the same thing as the corvus or korax. See
Arrian, Anabasis, IV.27.1 (vol. 1, p. 430); Diodo2rus Siculus, Bibliothe2ke2 historike2,
XII.61.3 (vol. 5, p. 56).
87
See the scholia on the Odyssey. Dindorf, Scholia Graeca, T.574 (vol. 2, p. 686):
“... ] au{t w" i{sthsi tou;" pelevkei" wJ" druovcou". druvocoi de; xuvla eijsi;n ojrqa; uJpokavt w
th'" trovpido", ejf w|n ejp ereivdetai, i{na mh; aujth;n hJ yavmmo" ejsqivh/. [correct] tine;" de;
druovcou" fasi; ta; prw'ta phgnuvmena xuvla eij" nauphgivan. [incorrect] tine;" de; tou;"
pelevkei" tou;" drui?nou" steleou;" e[cein eijwqovta". [incorrect] ... druovcou"] kurivw" me;n
tou;" passavlou", ejf w|n th;n trovp in iJsta'si tw'n kainourgoumevnwn ploivwn. eJxh'" de;
mavlista ou|toi tivqentai e{neka tou' i[shn genevsqai th;n nau'n: nu'n de;, ejf w|n ejt ivqei tou;"
pelevkea". [correct]”.
See also the scholia on the Argonautika of Apollo2n ios of Rhodes, I.723-4 (“...,
o{te prw'ton druovcou" ejpebavlleto nhov" Argou'", ...”), which glossed druovcou" as: “ejn
oi|" kataphvssetai hJ trovpi" xuvloi", tau'ta ou{t w kalou'sin. ... druvocoi ou\n ta; ejgkoivlia
th'" newv". [incorrect]”. Wendel, Scholia, p. 60. The oldest surviving manuscript of
Apollo2nios of Rhodes with scholia dates to the early eleventh century.
However, the scholia on Aristophanes’ Thesmophoriazusae, l. 52 in the tenth-
century Ravenna manuscript did understand the word: “druovcou": kurivw" druvocoiv
eijsin oiJ ejntiqevmenoi pavttaloi nauphgoumevnh" newv".”. See Rutherford, Scholia
Aristophanica, vol. 2, p. 446.
Both Pho2tios and the author of the Souda understood the correct meaning of the
word in the way that the scholia on Aristophanes understood it because they had
access to a similarly correct understanding as given by a scholion on Plato’s Timaeus.
See Pho2tios, Lexicon (Theodoridis), D.767 (p. 432): “druvocoi: druovcou" ejn Timaivw/
(81b) kalei' ta; sthrivgmata th'" phgnumevnh" new'".”; Souda, D.1547 (vol. 2, p. 143):
“Druvocoi: pavttaloi, oiJ ejntiqevmenoi nauphgoumevnh" newv". ... Druovcou" Plavtwn ejn
Timaivw/ kalei' ta; sthrivgmata th'" phgnumevnh" newv".”. Cf. Greene, Scholia Platonica,
Timaeus, 81b (p. 289): “druovcwn: ta; sthrivgmata th'" phgnumevnh" new;" druovcou"
fasivn.”. Plato himself had used the word analogically and the scholion had explained
its meaning for ship construction. See Plato, Timaeus, 81b (p. 216): “neva me;n ou\n
xuvstasi" tou' panto;" zwvou, kaina; ta; trivgwna oi|on ejk druovcwn e[ti e[cousa tw'n genw'n,
...”.
By the sixth (or ninth) century, in the Philoxenos Latin-Greek glosses of Paris,
Bibliothèque Nationale, MS. Lat. 7651, dryochos was understood only as something
wooden standing up, probably as a tree trunk. See Goetz, Glossarii Latini, vol. 2, p.
188, l. 42: “Stipites stele2chdruocoi kormoi [sic]”. See Note on citations of Greek and
that he had seen a ship which was claimed to have been Aeneas’s ship
in a ship shed or arsenal, newvsoiko" (neo2soikos), which had been built
to house it on the bank of the Tiber, said that poets used the word
dryochoi for what were clearly the ship’s frames or ribs fitted to the
keel.88 He was probably referring to Homer whose usage of the word
he did not understand. The Anonymous also classed the wales,
perivtona (peritona), in the same category as dryochoi.89
Why the nautical meaning of dryochon/dryochos had become
forgotten is unknown. In our opinion it may have had something to do
with the change from shell to skeletal construction. No matter how
one goes about building a wooden ship, the keel always has to be the
first timber laid down and it obviously has to be propped up off the
ground in some way. But perhaps skeletal construction required
methods of propping up the keel different to those required for shell
construction? However, no evidence concerning this question is
known to us, neither textual, nor pictorial, nor archaeological. We
have consulted maritime archaeologists on this question but they have
not been able to assist. The pitch applied to the outer surface of hulls
hides any evidence for how the keels may have been braced when
propped up and also the keels of wrecked ships are usually so worn
from use that there is little chance of finding stock impressions.90
The keel (tropis) was straight forward; however, because tropidia
were unexplained by Pollux, the Anonymous may have either guessed
at his explanation of them as “what are attached to the keel”, or
------------------------------
Latin glossaries, p. lxix above.
88
Prokopios, History of the Wars, VIII.xxii.12 (vol. 5, p. 280): “tav te paceva
xuvmpanta xuvla ej" th;n trovpin ejnarmosqevnta (a{per oiJ me;n poihtai; druovcou" kalou'sin,
...”.
This is the only known reference to Aeneas’ ship. It is not mentioned in any
other source. See Richardson, New topographical dictionary, p. 266. Prokopios may
have been fed a line by his tour guide in Rome. Nevertheless, it is clear that he saw
some sort of ship of a type with which he was not familiar. He said that it was 120
podes, 37.48 metres, long and that its strakes were single lengths of planking. He was
clearly amazed by the combination of its length and also strakes of single planks. His
description suggests an old ship of a type with which he was unfamiliar and which
had been built somewhere, or at some time, where or when long straight lengths of
timber had still been available, as they no longer were in his own time.
89
See below p. 200.
90
But note that large numbers of partly unexplained wooden pegs and treenails
were found driven into the sides and bottom of the keel of the Kyrenia wreck of the
fourth century B.C.E. The pegs on the bottom served to attach the false keel to the
keel. However the purpose of the pegs on the sides, which were driven downwards
obliquely and trimmed even with the sides, and the three treenails in the bottom,
which were also cut off flush with the bottom surface of the keel, is unexplained.
Possibly they had something to do with how the keel was set up on its stocks. See
Steffy, “Kyrenia ship”, p. 72.
196 CHAPTER FOUR
possibly he was following the Lexicon of Pho2tios.91 Pho2tios himself

was referring to the version of the word tropidei'on (tropideion) as
found in Plato’s Laws, where it was simply a diminutive of tropis.92
However, Pho2tios had extended the definition and had written that
tropidia were timbers eujq etou'nta (euthetounta), “arranged”, on the
keel. From where he derived this is unknown. It was neither in Pollux
nor in Hesychios nor in the scholia on Plato. The Anonymous said
that they were proshrmosmevna (prose2rmosmena), “fitted to”, the keel.
The replacement of euthetounta by prose2rmosmena would not be
remarkable in an author paraphrasing a previous work; however, since
this is the only point at which we have been able to establish a
possible indebtedness of the Anonymous to the Lexicon of Pho2tios,
his use of it can be regarded as no more than a possibilty.
If the Anonymous really did mean that tropidia was a technical
term for some kind of timbers attached to the keel, that is to assume
that he actually understood hull construction, the only parts of a hull
that they conceivably could have been were either the keelson, the
floor timbers, or the garboard strakes. These were the only
components of wooden hulls actually joined to the keel and, of them,
the most likely probability is the garboard strakes. No classical Greek
word to which the meaning of “keelson” can be definitively assigned
is known to us;93 moreover, since tropidia is a plural noun, it is
difficult to comprehend why it would have been used for the singular
keelson. Floor timbers were definitely known in Greek antiquity as
ejgkoivl ia (enkoilia) and the Anonymous knew that word; although, by
following a scholion on the Odyssey and equating the word to that for
a deck, ijkriva (ikria), he showed that he did not understand its
------------------------------
91
Appendix Three, §2.2: “Tropivdia de; ta; proshrmosmevna th'/ trovpei, ...”. Cf.
Pho2tios, Lexicon (Naber), vol. 2, p. 229: “tropivdia: ta; eij" trovpin new;" eujqetou'nta
xuvla: ... kai; oJ tovpo" ejf ou| tivqetai hJ trovpi": ou{tw" Plavtwn.”.
92
See Plato, Laws, VII.803a (vol. 2, p. 52): “oi|o n dhv ti" nauphgo;" th;n th'"
nauphgiva" ajrch;n kataballovmeno" ta; tropidei'a uJpogravfetai ‹ta;› tw'n ploivwn schvmata,
...”.
93
Jal thought that Pollux had defined the keelson as a deutevra trovpi" (deutera
tropis), a “second keel”, also known as a favlkh" (phalke2s) which was fastened to the
stei'ra (steira). See Pollux, Onomasticon (Bethe), I.85 (vol. 1, p. 27): “to; de; th'/ steivra/
proshlouvmenon favlkh", [ejf ou|] hJ deutevra trovpi".” and cf. I.86 (vol. 1, p. 28).
However, Jal misunderstood Pollux’s text because he thought that stei'ra meant a
keel. See Jal, Glossaire nautique, pp. 426, 1387, 1569.
Since steira actually meant the part of the stempost known as the cutwater, no
matter what Pollux thought that he understood by phalke2s, it obviously had nothing to
do in reality with any “second keel” in the sense of keelson. The word is a hapax
legomenon and its real meaning cannot be deduced.
classical meaning.94 Therefore, the garboard strakes are the most

likely possibility. That being said, it is much more probable that the
Anonymous either had no idea whatsoever what Pollux had meant by
tropidia and simply guessed that they must have been some things
somehow attached to the keel because of the form of the word as a
diminutive of tropis, or else he was following Pho2tios.
The Anonymous altered Pollux’s stei'ra (steira) to speira. Steira
meaning “cutwater” was Homeric,95 and its classical meaning was
clearly explained by Pollux: “Between the proemboliv" (proembolis)
and the e[mbolon (embolon) is called the steira”.96 The embolon had
been, of course, the waterline ram. The proembolis was the secondary
“ram” formed above on the stempost where the upper wales had come
together and been capped.97 Since the ram was at the water line, the
steira above it had been the lower part of the stempost. Above the
proembolis was the upper part of the stempost, known as the stovlo"
(stolos) or perikefalaiva (perikephalaia), the head of the prow. 98
However, as a result of the disappearance of the ram in the
intervening centuries, the Anonymous had no idea what Pollux had
meant by all of this.99 As a result he altered steira to speira and
------------------------------
94
See pp. 200-203 below.
95
Homer, Odyssey, II.427-8 (vol. 1, p. 66): “..., ajmfi; de; ku'ma steivr h/ porfuvreon
megavl i[ace nho;" ijouvsh": ...”. Cf. Homer, Iliad, I.481-2 (vol. 1, p. 38) (exactly the
same clause).
96
Pollux, Onomasticon (Bethe), I.85 (vol. 1, p. 27): “mevson de; th'" proembolivdo"
kai; tou' ejmbovlou hJ stei'ra kaloumevnh.”. Cf. Hesychios, Lexicon (Schmidt), S.1713
(vol. 4, p. 74): “stei'ra: ... kai; to; ejxevcon th'" prwvra" xuvlon kata; th;n trovpin”.
Hesychios appears to have acquired his information from the Homeric lexicon
of Apollo2nios Sophista (ca 100 C.E.). However, the Anonymous appears not to have
known this work because he did not make use of it at times when he might well have
done so. See Apollo2nios Sophista, Lexicon, p. 144, l. 20: “stei'ra to; ejxevcon th'"
prwvra" xuvlon, dia; to; stereo;n ei\nai kai; a[rrhkton.”.
97
Casson, Ships and seamanship, p. 85. Cf. Figures 1 and 2.
98
Pollux, Onomasticon (Bethe), I.86 (vol. 1, pp. 27-8): “to; de; metaxu; tou' stovlou
kai; th'" proembolivdo" ... : ªoJº stovlo" d ejsti;n uJp e;r th;n stei'ran, o}" kai; perikefalaiva
kalei'tai.”. There is a lacuna in the manuscripts after proembolivdo".
The head of the prow was also known as the ajkrostovlion (akrostolion). See
Athe2n aios of Naukratis (fl. ca 200 C.E.), quoting Kallixeinos of Rhodes [fl. ca 155
B.C.E.], Deipnosophistae (Gulick), 5.203.f (vol. 2, p. 420) [ajkrostovlion is
incorrectly translated by Gulick as “gunwale”] and the Scholia on the Argonautika of
Apollo2nios of Rhodes, I.1089 (pp. 96-7). It is significant that the major manuscript of
Athe2n aios’ Deipnosophiste2s in the Marciana Library is again a tenth-century
manuscript. See Athe2naios, Deipnosophistae (Kaibel), vol. 1, p. viii.
99
The well-read poet, historian, and lawyer Agathias (ca. 532-80), who was also
writing well after the replacement of the Roman liburna by the Byzantine dromon,
apparently did still understand how the prow of a liburna had been made up. Writing
of the construction of reed boats by the Huns at the siege of Cherso2n, he said that in
order to make them more seaworthy they curved the prows upwards in the likeness of
akrostolia and proembola. See Agathias, Historiae, V.21 (p. 192): “wJ" a]n de; aujtoi'"
198 CHAPTER FOUR
explained the word in terms of stones at the foot of pillars, of sails,

and of clothing and shrouds, explanations which he almost certainly
derived from the Lexicon of Hesychios.100 All that he understood of
proembolis was that it had had something to do with the bow.
However, he made it an extension of the keel rather than of the upper
wales, contrasting it to something at the stern called the podovsthma
(podoste2ma), by which he almost certainly meant the sternpost.101
The tropoi were explained by the Anonymous as oar-grommets,
tropwth're" (tropo2te2res); that is the rings or loops of twisted leather or
cordage that held oars to their tholes.102 In one way, he was quite
correct in this because tropoi and tropo2te2res were in fact synonyms in
classical Greek. He probably derived his knowledge of this
synonymity from a scholion on a line of Homer in the manuscript of
the Odyssey that he had,103 but it led him into yet another instance of
misunderstanding Pollux because tropos had a second, alternative
meaning in classical Greek: a cross beam or through beam. Athe2n aios
of Naukratis, quoting an earlier account by a certain Moschio2n,
perhaps a contemporary of Hiero2 II of Syracuse, referred to the cross
beams needed to brace the hull of any ship, which projected through
the hull of Hiero2’s great ship, as tropoi.104 Pollux cited tropoi in the
------------------------------
ploi>mwvterai ei\en, oiJ de; ta; ejmprovsqia touvtwn hjrevma pro;" to; metevwron ej" prwvra"
tuvpon periagagovnte" kai; uJpokavmyante" kai; w{sper ajkrostovlia kai; proevmbola
ejkmimhsavmenoi, ...”.
However it is clear that knowledge of this kind eventually faded from memory.
Steira became a term with which philologists had difficulty. For example, the scholia
on the Odyssey were clearly struggling to make some sense of the term and apparently
knew no more than that it had something to do with the keel. See Dindorf, Scholia
Graeca, B.428 (vol. 1, p. 117): “steivrh/º th'/ trovpidi, dia; to; sterea;n ei\nai, kai; dia; to;
stereou'sqai ejn aujth'/ th;n nau'n.”.
100
Appendix Three, §2.3. Cf. Hesychios, Lexicon (Schmidt) S.1445 (vol. 4, p. 64):
“Spei'ra: oiJ pro;" th'/ bavsei livqoi. kai; th'" new;" skeu'ov" ti. kai; suvstremma ejk scoinivo u, h]
rJavkh. kai; iJmavtia. kai; iJstiva. a[lloi ei\do" iJmativou eujmevgeqe" gunaikeivou.”.
101
Appendix Three, §2.5: “Tauvth" de; to; me;n e}n mevro" ejx ou| hJ prwv/r a dianivstatai
proemboli;" kalei'tai, to; de; pro;" th;n pruvmnan podovsthma , ...”.
Podoste2ma is an otherwise unknown word. However, it almost certainly meant
the same as podostavma (podostama) used from the twelfth century for a sternpost. See
Jal, Glossaire nautique, p. 1190; Koukoulès, “Nautiko;" bivo"”, p. 346; Kahane and
Tietze, Lingua Franca, §834 (p. 560).
102
Appendix Three, §2.4: “Tropoi; de; oiJ tropwth're": kai; ”Omhro": Hrtuvnanto de;
kwvpa" tropoi'" ejn dermativnoisin.”. Note that this quotation of Homer, Odyssey,
IV.782 does not correspond to the modern received editions. The Homeric “ejretmav”
for the oars has been replaced by “kwvpa"”.
103
See Dindorf, Scholia Graeca, D.782 (vol. 1, p. 234): “tropoi'"] toi'" legomevnoi"
tropwth'rsi.”.
104
Athe2n aios of Naukratis, Deipnosophistae (Gulick), 5.208 (vol. 2, p. 440):
“uJph'rcon de; kai; tw'n toivcwn eJkatevrwqen tropoi; proewsmevnoi, diavsthma suvmmetron
e[conte": ...”. On the interpretation of Athe2naios/Moschio2n’s description of Hiero2’s
context of a series of structural members of the ship and explained

them as: “The tropoi lying alongside on each side around the steira
[are] the first and the second, also the thalamios”.105 What exactly he
meant by this is unclear, but it certainly had nothing to do with oar-
grommets. He was probably groping towards an explanation of
through beams.
The hull was constructed, according to the Anonymous, of
stamivde" (stamides), a mistake for stami'ne" (stamines), or sthmonavria
(ste2monaria), xuvla diavtona (xyla diatona), and peritona. However,
once again, he derived these terms from lexicographical sources and
assembled them in a way which suggests that he had no real idea at all
about how the hull of a dromon was actually constructed.
The upper futtocks of frames were called stamines. This was a
common word in classical Greek and it passed into medieval Latin in
the West as stamenaria.106 However, the Anonymous had no idea
what stamines were, said that there were only three of them, made the
word into stamides, and seems to have understood them as deck
pillars supporting the deck.107 Again, this was probably a
misunderstanding of Pollux, who had written that the timbers on
which the sanivde" (sanides), the planks of the deck, were laid were
the kanovnia (kanonia), the deck beams, and stamines, off which the
kanonia ran.108
Xyla diatona were said by the Anonymous to have been cross
beams.109 This term was not a technical one and was not derived from
------------------------------
great ship see Casson, Ships and seamanship, pp. 191-9.
105
Pollux, Onomasticon (Bethe), I.88 (vol. 1, p. 28): “oiJ de; peri; th;n stei'ran
eJkatevrwqen parateinovmenoi tropoi; prw'to" kai; deuvt ero", oJ kai; qalavmio".”.
106
See Pryor, “Galleys of Charles I of Anjou”, p. 71; Jal, Glossaire nautique, p.
1383.
107
Appendix Three, §2.10: “Kai; trei'" de; stami'ne" [stamivd e": MS. A] h[goun
sthmonavria i{stantai kai; aujta; kata; stoi'con oi|" ejp ereivd etai to; katavstrwma.”.
Both the scholia on the Odyssey and also Hesychios had explanations of
stamines. See Dindorf, Scholia Graeca, E.252 (vol. 1, p. 267): “stamivnessi de; toi'"
ejpimhvkesi xuvloi" kai; sthvmono" tavxin ejpevcousin, a} parativqetai toi'" ijkrivoi" ejx
eJkatevrwn tw'n merw'n pro;" to; eJstavnai: ...”; Hesychios, Lexicon, (Schmidt), S.1633
(vol. 4, p. 71): “stamivne": parastavtai, kai; ta; ejpi; th'" scediva" ojrqa; xuvla, pro;" a} aiJ
sanivde" proshlou'ntai, h] pavssaloi. para; to; eJstavnai.”. It seems as though the author
of the scholion, followed by Hesychios and later the Anonymous, were drawing on a
common pool of knowledge but neither the Anonymous nor Hesychios appear to have
used the scholion directly here, nor the Anonymous, Hesychios.
108
Pollux, Onomasticon (Bethe), I.92 (vol. 1, p. 30): “ta; de; xuvla ejf w|n aiJ sanivde"
ejpivkeintai, kanovnia kai; stamivne".”. Cf. the scholia on the Argonautika of Apollo2n ios
of Rhodes, I.723-4: “kanovna" de; ta;" stavqma", oi|" ta; xuvla kai; tou;" toivcou" oiJ tevktone"
ejxisavzousi”. Wendel, Scholia, p. 60.
109
Appendix Three, §2.11: “Eijsi; de; kaiv tina xuvla diavtona dihvkonta ajpo; tou' eJno;"
toivcou th'" nho;" e{w" tou' eJtevrou, ejf w|n ejp ivkeitai: ...”.
200 CHAPTER FOUR
Pollux, Hesychios, or Pho2tios. It appears to have been a generalized

explanation, derived from classical Greek diavtono" (diatonos),
referring to any bracing from side to side or front to back. In classical
Greek the technical term for a cross beam of a ship was either tropos
or zugovn (zygon).110
According to the Anonymous, peritona were “[timbers] which
enclose the sides on the outside”. But, again, this seems to have been a
misreading of Pollux, who said that a peritovnaion (peritonaion) could
be something enclosing the upper part of the hull on either side.111 As
seen above, the Anonymous also said that a peritonon could be
classified, together with a dryochon, as any “long continuous timber
which fastens together many other short timbers”. Elsewhere, he also
said that peritona were some things on the hull between the lower and
upper bank of oars.112 This all seems to read as though both Pollux and
the Anonymous used peritonaion and peritonon for a wale on the
outside of the hull, the classical Greek word for which was zwsthvr
(zo2ste2r). Neither Pollux, Hesychios, nor Pho2tios cited zo2ste2r in the
context of a ship. Nor was the word used in any of the Naumachika. It
appears that peritonaion/peritonon replaced zo2ste2r as the technical
term in the vernacular for a wale some time between middle antiquity
and the tenth century, although zo2ste2r survived in literary language.113
In one of the most inexplicable passages of his treatise, the
Anonymous wrote that: “And somewhere there, when sailing
ijdiwtikw'" (idio2tiko2s), there is an opening, which is called a eudias
(bung hole), for the removal of water. Indeed as [the ship] sails, the
shores (hermata), that is, what are known as seats (hedrai), close this;
and [there is] the deck (ikria), which is called the floor timbers
(enkoilia)”.114
------------------------------
110
See Casson, Ships and seamanship, pp. 14, 193.
111
Appendix Three, §2.11: “...: ta; de; tou;" toivcou" e[xwqen sunevconta perivtona
kalou'ntai.”. Cf. Pollux, Onomasticon (Bethe), I.92 (vol. 1, p. 30): “to; de; sunevcon
a[nwqen eJkatevrou" tou;" toivcou" peritovnaion kalei'tai.”.
Note, however, that Pollux said that peritonaia could also be timbers projecting
out around the poop. Pollux, Onomasticon (Bethe), I.89 (vol. 1, p. 29): “ta; de; peri; th;n
pruvmnan prouvconta xuvla peritovnaia kalei'tai.”.
112
Appendix Three, §2.1: “Koinw'" me;n ga;r wjnovmastai druvo con a{pan xuvlon
ejpivmhke" sunevcon kai; oiJonei; proshlou'n e{tera braceva te kai; pollav. Noeivtw d a]n
ou{tw kai; to; perivtonon.”; §2.13: “Tauvth" de; a[nwqen th'" eijresiva" perivtonon, ei\ta sani;"
eJtevra, hJ legomevnh pevla, ei\ta perivtonon, ...”.
113
On zo2ste2r as the classical word for a wale, see Casson, Ships and seamanship, p.
86, n. 46 and p. 223. On its survival in literary language, see below pp. 410-11. On
peritona as wales see also Koukoules, “Nautiko;" bivo "”, p. 347; Hocker, “Galleys and
fleets”, p. 96.
114
Appendix Three, §2.8: “Plevo usa de; ijdiwtikw'" ejkei' dhvpou kai; quriv" ejstin eij"
Reading the various terms of this passage with what should have
been their meaning according to classical Greek, it makes only
minimal sense. It appears to have referred either to scuppers, through
which water shipped inboard on deck would drain off, or to bung
holes, which ships which could be beached would normally have in
their hulls in order to drain bilge water easily and to facilitate cleaning
of the insides of the hulls.115 In the form eujdivaio" (eudiaios), eudias
was an obscure word whose first known use was by Plutarch in an
analogy to the urethra and bowel, thus almost certainly referring to a
bung hole rather than a scupper.116 Pollux referred to it as: “the hole
(thyris) which can be opened to let out water” and Hesychios virtually
repeated him: “... from the holes (tre2mata) made in ships for rain.”.117
In both these cases, it could refer to either a scupper or a bung hole.
Since the word ceivmaro" (cheimaros) was used by Hesiod for a
bung,118 one might have suspected that eudiaios referred to a scupper
rather than a bung hole. However, both Hesychios and also the author
of the Souda confined the meaning to a bung hole by equating the
word to cheimarrous and referring to emptying the bilges, ajntliva
(antlia): “Eudiaios: the cheimarrous, the hole of the ship, through
which the bilges empty”.119 It appears most likely that by the tenth
century cheimar(r)o(u)s and eudia(io)s were synonyms for a bunghole
and bung.
The wording of the Anonymous was so close to that of Pollux, that
------------------------------
ejkroh;n tou' u{dato", h{ti" eujdiva" [eujdia;": MS. A] kalei'tai. Tauvthn de; dhlonovti th;n
plevousan sunevcousi ta; e{rmata, h[toi aiJ legovmenai e{drai: kai; ta; ijkriva, a} ejgkoivlia
kalou'ntai.”.
115
See Jal, Glossaire nautique, pp. 671, 1024, 1028. Jal gave the meaning of bung
and bung hole to eudiaios without providing any references. According to him, the
Greek term in his own day was mpou'ka (mpouka).
The reconstructed Athenian trie2re2s Olympias was provided with a bung hole so
that the bilges could be drained and washed out when she was hauled ashore, just as
Mediterranean fishing boats still are today. Communication from John Coates. All
boats which are built to be able to be beached have bung holes.
116
Plutarch, Table-Talk, VII.1 (699.F), in Plutarch, Moralia, vol. 9, p. 18: “...,
oujqe;n ijd ivou povrou dei'tai to; perivttwma th'" uJgra'" trofh'", ajll ei|" ajrkei' kai; koino;"
w{sper eujdiai'o " ajmfotevroi" eij" taujto; dia; taujtou' suneiskomizomevnoi": ...”.
117
See Pollux, Onomasticon (Bethe), I.92 (vol. 1, p. 31): “hJ d ajnoigomevnh quri;" eij"
ejkroh;n tou' u{dato" eujd ivaio" kalei'tai.”; Hesychios, Lexicon, ed. Schmidt, E.72 (vol.
2, p. 219): “eujdivaion: ... ajpo; tw'n ejn toi'" ploivoi" givnomevnwn trhmavtwn dia; tou;"
o[mbrou". oiJ de; to;n prwktovn, h] to;n ceivmar [r]on, eufhmizovmenoi”. Note that the
Anonymous distorted Pollux’s eudiaios to eudias, which literally meant “fair
weather”.
118
Hesiod, Works and Days, 626 in Hesiod, p. 48: “..., ceivmaron ejxeruvsa", i{na mh;
puvqh/ Dio;" o[mbro".”.
119
Souda, E.3415 (vol. 2, p. 444): “Eujd ivaio": ceimavrrou", kai; trh'ma th'" newv", di ou|
hJ ajntliva ejkrei'.”.
202 CHAPTER FOUR
it appears certain that he emulated the passage from him, rather than
from Hesychios or the Souda. But, his point about sailing idio2tiko2s,
which should mean sailing either “on private business” or “without
due care”, is entirely obscure. It is possible that he intended some
contrast between either public and private use of ships or professional
and unprofessional use of them. However, why either of such uses
should have anything to do with bung holes and bungs, or scuppers
also for that matter, escapes us. Alternatively, was he making a feeble
joke and referring to some “idiot” forgetting to put the bung in?
Anyone who has sailed small boats will know that it would not have
been the first time. It is curious that he added this to a text taken
entirely from Pollux and then added without any authority from
Pollux, or any other text known to us, that the eudias was closed when
sailing. This eliminates any possible meaning of the word as used by
the Anonymous being “scupper”, because they, of course, had to be
left open when at sea in order to fulfil their function. A bung hole on
the other hand obviously had to be closed when at sea.
The Anonymous said that the things that closed the bung hole were
the hermata. He then gave a parenthetical explanation of hermata as
hedrai. Not surprisingly, these terms are not found with this sense in
Pollux, Hesychios, Pho2tios, or the Souda. In classical Greek, hermata
could mean either the shores or legs used to keep a ship upright when
beached or the ballast which kept it upright when afloat. Hedrai were
seats of any kind. What the Anonymous meant by his use of these
terms is totally obscure; however, he seems to have been referring to
the bungs which were used to plug the bung holes when at sea.
Then the Anonymous equated ikria and enkoilia. In classical Greek
ikria meant a deck and enkoilia were the floor timbers of the frames
of a ship, the lowest sections of the frames which joined the keel.120
Leo VI recommended that dromons should include timbers, xuvla
(xyla), or enkoilia, amongst the spare parts and equipment that they
carried.121 However, enkoilia was not a common term in classical
Greek and its meaning may have been poorly understood in general.
------------------------------
120
Casson, Ships and seamanship, 221. Cf. Souda, E.1462 (vol. 2, p. 294):
“Enterovneia: ta; ejkoivlia, ta; ajpo; th'" trovpido" ajnercovmena xuvla, ejnterovneia kalei'tai:
...”; Athe2n aios of Naukratis (quoting Moschio2n), Deipnosophistae (Gulick), 5.206.f
(vol. 2, p. 435) [the translation of ejgkoivlia and stami'ne" is incorrect].
121
Appendix Two [a], §5: “Ecevtw de; kai; ejk perissou' xuvla tina; ejgkoivlia kai;
sanivda" kai; ...”. Significantly, Nike2phoros Ouranos deleted Leo VI’s parenthesis of
xyla with enkoilia and simply recommended taking spare xyla. Assuming that he
knew what enkoilia meant, he would have known that if the floor timbers of a dromon
needed to be replaced, it was unlikely to be still afloat. Cf. Appendix Five, §4.
The Anonymous’s erroneous identification of them with ikria was

almost certainly derived from a scholion in the manuscript of the
Odyssey to which he had access.122
(d) Prow
The prow, prwv/ra (pro2ra) housed a dromon’s main offensive weapons,

a flame thrower, sivf wn (sipho2n), for Greek Fire, above which,
according to Leo VI, was a fortified forecastle, yeudopavtion
(pseudopation), from which marines could fight,123 and a spur, which,
as has been seen, was suspended by a chain or coupling from the
stempost. In late antiquity and the Middle Ages spurs were not built as
integral parts of the hull, as they were in the Renaissance. Contracts
for the construction of galleys for Charles I of Anjou, King of Sicily,
specified neither the provision of spurs nor their dimensions,
indicating that the contractors did not have to build them into the
galleys. Western evidence from the twelfth and thirteenth centuries
suggests that they were long wooden spars, perhaps usually of oak,
and perhaps iron-clad in some cases, which were attached to the
stempost after construction of the hull, and that they could be bought
and sold separately.124
Although there is no Byzantine evidence for it known to us, if the
spurs of dromons were also iron-clad, it would be clear evidence that
they had a totally different construction to Greco-Roman rams, which
had been sheathed in bronze.125 In antiquity, and until the very late
------------------------------
122
Dindorf, Scholia Graeca, E.163 (vol. 1, p. 259): “i[[kria] ta; ojrqa; xuvla, ejf w|n ta;
th'" new;" katastrwvmata prosphvgnutai, ta; ejgkoivlia legovmena.” (“ikria: the straight
timbers, to which the half decks of the ship are fixed, the so-called enkoilia.”).
123
See Appendix Two [a], §6; Appendix Five, §5; Appendix Eight [a], p. 241, [b],
p. 21.
124
A contract for the sale of two spurs made of oak (robor), each 10.42 metres long
and 0.25 metres wide survives from Genoa in 1267. See Ferretto, Codice diplomatico,
p. 87. Laura Balletto kindly checked this reference for us in the Archivio di Stato di
Genova, Cart. 82, fol. 75r. The text reads: “... sperones duos galee de robore bonos
pulcros et sannos, longos god. [gode] XIIII pro quolibet et largos parmum [palmum]
unum ...”.
The continuation of the Itinerarium peregrinorum, the Itinerarium peregrino-
rum et gesta Ricardi Regis compiled by Richard de Templo, prior of the Augustinian
priory of The Holy Trinity in London, from various sources, including Ambroise,
between 1216 and 1222, suggested that the spurs (rostra, a classicizing affectation)
were “ironed”, ferrata. See Itinerarium peregrinorum (Stubbs), Itinerarium regis
Ricardi, II.42 (p. 208): “Rex vero ... jussit ut unaquaeque galea navem suis
perpungeret calcaribus, id est, rostris ferratis”. See also p. 144 and n. 59 above.
125
See Eisenberg, “Metallurgical analysis”; Morrison, et al., Athenian trireme, pp.
204 CHAPTER FOUR
Middle Ages, the metallurgical technology to cast from iron such

large and complex items as the rams of ancient galleys did not exist.126
Spurs must have been more simple constructions which medieval
metallurgical expertise was capable of creating in wrought iron.
Neither Leo VI nor Nike2phoros Ouranos mentioned the spur.
However, the analysis by Van Doorninck of the manœuvres described
in Leo’s Naumachika Leontos Basileo2s, §68 and paraphrased by
Nike2phoros Ouranos in his Peri thalassomachias, §62, proves
unequivocally that tenth-century dromons no longer had waterline
rams. Van Doorninck has demonstrated beyond question that the
interpretation of this paragraph by R. H. Dolley, who claimed that it
pointed to the continuing existence of waterline rams into the tenth
century was incorrect.127 The two paragraphs read as follows:
Leo VI, Naumachika Leontos Basileo2s, §68: “It is also possible to

capsize an entire enemy ship if --- having coupled it side by side to the
dromon, and the enemy rush to one side, as is their habit, to engage in
hand-to-hand-fighting and expect their own ship to lay against the dromon
--- another dromon were then to run at the side of the enemy vessel
towards the stern and strike it hard as they collide, and if the one [first]
dromon should be able to free itself from the coupling and back off a little
so that it is not laying against the enemy, and if the other [second] dromon
were to weigh down with all vigour, it will capsize the enemy ship and
her crew completely. You should organize the coupling so that it does not
hold the [enemy] ship evenly but leaves at the enemy ship’s stern some of
the sides a little exposed, where the dromon will be able to attack and
exert pressure to capsize the enemy ship”.128
------------------------------
167, 221-3.
The Souda said that emboloi were made of copper, but its author was merely
quoting Herodotos. See Souda, E.952 (vol. 2, p. 254): ““Embolo": cavlkwma
pepurwmevnon, peritiqevmenon kata; prwvran tai'" nausivn.”.
126
In order to cast objects, for example the Athlit ram, from a metal, it is necessary
to be able to heat the ore containing it beyond the melting point of the metal so that it
can run as a liquid. The melting point of bronze is only around 1100˚ Celsius, whereas
that of iron is 1537˚ Celsius. It was not until the late Middle Ages that furnaces
employing strong blasts of air which could raise the temperature of the charge of ore
and charcoal to make the iron “run” were invented. Until then, all iron was wrought
iron. The charge was heated until a spongy, soft mass of iron and slag (metallic
impurities and charcoal ash) was produced. This was then beaten with hammers to
drive out the slag. But such iron was still too viscous to be cast.
127
See Dolley, “Warships”, p. 49; idem, “Naval tactics”, p. 331; Van Doorninck,
“Waterline ram”. Van Doorninck himself modestly claims merely to have cast doubt
on underwater rams being involved, rather than to have proved that they were not. See
also Bonino, “Rams”.
128
Appendix Two [a], §68.
Figure 20
Longitudinal section of a dromon of the era of the Macedonian emperors.
© John H. Pryor
206 CHAPTER FOUR
Nike2phoros Ouranos, Peri thalassomachias, §62: “It is possible to

overturn an entire enemy [ship] if you couple it to the dromon side by
side, and the enemy rush to that part which is against the dromon, as is
their custom, to engage in hand-to-hand fighting, hoping that their ship
will lay against the dromon. Then another dromon should run at the
enemy [ship]’s side at the stern and should strike and push the enemy
[ship] severely. And the first dromon, the one coupled to the enemy
[ship], should be able to free itself from the coupling and back off a little
so that the enemy [ship] does not lay against it. The other dromon should
weigh down as much as it can and if it does this, it should up end the
enemy [ship] with the men in it. You should not couple the whole enemy
[ship] but only a little, so that at the enemy’s stern you leave the sides
bare, where the dromon can strike in order to overturn the enemy [ship]
with the enemy [crew]”.129
Leo VI’s text was syntactically very obtuse, although its intent was
clear enough. Nike2phoros Ouranos paraphrased it in order to make the
meaning clearer. Van Doorninck points out that the verb ajnatrevpein
(anatrepein) used by Leo VI meant “to capsize” in the context of a
ship. So also the noun suvgkrousi" (synkrousis) and the verb
sugkrouvein (synkrouein) which meant “collision” and “to collide”,
were used in both texts instead of ejmbolhv (embole2) and ejmbavllein
(emballein) respectively. Finally, the crucial action was for the second
dromon to “weigh down”, barhvsh/ (bare2se2), the enemy ship by
attacking at its stern. When the first dromon engaged the enemy ship
side by side, the enemy crew would pack the side to fight. The enemy
ship was only saved from listing because its hull was hard up against
the first dromon. The second dromon could then run up and over it
towards its stern with its spur and prow and, when the first dromon
disengaged, the weight of the enemy crew and of the second dromon’s
prow would roll it over completely.
In fact, this stratagem reads very much like the fire-side musings of
the emperor himself, and one may legitimately have reservations
about its practicability in battle for various reasons. First, it is very
probable that if the crew of any ship like a dromon or its Muslim
counterpart, a ship as shallow in draft and narrow in beam as they
were, unballasted and carrying as little as they did, all rushed to one
side, they would probably capsize the ship without any help from the
enemy, even if it was resting against the hull of the enemy ship.130
------------------------------
129
Appendix Five, §62.
130
On Olympias, the movement of even a single man weighing around 80
kilogrammes to the side above deck was enough to cause the ship to list by around
Secondly, given the fine coordination and timing required of the two
dromons, it is difficult to imagine how such a manœuvre could
possibly have been carried out successfully in the heat of battle with
galleys on all sides pitching and rolling with the seas. The text appears
to us to have been written by someone imagining three galleys
floating on a mill pond alone. Moreover, why could the crew of the
enemy galley not prevent the crucial unlinking at the last moment by
maintaining its own links? In any case, when the enemy crew saw the
second dromon bearing down, why could they not restore the balance
of their own ship by moving the crew to the other side? As we shall
see, dromons had a beam amidships of only around 4.5 metres and it
would have taken only a second or two for men to move from one side
to the other.
The difficulty that anyone unfamiliar with the precise import of this
paragraph might have with it is clearly illustrated by the garbled way
in which it was interpreted in the Arabic paraphrase of the paragraph
which was inserted by Muh5ammad ibn Mankalı3 into his Al-ahka2m al-
mulu2kyya wa ’l dawa2bit al-Na2musiyya: “When a ship rams another
ship, let it [the “ramming”] be at the side close to the stern, in order to
cut loose the chains and hooks from the enemy’s ships and thus
disentangle your ships from those of the enemy, and they would
perish”.131 It is possible to recognize here a pale reflection of Leo VI’s
stratagem which was incomprehensible to Ibn Mankalı3 or his source.
Perhaps it always had been incomprehensible if Leo’s stratagem was
merely the product of arm-chair musings.
Although excessive importance should not be attached to it, it is
interesting that the illustration of Byzantine dromons rolling over
Rho2s ships in the Bosporos in 941 in the Madrid manuscript of the
Synopsis historio2n of John Skylitze2s shows the Byzantines attacking
at the stern quarters of the Rho2s ships. [See Figure 9] If there is any
import to this beyond merely the way that the artist had to compose
his picture, it would be that the stern quarters would be the first point
of attack in any case, if the enemy allowed it, because that was where
the quarter rudders were and the stern quarters were the least
defensible parts of any galley. The entire objective of manœuvring, of
------------------------------
0.3˚. Thirty soldiers moving on the deck a mere 0.7 metres towards the sides would
cause the ship to list by about 4˚. See Morrison and Coates, Trireme reconstructed, p.
60; Morrison et al., Athenian trireme, pp. 160-1, 227.
131
See Appendix Eight, Part B, pp. 124-5. Christides, “Ibn al-Manqalı3”, p. 95, says
that this passage has no corresponding one in the Naumachika; however, we consider
that it is a clear reflection of Naumachika Leontos Basileo2s, §68.
208 CHAPTER FOUR
the formation of battle lines, and of battle tactics in general was to

keep enemy ships at the bow and to prevent them being able to attack
at the stern.132
Whatever the case, it is clear that the emperor did not envisage
attack with a waterline ram and other considerations also suggest that
tenth-century dromons no longer had waterline rams. First, as
Alexandres has also concluded on the basis of his reflections on the
battle tactics found in the Naumachika, particularly the extensive use
of missiles before engaging, exchanges with missiles would have been
unnecessary if the primary objective had been to sink by ramming. 133
The same comment can be made about the use of Greek Fire.
Secondly, the Naumachika also show clearly that the fundamental
battle technique of Byzantine fleets was to grapple with enemy ships
side to side and to link them together with iron rods so that the enemy
could not escape and marines could then engage the enemy crews.134
Again this would have been not only unnecessary but also actually
dangerous if rams had been used to flood the hulls of the enemy ships.
If dromons did not have waterline rams, then they must have had
spurs. But what were these called? None of the Naumachika contain a
clue. However, in the inventory for the Cretan expedition of 949,
amongst the equipment to be supplied by the Department of the
Vestiarion basilikon for 20 dromons, was specified: “20 perovnia
(peronia) for the kataprovswpa (kataproso2pa) together with their
katakovrake" (katakorakes)”,135 a phrase which has defied all efforts
to comprehend it. Peronion (pl. peronia) was a diminutive of, or a
derivative synonym for, perovnh (perone2), which could mean a pin, or
brooch, or buckle. It had many other senses in mechanical
engineering, and was derived from peronavw (peronao2), “pierce” or
“transfix”. Since only one of these peronia was to be supplied for
each dromon, they must therefore have been major pieces of
equipment and not pins, or bolts, or buckles. However, peronion in
the sense of something that pierces has the right sense for a spur and
------------------------------
132
Note, however, that this illustration to the manuscript was drawn by one of the
artists, Ce, who was drawing in a Western style not based upon the Byzantine
originals. See Appendix Seven. That being said, the tactical requirement would have
been equally as true of tenth-century Byzantine fleets as of twelfth-century Sicilian
ones.
133
See Alexandres, ÔH qalassiva, pp. 62-4.
134
135
Appendix Four [b], §IV.12: “perovnia kV, kataprovswpa su;n tw'n katakoravkwn
aujtw'n,” [= Haldon, “Theory and practice”, p. 227; Constantine VII, De cerimoniis,
II.45 (vol. 1, p. 672)]. See Appendix Four, n. 27 on the punctuation here and cf.
Haldon, “Theory and practice”, pp. 227 n. 83, 281-3.
speronus, one of the two medieval Latin words for the spur, was
almost certainly derived from it.136 Surely peronia were the dromons’
spurs. The specification was that there should be 20 peronia, “for the
kataproso2pa, together with their katakorakes”. Proso2pon had the
sense of the front, facade, or face of anything, in particular of a
ship,137 and one of the senses of korax was anything hooked for
grappling or holding something. Reading the “katav” prefixes simply
in their strengthening sense, we suggest that the real meaning of this
specification was: “Twenty spurs for the faces [of the bows], together
with their couplings”. Peronion was probably the Byzantine word for
the spur and katakorax that for the coupling to the head of the
stempost.138
Elsewhere in the inventories, amongst a list of additional
equipment to be provided by the Vestiarion basilikon, it was said that
130 perovnai (peronai) (sing. perone2) were to be provided for the
chelandia for the expedition.139 The composition of the actual fleet
was not detailed precisely, beyond the fact that there were 20
dromons. However, there were 150 ousiai, standard complements of
dromons or chelandia, specified for the total navy and it appears that
the specification of 130 peronai for the chelandia was merely an
ambit figure for the ships of the remaining 130 ousiai, collectively
called chelandia.140 It appears that both perone2 and its diminutive or
derivative synonym peronion were used for spurs.
------------------------------
136
See Jal, Glossaire nautique, under “Sperone” (p. 1378). Jal provided no
justification for this derivation. However, the identification of perovnh as a pin or
buckle with various medieval Latin forms such as “sporo, spero, speronus” for pins,
buckles, and the spurs of galleys is well established. See Niermeyer, Lexicon, p. 985
and the sources cited therein. In the transition to medieval Latin the Greek word took
on an initial “s”. Jal’s derivation thus appears to have been well justified and in the
Latin West the word became applied to the spur of a galley, just as we suggest it had
been in Byzantium.
137
This is made quite clear in manuscripts of Thucydides with scholia dated to the
tenth or early eleventh centuries. See Hude, Scholia, II.90.4 (p. 156): “to; de; shmeivo u
ajnti; tou' milivo u. ejpistrevyante" ... : to; me;n ejp istrevyante" ei\p en, o{ti h\n a[nw eijpwvn,
ejpeidh; kat eujqei'a n e[pleon: nu'n de; oujkevt i: to; de; metwphdo;n pavl in to; kat eujqei'a n
plevein ejstiv: to; ga;r mevtwpon th'" new;" hJ prwv/ra ejstiv. levgei ou\n o{ti ta; mevtwpa, o{ ejstiv ta;
provswpa tw'n new'n, parei'con toi'" ejnantivoi".”.
138
The Anonymous identified the katakorax, or katakoraka as he actually had it,
with the sipho2n; however, in our opinion, all that he knew was that a katakorax was
something at the prow and he therefore identified it with the sipho2n, which was the
only salient feature of the prow that he knew about. See Appendix Three, §2.14: “Epi;
de; th'" prwv/ra" oJ sivfwn o}" katakovrax [katakovraka: MS. A.] levgetai ejnergw'n o{tan
w\sin aiJ nh'e" ajntivprw/roi: ...”.
139
See Appendix Four [b], §VII.2: “perovna" tw'n celandivwn rlV, [= Haldon,
“Theory and practice”, p. 233; Constantine VII De cerimoniis, II.45 (vol. 1, p. 676)].
140
On the fleet’s composition see pp. 259-60, 372-3 and Appendix Four [b], §I.
210 CHAPTER FOUR
We do not, however, deny that elsewhere in the inventories for the

Cretan expeditions the terms peronion and perone2 apparently had
other meanings. The Department of the Vestiarion basilikon was also
to provide 120 peronai among a list of tools and spikes and to
attribute the sense of “belaying pins” to the word seems reasonable in
this context. In addition, twelve large iron peronia for a “wooden
castle”, xylokastron, were also to be supplied to the droungarios to2n
ploimo2n for the expedition.141 This specification was related to a
single xylokastron and seems to have been unconnected to the ships.
Perhaps it was a portable fortress for consolidating a beach-head.
Peronion and perone2 were terms which could be used in a variety of
contexts.
According to the Anonymous, there were also perivboloi
(periboloi) on either side at the prow, from which the anchors were
lowered.142 Periboloi were also listed in the inventories for the Cretan
expedition of 949 amongst equipment paid for by the Department of
the Eidikon for the karabia of the Rho2s.143 Our best suggestion for the
meaning of peribolos is “cathead”. On Greek trie2reis the ejpwtivde"
(epo2tides), the transverse “cheek timbers” of the outriggers at the
bows, had apparently served for this purpose;144 however, with their
passing something like catheads must have become necessary on
galleys. Other ships of any size must always have had something like
catheads. Leo VI equated the classical a[gkura (ankyra) for an anchor
to what appears to have been the vernacular tenth-century term for it,
an “iron”, sivdhron (side2ron), and said that it was the duty of one of
the two oarsmen at the bow to cast, bavllein (ballein), the anchors into
the sea. Nike2phoros Ouranos simply used the tenth-century term,
------------------------------
141
See Appendix Four [b], §§IV.21, VII.6: “perovna" kata; perivsseian sV.”;
“perovnia sidhra' megavla tou' xulokavstrou ibV ” [= Haldon, “Theory and practice”, pp.
227, 233; Constantine VII, De cerimoniis, II.45 (vol. 1, pp. 672, 677)].
142
Appendix Three, §2.14: “Th'" de; prwv/ra" plhsivon eJkatevroi" toi'" mevresi
perivboloi ejmpephgmevnoi i{stantai di w|n aiJ a[gkurai krevmantai, aiJ th;n nau'n iJstw'si
calwvmenai.”. This may have been derived from Pollux, Onomasticon (Bethe), I.93
(vol. 1, p. 31): “a[gkurai, ajmfivboloi, ajmfivstomoi, eJterovstomoi: ...”.
143
See Appendix Four [b], §VI.13 [= Haldon, “Theory and practice”, p. 231;
144
See Euripides, Ifigevneia hJ ejn Tauvroi", ll. 1350-1, in Euripides, Fabulae, vol. 2,
p. 297: “..., oiJ d ejpwtivdwn a[gkuran ejxanh'pton, ... ”. See also Casson, Ships and
seamanship, p. 86, n. 45. On epo2tides see below pp. 218-24.
Jal believed that in classical Greek pareiav (pareia) meant a cathead but there
seems to be no evidence for this. He also said that in his own day the vernacular
Greek for a cathead was kapovni and this does indeed appear to have been derived
from the Italian capone (and variants) for a cat tackle. See Jal, Glossaire nautique, p.
318; Kahane and Tietze, Lingua Franca, §155 (p.146).
side2ron,145 which had in fact been used for an anchor as early as the
sixth/seventh centuries in the anonymous Life of St Nicholas of
Sion.146 The emperor and the magistros both referred to anchors in the
plural, a clear reflection of the inefficient design and light weight of
ancient and medieval anchors, whose flukes were small and did not
grab well. Ancient and medieval ships needed many anchors. The
small seventh-century Byzantine coastal trader excavated at Yassı
Ada islet carried no less than eleven anchors and the eleventh-century
ship excavated in Serçe Limani, opposite Rhodes, nine. The anchors
of the Yassı Ada ship weighed between approximately 80 and 140
kilogrammes and those of the Serçe Limani ship between
approximately 50 and 65 kilogrammes.147 Byzantine dromons would
certainly have needed multiple anchors and this is confirmed by the
inventories for the 949 Cretan expedition.
These contain information about what appear to have been
anchoring systems. In three lists, they give a number of items in
sequence which correspond significantly to each other. These are all
given in Appendix Four; however, it is convenient to cite the Greek in
parallel here.
Table 6: Specifications for conjectured anchoring systems in the De cerimoniis
Appendix Four [b], IV Appendix Four [b], V Appendix Four [b], VII
13 sivdhra bolistika; rkV, 17 sidhrobolistika; ka-
ta; perivsseian nV,
14 sidhrobovlia rkV, 18 sidhrovbola nV, 23 scoiniva sidhrovbola
xV,
15 ajnagokatavgonta su;n 22 ajnagontiteva calka'
tw'n iJmantarivwn ieV,
aujtw'n kV,
16 peripetovmena ajna; 20 peripetovmena rV, 24 peripetovmena rmV,
kdV, oJmou' upV,
17 filourevai ajna; ibV, 19 filourevai rV,
oJmou' smV,
21 spartivna" rV, 25 spartivna" rmV,
22 leptavria sV, 26 leptavria skV,
25 skalodevmata uV. 27 skalodevmata tV,
28 koubavria rV,
------------------------------
145
See Appendix Two [a], §8: “Tw'n de; prw/raivwn ejlatw'n oiJ teleutai'oi duvo, oJ me;n
e[stw sifwnavtwr, oJ de; e{t ero" oJ ta;" ajgkuvra" bavllwn kata; qavlassan, h[goun ta; sivdhra:
...”. Cf. Appendix Five, §7. See also Appendix Two [b], §1.
146
Vita S. Nikolai, §38 (p. 66): “Bavlete ta; sivdhra tou' ploivo u eij" th;n gh'n, kai;
dhvsate to; ploi'on hJmw'n, kai; fevr ete to;n kavrabon: ...”.
147
See Pryor, “Naval architecture”, pp. 369-72; Bass, et al., Yassi Ada. Volume I, pp.
121-43; Steffy, “Medieval cargo ship”, p. 17; Bass et al., Serçe Limani, pp. 189-238.
212 CHAPTER FOUR
The inventories (Appendix Four [b], §§IV.13-14 and §§V.17-18)

specified that for the 20 dromons there should be 120 sivdhra
bolistikav (side2ra bolistika) with 120 sidhrobovlia (side2robolia), that
is, six of each per dromon, as well as another 50 sidhrobolistika;
(side2robolistika) with 50 sidhrovbola (side2robola). This has to be the
starting point. At [b], §VII.23 they also said that the Department of
the Vestiarion supplied another 60 “iron-throwing ropes”, scoiniva
sidhrovbola (schoinia side2robola), for the expedition. Since schoinia
side2robola clearly implied something made of iron and equivalent to a
rope, we conclude that the 120 and the 50 side2robolia/side2robola were
the same thing and that the 120 and 50 side2ra bolistika/side2robolistika
were something different. Side2ra bolistika literally meant “casting
irons” and these latter must therefore have been the anchors, which
were “cast” into the sea, as the Leo VI and Nike2phoros Ouranos both
said.
What then were the side2robolia/side2robola or schoinia side2robola
in this context? There is no doubt that one of the scholia on Lucian’s
Lexiphane2s used the word sidhrobovlion (side2robolion) for an
anchor;148 however, here we suggest that they were iron anchor chains
attached to the anchors. The rope anchor cables would have been
attached to them. When anchor cables are made of rope, as was the
case in the Middle Ages, it is necessary to have the last few metres
connected to the anchors made of a heavy chain because an anchor
works best when the drag of the ship on it is as close as possible to
parallel to the sea bottom. This causes the flukes to dig into the sea
bed. But, because rope is light and floats, it tends to pull anchors
upright and dislodge the flukes if attached directly to the anchors. Iron
chain, on the other hand, will sink, thus causing the entire anchor
system to form an arc with the anchor end as close to parallel to the
bottom as possible. All modern small craft using rope anchor cables
have a length of heavy chain connecting them to the anchors.
At both Appendix Four [b], §IV.17 and §V.19, the inventories
specified some things called filourevai (philoureai), of which there
should be twice as many as the number of anchors and anchor chains.
------------------------------
148
The scholion is on Lexiphane2s, §15: “... e{ktorav" tina" ajmfistovmou" kai; ijscavda"
sidhra'" ajfei;" kai; nausipevda" ...”. See Lucian, vol. 5, p. 312. Lucian appears to have
used parenthetically three obscure metaphors for an anchor: e{ktorav" tina"
ajmfistovmou" (hektoras tinas amphistomous), “double-fluked hold-fasts”, ijscavda"
sidhra'" (ischadas side2ras), “iron holders”, and nausipevda" (nausipedas), “ship-
fetters”. The scholion defined all three as sidhrobovliav. See Rabe, Scholia in
Lucianum, 46.15 (p. 200): “e{ktora" kai; ijscavda" kai; nausipevd a" ta; sidhrobovliav
fhsi”.
These were probably cables made from the inner bark of the Linden
tree, filuvra (philyra).149 Such cables were specified nowhere else in
the inventories and we suggest that this material had some properties
which made them especially suitable for use under water. They may
have been attached to the anchor chains. The other cables specified at
[b], §V.21 and §VII.25, spartivnai (spartinai), which were made from
the much more common spartum or esparto grass, may have been the
upper parts of the cables running from the philoureai to the catheads
and windlasses, by which the anchors were cast or weighed.
At Appendix Four [b], §IV.15, the inventories also specified
immediately after the entries for the anchor chains, that the 20
dromons should have “20 ajnagokatavgonta (anagokatagonta) with
their iJmantariva (himantaria)”, one per dromon. The word himantaria
appears to have been used as a generic for the entire anchor cable
systems. Casson has established that iJmavnte" (himantes) were the
“lifts” used on the yards of classical square-rigged ships.150
Himantaria in the same sense of a cable for lifting something fits the
anchor cables. At [b], §VII.22, the specification was “15 bronze
ajnagontiteva (anagontitea)”. The verbs ajnavgein (anagein) and
katavgein (katagein) could have the meanings of “to raise” and “to
lower” respectively, suggesting that anagokatagonta were devices for
raising and lowering some things. Because of their position in the list,
we suggest that they were windlasses or capstans by which the
anchors were cast and hoisted, terms for which in classical Greek
were periagwgeuv" (periago2geus), strofei'on (stropheion) and possibly
ejrgavth" (ergate2s).151 Whether the anagontitea of [b], §G.22 were the
same things as anagokatagonta is arguable because of the
specification that they were made of bronze. Why would anyone make
windlasses of bronze? To do so would have been extraordinarily
extravagant, even given that bronze is a non corrodable metal.
------------------------------
149
The Greek philyra was the same as the Latin tilia from which the modern name
of the genus, tiliaceae, is derived. Tilia Europea (syn. tilia vulgaris, tilia cordata) is
the common, small-leaved European linden tree from whose fibrous inner bark ropes,
nets, and matting were made for centuries. See Polunin, Trees and bushes, pp. 141 &
198; Edlin, What wood is that, pp. 126-8; Rackham, Ancient woodland, pp. 237-9.
See also the sixth- (or ninth-) century Philoxenos Latin-Greek glosses of Paris,
Bibliothèque Nationale, MS. Lat. 7651, in Goetz, Glossarii Latini, vol. 2, p. 472, l. 5
(“Filura tilia” [sic]). See “Note on citations of Greek and Latin glossaries”, p. lxix
above.
150
See Ships and seamanship, pp. 260-63.
151
See Lucian, “The ship or the wishes”, §5, in Lucian, vol. 6, p. 436: “... kai; pro;
touvtwn aiJ a[gkurai kai; strofei'a kai; periagwgei'" kai; ...”. On ergate2s and derivatives
for a capstan in Turkish, Arabic, and Modern Greek see Kahane and Tietze, Lingua
Franca, §769 (pp. 507-8).
214 CHAPTER FOUR
Normally windlasses were made of wood. Besides, the number of

them specified, fifteen, does not correspond to the number of dromons
and the specification referred to “extra” equipment supplied by the
Department of the Vestiarion basilikon.
At Appendix Four [b], §IV.16, the inventories specified 24
peripetovmena (peripetomena) per dromon. At [b], §V.20 and §VII.24,
it was specified, respectively, that 100 peripetomena should be
supplied for what may have been siege engines and another 140
peripetomena should be supplied for the fleet. The word appears to
have had the sense of something flying around. They may possibly
have been windlass bars by which the anagokatagonta were turned.
Against this, it is difficult to see why 24 of them would needed for
each windlass, especially given the light weight of Byzantine anchors.
At two places immediately after the mention of the spartum cables,
the inventories listed both 200 and 220 leptavria (leptaria). Leptaria
were also listed between schoinia (cordage) and some things called
chartaria amongst equipment purchased for the karabia of the
Rho2s.152 The meaning of leptaria is unknown but the specification
appears to have had some connection to the anchoring system. If the
word was derived from leptov" (leptos), which had the meaning
“light” amongst other things, the term might possibly have referred to
something light and hence to something which floated. We suggest
anchor buoys and their ropes: small floats attached by thin lines to the
anchors and used to mark their position on the sea bed so that that
could be easily ascertained when trying to raise them.153
Finally, as regards the anchoring systems, the inventories listed
among additional items provided from the Department of the
Vestiarion basilikon to the droungarios to2n ploimo2n (Appendix Four
[b], §VII.27, 28), immediately after the leptaria and before a list of
materials, 300 skalodevmata (skalodemata) and 100 koubavria
(koubaria). These items were probably mooring cables and perhaps
reels of some sort associated with their use.154
------------------------------
152
See Appendix Four [b], §§V.22, VI.19, VII.26 [= Haldon, “Theory and
practice”, pp. 229, 231, 233; Constantine VII, De cerimoniis, II.45 (vol. 1, pp. 673,
675, 677)].
153
In classical Greek a leptavrion was some sort of surgical instrument. The
reference appears to be to something thin, or light, or slight. Reiske suggested that
they were light cords or twine. See Constantine VII, De cerimoniis, vol. 2, p. 795.
However, there would be no need for such minor items to be mentioned in an
inventory such as this. We prefer to think that they were items of major importance of
some kind. A thin line for an anchor buoy would also accord with Reiske’s
etymology.
154
Our initial inclination was to follow Reiske on these terms: rope ladders for
(e) Poop
The poop, pruvmnh (prymne2), had a berth, kravb(b)at(t)o"

(krab(b)at(t)os), for the kevntarco" (kentarchos) or “captain” of the
dromon, a fleet commander, or for dignitaries such as the emperor.155
This would have been protected by a round tent or awning on a
wooden framework, skhnhv (ske2ne2), such as those depicted on Roman
galleys.156 The word krab(b)at(t)os was used in both Greek as krebavth
------------------------------
skalodemata and barrels for koubaria. However, we have been persuaded otherwise
by both Haldon and by the identification of kavdoi (kadoi) in the inventories
(Appendix Four [b], §III.20) as amphorae for water. See below pp. 361-3. It is
extremely unlikely that the fleet would have been using both amphorae and barrels as
water containers.
Haldon suggests that skalodemata for mooring ropes was derived from skavla
(skala), a ladder, thence a gangplank, thence a landing stage or quay (cf. below pp.
306-7) and devma (dema), a bond, rope, even a tow-rope. Hence a skalodema was a
rope bonding a ship to a landing; i.e., a mooring rope. In Modern Greek koubavri
(koubari) can mean a spool or reel. The Byzantine word may thus have meant
something like spools around which ropes, perhaps the skalodemata, were wound.
See “Theory and practice”, pp. 283-4.
155
Leo VI and Nike2phoros Ouranos both used the term “couch” (krabatos), which
meant a “berth”, or “cabin”, much as we would say “bed and breakfast” for “room
and breakfast”. It cannot have been a simple “couch” since Leo and Nike2phoros both
said that it provided protection for the commander from enemy missiles in battle. See
Appendix Two [a], §8: “Kai; oJ tou' nauavrcou dev, h[toi tou' kentavrcou, kravbato" ejpi; th'"
pruvmnh" ginevsqw, ... oJmou' de; kai; fulavttwn ejn kairw'/ sumbolh'" ajpo; tw'n rJiptomevnwn
belw'n para; tw'n ejnantivwn, ...”. Cf. Appendix Five, §7.
On the interpretation of krab(b)at(t)os as “berth” or “cabin” see also
Alexandres, ÔH qalassiva, pp. 60-61 and Koukoulès, “Nautiko;" bivo"”, pp. 352 and
384.
In the Praecepta imperatori Romano bellum cogitanti ... observanda, the author
said that when an emperor crossed from Constantinople to Pylai on the other side of
the Sea of Marmara for an Anatolian campaign he rose from his krabbatos when at a
sufficient distance from the capital and prayed that God would protect it while he was
absent. See Constantine VII, Three treatises, Text C, ll. 321-4 (p. 114): “... kai; ajpo;
iJkanou' diasthvmato" tou' basileivo u o{rmou genovmeno", w{ste aujt o;n ejp iskopei'n th;n
povlin, ejgeivr etai ajpo; tou' krabbavtou kai; i{statai kat ajnatola;" ta;" cei'ra" pro;" to;n
oujrano;n ai[rwn, kai; trivton th'/ ceiri; th;n povlin katasfragivsa" eu[cetai tw'/ Qew'/ levgwn
ou{tw".”.
Thietmar of Merseburg used the Latin lectus for the commander’s berth on the
poop of salandriae in close parallel to the Byzantine use of krab(b)atos. See Thietmar
of Merseburg, Chronicon, III.21 [12] (pp. 14-5): “... iterum equo comite in mare
prosiliens ad alteram salandriam, que sequebatur, tendit et, ab Henrico solum milite
eius, qui Sclavonice Zolunta vocatur, agnitus, intromittitur et, in lecto senioris
eiusdem navis positus, tandem ab ipso eciam cognitus, si imperator esset,
interrogatur.”.
156
Cf. Figures 1, 3, 12.
Only the Anonymous referred to the commander’s “tent” or “berth”. Appendix
Three, §2.5: “..., e[nqa dh; kai; skhnh; phvgnutai tw'/ strathgw'/ h] trihravrcw/ h[goun
kravbato". En oi|" de; oJ kravbatto" ejpereivdetai, ...”. This reference to a ske2n e2 was
almost certainly taken from Pollux, Onomasticon (Bethe), I.89 (vol. 1, p. 29): “The
timbers projecting arund the poop are called the peritonaia. The part which is called
216 CHAPTER FOUR
(krebate2) and in Latin as grabatus in sources for the arrest of Pope

Martin I and his transportation to Constantinople for trial on a charge
of treason in 653. After his arrival in Constantinople, the Pope was
left on the grabatus of his ship, exposed to the jeers of the mob.157
Right aft was a recurved stern ornament of some kind. In varying
forms, this ornament can be seen in the pictorial evidence from
antiquity, through the Byzantine period, to the High Middle Ages in
the West. In classical Greek and Latin it had been been known as the
a[flaston (aphlaston) and aplustre respectively but, what it was
actually known as in Byzantium in the tenth century is unclear.
According to the Anonymous, the krab(b)at(t)os was erected on
aphlasta and on trocanth're" (trochante2res), together known as
bovrdwne" (bordo2nes), a word which is otherwise unknown in the
context of ships.158 Since it would obviously have been impossible to
erect the krab(b)atos on stern ornaments, it is clear that the
Anonymous did not understand the classical meaning of aphlasta,
probably because his sources did not explain it. Pollux and Hesychios
both merely described it as the “peak” or “extremity” of the poop, as
also did the Souda, and this was insufficiently precise for the
Anonymous.159 Pho2tios had explanations derived from Apollodo2ros of
Athens and Didymos of Alexandria but the Anonymous appears to
have known neither of these nor, perhaps, the Lexicon of Pho2tios
itself.160 The word was not used in the Odyssey or by Thucydides and
------------------------------
the “ske2ne2” is constructed thereabouts for the strate2gos or trie2rarchos.” (“ta; de; peri;
th;n pruvmnan prouvconta xuvla peritovnaia kalei'tai. ejkei' pou kai; skhnh; ojnomavzetai to;
phgnuvmenon strathgw'/ h] trihravrcw.”).
It was also common practice to cover the poops of medieval Western galleys
with awnings. The ceremonial kadirga of the Turkish sult6a2ns in the Turkish Naval
Museum in Istanbul has an elaborately carved wooden ske2ne2 covering the poop. See
Basch, “Galley in Istanbul”. No doubt the wooden frame of the ske2ne2 of a Byzantine
war dromon would have been less elaborate but much more protective.
157
Anastasius Bibliothecarius, Collectanea, col. 592: “... reliquerunt eum a mane
ministri usque in horam decimam recubantem in grabato navis, eratque, sicut scriptum
est, spectaculum omnibus angelis et hominibus.”; Peeters, “Vie grecque”, §6 (p. 258):
“Kai; proswvr misan aujto;n plhsivo n Arkadianiw'n, ejavsante" aujto;n ajpo; e{wqen e{w" w{ra"
dwdekavth" keivmenon ejn th'/ krebavth/ tou' ploivo u.”.
158
Appendix Three, §2.5: “En oi|" de; oJ kravbatto" ejp ereivdetai, trocanth're" kai;
a[flasta, oiJ legovmenoi bovrdwne".”.
159
Pollux, Onomasticon (Bethe), I.90 (vol. 1, p. 29): “ta; de; a[kra th'" pruvmnh"
a[flasta kalei'tai, w|n ejnto;" xuvlon ojrqo;n pevp hgen, o} kalou'si stulivda: ...”; Hesychios,
Lexicon (Schmidt), A.8702 (vol. 1, p. 337): “a[flaston: to; ajkrostovlion, to; a[kron th'"
pruvmnh", ajpotetamevnon eij" u{yo"/.”; Souda, A.4631 (vol. 1, p. 432): “ “Aflaston: to;
a[kron th'" pruvmnh".”.
160
Pho2tios, Lexicon (Theodoridis), A.3369 (p. 308): “a[flasta: ta; ajkrostovlia
Apollovdwro". oJ de; Divdumo" to; ejpi; th'" pruvmnh" ajnatetamevnon eij" u{yo" ejk kanonivwn
platevwn ejp ikekammevnwn.”.
therefore the Anonymous had no scholia on it. The Iliad did have
scholia on the word but he had no manuscript of that. In fact, there
was confusion in post-classical times about the meaning of aphlaston.
One scholion on the Iliad, followed by one on Apollo2nios of Rhodes’
Argonautika went to some lengths to reject identification of it with the
akrostolion of the prow and to insist that it was at the poop.161
What the Anonymous meant by bordo2nes is unknown and there are
a number of possibilities. The Greek bovrdwne" may have been a
contraction from bouvrdwne" (bourdo2nes), a word associated with
“mules”. However, as used here the word appears to be a hapax
legomenon and no word with any similar form and meaning is known
to have passed into medieval Latin or the vernacular languages with
any nautical connotation and therefore its meaning cannot be
elucidated from later evidence. Perhaps the word was used in the
Anonymous’s own age for the stern ornaments but, against this, he
seems to have understood the word as applied to the substructure of
the krab(b)at(t)os, which would rule out the ornaments. Another
possibility is that there may have been a scribal error in the
Ambrosiana manuscript and that the word may have been bavndwne"
(bando2nes). In later medieval Latin and Italian, banda could mean a
parapet around the poop. Jal derived its etymology from Anglo-Saxon
bendan, to bend; however, surely a more likely derivation is from
Byzantine Greek.162 A more attractive alternative is to associate the
------------------------------
The ultimate source of all of this was probably Apollodo2ros, but Apollo2nios
Sophista also added to the debate. See Apo2llonios Sophista, Lexicon, p. 49, l. 6:
“a[flaston to; ajkrostovlion. kevklhtai de; kat ajntivfrasin ou[tw" to; ajsqenevstaton, oi|o n
eu[qlaston.”.
161
Erbse, Scholia graeca, O.717 (vol. 4, pp. 149-50): “a[flaston: ouj to;
ajkrostovlion, ajlla; to; ejp i; th'" pruvmnh" ei;" u{yo" tetamevnon ejk kanonivwn platevwn,
dihvkonto" di aujtou' platevo" kanonivo u, ojnomazomevnou me;n qranivtou, uJphreismevnou de;
tw'/ stulivskw/ tw'/ o[pisqen tou' kubernhvtou.”. Scholia on the Argonautika of Apollo2nios
of Rhodes, I.1089a, ed. Wendel, Scholia, p. 97: “..., e[stin ou\n a[flaston ‹ouj› to;
ajkrostovlion ‹to;› kata; th;n prwvran, ajll oJ poihth;" aujto; paradivdwsin ejpi; th'" pruvmnh"
levgwn ... e[stin ou\n a[flaston sanivd ion kata; th;n pruvmnan.”.
The sixth- (or ninth-) century Philoxenos Latin-Greek glosses of Paris,
Bibliothèque Nationale, MS. Lat. 7651 also associated it with the prow. See Goetz,
Glossarii Latini, vol. 2, p. 19, l. 43: “Aplustra asfalto" kaito2 akron th" prwa"” [sic].
See “Note on citations of Greek and Latin glossaries”, p. lxix above.
162
See Jal, Glossaire nautique, p. 234. In addition to the sources cited by Jal, the
word was also used in the registers of the chancery of the Angevin Kingdom of Sicily
during the reign of Charles I. See Pryor, “Galleys of Charles I of Anjou”, pp. 62-3.
There was a significant linguistic influence from Greek on the Latin used in the
Angevin chancery as far as maritime terminology was concerned; however, in the
context in which the word was used in the Angevin documents, banda appears
unlikely to have had anything to do with the poop. It was associated with the sides of
Western galleys, specifically with the outriggers for the oars. See also Kahane and
218 CHAPTER FOUR
word with whatever gave rise to the later Italian “bardone” and
Turkish “bardone/bradone” for a backstay. In classical Greek a
backstay had been an ejpivtono" (epitonos), as opposed to provtono"
(protonos), a forestay. It has been suggested that bardone/bradone
was derived from paravtono" [iJmav"] (paratonos [imas]), very
tentatively a rope stretched to the side, giving rise to medieval Latin
partanus, found in one entry in the Angevin chancery registers.
However, the latter was merely a manuscript error for peronus, itself
an alternative for prodanus (Lat.), prodano (It.), for a cable used to
lower a mast, all of which were derived from protonos because when
masts were lowered they were lowered towards the stern and the cable
used to control them during the process thus had to run forward.163 In
fact, neither forestays nor backstays could be used with lateen sails
because they would get in the way of manœuvring the yards.164 By the
age of the Anonymous both epitonos and protonos must have been
complete anachronisms. In any case, it is clear that the Anonymous
did not mean backstays by his use of bordo2nes because he used the
word in the context of the structure of the poop. Nevertheless, it is
certainly possible that because of its association with the stern,
whatever its actual meaning in Byzantine Greek, bordo2nes later gave
rise to the word for a backstay in Italian and Turkish once square sails
returned to the Mediterranean. Whatever the case, once again the
Anonymous appears to have been very confused by all of these terms.
According to the Anonymous, the aftermost part of the stern was
the parexeiresiva (parexeiresia) and there were paravptera
(paraptera), “side wings”, there, which were known as ejpwtivde"
(epo2tides). In a piece of wordplay derived either from a scholion on
Thucydides or from Hesychios, he explained the meaning of
parexeiresia as being derived from “outside the eiresia”, the oar-
bank.165 On classical Greek trie2reis this was in fact true because a
------------------------------
Tietze, Lingua Franca, §§15-16 (pp. 56-9) on the Italian “alla banda”, Turkish
“alavand/alabánda” and the Greek “bavnta/pavnta”.
163
Kahane and Tietze, Lingua Franca, §78 (pp. 95-6). Because they were
unfamiliar with the Angevin registers, the authors misunderstood the passage in
question. See Pryor, “Galleys of Charles I of Anjou”, p. 43.
Note that although the authors were familiar with the Anonymous in Dain’s edition,
they did not make a connection between bordo2nes in the Anonymous and the
Italian/Turkish bradone/bardone, even though Dain had mistranscribed bordo2nes as
bardo2nes. See Appendix Three, n. 38.
164
They are rarely shown in illustrations of lateen-rigged galleys and when they
appear to be it is always a question either of some other cable, such as a top-tackle, or
of artistic ignorance. See the numerous illustrations in Morrison, Age of the galley and
Quand voguaient les galères.
165
Appendix Three, §7.5: “Parexeiresiva dev ejsti to; o[pisqen mevro" th'" pruvmnh",
parexeiresia was an outrigger for the upper bank of oars. The epo2tides
were the lateral timbers at the bow of such outriggers which took the
shock of any head-on collision. According to Thucydides, Corinthian
ships at the second battle of Naupaktos in 413 B.C.E. rammed the
Athenian ships prow to prow and were able to smash their
parexeiresiai because they themselves had “reinforced outrigger
cheeks (epo2tides) for this very purpose”.166 However, many later
writers had no idea what either parexeiresiai or epo2tides were because
these terms had fallen out of use together with outriggers for oars,
which had been discontinued since the days of Roman liburnae.167
The scholion on Thucydides, and Hesychios also, had both located the
parexeiresia at the prow, pro2ra, in front of the oars, which was
incorrect. Assuming that the scholion in fact predated Hesychios, and
that the latter copied it rather than vice versa, this suggests that the
scholion was written after war galleys had ceased to have
parexeiresiai. The author of the Souda also repeated the location of
the parexeiresia at the prow derived from the scholion on
Thucydides,168 suggesting that he also did not understand
parexeiresiai. His, and the Anonymous’s, misunderstanding is clear
evidence that tenth-century galleys did not have outriggers either.
The Anonymous, however, located the parexeiresia at the poop,
prymne2, which was also incorrect. He appears to have done so on the
authority of another scholion on Thucydides. In reading Thucydides
on the battle of Pylos in 425 B.C.E., where the historian wrote that the
Spartan admiral Brasidas fell unconscious into the parexeiresia of his
ship,169 the Anonymous seems to have had before him a scholion
which located parexeiresiai at the poop as well as at the prow.170
------------------------------
e[nqa ta; paravptera tw'n new'n eijsin, a} ejpwtivde" kevklhntai: levgetai de; ou{tw" dia; to;
parekto;" th'" eijresiva" ei\nai to; phdavlion oiJo nei; ejr evtton kai; ijquvnon th;n nau'n.”. Cf.
Hude, Scholia, VII.34.5 (p. 385): “ta;" parexeiresiva": parexeiresiva ejsti;; to; kata; th;n
prwv/ran pro; tw'n kwpw'n, wJ" a]n ei[poi ti" to; pavrex th'" eijr esiva ".”; Hesychios, Lexicon
(Schmidt), P.834 (vol. 3, p. 282): “parexeiresivan: to; kata; th;n prw'/ran pro; tw'n
kwpw'n: wJsei levgoi ti", pavrex th'" eijresiva"”.
166
Thucydides, Peloponnesian war, VII.34.5 (vol. 4, pp. 64-5): “... ejp aujto; tou't o
pacutevra" ta;" ejpwtivda" ejcousw'n.”. See further Casson, Ships and seamanship, pp.
84-6.
167
See Casson, Ships and seamanship, pp. 143-6. Once again, in the sixth century
the well-versed Agathias had still understood the meaning of parexeiresia. See
Agathias, Historiae, V.21 (p. 192): “..., kwpwth'ra" ejf eJkatevra/ pleura'/ kai; oi|o n
parexeiresiva" aujtomavtou" ejmhcanhvsanto .”.
168
Souda, P.559 (vol. 4, p. 51): “Parexeiresiva": para; Qoukudivdh/ to; kata; th;n
prwvran pro; tw'n kwpw'n: wJ" a]n ei[poi ti" to; pare;x th'" eijr esiva"”.
169
Thucydides, Peloponnesian war, IV.12.1 (vol. 2, p. 230): “..., kai; traumatisqei;"
polla; ejlipoyuvchse te kai; pesovnto" aujtou' ej" th;n parexeiresivan ...”.
170
Cf. Hude, Scholia, IV.12.1 (pp. 234-5): “parexeiresiva ejsti;n oJ e[xw th'" eijresiva "
220 CHAPTER FOUR
Elsewhere, the Anonymous said that on either side of the poop, the
quarter rudders rested on “spreaders”, pevtasoi (petasoi), “dividers”,
scistav (schista), and the epo2tides. Then he said that the quarter
rudders, phdavlia (pe2dalia), were composed of tiller, oi[ax (oiax),
shaft, aujchvn (auche2n), and blade, ujperuvption (hyperyption). Where a
helmsman, kubernhvth" (kybertne2te2s), “leaned”, ejpiklivnetai (epikli-
netai), was an a[gklima (anklima). Once again, this was all derived
from Pollux.171 The anklima may have referred either to the helmsmen
“leaning” on their tillers or else to their being positioned at the sides
of the poop where the camber of the deck sloped off towards the hull.
Because Pollux had done so also, the Anonymous understood
correctly the terminology for the construction of rudders, which were
composed of blade, shaft, and tiller. His equations of the word
pe2dalion for the rudder as a whole with those for two of its composite
parts, auche2n for the shaft and oiax for the tiller, can also be accepted
as what was probably vernacular usage in the tenth century.172 But yet
again he probably derived the synonymity from Pollux or
Hesychios.173
------------------------------
th'" new;" tovpo", kaq o} mevro" oujkevt i kwvpai" kevcrhntai. e[sti de; tou't o to; ajkrovtaton th'"
pruvmnh" kai; th'" prwvra".”.
171
Appendix Three, §2.6: “Th'" de; pruvmnh" ta; mevrh pavlin eJkavtera pevtasoi kai;
scista; kai; ejpwtivd e" levgontai, ejn oi|" ejpivkeintai ta; phdavlia. Kai; to; me;n a[kron tou'
phdalivou h[toi tou' aujcevno" levgetai oi[ax: o{pou de; oJ kubernhvth" ejp iklivnetai a[gklima
kalei'tai. To; de; pa'n oi[ax te kai; phdavlion, to; de; teleutai'on uJperuvption, to; de; loipo;n
aujchvn.”. Cf. Pollux, Onomasticon (Bethe), I.89-90 (vol. 1, p. 29): “to; de; a[kron tou'
phdalivou [oi[ax: to; de; pa'n] oi[ax te kai; phdavlion [kalei'tai]. to; de; mevson aujtou' fqei;r h]
rJivza h] uJpovzwma, to; de; teleutai'on pteruvgion, to; de; loipo;n aujchvn. i{na de; kataklivnetai
oJ kubernhvth", a[gklima kalei'tai.”.
Uperuvption is an otherwise unknown word, probably a misreading of Pollux’s
pteruvgion for the classical ptevr ux for the blade of an oar.
On the development of rudders, see Mott, Development of the rudder.
172
Both Leo VI and Nike2phoros Ouranos recommended that dromons should carry
shafts, auche2n es, amongst their spares; however, they were probably using the word
for the entire quarter rudders. See Appendix Two [a], §5 and Appendix Five, §4. The
scholia on Aristophanes’ Peace, l. 142 in the tenth-century Ravenna manuscript also
said that the pe2dalion was: “... what we now call the auchenion.”. See Rutherford,
Scholia Aristophanica, Pax.142 (vol. 2, p. 47): “... phdavliovn ejsti to; nu'n kalouvmenon
aujcevnion.”. John Kaminiate2s also used the word aujchvn for the entire quarter rudder.
See p. 241, n. 252 below.
The Greek-Latin Cyril glosses of London, British Library, MS. Harley 5792;
used auche2n, oiax, and pe2dalion interchangeably. See Goetz, Glossarii Latini, vol. 2,
p. 252, l. 11: “Auchnploiou gubernaculum [sic]”; p. 379, l. 55: “Oiax clauus :
gubernaculum clauus singularitertan tum declinabitur [sic]”; p. 407, l. 17: “Phdalion
gubenaculum clauum serraculum [sic]. See “Note on citations of Greek and Latin
glossaries”, p. lxix above.
173
See Pollux, Onomasticon (Bethe), I.89 (vol. 1, p. 29) in n. 171 above. Cf.
Hesychios, Lexicon (Schmidt), O .175 (vol. 3, p. 181): “oi[ake": phdavlia h[toi
aujcevnia.”.
Petasoi and schista were terms mentioned by neither Leo VI nor

Nike2phoros Ouranos and they are not found in this sense either in
classical sources or in Pollux, Hesychios, Pho2tios or the Souda. The
word petasos was derived from petavnnumi (petannymi), having the
sense of opening or spreading out. In classical Greek a petasos could
be a distinctive kind of hat, a broad leaf, or an awning or canopy. In
the Harley manuscript copy of the Cyril glosses, a petasos was
understood to be a fatigium, for which is understood fastigium;
probably meaning a roof.174 Schista was derived from scivzw (schizo2),
having the sense of splitting, and schista were listed together with the
periboloi in the inventory for the Cretan expedition of 949 as items on
which 11 nomismata were spent in constructing, or reconstructing, the
11 karabia for the Rho2s .175 If our understanding of peribolos as
cathead is correct, and if the Anonymous actually knew the word to
have the same meaning, then it is possible that by analogy he used
schista for some things projecting from the hull at the stern and
associated with the rudders. That would almost certainly mean that
they were the through-beams to which the rudders were attached,
beams which had to “split” through the hull. Petasoi may then have
been coverings of the housings for the rudders, “spreading” the hull. If
we understand the housings for the quarter rudders as being composed
of “boxes” of some sort on the outside of the hull,176 then petasoi and
schista in the senses deduced above does seem to make some sense.
Box mountings for rudders can be seen in many illustrations of ships
from antiquity through to the High Middle Ages. However, whether
dromons of the tenth century actually had their rudders mounted in
boxes like this is impossible to know since the pictorial evidence from
late antiquity to the High Middle Ages shows a variety of mounting
systems for the rudders, including, “wings”, and “rings” on the
outside of the hull, as well as boxes.177 Exactly how the quarter
------------------------------
174
See Goetz, Glossarii Latini, vol. 2, p. 406, l. 29: “Petaso" fatigium
[fastigium]”. See “Note on citations of Greek and Latin glossaries”, p. lxix above.
175
See Appendix Four [b], §VI.13: “ejdovqh uJp e;r nauphghvsew" tw'n aujt w'n iaV
karabivwn tw'n te scistw'n kai; tw'n peribovlwn aujt w'n kai; loipw'n ,, iaV.” [= Haldon,
“Theory and practice”, p. 231; Constantine VII, De cerimoniis, II.45 (vol. 1, p. 675)].
176
See, for example, the mounts for quarter rudders of Rhodian and Roman galleys
in Casson, Ships and seamanship, plates 108, 114, 128, 129, 131, 170.
177
By way of example, amongst the figures here, Figures 1 and 12, and possibly 4,
6, 7, and 11 show “box” mounts. Figures 3, 5, 24, and 35 show “wing” mounts.
Figures 34, and 47 show “ring” mounts. Figures 8-19, 13-17, 22-3, 26, 36 48-54, and
57 either do not depict the rudders at all or else show no mounting system.
Figure 21 shows a through-hull mount on a galley representing the constellation
Argo in a manuscript of the Aratea attributed to Germanicus Caesar. The manuscript
has been dated to the eleventh century with a somewhat doubtful provenance of the
222 CHAPTER FOUR
rudders of dromons were mounted and what the terminology was for
the various elements of the mounts and housings is impossibly
indeterminate. Mott has argued that there were several different
systems for mounting rudders and it is indeed probable that different
terminology was used at different times and for different systems.178
The word paraptera used by the Anonymous is a hapax legomenon
otherwise unknown but its literal sense of “side wings” seems to be
quite appropriate for the whole complex of the housings for the
quarter rudders. What these were known as in classical Greek is
unknown. However, when the Anonymous used epo2tides for part of
the housing for the rudders, and equated epo2tides with paraptera, he
went horribly wrong because he was again relying on scholia on
Thucydides. Just as the meaning of parexeiresia became forgotten, so
also did that of epo2tides. The scholia on Lucian’s Zeus trago2dos, Zeus
rants, said that in one of the understandings of the chnivsko"
(che2niskos), the goose-head stern ornament on Greco-Roman sailing
ships, the epo2tides were joined to it.179 One of the scholia on the same
passage of Thucydides referring to the second battle of Naupaktos
said that the epo2tides were timbers on either side at the prow and the
author of the Souda repeated this.180 Only because his scholion had
understood epo2tides was the author of the Souda also able to
understand the term. However, when the Anonymous came to
Thucydides and the scholia that his manuscript had, he relied on other
scholia which located both the epo2tides and the parexeiresia at the
stern. In reading Thucydides on the continuation of the Peloponne2sian
war after the battle of Naupaktos, when the Syracusans emulated the
------------------------------
monastery of St Bertin. It is said to have been copied from an earlier manuscript from
Rheims. However, the pictures, including this one, are quite correctly said to have
been modelled, probably at more than one remove, on others from late antiquity,
probably the fourth century. The galley here is clearly drawn in a style similar to
others of late antiquity; for example the Dermech mosaic from Carthage (see Figure
35) and one of the Piazza Armerina mosaics (see Casson, Ships and seamanship, fig.
141), both of which are dated to the early fourth century. The backstays for the mast
give away a presumption of a square sail such as those carried by liburnae in the age
prior to the dromon. See Héliot, Manuscrits illustrés, pp. 687-8; Catalogue général,
pp. 7-8.
Other examples of “through-hull” mounts can be found in Mott, Development of
the rudder, figs 4.3 and 5.20-22.
178
Mott, Development of the rudder, pp. 9-69.
179
See Lucian, Zeus rants, §47, in Lucian, vol. 2, p. 162; Rabe, Scholia in
Lucianum, Zeu;" tragw/dov".47 (p. 82).
180
Souda, E.2848 (vol. 2, p. 400): “Epwtivde": Qoukudivdh": ... Epwtivde" eijsi ta;
eJkatevrwqen prwvra" ejxevconta xuvla ”. Cf. Hude, Scholia, VII.34.5 (p. 385): “ta;"
ejpwtivda": ejpwtivde" eijsi; ta; ejkatevrwqen th'" [th'" omitted in five MSS] prw'/ra"
ejxevconta xuvla.”.
Figure 21
Through-hull rudder mounts on a galley representing the constellation
Argo in a manuscript of the Aratea attributed to Germanicus Caesar
(Boulogne-sur-Mer, Bibliothèque Municipale, MS. 188, fol. 78),
eleventh century.
Corinthians by reinforcing the epo2tides of their ships,181 it appears that

he had before him another scholion which glossed epo2tides as: “The
[things] projecting on each side of the ship at the poop”.182 It must
------------------------------
181
Thucydides, Peloponnesian war, VII.36.2 (vol. 4, p. 68): “... kai; ta;" ejpwtivda"
ejpevqesan tai'" prwv/rai" paceiva", ...”.
182
Hude, Scholia, VII.34.5 (p. 385): “ta;" ejpwtivda": ta; [ejf] eJkatevr wqen th'" nho;"
pro;" th'/ pruvmnh/ ejxevconta.”.
224 CHAPTER FOUR
have been this scholion or one similar to it which suggested to the

Anonymous that epo2tides meant the housings for the quarter rudders.
The Anonymous said that the tillers, oi[ake" (oiakes), were bound to
something called a trochante2r.183 In the context of a ship, the word
trochante2r is known to have been used only rarely. Its sense was
given only by Hesychios, who said that trochante2res were a part of
the poop around the quarter rudders,184 and it seems certain that this
was the Anonymous’s source. However, in anatomy the word was
much better known. Gale2n referred to the processes or eminences on
either side of the neck of the femur which serve as points of
attachment for the muscles which control movements of the femur as
the great and the small trochante2r. They are still known in modern
anatomy as the greater and lesser trochanters and in insect entymology
the trochanter is that segment of the leg between the coxa attached to
the body and the femur. Again, muscles controlling the femur are
attached to it.185 In the second century C.E., Sextus Empiricus quoted
an epigram in which the word may have been used more loosely for
the hip joints, the sockets of the hip bones in which the heads of the
femurs sit, although his words will also stand a more technical
interpretation in accordance with Gale2n’s description.186
By analogy trochante2res may have been some sort of ball and socket
------------------------------
183
Appendix Three, §2.15: “Ta; de; th'" new;" scoiniva: kavloi, provtonoi, peivsmata,
ajpovgaia, prumnhvsia, kai; e[mboloi, oi} tou;" oi[aka" sunevcousi kai; di w|n eij" to;n
trocanth'ra ajpodesmw'ntai.”.
184
Hesychios, Lexicon (Schmidt), T.1523 (vol. 4, p. 181): “trocanth're": pro;" ta;
phdavlia. kalei'tai th'" pruvmnh" mevro"”. The word was also used by Nike2tas David
Paphlagon. See Nike2tas David Paphlagon, Vita S. Ignatii, coll. 516-7: “pro;" oi|" kai;
tw'n tou' patriavrcou monasthrivwn barbarikw'/ katadramovnte" oJr mhvmati kai; qumw'/,
pa'san me;n th;n euJreqei'san kthvsin ajfeivlonto, ei[kosi de; kai; duvo tw'n gnhsiwtevrwn
aujtou' kekrathkovt e" oijketw'n, ejf eJni; trocanth'ri ploivo u tou;" pavnta" ajxivnai"
katemevl isan.”.
Later the word appears to have changed its form and to have become trecanthvr i
(trechante2ri), which gave rise to Turkish tirhandil for a stern frame or transom. See
Kahane and Tietze, Lingua Franca, §863 (pp. 585-6). Note that the authors’
understanding of this passage of the Anonymous is completely wrong. They have
made emboloi into “pins” and trochante2r into a “sternpost”; which cannot be justified
since quarter rudders of the tenth century were located nowhere near the sternpost and
were not attached to anything by pins. The authors were thinking of post-medieval
sternpost rudders.
185
See Galen, De usu partium, 15.8 (vol. 2, pp. 370-74), trans. May, 15.8 (vol. 2,
pp. 676-79). See also Rosse, Hollinshead’s textbook of anatomy, pp. 314-5; Chapman,
Insects, pp. 134-6.
186
Sextus Empiricus, Against the professors, I.316-7, in Sextus Empiricus, vol. 4, p.
184: “ a[rqrw/ ejn ajspidoventi bebhkovta gui'a kaq o{lmou // blaisa; trocanthvrwn a[cri
peristrevf etai, ...”. Bury translated trochante2res here as “hip joints”; however, there
is no reason why in this text the word should not refer to the trochante2res as described
by Galen.
mountings for the rudders of ships. However, there is no pictorial

evidence for any such ball and socket mountings for the rudders of
classical galleys.187 More probably, they were projections on the
rudder shafts to which tackles were attached in order to control the
rudders. Tackles attached to projections on the shafts of the rudders
would parallel very closely the idea of muscles attached to the
trochante2res of the femur. Such tackles and their fastenings to the
rudders are shown in the Torlonia relief of ca 200 and in a third-
century mosaic from a house in Rome.188 This interpretation does not
fit either Hesychios’s description of the trochante2res as being part of
the poop or the Anonymous’s description of them as being some
things bound to the tillers by emboloi; although, in the latter case only
a small leap of imagination would be necessary. However, both
Hesychios and the Anonymous very probably knew only that they
were some things associated with the rudders at the poop.
Quarter rudders were huge blades with extremely long shafts, so
long that if hoisted to the mastheads, they could be used as bearers on
which to construct flying bridges from ships’ masts to surmount the
sea walls of besieged towns. Theophane2s the Confessor said that the
Muslims had intended to hoist them up and rest them against the
ramparts of Constantinople during the siege of 717 and John
Kaminiate2s also described them as being used for this purpose by the
Muslims at the siege of Thessalonike2 in 904.189 On galleys of the
Kingdom of Sicily in the later thirteenth-century, the earliest for
which accurate dimensions survive, the quarter rudders (temones) of
galleys were 6.06 metres long.190 It is not surprising that elsewhere the
Anonymous also listed amongst the ship’s cables e[mboloi (emboloi),
which he said restrained the tillers, oiakes, and by which these were
bound to the trochante2r.191 Whatever he may have understood by the
vocabulary he used, he was almost certainly referring to the blocks
and tackles which were needed to control the quarter rudders of
ancient and medieval ships.192 A variety of terms were used at various
times and in various places for these tackles.
------------------------------
187
See Mott, Development of the rudder, pp. 9-40, esp. pp. 35-9.
188
See Casson, Ships and seamanship, Ill. 144, 154 and, for discussion, Mott,
Development of the rudder, pp. 22, 39-40.
189
See Theophane2s, Chronographia, A.M. 6209 (vol. 1, p. 396). For John
Kaminiate2s see below pp. 240-42.
190
See Pryor, “Galleys of Charles I of Anjou”, p. 56.
191
See p. 224 & n. 183 above.
192
See Casson, Ships and seamanship, pp. 228-9; Mott, Development of the rudder,
pp. 29-30, 75-8.
226 CHAPTER FOUR
As early as the fifth century B.C.E., Euripides had written in Helen

that, in the preparation of a pente2konteros for sea, the rudders were let
down by zeuvglai (zeuglai), lit. “yokes”.193 This word appeared again
in a cognate form in the Acts of the Apostles when the crew of the ship
on which St Paul was travelling let go the zeukthrivai (zeukte2riai) of
the rudders before driving the ship ashore on Malta. Casson is surely
correct when he equates zeuglai and zeukte2riai to “pennants”, that is
tackles, of the rudders.194 In fact, Hesychios defined zeuglai as: “zygoi,
or parts of the zygos. And zeukte2res [are] binding straps. And
metaphorically tropo2te2res”.195 His meaning was clear. Just as
tropo2te2res were grommets for binding oars to tholes, zeukte2res and
zeuglai were “straps” for some other purpose; arguably for controlling
the rudders. The Argonautika attributed to the Orphic corpus, which
was derived from Apollo2nios of Rhodes, also said that the rudders,
oi[hke" (oie2kes), were fixed at the poop and tied off with straps.196
Vitruvius, following the pseudo-Aristotelian Me2chanika, referred
to a helmsman holding the tiller, ansa gubernaculi, which he said the
Greeks called oiax, and moving it with one hand carefully around the
centre [of the rudder shaft]. The Me2chanika said that the point at
which the rudder was attached to the ship was the uJpomovclion,
hypomochlion, the “fulcrum”.197 However, neither Vitruvius nor the
------------------------------
193
Euripides, ÔElevnh , l. 1536, in Euripides, Fabulae, vol. 3, p. 64: “phdavlia te
zeuvglaisi parakaqiveto .”.
194
Souter, Novum Testamentum, Pravxei" tw'n ÔAgivwn Apostovlwn, 27.40: “kai; ta;"
ajgkuvra" perielovnte" ei[wn eij" th;n qavlassan, a{ma ajnevnte" ta;" zeukthriva" tw'n
phdalivwn: ...”. See Casson, Ships and seamanship, p. 228, n. 17.
195
Hesychios, Lexicon (Schmidt), Z.120 (vol. 2, p. 256): “zeuvgla": zugoiv, h] mevr h
tou' zugou'. kai; zeukth'r e" iJmantovdesmoi. kai; tropwth'r e" metaforikw'".”.
196
Orfewv" Argonautika;, ll. 276-7 in Abel, Orphica, p. 13: “ejpi; d au\t oi[hka"
e[dhsan prumnovqen ajrthvsante", ejp esfivgxanto d iJma'sin.” (“on these they bound the
rudders from the poop and tied [them] off with straps.”).
197
Vitruvius, De architectura, X.iii.5, trans. Granger, vol. 2, p. 298:
“Quemadmodum etiam navis onerariae maximae gubernator ansam gubernaculi
tenens, qui oiax [oiax] a Graecis appellatur, una manu momento per centrum ratione
pressionibus artis [aptis] agitans, versat eam amplissimis et inmanibus mercis et pinus
[penus] ponderibus oneratam.”. “Aptis” and “penus’ were editorial emendations by
Krohn in his edition of 1912.
The middle clause of this passage almost defies understanding. Both Morgan
and Granger had to guess at a translation. We believe that its meaning was as follows:
“So also the helmsman of a very large cargo ship holding the tiller of the rudder,
which is called an oiax by the Greeks, moving [it] with one hand around the centre [of
the rudder shaft] in a trice with care [and] with small (skilful?) amounts of force, turns
it [the ship] loaded with very great and even enormous weights of merchandise and
pine [wood] (provisions?).”
Aristotle, Me2chanika, §5, pp. 354-6: “Dia; tiv to; phdavlion mikro;n o[n, kai; ejp
ejscavtw/ tw'/ ploivw/, tosauvthn duvnamin e[cei w{ste uJpo; mikrou' oi[ako" kai; eJno;" ajnqrwvpou
dunavmew", kai; tauvth" hjremaiva", megavla kinei'sqai megevqh ploivwn; h] diovti kai; to;
author of the Me2chanika appear to have been using technical language

here.
In a letter which was a translation of a now lost Paschal letter of
bishop Theophilos of Alexandria, St Jerome made an analogy to the
helmsmen of great ships who met the rush of oncoming seas by
turning the bows in alternate directions by tightening or slackening the
“lines” (funiculi) of the rudders (gubernacula). Once the seas had
passed, they let off the “straining bonds” (laborantia vincla) of the
rudders (clavi).198
(f) Deck and Castles
According to the Anonymous, from poop to prow at the centre line,

above the floor, ajmfimhvtrion (amphime2trion), and hold, kuvto" (kytos),
dromons were undecked, ajsavnidon (asanidon): “The middle of the
stern and the ship as far as the prow is undecked. ... And the bottom
[of the ship] is named the hold (kytos) and the floor
(amphime2trion)”.199 Yet again, this was based on Pollux; however, the
Anonymous misunderstood Pollux because he had at hand a
manuscript of the Onomasticon with a significant variant. In fact
Pollux wrote that: “The middle of the stern [is] a sanivdion (sanidion),
where that inside [is] the ejnqevmion (enthemion), [and] that attached
------------------------------
phdavliovn ejsti moclov", kai; mocleuvei oJ kubernhvth". h|/ me;n ou\n proshvrmostai tw'/ ploivw/,
givnetai uJpomovclion, to; de; o{lon phdavlion oJ moclov", ...” (“Why does the rudder, which
is small and at the end of the vessel, have so great power that it is able to move the
huge mass of the ship, though it is moved by a small tiller and by the strength of but
one man, and then without violent exertion? Is it because the rudder is a bar
(mochlos), and the helmsman works a lever? The point at which it is attached to the
ship is the fulcrum (hypomochlion), the whole rudder is the bar (mochlos), ...”).
198
Jerome, St, Epistolae, 100.14, col. 825: “Sicut enim gubernatores magnarum
navium, cum viderint immensum ex alto venire gurgitem, quasi venatores
ferocissimam bestiam, spumantes fluctus suscipiunt, eosque prorae objectione
sustentant, flectentes in diversum gubernacula, et prout ventorum flatus et necessitas
imperarit, stringentes funiculos, vel laxantes, cumque unda subsederit, ex utroque
navis latere laborantia clavorum vincula dimittunt, ...”. On this passage see Casson,
Ships and seamanship, pp. 228-9.
199
Appendix Three, §2.6: “To; de; mevson th'" pruvmnh" kai; new;" mevcri th'" prwv/ra"
ajsavnidon.”; §2.8: “Kai; to; me;n e[dafo" aujth'" kuvto" kai; ajmfimhvtrion ojnomavzetai.”. Cf.
Pollux, Onomasticon (Bethe), I.87 (vol. 1, p. 28): “kai; to; me;n e[dafo" th'" new;" kuvto"
kai; gavstra kai; ajmfimhvtrion ojnomavzetai.”.
On the amphime2trion, see Hesychios, Lexicon (Schmidt), A.4065 (vol. 1, p.
163): “ajmfimhvtria: ta; meta; th;n trovpin th'" new;" ejx eJkatevrou mevr ou" ejpitiqevmena.”;
repeated exactly by Pho2tios, Lexicon (Theodoridis), A.1354 (p. 141): “ajmfimhvtria: ta;
meta; th;n trovpin th'" new;" ejx eJkatevrou mevrou" ejp itiqevmena.”.
228 CHAPTER FOUR
[is] the ejpiseivwn (episeio2n)”.200 Although some manuscript variants

have either ajsavndion or a[sandron for sanivdion, Pollux’s latest editor,
Bethe, was surely correct to choose sanivdion, as attested to by the best
manuscripts.201 However, the Anonymous must have been using a
manuscript with a variant such as those rejected by Bethe since he
used ajsavnidon. Even if it is unclear what exactly Pollux meant by the
sanidion with its enthemion and episeio2n, it is neverthless certain that
he did not mean that the stern, or any part of a ship, was “undecked”.
Sanidion was, of course, a diminutive of, or derivative synonym for,
saniv" (sanis), meaning a board, or plank, or timber, or just about
anything made from wood; in particular, a platform, scaffold, or stage.
By sanidion, Pollux probably meant the platform of the poop, stepped
up from the level of the main deck. What he meant by enthemion and
episeio2n is arguable.
Jal thought that the enthemion was the internal part of the
asandion, but he was following Scheffer, who was merely repeating
Pollux. However, Jal then said that in literary Greek of the 1840s the
word was used for a dunette, a “poop”, known also as a kavssaron
(kassaron): the highest part of a quarter deck, where there was a cabin
for the officers.202 Hesseling claimed that enthemion was the same as
modern Greek despevntsa (despentsa) or koumpavnia (koumpania) and
meant a storeroom for provisions in the hold;203 however, he provided
no evidence for this and this was clearly not the meaning of the word
as used by Pollux. Since the word is unknown in nautical terminology
except in Pollux, the only other known use of the word being in the
Septuagint translation of the Old Testament in Exodus, where the
word meant a socket of a candelabrum,204 there is simply no way of
------------------------------
200
Pollux, Onomasticon (Bethe), I.90 (p. 29): “[to;] mevson de; th'" pruvmnh" sanivdion,
ou| to; ejnto;" ejnqevmion, to; d ajphrthmevnon aujt w'/ ejpiseivwn.”.
201
Asavndion is derived from a now-lost manuscript, at least one copy removed
from the manuscript of Arethas of Caesarea, from which the two extant fifteenth-
century manuscripts of Paris, Bibliothèque Nationale, MS. Gr. 2646 and Salamanca,
University Library, MS. I 2.3 stemmed. It is also found in the thirteenth-century
manuscript Paris, Bibliothèque Nationale, MS. Gr. 2647. “Asandron and ajsavndion are
found in the twelfth-century manuscript Heidelberg, Universitäts Bibliothek, MS. Pal.
375.
Dindorf relied on the Paris manuscripts for his edition of 1824 and consequently
has ajsavndion rather than sanivdion. See Pollux, Onomasticon (Dindorf), I.90 (vol. 1, p.
88).
202
See Jal, Glossaire nautique, pp. 607, 633, 877.
203
Hesseling, Mots maritimes , pp. 17, 21.
204
See Septuagint, Exodus, 38.16: “... kai; ta; lampa;dia aujtw'n, a{ ejstin ejp i; tw'n
a[krwn, karuwta; ejx aujtw'n: kai; ta; ejnqevmia ejx aujt w'n, i{na w\sin ejp aujtw'n oiJ luvcnoi, kai;
to; ejnqevmion to; e{bdomon ajp a[krou tou' lampadivou ejpi th'" korufh'" a[nwqen, ...”.
knowing what he may have meant by it. As for episeio2n, Casson has
argued that it was a pennant or flag of some kind on the basis of
Pollux, Onomasticon, I.91, where the parts of a mast were enumerated
from the keel to the masthead.205 A pennant or flag also fits the sense
of episeio2n here, where it was similarly attached, ajphrthmevnon
(ape2rte2menon) to the enthemion. However, all that Pollux actually
said was that something called an episeio2n was at the top of the mast
above everything else. Rather than a flag, it may have been a flag
pole. Then, in relation to the poop, Pollux may have meant that the
episeio2n was a flagpole which was set in an enthemion, which by
analogy to the Septuagint text could have been a socket for the pole.
The Anonymous interpreted Pollux’s wording in the way he did
because of the manuscript available to him and therefore his testimony
that tenth-century dromons were undecked at the centre line should be
questioned on textual grounds alone. Moreover, corroborating
evidence from other sources leads inescapably to the conclusion that
they were in fact fully-decked, just as they had been in the days of
Prokopios.206
First, the implications of an otherwise incomprehensible passage in
the Naumachika Leontos Basileo2s of Leo VI, which was reiterated
with minor changes by Nike2phoros Ouranos, suggest unequivocally
that this was the case. The passage says that: “Moreover, they will set
up the so-called wooden castles, xulovkastra (xylokastra), fortified
with planks, around the middle of the mast on the largest dromons,
from which men will throw into the middle of the enemy ship mill
stones or heavy iron [weights], like sword-shaped blooms, ...”.207 As it
------------------------------
205
Pollux, Onomasticon (Bethe), I.91 (vol. 1, pp. 29-30): “kai; to; me;n uJpodecovmenon
to;n iJsto;n lhnov" [kalei'tai], to; de; ejnarmozovmenon aujtw'/ ptevrna, to; de; teleutai'o n to;
pro;" th'/ keraiva/ hjlakavth kai; qwravkion kai; karchvsion, to; de; uJpe;r th;n keraiva n
a[trakto", ou| kai; aujto;n to;n ejpiseivo nta ajpartw'si.”. See Casson, Ships and
seamanship, p. 246, n. 86.
206
Alexandres, Dolley, and Eickhoff all accepted uncritically the evidence of the
Anonymous that dromons were only half-decked amidships and had full decks only at
the bow and the stern. See Alexandres, ÔH qalassiva, pp. 65-66; Dolley, “Warships”,
p. 50; Eickhoff, Seekrieg und Seepolitik, p. 138.
207
Appendix Two [a], §7: “Alla; kai; ta; legovmena xulovkastra peri; to; mevson tou'
katartivou ejn toi'" megivstoi" drovmwsin ejpisthvsousi periteteicismevna sanivsin, ejx w|n
a[ndre" tine;" eij" to; mevson th'" polemiva" nho;" ajkontivsousin h] livqou" mulikou;" h] sivdhra
bareva, ...”.
Note that Nike2phoros Ouranos altered “peri; ...” to “pro;" to; mevson tou'
katartivou”. Appendix Five, §6: “Alla; kai; ta; xulovkastra periteteicismevna uJpo;
sanivdwn i{na sthvkwsin eij" tou;" megavlou" drovmwna" pro;" to; mevson tou' katartivou pro;"
to; sthvkein a[ndra" ‹eij"› aujta; kai; rJivptein mevson eij" to; polemiko;n h] livqou" megavlou"
mulikou;" h] sivdhra bareva, ...”.
Cf. also Appendix Three, §3.2. The Anonymous also referred to xylokastra but
230 CHAPTER FOUR
stands this passage is meaningless. No commander worth his salt

would ever have gone into battle with wooden “castles” slung halfway
up the the masts of his galleys. Any shock of impact would bring them
crashing down and those in them would have been easy prey for an
enemy severing the shrouds of the masts with rigging cutters, which
were a normal part of the equipment of warships.
Vegetius wrote that: “The very sharp curved iron [blade] in the
likeness of a falx (sickle) is called a sickle, because, set in long pikes,
it quickly cuts the chalatorii by which the yard is suspended and, the
sails having collapsed, makes the liburna slow and useless”.208 In
Greek a sickle was a drepavnh (drepane2), or drevpanon (drepanon), and
logcodrevpana, longchodrepana, “lance-sickles’, are found among the
armaments of dromons in an inventory for the Cretan expedition of
949 in the De cerimoniis.209 They were no doubt similar to the
squarciavele, “sail cutters”, used on Angevin galleys in the thirteenth
century,210 which can be seen in the illustration of a galley in the
------------------------------
did not locate their position on the dromons.
208
Vegetius, Epitoma, IV.46 (pp. 164-5): “Falx autem dicitur acutissimum ferrum
curvatum ad similitudinem falcis, quod contis longioribus inditum chalatorios, quibus
antemna suspenditur, repente praecidit conlapsisque velis liburnam pigriorem et
inutilem reddit.”.
Casson believes that the chalatorii were “lifts”, that is slings, from the
mastheads to the ends of the yards of the square sails of liburnae. See Ships and
seamanship, p. 262, n. 11. If so, the “sickles” could have been used only by men on
the yards of an attacking ship and the operation must have been very hazardous.
However, the passage is a re-working of a passage in Caesar’s Gallic War describing
his fleet’s encounter with the ships of the Veneti and it is doubtful whether it has any
independent value. See Caesar, Gallic War, III.14 (pp. 156-8): “Una erat magno usui
res praeparata a nostris, falces praeacutae insertae adfixaeque longuriis, non absimili
forma muralium falcium. His cum funes, qui antemnas ad malos destinabant,
comprehensi adductique erant, navigio remis incitato praerumpebantur. Quibus
abscisis antemnae necessario concidebant, ...”. We owe this reference to Casson.
Caesar left little doubt that funes were halyards because when they were cut, the
yards collapsed. Vegetius appears merely to have substituted chalatorii for funes and
therefore the word must have meant halyards or tie tackles, in which case, the sickles
could have been used from the deck.
On thirteenth-century, lateen-rigged Western galleys “callati/collativi/collaturi”
were definitely tie tackles. See Pryor, “Galleys of Charles I of Anjou”, p. 58. In fact
no lateen-rigged ships could have “lifts” because the upper yard was much higher
than the masthead. But if the chalatorii were tie tackles or halyards and the sickles
were used from the deck, it then becomes difficult to understand why the defending
crew could not simply break them or push them away from the halyards or ties.
In any case, the passage of Vegetius suggests that liburnae went into battle with
masts and yards still raised. Although this was apparently the practice of the Veneti, it
was contrary to normal practice in the Mediterranean in antiquity and the Middle
Ages. The veracity of Vegetius’ entire passage must be questioned.
209
See Appendix Four [b], §II.11 [= Haldon, “Theory and practice”, p. 225;
210
manuscript of the De rebus Siculis carmen of Peter of Eboli. [See

Figure 54]
Masts were normally lowered before going into battle if at all
possible in order to prevent their smashing the hulls and causing loss
of life if they came crashing down.211 For example, in his description
of the battle of the Aegates islands in 241 B.C.E., Polybios wrote that
when the Carthaginians descried the Roman fleet they “lowered their
masts and cheering each other on in each ship closed with the
enemy”.212 A well-known example of what could happen if masts
were not lowered before battle occurred at the battle of Ayas in 1294.
There a much superior Venetian fleet of 68 galleys made the cardinal
mistake of going into battle against a smaller Genoese fleet with masts
still raised and sails unfurled. The result was a catastrophe. The
Venetian admiral Marco Basegio was killed and 25 of his galleys
were lost.
“Castles” halfway up a mast could have had no conceivable
purpose; however, the critical phrase in the texts, “around” (Leo VI),
or “towards” (Nike2phoros Ouranos) “the middle of the mast”
(“peri;”or “pro;"” “to; mevson tou' katartivou ”), stands easy emendment
to either “around the middle mast” (“peri; to; mevson katavrtion”) or
“around the middle [i.e., half way between] of the masts” (“peri; to;
mevson tw'n katartivwn”).213 There is also another possibility. The term
for the midships mast of Western sailing ships and galleys in the High
Middle Ages was “arbor de medio”, “the mast of the middle”.214 This
was not a “middle” mast between two others, but rather any mast
stepped amidships, irrespective of any other masts. The second mast
of these Western ships was the “arbor de prora”. If the Greek were to
be emended to mean “around the mast of the middle” in parallel to
later Western Latin usage, it would read: “peri; to; katavrtion tou'
mevsou ”.
Whatever the case, Vegetius had said that such castles had also
existed on larger Roman liburnae and Appian described them being
erected at the bows and sterns of Marcus Agrippa’s ships at the Battle
of Mylai in 36 B.C.E. and Dio Cassius by Marcus Antonius on his
------------------------------
211
Cf. Morrison, et al., Athenian trireme, p. 43.
212
Polybios, Histories, I.61 (vol. 1, p. 166): “oiJ de; Karchdovnioi katidovnte" to;n
diavploun aujtw'n prokatevconta" tou;" ÔRwmaivo u", kaqelovmenoi tou;" iJstou;" kai;
parakalevsante" kata; nau'n sfa'" aujtouv", sunevballon toi'" uJpenantivoi".”.
213
These emendations were suggested in Dolley, “Warships”, p. 51.
214
See Pryor, “Naval architecture”, p. 284; idem, “Galleys of Charles I of Anjou”,
pp. 40, 55.
232 CHAPTER FOUR
ships before Actium in 31 B.C.E.215 They are also seen on bas-reliefs

and other pictorial depictions of Roman galleys.216 The Po2s dei
diapleein potamous attributed to Maurice also recommended that
“castles” should be constructed on dromons for archers to fire from.217
From a structural point of view, it is very difficult to imagine how
such castles could have been raised on anything but full decks and the
testimony of Pollux was very clear on this point. He wrote that: “... if
the ship is fully decked, platforms for towers are built, and on these
are two towers, left and right, between which (is) the deck”.218 This
makes sense. Castles were built on both sides with a gangway on the
deck between them. It accords with the construction of decked Roman
liburnae to the age of Vegetius and, if the text of Pollux as we have it
reflects dromons of the ninth century, it indicates that they were fully
decked at that time and surely in the tenth century also.
Secondly, when Leo VI, and Nike2phoros Ouranos following him,
wrote that: “If you realize that any of the soldiers are cowardly, send
them to the lower oar-bank, and if any of the soldiers should be
wounded or fall you should fill his place from those below out of
necessity”,219 it again suggests that they were fully decked. Cowardly
marines were dismissed to the oar-bank “below” [the deck], where
they would be safe. The calibre of men there was such that they were
brought into combat above the deck only when the need was dire.
Thirdly, John Kaminiate2s wrote in his De expugnatione
Thessalonicae that when Leo of Tripoli’s fleet carrying him and his
fellow captives left Crete for Tarsos, he was on a “Roman warship”, a
die2re2s, which the Muslims had raised from the sea at Thessalonike2
after it had been sunk, and that: “ ... the barbarians with us put
themselves on the upper benches and left the lower [benches] to us,
which were full of deep gloom and evil smells, and could only be
------------------------------
215
Vegetius, Epitoma, IV.44 (p. 162). Cf. Appian, Civil wars, 5.106 (vol. 4, p.
554); Dio Cassius, Roman history, 50.23.3-3 (vol. 5, p. 484).
216
See, for example, Casson, Ships and seamanship, plate 130; Höckmann,
“Liburnian”, pp. 202-4; Morrison, Greek and Roman oared warships, pp. 356-8.
217
Maurice, Ek tou Maurikiou, §3: “ejmbavllein de; eij" tou;" toiouvtou" drovmwna"
crhsivmou" kai; gennaivo u" toxovta" kai; kastellw'sai aujtouv".”. Cf. Maurice,
Strate2gikon, XIIB.21.15-16 (p. 468): “Crhsivmou" de; toxovta" ejn aujtoi'" bavllein kai;
gennaivo u", kai; kastellw'sai aujtouv".”.
218
Pollux, Onomasticon (Bethe), I.92 (vol. 1, p. 30): “eja;n d h\/ katavfrakton to;
ploi'on, ejp inauphgou'ntai purgou'coi, kai; ejp aujt w'n purgiva duvo, dexio;n kai; eujwvnumon,
w|n mevson to; katavstrwma.”.
219
Appendix Two [a], §20: “Eij dev tina" tw'n stratiwtw'n ajnavndrou" ejp ignw'/" touvtou"
eij" th;n kavtw ejlasivan paravpempe, kai; ei[ potev ti" plhgh'/ h] pevsh/ tw'n stratiwtw'n, to;n
ejkeivnou tovpon ejk tw'n kavtw ejx ajnavgkh" ajnaplhrwvsei".”. Cf. Appendix Five, §18;
Appendix Eight [a], p. 244.
described as a floating grave”.220 Again this suggests that the deck was
a full deck and that conditions below it were dark and fetid as a result.
An argument ex silentio for dromons and chelandia having full
decks may also be drawn from the history of the word uJpovzwma,
hypozo2ma, “undergirdle”. Any ships as long and shallow as dromons
or chelandia were particularly prone to hogging and sagging as they
moved across waves. In classical trie2reis like Olympias the problem
of hogging was overcome as far as was possible by the use of
hypozo2mata, heavy ropes running from stempost to sternpost and back
again which were tensioned, most probably with deadeye tackles,221 so
that they acted in some respects as hogging trusses. Trie2reis had to
have hypozo2mata because they had only part decks, not full decks.
However, full decks removed the need for hypozo2mata because the
deck itself acted as a truss against both hogging and sagging.
Consequently it was important that it be as rigid longitudinally as
possible. By the thirteenth century, on galleys of the Kingdom of
Sicily, the deck was laid on deck beams which ran from the heads of
the frames at the hull and which were locked together longitudinally
by stringers run from bow to stern under the beams and which were
mortised onto them. Internal longitudinal stringers mortised onto the
frames also performed the same function.222 There can be little doubt
that the decks of Byzantine dromons and chelandia were constructed
similarly and the fact that the word hypozo2ma disappeared from the
language and became misunderstood suggests that part-decked war
galleys no longer existed. In fact they may have disappeared very
early. The word hypozo2ma did not pass into Latin. In classical Greek
literature it had been used by Plato in his Republic and Laws but it
was not scholiated in the medieval manuscripts of these works. At the
end of the second century Pollux misunderstood the word to mean a
------------------------------
220
John Kaminiates, De expugnatione Thessalonicae, §74.7-8 (p. 64): “h\men de;
pavnte", wJ" ei[rhtai, ejn mia'/ nhi; ÔRwmaiva/ polemikh'/ h{t i" h\n dihvrh", w|n th;n me;n a[nw
kaqevdran oiJ lacovnte" ei\nai meq hJmw'n bavrbaroi eJautoi'" ejklhrwvsanto, th;n kavtw de;
hJmi'n ejpafh'kan, skovtou" pollou' kai; duswdiva" plhvrh. kai; tiv a[llo ge crh; levgein h]
tavfon ejn u{dasi poreuovmenon; ...”. Cf. §§61.3, 73.12.
On the interpretation of this passage see also Livadas, “Medieval nautical
technology”. Livadas misunderstands kaqevdra as a deck rather than a rowing bench.
221
See Coates, “Spanish windlasses”; Coates and Shaw, “Speculations”.
Note that the rope hypozo2mata of Olympias could never be made to work and
that a wire hawser tensioned to 12 tonnes was used instead. See Coates,
“Reconstruction”, p. 22; Coates, et al., Trireme trials, pp. 6-8; Morrison et al.,
Athenian trireme, pp. 169-71, 220-2; Coates, “Development”, p. 74; Coates and
Shaw, “Speculations”.
222
See Pryor, “Galleys of Charles I of Anjou”, pp. 39, 48-9.
234 CHAPTER FOUR
part of a quarter rudder,223 and in the fifth century Hesychios did not
include it in his Lexicon at all. Nor did Pho2tios include it in his
Lexicon in the ninth century and towards the end of the tenth century
the author of the Souda misunderstood it to mean a timber of a ship.224
Finally, the evidence that at least some Muslim and Western
galleys of the tenth to twelfth centuries, all of which were ultimately
descended from Byzantine galleys, were fully decked,225 again leads
to the conclusion that the model on which they were based was also
fully decked.
The castles must have been built along either beam, with a
clearway between them, for two reasons. First, it should be noted that
in the texts of both Leo VI’s Naumachika Leontos Basileo2s, §7 and
Nike2phoros Ouranos’s, Peri thalassomachias, §6, “the largest
dromons” is plural, as is “castles”.226 Both texts can stand either of
two interpretations. Either there was only one castle and the whole
sentence refers to multiple dromons, or else there was more than one
castle on each dromon and the sentence refers to individual ships.
Dolley, Eickhoff, and others have thought either that there was only
one such castle and that it straddled a dromon from bulwark to
bulwark somewhere amidships or else that there were two castles, one
at the bow and one amidships, but again both straddling the ship from
bulwark to bulwark.227 However, in addition to the testimony of
Pollux, the Anonymous’s description of the mast and yard crutches
makes it certain that neither of these interpretations was the case. All
classical and medieval war galleys had a series of crutches, known in
Greek as iJstodovkai (histodokai), set up down the centre line, on
which masts rested when lowered before going into battle or for other
reasons.228 The only conceivable construction which could make it
------------------------------
223
Pollux, Onomasticon (Bethe), I.89: “to; de; a[kron tou' phdalivou [oi[a x: to; de; pa'n]
oi[ax te kai; phdavlion [[kalei'tai]]. to; de; mevson aujtou' fqei;r h] rJivza h] uJpovzwma, ...”.
224
Souda, U.493 (vol. 4, p. 669): “ÔUpozwvmata: xuvla th'" newv". ajnti tou' eijp ei'n
uJpozwvmata oJ skuteu;" pro;" mavgeiron paivzwn ei\pe zwmeuvmata, wJ" ajrtuvsewn e[mpeiron
kai; zwmeumavtwn. ajpeivrhto de; ajpo; Aqhnw'n ejxavgein xuvla kai; pivssan. ei\con de; kai;
Lakedaimovnioi trihvrei", ejp eidh; h\rcon kai; nhvswn tinw'n. Aristofavnh": tou'ton to;n
a[ndr ejgw; deivknumi kai; fhvm ejxavgein th'/si Peloponnhsivwn trihvresi zwmeuvmata.”.
225
See Pryor, “From dromo2n to galea”, pp. 107-12.
226
See above pp. 229 & n. 207; Appendix Two [a], §7; Appendix Five, §6.
227
See Dolley, “Warships”, p. 51; Eickhoff, Seekrieg und Seepolitik, pp. 138-9 and
plan p. 8.
228
Eickhoff assumed that the crutches were to take the masts and yards when in
harbour or during extended periods of rowing against the wind. He failed to
appreciate that on war galleys their main purpose was to permit their lowering before
going into battle. See Eickhoff, Seekrieg und Seepolitik, p. 138.
possible for a dromon both to have castles somewhere around a mast

and also for the masts and yards to be lowered onto the crutches when
necessary was for there to be two castles built to either side of the
centre line, just as Pollux said, allowing for a clear gangway down the
middle and for the crutches to be set up on the deck down the centre
line.
Secondly, in order for lateen-rigged ships to tack, the yards have to
be raised to the vertical and then swung across forward of the masts to
the other side.229 The foot of the yard has to be hauled in flush to the
mast so that the yard is suspended vertically by its halyard rove
through the block at the masthead. If a castle were built across the
whole beam of the ship anywhere forward of a mast, the foot of the
yard could not possibly be hauled in to the mast. If it was built behind
the mast, then they would get in the way of the sail. The only way in
which a ship with castles like this could tack would be if its yard was
shortened by at least twice the height of the castle, its upper yard
being shortened in order to balance the yard by at least the equivalent
of the shortening of the lower yard made necessary to clear the castle.
This would hardly be desirable from the point of view of the
performance of the ship under sail. Eickhoff realized that his
construction of the castles posed severe difficulties for the handling of
the yards and sails.230 In fact, it is impossible to imagine how galleys
could possibly have managed their lateen sails effectively if
constructed with superstructures of the type that Eickhoff and Dolley
thought dromons had. One glance at Dolley’s model shows
immediately that its midships sail could not possibly have been
tacked,231 and it is probable that Eickhoff’s could not have been either.
The problem is removed if the castles were not built across the beam
of the ships but rather to either side so that the yard could be brought
to the vertical.
The Arabic version of this paragraph of the Naumachika Leontos
Basileo2s included by Muh5ammad ibn Mankalı33 in his Al-adilla al-
rasmiyya fi ’l-ta’a2bı3 al-h5arbiyya translated Leo VI’s clauses as: “In
each ship there should be a tower beside the mast (al-s5a2rı3), ...”. The
corresponding clause was identical in his Al-ah5ka2m al mulu2kiyya wa
’l-d5awa2bit al-namu2siyya.232 Christides understood the Arabic term “al-
s5a2rı3”, for the mast, as “al-s5a2rı3 al-kabı3r”, the large mast located in the
------------------------------
229
See Pryor, “Mediterranean Round Ship”, pp. 67-9.
230
See Eickhoff, Seekrieg und Seepolitik, p. 141.
231
Dolley, “Warships”, plate V.
232
Appendix Two [a], §7. Cf. Appendix Eight [a], p. 242; [b], p. 21.
236 CHAPTER FOUR
middle of the ship,233 and therefore located the castle amidships. The
model of a dromon recently constructed under his direction for an
exhibition on Oinoussai of models of medieval Byzantine and Arab
ships perpetuates this.234 However, in fact, the largest of the masts on
all lateen-rigged medieval ships, whether sailing ships or galleys, was
always the foremast.235 The Arabic versions of Leo VI almost
certainly had it correct. Castles were on the deck adjacent to a mast,
but the mast to which al-s5a2rı3 referred was surely a foremast rather
than a midships mast. The castles would have been placed towards the
bow around the foremast where they would be most effective in
combat. The only illustration known to us of a Byzantine galley with
such a castle is in the Madrid manuscript of the Synopsis historio2n of
John Skylitze2s, fol. 31v. The illustration is of the arrival of the fleet of
the rebel Thomas the Slav at Abydos and the castle is depicted exactly
where it should be, towards the bow of Thomas’s flagship. [See
Figure 33]
In initial phases of combat, galleys almost always engaged by the
bow and castles amidships would have been of little use. The closer
castles were to the bow, the better the angles of fire would have been
for archers stationed in them and the easier it would have been to hurl
rocks or iron weights onto enemy ships. Broadside engagement such
as would have made castles amidships effective was avoided
whenever possible and would normally have occurred only when
formations had been broken up and battles had degenerated into
mêlées.236 In fact, since galleys in combat would normally approach
each other from opposite directions, even if they did end up engaged
side to side they would most probably do so engaged stern by bow
and vice versa.237 Castles would have been most useful at the bow,
from where the vulnerable poops of enemy galleys could be attacked.
One final issue concerning the castles must be addressed. In 1840
------------------------------
233
See Christides, “Ibn al-Manqalı3”, p. 86; idem, Conquest of Crete, p. 44, n. 53.
234
Andriotes, Buzantinav kai Arabikav istiofovra ploiva, fig. 1: Drovmwn-Dromon.
See also Christides, “Introduction”, pp. 29-30. The exhibition was organized in
conjunction with the Eighth International Congress on Graeco-Oriental and African
Studies: «Navigation and trade in the Mediterranean from the 7th c. – 19th c. A.D.»,
Oinoussai, 5-9 July 2000.
235
See below p. 243 & n. 259.
236
237
Consider the battle of Ponza of 14 June 1300 between the Angevin fleet under
Roger of Lauria and the Sicilian fleet under the Genoese admiral Conrad d’Oria. The
admirals’ galleys of Lauria and d’Oria slid along each other’s sides until engaged
stern to bow. See Pryor, “Naval battles of Roger of Lauria”, p. 210.
Figure 22
Chelandium on a medal forged by Alvise Meneghetti (1691-1768)
attributed to a Venetian Doge Pietro Candiano, as published by
Augustin Jal.
Augustin Jal reproduced a sketch of a medal with an engraving of a

chelandia on it with some sort of tower amidships and inscriptions
reading PETRUS CAND. DUX CsELANDIÆ PORTIS C. FECIT on
the obverse and SECURITAS VENETIÆ on the reverse, which he
said had been shown to him in Venice in 1834 by Giovanni Casoni, a
historian of the Venetian arsenal. The Doge referred to has been
variously identified since there were four Doges of Venice of the
Candiano family called Pietro in the ninth and tenth centuries.238 Jal’s
sketch gave rise to over a century of scholarly discussion of Byzantine
and Venetian chelandia in the ninth and tenth centuries. However, the
medal itself disappeared from view and no scholar actually saw it
until it was identified in the Archaeological Museum of Zagreb as one
of numerous medals and coins from the collection of Dr Giorgio Catti
of Fiume which was acquired by the Department of Archaeology of
the National Croatian Museum of Zagreb in 1894. In 1991 it was
published as one of many forgeries made by the Venetian jeweller and
antique dealer Alvise Meneghetti (1691-1768).239 The medal is a weld
------------------------------
238
Jal, Archéologie navale, pp. 246-8. Cf. Glossaire nautique, pp. 465 and 751.
Doges of the Candiano family by the name of Petrus/Pietro were: Pietro
Candiano I (887), Pietro Candiano II (932-9), Pietro Candiano III (942-59), and Pietro
Candiano IV (959-76).
239
See Gorini, Mirnik, and Chino, “Falsi di Meneghetti”, pp. 324-7 & no. 8 (p.
238 CHAPTER FOUR
of two laminae of copper alloy and the inscriptions should read

“PETRUS CAND. DUX CHELANDIÆ PORTIS C. FECIT”, and
“SECURITAS VENETIÆ”. The history of the medal and the
scholarship on it has now been traced by Reinhold Mueller,240 but it is
valueless for study of the construction of dromons or chelandia.
(g) Masts, yards, and sails
Although the texts of Leo VI and Nike2phoros Ouranos referred to

plural dromons and castles, they referred to a singular mast.241
However, all possible emendations of the texts made necessary by the
need to explain the castles imply that dromons must have had more
than one mast. “Around the middle mast” (peri; to; mevson katavrtion)
implied a mast between two others. “Around the middle of the masts”
(peri; to; mevson tw'n katartivwn) implied between two masts. “Around
the mast of the middle” (peri; to; katavrtion tou' mevsou) implied that
there was another mast distinguished from the midships mast.
All that can really be said about the masts is that dromons almost
certainly had at least two: a midships mast and a foremast. Dolley was
surely correct to point out that a foremast and foresail would have
been necessary in order to enable ships as large as these to be
manageable with lateen sails.242 Eickhoff drew his standard dromon
with two masts but he made the second one a small mizzen mast at the
stern, which is impossible. He also hypothesized, without any
evidence, that the largest dromons may have had a third mast, which
would have had to have been placed at the bow.243 However, his
reconstruction is historically inappropriate and the same is true of
Dolley’s replica, which also has the largest mast and sail amidships. It
is also true of the Oinoussai exhibition model of a dromon. Both
Eickhoff and Dolley were apparently influenced by the norm for post-
medieval square-rigged ships, which had the largest mast amidships
between a mizzen mast and a foremast. Their entire reconstructions of
the sailplans can be dismissed because they did not appreciate that
------------------------------
329).
240
Mueller, “Venetian ships”. We are indebted to Reinhold Mueller, who traced the
medal and its history at our request.
241
See Appendix Two [a], §7 and Appendix Five, §6.
242
Dolley, “Warships”, p. 52. Dolley claimed that with only one mast their bows
would have fallen off the wind. In fact they would have luffed up into the wind out of
control.
243
See Seekrieg und Seepolitik, p. 140.
from antiquity to the end of the Middle Ages all lateen-rigged ships of
all kinds, including war galleys, always had their largest mast and sail
towards the bow and a smaller one towards the stern. The
documentary evidence for this for Western galleys of the High Middle
Ages is very clear and there is no reason to suspect that it was not the
case also for similarly lateen-rigged Byzantine dromons.244 Although
he constructed his dromon replica with three masts, Dolley was quite
circumspect on this issue and did not actually claim that they had
three masts;245 however, following Dolley’s lead, and relying on a
graffito from a later period, Christides has claimed that they did have
three.246 It is indeed true that several graffiti from Byzantine churches
and monasteries show two- and three-masted galleys, frequently with
what appears to be a forked tongue, presumably Greek Fire, projecting
from the bow. However, none of these can be firmly dated to the era
of the Macedonian emperors. The dating of all of the graffiti is so
impossibly indeterminate as to make them virtually useless for the
study of the dromon in any period.247
------------------------------
244
See Pryor, “Galleys of Charles I of Anjou”, p. 55. It is true that the only
depiction of a two-masted galley known to us from the Middle Byzantine period, the
miniature of the manuscript of the Sacra Parallela, our own Figure 8, does show the
midships mast longer than the foremast. We attribute this, however, to the artist’s
need to cram the drawing into a very elongated space.
245
Dolley, “Warships”, p. 51. Alexandres also claims three masts. See ÔH qalassiva,
p. 72.
246
Christides, “Byzantine dromon and Arab s5h5ı3nı3”, p. 115; idem, “New light”, fig.
5 (p. 9); Andriotes, Buzantinav kai Arabikav istiofovra ploiva, fig. 1: Drovmwn –
Dromon.
Christides says that the graffito in question is from the “temple of Christ the
Saviour in Megara” and dates from the twelfth century. However, he attributes his
source to Alexandres, ÔH qalassiva, fig. 2.10 (p. 58), where it is ascribed to the
monastery of the Blatado2n at Thessalonike2, which was not established until ca 1355.
247
See Alexandres, ÔH qalassiva, fig. 2.1 (p. 58) [a two-masted galley from the
church of Galatision at Athens, which is dated to the late thirteenth century] and fig.
2.8 (p. 58) [a three-masted galley? (no oars are shown but the hull configuration is
that of a galley) from the church of the Holy Apostles at Thessalonike2, which is dated
to 1310-14].
See also Meinardus, “Medieval navigation”, fig. V.4 (p. 42) [a two-masted
galley from the church of St Luke of Stiris in the subterranean church of St Barbara,
which may date to the tenth century] and fig. IX.2 (p. 48) [a three masted galley? (no
oars are shown), from the Hephaisteion (church of St George), Athens, date
unknown].
See also Goudas, “Mesaiwnika; karavgmata”, fig. 19 (p. 336) [a two masted
galley? (no oars are shown) - from the Hephaisteion (Church of St George), Athens.
date unknown]
As Meinardus himself wrote: “..., there is no reason to assume that all of these
akidographemata were the pious expressions of Greek sailors. For that matter, it is
quite likely and even probable that some of these scratchings represent the ships of
Venetian and Genoese sailors.” (p. 32).
240 CHAPTER FOUR
One other argument ex silentio may be adduced against the

proposition that dromons had three masts. The Anonymous’s major
lexicographical source, Pollux, had discussed in clear terms the three
masts that some ancient ships had had, naming them as foremast,
ajkavteio" (akateios), midships mast, ejpivdromo" (epidromos), and a
smaller mast, presumably at the stern, dovlwn (dolo2n). Hesychios
described them similarly, except that he made an ejpivdromon
(epidromon) the mast at the stern.248 The Lexicon of Pho2tios did not
include any of these terms;249 however, significantly, the author of the
Souda did, but had no idea what they meant.250 Since the author of the
Souda did not understand Pollux’s terminology, perhaps neither did
the Anonymous, simply because by their age these terms for masts
had been forgotten. The Anonymous chose not to repeat Pollux and
made no mention of the number of masts that dromons had. Perhaps,
from the window of the library in Basil’s palace, he could see that
dromons sailing the Golden Horn or the Bosporos did not have three
masts and for once allowed reality to take precedence over philology.
Dolley claimed that the flying bridges used by Leo of Tripoli at the
siege of Thessalonike2 in 904 A.D., as described by John Kaminiate2s,
proved that the Muslim ships employed in the attack had two masts, a
foremast as well as a main mast.251 The passage of Kaminiate2s in
question reads:
..., coupling all the ships together in pairs, the one against the other, and
tightly holding the sides of each one to the other with strong cables and
iron chains so that they would not easily drift apart, they hauled by means
of the rigging hanging at the bow the timbers standing in the middle
which sailors are accustomed to call masts, then raising by means of the
ropes twisted around the mastheads the quarter rudders of each ship into
the air, pushing the blades out from the prow beyond the extent of the
ship. ... when the quarter rudders had been suspended aloft in the manner
described, they placed long strips of wood over them in rows, one next to
------------------------------
248
Pollux, Onomasticon (Bethe), I.91 (vol. 1, p. 30): “kai; oJ me;n mevga" kai; gnhvsio"
iJsto;" ajkavt eio", oJ de; katovp in ejpivdromo", oJ de; ejlavttwn dovlwn. kalei'tai dev ti kai;
lovggaso".”. Cf. Hesychios, Lexicon (Schmidt), A.2302 (vol. 1, p. 94), D.2185 (vol. 1,
p. 528), E.4760 (vol. 2, p. 152). These ships were not, however, specifically galleys of
any type. Some ancient and medieval sailing ships certainly had three masts.
249
The entries from ajdiavkrito" to ejp wvnumoi were missing from the manuscripts
from which Naber’s edition was compiled but were included in the thirteenth-century
manuscript Zavorda, monastery of St Nikanor (near Gravina in Northern Greece),
MS. 95 on which the new edition by Theodoridis has been based. These words,
however, do not appear in it either.
250
See Souda, A.819 (vol. 1, p. 77), D.1346 (vol. 2, p. 125), E.2310 (vol. 2, p. 355).
251
Dolley, “Warships”, p. 52.
the other, flooring in the intervening space by this ingenious method.
They then fenced in the edges on all sides with planks and secured the
ends of the shafts with very strong cables at the stern.252
This is in fact a very curious passage because there is actually no

mention of foremasts. On the one hand, the “rigging hanging at the
bow” might be considered to pre-suppose them but, if so, why was it
necessary to move the midships masts in order to raise the steering
oars aloft to construct the assault bridges to the walls? Why could they
not have been slung from the foremasts? On the other hand, if these
were single-masted ships, what was the “rigging hanging at the bow”?
And if they were single-masted ships, the masts could not have been
moved in any case because there would have been no second mast-
step for them and the decks would have been pierced in only the one
place.253 However, it appears to have escaped attention that this
passage was very similar to another in the tenth-century treatise De
obsidione toleranda which was based on Polybios’s account of the
Roman siege of Syracuse in 214 B.C.E. as preserved in the
encyclopedic Eklogaiv (Excerpta) produced for Constantine VII.
Polybios was known in literary circles in Constantinople in the tenth
------------------------------
252
John Kaminiates, De expugnatione Thessalonicae, §32.5-7 (p. 30): “..., kai;
pavsa" aujtw'n ta;" nh'a" kata; duvo suzeuvxante" eJt evran th'" eJtevra" ejcomevnhn, kai; tivsi
kavloi" stibaroi'" kai; sidhrai'" aJluvsesi ta;" eJkatevr wn pleura;" ejn ajllhvlai"
susfivgxante" pro;" to; mh; rJa/divw" aJf ivstasqai, ajneivlkusan dia; tw'n kata; prwv/ran
ajph/wrhmevnwn ejxartivwn ta; dia; mevsou probeblhmevna xuvla, a} toi'" nautillomevnoi"
katavrtia kalei'n e[qo". ei\ta tou;" tw'n eJkatevrwn nhw'n aujcevna" ejn touvtoi" dia; tw'n eij"
th;n kefalivda strefomevnwn scoinivwn mevson pou tou' ajevro" metewrhvsante", tav" te
spavqa" aujtw'n ejk tou' kata; prwv/ran mevrou" kai; pevra th'" tw'n nhw'n diamevtrou
probeblhkovte", ... uJyou' gavr, wJ" e[fhn, tw'n auJcevnwn metewrhqevntwn, e[ballon ejn aujtoi'"
ejpiqevnte" makrav tina xuvla stichdo;n a[llo kat a[llo, kai; to;n dia; mevsou tovpon
gewvsante" th'/ sesofismevnh/ tauvth/ mhcanh'/, ta; pevratav te pavntoqen sanivsi
katafraxavmenoi, kai; ta; a[kra tw'n aujcevnwn ejn toi'" kata; pruvmnan mevr esin a[lloi"
stibarwtavtoi" desmoi'" ejnasfalisavmenoi, ...”. The translation is ours.
In his translation of this passage, Dolley translated katartia as “yards” rather
than as “masts”. See Dolley, “Rig”, p. 52. Consequently, his understanding of the
construction of the flying bridges is inaccurate and his conclusion that these ships
necessarily had lateen yards cannot be sustained. They surely did have lateen yards,
but that is beside the point.
253
Livadas also interprets this passage to mean that the the Arab ships had two
masts because they had “foremast slings”, a term which he appears to have derived
from Dolley, “Rig”, p. 52, where it is unclear what Dolley actually meant by the term.
See Livadas, “Medieval nautical terminology”, p. 285. To interpret “kata; prwv/ran
ajph/wrhmevnwn ejxartivwn” as “foremast slings” is surely to stretch the meaning of the
Greek, whatever was meant by the English expression. In the context of a mast,
“slings” normally referred to ropes or chains which supported yards on the masts of
square-rigged ships; however, lateen-rigged ships do not have “slings”. They can not
have because the upper end of the yard is much higher than any “sling point” on the
mast could be.
242 CHAPTER FOUR
century. In neither the De obsidione toleranda nor Polybios were the

ships explicitly said to have had two masts and it is at least possible
that Kaminiates, who was probably influenced by Polybios, derived
his confused form of the account from him.254
Such flying bridges had certainly been used for a very long time.
Appian described their use by Mithridate2s VI Eupator at a siege of
Kyzikos in 74 B.C.,255 and Theophane2s the Confessor wrote that the
Muslim fleet under Maslama ibn ‘Abd al-Malik which besieged
Constantinople in 717 initially expected to attack the walls at the
shore and to raise the quarter rudders onto the ramparts.256 What
seems to have been a depiction of something very similar to the ship-
borne siege engines described by Kaminiate2s and the De obsidione
toleranda can be seen in an illustration at folio 40r, in a section
dealing with flying bridges, ejpibathvria (epibate2ria), run from the
mastheads of ships to the walls of a town, in the eleventh-century
manuscript Rome, Biblioteca Vaticana, MS. Gr. 1605 of the
Parangelmata poliorke2tika attributed to He2ron2 of Byzantium.257 Here
the bridges to the walls can clearly be seen constructed from long
spars decked in with planks and fenced along the sides, almost exactly
as Kaminiate2s described. They are raised by blocks from beams
crossing the mastheads of each pair of ships. However, there is a
critical difference. There are two masts here. Both the foremasts and
midships masts of each pair of galleys are in place and the bridges are
hoisted between all four masts. Such flying bridges continued to be
------------------------------
254
De obsidione toleranda, §§208-11 (pp. 220-21); Polybios, Histories, VIII.4.2-11
(vol. 3, pp. 452-5).
255
Appian, Mithridatic wars, §73 (vol. 2, p. 376): “... kata; de; tou;" limevna" duvo
penthvr ei" ejzeugmevnai puvrgon e{t eron e[f eron, ejx ou| gevf ura, oJpovte prospelavseian ej"
to; tei'co", uJpo; mhcanh'" ejxhvlleto.”.
256
Theophane2s, Chronographia, A.M. 6209 (vol. 1, p. 396): “ejbouvlonto ga;r th'/
aujth'/ eJspevra/ eij" ta; paravlia prosormivsai teivch kai; tou;" aujcevna" eij" ta;" ejpavlxei"
ejpiqei'nai.”.
He2ro2n, Parangelmata poliorke2tika, in Sullivan, Siegecraft, fig. 26.
257
Epibate2ria was a later form of ejpibavqra, epibathra, for gangways or boarding

bridges. See above pp. 193-4 & n. 86 and also Marsden, Technical treatises, pp. 85,
92, 97.
The treatise attributed to He2ro2n2 of Byzantium was in fact an anonymous treatise
compiled in the second quarter of the tenth century during the reign of Constantine
VII Porphyrogenne2tos from various earlier treatises on siege warfare. It was probably
another product of the encyclopedic movement associated with Constantine VII;
however, the author added comments of his own to his paraphrases of his sources and
on occasions these appear to have been based on some practical experience. The
attribution to a certain “He2ro2n” was added to the manuscript in the fourteenth or
fifteenth century. The section on assault bridges was paraphrased from a section of
the Peri me2chane2mato2n of Athe2naios Me2chanikos.
Figure 23
Ship borne siege engines in a treatise on poliorcetics attributed to He2ro2n of
Byzantium (Rome, Biblioteca Apostolica Vaticana, MS. Vat. Gr. 1605, fol.
40r), eleventh century.
used throughout the Middle Ages. Robert of Clari and Count Hugh of
St Pol described their construction by the Venetians for the assault on
Constantinople by the Fourth Crusade in 1203.258
In the absence of any Byzantine data the best estimate of the length
of the masts of dromons that can be made is by comparison to that of
the masts of thirteenth-century Sicilian galleys, whose middle masts
were 11.075 metres long and whose foremasts were 15.82 metres
long, being raked forward at approximately thirteen degrees to the
vertical, bringing their mastheads to approximately 14.20 metres
------------------------------
258
Robert of Clari, Conquête de Constantinople, §44 (p. 44): “Adont fist li dux de
Venice molt merveillex engiens faire et molt biaus, car il fist prendre les antaines qui
portent les voiles des nes, qui bien avoient trente toises de lonc ou plus; si les fist tres
bien loier et atakier a boines cordes as mas, et fist faire bons pons par deseure et bons
puis encoste de cordes; si estoit li pons si les que trois chevalier armé i pooient aler de
front. Et fist li dux les pons si bien warnir et couvrir as costés d’esclavinnes et de toile
que cil qui i montassent pour assalir n’eussent warde ne de quarriaus d’arbalestes ne
de saietes; et lanchoit li pons tant avant outre le nef qu’il avoit bien de hauteur du pont
dusques a tere pres de quarante toises ou plus; ...”. Hugh of St Pol, Epistola to Henry
of Louvain, in Tafel and Thomas, Urkunden, vol. 1, p. 308: “Dux Venetie vero super
quamlibet navim construxit de antennis pontem altissimum, in altitudine 100 pedes
habentem; et super quemlibet pontem poterant ire quatuor milites de fronte.”.
244 CHAPTER FOUR
above the water line. However, these galleys had an overall length
from stempost to sternpost of 39.55 metres and a beam amidships of
4.61 metres, whereas the estimates here for the corresponding
dimensions of dromons are 31.25 metres and 4.46 metres.259 In other
words, dromons of the tenth century were probably only around 75%
of the size of Angevin galleys. If the masts were scaled down
proportionately, the middle masts of dromons may have measured
around 8.3 metres long and the foremasts around 11.85 metres long
with a masthead height above the water line of around 10.65 metres.
The inventories for the Cretan expedition of 949 specified that the
Department of the Vestiarion basilikon was to supply for 20 dromons:
“20 calkivsia (chalkisia), together with the rest of the mavggana
(mangana)”.260 Chalkisia was surely derived from the classical Greek
karchvsion (karche2sion), which was well known to refer to a masthead
amongst other things.261 The word ought to have been known to the
Anonymous from Pollux and Hesychios and it is therefore curious that
he did not include it when enumerating the parts of masts and
rigging.262 In Latin the word became carchesium and was well known
to refer to the head of a mast.263 Jal identified karchvsion and
carchesium as the origin of the various medieval vernacular words
calces, calcez, calcese, calcet, cholzexe, for a “block mast”; that is, a
masthead with blocks inserted in it for the halyards to be rove
through.264 Since only 20 chalkisia were specified for the 20 dromons,
------------------------------
259
See Pryor, “Galleys of Charles I of Anjou”, pp. 55, 70, 74; idem, “From dromo2n
to galea”, p. 113; idem, “Naval architecture”, pp. 284-5.
260
See Appendix Four [b], §IV.9: “calkivsia kV meta; kai; ta; loipa; mavggana ,” [=
Haldon, “Theory and practice”, p. 227; Constantine VII, De cerimoniis, II.45 (vol. 1,
p. 672)].
261
See Lendle, “Das Karchesion”, esp. pp. 85-101.
262
See Pollux, Onomasticon (Bethe), I.91 (vol. 1, pp. 29-30): “to; de; teleutai'on to;
pro;" th'/ keraiva/ hjlakavth kai; qwravkion kai; karchvsion, ....”. Cf. Hesychios, Lexicon
(Schmidt), K.952 (vol. 2, p. 418): “kavrchsi. [karchvsia, ed. Latte, vol. 2, p. 419] : ta;
kevrata ta; ejpavnw tw'n katartivwn tw'n ploivwn. kai; ta; a[kra tw'n ijstw'n”.
263
See Isidore of Seville, Etymologiae, XIX.ii.9: “Carchesia sunt in cacumine
arboris trochleae, quasi F littera, per qua funes trahuntur.”.
See also the gloss on carcessium in an Abolita gloss in the eighth-century
manuscript Rome, Biblioteca Apostolica Vaticana, MS. Vat. Lat. 3321 in Goetz,
Glossarii Latini, vol. 4, p. 29, l. 35: “Carcessium est in summo malona‹uis› [sic]”.
See “Note on citations of Greek and Latin glossaries”, p. lxix above. Cf. the gloss on
carcena in the tenth-century Glossae Aynardi, Metz, Bibliothèque Publique, Cod.
Metensis 500 in ibid., vol. 5, p. 617, l. 19: “Carcena sunt loca in cacumine arboris
nauis ubi funus stant ad trahendum”; and that on carcessia in the ninth-century
Amplonianum primum glossary, Erfurt, Wissenschaftliche Allgemeinbibliothek,
Amplon. Fol. 42, in ibid., vol. 5, p. 353, l. 2: “Carcessia summitas mali”.
264
See Jal, Glossaire nautique, p. 385. The metathesis of “k” and “c” and the
liquids “l” and “r ” are both well known linguistic phenomena. See also Hesseling,
they were apparently used on only one of the masts, presumably the
largest of them: the foremast. They must therefore have been special
blocks of some sort. Much more probably, the word had already
become applied to the blockmast, as it was later in the medieval West.
Calkivsia was apparently the transmission word between karchvsion
and calces.
On other masts ordinary blocks (mangana) were apparently used to
raise the yards. We draw attention to the fact that some Byzantine and
Western illustrations of early medieval sailing ships show peculiar
mastheads “hooked” or “beaked” forward. There are too many of
these for it to be accidental. They can be seen in the ninth-century
illustrations of the ship in the Khludov Psalter and in the Paris,
Bibliothèque Nationale, manuscripts of the Sermons of St Gregory of
Nazianzos and the Sacra Parallela attributed to St John of Damascus,
[See Figures 8, 15, and 16] as well as at fol. 147r of the London,
British Library, MS. 40731 “Bristol” Psalter of the late tenth or early
eleventh centuries and at fols 117v and 201r of the London, British
Library, MS. Add. 19.352 Theodore Psalter of 1066.265 They can also
be seen in an eleventh-century manuscript of a Me2n ologion in the
monastery of the Esphigmenou of Mt Athos, in a twelfth-century
manuscript of the sermons of St Gregory of Nazianzos in the
monastery of the Pantelee2mon of Mt Athos, and in an enamel from the
Pala d’Oro of San Marco, Venice.266 In fact they were much more
ancient than this. Similar mastheads can be seen in a fourth-century
mosaic of a fishing boat from Roman Carthage, in a graffito of a
sailing ship from Corinth of the fifth or sixth centuries, and in the
painting from Kellia of ca 600-630.267 [See Figure 17] However, they
are never seen in any illustration of a ship which clearly has a square
sail. The conclusion to be drawn is that they were associated with
lateen sails. Pryor has suggested that masts with heads such as these
did not need blocks inserted in the mastheads and that lateen yards
could simply be raised with normal blocks slung from the hooks or
------------------------------
Mots maritimes, p. 19; Makris, “Griechischer lingua franca”, p. 221; Kahane,
“Byzantinoromanica”, pp. 316-17.
265
See Dufrenne, L’illustration, pl. 56; Der Nersessian, L’illustration, pl. 70, fig.
194 and pl. 112, fig. 317.
266
Paris, Bibliothèque Nationale, MS. Gr. 510, fol. 3r and MS. Gr. 923, fol. 207r;
Mount Athos, Esphigmenou, Cod. 14, fols 52a and 387a; Mount Athos, Pantelee2mon,
Cod. 6, fol. 37a. See Bass, History of seafaring, p. 149, pl. 3; Pelekanides, OiJ
Qhsauroiv, pll. 299 (p. 175), 329 (p. 208), and 348 (p. 266); Weitzmann, Sacra
Parallela, fig. 203 (pl. LIII).
267
Fantar, Mosaïque en Tunisie, p. 122; Basch, “Navires et bateaux coptes”, figs 4
and 25.
246 CHAPTER FOUR
beaks. The hooked mastheads presumably served to suspend the yards

forward of the masts so that they could be swung across the front of
the masts more easily when tacking. Apparently, it was only when
ships and their yards and sails grew large enough that blockmasts
became needed.268 Perhaps dromons of the tenth century needed
chalkisia only for their largest yards and sails and used ordinary
blocks slung from beaked mastheads for their other sails.
There is some support for this in the fact that the inventories for the
Cretan expedition of 949 specified that the Department of the
Vestiarion basilikon should have provided 20 “sails”, a[rmena
(armena), for the 20 dromons.269 But ships as large as Byzantine
dromons would have been unmanageable with only one sail. Besides
which, sails blew out constantly and no ship would ever have put to
sea without at least one spare set of sails. Leo VI himself said that
dromons should carry a duplicate set of sails.270 We should probably
understand the reference in the inventories as a generic reference to
“the” sail, that is the “mainsail”, the one used in fair weather on the
main mast towards the bow.271 It would be for this mast that the
blockmast was used.
Against this it should be pointed out that another section of the
inventories said that the Department of the Eidikon outlayed 1154
nomismata for cloth for eleven sails for nine karabia and two mone2ria
transporting Rho2s and prisoners.272 It appears that these ships really
were single-masted and that only one sail was provided for each of
them. However, these ships were smaller than dromons or chelandia.
The inventories for the Cretan expedition of 949 also specified in
the item immediately following this one referring to the blockmasts
that 20 “hoops”, yelliva (psellia), should be provided for the 20
dromons.273 Yevl(l)ion was a common word for a ring or hoop of any
------------------------------
268
See Pryor, “Mediterranean round ship”, p. 71.
269
See Appendix Four [b], §IV.3: “a[rmena kV,” [= Haldon, “Theory and practice”,
p. 227; Constantine VII, De cerimoniis, II.45 (vol. 1, p. 672)].
270
See Appendix Two [a], §5 and Appendix Five, §4.
271
That is to assume that it was not a much more simple case of some clerk in an
office thinking to himself: “One dromon, one sail”. This is a peculiar section of the
inventories where it was a case of what “should have been provided” by the
Department of the Vestiarion basilikon, whatever that means. The case of the verb
frontisqh'nai, “provide”, is aorist passive infinitive.
272
See Appendix Four [b], §VI.2-4 [= Haldon, “Theory and practice”, p. 229;
Constantine VII, De cerimoniis, II.45 (vol. 1, p. 674)]. The major point here is that
eleven sails were specified for the eleven ships; that is, one each. However, it must be
appreciated that 1154 nomismata was a huge amount of money. There simply has to
be something wrong, either in the number of sails or in the number of nomismata.
273
See Appendix Four [b], §IV.10: “yelliva kV,” [= Haldon, “Theory and practice”,
kind: an armband, anklet, etc. Elsewhere in the inventories, the word

was associated with siege engines of various kinds, including large
bow-ballistae.274 What exactly it was here can only be conjectured.
However, since only one of these per dromon was specified, it was
presumably a special piece of equipment and since this specification
follows immediately after that for the blockmasts, it is very tempting
to read the word as meaning a “parrel”, the classical Greek word for
which had probably been calinov" (chalinos).275 Yards had to be held
close to the masts by rings or hoops, which in later centuries were
made up of rope rove through wooden balls and spreaders, in order to
prevent the wind in the sails causing them to flail around and slam
back and forth against the masts.276 In medieval Italian and Latin these
were known as trozze, troçe, trosse, troce, or trocte.277 In English they
became known as “trusses” or “parrels” and parrell appears to fit the
meaning of psellion in this context. If this is correct, then psellion was
later replaced in Greek by trovtsa (trotsa), derived from the Italian
trozza.278 Either that or the inventorist was using a non-technical term.
Beyond this, the inventory for the Cretan expedition of 949
specified that the Department of the Eidikon should have provided
100 extra small sails, ajrmenovpoula (armenopoula), for the 20
dromons.279 One might well ask extra to what? Since these were
specifically said to be small, they were no doubt storm sails. Lateen
sails could not be shortened in heavy weather and it was necessary to
lower the yards, unbend the fair weather sails and bend on smaller
ones. In the West in the thirteenth century storm sails were normally
known as tertiarola because they were one third smaller than the
normal sails.280 An explanation of armenopoula as storm sails would
explain why each dromon was allocated five of them. Because they
were used in heavy weather, they would be much more prone to being
blown out than the fair weather sails.
------------------------------
p. 227; Constantine VII, De cerimoniis, II.45 (vol. 1, p. 672)].
274
See Haldon, “Theory and practice”, pp. 225: “..., kai; hJ touvt wn ejxovplisi". ...
yelliva kai; davktuloi, ...”; 229: “yelliva lV, kai; lovgw/ tw'n megavlwn toxobolivstrwn.”.
Cf. Constantine VII, De cerimoniis, II.45 (vol. 1, pp. 670, 672-3).
275
See Casson, Ships and seamanship, pp. 230, 260-3.
276
See Jal, Glossaire nautique, under “racage” (pp. 1250-51).
277
See Pryor, “Naval architecture”, p. 368; idem “Galleys of Charles I of Anjou”,
pp. 61-2.
278
See Hesseling, Mots maritimes, p. 32; Kahane, Lingua Franca, §679 (p. 450).
279
See Appendix Four [b], §III.8: “ajjrmenovpoula kata; perivsseian rV,” [= Haldon,
“Theory and practice”, p. 227; Constantine VII, De cerimoniis, II.45 (vol. 1, p. 671)].
280
See Pryor, “Naval architecture”, pp. 363-4 and Tables 3 & 4; idem, “Naval
architecture revisited”, pp. 261-66; idem, “Galleys of Charles I of Anjou”, pp. 57-8.
248 CHAPTER FOUR
As for the size of the sails, there is only one piece of evidence
known to us. The same section of the inventories for the Cretan
expedition referred to above, concerning the outlay by the Department
of the Eidikon of 1154 nomismata for eleven sails for the nine karabia
and two mone2ria transporting the Rho2s and prisoners, said that the
sails of the karabia were 30 pe2cheis and those of the mone2ria 28
pe2cheis. Assuming that these dimensions were of the same category as
those used for sails in the Latin West in the thirteenth century, where
the primary dimension specified was always the length on the yard, it
would mean that the lengths on the yard were respectively 14.04
metres and 13.10 metres. This can be compared to the 26.89 metres
for the foresail and 20.57 metres for the midships sail of Angevin
galleys of the 1270s.281 The latter had an overall length from stempost
to sternpost of 39.55 metres and a beam amidships of 4.61 metres,
whereas the estimates here for the corresponding dimensions of
dromons are 31.25 metres and 4.46 metres.282 Since dromons of the
tenth century were probably only around 75% of the size of Angevin
galleys, if we scale down the sails proportionately, the foresail of a
dromon may have measured around 20.17 metres on the yard and the
midships sail around 15.43 metres, somewhat larger than those of the
karabia and mone2ria, as one would expect. The peak of the foremast
sail may have been around 21 metres above the water line.
One of the passages penned by the Anonymous which has been
amongst the most difficult to make any sense of concerns the masts,
yards, and the so-called “stiffener keel”. He wrote that: “When [the
ship] is sailing, the mast step, travpeza (trapeza), in which the mast,
iJstov" (histos), that is, the katavrtion (katartion), is set up, is fixed in
the middle on to the keel. The lower part of the katartion which is
fixed in the mast step is called the heel, ptevrna (pterna); hence [the
expression] ‘is unheeled’ when it comes out of the mast step under
pressure from the wind”.283 He continued: “[There is] the ‘mast
receiver’, iJstodovkh (histodoke2), and the yard, keraiva (keraia), the
keratavrion (keratarion). The sail, iJstivon (histion), [is] the a[rmenon
(armenon). And what are known as the kaqormei'" (kathormeis) are
fixed (proshvl wntai) firmly (sterew'") in a row on the keel, trovpi"
------------------------------
281
See Pryor, “Galleys of Charles I of Anjou”, pp. 41, 55, 57.
282
See Pryor, “Galleys of Charles I of Anjou”, pp. 70, 74.
283
Appendix Three, §2.9: “ Th'" de; pleouvsh" mevson ejpi; th'" trovpio" prosarmovzetai
hJ travpeza, h|" ejnto;" oJ iJsto;" i{statai, h[toi to; katavrtion. Tou' de; katartivou to; me;n
proshlouvmevnon th'/ trapevzh/ katwvt eron mevro" ptevrna kalei'tai, ejx ou| kai; to;
ejxeptevrnisen, o{tan uJpo; ajnevmou biazomevnh e[xw th'" trapevzh" ejkbh'/.”. On the
interpretation of this passage see also Koukoules, “Nautiko;" bivo"”, p. 353.
(tropis), there being three of them, on which the keraia rests [when]
lowered”.284
In his edition of this passage Dain made a fateful misreading. For
“sterew'"”, an adverb meaning “firmly” and qualifying proshvl wntai
meaning “fixed”, he read “sterea'"”, an adjective meaning “firm” and
qualifying trovpio", the keel. In the scholarship, the words “trovpio"
sterea'"” have been understood ever since as referring to a trovpi"
stereav (tropis sterea), a “stiffener keel” on which the kathormeis
were fixed or mounted.285 But in fact the Anonymous said simply that
the kathormeis were fixed firmly on the keel and that the yards could
rest on them when lowered. The infamous tropis sterea never existed,
not even in terminology, let alone in reality.
Figure 24
Mosaic of a galley with a lowered mast from a sepulchre at Hadrumetum,
Tunisia, third century.
Katartion was a post-classical word for a mast, perhaps first

recorded in scholia on the Odyssey, and the Anonymous probably got
the synonymity from there, as also that for histion and armenon: “The
histion [is] the armenon, and the histos [is the] katartion, that is the
------------------------------
284
Appendix Three, §2.10: “ÔIstodovkh de; kai; keraiva to; keratavrion. ÔIstivo n de; to;
a[rmenon. Kai; oiJ legovmenoi kaqormei'" ejp i; th'" trovpio" sterew'" [sterea'": Dain]
proshvlwntai kata; stoi'con trei'" o[nte", ejf w|n hJ keraiva katagomevnh ejpivkeitai.”.
285
See, for example, Casson, Ships and seamanship, p. 151, n. 45; Dolley,
“Warships”, p. 50; Pryor, “From dromo2n to galea”, p. 104.
250 CHAPTER FOUR
timber standing in the middle.”286

The word histodoke2 was well known from Homer’s Iliad as a mast
receiver or crutch into which a mast was lowered.287 One can be seen
clearly at the stern of a ship in a mosaic of the third century from
Roman Hadrumetum.288 [See Figure 24] We understand the
Anonymous to have meant that on dromons the heel (pterna) of each
mast (histos or katartion) was set in a mast step (trapeza) on the keel
and that the mast could be lowered onto the histodoke2. Used in this
sense, trapeza, a “table” in classical Greek, appears to have been a
post-classical word for a mast step, first recorded in a scholion on the
Iliad in a manuscript which probably dates from the last quarter of the
tenth century.289 The classical words for a mast step were mesovdmh
(mesodme2) or lhnov" (le2nos).290 Trapeza is not found in this sense in
Pollux, Hesychios, or the Souda and from where the Anonymous
derived it is unknown. Pterna, on the other hand, appears to have been
derived by him from Pollux, or perhaps from Athe2naios of
Naukratis.291 This passage is another give away that the Anonymous
------------------------------
286
See Dindorf, Scholia Graeca, B.427 (vol. 1, p. 117): “iJstivon to; a[rmenon, iJsto;"
de; to; katavrtion, wJ" to; mevson iJstavmenon xuvlon.”.
287
Homer, Iliad, I.434 (vol. 1, p. 34): “iJsto;n d iJstodovkh/ pevlasan protovnoisin
uJfevnte" ...”. Erbse, Scholia Graeca, A.434 (vol. 1, p. 122): “iJstodovkh/ to; kata; th;n
pruvmnan ejxevcon xuvlon, kaq ou| klivnetai oJ iJstov". ta; de; para; to; devcw pavnta dia; tou' k2,
xeinodovko", iJstodovkh.”. See also Casson, Ships and seamanship, pp. 47 n. 30, 329,
and plate 191. Cf. also Hesychios, Lexicon (Schmidt), I.1029 (vol. 2, p. 374):
“iJstodovkh: iJstoqhvkh. to; dia; mevsou ... new;" fravgma, eij" o} katakleiovmeno" oJ iJsto;"
ejntivqetai.” .Cf. Latte, ed., vol. 2, p. 378: “ iJstodovkh: iJstoqhvkh. to; dia; mevsou ‹th'"›
new;" fravgma, eij" o} kataklinovmeno" ‹oJ› iJsto;" ejntivqetai.”. Hesychios’ source was
probably Apollo2nios Sophista. See Apollo2nios Sophista, Lexicon, p. 93, ll. 4-5: “
iJstodovkh to; dia; mevsh" th'" new;" fravgma, eij" o} kataklinovmeno" tivqetai oJ iJstov".”.
288
Another mosaic of a galley with a similarly lowered mast from the baths at
Themetra, dated to ca 200-220 C.E., is reproduced in Foucher, Navires et barques,
fig. 12 (p. 21).
289
Commenting on Iliad, XV.729 (“qrh'nun ejf eJptapovdhn, livpe d i[kria nho;"
eji?sh"”), the scholia in the Venetus A manuscript (Venice, Biblioteca Marciana, MS.
Gr. 452 [col. 822]) has: “qrh'nun: uJpopovdion. qrh'nun de; bevltion kalei'sqai uJpo;
ÔOmhvrou ta;" kaqevdra" tw'n ejretw'n: ajnacwrw'n ga;r ajpo; th'" new;" tw'n katastrwmavtwn
ejpi; tauvta" ajf iknei'tai. tine;" de; tovpon th'" new;" bavsin e[conta, ejf ou| to;n kubernhvthn
tou;" povda" tiqevnai, o} kai; ejdwvl iovn fasin. a[lloi de; th;n uJpodecomevnhn to;n iJsto;n
travpezan ei\pon.”. See Bekker, Scholia, O.729 (vol. 1, p. 436). This passage is not
reproduced in full in Erbse, Scholia graeca, O.729 (vol. 4, p. 152). The phrase
referring to the trapeza was most probably a post-classical addition to the scholion.
290
In later times these terms were often misunderstood. See the scholion on the
Argonautika of Apollo2nios of Rhodes, commenting on Argonautika, I.563 (“dhv rJa
tovte mevgan iJsto;n ejnesthvsanto mesovdmh/, ...”), which glossed mesovdmh/ as: “hJ iJstoqhvkh,
o{pou tivqetai oJ iJsto;" kai; klivnetai.”, in Wendel, Scholia, p. 48.
291
See Athe2naios of Naukratis, Deipnosophistae (Gulick), XI.474.f (vol. 5, p. 96):
“tou' ga;r iJstou' to; me;n katwtavtw ptevrna kalei'tai, h} eJmpivptei eij" th;n lhnovn, ...”.
Athe2naios was quoting Askle2p iade2s of Myrlea (lst century B.C.E.). Pollux seems to
had little familiarity with real ships. A mast step was fixed on the keel
permanently, not only when under sail. The mast might be unstepped
from it, but mast steps were large and complex pieces of carpentry. 292
They could not be dismantled and were always left in place.
Keraia and keratarion, a diminutive of kevra" (keras), were also
familiar classical terms for the yard of a sail.293 The two terms for a
sail used by the Anonymous in apposition, histion and armenon, were
classical and post-classical terms respectively and were used similarly
in the scholia on the Odyssey and in the glosses.294
The Anonymous said quite clearly that the three kathormeis were
fixed firmly in a row on the keel and that the yard rested in these
when it was lowered. Obviously, something like crutches must have
been necessary to take lowered yards, just as histodokai were for the
masts. One of the bas reliefs on Trajan’s column clearly shows a
galley with its sail furled and yard lowered onto crescent-shaped
crutches at the bow and stern. [See Figure 3] The mosaic of the galley
from the baths at Themetra also shows the yard, with its furled sail,
lowered onto similar crutches. [See Figure 4] What these crutches for
the yards were called in classical Greek, whether they were different
to the histodokai and from where the Anonymous derived the term
kathormeis for them is unknown. The most likely probability is that
the word was derived from classical Greek kaqormivzw (kathormizo2),
------------------------------
have derived his text from either Athe2n aios or Askle2piade2s. Cf. Pollux, Onomasticon
(Bethe), I.91 (vol. 1, p. 29): “kai; to; me;n uJpodecovmenon to;n iJsto;n lhnov" ªkalei'taiº, to;
de; ejnarmozovmenon aujtw'/ ptervna, ...”. The Anonymous appears to have taken over the
sense of this passage, merely changing le2nos to trapeza.
Describing the parts of the mast and yard, beginning at the bottom, the scholia
on the Argonautika of Apollo2nios of Rhodes, I.564-7c also referred first to the pterna.
See Wendel, Scholia, p. 49: “iJstov": ptevr nh, karchvsion, qwravkion, ...”.
292
See, for example, Steffy, Wooden ship building, esp. pp. 41, 64, 73, 74, 225,
231; Santamaria, “L’épave Dramont”, pp. 161-70.
293
Casson, Ships and seamanship, pp. 47, 232. See also Appendix Two [a], §5 and
Appendix Five, §4. Cf. Dindorf, Scholia graeca, E.254 (vol. 1, p. 268): “th;n keraivan,
to; plavgion xuvlon tou' iJstou', w|/ prosdevdetai to; a[rmenon.”. See also the hermeneumata
attributed to Dositheus in the Hermeneumata Monacensia in Goetz, Glossarii Latini,
vol. 3, p. 205, l. 16: “ceras antemna”. See “Note on citations of Greek and Latin
glossaries”, p. lxix. above.
294
Casson, Ships and seamanship, p. 233. Cf. Dindorf, Scholia graeca, B.427 (vol.
1, p. 117): “iJstivon to; a[rmenon, ...”. See the Greek-Latin Cyril glosses of London,
British Library, MS. Harley 5792 in Goetz, Glossarii Latini, vol. 2, p. 333, l. 30: “
Istion uelum”; the Philoxenos Latin-Greek glosses of Paris, Bibliothèque Nationale,
MS. Lat. 7651, ibid., vol. 2, p. 205, l. 3: “Vela a2rmena : ai2tou : i>t ou : oqo2nai [recte,
“Vela a[rmena aiJ tou' iJstou' ojqovnai”]”; and the glosses attributed to Dositheus in the
tenth-century manuscript of the Hermeneumata Vaticana, Rome, Biblioteca
Apostolica Vaticana, MS. Vat. Lat. 6925, ibid., vol. 3, p. 434, l. 17: “armenaistia
uela [sic]”. See “Note on citations of Greek and Latin glossaries”, p. lxix. above.
252 CHAPTER FOUR
“to bring a ship into harbour”, as a consequence of the practice of

lowering the yards when entering harbour.
Elucidating from the manuscript that the Anonymous never wrote
that the kathormeis were fixed on a “stiffener keel” but rather on the
keel itself has simply made whatever he intended to say even more
incomprehensible. Even if sheer logic did not demand it, the pictorial
evidence suggests that both histodokai and yard crutches were set up
on the deck, not the keel. It is true that the pictorial evidence does not
preclude the possibility that the ships in question had no decks and
that the crutches may therefore have been mounted on the keel. But if
that were so, the crutches would have been set as low in the hull as
possible for reasons of stability and therefore would not have been
depicted well above the level of the gunwales as they are. The pictures
surely show deck-mounted crutches. The Anonymous clearly
differentiated the crutches for yards from the histodokai but whether
there was in fact any difference between them is unknown. Why could
not one set of crutches have been used for both purposes? Whatever
the case, neither histodokai nor kathormeis could have been fixed on
the keel unless their posts were made to pass up through the decks as
the masts were. That might possibly have been done for reasons of
structural integrity since the weight of the masts and yards which they
had to carry was very considerable. However, it is far more likely that
all that the Anonymous actually knew was that yards were lowered
when dromons were entering harbour under oars. Since the crutches
onto which they were lowered were obviously aligned down the
centre-line of the ships to maintain stability, to him they were surely
mounted on the keel?
(h) Rigging
Dromons had running rigging and various cables for specific

purposes, some of which were catalogued by Leo VI and the
Anonymous. Amongst the spare equipment that Leo VI recommended
that dromons should carry, kavrua (karya) were listed between oar-
grommets, scoiniva (schoinia), and sails, a[rmena (armena).295 The
meaning of karya is unclear since karyon and its variants had no
nautical association in classical Greek. But it is possible that scholia
------------------------------
295
Appendix Two [a], §5: “ Ecevtwsan de; kai; pavnta pro;" ejxartismo;n drovmwno"
ajparaleipta; kai; dipla', ... kavrua, ...”. Note that Nike2phoros Ouranos deleted karya
from his version of the same list. Cf. Appendix Five, §4.
on Lucian’s Zeus trago2dos contain the explanation of it. In glossing

ajkavtia (akatia), which in classical Greek as ajkavteio" (akateios)
meant the foremast of a ship,296 one scholion said that some people
called karya, “by which the yard is hauled up”, akatia.297 Karya and
akatia would thus appear to have become alternative names for some
things by which the yard was hauled up. These may possibly have
been halyards. However, there is no corroborating evidence that either
akation or karyon meant a halyard and Casson has established that the
classical Greek word for a halyard was a[gkoina (ankoina). More
probably the scholion was referring to the sheaves of the blocks at the
mastheads.298
The Anonymous listed amongst the running rigging and ship’s
cordage several items known from classical Greek: brails of sails,
kavloi (kaloi), forestays, provtonoi (protonoi), mooring lines, peivsmata
(peismata), offshore or bow mooring lines, ajpovgaia (apogaia), stern
mooring lines, prumnhvsia (prymne2sia), and, as seen above,299 e[mboloi
(emboloi) which he thought restrained the tillers, oi[ake" (oiakes), and
by which these were bound to trocanth're" (trochante2res). However,
once again, he took most of these terms from a list that he found in
Pollux,300 and it is clear that either he did not understand them or some
of the words had changed meaning. Brails were ropes which in
antiquity had been sewn into the feet of square sails, run through
fairleads on the front of the sail and then up and over the yards down
to the deck aft of the mast. By hauling them in, the sails could be
quickly shortened in heavy weather. But, it is impossible to brail
------------------------------
296
See Pollux, Onomasticon (Bethe), I.91 (vol. 1, p. 30): “kai; oJ me;n mevga" kai;
gnhvsio" iJsto;" ajkavteio", oJ de; katovpin ejpivdromo", oJ de; ejlavttwn dovlwn.”.
297
See Rabe, Scholia in Lucianum, Zeu;" tragw/dov".46 (p. 78): “eijsi; d oi} kai; ta; par
hJmi'n legovmena kavr ua ajkavtiav fasin, oi|" hJ keraiva ajnevlketai.”.
298
H. and R. Kahane suggested that karyon, meaning a “nut” in classical Greek,
acquired the meaning of a pulley by analogy during the Byzantine period and from
there gave rise to the medieval Latin and Italian car for the lower of the two spars of
which the yards of lateen-rigged ships were composed. The second part of this
hypothesis is to draw a very long etymological bow indeed, but the first part is
certainly possible. See Kahane, “Massaliotica”, p. 321 and n. 7; idem, “Eléments
byzantins”, p. 468; idem, Lingua Franca, §176 (p. 157).
Alexandres also suggests that karya were the sheaves of the blocks at the
masthead through which the halyards or ties were rove. See Alexandres, ÔH qalassiva,
p. 72. See also Casson, Ships and seamanship, pp. 230, 260-3.
299
See above pp. 134 n. 35, 224 & n. 183.
300
Appendix Three, §2.15: “Ta; de; th'" new;" skoiniva: kavloi, provtonoi, peivsmata,
ajpovgaia, prumnhvsia, kai; e[mboloi, oi} tou;" oi[aka" sunevcousi kai; di w|n eij" to;n
trocanth'ra ajpodesmw'ntai.”. Pollux, Onomasticon (Bethe), I.93 (vol. 1, p. 31): “...
iJstov", iJstodovkh, keraiva, scoiniva, kavloi, provtonoi, kalwv/dia, peivsmata, ajpovgua,
[ejpivgua], prumnhvsia: ...”.
254 CHAPTER FOUR
lateen sails in this way. Similarly, it is not possible to use forestays

with a lateen rig since they would prevent the yard from being hauled
across the front of the mast when tacking. How anyone familiar with
dromons sailing the Bosporos could have included brails and forestays
amongst their rigging defies the imagination. Against this it should be
acknowledged that words mutated in meaning. Kalos seems to have
become just another word for a rope. In his novel Leukippe2 and
Kleitopho2n, Achilles Tatios (fl. late second century C.E.), used kalos
for a cable used to tow a ship’s boat.301 Similarly, in one of the glosses
kalos seems to have become understood as just another word for a
rope of any kind and protonos as another word for bow mooring lines:
apogaia. With the passing of brails and forestays the words may well
have acquired new meanings by association to their old ones.302
(i) Crews
In the tenth century, the term used for both the crews of single
ships and also those of fleets was stratov" (stratos), lit. “people”, host,
band.303 In modern English, it may have all of the senses of “crew”,
“crews”, or “[men of the] fleet”.
For the standard bireme dromon, the starting point is the two
treatises of Leo VI and Nike2phoros Ouranos, which fundamentally
agree with each other.304 There were two oar-banks, ejl asivai (elasiai),
one below and one above deck. Used in this sense, elasia was not a
classical Greek term and appears to have been the tenth-century
equivalent of eijresiva (eiresia), which could have many meanings in
------------------------------
301
Achilles Tatios, Leukippe2 and Kleitopho2n, III.3.2 (p. 140) and cf. III.4.1-2 (p.
142).
302
See the Greek-Latin Cyril glosses of London, British Library, MS. Harley 5792,
in Goetz, Glossarii Latini, vol. 2, p. 337, l. 47: “Kaloshscoino" funis rudes [recte,
“kavlo" hJ scoi'no" funis rude[n]s”]; p. 235, l. 60: “Apogionscinion retinaculum [sic]”;
and p. 424, l. 14: “Protonoita apogiascoinia rudentes” [recte, “provtonoi ta; ajpovgeia
scoiniva rudentes”]. All this is to assume, of course, that the compiler of the glosses
actually knew what he was talking about. See “Note on citations of Greek and Latin
glossaries”, p. lxix above.
303
Appendix Two [a], §§24, 42, 75, 76; Appendix Five, §§16, 22, 40, 67, 68. See
also Ahrweiler, Byzance et la mer, under stratov".
304
Appendix Two [a], §§7-8: “”Ekasto" de; tw'n dromwvnwn eujmhvkh" e[stw kai;
suvmmetro" e[cwn ta;" legomevna" ejlasiva" duvo, thvn te kavt w kai; th;n a[nw. ÔEkavsth de;
ejcevt w zugou;" to; ejlavciston keV ejn oi|" oiJ kwphlavtai kaqesqhvsontai, wJ" ei\nai zugou;"
tou;" a{panta" kavt w me;n keV, a[nw de; oJmoivw" keV, oJmou' nV. Kaq e{na de; aujtw'n duvo
kaqezevsqwsan oiJ kwphlatou'nte", ei|" me;n dexiav, ei|" de; ajristerav, wJ" ei\nai tou;"
a{panta" kwphlavta" oJmou' tou;" aujtou;" kai; stratiwvta" touv" te a[nw kai; tou;" kavtw
a[ndra" rV.”. Cf. Appendix Five, §§6-7.
classical Greek, ranging from an oar, to oarsmen, to rowing benches,

to rowing, and to a bank of oars. The Anonymous used eiresia for an
oarage [system], a bank of oars, and an oar.305 Each elasia was
composed of 25 oarsmen per side, for a total of 100 oarsmen. Between
them both, Leo VI and Nike2phoros Ouranos made it clear that the
total number of oarsmen was 100 and that they also doubled as
soldiers, stratiw'tai (stratio2tai), or marines. Nike2phoros’s syntax was
somewhat clearer about this than that of Leo VI; although he certainly
understood his imperial predecessor correctly.306 In four cases in the
illustrations of the Madrid manuscript of the Synopsis historio2n of
John Skylitze2s some or all of the oarsmen are shown wearing armour;
however, there does not appear to be any particular reason why they
are in these cases but not in all the others.307
Beyond that, Leo VI and Nike2phoros following him, both said that
on larger dromons there could be 200 men, of which 50 would serve
the lower oar-bank and the other 150 should be stationed above deck
and should be able to fight.308 Exactly how such a distribution of
oarsmen and marines was arranged they did not specify.
The testimonies of Leo VI and Nike2phoros Ouranos have to be the
starting point; however, their specificity of the number of oarsmen at
25 per oar-bank per side, for a total for the two banks of 100 in all for
a standard dromon, need not be taken as more than an approximation.
The inventories for the Cretan campaign of 949 provide more precise
evidence and raise the vexed issue of what was meant by the arguable
term oujsiva (ousia) and its adjectival derivative ousiakos, literally
meaning substance, essence, or property, an issue which has
bedevilled scholarship on the Byzantine navy of the Macedonian era.
It used to be thought that an ousia or an ousiakos was an actual ship
type. But in fact, an ousia was not a type of ship but rather a standard
complement of 108 (or 110) oarsmen for a chelandion or dromon.
------------------------------
305
See Appendix Three, §§1.2, 2.7 & 13, 7.5.
306
Cf. also Appendix Two [a], §§14, 73; Appendix Five, §§12, 66. Leo VI’s syntax
was so unclear that Ibn Mankalı3’s translator did not understand him. See Appendix
Eight [a], pp. 242-3
307
See John Skylitze2s, Suvnoyi" iJstoriw'n, folios 31v, 38r, 38v, 39v [= Appendix
Seven, Table Two, nos 7, 12, 13, 15; EstopanNan, Skyllitzes Matritensis, figs 68, 86,
88, 90]. It is noticeable, although what significance may drawn from it is debatable,
that armoured oarsmen do not appear on any galley in those illustrations drawn in any
of the Western or Muslim styles.
308
Appendix Two [a], §9: “Kai; e{teroi de; drovmwne" kataskeuazevsqwsavn soi
touvtwn meivzone", ajpo; diakosivwn cwrou'nte" ajndrw'n h] plevon touvtwn h] e[latton kata; th;n
creivan th;n devo usan ejpi; kairou' kata; tw'n ejnantivwn: w|n oi{ me;n nV eij" th;n kavtw ejlasiva n
uJpourghvsousin, oiJ de; rV kai; nV a[nw eJstw'te" a{pante" e[noploi macevsontai toi'"
polemivoi".”. Cf. Appendix Five, §8.
256 CHAPTER FOUR
This understanding of ousia as a ship’s complement rather than a ship

type was a view first advanced by Jenkins in his translation of, and
commentary on, the De administrando imperio.309 Although his view
found acceptance for some time, Ahrweiler came to the conclusion
that the word ousia did eventually become applied to actual ships and
Alexandres, Treadgold, Eickhoff, and Hocker have also concluded
that ousiai or ousiakoi were actual ships. Haldon, notably, has not.310
Most recently, Makrypoulias has attempted to return to the pre-
Jenkins view that ousiai and ousiakoi were actual ships.311 However,
his view is based on certain passages in the inventories for the Cretan
expedition of 949 which he misconstrues. First, he says that: the term
ousia must mean a ship “Otherwise, we would be unable to explain
how a pamphylos (with a crew of 120 or 150 men) could be fitted with
twenty-four 108-men complements (2,592 men)”. But this conclusion
is based on a misinterpretation of some passages: “As defence for the
God-guarded city, one pamphylos and 24 ousiai” (Appendix Four [b],
§I.1); “For the defence of the City, the strate2goi of Aigaion Pelagos
were left with six chelandia pamphyla, each of 120 men ...”
(Appendix Four [b], §I.5); “The strate2gos of Samos with six
chelandia pamphyla, each of 150 men ...” (Appendix Four [b], §I.7);
and “The strate2gos of the Kibyrrhaio2tai with six chelandia pamphyla,
each of 150 men ...” (Appendix Four [b], §I.9). Obviously, chelandia
pamphyla could be crewed by either 120 or 150 (or 130 or 160) men.
But there are no grounds on which to extrapolate that the single
pamphylos mentioned in the first passage had 24 ousiai. Ships’
companies have always been used for many purposes. In this case, the
24 ousiai were obviously left behind as a marine guard for
Constantinople.
Secondly Makrypoulias refers cryptically to: “... why crews should
have travelled from Constantinople to Calabria or Spain without their
ships!!! (sic)”. This reference is to the following passages: “The
imperial fleet, 150 ousiai, of which 6 [were] hand-picked and 2
recently mobilized. 100 ousiaka chelandia, of which 100 ousiai, 7
ousiai in Dyrrachion and Dalmatia. 3 ousiai in Calabria, 3 ousiai with
the ostiarios and nipsistiarios Stephen for service in Spain”.
------------------------------
309
See Constantine VII, De administrando imperio, §51, ll. 41, 91 (pp. 248, 250)
and Jenkins’s commentary in Volume II: commentary, pp. 195, 198.
310
See Ahrweiler, Byzance et la mer, pp. 416-17; Alexandres, ÔH qalassiva, pp. 73-
4; Treadgold, “Army”, p. 134; Eickhoff, Seekrieg und Seepolitik, p. 137; Hocker,
“Galleys and fleets”, p. 94. See Haldon, “Theory and practice”.
311
See Makrypoulias, “Navy”, pp. 154-5.
(Appendix Four [b], §I.1). But in fact, all that this text meaned was
that the entire imperial navy was composed of 150 ousiai, of which
three were in Calabria and three were delegated to Stephen for the
embassy to the Umayyad court in al-Andalus.312 It says nothing about
the actual ships and there is no reason to ask the rhetorical question of
why crews should have travelled without their ships to Calabria and
al-Andalus (or Dyrrachion or Dalmatia). It is true that there are
references to actual ships (pamphyloi and ousiaka chelandia);
however, the entire text is problematical. The figures do not add up,
no matter how one interprets the technical terminology. The passages
following on from “The imperial fleet, 150 ousiai, ...” were not meant
to be inclusive. They were merely parenthetical references to some
important deployments.
Finally, Makrypoulias raises the question of why it is that there
should be a reference to “20 dromons, each with two ousiai” in the
actual fleet sent to Crete, whereas later “these 40 ousiai are termed
ousiaka chelandia”. He concludes that: “Whatever the meaning of the
term, it is certainly not ‘complement’”. But in fact the two texts he
refers to are firstly “20 dromons, each of two ousiai. 40 ousiai.”
(Appendix Four [b], §I.2) and secondly “80 sipho2nia for 40 ousiaka
[ships]” (Appendix Four [b], §V.13), not for 40 ousiaka chelandia.313
It is quite possible that chelandia were intended by the understood
[ships] after ousiaka as he assumes, but, in any case, this text merely
referred to the provision of 80 sipho2nia for 40 ousiaka [ships].
Whatever may have been intended by the 40 ousiaka [ships], it had no
reference to the 40 ousiai of the 20 dromons of the fleet. Elsewhere,
the inventory said clearly that the 20 dromons should have three
sipho2nia each.314
There are four particular reasons why the words ousia and ousiakos
cannot have referred to real ships. First, the De administrando imperio
said that from the time of Leo VI, an ousia was assigned to the
hippodrome to guard the palace when the regiment of the Arithmos,
the imperial guards, accompanied the emperor on expeditions.315 The
------------------------------
312
See above p. 71.
313
Appendix Four [b], §§I.2, V.27: “ta; mevllonta taxeideu'sai ejn Krhvth/ pavmfuloi
zV, oujsiaka; celavndia lgV, oJmou' celavndia mV. drovmone" kV ajna; oujsiw'n bV.”; “ta; mV
oujsiaka; sifwvnia pV,” [= Haldon, “Theory and practice”, pp. 219, 229; Constantine
VII, De cerimoniis, II.45 (vol. 1, pp. 664, 673)].
314
See Appendix Four [b], §IV.1 [= Haldon, “Theory and practice”, p. 227;
315
See Constantine VII, De administrando imperio, §51, ll. 40-41 (p. 248):
“Pollavki" ga;r ejxercomevnou aujtou' eij" ta; plhsivo n provkensa, th;n mivan oujsiva n
258 CHAPTER FOUR
concept of an ousia galley floating around the hippodrome is curious,

to say the least! The De cerimoniis also recorded that the oarsmen,
ejlavtai (elatai) of the imperial dromon, who received a donative at the
festival of the broumalia before Christmas, were its periousia.316 The
identification of oarsmen with ousia was quite clear. “Peri” was
simply a strengthening prefix here, without any other meaning.
Secondly, no one appears to have noticed the use of the verbal form of
ousia, oujsiwvsei" (ousio2seis) and oujsivwson (ousio2son), as used by Leo
VI and Nike2phoros Ouranos. Both the emperor and the magistros,
who followed him said that a strate2gos should “ousia” dromons to
match the crews (stratoi) of the enemy.317 They used the verb oujsiovw
(ousioo2) in the sense of crewing dromons with men. No other meaning
can possibly be ascribed to their use of the word. Thirdly, in a
paragraph referring to a reception in the Magnaura hall in the imperial
palace, the De cerimoniis reported that the ousia of the droungarios
tou ploimou and the pamphylon should be in attendance with their
weapons, and the commanders of the ships should also be there with
their ousiai. Ousiai can have had no other meaning here but crews.318
Finally, perhaps the most conclusive argument of all against the
interpretation of ousia as having ever referred to an actual ship type is
the fact that all known medieval Byzantine terms for ship types,
including dromo2n, chelandion, karabos, akatos, pamphylos, and
sage2na eventually found their way into either Latin or Arabic or both
as terms for ship types in those languages. Ousia did not, because it
referred not to a ship but rather to a ship’s complement.319
------------------------------
katelivmpanen eij" to;n iJppovdromon pro;" fuvlaxin tou' palativou ...”.
316
See Constantine VII, De cerimoniis, II.18 (p. 601).
317
See Appendix Two [a], §75: “Ea;n ga;r sunora'/" e[cein tou;" polemivo u" ploi'a
pleivona strato;n uJpodecovmena, oujsiwvsei" kai; aujto;" tou;" i[sou" drovmwna" ejn plhvqei.”.
Cf. Appendix Five, §68.
318
Constantine VII, De cerimoniis, II.15 (pp. 578-9): “... oiJ de; tou' trapezivou ta;
yeudoxeva kontomavnika. e[xwqen de; th'" calkh'" puvlh" tou' triklivnou tw'n kandidavtwn
e[sth e[nqen kajkei'qen hJ tou' drouggarivo u tw'n ploi?mwn oujsiva kai; oJ mevga" pavmfulo",
bastavzonte" dovrka" kai; ta; eJautw'n forou'nte" spaqiva. ejn de; th'/ prwvth scolh'/ kai; tw'/
triklivnw/ tw'n ejxkoubivt wn e{w" tou' aujtou' tribounalivo u e[sthsan e[nqen kajkei'qen aiJ
oujsivai tw'n pamfuvl wn, bastavzonte" dovrka" kai; ta; eJa utw'n forou'nte" spaqiva. oiJ de;
a[rconte" tw'n ploi?mwn e[sthsan kai; aujtoi; e[nqen kajkei'qen, e{kasto" eij" th;n ijdiva n
oujsiva n. ...”.
319
It is significant that although oujsiva appeared frequently in the Greek-Latin
glosses, it was always glossed in its meaning of essence, substance, etc. and never
with reference to ships. See Goetz, Glossarii Latini, vol. 7, p. 600 and the cross
references cited therein.
Alexandres has claimed that ousia was derived from the Venetian “huissier”, a
term which became widely used in many vernacular and medieval Latin forms in the
West in the late twelfth and thirteenth century, in particular for horse transport
The inventories for the 949 expedition recorded that the imperial
fleet was composed of 150 ousiai. Of these 6 were pamphyloi, hand
picked, and two had been recently mobilised. Then there were 100
ousiaka chelandia, that is chelandia of one ousia. Seven ousiai were
on duty in Dyrrachion and Dalmatia, three in Calabria, and three had
been sent to al-Andalus under the ostiarios and nipsistiarios Stephen.
One pamphylos and 24 ousiai were left to guard Constantinople.320
Those actually sent on the campaign to Crete amounted to 7
pamphyloi and 33 ousiaka chelandia, totalling 40 chelandia, and 20
dromons with two ousiai each, for a total of another 40 ousiai.321
Other ships and ousiai were either deputed to other tasks or left in
place as home guards. Of these the most important were the
following. The strate2goi of the thema of Aigaion Pelagos were left
with 6 chelandia pamphyla of 120 men each and 4 chelandia ousiaka
with 108 men each. The strate2gos of Samos was apparently left with 6
chelandia pamphyla of 150 men each and 6 chelandia ousiaka of 108
men. The pro2tospatharios John was sent to Africa with 3 chelandia
and with 4 dromons of 220 men each. The strate2gos of the
Kibyrrhaio2tai was also apparently left behind with 6 chelandia
pamphyla of 150 men and 6 chelandia ousiaka of 110 men. Two
pamphyloi and 4 [chelandia] ousiaka were left to guard the thema.
One ousia and also 4 dromons of 220 men each were left in Rhodes to
guard the imprisoned brother-in-law of the emperor, Stephen
Lekape2nos.322 Other ousiai were deputed to cut wood and there were
------------------------------
galleys. See Alexandres, ÔH qalassiva , p. 73. There are no sources, however, for the
use of any form of the term “huissier” or the medieval Latin “uscerius” in the West
earlier than the twelfth century and it is not possible that ousia was derived from it.
One might suggest the reverse of what Alexandres claimed, namely that “huissier”
and “uscerius” were derived from ousia. However, in fact the various forms of
“huissier” and “uscerius” were almost certainly derived from the Arabic ‘usha2rı3 for a
transport galley and had nothing to do with ousia. ‘Usha2rı3, which was an Arabic term
not derived from Greek, appeared in Egyptian sources as early as the ninth century.
See Fahmy, Muslim naval organisation, pp. 150-51.
320
Appendix Four [b], §I.1: “To; basiliko;n plovi>mon oujsivai rnV, ejx w|n pavmfuloi ıV
kai; oiJ ajrtivw" kataskeuasqevnte" bV. oujsiaka; celavndia rV. ejx aujt w'n rV oujsivwn [tw'n
Jrousivwn, Reiske] e[n te Durracivw/ kai; ejn Dalmativa/ oujsivai zV, ejn Kalabriva/ oujsivai gV,
meta; tou' ojstiarivo u Stefavnou kai; niyistiarivou eij" th;n jIspaniva n douliva oujsiva i gV. eij"
fuvlaxin th'" qeofulavktou povlew" pavmfulo" aV kai; oujsivai kdV.”. [= Haldon, “Theory
and practice”, p. 219; Constantine VII, De cerimoniis, II.45 (vol. 1, p. 664)].
321
Appendix Four [b], §I.2: “ta; mevllonta taxeideu'sai ejn Krhvth/ pavmfuloi zV,
oujsiaka; celavndia lgV, oJmou' celavndia mV. drovmone" kV ajna; oujsiw'n bV. oujsivai mV.”. [=
p. 664)].
322
Appendix Four [b], §§I.5-11: “ejavqhsan eij" fuvúlaxin th'" povlew" oiJ strathgoi;
‹tw'n ploi>moqemavtwn: oJ strathgo;"› tou' Aijgaivou pelavgou" meta; celandivwn pamfuvl wn "V
ajna; ajndrw'n rkV kai; celandivwn oujsiakw'n dV ajna; ajndrw'n rhV. ... oJ strathgo;" th'" Savmou
260 CHAPTER FOUR
also galeai involved, some of which were again left behind as home
guards.323
The two figures for crews of dromons, for those sent to Africa and
for those left in Rhodes to guard Stephen Lekape2nos, were the same.
Their complements were two ousiai or 220 men. One ousia therefore
equalled 110 men. This is as consistent as could reasonably be
expected with the figure of 108 men for the chelandia ousiaka or
chelandia of one ousia specified on all but one occasion and 110 men
on the other.324 Chelandia pamphyla could apparently have crews of
between 120 and 150 men. According to the inventory for the Cretan
expedition of 911 pamphyloi could have crews of either 130 or 160
men.325 Crews were surely tailored to meet specific needs from time to
time and may well have varied from thema to thema according to
local custom. Nor is there any reason to assume that all dromons, or
chelandia, or pamphyloi, were necessarily exactly the same size. As
Leo VI said, some dromons were larger and could carry larger crews.
Some variation in the crew figures perhaps reflected the size of the
actual ships of various classes available in various places at the time
or perhaps reflected the particular needs of individual expeditions.
Figures of between 100 and 108 or 110 for the oarsmen and 120-
160 for the total crews are consistent with Western evidence for those
of standard light galleys. In the early eleventh century, Thietmar of
Merseburg wrote that salandrie had two banks of oars per side and
total crews of 150 nautae (sailors).326 In the 1260s-1280s bireme
galleys of the Kingdom of Sicily still had 100-108 oarsmen but they
------------------------------
meta; celandivwn pamfuvl wn ıV ajna; ajndrw'n rnV kai; celandivwn oujsiakw'n ıV ajna; ajndrw'n
rhv. ajpestavlhsan de; meta; tou' prwtospaqarivou Iwavnnou kai; ajshkrhvth" ejn Afrikh'/
celavndia gV kai; drovmone" dV ajna; ajndrw'n skV. oJ strathgo;" tw'n Kiburraiwtw'n meta;
celandivwn pamfuvlwn ıV ajna; ajndrw'n rnV kai; celandivwn oujsiakw'n ıV ajna; ajndrw'n riV:
kateleivfqh de; kai; eij" fuvlaxin tou' qevmato" pavmfuloi bV, oujsiaka; dV. ... kateleivfqh de;
kai; eij" fuvlaxin tou' kurou' Stefavnou tou' gunaikadelfou' tou' basilevw" ejn ÔRovdw/ oujsiva
aV kai; dromovnwn dV ajna; ajndrw'n skV.”. [= Haldon, “Theory and practice”, p. 219;
323
Appendix Four [b], §§I.6, 10, 12-14: “[from the thema of Aigaion Pelagos]
kateleivfqh de; kai; miva oujsiva eij" to; kovyai th;n th'" ojgdovh" ijndiktivo no" xulhvn. ... [from
the Kibyrrhaio2tai] kateleivfqh de; kai; eij" to; kovyai th;n th'" ojgdovh" ijndiktivono" xulh;n
oujsivai bV. ... galevai th'" Attaliva" ieV. ejx aujt w'n kateleivfqh eij" fuvlaxin tou' qevmato"
galevai ıV. galeva i th'" Antioceiva" bV. kateleivfqhsan kai; au|tai eij" fuvlaxin tou' aujtou'
qevmato". galevai th'" karpavqou. kateleivfqhsan eij" fuvlaxin th'" nhvsou Karpavqou
galeva aV.”. [= Haldon, “Theory and practice”, pp. 219, 221; Constantine VII, De
cerimoniis, II.45 (vol. 1, p. 665)].
324
See Appendix Four, n. 11.
325
Appendix Four [a], §§2-5 [= Haldon, “Theory and practice”, pp. 203, 205;
326
See p. 190, n. 70 above.
also carried officers, helmsmen, soldiers, and others, to a total figure

of around 150 men.327 The normal number of oarsmen for
complements of standard light galleys seems to have remained at 50-
54 per side in two banks or files of 25-27 benches from the tenth to
the thirteenth centuries, irrespective of how the benches were
arranged. Total complements obviously varied according to size and
mission, but an average of around 140-150 was the norm.
This being the case, how are the very large figures for some crews
to be explained? Both Leo VI and Nike2phoros Ouranos directed that
some larger dromons should be constructed on which 50 men should
serve the oars below deck and 150, all of whom were armed, should
be stationed above deck to fight. But the 150 above deck were not said
to have served two or three files of oars, merely to have been armed to
fight.328 In the inventory for the 911 expedition to Crete, the crews
specified for the dromons of the imperial fleet and also for those of
some provincial themata had 230 oarsmen, a[ndre" kwphlavtai (andres
ko2pe2latai) and 70 marines, polemistaiv (polemistai),329 and this
specification was repeated in a slightly altered form in the paragraph
concerning the arming of a dromon in the inventory for the 949
expedition,330 qualifying the specifications earlier in the inventory that
each dromon should have two ousiai or 220 men.331 It is important to
appreciate here that 230 men of the ships for the expedition of 949
were described as “oarsmen, that is soldiers” (ploimoi ko2pe2latai e2toi
kai polemistai).332 They could double in both roles.
On the one hand, the figures ranging between 108 and 160 men or
oarsmen, the usage is variable, are not a problem. Dromons with an
ousia of 108 men, of whom 100 pulled at the oars at any one time in
------------------------------
327
See Pryor, “Galleys of Charles I of Anjou”, pp. 81-4 and Table Three. Eickhoff
also realized that 25 or so oar benches per side remained the norm for all
Mediterranean light galleys through to the eighteenth century. See his Seekrieg und
Seepolitik, p. 137, n. 8.
328
Appendix Two, §9; Appendix Five, §8.
329
Appendix Four [a], §2: “Drovmwne" xV e[conte" ajna; ajndrw'n kwphlatw'n slV kai;
ajna; polemistw'n oV:... ” and cf. §§3-6 [= Haldon, “Theory and practice”, pp. 203-5:
330
Appendix Four [b], §II.22: “oJ drovmwn ojf eivlei e[cein a[ndra" tV, oiJ me;n slV plovi>moi
kwphlavtai h[toi kai; polemistaiv, kai; oiJ e{t eroi oV a[ndre" polemistai; ajpo; tw'n
kaballarikw'n qemavtwn kai; ajpo; tw'n ejqnikw'n.” [= Haldon, “Theory and practice”, p.
225; Constantine VII, De cerimoniis, II.45 (vol. 1, p. 670)].
331
Appendix Four [b], §§I.2, 11 [= Haldon, “Theory and practice”, p. 219;
332
h[toi, which may have a range of meanings in the sense of “either ... or ...”, or
“both ... and ...”, etc., in this context is used to introduce an explanation, having the
sense of “that is”.
262 CHAPTER FOUR
two files per side on two banks, could never have been more than a
norm and variations in crews of up to around 40-50 more could
obviously have been accomodated within the tolerances of oarage
systems or by ships of somewhat larger dimensions.
On the other hand, there are only two explanations for the very
large figures of 220 or 230 oarsmen for some dromons. If they could
all row together, these must have been quadriremes larger than the
norm. But if quadrireme dromons did exist, it is inconceivable that the
many contemporary sources, which were normally addicted to the
spectacular, would not have mentioned them. Moreover, although
oarsmen could double as marines, only those serving the oars above
deck were armed and there is no evidence that any Byzantine dromons
had four files of oars. In fact, that all 230 oarsmen could row on oars
at the same time is impossible and the evidence of the inventories for
the Cretan expedition of 949 that the same 20 dromons which carried
two ousiai had only 120 oars is conclusive evidence that they could
not all row at the one time.333 Many crew must have been taken aboard
at various times as supernumeraries who could be used either as
oarsmen in watches, so that fleets could continue under way under
oars around the clock if necessary, or as marines and landing forces. It
should be borne in mind that the Cretan expeditions were assaults
against an island held by a formidable enemy with a long history of
naval prowess. It would not be surprising if the dromons were packed
to their gunwales with supernumeraries who could both participate in
the assault on the island and also fight if the enemy engaged at sea.
Doubling crews by taking aboard supernumerary oarsmen or
marines would have created significant problems. Galleys such as
dromons were finely tuned pieces of machinery with oarage systems
which had evolved to deliver maximum performance. Upsetting the
oarage balances beyond allowable tolerances would have affected
their performance capabilities badly and, in extreme cases made them
unworkable. We demonstrate below that a dromon with a standard
complement of one ousia of 108-110 men plus a normal complement
of officers, soldiers, sailors, etc., would have been designed to have a
freeboard at the lower oarport above the plane water line of around
0.36 metres amidships. To double the ousia with another 110 lean but
muscly men weighing around 85 kilogrammes each would add
another 9.35 tonnes in weight. The plane area at the waterline was only
------------------------------
333
Constantine VII, De cerimoniis, II.45 (vol. 1, pp. 672)] and cf. below pp. 300-304.
Figure 25
Section of Figure 32 to demonstrate the effect of overloading by ca 9.35
tonnes.
© John H. Pryor
approximately 95 square metres and another 9.35 tonnes would sink

the ship by almost 10 centimetres and impede the working of the
oarage system badly. Ships such as these had evolved ergonomically
to work in the most efficient way. The angle to the horizontal of the
oars could not be changed significantly without making the entire
stroke inefficient and the recovery in particular extremely difficult. Of
course there would have been some tolerances, but how great would
they have been? After all the considerations worked through below,
the oarage system which we have come to consider most probable is
as shown in Figure 32. If the waterline in this figure is sunk by 10
centimetres, the lower oarsmen would have to lower their arms by five
centimetres during the pull and would be unable to raise the bottoms
of their blades more than around 70 centimetres out of the water
during the return, making them unworkable in more than the light
breezes of Beaufort Scale Three, 7-10 knots, which raise wavelets up
to 60 centimetres, which do not even break.334 The conclusion is that
------------------------------
334
Haldon, “Theory and practice”, p. 254, has also addressed the sinking effect
caused by taking on extra crews and/or supernumeraries, although not from the point
of view of the funcionality of the oarage systems. John Coates, the designer of
Olympias, informed John Pryor that sinking the ship in the water by an additional ten
centimetres more than the level at which it was designed to float would have thrown
264 CHAPTER FOUR
if dromons took aboard a second ousia, let alone the 230 oarsmen and
70 soldiers of the Cretan expeditions, they would have had to have
been stripped of provisions, water, spare gear, or armaments in order
to compensate for the extra weight. Alternatively, the lower oar-bank
would have had to have been shut down and the oar ports sealed. But
if that was done the ships would have been dangerously low in the
water and vulnerable to any sort of a sea at all.
The inventory for the expedition of 949 included a “portulan”, a
stadiodromikon, which, if we can believe it, gave the distances from
Constantinople to Crete, specifying fourteen traverses en route.335
Purposely excluding here discussion of the vexed issue of how the
stadiodromikon was compiled and how tenth-century Byzantines
could have measured spatially across open water when they had no
technology capable of dead-reckoning distances at sea, no traverse
was given at more than 100 Byzantine milia, about 85 English miles,
and such short traverses may well have reflected stripping the ships
bare in order to accommodate supernumerary crews. These were all
traverses which the fleet would make before the prevailing north to
north-easterly winds of summer and at the average speed of around
two knots maintained around the clock which medieval galley fleets
were capable of in all conditions,336 none should have taken more than
two days. As long as the increased weight of the crews was
compensated for somehow, the ships could have carried far larger
crews than they normally would have on extended cruises.
------------------------------
its oarage system into chaos.
335
Haldon, “Theory and practice”, p. 235; Constantine VII, De cerimoniis, II.45
(vol. 1, p. 678). On the stadiodromikon see Pryor, “Stadiodromikovn”; Huxley,
“Porphyrogenitan Portulan”. Huxley reached the same conclusions as those here,
namely that the distances estimated in the Stadiodromikon are consistently too high,
and also that the compiler of it was a “bureaucratic landlubber” rather than a
“practical mariner”, which was no doubt true. See also Christides, Conquest of Crete,
Appendix D: “Nicephoros Phocas’ sailing venture to Crete (960). The
stadiodromikon” (pp. 221-4); Haldon, “Theory and practice”, pp. 301-2.
In order the places mentioned were: He2rakleia, Proikonne2sos, Abydos, Ta
Peukia, Tenedos, Mityle2ne2, Chios, Samos, Phournoi, Naxos, Ios, The2ra-The2rasia, Ta
Christiana, Dia, Crete. As stop-overs, some of these would have been unnecessary,
for example Ta Peukia and the Phournoi islets, and others completely useless. If Ta
Peukia can identified with Pefkhia near Ophryneion, then it was only around 19
kilometres from Abydos and Tenedos a further 29 kilometres on. Tenedos itself is only
around 48 kilometres from Abydos.
The longest traverses were from Proikonne2sos to Abydos, Tenedos to Mytile2n e2,
Mytile2ne2 to Chios, and Chios to Samos, all of which were reckonned at 100 milia,
which should be around 85 English miles. The distances are in fact approximately 70
English miles in each case.
336
See below pp. 338-53 and Table 7; Pryor, Geography, technology, and war, pp.
71-5.
This is to assume that the traverses were from mooring to mooring

not merely from landmark to landmark. In support of the first
interpretation is the fact that He2rakleia, Proikonne2sos, Abydos,
Tenedos, Mityle2ne2, Chios, and Naxos were indeed moorings for
Byzantine squadrons and would have been logical stop-overs.
However, Ta Peukia/Pefkhia, the Phournoi islets, The2ra-The2rasia, Ta
Christiana, Dia, and even Samos had no anchorages for large fleets.
Moreover, one major purpose for making stopovers would have been
to take on water but the Phournoi islets, The2ra, Ta Christiana, and Dia
had little or no water. Ta Christiana is an isolated, waterless,
uninhabited islet. The Phournoi islets were uninhabited.337
In fact Byzantine fleets operating against Crete did not traditionally
sail direct from Constantinople with all forces aboard but rather made
a rendezvous with cavalry and other forces on the south-west coast of
Asia Minor, leaving only a short passage to Crete. Almost a century
earlier the Caesar Bardas had brought the Byzantine forces for his
own assault on Crete overland to Ke2poi at the mouth of the Maeander
river, some 40 miles south of Phygela, to rendezvous there with the
fleet.338 Then, one of the inventories for the expedition of 911
mentioned that nails for gangways and mangers should be sent to
Phygela, obviously for fitting out the cavalry transports. The final
assault on Crete by Nike2phoros Pho2kas in 960 was also launched from
Phygela.339 It is not known from where the 949 expedition was
launched but all the evidence suggests that it was from Phygela or
somewhere in its proximity. The Stadiodromikon, therefore, was
nothing but some landmarks for the fleet, if anything at all. But the
point is that Phygela or other aple2kta in its vicinity were the normal
staging posts for assaults on Crete and from there to Crete was a mere
350 or so kilometres via Samos, Naxos, Ios, and The2ra. The longest
passages without being able to take on water would be Samos to
------------------------------
337
Haldon, “Theory and practice”, p. 301; Denham, Aegean, p. 255; Malamut, Iles
de l’Empire byzantin, vol. 1, p. 42, vol. 2, p. 544.
338
Genesios, Basileiai, D.20-23 (pp. 73-6); John Skylitze2s, Synopsis historio2n ,
Basivleio" oJ Makedw'n.12 (p. 128); Theophane2s continuatus, V.17 (pp. 235-8).
339
Appendix Four [a], §16: “peri; tou' eJtoimasqh'nai karfivo n pentadaktulai'on lovgw/
th'" strwvsew" tw'n dromonivwn, eij" ta;" skavla" kai; eij" ta;" pavqna" ciliavda" lV, kai;
katevlqwsin eij" ta; Fuvgela.”. [= Haldon, “Theory and practice”, p. 211; Constantine
VII, De cerimoniis, II.44 (vol. 1, p. 658)]. See also below p. 306. See also Leo the
Deacon, Historiae, I.gV-qV, II.ıV-hV (pp. 7-16, 24-29); John Skylitze2s, Synopsis
historio2n, Rwmano;" oJ Nevo".4 (pp. 249-50); John Zo2naras, Epitome2 historio2n, XVI.23
(vol. 4, pp. 72-3); Pseudo Symeon magistros, Chronographia, pp. 758-60;
Theophane2s continuatus, VI.Basileiva Rwmanou' uJiou' Konstantivnou tou'
Porfurogenhvtou.7-12 (pp. 473-8).
266 CHAPTER FOUR
Naxos and Ios to Chandax, both of around 125 kilometres. For such
short distances water supplies could be cut to a minimum to
compensate for the weight of supernumerary crews.
The command structure of Byzantine fleets and of dromons in the
tenth century is as problematical as the question of their crews. In
what follows, we emphasize that we are concerned only with the
operation of fleets at sea. Many of the command structures of the
Empire amalgamated both administrative institutions on the one hand,
and military and naval functions on the other. While we are perfectly
conscious of the overlap between all of these, we have confined our
analysis to what can be discerned of the operational command of the
various fleets.
About the only thing upon which Leo VI, Nike2phoros Ouranos, and
the Anonymous were agreed in this respect, was that the general term
for “admirals” of fleets was strate2gos, the same term as was used for
“generals” of armies and for governors of themata.340 This term,
however, could have both a general meaning as well as one specific to
a rank or title. On the one hand, in the time of Leo VI, the “admiral”
of the imperial fleet, basiliko;n plwvi>mon (basilikon plo2imon), based at
Constantinople bore the title of droungarios of the ship(s),
droungarios tou ploimou or to2n ploimo2n.341 The title still existed at the
time of the Cretan expedition of 949 and was even used under Alexios
I Komne2nos for the commander of the emperor’s personal squadron at
Constantinople; however, the position declined in importance from
the late tenth century.342 On the other hand, the title of the “admirals”
of the fleets of the three great naval themata of the tenth century,
Aigaion Pelagos, Samos, and the Kibyrrhaio2tai, seems to have
remained strate2gos throughout the century.343 Their “admirals” were
no doubt the same men who held the governorships of the themata as
their strate2goi.
------------------------------
340
Appendix Two [a], §2 ff.; Appendix Three, Pref.4 and §2.5; Appendix Five, §1
ff. See also Oikonomides, Listes de préséance, pp. 341-6.
341
Appendix Two [a], §27; Constantine VII, De administrando imperio, §51 (pp.
246-57). See also Oikonomides, Listes de préséance, p. 340; Guilland, “Drongaire”,
pp. 535-42.
342
See Haldon, “Theory and practice”, p. 231: “ejdovqh uJp e;r ajgora'" calkwvmato"
diafovrou tw'/ doqevnti lovgw/ th'" uJpourgiva" tou' drouggarivo u tou' ploi?mou ...”. Cf.
Constantine VII, De cerimoniis, II.45 (vol. 1, p. 676) and cf. Appendix Four [b],
§VII.rubric. See also Ahrweiler, Byzance et la mer, esp. pp. 118 ff., 209-10.
343
Appendix Two [a], §27; Appendix Three, §4.2; Appendix Five, §25. See
Appendix Four [a], §1: “ejdeJxato oJ strathgo;" tw'n Kiburraiwtw'n e[cein strato;n ÀecV,
...”; [b] §I.9: “oJ strathgo;" tw'n Kiburraiwtw'n meta; celandivwn pamfuvlwn ıV …”. [=
Haldon, “Theory and practice”, pp. 203, 219; Constantine VII, De cerimoniis, II.44
(vol. 1, p. 651), II.45 (vol. 1, p. 665)].
Under the strate2goi of the naval themata served other droungarioi

and also tourmavrcai (tourmarchai). In the tenth century, a
tourmarche2s was a governor of a tourma, one of the two or three
principal subdivisions of a thema, and the droungarioi seem to have
become governors of regional subdivisions of tourmai.344 In a much
discussed passage whose syntax is very obscure but whose meaning
was elucidated by Nike2phoros Ouranos, Leo VI wrote that in the past
the governors of the maritime themata and thus the “admirals” of their
fleets had borne the title of droungarios but that in his own day the
title of droungarios had been elevated to that of strate2gos and that
tourmarchai and droungarioi now served under the strate2goi.345 To
what the emperor appears to have been referring is that originally,
when the navy was organized as the fleet of the Karabisianoi, the
droungarioi were subject to strate2goi.346 However, the term
Karabisianoi is last recorded in 711 and the organization of the navy
was changed by the erection of the Kibyrrhaio2tai into a naval thema,
to which were also added later the themata of Samos and Aigaion
Pelagos, under droungarioi. But from the middle of the ninth century,
these were also elevated to the rank of strate2goi. A strate2gos of the
Kibyrrhaio2tai is first mentioned in 730; although Samos and Aigaion
Pelagos did not have strate2goi until much later. The Anonymous also
said that in the dromons of the themata, tourmarchai and droungarioi
served under the strate2goi, almost certainly having taken this from
Leo VI.347 These positions of “vice-admiral” and “rear-admiral”
certainly still existed in the mid century since after the arrival of the
fleet of 949 in Crete the tourmarchai and droungarioi were paid 30
nomismata and 20 nomismata respectively.348 This fleet was divided
into four themata or squadrons, most probably for logistical reasons,
and tourmarchai or droungarioi may have been in command of the
------------------------------
344
See Oikonomides, Listes de préséance, p. 341.
345
Appendix Two [a], §§26-7: “[ejpi; de; tw'n qematikw'n dromwvnwn] kai; drouggavrioi
ejpisthvsontai kai; tourmavrcai, kai; aujtoi; tw'/ strathgw'/ uJpotaghvsontai kai; toi'" ejkeivnou
paraggevlmasin uJpakouvsousin. Oujk ajgnow' de; o{ti kata; th;n oJmoivwsin tou' basilikou'
plwi?mou kai; oiJ tw'n a[llwn qemavtwn plwvi?moi strathgoi; drouggavrioi ejkalou'ntov pote
toi'" prwvh n crovnoi" kai; oiJ uJp aujtou;" kovmhte" movnon kai; kevntarcoi: ajlla; nu'n eij"
strathgivda hJ eJkavstou tw'n drouggarivwn ajrch; ajnabevbhken kai; ou{t w kaloumevnh tai'"
strathgikai'" katamerivzetai tavx esin.”. Cf. Appendix Five, §§24-5.
346
See also Antoniadis-Bibicou, “Thème des Caravisiens”, esp. pp. 80-86.
347
Appendix Three, §4.2: “Epi; de; tw'n qematikw'n dromwvnwn drouggavrioi kai;
tourmavrcai uJpo; th;n tou' strathgou' cei'ra kai; aujtoi; telou'nte".”.
348
See Haldon, “Theory and practice”, p. 215: “ijstevo n, o{ti ta; qematika; plovi>ma ejn
th'/ kata; Krhvth" ajfivxei ejrogeuvqhsan ou{tw": oiJ tourmavrcai ajna;: lV, ... oiJ drouggavrioi
ajna: kV, ...” [expedition of 949, the text is misplaced in the manuscript]; Constantine
VII, De cerimoniis, II.44 (vol. 1, p. 662).
268 CHAPTER FOUR
squadrons.349
The commanders of individual ships were known as kevntarcoi
(kentarchoi). That much at least is clear from Leo VI and Nike2phoros
Ouranos.350 In another piece of his familiar classicizing affectation
derived from Pollux, in which he also attempted to show off some
knowledge of Latin, the Anonymous provided a false etymology for
kentarchos from the Latin centum, for 100, and then concluded that he
commanded a hundred men, while at the same time using the classical
terms trihvrarco" (trie2rarchos) and eJkatovntarch" (hekatontarche2s)
for the commander of a trie2re2s and a ship of 100 oarsmen
respectively.351
In the command chains of fleets, there were squadron commanders
between the “admirals” or fleet commanders and individual ship
commanders. Leo VI, and Nike2phoros Ouranos and the Anonymous
both following him, all wrote that they should be in command of
either three or five dromons, employing the ubiquitous word kovmh"
(kome2s) for the rank.352 Leo VI explained what he meant by the term
in this context by reference to the classical term for an admiral or fleet
commander, nauvarco" (navarchos), a term which was no longer used
as a rank or title by the tenth century,353 and to the non-technical term
for a leader or “officer”, hJgemwvn (he2g emo2n). Elsewhere, the emperor
used navarchos in the sense of a commander subordinate to a
strate2gos.354 Significantly, Nike2phoros Ouranos deleted the reference
to navarchos and changed he2gemo2n to an even less specific term for a
“leader”, ajrchgov" (arche2gos). Kome2tes were enumerated among the
------------------------------
349
See Haldon, “Theory and practice”, p. 223: “Dia; tw'n tessavrwn qemavtwn tou'
basilikou' ploi?mou, ...”; Constantine VII, De cerimoniis, II.45 (vol. 1, p. 667).
350
Appendix Two [a], §8: ““Exw de; touvtwn to;n kevntarcon tou' drovmwno" kai; ...”. Cf.
§14 and Appendix Five, §§7, 12. See also Oikonomides, Listes de préséance, pp. 340-
41.
351
Appendix Three, §§4.2-3: “ÔEkatontavrch" oJ ejpi; mia'" nho;" eJkato;n ajndrw'n
hJgouvmeno" o{sti" kai; trihvrarco" kevklhtai. “Esti de; oJ legovmeno" kevntarco": kevntoum
ga;r para; ÔRwmaivo i" oJ eJkato;n ajriqmo;" proshgovr eutai kai; kevntarco" oJ eJkato;n ajndrw'n
hJgouvmeno". ... Ef eJkavsth" de; tw'n new'n a[rcontev" eijsi;n oi{de: trihvrarco" kai; ...”.
Such classicizing terms were, of course, commonly used elsewhere for officers
of armies. Nike2phoros Pho2kas used dekavrch" (dekarche2s), penthkontavrco"
(pente2kontarchos), and eJkatontavrco" (hekakontarchos). See Nike2phoros Pho2kas,
Praecepta militaria, I.1, in McGeer, Dragon’s teeth, p. 12 et passim.
352
Appendix Two [a], §25: “Oujc wJ" e[tucen aJpavntwn tw'n dromwvnwn poreuomevnwn,
ajll ejpisthvsei" aujtoi'" a[rconta" h] kata; pevnte h] kata; trei'" drovmwna", e{na to;n
legovmenon kovmhta, o{sti" nauvarcov" te kai; hJgemw;n tw'n uJp aujt w'n dromwvnwn uJpavrcwn
frontivsei prosecevsteron peri; pavntwn eujkovlw" kai; diatavxei pro;" e{kasta.”. Cf.
Appendix Three, §4.1; Appendix Five, §23.
353
It is not mentioned in Oikonomides, Listes de préséance.
354
Appendix Two [b], §§3, 4.
officers of the fleet paid after the landing in Crete in 949. They were
paid only 6 nomismata each, as compared to the 20 nomismata for the
droungarioi and the 30 for the tourmarchai;355 however, since the
kentarchoi were not mentioned in this list of payments, it is not
possible to deduce from it the position of the kome2tes in the chain of
command.
Elsewhere Leo VI wrote that the berth or krabatos of the
navarchos, that is the kentarchos, should be at the poop. Nike2phoros
Ouranos deleted the parenthetical reference to a navarchos but the
Anonymous said that the berth might be for either the trie2rarchos or a
strate2gos.356 Did this reflect the operational command structure of
Byzantine fleets? When a fleet or squadron commander sailed on a
particular dromon, did he take over the operational command of that
dromon from its kentarchos? Or was it the case that, as was the
practice in navies of later centuries, when a fleet or squadron
commander “hoisted his flag” in a particular ship, he had the overall
command of the fleet but the operation of the ship from which he
commanded was left to its own commander. Surely Byzantine fleet
and squadron commanders would have had so many other problems to
consider, and tasks to fulfill, that the operational command of the
ships on which they sailed would best be left to their kenatarchoi. The
text of the Anonymous does not allow resolution of whether the
krabatos was for the kentarchos, who might be a strate2gos if he held
that rank, or whether it was for both the kentarchos and also a
strate2gos if one was aboard ship. However, the fact that Nike2phoros
Ouranos deleted Leo VI’s parenthetical reference to the krabatos
being for a navarchos, that is a kome2s or squadron commander, and
did not replace it with kome2s, a word which he did take over
elsewhere, suggests that kentarchoi did remain in operational
command of their own ships even when superior officers were aboard.
The command structure of individual ships was headed by their
kentarchoi or “captains”. We can dismiss most of the evidence of the
Anonymous for this structure because he simply lifted it from what he
could understand of Pollux.357 In all probability, the most senior
“officers”, or perhaps rather “petty officers”, of Byzantine warships
------------------------------
355
See Haldon, “Theory and practice”, pp. 215, 217: “ijstevo n, o{ti ta; qematika;
plovi>ma ejn th'/ kata; Krhvth" ajfivxei ejrogeuvqhsan ou{tw": ... oiJ kovmhte" ajna;: ıV ...”
[expedition of 949, the text is misplaced]; Constantine VII, De cerimoniis, II.44 (vol.
1, p. 662).
356
Appendix Two [a], §8; Appendix Three, §2.5; Appendix Five, §7. See also
above pp. 215-17.
357
See Appendix Three, §4.3 and n. 70.
270 CHAPTER FOUR
below the kentarchos were those officers called in the tenth century
prwtokavraboi (pro2tokaraboi), a word which we should understand as
something like “first ship [man]”, or “first mate”, since it appears to
have been used for men who had risen from the ranks. Karabo2s and its
diminutive karabion were post-classical words usually used with the
sense of a ship’s boat; although, in Greek papyri from Muslim Egypt
and in some Byzantine texts they appeared as a term for a warship.358
Leo VI equated pro2tokaraboi to the classical word for helmsmen,
kubernh'tai (kyberne2tai). Nike2phoros Ouranos followed him but
deleted the equation with kyberne2tai. The Anonymous preserved the
meaning of kyberne2tai; however, he simply got it from Pollux.
Elsewhere he equated a navarchos with a pro2tokarabos, but his
equation can be dismissed since it is quite clear that a pro2tokarabos
was subordinate to a kentarchos whereas a navarchos, whatever may
have been intended by the term, was clearly superior to one.359 By the
tenth century the classical term for a helmsman, kyberne2te2s, was no
longer in vernacular usage and pro2tokarabos was used instead.
In a very curious chapter of the De administrando imperio which
gave an account of the development of the personal flotillas of the
emperor and empress, it was recorded that two men named Podaro2n
and Leo the Armenian, who had been the “first oarsmen”, prwtelavtai
(pro2telatai), of Nasar, the patrikios and droungarios tou ploimou,
were promoted to become the first oarsmen of the imperial crimson
barge during the reign of Basil I.360 Pro2telatai were almost certainly
the stroke oarsmen.361 Then, when Leo VI constructed two imperial
dromo2nia to use in imperial progresses, he promoted these men to
become the pro2tokaraboi of the dromo2nia.362 That pro2tokarabos
meant “helmsman” here was confirmed later on in the chapter where
it was said that when the pro2tokaraboi of the first imperial dromo2nion
------------------------------
358
See above pp. 164-5, 188-9. See also Appendix Four [b], §§VI.1-2, 7, 11, 13-14
[= Haldon, “Theory and practice”, pp. 229, 231; Constantine VII, De cerimoniis, II.45
(vol. 1, p. 674)].
359
See Appendix Two [a], §8; Appendix Three, §§2.6, 4.3; Appendix Five, §7. See
Ahrweiler, Byzance et la mer, under prwtokavrabo".
360
Constantine VII, De administrando imperio, §51, ll. 73-7 (pp. 248-50).
Nasar was droungarios tou ploimou towards the end of the reign of Basil I.
Placed in command of the entire navy of the Empire, he won a notable victory over
the Muslims off Punta di Stilo on the south coast of Italy in 880 and from the imperial
fleet Podaro2n and Leo the Armenian were promoted to be chief oarsmen of the
crimson imperial barge, the rousion agrarion. Cf. above pp. 65-6.
361
The word survived in medieval Latin with this meaning as protelati and
portoladi. See Pryor, “Galleys of Charles I of Anjou”, p. 81.
362
Constantine VII, De administrando imperio, §51, ll. 77-80 (p. 250). On the
construction of the two imperial dromo2nia, see above p. 164, n. 7.
were seconded to the fleet for a naval expedition under the patrikios
Eustathios, a man named “old Michael”, who had previously been a
pro2telate2s, “steered”, ejkubevrna (ekyberna), the dromo2nion.363 “Old
Michael” was later made one of the pro2tokaraboi of the emperor’s
dromo2nion when Podaro2n and Leo the Armenian were promoted to be
topothrhtaiv (topote2re2tai), of the imperial ships.364 Later he was
promoted to pro2tospatharios te2s phiale2s but still used to sail on the
emperor’s dromo2nion and would instruct his successor pro2tokaraboi
on how to “manage the quarter rudders and steer”.365 That the
functions of pro2tokaraboi were those of helmsmen was clear.
According to Liudprand of Cremona, the future emperor Ro2manos
Lekape2nos’s first major promotion to prominence in the fleet was to
pro2tokarabos, even though Liudprand misunderstood the position to
be that of a commander of ship.366
There was at least one armed bow-hand in command of the
foredeck and still known in the tenth century as prw/reuv" (pro2reus) as
------------------------------
363
Constantine VII, De administrando imperio, §51, ll. 80-90 (p. 250). Most
probably the expedition referred to was the one led by the strate2gos of Calabria
Eustathios to Sicily in 902. See above p. 68.
364
What topote2re2te2s meant in this context is unclear. The word literally meant
“warden of a place”. It became used quite widely for a range of subordinate officers
and mutated in meaning. The only text known to us which dates from the
Porphyrogenne2tan period which reports a topote2re2te2s with an active role as a field
commander is found in the continuation of the chronicle of George Hamartolos
sometimes attributed to Symeon Logothete2s, where a certain topote2re2te2s, Michael,
was reported on active service against the Bulgars during the reign of Ro2manos I
Lekape2nos. See George Hamartolos, Chronikon syntomon, col. 1152.
Nike2phoros Pho2kas wrote that the commander of the cavalry scouts known as
the prokoursavtore" (prokoursatores) should be either a topote2re2te2s or a strate2gos.
See Nike2phoros Pho2kas, Praecepta militaria [McGeer], IV.2 (p. 38). However, note
that in the corresponding passage of his Taktika, which was paraphrased from
Nike2phoros Pho2kas, Nike2phoros Ouranos deleted the mention of a topote2re2te2s. See
Nike2phoros Ouranos, Taktika [McGeer], ch. 61.2 (p. 118).
In this maritime context the word is normally translated as “vice admiral”, the
position being understood as one of the deputies of the droungarios to2n ploimo2n. In
the treatise on precedence or Kle2torologion of Philotheos of 899, however,
tourmachai to2n ploimo2n appeared between the droungarios to2n ploimo2n and the
topote2re2te2s. See Oikonomides, Listes de préséance, pp. 144 & 150. If the droungarios
to2n ploimo2n and topote2re2te2s of the ships were associated with the fleet of the Stenon,
as the De administrando imperio makes quite clear, it is difficult to believe that the
tourmachai to2n ploimo2n were not, and they clearly ranked ahead of the topote2re2te2s. It
is rather tempting to consider the office of topote2re2te2s in this context as being similar
to that of a “port admiral”, a post entrusted to experienced sailors whose days at sea
were behind them.
365
Constantine VII, De administrando imperio, §51, ll. 103-112 (p. 250), 137-148
(p. 252); esp. ll. 146-8: “..., a{ma de; kai; toi'" tovte prwtokaravboi" uJpotiqevmeno" kata;
th;n duskrasivan kai; pneu'sin tw'n aJnevmwn th;n basivleion nau'n phdalioucei'n te kai;
kuberna'n.”.
366
Liudprand of Cremona, Antapodosis, III.25 (p. 83).
272 CHAPTER FOUR
in antiquity. Of the two oarsmen at the bow on the upper oar-bank,

one was the sifwnavtwr (sipho2nato2r) and operated the sipho2n at the
bow. The other was in charge of the anchors. Why it was necessary to
have a crewman deputed to look after the ship’s standard, flavmoulon
(phlamoulon), appears inexplicable, unless perhaps he was also the
signaller.367 The Anonymous identified the standard keeper as
keleusthv" (keleuste2s); however, this he derived from Pollux, who did
not explain the word. In fact its classical meaning was the master of
oars, in charge of rowing training, giving the beat to the oarsmen, and
transmitting orders from the officers of the poop. This appears to have
been a word and a rank which fell out of use during the post-
Hellenistic period. The Romans took it over as celeusma for the
command used to set the stroke but used hortator or pausarius for the
master of oars and portisculus for his hammer.368 Scholia on
Thucydides misunderstood the word to mean either strate2gos or
kyberne2te2s.369 It does appear to have been properly understood in
glosses of the seventh to ninth centuries, however.370 Obviously,
Byzantine galleys must have had some such master of oars, probably
two in fact, one for the upper and one for the lower bank of oars, but
what they were called is unknown. Perhaps they were still known as
keleustai, or some variant of that, and the Anonymous merely got
their functions wrong. There is support for this in the hypothesis that
the medieval Latin and vernacular terms for a galley’s crew or
oarsmen, the Western equivalent of ousia, were supposedly derived
------------------------------
367
See Appendix Two [a], §8; Appendix Three, §4.3; Appendix Five, §7. See
Casson, Ships and seamanship, pp. 303, 318-19. Interestingly, Ibn Mankalı3’s
translator added that the sipho2nator should have an “elite squad” with him. See
Appendix Eight [a], p. 243.
368
See Morrison, Greek and Roman oared warships, p. 350.
369
See Appendix Three, §4.3; Hude, Scholia, II.84.3 (p. 153): “tw'n keleustw'nÚ tw'n
strathgw'n kai; tw'n kubernhtw'n.”. See also Casson, Ships and seamanship, pp. 300-
310.
370
See Isidore of Seville, Etymologiae, XIX.ii.13: “Porticulus malleus in manu
portatus, quo modo signum datur remigantibus.”. See also the Greek-Latin Cyril
glosses of London, British Library, MS. Harley 5792 in Goetz, Glossarii Latini, vol.
2, p. 347, l. 28: “Keleusth" iussor [sic]”; the Philoxenos Latin-Greek glosses of Paris,
Bibliothèque Nationale, MS. Lat. 7651, ibid., p. 154, l. 7: “Porticulus [recte,
portisculus] keleusth2" [sic]”; and the Glossae Nonii of the eighth-ninth-century
manuscript Leiden, Bibliothek der Rijksuniversiteit, MS. BPL 67F, ibid., vol. 5, p.
645, l. 34: “Portisculus hortator remigum”. See “Note on citations of Greek and Latin
glossaries”, p. lxix above. The gloss of Leiden 67F was based on the actual text of
Nonius Marcellus, probably from the early fourth century. See Nonius Marcellus, De
compendiosa doctrina, II.151 (vol. 1, p. 221): “portisculus proprie est hortator
remigum, id est, qui eam perticam tenet, quae portisculus dicitur, qua et cursum et
exhortamenta moderatur.”.
from celeusma, thence clusma, thence ciurma and many other forms
in various dialects and languages, such as çurma, çörme, and then
eventually back into Greek as tsouvrma (tsourma) or tzouvrma
(tzourma) and variants. The earliest known medieval use of the word
is the Venetian zurma in 1278, but it had obviously remained alive
over the centuries before that. The Venetian term may well have been
derived from a Byzantine one.371 Whatever the case, the development
of the meaning of the word explains why the Byzantine ousia never
passed into the other languages as a term for a galley’s crew.
The flautist of a classical trie2re2s, trihrauvlh" (trie2raule2s), said by
the Anonymous to be the ship’s trumpeter, ijbukinavtwr (ibykinato2r),
correctly boukinavtwr (boukinato2r), was also no doubt a real officer of
some sort. The Strate2gikon attributed to Maurice mentioned
boukinato2res, and Leo VI and Nike2phoros Ouranos both implied that
Figure 26
Dromons in the Kyne2getika of Pseudo-Oppian (Venice, Biblioteca Marciana,
MS. Gr. 479 [coll. 881], fol. 23r), eleventh century.
a trumpet, bouvkinon (boukinon), was used for giving orders; although

not in battle.372 The figures at the sterns of the two ships in an
illustration of naval warfare in the eleventh-century manuscript of the
Kyne2getika of pseudo-Oppian in the Marciana library appear to be
------------------------------
371
Tafel and Thomas, Urkunden, vol. 3, p. 257: “ ... cum uno ligno, quod armaverat
in Ania, et zurinam 4 suam, Latinos et Grecos, ...”.
4
leg. zurmam s. ziurmam
Cf. Kahane, Lingua Franca, §723 (p. 475); Jal, Glossaire nautique, p. 477.
372
See Appendix Two [a], §45; Appendix Three, §4.3; Appendix Five, §43 and cf.
Nike2phoros Ouranos, Ek to2n taktiko2n, 119.26 (p. 98); Maurice, Ek tou Maurikiou, §3
(p. 41). Cf. Maurice, Strategikon, XIIB.21.12 (p. 468).
274 CHAPTER FOUR
playing such flutes or trumpets.

Whether there was any real difference between the instrument that
on Greek trie2reis had been called a “flute”, aujlov" (aulos), and that
which was called a bucina on Roman liburnae, and a boukinon on
Byzantine dromons is debatable.
Whatever the case, there can be little doubt that dromons must have
had some such officers as keleustai and trie2raules. The sea trials of
Olympias revealed that there were significant problems in setting the
stroke and communicating commands throughout the ship. High-
pitched sound, such as that from a pipe, carried best within the ship.373
On dromons, how were commands communicated simultaneously to
crews rowing both below and above deck? Obviously, there must
have been some way of doing so, perhaps by having a hatch in the
deck immediately forward of the poop with two keleustai or
boukinato2res stationed there within hearing of the kentarchos, one
above and the other below deck.
The other officers referred to by the Anonymous, the
penthkovntarco" (pente2kontarchos), eJkatovntarco" (hekatontarchos),
nauarchos and ejpistoleuv" (epistoleus), can all be dismissed as
derivatives from Pollux with no relevance to the tenth century. In
antiquity a pente2kontarchos had been a sort of commander’s secretary
or “purser” of a trie2re2s.374 Hekatontarchos was merely a generic for a
commander of 100 men. Navarchos, as we have seen, meant an
admiral and epistoleus was a vice admiral. However, there is no
evidence that any of these terms were used for ranks in the Byzantine
navy.
Finally the crews. The oarsmen were known as kwphlavtai
(ko2pe2latai) or ejlavtai (elatai) and the marines as either “soldiers”,
stratiw'tai (stratio2tai), or “warriors”, polemistaiv (polemistai).375
These were soldiers from various army corps, not special sea-soldiers
or “marines” in the modern sense. The word for “sailors”, nau'tai
(nautai), seems to have been reserved for the crews of the baggage or
supply ships, skeuofovra (skevophora) or fortikav (phortika), and
horse transports, iJppagwgav (hippago2ga), of the “baggage train”,
------------------------------
373
See Morrison and Coates, Trireme reconstructed, pp. 30-32, 63, 101-2; Rankov,
“Rowing Olympias”, pp. 53-5; Morrison, et al., Athenian trireme, pp. 250-2.
374
See Morrison, et al., Athenian trireme, pp. 112, 124.
375
See Appendix Two [a], §§5, 6, 8, 13, etc.; Appendix Three, §3.1, 5.rub.;
Appendix Five, §§4, 5, 7, 11; Haldon, “Theory and practice”, pp. 202-35 passim and
Constantine VII, De cerimoniis, II.44 & II.45 passim (a[ndre" kwphlavtai, polemistaiv,
stratiw'tai, etc.).
suggesting that these ships were sailing ships rather than galleys.376
To this point we have considered command structures almost
entirely by internal examination. However, it is necessary to modify
some of the arguments above in the light of external evidence.
First, the helmsmen. Leo VI and Nike2phoros Ouranos both referred
to only two helmsmen per dromon. However, all medieval
Mediterranean ships, whether sailing ships or galleys, always had two
quarter rudders, one on each side. If there were only two helmsmen,
they would have had to have manned them without relief. In the
earliest evidence known from the West for the crews of war galleys,
that of the chancery registers of the Kingdom of Sicily in the reign of
Charles I of Anjou, there were always four helmsmen, nauclerii, per
galley.377 No doubt, they stood alternate watches on the rudders and
this must surely have been the case on Byzantine dromons also. There
is supporting evidence for this in the inventory for the Cretan
expedition of 949. Although Leo VI mentioned only two helmsmen,
one sipho2n operator, and one bowman, the inventory said that twelve
light corselets should be provided for these men.378 So, obviously
there were more of them than the four men in all mentioned by the
emperor; most probably four of each kind, to make a total of twelve.
We should understand Leo VI and Nike2phoros Ouranos to have
been referring to the command structure as it would have been found
in place at any point in time rather than to the total number of crew.
There would always have been one bowman on watch at the prow;
however, in battle there would have been many more than one
stationed there and at other times watches were no doubt alternated as
they were for the helmsmen. Similarly for the sipho2n operators. Leo
VI said explicitly that a particular oarsman, one of the two on the last
benches at the bow, should operate the sipho2n. However, he also said
that he had invented hand-sipho2nes, the Anonymous said that two
more sipho2nes were used at the sides, and the inventory for the Cretan
expedition of 949 said that each dromon had three sipho2nes.
It is also possible that the same considerations applied to the
kentarchoi. None of the Byzantine texts suggest that there was more
than one kentarchos per dromon; however, the Angevin chancery
documents are quite clear that the galleys of the Kingdom of Sicily
had two commanders, comiti, each. Again, they no doubt stood
------------------------------
376
See Appendix Two [a], §13; Appendix Five, §11.
377
See Pryor, “Galleys of Charles I of Anjou”, Table Three (p. 82).
378
276 CHAPTER FOUR
watches in command of the ships when at sea. How general questions

of command were resolved between them is not known because they
both had the same rank and were paid the same.379 Assuming that such
questions could be resolved somehow, it would certainly make sense
to be able to share the operational command of dromons in watches
between two kentarchoi and, given the interpretation of the texts as
referring to the command structure as it would have been found in
place at any particular time, there is no reason per se to discount the
possibility. No text known to us has a correlation between a number of
dromons and a number of kentarchoi.
(j) Oarage system and dimensions
On either side of a bireme dromon there were the rowing benches, the
“thwarts”, zugovi (zygoi), arranged in two banks, one “above” (a{nw)
and the other “below” (kavtw).380 Above and below what is not
indicated in any of the sources; although, the deck must surely have
been meant. The syntax of the Anonymous at §1.7 where he discussed
the oar-banks was very obscure; however, he appears to have said that
the thwarts below deck were called zugav (zyga) and the lower oarsmen
zuvgioi (zygioi). He added that the benches above deck were known as
“benches”, qra'noi (thranoi), and their oarsmen as qrani'tai
(thranitai). Then he added that if there was a third oar-bank, those on
it were known as qalavmioi (thalamioi).381 In all this, however, he was
once again merely following his sources, probably Pollux,382 who was
------------------------------
379
See Pryor, “Galleys of Charles I of Anjou”, Table Three (p. 82) and Table Six
(p. 90).
380
See Appendix Two [a], §§7-8: “”Ekasto" de; tw'n dromwvnwn eujmhvkh" e[stw kai;
suvmmetro" e[cwn ta;" legomevna" ejlasiva" duvo, thvn te kavtw kai; th;n a[nw. ÔEkavsth de;
ejcevt w zugou;" to; ejlavciston keV ejn oi|" oiJ kwphlavtai kaqasqhvsontai, wJ" ei\nai zugou;"
tou;" a{panta" kavtw me;n keV, a[nw de; oJmoivw" keV, oJmou' nV .”. Cf. Appendix Five, §§6-7.
381
Appendix Three, §1.7: “Ta; de; eJkatevrwqen tw'n toivcwn katavstega katavstrwma
levgetai kai; qra'no" kai; sanidwvmata, w|n a[nwqen hJ prwvth eijresiva kai; oiJ oJpli'tai kai;
toxovtai kai; peltastaiv, kavtwqen de; tou' sanidwvmato" hJ deutevra h{ti" di o{lou ejr evttei,
tucovntwn ejpi; tou' katastrwvmato" a[nwqen polemouvntwn. Kai; oiJ me;n ejp i; tou' qravnou"
kaqhvmenoi qrani'tai levgontai, oiJ de; eij" ta; zuga; zuvgioi: kai; qalavmioi de; e[stin o{te eij
e[cei trei'" eijr esiva " hJ nau'".”.
382
Pollux, Onomasticon (Bethe), I.87 (vol. 1, p. 28): “kai; to; me;n e[dafo" th'" new;" ...
kaloi'to d a]n kai; qavlamo", ou| oiJ qalavmioi ejrevttousi: ta; de; mevsa th' new;" zugav, ou| oiJ
zuvgioi kavqhntai, to; de; peri; to; katavstrwma qra'no", ou| oiJ qrani'tai.”.
All this was widely known and the Anonymous may have got it from anywhere.
See, for example, the scholion on Aristophanes’ Acharnenses, l. 162 in the tenth-
century Ravenna manuscript in Rutherford, Scholia Aristophanica, Acharnenses.162
(vol. 2, p. 282): “tw'n ga;r ejrettovntwn oiJ me;n a[nw ejrevttonte" qrani'tai levgontai, oiJ de;
Figure 27
Midships oars of a dromon of the era of the Macedonian emperors, drawn
at a right angle to the centre-line of the ship.383
© John H. Pryor
himself only rehearsing the standard arrangement and nomenclature

for the oar-banks of classical Greek trie2reis. In early Greek, before the
development of trie2reis, zygon or zygos meant simply a bench or
thwart from which an oarsman rowed. When trie2reis were developed,
the term zuvgio" (zygios) or zugivth" (zygite2s) became used for an
oarsman in the middle oar-bank. Qalavmio" (thalamios) or qalamivth"
(thalamite2s) became used for an oarsman in the oar-bank below the
zygioi because their oars were worked through oar ports known as
qalamiavi (thalamiai). Qranivth" (thranite2s) became used for an
oarsman in the oar-bank above the zygioi, apparently because they
were seated on small benches known as qra'noi (thranoi). However,
once again in the grip of a piece of classicizing lexicography, these
terms were only vaguely understood by the Anonymous from Pollux
------------------------------
mevsoi zeugi'tai, oiJ de; kavtw qalavmioi.”. See also the scholion on Thucydides, VI.31.3
in Hude, Scholia, (p. 343): “qrani'tai: oiJ th;n ajnwtevran tavxin ejn th'/ trihvrei tw'n nautw'n
e[conte": oiJ ga;r mevsoi zeugi'tai: oiJ de; katwvt eroi qalavmioi.”.
383
Dimensions based on those calculated in this section, especially Figure 32.
278 CHAPTER FOUR
or elsewhere. Leo VI and Nike2phoros Ouranos made it clear that in

the tenth century zygos was used for all the thwarts or benches on
which oarsmen sat, irrespective of any vertical arrangement of the oar-
banks.
The emperor, and Nike2phoros Ouranos following him, said that
there should be 25 thwarts in each oar-bank, for a total of fifty, each
with two oarsmen, one on the right and the other on the left.384
However, we must understand them to have meant 25 thwarts per
side. No one would ever have constructed a war galley with thwarts
stretching from one beam to the other because they would have
impeded the free access from stern to bow down the centre line which
was essential in battle. Moreover, it would have been impossible to do
so in any case because the oarsmen’s thwarts were almost certainly
angled forward at around 18.4˚ from the right-angle to the
centreline.385
Following Pollux again,386 the Anonymous said that the oars of the
lower oar-bank were rowed from tholes, skalmoiv (skalmoi), through
oar-ports, trivmata (trimata), corrrectly trhvmata (tre2mata),387 in one of
the hull strakes which was known as the qureovn (thyreon). They were
hung from the skalmoi by oar-grommets, tropwth're" (tropo2te2res).388
------------------------------
384
See Appendix Two [a], §8: “ÔEkavsth de; ejcevtw zugou;" to; ejlavciston keV ejn oi|" oiJ
kwphlavtai kaqasqhvsontai, wJ" ei\nai zugou;" tou;" a{panta" kavtw me;n keV, a[nw de; oJmoivw"
keV, oJmou' nV. Kaq e{na de; aujtw'n duvo kaqezevsqwsan oiJ kwphlatou'nte", ei|" me;n dexiav,
ei|" de; ajristerav, wJ" ei\nai tou;" a{panta" kwphlavta" oJmou' tou;" aujtou;" kai; stratiwvta"
touv" te a[nw kai; tou;" kavtw a[ndra" rV.”. Cf. Appendix Five, §7.
385
Cf. below pp. 284-91.
386
Appendix Three, §2.12: “ÔH de; sani;" di h|" aiJ kw'pai ejxevrcontai qureovn, kai;
o{qen me;n ejkdevd entai skalmov", w|/ de; ejndevd entai tropwthvr. To; de; ejpi; tw'n skalmw'n
ejpiskalmiv". Di w|n de; ei[retai hJ kwvph trhvmata [trivmata: MS. A]. To; de; pro;" aujt w'/ tw'/
skalmw'/ devr ma a[skwma, to; par hJmi'n manikevllion.”. Cf. Pollux, Onomasticon (Bethe),
I.87-88 (vol. 1, p. 28): “kai; o{qen me;n aiJ kw'pai ejkdevd entai, [skalmov", w|/ de;
ejkdevdentai], tropwthvr, ... to; d uJpo; to;n skalmo;n [ejp iskalmiv"]. ... di w|n de; dieivretai hJ
kwvph, trhvmata. to; de; pro;" aujtw'/ tw'/ skalmw'/ devrma a[skwma .”.
387
Tre2ma was a post-classical word derived from the classical truvphma (trype2ma)
for an oarport. See Hesychios, Lexicon (Schmidt), T.1315 (vol. 4, p. 171): “trhvmata:
truphvmata”.
Almost all ancient and medieval pictorial representations depicted the oar ports
as as being round; however, in fact they must have been elongated as shown on the
Victory of Samothrace monument in order for the oars to work through them on a
lateral plane around the skalmoi inside the hull. See Casson, Ships and seamanship,
fig. 118.
388
Leo VI and Nike2phoros Ouranos both recommended that dromons should carry
spare skalmoi and oar-grommets. Leo used the classical scoiniva (schoinia) for ropes
or cords for the oar-grommets but Nike2phoros corrected it to the technical tropo2te2res.
It has been almost universally assumed, on the basis of classical evidence, that
oar-grommets were made of leather. See p. 198 and n. 102 above. However, on
The ports were sealed against water by leather sleeves around the
oars, a[skwmata (asko2mata) in classical Greek but manikevllia
(manikellia) in the tenth-century.389
Thus far he got it right, but then the Anonymous made a “howler”
because he did not understand Pollux. Pollux wrote that skalmoi were
set in and on something, presumably a timber of some kind, called an
ejpiskalmiv" (episkalmis): “to; d uJpo; to;n skalmo;n [ejpiskalmiv"]” (“what
is under the thole [is] the episkalmis”). The sense of the word was
confirmed by Hesychios,390 and it was also used in the plural
ejpiskalmivde" (episkalmides) by Agathias with the sense of some
things to which makeshift oars were attached by the Huns at the siege
of Cherso2n, surely as a synonym for skalmoi in his case.391 But,
deceived by the “above” or “upon” sense of the prefix “ejpiv”, the
Anonymous altered Pollux’s text to read: “To; de; ejpi; tw'n skalmw'n
ejpiskalmiv"” (“What is on the tholes is the episkalmis”), making
nonsense of it. That tholes were set in something called an episkalmis
must have been correct.392 They had to be set in something. But that
------------------------------
Olympias it was found that leather stretched too much and it had to be replaced by
rope tied in such a way as to enable the grommets to be tightened when necessary.
See Morrison and Coates, Trireme reconstructed, pp. 74, 100; Coates, et al., Trireme
trials, p. 56; Coates, “Oar”, p. 49; Morrison, et al., Athenian trireme, pp. 241-2.
Note also that, contrary to what one might expect, the oars of Olympias were
actually rowed against the oar-grommets rather than against the tholes themselves.
The oars passed forward of the tholes, rather than aft of them. See Morrison and
Coates, Trireme reconstructed, p. 100; Coates, “Oar”; Morrison, et al., Athenian
trireme, pp. 215, 242.
Given our oft-stated doubt about the Anonymous and his treatise, we would not
wish to make too much of either his technical expertise or the precise import of his
language. However, that being said, it should be noted that when he referred to the
oars, he said that they were “hung from” the tholes by the oar grommets, rather than
that they were “attached to” them. See Appendix Three, §2.12: “ÔH de; sani;" di h|" aiJ
kw'pai ejxevrcontai qureovn, kai; o{qen me;n ejkdevdentai skalmov", w|/ de; ejndevdentai
tropwthvr.”. The use of the word ejkdevw, from which the verb ejkdevdentai meaning
“hung from”, is curious. One might have expected some word relating to “attaching
to”. It does suggest that the oars were secured to the tholes by an oar grommet but that
they were rowed against the grommet rather than the thole.
389
The meaning of asko2mata may have been forgotten in practice by the tenth
century. The scholia on Aristophanes’ Acharnenses, l. 97 in the tenth-century
Ravenna manuscript confused the asko2mata with tropo2te2res. See Rutherford, Scholia
Aristophanica, Acharnenses.97 (vol. 2, p. 274): “a[skwma oJ iJma;" oJ sunevcwn th;n kwvphn
pro;" tw'/ skalmw'/: ...”.
390
Hesychios, Lexicon (Schmidt), E.5188 (vol. 2, p. 167): “ejp iskalmiv": to; uJpo; tw'/
skalmw'/ sanivdion”.
391
Agathias, Historiae, V.22.2 (p. 192): “ejmbavnte" dh; ou\n ejn aujtai'" a[ndre" ej"
eJxakosivo u" kai; ptuva wJ" plei'sta tai'" ejpiskalmivsin ejntropwsavmenoi ... ”.
392
On Olympias, the tholes were set on carlings between the frames. Personal
communication from John Coates to John Pryor. However, there is no reason why a
single continuous stringer run along the inside of the frames could not have served the
280 CHAPTER FOUR
anything, whether an episkalmis or anything else, could be set “on the

tholes” would obviously have been impossible since if it were, the
oars could not then be hung from the tholes by attaching the oar-
grommets to them and then slipping the grommets over the tholes.
The only independent information that the Anonymous appears to
have added to Pollux is the use of manikellia for asko2mata, and the
curious use of thyreon for the strake which had the oar-ports. This was
not a classical word and the Anonymous’s source is unknown. It was
probably suggested from the idea of a row of oar ports and hence
derived from quriv" (thyris) an opening. The classical Greek word for
an oarport was qalamiav (thalamia); 393 however, it was a word little
used in classical Greek literature and the Anonymous appears not to
have known it. The corresponding Latin word was columbarium,394 a
word which survived with the same meaning into the Middle Ages,
but it also was apparently unknown to the Anonymous.
Manikellia for oar sleeves is confirmed by an inventory for the
Cretan expedition of 949, which specified 50 manikelia, together with
their gonativa (gonatia), for each dromon.395 Gonatia must have had
the sense of some kind of “joints”. We suggest that they were the
means by which the oar sleeves were attached either to the hull or to
the oars.396 Although it obviously proves nothing about Byzantine
dromons, the construction of the oar sleeves for Olympias is
interesting in this respect.397 The sleeves were made from four
trapezoidal pieces of leather 47 centimetres long sewn together to
form a tapering cone. The large ends were in the form of a square
equal to the clear distance between the hull frames; i.e. 37
centimetres. At the large end about 25 centimetres was left unsewn,
------------------------------
same purpose. Such a stringer may have been the episkalmis.
393
See Casson, Ships and seamanship, pp. 83 & 87, n. 52.
394
See, for example, Isidore of Seville, Etymologiae, XIX.ii.3: “Columbaria in
summis lateribus navium loca concava per qua eminent remi; dicta, credo, quod sint
similia latibulis columbarum in quibus nidificant.”.
395
Appendix Four [b], §IV.7: “manikevlia ajna; nV, oJmou' Àa su;n tw'n gonativwn aujtw'n,”
[= Haldon, “Theory and practice”, p. 227; Constantine VII, De cerimoniis, II.45 (vol.
1, p. 672)]. Manikel(l)ion passed into medieval Latin in the West as manichilium for
the same thing. See Pryor, “Galleys of Charles I of Anjou”, pp. 80-81.
396
Reiske suggested that gonatia were leather guards used by the oarsmen to
protect their knees against abrasion from the benches when pushing against them in
order to impart maximum force to the rowing stroke. See Constantine VII, De
cerimoniis, vol. 2, p. 795. However, this is not plausible. Reiske apparently knew little
about rowing. Oarsmen never used their knees to push against benches to impart
greater force to their stroke. They used their feet against footrests.
397
The following information was supplied by John Coates to John Pryor in
personal communications.
thus forming four flaps. The two side flaps were secured by battons
nailed to the frames either side of the oarports. The top and bottom
flaps were similarly secured by battons, but nailed to the inside of the
planks above and below the oarports. At the small end of the sleeves
the leather was cut 14.5 centimetres square and, when sewn together,
the hole was just large enough to pass the blades (18.8 centimetres
wide) and looms of the oars through them. They were secured to the
oars by pushing the oars through them and then pulling the oars and
sleeves inboard so that the sleeves were inside out and then tying the
sleeves to the oars with thin cords. The oars and sleeves were then
pushed back out through the oarports. The result was reported to be
satisfactory. This practical experiment leads to the suggestion that the
“joints” or gonatia of the manikellia for the dromons may have been
something like the battens used on Olympias to attach the sleeves to
the hull.
According to the Anonymous, above the lower bank of oars was
the wale called the peritonon, then another strake called the pevla
(pela), then another wale, and then another thyreon for the upper oar-
bank, which presumably also had tre2mata for the oars to come
through as the lower one had.398 Apparently the tre2mata of the upper
bank did not need manikellia since the inventory for the Cretan
expedition of 949 specified 120 oars per dromon but only 50
manikelia;399 presumably because the tre2mata of the upper thyreon
were much higher above the waterline than those of the lower, and
were above the deck in any case so that any water coming through
them would run off through scuppers. What the Anonymous intended
by the plank he called a pela is totally obscure. The word as such
appears not to have been known in either classical or medieval Greek
and no word with any similar form and nautical connotation is known
to us in medieval Latin or Western vernacular languages.400
------------------------------
398
Appendix Three, §2.13: “Tauvth" de; a[nwqen th'" eijresiva" perivtonon, ei\ta sani;"
eJtera, hJ legomevnh pevla, ei\ta perivtonon, ei\ta pavlin qureovn, e[nqa hJ a[nwqen eijresiva.
“Anwqen de; pavntwn hJ ejphgkeniv", to; a[rti legovmenon katapathtovn: ejkei'sev pou kai; to;
kastevllwma givnetai, e[nqa ta;" ajspivda" oiJ stratiw'tai kremw'si.”.
399
Appendix Four [b], §§IV.7-8 [= Haldon, “Theory and practice”, p. 227;
400
No such term is found in either Jal, Glossaire nautique or Kahane, Lingua
Franca.
Pevlla (pella) was used in the De administrando imperio at §9, l. 18, where it
appears to have had the meaning of oars since it was used in conjunction with
skarmoiv (skarmoi) for tholes. See Constantine VII, De administrando imperio, §9, l.
18 (p. 58). This usage is a hapax legomenon in medieval Greek. Du Cange
hypothesized that pella was derived from the Latin pala, which in medieval Latin
could mean a blade of an oar. See Du Cange, Glossarium, col. 1144; Kahane, Lingua
282 CHAPTER FOUR
The hull was topped by a “gunwale”, ejphgkeniv" (epe2nkenis) in

classical Greek but apparently known as a katapathtovn (katapate2ton)
by the tenth century. Epe2nkenis was well known in Greek from
Odyssey, V.253 as meaning a gunwale.401 According to Eustathios of
Thessalonike2 in his commentary on the line, this interpretation of
epe2nkenis, which presumably was therefore common, was incorrect
and the term referred either to a [timber] to which the skalmoi were
attached, or to a peritonon.402 However, there is no guarantee that
Eustathios actually understood all these terms any better than did the
Anonymous, whose text does appear to make internal sense at this
point. Indeed, the only timber to which skalmoi could possibly be
attached would be the one below them in which they were set, and
Pollux and Hesychios were agreed that that was called an episkalmis.
It seems certain that Eustathios was wrong about this unless the term
epe2nkenis had changed meaning during the intervening centuries or
unless he was thinking only of small boats. In small boats, of course,
the timber in which tholes, or rowlocks, would have been set was the
gunwale. It still is.
There was a pavesade, kastevllwma (kastello2ma), to which marines
could attach their shields, as can be seen clearly in the illustrations of
the Ilias Ambrosiana manuscript and the Marciana Library manuscript
of Pseudo-Oppian’s Kyne2getika. [See Figures 7 and 26] In the case of
the Madrid manuscript of John Skylitze2s’ Synopsis historio2n, shields
are depicted only once, on folio 34v, on the imperial dromon attacking
the fleet of Thomas the Slav, and then they are misplaced on the lower
hull. [See Figure 57] This word for a pavesade appears to have been
the name for the bulwark of Byzantine war galleys. Theophane2s the
Confessor wrote that when the future emperor He2rakleios sailed from
Africa in 610 to seize the imperial throne, he did so with a fleet that
included pavesaded ships, ploi'a kastellwmevna (ploia kastello2mena),
and the strate2gikon attributed to Maurice said that commanders of
dromons engaged in forcing crossings of rivers should similarly
------------------------------
Franca, §457 (p. 328). But, whatever the Anonymous understood by the term pela, it
was clearly not oars but rather a name for some plank or strake of the hull.
401
Homer, Odyssey, V.253: “... : ajpa;r makrh'/sin ejphgkenivdessi teleuvt a.”.
402
Eustathios of Thessalonike2, Commentarii ad Homeri Odysseam, 1533.38-43
(vol. 1, p. 213), to Odyssey, E.253: “Tou' de; qamevsi, dokei' eujq ei'a ei\nai oJ qamh;" dia;
tou' h'. h] dia; dicrovnou, hJ qamiv". Ephgkenivd e" de;, sanivde" ejk prwvra" eij" pruvmnan
tetamevnai kai; ejpenhnegmevnai. o{qen kai; ejt umologei'*tai. para; ga;r to; ejpenegkei'n,
ejpenegki;" givnetai. kai; kata; metavqesin, ejp egkeni;" kai; kata; e[ktasin, ejphgkeniv". e[sti
de; ejphgkeni;" h] kaq h}n oiJ skalmoi; phvgnuntai, h] o{per koinw'" perivt onon levgetai para;
to; diovlou teivnesqai. para; de; toi'" palaioi'" fevretai, kai; o{ti ejphgkenivde", makra; xuvla
th'" scediva". h] ta; paraqevmata.”.
pavesade, kastellw'sai (kastello2sai), them.403 On Western galleys of

the twelfth and thirteenth centuries the shields were also hung along
the pavesades or bulwarks in a similar way. The anonymous
Itinerarium peregrinorum of the Third Crusade said that they were
arranged on the upper deck, overlapping in a circle or curve; that is,
following the curve of the deck.404 Theodore Prodromos, a leading
poet at the court of John II Komne2nos and that of the early years of
Manuel I, in his description of the preparations for a naval battle in his
novel Rhodanthe and Dosikles, which was probably written in the
1140s, wrote that marines struck out at the enemy from between each
two shields and then withdrew behind them for protection and that the
positioning of the shields was like the tops, the crenellations, of walls
and turrets, from which archers fired.405 Prodromos did not name the
types of ships involved but he was clearly referring to Byzantine
warships of the early twelfth century and dromons or chelandia can be
assumed.
Dromons did vary in size and with that must also have come
variation in oarage systems, although not necessarily in fundamental
or proportional structures.
In the case of dromons smaller than the norm, the galeai which
were described by the continuator of Theophane2s the Confessor in the
first part of the Theophane2s Continuatus as myoparo2nes and
pente2kontoroi, and parenthetically as mone2reis by Leo VI, and which
presumably had only one oar-bank, had only 50 oarsmen. Was this
oar-bank still located below deck as it had been on the dromons of
Prokopios back in the sixth century, or was it now located above
------------------------------
403
Theophane2s, Chronographia, A.M. 6102 (vol. 1, p. 298): “Touvt w/ tw'/ e[t ei mhni;
Oktwbrivw/ dV, hJmevra/ bV, ijndiktiw'no" idV, h|ken ÔHravkleio" ajpo; Afrikh'" fevrwn ploi'a
kastellwmevna, ...”; Maurice, Ek tou Maurikiou, §§2, 3 (p. 41).
404
See Itinerarium peregrinorum (Mayer), p. 323: “In superioribus vero tabulatis
clipei per girum disponuntur consertis, ...”. See also the illustration in the De rebus
Siculis carmen by Peter of Eboli in Peter of Eboli, De rebus Siculis, p. 104.
405
Theodore Prodromos, Rhodanthe and Dosikles, bk. 5, ll. 457-68 (p. 89):
ajll hjremoi'en ejmparevnta toi'" pivloi".
a[nw d ejp aujt w'n tw'n teqeimevnwn pivlwn
plhqu;" parh/wvrhto makrw'n ajspivdwn,
kai; tou'to boulh'" ajndro;" eujephbovlou.
ajnh;r ga;r ajmfoi'n ajspivd wn eJstw;" mevso"
plhvttein ejkei'qen ei\ce tou;" ejjnantivou",
aujto;" d e[swqen eijsiw;n tw'n ajspivdwn
ajtraumavtisto", ajblabh;" ejfestavnai:
tuvpon ga;r ei\cen hJ qevsi" tw'n ajspivdwn
oi|on ta; teicw'n a[kra kai; tw'n purgivwn,
ajf w|n oJisteuvo usin a[ndre" toxovtai
(teicw'n ojdovnta" tau'ta to; plh'qo" levgei)”.
284 CHAPTER FOUR
deck? On the one hand, both Leo VI and Nike2phoros Ouranos, and
also the Anonymous following the emperor, wrote that these galeai
were to be used as scout ships and it might therefore be suggested that
they were not intended to form part of a line of battle and that
therefore there would have been no need to protect their oarsmen
below deck. On the other hand, in other passages Leo VI and
Nike2phoros Ouranos both distinguished galeai from “small and fast
dromons” not armed for battle but used as scouts and for conveying
messages by saying that the galeai should be armed against normal, or
many, “eventualities”. This would suggest that they were intended to
go into battle and that the oarsmen would therefore be best protected
below deck.406 That being said, we will argue in Chapter Six that the
major development made to Western galeae in the eleventh and
twelfth centuries, the critical change which gave them a technological
superiority and which led to the demise of the dromon and the pre-
eminence of the Western galea as the battle galley par excellence in
the Mediterranean, was the development of two files of oarsmen, both
rowing from benches above deck: the so-called alla sensile oarage
system in which the oarsmen were not fully seated but rather used a
“stand-and-sit” stroke. It is also clear that these early Western galeae
were emulated from Byzantine galeai, most probably from those
encountered by the Normans and others in South Italy. It is therefore
tempting to believe that, by the eleventh century at least, Byzantine
galeai were rowed from above the deck, whether they had been in the
age of Leo VI or not.
In the case of dromons larger than the norm, there is no hard
evidence that tenth-century dromons did have a third file of oars and
the indirect evidence suggests that they did not. In fact the
Anonymous was the only Byzantine author to appear to say that some
Byzantine galleys could have three oar-banks.407 However, he was
merely extrapolating from the oarage system of classical trie2reis
which he knew indirectly through Pollux. To rely on his evidence for
a third oar-bank without any corroborating evidence would be
injudicious.
Although they wrote that some dromons could be larger than the
norm, neither Leo VI nor Nike2phoros Ouranos actually said that any
had three oar-banks. One of the inventories for the Cretan expedition
of 949 specified 1,000 manikelia for 20 dromons, 50 each, certainly
------------------------------
406
See Appendix Two [a], §§10, 82; Appendix Three, §3.2; Appendix Five, §§9,
74.
407
See above pp. 276-8 and Appendix Three, §1.7.
for the lower oars. On these dromons, the lower oar-bank must have
had 25 oars per side, just as Leo VI and Nike2p horos Ouranos
specified. However, the total number of oars for the 20 dromons was
120 each, a total of 2,400.408 Obviously, it would not have been
possible to have three files of oars per side if the total number of oars
was only 120 and the lower bank had 50. These dromons had only two
oar-banks and the extra 20 oars were spares, not a complete duplicate
set as recommended by Leo VI and Nike2phoros Ouranos, but certainly
spares.409
Makrypoulias has argued that the dromons of the 949 expedition
rowed 50 oars from the lower bank and 70 from the upper, basing his
argument on the fact that the armaments specified included 70
lamellar cuirasses, klibavnia (klibania), and 70 sewn shields,
skoutavria rJapta; (skoutaria rhapta).410 He argues that because of the
coincidence of numbers, and because the upper oarsmen doubled as
marines, this indicates that all the other 70 oars could, at least
possibly, have been rowed at the same time by 70 oarsmen and that
the 70 shields would all have been hung on the pavesade (kastello2ma).
However, this construction is flawed.
Throughout his study Makrypoulias assumes that all the figures in
the various Byzantine texts must be inclusive. If a text refers to 120
oars, then it must have been possible for them all to have been rowed
at the same time. But this was not necessarily the case. Oars break,
even under conditions of normal use. One oarsman “catching a crab”
could easily lead to chaos in an oar-bank and the smashing of oars.
More importantly, they could be expected to be broken in large
numbers in battle. Anyone outfitting a fleet for an expedition against
which the enemy could be expected to engage at sea would supply the
dromons with extra oars. Moreover, Makrypoulias fails to include in
his construct other figures amongst the armaments that do not support
his conclusion. A hundred swords, spaqiva (spathia), another 30
“Lydian” shields, skoutavria Ludiavtika (skoutaria Lydiatika), 80
trident pikes (corseques), kontavria meta; tribellivwn (kontaria meta
tribellio2n), 100 pikes, menauvl ia (menaulia), 100 throwing javelins,
rJiktavria (rhiktaria), 50 “Roman” bows, toxarevai ÔRwmaiai (toxareai
------------------------------
408
Appendix Four [b], §§IV.7-8 [= Haldon, “Theory and practice”, p. 227;
409
See Appendix Two [a], §5; Appendix Five, §4. Cf. Haldon, “Theory and
practice”, p. 337, n. 386.
410
See Makrypoulias, “Navy”, pp. 164-5 and Appendix Four [b], §§II.1, 8 [=
p. 669)].
286 CHAPTER FOUR
Rho2maiai), and 50 surcoats, ejpilwvrika (epilo2rika), are all

mentioned.411 These figures for other armaments indicate clearly that
there was no strict numerical relationship between the armaments
specified for oarsmen or marines and the number of oars rowed from
the upper bank.
In addition, no one would ever have constructed a galley with only
25 benches in the lower bank and 35 in the upper. A galley might have
had a few more benches in the upper bank because of the curvature of
the hull, just as Olympias had. However, surely not ten more. No
galley known to us from any period of history had anything like 25
and 35 benches on the lower and upper decks respectively.
For the Cretan expeditions of both 911 and 949 the largest crews
specified for the dromons of the imperial fleet and those of the
themata were 230 oarsmen and 70 marines. Makrypoulias has applied
his same inclusive methodology to the oarsmen that he applied to the
rowing benches. Because 230 oarsmen in total were mentioned, then
in order to have as many of them as possible being able to row at the
same time he concludes that the upper oars must have been rowed by
two oarsmen each.412 Thus there were 50 lower oarsmen on 25
benches and 140 upper oarsmen on 35 benches. The remaining 40 are
left unexplained.
However, there is no hard evidence from anywhere in the
Mediterranean for the use of oars with multiple oarsman between
antiquity and the sixteenth century. It is true that William of Tyre did
write that in a Venetian fleet sent to Outremer in 1123 there were
some galleys larger than galeae which were known as gati, and that
these had 100 oars, each pulled by two oarsmen.413 But this was
almost certainly a mistake. William’s is the only known report of oars
with multiple oarsmen from the Middle Ages. His text was corrected
by the author of the Eracles, his Old French translator, who is known
to have been familiar with matters naval and maritime, to 100
oarsmen.414
------------------------------
411
Appendix Four [b], §§II.7, 9-10, 12-14, 20 [= Haldon, “Theory and practice”, p.
225; Constantine VII, De cerimoniis, II.45 (vol. 1, pp. 669-70)].
412
Alexandres suggests the same thing. See Alexandres, ÔH qalassiva, pp. 66-67.
413
See William of Tyre, Chronicon, 12.22.18-21 (vol. 1, p. 574): “Erant sane in
eadem classe quedam naves rostrate, quas gatos vocant, galeis maiores, habentes
singule remos centenos, quibus singulis duo erant remiges necessarii.”.
414
See Eracles, XII.22 (p. 546): “En cele navie, si com je vos ai dit, avoit nés que
l’en claime chaz qui ont bés devant ausint comme galies, mès eles sont gregneurs; en
chascune ot deus gouvernaux et cent nageeurs.”. On the Eracles author’s knowledge
of matters maritime see Pryor, “Eracles”. See also Pryor, “From dromo2n to galea”, p.
109.
Moreover, consideration of the mechanics of rowing a bireme

galley with two superimposed oar-banks leads inescapably to the
conclusion that Makrypoulias’s arrangement of the oarsmen is
ergonomically impossible.
Since bireme dromons had one bank of oars below deck, the
oarsmen of the lower oar-bank cannot possibly have used a “stand-
and-sit” stroke because of the height limitations below deck. They
must have been fully seated. Therefore, in order to coordinate and
synchronise the rowing strokes of the upper oarsmen with those of the
oarsmen below, it must have been necessary for them to have been
fully seated also.
The space or distance between any two tholes, skalmoi, is referred
to in a Latinization as an interscalmium. Because there are no
Byzantine data for what this space or distance may have been on a
dromon, we must proceed by analogy to other available data. On the
one hand, the trie2re2s Olympias, on which the oarsmen were also fully
seated, was built with interscalmia measuring only 0.888 metres;
although, it was subsequently realized that this should have been 0.98
metres.415 On the other hand, thirteenth-century galleys of the
Kingdom of Sicily, on which the oarsmen would have used a “stand-
and-sit” stroke, had interscalmia of approximately 1.20 metres.416 But,
without a moveable seat, it is virtually impossible for a fully-seated
man to pull an oar handle through much more than a metre.
Approximately a metre of the longitudinal axis of the ship should have
been close to the mark for the interscalmia for the fully-seated
oarsmen of dromons.417
------------------------------
It must be admitted that William of Tyre also ought to have been familiar with
matters maritime, having made several voyages by sea. His statement is
incomprehensible. However, the ship type, or possibly just a ship name, to which he
referred, the g(c)at(t)us, was a type/name which had been quite widely mentioned
early in the twelfth century but which had disappeared by his own day. Cf. below p.
412. It had long since ceased to be mentioned in any other sources. The word had
probably been derived from the Arabic qit6‘a. This word and its possible ship type has
never been subjected to research but it is possible that William’s report was based on
later reports of something no longer actually known.
415
Olympias was built with interscalmia of 0.888 metres, twice the classical Attic
cubit of 0.444 metres. However, it proved to cramp the oarsmen’s stroke excessively.
It was realized that the archaic or Doric cubit of approximately 0.49 metres should
have been used, which would have made the interscalmia 0.98 metres. See Coates and
Morrison, “Sea trials”, pp. 138, 140; Rankov, “Reconstructing the past”; Morrison,
“Lessons”; Morrison, “Triereis”, pp. 12-13, 18-19; Morrison, et al., Athenian trireme,
pp. 245-6, 268-9; Morrison, Greek and Roman oared warships, pp. 281-2.
416
See Pryor, “Galleys of Charles I of Anjou”, pp. 64-8; idem, “From dromo2n to
galea”, p. 110. See also below p. 435-6 & n. 22.
417
See Coates, “Naval architecture”, p. 2; Shaw, “Oar mechanics”, p. 169.
288 CHAPTER FOUR
It is important to bear in mind that in ship design everything was

always a compromise. One range of objectives could be achieved only
at the expense of others. Human comfort, stowage capacity, and sea-
worthiness had to be sacrificed if one wanted to build a warship which
would develop maximum power from the oars for short-term speed in
battle, which would have maximum manœuvrability, which would be
able to carry maximum crews for battle, and which would also have
sufficient sturdiness to withstand attack. And these were the
requirements of which battle dromons had to be capable.
Consequently, one would expect that oarsmen below deck would have
been packed into the hull with mimimum headroom between their
heads and the deck beams above them. And, there is also another
reason why this should be so. Oars are levers and they develop their
greatest mechanical advantage when as close to parallel to the water
as other factors will allow. Therefore, in order to maximize the power
generated by the oarsmen above deck, it would have been necessary to
minimize the height of the deck above the waterline in order to reduce
the angle to the waterline of both banks of oars to the minimum
possible. For the same reason, the oarsmen above deck must have had
their legs stretched out as straight as possible in order to minimize the
height of their benches above the deck and it would have been
necessary for the oarsmen below deck to have had a similar posture in
order to synchronize the strokes. We have allowed for a mere 25
centimetres from the top of the oarsmen’s benches to the bottom of
their feet stretchers.
Because the height of the top of the head of a six-foot (1.83 metres)
seated man is only around 0.95 metres above his seat, even when he is
fully erect, we can assume with confidence that the deck beams of a
dromon would have been no more than a metre or so above the
benches of the lower oarsmen. When these leaned into their stroke,
their shoulders will have been no more than around 70 centimetres
above their benches and their hands no more than around 40
centimetres above the tops of their thighs. These figures could be
reduced further if we assumed that oarsmen were shorter than six feet;
however, for the sake of argument we have proceeded with the figure
of six feet. Reducing it would make no difference to the conclusions
to be drawn. Given deck beams and planks of a total thickness of
around 15 centimetres, and ignoring for the moment the camber of the
deck, in a dromon the minimum difference in height between the mid
points of the hands of the upper and lower oarsmen on the handles of
their oars cannot have been less than approximately 1.40 metres, and
Figure 28
The oarage system of a bireme dromon of the era of the Macedonian
emperors, I: Interscalmia.
© John H. Pryor
that is to stretch the argument as far as is possible. In Olympias the

corresponding measurement between the lowest, thalamian, and
highest, thranite, oars was around 1.45 metres.
One might have thought that this height might have been reduced
by having one deck beam per interscalmium and placing the lower
oarsmen between the beams so that their heads moved within the
space between any two beams. However, this would not have been
possible because the number of deck beams required would have
precluded the possibility of one beam per interscalmium. Thirteenth-
century galleys of the Kingdom of Sicily had 55 deck beams for the
27 interscalmia of the rowing platform, two per interscalmium when
the staggering of the two oars per bench is taken into account.418
------------------------------
418
See Pryor, “Galleys of Charles I of Anjou”, pp. 39, 50, 54-6.
290 CHAPTER FOUR
Something similar would certainly have been necessary for the

structural integrity of dromons and chelandia and this would have
meant that the tops of the lower oarsmen’s heads must have been
below and clear of the deck beams.
We have no data for the lengths of oars of Byzantine dromons and
must again proceed by analogy to those designed for Olympias after
experimentation and to the recorded lengths for oars of galleys of the
Kingdom of Sicily. The latter had oars 6.86 metres long except for a
few of 7.91 metres at bow and stern.419 However, these were galleys in
which both files of oars were rowed from above deck using a stand-
and-sit stroke. The oars of Olympias were designed on the basis of the
lengths for oars specified in Athenian naval inventories: 9 and 9.5
cubits. Using the archaic cubit of 0.49 metres, this would give them
lengths of 4.41 metres and 4.655 metres.420 After experimentation the
latest oars designed for Olympias, which would be used on any future
Olympias Mark II, would have a total length of 4.66 metres, with
lengths from the end of the handle to the thole of 1.265 metres and
from the thole to the tip of the blade of 3.395 metres, a ratio of 1:2.68.
Their gearing would be 1:3.0 and their weight in hand 3.60
kilogrammes. The angle of the thalamian oars, the lowest oars, to the
waterline when the blades were below water during the stroke would
be approximately 11 degrees. The oar benches would be canted or
angled outboard towards the bow. On later medieval galleys they were
also canted towards the bow and for Olympias Mark II, it has been
proposed to cant them at an angle of 18.4˚ from the right angle to the
centre line, the angle whose tangent is closest to one third.421 During
the stroke the oars of Olympias Mark II would move through an arc of
approximately 54˚ from the dead point at the beginning of the stroke
before the catch in the water to the dead point at the end of it, at which
point the oars would be approximately 4˚ aft of a right angle to the
centre line of the ship. The mid-point of the oarsmen’s hands on the
handles of the oars would move forward and backward with each
------------------------------
419
420
See Morrison, et al., Athenian trireme, p. 269.
421
Alertz, “Naval architecture”, pp. 159, 162; Bondioli, et al., “Oar mechanics”, pp.
173, 176, 182-9; Morrison, et al., Athenian trireme, p. 271; communications from
John Coates to John Pryor.
The gearing of an oar is the ratio of its length from the mid-point of the
oarsman’s hands on the handle to the thole to that from the thole to the centre of water
pressure on the blade. This explains the difference between the ratio of 1:2.68 and the
gearing of 1:3.0. Its weight in hand is the downward force on the mid-handle needed
to raise the oar from the water and to balance it at the pivot at the thole. See also
Morrison, Greek and Roman oared warships, pp. 333-6.
stroke approximately 0.949 metres, which is just about the maximum

physically sustainable from a fixed seat, at an average angle of 23˚ to
the centre line, in order to move the oar handle around 0.874 metres
forward and back along the longitudinal axis of the ship. With
interscalmia of 0.98 metres, that would allow just sufficient tolerance
to allow the butt of each oar handle at the beginning of a stroke to pass
past where the body of the next oarsman aft was at the end of the
previous stroke in order to enable the oar stroke to be as long as
physically possible, an important factor in achieving maximum oar
power.
Given the total lack of any empirical evidence from Byzantine
sources, this is the only reasonable point of departure for
reconstructing the most probable oarage system of a dromon since the
mechanics of rowing from benches below deck must have been
similar in the cases of trie2reis and dromons but totally different in the
case of the later Western galeae which were rowed entirely from
above deck.
Olympias has 27 benches each side for the thalamian and zygian
oars and 31 benches for the thranite oars. She has an overall length
from stern to stempost, excluding the ram, of approximately 34.5
metres and an overall beam amidships, including the parexeiresiai, of
5.45 metres. However, if Olympias Mark II were to be built using the
archaic cubit and the number of oars were to remain the same but the
length of the interscalmia be raised to 0.98 metres and all other
lengths remain unchanged, she would have an overall length of 37.35
metres, excluding the ram. The beam would be more complex because
for reasons associated with performance under oars it is proposed to
eliminate the intermeshing of the blades of the three oar-banks in the
water by increasing the beam at the parexeiresiai to 5.62 metres while
keeping the beam at the waterline the same. The beam at the top of the
hull proper would be approximately 4.6 metres.422 Our best estimate
for the overall length of a 100-oared bireme dromon is only 31.25
metres. Assuming that the oarsmen of the upper and lower banks were
staggered on average by a half an interscalmium, a metre each for the
twenty-five interscalmia would make 25.5 metres and one must then
allow for prow and poop to the extremities of the stempost and
sternpost. On thirteenth-century Sicilian galleys this increased the
total length by around 22.5%,423 which would give standard bireme
------------------------------
422
See Morrison, et al., Athenian trireme, pp. 269-72.
423
See Pryor, “From dromo2n to galea”, pp. 110-14.
292 CHAPTER FOUR
dromons of 100 oars an overall length of approximately 31.25 metres,

or at least somewhere between 31 and 32 metres. This is as accurate
an estimate of their length as is possible and seems to be reasonable.
As for the beam amidships, the ratio of maximum beam at the
parexeiresiai amidships to overall length from stempost to stern of
Olympias Mark II would be approximately 1:6.65. That at the hull
proper would be approximately 1:8.21, whereas that of thirteenth-
century Sicilian galleys was 1:8.57. However, trie2reis with
parexeiresiai had a completely different oarage system to medieval
Western galleys, whose outrigger began to curve outboard only above
deck. Given the lack of any empirical data for dromons, it is necessary
to make a choice. Either the hull began to flare outboard upwards
from the lower oarports in a manner parallel to, but not the same as,
trie2reis with parexeiresiai, or it did not and dromons were straight-
hulled. For reasons examined below, we believe it most likely that the
upper hulls of dromons did flare outboard from above the lower
oarports. However, in order to demonstrate the reasoning, we proceed
from the beginning on the assumption that they were not and that the
hulls were straight sided. Using the hull beam:length ratio of
Olympias Mark II of 1:8.21 rather than the higher figure of 1:8.57 of
the Sicilian galleys on the grounds that the oarage system of the
former must have been more similar to that of dromons than the latter,
a maximum beam amidships of around 3.80 metres for a straight-sided
dromon should not have been too far from the truth.424
Olympias, of course, did not have a full deck and therefore
calculating a depth in hold from floor to deck is not possible for her.
However that of the thirteenth-century galleys of the Kingdom of
Sicily, which were similar to dromons at least in the respect that they
were also fully decked, was 2.04 metres and, by comparison, that of
the smaller dromons was probably around 1.75-1.8 metres to the deck
beams and 1.85-1.90 metres to the planks of the deck.425
Reducing the size of the oars of Olympias proportionately to the
beam, which would obviously be the critical dimension in such an
exercize, would give a dromon lower oars of only 3.85 metres. With a
gearing ratio proportional to that of the oars of Olympias, the inboard
------------------------------
424
The estimates of Dolley in “Warships”, pp. 48-9, which were based on
guesswork, were 130 feet (39.62 metres) long by 17-18 feet (5.18-5.49 metres) wide.
His estimates were too high and a dromon as he reconstructed it could not possibly
have been capable of any speed. Admiral Serre’s estimates of 36.0 metres in length
and 4.40 metres in the beam, were closer to the mark although the length was still too
high. See Serre, Marines de guerre, vol. 1, p. 91.
425
Figure 29
emperors, II: straight-hulled lower midships section with thalamian oars of
Olympias Mark II and scaled-down version.
© John H. Pryor
length of the oars of a dromon would then be around 1.045 metres and
their outboard length around 2.805 metres. By analogy to the latest
294 CHAPTER FOUR
thalamian oars designed for Olympias, which have eliptical blades,

the blades of the oars of a dromon would be around 15 centimetres by
81 centimetres. The mid-point of the handles would be around 16.5
centimetres from the butt and the centre of water pressure on the blade
a similar distance from the tip. Assuming that the angle to the
waterline of the lower oars was also around 11 degrees as in
Olympias, this would enable us to reconstruct the midships section of
such a bireme dromon within quite close parameters.
One has to bear in mind that the blades of oars are wider than their
looms. Thus, even if the length of a blade is, say, 50 centimetres from
the tip, the point at which the loom of the oar enters the water will be
above that, more or less above that according to the less or more acute
angle of the oars to the water. For Olympias Mark II the thalamian
oars would cross the waterline approximately 1.165 metres from the
tip of the blades, even though their blades would be only 0.98 metres
long. The figure for a dromon’s oars, if scaled down, would be
approximately 0.81 metres for the blades and 0.96 metres to the
waterline. That would make the height of the oar above the waterline
at the thole 0.365 metres and the height of the lower rim of the oarport
only approximately 0.3 metres. In any sort of swell the oarsmen would
end up rowing below water. Even light breezes of only 7-10 knots
raise wavelets of up to two feet or 61 centimetres.
Moreover, there is another obvious problem. As can be seen from
Figure 29, with scaled-down oars the mid point of the hands of the
oarsmen on the handles would move forward and back with each
stroke only 0.72 metres, using only around 72% of their interscalmia,
or around 79% if a similar tolerance to that of the oars of Olympias
Mark II is allowed for. Obviously, that would be extremely inefficient
and no war galley would ever have evolved with oars that delivered
less power and speed than what was ergonomically possible.
In fact, even though dromons were smaller than Olympias Mark II
would be, with interscalmia of 0.98 metres, say one metre, there is no
reason why they could not have used oars of the same length as those
of Olympias Mark II, thus enabling them to develop more power per
tonne of ship than a trie2re2s, as long as other factors were equal of
course. With oars the same length as those of Olympias Mark II the
full interscalmia would have been used and the bottom rim of the
lower oarports would have been approximately 36 centimetres above
the water line, a more acceptable configuration. The bottom rim of the
thalamian oarports of a future Olympias Mark II would also be
approximately 35 centimetres above the water line.
Figure 30
emperors, III: straight-hulled midships section with full sized oars above
and below.
© John H. Pryor
While varying with stature, with height, arm length, and torso
flexibility, a seated man is indeed able to move the mid point of his
hands on the handle of an oar forwards and backwards up to around a
296 CHAPTER FOUR
metre with each stroke. Therefore, it would be ergonomically

inefficient to design a war galley for seated oarsmen with interscalmia
less than around a metre, as the experience of Olympias has shown.
Interscalmia would never be made larger than necessary because that
would lead to a decrease in the power:size ratio of the ship; however,
they would also never be made smaller than what was necessary to
allow oarsmen to develop their stroke and power to the maximum
possible ergonomically. Around a metre seems to be just about right
and that being the case the length of the oars of Olympias Mark II also
seems to be just about right. The figures could be varied a little if
required, but the general parameters seem to be unarguable.
However, the same problems are then encountered with the upper
oars rowed from above deck as with scaled-down oars. Because of the
less acute angle of the upper oars to the water, the length of the oar
required to submerge the shoulders of the blades of the oars would
have been less than it was in the case of the lower oars. In Olympias
Mark II, the length required for the upper thranite oars, which have
shorter blades than the thalamian oars, would be only around 60
centimetres, making the length from waterline to thole 2.795 metres.
In a dromon the minimum difference in height between the mid points
of the hands of the upper and lower oarsmen on the handles of their
oars cannot have been less than around 1.40 metres and the lowest
angle to the waterline that could have been possible on a straight-sided
dromon with upper oars the same length as those of the lower ones
would have been 32 degrees, almost exactly the same as the case of
the thranite oars of Olympias Mark II. [See Figure 30]
The blades of the upper oars of such a dromon would have had to
intermesh below water with those of the lower oars if the oars were of
equal length, probably unworkably so in fact, and therefore the oar
strokes would have had to have been synchronised. However, it is
clear from the plane section of Figure 30 that it would simply not have
been possible for the upper oars to have remained synchronised with
the lower oars unless the upper oarsmen used only around 75.5% of
the interscalmia available to them, moving the mid point of their
hands on the oar handles forward and back only around 0.755 metres
during each stroke. The problem of the blades intermeshing could be
overcome by extreme flaring of the upper hull, as can be deduced
from the closely dotted alternative upper oars in Figure 30. But even
this would do nothing to overcome that of ergonomic inefficiency.
But the problems must have been overcome somehow for to have
built war galleys in which the upper oarsmen could use only 75.5%
Figure 31
emperors, IV: straight-sided midships section with oversized and full sized
oars above and below.
© John H. Pryor
of the ergonomic capacity available to them would obviously have

been unacceptable. Whatever they were, dromons were originally fast
ships and they remained the premier war galleys of the Empire for
298 CHAPTER FOUR
centuries. It is inconceivable that their design would have delivered

less than optimum ergonomic efficiency.
There has to be another solution and the obvious one is that the
upper oars must have been longer than the lower ones. [See Figure 31]
With the tholes of the upper oars set approximately 16 centimetres
outboard of those of the lower oars, and that is nothing more than an
estimate based on a fair curve since we have no information about the
upper hulls and bulwarks of dromons other than that they had to be
sufficiently vertical to hang shields on them to provide effective
protection, the upper oars could have been lengthened to 5.178 metres
if the gearing of 1:3.0 were maintained. The angle to the waterline
could have been lowered from 32˚ to 28˚, indeed it would have had to
have been lowered. This would have had two benefits. It would have
increased the mechanical advantage of the oars and would also have
ameliorated a little the difficulty of rowing the upper oars at such
steep angles.426 At 5.178 metres, the gearing could have been
maintained with handles of 44 centimetres and a mid point of the
hands 22 centimetres from the butt, an inboard thole to butt length of
1.406 metres, an outboard length of 3.774 metres to the tip of the
blade, and a blade immersion of 0.68 centimetres. With such oars the
upper oarsmen would have worked in a horizontal plane through the
same horizontal space as the lower oarsmen. The mid point of their
hands on the handles would have moved through 94.9 centimetres at
an average angle of 23˚ to the centreline, moving the mid point of the
handles 87.4 centimetres forward and back with each stroke. Such an
oarage system with longer upper oars would not have freed the blades
from intermeshing below water but apart from that it would have had
no particular drawbacks. If necessary, the weight in hand of the oars
could have been maintained at whatever weight it was for the lower
oars by re-shaping the thickness and tapering of the looms.
This is the appropriate point at which to return to the issue of the
shape of the upper hull: of whether it was straight sided as assumed so
far or whether it was flared outboard. A major series of problems
encountered in the sea trials of Olympias resulted from the design of
the hull and oarage system, in which the middle (zygian) and lower
------------------------------
426
On Olympias, whose upper or thranite oars were at an angle of about 32˚ to the
water when fully immersed at the end of the stroke, the oarsmen’s hands were held so
high that they had difficulty applying downward force to lift the blades from the
water. Most changed to to an underhand grip on the handle with their inboard hand in
order to spread the load on their bodies more evenly. See Morrison and Coates,
Trireme reconstructed, p. 40.
(thalamian) oars intermeshed in the water. The upper tips of the

blades of both also touched the bottom edge of the blades of the upper
(thranite) oars. Particularly in the early stages of learning to row the
ship the result was that oars clashed and were broken and the
thalamian oars were often held under water by zygian oars fouling
them from above, thus preventing recovery and actually endangering
the oarsmen. Another factor contributing to the problem was that in
Olympias the thalamian oarsmen could not see their oar blades
because the oar sleeves, asko2mata, covered the oarports. They had to
row blind, by sound and touch. The same would have been true, of
course, of the lower oars of a dromon. On Olympias, some of these
problems were overcome or alleviated by adjustments made to the
oars during the sea trials; however, one which could not be overcome
was that the turbulence created in the water by the massed oars of
Olympias reduced their effectiveness by increasing the slippage of the
oar blades through the water. Water operates most efficiently as a
fulcrum for an oar lever if it is still and clean. If it is disturbed, the
blade slips sideways through the water more, reducing the
effectiveness of the stroke.427
All of the problems resulting from the intermeshing oar blades and
associated factors in Olympias led to the conclusion that the hull and
oarage designs should be modified to ensure that in any future
Olympias Mark II the oar blades would not intermesh.428 We believe
that for similar reasons Byzantine dromons would also have had hull
and oarage systems which avoided having the blades of the oars of the
two banks intermeshing. And, this could in fact have been achieved by
a quite moderate flaring of the upper hull above the lower oarports. As
a result the beam of the ships at the deck would be increased from
approximately 3.80 metres to 4.46 metres. The flare on either side
would only be around 33 centimetres, hardly a matter of great
moment.
There would not appear to have been any particular disadvantages
to doing this, and it would give the great advantage of having the two
oar-banks rowing clear of each other, especially in stressed conditions
such as battle. There is little or nothing in the iconography to support
this interpretation, but there is nothing to contradict it either. The only
probable representations of dromons that we know of are those in the
manuscripts of the Roman Vergil, Ilias Ambrosiana, Sacra Parallela,
------------------------------
427
See Shaw, “Meshing”; Morrison, et al., Athenian trireme, Ill. 56 (p. 198), pp.
236-41.
428
Morrison, et al., Athenian trireme, p. 272.
300 CHAPTER FOUR
Synopsis historio2n of John Skylitze2s, and the Pantelee2mon manuscript

of the Sermons of Gregory of Nazianzos. [See Figures 4, 6-8, 13-14,
34, 47, 51-3, 57] None may be said to show clearly a flaring in the
upper hull; although, the first three do show a separate “band” at the
top of the hull with oar ports in it. It is not inconceivable that the
artists intended to depict a bank of oars rowed from above deck where
the hull flared outboard. That being said, only three of the Skylitze2s
illustrations show two banks or files of oars and all three are drawn in
Western styles representing twelfth-century Western galeae.429
In any oarage system with banks of oarsmen superimposed
vertically the parallel horizontal distance between the upper and lower
oars of any interscalmium cannot be the same at the beginning and
end of a stroke because the oar blades are pushed far forward at an
acute angle to the centre line at the beginning of a stroke but end up
only slightly aft of a right angle to the centre line at the end of it. The
position of the upper and lower oarsmen within interscalmia can be
adjusted to make the parallel horizontal distance between the oars
more or less equal at either the beginning or end of the stroke, but not
at both. One has to make a choice, or a compromise. However, it is
more important for oar blades to be as evenly spaced as possible at the
end of the stroke since the potential for oar clashing is greater then,
when the blades are being lifted from turbulent water, than at the
beginning, when the blades are moving into position through the air.
As in any ship design, no doubt there would have been some degree of
compromise in a dromon and there is no need to be absolute in such
matters. Given our conclusion that the upper hulls of dromons were
almost certainly flared outboard above the lower oarports in order to
prevent intermeshing of the blades of the upper and lower oars below
water and to separate the tracks of the two banks of oars in the water,
and therefore because the tholes of the upper oars are approximately
47 centimetres outboard of those of the lower oars, the positioning of
the tholes which gives the clearest operation for the two banks of oars
is with the tholes of the upper oars only about 33 centimetres forward
of those of the lower oars. Such a positioning gives a parallel
clearance between the two banks of oars of approximately 33
centimetres at the end of the stroke and 15 centimetres at the
beginning.
To return to the issue of multiple oarsmen for the upper oars. A
mere glance at Figures 30-32 makes it apparent immediately that any
------------------------------
429
See below pp. 426-30 and Appendix Seven, p. 637
Figure 32
emperors, V: midships section with oversized and full sized oars above and
below and flared upper hull
© John H. Pryor
arrangement of oarsmen with two oarsmen manning each upper oar as

suggested by Makrypoulias is ergonomically impossible. We adapt
302 CHAPTER FOUR
Figure 32 to make the point since with the oversized upper oars of
Figures 31 and 32 the upper oarsmen would be proportionately closer
to the centre line than with the full-sized oars of Figure 30 and
therefore there is more room for a supposed second oarsman. The
arrangement is impossible in any case, but it would be even more so
(sic!) without oversized oars.
The consequences are immediately apparent. Oarsmen have
reasonably broad shoulders. We have allowed 50 centimetres, which
would be a bare minimum. We have also allowed 20 centimetres
separation between the two oarsmen. That might possibly be reduced
a little, but doing so would make no difference to the conclusion
reached. The mid-point of the inner oarsman’s hands on the handle
would be 19.6 centimetres from the butt when viewed in plane. The
mid-point of the outer oarsman’s hands on the loom would then be a
mere 34.6 centimetres from the thole. His outer shoulder would be a
ridiculous 10 centimetres or so from the thole. He would be able to
move his hands forward and back only 31.4 centimetres, the oar
actually moving forward and back parallel to the centre line a mere
28.9 centimetres. He would also be rowing during the pull of the
stroke with the mid point of his hands on the handle of the oar
somewhere down around his navel. In other words, a second oarsman
added to an oar above deck would be so cramped in his stroke as to be
effectively useless.
To have added a second oarsman to oars above deck would have
necessitated complete redesign, not only of the oars but of the entire
ship and, even if this were done, multiple-oarsmen oars pulled above
deck could not have been synchronized in the stroke with single-
oarsman oars below deck. It would have been simply impossible.
Finally, if 70 oarsmen did indeed row at the same time from the
upper oar-bank, it would have meant that another ten bench positions
would have been needed and that those dromons would have had to
have been around 41-42 metres long, around four metres longer than
Olympias Mark II and long even by the standards of late-thirteenth-
and fourteenth-century galleys of the Latin West.430 It is almost
inconceivable that tenth-century dromons were as long as the latter,
which were bireme and trireme galleys at the high point of their
development.
We conclude both on the basis of analysis of the Byzantine
evidence itself and also on that of comparison to what is known of
------------------------------
430
Figure 33
emperors, VI: midships section with oversized and full sized oars above and
below, flared upper hull, and two oarsmen for oversized oars above deck.
© John H. Pryor
galley design in general in the Mediterranean over the centuries that

the dromons of the Cretan expeditions were almost certainly biremes,
and biremes only, and that they rowed single-oarsman oars with 50
304 CHAPTER FOUR
oarsmen on the lower benches and another 50 on the upper ones, just
as Leo VI and Nike2phoros Ouranos wrote. Their upper hulls were
almost certainly flared and their upper oars almost certainly longer
than the lower ones. There is no evidence in the Byzantine sources
that the upper oars were in fact longer than the lower ones; however,
there is none that they were not either. If they were the same length as
the lower ones, the problem of intermeshing blades could have been
overcome by extreme flaring of the hull or pavesade outboard but that
would not have solved the problem of the ergonomic inefficiency of
the upper oars. In fact it must have been the case at all times that
galleys with superimposed oar-banks either used oars of different
lengths for the different banks or, if they did not, the ergonomic
efficiency of the upper oarsmen must have been impeded, as was the
case with Olympias.431 And, although there is no mention of different
upper and lower oars in the Byzantine sources, in fact they must have
had differently-shaped blades because of the different angles at which
they met the water and, that being the case, there would have been no
reason for them not to have been of different lengths also.
Their tonnage can only be roughly estimated at best. Western
galleys of the Kingdom of Sicily in the later thirteenth century had a
deadweight tonnage, the weight of maximum cargo or military
equipage, of around 40 tonnes but their overall length was 39.55
metres.432 If bireme dromons had hull configurations below water not
greatly dissimilar to the later Western galleys, then their deadweight
tonnage ought to have been in the order of around 29.5 tonnes.
(k) Horse transports
Further research into the transportation of horses by sea during

antiquity, the Middle Ages, and beyond, including the Byzantine
Empire, remains an urgent desideratum.433 There is no doubt that in
antiquity political powers developed the technological capability to
transport horses by sea and that they continued to have it during the
Middle Ages. However, there is little evidence to suggest that this
------------------------------
431
See Morrison, et al., Athenian trireme, Ill. 81 (p. 271).
432
See Pryor, “From dromo2n to galea”, pp. 110, 114.
433
This issue was first canvassed for the Middle Ages by John Pryor in
“Transportation of horses by sea”, here pp. 9-11. Since then, in spite of an obvious
need for it to be addressed both in further detail and also for antiquity, no one has
done so.
capability extended to more than short traverses from one station to

another until the twelfth century when, first Venice, and then other
Western powers, developed a capability to transport horses for long
distances for the Crusades using both sailing ships and also galleys
known as taride or chelandre with the same bireme alla sensile oarage
system as the galea, in which both files of oars were rowed from
above deck, thus leaving the hold clear for cargo or horses.434
According to Thucydides, it was Perikle2s who first converted some
old ships, nh'e" (ne2es), into horse transports (ne2es hippago2gai) in 430
B.C.E. to transport cavalry[men], iJppei'" (hippeis), to the Pelopon-
ne2sos.435 In the context, these were most probably sailing ships rather
than trie2reis or other galleys. That they had to be converted indicates
clearly that these cavalry took their horses with them; however, it
appears that at this time the Athenians could not transport horses by
sea for long distances. Putting his own understanding into the mouth
of Athe2nagoras of Syracuse in a speech to the Syracusan Assembly,
Thucydides wrote that the Athenians would not be able to bring horses
with them for their invasion of Sicily in 415 B.C.E. They did transport
thirty hippeis; however, there is no suggestion that these took their
horses with them. Indeed, in initial battles the Athenians suffered
heavily because they had no cavalry to resist that of the enemy.
During the winter of 415-14 B.C.E. they sent to Athens requesting
more money and also cavalry. Two hundred and fifty cavalry were
sent with their equipment but without horses for it was expected that
horses would be obtained in Sicily.436 The Athenians did not attempt
to transport horses by sea all the way to Sicily because they did not
have the capability to do so.
Leo VI and Nike2phoros Ouranos included horse transports, nh'e"
iJppagwgoiv (ne2es hippago2goi) or ploi'a iJppagwgav (ploia hippago2ga),
alongside transport ships, forthgoiv (phorte2goi), or supply ships,
skeuofovra (skevophora), in a “baggage train”, tou'l do" (touldos) or
tou'ldon (touldon).437 They reserved the word “sailors”, nauvtai
(nautai), for the crews of these ships as though they envisaged them
being sailing ships rather than galleys.
In the Cretan expeditions of 911 and 949 large numbers of cavalry
were involved. The figures will continue to be debated but, whatever
------------------------------
434
See Chapter Six.
435
Thucydides, Peloponnesian war, II.56.1-2 (vol. 1, p. 356).
436
Thucydides, Peloponnesian war, VI.37.1, 43, 64.1, 71.2, 74.2, 93.4, 94.4 (vol. 3,
pp. 252, 262, 296, 308, 312, 354, 354-6).
437
See Appendix Two [a], §§11, 13, 23; Appendix Five, §§10, 11, 21.
306 CHAPTER FOUR
they actually were, there is no doubt that considerable numbers of

horses must have been transported, as can be inferred from the 40,000
modioi of barley collected for the expedition of 911.438 This can only
have been for horses and would have sufficed for around 245,000
horse-feed-days at sea or 185,000 horse-feed-days on campaign.439
With, say, a four-day voyage from Phygela to Crete, it would have
sufficed for 10,000 cavalry for the voyage and a campaign of around
15 days before re-supply or living off the land became necessary. In
all probability that number of cavalry would have been unnecessary
and the barley would have lasted much longer. 40,000 modioi
suggests a supply for an extended campaign by a smaller force.
In the inventories for the Cretan expeditions of 911 and 949 there is
only one reference to horse transports, probably because the compiler
had no documents from the departments of the logothete2s to2n agelo2n
and the logothete2s tou dromou, which would have been responsible
for supplying the ships.440 However, in the passage relating to an
undertaking by the strate2gos of Samos to provide a large quantity of
nails or spikes for the expedition of 911, an inventarist wrote:
“Concerning preparing a nail 5-fingers [long] for the fabric, strw'si"
(stro2sis), of the dromons, as regards the gangways, skavl ai (skalai),
and as regards the mangers, pavq nai (pathnai), 30,000, and they [i.e.,
the nails] should “go” [i.e., be sent] down to Phygela”.441
Skala, which was derived from the Latin scala, had long been
------------------------------
438
As discussed by Haldon in “Theory and practice”, pp. 288, 295, 299-301. See
Haldon, “Theory and practice”, p. 211: “peri; tw'n ojf eilovntwn eJtoimasqh'nai eij"
Qra/khsivou", h[goun tw'n kV ciliavdwn tou' kriqarivou ... peri; tou' devxasqaiv tina
basiliko;n to;n o[nta eij" Anatolikou;" eJtoimavsai kriqavrion ciliavda" kV ...”;
Constantine VII, De cerimoniis, II.44 (vol. 1, pp. 658-9).
The barley can only have been for horses since in both entries the inventory
went on to specify quantities of wheat, si'to" (sitos), biscuit, paxamavtion
(paxamation), and flour, ajreuvrion (areurion). These must have been for the crews of
the ships and the army. Cf. Haldon, “7. Bread for the army”, in “Theory and practice”,
pp. 294-302.
A diet of mainly barley would not have been good for horses for any length of
time, but for short traverses it would not have had an adverse effect. See Hayes,
Horses on board ship, pp. 167-79; Martin, Transport of horses, pp. 33-5; Smith,
Manual of veterinary hygiene, pp. 916-17. Cf. Hyland, Medieval warhorse, pp. 146-7.
439
The modios in question was most probably the the “sea” (thalassios) or
“imperial” (basilikos) modios, of around 16-17 litres. 40,000 modioi would be roughly
660,000 litres of barley. Medieval and other evidence suggests that consumption by
horses aboard ship was around 2.7 litres per horse by day and that that should be
increased by around a third for horses on campaign under moderate conditions of
work. See Pryor, “Modelling Bohemond’s march to Thessalonike2”, pp. 16-18.
440
Haldon, “Theory and practice”, p. 256. Cf. Constantine VII, Three treatises, pp.
161, 184.
441
See above p. 265 n. 339.
accepted into Greek for a gangway or boarding ramp, also giving rise
to the additional meaning of the word as a port or landing place.442 In
medieval Latin and Italian scala was used for a boarding gangway,
although the word pons, a bridge, was also used for ramps of horse
transports.443 Pathne2 for a manger or feed trough was a post-classical
form of favtnh (phatne2) for the same.444 The combination of pathnai
and skalai suggests strongly that horse transports were in question
here, that the inventorist used skalai for boarding ramps, and that at
least some dromons or chelandia were used as horse transports. It also
indicates clearly that the horses were brought overland to Phygela and
only there were the ships fitted out as horse transports.
What type of ships would the Byzantines have used for
transporting horses? On the one hand, sailing ships would have been
able to carry many more horses per ship, and to carry them more
efficiently, than galleys. On the other hand, sailing ships of any size
had severe limitations at destination. In the non-tidal Mediterranean,
they could not be beached without wrecking them. Galleys could be.
Sailing ships would be more suitable if an expedition’s destination
was a friendly port which had docks. Galleys would be more suitable
if the destination was a defended enemy coastline. Most probably
Byzantines used both sailing ships and galleys according to the needs
of the occasion as Western powers did later.
We know of only four pieces of Byzantine evidence for their use of
horse transports. First, Theophane2s the Confessor wrote that in 763
Constantine V put together a fleet of 800 chelandia carrying 12 horses
each for an expedition against the Bulgars.445 Later, in 773 or 774,
another fleet carrying 12,000 “cavalry”, kaballarikovn (kaballarikon),
was sent against the Bulgars; however, Theophane2s did not record
whether the horses went by sea also.446
Secondly, the anonymous Life of St Antony the Younger, who in
his earlier career was John Echimos, the ejk proswvpou (ek proso2pou),
acting strate2gos, of the thema of the Kibyrrhaio2tai, reported that
around 823-5 a large Muslim fleet of “trie2reis” attacked the capital of
------------------------------
442
See Kahane and Pietrangeli, “Cultural criteria”, p. 528; idem, Lingua Franca,
§841 (pp. 568-72); Jal, Glossaire nautique, p. 1357.
443
See Jal, Glossaire nautique, pp. 1198, 1323, 1357; Pryor, “Galleys of Charles I
of Anjou”, p. 55 & n. 67.
444
See also Haldon, “Theory and practice”, p. 270.
445
Theophane2s, Chronographia, A.M. 6254 (vol. 1, pp. 432-3): “th'/ de; iıV tou'
Iounivo u mhno;" ejxh'lqen oJ basileu;" ejpi; th;n Qrav/khn ajposteivla" kai; plwvi>mon dia; tou'
Eujxeivnou Povntou e{w" wV celandivwn ejpiferomevnwn ajna; ibV i{ppwn.”.
446
Theophane2s, Chronographia, A.M. 6266 (vol. 1, pp. 447-8).
308 CHAPTER FOUR
the thema, Antalya, and that 60 horsemen deployed from them. The
Muslim commander was mounted.447 By this time Muslims certainly
did have the capability to transport horses and cavalry[men] by sea,448
but the report in the Life reflected a Byzantine knowledge and
experience. Whether or not the author really intended to convey by his
use of the word trie2reis that the horses were transported on oared
galleys rather than generic “ships” of some kind is more
problematical. However, if the commander was mounted when he
disembarked, as the Life stated, then he must have done so from a
galley because sailing ships of any size could not be beached.
Thirdly, the author of the first part of the Theophane2s Continuatus
wrote that when Thomas the Slav advanced on Constantinople in 821,
he “equipped both bireme ships and other rounded corn-transporting
[ships] together with horse-transporting [ships], ...”, which he
assembled at Mityle2ne22 before advancing on Abydos. This account was
later repeated almost verbatim by John Skylitze2s. At folio 31 verso of
the Madrid manuscript of his Synopsis historio2n, an artist working in a
Byzantine style depicted Thomas’s fleet advancing on Abydos with
one of the ships carrying horses. This particular picture was almost
certainly copied from an original Byzantine one and thus represents
indirectly the only surviving Byzantine illustration of a horse
transport.449 In order to show the horses’ heads, the artist depicted
them over the gunwale, suggesting an open boat, but not revealing by
that anything more than artistic licence. However, it certainly is
significant that the ship was depicted as an oared galley and not as a
sailing ship. Byzantine galleys could certainly transport horses.
Finally, in his description of the last and finally successful assault
on Crete in 960-61 by Nike2phoros Pho2kas, Leo the Deacon described
the ramps used for unloading horses from the horse transports as
“gangways”, klivmake" (klimakes). He referred to the ships of the fleet
------------------------------
447
Vita Antonii junioris, p. 199.
448
It was certainly a commonplace by the ninth century. The anonymous but
contemporary author of the Life of Pope Sergius II (844-7) reported that in 846 the
Muslims who assaulted Rome with 73 ships brought 500 horses with them. Liber
Pontificalis, 104 (Sergius II), §44 (vol. 2, p. 99).
449
Theophane2s continuatus, II.13 (p. 55): “... nau'" te eJxartuvwn dihvrei" kai; eJt evra"
strogguvla" sitagwgou;" eJpomevna" aujtw'/ kai; iJppagwgouv", ...”; John Skylitze2s, Synopsis
historio2n, Micahvl oJ Traulov".7 (p. 32): “ ... kai; loipo;n aJdeiva" tucw;n dievq eto ta; kaq
eJauto;n krataiovt eron, nau'" eJxartuvwn polemika;" kai; eJt evra" sitagwgou;" kai;
iJppagwgouv". ...”.
The illustration in question is number seven of Appendix Seven, Table Ten,
reproduced in Estopañan, Skyllitzes Matritensis, fig. 68 (p. 246); Grabar and
Manoussacas, L’illustration, fig. 20; Skylitze2s, Suvnoyi" iJstoriw'n, fol. 31v; Tsamakda,
Ioannes Skylitzes, fig. 61.
by the classical term “fire-bearing trie2reis”, which he then glossed as

“dromons”. Then in the next sentence he said that the ramps were run
from “transports”, porqmeiva (porthmeia). The army was thus able to
Figure 34
The fleet of Thomas the Slav advancing on Abydos and carrying horses
aboard a galley in the Synopsis historio2n of John Skylitze2s (Madrid,
Biblioteca Nacional, vitr. 26-2, fol. 31v), ca 1160, based on an eleventh-
century original.
be landed, armed and mounted, from the sea to the land.450 It is

unclear whether the porthmeia were dromons or whether they were
some other ships; however, the fact that they were able to close with
the shore and unload the horses across ramps means that they were
galleys rather than sailing ships. Leo’s klimakes were no doubt the
same as the skalai of the inventory for the Cretan expedition of 91l.
There can be no doubt that even if Byzantines did use sailing ships
as horse transports, by the ninth or tenth centuries at least, they were
also capable of transporting them on galleys: dromons or chelandia.
Ramps had been used since antiquity for the loading and unloading
of horses. Two early fourth-century mosaics from the Dermech
district of Roman Carthage and from Piazza Armerina in Sicily,
------------------------------
450
Leo the Deacon, Historiae, c. 3 (p. 7): “kai; tacuplohvsa", purfovrou" te trihvrei"
pleivsta" ejpagovmeno" (drovmwna" tauvta" ÔRwmai'oi kalou'si), th'/ Krhvth/ proswvrmisen.
ejpei; de; th'" ajpobavsew" ejdovkei kairo;", e[d eixe praktikw'", h}n ei\cen ejmpeiriva n e[rgwn
polemikw'n. klivmaka" ga;r ejpi; tw;n porqmeivwn ejpiferovmeno", tauvta" ejpi; th;n hji>ovna
prosufaplw'n, th;n strati;a n e[noplovn te kai; e[f ippon ajpo; th'" uJgra'" ejpi; th;n xhra;n
metebivbaze.”.
310 CHAPTER FOUR
possibly created by the same team of craftsmen, show hunters

escaping pursuing beasts by riding up ramps into galleys. In the case
of the Dermech mosaic (Figure 35) the ramp leads to the bow of the
ship but in that of Piazza Armerina (Figure 36) it is at the stern. In both
Figure 35
Mosaic of a galley from the Dermech district of Roman Carthage, early
fourth century.
cases the ramps appear to be being hauled inboard by the crews over
the gunwales rather than through any port in the hulls. As long as the
ramps were run over the gunwales, they could no doubt be put
anywhere. Another mosaic from Piazza Armerina of a galley loading
exotic animals, presumably for the games, has ramps at both ends.451
In such cases where ramps were run over the gunwales and horses
were loaded into and out of the ships in that way, the galleys must
either have been open boats without decks or else they must have had
large hatches in the deck with gently sloping ramps or brows leading
down into the holds.
What is primarily at issue here is the question of in what ways
horses could possibly have been transported on dromons or chelandia
of the Macedonian era. Were they already equipped with the famous
ports in the hull at the stern which Western transport galleys, known
as taride or chelandre, had by the thirteenth century at least, and
------------------------------
451
See Mahjoubi, “Nouvelle mosaïque”, esp. plate p. 265; Dunbabin, Mosaics of
Roman North Africa, pl. XIII, N o 26 [incomplete, does not show the stern of the ship];
Pace, Mosaici, fig. 25 [incomplete, shows all of the ship but only part of the ramp and
horse]; Casson, Ships and seamanship, fig. 141.
through which the 30 or even 40 horses and cavalry which they could
carry could be embarked, and disembarked already mounted, via
landing bridges thrust out from the ports?452
Perhaps the most well known account of such an action is Robert
Figure 36
Mosaic of a galley from Piazza Armerina, Sicily, early fourth century
of Clari’s description of the landing of the Fourth Crusade outside

Constantinople in July 1203: “..., then the fleet landed, and when they
had arrived [at the shore], the knights issued forth from the horse
transports (uissiers ) all mounted; for the horse transports were made
in such a way that they had a door (wis) that one could open easily;
and then one thrust out a bridge by which the knights could issue forth
onto land all mounted.”453
------------------------------
452
See Pryor, “Transportation of horses by sea”, esp. pp. 23-4, 103-120. The
conclusions reached in this study were later much modified in “Naval architecture
revisited”, pp. 255-9. See also idem, “Crusade of Emperor Frederick II”, pp. 124-7;
idem, “From dromo2n to galea”, pp. 115-6.
453
Robert of Clari, Conquête de Constantinople, §43 (pp. 161-2): “..., tant que li
estores arriva, et quant il furent arrivé, si issirent li chevalier hors des uissiers tot
monté; que li uissier estoient en tele maniere fait que il i avoit wis que on ouvroit
bien, si lanchoit on un pont hors, par ou li chevalier pooient issir hors a tere tot
monté.”. Note that although Robert of Clari implied that the word (h)uissier was
312 CHAPTER FOUR
Such ports, without landing bridges, can be seen in the fourteenth-

century Paris, Bibliothèque Nationale manuscript of Les livres des
histoires du commencement du monde. [Figure 37] The wreck of what
was most probably a horse transport of this type dated tentatively to
the twelfth century was discovered off Camarina, Sicily, in 1989.454
Alexandres claims that dromons of the Macedonian era could have
had such ports and that they were known as “openings”, qurivde"
(thyrides).455 However, the only Byzantine usage known to us of the
word quriv" (thyris) for such ports occurs in a poem of Manganeios
Prodromos describing the Norman Sicilian fleet supposedly
demonstrating before Constantinople, probably in 1157.456 Whether
the usage reflected a Byzantine experience is arguable; although, it
probably did from the ninth or tenth centuries at least because landing
bridges to the shore would have required them.
Western horse transports of the thirteenth century were
significantly larger than Byzantine dromons or chelandia of the tenth
century. They were also shorter but beamier and deeper than
contemporary Western war galleys, galeae. In contracts for the
construction of taride to carry 30 horses each for Charles I of Anjou,
King of Sicily, in the 1270s and 1280s it was specified that they
should have an overall length of 18 Neapolitan canne (37.97 metres)
as opposed to the 18.75 canne (39.55 metres) of contemporary galeae
and the approximate 31.25 metres calculated above for dromons.
Their maximum beam amidships at the deck beams was either 18.5 or
19 palmi (4.88 or 5.01 metres), as opposed to the 17.5 palmi (4.61
metres) of galeae and the approximate 3.80 metres which we have
calculated above for straight-hulled dromons and 4.46 metres for
flared-hulled dromons. But in the context here the latter dimension is
irrelevant because what is important is the dimension between the
centre line and the hull at the level of the lower oarsmen. Their beam
------------------------------
derived from Old French w(u)is for a door, in fact it was derived from the Arabic
‘usha2rı3 for a transport galley. See p. 258, n. 319 above.
454
Di Stefano, “Antichi relitti”, pp. 130-34.
455
See Alexandres, ÔH qalassiva, p. 74. Alexandres provides no evidence for his
claim.
456
We are indebted to Michael Jeffreys for the following text of his forthcoming
edition (with Elizabeth Jeffreys) of the poems of Manganeios Prodromos. Manganeios
Prodromos, Poem 11, ll. 158-60:
“Ouja i; th'/ povlei! Pevfqaken eij" ta;" Kriqa;" oj stovlo",
kai; to; poso;n tw'n trihrw'n tri;" eJkato;n trihvr ei",
pro;" touvtoi" a[katoi pollai; kai; quvrai tai'" ajkavtoi",
...”
The previous edition is Manganeios Prodromos, Poems, poem 12.
Cf. above p. 114 & n. 206.
Figure 37
Horses unloaded from ports at the sterns of galleys in a manuscript of Les
livres des histoires du commencement du monde (Paris, Bibliothèque
Nationale, MS. Fr. 301, fol. 58v), fourteenth century.
on the floor (planum), was either 13.5 or 14 palmi (3.56 or 3.69

metres), as opposed to the 11.25 palmi (2.97) metres of galeae
amidships. Their depth in hold amidships was 8 palmi (2.11 metres) to
314 CHAPTER FOUR
the actual deck as opposed to 2.04 metres for galeae and 7.5 palmi
(1.98 metres) below the deck beams as opposed to the 1.75-1.8 metres
calculated above for dromons. The horses were stabled fore-and-aft in
groups of three abeam over a length of 12 canne (96 palmi, 25.31
metres) in the hold. Each group of three was allocated a space of 7.5
palmi (1.98 metres) with a “cat[h]ena mortua” half a palmus (13
centimetres) wide and twice as deep between each group of three. The
ports in the stern quarters were 8.5 palmi (2.24 metres) high by 5.5
palmi (1.45 metres) wide and embarkation bridges of the same width
and 14 palmi (3.69 metres) long could be thrust out from them.457
Taride constructed by Genoa in 1246 for the Crusade of Louis IX
of France had even more depth in hold, 2.23 metres, although they
were somewhat shorter (35.71 metres) and their beam at the deck
amidships is not known.458 On sailing ships, the Statutes of Marseilles
of 1253 specified that each horse should be allowed a space three
Marseillese palmi (75.6 centimetres) wide.459
------------------------------
457
Filangieri, Registri, vol. 12, pp. 161-3, 175-6, 242-5; vol. 13, pp. 242-3; vol. 18,
pp. 302-5; vol. 24, pp. 33-7.
The surviving versions of these documents are transcriptions made by
nineteenth- and twentieth-century historians; Giuseppe del Giudice, Camillo Minieri
Riccio, and Erasmo Ricca. The registers were destroyed during the allied invasion of
Italy in the Second World War.
The script used in the Angevin chancery was a highly abbreviated late medieval
chancery gothic minuscule which was difficult to read, especially when it came to
technical terminology. All historians who made transcriptions had difficulty with the
technical terminology. None of the transcriptions is accurate. However, the
documents followed a common form and the following is based on vol. 12, pp. 242-3
with emendations based on readings from the other documents: “Quelibet terida erit
longitudinis cannarum XVIII [et altitudinis palmorum VIII], ... [Item a tabula sentine
usque ad tabulam cohoperte altitudinis palmorum VIII.] ... item in plano latitudinis
palm[orum] XIII et medii [XIV]; ... item debet esse altitudinis a paliolo ubi equi
debent tenere pedes palm[orum] VII et med[ii de canna] in minori vel minus basso
loco teride subtus laccas; [ita] quod grossicies laccarum non comprehendatur in isto
numero; ... item fiat porta una in puppi [cuiuslibet teridarum] pro introitu et exitu
hominum et equorum, que porta debet esse altitudinis palm[orum] VIII et med[ii] et
amplitudinis palm[orum] V et med[ii], ...item quelibet terida sit rotunda in puppi ad
modum conduri ad hoc quod equus possit intrare et exire insellatus et armatus; item in
laccis de punta in puntam sit longitudinis [latitudinis] palm[orum] XIX [XVIII et
medii]; ... item in qualibet terida sint impaliolate canne XII pro equis recipiendis,
numerando a puppi usque proram, ... item de VIII ad VIII palmos sit catena una
mortua, que sit altitudinis in duplum quam in latitudine, ita quod infra duas cat[h]enas
sint equi tres, cum equi III debeant morari infra palmos VII et med[ium], et reliquo
medio palmo erit cat[h]ena; et sic oportet poni cat[h]enas [quod] infra X cannas
longitudinis morari possint ad minus habiliter et bene [in terida ipsa] equi XXX; ...
item pontem unum pro recipiendis equis, latitudinis sicut est porta teride et
longitudinis palm[orum] XIV; ...”
For the dimensions of galeae see Pryor, “From dromo2n to galea”, pp. 110.
458
See Pryor, “From dromo2n to galea”, p. 115.
459
Pernoud, Marseille, IV.25 (p. 158): “... et pro equo detur platea in latitudine 3
By the thirteenth century all Western galleys used the bireme

oarage system with both oars rowed from above the deck: the alla
sensile system. There were no oarsmen below deck.460 Consequently,
the whole of the holds could be used for stabling horses. Was this the
case for Byzantine dromons or chelandia when used as horse
transports? Assuming for the moment that such ships used for
transporting horses were of the same design as battle dromons as
reconstructed above, did the Byzantines remove the lower oar-bank
and dismantle the thwarts? Or did they somehow try to fit the horses
in below deck, leaving the lower oar-bank in place? They certainly
would not have tried to stable horses above deck because, first, the
weight of horses placed so high above the centre of gravity of such
shallow-drafted and narrow ships would have created severe problems
of stability. Secondly the animals would have been hopelessly
exposed to missile attack by an enemy. But thirdly, and most
importantly, the rolling of the ships would have endangered the
animals themselves and made them extremely prone to injury. Horses
were always shipped as close to the keel as possible for their own
safety, even into the nineteenth century.461
If the lower oar-banks were removed, the ships would have been
left with only one oar-bank and could not possibly have kept up with
the fleet when under oars. Tactically, this would surely have been very
dangerous. It is true that the Strate2gikon of Maurice, and Nike2phoros
Ouranos following it, did say that the touldos could not keep up with
the dromons.462 However, even if we can accept that transports and
horse transports would normally be heavier and slower than battle
galleys, we find it difficult to accept that ships whose purpose was to
assault enemy coastlines would have been so disabled as to remove
half their motive power. If the lower oar-bank was not removed, then
any horses would have had to have been stabled in a row down the
centre lines of the ships between the two files of oars.
Now, horses cannot be seasick, cannot vomit, but they are affected
------------------------------
palmorum; ...”.
460
See below, Chapter Six.
461
See Shirley, Transport of cavalry, p. 33. The projected conversion of a trie2re2s to
a horse transport as envisaged by the architects of the Olympias project, with the
horses stabled abeam at the level of the zygian oars, would have been completely
insane. The requirements of the horses themselves were completely ignored. See
Morrison, et al., Athenian trireme, 227-30.
462
Maurice, Ek tou Maurikiou, §5 (pp. 41-2): “Epeidh; de; aiJ sagh'nai kai; ta;
forthga; ploi'a kai; baruvt era ouj duvnantai sunakolouqei'n toi'" drovmwsin, ...”. Cf.
Maurice, Strate2gikon, XIIB.21.21-3 (p. 468); Nike2phoros Ouranos, Ek to2n taktiko2n,
§122.5 (p. 101).
316 CHAPTER FOUR
badly by sea travel. For transporting horses, ships need to be as stable

as possible and minimizing the effects of their pitching and rolling on
the horses is very important. They should be as beamy as possible and
have as great a tonnage as possible.463 Long, low ships such as
medieval galleys, designed to cut through the water rather than to ride
the waves, would have been far less susceptible to pitching than to
rolling and, therefore, one might have expected that horses would
have been stabled abeam since they do have the ability to brace
themselves against any up-and-down movement between their heads
and rumps by spreading their legs.464 On British troop transports of the
late-nineteenth and early-twentieth centuries they were stabled abeam
since the ships used for oceanic transport at that time also rolled more
than they pitched.465 However, the historical evidence is clear that in
the Middle Ages horses transported on medieval Mediterranean horse-
transport galleys were stabled fore-and-aft rather than abeam.466 There
can be little doubt that the mysterious “cat[h]ena mortua”, once
thought to have been a “dead” or “standing” beam, was in fact a
manger or feed trough. The words found their way into the modern
record in the form that they have through a tortuous transmission
process reaching back through vernacular South Italian to the Greek
phatne2 and pathne2 for a feed trough or manger.467 Identification of
------------------------------
463
On sea sickness see Hyland, Medieval warhorse, p. 148.
The British government allowance in the late nineteenth and early twentieth
centuries was 10 tons (10.16 metric tonnes) gross register tonnage per horse and man.
See Hayes, Horses on board ship, p. 29; Martin, Transport of horses, p. 13; Smith,
Manual of veterinary hygiene, p. 899. Medieval oared horse transports, no matter of
what vintage, could have come nowhere near this figure.
Veterinary Lieutenant Martin of the British army Veterinary Department
recommended that no ship should be used for transporting horses which did not have
bilge keels to counteract rolling. See Martin, Transport of horses, p. 3.
464
Martin, Transport of horses, pp. 23-4; Smith, Manual of veterinary hygiene, pp.
908-9.
465
Martin, Transport of horses, p. 14; Smith, Manual of veterinary hygiene, pp.
899-900, 905, 908-9.
466
See Pryor, “Naval architecture revisited”, esp. pp. 258-9 and cf. Harris,
“Frederic af Chapman”. See also Pryor, “Transportation of horses by sea”, pp. 114-
116 (esp. docs III & V) and nn. 88-92; idem “Naval architecture revisited”, esp. pp.
258-9; idem, “From dromo2n to galea”, p. 116. Much of the reconstruction of stabling
in these three studies is now obsolete.
467
Contracts for the construction of taride (and also galeae) recorded by chancery
scribes for the Angevin court contained technical terms in Latin forms which were
Latinizations of vernacular South Italian, which was heavily influenced by Greek as a
legacy of the long occupation of South Italy by Byzantium and the close contact
between Byzantine-Greek society and that of South Italy.
No doubt the original vernacular term which the historians transcribed as catena
or cathena was something like pathena. Publius Vegetius, the author of the
Mulomedicina, who was probably the same man as Flavius Vegetius Renatus the
cat[h]ena as a manger proves in fact that the horses were stabled fore-
and-aft since no other arrangement would have been possible. In the
Byzantine case, of course, unless the lower oar-bank was removed, the
horses could not possibly have been stabled abeam on dromons or
chelandia.
By the early twentieth century it had been learned that using slings
under the bellies of horses in horse stalls was not a good idea and their
use was discontinued.468 However, there can be little doubt that such
under-belly slings were used on medieval and later horse transports.
Documents from the thirteenth-century Angevin registers referred to
ring bolts (anuli) or belaying cleats (castaneole) from which horses
were “suspended” on horse-carrying taride, and they could only have
------------------------------
author of the Epitoma rei militaris, used patena for a manger. See, Vegetius,
Mulomedicina, I.56.3-4 (p. 81): “Patena quae apellatur [illa quae apellatur patena,
MS. P], hoc est alveus ad hordeum ministrandum, sit munda semper, ne sordes
aliquae cibariis admisceantur et noceant; loculis praeterea vel marmore vel lapide vel
ligno factis distinguenda est, ut singula iumenta hordeum suum ex integro nullo
praeripiente consumant.”.
What “mortua” may have been originally is unknown. All the transcriptions
have the word in this form; however, we suggest that the original abbreviated form
may have been intended to represent a word such as “maniura/maniera/
manieria/maneria/maniaora”, varieties of the same word used in thirteenth-century
Latin for a manger or nose bag for horses; pathena and maniura/maniera/manieria/
maneria/maniaora thus being used in apposition. Anyone familiar with late
thirteenth-century notarial and chancery scripts will know that to read “mortua” for
“maniura” is not as far-fetched as it seems.
The only problem remaining with this reconstruction is the very small width of
the troughs. It is hard to envisage how horses could have got their muzzles into
troughs as narrow as 13 centimetres in order to eat. Captain Hayes recommended that
mangers or feeding troughs should be at least 13 inches (33.02 centimetres) wide at
the top. He recorded that at the turn of the nineteenth and twentieth centuries the
British Admiralty specification for the width of feed troughs was 12 inches (30.48
centimetres). See his Horses on board ship, pp. 95 & 130. However, if the horses
were face to face and shared feed troughs, then these would have been 26 centimetres
wide and that would have been at least adequate. This is probably the solution to the
problem.
468
Martin, Transport of horses, pp. 23-6; Smith, Manual of veterinary hygiene, pp.
911-12.
The history of this issue is obscure. As early as the late eighteenth century some
veterinarians and cavalry officers were beginning to question the use of slings. Philip
Astley, the eighteenth-century cavalry sergeant-major and circus master was opposed
to the use of them for transporting horses by sea. In 1797 a London hay and corn
merchant, S. Lawson, published a book in which he said that animals were safer and
better off on their feet in pens rather than in stalls in slings. See Smith, Early history,
vol. 2, pp. 135, 229. In 1848 the veterinary surgeon J. S. Mellows expressed his
opinion that if horses were actually suspended off their feet it impeded evacuation,
contributing to the disease known as ship staggers. See Mellows, “Observations”, pp.
102-3 and cf. Shirley, Transport of cavalry, pp. 23-6. Nevertheless the practice of
transporting horses by sea in stalls with under-belly slings continued routinely until
the Boer War.
318 CHAPTER FOUR
been suspended by under-belly slings of some sort.469 Amongst the list

of equipment that keepers of arsenals, tarsienatus, were held
responsible for in a formulary for their appointment, there was a
reference to lanzones for transporting horses upon the sea. In this form
the word is unknown in medieval Latin; however, it must have been a
Latinization of some vernacular word such as lenzo, meaning a cloth
sheet or sling.470 There is also an unverifiable record of a clause in
another document referring to a “belt of cloth and rope for placing
under the stomachs of horses”.471
Knowledge of such matters has advanced greatly in recent years
and it is now known that the slings would not have suspended the
horses off their feet for more than short periods. That would have
killed them. Horses can be suspended off their feet for short periods,
for example, for loading and offloading, and during operations and
convalescence,472 but not for long periods. The horses would still have
had their feet on the floor but would have been restrained against
being thrown off their feet or attempting to lie down by slings which
were braced up firmly under their bellies but did not lift them off their
feet.473 That this was the case is suggested by Ambroise’s report that,
------------------------------
469
Five separate documents in the registers, transcribed by four different historians,
referred to these ring bolts (anuli) or belaying cleats (castaneole) from which horses
were “suspended”. See Filangieri, Registri, vol. 12, p. 162: “ ... Item debet habere
anulos necessarios pro ligandis et appendendis equis ipsis”; ibid., p. 176: “... item
debet habere castaneolas necessarias de ligno pro ligandis et appendendis equis; ...”;
ibid., p. 244: “Item debet habere castanidas de ligno necessarias pro ligandis et
appendendis equis; ...”; vol. 18, p. 304: “Item debent habere castaniolas de ligno ...
pro ligandis et appendendis equis”; vol. 24, p. 36; “item debet habere castinuolas de
ligno necessarias pro ligandis et apponendis equis”. Apponendis in the last document,
transcribed by Ricca, is a clear misreading of appendendis.
470
Filangieri, Registri, vol. 31, p. 65: “… lanzones pro equis transvehendis super
mare ...”. The document is in the manuscript, Formularium curie Caroli Secundi regis
Sicilie, which was a copy made for the Papacy of what was Register Nine of the
reconstructed registers of Charles II. The manuscript was compiled in 1306-7 but
from much older materials.
471
Unfortunately, when Bevere cited a clause from a document in the Angevin
registers which read: “Cynta de tela et cordis ad ponendum sub ventribus equorum.”,
he gave no reference. See Pryor, “Transportation of horses”, p. 113; Bevere, “Ordigni
ed utensili”, p. 720. A systematic search of the reconstructed Angevin registers has
failed to locate this citation; however, there is no reason to doubt Bevere’s veracity.
472
See, for example, the illustrations from a manuscript of an Italian translation of
the Medicina equorum of Jordanus Ruffus (ca 1250), from the fourteenth-century
Libro de Menescalcia de albeiteria et fisica de las bestias of Juan Alvares de
Salamiella, and from J. B. von Sind’s, L’art du manège pris dans vrais principles,
suivi d’une nouvelle méthode pour l’embouchure des chevaux (Bonn, 1762), in
Dunlop and Williams, Veterinary medicine, pp. 186, 226, 335. Cf. Hyland, Medieval
warhorse, p. 145.
473
Hayes, Horses on board ship, pp. 143-4; Martin, Transport of horses, p. 23;
Shirley, Transport of cavalry, 24.
after landing, the horses transported by Richard Coeur de Lion from

Messina to Cyprus for the Third Crusade in 1191 had to be: “...
exercized, / for they were all benumbed / and dazed and exhausted /
From the month that they had been on the sea / And always without
being able to lie down”.474 The text suggests that the horses were
prevented from lying down by slings but were on their feet, not
suspended off them. In 1340 the English were still using canvas slings
of some sort in the stalls of horse transports,475 and the practice
continued for centuries. It appears that on Mediterranean transport
galleys the horses would have been able to brace themselves against
the minor amount of pitching that occurred but that slings were used
to help them resist the effects of the ships’ rolling and to prevent them
trying to lie down.
Also contrary to earlier opinion, it is now known that medieval
warhorses of the Byzantine and Crusader periods, before the days of
heavy plate armour, were not especially large animals by modern
standards.476 Because the builders and fitters-out of ships would not
have allowed more space for horses than was necessary, some
estimate of their size can be gained from the dimensions specified for
the stabling arrangements and the stern quarter ports on horse
transports of the thirteenth century. Presumably the width of 75.6
------------------------------
Captain Hayes and lt colonel Shirley considered that slings were of use only in
fine weather to enable horses to ease part of their weight on them. They should be
loose enough to allow the flat of a hand to pass between the sling and the belly and
should not exert pressure on the belly or chest.
474
Ambroise, L’estoire de la guerre sainte, ll 1565-1576 (coll. 42-3):
Li reis la nuit sanz plus targer That night, without delaying further, the king
Fist tanz de chevalz descharger Had as many horses disembarked
Cum enz es eneques avoit. As he had in his eneques.
L’empereres mot ne savoit The emperor knew nought
Qu’il en eust nul amené. That he had brought any of them.
Li cheval furent demené, The horses were exercized,
Car il erent tut engurdi For they were all benumbed
E deboistié e esturdi and dazed and exhausted
D’un mois qu’orent en mer esté From the month that they had been on the sea
E sanz jesir toz jorz esté. And always without being able to lie down.
Sanz plus de sejor qu’il eussent, Without more of the rest that they ought,
Que par raison aveir deussent, … Which by right they should have had, …
The story was repeated in the Itinerarium peregrinorum et gesta regis Ricardi.
See Itinerarium peregrinorum (Stubbs), II.33 (p. 192): “Nocte quoque eadem, rex in
papilionibus suis moratus, fecit equos suos educi ab esneckis. ... Verum quoniam equi
nostri vexatione marina per unum mensem semper stando, plurimum fuerant turbati,
equis ipsis parcentes modestius insecuti sunt, ...”
475
See Hewitt, Organization of war, Appendix I (p. 180).
476
Davis, Medieval warhorse; Hyland, Medieval warhorse.
Disregard now the assumptions of Pryor on the size of horses during the age of
the Crusades in “Transportation of horses by sea”, p. 106.
320 CHAPTER FOUR
centimetres specified in the statutes of Marseilles was just enough to

stable a horse and to allow a groom to squeeze around it when
necessary. Then, the minimum depth in hold below the deck beams of
Angevin taride was 1.98 metres and this must have been considered
the minimum height necessary for the horses being used. Each group
of three was also allocated a length space of 1.98 metres. Such
dimensions actually accord quite closely to those specified by the
British Admiralty and advocated by veterinary surgeons at the turn of
the nineteenth and twentieth centuries.477 The height of the ports at the
stern, 2.24 metres, was presumably just enough to allow mounted
cavalry to disembark when leaning forward along a horse’s neck. This
data suggests that the warhorses used in the Mediterranean in the
thirteenth century stood around 15 hands (1.524 metres) at the
withers, were around 1.825 metres at the ears when in a normal erect
stance, 66 or more centimetres wide in the belly or barrel depending
on condition, and around 1.90 metres from nostril to the point of the
hock.478
Figure 38 shows immediately that horses of this kind could not
possibly have been stabled below the deck of dromons or chelandia as
we have reconstructed them with the lower bank of oarsmen still in
place. Between the inner shoulders of the oarsmen below deck there
------------------------------
477
In 1901 veterinary lieutenant Martin wrote that British Government regulations
permitted three sizes, all 26.5 inches wide (67.31 centimetres) wide, but 6 feet 1 inch
(1.855 metres), 6 feet 5 inches (1.955 metres), and 6 feet 9 inches (2.06 metres) long.
He himself recommended stalls 8 feet (2.44 metres) long by 2 feet 4 inches (71
centimetres) wide with the height of the tweendeck to the beams of the deck above at
least 7 feet 3 inches (2.16 metres) and from deck to deck at least 8 feet (2.44 metres).
See his Transport of horses by sea, pp. 6-7, 14-16.
In 1902, according to captain Hayes, the Admiralty specifications were:
minimum clear length between breast and haunch boards, 6 feet 9 inches (2.06
metres); minimum clear breadth between the side division boards, 2 feet 4 inches (71
centimetres). See his Horses on board ship, p. 127. Hayes himself recommended that,
for horses of ordinary size (15 hands), the stall should be no less than 6 feet 6 inches
(1.98 metres) long, but preferably up to eight feet (2.44 metres). The height from deck
to the beams of the deck above should be at least 6 feet 9 inches (2.06 metres), of
which 3 inches would be taken up by the false floor on which the horses stood,
leaving 6 feet 6 inches (1.98 metres) clear height. See his Horses on board ship, pp.
31 & 124-5.
In 1905, Colonel Smith, who was mostly following lieutenant Martin, simply
recommended a deck height of 8 feet (2.44 metres), but he acknowledged that it was
frequently less than that on horse transports. He said that the official British
government stall length was 6 feet to 6 feet 6 inches (1.82-1.98 metres). However, he
considered this inadequate and recommended a length of 7 feet 6 inches to 8 feet
(2.29-2.44 metres) and a width of 2 feet to 2 feet 4 inches (0.61-0.71 metres). See his
Manual of veterinary hygiene, pp. 900, 909.
478
See the calculations of Ann Hyland using her own horses in her Medieval
warhorse, pp. 143-8.
Figure 38
Horse transports of the era of the Macedonian emperors, I: standard bireme
dromon with a fifteen-hand horse.
© John H. Pryor
would have been around about 75 centimetres clearance amidships,

just enough one might suggest to stable a horse 66 centimetres in the
barrel. However, grooms would have no way of getting to the horses
to tend to them and the headroom in the hold would have been at least
10-15 centimetres too little, even if measured from the floor to the
deck beams. As the hulls narrowed towards the bow and stern, the
clearance between the oarsmen would have decreased and it would
have become impossible to stable horses towards the extremities of
the hulls.
Moreover, horses could not have been stabled directly on the floors
of the ships. They slip around on their hooves and defecate and urinate
in large volumes. They would have had to have been stabled on false
floors of battens and planks, perforated to allow urine to run to
collection points for bailing out.479 The headroom would in fact have
------------------------------
479
Captain Hayes, quoting British Admiralty regulations said that the “platforms”
should made with 1.5 inch deal boards with one inch spaces between them laid on 2-
322 CHAPTER FOUR
been at least 20-25 centimetres too little. One might suggest that
Byzantine cavalry horses were smaller than 15 hands or that the depth
in hold of dromons may have been greater than that which we have
calculated, or a combination of both. The first alternative is a
possibility; however, in order to fit the horses in, even amidships, they
would have to have stood only around 12.15 hands. They would have
been unusually small, even for Roman horses, mere ponies in fact.480
The second alternative would not have been possible without altering
the ships in fundamental ways.
If dromons and chelandia were used as horse transports, and
chelandia at least certainly were, then the ships must have been
specially constructed to carry horses and have been different to battle
dromons. If a lower oar-bank was left in place, the ships would have
had to have been significantly wider in the beam. Whatever the case,
they would have had to have been deeper in the hold.
In fact their dimensions must have been entirely different. On the
taride of Charles I of Sicily the length of stalls for the horses was 1.98
metres. So, if the length of chelandia or dromons on the floor was
around 25 metres or a little more, as it must have been in order to
accommodate the 25 oarsmen of the lower oar-banks, then around 12
horses seems just about right. This does at least make some sense of
Theophane2s the Confessor’s figure of 12 horses per chelandion. At the
very least, the coincidence is striking. However, the taride of Charles
I of Sicily were shorter but beamier and deeper than the war galleys,
galeae, and we should assume that the Byzantines similarly modified
the design of dromons or chelandia for carrying horses and make any
comparison to Angevin taride rather than galeae. The starting point
has to be sufficient beam and depth in hull to stable a row of 15-hand
horses down the centre line.
A 15-hand horse will stand around 1.825 metres tall at the ears
------------------------------
inch battens, the platform thus being 3.5 inches (8.89 centimetres) thick. See his
Horses on board ship, p. 129. Colonel Smith, following lieutenant Martin, said that on
British horse transports these “foot-boards” were made of planks set an inch apart to
allow urine to run off, resting on battens to keep them off the deck, and crossed on the
top by additional battens to prevent the horses from slipping. See Smith, Manual of
veterinary hygiene, p. 906; Martin, Transport of horses by sea, pp. 19-21. Lt colonel
Shirley recommended 12-18 inches of shingle flooring rather than planks and battens.
See Transport of cavalry, pp. 26-9.
On Angevin taride a length of 12 canne was impaliolate, that is provided with a
false floor, paliolus, on which the horses stood. This was made of oak to resist the
wear and tear of hooves. See n. 457 above and Pryor, “Naval architecture revisited”,
pp. 257-8. Such false floors must have been at least 10 centimetres in height.
480
The archaeological evidence collected by Hyland suggests a range from around
13.5 hands to 15.25 hands for Roman horses. See Hyland, Equus, p. 68.
Figure 39
Horse transports of the era of the Macedonian emperor, II: modified
dromon/chelandion with a fifteen-hand horse.
© John H. Pryor
when naturally erect, so the depth in hull must have been at least 1.86
metres below the deck beams, plus around 10 centimetres for the false
floor; say 1.95 metres in all. The width of three Marseillese palmi, or
75.6 centimetres, specified in the statutes of Marseilles should be
understood to have been the internal width of a stall. At just under 2
feet 6 inches, this equates closely, if a little generously, to what was
the recommended width for stalls on British transports at the turn of
the nineteenth and twentieth centuries. The horses could not have been
confined too closely for health reasons.481 They would also have had
to have been separated from the oarsmen by strong rails set in upright
stanchions in order to prevent injuries to themselves and the crews
from them kicking out or, even worse, breaking loose if panicked and
causing dangerous chaos in which both horses and men would have
been liable to be severely injured. The stanchions at the corners of the
stalls of horses on British transports at the turn of the nineteenth and
------------------------------
481
324 CHAPTER FOUR
twentieth centuries were 4 inches (10.16 centimetres) wide.482 This

suggests that a minimum beam for stabling horses fore-and-aft aboard
medieval transport galleys should have been around a metre. Allowing
a minimum ten centimetres or so clearance between the oarsmen and
the stall stanchions would mean that the oarsmen would have had to
have been around 1.20 metres apart.
The beam of horse-transporting chelandia or dromons may have
been around 4.85 metres at the deck amidships, their depth in hold
around 1.95 metres, and their beam on the floor around 1.20 metres.
That would just about make it possible to have a file of horses of
around 15 hands stabled down the centre. But conditions for both the
horses and the oarsmen must have been not only “aromatic” but also
dangerous. It is very hard to imagine how horses could have been
transported for anything more than very short distances under such
conditions. Perhaps this explains why the inventories for the
expeditions to Italy in 934 and 935 included in the De cerimoniis,
mentioned cavalry[men] but made no explicit mention of horses.483
The imperial authorities must have expected Hugh of Provence, the
King of Italy, to supply the horses for the cavalry[men] which they
were sending to him to assist in his campaign against the Lombard
princes in South Italy?
A final consideration is the sheer weight of the horses. Medieval
war horses of around 15 hands would have weighed around 550
kilogrammes when in good condition. Twelve horses would have
weighed around 6.6 metric tonnes. Together with their equipment and
supplies, they would have weighed at least 8.0 tonnes if food and
water for any extended period were taken aboard. With ships of
deadweight tonnage as low as that we have calculated for dromons,
that dimension of extra cargo capacity could not easily be found
unless other accomodations were made. If the lower bank of oarsmen
------------------------------
482
Captain Hayes, quoting Admiralty regulations, said that the upright stanchions at
each corner of stalls should be 6 by 4 inches (15.24 by 10.16 centimetres). See his
Horses on board ship, pp. 127-8. Robustness of the stalls was emphasised by both
lieutenant Martin and colonel Smith. Martin, Transport of horses, pp. 18-21; Smith,
Manual of veterinary hygiene, pp. 905-10.
483
Haldon, “Theory and practice”, pp. 213, 215 and commentary at p. 257. Cf.
Constantine VII, De cerimoniis, II.44 (pp. 660-62).
Runciman related some passages in the De administrando imperio which
referred to the thema of the Peloponne2sos providing 1,000 horses for some service in
“Lombardy” to the expeditions of 934 and 935. Runciman, Romanus Lecapenus, p.
74. However, mentions in the text of the proto2spatharios John Pro2tevon as strate2gos
of the thema date the reference of the passages to 921. See Constantine VII, De
administrando imperio, §§50.26-7, 51.199-204, 52.1-15; Volume II: commentary, p.
204.
was left in place, the hull must have had to have been designed very
differently so that the extra weight could have been borne without
affecting the level of the water line. Most importantly, that itself could
not have been changed without a complete reworking of the oarage
systems for both banks of oars.
We conclude that, at least originally, chelandia were specialized
horse transports, the name applied to them being derived from the
“courser” rather than the “galley” meaning of the word kele2s . They
must have been constructed differently to dromons for if the latter
were capable of carrying 12 horses they would have been very
inefficient battle galleys, and the evidence is clear that that was not the
case. When the Anonymous wrote that chelandia and dromons were
“both ... constructed from the same ship’s timbers, even if they differ
in their overall nomenclature, the one being called dromo2n and the
other chelandion”,484 that may have been true in a generic sense.
However, it does not necessarily mean that he regarded the two types
as indistinguishable, even if, given the limited nature of his
understanding of Byzantine war galleys, he himself may indeed have
been unable to distinguish between them.
Evidence for the capability of any maritime power to transport
horses over anything more than short distances before the twelfth
century is very meagre. The only Byzantine naval expedition known
in any detail which must have involved the transportation of horses
over long distances beyond the frontiers of the Empire was
Belisarios’s expedition to Vandal Africa in 533. Virtually nothing is
known about the later expedition under the patrikios John sent by the
emperor Leontios in 697 to recover Africa from the Muslims except
that after capturing Carthage he was forced to return to Crete for
supplies and reinforcements after being overwhelmed by a Muslim
relief fleet.485 Prokopios wrote that Belisarios took 5,000 hippeis,
cavalry[men], with him; however, that figure was an ambit one,
contrasted to 10,000 stratio2tai, soldiers, and it seems to us highly
unlikely that the Byzantines possessed the technology to transport
such an enormous number of horses all the way to Africa, a voyage
which Prokopios said took three months. Certainly, some horses at
least were transported all the way because at He2rakleia a number of
horses from imperial herds in Thrace were embarked,486 and
------------------------------
484
Appendix Three, §2.16.
485
See above, pp. 27-8.
486
Prokopios, History of the wars, III.xi.2 (vol. 2, pp. 100-102): “h[dh de; xu;n aujtoi'"
kai; th;n ej" Karchdovna strateivan ejn paraskeuh'/ ei\ce, pezou;" me;n stratiwvta" muvr iou",
326 CHAPTER FOUR
Prokopios put a mention of landing horses into the mouth of

Belisarios in an address to his commanders before the landing. He
then wrote that during the digging of a stockade at the place of landing
they struck enough water to suffice for all the men and animals.487
Cavalry were certainly deployed in the ensuing campaign and the final
battle at Tricamaron was a cavalry engagement. However, in the
following year, at the outset of the Gothic War in Italy, the
Ostrogothic queen regent Amalasuntha claimed to have assisted
Belisarios’s invasion of Africa by furnishing a great number of horses,
to which she attributed his final victory. Many of Belisarios’s
cavalry[men] may have in fact acquired their horses in Sicily.488 He
also captured horses from the Vandals in Africa.489 In sum, it is
unclear how many he actually transported all the way from the region
of Constantinople.
Byzantines could certainly transport horses for short distances, as
was shown during the subsequent Gothic War. In 545 Belisarios sent a
message to Justinian begging for a new army and horses, and then in
547 he sailed from Rome for Sicily and then Taranto with 700 cavalry
and 200 foot. He was forced by to put in to Crotone and stayed there
with the infantry but sent the cavalry ahead to secure the passes and
supplies for themselves and the horses. In the following year he sailed
from Otranto to the relief of Rossano but the fleet was scattered by a
storm and after regrouping was detered from landing by Totila’s
cavalry lining the beaches, which in itself suggests that Belisarios’s
ships did not yet have the stern ports and ramps which would later
make it possible to land cavalry in the face of opposition. After
retiring to Crotone, a council decided to land the men and horses and
march overland to Picenum.490
Transporting horses for short distances was one thing, but doing so
for long distances was quite another. In 533 Belisarios’s horses were
most probably transported on sailing ships. The transport fleet was
said by Prokopios to have been composed of 500 ships ranging in
capacity from 3,000 to 50,000 medimnoi, 50-825 metric tonnes, and
------------------------------
iJppeva" de; pentakiscilivou", ...”; xii.6 (vol. 2, p. 112): “..., ejpei; basileu;" i{ppoi" o{ti
mavlista pleivstoi" to;n strathgo;n ejntau'qa ejdwrei'to ejk tw'n basilikw'n iJppoforbivwn, a{
oiJ nevmontai ej" ta; ejp i; Qrav/kh" cwriva.”. Cf. Theophane2s, Chronographia, A.M. 6026
(vol. 1, p. 190).
487
Prokopios, History of the wars, III.xv.29, 35 (vol. 2, pp. 138-40).
488
Prokopios, History of the wars, V.iii.22-4 (vol. 3, p. 30).
489
Prokopios, History of the wars, III.xvi.12, xxv.15 (vol. 2, pp. 146, 202). Cf.
Theophane2s, Chronographia, A.M. 6026 (vol. 1, p. 191).
490
Prokopios, History of the Wars, VII.xii.1-3, xxvii.13-17, xviii.3-7, xxx.1, 9-15
(vol. 4, pp. 248, 390, 394, 406, 408-10).
for such a long voyage sailing ships would certainly have been
preferable to galleys.491 But the campaigns to recover Crete in the
ninth and tenth centuries make it clear that even then the Byzantines
did not transport their horses all the way from Constantinople but
rather embarked them at aple2kta in south-west Asia Minor. The
cavalry and their horses were marched overland to the aple2kta.
Krateros’s expedition of ca 825-6 was launched from the thema of the
Kibyrrhaio2tai. In 866 the Caesar Bardas used Kepoi at the mouth of
the Maeander river. In 911 Phygela was used and for the final assault
in 960 Nike2phoros Pho2kas again used Phygela. The sources do not
reveal whether aple2kta were used for the other Byzantine attempts to
reconquer Crete but they certainly would have been.492
Other considerations also need to be taken into account. Horses
need large amounts of water to stay in good condition, anything from
4-10 gallons (18.2-45.5 litres) per day according to conditions and
what activities they are required to perform.493 Aboard ship they
would have nothing to do, but conditions would have been very
cramped, hot, and humid below decks at sea on the Mediterranean in
the summer and around 8 gallons or 36 litres would have been needed
per horse per day.
In a contract drawn up between Louis IX of France and Venice in
1268 for transportation of Louis’s projected Crusade to Tunis, the
king of France’s agents specified that the rations which the Venetians
should supply should include 15 quartae of water by the measure of
Paris, 28 litres, per horse per day.494 In the Informationes pro passagio
------------------------------
491
Prokopios, History of the wars, III.xi.13 (vol. 2, p. 104).
492
See above, pp. 46-7, 72, 305-9.
493
Information supplied by the Department of veterinary anatomy and pathology,
University of Sydney. Haldon’s estimates are similar: 5-8 gallons (22.75-36.4 litres)
per day. See Haldon, “Expeditionary force”, p. 127; Haldon, “Theory and practice”, p.
299, n. 237. Hyland concurs that horses’ minimum requirement is around 4 gallons
per day but that conditions on board ship would have forced consumption up. See
Medieval warhorse, p. 146.
Lieutenant Martin reported that horses aboard ship required 8 gallons per day
and Captain Hayes that the Government allowance was 10 gallons per horse per day,
which allowed for wastage, which could often be large. He wrote that actual
consumption even from Bombay to Liverpool in summer was no more than 5 gallons
per day. See Martin, Transport of horses, p. 28; Hayes, Horses on board ship, p. 155.
Lt colonel Shirley recommended 6 gallons a day, but his experience was on the cold
North Atlantic run to Canada. See Shirley, Transport of cavalry, p. 29. Colonel, later
general, Smith recommended 7-8 gallons per day and general Wolseley a minimum of
6 gallons per day. See Smith, Manual of veterinary hygiene, p. 27; Wolseley,
Soldier’s pocket-book, p. 74. Around 8 gallons a day should have been about right in
the cramped, hot, and humid conditions below deck on medieval ships in the summer.
494
Du Chesne, “Contractus navigii domini regis cum Venetis factus anno Domini
328 CHAPTER FOUR
transmarino of Marseilles drawn up in 1318 for a projected Crusade

by Count Louis of Clermont, 2,400 millayrole of water were to be
allowed for 60 days for 120 horses: 21.15 litres per horse per day.495
However, in both cases these horses were to be transported on very
large sailing ships and the water requirements would have been lower
in the conditions aboard them than in those which would be
encountered on a dromon or chelandion. More pertinent are the
specifications of a contract drawn up in 1246 between Genoese
contractors and the representatives of Louis IX of France for the
construction of twelve taride for the French fleet for the Sixth
Crusade. According to the contract, these transport galleys were to
have 150 oars each and to be capable of carrying up to 20 horses.
Each of them was to be fitted out with butts, bote, for carrying water
with a total capacity for the twelve ships of up to 250 mezeroliae,
approximately 37 tonnes, 3.10 tonnes per ship, and to have 25 barrels
for loading water. The contract specified that the ships were to be
prepared to accompany the royal fleet for up to two years to its
destination in the East. The elapsed time for which the water supplies
were required to last is unknown; however, it is clear that meeting the
water requirements of the horses was a major consideration.496
------------------------------
M.CC.LXVIII”, in his Historiae Francorum scriptores, vol. 5, 435-7; here p. 437.
495
De Boislisle, “Projet de Croisade”, pp. 253-4.
The text is extremely corrupt. It was published by de Boislisle from the
fourteenth-century manuscript Paris, Bibliothèque Nationale, MS. Fr. 12814, fol.
217v ff., one volume of a collection of Memoriaux which came from the monastery of
St Germain des Prés during the French Revolution. The fourteenth-century scribe
clearly had great difficulty with the technical language and the Latinizations of
Marseillese Provençal in the text which he had to copy.
The only place at which horses are mentioned in the text is in a section headed:
“Hee sunt mensure navium de tribus copertis” (“These are the dimensions of ships
with three decks”). Part of the text then reads: “Item, si voluerint portare equos,
portabit CXX cavallos.” (“Item, if they wish to carry horses, it [the ship] will carry
120 knights.”). Then in a section which is separated from this one in de Boisisle’s text
but which may well originally have been joined to it, the provisions for men and
horses are specified, including “Primo, levabit necessaria pro LX diebus pro equis,
videlicet: ... Item, aquam pro equis, MM.CCCC. millayrolas.” (“First, it [the ship] will
load necessaries for 60 days for the horses, namely: ... Item, water for the horses,
2,400 millayrole.”).
496
Champollion-Figeac, “Traités passés en l’année 1246 entre les commissaires du
roi Saint Louis et ... Gênes ...”, in his Documents historiques, vol. 2, part 2, No XXIX
(pp. 54-61), here XXIX.vii (pp. 59-60): “Botas pro aqua portanda usque ad mezerolias
CCL, et barulos pro aqua levanda XXV, …”; “…, boutes pour aigue porter jusques à
CCL mizeroles, et XXV barris pour aigue lever, …”. The Old French version of the
contract given at N o XXX.viii (pp. 66-7) agrees with the Latin version in all specifics.
There are problems with this contract. The number of oars specified for each
tarida was 150 but no galley of any kind rowed that many oars at the time. Unless,
perhaps, around a third of those specified were intended to be spares. Even so, the 20
Even the 12 horses per chelandion reported by Theophane2s the

Confessor would have consumed around 430 litres of water per day.
For a voyage of, say, four days, they would have consumed around
1.73 tonnes of water and that raises the question of where the large
amounts of water that the horses needed could be stowed aboard
dromons or chelandia if they still had a bank of oarsmen below deck.
We will return to this problem when we address the problem of water
supplies in general.497
Belisarios’s fleet weighed anchor around the spring equinox (20
March), at the very beginning of the sailing season,498 probably with
the intention of completing the voyage before mid summer in order to
avoid the worst of the heat. At 36 litres of water per horse per day, it
would have meant that if there really were 5,000 horses, some 180
metric tonnes of water would have been needed for the horses alone
every day. During the voyage they would have consumed around
16,200 tonnes of water. Leaving aside the additional water needs of
the men, which should have been at least equal to this again, one has
to question how that amount of water could have been manhandled
with buckets or amphorae from the springs, small streams, and wells
to be found en route. Such considerations suggest that the report that
Amalasuntha supplied Belisarios with horses in Sicily was correct and
they also help to explain why the Byzantines took their horses
overland to aple2kta in south-west Asia Minor for the expeditions to
Crete before making only very short dashes of no more than 400
kilometres to the island.
One other consideration contributes to an understanding of why it
was very difficult to transport horses for long distances. They suffer
------------------------------
horses to be catered for suggests quite small transport galleys. Thirty horses had been
quite common earlier in the century and would be again not so many years later. And,
it is clear from the oars specified that these were galleys rather than sailing ships, but
only 20 sailors, marinarii, were required for each. This figure obviously does not
agree with that for the number of oars, unless, perhaps, the marinarii were not
oarsmen but sailors operating as deck hands. Whatever the case, it is clear that there
are serious textual problems created in the transmission of the text over the centuries
and the 250 mezerolie of water were probably intended to be for the whole 12 taride,
giving a water supply per tarida of 3.093 tonnes. This figure is far more realistic than
the 37 tonnes otherwise demanded. No medieval galley could take on board that
weight of water and remain seaworthy.
497
See below p. 371.
498
According to Vegetius, the sea was closed from 11 November to 10 March.
From 10 March to 15 May navigation could be resumed, but only at risk. War fleets,
he said, should be more cautious than sailing ships, and for good reason. Low-lying
galleys, designed to cut through the waves rather than to ride them, were much more
prone to being swamped in rising seas than sailing ships. Only from 27 May to 24
September did the seas become truly safe. See Vegetius, Epitoma, IV.39 (pp. 156-8).
330 CHAPTER FOUR
badly from cramped conditions and poor ventilation

Most illnesses suffered by horses aboard ship, such as azoturia,
heat apoplexy, pneumonia, laminitis or fever in the feet, and
constipation are caused by hot weather, immobility, overcrowding, or
inadequate ventilation. Azoturia is a build-up of lactic acid, causing
swelling of the muscle fibres, followed by degenerative changes,
locomotor inability, passage of myoglobin into blood plasma and its
excretion into the urine. It is induced by severe physical stress, and is
alternatively called tying-up syndrome. Lack of oxygen can cause
horses to go blue and they can die of suffocation.499
Horses breathing in pass on to their lungs air containing 20.96%
oxygen and 0.04% carbon dioxide and 0.02% impurities. Their
expired air contains 19.96% oxygen and 4.04% carbon dioxide and
impurities. The nitrogen remains constant. The air loses in the lungs 4-
5% of its oxygen and gains 3-4% carbon dioxide plus impurities. An
average inhalation for horses at rest is around 4 litres and they will
absorb 2.38 cubic metres of oxygen and produce 2.04 cubic metres of
carbon dioxide in 24 hours. They breathe around 11.5 times per
minute and take in around 2.8 cubic metres of air per hour, making
around 0.7 cubic metres of air completely unbreathable every hour.
Around 150 times more air must be introduced per hour than actually
passes through their lungs. Fresh air per hour per horse needed to keep
the critical inorganic impurities at .02% was calculated by Frederick
Smith, following Francis de Chaumont, as the equation e/p = d, where
e = the amount of carbon dioxide plus impurities in cubic feet exhaled
per hour, viz 3, p = the limit of permissible organic impurity per cubic
foot, viz 0.0002, and d = the amount of fresh air required in cubic feet
per hour; i.e., 15,000 cubic feet, 425 cubic metres.500
Following De Chaumont/Smith’s equation the 12 horses per
chelandion reported by Theophane2s the Confessor would have
required 5,100 cubic metres of fresh air per hour but the volume of the
hold of such ships could not possibly have been more than around 150
cubic metres, probably somewhat less. It would have been necessary
to change the air over 30 times an hour for the horses alone, not to
mention the oarsmen. The decks must have had extensive gratings and
------------------------------
499
See Hayes, Horses on board ship, pp. 225-6; Martin, Transport of horses, pp.
38-45; Mellows, “Observations”, pp. 105-6; Smith, Manual of veterinary hygiene, pp.
917-18. Cf. Hyland, Medieval warhorse, pp. 102, 169, 182, n. 171.
500
De Chaumont, “On ventilation”, p. 1031; Smith, Manual of veterinary hygiene,
pp. 40-41, 53-6. We have translated Smith’s imperial measures into metric ones. De
Chaumont’s equation was developed for men and adapted by Smith for horses.
forced ventilation would have had to have been used if horses were
transported below deck in dromons or chelandia for more than very
short distances. The holds of the ships would have had to have been
ventilated by something like windsails and cowls, the windsails turned
into the wind acting as inlets and the cowls as outlets.501 Forced
ventilation would have been necessary for the lower oarsmen also.
Even so there would have been high rates of illness among the
horses during extended voyages and they would have required
considerable recovery time before they could have been put to work.
It was one thing for the knights of the Fourth Crusade to go mounted
directly into battle from the horse transports outside Constantinople
after crossing the Bosporos from Chalke2do2n, as Robert of Clari
reported. But as Ambroise reported, it was quite another for the horses
of Richard Cœur de Lion when they were landed in Cyprus. Dio
Cassius reported that in 46 B.C.E. Julius Caesar’s cavalry in Africa
was driven back by that of Marcus Petreius and Titus Labienus
because the horses had not yet recovered after the short voyage from
Sicily.502 Laminitis will occur in many horses if they are worked even
moderately straight after landing after a long voyage. Considerable
recovery time is necessary.503 Even if launched from aple2kta in south-
west Asia Minor, for the Byzantine campaigns against Crete, it would
have been highly desirable to have unloaded the horses on Naxos or
Ios to allow complete recovery before the final passage to Crete.
There is no doubt that by the tenth century maritime powers could
transport cavalry[men] and horses for short distances, but long
distances were another matter. Fulcher of Chartres was quite explicit
about this. He commented in a chapter of his Historia
Hierosolymitana written between 1102 and 1106 and contained in
manuscripts of the first redaction of the chronicle completed by 1124,
referring to the first year of the reign of Baldwin I of Jerusalem in
1101, that: “For whom [the Franks] there would have been nothing
lacking, if only men and horses should not fail. Wherefore, we could
not go on an expedition, except if we campaigned locally or towards
------------------------------
501
See Hayes, Horses on board ship, p. 43; Martin, Transport of horses, pp. 4-6;
Mellows, “Observations”, pp. 105-6; Shirley, Transport of cavalry, pp. 21-3; Smith,
Manual of veterinary hygiene, pp. 900-904.
502
Dio Cassius, Roman history, XLIII.ii.2 (vol. 4, p. 212). See also above pp. 311
& n. 453, 318-19 & n. 474. In fact the crossing of the Bosporos in 1203 was not made
from Chalke2do2n as reported by Robert of Clari but rather from Chrysopolis.
503
Hayes, Horses on board ship, pp. 208-10; Martin, Transport of horses, pp. 46-7;
Shirley, Transport of cavalry, p. 31; Smith, Manual of veterinary hygiene, p. 921. See
also Hyland, Medieval warhorse, p. 148.
332 CHAPTER FOUR
Ascalon or Arsuf; indeed those who came by sea to Jerusalem could

by no means bring horses with them”.504 Later, Fulcher suggested
clearly just how difficult transporting horses by sea over long
distances was and how big a problem water supplies were. In 1123 the
Venetians sent a new Crusading fleet to the Holy Land. In all
probability this was the first attempt made by a Western power to
transport horses by sea across the length of the Mediterranean. Fulcher
described the fleet and its voyage in the following terms:
Who [the Venetians], having left their own land the year before,
wintered on the island called Corfu, awaiting a favourable season. Their
fleet was of 120 ships, not counting small boats or skiffs, of which [ships]
some were spurred (rostratae), some indeed were transport ships, and
some were triremes. ... After the routes were opened to ships in the spring
season, they did not delay in fulfilling what they had long vowed to God.
... In which [ships] were 15,000 armed men, Venetians as well as the
pilgrims joined to them. In addition they conveyed 300 horses with them.
... And since it was necessary that they proceed together and not
scatteredly, and because the winds also veered from time to time, they
carefully controlled their voyage lest they quickly become separated from
each other. Therefore, sailing by short stages, by day and not by night, by
necessity they put in daily at the ports which they found frequently, lest
both they and their horses, suffering lack of fresh water, be oppressed by
thirst.505
------------------------------
504
Fulcher of Chartres, Historia Hierosolymitana, II.vi.12 (p. 390): “quibus nulla
inopia esset, si tantummodo gens et equi non defuissent. quamobrem in expeditionem
ire nequibamus, nisi prope vel versus Ascalonem vel Arsuth equitaremus; et qui per
pelagus Hierusalem veniebant, equos secum adducere nequaquam poterant.”.
Note that this text is as given in Hagenmeyer’s notes for the manuscripts of the
first redaction, not as printed in his text from those of the second redaction and as
translated by Fink and Ryan.
505
Fulcher of Chartres, Historia Hierosolymitana, III.xiv.1-2, xv.1-4 (pp. 656-8):
“qui [Venetici] anno precedenti de terra sua egressi, in insula, quae Curpho
nuncupatur, tempus exspectantes opportunum hiemaverunt. classis quippe eorum
CXX navium fuit, exceptis carinis et carabis, quarum aliae rostratae aliae quidem
onerarie, aliae vero triremes fuerunt. ... Igitur postquam verno tempore patescunt viae
ratibus, quod Deo diu devoverant explere non torpuerunt. ... quibus ter quina
hominum armatorum milia tam de Veneticis quam peregrinis sibi adiunctis inerant.
porro equos secum CCCos convehebant. ... et quia necesse erat, ut simul nec sparsim
incederent, flabris etiam interdum alternantibus, nisi provide iter suum modificarent,
alii ab aliis cito discreparent, propterea dietis brevibus die non nocte velificantes,
portibus frequenter inventis necessario cotidie applicabant, ne recentis aquae
penuriam patientes tam ipsi quam eorum equi siti gravarentur.”.
Again this is the text of the first redaction as given in most of the manuscripts
according to Hagenmeyer’s notes. There is a significant variant in the third last line of
“diebus” (which makes much less sense) for “dietis” in the text as printed from the
manuscripts of the second redaction.
We believe that Fulcher was not exaggerating. The Venetian fleet

took from early spring to mid-May to reach the Holy Land from
Corfu, probably around two months for the 2,255 kilometres at an
average speed of only around 0.85 knots, extraordinarily slow
considering that the prevailing winds would have been astern on the
port quarter for the most part. It suggests that the fleet did indeed sail
only by day and put in wherever possible to take on water. Watering,
in particular watering the horses, must have been a laborious and
time-consuming exercize. We also believe that the Venetians probably
moored whenever possible to open the hatches and ventilate the holds
where the horses were stabled and that they probably also landed to
exercize them from time to time.
(l) Performance capabilities, water supplies, and logistics
According to his Life of St Theoktiste2 of Lesbos, the narrator, Nike2tas

Magistros, was told on Paros by a hermit that from Paros he would
sail to Naxos, lie there in harbour for one day, sail for Crete on the
second day and reach it on the third. Since at the time Nike2tas was
accompanying the expedition of Himerios to the Levant and Crete, he
was presumably on a war galley, a dromon or chelandion, and the
voyage predictions read like a reflection of what a galley could expect
to accomplish by sailing before the prevailing northerlies of summer
south to Crete. Naxos to Chandax, via Ios and The2ra is only around
195 kilometres; easy sailing before the prevailing northerlies of
summer in 24-36 hours, depending upon the time of the second day
that they reached Crete, at an average speed of around 2.8-4.3 knots.506
This is one of the very few pieces of arguably reliable “data” which
survive in Byzantine sources for the performance capabilities of what
may have been a dromon. A search of the surviving historical record
for data recording the duration times of voyages made by dromons has
proved almost fruitless. Where data appears to survive, for example in
Theophylaktos Simokatte2s’ account of the transmission of the news of
the murder of the emperor Maurice from Constantinople to
Alexandria,507 there is invariably some question of unreliability. Either
------------------------------
506
Nike2tas Magistros, Vita S. Theoktistae, §13, trans. Hero (pp. 107-8).
507
Theophylaktos Simokatte2s, Historiae, VIII.13.7-14 (pp. 309-11). The story was
later repeated by Theophane2s the Confessor. See Theophane2s, Chronographia, A.M.
6095 (p. 291).
Under the year A.M. 6095, Theophylaktos recounted a story of a pious copyist
334 CHAPTER FOUR
the narrator was reporting something miraculous, or the type of ship

was not specified, or, as in Theophylaktos’s case indeed, no ship was
mentioned at all even if it was no doubt presumed that the news was
transmitted by sea.
Another possible case of one voyage duration that may stand up to
scrutiny is that of Nicholas Mouzalo2n, who wrote that he was virtually
ordered to the archbishopric of Cyprus by Alexios I Komne2nos early
in the twelfth century. It is tempting to think that he may have been
provided with an imperial dromon for the voyage. In similar
circumstances St Symeon of Mytile2ne2 had been given one by the
Empress Theodo2ra when he was appointed to the bishopric of
Mityle2ne2.2 508 Unfortunately, Mouzalo2n simply used the generic for a
ship, ploi'on, when he wrote that he made Cyprus from Constantinople
in 10 days.509 The future patriarch was, however, making a point when
he wrote that he made the voyage in the best possible anticipated time
because his ship had excellent sailors and the Holy Spirit filled its
sails. We do not know at what time of the year he made the voyage,
but assuming that it was early summer with around 14.75 hours
------------------------------
in Alexandria who saw in a vision the statues of Emperor Maurice and his family
being dragged from their pedestals in Constantinople. Nine days later news of the
murder of Maurice by Pho2kas reached Alexandria. Pho2kas seized the throne on 23
November 602 and, even though no ship was actually mentioned by Theophylaktos,
some ship was presumed to have covered the 1,600 kilometres from Constantinople to
Alexandria at the onset of winter in 9 days, an average speed of around 4.35 knots if
sailing around the clock. But noone would have tried to navigate the Dardanelles or
the East coast of the Aegean by night at the onset of winter. From Rhodes to
Alexandria they obviously would have had to but that is only about 645 kilometres.
For the rest, at that time of the year they would have had a maximum of around 9.5
hours of daylight. See USNO, Sun and moon, accessed 21/02/2005. On 30 November,
1800, the sun rose at 0551 hours and set at 1618 hours at Constantinople and rose at
0641 and set at 1658 at Alexandria, giving an average for the voyage from sunrise to
sunset of around 9.5 hours.
The actual average speed when under way would therefore have had to have
been an incredible 8.36 knots. To put this in perspective, in 1798 Nelson sailed for
Alexandria from Syracuse in pursuit of the French at top speed in mid summer on 25
July. He reached Alexandria via Koro2ne2 on 1 August in eight days: 1,610 kilometres
at an average speed of 4.75 knots. If we are to believe Theophylaktos, we must
believe that a seventh-century ship almost doubled the speed of what was the fastest
squadron in the British Mediterranean fleet in 1798.
What then are we to make of Theophylaktos’s story. Well, it was a miracle, a
“miraculous narrative” as he wrote, and no doubt he intended his audience to
recognize it for what it was. It wasn not to be taken seriously. At the beginning of the
seventh century no ship could reach Alexandria from Constantinople in November in
nine days.
508
See above p. 172.
509
Doanidu, “JÔH paraivthsi" Nikolavou tou' Mouzavlwno"”, p. 119.
daylight,510 the voyage was made at an average of around 5.0 knots,

which would certainly be at the upper level of expectations, but not
impossible. Did his mention of excellent sailors point to an imperial
dromon?
If all conditions remained continuously favourable and it could use
its sails, in early summer a dromon ought to have been able to reach
Paphos from Constantinople in around 10-12 days. Any galley could,
of course, make way either under sail or under oars. In favourable
light conditions it might be possible to use both oars and sails, but that
would have been unusual.511 With a favourable light breeze from
astern, say around Beaufort Scale Three, 7-10 knots with large
wavelets up to 2 feet or 61 centimetres, a dromon could no doubt bowl
along quite nicely under sail. In fine conditions it might make a
voyage such as that from Constantinople to Rhodes in as little as 8-10
days or to Paphos in 10-12 days.
Before the prevailing North to East light winds of summer, such a
dromon sailing from Constantinople at dawn, around 0430 hours, on a
fine summer’s day in early July might well drive across the Sea of
Marmara on a west-south-west course with the wind only two to six
points on the starboard stern quarter and even make Rhaidestos, 130
kilometres away, by evening around 1945 hours at an average speed
of around 5 knots. He2rakleia, around 105 kilometres from
Constantinople, would have been even more reachable at an average
speed of around 4 knots.512 Proikonne2sos was only around 50
kilometres south-south-west of He2rakleia, an easy run, and Abydos
115 kilometres from Proikonne2sos. Even the 155 kilometres from
Rhaidestos to Abydos would not have been impossible with the
assistance of the current through the Dardanelles.513 The run from
------------------------------
510
See USNO, Sun and moon, accessed 21/02/2005. On 30 June 1800 the sun rose
at Istanbul at 0434 hours and set at 1940 hours. It rose and set at Paphos at 0439 and
1907 hours respectively, giving an average for the voyage of 14 hours, 47 minutes
from sunrise to sunset.
511
The sea trials of Olympias have shown that both oars and sails may be used
together but only in light breezes from astern or on the quarter. Obviously oars could
not be used when a ship was heeling under sail with a wind from abeam. See Coates
and Morrison, “Sea trials”, p. 139; Coates, et al., Trireme trials, p. 39; Morrison, et
al., Athenian trireme, pp. 258-9.
512
In July at Istanbul the wind prevails from the north to east around 70% of the
time and calms represent another 12%. The mean wind speed is only 11 knots. Winds
from the west to south occur less than 3% of the time. See Great Britain, Black Sea
Pilot, p. 72. On 30 June 1800, the sun rose at 0434 hours at Istanbul and set at 1946
hours at Rhaidestos and at 1944 hours at He2rakleia. See USNO, Sun and moon,
accessed 21/02/2005.
513
A run from Rhaidestos to Abydos would have been almost directly south-west.
336 CHAPTER FOUR
Abydos to Tenedos was only around 50 kilometres. Heading south into

the Aegean with the wind prevailing strongly from the north,514
daylight passages of around 130 kilometres to Mityle2ne2, 115
kilometres to Chios, about the same to Samos, perhaps two days with
a stop over somewhere among the scattered islets of the Sporadhes for
the 260 kilometres to Kos, and then a final day’s sail of around 130
kilometres to Rhodes would bring a dromon into Rhodes from
Constantinople in 8-10 days. At Rhodes in July the wind averages
76.5% from west to north at an average speed of 18 knots, driving any
ship straight into it from Kos.515 The 400 kilometres from Rhodes to
Paphos ought to have been coverable before the prevailing west to
north winds in three days with a first-night lay-over somewhere
around Patara or Phoinikous.516
The scenario would, however, be very different if the wind rose to
Beaufort Scale Four-Five (16-17 knots). That would raise waves of
around 4.75 feet, 1.45 metres. All galleys at all times were designed to
cut through the water rather than to ride the waves and such a wind,
which is just a “moderate” to “fresh” breeze on the Beaufort Scale,
nothing out of the ordinary, would send waves washing over the deck
of any dromon. Even if the wind were astern, she would still be forced
to run for the coast. If the wind were ahead, it would be worse because
that would mean that the ship was attempting to beat to windward and
therefore would be heeling over with one gunwale continuously under
water. And in the Aegean in summer the meltemi can rise to become a
very strong wind indeed, even up to Beaufort Scale Seven (28-33
knots), particularly in the channels between the islands, raising short,
------------------------------
At Çanakkale near Byzantine Abydos, the wind in July prevails from the north to east
88% of the time at an average of 10.5 knots. See Great Britain, Black Sea Pilot, p. 71.
514
In July in the Aegean the wind prevails strongly from the north to east-north-
east in the northern sector, swinging to the north-west to north-north-east in the
central Aegean and then to the west-north-west to north-west in the south approaching
Rhodes. The strength is commonly in the order of Beaufort Scale One-Four, 0-16
knots, in the north, strengthening to Three-Five, 7-21 knots, in the central Aegean and
to Four-Five, 11-21 knots, in the south towards Rhodes. See Great Britain,
Mediterranean Pilot. Vol. IV, fig. 6.
515
Great Britain, Mediterranean Pilot. Vol. IV, p. 33.
516
In 1102 the small Byzantine coaster on which the Anglo-Saxon pilgrim Saewulf
was travelling made Patara from Rhodes in late September in a day’s sailing, around
95 kilometres in 11.2 hours of daylight at around 4.9 knots: good sailing. [On 30
September 1800, the sun rose at Rhodes at 0603 hours and set at 1755 hours, giving
11 hours and 12 minutes of daylight. See USNO, Sun and moon, accessed
21/02/2005]. See Huygens, Peregrinationes tres, pp. 59-61. At Paphos the wind in
July prevails from the west to north around 66% of the time at a mean wind speed of a
gentle 6 knots. See Great Britain, Mediterranean Pilot. Vol. V, p. 29.
steep seas.517 Scale Seven winds would raise seas up to 13.5 feet
(4.115 metres) and no dromon would stand a chance of continuing its
voyage in such conditions. The authors of the Olympias project have
concluded that a trie2re2s would be swamped in waves above 0.85
Figure 40
Bireme dromon of the era of the Macedonian emperors heeling under sail to
ten degrees.
© John H. Pryor
metres,518 and we believe that in all probability a dromon would have

been also. The meltemi becomes strongest from July through to
September and from noon through to evening, so the optimum time
for a voyage from Constantinople to Rhodes would be spring, from
April to early July, sailing before sunrise as soon as the light made
navigation possible.
However, galleys were simply not designed to be sailed and
throughout history they were always notoriously poor sailers. Because
their lack of deep keels meant that they made excessive leeway when
beating into the wind, because their shallow draft and low freeboard
meant that they could not heel under sail very much, because their
------------------------------
517
See Denham, Aegean, pp. xxv-xxvi.
518
Morrison, et al., Athenian trireme, p. 197. Cf. Shaw, “Oar mechanics”, p. 166.
338 CHAPTER FOUR
narrow beam and low depth in hold meant that their hulls did not have
the structural strength to carry a large press of sail, and because their
extreme length:beam ratio and lateen sails meant that they carried
pronounced weather helm, constantly griping, the bows coming up
into the wind, galleys were always notorious for poor upwind
performance under sail.519 That is nothing to be wondered at for they
were not designed to do that. And dromons may have been even worse
sailers than later medieval Western galleys because with oarsmen
below deck there would have been nowhere to stow ballast and
without ballast any ship heeling under sail would be extremely
unstable.520 Moreover, a heel under sail of a mere ten degrees or so
would put the lower rims of the lower oar ports at the flat water line
and at that point it is highly questionable whether the oar sleeves
would have prevented water from entering the hull, even if they were
tied off.
The speed that oared ships of all kinds could maintain under oars is
a matter of considerable scholarly debate. Different scholars have
directed their attention to different periods and various types of oared
ships and have produced results which are very difficult to reconcile.
What one would like, of course, is reliable historical data for
voyages made by dromons in pressing circumstances in conditions
which would suggest that the voyages were made under oars in calm
conditions or at worst against against light breezes. However, little
data can be found in the Byzantine sources. For the most part we are
compelled to have recourse to those from the sources for classical
antiquity and the Western Middle Ages.
At one end of the scale are the estimates of the capabilities of
trie2reis crews made by various scholars associated with the Olympias
project. These estimates vary somewhat but may be represented by
those of John Coates.521 We, however, find it very difficult to credit
that a crew of any galley at any time could maintain a speed such as
this, around 7.5 knots under oars for 10 hours, covering some 130
------------------------------
519
Pryor, Geography, technology, and war, pp. 71-3; Bragadin, “Navi”, pp. 393-4;
Guilmartin, Gunpowder and galleys, pp. 205-6.
520
It is well known that classical trie2reis carried no ballast and therefore floated
ashore even when sunk. Cf. below p. 392 & n. 634.
521
Coates, “Naval architecture and oar systems”, p. 129. Cf. Morrison, et al.,
Athenian trireme, pp. 262-7. The estimate of Coates, Platis, and Shaw in Coates, et
al., Trireme trials, Annex F.9 (Table D), p. 85, is even higher: an average of 8.0 knots
maintainable for 12 hours. That of Cotterell and Kamminga is similar: around 7.5
knots or a little higher. See Cotterell and Kamminga, Mechanics of pre-industrial
technology, p. 259. However, this is not surprising since Cotterell and Kamminga
were dependent upon Coates et al. for their data.
kilometres in a day. There is simply too little evidence to support it. In

fact the various estimates of those associated with the Olympias
project were primarily based on two voyages reported by Thucydides
and one statement of Xenopho2n, analysis of which has given rise to an
expectation that trie2reis could maintain speeds of this order under
oars. Other voyages analyzed have begun from the assumption that
these speeds were possible. But when the evidence is analyzed again,
Figure 41
Curve of sustainable speed against time for a trie2re2s.
© John Coates
the expectation largely evaporates.

In his Anabasis, Xenopho2n stated that from Byzantion to He2rakleia
Pontike2 was a “long day’s voyage for a trie2re2s under oar”. This was
not a report of an actual voyage but rather an estimate of distance. In
the two oldest manuscripts, both of the twelfth century, the reading
was in fact an “exceedingly long day”, possibly reflecting the thinking
of the scribes, who could not believe it.522 He2rakleia Pontike2 was
------------------------------
522
Xenopho2n, Anabasis, VI.4.2 (p. 468): “Kai; trihvrei mevn ejstin eij" ÔHravkleian ejk
Buzantivou kwvpai" hJmevra" [mavla] makra'" plou'": ...”.
The two manuscripts which have the added word mavla, “very” or “exceedingly”,
are Rome, Biblioteca Apostolica Vaticana, MS. Gr. 1335 and Venice, Biblioteca
Marciana, MS. 511. These manuscripts are the foundation of the so-called “second
family” of manuscripts of the Anabasis and both are dated to the twelfth century,
Vaticana 1335 being earlier than Marciana 511.
The manuscripts of the second family are somewhat longer than those of the
first family, being based on the manuscript Paris, Bibliothèque Nationale, MS. Gr.
1641, which is dated 1320 but which was a copy of one thought to have been
produced in the late ninth or early tenth centuries. It is not known whether the
additions of the second family, such as mavla here, were revisions made by Xenopho2n
340 CHAPTER FOUR
around 240 kilometres from Byzantion, of which the first 30 or so

involved a hard pull up the Bosporos against the current (up to 6-7
knots) and prevailing winds. Even assuming a mid-summer voyage
with 15 hours daylight,523 this would be a voyage made at an
incredible average of over 9 knots. It is simply not believable.
Then there was Thucydides’ report of the famous dash by a trie2re2s
from Piraeus to Mityle2ne22 in 428 B.C.E. to cancel the order carried by
a first trie2re2s to put to death the men taken prisoner at the fall of
Mityle2ne2 to the Athenians. Morrison calculated that this dash was
accomplished in 24 hours at sea in one continuous voyage with a crew
alternating at the oars in shifts at an average speed of around 7.7
knots.524 Thucydides wrote specifically that there were no head winds
and, whatever the actual speed was, it must surely have been at the
upper limit of what a trie2re2s could achieve under oars. However, in
our opinion, Morrison’s reconstruction forces what Thucydides
actually wrote beyond its sustainability and in fact no estimate of the
speed of either trie2re2s involved is possible. He himself virtually
admits as much.525
More believably, Thucydides reported that in the summer of 411
B.C.E. the Spartan admiral Mindaros took a Peloponne2sian fleet of 73
ships north from Chios to Rhoiteion, just inside the Dardanelles, in
two days. This must have been a voyage made under oars, most
------------------------------
himself in a second draft or were additions of later readers and copyists.
523
At Istanbul on 30 June 1800, the sun rose at 0434 hours and set at 1940 hours,
giving 15 hours and six minutes of daylight. See USNO, Sun and moon, accessed
21/02/2005.
524
Thucydides, Peloponnesian war, III.36.2-3, 49.2-4 (vol. 2, pp. 56, 84-6). See
Morrison, “Trireme”, pp. 57-9; Morrison, et al., Athenian trireme, pp. 95-6, 104.
525
The Athenians’ first decision to send a trie2re2s to announce the decision to
execute the prisoners was taken “after a debate”. Thucydides did not say that it was
after a meeting of the Athenian Assembly and he did not say at what time of day the
ship set sail. Next day, after a meeting of the Assembly was held, a second trie2re2s was
despatched in all haste. It left “a day and a night” later. Meetings of the Assembly
began at daybreak or in the early morning and could go on until nightfall but most
Assembly meetings were probably over by midday. See Hansen, Athenian democracy,
pp. 136-7. So the first trie2re2s probably had left the previous morning. The Mityle2ne2an
envoys in Athens provided wine and barley for the crew of the second trie2re2s, who
rowed continuously in shifts and ate when off the oars. There was no contrary wind.
Since the earlier ship had not been in a hurry, the second arrived shortly after the first
and was able to countermand the decision to execute the prisoners.
This is all that Thucydides reported and Morrison’s re-construction of the speed
of the second, chasing trie2re2s is entirely dependent upon his re-construction of the
voyage of the first. Of this he guesses that it left Piraeus around midday and arrived at
Mityle2n e2 round midday on the third day, having bivouacked on land twice overnight
and having taken a midday break on land on the second day. However, there is
absolutely nothing in Thucydides’ account to support this.
probably in light conditions with slight onshore and offshore breezes

by day and night respectively but with no influence from the meltemi
which would have prevailed out to sea in the central Aegean. The
situation was urgent and they left Chios early on the first day,
bivouacked overnight at Arginousai and sailed again “while it was
still night proper”; i.e., the Greek implies, just at the first pale glimmer
of light when dawn proper was still a long way off.526 They reached
Harmathous at “breakfast” on the second day and Rhoiteion “before
the middle of the night” on the second day, although some of the ships
apparently failed to make Rhoiteion on the second day and put in at
Sigeion and other harbours around the entrance to the Dardanelles.527
Morrison’s reconstruction of this voyage presses the point of
coastal navigation by galley fleets and of the speeds attainable by
trie2reis under oars to an unrealistic extreme. His calculations of the
distances involved are 65 sea miles (110 kilometres) and 124 sea
miles (210 kilometres) respectively for the two days. However, he has
opted for extreme coastal routes, never taking the fleet more than a
couple of kilometres offshore except for the crossing of the entrance
to the Gulf of Izmir. But Thucydides did not say that this was so. In
fact, since the situation was pressing, extreme coastal routes would
have been counter-productive. We acknowledge, of course, the fact
that galley fleets did navigate by coastal routes. But the question is
one of sense and sensibility. To have made an extended detour around
the shores of the Gulf of Çandarli when the entrance to it is barely
eight kilometres across would have been unnecessary. Given the
pressing nature of the voyage, even more incomprehensible would
have been a run 25 kilometres north-east into the Gulf of Edremit
before heading west to Harmathous. This would have added some 80
kilometres to a direct crossing from Eg°ribucak Point west of Ayvalik
and then around the north-east coast of Lesbos to Harmathous, totally
unnecessarily since even the direct crossing would not have taken a
ship more than three kilometres or so offshore. Whereas Morrison’s
calculation of the speed achieved on the second day was 6.9 knots, our
own calculations of the distances reduce that from Chios to
Arginousai to around 98 kilometres at an average speed of
approximately 3.65 knots and Arginousai to Rhoiteion to around 135
kilometres at an average speed over 18 hours of approximately 4.05
------------------------------
526
Anyone who has been at sea on a cloudless and starlit night and has watched the
dawn come up over a gentle sea will know what Thucydides meant. First light can
precede dawn by an hour or more if the atmospherics are right.
527
Thucydides, Peloponnesian war, VIII.101 (vol. 4, pp. 378-80).
342 CHAPTER FOUR
knots. These were reasonable speeds for hardened crews to have

maintained under oars for extended periods of time, although the fact
that some crews were unable to maintain them to the finish at
Rhoiteion indicates that they were at the upper limit of oarsmen’s
capabilities.528
The estimates of those associated with the Olympias project of the
speeds which trie2reis could sustain for extended periods were too high
and even the the classical evidence does not support them. Trie2reis
were no doubt fast ships but it is implausible to argue that they were
the fastest galleys ever built and could outperform war galleys of later
periods. Liburnae were later adopted into Roman fleets because they
were faster than other contemporary galleys. Dromons were
developed because they were faster than liburnae. Galeae would later
spread rapidly in the West and then across the Mediterranean because
they had fine lines and were fast. The expressed disappointment at the
performance of Olympias in sea trials, even acknowledging the now-
known flaws in the ship’s design, is excessive.529 The ship’s
performance was probably closer to realistic expectations than its
designers believed.
Consideration of recorded voyage times of ancient and medieval
galleys and galley fleets leads to the conclusion that the capabilities of
galleys under oars, including dromons, were more realistic than the
expectations of some members of the Olympias project. That being
said, the evidence of the various records varies enormously and few
reports are verifiable. Invariably they had some literary, didactic, or
polemical purpose and may not be read as “shipping notices”; for
example, Pliny the Elder’s reports of voyages to Egypt.530 In many
------------------------------
528
See Morrison, et al., Athenian trireme, pp. 97-8 and map 11, 104-5.
529
Morrison, et al., Athenian trireme, pp. 264-7.
530
Pliny, Natural History, XIX.i.3 (vol. 5, pp. 420-22).
According to Pliny, senator Valerius Marianus (or Marinus) made Alexandria
from Pozzuoli in summer on the ninth day (in 8-8.5 days) “with a very gentle breeze”
“... lenissumo flatu.”: 1,930 kilometres at an average speed of 5.25 knots even if the
shortest high-seas route was taken. Two imperial praefecti Aegypti, Gaius Galerius,
praefectus under Tiberius, and Claudius Balbillus, praefectus in 55 C.E., supposedly
did even better. They made Alexandria from the Straits of Messina in 6-6.5 and 5-5.5
days respectively: 1,610 kilometres at 5.8 and 9.15 knots respectively.
To put this in perspective, Cutty Sark made Sydney from London in 1885 in 78
days: 24,140 kilometres at an average speed of around 7.25 knots. Pliny asks us to
believe that Roman ships could match or better the performance of one the fastest
nineteenth-century clipper ships ever built, even though the latter’s average speed was
greatly increased because she drove for days through the Roaring Forties at up to 15
knots. One must be careful about accepting such data. Historians are too credulous of
narrative sources. Pliny’s purpose in recounting such voyage times was to show what
a “marvel” (“miraculum” was his word) the flax plant used to make linen for sails was
cases they may have represented what authors thought ought to have
been possible rather than what actually happened; although that can
actually be valuable. Longer voyages also incorporated time spent in
ports of call and even shorter ones sometimes incorporated lay-overs
by night. The data is also skewed by some reports of very short
voyages at almost impossibly high speeds. It is that which explains the
very high average of 5.2 knots for voyages under oars in all conditions
calculated in Table 7. Excluding those short voyages, made at an
average of 6.1 knots according to the reports, the average for voyages
under oars of more than one day becomes 4.0 knots. The data which
suggests that average rates of speed decreased according to the length
of voyages must have reflected reality. There is sufficient consistency
in the data to suggest that in favourable conditions fleets could
maintain around four knots while at sea under oars. When layovers and
watering are taken into account, average speeds for extended voyages
would have fallen to no more than two knots. There is no evidence to
suggest that dromons, or indeed any other types of galleys from other
eras, were capable of speeds greater than this except for short sprints.
When winds were adverse and they could not use their sails,
galleys could use their oars. However, an issue invariably overlooked
when discussing the use of oars against adverse winds is the
limitations on doing so created by waves. All winds raise waves. In
the case of a standard dromon, the optimum position for a seated
oarsman would have been to have the handle just below the level of
his shoulders when his arms were fully extended to begin the stroke
but at the end of the stroke he would have to lower his hands to lift the
blade clear of the water for the return. As shown above, between the
shoulders and the top of the legs of a seated man is only around 40
centimetres and as a result, the blades of the lower oars of a dromon
simply could not have been raised more than around 80 centimetres
above the calm waterline, meaning that in waves above 1.60 metres
the lower oars could not have been used at all because the oarsmen
could not have achieved a return stroke. Winds of Beaufort Scale
Four, “moderate breezes” of 11-16 knots, will raise waves of that
height at the top of the range. Even in waves between 0.80 and 1.60
metres, part of the looms would have been below wave crests during
the return and that would have made rowing extremely difficult. In
more than light to moderate breezes the lower oars of dromons could
not have been used and the upper oars alone would have been
------------------------------
but his “data” is not to be believed.
344 CHAPTER FOUR
Table 7:
Some reported voyages of ancient and medieval galleys and galley fleets531
Aw Anonymous, African war L Livy, Ab urbe condita

App Appian, Civil wars Lu Lucan, Civil war
Bc Anonymous, Breve chronicon Pl Plutarch, Dion
C Caffaro, De liberatione Po Polybios, Histories
D Doanidu, “ÔH paraivthsi" Nikolav- Pr Prokopios, History of the wars
ou tou' Mouzavlwno"” RL Pryor, “Roger of Lauria”
DS Diodo2rus Siculus, Bibliothe2ke2 his- SL Dotson, “Simone Leccavello”
torike2 Th Thucydides, Peloponnesian War
H Herodotos, Histories The Theophane2s, Chronographia
Ip Itinerarium peregrinorum (Stubbs) X Xenopho2n, Hellenika
Source: Fleet: From: Current:

Date of voyage: Commander: To: Oars or sails:
Composition: Approximate Time taken:
distance: Approximate
Voyage objective: average speed
Degree of haste:
Part A: Voyages before generally favourable prevailing winds
H, VIII.66: Persian: Euripos: Generally

480 B.C.E.: Not stated: Phale2ron: favourable:
Not stated: 185 kms: Probably sails then
Invasion force: oars:
Moderate: 3 days:
1.39 knots:
Xh, II.1.30: Mile2tan: Aigos potamoi: Neutral:
405 B.C.E.: Theopompos: Lakedaimo2n: Probably both:
Single ship: 675 kms: after the battle to the
Report of victory: third day, ca 42-
High: 54 hours:
6.75-8.6 knots:
Po, V.110.5: Macedonian: Saso2 island: slightly adverse:
216 B.C.E.: Philip V: Kephalle2nia: Probably both:
100 lemboi: 305 kms: “on the second day”
Fleeing Romans: (1-1.5 days):
Highest: 4.6 - 6.9 knots:
L, XXVI.19.11, Roman: Mouth of the Ebro: Slightly favourable:
42.6: Publius Cornelius Cartagena: Probably both:
209 B.C.E.: Scipio jr: 475 kms: “On the seventh
30 quinquiremes: Coasting, accompan- day” (6-6.5
ying land forces: days):
Low: 1.65 - 1.8 knots
------------------------------
531
We acknowledge that interpretation of the ancient and medieval reports is
problematical and our own may be subject to challenge. Distances have been rounded
to the nearest five kilometres.
(Table 7 continued)


Degree of haste:

L, XXIX.27.6-8: Roman: Lilybaion: Neutral:
204 B.C.E.: Publius Cornelius Cape Bon: Sails:
Scipio jr: 145 kms: 1 day (24 hours):
50 quadriremes and Invasion: 3.25 knots:
quinquiremes, Moderate:
400 transports:
L, XLV.41.3: Roman: Brindisi: Slightly adverse:
167 B.C.E.: Lucius Aemilius Corfu: Probably both:
Paulus: 225 kms: 9 Roman hours (10.5
Not stated: Opening campaign: hours):
Moderate: 11.6 knots
App, II.89: Roman: Rhodes: Neutral:
48 B.C.E.: Julius Caesar: Alexandria: Sails:
Unknown number of 565 kms: Three days:
triremes: Opening campaign: 4.2 knots:
Moderate:
A, 2: Roman: Aponiana: Neutral:
47 B.C.E.: Julius Caesar: Africa: Sails:
“a fast ship and a 155 kms: “After the fourth
few warships”: Opening campaign: day” (3-3.5
Moderate: days):
1.0-1.15 knots:
Lucan, IX.1004-5: Roman: Troy: Neutral:
48, B.C.E.: Caesar: Alexandria: Probably sails:
Not stated: 1210 kms: 7 days:
Opening campaign: 3.9 knots:
Moderate:
346 CHAPTER FOUR
(Table 7 continued)


Degree of haste:
The, A.M. 6026: Byzantine: Kaukana: Neutral:

533: Belisarios: Kephale2 Brachous Probably sails:
Not stated, but [via Malta]: “On the third day”
included 90-92 400 kms: (2-2.5 days):
dromons: Invasion force: 3.6-4.5 knots:
Moderate:
Pr, III.25.21: Vandal: Cagliari: Neutral:
534: Tzazo2n: African coast: Probably sails:
Unspecified nh'e": 210 kms: “on the third day”
Responding to news (2-2.5 days):
of Vandal defeat: 1.9 - 2.35 knots:
Highest:
D, p. 119: Byzantine: Constantinople: Slightly adverse:
1107: Not known: Cyprus: Sails:
Possibly an imperial 1370 kms: 10 days:
dromon: To take up 3.1 knots:
archbishopric:
Moderate:
C, p. 102: Genoese: Genoa: Neutral:
1097: Not known: St Symeon: Probably both:
12 galeae and a 3540 kms: Approx. 4 months
sandanum: First Crusade: from mid July to
High: around 20
November:
Approx. 0.65 knots
Ip, II, 26-31: English: Messina: Neutral:
1191: Richard Cœur de Limassol: Probably mainly
Lion: 2175 kms: sails:
Some Mediterranean On Crusade: 21 days at sea (20-
galeae: High: 20.5 days):
2.4 knots:
(Table 7 continued)


Degree of haste:
Ip, II, 26-28: French: Messina: Neutral:

1191: Philip II: Acre: Probably mainly
Some Mediterranean 2495 kms: sails:
galeae: On Crusade: On the 22nd day (21-
High: 21.5 days):
2.65 knots:
Bc, p. 898: Sicilian: Brindisi: Neutral:
1228: Frederick II: Limassol: Probably mainly
40 galeae: 1930 kms: sails:
On Crusade: On the 24th day (23-
High: 23.5 days):
1.85-1.89 knots:
Average under sails: 2.85 knots

Average under sails and oars: 4.3 knots
Overall average: 3.45 knots
Part B: Voyages with neutral prevailing winds
Th, III.36.2-3, 49.2- Athenian: Piraeus: Mostly neutral:

4: Unknown: Mytile2n e22: Oars:
427 B.C.E.: Single trier2 e2s: 380 kms: 24 hours:
Rescind command to 7.7. knots?:
put prisoners to
death:
Highest:
Pl. XXV.2: Athenian: Zakynthos: Neutral:
357 B.C.E.: Dio2n: Cape Passero: Probably both:
32 triakontoroi and 3 885 kms (by the “on the 13th day”
sailing ships: open sea): (12-12.5 days):
Invasion force: 1.0 – 1.05 knots:
Probably high:
348 CHAPTER FOUR
(Table 7 continued)


Degree of haste:
Part B: Voyages with neutral prevailing winds
DS, XX.49.1-50.6: Egyptian: Kition: Neutral:

306 B.C.E.: Ptolemy I Soter: Salamis of Cyprus: Probably oars:
140 pente2reis and 80 kms: Probably by night
tetre2reis, 200 Approach to battle: (10 hours):
transports: High: 4.35 knots:
Po, III.41.4-5: Roman: Pisa: Slightly favourable:
218 B.C.E.: Publius Cornelius Marseilles: Probably both:
Scipio snr: 420 kms: On the fifth day (4-
60 nh'e": Coasting voyage: 4.5 days):
Probably low: 2.1 – 2.35 knots:
Po, V.2.11: Macedonian: Lechaion: Neutral:
218 B.C.E.: Philip V of Patras: Probably oars:
Macedon: 130 kms: By night (10 hours):
Not stated: In transit: 6.95 knots
Probably low:
Pr, III.13.22: Byzantine: Zakynthos: Neutral:
531: Belisarius: Mt Etna: Probably both:
Not stated: 755 kms: “on the 16th day”
Opening campaign: (15-15.5 days):
Moderate: 1.1 - 1.135 knots:
Average under oars: 6.33 knots

Average under oars and sails: 1.46 knots
Overall average: 3.9 knots
(Table 7 continued)


Degree of haste:
Part C: Voyages against generally unfavourable prevailing winds
Th, VIII.101: Peloponne2sian: Chios: Slightly favourable:

411 B.C.E.: Mindaros: Rhoiteion: Oars:
73 nh'e" (presumably 235 kms: Approx. 32.5 hours:
trie2reis): Approach to battle of 3.9 knots:
Kynos se2ma:
High:
Xh, I.1.13: Athenian: Parion: Slightly adverse:
410 B.C.E.: Alcibiades: Proikonne2sos: Oars:
86 trie2reis: 55 kms: By night (10 Hours):
Approach to battle of 3.05 knots:
Kyzikos:
High:
DS, XIII.49.2-51: Athenian: Heleous: Mostly adverse:
410 B.C.E.: Alcibiade2s: Proikonne2sos: Oars:
not stated: 145 kms: By night (10 hours):
Approach to battle of 7.8 knots:
Kyzikos:
High:
Pl, XXV.4-5: Athenian: Gulf of Sidra: Adverse:
357 B.C.E.: Dio2n: He2rakleia Mino2a : Probably oars:
32 triakontoroi and 3 1130 kms: Five days:
sailing ships: Invasion force: 5.1 knots:
Moderate:
DS, XX.6.1: Syracusan: Syracuse: Generally neutral:
310 B.C.E.: Agathokle2s: Libya: Probably both:
60 nh'e": 645 kms (via Nth Till dawn on the
coast of Sicily): seventh day (6-
Invasion force: 6.5 days):
Probably high: 2.23-2.415 knots:
350 CHAPTER FOUR
(Table 7 continued)


Degree of haste:
Part C: Voyages against generally unfavourable prevailing winds
L, XXXI.23.4: Roman: Cape Sounion: Adverse:

200 B.C.E.: Gaius Claudius Chalkis: Probably oars:
Cento: 130 kms: By night (10 hours):
20 triremes: Approach to 6.95 knots:
engagement:
Moderate:
L, XLII.48.9: Roman: Straits of Messina: Slightly adverse:
171 B.C.E.: Gaius Lucretius Kephalle2nia: Probably both:
Gallus: 690 kms: On the fifth day (4-
40 quinquiremes: In transit: 4.5 days):
Moderate: 3.45 – 3.9 knots:
A, 92: Roman: Utica: Neutral:
46 B.C.E.: Julius Caesar: Cagliari: Probably oars:
Not stated: 250 kms: “after the third day”
Returning to Rome: (2-2.5 days):
Moderate: 2.25 - 2.80 knots:
RL, 196 Aragonese/Sicilian: Messina: Generally adverse:
1285: Roger of Lauria: Barcelona: Mostly oars:
Galleys: 1770 kms: Approx. 30 days:
Urgent recall: Approx. 1.45 knots:
Highest:
SL: Genoese: Cape Skillaion: Generally adverse:
1351 Simone Leccavello: Chios: Probably oars:
Single galley: 320 kms: 24-30 hours:
Carrying dispatches: 6.35-7.94 knots:
high:
Average under oars 4.75 knots

Average under oars and sails 3.0 knots
Overall average 4.4 knots
(Table 7 continued)
Part D: Overall averages
Average under oars in all conditions 5.2 knots

Average under sails in all conditions 2.85 knots
Average under sails and oars in all conditions 3.8 knots
Overall average in all conditions 3.8 knots
Average for voyages of 24 hours or less 6.5 knots

Average for voyages of 2-5 days 3.6 knots
Average for voyages of 6-20 days 2.2 knots
Average for voyages of more than 20 days 1.8 knots
inadequate to make headway against 16-knot breezes. On 1 August

1988, during the first circumnavigation of Poros island by Olympias,
rowing into a wind of 20-25 knots at 20˚ off the starboard bow, waves
of up to one metre meant that the thalamian oars could not be used
and the conditions were estimated to be just about the limit for the
zygian oars. Occasional larger waves seriously disrupted the stroke.532
Thucydides reported that in the lead-up to the first battle of
Naupaktos in 430 B.C.E., the oarsmen of the Athenian fleet under
Phormio2n were unable to recover their oars when the waves became
choppy because they were ill trained. Against this, according to
Polybios, at the battle of the Aegates islands in 241 B.C.E., the Roman
admiral Gaius Lutatius Catulus successfully took the Roman fleet to
sea against the wind in a heavy chop trusting in the skill of his crews,
who he had drilled throughout the summer at Lilybaion.533
Rowing into head winds against any sort of a sea would have
exhausted crews before many hours. On one occasion in 1992 the
crew of Olympias could sustain a speed of three knots against
headwinds gusting up to 25 knots for only 70 minutes before the crew
became exhausted.534 No doubt the hardened crews of classical
trie2reis, Byzantine dromons, and medieval galeae could perform
better than that, but by how much?
We conclude with some observations on the performance attained
by Olympias under oars during sea trials in 1988 when manned by a
------------------------------
532
Coates, et al., Trireme trials, p. 45.
533
Thucydides, Peloponnesian war, II.84.3 (vol. 1, p. 416); Polybios, Histories,
I.60.9 (vol. 1, p. 166).
534
Coates and Morrison, “Sea trials”, p. 139.
352 CHAPTER FOUR
crew of cadets from the Hellenic Navy Petty Officers’ Academy: fit
young men but with little experience as oarsmen. After three weeks’
training, on Saturday 9 June on the first leg of a voyage around the
Saronic Gulf the crew managed a speed of 3.0-3.2 knots under oars
for four hours in seas with a gentle zephyr breeze of four knots at 30˚
off the port bow. The crew also had to be resupplied with water from a
tender because their consumption increased beyond what they had
taken with them. On July 13 in Tselevinia Strait a sprint under oars
was attempted when the wind dropped to a mere one knot. The ship
attained 7.2 knots for a short while. On 14 July rowing against a
headwind of 3.3 knots, which increased to 12 knots after one hour and
35 minutes, the ship covered 21.5 kilometres in 4 hours, 45 minutes at
an average speed of 2.8 knots using only one of the three banks of
oars but alternating the oarsmen. On July 17 the ship sailed from
Epidauros. Commander Platis’s “log” reports:
On Sunday the 17th of July at 6.30 the ship sailed for the last leg of the
Saronic Gulf voyage from Epidaurus to Poros. Poor [i.e., adverse] wind
conditions prevailed for the most part of this leg, where the ship was
rowed by two files [banks] of oars [the thranite and zygian oars] for four
hours, reaching an average speed of 3.5 knots. In one part of this voyage
we had a wave height [trough to crest] of 0.8 metres from the bow [i.e.,
the wind was ahead] and useful observations about rowing conditions in
rough water [sic] were made.
It was really difficult for the oarsmen to synchronize their stroke along
the length of the ship since there were sections of the ship where the oars
were catching water since they were at the crest of a wave while at other
sections the oars were in the air being at the trough of a wave. The speed
under these circumstances was reduced to about two knots.535
Commander Platis concluded that the ship proved to be safe within the
conditions for which it was designed; that is, in wave heights of up to
about one metre.
In 1990 Olympias made a voyage under oars of 26 kilometres from
Ververouda to Tolo into a very light wind about 60˚ off the starboard
bow in four hours, 45 minutes at an average of 3.4 knots. During the
return voyage from Tolo to Poros three days later the best that the
crew could manage in light breezes, sometimes using both oars and
sails, was 52 kilometres in 6 hours, 48 minutes at an average of 4.2
knots. The ship had to be taken in tow to make its destination by
------------------------------
535
Platis, “Greek crew trials”, p. 343. We have emended tacitly commander Platis’s
English grammar and expression. Eplanations in square brackets are our own.
Figure 42
The oars of a dromon of the era of the Macedonian emperors, drawn in the
middle of the return stroke at 67˚ to the centre line.
© John H. Pryor
nightfall.536
Over any extended period of time the best that Olympias could
accomplish under oars, even in virtually millpond conditions, was
around 3-4 knots. Even in what by the Beaufort Scale are at worst
light to moderate breezes ahead, speed under oars was reduced to
around two knots.
Our conclusions may appear to stretch the limits of credibility.
What sort of a ship was it that was capable of what appears to have
been such minimal performance and yet was a renowned warship at
the front-line of the defence of the Empire for centuries? The simple
answer, of course, to reiterate, is that ships designed for a specialized
purpose, to pack the maximum punch in battle in calm sea conditions,
would not have been suitable for other purposes. Would an emperor
wanting to send a message from Constantinople to Cherso2n in
October, or a strate2gos of the Kibyrrhaio2tai at Antalya wanting to
send a message to Constantinople in March, use a dromon? We think
------------------------------
536
Shaw, “Voyage and speed trials”, pp. 40, 42.
354 CHAPTER FOUR
not. They would have used sailing ships capable of riding the waves
and holding the sea in rough weather and of pointing into the wind on
a tack. There can be no doubt that dromons were superb for the
purposes for which they were designed. They would not have become
and remained for centuries the battle galleys par excellence of the
Empire had they not been. But that purpose was for battle and battle
alone. No wonder that Leo VI recommended that a strate2gos should
take his fleet out to engage the enemy only in calm conditions.537
According to the fourth part of the Theophane2s continuatus,
probably written by Theodore Daphnopate2s , who should have known,
when Ro2manos II proposed to send an expedition to Crete under
Nike2phoros Pho2kas in 960 resistance to the proposal in the Senate was
only overcome by the parakoimo2menos Joseph Bringas, who urged
that the length of the journey should not be feared.538 In the tenth
century even a voyage from Constantinople to Crete was regarded as a
long-range expedition, a major undertaking, for a large fleet. Naval
warfare was a matter of coasting for very limited distances and
developing strategies which combined possession of the coasts and
islands with what naval forces could achieve as regards control of
coastal sea lanes. Even though we do not believe that the
Stadiodromikon for the Cretan expedition of 949 actually reflected
passages from mooring to mooring, it is nevertheless significant that
no passage mentioned was longer than around 115 kilometres and
above and beyond all other considerations it was access to fresh water
supplies which determined that this should have been so.539
Supplies of fresh water were vitally important because it was the
“fuel” which drove any galley. Unless a galley could use its sails, it
would come to a stop within hours if its water ran out because
dehydration would quickly enfeeble the oarsmen. Human beings have
very poor resistance to dehydration. Naval forces had to provide for
fresh water and provisions in advance or ensure that they could obtain
them en route, and water could be a precious commodity in many
parts of the Mediterranean during summer. Few ports in the Eastern
Mediterranean were on large rivers and many had no river at all and
were dependent upon wells. Some did not even have those; for
example, Metho2ne2, which depended on cisterns. Moving into waters
------------------------------
537
Appendix Two [a], §31. Cf. Appendix Five, §29.
538
Theophanes continuatus, VI.Basileiva Rwmanou' uiJo u' Kwnstantivnou tou'
porfurogennhvtou.9 (p. 475): “... kai; mh; dedievnai th'" oJdou' to; mh'ko" kai; ...”. On
Theodore Daphnopate2s see above p. 188, n. 62.
539
See above pp. 264-6.
off enemy shores deprived fleets of water unless they could take it by
force, which was usually not easy to do since most significant coastal
water sources had been incorporated into fortified habitations.
“Foraging” for water from small streams in deserted coves or wells in
isolated villages taken over by force might be possible, even if
dangerous, for single ships or small flotillas, but such sources would
be inadequate for large fleets.
Galeae replaced dromons and chelandia in the late eleventh
century in the West almost certainly because it was discovered that
bireme galleys could be rowed from two bench positions above deck
rather than from two superimposed benches. Among other problems
overcome would have been that of ventilation of the hold. Fifty
oarsmen working below deck would have emitted large amounts of
body heat, carbon dioxide and impurities, and sweat, even if there
were no horses aboard. Removal of this and replacement of oxygen
would have required forced ventilation.
In human beings only around 24% of the potential energy stored in
fuels in the body is converted to mechanical work. The remainder is
expended as heat. In the Daedalus Project, which culminated in a
man-powered flight of four hours from Crete to The2ra, it was
calculated that the pilot would need to produce approximately 3-3.5
watts of mechanical power per kilogram of body weight, which would
require 14.6 watts per kilogramme of fuel oxidation and this would
require 44 mililitres of oxygen per minute per kilogramme. It was
calculated that the 68-kilogramme pilot with a mechanical efficiency
of 24% would produce about 13 watts per kilogramme of metabolic
heat, around 900 watts, of which around 225 was in the form of work
and the remaining 675 watts needed to be dissipated.540
Coates calculates that men working hard and producing a
maximum of 400 watts on the oar breathe out 100 litres of air a
minute containing 4-4.5% carbon dioxide and that their thermal
efficiency as heat engines is about 20%. They produce about 1,500
watts of heat which must be removed by evaporating water in the
lungs and sweating at up to two litres an hour. A normal workload of
150 watts requires removal of only around 600 watts of heat, with a
corresponding reduction in sweat. Extrapolating from Coates’s
calculations, to ventilate the 50 men of the lower oarcrew of a
dromon, each working at about 150 watts on the oars, 5 cubic metres
of air would have to be drawn into and expelled from the ship’s hull
------------------------------
540
Nadel and Busolari, “Daedalus project”, pp. 351, 359.
356 CHAPTER FOUR
every minute to keep the concentration of carbon dioxide down to a

practicable upper limit of 2%. But, in the very humid conditions
aboard, about 5 times that amount would be needed to get rid of the
evaporating sweat; total 25 cubic metres per minute, 1,500 per hour.
Men working at 150 watts in still air can maintain the rate for only
around a half an hour but for much longer in moving air. Because of
the low height in the hold, about 8 square metres of entry area and a
little more exhaust area would have been needed and there would have
had to have been forced ventilation.541 Merely having hatches of that
area in the deck would not force fresh air down into the hold. The sea
trials of Olympias, have shown that even though that ship does not
have a full deck and is open to the sky at the centre line, ventilation at
the level of the lowest bank of oars, the thalamian oars, is “barely
adequate”.542 In a dromon or any galley with both a full deck and also
a bank of oarsmen below deck there must have been forced ventilation
of the hold by some contrivances such as windsails and/or cowls.
In fact Coates’s estimates would probably have been too low. If we
consider the need to maintain the level of carbon dioxide plus
impurities including water vapour at no more than 0.06% so that the
lower oarsmen could work effectively for an extended time, then
using De Chaumont/Smith’s equation, according to which human
beings expel 0.6 cubic feet of carbon dioxide plus impurities per hour,
as opposed to the 3 cubic feet of horses, each man would have
required 85 cubic metres of fresh air per hour.543 Fifty men would
have required 4,250 cubic metres of fresh air per hour and the volume
in the hold of a dromon, calculated at around 140 cubic metres would
have had to have been changed completely around 30 times an hour.
Oarsmen working hard would have needed large amounts of water.
Estimates of the amount of water crews require have been revised
upwards dramatically in the past few decades and around eight litres
per day for galley crews is increasingly supported by a range of
evidence from antiquity through to the seventeenth century.544 The
Informationes pro passagio transmarino of Marseilles of 1318
------------------------------
541
See Morrison, Greek and Roman oared warships, pp. 286, 326.
542
See Coates and Morrison, “Sea trials”, p. 140; Morrison, et al., Athenian trireme,
pp. 238, 274.
543
See De Chaumont, “On ventilation”, p. 1031.
544
Earlier estimates of water requirements, as low as four pints (2.25 litres) per day,
are now regarded as hopelessly inadequate. See Sleeswyk and Meijer, “Water supply
of the Argo”, pp. 133-5; Dotson, “Economics and logistics”; Pryor, Geography,
technology, and war, pp. 75-85; idem “From dromo2n to galea”, p. 114; idem,
“Geographical conditions”, p. 210.
specified 3.75 millayrole (238 litres) of water per man for 60 days: 4
litres per day.545 This was for men who were passengers only and must
therefore have been a bare minimum.
During sea trials of Olympias in June-July-August the oarsmen
sweated profusely and needed a litre of water per hour, just for
drinking, to prevent dehydration.546 During the Daedalus project, one
litre of water an hour was needed by a man producing 210 watts for
four hours. The consumption rate was 0.005 litres/watt.hour.547 So an
oarsman producing 150 watts per hour for 8 hours, a sustainable
figure if there was adequate ventilation, would need 6 litres of water
for drinking alone. To that one should add another two litres per man
per day for other needs, especially for the the soupy stew of salt meat
and legumes that was the staple accompaniment to biscuit in the diet
of medieval crews. On French galleys of the seventeenth century the
allowance was 7 litres per man per day. Eight litres per day is also an
accepted requirement for troops in moderate work.548
If we consider the standard ousia of 108 men of Byzantine
dromons or chelandia, the water requirement can be expected to have
been a minimum of 108 x 8 = 864 litres per day. It would be
reasonable to increase that to at least 1,000 litres or one tonne of water
per ship per day when officers and marines are taken into account.
And this would be to discount supernumeraries, dromons with two
ousiai, and the various higher figures for crews discussed above. One
tonne of water per day may be expected to have been an absolute bare
minimum for galleys moving under oars in summer for eight hours a
day, although such a figure would obviously vary according to
whether the sky was overcast, whether there was a cool breeze,
------------------------------
545
De Boislisle, “Projet de Croisade”, pp. 253-4.
546
Coates and Morrison, “Sea trials”, p. 138; Rankov, “Reconstructing the past”, p.
138. Cf. Platis, “Greek crew trials”, p. 340; Morrison, et al., Athenian trireme, p. 238.
During a passage in a calm on a very hot afternoon on 4 August 1988, the
thalamian oarsmen of Olympias suffered particularly badly from the heat and
dehydration during a pull under oars of some 4.5 miles down the east side of Poros
island. See Coates, et al., Trireme trials, p. 46.
547
See Nadel and Bussolari, “Daedalus project”. Cf. Morrison, Greek and Roman
oared warships, pp. 326-7. But note that in the Daedalus project the pilot’s water was
loaded with 10% glucose and 0.4 grammes per litre of sodium. Without these
additives, pure water alone would not have permitted the pilot to sustain the flight.
One wonders whether ancient and medieval oarsmen learned to add very small
amounts of seawater to their drinking water, or whether they simply gained the salt
they needed from the salted meat and fish which was part of the normal diet. The
replacement of the glucose, however, would not have been so easy.
548
See Burlet, et al., “Comment pouvait-on ramer”, pp. 152-3; Wolseley, Soldier’s
pocket-book, p. 95.
358 CHAPTER FOUR
whether sails could be used, and a host of other factors. The lower
oarsmen of dromons, rowing in an enclosed space below deck, would
undoubtedly have consumed more water than those above deck and
probably more than even the thalamian oarsmen of Olympias.549
That for ships to run out of fresh water at sea was common is
suggested by two tales included in the Spiritual meadow of John
Moschos. In the first, an anchorite named Theodore, bound for
Constantinople by ship, turned sea water into fresh when supplies ran
out. In the second, a pious naukle2ros, ship master, bound for
Constantinople, prayed for rain for four days to relieve the distress of
crew and passengers who had foolishly exhausted their water supplies.
He was rewarded by a shower confined to the area of the ship, whose
course the cloud followed.550 The author probably intended these ships
to be understood to have been sailing ships and if such distress could
be occasioned by their running out of water, the dimensions of the
problem must have been infinitely greater for galleys.
In 306 B.C.E. the Macedonian fleet, under the command of the
later king De2me2trios I Poliorke2te2s, sailed from Gaza for Egypt.
Caught by a storm off an inhospitable and enemy-controlled shore, it
was forced to ride it out at anchor off the harbourless Kasion, which
was only about 140 kilometres from Gaza. By then they were already
out of water. Diodo2ros Siculus did not say how many days they were
at sea, but the short distances involved suggest that the fleet was
carrying only very limited water supplies.551 In the twelfth century,
Fulcher of Chartres reported that in 1126 a Fa2t6imid fleet raiding the
shipping lanes and coasts of the Kingdom of Jerusalem ran out of
water off Beirut and the crews were forced to land to try to fill their
“buckets”, situle, from springs and streams. They were cut to pieces
by forces from Beirut and forced to evacuate and flee towards Cyprus,
no doubt to try to water in some deserted cove.552
When Genoa was at war with Mamlu2k Egypt in 1383, Pietro
Piccono was sent with four galleys on an embassy from Famagusta to
Beirut on 31 July, arriving on 2 August, around tierce. Being asked to
wait some days for a response but his request for reprovisioning being
------------------------------
549
Rankov reports that when rowing Olympias the ship was hotter and stuffier the
lower down in it an oarsman was, even though Olympias did not have a full deck, and
that the thalamian oarsmen also suffered from a shower of sweat from the oarsmen
above them. See his “Reconstructing the past”, p. 138. Cf. Morrison, et al., Athenian
trireme, p. 238.
550
John Moschos, Spiritual meadow, §§173, 174 (coll. 3041, 3041-4).
551
Diodo2rus Siculus, Bibliothe2ke2 historike2, XX.74.1-3 (vol. 10, pp. 338-40).
552
Fulcher of Chartres, Historia Hierosolymitana, III.56 (pp. 804-5).
refused, he left and tried to take on water somewhere 10 miles from

Beirut but was repulsed. He was compelled to take it on at the Nahr
al-Ibra2hı3m. Although again refused by the Muslim authorities, he took
the water by force and returned to Famagusta, arriving on 4 August
around tierce. Piccono’s report on what happened, dictated at
Famagusta on 4 August, especially his report on the Muslim
authorities’ refusal to allow him to water anywhere, was confirmed in
a report of his superior Niccolò Maruffo to the Doge of Genoa on 5
September. Ashtor and Kedar explain Piccono’s behaviour in
attempting to water by force as an attempt to provoke a casus belli,
considering that his galleys could not possibly have run out of water
in such a short time.553 However, Beirut is about 240 kilometres south-
east of, and down-wind from, Famagusta. Piccono’s galleys had taken
two days and a few hours to reach Beirut, most probably using their
sails. But 2 August was occupied with negotiations with the governor
of Beirut and it was midsummer and the return could be expected to
have to be made under oars against head winds once away from the
coast and the diurnal south-westerly sea breezes. Piccono would have
to count on at least two and, to be safe, three or four days to return to
Famagusta. By the fourteenth century there is no doubt that Western
galleys could carry water for at least five days and probably more if
necessary, but whether Piccono’s galleys were carrying maximum
supplies is another question. They were not on campaign but rather on
an embassy. It is quite possible that by 2 August Piccono no longer
had sufficient water to be confident about the return voyage and this
may explain his actions. As it eventuated, he appears to have made the
return voyage as quickly as the outward one, which was fortuitous to
say the least.
The weight of fresh water must have been a real problem.
Thirteenth-century Sicilian galleys, the earliest for which we have
construction details, had only around 50 centimetres freeboard
amidships and one tonne of water would sink them by a centimetre or
so.554 No more than a quarter or so of their 40 tonnes of deadweight
tonnage could have been used for water because of the weight of the
food, armaments, equipment, spare gear and myriad other essentials
required. The weight of the crews would have been another 10 tonnes
or so. Later evidence from Genoa suggests that light war galleys could
carry between 4 and 8 tonnes of water. An inventory for the galley of
------------------------------
553
Ashtor and Kedar, “Una guerra”, pp. 14-16, 38-9, 40.
554
Pryor, “From dromo2n to galea”, pp. 112-3.
360 CHAPTER FOUR
Simone Leccavello undertaking a long voyage to the Aegean in 1351

specified 54 water barili, each of 79.5 litres, total 4.3 tonnes.555 Three
other inventories of 1400 and 1402 specified 70 barili (5.5 tonnes), 72
barili (5.7 tonnes), and 48 barili (3.8 tonnes).556 The highest figure
known to us is for 10 galleys of a fleet equipped at Savona in 1476
carrying on average 99.8 Genoese barili each: 7.9 tonnes.557
Our estimate of the deadweight tonnage of the smaller Byzantine
dromons is only around 29.5 tonnes, of which around 8.5 tonnes
would be the weight of the oar crew, the ousia, alone. It is extremely
improbable that they could have carried much more than around 4.5-5
tonnes of water, enough for no more than four days or so if moving
under oars, more if the sails were being used, less if carrying more
than a single ousia.
It is no wonder that Syrianos Magistros, followed by Nike2phoros
Ouranos recommended that not only strate2goi but also each and every
ship in a fleet should have aboard seamen familiar with the coasts and
where fresh water could be obtained. As Nike2phoros wrote:
It is appropriate for a strate2gos to have with him men who have accurate
knowledge and experience of the sea in which he is sailing, which winds
cause it to swell and which blow from the land. They should know both
the hidden rocks in the sea, and the places which have no depth, and the
land along which one sails and the islands adjacent to it, the harbours and
the distance such harbours are the one from the other. They should know
both the countries and the water supplies;558 for many have perished from
lack of experience of the sea and the regions, since winds frequently blow
and scatter the ships to one region and another. And it is appropriate that
not only the strate2gos should have men with this knowledge we have
discussed but also each and every ship should have someone knowing
these things to advise well when appropriate.559
------------------------------
555
Archivio di Stato di Genova, Antico Comune, Galearum introytus et exitus, No.
690. Reference courtesy of M. Balard.
556
Musso, “Armamento”, pp. 39-41, 41-3, 43-6, 59-60, 71-6.
557
Varaldo, “Inventario”, p. 91.
558
Nike2phoros Ouranos and Syrianos Magistros almost certainly meant “fresh
water” by u{data (hydata). The knowledge required was that of where to obtain
precious fresh water, rather than that of the “waters”; i.e., the seas.
559
Nike2phoros Ouranos, Ek to2n taktiko2n, §119.1.1-3 (p. 93), (checked by us against
the manuscript, Oxford, Bodleian Library, MS. Baroccianus Graecus 131): “ÔArmovzei
to;n strathgo;n e[cein meq eJautou' tou;" ginwvskonta" ajkribw'" th;n pei'ran th'" qalavssh"
eij" h}n plevei, to; poi'o i a[nemoi kumaivnousin aujth;n kai; to; poi'o i fusw'sin ajpo; th'" gh'": i{na
de; ginwvskwsi kai; ta;" kruptomevna" pevtra" eij" th;n qavlassan kai; tou;" tovpou" tou;" mh;
e[conta" bavqo" kai; th;n parapleomevnhn gh'n kai; ta;" parakeimevna" aujth'/ nhvsou", tou;"
limevna" kai; to; povson ajpevcousi oiJ toiou'toi limevne" ei|" ajpo; tou' a[llou: i{na de;
ginwvskwsi kai; ta; cwriva kai; ta; u{data: polloi; ga;r ejk tou' e[cein ajp eirivan th'" qalavssh"
Figure 43
The cista Ficoronica: Jason and the Argonauts watering at the spring of
the Bebrycians.
Some evidence suggests that Greek trie2reis did not actually carry
water supplies at all for more than the duration of a single passage.560
In what containers did Byzantine galleys carry their water supplies,
if any? It is curious that neither amphorae nor barrels were mentioned
by Leo VI or Nike2phoros Ouranos. However, according to one of the
inventories for the Cretan expedition of 949, the Department of the
------------------------------
kai; tw'n tovpwn ajpwvlonto, ejpeidh; fusw'si pollavki" a[nemoi kai; skorpivzousi ta; ploi'a
eij" a[llon kai; a[llon tovpon. Kai; aJrmovzei i{na mh; movnon oJ strathgo;" e[ch/ tou;"
ginwskonta" o{per ei[pamen, ajlla; kai; e}n e{kaston ploi'o n i{na e[ch/ to;n tau'ta
ginwvskonta, pro;" to; bouleuvesqai kalw'" to; sumfevron.”. Cf. Appendix One, §5.1-3.
560
Herodotos, Histories, 8.22 (vol. 4, p. 20); Thucydides, Peloponnesian war,
VI.34.5, 42.1 (vol. 3, pp. 246, 260). See also Morrison, et al., Athenian trireme, pp.
95, 102.
362 CHAPTER FOUR
Eidikon should have provided 100 kavdoi (kadoi) for 20 dromons.561

There is no doubt that kadoi were amphorae for water supplies in
classical Greece. Those depicted on the “cista Ficoronica” of ca 300
B.C.E., in a representation of Jason and the Argonauts watering at the
spring of the Bebrycians, are estimated to have weighed around 18
kilogrammes for a capacity of 27 litres.562 Sleeswyk and Meijer
provide no authority for their estimate of the capacity of the kadoi of
the cista Ficoronica. However, they are almost certainly correct. It
seems that amphorae with a capacity of around 25-28 litres, weighing
dry around 15-17 kilogrammes, for a total weight of 40-45
kilogrammes, were at the limit of what a single man could lift and
carry on his shoulders, given their awkward shape.563
The water pithos found in the seventh-century Yassı Ada wreck
was larger than these but had no handles and was not portable.564 It
was designed to be installed in place and then filled. The kadoi of the
cista Ficoronica had two handles so that they could be carried and
ergonomic considerations would have limited how large they could
have been. The five kadoi per dromon that the inventory for the 949
expedition to Crete allowed could have contained no more than
around 135 litres and it would not have mattered much what century
was in question, a mythological Greek past or the tenth century C.E.,
they could not possibly have been main storage vessels for water.
There is no other mention in the inventories of amphorae which
might have been for water and the only items which might possibly
have had the sense of “barrels” were 30 bronze boutiva (boutia) listed
among things provided by the Department of the Vestiarion basilikon
to the droungarios tou ploimou for the 949 expedition.565 The word
------------------------------
561
See Appendix Four [b], §III.20 [= Haldon, “Theory and practice”, p. 227;
562
Sleeswyk and Meijer, “Water supply of the Argo”, p. 133.
The “cista Ficoronica” is a bronze water urn from Praeneste, south-east of
Rome, dated to the second half of the fourth century B.C.E. and of Greco-Etruscan
workmanship. It was acquired by the antiquarian Francesco Ficoroni in 1738. It is
now in the Villa Giulia museum of Etruscan antiquities in Rome, Inv. No 24787. See
Dohrn, Ficoronische ciste.
563
See Wallace, “Amphora capacities”; Wallace Matheson, “Rhodian amphora
capacities”; Peacock and Williams, Amphorae and the Roman economy, Table 1 (p.
52).
564
Bass and Van Doorninck, Yassi Ada. Volume I, pp. 186-8.
Unfortunately no estimation of the contents, volume, and weight of the pithos
was made when it was reconstructed from the fragments found and the director of the
museum at Bodrum then put it on display in the open air, where it promptly
disintegrated and had to be thrown away.
565
See Appendix Four [b], §VII.21 [= Haldon, “Theory and practice”, p. 233;
boutia may have been a form of bou'tzi (boutzi) or bou'tti" (bouttis)

later used for a barrel. However, since these boutia were made of
bronze, they were obviously not water barrels. They were probably
buckets.
By the tenth century it is probable that Byzantines were still using
amphorae as water containers but may well have begun to use barrels
also. There is a reference in a Cairo geniza letter dated to the middle
of the eleventh-century, associated with the Maghribin merchant
Nahray ben Nissı3m, who had settled in Egypt, to buckets which held
the equivalent of “half a Byzantine barrel”.566 Here the Arabic word
used for a “barrel” was bty [battiya/bittiya/buttiya]. It is clear that by
that period Byzantines were well known to use barrels. How much
earlier than this they had been doing so is unclear; however, the tenth-
century Parangelmata poliorke2tika attributed to He2ro2n of Byzantium
has at folio 7v a clear picture, the earliest known to us, of a barrel,
even though the author had no word, or at least no classical word, for
the object. The caption to the illustration reads “cylindrical vessel”,
aggei'on kulindrik[ovn] (angeion kylindrik[on]).567
It is true that the selective and incomplete nature of the inventories
for the Cretan expeditions preserved in the De cerimoniis means that
their failure to mention either barrels or amphorae cannot be regarded
as hard evidence that neither were in fact used for water supplies.
Nevertheless, it leaves open the possibility that there may have been
some other kind of water containers. In the inventories for the Cretan
expedition of 949, it was specified that the Department of the Eidikon
should supply for 20 dromons: “5 sheets of lead each for the
------------------------------
Reiske translated the 100 koubavria (koubaria) also mentioned in the
inventories as cupae, casks or barrels. However, as we have seen above, these were
almost certainly not barrels but rather reels or windlasses of some sort associated with
the mooring cables. See above p. 214 and n. 154.
566
Goitein, Mediterranean society, p. 321.
The letter is in the Taylor-Schechter collection of Cambridge University
Library, MS. TS 12.241. It is written in Judaeo-Arabic, medieval Arabic written in
Hebrew script. The relevant lines are recto 6-7, transliterated as follows: “... n‘ml fy
’l’nbb’ b’lnwb’ kmsyin dlw w’ldlw ’ldy ystq’ bh ns5f bty’ rwmy’ ...”, translated as “...
we laboured at bailing by turns of fifty buckets and each bucket holds half a
Byzantine barrel ...”. The Arabic dlw meant a “bucket”, bty’ a “barrel”, and rwmy’
“Roman” or “Byzantine”. We are indebted to Mr. Ben Outhwaite of the Taylor-
Schechter Unit of Cambridge University Library for his assistance.
567
He2ro2n, Parangelmata poliorke2tika, §5 (pp. 36-7 and fig. 1). The anonymous
author described barrels as: “... kai; oi|a ta; ejk sanivdwn kukloterw'" sunhrmosmevna kai;
desmoi'" e[xwqen perieilhmmevna ta; pro;" uJpodoch;n oi[nou ginovmena ejl aivou te kai; panto;"
uJgrou: ...” (“... and the kind [of thing] ‹made› of planks fitted together in a circle and
surrounded with bands on the outside which are for storage of wine and oil and every
liquid, ...”).
364 CHAPTER FOUR
kalumbovmatoi (kalymbomatoi), total of 100 sheets, weighing 3,000

litrai” and also “20 hides for the kalubovmatoi”. Elsewhere another
inventory recorded that the Department of the Eidikon supplied for the
kolymbomatoi of the chelandia of the fleet: “5 sheets of lead each, 100
sheets, that is 3,000 litrai”.568 This may have been merely a repetition
of the specification for the dromons revealing the confusion between
chelandia and dromons. Many more than 20 chelandia were involved.
Ka(o)ly(m)bomatos is an unknown word, for which many meanings
have been suggested, including sheathing for the hull. However, this
makes no sense. Five sheets of lead per dromon would be of no use
for a hull except, perhaps, for patching.
We have several reasons for this opinion. First, the lead sheathing
found on the Kyrenia wreck of the fourth-century B.C.E., a small
merchant ship only around 15 metres long, weighed over 1,200
kilogrammes.569 Secondly, the 150 Byzantine litrai, which the five
sheets of lead weighed, were equivalent to only around 48
kilogrammes. One kilogramme of lead has a volume of approximately
84 cubic centimetres, 48 kilogrammes of 4,032 cubic centimetres.
Now, the lead sheathing found on ancient wrecks was normally one to
two millimetres thick.570 It is almost impossible that Byzantine
technology would have been capable of producing sheet lead less than
one to two millimetres in any case. Taking the lowest conceivable
figure for the thickness, one millimetre, the 48 kilogrammes of lead
would cover an area of only 4.32 square metres. Obviously, sheets of
lead measuring only around 4 square metres cannot have been used
for sheathing hulls. Thirdly, as we noted above,571 sheathing of hulls
with lead appears to have disappeared from the end of the first century
C.E. None of the later wrecks so far excavated had it. Fourthly, why
would anyone sheathe with lead the hull of a warship, which above all
required a light hull for speed and manœuvrability in battle? Enough
lead to sheathe the hull of a dromon some 30 metres long and 3.8
metres in the beam would weigh a great deal and would sink the hull
in the water considerably, slowing the ship down commensurately.572
Our calculation suggests that over two tonnes of lead would have been
------------------------------
568
Appendix Four [b], §§III.1-2, VI.28 [= Haldon, “Theory and practice”, pp. 227,
233; Constantine VII, De cerimoniis, II.45 (vol. 1, pp. 671, 676)].
569
See Hocker, “Lead hull sheathing”, p. 200.
570
See Hocker, “Lead hull sheathing”, p. 197; Blackman, “Hull sheathing”;
Blavatsky and Peters, “Donuzlav wreck”, p. 25; Steffy, “Kyrenia ship”, pp. 83-4;
Fitzgerald, “The ship”, p. 168.
571
See above pp. 147-9 and n. 69.
572
Cf. Morrison, et al., Athenian trireme, p. 186.
needed. Fifthly, the following entry in the inventory also specified

twenty hides for the kaly(m)bomatoi. But no one would ever have
wrapped the underwater hull of any warship in hides. They would
have soaked up water and slowed the ship immeasurably.
Reiske associated kaly(m)bomatos with kaluvbh (kalybe2), a hut,
suggesting that it was the commander’s berth and that the lead and
hides were to roof it and protect its sides in battle. However, in fact
krabatos was used for the berth. It is possible that the inventorist was
employing a non-technical term; however, why would he do so when
the rest of the inventory was replete with technical terminology? It is
certainly probable that in battle the commander’s berth would have
been protected in some way but noone would use a metal as weak but
heavy as lead for this. Bronze, or virtually any other metal, would
have been far better for the purpose. Moreover, the area that the 48
kilogrammes of lead would have covered at one millimetre thickness
would have been only around 2 by 2 metres. And, in any case, a lead
roof only one millimetre thick would not withstand much of an
impact. Any javelin falling from a height would go straight through it.
Even at, say, two millimetres thickness, it would provide little defence
against heavy missiles and this would reduce the area protected to a
mere 2.016 square metres. Reiske’s suggestion is extremely
improbable, and in fact krabatoi were almost certainly made of
wooden frames with cloth tent coverings, as discussed above.573
We believe that Haldon has provided the clue to understanding
kaly(m)bomatos. He suggests that the word should be kolumbovmaton,
derived from kovlumbo" (kolymbos), for a well or sump, and may thus
have referred to a lead sump in the ship’s bilges.574 A passage in one
of the Miracles of St Artemios, a miracle which was probably written
down in its surviving form between 658 and 668, although possibly as
late as ca 692-700, referred to a ship’s carpenter who repaired a
problem with a ship’s keel at the place where there was a kovlumbo"
(kolymbos).575 The root koluvmb-, when combined with various
suffixes, had reference to “swimming”, “diving”, “drawing water”,
------------------------------
573
See Constantine VII, De cerimoniis, vol. 2, p. 794. Cf. above pp. 215-16 & n.
156.
574
See Haldon, “Theory and practice”, pp. 277-8.
575
Crisafulli and Nesbitt, Miracles of St Artemios , miracle 27 (p. 153): “h\n de; pro;
ojlivgou to; ploi'o n, ejn w|/ e[pleen, uJpomei'navn ti kata; th;n trovp in, kai; e[tucen katelqei'n
aujto;n to;n nosou'nta koluvmbw/ kai; poih'sai o{per e[crh/zen to; ploi'on.”.
The original composition of the collection of miracles of St Artemios may go
back to the late fourth to early fifth centuries, and with that the use of the word
kolymbos.
366 CHAPTER FOUR
etc. A kolumbhvqra (kolymbe2thra) could be a “swimming place”,

“vat”, “cistern”, “reservoir”, or even a “baptismal font”. This is surely
the clue. A kaly(m)bomatos must have been a tank of some sort.
What sort of tank could be made from five sheets of lead and a
hide? Five sheets of lead suggest a open-topped tank. The hide may
have been used to cover the top at sea in order to prevent spillage.
We conclude that kaly(m)bomatoi may have been water tanks.
There is support for an hypothesis of water tanks being built from lead
in Moschio2n’s account of the great ship of Hiero2 II of Syracuse as
preserved by Athe2naios of Naukratis. According to Athe2naios,
Moschio2n wrote that: “There was an enclosed water tank at the bow,
holding 20,000 metre2tai, constructed from planks and pitch and cloth.
Alongside it was constructed from lead sheets and planks an enclosed
fish-rearing [tank]; ...”.576 Although it was the fish tank rather than the
water tank which was constructed of lead, obviously the fish tank also
held water and there is therefore no reason why a water tank for
drinking purposes could not have been constructed in the same way. A
lead water tank, possibly part of a bilge-pump system, was found on
the Chiessi wreck of ca 60-85 C.E. off Elba.577
If the lead was hammered out to a thickness of one millimetre,
supported by a wooden frame, the 4.032 square metres of sheet lead
could make a tank around 90 by 90 by 90 centimetres, or any variation
of that, containing around 730 litres of water. For crews of around 108
oarsmen, this would have provided a water supply of around 6.75
litres per man, around a daily allowance. If it was also shared by the
officers and marines, then the allowance would obviously have had to
have been less. A figure of this order may have been just about right if
dromons moved for seven to eight hours per day on their voyage to
Crete. We hypothesize that a kaly(m)bomatos was a lead-lined tank
with a hide “lid”, containing a day’s water supply for the oarsmen.
The only other hypothesis conceded as a real possibility for the
meaning of ka(o)ly(m)bomatos is a tank or “sump” in the hold from
where bilge water was pumped out, or by extension a pump itself.
Wooden ships always had limbers, or limber holes, cut in the
underside of the floor timbers near the centre line to allow bilge water
to drain down to the lowest point, where there was a well, or sump, or
------------------------------
576
Athe2n aios of Naukratis, Deipnosophistae (Gulick), 5.208 (vol. 2, p. 440): “h\n de;
kai; uJdroqhvkh kata; th;n prw'/ran kleisthv, discilivo u" metrhta;" decomevnh, ejk sanivdwn
kai; pivtth" kai; ojqonivwn kateskeuasmevnh. para; de; tauvthn kateskeuvasto dia;
molibdwvmato" kai; sanivdwn kleisto;n ijcquotrofei'on: ...”.
577
Parker, Ancient shipwecks, p. 301.
bilge box which could be pumped dry. However, there is a

considerable amount of archaeological evidence for chain pumps and
bilge sumps surviving from antiquity and the early Middle Ages and
in all cases the pumps and sumps were wooden. That being said, there
is also evidence for lead evacuation tanks at the heads of pumps above
deck, connected to lead pipes which may have served to drain bilge
water off to one side or the other as the deck happened to be canted at
any time.578 These are always found in the upper strata of the wrecks,
frequently on top of the cargo, indicating that they have fallen down
from above when the deck finally collapsed. They are never found in
the lowest strata at the keel. It is possible that the lead for the
ka(o)ly(m)bomatoi may have been intended for this purpose.
Nevertheless, we prefer the association of kaly(m)bomatoi with
water tanks rather than bilge pumps firstly because we cannot imagine
what the hides could have been used for if the latter was the case and
secondly because there were perfectly well known words in Greek for
pumps: kocliva" (kochlias) for an Archimedes’ screw-pump and
ajntlhthvrion (antle2te2rion) for a bilge pump or bilge bucket.579
There is no mention in the Byzantine sources related to the Cretan
expeditions or naval warfare of anything that might have been a water
skin, such as buvrsa (byrsa), flavskh/flaskivon (phlaske2/phlaskion), or
ajskov"/ajskodau'la (askos/askodaula) and no way of estimating what
their contents may have been even if there was.580
------------------------------
578
Carre and Jézégou, “Pompes à chapelet”; Foerster-Laures “Bilge and pump”;
Parker, Ancient shipwecks, pp. 143, 204, 237, 247, 373; Santamaria, “L’épave
Dramont”, pp. 171-4.
579
The dictionary definitions invariably give “bucket” rather than “pump”.
However, pump fits the context of many of the passages in question much better than
bucket, as it does also for the Latin equivalent: sentinaculum. Jal appreciated that the
word might mean a pump. See Casson, Ships and seamanship, p. 176; Carre and
Jézégou, “Pompes à chapelet”, p. 136; Jal, Glossaire nautique, p. 146.
The term also appears to have been understood as a “water raiser” in the Greek-
Latin Cyril glosses of London, British Library, Ms. Harley 5792 in Goetz, Glossarii
Latini, vol. 2, p. 231, l. 5: “Antlhthrion [h]auritorium [sic]”. See “Note on citations of
Greek and Latin glossaries”, p. lxix above.
580
Buvrsa and ajskov" were both classical words for wine or water skins. Askodau'l a
is found in two Byzantine contexts which suggest that it may have been a word for
water skins. In the first it is found as ajskodavblai in the Praecepta imperatori Romano
bellum cogitanti ... observanda amongst a list of equipment in the imperial wardrobe
when on campaign. See Constantine VII, Three treatises, Text C (p. 106). The word
is not qualified there, although Haldon does translate it as water-skins, but in the
Taktika of Leo VI it is quite clear that ajskodau'la referred to water skins. See Leo
VI, Taktika (Vari), XII.123 (vol. 2, p. 99): “Crh; ou|n kai; ejn hJmevra/ tou' polevmou
e}kaston stratiwvthn ejn tai[" sevllai" aujt w'n ejpifevr esqai u{d wr eij" ta; legovmena
flaskiva, kai; paxamavtin ejn tw'/ sellopouggivw/, ...” (“And then in time of war each
soldier ought to have on his saddle, water in the so-called phlaskia, and biscuits in his
368 CHAPTER FOUR
On bireme dromons stowage of water, as well as its weight, must

have been a problem. With two files of oarsmen below deck, where
could water have been stowed? A 245 litre (54 gallon) capacity barrel
would have a diameter at its pitch of 28 inches (71 centimetres), a
head diameter of 23 inches (58.5 centimetres), a height of 36.5 inches
(92.5 centimetres), and would occupy a cylindrical space of around
370 litres.581 The only place in the hold of a dromon where barrels
such as that could have been stowed would have been on the floor
down the centre-line, but that space would surely have had to be
reserved for the long spare gear that had to be carried: rudders, oars,
yards, even masts.582 Such gear could obviously not have been carried
above deck on a warship, unless it was abandoned before battle.
In the case of galleys such as dromons with files of oarsmen below
deck, in addition to the possibility of a lead water tank, we are driven
to the conclusion that oarsmen carried their own water supply in a
kados, or small barrel, or perhaps a skin. The ability to carry much
larger amounts of water in a centralised water supply would in fact
have been another huge advantage that the galea would have had over
galleys like dromons when it was developed in the West towards the
end of the eleventh century. If oarsmen had their own small amphora,
barrel, or skin, it would explain why these are not mentioned in the
Cretan inventories. They may have supplied them themselves.
The dimensions of the kadoi of the cista Ficoronica estimated by
Sleeswyk and Meijer at 27 litres capacity were approximately 39
centimetres in diameter and 62 centimetres in height. The dimensions
of Hellenistic Rhodian amphorae with a capacity of around 25-26
litres were approximately 35 centimetres in diameter and around 77.5
centimetres high, slightly narrower and somewhat taller.583 For what
follows we have used the shape of the kadoi of the cista Ficoronica;
however, a shape closer to that of the Rhodian amphorae would make
no difference to the argument. For a capacity of 27 litres, such kadoi
would weigh around 45 kilogrammes when full and would occupy a
cylindrical space of around 75,000 cubic centimetres, 0.075 cubic
metres, or 75 litres. Their efficiency in terms of capacity to space
occupied would have been only in the order of 0.36:1. Their dry
------------------------------
saddle bag.”). For “ta; legovmena flaskiva ” some manuscripts have “ta;" legovmena"
ajskodau'la"/ajskodavbla"”, so the word was an alternative to flavskh/flaskivo n for a
wine- or water-skin.
581
Kilby, Cooper, p. 61.
582
See Appendix Two [a], § 5.
583
See Wallace Matheson, “Rhodian amphora capacities”, esp. pp. 295-6.
Figure 44
Stowage of barrels or amphorae aboard a bireme dromon of the era of the
Macedonian emperors.
© John H. Pryor
weight to capacity ratio was around 1:1.5.

Barrels would have been much more efficient than kadoi. It is
tempting to associate an optimum size for a small portable barrel with
that of the Genoese quartarolo (39.75 litres) or the Neapolitan barile
(43.625 litres). Barrels much larger than 40 litres capacity would have
370 CHAPTER FOUR
been too heavy and large to be handled by one man in any case. Forty
litres is also about double the size of a normal bucket and well
buckets larger than that and weighing over 20 kilogrammes would
become difficult to handle.
Forty litres would have weighed 40 kilogrammes, plus around 10
kilogrammes for the weight of the barrel. The barrel would have
measured approximately 35.5 centimetres across the head, 45
centimetres in diameter at the pitch, and 53 centimetres in height.584 A
man could not get his arms around anything much bigger than that to
lift it in any case. Such a barrel would have occupied a cylindrical
space of around 84,000 cubic centimetres, 84 litres, for an efficiency
rating of capacity to space occupied of 0.48:1. Barrels would also
have been much more efficient than amphorae in terms of dry weight
to capacity, between 1:3.5-4.5, in this case 1:4.
It would have been just possible to stow two 27-litre kadoi or two
40-litre barrels alongside the thwarts of the oarsmen of the lower bank
between them and the hull, two for each oarsman. Half of the barrels
or kadoi may well have been stowed similarly above deck for the
oarsmen of the upper bank but obviously they could not have stayed
there during battle and there must have been room to stow them below
if necessary. Either that or they were jettisoned before battle.
The only logical conclusion to the problem of the water supply of
dromons, a supply which then governed their cruising range, is that
they may have been able to stow away around 100 40-litre barrels
weighing around 5 tonnes when full or around 100 27-litre amphorae
weighing around 4.5 tonnes. This would give a dromon a minimum
range under oars in summer using one tonne of water per day of 3-4
days. With an average speed in favourable conditions of around 4
knots and an average of around 14 hours of daylight during mid-
summer campaigning seasons, 3-4 days’ water supply would have
given Byzantine fleets a range of no more than 330 kilometres under
oars. All things would have been variable of course. Conditions would
have made all the difference, as also would have using the sails when
possible, cool weather, and human enduranceand skill. Fleets could
also have proceeded by night if out to sea away from coasts and
islands or if the skies were clear and the moon was full or even if the
need was great. But in normal circumstances, Byzantine fleets could
not have ranged much more than around 330 kilometres under oars
and in daylight without watering. If packed to the gunwales with
------------------------------
584
Kilby, Cooper, p. 61.
supernumeraries as for the assaults on Crete, that figure would have to

have been lowered. In fact we conclude that the 400 kilometres from
Phygela to Chandax would almost certainly have been beyond the
range of Byzantine fleets under oars and in daylight. Either they
watered at Naxos or Ios, which has a better harbour and better water
supplies for a large fleet, or they used their sails and also sailed by
night. In fact Ios has a large anchorage sheltered from the meltemi and
capable of accomodating any medieval fleet, with shallows at the head
of the inlet ideal for beaching galleys and a small stream. It would
probably have been the best final aple2kton before the dash to Crete.
The water requirements of horses would have added to the
problems of those of the men and in themselves are sufficient to
explain why the Byzantines never attempted to transport horses to
Crete from Constantinople but rather brought them overland to south-
west Asia Minor. Even carrying only the 12 horses that Theophane2s
the Confessor recorded, a horse-transporting chelandion would have
needed another 290 litres per day for the horses, adding around
another tonne of water for a voyage of around 3-4 days and requiring
25 more 40-litre barrels or 37 more 27-litre kadoi, all of which would
have had to have been stowed somewhere.
On the basis of the specifications for the amount of wheat, flour,
and barley to be supplied for the Cretan expedition of 911 from the
themata of Thrake2sio2n, Anatolikon, and Kibyrrhaio2tai, and the
numbers of troops and horses specified for the same, Haldon has
calculated that the troops might have been supported for around 18-24
days and the horses for between 18 and 28. He suggests that it is
probable that Byzantine naval expeditions operated on a standard of
carrying supplies for no more than around 24 days, which was similar
to that for expeditions by land.585 This may well have been the case, at
least for short-range expeditions, but naval expeditions would
certainly have had to have watered much more frequently than this.
Consider what the problems of water supplies must have really
involved. The size of the fleet of 949 has been the subject of great
debate.586 It does not matter a great deal here since our conclusions
stand irrespective of the exact size of the fleet, nevertheless we
present our own interpretation. According to the inventories, the
imperial fleet numbered a total of 150 ousiai; however, many of these,
------------------------------
585
Haldon, “Theory and practice”, pp. 294-300.
586
See Appendix Four [b], §§I.1-15 [= Haldon, “Theory and practice”, pp. 218-21;
Constantine VII, De cerimoniis, II.45 (vol. 1, pp. 664-5)]. See also, in particular,
Makrypoulias, “Navy”; Haldon, “Theory and practice”, pp. 334-9.
372 CHAPTER FOUR
as well as some ships, were delegated for other duties and did not sail
to Crete. Most probably 33 chelandia of one ousia, 7 larger
pamphyloi, and 20 dromons carrying two ousiai each, sailed from
Constantinople. In addition, the thema of Aigaion Pelagos supplied 6
chelandia pamphyla of 120 men and 4 chelandia ousiaka of 108 men
which may either have joined the expedition en route or have been
sent to Constantinople as a home guard, depending on how the text is
read.587 Samos supplied 6 chelandia pamphyla of 150 men and 6
chelandia ousiaka. The Kibyrrhaio2tai supplied 6 chelandia pamphyla
of 150 men and 6 chelandia ousiaka of 110 men. Nine galeai from
Antalya and an unknown number from Karpathos also joined the fleet.
Four chelandia also came from the Peloponne2sos. The total fleet
destined for Crete probably numbered 20 dromons, 49 or 53 chelandia
of one ousia, 19 or 25 larger chelandia pamphyla, and more than 9
galeai, as well as, perhaps, some sailing ships about which nothing is
known. A minimum figure of some 95 ships is used below. The fleet
was almost certainly larger than that, but the figure of 95 ships will
make the argument perfectly well.
The number of men involved is even more arguable for the reason
that the word ousia for a standard ship’s complement did not include
officers, supernumeraries, and marines or soldiers. It referred only to
the oarsmen. Nor is the size of the crews of galeai specified
anywhere; although they probably had half-complements of 54
oarsmen. We estimate the fleet to have had between 13,000 and
14,000 ordinary seamen, 13,500 for the sake of argument. For the
purposes of logistical analysis, to those we would need to add the
ships’ officers, servants, and perhaps some deck hands, adding in the
order of another 40 men per large ship and 20 per galea as suggested
above.588 Ignoring supernumeraries, troops and horses being
------------------------------
587
The text as preserved in the Leipzig manuscript reads: “... ejavqhsan eij" fuvlaxin
th'" povlew" oiJ strathgoi; tou' Aijgaivou pelavgou" meta; celandivwn pamfuvlwn ıV ajna;
ajndrw'n rkV kai; celandivwn oujsiakw'n dV ajna; ajndrw'n rhV.” (“... For the guard of the
City, the strate2goi of Aigaion Pelagos with six chelandia pamphyla, each of 120 men
and 4 ousiaka chelandia, each of 108 men.”).
There are some problems with the text as it stands, in particular why Aigaion
Pelagos should have had more than one strate2gos, assuming that the plural is not just
a scribal error for what should have been the singular. Haldon emends the text to read:
“... ejavqhsan eij" fuvlaxin th'" povlew" oiJ strathgoi; ‹tw'n ploi>moqemavtwn: oJ strathgo;"›
tou' Aijgaivo u pelavgou" ....” (“... For the guard of the City, the strate2goi ‹of the naval
themata: the strate2gos› of the Aegean Sea ...”). See “Theory and practice”, pp. 218-9.
It is then unclear whether the squadron of Aigaion Pelagos sailed to Crete or went to
Constantinople; although Haldon believes that the latter was the case. See “Theory
and practice”, p. 306.
588
See above pp. 260-64.
transported, and the crews of any sailing ships which may have been
involved, and assuming that non-oarsmen and the second ousia of
each dromon required only half of a full water ration since they would
not have been labouring like oarsmen, the fleet would have required
between 110 and 120 metric tonnes of water per day, let us say 115.
No rivers anywhere en route were large enough for fleets to sail up
them beyond the salt water zone so that they could water by lowering
buckets overboard. Fleets were dependent on ports for water but in the
tenth century most of these would have been merely sheltered
roadsteads and developed port facilities few and far between. A fleet
as large as this would have had to anchor offshore or, at best, come in
by turns in small numbers to whatever docks existed. It is true, of
course, that galleys could be beached. However, in the largely tideless
Mediterranean, beaching galleys and then loading tonnes of water
onto them would not be such a bright idea. The water would have had
to have been taken aboard while the galleys were afloat. Moreover, we
doubt whether any ports, even Constantinople, had reticulated water
supplies fed to any docks that did exist. The earliest attempt to do this
known to us was Genoa’s building of an aqueduct along her docks
during the thirteenth century. Water would have had to have been
loaded manually by bucket into portable barrels or amphorae from
wells, springs, or streams, and then transported to the ships.
Admittedly, there were many men in the crews but that would be
useful only if there were reasonably large streams. Only one barrel at
a time can be filled from a well. Even of those places mentioned in the
Stadiodromikon for the Cretan expedition of 949 which had at least
some water, He2rakleia, Proikonne2sos, Abydos, Tenedos, Mityle2ne2,
Samos, Naxos, The2ra, and Dia all had either unsuitable anchorages or
unreliable streams or were dependent on wells. Only Chios and Ios
had good anchorages and reliable streams. Phygela was also ideal,
having over two kilometres of gently shelving beach and a large
stream running into the bay.589
When we consider questions of watering and of aple2kta, we also
need to consider the requirements of a fleet. Even if they consider the
spatial requirements at all, and most do not, historians write as though
fleets could simply be parked like cars in a multi-storey carpark,
bumper to bumper with just enough room between them to open the
door. But, of course it was not like that. As reconstructed, standard
dromons and chelandia were probably around 31.25 metres long. To
------------------------------
589
See Pryor, “Stadiodromikovn”, p. 106.
374 CHAPTER FOUR
anchor such ships in even shallow water, let us say 2 fathoms (12 feet,
3.66 metres), in order for the anchor to hold the ship, the cable(s)
should make an average angle to the horizontal of no more than
around 33˚, preferably less. As shown above, the cables were
connected to the anchors by lengths of chain which lay flat on the sea
bed and enabled the anchor flukes to hold, just as on modern small
craft using rope cables.590 That would mean that a clear circle of
approximately 80 metres diameter would have to be allowed for the
ship to swing in. Taking the figure of 95 ships for the fleet would
mean that a commander would require a roadstead at least 7.6
kilometres long to anchor the fleet in a single line so that each ship
could swing safely with the wind at its anchors. If the depth of the
water was greater, then the required space would increase
proportionately. At four fathoms he would need 8.9 kilometres. Of
course, the commander could take a chance that all the ships would
swing in unison, which in actuality ships never do, and cut the
clearance to a radius of 40 metres, say 3.8 and 4.45 kilometres
respectively, but he would have been foolish to do so in the season of
the meltemi. It is not known in which month the fleet of 949 sailed;
however, in 960 Nike2phoras Pho2k as sailed from Constantinople in
July,591 right at the beginning of the meltemi season, which normally
continues until mid September.
Even if each of the ships were moored midway between two
anchors in order to cut down the distance they would swing, they
would still need around 65 metres to swing in and the whole fleet
would need around 6.2 kilometres. They could, of course, be moored
bow and stern by laying out a stern anchor as well as the bow anchor,
and that would eliminate all swing. However, mooring ships bow and
stern is a hazardous practice because if the wind springs up fresh from
abeam, the anchors will almost certainly drag. Mooring ships bow and
stern with the inefficient anchors of the Middle Ages during the
meltemi season in the Aegean would not have been recommended.
With its oars run out under way, a dromon would occupy a lateral
distance of around 11 metres. So a mathematical minimum shore line
to run the 95 ships ashore directly by the bow would be around 1.05
kilometres. However, they would almost certainly not be driven
ashore by the bow but rather beached in the time-honoured
Mediterranean manner. That is, a bow anchor would be dropped off
------------------------------
590
See above p. 212.
Theophane2s continuatus, VI.Basileiva Rwmanou' uiJo u' Kwnstantivnou tou'
591
porfurogennhvtou.10 (vol. 1, p. 475).

Figure 45
Anchoring, mooring, and beaching galleys.
© John H. Pryor
376 CHAPTER FOUR
shore and the ships then rowed over it, reversed by backing water on
one side, made ground by the stern by backing water, and made fast to
land with a stern cable. That would make disembarking and
embarking crews and loading supplies much easier, as well as being
much better defensively in case of surprise attack. However, it would
require much more length of beach, a mathematical minimum of 2.97
kilometres in order for the fleet to carry out the manœuvre
simultaneously. The ships could be packed up more closely if they
came in one at a time; however, that would stretch out inordinately the
time needed to complete the manœuvre. Even a skilled crew would
take a few minutes or so to complete it and a fleet of 95 ships would
take many hours. Obviously no commander would order beaching in
such a way except in the most confined geographical conditions which
may have necessitated it.
Such mathematical minima are, of course, unrealistic. In practical
conditions, having to contend with winds, waves, and currents and
needing to allow reasonable clearance between adjacent ships,
anchoring areas of perhaps 1-1.5 million square metres, mooring areas
of perhaps around 750,000 square metres, or a beaching shore line of
4 or more kilometres would have been much more realistic.
If the fleet anchored in four squadrons, it would still have required
1.92 kilometres of shore line in 2 fathoms and 2.26 kilometres in 4
fathoms, or 1.56 kilometres in both if moored. A mathematically
minimum space to anchor all 95 ships in a packed circle would be
around 610,000 square metres in 2 fathoms and 840,000 in 4 fathoms,
or 400,000 square metres in both if moored.
Both anchoring or mooring, and also watering, large fleets must
have been both difficult and also laborious and time-consuming,
except when a very long beach or large sheltered harbour with a
substantial stream running into it was available. Chios, Ios, and
Phygela would have been the most attractive aple2kta from a large
fleet’s logistical point of view.
However, by 949 the maritime thema of Samos with its own
strate2gos was well established with its headquarters at Kastron Samos,
modern Pythagorion, in the South-East corner of the island. Was this
why Samos was included in the Stadiodromikon. Would Samos have
been a suitable aple2kton? In fact the answer to this question is clearly
negative. The entrance to the harbour is only around 365 metres
across from Cape Foniás to the breakwater, although the natural
harbour entrance is about 715 metres across. It is roughly circular with
an inner radius of around 365 metres. The capacity is around 420,000
square metres, with an average depth of 3-4 fathoms, 5.5-7.3 metres.

There is no way in the world that the entire fleet of 949 could have
moored here and nor are the beaches around from the rocks at the
northern point of the bay to the natural breakwater point anywhere
near long enough to have beached all the ships; although, if some
were beached and some moored, the fleet might just have been able to
squeeze in. However, there was no sizable stream in the cove and in
Roman times there had been an aqueduct bringing water to the town.
It may have been still operative in the tenth century but that is not
known for sure. If not, watering would have had to have been done
from wells. Kastron Samos may have been a well-sheltered anchorage
for a small squadron of a thema, but it was hopelessly inadequate as
an aple2kton for a fleet such as that of 949. The problem of rotating the
ships to get them close to the shore for watering would have been
almost insurmountable.
Figure 46
The harbour of Kastron Samos, adapted from Great Britain, Admiralty,
Hydrographic Office, Chart No 1568 of 1967.
© John H. Pryor
With portable kadoi of 27 litres or barrels of 40 litres, watering the

crews of the fleet of 949 with 115 tonnes of water from wells would
have needed around 4,250 kados-manlifts or 2,875 barrel-manlifts per
378 CHAPTER FOUR
day. A supply of 345 tonnes for three days would have needed 12,750
kados-manlifts or 8,625 barrel-manlifts. Yes, there were large
numbers of men in the crews, but they would all have to have taken
their turn at limited access to springs, small streams, or wells.
How big were well buckets? Again there would have been an
ergonomic optimum. The bigger and heavier the full bucket, the more
men, or the longer the time, needed, to raise and empty it. An
optimum size for a bucket may have been around 20 litres. A bucket
with an internal diameter of 28 centimetres and a height of 32.5 would
have a capacity of 19.7 litres and its size seems to be about right. If so,
the number of well-lifts required would be double that of barrel-lifts.
How long would 17,250 well-lifts in Samos harbour have taken?
These figures could be varied considerably without affecting the
obvious conclusion. Watering large fleets must have been extremely
laborious and time consuming and this helps to explain why extended
expeditions were regarded as such monumental undertakings. In fact
we believe that we barely begin to comprehend the enormousness of
the logistics of galley warfare in the Middle Ages.
(m) Armaments
Dromons had an array of weapons, of which the most important

were the flame-throwers (sipho2nes or sipho2nia), tubes from which
Greek Fire could be projected. [See Appendix Six]
As well as the siphone2s, both Leo VI and Nike2phoros Ouranos
referred to “cranes”, geravnia (gerania), shaped like a capital letter
gamma, that is like a “G”, which could turn on their upright post until
the arm was over an enemy ship and then pour combustibles, either
flaming pitch or the same “processed” fire, pu'r ejskeuasmevnon (pyr
eskevasmenon), used in the siphone2s, onto them.592 Then they said that
the combustibles were poured on the enemy ship when the mavgganon
(manganon) was turned or tipped. Presumably this was at the end of
the arms of the cranes. Manganon was a word with many connotations
------------------------------
592
Appendix Two [a], §67; Appendix Five, §61.
These gerania were quite probably the same things as the khlwvneia (ke2lo 2neia)
mentioned by the author of the Vita Basilii in the Theophane2s continuatus as having
been used by the droungarios tou ploimou Nike2tas O›oryphas to suspend alive
Christian apostates captured in the fleet of the Muslims of Crete at the battle of the
Gulf of Corinth in 879 before dropping them into cauldrons of boiling pitch. See
Theophane2s continuatus, V.61 (p. 301). In classical Greek a khvlwn (ke2lo2n) or
ke2lo2n eion was a pivoted swing beam.
in Byzantine military practice. Its original meaning was a block of a

block-and-tackle. However, by extrapolation it appears to have
become applied to many types of engines of war which used blocks-
and-tackles. Perhaps it should be understood here as something like a
hinged cauldron which could be tipped over by means of some trip
mechanism operated from the post of the geranion. Whatever the
case, geranion or gevrano" (geranos) had clearly become associated
with pivoting cranes. In the Peri me2chane2mato2n attributed to
Athe2naios Me2chanikos a machine referred to as a karche2sion was
described. This had a pivoting arm named as a geranos, at the end of
which was a hooked ladder for grappling onto walls.593
In addition to the sipho2nes and gerania, normal weapons included
“bow-ballistae” at the prow, stern, and along the sides called
toxobalivstrai (toxobalistrai), which could fire quarrels known as
“mice”, muve" (mues), or “flies”, mui'ai (muiai).594 Before the invention
of Greek Fire, the Strate2gikon attributed to Maurice had recommended
the mounting of small ballistae protected by mats at the prow.595 And,
according to John Kaminiate2s, the Muslim ships which assaulted
Thessalonike2 also had catapults or ballistae which they used to fire
rocks against the defenders.596 In specifying siege machinery required
for attacks on fortresses the inventory for the Cretan expedition of 949
referred to the “blocks”, trociliva (trochilia), of large bow-ballistae,
suggesting that they were so powerful that the strings had to be drawn
by block and tackle systems.597 Other projectiles familiar in Latin and
------------------------------
593
See Wescher, Poliorcétique, pp. 35-7.
594
Appendix Two [a], §60; Appendix Five, §57. See also Theophane2s continuatus,
V.59 (p. 298): “..., kai; toi'" petrobovloi" ojrgavnoi" kai; toxoballivstrai" kai; tai'" ejk
ceirw'n tw'n livqwn ajf evsesi ...”.
We use the literal, if clumsy, “bow-ballistae” here because the Byzantines, like
their Greek and Roman forebears, used the words ballista and ballivstra (ballistra)
indiscriminately for two distinct types of engines: bow-ballistae, large crossbows
firing bolts or arrows on the one hand, and catapults hurling rocks on the other. The
bow-ballistae here were engines mounted on ships.
595
Maurice, Ek tou Maurikiou, §5 (p. 41): “Eij" e{kaston de; drovmwna, eij ejndevcetai,
poih'sai toxoballivstra" ajpo; kilikivwn skepomevna" ejn tai'" prwv/rai" i{na tou;"
ejpercomevnou" ejcqrou;" makrovqen ajpodiwvkwsin: ...”. Cf. Maurice, Strate2gikon,
XIIB.21.12-15 (p. 468).
596
John Kaminiate2s, De expugnatione Thessalonicae, 29.3 (p. 27): “a[lloi toi'"
petrobovloi" ejgkaqhvmenoi ta;" uJpermegevqei" ejkeivna" tw'n petrw'n calavza"
metewrivzonte" e[pempon.”. Such engines for firing rocks may have been either bow-
ballistae which used a bow mechanism but fired rocks instead of bolts or they may
have been catapults.
597
See Haldon, “Theory and practice”, p. 225: “..., toxobolivstrai megavlai meta;
trocilivwn kai; kovrdwn metaxotw'n, ...”; Constantine VII, De cerimoniis, II.45 (vol. 1, p.
670).
380 CHAPTER FOUR
Arabic sources also appear in the sources: pots containing quicklime,

Greek Fire, venomous reptiles which were hurled by catapults and
which smashed on impact, small iron caltrops to impede enemies’
movement on their decks, large caltrops wrapped in combustibles and
hurled aflame onto enemy ships.
The large catapults or bow-ballistae were presumably fixed on
swivel mounts of some sort so that they could be aimed. In addition to
them, the inventories for the Cretan expedition of 949 also mentioned
“hand-bow-ballistae”, ceirotoxobolivstrai (cheirotoxobolistrai),
which were said to have had silk bow strings, covrdai; metaxotaiv
(chordai metaxotai), and some things known as nauvklai (navklai).598
These were apparently crossbows. Navkla is an unknown word. It
clearly meant some part of a crossbow and had nothing to do with the
Latin navicula and small boats as such. It may well have referred to
whatever mechanism was used to draw the strings on such ballistae or,
by analogy to the hull of a ship, to a cup or socket in which the bolts
rested before release.599
Metaxotaiv appears to have referred to silk, mevtaxa (metaxa) thence
to silk bowstrings. Even though silk would have been a suitable fibre
for bow strings because of its strength and hard-wearing qualities, we
have reservations about its widespread use for such purposes because
of its cost and because no sources known to us from the Latin West in
the High Middle Ages, when the crossbow became widely
disseminated, referred to silk bowstrings, even though silk was freely
available in the West by that time. However, the Parangelmata
poliorke2tika attributed to He2ro2n of Byzantium also referred to ropes
under tension which might be made from silk as well as from other
materials. Elsewhere in the inventories for the Cretan expedition of
949 both silk, and also silk and spartum, bowstrings were
mentioned.600 So, the evidence that the Byzantines really did use silk
for bowstrings appears to be undeniable. Reiske suggested the
possibility of hemp strings reinforced with silk for additional strength,
which is also a possibility.601
------------------------------
598
See Appendix Four [b], §II.15: “nauvkla" meta; ceirotoxobolivstrwn kai; covrdwn
metaxotw'n kV ” [= Haldon, “Theory and practice”, p. 225; Constantine VII, De
cerimoniis, II.45 (vol. 1, p. 670)].
599
Haldon suggests a boat-shaped platform or frame upon which the bow-ballistae
were mounted. See “Theory and practice”, pp. 271-2.
600
See Appendix Four [b], §§III.17, VI.27 [= Haldon, “Theory and practice”, pp.
227, 233; Constantine VII, De cerimoniis, II.45 (vol. 1, pp. 671, 676)]. See also
He2ro2n, Parangelmata poliorke2tika, §44 (p. 90).
601
See Constantine VII, De cerimoniis, vol. 2, p. 791.
There is debate about what cheirotoxobolistrai were, about what

the prefix cheiro- (hand-) may have referred to. It may be thought to
have referred to a “hand-held” bow-ballista; that is, a “crossbow”.
However, Anna Komne2n e2 later wrote that the crossbow of the Latins,
which she called a tzavggra (tzangra), was unknown to the
Byzantines.602 Kolias concluded that cheirotoxobolistrai were not
crossbows but rather larger ballistae shooting the mues or muiai,
pointing to the first use of tzangra in Byzantine sources in the
eleventh-century Parekbolai and the Strate2gikon of Kekaumenos as
tzarcw'n (tzarcho2n) and tzavgrai (tzagrai) respectively. Haldon has
suggested that the cheiro- prefix may have referred to the bow-
ballistae being hand-drawn as opposed to those which had to be
drawn by some mechanism.603 However, Dennis believes that the
cheirotoxobolistrai were in fact hand-held crossbows.604 If the
cheirotoxobolistrai were actually crossbows, then perhaps their use
had been discontinued by the age of Anna Komne2ne2, as Dennis
suggests, or else the tzangra was a more effective and powerful
version of the crossbow, it may even have been Persian in origin.605
We ourselves are inclined to the latter alternative; namely, that
Byzantines did have hand-held and hand-drawn crossbows but that the
Latin weapon which appeared in the later eleventh century was more
powerful and different in some way.
According to Leo VI, §73, and Nike2phoros Ouranos, §67, a fully
armed, ejxwplismevno" (exo2plismenos), dromon should have soldiers,
stratiw'tai (stratio2tai), fully armed as katavf raktoi (kataphraktoi).
The oarsmen of both upper and lower oar banks were all soldiers
according to Leo, §8, and Nike2phoros, §7; however, only those above
deck were armed as kataphraktoi. According to Leo, §14, and
Nike2phoros, §12, as well as the oarsmen all other men stationed above
deck, from the kentarchos down to the last man, were also armed as
kataphraktoi. Those not fully armoured, including the archers, wore
padded felt jackets known as neurikav (neurika) rather than the mail
corselets, lwrivkia (lo2rikia), or lamellar cuirasses (coats of armour of
overlaid plates), klibavnia (klibania), of the kataphraktoi.606
------------------------------
602
See Anna Komne2ne2, Alexiade, X.viii.6 (vol. 2, pp. 217-218).
603
604
Kolias, Byzantinische Waffen, pp. 239-53, esp. pp. 245-6. Dennis also discusses
the alternative use of muve" and mui'ai for the quarrels. See Dennis, “Flies, mice, and
the Byzantine crossbow”.
605
There is enormous disagreement about the meaning of all of these terms. See
Chevedden, “Artillery”, esp. pp.146-52.
606
382 CHAPTER FOUR
As well as body armour, marines and oarsmen were variously

armed with shields, skoutavria (skoutaria), helmets, kassivde"
(kassides), vambraces, ceirovyella (cheiropsella), swords, spaqiva
(spathia), bows, tovxa (toxa) or toxavria (toxaria), arrows, sagivtai
(sagitai), pikes, mevnaula (menaula), and throwing javelins, rjiptavria
(rhiptaria). The Anonymous merely repeated most of this from Leo’s
Naumachika Leontos Basileo2s, §14, changing contemporary terms to
pretentious classical equivalents, but he did add leggings or greaves,
knhmi'de" (kne2mides), and rigging cutters, drevpana (drepana).607
(n) Tactics, strategy, and techniques
In Chapters Seven and Eight of his treatise, the Anonymous turned to

tactics and strategies. Chapter Eight breaks off in the manuscript at the
point where he was about to discuss battle formations but, even so, he
had already suggested meanings for sailing around, perivplou"
(periplous), sailing past, paravplou" (paraplous), sailing through,
dievkplou" (diekplous), to outflank, uJperkera'sai (hyperkerasai), and
was in the midst of discussing encircling, kuklikovn (kyklikon).608 Of
these diekplous was used for the battle manœuvre of breaking the line
by Herodotos, Xenopho2n, and Thucydides, was scholiated in many
manuscripts of Thucydides, and the scholion appears to have been
known to the Anonymous.609 Periplous was used for a battle
manœuvre of encircling by Xenopho2n,610 but there is no evidence that
the Anonymous knew Xenopho2n. Kyklikon, an adjective, was most
probably derived by the Anonymous from kuvklo", kyklos, “the
circle”, used by Thucydides for the defensive ring adopted by the
Peloponne2sian fleet at the first battle of Naupaktos in 430 B.C.E. to
prevent the Athenians under Phormio2n breaking their line by a
diekplous. Thucydides described the Peloponne2sians forming their
fleet into a defensive circle with the prows facing outwards while the
Athenians sailed around them in a circle. The Anonymous appears to
have changed Thucydides’ noun for the formation, kyklos, to the
adjective kuklikov", kyklikos, which should have meant “encircling”.
------------------------------
607
608
Appendix Three, §§7.4, 8.1.
609
Herodotos, Histories, VI.12, VIII.9 (vol. 3, p. 158; vol. 4, p. 10); Xenopho2n,
Hellenika, I.vi.31 (vol. 1, p. 62); Thucydides, Peloponnesian war, I.49.3 (vol. 1, p.
82); Hude, Scholia, I.49.3 (p. 44). See Appendix Three, n. 83.
610
Xenopho2n, Hellenika, I.vi.31 (vol. 1, p. 62).
No doubt the rest of the Anonymous’s sentence would have made

clear how he interpreted Thucydides if the subsequent folios of the
manuscript had not been lost; however, it is most probable that in fact
he was merely paraphrasing Thucydides and meant nothing at all by
the change.611 Neither paraplous, nor its verbal equivalent, nor
hyperkerasai, were used with the sense of the battle manœuvres
assigned to them by the Anonymous in any classical source known to
us. Nor were his supposed meanings of any of these words derived by
him from Pollux, Hesychios, or Pho2tios. He appears to have been
indulging in elementary etymological speculation, extrapolating from
the known meaning of diekplous. In any case, any tactics to which
diekplous and these other words might possibly have referred had
passed away long ago with the waterline ram. The Anonymous is of no
value for naval tactics and strategies in the tenth century.
For these, since the chronicles and other sources are devoid of all
but the most spare and sketchy descriptions of battles, we are
primarily dependent on the Naumachika of Leo VI and its paraphrase
by Nike2phoros Ouranos, even though there must be serious doubt
about the practicality of many of the emperor’s recommendations.612
It is clear that by the heyday of the dromon naval tactics and
strategies were very different to those of the age of the trie2re2s.
Reduced to a fundamental, these differences can be attributed to the
disappearance of the only “ship-killing” weapon ever known before
the invention of explosive projectiles: the ram. For all its potency in
some circumstances, Greek Fire was never the ship-killer that the ram
had been and no system of battle tactics was ever built around it.613
The weapon was obviously effective in certain circumstances but
there are many uncertainties about what those were. Byzantine
chroniclers frequently attributed fleet victories to the use of Greek
Fire; however, they rarely mentioned its use when fleets were
defeated, even though it must obviously have been used on many such
occasions. Only some twenty years after Kallinikos’s Greek Fire
scattered the Umayyad armada at Constantinople in 672-8, the fleet
------------------------------
611
See Appendix Three, §8.1 and nn. 85-6; Thucydides, Peloponnesian war, II.84.3
(vol. 1, p. 416).
612
We are well aware of the fact that in some cases the strategies and tactics
recommended in the tenth century had a long history in the tactical literature
stretching back to antiquity. However, to have traced the textual traditions would have
been beyond the scope of this work. We have limited the analysis of that tradition to
those works which we can demonstrate that Leo VI actually knew directly.
613
To characterise Greek Fire as comparable to “the atomic bomb in our own day”,
as did Ellis Davidson, is grossly misleading. See Ellis Davidson, “Secret weapon”, p.
61.
384 CHAPTER FOUR
sent to Africa under the patrikios John was unable to resist naval
forces sent from Egypt and he had to abandon Carthage. One
occasion on which it was used by a defeated force was during the
revolt of Thomas the Slav, when the fleet of the Kibyrrhaio2tai, which
had joined the revolt, had it.614 The flame-throwing sipho2n form of
Greek Fire had a limited range and required both calm conditions and
a following wind. Ships had to be closely engaged before the weapon
could be used. Conventional missiles and projectiles would have had a
much longer range and Greek Fire probably never displaced them.
Moreover, enemy ships would almost have had to have been willing
to allow the Byzantines to close up to use the weapon. If they chose to
stay out of range, it would have been ineffective.
As demonstrated above, the spur was not designed to puncture a
hull and sink a ship but rather to destroy its motive power by
smashing its oars. No other weapons, neither projectiles nor any other,
now had ship-killing capabilities. Battle tactics therefore changed.
Objectives changed from attempts to deliver a knock-out blow to
degrading attrition. Rather than manœuvring to obtain a position to
ram and sink, tactics became to degrade an enemy ship’s ability to
resist so that it could be boarded and captured. These objectives
remained unchanged until the days of galley warfare in the
Mediterranean were over. The preliminary phases of battle therefore
became extensive exchanges of missiles of various types.
The Byzantine Empire, especially in its heyday from the seventh to
tenth centuries, has had something of a reputation as a power with a
major maritime focus but a close scrutiny of the record does not really
support this.
In spite of the fact that some crews in Byzantine fleets at various
times were well regarded, for example the Mardaites of the thema of
the Kibyrrhaio2tai, there is little evidence to suggest that in general
Byzantine seamen were so skilled that this gave Byzantine fleets any
edge over their oponents. It is true that Byzantine squadrons managed
to defeat the Rho2s on all occasions when they attacked
Constantinople: in 860, probably in 907 under Oleg of Kiev, in 941
under Igor, and in 1043 under Jaroslav I. A fleet also defeated the
Rho2s on the Danube in 972. However, rather than being attributable to
any qualities of Byzantine seamen, these victories were due to the
triple advantages of Greek Fire, dromons and chelandia being much
------------------------------
Genesios, Basileiai, B.5-6 (pp. 28-9); Theophane2s continuatus, II.14 (p. 60);
614
John Skylitze2s, Synopsis historio2n, Micah;l oJ Traulov".8 (p. 35).

larger than the Norse river boats of the Rho2s, and (except in 972)
being able to fight in home waters against an enemy far from home.
The last is true also of the defeat of the Muslim asssaults on
Constantinople in 672-8, and 717-18. In both cases it was the
advantage of campaigning in home waters rather than hundreds of
miles from sources of supplies, the problems faced by the Muslims of
surviving through the winters, and Greek Fire, that proved decisive.
With the exception of the factor of Greek Fire, the same is probably
true of the victories over the fleets of Thomas the Slav in 822-3.
In general the record of Byzantine fleets from the seventh to the
tenth centuries was hardly impressive.615 To be sure, they did achieve
some notable victories: the Veneto-Byzantine victory at Syracuse in
827-8, the defeat by storm of the Muslim fleet off Cape Chelidonia in
842, the victory of Nike2tas Ooryphas over the Cretans in the Gulf of
Corinth in 879, Nasar’s victories off western Greece and off Punta di
Stilo in 880, the victory of Himerios on the “Day of the Apostle
Thomas”, probably in 905, the defeat of Leo of Tripoli off Lemnos in
923, the victory of Basil Hexamilite2s over the fleet of Tarsos off Lycia
in 956, and the defeat of an Egyptian squadron off Cyprus in 965.
Against that record, however, have to be balanced many disastrous
defeats: of Constans II at the battle of the masts off Phoinikous in 655,
the defeat of Theophilos, the strate2gos of the Kibyrrhaio2tai, off
Antalya in 790, a defeat off Thasos in 839, the defeat of Constantine
Kontomyte2s off Syracuse in 859, the annihilation of a fleet off
Milazzo in 888, a defeat off Messina in 901, the disastrous defeat of
Himerios north of Chios in 912, the defeat of an expedition in the
Straits of Messina in 965, and defeats off Tripoli in 975 and 998.
Although the tide of Byzantine naval success ebbed and flowed
over the centuries as other circumstances dictated, nothing suggests
that the quality of the empire’s seamen was in any way decisive.
Indeed, there are occasional pieces of evidence which suggest that all
was not always happy in the fleets. Sometime between 823 and 825
John Echimos confiscated the properties of seamen of the
Kibyrrhaio2tai. After he had become a monk and taken the name
Antony, he was interrogated as to his reasons for doing so on the
orders of the emperor Theophilos. According to the author of his Life,
his explanation was that they had been supporters of Thomas the Slav
and were “hostile to Christians”, thus implying that they were
------------------------------
615
It is not possible to be exhaustive here. What follows is a limited, but balanced,
summary of the more important Byzantine victories and defeats in fleet engagements
as canvassed in Chapter One.
386 CHAPTER FOUR
iconoclasts, and that he had confiscated their property and given it to

supporters of Theophilos’s father Michael II. In spite of this
explanation, the emperor initially imprisoned him and had him
interrogated, suggesting that there was more to the story and that he
rejected the explanation.616 The fleet of the Kibyrrhaio2tai had joined
the rebellion of Thomas the Slav, as it did later those of Bardas
Skle2ros in 976-9 and Bardas Pho2kas in 987-9, and it is clear that at
times there must have been serious disaffection in what was the front-
line fleet of the empire in the ninth and tenth centuries.
In 880 the expedition sent under the command of Nasar, the
droungarios tou ploi mou, to counter the attack in the Ionian sea by
the Aghlabid fleet was forced to a temporary halt at Metho2ne2 by the
desertion of a large part of the crews.617 Why they deserted is
unknown, but we can be fairly sure that it was not a simple question of
their “being terrified in the face of danger” as the Vita Basilii of the
Theophane2s continuatus suggested. In the tenth century, according to
the Life of St Neilos of Rossano, the populace of Rossano in Calabria
destroyed the fleet of their thema to avoid having to serve in it.618 How
many other instances of dissent do we know nothing about?
In Constitution XVIII of Leo VI’s Taktika, the emperor advised his
strate2goi to attack the Muslims at sea if they were invading by land
and to assault their territory left undefended by any of their naval
expeditions. Byzantine squadrons should assault Tarsos and Adana
while the army advanced through the Tauros passes.619 This seems to
have referred to the ages-old, incessant campaigning across the
frontiers; however, inserted in Constitution XX was a paragraph
recommending that the naval strate2goi, plwivmoi strathgoiv (plo2i moi
strate2goi), should strike pre-emptively from Cyprus against the fleets
of Egypt, Syria and Cilicia before they could unite. Since Cyprus was
not recovered for the empire definitively until 965, the paragraph was
probably a result of the temporarily-successful expedition of Himerios
to Cyprus and the Levant in 910-11, before his catastrophic defeat off
Chios in April 912.620 Since Leo’s Taktika is generally thought to have
------------------------------
616
Vita Antonii junioris, p. 209.
617
618
Vita S. Nili, IX.60 (col. 105).
619
Leo VI, Taktika (PG), XVIII.138-40 (coll. 979-80).
620
Leo VI, Taktika (PG), XX.212 (col. 1071). However, we have edited it from the
better text in the manuscript, Milan, Biblioteca Ambrosiana MS. B 119-sup. [Gr. 139]
at folio 314v, where it is numbered as Constitution XIX, §210: “siV ejavn polemh'/" pro;"
ajnqrwvpou" ejk pollw'n tovp wn sullegomevnou", w| strathgev, devo n se mh; perimevnein e{w"
ou| eij" e}n sunacqw'sin, ajlla; e[ti ejsparmevnoi" aujtoi'" kai; disparmevnoi", h] kata; th'"
ijdiva" cwvra" e{kaston h] eij" eJt evrou" tovpou" pro; tou' sunelqei'n aujt ou;", ejpitivqou: kai;
been compiled ca 905-6, this paragraph must have been added to it

later, after 910 but before the defeat of Himerios and the death of the
emperor on 11 May 912.
Naval warfare became more unpredictable. No longer could any
power hope to have such an advantage in weaponry or the skill of
crews that success could be expected. Weaponry and skill could still
make a difference of course, but rarely a decisive one.621 More often
than not victory or defeat became a matter of circumstances,
admiralship, and numbers. To commit to battle was to risk the
unpredictable fortunes of war and really decisive victories were hard
to achieve in any case. A victor could rarely prevent large sections of
defeated fleets escaping the field and the Mediterranean powers could
replace ships, if not skilled crews, remarkably quickly in any case.
Caution became the first priority. Syrianos Magistros advised that a
fleet should always proceed with four light and fast scout ships out
ahead, with two up to six miles or so out and the other two between
them and the fleet. A strate2gos should always have good intelligence
of the enemy and should engage the enemy only if he had superior
numbers, not even if the forces were equal unless the enemy forced
the engagement. He should not engage at all unless the enemy posed a
danger. Some of these recommendations were later repeated by
Nike2phoros Ouranos.622 Leo VI advised planning attacks with
forethought, being wary of committing to general engagements, only
doing so when confident of superiority over the enemy, and against
becoming over confident.623 The first priority of a strate2gos was to
preserve his own forces intact and then to search for any opportunity
------------------------------
nu'n de; toi'" ejx Aijguvptou kai; Suriva" kai; Kilikiva" sunagomevnoi" barbavroi" pro;" th;n
kata; Rwmaivwn ejkstrateiva n dei' tou;" plwivmou" strathgou;" su;n tw'/ nautikw'/ stolw'/ th;n
Kuvpron katalabovnta", pro; tou' eJnwqh'nai ta;" barbarika;" nau'", ajp oseivlai kat aujtw'n
plwvimon duvnamin iJkanh;n katagwnivsasqai th;n barbarikh;n naumacivan e[ti dihrhmevnhn,
h] ta;" nau'" ejkeivnwn ejmprh'sai pro; tou' ajpopleu'sai th'" ijdiva".”.
The attribution of the paragraph to being a product of Himerios’s expedition was
that of Vasiliev. See Vasiliev/Canard, Byzance et les Arabes. Tome II, part 1, p. 211.
621
For a later exception see Pryor, “Roger of Lauria”.
622
See Appendix One, §§6.1-3, 9.8, 9.10-11, 9.14; Nike2phoros Ouranos, Ek to 2n
taktiko2n, §§119.4-6.
Such recommendations were not limited to naval warfare. Nike2p horos Pho2kas
recommended that general engagements should be avoided except when the enemy
had fled or was crippled. Even enemy forces of strength equal to one’s own were to
be avoided. See Nike2phoros Pho2kas, Praecepta militaria, IV.19 (p. 50). Cf.
Nike2phoros Ouranos, Praecepta militaria, ch. 61.19, in McGeer, Dragon’s teeth, p.
132.
623
See Appendix Two [a], §§17, 36-7, 40, 74-5. Cf. Appendix One, §§15, 34-6, 67-
8; Appendix Three, §3.1. On the recommendation to ensure superiority in numbers cf.
above p. 181.
388 CHAPTER FOUR
or strategem that would enable him to attack the enemy with the least
risk to his own forces. Thus Leo VI recommended giving battle in
waters of one’s own choice off enemy coasts and laying ambushes.624
The recommendation to engage off enemy coasts so that the crews of
the enemy would not fight to the death but would seek safety in flight
confirms the fact that almost all medieval galley warfare was coastal.
Both Syrianos Magistros and Nike2phoros Ouranos also made that
perfectly clear.625
Expeditionary objectives could frequently be achieved best by
preserving one’s forces intact and actually avoiding battle since naval
warfare was essentially amphibious warfare whose purpose was to
secure control of terrestrial objectives rather than to attempt to control
maritime space. The latter was an unrealistic and vain hope given the
limitations of medieval naval technology with respect to the vast
expanses of the sea. Limited water supplies and cruising ranges, lack
of any weapon capable of quickly destroying enemy ships,
performance capabilities inadequate to force an enemy to engage if he
did not wish to do so, and limitations of visibility, meant that control
of maritime space was never achievable.
The masthead height of the foremast of a standard dromon as we
have reconstructed it was only around 10.65 metres above sea level.
[See Figure 20] There were, admittedly, larger dromons; however, for
what follows a couple of metres more of masthead height would make
no difference to the conclusions reached. With a foremast height of
10.65 metres above sea level, the theoretical horizon of a lookout at
the masthead would have been only around 11.8 kilometres.
Theoretically, the peak of a lateen sail 21 metres above sea level could
be seen a further 51.7 kilometres away but, of course, no man could
see 63.5 kilometres with unaided sight. In all probability, around 15-
20 kilometres would have been the limit of visibility from the
------------------------------
624
See Appendix Two [a], §§40, 53; Appendix Five, §§38, 51; Appendix Eight [a],
p. 246. Note that in §40, even though following Syrianos Magistros quite closely at
this point, Leo VI actually reversed Syrianos’s advice, who had advised setting up
battle close to shore if off one’s own territory so that there would be a refuge if
defeated but out to sea if off enemy territory. See Appendix One, §9.42-4. Leo seems
to have been influenced by another sentence of Syrianos which said that off foreign
territory ships positioned at the seaward end of a line would be most likely to desert
while off one’s own territory it would be those at the landward end. Ibid., §9.23.
Polybios commented on the Roman victory over Hasdrubal’s forces at the
mouth of the Ebro in 217 B.C.E. that having Hasdrubal’s land forces occupying the
shore line was in fact a disadvantage to the Carthaginian fleet because it ensured a
safe and easy retreat for the ships’ crews, who abandonned the fight with little
resistance. See Polybios, Histories, III.96.2-5 (vol. 2, pp. 236-8).
625
Cf. above pp. 369-61 and nn. 559.
masthead of a dromon.626 Scout ships could not, therefore, patrol a

space more than 30-40 kilometres in advance of a fleet and probably
no more than 30, since they were always said to have been smaller
than standard dromons and would have had lower mastheads. In fact,
in order to be able to actually read signals with unaided eyesight and
communicate them back to the fleet, distances must have been even
less than this. Syrianos Magistros advised that a fleet should always
proceed with scout ships out ahead, up to six milia or so. Two scout
ships should be 6 milia ahead and another two should be between
them and the fleet to relay any messages.627 Six milia was only around
8 kilometres. If the forward scout ships then had a range of visibility
of another 8-16 kilometres, then the real maritime space that could be
observed was only around 25 kilometres at best.
Moreover, even if scouts descried an enemy fleet 25 kilometres
away, it would take hours for the fleets to come to engagement, even
if they both cooperated and sailed at full speed to engage. If the
weather conditions were favourable for one fleet, they would
invariably be unfavourable for the other. If either fleet sought to avoid
engagement, it could never be forced to do so unless trapped
somehow.
Even narrows such as the Straits of Otranto are approximately 110
kilometres wide and the entrance to the Aegean between Crete and
Rhodes is approximately 180 kilometres wide; although, Karpathos
does straddle the gap. No medieval power could ever hope to control
ingress and exit through such maritime spaces.
In such circumstances naval forces could rarely be more than an
adjunct to land forces, sea power to land power.628 We are accustomed
to think of the Byzantine Empire as a great naval power, at least in the
various periods of its prosperity. And in the sense that it was an
empire for which the sea lanes in the Aegean, southern Adriatic, Black
Sea, and along the south coast of Asia Minor were vitally important
that is true. However, in reality Byzantine naval forces were always
very secondary to the land armies and the use of sea power was
------------------------------
626
From the masthead of Olympias, which was approximately 11.5 metres above
sea level, the horizon was 11.25 kilometres distant. A lookout could just see the deck
of a similar low-hulled ship at a range of 16.1 kilometres. See Coates in Morrison,
Greek and Roman oared warships, p. 258. No data are given for the visibility of sails
over the horizon.
627
Appendix One, §§6.1-3.
628
This has been appreciated by Treadgold and Dvornik also, but not we believe for
the correct reasons. See Treadgold, Byzantium and its army, p. 91; Dvornik,
Intelligence services, pp. 153-4.
390 CHAPTER FOUR
merely an adjunct to that of the land.629 By ithemselves they rarely

achieved very much. The reconquest of Crete in 960-61 was an
exception to the rule, but that followed at least four previous failed
attempts. The history of the Empire reveals that most naval
expeditions accompanied terrestrial expeditions. Naval forces ferried
land forces and protected their maritime flanks and supply lines. Most
naval engagements occurred in these circumstances rather than in
those of opposing fleets seeking each other out. Control of the land
meant control of the sea because control of the land carried with it
both control of the refuges to which all galley fleets had to have
recourse in inclement weather and also control of the water supplies
without which they could not operate for more than a few days. It also
helps to explain the extensive record of fleets lost when caught at sea
in inclement weather off coasts that were either geographically or
humanly hostile. Reading the record of Byzantine and also Muslim
fleets destroyed at sea by storms, and contemplating the horrific loss
of human life involved, gives a sobering perspective on the essential
futility of naval warfare in the Byzantine-Muslim period. Rarely did
naval victories lead to long-term or extensive political gains.
It is no accident that only two Byzantine “admirals” ever became
emperor: Apsimaros (Tiberios III), who had been droungarios of the
Kibyrrhaio2tai, and Ro2manos I Lekape2nos, who had been droungarios
tou ploimou. Nor is it an accident that only once did an emperor take
command of the navy as a whole and attempt to seek out and destroy
an entire enemy fleet. Constans II did that at the Battle of the Masts
off Phoinikous in 655, with disastrous consequences.630 As opposed to
this, emperors took command of land armies on many occasions. The
secondary character of the fleets in the Byzantine polity is reflected in
the rank accorded to their commanders in the various lists of
precedence compiled in the ninth and tenth centuries. In the Taktikon
Uspenskij, composed around 842-3, the droungarios tou ploimou of
the imperial fleet in Constantinople ranked only in eighty fourth
position among the officers of state and the strate2gos of the
Kibyrrhaio2tai, although ranked twenty fifth overall, ranked only
------------------------------
629
This has also been appreciated by Treadgold. As he says: “It should also be
recognized that behind the Byzantines’ excessive indifference to maritime defence lay
the correct belief that the empire was essentially a land power. Arabs on Sicily and
Crete were a major nuisance, but not a major threat to the empire’s vital interests.”.
See Treadgold, Byzantine revival, p. 260. However, Treadgold does not perhaps
appreciate the technical reasons why sea power always remained an adjunct to land
power in the empire; namely, the limitations of medieval naval technology.
630
See above p. 25.
eleventh among the 18 strate2goi of the themata. In the Kle2torologion

of Philotheos of 899 he ranked thirty eighth out of 60 and the
strate2gos of the Kibyrrhaio2tai, although ranked twenty first overall,
was only the fifteenth of 25 strate2goi of the themata. The other two
strate2goi of naval themata, those of Samos and of Aigaion Pelagos,
were third- and fourth-last respectively among the strate2goi, ranking
only above those of far-off Dalmatia and Cherso2n. According to the
De cerimoniis, during the reign of Leo VI the annual cash salaries,
rJovgai (rhogai), paid to the strate2goi of the Kibyrrhaio2tai, Samos and
Aigaion Pelagos were only 10 pounds of gold and they came last in a
list headed by the strate2goi of Anatolikon, Armeniakon, and
Thrake2sio2n, who received 40 pounds. Others received 30 or 20
pounds. Only kleisourarchai in command of territories smaller than
themata received less, as well as the strate2goi of the West, who were
self-supporting. In the Taktikon Benes°evic° of 934-44, the strate2gos of
the Kibyrrhaio2tai was only the twenty first of 32 strate2goi and those
of Aigaion Pelagos and Samos were fourth- and fifth-last respectively.
The droungarios tou ploimou ranked a further 17 places below the
strate2gos of Aigaion Pelagos. By the time of the Escorial Taktikon of
ca 971-75, the strate2goi of the Kibyrrhaio2tai, Samos, and Aigaion
Pelagos ranked fifty fifth, sixty seventh, and sixty eighth respectively,
with the droungarios tou ploimou in a miserable hundred and thirtieth
position.631
Appreciation of the fact that all medieval naval warfare was
essentially coastal and amphibious warfare is important since many of
the recommended strategies and tactics were devised in that context.
Ambushes, for example, are easily comprehensible in coastal warfare.
Reserve squadrons might be hidden behind islands or promontories.
They are more difficult to envisage on the high seas. How could one
hide reserve squadrons on the high seas except behind fog banks,
which are unusual in the Mediterranean, or if one came out of the sun
with it at one’s back and catching the enemy with it in his eyes? It
should be added, however, that to do so was indeed a favourite
tactic.632
Leo VI also appears to have recommended attacking when enemy
fleets had been “shipwrecked” or scattered by squalls or when they
were caught ashore making repairs, attacking even during a storm or
------------------------------
631
See Constantine VII, De cerimoniis, I.50 (vol. 1, pp. 696-7); Oikonomides,
Listes de préséance, pp. 57, 102-4, 246, 264-8.
632
Roger of Lauria did so at the battle of the Gulf of Naples on 5 June 1284. See
Pryor, “Roger of Lauria”, p. 192.
392 CHAPTER FOUR
at night, and to have warned against lack of vigilance when moored

because the enemy might attack by day or night.633 However, there
must be serious doubts about the practicality of some of this advice. If
an enemy fleet was in difficulties in a storm, surely one’s own would
be also? And, even if it is true that ancient and medieval galleys still
floated when waterlogged because they carried as little ballast as
possible and were buoyant,634 nevertheless, what would be the point of
attacking a waterlogged wreck? Moreover, engaging at night was
highly dangerous because of the impossibility of maintaining
formation, of signalling effectively, and of manœuvring in squadrons.
Battles at night would inevitably degenerate into ship-to-ship mêlées
whose outcomes would be totally unpredictable. Only two naval
battles which took place at night are known to us from the Middle
Ages. The second occurred by accident rather than design: the battle
of Las Hormigas in 1285. However, it does appear that Nasar did
succeed in destroying the Aghlabid fleet somewhere off western
Greece in 880 by deliberately attacking at night and that he gained a
major victory because the Muslims were unprepared and could not
rally to organize themselves.635 He must have been extremely
confident and have taken a fearful gamble.
The primary consideration remained to avoid being caught
unawares and taken by surprise. In 822 one of the fleets of Thomas
the Slav was caught unawares by the imperial fleet in port somewhere
in Thrace at the Byrides and was destroyed by Greek Fire.636 In 879
Nike2t as O›oryphas transported his fleet by land across the isthmus of
Corinth to take the Cretan Muslims by surprise in the Gulf of Corinth
and defeat them before they had a chance to rally.637
Both caution and common sense required close attention to the
weather. As we have seen, Constantine VII possibly had a treatise on
these matters compiled from submissions sent to him.638 Leo VI
------------------------------
633
See Appendix Two [a], §§33, 57; Appendix Two [b]), §§1, 5. Cf. Appendix
Five, §§31, 54; Appendix Eight [a], p. 245.
634
There are many examples from antiquity. See Morrison, et al., Athenian trireme,
pp. 127-8; Morrison, Greek and Roman oared warships, p. 30. After the battle of
Sybota of 433 B.C.E. between the Corinthians and the Korkyraioi, the victorious
Corinthians were able to recover their own wrecks, nauagivai (navagiai), and a
relieving Athenian squadron had to make its way up Corfu Strait through floating
corpses and wrecks. See Thucydides, Peloponnesian war, I.50.5-I.51.4 (vol. 1, pp. 86-
8).
635
Pryor, “Roger of Lauria”, pp. 195-200; Theophane2s continuatus, V.63 (p. 304).
On the battle see Vasiliev/Canard, Byzance et les Arabes. Tome II, part 1, pp. 95-8.
636
Theophane2s continuatus, II.16 (p. 64).
637
638
enjoined his strate2goi to acquire practical experience of meteorology

and astronomy and to take account of relationships between weather,
the seasons, the stars, and signs of the zodiac. Voyages and
engagements should only be undertaken in calm weather.639 Once
again, however, the practicality of some of this advice is questionable.
It was all very well to advise a strate2gos to be acquainted with
practical meteorology but, unless he had grown up as a sailor all his
life, it was not a skill he was likely to learn overnight. Much more
probably, Byzantine aristocrats of the type who were habitually
appointed to command of fleets would have depended for such
knowledge upon veteran mariners such as their pro2tokaraboi, just as
Nike2phoros Ouranos recommended.640
If we really can trust Leo VI, Byzantine naval expeditions appear to
have been rather formally organized, perhaps too formally. They may
have been only too predictable to enemies and too easy for spies to
find out about. The Byzantines themselves, of course, both employed
spies and also sent out scouting expeditions to gather intelligence
about the deployment of Muslim naval forces. Muslim powers did the
same, employing their Christian as well as Muslim subjects. All
powers at all times used spies. Prokopios referred to both the
Byzantines and Persians using them.641 Merchants and political envoys
were no doubt particularly useful and that they might act as spies was
a commonplace.642
According to Theophane2s the Confessor, when the later caliph
Mu‘a2wiya ibn Abı3 Sufya2n, who was then still governor of Syria,
raided Cyprus in 648, he was able to break off the assault on the island
and sail for Arados when he was informed by spies or by scout ships
that a large relieving force under the koubikoularios Kakorizos had set
out against him.643 The Vita Basilii of the Theophane2s continuatus
recounted the story of an Muslim spy sent from Syria around 880 in
preparation for an attempted Egyptian/Syrian naval expedition against
Byzantium who was so impressed by the naval forces gathered at
Constantinople that his report to home dissuaded his masters from
------------------------------
639
See Appendix Two [a], §§2, 31, 40, 49; Appendix Two [b], §4. Cf. Appendix
Five, §§2, 29, 38, 47; Appendix Eight [a], p. 245.
640
See above p. 360.
641
Prokopios, History of the Wars, I.xxi.11 (vol. 1, p. 196).
642
On military intelligence and espionage in general in Byzantino-Muslim warfare
see Christides, “Military intelligence”; Dvornik, Intelligence services, pp. 235-61;
Koutrakou, “Diplomacy and espionage”; idem, “Spies of towns”.
Here we distinguish between “spies” gathering intelligence and “scouts” of
military expeditions, even though the two are frequently confused in the sources.
643
Theophane2s, Chronographia, A.M. 6140 (vol. 1, pp. 343-4).
394 CHAPTER FOUR
their attempt.644 The story reads like a didactic fancy but it surely
reflects a well-known reality of Muslim espionage against Byzantine
forces. It should be emphasized that assembling naval expeditions in
particular was not something that could be done overnight.
Preparations might take months or even years and such preparations
would come to the notice of enemy powers via reports of spies or
merchants, or merchants who were spies, or political envoys who
were also spies.
In the ninth century, according to the Life of St Gregory of
Dekapolis, Gregory was suspected by the men of Otranto of being a
spy for the Muslims since they put a turban on his head. Also in the
ninth century, according to the Life of St Elias the Younger, the saint
and one of his disciples were arrested and imprisoned as “spies of
towns”, katavskopoi tw'n povlewn (kataskopoi to2n poleo2n), near
Bouthro2ton in Epiros by agents of the strate2gos of the thema of
Nikopolis on suspicion of spying for a nearby Muslim force. In the
eleventh century, Kekaumenos related the story of a flotilla of 5
Muslim ships which put in to De2me2trias pretending to wish to trade
but which then sacked the town when a traitor helped them.645
Leo VI stressed the need for secrecy when it came to the
technology and stratagems of naval warfare.646 However, preparations
for large-scale naval expeditions would almost certainly have been
impossible to keep hidden from enemy eyes and ears.
Squadrons were collected from ships of the thematic and imperial
fleets and then assembled at various aple2kta, depending upon
objectives. Leo VI advised that when the assembled fleet set out it
should proceed in squadrons, “according to the formation which has
been exercized”, with sufficient distance between each ship to prevent
collisions under oars. This latter he derived from the excerpt on
crossing rivers from the Strate2gikon attributed to Maurice.647 A
reading of §30 of the Emperor’s Naumachika Leontos Basileo2s rather
gives the impression of the ships moving on the sea as though they
were pieces on a chess board and, impressed by this paragraph, R. H.
Dolley once wrote that: “... preliminaries over, the fleet weighed
anchor and stood out to sea. This operation had to be carried out in
------------------------------
644
See Theophane2s continuatus, V.68 (pp. 308-9). Cf. Dvornik, Intelligence
services, pp. 147-8; Koutrakou, “Diplomacy and espionage”, p. 132.
645
Vita S. Gregorii tou Dekapolitou, §13 (p. 58); Vita di Sant’ Elia il giovane, §28
(pp. 42-5); Kekaumenos, Strate2gikon (Spadaro), §84 (pp. 124-7).
646
647
See Appendix Two [a], §§25, 30; Appendix Two [b], §4. Cf. Appendix Five,
§§23, 28; Maurice, Ek tou Maurikiou, §4 (p. 41).
perfect order, the different squadrons weighing in regular succession

and keeping station to the best of their abilities”.648 However, the
meaning of the Emperor’s instructions requires consideration. What
did he mean by “according to the formation which has been
exercized”? It cannot have been battle formation because no fleet
would make a voyage in battle formation. Even if weather conditions
made it possible to do so, there would be no purpose in maintaining a
fleet in battle formation until the final approach to battle and
attempting to do so would only lead to the very collisions that Leo and
Maurice envisaged and warned about. Moreover, later experience
showed clearly that even attempting to move and make progress while
in battle formation slowed fleets down so that approaches to battle
became played out in slow motion.649 No commander would wish to
move so slowly during transit voyages. Additionally, Syrianos
Magistros specifically advised against drawing up in battle formation
too much in advance of engagement so that the enemy would not have
time to adopt an appropriate counter formation.650 But why would a
strate2gos exercize a formation for a voyage in transit? During transit
voyages from one aple2kton to another, squadrons would surely have
proceeded independently, if perhaps in loose contact. This conclusion
is supported by the specification in the Ek tou kyrou Leontos tou
Basileo2s, §3, from Constitution XX.220 of Leo’s Taktika, that
destinations should be given to squadron commanders in sealed orders
not to be opened until after they were at sea, and in the specification
of the inventory for the 949 Cretan expedition that the fleet should
proceed in four themata or squadrons.651
Dromons and chelandia formed the main body of a fleet, with a
touldos or touldon, a baggage train of horse transports and supply
ships bringing up the rear.652 These appear to have been sailing ships,
except if perhaps some of the horse transports were galleys designed
for amphibious landings.653 They were sent off to safety in the event of
battle. Apparently preceding and shadowing the main body of the
fleet, according to Leo VI, following Syrianos Magistros, were
------------------------------
648
Dolley, “Naval tactics”, p. 329.
649
See Guilmartin, Gunpowder and galleys, pp. 54 (Prevesa, 1538), 201-3, 248
(Lepanto, 1571).
650
Appendix One, §§9.31, 39.
651
Appendix Two [b], §3. See also above p. 267-8 and n. 349.
652
See Appendix Two [a], §§11, 13, 22, 23; Appendix Five, §§10, 11, 20, 21. This
may have come from the Strate2gikon attributed to Maurice. See Maurice, Ek tou
Maurikiou, §5 (pp. 41-2). Very interestingly, Ibn Mankalı3’s translator knew that horse
transports hippagogoi, were tara’ı3d in Arabic. See Appendix Eight [a], p. 243.
653
See above pp. 274-5, 305.
396 CHAPTER FOUR
numbers of light dromons, faster than the norm, variously called

mone2reis or galeai, used for purposes of scouting and
communications.654 However, there is a serious problem with this. It is
a landlubbers’ misconception that smaller and lighter ships will be
faster than larger and heavier ones. But this is not in fact the case if
the designs are the same and the ships differ only in scale.655 Either
that or even though they were popularly known as a type of dromon,
galeai really did have some different design characteristics which
made them faster than standard dromons and chelandia. As early as
the fourth century Vegetius reported that in order to avoid detection
Roman scout ships, scaphae, rowing only 40 oars, had their sails,
rigging and pitch on their hulls dyed blue, and that the crews also
wore clothes dyed blue.656 The expedition of the tourmarche2s Melito2n
to Crete around 920-21, with four chelandia, found him scouting
around Kythe2ra and in 960 Nike2phoros Pho2kas used the strate2gos of
Thrake2sio2n to reconnoitre, kataskovpevw (kataskopeo2), Crete.657
Signals were important. In battle, when voice or trumpet could not
be heard, flags were used. Leo VI’s discussion of signalling suggests
that the same essential principle was used as in signalling-by-flags in
the modern era, different flags and different positions conveying
different messages. It is puzzling, however, that the emperor
mentioned neither the smoke signalling nor the signalling by mirror,
kavtoptron (katoptron), that were mentioned in the Naumachiai of
Syrianos Magistros, a text with which he was familiar as we have
seen, and which were later repeated by Nike2phoros Ouranos.658
Quite complex orders to a fleet could apparently be conveyed from
the flagship by use of a signal flag or banner, called a kamelauvkion
(kamelaukion) by Leo VI, by raising or lowering it, by inclining it to
right or left, shifting it to the right or left, by waving it, and by
changing its patterns or colours or those of its “head”, kefalhv
(kephale2).659 The emperor appears, however, to have derived this from
a passage in the Strate2gikon attributed to Maurice dealing with the use
of flags, bavnda (banda), by divisional army commanders, meravrcoi
------------------------------
654
See Appendix Two [a], §§10, 33, 76-77, 81; Appendix Three, §3.2; Appendix
Five, §§9, 31, 69-70, 74. Cf. Appendix One, §6.2, Appendix Eight [a]. p. 243.
655
See above pp. 130-31 and n. 26.
656
Vegetius, Epitoma, IV.37 (pp. 153-4).
657
See Vita S. Theodori, pp. 287 and cf. above p. 191; Leo the Deacon, Historiae,
I.3 (p. 9).
658
See Appendix One, §7.1; Nike2phoros Ouranos, Ek to2n taktiko2n , §119.2 (5) (p.
94).
659
See Appendix Two [a], §§41, 44-8; Appendix Five, §§39, 42-6; Appendix Eight
[a], pp. 246-7, [b], p. 122.
(merarchoi), for conveying orders.660 The “head” of a bandon, and

hence no doubt of a kamelaukion, was the main part of the flag, its
field, as opposed to the tails or streamers, flavmoula (phlamoula),
attached to its fly.661 [Cf. Figure 26]
A fleet was to be exercized in carrying out commands signalled to
it in these ways. Leo VI wrote that the signals included those to
engage with the enemy and to disengage, to slow or speed up an
advance, to set up an ambush or come out from one, and to come to
the assistance of a section in difficulties. Clearly the emperor was able
to envisage signals which were more complex than general orders to
the whole fleet to attack or retreat, etc. Squadrons could be identified
both to do something and to have something done for them; as, for
example, “Squadron one reinforce squadron five”. This has
implications for what signal flags must actually have been. Squadrons
were probably under the command of kome2tes in the tenth century,
even though Leo VI and Nike2phoros Ouranos also used other classical
or non-technical terms for such commanders: navarchos, he2gemo2n,
arche2gos. The Strate2gikon attributed to Maurice referred to such
squadrons as tagmavta (tagmata) or mevrh (mere2) and to their
commanders as moiravrcai (moirarchai), meravrcai (merarchai), or
a[rcwnte" (archo2ntes). However, these were all terms used in land
armies in the sixth and seventh centuries and whether they were used
in fleets in the tenth century is questionable. But, that the ships of
different squadrons were identified by their own squadron flags,
bavnda (banda), as the Strate2gikon said, can hardly be doubted.662
Leo VI equated what was the contemporary term for a signal flag,
kamelaukion, apparently by analogy to the imperial “cap” or diadem,
kamhlauvkion (kame2laukion), to the classical term for a battle flag,
foinivki" (phoinikis).663 In classical Greece, the phoinikis had been a
red or purple banner. Leo indicated by his use of the imperfect “they
------------------------------
660
See Maurice, Strate2gikon, IIIB.16 (pp. 260-62): “Crh; to; ijdiko;n tou' meravrcou
bavndon mh; movnon ejxhllagmevnon to; ei\do" para; ta; a[lla, ta; ujp aujt o;n bavnda, poih'sai,
i{na eujepivgnwstovn ejsti pa'si toi'" uJp aujto;n bandofovroi", ajlla; mh;n kai; dia; kinhvsewv"
tino" xevnh" ejn tw'/ i{stasqai, oi|on h] a[nw h] kavtw h] dexia; h] ajristera; sunecw'" ejpiklivnein
kai; ejgeivr ein th;n kefalh;n tou' bavndou h] puknw'" tinavssein ojrqovn, w{ste kai; ejnteu'qen
aujto; ejn tai'" sugcuvsesin eujkovlw" uJpo; tw'n loipw'n bavndwn gnwrivzesqai.”.
661
On Byzantine battle flags in general see Babuin, “Standards”; Dennis,
“Byzantine battle flags”.
662
Maurice, Ek tou Maurikiou, §§2, 3 (p. 41).
663
See Appendix Two [a], §47: “En ga;r polevmou kairw'/ shmei'on ei\con th'"
sumbolh'" ai{ronte" eij" u{yo" th;n legomevnhn foinikivda: h\n de; to; legovmenon
kamelauvkion ejpi; kontarivou uJyouvmenon, mevlan th;n crovan kai; a[lla tina; kata; to;n
o{moion trovpon uJpodeiknuvmena .”. Cf. Appendix Five, §45.
398 CHAPTER FOUR
used to have”, ei\con (eichon), that this was an ancient device no

longer used but one which he equated to the contemporary
kamelaukion. However, the kamelaukion was black, not red. In his
version of the emperor’s text, Nike2phoros Ouranos omitted the
equation of the kamelaukion to the phoinikis but confirmed that its
colour was black. Again there is something rather puzzling here.
Black would be an extremely poor colour for a signal flag, very
difficult to see against the dark blue background of sky and sea,
especially in poor weather. Syrianos Magistros had said that signals
were made with very white fabric waved around and Nike2phoros
Ouranos paraphrased him but limited the context to those of scout
ships sent on ahead, almost as though he knew that Syrianos rather
than the emperor was correct but wished to avoid a clash with Leo
VI’s recommendations which he was obliged to repeat.664
A kamelaukion had a head, kephale2, whose appearance and colour
could be altered, and presumably tails or streamers, phlamoula,
although these were not named by Leo VI.665 As we saw above in the
context of the helmsmen, sipho2n operators, and bow-hands, Leo VI’s
specification of the numbers of personnel was not necessarily the total
on board. Should we assume that each dromon had only one
kamelaukion, or may we read him as referring to a generic and that
dromons may have had many such kamelaukia? Here we point out
that one of the inventories for the Cretan expedition of 949 said that
each dromon should have 50 kamelaukia.666 This text has invariably
been interpreted as referring to kame2laukia caps, but why should it
be? Indeed, what could have been the purpose of 50 soft caps in the
context of an inventory of a dromon’s armaments? We do not deny
that elsewhere the meaning of kamelaukion obviously was a cap of
some sort,667 but here they cannot have been caps for the lining of
helmets because the same inventory specified 80 helmets. Why should
the specification not have referred to a sophisticated system of
signalling flags used by the Byzantines? How could a commander
------------------------------
664
See Appendix One, §7.1: “Shmei'a de; kata; me;n qavlattan ta; leukovtera tw'n
uJfasmavtwn kinouvmena,...”; Nike2phoros Ouranos, Ek to2n taktiko2n, §119.2 (5) (p. 94),
checked by us against the manuscript Oxford, Bodleian Library, MS. Baroccianus
Graecus 131: “Shmavdion poiou'sin eij" th;n qavlassan pro;" tou;" ojpivsw ta;
proapostellovmena ploi'a eij" bivglan livna leuka; kinouvmena ...”.
665
See Appendix Two [a], §§44-6; Appendix Five, §§42-4.
666
667
Nike2phoros Pho2kas, followed by Nike2phoros Ouranos, clearly used the term
with this meaning. See Nike2phoros Pho2kas, Praecepta militaria, I.3 (p. 12); cf.
Nike2phoros Ouranos, Taktika, c. 56.3, in McGeer, Sowing the dragon’s teeth, p. 90.
change the “colour” and “appearance” of a flag without actually

changing the flag. And, that is what Nike2phoros Ouranos appears to
have said. Where Leo VI had written that it might be changed by
having its head sometimes made to look different, Nike2phoros altered
that to read that it should actually be changed or exchanged, for
another flag is the implication. We suggest that Byzantine dromons
may have carried a whole wardrobe of kamelaukia for signalling. If it
was not beyond the ability of the British navy in the eighteenth
century to develop such a wardrobe, why should it have been beyond
that of the Byzantines? They had a very long tradition of naval
warfare to draw upon. Such a wardrobe of kamelaukia flags would
have been necessary to convey the kind of complex messages Leo VI
envisaged. If it was not beyond the ingenuity of Kleoxenos,
De2mokleitos, and Polybios himself to devise a system of signalling
with fires which could represent letters of the Greek alphabet, why
should it have been beyond the ability of others to have devised a
similar system with flags? 668
In the approach to battle it was essential to draw up a fleet in
formation. Syrianos Magistros emphasized the importance of this and
discussed how the commander should maintain the formation.669 A
disorganized fleet dared not engage because its ships would be unable
to lend support to each other and would be overwhelmed. This was the
cardinal sin that Constans II supposedly committed in 655 when he
went into the Battle of the Masts without bringing his fleet into
formation and was annihilated, barely escaping with his life.670 In 904
the droungarios tou ploimou Eustathios Argyros had to break off his
attack on the fleet of Leo of Tripoli because he had not been able to
draw up in a counter formation, ajntitavxasqai (antitaxasthai), his own
fleet.671 Leo then went on to sack Thessalonike2. The reverse occurred
in 956 or 957 when Basil Hexamilite2s, the strate2gos of the
------------------------------
668
Polybios, Histories, X.43.1-X.47.11 (vol. 4, pp. 206-218). A possibility which
escaped Babuin, “Standards”, p. 22; Kolias, “Kamelaukion”.
669
Appendix One, §9.4-7.
670
Theophane2s, Chronographia, A.M. 6146 (vol. 1, p. 346): “tou' de; basilevw"
mhde;n poihsamevnou pro;" paravtaxin naumaciva", ...”; al-T4abarı3, Ta’rı3kh (Yar-Shater),
A.H. 31 (vol. 15, p. 76).
We say “supposedly” because according to al-T4abarı3 both fleets were in fact
drawn up in tight formation. However, whether the Byzantines were actually in
formation or not is unimportant. What is important is that both Theophane2s and al-
T4abarı3 knew that they ought to have been in formation.
671
See Theophane2s continuatus, V.20 (pp. 366-7): “ajpostevllei ou\n oJ basileu;" to;n
Eujstavqion to;n thnikau'ta drouggavrion meta; stovlou kata; tou' Tripolivtou: o}" mh;
dunhqei;" ajntitavxasqai touvtw/ ajntestravfh kenov".”.
400 CHAPTER FOUR
Kibyrrhaio2tai, successfully beat back a fleet from Tarsos larger than

his own because he managed to form his own fleet into a counter
formation, ajntipavrataxi" (antiparataxis), before engagement.672 One
of the classic battle tactics was to disorganize an enemy’s formation
by feigning flight until the enemy ships in pursuit became strung out
and then either to send in fresh reinforcements against the
disorganized enemy or to turn around in formation and overwhelm the
disorganized enemy ships one by one.673 The Carthaginians employed
the tactic at the battle of Eknomos in 256 B.C.E. and a millennium and
a half later Roger of Lauria used the tactic to great effect at the Battle
of the Gulf of Naples on 5 June 1284.674
It was also essential to maintain formation as long as possible.
According to Theophane2s the Confessor, in 790 a Muslim fleet
moving north from Cyprus in fair weather was carried about at sea.
There is a sense in the Greek that the Muslim fleet was disorganized.
Perhaps as a consequence, one of the two Byzantine commanders,
Theophilos, the strate2gos of the Kibyrrhaio2tai, was over confident,
moved out ahead to engage by himself, and was captured by the
enemy.675
According to Leo VI, the standard formation was the line abreast in
a shallow, crescent-moon semi-circle, with the flagship at the centre
of the line in its “deep”, and the stronger and larger dromons at the
ends of the line. This was also supposedly the best formation for
making a fighting retreat by backing water.676 Other formations may
also have been used in various circumstances: a straight line, or
several lines or squadrons, some of which could attack from the flanks
or the rear once the enemy was engaged by the main formation.677
------------------------------
See Theophane2s continuatus, VI.Autokratoriva Kwnstantivnou.29 (p. 453).
672
673
See Appendix Two [a], §§54, 56; Appendix Five, §§52-3; Appendix Eight [a], p.
248, [b], p. 123.
674
Polybios, Histories, I.27.7-10 (vol. 1, p. 76); Pryor, “Roger of Lauria”, pp. 189-
95.
675
Theophane2s, Chronographia, A.M. 6282 (vol. 1, p. 465): “oiJ de; “Arabe"
kinhvsante" ajpo th'" Kuvprou, kai; eujdiva" aujtou;" katalabouvsh", periefevronto ejn tw'/
pelavgei. ajnafanevntwn de; aujtw'n th;n gh'n. ei|don aujtou;" oiJ strathgoiv, kai;
parataxavmenoi hJtoimavsqhsan tou' polemei'n. Qeovfilo" dev, oJ tw'n Kiburaiwtw'n
strathgov", rJwmalevo " ajnh;r kai; iJkanwvtato" w[n, qarshvsa" kai; pavntwn proexelqw;n
touvtoi" te sumbalwvn, ejkrathvqh uJp aujtw'n, ...”.
According to the Me2nologion of Basil II (Rome, Biblioteca Apostolica Vaticana,
MS. Vat. Gr. 1613), Theophilos attacked in his dromon but was abandoned by the
three other strate2goi with him because they were envious of his valour. See
Anonymous, Menologion, coll. 285-8.
676
See Appendix Two [a], §§2, 28, 42, 49, 50, 78; Appendix Two [b], §2;
Appendix Five, §§26, 32, 47, 48, 71; Appendix Eight [a], p. 247, [b], p. 123.
677
See Appendix Two [a], §§51-2; Appendix Five, §§49-50; Appendix Eight [a],
This latter would seem to have been dependent upon having an

overwhelming superiority in numbers. As Leo VI himself wrote, in
what must be something of a classic of understatement, the same was
true of the tactic to disengage a formation that had fought the enemy
to a standstill and then to send in reinforcements.678
There is no doubt that the crescent-moon line abreast formation
was that which was used normally. Evidence for it goes back to
antiquity and it remained the standard battle formation until the end of
the days of galley warfare in the Mediterranean. Its objective was to
overwhelm the ends of an enemy line so that galleys at the ends of
one’s own line could turn in the enemy’s galleys and attack their
exposed flanks where they were most vulnerable.
There are obvious problems with some of the rest of the emperor’s
advice. Much of it was paraphrased from Syrianos Magistros and
Maurice.679 To use the crescent formation for a fighting retreat as he
advised would be ludicrous. It would certainly prevent the enemy
overwhelming isolated ships; however, backing water is both
extremely tiring and extremely slow and also makes it very difficult to
hold a course because rudders are ineffective.680 A fleet worsted in
battle would have great difficulty holding formation if backing water
and would never succeed in disengaging from the enemy and escaping
by doing so. The oncoming enemy would simply keep pressing onto
the retreating ships backing water until their crews were exhausted
and could be overwhelmed. Almost certainly, the emperor got the idea
for this recommendation from Thucydides, either directly or
indirectly. But Thucydides’ context was one of a small Athenian
squadron of twelve ships retreating by backing water into harbour in
the face of superior Peloponne2sian forces in order to cover their retreat
and that of their worsted Korkyraioi allies.681 The tactic could work if
retiring back into a protected position, but not if caught exposed at
sea. In 87 B.C.E. the Rhodian fleet also used retiring by backing water
to effect a retreat back into the safety of Rhodes harbour in the face of
a superior fleet of Mithridate2s VI Eupator of Pontos.682
------------------------------
pp. 247-8, [b], p. 123.
678
See Appendix Two [a], §55; Appendix Five, §52; Appendix Eight [a], p. 248,
[b], p. 123.
679
See Appendix One, §§9.30-41; Maurice, Ek tou Maurikiou, §§3, 7 (pp. 41-2).
680
See Shaw, “Rowing astern”. Cf. Morrison, et al., Athenian trireme, p. 247.
Academic references are unnecessary. Pryor has experienced it himself on many
occasions on the water.
681
See Thucydides, Peloponnesian war, III.78.3 (vol. 2, p. 136).
682
Appian, Mithridatic Wars, 24 (p. 282).
402 CHAPTER FOUR
To adopt a straight front when one wanted to use the sipho2nes for
Greek Fire, as the emperor recommended in §51, makes no apparent
sense. When two opposing crescent formations clashed head on, they
would inevitably straighten out in any case as the galleys engaged
successively from the ends of the lines towards their middles. Why
would maintaining a straight line rather than a crescent in the
approach make any difference if one intended to use the sipho2nes? It
is possible that what the emperor had in mind was that in order to use
the sipho2nes one would have to engage with the enemy ships more
closely than would be necessary for a missile exchange and, if so, it
may be that he had in the back of his mind some passages from
Syrianos Magistros which he did not reproduce but which had the idea
of an engagement from the convex side of a crescent. According to
Syrianos, beginning in a straight line, the fleet engaged at the ends of
the line and then its centre pressed forward forming a convex crescent
until the whole fleet was engaged. The objective was to break through
the enemy line at the centre and split it into two by positioning one’s
strongest ships in the centre of one’s own line.683 It is just possible that
this passage gave Leo VI the idea of how to engage closely in order to
bring the sipho2nes into play, but it makes little apparent sense.
For lack of any ship-killing weapon it is highly improbable in fact
that any tactical manœuvres whatsoever could have proved decisive.
Medieval naval battles became a matter of approach in formation,
attempts to hold formation above all costs in order to protect the
vulnerable sides and sterns of the ships, and then an inital phase of
engagement by extensive exchanges of missiles designed to degrade
the enemy’s manpower before close engagement and boarding.
Missile exchange at a distance continued to be the initial phase into
the High Middle Ages throughout the Mediterranean.684 That was why
dromons had a forecastle, a pseudopation, at the prow, from which
marines could hurl missiles against an enemy ship.685
Such missiles employed the same processed fire material as used in
the sipho2nes but hurled by catapult, either in pottery jars or in the
form of caltrops wrapped round with tow and soaked in it.686 There
can be no doubt that the former at least were used because examples
------------------------------
683
See Appendix One, §§9.35-40.
684
See Alexandres, ÔH qalassiva, p. 62; Pryor, “Roger of Lauria”, pp. 179, 186-7,
207.
685
See Appendix Two [a], §6; Appendix Five, §5; Appendix Eight [a], p. 242, [b],
p. 21. Cf. above p. 203.
686
See Appendix Two [a], §§63, 65; Appendix Five, §60.
survive.687 Then there were the other more conventional missiles that
one would expect: ordinary caltrops, arrows shot by bows, rocks, and
the small “arrows” or bolts known as “flies” or “mice” and shot by
bow-ballistae.688 Quite probably the jars of unslaked lime mentioned
by Leo VI were also used but one may have one’s doubts about the
practicality of jars full of poisonous reptiles.689 From his tone, even the
emperor seems to have had doubts about that one. The most effective
missiles and those which formed the bulk of those exchanged were
rocks, caltrops, arrows from bows, “flies” or “mice” from the bow-
ballistae, and then javelins when closed up somewhat more. The large
numbers of such missiles mentioned in the inventories for the Cretan
expedition of 949, and the absence of more “exotic” projectiles in
them, show that this was so: in particular, 10,000 caltrops, 50 bows
and 10,000 arrows, 20 hand-held bow-ballistae and 200 “mice”, and
100 javelins per dromon.690 In the spring of 822 the fleet of Thomas
the Slav opened its engagement with the imperial fleet in the Golden
Horn by hurling rocks.691
The importance of proper management of the preliminary missile
phase was indicated by the emperor’s insistance on using them
effectively, not wasting them against an enemy protected by shields,
and ensuring that neither supplies were exhausted nor the crews
exhausted themselves in hurling them. The Muslims of Cilicia, he
wrote, were well trained in naval warfare and covered up with their
shields until an enemy had exhausted his missiles before engaging.692
He appears to have appreciated that battles were not won in missile
phases and that, although these might influence the outcome, hand-to-
hand combat decided it.
In the final phase of battle opposing ships grappled. The words
used by the Anonymous, Leo VI, and Nike2phoros Ouranos to describe
this phase were desmov" (desmos), a bond, and desmei'n (desmein), to
bind or fetter, both connected to desmeuvein (desmeyein), to bind or
fetter or tie together. We have chosen “couple” and “to couple” as the
------------------------------
687
See Christides, “New light”, pp. 19-25.
688
689
See Appendix Two [a], §§60-61; Appendix Five, §§57-8; Appendix Eight [b], p.
124. Or did the emperor have a recollection of the report of John Malalas that the asp
which killed Cleopatra was one of those which she carried in her ships for purposes of
battle. See John Malalas, Chronographia, trans. Jeffreys, et al., p. 116.
690
See Appendix Four [b], §§II.13-18 [= Haldon, “Theory and practice”, p. 225;
691
Theophane2s continuatus, II.15 (p. 62).
692
See Appendix Two [a], §§15-17; Appendix Five, §§13-15; Appendix Eight [a],
p. 244; Leo VI, Taktika (PG), XVIII.121-2 (coll. 973-6).
404 CHAPTER FOUR
closest translation of what we believe was intended. When the crew of

an enemy ship was sufficiently degraded to make boarding and
capturing realistic, iron rods, kamavke" sivdhrai (kamakes side2rai), no
doubt with grappling hooks at both ends, were used to grapple with it
and couple it so that it could not escape. The tactic used in defence
against enemy ships trying to do the same was to keep the ships apart
by using even longer poles: ajkovntia (akontia) or kontavria (kontaria).
The bronze poles, ajkovntia calka' (akontia chalka), mentioned in the
inventories for the Cretan expedition of 949, may possibly have been
for this purpose; although, they probably were not. Bronze would not
have been a very good material since the poles would have been
heavy to wield and bronze is also soft, but brittle, easily broken by an
iron axe.693 Coupling and preventing coupling were apparently
procedures which required considerable practice and exercize, to
judge from Leo VI’s insistence on the point and his words of warning
that the procedure was not always advantageous.694
From this point the fully-armed soldiers on the upper oar banks of
the dromons came into play.695 During the missile phase, they were
almost certainly stationed on those parts of the decks called by the
Anonymous, by analogy to the half-decks of trie2reis, katavstrwmata
(katastro2mata), along the sides behind their shields slung on the
kastello2mata; although, neither of these were words used by Leo VI
and Nike2phoros Ouranos. At this point fights must have degenerated
into hand-to-hand mêlées. The only potentially decisive weapons left
at this point were the “cranes” called geraniva (gerania), which if we
can believe Leo VI could pour Greek Fire already alight onto the deck
of an enemy ship coupled alongside, and the rocks or iron weights
hurled from the xylokastra in attempts to smash the deck and
ultimately the hull of the enemy ship.696 However, we have
reservations about the practicality of both of these suggestions of the
------------------------------
693
See Appendix Two [a], §§28, 68; Appendix Five, §§26, 62. Cf. Appendix Three,
§5.2. See also Appendix Four [b], §VII.18 [= Haldon, “Theory and practice”, p. 233;
Constantine VII, De cerimoniis, II.45 (vol. 1, p. 677)]. It is interesting that Ibn
Mankalı3 chose not to include the recommendations on coupling in his treatises.
Note also that the akontia chalka appear in the inventories immediately before
what appear to be items of cooking equipment. They may have been rods for
suspending pots over fires.
694
See Appendix Two [a], §§28, 37, 68; Appendix Five, §§26, 35, 62. Cf.
695
See Appendix Two [a], §§9, 14, 20, 73; Appendix Five, §§8, 12, 18, 66. Cf.
Appendix Three, §§2.7, 5.pr., 1.
696
See Appendix Two [a], §§7, 67; Appendix Five, §§6, 61. Ibn Mankalı3 included
the towers, rocks and weights, but not the cranes. See Appendix Eight [a], p. 242, [b],
p. 22.
emperor. Both sinking and, especially, setting fire to an enemy ship

coupled alongside would pose obvious dangers to one’s own ship.
Finally, we consider two techniques recommended by Leo VI
which appear to us to be once again the fire-side musings of an arm-
chair sailor. In one paragraph, the emperor recommended thrusting
pikes, mevnaula (menaula), through the oar-ports, trype2mata, of the
lower oar-bank of a dromon and thus slaughtering the enemy.697 He
said that he had recently devised this technique himself, which
immediately arouses suspicions in any case. First, in order for
pikemen to thrust pikes out through the lower oar-ports, the oar
sleeves would have to be removed so that they could see. This would
leave the dromon prone to flooding. Secondly, one would also have to
remove the oars, thus robbing the ship of motive power. Thirdly, what
could such pikes actually hit? They obviously could not hit anyone on
the deck of the enemy ship because they would be right down close to
the waterline. Therefore, in order to actually hit anyone on the enemy
ship, they would have to be thrust through its own lower oar-ports.
But these were covered by their own oar sleeves and so the pikemen
would be operating blind. Fourthly, ships move constantly in relation
to each other, even in battle. Even if coupled together, they would still
pitch and roll relative to each other. A pike with its haft inside the oar-
port of the attacking ship and its head through the oar-port of the
other, would either cause chaos on the thwarts of the attacking ship as
the haft was thrown around all over the place or it would be instantly
snapped if the two opposed oar-ports changed their relative positions
sufficiently.
In the very next paragraph, Leo VI also recommended holing the
hull of an enemy ship from the lower oar-bank and Nike2phoros
Ouranos added that pikes were to be used for the purpose.698 However,
pikes would make very ineffective boring instruments. The magistros
was probably just trying to guess at how the emperor had thought that
it could be done by analogy to the preceding paragraph. Moreover,
below the water line the hull of any galley curved sharply in towards
its keel, and thus away from any ship lying alongside it. It is difficult
to imagine how anyone operating from the lower oar-bank of a
dromon could make any hole in the hull of an enemy ship below the
------------------------------
697
Dolley also doubted, correctly, whether this technique was “really practical”.
See his “Naval tactics”, p. 333. Ibn Mankalı3 chose not to include it.
698
See Appendix Two [a], §70; Appendix Five, §64. Again, Ibn Mankalı3 chose not
to include this recommendation.
406 CHAPTER FOUR
water line. The angle of approach to the hull of any instrument used
for the purpose would surely be so acute as to make it impossible to
get a grip on it in order to penetrate it. Moreover, with both ships
constantly moving with the seas, how could the instrument be kept in
one place long enough to perform the task?
CHAPTER FIVE
THE DEMISE OF THE DROMON
Detailed sources for dromons gradually disappeared after the tenth

century as the intellectual movements which gave rise to the
encylopaedic literature of which the military manuals, De cerimoniis
inventories, and other compendia were a part, waned. The term
dromo2n continued to be used in Byzantine narrative and documentary
sources through to the end of the twelfth century; however, few details
emerge from these mentions to add to what is known from the tenth
century. From the twelfth century the references increasingly have the
appearance of literary anachronisms and although pictorial
representations of Byzantine galleys begin to survive again, they are
never sufficiently realistic to found any details of reconstruction on
them. None show more than a single mast or more than one file of
oars.1
Considerable numbers of graffiti of ships survive in churches and
other buildings dating from the twelfth century and later; however, as
discussed above, their dating is impossibly indeterminate for the most
part and in many cases they may not even have been intended to
represent Byzantine ships. They may well have represented Western
ships.2 There are also two surviving seals dating from the later twelfth
or early thirteenth centuries which do depict what must have been
intended to be Byzantine galleys. The first belonged to a Manuel
Raoul, of the well-known Byzantine family of the Raoul, descended
from a Latin who had settled in the Empire. It depicts a galley with
three oarsmen and one helmsman and flying a tri-streamered flag. The
other belonged to a Theodo2ra Komne2n e2, wife of an Isaakios who was
a megas doux but who is otherwise unidentified. The galley has four
oarsmen, probably a helmsman, and again a tri-streamered flag.
------------------------------
1
Christides has reproduced what he says are illustrations of bireme dromons in
“Ibn al-Manqalı3 (Manglı3)”, pp. 89-93 and figs 4, 5, 6, which he has taken from
Anderson, Oared fighting ships, fig. 11 and plate 7B, where the ships were, indeed,
described as dromons. However, Anderson cited no sources and was in fact being
loose with his terminology. The illustrations were not of dromons at all. They were of
Genoese and Sicilian galeae of the twelfth and early thirteenth centuries in the
Annales Ianuenses of Genoa and the De rebus Siculis carmen of Peter of Eboli. See
Figures 50, 54.
2
See above, pp. 239 & nn. 246-7.
408 CHAPTER FIVE
Neither seal is sufficiently detailed to deduce anything about the

construction of the galleys.3
The chronicle ascribed to Pseudo Symeon magistros said that the
fleet led by Nike2phoros Pho2kas against Crete in 960 was composed of
2,000 chelandia equipped with Greek Fire, 1,000 dromo2nes, and 360
transports (karabia).4 However, use of the words gradually became
less frequent. Kekaumenos equated long ships, makrai; nh'e" (makrai
ne2es), with chelandia and wrote that they should be manned by
archers, toxovtai (toxotai). He also referred to dromo2n es in the hands
of the strate2gos of Ragusa, Katakalo2n Klazomenite2s, in the context of
raids in the Adriatic by the Serbian toparche2s of Zenta and Stamnos.5
However, use of the words gradually became less frequent. Michael
Attaleiate2s (ca 1020 - post 1085) did not use the words dromo2n and
chelandion at all. He invariably used the classical word trie2re2s to refer
to a warship. John Skylitze2s (fl. second half of the eleventh century)
did refer to dromo2nes, in particular the “imperial dromon”, basiliko;"
drovmwn (basilikos dromo2n), on several occasions, but without giving
any details of the ships,6 as did Nike2phoros Bryennios (ca 1064-
1136/7), who also mentioned an “imperial dromon” for the use of the
emperor on four occasions.7 By the twelfth century, George Kedre2nos
could still refer to chelandia in action against the Bulgarians in the
fleet of Constantine V; however, he was just copying his source, in
this case probably Theophane2s the Confessor.8 John Zo2naras († post
1159?) did not use the word dromo2n. When referring to ships and
fleets, he invariably used generic or classical words such as ploia,
ne2es, stolos, and trie2reis.
The Alexiad of Anna Komne2n e2 is more problematical since the first
draft of it, and also large sections of what now survives, were
arguably written by her husband, the Caesar Nike2phoros Bryennios.
Yet other sections betray their origins in reports of various military
------------------------------
3
Zacos, Byzantine lead seals, plate 189, nos 2751, 2751a.
4
Pseudo Symeon magistros, Chronographia, p. 758. Cf. Theophane2s continuatus,
VI.Basileiva ÔRwmanou' uiJou' Kwnstantivnou tou' Porfurogennhvtou.10 (p. 475). The
figures are obviously inflated in both cases.
5
Kekaumenos, Strate2gikon (Spadaro), §74 (pp. 108-11).
6
See John Skylitze2s, Synopsis historio2n, Kwnstanti'no" oJ uiJov" Levonto".13 (p.
210), ÔRwmano;" oJ Lakaphnov".6, 18 (pp. 215, 223-4), Basivleio" kai; Kwnstani'no".7 (p.
258), Kwnstanti'no" oJ Monomavco".6 (p. 431).
7
Nike2phoros Bryennios, Hyle historias, I.4, 21, III.22, 24 (pp. 81, 125, 249, 251).
8
George Kedre2nos, Synopsis historio2n, vol. 2, p. 15, l. 6. Cf. Theophane2s,
Chronographia, A.M. 6257 (p. 437). But see also George Hamartolos, Chronikon
syntomon, IV.ccliii.33 (col. 944).
DEMISE OF THE DROMON 409
Figure 47
Dromon in a manuscript of the Sermons of St Gregory of Nazianzos (Mount
Athos, Pantelee2mon, Cod. 6, fol. 138r), twelfth century.
commanders.9 Which words were Anna’s and which were those of her
sources, particularly in the sections dealing with military campaigns,
is problematical. In one passage the surviving text equated Venetian
dromo2nes to trie2reis.10 In another, identified as having the literary
footprint of Nike2phoros, the imperial fleet under Nicholas
Maurokatakalo2n in 1096 was referred to as being composed of
“die2reis, trie2reis, and some dromades nee2s”,11 which, in the context,
may equally have meant either “swift” ships or ships “of the type of
dromons”. In yet another, most probably derived from reports of
Landulf and Tatikios, the commanders of the Byzantine fleet, the
Pisan fleet of the First Crusade was described as being composed of
die2reis, trie2reis, and dromo2ne, as well as other fast-sailing ships.12 In
------------------------------
9
Howard-Johnston, “Anna Komnene”. Cf. Macrides, “The pen and the sword”.
10
Anna Komne2ne2, Alexiade, VI.v.9 (vol. 2, p. 54): “Kairou' d ojlivgou
parerruhkovto" drovmwnav" te kai; trihvrei" eujtrepivsante" oiJ Benevtikoi ...”.
11
Anna Komne2ne2, Alexiade, X.viii.3 (vol. 2, p. 216): “..., ta;" tou' o{lou stovlou
dihvrei" kai; trihvr ei" kai; tina" dromavda" ajnalabovmeno" nau'" ...”. Cf. Howard-
Johnston, “Anna Komnene”, n. 50 (p. 283).
12
Anna Komne2ne2, Alexiade, XI.x.1 (vol. 3, p. 42): “... dihvr ei" te kai; trihvrei" kai;
drovmwna" kai; e{t era tw'n tacudrovmwn ploivwn ...”. Cf. Howard-Johnston, “Anna
Komnene”, n. 64 (p. 292).
410 CHAPTER FIVE
his own Hyle historias, Nike2phoros Bryennios only once used the
classical term trie2reis. Elsewhere, except for when he referred to the
“imperial dromon”, he used the generic ne2es for “ships”, even when it
is clear that the ships in question were war galleys.13 It is most
probable that Anna replaced generic and contemporary terms used by
her sources with the classical terms die2reis and trie2reis.
In another passage of the Alexiad describing the battle of Corfu in
1084 and referring to Venetian ships, nh'e" (ne2es), which would
probably have been galleys of a Western rather than Byzantine type
by that time, but which presumably reflected Anna’s, or her source’s,
understanding of Byzantine galleys, she suggested that they had
multiple wales, at least two of which were normally underwater. She
wrote that because the Venetian ships had been unloaded, they were
sailing light and the water did not come up to even the second wale,
zo2ste2r.14 That Byzantine galleys did indeed have multiple wales is
confirmed by a passage in Rhodanthe and Dosikles. Theodore
Prodromos wrote that: “As much of them [the ships] as were not
submerged but rode above the waves of the sea, from the second wale
to the third, ...”.15 Prodromos also referred to the hulls of trie2reis being
covered from the second to the third wales with thick, matted felt in
which incoming enemy missiles would stick, so that they could not
come inboard, thus avoiding injury to the crews.16 Since there would
have been solid hull between the second and third wales, this does not
make any sense; however, there is at least a clear reference to three
wales. Nike2tas Cho2niate2s, who is considered to have begun writing his
Historia at Constantinople under the Angeloi emperors but who
completed it in exile at Nicaea after 1204, also indicated that
Byzantine galleys had at least three wales. Describing the Sicilian
------------------------------
13
Nike2phoros Bryennios, Hyle historias, III.3 (p. 215). Cf. II.27, III.3 (pp. 199,
215).
14
Anna Komne2ne2, Alexiade, VI.v.7 (vol. 2, p. 53): “... wJ" mhvd a[cri deutevrou
zwsth'ro" tou' u{dato" fqavnonto", ...”. In our opinion, the passage in which this occurs,
describing the naval battle of Corfu between the Venetians and the forces of Robert
Guiscard, also bears the literary imprint of Nike2phoros Bryennios; although, it is not
one of those identified as such by Howard-Johnston.
15
Thodore Prodromos, Rhodanthe and Dosikles, bk. 5, ll. 449-51 (p. 89): “o{son
ga;r aujtai'" oujk ejbaptivsqh kavtw, / ajll uJperevplei th'" qalavssh" th;n rJavcin, / ejk
deutevrou zwsth'ro" a[cri kai; trivtou / ...”.
16
Theodore Prodromos, Rhodanthe and Dosikles, bk. 5, ll. 451-9 (p. 89): “ejk
deutevrou zwsth'ro" a[cri kai; trivtou / pivloi" kateskevpasto nastoi'", pacevsi: / boulh'"
sofh'" eu{rhma kai; strathgiva", / wJ" a]n ta; plei'sta tw'n tetamevnwn belw'n / ejkei'
paraklwvqointo, mhd ej" to; provsw / e[coien ejlqei'n kai; balei'n tou;" ejn mevsw/, / ajll
hjremoi'en ejmparevnta toi'" pivloi". / a[nw d ejp aujtw'n tw'n teqeimevnwn pivlwn / plhqu;"
parh/wvrhto makrw'n ajspivdwn, ...”.
fleet that attacked Thebes and Corinth in 1147, he wrote that the
Sicilian commander ordered his ships to be so loaded with booty that
they sank up to the third wale. Again his comment is likely to have
been based on whatever knowledge he had of Byzantine, rather than
Sicilian, galleys. When Nike2tas referred to Sicilian trie2reis being so
overloaded that they were submerged nearly to the level of the upper
eiresia, file of oarsmen, thus implying a construction similar to that of
tenth-century bireme dromons with superimposed banks of oarsmen,
he was undoubtedly engaging in some classical allusion.17 Whether
they were classical Greek trie2reis or tenth-century dromons or any
other galleys, the only way that galleys with such an oarage system
could be submerged to the level of the upper bank of oarsmen would
be if they had been sunk and the hulls entirely flooded. Taken
together, these three passages suggest that Byzantine galleys had at
least three wales, one below water, a second at or around the water
line, and a third on the upper hull.18
In the mid twelfth century, in a eulogy for the emperor Manuel I
probably delivered after the Norman attack on the Empire in 1157,
Michael the Rhetor mentioned dromo2nes amongst other types of ships
in a fleet put together by Manuel to counter the attack,19 and
Eustathios of Thessalonike2 (ca 1115-1195/6), in another eulogy for
Manuel dated to Lent 1176, mentioned dromo2nes, horse transports,
and trie2reis in a fleet raised by Manuel to counter the Venetian threat
in 1172.20 Once again, these sources were orations composed in a
classicizing style and neither of them can be relied upon, even so far
as to maintain that galleys called dromo2nes still actually existed in
Byzantine fleets of the mid twelfth century.
When the word dromo2n began to be used in Western literary
sources in various transliterations such as dromundus, dermundus,
etc., it became applied indiscriminately to large ships of any kind. It
------------------------------
17
Nike2tas Cho2niate2s, Historia, p. 74, ll. 33-4: “... ajll o{ron tiqei;" th'" ejfevsew" to;
kai; eij" trivton zwsth'ra th'/ oJlkh'/ tw'n crhmavtwn ta;" pavsa" h] ta;" pleivou" nh'a"
baptivzesqai, ...”; p. 76, ll 94-5: “... kai; th'" a[nw eijr esiva" ejgguv" pou baptomevna" tw'/
rJeuvmati.”. Cf. Heliodo2ros, Aithiopika, I.i (p. 3): “to; ga;r a[cqo" a[cri kai; ejp i; trivton
zwsth'ra th'" new;" to; u{dwr ajnevqliben: ...”.
It is just possible that in the early twelfth century some Western galleys still had
superimposed banks of oarsmen emulated from the dromon. See below pp. 424-6.
However, it is more probable that they did not and that Cho2niate2s’ mental model was
a Byzantino-Greek classical conceit.
18
Western galleys of the thirteenth century normally had five wales. See Pryor,
“Galleys of Charles I of Anjou”, pp. 48-9.
19
Michael Rhetor, “Oratio ad Manuelem imperatorem [2]”, p. 156.
20
Eustathios of Thessalonike2, “Oratio ad Manuelem imperatorem [2]”, p. 37.
412 CHAPTER FIVE
was known to the Anglo-Saxon earldorman Æthelweard, probably

from Isidore of Seville, as early as the 980s. He used dromo to
translate the Latin longae naves or the Anglo-Saxon words ceol
(“keel”) and scip into Latin for his chronicle when referring to the
ships of the Danes and Anglo-Saxons when they first arrived in
Britain.21 Geoffrey Malaterra used the word in contradistinction to
what were other names known to him for types of oared galleys:
galeae, catti, and golafri.22 The Liber Maiolichinus de gestis
Pisanorum illustribus on the Pisan expedition to the Balearics of
1114-15 described the fleet as being composed of:
Gatti, drumones, garabi, celeresque galee,

Barce, currabii, lintres, grandesque sagene,
Et plures alie variantes nomina naves.23
Sometimes, however, the word was definitely applied specifically to

sailing ships. For example, the Anglo-Saxon pilgrim Saewulf, who
went to the Holy Land in 1102-3, referred to the ship on which he left
for home from Jaffa as a dromundus but his evidence makes it clear
that this was a sailing ship.24
This Western literary usage may have one or both of two
explanations. Westerners may simply have adopted the word in
transliteration for large ships of any kind because dromo2n had become
the standard nomenclature for major units of Byzantine fleets and the
------------------------------
21
Æthelweard, Chronicle, I.3 (p. 7): “Advecti igitur sunt praefati iuvenes
explentes petitionem regis senatusque, cum tribus dromonibus armis ornati, ...”
Æthelweard may have taken this passage directly from the Latin text of the Historia
ecclesiastica gentis Anglorum of the Venerable Bede, I.xv: “Tunc Anglorum sive
Saxonum gens invitata a rege praefato, in Brittaniam tribus longis navibus
advehitur.”, with “tribus longis navibus” glossed as “tribus dromonibus” on the
authority of Isidore of Seville. See Bede, Historia ecclesiastica, p. 68. On Isidore see
pp. 126 & n. 14, 128, 134-5 above.
Alternatively Æthelweard may have derived this passage from a manuscript that
was an ancestor of that now known as Ms. E of the Anglo-Saxon chronicle, Oxford,
Bodleian, MS. Laud Misc. 636, where the word for ship was ceol. This passage
appears only in MS. E. See Anglo-Saxon chronicle, p. 10. Æthelweard also used
dromones in other passages apparently derived from MS E of the Anglo-Saxon
Chronicle which post-dated Bede’s Historia ecclesiastica, where the word for ship
was scip. See Plummer and Earle, Saxon chronicles. See Æthelweard, Chronicle, III.1
(p. 26) [= MS. E, annus 787], III.4 (p. 31), III.4 (p. 33), IV.3 (p. 41) [= MS. E, annus
875], IV.3 (p. 44) [= MS. E, annus 882].
22
Geoffrey Malaterra, De rebus gestis, II.8 (p. 32): “Nostri denique tantum modo
germundos et galeas, Sicilienses vero cattos et golafros, sed et dromundos, et diversae
fabricae naves habebant.”.
23
Liber Maiolichinus, ll. 106-8 (p. 10).
24
See Pryor, “Voyages of Saewulf”, pp. 49-51.
word had then become diffused as such in the nautical lingua franca of
the Mediterranean. However, even if so, by the twelfth century the
word may have become applied already to transports sailing ships as
well as galleys, both in the Empire and across the Mediterranean at
large. No eleventh- or twelfth-century Byzantine sources elucidate the
issue because they simply used the word without ever attributing to it
any specific characteristics of either sailing ships or galleys.
The word became widely used in Old French literature, probably
for the first time in the surviving literature in the Chanson de Roland.
In the Old French version of the Oxford manuscript, Bodleian Library,
MS. Digby 23, the Muslim amı3r Valdabron was said to be the master
of 400 drodmunz. The amı3r Baligant summoned his men from forty
kingdoms and commanded his great drodmunz to be made ready and
later was said to have: “I do not know to tell you how many drodmunz
...”.25 In various forms (dromon, dromont, dromunt, dromund) the
word appeared also, for example, in La chevalerie d’Ogier de
Danemarche,26 in the Chanson de Guillaume,27 in Le couronnement de
Louis,28 in the Charroi de Nimes,29 in Aliscans,30 in La Fin d’Elias,31 in
the Anglo-Norman Roman de Rou of Wace,32 in the Roman
d’Auberon,33 and in Blancandin et l’orguielleuse d’amour.34 Benoit de
Sainte-Maure used it in his Roman de Troie of ca 1160-70, as did
Chrétien de Troyes in his romance Cligés of ca 1176.35 Chelandion
also found its way into Old French as calant and chalant; although,
these terms were used less widely than the various forms of dromo2n .36
Use of the word spread as far as Norway and Iceland but not, to the
best of our knowledge, to Germany. In the Old French and Anglo-
Norman versions of the twelfth-century chanson of Bueve de Hantone
------------------------------
25
Chanson de Roland. Vol. 1: La version d’Oxford, ll. 1564, 2624, 2730. The
word also appears in other forms (dormun, dromon) in other MSS. See vol. 2, ll.
2810, 2918; vol. 4, ll. 4537, 4728; vol. 7, ll. 931, 2310.
26
Ogier de Danemarche, ll. 2325, 2348, 3070.
27
Chanson de Guillaume, vol. 2, ll. 213, 2368, 3008, 3059, 3517.
28
Couronnement de Louis, l. 1327.
29
Charroi de Nimes, l. 212.
30
Aliscans, ll. 18, 2268.
31
Fin d’Elias, l. 1254.
32
Roman de Rou, pt II, l. 2002.
33
Roman d’Auberon, ll. 2421, 2424.
34
Blancandin, ll. 2743, 2948, 3188, 3864, 4308.
35
Cligés, l. 6575; Roman de Troie, l. 27, 566.
36
Blancandin, ll. 2134, 2187, 2752, 3952, 5300, 5311; Ogier de Danemarche, l.
2325; Le Chevalier au Cygne, l. 142; Fin d’Elias, l. 1256; Aliscans, ll. 17, 2267;
Mortier, ed., Chanson de Roland, vol. 4, ll. 4588, 4726; vol. 7, ll. 2253, 2309; Roman
de Rou, pt II, l. 4039; Chanson de Guillaume, ll. 1725, 2354, 3517, 3522.
414 CHAPTER FIVE
(Old French) or Boeve de Haumtone (Anglo-Norman), dromont (Old

French) and dromoun (Anglo-Norman) was used for ships, as also was
calant. In the Anglo-Norman version dromoun was used for a Muslim
ship.37 In the thirteenth century, the chanson was rendered into Old
Norse as Bevers Saga and in it the hero was taken by ship to Egypt on
a drómundr full of heathens.38 Even earlier, in the Saga inga konungs
og brå∂ra hans of the Heimskringla of Snorri Sturluson, nine ships
under Jarl Rögnvald of the Orkneys and Erling Skakki on their way to
the Holy Land came upon a large drómundr. It was not said whether
this ship was a sailing ship or a galley but from the fact that the
“heathen” (hei∂nir) aboard the drómundr were able to rain weapons
and stones and pots of boiling pitch and oil down upon the Norse
ships, most probably a large sailing ship was intended.39 In Grettis
saga Ásmundarsonar, Grettir’s brother Thorstein Asmundson was
nicknamed “Dromund” because he was tall but slow of mien,
suggesting that these were qualities that the Norse associated with the
word and what they understood of the ship type by that time.40
Forms of both dromo2n and chelandion also continued to be used in
Arabic sources. Writing in the early eleventh century, ante 1034,
Yah5y a2 ibn Sa‘ı3d al-Ant6a2ki, wrote under the year 999 that when Basil
II besieged Tripoli, two shalandiyya2t supplied his forces from the sea.
In his monumetal Al-Ka2mil fı3 ’l-ta’rı3kh, Ibn al-Athı3r used the word
shalandiyya2t on four occasions. Under the year A.H. 201 he reported
that the Muslims captured nine large mara2kib ships with their men as
well as shalandiyya2t, which in this context he may have intended to
refer to the ships’ boats. In A.H. 244, the governor of Syria, Al-
‘Abba2s ibn al-Fadl ibn Ya‘qu2b, was said to have sent a fleet against
Syracuse which encountered a Christian fleet of 40 shalandiyya2t.
When the capture of Enna was announced to the Byzantine Emperor,
he was said to have sent a fleet of 300 shalandiyya2t to Sicily under the
command of a patrikios. Finally, the fleet sent to Ifrı3qiya by the
Almohad Caliph ‘Abd al-Mu’min in A.H. 554 included 70 shawa2nı3,
------------------------------
37
Bueve de Hantone, ll. 1603, 8248; Boeve de Haumtone, l. 354 and cf. l. 2744.
38
Bevers saga, §5 (p. 295): “Nú taka teir sveininn og fluttu hann út til hafsins og
fundu tar einn drómund fullan af hei∂ingjum.”.
39
Snorri Sturluson, Heimskringla, Saga Inga konungs og bræ∂ra hans, §17 (vol. 2,
p. 780): “En Rögnvaldur og Erlingur skakki hittu á drómund einn í hafi og lög∂u til
níu skipum og bör∂ust vi∂ tá. En a∂ lyktum lög∂u teir snekkjurnar undir drómundinn.
Báru ta hei∂nir menn ofan á tá bå∂i vopn og grjót og gry)tur fullar af vellanda biki og
vi∂smjörvi.”. Cf. Orkneyinga Saga, §88 (p. 224).
40
Grettis Saga, ch. 13 (p. 34): “Tu Rannveig áttu tann son, er Torsteinn hét,
manna frí∂astr ok sterkr ma∂r, raddma∂r mikill ok hár ma∂r á vouxt ok noukkut seinligr í
vi∂brag∂i; tví var hann drómundr kalla∂r.”.
t6ara2’id, and shalandiyya2t.41

Four texts from the late-twelfth and early-thirteenth centuries
suggest that by that time at least the word dromo2n had indeed become
applied to transport ships and that Byzantines had begun to use a
bireme galea, probably as developed in the Latin West, as their main
battle galley. One of the last references to dromons to add any detail
to the mere mention of the ships occurs in the chronicle of the
historian of the Crusader States, archbishop William of Tyre. Writing
of the Byzantine fleet sent to the Holy Land in 1169 by Manuel I
Komne2nos to join the Franks of Outremer in a combined assault on
Egypt, William wrote that it was composed of 150 “longae naves
rostrate, arranged with twin ordines of oars”, which he said were
commonly known as galee, as well as 60 “[naves] maiores ad
deportandos equos deputate”, horse transports equipped with stern
ports and ramps for embarking and disembarking, and 10 or 12
“[naves] maxime que dromones dicuntur”, very large transports called
dromones.42
Either William did not understand the traditional Byzantine
meaning of dromo2n or else by the time that he was writing the word
dromo2n really had become applied to transport ships. On the one
hand, if in fact he misunderstood the Byzantine terminology, he may
be read as saying that the Byzantine fleet had 150 battle galleys,
presumably dromons, which he then equated with Western galee with
which he was familiar, and 60 specialized horse transports as well as
10 or 12 general transports. William used the classical “rostrate”,
“rammed”/“with rams”, to describe the longae naves or galee;
however, this was merely conventional classicizing affectation and the
reference must have been to spurs. Then, he said that they had twin
ordines of oars. If this was not yet another literary allusion,43 it may
------------------------------
41
Yah5ya2 ibn Sa‘ı3d, Histoire, p. 459; Ibn al-Athı3r, Al-Ka2mil (Tornberg), vol. 6, p.
339. Ibn al-Athı3r, Al-Ka2mil (Fagnan) pp. 227, 228, 585. Cf. above pp. 48, 98. We
have not attempted to comb systematically the Arabic sources for the use of the words
since there is little point in doing so. These are merely two instances which have come
to our attention.
42
William of Tyre, Chronicon, 20.13.7-17 (vol. 2, p. 927): “Erant sane in prefato
exercitu naves longe, rostrate, geminis remorum ordinibus instructe, bellicis usibus
habiliores, que vulgo galee dicuntur, centum quinquaginta, item his maiores ad
deportandos equos deputate, ostia habentes in puppibus ad inducendos et educendos
quoque eos patentia, pontibus etiam, quibus ad ingressum et exitum tam hominum
quam equorum procurabatur commoditas, communite, sexaginta, item harum maxime,
que dromones dicuntur, alimentis varii generis armisque multiplicibus, machinis
quoque et tormentis bellicis usque ad summum referte, decem aut duodecim.”. Cf.
above, pp. 109, 114-15.
43
William’s mental model may well have been derived from Isidore of Seville.
416 CHAPTER FIVE
have been intended to refer either to two files of oars, both rowed
from a single bank above deck, or to two superimposed banks of oars.
Ordo could have either meaning. If William did misunderstand the
Byzantine terminology, then the latter meaning of ordo, two
superimposed banks, is possibly correct and William may therefore be
a witness to the survival of the traditional dromon to the late twelfth
century. On the other hand, if William recorded the composition of the
Byzantine fleet correctly, then we have the first evidence for the
passing of the dromon as a battle galley and the adoption by the
Byzantines of some type of bireme galea in its place. In this case ordo
almost certainly referred to two files of oars rowed from the same
benches above deck.
There are four reasons for believing that William did in fact record
the composition of the fleet correctly and that therefore he is the first
witness to the demise of the dromon as a war galley. First, William
had been to Constantinople on two occasions and must have been
familiar with Byzantine war galleys. It had been he who had been sent
to Constantinople in 1168 by Amalric I of Jerusalem to negotiate with
Manuel I Komne2nos for this very same combined assault on Egypt
and he wrote that the terms of the agreement for the expedition were
drawn up in documents which he himself brought back to Jerusalem.
These would surely have contained the specifications for the fleet.
William also spent another seven months in Constantinople later, in
1179-80, on his way home from attending the Third Lateran Council.44
Secondly, in his description of this Byzantine expedition to Egypt
Nike2tas Cho2niate2s also wrote that the fleet was composed of 200 long
ships, makrai ne2es, of which 60 were trie2reis sent to Acre under
Theodore Maurozome2s to embark the Frankish cavalry, amongst other
purposes. The identity in the figures is too striking to ignore and there
is no evidence that Nike2tas either had access to a manuscript of
William of Tyre or could even read Latin. Nike2tas is therefore an
independent witness to the veracity of William’s account, at least in so
far as the number of horse transports are concerned. This being so,
there is no reason per se to question the rest of it.45 And, thirdly, the
------------------------------
See Isidore of Seville, Etymologiae, XIX.i.23: “Biremes autem naves sunt habentes
remorum ordinem geminum.”. William’s use of the less expected gemini rather than
the more obvious duo to qualify ordines, suggests that he had Isidore in mind.
44
William of Tyre, Chronicon, 20.4, 22.4 (vol. 2, pp. 916, 1009-1010).
45
Nike2tas Cho2niate2s, Historia, Basivl eiva Manouh;l tou' Komnhnou'.EV (p. 160, ll. 36-
44): “..., stovlon katartuvei baru;n kata; tou... ajpo; de; toutwni; tw'n trihrevwn eJxhvkonta
tw'/ Maurozwvmh/ paradou;" Qeodwvrw/ pro;" to;n rJh'ga ejxevp emye, prokataggelou'nta me;n
kai; th;n o{son oujd evpw tou' loipou' stovlou ajnagwgh;n kai; th;n ejkei'se tou' Kontostefavnou
anonymous author of the Old French translation of William of Tyre,

known as the Eracles, did not change William’s specifications for the
expedition of 1169, even though he was generally well versed about
naval matters and did not hestitate to alter or gloss William’s text on
other occasions when he thought him to be wrong or unclear.46
Nike2tas Cho2niate2s used the word dromo2nes only in one instance. In
all other cases he used either various circumlocutions or else the
classical word trie2reis. The one exception occurred in his description
of the Venetian fleet of the Fourth Crusade of 1202-4, which he said
consisted of “110 horse-carrying dromo2n es and sixty long ships, nh'e"
de; makraiv (ne2es de makrai)”.47 It is well known from Latin sources
that the horse transports for the Fourth Crusade were usserii/huissiers/
oxerii, which were transport galleys of the type of taride, emulated
from Muslim t6ara2’id, or chelandra, derived from the Byzantine
chelandion.48 These 110 “dromo2n es” were not battle galleys but rather
horse transports with stern ports and ramps for embarking and
disembarking cavalry. The sixty long ships referred to by Nike2tas
were the galleys of the Venetian battle fleet. It should be borne in
mind that Nike2t as was in Constantinople in 1203-4 and was an eye
witness to the assault. His apparent misapplication of the term dromo2n
is further evidence that by the time he was writing, it had indeed
become applied to transport ships and that Byzantines had adopted
some other type of galley as their battle galley.
The third instance in the same period of the application of the term
dromo2n to some type of ship distinguished from a war galley occurs in
the Itinerarium peregrinorum. Describing a relief fleet of 15 ships
sent by S4ala2h5 al-Dı3n to the relief of Acre in 1190 during the Third
Crusade, the author described it as being composed of three larger
ships, “which they call dromones in the vernacular”, following behind
lighter and swifter galee. Again, a contemporary eye-witness used a
------------------------------
a[fixin, dianasthvsonta de; kajkei'non wJ" ei[h eJtoimasavmeno" ta; pro;" e[xodon, a{ma d
ajpokomivsonta kai; ta; tw'n ÔIerosolumitw'n iJppevwn ojywvnia, o{soi sunekstrateuvein
h[mellon tw'/ rJhgi; sunefaptomevnw/ tou' polevmou kai; kat Aijguptivvwn tiqemevnw/ th;n
kivnhsin, kaq w|n oJ stovlo" oJ ÔRwmai>ko;" sugkekrovthto.”.
It should be noted that Nike2tas did not derive these figures from John Kinnamos.
Kinnamos described the fleet as being composed of “vessels, horse transports, and
very numerous warships” (“..., stovlon new'n oJ basileu;" tekthnavmeno" iJppagwgw'n te
kai; polemisthrivwn ...”) but did not give any figures. John Kinnamos, Historiae, VI.9
(p. 278).
46
Eracles, XX.13 (p. 961). Cf. above p. 286 and nn. 413-14.
47
Nike2tas Cho2niate2s, Historia, Basileiva Alexivou tou' Aggelou'.BV (p. 539): “...
dromwvnwn me;n iJppagwgw'n eJkato;n devka, nhw'n de; makrw'n eJxhvkonta, ...”.
48
See Pryor, “Transportation of horses by sea”, p. 21; Idem, “Crusade of Frederick
II”, pp. 125-7.
418 CHAPTER FIVE
Latin form of dromo2n to describe ships that were larger and slower
than those that he equated with Western battle galleys.49
Finally, in Old French, the continuation of the chronicle of William
of Tyre attributed to Ernoul, similarly identified the term dromo2n in
its Old French form of dromont with a large sailing ship. The report
that while Richard Cœur de Lion was en route from Cyprus to Acre in
1191, his fleet fell in with and sank a large Muslim sailing ship sent
by S4ala2h5 al-Dı3n from Beirut or from Egypt in a last attempt to relieve
Acre was one recorded in several sources, both Arabic and Latin.50 In
the chronicle attributed to Ernoul the ship was said to have been a
dromont.51 In his chronicle of the Fourth Crusade, Robert of Clari on
one occasion used the word dromon in apposition to huissier to refer
to the horse transports constructed by Venice for the Crusade,52 just as
Nike2tas Cho2niate2s did.
The texts of William of Tyre, Nike2tas Cho2niate2s, the Itinerarium
peregrinorum, Ernoul, and Robert of Clari suggest clearly that by the
turn of the twelfth and thirteenth centuries the term dromo2n and its
Latin equivalents had become applied to transport ships and was no
longer used for battle galleys. No other text known to us suggests
anything to the contrary. In his Chronike2 syngraphe2, George
Akropolite2s, the historian of the Empire of Nicaea, used dromo2n
amongst various other words for ships but without making it possible
to tell whether he was referring to a warship type still currently in
use.53
From the thirteenth century even the mere use of the word dromo2n
became infrequent. It was gradually replaced by other terms,
especially kavtergon (katergon), which appears to have been derived
------------------------------
49
Itinerarium peregrinorum (Mayer), p. 348: “Tres maiores naves subsequuntur,
quas vulgo dromones appellant, galee vero leviores et ad quelibet attemptanda
agiliores precedunt.”.
50
See Pryor, Geography, technology, and war, pp. 120-21.
51
Morgan, Continuation, §120 (p. 121): “Dedenz ce qu’il ariva devant la cité
d’Acre, Salahadin faiseit venir une grant nave d’Egipte que l’on diseit le dromont, ...”.
52
Robert of Clari, Conquête de Constantinople, §10 (p. 130): “Quant li pelerin
furent tot asanlé en Venice et il virrent le rike navie qui faite estoit, les rikes nes, les
grans dromons et les uissiers a mener les chevax, et les galies, ...”.
53
George Akropolite2s, Opera, §48 (vol. 1, p. 87): “ta; me;n ou\n creiwvdh eJautoi'"
periepoiou'nto, spavnin de; tw'n ajnagkaivwn toi'" ÔRwmaivoi" ejk touvtou sunevbh givnesqai,
ejf w|/ tugcavnein drovmwna" kai; e{t era xuvla toi'" Genoui?tai" peiratikav.”; §85 (vol. 1, p.
181): “e[p eisen ou|n aujtou;" eijsiovnta" ejn o{sai" ei\con trihvr esi kai; tisin eJtevroi"
ploivoi" lembadivoi" oi\on kai; drovmwsi, ...”. See also Akropolite2s’ funeral oration for
John III Doukas Vatatze2s. Ibid., vol. 2, p. 15: “oJ de; to; skavfo" paralabw;n smikrovtaton
pavnu kai; o{son oujc aJplw'" koivlhn nh'a ajll oujde; drovmwna h] levmbo" logivzesqaiv te kai;
faivnesqai, ...”.
from katav and e[rgon, having the sense of any works or service or
anything else owed or needed for naval warfare. It was originally
applied not to ships per se but rather to crews, to populations owing
military service, and even to armaments.54 Only from the twelfth
century did it become used to describe an actual category of ships, and
even then it appears to have become used as a generic for a warship
Figure 48
Graffito of a katergon? From the monastery of the Blatadon at Thessalonike,
post 1355.
rather than with any specific reference to a ship type. Anna Komnene
referred to the personal galley of Nicholas Maurokatakalon’s vice-
admiral in 1096 as having been called by the crew the “katergon
exkoussaton”.55
------------------------------
54
See Haldon, “Theory and practice”, p. 211: “peri; tou' eJtoimasqh'nai karfivo n
aJrpavgion koinostomai'on lovgw/ celwnw'n kai; skalw'n kai; loipw'n katevrgwn ciliavda" gV,
...” and commentary at p. 270. Cf. Constantine VII, De cerimoniis, II.44 (vol. 1, pp.
658-9).
The people of Samos were referred to as katergon in a chrysobull of Manuel I of
1158. See Diplomata et acta monasterii Sancti Ioannis Theologi in Patmo insula, No
XXVIII, in Miklosich and Müller, Acta et diplomata, vol. 6, p. 111.
According to Michael Choniates, katergokistai were responsible for the
administration of the obligation called ktisis katergon or katergoktisia which was
imposed on coastal populations. See Michael Choniates, Ta Sozomena, vol. 2, p. 107.
55
See Anna Komnene, Alexiade, X.viii.3 (vol. 2, p. 216): “... deuvt eron kovmhta
meta; tou' ijdivo u katevr gou ejxkoussavtou ...”. Cf. XII.viii.8 (vol. 3, p. 81). Exkoussatos
was not a Greek word. It appears to have been Anna’s rendering of the Latin
excusatus, having the sense here of “reserved [for the use of]”.
Howard-Johnston identifies X.viii as the work of Nikephoros Bryennios and
XII.viii as a passage based on military reports, perhaps by the megas doux Isaac
Kontostephanos. See Howard-Johnston, “Anna Komnene”, pp. 279, 283.
See also the continuation of the chronicle of George Hamartolos. George
420 CHAPTER FIVE
A note added in a later hand to folio 18 of a tenth-century

manuscript of a Me2naion, Offices of the Greek Church, for April and
early May, recorded the death in 1179 of a certain Constantine
Doukas and his service at a siege of Ancona, and mentioned that he
had been surrounded there by German and Venetian katerga.56 The
Greek version of the Chronicle of the Morea recorded that a fleet sent
to the Morea in 1263 by Michael VIII Palaiologos was composed of
katerga, karabia, and tarevte" (taretes). Taretes was an adoption into
Byzantine Greek of the Latin taride for horse transports, suggesting
that by the thirteenth century the Byzantines had not only abandoned
dromons in favour of galeae but had also abandoned chelandia as
horse transports in favour of taride. A chrysobull of Alexios III
Komne2nos of Trebizond for the Venetians, dated to 1364, also
referred to katerga.57 Some of the fourteenth-century “short
chronicles” consistently used katergon for both Byzantine and
Western galleys.58 The text known as the Traité des offices attributed
to pseudo-Kodinos and composed ca 1350-1360 referred in many
places to katerga and even to an “imperial katergon” but never to
either dromons or chelandia.59 By the fifteenth century, in the
anonymous verse chronicle known as the Chronicle of the Tocco,
katergon (and katirgon) was clearly used as a Greek equivalent for the
contemporary Latin galea.60 The references could be multiplied but
there is little point in doing so. Eventually, the word gave rise to the
Ottoman Turkish kadirga for a war galley.61
We conclude with a comment on change of terminology in
manuscripts of the “letter” attributed to Pseudo-John of Damascus and
the Letter of the three Patriarchs. The version of the Letter of the
------------------------------
Hamartolos, Chronikon syntomon, VI.viii.2 (col. 1228B): “..., kai; meta; suntomiva"
e[pemyen kavterga eujq evw", ...”.
56
The manuscript is Paris, Bibliothèque Nationale, MS. Gr. 1564. Previously it
was Cod. Reg. 2476 in the French royal library and this note was excerpted from it by
B. Montfaucon in his notice on the manuscript. Montfaucon, Palaeographia Graeca,
pp. 47-8. Constantine Doukas is known from no other source. See Polemis, The
Doukai, p. 191. The dating to 1179 is incorrect and should probably refer to the siege
of Ancona in 1173. The note says that Constantine died seven days after returning
home from the siege. See also Ahrweiler, Byzance et la mer, pp. 260-61.
57
Chronicle of the Morea, l. 4579 (p. 302); Miklosich and Müller, Acta et
diplomata, vol. 3, No. XXXIII (p. 131): “..., tw'n ajrcovntwn kai; kefalavdwn mou, tw'n
katevrgwn, karabivwn kai; eJteJr wn toiouvtwn xuvlwn mou, ...”.
58
See Schreiner, Kleinchroniken, vol. 1, pp. 65, 68, 80, 85, 86.
59
Pseudo-Kodinos, Traité des offices, pp. 167, 186, 236, 237, 286, 287.
60
Chronicle of the Tocco, ll. 335, 478, 533, 546, 562, 599, 611, 1136, 1144, 1838,
1895, 3621, 3747, 3774, 3800-1, 3804.
61
See Kahane and Tietze, Lingua Franca, §785 (pp. 523-6).
three Patriarchs found in the manuscript of the monastery of the

Iviron, Mt Athos, Codex 381, which is dated to 1426, contains the
following clause: “... and capturing these areas in a naval manner by
means of katerga, in number 120, ...”.62 Here the word for ships was
katerga. However, this manuscript was a late reworking of an earlier
version of the Letter of the three Patriarchs which had many variants,
including an ending incorporating material from the letter of Pseudo-
Figure 49
Graffito of a katergon? From Hagia Sophia, Trebizond, probably fourteenth
century.
John of Damascus, where the corresponding word was “drovmorsi”,

editorially emended to “drovmwsi”.63 The use of katerga by the scribe
of the Iviron manuscript either represented his inability to make sense
of the reference to dromons in the earlier manuscript or his knowledge
that dromons had been superseded by katerga.
------------------------------
62
Munitz, Letter of the three Patriarchs, §15 (p. 101): “... kai; katalabw'n ta;
ejkei'se naustolikw'" dia; katevrgwn to;n ajriqmo;n rkV , ...”.
63
CHAPTER SIX
THE TRIUMPH OF THE GALEA
The question remains. Why did the galley which had become known
as the dromo2n disappear? Or to rephrase the question, why did the
word dromo2n cease to be used for war galleys? Byzantines, and
others, continued to have war galleys but they discontinued the use of
the word dromo2n and its variants for them. The dromon itself had
developed in late antiquity because it had some significant
performance advantages over the Roman liburna which have never
been explained. We have suggested that these were related to the
replacement of polyremes by monoremes, of the ram by the spur, of
the square sail by the lateen, of part decks by full decks, and to the
development of new hull design characteristics, particularly at the
bow, which gave greater speed, especially in battle. In its hey-day in
the tenth century, the dromon had been one of the bulwarks of the
Byzantine Empire, together with the armies of the themata. Yet from
the twelfth century it gradually disappeared as a battle galley, its name
became applied to transport ships, and eventually even its name faded
from use. These developments must have been a product of the
evolution of some new form of naval technology which gradually
replaced the dromon as the paramount battle galley in the
Mediterranean. The only convincing candidate for this historical role
is the bireme galley, becoming known as galea, which was developed
and improved over time in the Latin West from the eleventh century
until it attained a standard form by the late thirteenth. There is no
evidence to suggest that any form of galley developed in the Muslim
world ever had performance characteristics of such superiority that it,
rather than the Latin galea, could have played this historic role.
Early Western galeae were almost certainly modelled originally on
Byzantine galeai, or at least that is where the word came from, since,
as we have seen, galeai was used for monoreme dromons at least as
early as 905-6 by Leo VI, whereas the earliest known use of the Latin
term is in late eleventh-century Italo-Norman chronicles.1 The very
------------------------------
1
See Leo Marsicanus, Chronica, III.25 (p. 716): “..., duabus galeis armatis
insulam ingressus est, ...” [written ca 1087-1105]; William of Apulia, Gesta, V.339
(p. 297): “Quamque magis celerem cognoverat esse galeram / Scandit; ibi posito
Roberti corpore transit, ...” [written ca 1095-99]; Geoffrey Malaterra, De rebus gestis,
424 CHAPTER SIX
fact that the term does first appear in Latin in these sources from
South Italy adds to the weight of evidence suggesting an adaptation of
the ship type and an adoption of the use of the term for it in South
Italy from the Byzantine originals no doubt encountered there by the
Normans and others. Very little is, in fact, known about early Western
galeae even though references to them proliferated extremely rapidly
in the chronicles from the early twelfth century. Although frequently
mentioned, they were never described in any detail and documentary
sources recording construction specifications for them do not survive
before the late thirteenth century. All that is known about early
Western galeae is that they were fast and had fine lines.2
It is not even clear whether they were monoreme galleys at this
time, as Byzantine galeai had been, or whether they were already
biremes. Pictorial evidence does not help a great deal since Western
art depicted no more than schematic monoreme “banana boats” before
the mid twelfth century. The first clear evidence for the construction
of galeae occurs in marginal miniatures in the Paris, Bibliothèque
Nationale, MS. Suppl. Lat. 773 manuscript of the Annales Ianuenses
of Genoa. These annals were commenced as a private record by the
Genoese consul Caffaro around 1100, were adopted officially by
Genoa in 1152, and were continued by him to 1163. Thereafter,
various scribes continued them until 1294.3 Seven miniatures
accompanying references to galeae in twelfth-century entries in the
annals show galleys with marked stern ornaments, pronounced spurs,
and either one or two rows of oar ports.
Although Caffaro’s editor, Belgrano, thought that the miniatures
------------------------------
IV.25 (p. 103): “..., navicula in qua episcopus erat, sociis armis carentibus, a duabus
piratarum navibus, quas galeas appellant, hostiliter aggreditur.” [written ante 1099];
Anonymous chronicle of Bari, p. 153: “Capta est galea Petri de Gira a Saraceni in
Malea. Et galea quatuor Barenses compraehensae sunt a stolo imperatore.”, cf. p. 155
[written ca 1115 but based on much earlier sources].
If the famous Chanson de Roland really can be dated to the late eleventh
century, then we can also add it to the Italo-Norman texts. The word galies/galees is
used twice in the Oxford manuscript of the text for ships among the fleet of the emir
Baligant. See Chanson de Roland, vol. 1, ll. 2625, 2729.
The text “Tunc rex Aelfredus iusssit cymbas et galeas, id est longas naves,
fabricari per regnum ...”, in Asser’s Life of king Alfred, has been shown to be a later
interpolation from Matthew Paris. See Asser, Life of King Alfred, §50c (p. 39) and n.
to §50c.
2
See Amatus of Monte Cassino, L’ystoire de li Normant, V.14 (p. 151): “... et fist
armer de moult sollempnels mariniers .ij. galéez subtilissime et moult vélocissime; ...”
[written ca 1078-83].
On early galeae and the use of the word see Pryor, “From dromo2n to galea”, pp.
108-10; Kahane, “Two nautical terms”, pp. 428-39.
3
See Face, “Caffaro”.
TRIUMPH OF THE GALEA 425
Figure 50
Galleys in the Annales Ianuenses of Genoa (Paris, Bibliothèque Nationale,
MS. Suppl. Lat. 773), ca 1160-1200.
(a) accompanying the entry for 1125
(b) accompanying the entry for 1136
(c) accompanying the entry for 1165
(d) accompanying the entry for 1168
(e) accompanying the entry for 1170
(f) accompanying the entry for 1175
(g) accompanying the entry for 1191
© John H. Pryor
426 CHAPTER SIX
were all drawn by the same artist, it is important to note that the three
earliest miniatures of galeae accompanying entries for 1125, 1136,
and 1165, which date from the lifetime of Caffaro, have two rows of
oar ports, whereas all of those later than this, accompanying entries
for 1168, 1170, 1175, and 1191, have only single rows of oar ports.4
The style of depiction of the galleys also varies considerably,
leading to the conclusion that either the miniatures were not all done
at the one time by the same artist or that if, as they now are in the
Bibliothèque Nationale manuscript, they were all done at the one time
by the same artist, then the originals which he copied were not. The
differences between them may therefore provide evidence of changes
in galea design over the twelfth century. The first three miniatures
show the upper row of oarports in a band at the top of the hull above
the spur and the lower in another band either at the level of the spur or
below it. The evidence of these three miniatures is admittedly
exiguous and no corroborating corollary evidence from any other
pictorial sources dated to the first half of the twelfth century is known
to us. Nevertheless, they may sustain a tentative hypothesis that in the
early twelfth century Genoese galeae were biremes with two
superimposed banks of oars, both rowed through oarports, one above
deck and the other below it, just as Byzantine dromons had been. If
this was the case, then it raises a second possibility that the Byzantine
galeai on which early Western galeae were modelled had also become
biremes by the late eleventh century, whereas in the Macedonian age
they had been monoremes distinguished from bireme dromo2nes and
chelandia. Such a development would have been yet another instance
of the gradual evolution over time of ship types and the names applied
to them, particularly of the tendency of ship types to grow larger,
parallelling that by which biremes also became called dromo2nes
between the sixth and tenth centuries. The four later Genoese
miniatures suggest a possibility that from some time in the second half
of the twelfth century a change was made to another oarage system
which required only one row of oar ports in the hull. That being said,
the evidence is inadequate to sustain the thesis and it is more probable
that Western galeae were from the beginning associated with a new
and different oarage system and that the differences in the miniatures
of the Annales Ianuenses were a product of artistic style only.
More revealing than the miniatures of the Annales Ianuenses are
three illustrations of galleys in the Madrid, Biblioteca Nacional, vitr.
------------------------------
4
Caffaro, Annali, vol. 1, pp. xxv-xxvi.
Figure 51
Bireme Muslim galley in the Synopsis historio 2n of John Skylitze2s (Madrid,
Biblioteca Nacional, vitr. 26-2, fol. 111v), ca 1160.
Figure 52
Bireme galleys in the Synopsis historio2n of John Skylitze2s (Madrid,
Biblioteca Nacional, vitr. 26-2, fol. 145r), ca 1160.
428 CHAPTER SIX
26-2 manuscript of the Synopsis historio2n of John Skylitze2s.5 These

show clearly for the first time,6 bireme galleys which have a different
oarage system. One file of oars is rowed through oarports but the other
is worked from above the gunwale.
Figure 53
Bireme galleys in the Synopsis historio2n of John Skylitze2s (Madrid,
Biblioteca Nacional, vitr. 26-2, fol. 146v), ca 1160.
The first of these three important illustrations (folio 111v) purports to

show four Muslim galleys, drawn in a style similar to that of the
Annales Ianuenses, of which the bottom-most has a second file of
three oars at the stern rowed from above the gunwale in addition to a
file rowed through oarports in the hull. [See Figure 51]
The second (folio 145r) shows three bireme galleys with this same
oarage system drawn in a different and extremely distinctive style
found in none of the other 49 illustrations of galleys in the manuscript.
[See Figure 52] The third occurs in a series of illustrations by an artist
drawing in eclectic styles. In one of these (folio 146v), he depicted
four galleys in a Western style similar to that of the Annales Ianuenses
------------------------------
5
See Appendix Seven.
6
The Roman liburnae of Trajan’s column also had a file of oars worked over the
gunwale. See Lepper and Frere, Trajan’s column, plates 25, 26, 34, 35, 58, 59, 61, 63.
but again, in two of them, with one file of oars rowed through oarports
and the other from above the gunwale. If, as Wilson has argued, the
original Constantinopolitan manuscript was brought to Sicily in 1158,
and the extant Madrid copy was made shortly after that, then these
illustrations are the earliest depictions of this distinctive new bireme
oarage system for medieval Western galleys. They would predate the
earliest of the Annales Ianuenses miniatures showing only a single
row of oar ports, which presumably depict this same oarage system,
even though they do not show the files of oars, by around a decade
and a half.
Figure 54
Sicilian galley in a manuscript of the De rebus Siculis carmen of Peter of
Eboli (Berne, Burgerbibliothek, MS. 120, fol. 119r), early thirteenth
century.
There is no doubt that all three of these Skylitze2s illustrations were

based on Western models. The artists of folios 111v and 146v were
drawing galleys in the style of the Annales Ianuenses but with this
new type of bireme oarage system. The galleys of folio 145r are
extremely similar in design to one shown in the South Italian or
Sicilian manuscript of the De rebus Siculis carmen by Peter of Eboli,
dated to ca 1200. For the first time, this illustration shows a bireme
galley with this same oarage system, but with both files of oarsmen
430 CHAPTER SIX
clearly shown above deck.7 It is almost impossible to believe anything

else than that the Skylitze2s artist of folio 145r and the illustrator of
Peter of Eboli belonged to the same artistic tradition and were
depicting a galley type which had become standard in the Latin West,
or at the very least in Sicily and South Italy, by the sixth or seventh
decades of the twelfth century.
This bireme oarage system was in fact what later became known in
the late Middle Ages as the alla sensile system. Two oarsmen each
rowed single oars from the same bench above deck. They used a
stand-and-sit stroke as opposed to the fully seated stroke that had been
used on classical and Byzantine galleys. The inboard oar was rowed
through an oarport in an outrigger, which now reappeared on war
galleys in the Mediterranean for the first time since the replacement of
trie2reis by liburnae. The outboard oar was rowed from a thole
mounted on the outrigger’s “gunwale”, the apostis. The earliest
documents which confirm that Western galleys were fully decked and
used this alla sensile oarage system date from between 1269 and 1284
from the chancery of the Angevin Kingdom of Sicily during the reign
of Charles I of Anjou.8
Comparing the illustrations of the Skylitze2s manuscript, the four
later miniatures of the Annales Ianuenses, and the Peter of Eboli
illustration, it is apparent that on folios 145r and 146v of the Skylitze2s
manuscript the lower banks of oars are rowed through oarports in the
hull below the level of the spur and a narrow band on the upper hull
which can be presumed to have represented the outrigger or bulwark
above the deck. However, on folio 111v of the Skylitze2s manuscript
and in both the Annales Ianuenses miniatures and the Peter of Eboli
illustration the oar ports are in this upper band of the hull. In the latter
cases, there can be no doubt that what is represented was the alla
sensile oarage system. In the case of folios 145r and 146v of the
Skylitze2s manuscript it is just possible that a transitional oarage
system between the new alla sensile system and the older one was still
in use and that the lower bank of oars was rowed from below deck.
However, more probably the artists of these folios were simply being
inaccurate and the artist of folio 111v had it right. The illustrations of
------------------------------
7
On the interpretation of this illustration see Pryor, “Galleys of Charles I of
Anjou”, pp. 63, 71, 73; idem, “From dromo2n to galea”, p. 110.
8
See Pryor, “Galleys of Charles I of Anjou”, pp. 69-73. The reconstruction there
was based on the assumption that the oarsmen still used a fully seated stroke. Pryor
later changed his opinion on this matter and accepted that a stand-and-sit stroke was
used on these Angevin galleys, presenting a revised reconstruction of the oarage
system in “From dromo2n to galea”, pp. 112-114.
the Skylitze2s manuscript, which have been hitherto almost unnoticed

by maritime historians,9 are thus extremely important and provide the
oldest evidence for the appearance of the alla sensile oarage system.
They give a far more definite picture of it than do the miniatures of the
Annales Ianuenses and the best evidence for it before the Peter of
Eboli illustration.
Even if early Western galeae had two superimposed banks of oars,
by the mid twelfth century this was certainly no longer the case and by
the turn of the twelfth and thirteenth centuries only one text known to
us suggests the use of any oarage system different to the alla sensile:
the anonymous Itinerarium peregrinorum and the chronicles derived
from it. In a very curious passage, its author, while reflecting on the
contrasts between ancient and “modern” naval warfare, clearly used
“ordo” in the sense of superimposed banks of oars.10 However,
whether he really meant that Western galleys of the end of the twelfth
century had superimposed banks of oars is debatable because he was
indulging in a passage of classicizing erudition based on Vegetius.11
That being said, he did continue on and in two places definitely
referred to two superimposed banks of oars on some Crusader galleys
at Acre in 1190. First, he said that on galleys which sortied to engage
the Egyptian fleet coming out to fight from Acre, the shields were
arranged around the “upper benches”, and that the oarsmen “sat in the
lowest part, so that those who were in the highest part for fighting
might fight in freer space”.12 This could only mean that the decks and
their oar benches were cleared for marines and that the oarsmen rowed
from below deck. Secondly, in his description of the ensuing battle, he
said of one galea that:
By now the enemy had boarded another [galley] and having dislodged the
marines [was] master of the upper deck. But those to whom the lower post
had been assigned tried to escape with the help of the oars. Extraordinary
------------------------------
9
However, see now Babuin, “Illuminations”.
10
Itinerarium peregrinorum (Mayer), p. 322: “Apud veteres siquidem in
huiusmodi navibus [liburne] numerosior exigebatur ordo remorum, quibus gradatim
per tabulata distincta surgentibus undas alii longissimo, alii breviore vexabant
impulsu. Ternos autem vel quaternos ordines sepius habebant et quinos interdum, sed
et senos naves quedam in Actiaco prelio, ... Ceterum omnis illa vetustatis
magnificentia imminuta defluxit, nam classis bellica, que senis olim decurrebat
ordinibus, nunc binos raro excedit.”.
11
Cf. Vegetius, Epitoma, IV.33, 37 (pp. 151, 153).
12
Itinerarium peregrinorum (Mayer), p. 323: “In superioribus vero tabulatis clipei
per girum disponuntur conserti, et in imo considunt remiges, ut spatio liberiore
dimicent, qui ad pugnam in suppremo consistunt.”.
432 CHAPTER SIX
and terrible was the conflict; for the oars being pulled in opposite ways,
the galley was driven sometimes this way by our [men], sometimes that
way as the Turks drove [it]. At length our men prevailed, and being
dislodged by the onset of the Christians the enemy rowing above was
overcome.13
This tale of a galley being driven now one way and now the other by
two banks of oars being rowed in opposite directions, one below deck
and the other above it, could only have been possible if the galley had
two banks of oars arranged in the same way that Byzantine dromons
had had. But the entire story is a mere fancy. Even if oars were on two
superimposed banks, they could not possibly have been rowed in
opposite directions without becoming hopelessly entangled. Certainly,
the oars of the three banks of Olympias could not have been because
their blades intermeshed at the waterline.14 Moreover, even if a galley
did have banks of oars whose blades were normally clear of each other
at the waterline, it would not have been possible for antagonists to row
them in opposite directions unless each played the game and
cooperated. In this case, the Muslims on the deck could easily have
fouled the oars of the Christians below by the simple expedient of
rising from their benches until the angle of their oars to the water was
sufficient for the blades to intermesh with those of the Christians.
Equally, the Christians below could have fouled the oars of the
Muslims above by simply raising theirs out of the water until they
intermeshed. No matter which of the two antagonists was trying to
escape, the other could easily have prevented it and smashed oars and
chaos would have been the result in either case. We conclude that this
tale was nothing more than a raconteur’s fanciful entertainment.
This conclusion from the logic of the text is confirmed by
consideration of the same incident as it was related by Ambroise.
Ambroise had been on the Third Crusade and his poem was dependent
either on the Itinerarium peregrinorum or on a now-lost chronicle
common to both. He wrote that: “On the fleets was the din of battle, /
------------------------------
13
Itinerarium peregrinorum (Mayer) p. 324: “Aliam vero iam hostis victor
superioris tabulati bellatoribus depulsis invaserat. At hii, quibus inferior statio fuerat
deputata remorum auxilio elabi contendunt. Mirum quidem et miserandum certamen,
nam remis in diversa tendentibus, nunc huc nostris nunc illuc Turcis agitantibus galea
depellitur, nostri tamen prevalent, et hostis superius remigans christicolarum
superventu detrusus succumbit.”.
14
The modified design for any future Olympias Mark II does, however, allow for
the blades of the thranite oars to be clear of those of the thalamian oars in the water. It
might theoretically be just possible for the two banks of oarsmen to row in opposite
directions if both cooperated with each other. Information supplied by John Coates.
Each was often driven back, / Often together did they come”.15 This
passage incorporated the same essential idea of ships being driven
back and forth as that of the Itinerarium peregrinorum but is
believable because it did not connect it with the conception of a single
galley being rowed in opposite directions by two banks of oars.
There remains the story of the benches above deck being evacuated
to give marines more freedom of action while the galleys were rowed
from “below”. This cannot be definitively disproved; although, three
considerations suggest that it also was a fabrication. First, since the
anonymous Templar chaplain was extrapolating from a comparison to
Roman liburnae with superimposed banks of oars as described by
Vegetius, it is probable that this first part of his story was also a
classicizing literary affectation. Secondly, if the Crusaders vacated the
upper oar benches before going out to face the Muslim fleet so that the
marines would have more freedom of action, not intending to use the
benches above deck for rowing, why did they apparently leave the
oars for them on board where the Muslims could gain access to them?
Thirdly, Ambroise made no mention of this story.
Having cast doubt on the veracity of the testimony of the
Itinerarium peregrinorum that at least some galleys at Acre during the
Third Crusade had superimposed banks of oars, there is no other
evidence to suggest that Western galleys used any oarage system other
than the alla sensile by that time, irrespective of whether they were
monoremes or biremes. There is a possibility that they may have done
so in the eleventh and early twelfth centuries; however, the evidence
for this is extremely scanty.
The development of the alla sensile oarage system permitted the
replacement of the fully-seated stroke of the dromon, which had been
necessitated by the fact that dromons had a file of oars below deck, by
a stand-and-sit stroke which could be used above deck because of the
absence of height limitations. Because oarsmen using the stand-and-sit
stroke pulled the oar through the stroke by falling back onto the
bench, the whole weight of their bodies and the power of their legs
could be thrown into the stroke. In the fully-seated stroke, on the other
hand, oarsmen used mainly their arms and upper body to power the
stroke. Even though there were footrests for the oarsmen to brace their
feet against, because the benches and footrests were fixed in place,
their legs were constrained in the one position and less drive could be
------------------------------
15
Ambroise, L’estoire de la guerre sainte, ll. 3315-17 (col. 89): “As estoires iert la
huee, / Chascone iert sovent remuee / Sovent ensemble s’ajostouent”.
434 CHAPTER SIX
obtained from them.

The stand-and-sit stroke should have permitted more power to be
applied to the oars and consequently have resulted in an increase in
speed and in the endurance of oarsmen. However, there is no hard data
to confirm this and, indeed, some recent research suggests the
opposite. On the one hand, examining the oar mechanics and power of
classical galleys such as Olympias using fully-seated oarsmen, Shaw
calculated that a cruising speed of 7.5-8 knots could be maintained all
day by the oarsmen applying 135-160 watts of power to their oars. On
the other hand, similarly examining the oar mechanics and power of
Renaissance galleys using the a scaloccio oarage system,16 Bondioli
and his colleagues calculated that for these galleys to maintain six
knots in zero wind conditions the lead oarsman would need to apply
289 watts of power to the oar.17 A scaloccio rowing is considered to
have been inefficient, delivering only 25% efficiency of the actual
human power applied to effective propulsion power. The efficiency of
fully-seated oarsmen is considered to have been around twice that
figure.18 However, no similar research has been conducted for the
oarage system which lay chronologically between these other two: the
alla sensile oarage system. With respect to the question raised here,
namely, why was the dromon succeeded by the Western bireme galea,
there is an obvious need for similar modelling and power calculations
to be applied to their oarage systems.
Even if the power advantage of the stand-and-sit stroke over the
fully seated stroke can be questioned, there is no doubt that the alla
sensile system certainly did involve other changes that must have
produced an increase in power and speed. The following discussions
of these changes are couched in theoretical terms and we realize, of
course, that in practical terms there would have been trade-offs and
compromises and that the various factors had to be harmonized in
concert with others to produce the best overall results.19
------------------------------
16
A scaloccio: the oarage system which succeeded the alla sensile system in the
Renaissance. A stand-and-sit stroke was still used but instead of each oarsman rowing
a single oar, multiple oarsmen rowed on one larger oar.
17
Shaw, “Oar mechanics”, p. 169; Bondioli, et al., “Oar mechanics”, Table 12/2
(p. 201).
18
Communication from John Coates to John Pryor. See also Coates, “Naval
architecture”, p. 5.
19
The following discussion of the alla sensile oarage system, especially as it
pertains to the galleys of Charles I of Anjou, supersedes all previous discussions of it
by John Pryor, especially Pryor, “From dromo2n to galea”, pp. 112-14; idem, “Galleys
of Charles I of Anjou”, pp. 69-73. It should be stressed that with the exception of a
few fundamental dimensions, the width of the deck, the depth in hold, and the length
The gearing ratio of oars may be expressed as the ratio A:B, where A
is the length from the mid point of the oarsman’s hands on the handle
to the thole, and B is that from the thole to the centre of water pressure
on the blade.20 The mechanical advantage of an oarsman then becomes
the ratio C:B where C is the overall length of the oar from the mid
point of the hands on the handle to the centre of water pressure on the
blade. Therefore the greater A can be made with respect to B, the
higher the mechanical advantage of the oarsman becomes and the
more effective power he can deliver to the oar. The addition of an
outrigger to the hull makes one or both of two things possible. Either
A can be increased with respect to B and therefore the mechanical
advantage improved, or B can also be increased proportionately so
that the entire oar becomes longer and heavier and can deliver more
power without requiring any increase in the effort required of the
oarsman as long as the balance and weight in hand of the oar remains
the same. A compromise combination of both can also be achieved.
Thus the development of galleys with outriggers must have produced
an increase in effective power over what the oarsmen of galleys
without outriggers could deliver.21
Both outriggers and a stand-and sit stroke would also increase
speed for other reasons. If the maximum length of a seated oarsman’s
stroke is D, an oar of length E with a thole at the hull will move the
hull the distance F for each stroke. [Figure 55 (a)] A longer oar of
length G will move the hull a greater distance H. The further the
centre of water pressure on the blade of the oar is from the thole, the
greater the distance the hull will be moved by a stroke of the oar. This
is the first consideration. It is desirable to make the distance between
the centre of pressure on the blade and the thole as great as possible,
balancing that against the increase in the weight of the oar and the
difficulty for an oarsman to manage it the longer the oar becomes.
Consider the stand-and-sit stroke. As shown above, the
approximate length of the interscalmium of a dromon, and of the
length of an oarsman’s stroke, must have been around a metre.
However, that of the oarsmen of Angevin galleys of the late thirteenth
century using a stand-and-sit stroke is calculated to have been
------------------------------
of oars, the oarage system of thirteenth-century alla sensile bireme galleys is totally
unknown and that what follows is the product of comparison to classical evidence, the
experimentation with Olympias, and deduction.
20
See above p. 290 and n. 421.
21
These superior features of the galea over the dromon were pointed out by
Dotson in “Galley design”, p. 22.
436 CHAPTER SIX
approximately 1.20 metres.22 If the length of the stroke is increased

------------------------------
22
The dimensions of the galleys of Charles of Anjou here are based on a document
dated 20 January 1275 which referred to galleys to be constructed “according to the
plan and size and gallipum (model) of a certain “red galley of our court which came
from Provence”. This has been collated with another, dated 10 November 1278, which
referred to galleys with the same specifications, presumably also based on the “red
galley” of Provence. See Filangieri, Registri, Reg. 63, no. 486 (vol. 12, pp. 126-9),
Reg. 89, no. 88 (vol. 21, pp. 264-6). Here the text concerns the hull only. The full text
may be found in Pryor, “Galleys of Charles I of Anjou”, pp. 38-44. There is another
document, dated 15 December 1283, which for some reason did not find its way into
the registers reconstructed by Filangieri and the other archivists, which refers to
galleys constructed “according to the model of a galley of the court”. These were
marginally larger than the red galley of Provence and had oars which were 27 palmi
long rather than 26. See Minieri Riccio, Saggio, vol. 1, no. 204 (pp. 207-8).
“In primis dicta galea rubea est longitudinis de palma in palmam cannarum XVIII
et palmorum VI; et in carina est longitudinis cannarum XIII et palmorum III. Item
puppis eiusdem galee armat in altitudine palmorum XIV minus tertio; et in prora
armat in altitudine palmorum XI et tertium. In medio armat per altitudinem de tabula
in tabulam palmorum VIII minus quarto. … Altitudo [Latitudo] de cinta in cintam in
medio corporis eiusdem galee est palmorum XIV; et in dalfino amplitudo de cinta in
cintam in prora est palmorum VI et tertii, et in puppi palmorum VI. Item a prima
tabula fundi superius usque ad cintam in medietate ipsius sunt palmi V. Item habet
predicta galea a banco puppis usque ad iugum prore lactas LV, numerato iugo,
quarum quelibet est longitudinis palmorum XVII et medii.* … Item palleria in medio
ipsius galee est altitudinis palmorum III [tertiorum]. Altitudo de tabula fundi usque ad
tabulam cohoperte est palmorum VIII minus quarti. Item amplitudo sive planum fundi
in medio corporis ipsius galee ab uno genu ad aliud ad cordam tesam est palmorum XI
et quarti. Item distantia de una cursia est amplitudinis palmorum II et medii et
altitudinis palmorum I et medii. … Galea ipsa navigat cum remis CVIII, quodlibet de
palmis XXVI; verumtamen, debeant fieri aliqui de palmis XXX in puppi et prora.”.
*Missing from the text as published in Filangieri, Registri.
Translation: “Firstly the said red galley is 18 canne and six palmii [39.55 m.] long
from the extremity of the stempost to that of the sternpost; and it is 13 canne and 3
palmi [28.21 m.] long on the keel. Item, the poop of the same galley rises in height by
14 palmi minus a third [3.60 m.]; and at the prow it rises in height by 11 palmi and a
third [2.99 m.]. Amidships it rises in height from the planks of the floor to those of the
deck by 8 palmi minus a quarter [2.04 m.]. … The beam from wale to wale in the
midships of the same galley is 14 palmi [3.69 m.]; and at the dalfinum* the beam from
wale to wale at the prow is 6 palmi and a third [1.67 m.] and at the poop 6 palmi [1.58
m.]. Item, from the first plank of the floor up to the wale in the midships of the same
[galley] there are 5 palmi [1.32 m.]. Item, from the [aftermost] bench of the poop to
the yoke** of the prow the aforesaid galley has 55 deck beams, including the yoke,
each of which is of a length of 17 palmi and a half [4.61 m.]. … Item, the ceiling in
the midships of the same galley is one third of a palmus [0.09 m.]. The height from
the planks of the floor to the planks of the deck is 8 palmi minus a third [2.02 m.].
Item, the beam or plane of the floor in the midships of the same galley is 11 palmi and
a quarter [2.97 m.] from one genu*** to the other [measured] with a taught line. Item,
the dimension of a centre gangway is 2 palmi and a half in beam and 1 palmi and a
half in height. The same galley sails with 108 oars, each of 26 palmi [6.86 m.];
however, there should be some at the poop and the prow of 30 palmi [7.91 m.].”.
*Dalfinum (or delfinum): probably the frame section at the extremities of the rowing
platform. **Iugum prore: The transverse beam which carried the forward ends of the
aposticii, the outboard beams defining the rowing platform. ***Genu: the internal
turn of the frames at the maximum beam of the floor.
from D by 20% to I, then the hull will move not F for each stroke but
the greater distance J. There should be an appropriate increase in
Figure 55
The alla sensile bireme oarage system.
© John H. Pryor
------------------------------
Calculation of the interscalmium at approximately 1.20 metres is based on analysis
of the frames, deck beams, and overall dimensions. It is not specified in the document.
See Pryor, “Galleys of Charles I of Anjou”, pp. 64-8. However, it agrees with the
lengths of interscalmia of Renaissance trireme alla sensile galleys, 1.2-1.25 metres.
438 CHAPTER SIX
speed. [Figure 55 (b)]

If an outrigger is added at a distance K from the gunwale of the hull
and the oar handle and centre of pressure on the blade is moved an
equivalent distance outboard, even with the distance between the thole
and centre of pressure on the blade being the same, B, and the length
of the stroke being also the same, I, each stroke moves the hull at the
gunwale through the water not the distance J but rather the greater
distance L, with a corresponding increase in speed. [Figure 55 (c)]
Compare the alla sensile system in which both oars were rowed
from above deck to the superimposed banks system. An oar with a
gearing ratio A:B at any particular angle to the water will have a
distance B from the centre of water pressure on the blade of the oar to
the thole. [Figure 55 (d)] It is desirable not to decrease B because each
stroke would then move the hull a shorter distance. Therefore, for an
oar of any given length M, the greater the angle of the oar to the water
the higher must become the gearing A1:B and the lower the
mechanical advantage C1:B. This means that the lower a galley rides
in the water, and therefore the lower the angle of the oars to the water,
the higher will be the mechanical advantage of the oarsmen and the
greater the effective power that they can apply to their oars.
Finally, the stand-and sit stroke must also have produced one other
huge advantage. Because the buttocks of the oarsmen were not fixed
in one place but rather could move, the result would have been similar
to that of having had a moveable seat. The interscalmia of Angevin
galleys of approximately 1.2 metres must have evolved for ergonomic
reasons, just as those of around 1.0 metres for fully-seated-oarsmen
galleys had in the past. In fact, if a seated man rises to his feet and
keeps one foot anchored under and behind his seat and then stretches
the other forward to the maximum extent of his arms, that foot will in
fact move approximately 1.20 metres. A man of average height rising
off his seat to stand for the pull would have his fists around the oar
handle at about 1.25 metres above deck and the distance from his
shoulder-blades to his grip on the oar at the end of his stretched-out
arms would have been about 0.75 metres. With interscalmia of 1.20
metres and thwarts of approximately 20 centimetres, as suggested for
Olympias Mark II, when he thrust his inboard foot forward to just
under the bench in front, his grip on the oar would be approximately
1.5 metres aft of the centre of his own bench. Then, when he fell back
onto it at the end of the pull and closed his arms and the oar into his
chest, his shoulder-blades would have been some 40 centimetres
forward of his thole and the grip on the oar would have been pulled
through a distance of some 1.6 metres. [Figure 55 (e)] The resulting

increase in stroke length over what could be achieved with a fixed-
bench, fully seated stroke would have been extremely significant.
The galleys built according to the model of the red galley of
Provence had oars which were 26 Neapolitan palmi, 6.86 metres, long,
except for some at the poop and prow of 30 palmi, 7.91 metres,
suggesting that the rowing platform, telaro,23 narrowed somewhat
towards the stern and the bow. Nothing in the Angevin documents
suggests that oars used from an inboard position and those used from
an outboard position were different in length and we must therefore
conclude that they were all the same length. They may have had
different gearings since from the fourteenth century, there is no doubt
that oarsmen pulling individual oars alla sensile did use oars of
different lengths and gearings which actually necessitated rowing in
different ways.24 However, since the Angevin contracts did specify
that the oars at bow and stern were different but not that the inboard
and outboard oars were, we have assumed that they had the same
gearing.
The rowing benches would have been canted or angled outboard
towards the bow. For Olympias Mark II, it is proposed to cant the
benches at 18.4˚ to the centre line, the angle whose tangent is closest
to one third, and on later medieval galleys they were also canted.25
Canting the benches outboard on an alla sensile galley would in fact
work only if the tangent of the angle of cant was indeed one third.
Any higher or lower angle would not work, in theory. Of course,
within the parameters of the technology of medieval shipbuilding,
with plastic materials, inexactitude of measurements, and the variable
ergonomics of human movement, any figure between around 16˚ and
20˚ would no doubt have been possible. There may have been perhaps
even more latitude than that since oars would have moved quite
significantly inboard and outboard of the thole during the stroke.
The actual positioning of the benches could be varied somewhat
and in actuality may well not have corresponded exactly to that
reconstructed here. However, it must have been at least similar.
Assuming an angle of cant of 18.4˚, and that when the outboard
------------------------------
23
This word was not used in the Angevin documents and there was no term for the
rowing platform as a whole which was. However, the vernacular telaro was certainly
used elsewhere for it.
24
Alertz, “Naval architecture”, esp. p. 149; Bondioli, et al., “Oar mechanics”, esp.
p. 179.
25
Morrison, et al., Athenian trireme, p. 271; Alertz, “Naval architecture”, pp. 159,
162; Bondioli, et al., “Oar mechanics”, pp. 173, 176, 182-9.
440 CHAPTER SIX
oarsmen rose to their feet and put their outboard legs forward, the mid
point of their outboard feet should end up no closer to the edge of the
deck than 15 centimetres, that the length of their pace from the mid-
point of the bench to the mid-point of their outboard foot was 1.2
metres, and calculating that their shoulders and benches were some 50
centimetres across, the centre point of their benches should have been
some 0.77 metres inboard of the edge of the deck.
At the edge of the deck, the height of the galley above the waterline
amidships would have been 0.55 metres. The deck itself is calculated
to have had a camber of 0.27 metres from the centre-line to the edge
and the oarsmen’s feet would have been approximately 15 centimetres
above the edge of the deck. The oars were 6.86 metres long, would
have had handles of 0.4 metres and blades 0.8 metres by 0.15 metres,
the loom and blade being submerged by around one metre during the
pull. The centre of pressure on the blades would have been 0.2 metres
from the tip and if they had a gearing of 1:3.25, which is at the high
end of the range, and an angle to the water of around 20˚, which must
have been the case, an outboard oarsman using a stand-and-sit stroke
would have been capable of moving the oar through an arc of 66˚. The
tholes would have been some 0.88 metres outboard of the edge of the
deck. [See Figure 56] At the end of the stroke the thole would have
moved approximately 4.8 metres and the mid-point of the oarsman’s
hands approximately 6.4 metres.26
With inboard and outboard oarsmen both using a stand-and-sit
stroke from above deck, and both using oars with the same length and
gearing, it is in fact impossible for them to have both used the same
bench, even if it was canted. Because of the width of the human torso
and the need for some clearance between the oarsmen, the inboard
oars would have begun the stroke forward of the outboard oars but
have finished it aft of them, which would have been impossible. Even
lowering the gearing of the inboard oars massively, which would have
thrown out completely the synchronicity of the strokes, would not
have obviated this. In fact the only way that the system could have
worked with oars of the same length and gearing is if they occupied
------------------------------
26
Discounting slippage, one of the practical considerations left aside in this
theoretical discussion, which would reduce this figure somewhat. In practice all oar
blades move somewhat in the water because the weight applied to them by oarsmen
does force some water to move. Slippage increases when water is disturbed, which is
a compelling reason to have an oarage system in which the blades of different files of
oars do not intermesh with each other in the water but rather are in water clear of each
other.
Figure 56
Galleys of Charles I of Anjou, ca 1269-1284, and the bireme alla sensile
oarage system.
© John H. Pryor
442 CHAPTER SIX
separate benches and the inboard oarsmen were only slightly inboard
of the outboard ones. It is significant that the telaro of the Angevin
galleys, which rowed 54 oars on each side, 27 in each file, stretched a
distance of 55 deck beams from the aftermost bench to the yoke of the
prow. Since the tholes were set to correspond to alternate deck beams,
there was a half an interscalmium in excess and that is to be explained
by alternate staggering of the oarsmen’s benches. It also explains why
the Peter of Eboli illustration shows the oars of the two files staggered
and inboard oarsmen visible between the outboard ones. [See Figure
54] This was not artistic licence but rather a careful representation of
reality.
The conditions governing the positioning of the inboard oars would
have been, first, that the gearing would have been the same as that of
the outer oars so that the stroke could be synchronised. Secondly, the
oars should have been as far outboard as possible for maximum
power. If possible, thirdly, the blades should have rowed in clear
water and not intermeshed with those of the outer oars. Positioning the
inboard oarsmen approximately 95 centimetres inboard of the edge of
the deck would have allowed all conditions to be met. The blades
would have been clear of each other in the water and the stroke of
both inboard and outboard oars would be virtually identical. The
columbaria, oarports, of the inboard oars in the fabric of the telaro
would have needed to be elongated to allow the oars to work against
the thole thongs and then through them, but that was almost certainly
the case in all periods, as is shown in some depictions occasionally.27
Marino Sanudo Torsello wrote between 1310 and 1320 that: “It
should be known that in the year of the Lord 1290, two oarsmen used
to row on a bench on almost all galleys which sailed the sea. Later
more perceptive men realized that three oarsmen could row on each of
the aforesaid [benches]. Almost everyones uses this nowadays”.28
Here Sanudo referred to the trireme alla sensile oarage system which
replaced the bireme one of the twelfth and thirteenth centuries because
------------------------------
27
Most notably the Victory of Samothrace monument in the Louvre. See
Morrison, Greek and Roman oared warships, fig. 20 (pp. 219-21). More normally
oarports were depicted in a merely stylized way as circular openings. But see also
Odysseus’ ship on an Attic red-figure wine jar in the British Museum in Morrison, et
al., Athenian trireme, fig. 47 (p. 168)
28
Sanudo Torsello, Secreta fidelium crucis, II.iv.5 (p. 57): “Sciendum quod in
M.CCXC. anno Domini, quasi in omnibus galeis quae transfretabant per mare, duo in
banco remiges remigabant: postmodum perspicaciores homines, cognoverunt quod
tres possent remigare remiges super quolibet praedictorum, quasi omnes ad praesens
hoc utuntur.”.
he clearly stated that three oarsmen shared the same bench. The stand-
and-sit stroke was still used but the oars had different lengths and
gearings. It was this that made it possible for three oarsmen to share
the same bench, all rowing from tholes set in the apostis and none
through oar ports in the fabric of the telaro.29
Because dromons had a bank of oars below deck, their decks must
have been higher above water than the decks of alla sensile galleys
needed to be. Amidships, the freeboard of the hulls of Angevin galleys
at the deck was only in the order of 0.55 metres and the oars had an
angle to the water of around 20˚.30 As shown above,31 the decks of
dromons amidships must have been a minimum of around 0.95-1.0
metres above the plane waterline and the minimum angle of the upper
oars to the waterline can not have been less than around 28˚. The alla
sensile system must have produced an increase in mechanical
advantage and power efficiency over what dromons were capable of.
The alla sensile system almost certainly delivered other advantages
as well. Having all oarsmen above deck to double as marines in battle
may have outweighed the advantage of having half of them protected
from missile attack below deck. Having all of them above deck in
fresh air rather than having half of them working in fetid conditions
below deck must have contributed to an increase in endurance. The
hold would also have been freed up for armaments and spare gear, and
especially for provisions and water, thus undoubtedly increasing
crusing range.
Discussion of the advantages of the alla sensile system has been
couched in theoretical terms. In practical terms, all advantages would
have had their disadvantages. Increases in the length and weight of
oars would have increased the difficulty for oarsmen to manage them,
unless the gearing, balance, and weight in hand was maintained
somehow. Moving the hull a greater distance for each stroke by either
using a longer stand-and-sit stroke or by the use of an outrigger would
have demanded greater effort from the oarsmen unless the mechanical
advantage was also improved. No doubt, the lengths and weights of
oars, their balance, their gearing, their mechanical advantage, and their
------------------------------
29
This second transition from the bireme alla sensile oarage system of the
eleventh-thirteenth centuries, using oars of the same length pulled from alternating
benches through columbaria and from tholes on the apostis, to the trireme alla sensile
system of the fourteenth-fifteenth centuries, using oars of different lengths all pulled
from the same bench against tholes on the apostis, has never been researched. It
remains a desideratum but is beyond the scope of this book.
30
See Pryor, “From dromo2n to galea”, p. 114.
31
See above pp. 288-304, esp. Figure 32.
444 CHAPTER SIX
design, would all have been elements of formulae which would have
been experimented with over time to eventually produce the best
possible compromise. Whatever the case, it was almost certainly the
development of the alla sensile oarage system which led to the demise
of the system of superimposed banks of oarsmen on more than one
level which had dominated naval warfare in the Mediterranean for
1500 years and which had culminated in the Byzantine dromon.
The technological advantages of the new bireme galea may well
have given the West a technological edge over the Muslim and
Byzantine worlds in the crucial period ca. 1075 to 1150. When the
new Western design with its superior features began to be emulated in
the Byzantine and Muslim worlds is unclear; however, the Byzantine
evidence at least suggests that emulations of the galea as katergon
replaced the dromon in the course of the twelfth century. Such
technological factors have not been considered by those who have
addressed the issue of the rise of Western sea power in the twelfth and
thirteenth centuries and the apparent decline of the navies and
merchant marines of the Byzantine and Muslim worlds.32
------------------------------
32
See, for example, Abu Lughod, Before European hegemony; Lewis, “Balkan
peninsula”; Lewis and Runyan, Naval and maritime history, chs 3-4; Rose, “Islam
versus Christendom”; Santamaria-Arandez, “Reconquista”; Tu™ma, “Puzzle of a
decline and a rise”. I also failed to consider it in Geography, technology, and war
[JHP].
CONCLUSION
We began this study with two questions to which we had been led in
the course of our research into the war galley known to the Byzantines
as the dromo2n. The first of these was whether there ever actually
existed a particular and distinctive ship type to which this term
corresponded, or whether, rather, it was applied to a series of ship
types evolving over the centuries? If so, was it possible for us to
ascertain from the sources how these ship types actually did evolve?
The second of these questions was whether or not Byzantine authors
ever really intended to refer specifically to ship types known to them
in their own ages by the various terms which they used, such as
dromo2n, chelandion, trie2re2s, etc.? A corollary to this second question
concerned the degree to which Byzantine authors either consciously or
subconsciously used classical terminology in reference to the events
and technology of their own ages. Consideration of these questions led
to the examination of the relationships between the use of terminology
in texts and the physical reality of Byzantine war galleys which was
the central enquiry of this book. Ultimately the objective remained
what it had been at the outset; namely, to recover for the use of
historians as much as is possible of the physical reality of the galleys
referred to as dromo2nes over a period of some seven centuries from
ca. 500 to ca. 1200. Beyond that we have attempted to explore the
relationships between the technology available to the Byzantines, the
physical world in which their naval forces had to operate, and the
objectives to which they aspired.
We have been able to demonstrate conclusively that the
terminology of Byzantine texts is a maritime historian’s minefield.
They can rarely be accepted at face value and their testimony must
always be weighed against other evidence. Unless there is
corroborating evidence elsewhere, it must always be regarded with
suspicion. Theophane2s the Confessor and Anastasius Bibliothecarius
used classical terminology parenthetically to dromo2n/dromon without
intending anything technical by their use of it. The Anonymous’s
Naumachika syntachthenta para Basileiou patrikiou kai
parakoimoumenou is replete with classical anachronisms and even the
Naumachika of Leo VI and the Peri thalassomachias of Nike2phoros
Ouranos show traces of the same thing. The chronicles are equally
suspect. The various tactical manuals are also replete with impractical
446 CONCLUSION
advice based on arm-chair sailoring. Even the fleet inventories for the
Cretan and Italian expeditions contained in the De cerimoniis
attributed to Constantine VII are suspiciously prone to a charge of
bureaucratisation. Apart from that, they are certainly incomplete and
maritime historians must use them with care.
What is actually known about the galleys called dromo2nes remains
frustratingly little. Unless new sources are discovered, or unless
maritime archaeology comes to the rescue, we will almost certainly
continue to see this most famous of early medieval warships through a
glass darkly. However, we believe that we have been able to establish
at least a few certainties, some probabilities, and a wide range of
possibilities.
The early use of the term dromo2n was almost certainly philological
rather than technical in import. The word began to be used with
reference to some Romano-Byzantine war galleys, especially
liburnae, at the latest by the late fifth or early sixth centuries, surely
because they had unusual speed or manœuvrability of some kind. This
was most probably a consequence of changes in construction
characteristics already under way at the time.
The evidence for such changes in the construction characteristics of
war galleys from the sixth century is conclusive: the lateen sail
replacing the square sail, the spur replacing the ram, a bank of oars
rowed from below a full deck, and changes in hull design, especially
at the bow. It is also probable that some degree of change from shell
to skeletal hull construction was involved, but that ought not to have
affected performance. The chronological coincidence between the
evidence for these changes and the appearance in the texts of the term
dromo2n is so striking, given the fact that no other new word for a type
of ship appeared at the same time, that it is reasonable to draw the
conclusion that the changes eventually differentiated these galleys
from earlier Roman warships and that the word became applied to the
now differentiated galley type.
Adoption of the word dromo2n into Arabic and Latin in various
forms between the sixth and eighth centuries makes it clear that the
new galley type was common to the Romano-Byzantine world from
the beginning and that it was adopted immediately by the Muslims
when they took to the sea.
There never was a single dromo2n. The term was applied to galleys
which evolved over the centuries from what they had been when the
term was first applied to them in the late-fifth and early-sixth
centuries, to what they became in their heyday in the fleets of the tenth
CONCLUSION 447
century under the Macedonian emperors. Between the sixth and tenth
centuries, the primary reference of the term changed from being to
monoreme galleys of 50 oars to being to bireme galleys of 100-108
oars. The evolution of the ships between the sixth and tenth centuries
remains almost a complete unknown. Thereafter, however, the term
remained applied to galleys with two superimposed banks of oars.
Dromons became obsolescent as battle galleys from the twelfth
century almost certainly as a result of the progressive development of
the bireme galea in the Latin West and then its emulation in the
Byzantine and Muslim worlds. When the new galea of the West
became adopted into Byzantine fleets in the twelfth century, use of the
terms dromo2n and chelandion gradually became anachronistic and
was discontinued and replaced by katergon and taretes.
Ironically, in this self-same last period of life of dromons and
chelandia, the words themselves became widely emulated in Latin,
Old French, Norse, and Arabic literature but increasingly with
reference to large transport ships. By the mid twelfth century
Byzantines themselves appear to have been using the word dromo2n
with reference to transport ships.
The evidence for the construction of dromons must be regarded
with much more circumspection than has hitherto been the case. The
anonymous treatise commissioned by Basil the parakoimo2menos has
little credibility as a guide to the real construction of dromons and
chelandia in the tenth century because it has been shown to be
primarily an exercise in antiquarian philology. We have identified a
number of sources which the Anonymous used but there may well be
yet others which have escaped us. He himself is not to be condemned
for this since he was merely doing what any moderately well educated
Byzantine writer of the period would have done to impress a patron.
Rather, it is those modern scholars, and maritime historians in
particular, who have assumed that he was actually describing the
construction of a tenth-century dromon who have been at fault. Only
where what he says can be corroborated from other sources or is in
accordance with the common characteristics of ships of all kinds can
he be relied upon. Beyond that, he can at least be tentatively relied
upon where he appears to have supplied unique information; for
example, in the use of katapate2ton for a gunwale, bordo2nes for some
part of the poop, kathormeis for the yard crutches, and manikellia for
the leather oar sleeves.
In the cases of other authors who used the term dromo2n, and indeed
chelandion and other terms also, we can rarely see beyond their use of
448 CONCLUSION
the terms. Even a close scrutiny of the treatises of Leo VI and

Nike2phoros Ouranos reveals very little about the construction
characteristics of the dromons to which they referred, barely enough
to enable us to distinguish the actual galleys of their age from their
classical predecessors or medieval successors. Yet these galleys did
exist and undoubtedly were different. The problem remains that of the
relationship between the terminology of the sources and the tangible
objects to which they refer.
We can be confident that standard dromons had two banks of oars
with 25 benches in each file, that one was rowed from below deck and
one from above, that they had lateen rather than square sails, and spurs
rather than rams. Spurs were not built as integral parts of the hull but
rather were separate beams held to the stemposts by couplings,
probably iron chains. Dromons certainly had two masts, the larger
being the foremast, which had a blockmast, and the smaller the
midships mast, which had a “beaked” masthead. Both, together with
their yards, could be lowered onto crutches before battle or when not
under sail. The yards were held to the masts by parrels. The shields of
the marines could be ranged along a pavesade on each bulwark and
some dromons at least had castles on either side aft of the foremast.
There was a berth for the commander or dignitaries at the poop,
sheltered by a round tent, some kind of stern ornament, and two
quarter rudders attached to a through beam in housings outboard and
controlled by tackles. The poop was probably stepped up from the
deck. The prow had a fortified foredeck above the main sipho2n and
catheads for the anchors. We can also be confident of some details of
construction: a keel, frames composed of floor timbers and futtocks, a
full deck, a bung-hole, thwarts on either side for the oarsmen, both
above and below deck, oars held to tholes by oar-grommets, oarports
in the hull, the oarports of the lower bank sealed by leather oar
sleeves, a flared upper hull and upper and lower oars of different
lengths, all oars pulled by a single oarsman, at least three external
wales, and a pavesade topped by a gunwale above deck. The main
armaments were bow-ballistae, cranes, and sipho2nes for Greek fire, in
addition to crossbows and normal bows, javelins, swords and other
hand-held weapons. However, beyond this, virtually nothing can be
certified. Even the estimated dimensions of 31.25 metres in length and
4.46 metres in beam can only be deduced from the need for maximum
performance and the seated oarage system.
As well as with reference to standard biremes, the word dromo2n
was also used with reference to monoreme galeai and there were
CONCLUSION 449
certainly some dromons which were larger than the standard biremes.
However, there is no evidence that will withstand scrutiny that
dromons were ever triremes or had had three masts.
Dromons had evolved in the way that they had to deliver optimum
performance in battle. That was their purpose. They were designed to
give maximum short-term speed and manœuvrability in calm
conditions. Even if from astern, winds of more than Beaufort Scale
Four-Five, moderate to fresh breezes, would force them to run for
land. Their sails could almost certainly be used only when the wind
was astern and against the wind under oars the lower oars could not
have been used in more than light breezes and the ships would quickly
have come to a standstill. Average speed under oars in all conditions
was probably around four knots and average speeds for extended
voyages no more than around two knots, calculated round the clock.
Water requirements for the crews would have been high and stowage
space available for it low. We conclude that oarsmen carried their own
water supplies and that it would have been sufficient for no more than
3-4 days or so under oars, giving dromons a range of around 330
kilometres under oars. Stowage of water for horses would have
exacerbated all problems. Taking on the large quantities of water
necessary for men and animals from wells and small streams would
have been extremely time consuming and finding anchorages or
beaches large enough to accommodate fleets as large as those of the
Cretan expeditions would not have been easy.
Chelandia are even more problematic. There can be no doubt that
they were developed originally as an adaptation of dromons for the
purpose of transporting horses by sea and landing them onto beaches
in the face of opposition. However, if dromons had evolved to be
battle galleys with maximum performance capabilities, as they surely
had done, then chelandia can not possibly have been exactly the same
in design. They must have had more depth in hold and width in the
beam. However, the confusion of terminology in the sources and the
use of the word chelandia for battle galleys by some of them,
including Latin and Arabic texts, suggests clearly that even if they
originated as horse transports, chelandia did not remain confined to
that role alone. They were emulated in the Latin West and in the
Muslim world, in both roles, as early as the ninth century.
Byzantines could certainly transport horses by sea for short
distances but for long distances there is only the evidence of
Belisarios’s expedition to Vandal Africa and how many horses were
transported the whole distance and in what sort of ships is not known.
450 CONCLUSION
The evidence for the Vandal and Gothic wars suggests that at that time
the Byzantines did not transport horses on galleys equipped with ports
in the hull and landing bridges. The earliest evidence for that is the
inference of the ninth-century Life of St Antony the Younger.
However, the Byzantines certainly did have such horse transports by
the time of the 911 expedition to Crete. The horses would have been
stabled below deck in fore-and-aft stalls with slings to prevent them
being thrown about and there is no evidence to suggest that chelandia
could carry any more than 12 horses each. By the ninth and tenth
centuries, transporting horses in any numbers for more than short
distances was avoided, as the evidence of the Cretan expeditions
makes clear. Stowing away water supplies for the horses would have
been a major problem, as would have been ventilating the holds to
avoid the illnesses which are caused in horses by lack of oxygen and
build-up of carbon dioxide and impurities.
Crews included a standard ship’s complement or ousia of 108
oarsmen arranged in two oar banks, elasiai, of 50 oars, one above and
one below deck, and these oarsmen, certainly those above deck, were
armed and could also fight as marines. There must have been forced
ventilation of the holds by wind sails and cowls for the lower
oarsmen. Total crews varied from around 120 to around 160 when
“captains” (kentarchoi), helmsmen (pro2tokaraboi), bow-hands,
trumpeter, sipho2n operators, and marines were taken into account.
Supernumerary crews or marines could also be taken aboard, but only
if their weight was compensated for by stripping the ships of water,
provisions, armaments, or spare gear. Strate2goi were normally in
command of fleets, except for the droungarios tou ploimou in
command of the imperial fleet at Constantinople. “Vice-admirals” and
“rear admirals”, tourmarchai and droungarioi, served under strate2goi,
and there were also squadron commanders, kome2tes.
On the one hand, there is little in the naval record to suggest that
Byzantine crews were ever markedly superior in skills to those of their
opponents. On the other hand, there is a tantalyzing smattering of
evidence to suggest that at times there was considerable disaffection
amongst the crews of various fleets.
The disappearance of the waterline ram and its replacement by the
spur led to a complete revolution in battle tactics because Greek Fire
was never a ship-killing weapon and no tactical system was ever built
around it. It required both calm conditions and following winds to be
effective. Engagements became characterized by extensive exchanges
of missiles intended to degrade enemy crews. Maintenance of
CONCLUSION 451
formations during initial stages of engagements became of paramount

concern and we have argued that the Byzantines developed a
sophisticated system of signalling flags to control battle formations as
well as for other purposes. The crescent moon formation became the
battle formation par excellence, designed to prevent the enemy
breaking the line and attacking the poorly defended sterns of the
galleys. When missile exchanges had been exhausted or enemy crews
had been degraded sufficiently, the formations closed and grappled,
using iron grappling rods to prevent defeated enemy ships escaping.
Only then did the marines come into play, boarding and taking enemy
ships.
In fact battle was to be avoided if at all possible and strategic
objectives were to be attained by other means. To a degree this was a
product of the difficulty of obtaining, and the ineffectiveness of,
victory in battle at sea. Maritime space could not be controlled and
naval forces could be quickly rebuilt. Sea power was always only an
adjunct to land power. Fleets and their “admirals” remained secondary
to armies and their “generals” throughout the history of the Empire
and this was reflected in relative success in seizing the throne.
Leo VI stressed the need for secrecy when it came to the
technology and stratagems of naval warfare; however, preparations for
large-scale naval expeditions would almost certainly have been
impossible to keep hidden from enemy spies, who were ubiquitous.
Espionage must have played a major role in negating the effectiveness
of naval expeditions.
Operationally, the epoch separates quite neatly into five periods,
beginning with an inital one in which Romano-Byzantine hegemony
at sea was challenged by the Goths and Vandals, who were both
overcome in the mid sixth century. In the 150 years or so of this first
period there really only two major fleet encounters, the defeat of
Flavius Basiliskos off Cape Bon in 468 and the victory of John over
the Goths off Senogallia in 551. This initial period was followed by
one marked by peace at first and then an assault by the Umayyads
lasting a hundred years until it was fought to a standstill by the
Byzantines ca 750. In this period the great fleet engagement was the
disastrous Byzantine defeat at the Battle of the Masts off Phoinikous
in 655 but the decisive turning points were the defeat of the Muslim
assaults on Constantinople in 672-8 and 717-18, in both of which
Greek Fire proved to be a decisive weapon. The period ca 750-875
was one of chaos on all sides with both Muslims and Byzantines
having successes at various times. It was also marked by the advent of
452 CONCLUSION
new powers, especially the Aghlabids in Ifrı3qiya, the Umayyads in al-

Andalus, and the Carolingians in the West to some degree in the ninth
century. The two major strategic changes were the loss of Crete to
Andalusi corsairs and the loss of Sicily to the Aghlabids, who
controlled most of the island by 875. However, the Empire responded
brilliantly in the fourth period to the death of Basil II in 1025. Crete
and Cyprus were both recovered, Rho2s attacks on Constantinople were
beaten off, and the Muslims were expelled from the Italian mainland.
In spite of important defeats, such as that of the Kibyrrhaio2tai by the
Muslims of Tarsos in 898 and of Himerios by Leo of Tripoli off Chios
in 912, the Byzantines also had major victories, especially those of
Nasar in the Ionian in 880 and Basil Hexamilite2s off Cilicia in 956.
However, ironically it was the self-same imperial sucess which then
led to neglect of naval forces in the eleventh century in the final
period. By the accession of Alexios I Komne2nos in 1081, the Empire
had virtually no naval forces. This last period saw the triumph of the
Latin West in spite of the efforts of the first three Komne2noi emperors
to rebuild Byzantine fleets and in spite of the efforts at sea of the
Almoravids and Almohads in the Maghrib. Subsequently, the
appalling neglect of the entire apparatus of state by the Angeloi
emperors between 1185 and 1203-4 resulted in the almost complete
disappearance of imperial naval forces and those few that were left in
1203-4 were overwhelmed by the Venetians. In this last period the
Byzantines had no naval victories at all except for a possible defeat of
the Sicilians off Cape Malea in 1149 and that of the Venetians in
1171-2, a victory gained by strategy and disease rather than by action.
It was the development of the bireme galea in the Latin West
which precipitated the demise of the dromon. Since the Latin galea
first appeared in Norman Italy, it is difficult to believe anything other
than that it was adapted in some way or ways from the earlier
Byzantine monoreme galea, but exactly how remains unknown. The
Latin ship was obviously superior to all battle galleys existing at the
time since it proliferated very rapidly across the West, yet we know
very little about it until the late 1260s when surviving Angevin
chancery registers contain specification details. Until then we are
reliant on the interpretation of a few pictures, especially the depiction
of a galea in the Berne manuscript of the De rebus Siculis carmen of
Peter of Eboli. No specifications for them survive from the maritime
republics before the fourteenth century. The critical design change
was the development of the alla sensile oarage system in which two
oars were rowed from the same bench position above deck using the
CONCLUSION 453
stand-and-sit stroke and this was made possible by the introduction of

an outrigger, the telaro. The stand-and-sit stroke lengthened the stroke
by around 50% and must have produced a great increase in power;
however, the ergonomics of alla sensile rowing by comparison to
fully-seated rowing remains to be investigated scientifically. Alla
sensile rowing of both oars from above deck must also have delivered
other significant advantages, particularly in terms of fresh air
producing an increase in oarsmen’s endurance and of freeing up the
hold for provisions, water, armaments, and spare gear. Cruising ranges
must have increased dramatically. These technological advantages
remain to be drawn into historical discussion of the rise to maritime
predominance of the Latin West in the twelfth and thirteenth
centuries. They may well, for example, have been critical in enabling
the Latin West to project its sea power into Levantine waters and by
doing so to have facilitated the establishment and development of the
Crusader states in the twelfth century and their defence in the
thirteenth.
Postscript
At the very eleventh last hour, and fifty nine minutes, when this book
was in the final stages of production, there came the news that at long
last some wrecks of Byzantine-period ships had been discovered in
Istanbul. In communications received from Professors Cemal Pulak
and Felipe Castro of the Institute of Nautical Archaeology at Texas
A&M University, the news has been relayed that early in 2005, during
excavations for a new metro extension and underground railway
system in Istanbul, seven wrecks of ships, dated to the late tenth or
early eleventh centuries on the grounds of the amphorae found with
one of them, have been found in the Theodosian harbour or harbour of
Eleutherios on the south, Sea of Marmara, coast of the city. Even
more recently a wreck probably to be dated to the sixth-eighth
centuries has been found. More ships are quite likely to be discovered
in the future since only a fraction of the site has been excavated.
Preliminary excavations are being conducted by Professor Pulak
under the directon of the Istanbul Archaeological Museums. He has
been excavating, recording, and recovering a small merchantman
around 11-12 metres in length, of which the keel and a considerable
amount of the hull amidships is preserved. Unfortunately neither the
bow nor the stern remains. There are six other wrecks. Two of these
454 CONCLUSION
have their keelsons preserved and one small boat has its mast step still
in place. Judging from its position in the boat, this boat was sprit-
rigged. Two of the ships are lying on their sides and may well have a
considerable amount of hull planking above the turn of the bilge
preserved.
Most excitingly, two of the ships appear to have been “ long ships”,
one them being around 20 metres long. They appear to have been
carefully constructed and to have had rather light timbers by
comparison to the merchantman, suggesting that they were special
craft for some purpose, perhaps for navigating the Bosporos. They
appear to have been too short to have been been war galleys such as
dromons or chelandia and since no decorative carving has been found
on them, they were probably not ceremonial barges either. A quantity
of hay is reported to have been found on one of them. If this proves to
have been a Byzantine horse transport, it will be a sensational
discovery.
(Information courtesy of Cemal Pulak)
John H. Pryor, 16 February, 2005.

APPENDIX ONE
SYRIANOS MAGISTROS, NAUMACIAI SURIANOU MAGISTROU ,

EDITION AND TRANSLATION1
Technical terms, the understanding and translation of which are

discussed elsewhere in the text, are asterisked the first time they are
used. They may be accessed through the Index.
------------------------------
1
Edited from a microfilm of folios 333r-338v of the tenth-century manuscript
Milan, Biblioteca Ambrosiana, MS. B 119-sup. [gr. 139], referred to by Dain and
hereafter here also as MS. A. See pp. 179-81 above. A text was published in Dain,
Naumachica, pp. 43-55; however, Dain’s photographs of the Ambrosiana manuscript
were destroyed in World War Two and he was compelled to rely on his notes made in
1931. His transcription was relatively speaking quite accurate; however, it did contain
some errors which have been corrected here tacitly.
456 APPENDIX ONE
Naumacivai Surianou' Magivstrou

dV
12 ... tetagmevnoi, teleutai'oi de; oiJ prw/ra'tai: ejpibaivnousi de;
toujnantivon.
2 Crhvsimon de; tou'to oJphnivka ajllotriva/ gh'/ prospelavsai
boulovmeqa: oiJ ga;r tauvth" oijkhvtore" eijdovte" povrrwqen to;n
stovlon katavgonta pollavki" eij" ajllhvlou" sunevrcontai,
ejpitrevcousi de; kata; tw'n hJmetevrwn ejxiovntwn ploivwn kai;
tw'n me;n w|de, tw'n d ajllacou' wJ" e[tuce feromevnwn ejpiqumiva/
th'" tw'n polemivwn periousiva", oujk eijdovte" ta; par ejkeivnoi"
telouvmena.
3 Dio; dh; ajnagkai'on aujtou;" ajpobavnta" tw'n ploivwn kaqavper
ejn favlaggi suntetavcqai, e[st a]n diav tinwn shmeivwn
katamhnuvswsin aujtoi'" [oi|"] oiJ skopoi; ei[te to; ajmevrimnon
kai; ajnuvpopton dia; th'" hJsuciva", ei[te th;n eij" mavchn
eJtoimasivan kai; to; plh'qo" aujtw'n dia; th'" fwnh'" th'"
savlpiggo" h[ tino" tw'n a[llwn shmeivwn.
4 Anagkai'on de; kai; ta[lla tou;" ejrevta" paideuvein o{sa
poiei'n uJpo; tw'n kubernhtw'n diakeleuvontai, kai; prov ge
touvtwn to; nhvcesqai, ouj movnon fainomevnou", ajll e[stin o{te
kai; duvnonta": ajnh;r ga;r duvth" pote; ejpi; makro;n kata; bavqo"
dianhxavmeno" kai; ta; peivsmata tw'n ajgkurw'n diatemw;n tw'n
Persw'n ta;" nau'" ejpiovntwn tw'n ajnevmwn sunevtriyen, a[llo"
de; uJpo; polemiva" nho;" diwkovmeno" kai; nu'n me;n w|de
kataduovmeno", nu'n de; ajllacou' makrovq en ajnafainovmeno",
ta;" tw'n polemivwn ejx evkline cei'ra".
eV ”Oti crh; to;n strathgo;n e[cein meq eJautou' pavntote tou;"
pepeiramevnou" tw'n kata; qavl attan kai; ta; parakeivmena touvtoi"
cwriva.
1 ”Oti me;n ou\n crh; pavntw" e[cein meq eJautou' to;n strathgo;n
tou;" eijdovta" ta; kata; qavlattan di h|" kai; pro;" h}n ajpago-
meqa, fanerovn: levgw dh; thvn te th'" qalavssh" pei'ran, o{pw"
katapneomevnh kumaivnetai kai; tou;" ajpogeivou" ajnevmou" kai;
------------------------------
2
Cf. Appendix Two [a], §8.
NAUMACIAI SURIANOU MAGISTROU 457
Naval battles of Syrianos Magistros

4
1 ... drawn up, and last the officers commanding the prow.
They stand opposite.
2 This is useful when we wish to put into a foreign land, for
the inhabitants there, realizing from a distance that the fleet
is approaching, often gather together and hasten to attack
our ships as they go out3 while [our ships] are carried
haphazardly some in one direction and some in another in
their eagerness for the possessions of the enemies, and
they [the enemies] do not know what is going on.
3 And so they must disembark from the ships and draw
themselves up as though in line abreast,4 until scouts
indicate to them by certain signals either that [their arrival]
is unexpected and unsuspected because of the quiet, or [the
enemy’s] readiness for battle and their numbers, through
trumpet calls or some other signs.
4 It is necessary also to train the oarsmen in all they have to
do under the instructions of the helmsmen and above all
how to swim, not only on the surface but also, on occasion,
under water. For a diver once swam under water for some
distance and cut the anchor cables and was able to destroy
the Persian ships when the wind changed; and another who
was pursued by an enemy ship was able to escape the
enemy’s clutches by diving first at one spot and then
reappearing in another a long way off.
5 That a commander must always have with him men with
knowledge of the sea and the adjacent districts.
1 It is indeed obvious that a strate2gos* should always have

with him men who know the characteristics of the sea
through which and towards which we are sailing; I mean,
experience of the sea, what waves it produces in a gale,
------------------------------
3
The verb ejxevrcomai meaned “to go out”, or “to march out” and here referred to
the ships. However, the sense seems to demand the meaning of the men disembarking
from the ships.
4
Paragraph 9.30 below makes it clear that what Syrianos meant by the
classicizing affectation “phalanx” had nothing to do with the ancient Macedonian
phalanx but rather was a line abreast or abeam, whether straight, crescent-shaped, or
convex. We have translated the Greek “favlagx” by “line abreast” or ‘line abeam”
throughout. On Syrianos’s use of favlagx, see Zuckerman, “Military compendium”, p.
212.
458 APPENDIX ONE
ajnevmou" kai; tou;" uJf avlou" livqou" kai; tou;" ajbaqei'" tovpou",
oJmoivw" de; kai; th;n parapleomevnhn gh'n kai; ta;" para-
keimevna" aujth'/ nhvsou", tou;" limevna", ta; ejx eJtevrou touvtwn
eij" e{tera diasthvmata, ta; cwriva, ta; u{data: polloi; ga;r
ajpeiriva/ th'" qalavssh" kai; tw'n tovpwn ajpwvlonto, kaqavper
kai; plei'stoi tw'n a[llwn.
2 Crh; de; ouj movnon ejkeivnh" th'" qalavssh" pei'ran e[cein

aujtou;" ajlla; kai; tw'n parakeimevnwn aujth'/ cwrivwn: polla-
ki" ga;r a[nemoi katapneuvsante" a[llhn ajllacou' tw'n nhw'n
dieskevdasan.
3 Oujkou'n crh; ouj movnon to;n strathgo;n e[cein aujtouv", ajlla; dh;
kai; tw'n nhw'n eJkavsthn e[cein tina; to;n tau'ta eijdovta, w{ste
ta; periv touvtwn eijdovta kalw'" to; sumfevron bouleuvesqai:
laivlapo" ga;r pollavki" katalabouvsh" ou[t tw'/ strathgw'/
ou[te ajllhvlai" ajkolouqei'n duvnantai.
4 Anagkai'on de; kai; to; duvo pavntw" tina;" tw'n ejretw'n ei\nai
kaq eJkavsthn nau'n tou;" dunamevnou" ajnaneou'n ta; diav tina
tuvchn ejpisumbaivnonta tai'" nausi; trhvmata te kai;
qrauvsmata, pavnta" te eijdevnai tou;" ejrevta" o{pw" kai; pro;
th'" tevcnh" aujtoi; ejmfravttein ta; trhvmata kata; qavl attan
duvnantai dia; tw'n proceivrwn iJmativwn h] strwmavtwn, ajlla;
mh; tou;" a[llou" kalei'n povrrwqen h] pro; kairou' th'"
swthriva" ajpoginwvskein.
ıV Peri; skopw'n.
1 Pollavki" ajgnoou'nte" tou' potev eijsin oiJ polevmioi
ajparavvskeuoi sunantw'men aujtoi'". Oujkou'n ajnagkai'on katav
te gh'n diercomevnou" kai; qavlattan proporeuvesqaiv tina"
tw'n hJmetevrwn kataskophvsonta" kai; ajpaggelou'nta" th;n
tw'n ejcqrw'n ejpifavneian kai; provteron me;n dia; shmeivwn
katamhnuvein aujthvn, e[peita de; kai; dia; stovmato" tavcion
ejpistrevfonta" kai; levgonta" kai; to;n tovpon ejn w|/ katei'don
ajutou;" kai; to; plh'qo" o{son touvtwn kaqevsthken.
25 Kai; kata; qavl attan me;n ta;" koufotevra" kai; tacutevra" tw'n
nhw'n ajpostevllein, dunatwtevrou" mavl ista kai; kar-
terikou;" h] ajndreiotevrou" tou;" ejrevta" ejcouvsa": ouj ga;r
polemei'n, ajlla; manqavnein kai; ajpaggevllein toi'" ajpostei-
lasin eij" creivan katevsthsan.
------------------------------
5
what are the off-shore winds, the hidden rocks, and the
places which have no depth, likewise the coast along
which one sails and the islands adjacent to it, the harbours,
the distances from each to the others, the area and the
[fresh] water. For many have perished through lack of
knowledge of the sea and the [surrounding] areas, as have
very many of the other [men].
2 They must have experience not only of that [area of the]
sea but also the adjacent districts for winds often get up
and scatter the ships in different directions.
3 And so not only should the strate2gos have such

[experienced] men but each of the ships should have
someone with this knowledge so that he can give good
advice on these [matters] when necessary; for frequently
when a squall springs up [ships] can follow neither the
strate2gos nor each other.
4 It is also necessary that there should always be in each ship
two oarsmen capable of repairing the holes and damage
that happen accidentally in ships, and that all the oarsmen
should know how they can block holes at sea with the
clothes or bedding that are at hand before a repair [is
made], but should not summon the others from a distance
or despair of safety prematurely.
6 Concerning scouts
16 Often when we are unaware of the position of the enemy
we encounter them without preparation. Thus it is
necessary, when proceeding both by land and sea, for some
of our men to set off in advance to reconnoitre and to
report the situation of the enemy, and to pass this
information back first through signals and then orally,
returning quickly and reporting the place in which they had
seen them and how large is their number.
2 At sea the lighter and faster of the ships should be sent,
manned with the more capable and strong or brave
oarsmen. Their duties are not to fight but to find out and
report back to those who sent them.
------------------------------
6
460 APPENDIX ONE
3 Epei; de; pollavki" oiJ ejcqroi; kata; pleura;n ajkrwthrivou h]

potamo;n h] limevna h] nh'son eJautou;" katakruvyante", ei\ta
ejkei'q en ejx iovnte" th;n proporeuomevnhn sunevl abon, crh;
tevttara" aujta;" ei\nai, duvo me;n ajpecouvsa" tou' panto;"
stovlou wJsei mivlia e}x kai; metaxu; touvtwn eJtevra" duvo, i{na aiJ
deuvterai uJpo; tw'n protevrwn diav tinwn shmeivwn, oi|" a]n
ajllhvlai" suntavxwsin, mhnuqei'sai th;n tw'n ejcqrw'n
ejpifavneian kai; au\tai ta; o{moia pro;" to;n stovlon poihvsasai
wJ" h[dh tw'n polemivwn parovntwn pro;" povl emon a{pante"
paraskeuasqhvsontai.
4 Kata; de; gh'n ajpostevllein tou;" mavlista koufotevrou" te
kai; tacutevrou" tw'n a[llwn. Dei' de; pro;" touvtoi" aujtou;"
ei\nai ojxuderkei'", eujhkovou", ejpithdeivou" pro;" te
kataskoph;n kai; ajpaggelivan tw'n oJraqevntwn h]
ajkousqevntwn, movna" ta;" macaivra" ejpiferomevnou", ei\nai de;
kai; aujtou;" oJmoivw" tevttara", duvo me;n proporeuomevnou"
kai; met ejkeivnou" eJtevrou" duvo, tosou'ton7 tw'n protevrwn
ajpevconta" o{son oJra'n aujtou;" h] ajkouvein touvtwn duvnantai,
ouj movnon dia; ta;" kampa;" tw'n cwrivwn, ajlla; kai; to; ejk
makrou' dei'n protrevcein aujtouv", w{ste tw'n shmeivwn
povrrwqen kata; diadoch;n ginomevnwn kai; tou' strathgou'
tau'ta manqavnonto" oijkonomei'n aujto;n duvnasqai to;
sumfevron.
zV Peri; shmeivwn oi|" skopoi; kevcrhntai.
1 Shmei'a de; kata; me;n qavlattan ta; leukovtera tw'n
uJfasmavtwn kinouvmena, mavlista de; kapno;" baqu;" eij" u{yo"
aijrovmeno": to; me;n ga;r ejn u{dasi faivnetai, to; de; ejn ajevri,
kai; to; me;n bracu; kai; cqamalo;n kai; dia; tou'to povrrwqen
dusqewvrhton, to; de; dia; to; mevgeqo" kai; to; u{yo" povrrwqen
ejxelevgcetai. Eij de; kai; kata; nwvtou to;n h{l ion e[cwsin,
dunato;n kai; dia; katovptrou h] kai; spavqh" sucna; kinou-
mevnh" didavxai povrrwqen to; zhtouvmenon.
2 Kata; de; gh'n shmei'ovn ejstin hJ tw'n salpivggwn polufwnotev-
ra: tauvth/ ga;r crhsovmeqa plhvqou" polemivwn ajnafanevnto":8
ojjlivgwn ga;r o[ntwn aujtw'n h] ajgnoouvntwn th;n hJmetevran
parousivan ouj crh; tauvtai" kecrh'sqai, i{na mh; kai; ma'llon
uJp aujtw'n ejlegcwvmeqa, ajll uJpostrevfonta" aujtivka ajpag-
------------------------------
7
tosou'ton, thus Dain: tosouvton, MS. A.
8
plhvqou" polemivwn ajnafanevnto", thus Dain following K. K. Müller, Eine
griechische Schrift über Seekrieg (Würzburg, 1882): plh'qou'" polemivwn ajnafanevntwn,
MS. A.
3 Since the enemy often conceal themselves behind a

headland or in a river or harbour or island and then emerge
to capture the ship sent in advance, there should be four of
these, two keeping about six miles ahead of the main fleet
and the other two in between so that the second group are
informed of the disposition of the enemy by the former
through certain signals which they will have arranged with
each other, and should have done the same with the fleet.
Thus, when the enemy comes up, everyone will be
prepared for war.
4 On land you should send out the lightest and most fast-
moving troops. In addition they should be sharp-eyed and
with good hearing and ready to observe and report what
has been seen and heard. They should be armed with
daggers only. Again there should be four of them, two sent
out in front and the other two behind them but keeping a
distance that enables them to see and hear them; this is not
only to allow for irregularities in their terrain but also for
the need to run ahead for some way so that signals can be
passed over a distance by relay and when the strate2gos has
the information, he can then make the appropriate
dispositions.
7 Concerning the signals which scouts use

1 At sea signals are made with very white fabric waved
around and especially with thick smoke rising up high. The
first is visible over water and the second in the air. The
first is brief and low and thus difficult to see from a
distance. The second can be distinguished from far off
because of its size and height. If they9 have the sun at their
back, it is possible to send the required information for a
distance by means of a mirror or a quickly moving sword.
2 On land a signal is given by a trumpet with many tones.
We will use these [signals] when a mass of the enemy
appears. But if there are few of them or they are unaware
of our presence, we should not use the trumpets to prevent
their noticing us and immediately turning back to make a
------------------------------
9
“They” must refer to the main fleet here.
462 APPENDIX ONE
gevllein. “Eti shmei'on ouj movnon hJ th'" savlpiggo" fwnhv,

ajlla; kai; hJ tauvth" hJsuciva, hJ me;n th;n tw'n ejcqrw'n
parousivan mhnuvousa, hJ de; to; ajmevrimnon kai; ajnevtoimon.
hV Peri; strathgikw'n shmeivwn.

110 Ta; de; strathgika; shmei'a dei' pavntw" para; pavntwn
gnwrivz esqai tiv touvtwn e{kaston bouvl etai, w{ste eJnov" tino"
touvtwn uJpodeicqevnto" kai; tou' stovlou tiv potev ejstin tw'n
shmeivwn gnwrivzonto" rJa/divw" aujto;n pravttein ta;
keleuovmena.
qV Pw'" dei' suntavttein ta;" nau'" polemei'n mevllonta".
1 Mevllonte" de; peri; naumaciva" didavxai ajnagkai'on eijpei'n
provteron peri; suntavxew" new'n h|/ dh;11 qalavttio" favl agx
ejstivn. Kai; ga;r w{sper ejn tai'" pezikai'" favlagxin dia; th'"
eujtaxiva" tou' strateuvmato" ma'llon h] tw'n a[llwn to; kravto"
hJmi'n perigivnetai, ou{tw kajn tai'" naumacivai": to; ga;r
a[takton eJtoimovteron eij" diavlusin.
2 Kai; provterovn ge rJhtevon o{ti tw'n polemikw'n new'n aiJ mevn
eijsin mevgistaiv te kai; poluavnqrwpoi kai; dia; tou'to
ajrgovterai tw'n a[llwn kai; ajsfalevsterai, aiJ de; mikraiv te
kai; kou'fai kai; ojligavnqrwpoi, aiJ de; mevsw" pw" e[cousin
pro;" eJkatevra" aujtw'n.
3 Crh; tai'" me;n megivstai" mavlista me;n kecrh'sqai ejn tai'"
kata; qavl attan mavcai", e[stin d o{te kai; kata; livmna"
pollavki", ouj mh;n kai; kata; potamouv": ouj ga;r rJa/divw" dia; to;
bavro" ajnafevresqai duvnantai kai; mavlista o{tan hJ gh' uJpo;
tw'n polemivwn despovzhtai. Tai'" de; mevsai" kai; bracutevrai"
oujde;n kwluvei kai; kata; potamou;" kecrh'sqai.
4 Mevllonte" de; naumacei'n ta;" me;n ijscurotevra" te kai;
poluanqrwpotevra" tw'n new'n tw'n a[llwn protavxomen12 kata;
pleura;n suntattomevna" ajllhvl ai". Afivstasqai de;
tosou'ton eJtevran th'" eJtevra" o{son mh; uJp ajllhvlwn to;n
ajgw'na kwluvesqai mhvte sumfuvresqai pro;" eJauta;" kai; to;n
kaqoplismo;n de; tw'n ejn aujtai'" stratiwtw'n ajsfalevsteron
tw'n a[llwn ei\nai: kai; ga;r kai; ejn tai'" pezikai'" favl agxi
tou;" prwtostatou'nta" ajsfalevsteron kaqoplivzomen a{te dh;
prwvtou" kai; eij" cei'ra" ajnadecomevnou" th;n mavchn. Ta;" d
a[lla" tw'n new'n katovpin ajkolouqei'n suntetagmevna" oJmoivw"
------------------------------
10
11
h|/ dh, thus Dain on the authority of A. M. Desrousseaux: h[dh, MS. A.
12
protavxomen, thus Dain: protavxwmen, MS. A.
report. Thus a signal is made not only with a trumpet-call

but by its absence, the one announcing the arrival of the
enemy, the other that they are unprepared and off their
guard.
8 Concerning signals of strate2goi.
1 What each of the signals of the strate2gos means should
always be understood by everyone, so that when one of
them is given, the fleet can recognize which it is and easily
carry out the order.
9 How to form up ships in preparation for war.

1 Before giving instructions on naval warfare, we should
discuss the formation of ships which makes up a naval line
abeam. For in naval warfare, as in infantry lines abreast,
the survival of our authority depends more on the good
discipline of the expedition than on anything else. A
disorderly [force] is more liable to collapse.
2 First it must be said that some warships are very large and
heavily crewed and because of this are slower than the
others and safer. Others are small, light and with few crew,
while others are between these extremes.
3 Use should be made of the largest [ships] especially in war

at sea and also on many occasions in harbours, but not in
rivers since because of their size they cannot be
manœuvred easily, particularly when the shore is
controlled by the enemy. But there is nothing to prevent
the use of the mid-sized and smaller [ships] in rivers.
4 When about to engage in naval warfare, we should arrange
the stronger and more heavily crewed of the ships
alongside each other and place them in front of the rest.
Each should be far enough away from the next so as to
prevent their obstructing each other during the conflict and
their colliding with each other, and the equipment of the
soldiers aboard should be better than that of the rest. For in
infantry lines abreast we arm those in the front ranks better
because they are the first to engage in battle and at close
quarters. The remaining ships should follow behind, drawn
464 APPENDIX ONE
tai'" prwtostatouvsai", ajll oujc wJ" e[tucen ejpiferomevna".

5 Fulavxomen de; th;n tavxin ouj movnon ejn aujtw'/ tw'/ ajgw'ni, ajlla;
ga;r kai; ejn tw'/ poreuvesqai pro; th'" tw'n ejcqrw'n parousiva"
eujkaivrw" tauvthn thrhvsomen, ejpei; kai; ejn tai'" pezomacivai"
tou'to poiou'men: to; d ai[tion i{na provteron ejqivzonta ta;
strateuvmata th;n ejn polevmoi" eujtaxivan ejn kairw'/ tauvthn
e[coien.
613 To;n dev ge panto;" tou' stovlou hJgouvmenon ajnagkai'on
proporeuovmenon tou' panto;" stovl ou tosou'ton o{son oJra'n to;
pa'n mh'ko" th'" favlaggo" duvnasqai, diorqou'n aujth;n eij" ei[ ti
kai; aJmartavnei, fevrein te meq eJautou' ejk tw'n tacutevrwn
new'n duvo ajpoferouvsa" ta; ejkeivnou prostavgmata, aujto;n de;
ajpevnanti tou' mevsou proporeuvesqai mikro;n ejf eJkavtera ta;
mevrh metaferovmenon, i{na mh; aujto;" kaq e{kaston mevro" th'"
favlaggo" diatrevcwn to; me;n eij" eujtaxivan a[gei dia; th'"
parousiva" aujtou', to; de; eij" ajtaxivan meqivstatai parel-
qovnto" aujtou'.
7 “Ergon ga;r touti; mevgiston kajn tai'" melevtai" dia; panto;"
prostavttein to;n strathgovn, w{ste paratetagmevnwn tw'n
nhw'n ta;" me;n prolambanouvsa" ajmeleiva/ tw'n hJgemovnwn
aujtw'n sunevcein ta;" a[lla" ejkdecomevna", ta;" de; bradu-
nouvsa" ejpelauvnein e[st a]n tw'n a[llwn nhw'n ejn i[sw/
gevnwntai.
814 Dei' de; ta; pro;" povlemon eujtrepisqevnta kalw'" pro; panto;"
a[llou kaq eJauto;n to;n strathgo;n skevyasqai kai; meta; tw'n
crhsimwtevrwn bouleuvsasqai eij dei' pavntw" polemh'sai h]
mhv. Anavgkh de; tw'/ mevllonti peri; polevmou bouleuvsasqai
eijdevnai kalw'" thvn te hJmetevran thvn te tw'n ejnantivwn duvna-
min, povsa te par hJmi'n ejstin ploi'a kai; povsa tw'n ejnantivwn:
ei\ta povsa megavl a kai; poluavnqrwpa kai; o{sa mikrav te kai;
ojligavnqrwpa, i{na mh; pollavki" ejx ajgnoiva" kata; pleiovnwn
ferovmenoi rJa/divw" uJp aujtw'n katapolemwvmeqa:
9 ei\ta to; stravteuma h] neovlekton h] pepeiramevnon polevmou:

ei\ta to;n kaqoplismo;n kai; th;n tou' laou' provqesin eij" to;n
prokeivmenon povl emon. Manqavnomen de; tau'ta e[k te tw'n
hJmetevrwn kataskovpwn kai; tw'n prosfuvgwn kai; oujc ejniv tini
------------------------------
13
14
Cf. Appendix Two [a], §§36, 74.
up like those in front but not haphazardly.

5 We should keep the formation not only during the conflict
itself but should preserve it satisfactorily also during the
manœuvring before the arrival of the enemy, since this is
what we also do in infantry warfare. The reason for this is
to enable the expedition which has become used to its
good battle formation to maintain it in a crisis.
6 The commander of the entire fleet should proceed in front
of the entire fleet as far as is necessary to enable him to
view the whole length of the line abeam. He should correct
any fault and have with him two of the faster ships to carry
his instructions. He himself should proceed in front of the
middle, veering a little to each side so that, by passing by
each side, he prevents the one remaining in good order
because of his presence and the other falling into disorder
because of his departure.
7 Even during exercizes, it is especially important for the

strate2gos always to give orders that, when the ships are in
formation, those that are ahead through the negligence of
their commanders should keep with the others that are
following, and that those that are lagging behind should
move forward until they are up with the other ships.
8 The strate2gos should above all else consider carefully
within himself the preparations for war, and amongst the
more useful points for him to deliberate is whether or not it
is entirely necessary to make war. A [strate2gos] who is
considering war should have good knowledge of both our
strength and that of the opposition, how many ships we
have and how many the opposition has, then how many are
large and heavily crewed and how many are small and
lightly crewed, to prevent our frequent and easy defeat by
[the enemy] when for lack of information we attack greater
numbers.
9 And then [he should know whether] the [enemy]
expedition is newly recruited or experienced in battle; and
next the equipment and the attitude of the men towards the
proposed war. We learn these things both from our own
466 APPENDIX ONE
levgonti pisteuvonte", ajlla; polloi'" sumfwnou'sin.
10 ÔEkatevran de; tw'n dunavmewn parabavllonte" thvn te

hJmetevran thvn te tw'n polemivwn, eij me;n uJperbavl wmen th'/
dunavmei tw'n ejnantivwn, polemw'men aujtou;" ouj katafro-
nou'nte" aujtw'n dia; th;n uJperbolh;n th'" dunavmew": polloi;
ga;r tw'/ plhvq ei qarrhvsante" uJpo; ejlattovnwn hJtthvqhsan.
11 Eij de; ijsavzei ajmfotevrwn hJ duvnami" katav te rJwvmhn swvmato"
kai; ajndreivan kai; kaqoplismo;n kai; ta\lla, eij me;n kaq
hJmw'n ouj proevrcontai oiJ polevmioi, i{na kai; hJmei'"
mevnwmen fulavttonte" eJautou;" kai; ta; i[dia, ajlla; mh;
polemw'men aujtouv": eij de; ejpevrcontai kaq hJmw'n h] th;n
hJmetevran lhivzontai, polemw'men aujtouv". 15
12 Eij de; pollw'/ plevon hJmw'n katiscuvousin oiJ polevmioi, mevga"
de; tai'" hJmetevrai" povlesin ejphvrthtai kivnduno", paraitou-
mevnou" to;n povl emon sofiva/ ma'llon h] dunavmei tw'n polemivwn
katagwnivsasqai, a[llav te polla; ejpiskopou'nta" kai; dh; kai;
kairo;n kai; crovnon kai; tovpon, di w|n pollavki" oiJ ceivrone"
tw'n kreittovnwn periegevnonto: crovnon me;n kaq o}n
prosbavllonte" toi'" ejcqroi'" tou;" ajnevmou" summavcou"
kekthvmeqa, w{sper wJ" ta; polla; givnetai ejpiv te tw'n ejthsivwn
kai; ajpogeivwn ajnevmwn: tovpou" de; th;n metaxu; duvo gaiw'n
qavlassan h] potamo;n kaq h}n to; plh'qo" tw'n polemivwn dia;
th;n th'" qalavssh" stenovthta a[crhston eij" povl emon
givnetai. Givnetai de; tou'to tricw'" h] metaxu; duvo nhvswn h]
metaxu; hjpeivrou te kai; nhvsou h] metaxu; duvo hjpeivrwn.
13 “Esti de; kai; a[llw" perigenevsqai tw'n pleiovnwn, ejpeida;n
eij" diavfora diaireqw'si susthvmata w{ste summacei'n
ajllhvloi" mh; duvnasqai. Givnetai de; tou'to o{tan ejk diafovrwn
tovpwn makra;n ajllhvl wn ajf esthkovnte" eij" ajllhvlou"
sunevrcwntai, h] o{tan ejx eJno;" tovpou eij" diafovrou"
ejpanastrevf wsin, h] kai; a[llw" kata; th;n ajllotrivan eij"
diavf ora susthvmata diairouvmenoi: tine;" de; touvtwn toi'" me;n
provteron, toi'" de; u{steron sumplakevnte" ajmfotevrwn
kathgwnivsanto.
14 Kai; tau'ta me;n ei[rhtai pleivsthn kaq hJmw'n duvnamin
ejcovntwn tw'n polemivwn kai; kinduvnwn ejpikeimevnwn kata;
tw'n hJmetevrwn pragmavtwn ejn tw'/ to;n povlemon hJma'" pa-
raitei'sqai. Eij de; mhdei;" hJmi'n e{tero" ejphvrthtai kivnduno"
------------------------------
15
eij de; ejp evrcontai ... podemw'men aujtouv" omitted by Dain.
scouts and from deserters, and we do not believe what one

man says but what many agree upon.
10 If, after comparing each of the forces, both ours and that of
the enemy, we are superior to the force of the opposition
let us make war on them, but without despising them
because of the superiority of our force. Many have trusted
in numbers and have been defeated by a smaller [force].
11 But if the forces on each side are equal in strength of body,
bravery, equipment and the rest, and if the enemy does not
come out against us so that we can continue protecting
ourselves and our property, we should not make war on
them. If they come out against us and ravage our territory,
we should make war on them.
12 If the enemy is overwhelmingly stronger than us and a
great danger hangs over our cities, then we should avoid
war and overcome the enemy by wisdom rather than
might, taking into account many other matters, especially
the weather, time and place, factors through which the
weaker often get the better of the stronger: time, by
attacking the enemy at a moment when we have the winds
as allies, as happens frequently with Etesian and off-shore
winds; place [by using] the sea between two pieces of land,
or a river, [areas] in which the numbers of the enemy are
useless because of the narrowness of the sea. This happens
in three ways: between two islands, or between the
mainland and an island, or between two mainlands.
13 There is also another way to overcome larger numbers
when they are divided into different groups and so cannot
support each other. This happens when they come together
again after being positioned in several widely spaced
places, or when they are sent to different places from being
in one, or when in other ways they are scattered in
unfamiliar territory in different groups. Some, having
engaged first one and then the other group, overcome both.
14 This is relevant when the enemy has a vastly superior force

and danger threatens our affairs while we are avoiding war.
If no other danger hangs over us when we are considering
war, there should be no war. It is better to enter enemy
468 APPENDIX ONE
paraitoumevnoi" to;n povl emon, ouj dei' polemei'n. “Ameinon de;

kata; th'" polemiva" ajnteisavgesqai, wJ" mhvte deilivan hJmw'n
tou;" polemivou" kataginwvskein paraitoumevnwn to;n
povlemon kai; hJma'" tou;" polemivou" ajntilupei'n ta; i[sa
duvnasqai: e[sti d o{te touvtou genomevnou ajf evnte" th;n
ajllotrivan oiJ polevmioi ejpi; th;n ijdivan ajnevstreyan.
1516 Th'" toivnun sugkrivsew" eJkatevra" dunavmew" ou{tw pw"
genomevnh" kai; to;n povl emon hJmw'n ejcomevnwn crh; to;n
strathgo;n pavnta" eij" eJauto;n sugkalevsanta protreptiko;n
eij" povl emon poihvsasqai lovgon, diabavllonta me;n tou;"
ejnantivou" meta; tou' piqanou', ejpainou'nta de; tou;" ijdivou".
16 Pro;" touvtoi" ejpifevrein wJ" ei[ ti" tw'n pavntwn ajpor-
ragei;" th'" ijdiva" tavx ew" pro;" leipotaxivan ejkklivnein,
aujto;" ta;" ejscavta" uJpomevnei timwriva": eij de; gunh; kai; oiJ
pai'de" kai; ei[ ti" a[llo" kata; to;n oi\kovn ejstin, tw'n ijdivwn
oi[kwn kai; th'" ijdiva" patrivdo" ajpelasqhvsetai kai;
katoikhvsousi gh'n pollw'n kakw'n gevmousan. Ei\ta meta;
tau'ta tiv" uJpe;r gunaikw'n kai; paivdwn kai; gonevwn ouj
prokinduneuvei kai; to;n qavnaton tou' zh'n ouj prokrivnei…
Tou'to ga;r kai; hJ a[logo" bouvletai fuvsi": pollavki" ga;r hJ
tekou'sa uJpe;r tw'n neossw'n eij" cei'ra" eJauth;n divdwsi
tou' ajgreuvonto": deino;n de; th;n me;n a[logon fuvsin tw'n
oijkeivwn uJperaspivz esqai hJma'" de; logikou;" o[nta"
katafronei'n tw'n ijdivwn.
17 Ei\ta pavl in ejpistrevfein pro;" th;n ajpeilh;n ejpifevronta kai;
ta;" tw'n ajpeiloumevnwn aijtiva", oi|on o{ti crh; dia; tou'to tou;"
leipotavkta" pollai'" provteron tai'" timwrivai" kaqupo-
bavllonta puri; ma'llon h] xivfei eij" e[scaton paradivdosqai,
prw'ton me;n o{ti Qeou' katefrovnhsan kai; th'" ijdiva" hjl ovgh-
san pivstew", ei\ta gunaikov", paivdwn, gonevwn, ajdelfw'n kai;
tw'n oJmopivstwn, kai tau'ta dunamevnou" perigenevsqai tw'n
ejnantivwn.
18 Dei' de; pro;" touvtoi" ejperwta'n to;n strathgo;n kai; to; plh'qo"
eij kai; aujtoi'" tau'ta dokoivh, sumfwnouvntwn de; aujtw'n tw'/
strathgw'/ kai; yhfizomevnwn qavnaton kata; tw'n leipotak-
touvntwn ejpi; to;n e[painon pavlin to;n strathgo;n strevf esqai,
tiv levgonta: ‹‹Egw; de; oi\mai wJ" oujk a[n ti" tw'n pavntwn
leipotakthvseien oJrw'n tw'n a[llwn to; provqumon, ajlla; kaiv,
ei[ ti" par uJmi'n ejstin pro;" leipotaxivan ejkklivnwn th;n
------------------------------
16
For §§15-18 cf. Appendix Two [a], §35.
14 territory as a result of provocation, so the enemy cannot

accuse us of cowardice when avoiding war and we can
inflict equal damage on the enemy. Sometimes when this
happens, the enemy abandon foreign territory and return to
their own.
15 When the comparison between each force has been made

in this sort of way and we are at war, the strate2gos must
summon everyone to him and make a speech exhorting the
men to war, belittling the enemy in a credible manner and
praising our own men.
16 In addition he should point out that if any of the men
should break rank and desert, he must expect the ultimate
vengeance. If his wife, children, or anyone else is found at
home, they will be driven from their homes and native
country to inhabit a land full of misery. Who after that
would not hazard all for wives, children, and parents and
prefer death to life? This is what irrational beasts choose,
for often the mother abandons herself to the hands of the
hunter to save her chicks. It would be strange for irrational
beasts to protect their offspring and for us rational beings
to have no concern for ours.
17 Then he should return to threats and add reasons for what

had been threatened; for example, that he should have
subjected the deserters to many punishments earlier and
then handed them over for the ultimate penalty by fire
rather than the sword, first because they had rejected God
and disregarded their own faith, and also their wife,
children, parents, brothers, and fellow believers, and this
when they were able to get the better of the opposition.
18 In addition the strate2gos should ask the mass if this is their
opinion, and when they agree with the strate2gos and vote
for death for those who desert, the strate2gos should revert
to praise, saying something like this: “I think that not one
man from the whole [force] will desert when he sees the
enthusiasm of everyone else, but should any of you have
any inclination towards desertion, he will be stirred up to
470 APPENDIX ONE
gnwvmhn, pro;" to;n o{moion tw'n a[llwn dianasthvsetai zh'l on››.

19 Eujch'/ de; teleutai'on to;n lovgon sfragivsanta tw'n nhw'n
ejpibh'nai protrevpein kai; eij" tavx in polevmou kaqivstasqai
kai; ou{tw" kata; ta; provteron eijrhmevna th;n o{lhn sunta-
xanta favlagga tou;" ejnantivou" ejpizhtei'n.
20 “Hdh de; tw'n polemivwn plhsiazovntwn crh; to;n strathgo;n

diatrevconta pro; tou' metwvpou th'" favlaggo" protrevpein e[ti
proqumotevran th;n mavchn poihvsasqai kai; eij" crhsta;"
ejlpivda" a[gein to; stravteuma, ei\ta kata; nwvtou panto;" tou'
stovlou genovmenon, a[n te kaq e{na zugovn, a[n te kata; duvo hJ
favlagx tw'n nhw'n h\/, ta;" bradunouvsa" sunelauvnein kai;
ajpokaqista'n ejpi; tou;" ijdivou" zugouv".
21 Ka]n me;n ojknhrotevrou" tou;" stratiwvta" oJra'/ pro;" th;n
mavchn, pro; tw'n a[llwn aujto;n mikro;n ejxiovnta toi'" ejnantivoi"
sumplevkesqai par eJkavtera ta;" iJscurotevra" kai;
poluanqrwpotevra" tw'n nhw'n meq eJautou' e[conta. Eij de;
proqumotevrou" tou;" stratiwvta" oJrwv/h pro;" th;n mavchn,
aujto;n me;n kata; to; mevson th'" favl aggo" ejpakolouqou'nta
katasalpivzein te kai; proqumotevrou" tou;" oijkeivou"
poiei'n, tou;" de; ojknhrotevrou" ajpeilei'n foberwvt eron, ejf
eJkatevrou de; tw'n a[krwn th'" favl aggo" eJpisth'saiv tina" tw'n
ajsfalestevrwn eij" to; sunevcein th;n favl agga.
22 “Estin de; o{te kai; oiJ kata; to; mevtwpon th'" favl aggo"
tetagmevnoi leipotaxiva" dovx an ejmfaivnousin, oJpovte ta;"
kwvpa" hJremouvsa" katevcousin h] ojknhrovteron aujta;"
e{lkousin, pote; de; oiJ katovpin ejpiferovmenoi: dio; dh; ejkei'
to;n strathgo;n spoudaivw" paragenovmenon h] ajnt aujtou'
e{teron tou;" me;n rJa/quvmou" diegeivrein, tou;" de; ajtavktou" eij"
tavxin a[gein.
2317 Eij de; mhde;n touvtwn oJra'tai ginovmenon, tina; de; mevrh
leipotaxiva" uJpovlhyin e[cousin, ejp ejkei'na to;n strathgo;n
to;n ojfqalmo;n e[cein kai; katamanqavnein kai; ajpostevl lein
pro;" aujtouv" tina" tw'n koufotevrwn nhw'n ejpapeilou'nta
qavnaton ejk tou' parautivka ei[ ti" tw'n a[llwn polemouvntwn
leipotakthvseien. ÔUpovlhyin de; leipotaxiva" e[cousin ejn
me;n ajllotriva/ oiJ kata; to; pevlago" tetagmevnoi, ejn de; th'/
hJmetevra/ oiJ kata; th;n h[peiron.
24 Crhsimwvtaton de; kai; to; proaforivzein tina;" kaq e}n h] kai;
------------------------------
17
emulate the zeal of others”.

19 Then, marking the end of his speech with a prayer, he
should hasten on board the ships and establish the battle
formation and thus, having set up the whole line abeam in
the way described previously, he should seek out the
opposition.
20 When the enemy is quite near, the strate2gos should move
rapidly before the front line of the line abeam and urge
them to prepare for war even more enthusiastically and
instil good hope into the expedition. Then he should go to
the back of the whole fleet, whether the line abeam of
ships is arranged in one rank or two, and stir up those that
are lagging behind and see that they are in their own ranks.
21 If he sees that the soldiers are rather reluctant for battle he
should go out a little in front of the others and engage with
the opposition, having with him on each side the strongest
and most heavily crewed of the ships. Should he see that
the soldiers are rather eager for battle, he should follow in
the middle of the line abeam and sound the trumpet and
make his own side even more enthusiastic while he
threatens the reluctant fighters more terribly. He should
position some of the most reliable on each wing of the line
abeam to hold the line abeam together.
22 Sometimes those stationed in the front of the line abeam
show a tendency to break rank when they hold their oars at
rest, or use them less vigorously, or when those behind
bear down on them. For this reason the strate2gos, or
someone else in his place, should be there energetically,
and should rouse up those who are slacking and bring the
disorderly back into formation.
23 If none of these things are seen to be happening, the
strate2gos should have his eye on those sections where he
suspects there might be desertion and notice [what is
happening] and send some of the lighter ships to them,
threatening immediate death if any of the others who are
fighting should desert. In foreign territory those positioned
on the open sea are more likely to desert while in our
territory it is those positioned by the coast.
24 It is very useful to position in advance at one or the other
472 APPENDIX ONE
kaq eJkavteron a[kron th'" o{lh" favl aggo", ai} mevsw" pw"
e[cousin prov" te ta;" meivz ona" tw'n polemivwn kai; ta;"
koufotevra" aujtw'n, wJ" mhvte uJpo; tw'n meivz ovnwn
katalambavnesqai tw'n ejcqrw'n feugouvsa", mhvte uJpo; tw'n
ceirovnwn katagwnivz esqai.
25 Protrevpein de; aujtav", ejpeida;n i[doien eij" cei'ra" hJkouvsa"
ta;" favlagga", uJperfalaggivsai te kai; kata; nwvtou
gevnesqai tw'n ejnantivwn: ajsqenevsteron ga;r e{xousin eu\
oi\d o{ti oiJ polevmioi pro;" th;n mavchn diairouvmenoi,
tw'n me;n kata; tw'n e[mprosqen ajgwnizwmevnwn, tw'n de; a[l lwn
a[llote kata; tw'n o[pisqen ejpistrefomevnwn, i{na mh; kata;
nwvtou touvtwn macevswntai.
26 ÔW" a]n de; oiJ polevmioi oJrw'nte" kata; provswpon tou;"
uJperfalaggivzonta" mh; sumparekteivnwsin kai; aujtoi; th;n
ijdivan favlagga kai; kwluvswsin aujtw'n th;n diavbasin, crh; ta;"
eijrhmevna" nau'" mh; kata; provswpon tw'n uJpenantivwn, ajlla;
kata; nwvtou fevresqai tw'n ijdivwn, e[st a]n ajmfotevrwn aiJ
favlagge" eij" cei'ra" ajllhvl wn h{xwsin: ejpeida;n de;
sumplakeivsa" ta;" favlagga" i[dwsin, tovte kai; aujtou;"
uJperfalaggivsanta" kata; nwvtou genevsqai tw'n ejnantivwn,
tosou'ton ejkeivnwn ajpevconta" o{son mh; katalambavnesqai
uJpo; tw'n ijscurotevrwn aujtw'n duvnasqai, plhsiavz ein de;
mavlista kai; kataqorubei'n ejkeivnou" oi} kata; tw'n hJmetevrwn
qermovteron uJpembaivnousin.
2718 Kalo;n de; ta;" toiauvta" nau'" kata; ta; a[kra proaforivz ein ouj
movnon tou' poih'sai, ajlla; kai; tou' mh; paqei'n e{neka: tw'n ga;r
polemivwn tou'to pravttein ejpeigomevnwn kai; aujtw'n
ajntexagomevnwn pro;" th;n ejkeivnwn ajpavnthsin, a{te dh; ejpi;
tou'to proaforisqeivsa" aujta;" to; ajmevrimnon toi'" ijdivoi"
poihvsousin. 19 To; de; toiou'ton givnetai o{tan nausi; tw'n
ejcqrw'n pleonavz wmen.
28 Tine;" de; th;n oJrmh;n tou' stovlou ojxutevran ei\naiv fasi
proqumiva"20 te tw'n oijkeivwn e{neka kai; deiliva"21 tw'n
ejnantivwn: tine;" de; ajsfalevsteron e[doxan hjrevma
kinoumevna" toi'" polemivoi" sumplevkesqai, oiJ de; kai; mh;
kinoumevna". Emoi; de; ajsfalevsteron ei\nai dokei' th;n tou'
------------------------------
18
19
poihvsousin, thus Dain, following Müller, Griechische Schrift: poihvswsin, MS.
A.
20
proqumiva", thus Dain, following Müller, Griechische Schrift: proqumiva, MS. A.
21
deiliva", thus Dain, following Müller, Griechische Schrift: deiliva , MS. A.
wing of the whole line abeam, some ships which are

between the larger and the lighter of the enemy ships, so
that they are not captured by the larger enemy ships if they
flee nor are overcome by the inferior ones.
25 These should be encouraged, when they see that the lines

abeam are coming to blows, to make an outflanking
movement and approach the rear of the opposition. For I
am well aware that the enemy will be in a weaker position
when divided in the battle, with some fighting those in
front and others turning sometimes towards those behind
so as not to be fighting at their backs.
26 So that the enemy also do not extend their own line abeam
and prevent the manœuvre when they see the outflanking
movement in front of them, the ships just mentioned
should not come to hands with their opponents but be
positioned at the rear of their own fleet until both lines
abeam engage with each other. When they see that the
lines abeam are engaged, then they should make the
outflanking movement and come behind the opposition,
keeping as far away from them as will prevent their
capture by the stronger [enemy ships], but still coming
close and shouting down those who are making the most
vigorous onslaughts on our men.
27 It is a good idea to position such ships on the wings
beforehand not only to enable them to act but also to
prevent damage, for while the enemy is being compelled to
act and is coming out to meet them, since they have been
positioned for this purpose, these will relieve the pressure
on our men. This manœuvre takes place when we have
more ships than the enemy.
28 Some say that the onslaught of the fleet is made more
effective through the enthusiasm of our men and the
cowardice of the opposition. Some say that the [ships]
engage with the enemy more safely when moving steadily,
others when they are not moving.22 It seems to me safest to
------------------------------
22
How one might engage the enemy while remaining stationary escapes us.
474 APPENDIX ONE
laou' oJrw'nta" diavqesin pro;" to;n povlemon oijkonomei'n to;

sumfevron:
29 oi|on eij me;n ojknhrotevrou" pro;" th;n mavchn oJrw'men tou;"
ijdivou", su;n boh'/ krativsth/ kai; qoruvbw'/ pollw'/ kai; tavcei
kata; tw'n ejcqrw'n oJmou' sunelauvnein aujtav": eij de;
proqumotevrou" oJrw'men aujtou;", hjrevma kinei'sqai th;n tavxin
fulavttonta": eij de; pro;" th;n polemikh;n suvntaxin duskovlw"
e[cousi, kaqovlou hjremei'n th;n tavxin fulavttonta" kai; tou;"
ejnantivou" ejkdecomevnou": ejn ga;r tw'/ kinei'sqai tauvthn
ajpovllusin. Epeida;n de; aujtou;" i[dwsin plhsiaivteron
genomevnou", tovte kai; aujtou;" kinhqevnta" ojxuvteron su;n
boh'/ pollh'/ toi'" ejnantivoi"23 sumplevkesqai.
3024 Ei|" me;n dh; trovpo" ou|to" paratavx ew" kaq o}n th;n favl agga
teivnonte" toi'" ejnantivoi" sumplekovmeqa. “Estin d o{te kai;
th;n ejuqei'an koilwvsante" favl agga mhnoeidh' tauvthn
poiou'men. Givnetai de; tou'to o{tan ijscurotevrou" oJrw'men
tou;" ejnantivou" kai; th;n tavxin fulavttonta", a[llw" te
paraitei'sqai to;n povl emon ouj dunavmeqa feidoi' tw'n ijdivwn.
Tovte ga;r dh; tovte kata; ta; a[kra tou' schvmato" ta;" mavl ista
ajsfalestevra" suntavx ante" ta;" me;n mevsa" met ejkeivna"
tavxomen kai; met aujta;" ta;" ajsqenestevra": ajnavgkh ga;r ta;"
polemiva" nh'a" th'" eijsovdou fulavttesqai tou' mh; paqei'n
e{neka eJkatevrwqen beblhmevnwn.
31 To; de; sch'ma mh; livan e[stw baquv, ajll e[l atton pavntw"
hJmikuklivou, i{na tw'n polemivwn sunercomevnwn kata; tw'n
a[krwn th'" favl aggo" kai; oiJ kata; to; bavqo" aujtivka fqavnein
duvnantai toi'" oijkeivoi" summachvsonte". To; de; toiou'ton th'"
favlaggo" sch'ma ouj dei' ejk makrou' a[gein, ajlla; plhsia-
zovntwn tw'n polemivwn, i{na mh; povrrwqen ijdovnte" to; sch'ma
th'" favlaggo" oiJ polevmioi kai; aujtoi; pro;" to; sumfevron
aujtoi'" th;n ijdivan diatupwvsousi favl agga,
32 ta;" me;n ijscurotevra" kata; ta; a[kra tavttonte", ta;" de;

ajsqenestevra" kata; to; mevson kai; h] diascisqevnte" kai; to;n
e[xw tovpon lambavnonte" h] kata; duvo zugou;" sunercovmenoi,
kai; to;n me;n cwrei'n kata; tou' bavqou" th'" hJmetevra"
protrevponte", to;n de; katovpin ajkolouqein, kai; tou;" me;n
------------------------------
23
toi'" ejnantivo i", thus Dain, following Müller, Griechische Schrift: tou;" ejnantivou",
MS. A.
24
consider the attitude of the men to the battle and then take
appropriate action.
29 For example, if we see that our men are reluctant for battle,
we should drive them against the enemy with loud shouts
and much noise and speed. If we see that they are quite
enthusiastic, they should advance steadily, keeping the
formation. If they are in a difficult position facing the
enemy formation, they should maintain the formation and
stay completely still, waiting for the opposition, for they
lose it by moving. When they see them coming close, then
they should start moving and engage the opposition with
loud shouts.
30 This is one method of drawing up a formation in which we
employ a line abeam and engage the opposition. There are
also times when we can hollow out the straight line abeam
and make it crescent-shaped. This can be done when we
see that the opposition is stronger and is maintaining his
formation, especially when we cannot force battle out of
consideration for our own men. Particularly then, we
should position the safest [ships] at the wings of the
arrangement and place the mid-sized next to them and after
them the weaker ones, for we must keep the enemy ships
from the entrance to prevent damage when they are
attacked from both sides.
31 The arrangement should not be too deep but certainly less
than a semi-circle so that when the enemy attacks the
wings of the line abeam those at the deepest point can
come to support their own side. This arrangement of line
abeam should not be set up too far ahead but only when
the enemy are approaching, to prevent their seeing the
arrangement of the line abeam from a distance and then
deploying their own line abeam in a way that suits
themselves,
32 by putting the stronger [ships] on the wings and the weaker
towards the middle and either dividing themselves and
taking the outside position or marshalling in two ranks and
making one advance to the deepest part of our line and the
other follow behind. The wings of the second rank then
476 APPENDIX ONE
a[krou" tou' deutevrou zugou' toi'" a[kroi" sumplevkesqai, ta;"

de; mevsa" tai'" prolabouvsai" ajkolouqei'n, i{na tauvta"
oJrw'nte" oiJ par eJkavtera th'" mhnoeidou'" favl aggo"
tetagmevnoi mh; sunevrcwntai kata; tw'n prolabovntwn, wJ" mh;
kata; nwvtou aujtw'n genevsqai tou;" kata; to;n deuvteron zugo;n
tetagmevnou".
33 Dio; dh; tou'ton to;n trovpon ajntiparatattomevnwn tw'n
ejnantivwn ouj crh; ta;" plhsiazouvsa" tai'" a[krai" kata; tw'n
prolabovntwn tw'n polemivwn sunevrcesqai, ajll ajnamevnein
tou;" ejpi; tou' deutevrou zugou' tw'n polemivwn, katalam-
banouvsa" de; h] sumplevkesqai aujtai'" h] th;n ei[sodon
paracwrouvsa" kata; nwvtou tw'n ejcqrw'n genevsqai. 25
34 Dei' de; pavntw" to;n ejnto;" feuvgein tovpon, i{na mh; uJpo; tw'n
ejkto;" sunwqouvmenoi kai; puknouvmenoi ouj movnon ejnergev-
stera ta; bevlh tw'n ejcqrw'n kaq hJmw'n gevnwntai, ajlla; kai;
uJf eJautw'n dia; th;n puvknwsin suntribwvmeqa.
3526 “Esti de; ouj movnon kata; th;n koivlhn ejpifavneian tou'
mhnoeidou'" schvmato" toi'" polemivo i" sumplevkesqai, ajlla;
kai; kata; th;n kurth;n aujtou' ejpifavneian ejpi; tou;" polemivou"
tauvthn strevyanta", oujk ejpi; to;n aujto;n tovpon eJkavsth" tw'n
nhw'n tetagmevnh", kaqavper ejpi; th'" mhnoeidou'" ejl evgomen
favlaggo", ajlla; ta;" me;n ijscurotevra" te kai; poluanqrwpo-
tera" tw'n nhw'n tetagmevna", kata; tou' mevsou, tw'n de; me;swn
met ejkeivna" kai; pro;" toi'" a[kroi" tw'n eujtelestevrwn, i{na
prohgoumevnw" me;n aiJ mevgistai toi'" ejnantivoi" sumplevkw-
ntai, aiJ d eujtelevsterai fulavttointo ejn ajpostavsei pro;"
toi'" a[kroi" ferovmenai.
36 Dei' de; kai; ejp aujtw'n tw'n a[krwn tavttein ajna; duvo tina;" tw'n
krativstwn eij" fulakh;n tw'n ajsqenestevrwn. Eij de;
pleonavzomen tw'/ ajriqmw'/ tw'n nhw'n pro;" ta;" tw'n polemivwn,
touvtwn to; plevon katovpin tavttein kata; tou' mevsou th'"
favlaggo", i{na prohgoumevnw" me;n tw'n megivstwn
sumplekomevnwn toi'" ejnantivoi" katovpin aujtw'n kai; au\tai
ferovmenai summacw'sin aujtai'" h] ejkeivnai" a}" ma'llon oJrw'si
katapolemoumevna" tw'n a[llwn.
37 Givnetai de; kai; tou'to to; sch'ma provteron me;n eij" eujq ei'an
------------------------------
25
genevsqai, thus Dain: givnesqai, MS. A.
26
For §§35-41 cf. Appendix Two [a], §40.
32 engage with the wings while those in the middle follow

those that have moved up already. So, on seeing these
[those moving up], those drawn up on each side of the
crescent-shaped line abeam do not attack those who have
come up there, so as not to have those positioned in the
second rank at their backs.27
33 So, when the opposition draw themselves up in this
defensive manner, the [ships] that are approaching the
wings should not attack the [ships] of the enemy that arrive
first at the wings but should wait for the second rank of the
enemy. When they have caught up they should either
engage them or allow them an entrance and come to the
rear of the enemy.
34 They should always avoid the inner position, otherwise
they will be crowded by those outside and hemmed in and
not only will the weapons of the enemies land on us more
violently but we will be crushed by our own in the
confusion.
35 It is possible to engage the enemy not only on the concave
side of a crescent-shaped arrangement but also on the
convex side, turning this against the enemy, though not
with each of the ships positioned in the same place that we
proposed in the crescent-shaped line abeam but with the
stronger and heavily-crewed ships drawn up in the middle,
the mid-sized [ships] next, and the inferior [ships] towards
the wings, so that the largest can start the engagement with
the opposition while the inferior ones keep their distance
on the wings.
36 You should also position on each of the wings two of the

strongest [ships] to protect the weaker. If we are superior
to the enemy in numbers of ships, you should position
more of these [the strongest] behind the middle of the line
abeam, so that when the largest start the engagement with
the opposition these are behind them and can be brought
up in support of whichever group they see is being hard
pressed by others.
37 This arrangement develops first of all when [its ships] are
------------------------------
27
What this paragraph is trying to say escapes us.
478 APPENDIX ONE
suntattomevnwn kai; mevson me;n tetagmevnwn tw'n megivstwn te

kai; poluanqrwpotevrwn, ei\ta met ejkeivna" tw'n mevswn kai;
meta; tauvta" tw'n ajsqenestevrwn: kai; ou{tw" tw'n me;n a[krwn
uJpokratoumevnwn th'" favlaggo", tw'n de; mevswn proagovntwn,
eJpomevnwn de; kai; tw'n a[llwn mecriv tw'n a[krwn th;n kata; to;
plavto" tavx in ouj parercomevnwn.
38 Dei' de; kata; tou'to to; sch'ma toi'" ejnantivoi"
sumplekomevnou" hJma'" ajpevcein kai; ma'llon ajllhvlwn
poihvsomen ta;" nau'", w{ste mei'zon to; sch'ma kata; mh'ko"
genevsqai, fulattomevnou" th;n metaxu; tou' schvmato"
puvknwsin uJpo; tw'n polemivwn sunwqoumevnou", to;n de;
strathgo;n ejnto;" periercovmenon th'" favlaggo" katasal-
pivzein kai; proqumotevrou" tou;" ijdivou" poiei'n kai; mavl ista
kaqo; mevro" oJra'/ th;n mavchn ajkmavzousan.
39 Dei' de; kai; tou'to to; sch'ma mh; povrrwqen a[gein, i{na mh;
metapoiei'n oiJ polevmioi ta;" nau'" duvnwntai pro;" to;
crhsimovteron aujtoi'" th'" mavch" katepeigouvsh". Crwvmeqa
de; kai; touvtw/ tw'/ schvmati o{tan diascivsai th;n favlagga tw'n
ejnantiwn dianowvmeqa kai; dialu'sai th;n tavx in aujtw'n.
40 Sumbavlletai de; hJmi'n tou'to mavlista oJpovtan oiJ ejnantivoi th'/
mhnoeidei' kevcrhntai favlaggi, w{ste dia; mevsou ejkeivnh"
cwrouvsh" th'" kurtoeidou'" favl aggo" ejn tavxei th;n mavchn
poiei'n ajntiva tw'n kata; th;n kurth;n ejpifavneian tetagmevnwn
nhw'n ajf eJkavsth" sumplekomevnh" tai'" ejnantivai".
41 Istevon de; o{ti tw'n polemivwn crwmevnwn th' mhnoeidei'28
favlaggi kai; hJmw'n th'/ ejnantiva/ oujkevti kata; ta; proeirhmevna
tavxomen ta;" me;n poluanqrwvpou" kata; to; mevson th'"
kurtoeidou'" favlaggo" kai; met ejkeivna" ta;" a[lla", ajlla;
ta;" me;n poluanqrwvpou" kata; tw'n poluanqrwvpwn kai; ta;"
ajsqenestevra" kata; tw'n asqenestevrwn kai; mevsa" oJmoivw"
kata; tw'n i[swn.
42 ÔRhtevon de; kai; peri; tw'n tovpwn th'" qalavssh" kaq ou}"
ojfeivlomen naumacei'n. Eij me;n ga;r th;n polemivan parap-
levonte" th;n naumacivan poiei'n mevllomen, kata; pevl ago"
tauvthn poiw'men paraitoumevnou" th;n provsgeion mavchn: eij
de; th;n hJmetevran, ouj povrrw makra;n ajf esthkovnte" th'" gh'",
i{na hJtthqevnte" kai; swv/zesqai dia; pelavgou" ouj
sugcwrouvmenoi ejpi; th;n gh'n katafeuvgwmen.
------------------------------
28
mhnoeidei', thus Dain, following Müller, Griechische Schrift: monoeidei', MS. A.
formed up in a straight line, with the largest [ships] with

the heaviest crews in the middle, then the mid-sized next to
them and the weaker next to those. Thus the wings of the
line abeam are held back, those in the middle press
forward with the others as far as those on the wings,
moving along the breadth of the line.
38 When we engage the opposition using this arrangement we
must keep our distance from each other, and we shall place
the ships so that the arrangement is greater in its length,
while we maintain close order within the arrangement as
we are pushed together by the enemy. And the strate2gos,
who is positioned in the line abeam, should sound the
trumpet and urge our men on, especially in any place
where he sees the battle is at its height.
39 Do not set up this arrangement too far ahead in case the
enemy is able to reposition his ships more conveniently
whilst the battle is getting under way. We use this
arrangement when we intend to split the line abeam of the
opposition and break up their formation.
40 We can achieve this whenever the opposition has used the
crescent-shaped line abeam and as a result the convex line
abeam can sail through its middle in formation and give
battle to those drawn up opposite in the convex line, each
ship engaging those in front.
41 You should know that when the enemy employs the
crescent-shaped line abeam and we use the opposite [for-
mation], we shall not form up the ships as described above,
with the heavily crewed ships in the middle of the convex
line abeam and the others next to them, but with the heav-
ily crewed ships against heavily crewed, weaker against
weaker, and the mid-sized ships against their equals.
42 Some comments must be made about the areas of the sea
in which we have to make naval war. If we are about to
make naval war whilst sailing past enemy territory, let us
set it up on the open sea, even though they wish to give
battle close to the shore, if we are sailing past our territory,
whilst at no great distance from land so that if we are
defeated and cannot escape towards the open sea, we can
take refuge on land.
480 APPENDIX ONE
43 Oujkou'n crh; dia; tau'ta kai; toi'" ejpi; th'" sterea'"

katamhnuvein to;n tovpon ejn w|/ polemei'n mevllomen, kai; ouj
tou'to movnon, ajlla; dh; kai; provteron th;n h[peiron parap-
levonte" kai; manqavnonte" pou' potev eijsin oiJ polemivoi
mhnuvein toi'" hjpeirwvtai", kajkeivnou" pavl in ta; o{moia
manqavnonta" katamhnuvein tw'/ stovl w/ to;n dunato;n trovpon:
pollavki" ga;r oiJ ejcqroi; to;n me;n stovlon lanqavnousin o{poi
eijsin, uJpo; de; tw'n hjpeirwtw'n ginwvskontai, h] toujnantivon.
4429 Tine;" de; kaqovl ou th;n provsgeion diabavllousi mavchn, tiv

levgonte": wJ" oiJ polloi; to;n povlemon dedoikovnte" ejpi th;n
gh'n katafeuvgousin: ejgw; de; oujk a]n tou'to oi\mai tolmh'saiv
tina", tou' strathgou' ta; proeirhmevna fulavttonto".
iV Pw'" dei' to;n strathgo;n meta; th;n mavchn peri; tou' stovlou
oijkonomei'n.
1 Tou' toivnun polevmou krathqevnto", eij me;n tw'n polemivwn
katiscuvsomen a[n te kaqovlou a[n te ejpi; mevrou", ouj crh; to;n
strathgo;n a{te dh; tou;" polemivou" nenikhkovta ajdeevsteron
diativq esqai, ajll ejkeivnh/ th'/ ajsfaleiva/ kecrh'sqai h{/tini kai;
pro; tou' polevmou ejkevcrhto.
2 Eij de; uJpo; tw'n ejcqrw'n nenikhvmeqa, mhd ou{tw" ajpogi-
nwvskein, ajll ejpisunavgein ta;" uJpoleifqeivsa" kai; kairo;n
deutevra" ejpizhtei'n mavch".
------------------------------
29
43 For this reason, therefore, we must indicate to those on

land the place where we intend to fight, though not only
this. When we have previously sailed past the mainland
and learned where the enemy are, we should also inform
those on the mainland and they should also, when they
have similar information, inform the fleet as far as is
possible. For the enemy frequently do not know where the
fleet is and find this out from those on the mainland, and
vice versa.
44 Some avoid battle close to the shore entirely, giving
reasons such as this, that most of the men fear war and
would flee to the land. I, however, do not think that any
man would dare to do this if the strate2gos takes the
precautions outlined above.
10 How the strate2gos should deal with the fleet after the battle.
1 If, when the war has taken place, we have got the better of
the enemy, either completely or in part, the strate2gos
should not be less vigilant because he has defeated the
enemy but should act as cautiously as he did before the
war.
2 If we have been defeated by the enemy, we should not
despair but collect up the surviving [ships] and seek an
opportunity for a second battle.
APPENDIX TWO [a]
LEO VI, NAUMACIKA LEONTOS BASILEWS,

AND
APPENDIX TWO [b]
LEO VI, EK TOU KUROU LEONTOS TOU BASILEWS,

EDITION AND TRANSLATION

------------------------------
1
Edited from a microfilm of folios 323r-331v of the tenth-century manuscript
Milan, Biblioteca Ambrosiana, MS. B 119-sup. [gr. 139], referred to by Dain and
hereafter here also as MS. A. In this manuscript, Constitution XIX of Leo VI’s
Taktika, Peri; qalassomaciva", was excerpted from the rest of the Taktika and
transferred to the beginning of the section on naval warfare in the manuscript under
the heading Naumacika; Levonto" Basilevw". See above pp. 180-81.
A text was published in Dain, Naumachica, pp. 19-33; however, Dain’s
photographs of the Ambrosiana manuscript were destroyed in World War Two and he
was compelled to rely on his notes made in 1931. It is clear that at some points he
confused the text of the Ambrosiana manuscript with those of Constitution XIX in
other manuscripts of the Taktika.
We have compared this text to the composite one published by Migne from the
edition of Joannes Meursius the elder in PG, 107. Although the wording is frequently
different, the PG edition adds nothing to the understanding of this text.
484 APPENDIX TWO
Naumacika; Levonto" Basilevw"

Peri; naumaciva" boulovmeqa diatavxasqai, peri; h|" oujde;n me;n ejn
toi'" palaioi'" taktikoi'" kekanonismevnon eu{romen: ajf w|n de;
sporavdhn ajnevgnwmen kai; di ojl ivgh" peivra" tou' nu'n kairou' para;
tw'n plwi?mwn hJmw'n strathgw'n ajnemavqomen, ta; me;n pepoihkovtwn,
ta; de; paqovntwn, ajnalexavmenoi mikrav tina kai; o{son ajformh;n
dou'nai toi'" ejpi; qalavssh" mavcesqai mevllousin dia; tw'n pote
legomevnwn trihrw'n, nu'n de; dromwvnwn kaloumevnwn, ejn ojlivgoi"
diorisovmeqa.
bV2 Prw'ton me;n ou\n, w\ strathge; th'" nautikh'" dunavmew", dei' ei\naiv
se ejpisthvmona th'" naumacikh'" ejmpeiriva" kai; tavx ew", kai;
proskopei'n, kai; proginwvskein ta;" tw'n ajevrwn kai; tw'n
pneumavtwn kinhvsei" dia; th'" tw'n fainomevnwn ajstevrwn kai; ejn
a[stroi" shmeivwn peivra" kai; tw'n kata; to;n h{lion te kai; th;n
selhvnhn ginomevnwn ejpishmeiwvsewn: ejpiginwvskein de; kai; th'"
tw'n kairw'n ejnallagh'" th;n ajkrivbeian, wJ" a]n e[cwn peri; tau'ta
ejmpeivrw" diafulavtth/ ajsfalh;" kai; ajkivnduno" ajpo; tw'n th'"
qalavssh" ceimwvnwn.
gV Kataskeuasqh'nai de; dei' kai; drovmwna" ajrkou'nta" pro;"
naumacivan kata; tw'n ajntistrateuomevnwn plwi?mwn polemivwn kai;
pro;" th;n ejkeivnwn katavstasin kai; tw'n sw'n poihvsasqai th;n
kataskeuh;n dunath;n pro;" a{panta ejkeivnoi" ajntimavcesqai.
dV ÔH de; tw'n dromwvnwn kataskeuh; mhvte pavnu e[stw pacei'a, i{na mh;
ajrgoi; gevnwntai ejn tai'" ejlasivai", mhvte livan eij" leptovthta
ejxeirgasmevnh, i{na mh; ajsqenh;" ou\sa kai; saqra; tacevw" uJpo; tw'n
kumavtwn kai; th'" tw;n ejnantivwn sugkrouvsew" dialuvhtai: ajlla;
suvmmetron ejcevtw th;n ejrgasivan oJ drovmwn, i{na kai; ejl aunovmeno"
mh; livan ejsti;n ajrgo;" kai; kludwnizovmeno" h] para; tw'n ejcqrw'n
sugkrouovmeno" ijscurovtero" diamevnh/ kai; a[rrhkto".
eV Ecevtwsan de; kai; pavnta pro;" ejx artismo;n drovmwno" ajpara-
leipta; kai; dipla', oi|on aujcevna", kwvpa", skarmou;", scoiniva,
kavrua, kai; ta; a[rmena de; aujtw'n kai; keratavria kai katavrtia
kai; oJpovsa a[lla hJ nautikh; tevcnh pro;" creivan ajpaitei'. Ecevtw
de; kai; ejk perissou' xuvl a tina; ejgkoivlia kai; sanivda" kai;
------------------------------
2
The numbering of this paragraph in the manuscript is bV [2]. There is no
numbering for the first paragraph. From here to paragraph 59 Dain’s Greek
numbering is out by one in each case. From paragraph xV [60] Dain’s Greek
numbering coincides with that of the manuscript. Paragraphs 60 and 61 in Dain’s text
are a single paragraph xV [60] in the manuscript.
NAUMACIKA LEONTOS BASILEWS 485
The Naval Warfare of the emperor Leo

We wish to discuss naval warfare, [a topic] on which we have
found no prescription in the old tactical [manuals]. We have
read these at random and we have acquired a little experience of
the present time from our naval strate2goi*, both of what they
had done and what they had suffered, [and] having selected
some brief [examples] to give something of a starting point for
those about to fight at sea in [what were] once called trie2reis*
(triremes) but [are] now known as dromons, we will set [this]
out succinctly.
2 First, strate2gos of the naval force, you must have knowledge of
naval practice and formation, and [know how] to look out for
and anticipate shifts in the airs and breezes through experience
of the stars that appear and of the signs in the stars and [through
experience of] the marks that happen on the sun and the moon.
And [you must] have a precise knowledge of the change of the
seasons, so that, being experienced in these, you may be
preserved safe and sound from storms at sea.
3 [You] must equip dromons that are adequate for naval warfare
against the enemy ships campaigning against you and against
their condition, and [you must] make your [ships’] equipment
able to withstand them in all respects.
4 The construction of the dromons should be neither too heavy, or
they will be sluggish when under way, nor built too lightly, or
they will be weak and unsound and quickly broken up by the
waves and the attacks of the opposition. Let the dromon have
suitable workmanship so that it is not too sluggish when under
way and remains sturdy and unbroken when in a gale or struck
by the enemy.
5 There should be a complete supply in duplicate of the fittings of
a dromon, such as rudders (auche2n es*), oars, tholes (skarmoi*),
oar-grommets (schoinia*), blocks (karya*), and their sails, and
yards (kerataria*), and masts (katartia*), and everything else
the nautical art considers necessary. [The ship] should also have
486 APPENDIX TWO
stuppiva kai; pivssan kai; uJgrovpisson: kai; nauphgo;n meta;

pavntwn tw'n ejrgaleivwn aujtou' e{{na tw'n ejlatw'n, oi|on skepavrnou,
trupavnou, privono" kai; tw'n oJmoivwn.
ıV Ecevtw de; pavntw" to;n sivf wna kata; th;n prwv/ran e[mprosqen
calkw'/ hjmfiesmevnon, wJ" e[qo", di ou| to; ejskeuasmevnon pu'r kata;
tw'n ejnantivwn ajkontivsei. Kai; a[nwqen de; tou' toiouvtou sivf wno"
yeudopavtion ajpo; sanivdwn, kai; aujto; periteteicis-mevnon
sanivsin, ejn w|/ sthvsontai a[ndre" polemistai; toi'" ejpercomevnoi"
ajpo; th'" prwv/ra" tw'n polemivwn ajntimacovmenoi h] kata; th'"
polemiva" nho;" o{lh" bavllonte" di o{swn a]n ejpinohvswsin o{plwn.
zV Alla; kai; ta; legovmena xulovkastra peri; to; mevson tou'

katartivou ejn toi'" megivstoi" drovmwsin ejpisthvsousi peri-
teteicismevna sanivsin, ejx w|n a[ndre" tine;" eij" to; mevson th'"
polemiva" nho;" ajkontivsousin h] livqou" mulikou;" h] sivdhra
bareva, oi|on maziva xifoeidh', di w|n h] th;n nau'n diaqruvyousin, h]
tou;" uJpokeimevnou" sunqlavsousi, sfodrw'" kataferovmena, h]
e{terovn ti ejpiscuvsousin h] ejmprh'sai dunavmenon th;n nau'n tw'n
polemivwn h] tou;" ejn aujth'/ qanatw'sai. ”Ekasto" de; tw'n dromwvnwn
eujmhvkh" e[stw kai; suvmmetro", e[cwn ta;" legomevna" ejlasiva" duvo,
thvn te kavtw kai; th;n a[nw.
hV ÔEkavsth de; ejcevtw zugou;" to; ejl avciston keV ejn oi|" oiJ kwphlavtai
kaqesqhvsontai, wJ" ei\nai zugou;" tou;" a{panta" kavtw me;n keV,
a[nw de; oJmoivw" keV, oJmou' nV. Kaq e{na de; aujtw'n duvo kaqezevs-
qwsan oiJ kwphlatou'nte", ei|" me;n dexiav, ei|" de; ajristerav, wJ"
ei\nai tou;" a{panta" kwphlavta" oJmou' tou;" aujtou;" kai;
stratiwvta" touv" te a[nw kai; tou;" kavtw a[ndra" rV. “Exw de;
touvtwn to;n kevntarcon tou' drovmwno" kai; to;n to; flavmoulon
katevconta kai; touv" duvo kubernhvta" tw'n tou' drovmwno"
aujcevnwn, ou}" kalou'si prwtokaravbou", kai; ei[ tina e{teron devon
eij" th;n tou' kentavrcou uJphresivan. Tw'n de prw/raivwn ejlatw'n oiJ
teleutai'oi duvo, oJ me;n e[stw sifwnavtwr, oJ de; e{tero" oJ ta;"
ajgkuvra" bavllwn kata; qavlassan, h[goun ta; sivdhra: e[stw de; kai;
oJ prw/reu;" a[nw th'" prwv/ra" kaqhvmeno" e[noplo". Kai; oJ tou'
nauarcou dev, h[toi tou' kentavrcou, kravbato" ejpi; th'" pruvmnh"
ginevsqw, oJmou' me;n ajf wrismevnon deiknuvwn to;n a[rconta, oJmou' de
some extra timbers, [that is] floor timbers (enkoilia*), and

planks, and tow, pitch, and liquid pitch. And, one of the
oarsmen [should be] a shipwright with all the tools, such as an
adze, an auger, a saw, and the like.
6 Most importantly [the ship] should have a sipho2n* (flame-
thrower) in front at the prow, bound in bronze, as is usual,
through which processed fire can be hurled against the enemy.
Above this sipho2n [there should be] a false floor
(pseudopation*) of planks, itself fortified with planks, on which
marines can stand to fight those attacking from the prow of the
enemy or to throw whatever weapons they can devise against
the whole enemy ship.
7 Moreover they will set up the so-called xylokastra* (wooden
castles), fortified with planks, around the middle of the mast on
the largest dromons, from which men will throw into the middle
of the enemy ship mill stones or heavy iron [weights], like
sword-shaped blooms,3 with which they will either break up the
ship or crush those underneath as [the weights] crash down
heavily, or they will be able to achieve some other result, either
setting fire to the enemy ship or killing those in it. Each of the
dromons should be long and [well] proportioned with two
elasiai* (oar-banks), one below and one above.4
8 Each [oar-bank] should have at a minimum twenty-five zygoi*
(thwarts), on which the oarsmen will be seated, so that in all
there are twenty five thwarts below and similarly twenty five
above, making a total of fifty. Two oarsmen should sit on each
of these [thwarts], one on the right and the other on the left, so
that all the oarsmen together themselves [are] also soldiers,
both those above and those below, [total] one hundred men.5
Apart from these [there should be] the kentarchos* (“captain”)
of the dromon and the one who keeps the standard and the two
kyberne2tai* (helmsmen) [in charge] of the rudders of the
dromon, who they6 call pro2tokaraboi*, and anyone else who is
required in the service of the kentarchos. Of the last two
oarsmen at the prow, one should be the sipho2nato2r* (operator
of the flame-thrower), the other should be the one who throws
the anchors, that is the “irons”, into the sea. The bowman sta-
------------------------------
3
Mazivon: literally, a “lump” or “mass”.
4
This sentence belongs with §8 but is included here in §7 in the manuscript.
5
Cf. above pp. 254-6, 260-61.
6
That is, people of the present day, the tenth century.
488 APPENDIX TWO
kai; fulavttwn ejn kairw'/ sumbolh'" ajpo; tw'n rJiptomevnwn belw'n

para; tw'n ejnantivwn, ejx ou| kai; e{kasta blevpwn pro;" th;n creivan
mavlista keleuvsei oJ a[rcwn to;n drovmwna.
qV Kai; e{teroi de; drovmwne" kataskeuazevsqwsavn soi touvtwn

meivzone", ajpo; diakosivwn cwrou'nte" ajndrw'n h] plevon touvtwn h]
e[latton kata; th;n creivan th;n devousan ejpi; kairou' kata; tw'n
ejnantivwn: w|n oiJ me;n nV eij" th;n kavtw ejl asivan uJpourghvsousin, oiJ
de; rV kai nV a[nw eJstw'te" a{pante" e[noploi machvsontai toi'"
polemivoi".
iV “Eti de; kai; kataskeuavsei" drovmwna" mikrotevrou" gorgotavtou",
oiJonei; galeva" h] monhvrei" legomevnou", tacinou;" kai; ejlaffouv",
oi|sper crhvsh/ ejn te tai'" bivglai" kai; tai'" a[llai" spoudaivai"
creivai".
iaV Kai; eJtevra" de; nau'" poihvsei" forthgou;" kai; iJppagwvgou"
touvldou divkhn, ai{tine" th;n ajposkeuh;n a{pasan tw'n stratiwtw'n
bastavsousin, i{na mh; di aujth;n barou'ntai oiJ drovmwne" kai;
mavlista ejn ajgw'no" kairw'/, o{te de; creiva mikra'" dapavnh" h]
o{plwn h] a[llh" u{lh" ejkei'q en ajnalambavnwsi ta;" dioikhvsei".
ibV To;n de; tw'n dromwvnwn ajriqmo;n kai; tw'n ejn aujtoi'" stratiwtw'n
ajneikastovn ejstin kai; a[dhlon diorivsasqai:7 hJ ga;r kata; to;n
kairo;n creiva pro;" th;n tw'n ajntimacomevnwn polemivwn duvnamin,
wJ" a]n ajpaithvsh/ kai; to; plh'qo" tw'n dromwvnwn. Kai; pavlin to;n
ajriqmo;n tou' ejn aujtoi'" laou' kata; to; mevgeqo" tw'n ploivwn kai;
th;n devousan ejn aujtoi'" polemikh;n o{plisin ou{tw kai; poihvsei".
igV Prosevti de; kai; ta; skeuofovra kai; iJppagwga; ploi'a tou;" ejn
aujtoi'" ajrkou'nta" e{xousi nauvta", oujde; aujtou;" ajnovplou", ajlla;
kai; tovxa e[conta" kai; sagivta" kai; rJiptavria kai; ei[ ti creiw'de"
pro;" povlemon e{teron dia; ta;" ajnagkaiva" peristavsei".
Epiferevsqwsan de; kai; perissa; o{pla: pote; gar kai; leipovntwn
o{plwn, ejkei'q en oiJ stratiw'tai lavbwsin. Ta; de; toiau'ta ploi'a
kai; a[rmata ejcevtwsan kai; mavggana kai; ta; a[lla o{pla pro;"
creivan, eij tuvch/ mhv pote ejpileivpwsin katadapanwvmena ejn tai'"
mavcai".
------------------------------
7
Cf. Appendix Three, §3.1.
tioned above the prow should be under arms. The krabatos* of

the navarchos* or the kentarchos should be at the stern, both
indicating [that] the archo2n* (commander) is set apart, and also
protecting him at a time of attack from the missiles thrown by
the opposition. And from here the commander can see
everything and give orders for the dromon as necessary.
9 Other dromons should also be constructed for you [which are]
larger than these, with space for two hundred men, perhaps
more or fewer according to the compelling need of the moment
against the opposition. Of these, fifty should serve on the lower
oar-bank and one hundred and fifty, stationed above and all
armed, will fight the enemy.
10 As well, you will also construct smaller, very fast dromons, that
is [those] known as galeai* or mone2reis (monoremes), speedy
and light, which you will use as sentinels and for other essential
tasks.8
11 And you will build other ships [as] phorte2goi* (supply) [ships]
and horse transports, like a baggage-train, which will carry all
the equipment of the soldiers, so that the dromons are not
burdened with it; and especially in time of battle, when there is
need of a small supply of weapons or other materiel, [these]
undertake the distribution.
12 About the number of dromons and the soldiers in them, it is
impossible and unrealistic to be prescriptive. For the number of
dromons required varies according to the needs of the moment,
when facing the opposing enemy forces. Once again, you will
supply the number of the force in them according to the size of
the ships and the warlike armament required in them.
13 As well the skevophora* (supply [ships]) and the horse
transports will have sufficient nautai* (sailors) on board, and
these [should] not be unarmed but [should] have bows and
arrows and javelins and anything else [that might be] necessary
in another battle in difficult circumstances. Extra weapons
should also be loaded, for sometimes when there is a shortage
of weapons the soldiers can draw on these. Ships of this sort
should also have arms, mangana,* 9 and other weapons as
needed, so that they should never run short when used up in
battle.
------------------------------
8
Cf. Appendix One, §6.1-2; Appendix Three, §3.2.
9
Cf. §67 below.
490 APPENDIX TWO
idV Ekto;" de; tw'n stratiwtw'n h[toi tw'n a[nw ejl atw'n, o{soi a[n eijsin,
ajpov te tou' kentavrcou kai; ejfexh'" e{w" tou' ejscavtou,
katavfraktoi e{sontai o{pla e[conte" oi|on skoutavria, mevnaula,
tovxa, sagivta" ejk perissou', spaqiva, rJiptavria, lwrivkia,
klibavnia, eij kai; mh; o[pisqen, ajlla; pavntw" e[mprosqen pevtala
e[conta, kassivda", ceirovyella, kai; mavl ista oiJ e[mprosqen ejn th'/
prosbolh'/ th'" mavch" kata; cei'ra" sumplekovmenoi kai;
ajgwnizovmenoi. OiJ de; mh; e[conte" lwrivkia h] klibavnia, pavntw"
foreivtwsan ta; legovmena neurikav, a{per ajpo; diplw'n
kentouvklwn givnetai. Kai; ou|toi o[pisqen tw'n a[llwn skepovmenoi
tovxoi" crhvsontai. Kai; livqou" de; dunamevnou" ajpo; ceirw'n
rJivptesqai pleivstou" ejcevtwsan h[toi kovclaka" ejn toi'"
dromwnivoi", ou{sper kata; polemivwn bavllonte" oujk e[latton tw'n
a[llwn o{plwn aujtou;" katablavyousin: o{pla gavr eijsin oiJ livqoi
eujpovrista kai; ajnelliph'.
ieV Mh; mevntoi ou{tw" ballevtwsan tou;" livqou" movnon w{ste th;n
duvnamin aujtw'n ejn touvtoi" ejkdapanh'sai kai; sth'nai tou' loipou'
h] kai; ta; o{pla ta; ballovmena ajpokenw'sai, mhv pote oiJ ejnantivoi
suvskouta poihvsante" kai; ta;" bola;" oJpwsou'n dexavmenoi, ei\ta
tw'n belw'n plhrwqevntwn kai; tw'n balovntwn ajpokamnovntwn,
ajqrovoi ajnastavnte" ajparxwntai tai'" spavqai" kai; toi'"
menauvloi" ajmuvnesqai, kai; ajkopivastoi10 th'/ ajqrova/ kinhvsei
ajnafanevnte" kai; toi'" kekopiakovsi stratiwvtai" ejpitiqevnte"
ijscurovteroi gevnwntai kai; eujkovlw" aujtou;" katapolemhvswsin.
Filei' ga;r ta; toiau'ta to; bavrbaron.
iıV ÔUpomevnousi ga;r Sarakhnoi; th;n bivan th'" prosbolh'" kai; o{tan
ajpokamovnta" i[dwsin kai; tw'n o{plwn kenwqevnta" h] sagitw'n h]
livqwn h] eJtevrwn tinw'n, tovte ajnaphdw'nte" oJmou' te kataplhvt-
tousin kai; tai'" ejk ceiro;" ajpo; spaqivwn kai; menauvlwn
prosbolai'" eujrwvstw" te kai; ajkmaiovteron ejpevrcontai.
izV Dio; fulavttesqai crh; ta; toiau'ta kai; meta; tou' devonto" skopou'
poiei'sqai th;n prosbolh;n, i{na ma'llon oiJ polevmioi pavqwsi ta;
pro;" blavbhn ginovmena h] oiJ hJmevteroi stratiw'tai. Dei' ga;r
aujtou;" th;n oijkeivan duvnamin kai; ta;" boula;" fulavttein ajp ajrch'"
a[cri tevlou" th'" mavch" kai; metrei'n tw'n ejnantivwn th;n diavq esin
kai; ou{tw" th;n mavchn diaskeuavzein. 11
ihV Pro;" touvtoi" frontivsei", w\ strathgev, kai; th'" deouvsh" tw'n
------------------------------
10
ajkopivastoi, thus Dain: ajkopivatoi MS. A.
11
diaskeuavzein, thus Dain: diaskedavzein MS. A.
14 Apart from the soldiers or the upper oarsmen, [all others]

however many there might be, from the kentarchos down to the
last [man], should be kataphraktoi* - having weapons such as
shields, pikes, bows, extra arrows, swords, javelins, corselets,
lamellar cuirasses (certainly with plates in front even if not at
the back), helmets, [and] vambraces - especially those engaged
in fighting hand-to-hand in the front line of attack in battle.
Those who do not have corselets or lamellar cuirasses should
certainly wear what are known as neurika*, which are made
from double layers of felt. And these, protected behind the
others, should use bows. There should also be in the dromons
large quantities of stones or pebbles that can be hurled by hand;
[when] they throw these at the enemy, they can hurt them no
less than [with] other weapons, for stones are weapons that are
easily obtained and abundant.
15 But they should not just throw the stones in such a way that
they expend their energy on these and do nothing thereafter, or
consume the throwing weapons, in case the opposition links
shields and absorbs the missiles however they might, then,
when these are used up and those throwing them are exhausted,
come out all together and begin to counter-attack themselves
with swords and pikes, and seemingly unwearied in their mass
movement and attacking soldiers who are already weary, they
are stronger and easily overpower them. Barbarians like [doing]
such [things].
16 For the Saracens endure the impetuosity of the attack and, when
they see [that their attackers] are tiring and have used up their
weapons, or arrows, or stones, or whatever else, then they rush
out all together and both take them by surprise and [also] in
hand-to-hand fighting with swords and pikes attack vigorously
and more energetically.
17 So precautions should be taken against such situations and the
attack should be made with the necessary forethought, so that it
is the enemy who suffers harm rather than our soldiers. They
must preserve their own energy and projectiles from the
beginning to the end of the battle and measure the condition of
the enemy and make ready for the battle accordingly.
18 In addition, strate2gos, you will consider the essential supplies

492 APPENDIX TWO
stratiwtw'n dapavnh", w{ste e[cein aujtou;" ta; ajnagkai'a, i{na mh;

touvtwn leipovmenoi h] stasiavswsin h] ejn th'/ ijdiva/ cwvra/ o[nte"
tou;" suntelesta;" kai; uJphkovou" hJmw'n turannw'sin kai;
ajdikw'sin th'/ spavnei tw'n ajnagkaivwn ajnagkazovmenoi. All, ei[
ge dunatovn, ejn tavcei th;n polemivan katalavbh/" gh'n kai; ejx
aujth'" a{panta ta; ejpithvdeia proslavbh/".
iqV Paraggeivlh/"12 de; kai; toi'" a[rcousi mhdevna tw'n uJp aujtou;"
stratiwtw'n ajdikei'n h] to; oiJonou'n dw'ron par aujtw'n lambavnein
h] ta;" legomevna" sunhqeiva". Peri; ga;r th'" sh'" ejndoxovthto" tiv
crh; levgein wJ" oujde; ejnqumhqh'nai ti toiou'ton devon mhvti ge
diapravx esqai, mhvte dw'ron to; oiJondhvpote ajpo; mikrou' h] megavlou
ajnqrwvpou tou' uJpo; se; telou'nto" lambavnein to; suvnolon.
kV Tou;" de; stratiwvta" ajndreivou" ejpilevgou kai; proquvmou" kai;

mavlista tou;" eij" ta; a[nw tou' drovmwno" tassomevnou", oi{tine"
kai; ajpo; ceiro;" toi'" polemivoi" sumplevkontai. Eij dev tina" tw'n
stratiwtw'n ajnavndrou" ejpignw'/", touvtou" eij" th;n kavtw ejlasivan
paravpempe, kai; ei[ potev ti" plhgh'/ h] pevsh/ tw'n stratiwtw'n, to;n
ejkeivnou tovpon ejk tw'n kavtw ejx ajnavgkh" ajnaplhrwvsei".
kaV Crh; gavr se pavntw" eijdevnai th;n eJkavstou tw'n uJpo; se; stratiwtw'n
e{xin kai; diavq esin kai; th;n a[llhn pro;" ajndreivan poiovthta,
w{sper oiJ kunhgevtai tw'n kunw'n eJkavstou ta;" ejpithdeiovthta"
ejpiginwvskonte" e[cousin eujkaivrou" aujtou;" pro;" o} bouvl ontai.
kbV Ou{tw" ou\n diaqhvsei" e{kasta kaqw;" a]n sunivdh/" ajrkou'nta pro;"
th;n prokeimevnhn ejkstrateivan, touv" te drovmwna" kai; tou;" ejn
aujtoi'" stratiwvta", tav te o{pla kai; ta;" dapavna" kai; th;n
a[llhn ejn eJtevroi" ploivoi" ajposkeuhvn, h{ntina oiJonei; tou'l-
don ejn ajsfalevsi tovpoi" se crh' kaqista'n, o{tan kairo;"
ejlpivzhtaiv soi mavch".
kgV Kai; prosevti, ei[ ge creiva toiauvth kalevsei, kai; i{ppou" ejn toi'"
iJppagwgoi'" ploivoi" prosepirrivptein w{ste kat;a th'" polemiva"
e[cein kaballarivou": kai; aJplw'" pavnta ajpartivsa" oJdoiporhvsei"
deovntw".
kdV Kai; prw'ton me;n pro; tou' ajpokinh'sai aJgiasqhvtwsan a{panta ta;
flavmoula tw'n dromwvnwn dia; qeiva" tw'n iJerevwn iJerourgiva" kai;
euJch'" ejktenou'" pro;" to;n tw'n o{lwn Qeo;n uJpe;r eujodwvsew" tou'
stratou' kata; tw'n polemivwn. “Epeita kai; dialalhvsei" pro;"
------------------------------
12
Paraggeivlh/", thus Dain: Paraggeivl ei" MS. A.
of the soldiers, for them to have what is necessary, so that they

do not rebel for lack of these things or, if in their own territory,
oppress and mistreat the tax payers and our subjects, compelled
by scarcity of what is necessary. But, if possible, you should
quickly capture enemy land and obtain from it everything that
is required.
19 You should also instruct the commanders that they are not to
wrong any of the soldiers under them or to accept any gift
whatever from them or what are known as the customary
perquisites.13 But concerning your Gloriousness, what should
[I] say as you have neither considered doing any such [thing],
nor taken any gift whatsoever from any man great or small
serving under you?
20 Choose courageous and vigorous soldiers, especially those
stationed on the upper [part] of the dromon, who engage the
enemy in hand-to-hand fighting. If you realize that any of the
soldiers are cowardly, send them to the lower oar-bank, and if
any of the soldiers should be wounded or fall you should fill his
place from those below out of necessity.
21 You should above all be aware of the condition and general
level of bravery of each soldier under you, as huntsmen know
the capabilities of each of their dogs and have them ready for
their requirements.
22 You will arrange everything as you see is sufficient for the
proposed expedition: the dromons, and the soldiers in them, the
weapons and the supplies and the remaining equipment in other
ships, which you should station like a baggage train in safe
places at whatever time you anticipate a battle.14
23 In addition, if such a need arises, you should load horses onto

the horse-transport ships so that you have cavalry [to use]
against the enemy. Then, [to put it] simply, having completed
all preparations, you will proceed suitably.
24 First, before moving off, all the standards of the dromons
should be blessed during a celebration of the Liturgy by the
priests, and by a lengthy prayer to the God of all for the
successful venture of the stratos* against the enemy. Then you
------------------------------
13
It was a long practised custom in the Greco-Roman and Byzantium worlds for
offices to be acquired by payment of a fee or perquisite (sunhvq eia), to those who had
the dispensation of them.
14
Cf. Maurice, Ek tou' Maurikivo u, §§1, 5, (pp. 41-2).
494 APPENDIX TWO
a{panta to;n lao;n kai; pro;" tou;" a[rconta" ijdivw" ta; devonta kai;
aJrmovz onta tw'/ kairw'/ kai; ou{tw" proqumopoihvsa" to;n strato;n
ajpokinhvsei", ejpithdeivou ajnevmou pneuvsanto" kai; mh; ejnantivou.
keV Oujc wJ" e[tucen aJpavntwn tw'n dromwvnwn poreuomevnwn, ajll

ejpisthvsei" aujtoi'"15 a[rconta" h] kata; pevnte h] kata; trei'" drovmw-
na", e{na to;n legovmenon kovmhta, o{sti" nauvarcov" te kai; hJgemw;n
tw'n uJp aujtw'n dromwvnwn uJpavrcwn frontivsei prosecevsteron
peri; pavntwn eujkovl w" kai; diatavx ei pro;" e{kasta.
kıV OiJ de; eijrhmevnoi a[rconte" uJpo; se; telou'nte" ajpo; sou' kai; ta;
paraggevl mata devxontai kai; toi'" uJp aujtou;" metadwvsousin. Kai;
tau'ta me;n ejpi tou' basilikou' legomevnou plwi?mou: [ejpi; de; tw'n
qematikw'n dromwvnwn16 kai; drouggavrioi ejpisthvsontai kai;
tourmavrcai, kai; aujtoi; tw'/ strathgw'/ uJpotaghvsontai kai; toi'"
ejkeivnou paraggevlmasin uJpakouvsousin.
kzV Oujk ajgnow' de; o{ti kata; th;n oJmoivwsin tou' basilikou' plwi?mou
kai; oiJ tw'n a[llwn qemavtwn plwvi?moi strathgoi; drouggavrioi
ejkalou'ntov pote toi'" prwvhn crovnoi" kai; oiJ uJp aujtou;" kovmhte"
movnon kai; kevntarcoi: ajlla; nu'n eij" strathgivda hJ eJkavstou tw'n
drouggarivwn ajrch; ajnabevbhken kai; ou{tw kaloumevnh tai'"
strathgikai'" katamerivzetai tavx esin.
khV Gumnavsei" de; diafovrw" touv" te plwi?mou" stratiwvta" kai;
aujtou;" tou;" drovmwna", pote; me;n kaq e{na e{kaston a[ndra, pote;
de; kata; pleivona", w{ste katevnanti ajllhvlwn ejpevrcesqai
spaqivoi" kai; skoutarivoi" crwmevnou": kai; aujtou;" de; o{lou"
drovmwna" kat ajllhvlwn wJ" ejpi; paratavx ew" ejpercomevnou" kai;
pote; desmou'nta", pote; de; ajpoluvonta" kai; diafovrw" kat
ajllhvlwn prosbavllonta", pote; de kai;; ajkontivoi" wjqou'nta" ta;
ploi'a tw'n ejnantivwn, w{ste mh; plhsiavzonta" desmei'n: ouj ga;r ajei;
to; dia; kamavkwn sidhrw'n desmei'n ajllhvlou" tou;"
ajntipolemou'nta" crhvsimon dia; tou;" ajf euvktou" kai; ajnagkaivou"
kinduvnou".
kqV Kai; ejtevrw" de; gumnazevsqwsan wJ" a]n nohvsh/ hJ sh; ejndoxovth" ta;"
kata; tw'n ejnantivwn ejndecomevna" ejpinoiva", wJ" a]n ejnteu'q en
ejqivzwntai pro;" tou;" ktuvpou" kai; boa;" kai; th;n a[llhn kivnhsin
tou' polevmou, kai; mh; taravsswntai wJ" ajgumnavstw" kai; ajqrovon
kai; para; dovxh/ ejpi; tau'ta ejrcovmenoi.
------------------------------
15
aujtoi'", thus Dain: aujtou;" MS. A.
16
[ejpi; de; tw'n qematikw'n dromwvnwn], thus Dain. MS. A omits this. However Dain’s
emendation is required by “me;n” in the previous clause.
will address the entire force and the commanders with suitable
words fitting to the occasion and, having thus inspired the
stratos, you will move off when a favourable wind, and not an
adverse one, has arisen.
25 The dromons should not proceed haphazardly, but you will put
in them commanders [in charge] of every five or three dromons,
a so-called kome2s* (count), who as navarchos and he2g emo2n* of
the dromons under them, will have particular responsibility in
all matters and make every arrangement.
26 These commanders who have just been mentioned serve under
you and will receive instructions from you and will pass them
on to those under them. This is the system in what is known as
the imperial fleet. [In the thematic dromons] both droungarioi*
and tourmarchai* will be appointed and they will be
subordinate to the strate2gos and obey his instructions.
27 I am not unaware that by analogy with the imperial fleet the
naval strate2goi also of the other themes were once in previous
times called droungarioi and those under them were only
kome2tes and kentarchoi. But now in a strate2gos’s command the
office of each of the droungarioi has risen and is classed, under
this name (i.e., strate2gos), in the ranks of strate2gos.17
28 You will exercize both the naval soldiers and the dromons in
different ways, sometimes as each individual man and
sometimes in groups, when they attack each other with swords
and shields. And [you will make] all the dromons attack each
other as if in formation, sometimes coupled together, some-
times not coupled and attacking each other in different ways,
sometimes also pushing the ships of the opposition away with
poles so that they do not come close and couple. For it is not
always advantageous for [those] warring to couple themselves
together with kamakes side2rai* (iron rods) because of the
unavoidable and inevitable dangers.
29 They should also be exercized in other ways, as your
Gloriousness perceives the techniques [to be] expected against
the opposition, so that thereafter they are accustomed to the
blows, cries, and general commotion of war and will not be
confused through being untrained should they encounter these
things all at the same time and unexpectedly.
------------------------------
17
Cf. above pp. 267-8.
496 APPENDIX TWO
lV Ou{tw" ou\n gumnasqevnte" kai; diateqevnte" pleuvsousin ejn tavxei

sunhgmevnoi tosou'ton ejf o{son ajllhvloi" mh; ejmpodivz ein ejn tai'"
ejlasivai" kai; ejn tai'" wJ" eijko;" kata; qavl assan uJpo; tw'n ajnevmwn
bivai": ajll oiJoneiv ti" paravtaxi" gegumnasmevnh, ou{tw
poreuevsqwsan. Kai; ejn tai'" oJrmhsivai" de; tw'n ajplhvktwn
eujtavktw" to;n katavploun poieivtwsan kai; katagevtwsan
ejnordivnw" ejxormw'nte" pro;" th;n xhra;n h] eij" limevna pavntw", h]
eij" u{formon tovpon ejn w|/ zavl h" sumbaivnoush" ouj klusqhvsontai.
laV Dei' dev se kai; th;n tou' ajnevmou ejpifora;n proeidevnai dia; tw'n
shmeivwn kata; to;n kairovn kai; pro;" tauvthn kai; to;n tovpon th'"
oJrmhsiva" ejklevxasqai, kai; eij mhv ti" katepeivgh/ ajnavgkh, mh; a[neu
pneuvmato" aijsivou kai; galhvnh" kai; ajsfalou'" ejlpivdo" swthriva"
ejpirrivptein seauto;n eij" ajnepithvdeion plou'n, ajll uJfora'sqai
kai; ta;" legomevna" tw'n nautikw'n parashmasiva" tw'n a[strwn kai;
o{sa a[lla sumfevronta, kai; ou{tw" poiei'sqai th;n poreivan.
lbV En de; toi'" ajplhvktoi", eij me;n ejn th'/ ijdiva/ oJrmei'" cwvra/ kai; mhdevna
fovbon e[cei" ajpo; tw'n polemivwn, kai; ou{tw" meta; eujtaxiva"
ajnapauvesqai to;n strato;n kai; ejn nukti; kai; ejn hJmevra/, mhdevna
tw'n ejpicwrivwn blavptonta" h] ajdikou'nta" h] karpou;" aJrpavzonta"
h] fqeivronta".
lgV Ei; de; ejn th'/ polemiva/ gh'/ plhsiavzei" h] polemivou" parei'naiv pou
ejlpivzei", pavntw" crhv se bivgla" e[cein makrovq en kai; kata; gh'n
kai kata; qavl attan, kai; ajgruvptw" diatelei'n kai; kathsfalis-
mevnon kai; e{toimon ei\nai eij" paravtaxin: pollai; ga;r aiJ tw'n
polemivwn ejpiboulaiv. Kai; ga;r h] kata; gh'" euJrovnte" se oJrmou'nta
biavsontai, eij tuvcoi de; kai; ta;" nau'" ejmprhvsousin, h] dia;
qalavssh" ajnafanevnte" prosbolh;n poihvsousi nukto;" kai;
hJmevra". Kai; eja;n ajnevtoimo" ejn eJtoivmoi" euJrevq h/", proterhvsousin
oiJ ejnantivoi kata; sou', eij dev se e{toimon euJrhvsousin, a[prakto"
aujtoi'" hJ ejpiboulh; genhvsetai.
ldV Epei; de; touvtwn summevtrw" ejmnhvsqhmevn te kai; dietaxavmeqa,
fevre loipo;n kai o{pw" paratavxei" kai; ta;" prosbola;" ta;" ejn
tai'" mavcai" poihvsei" wJ" ejn sunovyei dioriswvmeqa, kaq o}n
trovpon kai; ejn tai'" kata; gh'n polemikai'" prosbolai'" dietaxa-
meqa.
leV ”Otan toivnun ejlpivzhtaiv soi polevmou kairov", w\ strathgev,
sunelqovntwn tw'n stratiwtw'n kata; ta;" tavx ei" eJkavstwn dih/rh-
30 When they have been exercized and organized, they will sail in
formation, with a sufficient distance between each [ship] to
prevent their colliding when rowing,18 and in the wind gusts to
be expected at sea. Moreover, they should proceed according to
the formation which has been exercized. In the moorings of the
aple2kta* they should make their kataplous (landing) in good
order,19 and they should put in to shore in a regular manner,
making for dry land, or especially to a harbour, or to a mooring,
in which they will not be battered should a squall arise.
31 You should anticipate the direction of the wind through the
seasonal signs and then choose the mooring place accordingly.
If there is no urgent need, do not throw yourself into an
inauspicious voyage without a favourable wind, a calm [sea]
and a secure expectation of safety, but also take into account
what are known by sailors as the stars’ signs and all other
relevant matters, and then proceed appropriately.
32 In the aple2kta, if you moor in [our] own territory and have no
fear of the enemy at all, [you may] thus [allow] the stratos to
rest in good order by both day and night, harming none of the
local inhabitants or wronging them or seizing their produce or
doing any damage.
33 But if you approach enemy land or you expect the enemy to
appear somewhere, you must certainly have scouts some way
off on both land and sea, and [you] should remain vigilant and
alert and ready for [drawing up] the formation. For the devices
of the enemy are many. Either, finding you moored to the land,
they will attempt to burn the ships or, appearing by sea, they
will make an attack night and day.20 And if you find yourself
ill-prepared amongst the prepared, the opposition will get the
better of you; but if they find you prepared, their devices will
achieve nothing.
34 Since we have now recalled these matters adequately and
discussed [them], let us then briefly indicate how you will
organize formations and attacks in battles, in the way in which
we discussed attacks in battles on land.
35 Whenever, strate2gos, you anticipate a period of fighting, when

the soldiers have come together, each drawn up in their forma-
------------------------------
18
Cf. Maurice, Ek tou' Maurikivo u, §7 (p. 42).
19
20
498 APPENDIX TWO
mevnwn, uJpanagnwsqhvsetai aujtoi'" ta; stratiwtika; ejpitivmia a{per

hJmi'n ejn tw'/ peri; th'" kata; gh'n stratiwtikh'" gumnasiva" ei[rhtai,
kai; ejpirrwvsei" aujtou;" kai; ejniscuvsei" lovgoi" proshvkousi
parormw'n kai; ejpaleivf wn pro;" tou;" ajgw'na", i{na to; me;n dia; to;n
fovbon tw'n ejpitimivwn, to; de; dia; th;n th'" sh'" ejndoxovthto"
paraivnesin ajndrei'oi kai; eu[tolmoi gevnwntai kai; ejn toi'"
mevllousi polemikoi'" kinduvnoi" ejk ceiro;" ajgwnizovmenoi.
lıV Dei' dev se ma'llon di ejfovdwn me;n kai; a[llwn ejpithdeumavtwn te
kai; strathghmavtwn meqodeuvein kata; tw'n polemivwn h] di o{lou
tou' uJpo; se; plwi?mou stovlou, h] dia; mevro" aujtou'. Mh; mevntoi
cwri;" ajnavgkh" megavl h" ejpi; tou'to katepeigouvsh" eij" dhmovsion
povlemon seauto;n ejpirrivptein: poll;a; ga;r ta; th'" legomevnh"
tuvch" ejnantiwvmata kai; ta; tou' polevmou paravdoxa.
lzV Dia; tou'to crhv se ajei; parafulavttesqai kai; mh; pro;" dhmosiva",
wJ" ei[rhtai, paratavxei" ajpoqrasuvnesqai, mavlista ejn ploivoi",
o{pou desmouvntwn ajllhvlou" a[f eukto" kai; biaiva hJ ejk ceiro;"
mavch givnetai kai; oujk e[sti dunato;n tou' sumfevronto"
ejpilabevsqai.
lhV Kai; tau'ta me;n fulavttesqai eij mh; a[ra qarrei'" kai; tw'/ plhvq ei
tw'n dromwvnwn kai; th'/ ajndreiva/ kai; oJplivsei kai; proqumiva/ tw'n
stratiwtw'n ejpikratevstero" ei\nai tw'n polemivwn.
lqV Ou[te ga;r plh'qo" ploivwn ou[te mevgeqo" katorqwvsei povlemon, eij
mh; tou;" ejn aujtoi'" polemou'nta" e[cousin eujyuvcou" kai;
gennaivou" kai; proquvmou" eij" th;n kata; tw'n ejnantivwn
ejgceivrhsin, kai; prov touvtwn eij mh; th;n qeivan eujmevneian kai;
summacivan e[cousi dia; kaqarovthto" bivou kai; dikaiosuvnh" pro;"
te tou;" suntelesta;" kai; pro;" tou;" polemivou", ei[21 tiv" ejsti to;
mhde;n ajnovsion ejn toi'" aijcmalwvtoi" diapravttesqai h] aijscro;n h]
ajfilavnqrwpon, kai; to; mh; ajdikouvmenon mh; ajdikei'n, tou;" de;
ajdikou'nta" meta; th'" tou' Qeou' bohqeiva" ajmuvnesqai.
------------------------------
21
ei[, thus Dain: h{ MS. A.
formations, the military code of penalties, which we have

discussed in the [handbook] on land-based military training,22
should be read out to them, clause by clause. And you will
encourage and strengthen them with appropriate speeches,
rousing and inciting [them] to the contest, so that, partly out of
fear of punishment and partly because of your Gloriousness’s
admonitions, they become brave and daring, even when
fighting hand-to-hand in the coming dangers of engagement.23
36 You must indeed deal with the enemy through attacks and other
practices and stratagems, either with the whole of the naval
fleet under you or with part of it. However, without some
urgent compelling reason for this, you should not rush into a
general engagement.24 For there are many obstacles [in the
workings] of so-called Tyche25 and events in war [are] contrary
to expectation.
37 You must therefore always be on guard and, must not be over
confident, as has been said, about general formations, especially
where ships are coupled to each other, when fierce hand-to-
hand fighting is inevitable and it is not possible to gain any
benefit.
38 [You should] take these precautions if indeed you are not
confident of being superior to the enemy in the number of the
dromons and the bravery, armament, and enthusiasm of the
soldiers.
39 For neither the number nor the size of the ships will bring an
engagement to a successful conclusion if they do not have
fighting in them [men] of good spirit, and [who are] stalwart
and enthusiastic in attacks on the opposition, and more
important than this, if they do not have divine favour and
support through the purity of [their] lives and [their] just
behaviour both to the tax payers and to the enemy; if they do
nothing contravening divine laws or disgraceful or inhuman to
the prisoners, do not injure when no injury has been done, and
wrong-doers are dealt with through God’s assistance.
------------------------------
22
The emperor referred here to regulations included in his Taktika at Constitution
VIII, §§19-27. See Leo VI, Taktika (PG), coll. 765-8. These regulations were based
on much older material which, in differing versions, dated back to the age of Justinian
I and before. See also Ashburner, “Byzantine mutiny act”.
23
Cf. Appendix One, §9.15-18.
24
Cf. Appendix One, §9.8.
25
Tyche: personification of “Fate”, often used in Byzantine thinking as a substitute
for divine intervention.
500 APPENDIX TWO
mV Ea;n de; pavntw" ajpaitei'tai26 kai; mavch" kairov", diatavx ei" tou;"
drovmwna" poikivlw"27 kai; diafovrw", kavq w" a]n o{ te kairo;" kai; oJ
tovpo" ajpaith'/. 28 ”Wste eja;n qarrh'/" ejpikratevstero" ei\nai tw'n
polemivwn, w{" ei[rhtai, kai; dia; tou'to pro;" mavchn sumbavllein wJ"
ejlpivz wn nikhvsein aujtouv", mh; ejn th'/ ijdiva/ sou gh'/ plhsivon
poihvsei" th;n mavchn, ejn h|/ ejlpivsousin oiJ stratiw'tai to; dh;
legovmenon kataxulwvsante" swqh'nai, ajlla; ma'llon plhsivon th'"
tw'n ejnantivwn gh'", i{na aujtoi; th;n swthrivan ejlpivsante" e;n th'/ ijdiva/
gh'/ th;n fugh;n para; tou;" ajgw'na" protimhvswntai. Stratiwvth" ga;r
eij" deilivan ejn ajnavgkh/ polevmou peripivptwn th;n swthrivan dia;
th'" fugh'" ejlpivsei kai; tacevw" rJivyei ta; o{pla kai; oujde;n aujth'"
protimhvsetai: ojlivgoi ga;r oiJ ejn kairw'/ paratavxew" to;
ajpoqanei'n tou' ajdovx w" fugei'n prokrivnonte", ei[te ejn toi'"
barbavroi" ei[ph/", ei[te ejn toi'" ÔRwmaivoi".
maV Pro; de; th'" tou' polevmou hJmevra" crhv se bouleuvesqai meta; tw'n
uJpo; se; ajrcovntwn tiv dei' pra'xai, kai; o{per ajnafanh'/ dia; th'"
koinh'" boulh'" crhvsimon tou'to stoiceiw'sai. 29 Kai; paraggei'lai
toi'" a[rcousi tw'n dromwvnwn, w{ste ei\nai aujtou'" eJtoivmou"
ejkplhrw'sai ta; bouleuqevnta, ei[per mh; ejnantivon ti vajpanthvsh/30
ejk th'" ejfovdou tw'n polemivwn. Alla; kai; tovte eJtoivmou" ei\nai
pavnta" ajpoblevponta" eij" to;n so;n drovmwna, w{ste ejx aujtou'
labei'n shmei'on tiv a[ra poih'sai proshvkei, kai; touvtou doqevnto"
suntovmw" givnesqai to; uJpodeicqevn.
mbV Pavntw" ga;r dei' se, w\ strathgev, drovmwna ei[cein to;n i[dion ejx
a{panto" tou' stratou' ejpilevktou" e[conta tou;" stratiwvta"
megevq ei swvmato" kai; ajndreiva/ kai; ajreth'/ kai; th'/ a[llh/ panopliva/
diafevronta": kai; to;n drovmwna de; megevq ei kai; gorgovthti tw'n
a[llwn aJpavntwn diafevronta, wJ" kefalhvn tina th'" paratavx ew"
aJpavsh": kai; katasth'sai to;n th'" sh'" ejndoxovthto" toiou'ton
drovmwna, to;n dh; legovmenon pavmfulon.
mgV ÔOmoivw" de; kai; tou;" a[llou" uJpo; se; a[rconta" o{soi e[cousin uJp
auJtou;" tivna" drovmwna" ejx aujtw'n ejpilevx asqai a[ndra" kai; e[cein
ejn toi'" oijkeivoi", w{ste kai; aujtou;" diafevrein tw'n a[l lwn. Kai;
touvtou" de; pavnta" kai; tou;" loipou;" pro;" to;n so;n ajpoblevpein
drovmwna kai; par aujtou' rJuqmivzesqai kai; kanonivz esqai kata;
------------------------------
26
ajpaitei'tai, thus Dain: ajpaith'tai MS. A.
27
poikivlw", thus Dain: poikivlou" MS. A.
28
ajpaivth'/, thus Dain: ajpaitei' MS. A.
29
stoiceiw'sai, thus Dain, following Desrousseaux: stoicei'sai MS. A.
30
ajpanthvsh/, thus Dain: ajpanthvsei MS. A.
40 If, even so, the time requires a battle, you will form up the
dromons in a variety of ways, as the time and place requires.
Thus, if you are confident of your superiority over the enemy,
as has been said, and because of this you are engaging in battle
since you expect to defeat them, do not set up the battle near
your own land, in which [any] soldiers who, so to speak,
abandon ship, expect to take refuge, but rather near the land of
the opposition, so that they, expecting to find safety in their
own land, might prefer flight to the conflicts.31 For a soldier,
succumbing to cowardice under the pressure of battle, will hope
for safety in flight and will quickly abandon his weapons and
prefer nothing to it [safety]. [There are] few who in the time of
battle prefer death to an inglorious retreat, whether you speak of
the barbarians or the Romans.
41 Before the day of the engagement you must discuss with the
commanders under you what should be done, and what appears
useful to the general intention should become the basic [plan].
[You must] issue instructions to the commanders of the
dromons so that they are ready to carry out what has been
planned, unless indeed a contrary decision emerges after an
enemy attack. But then everyone [must] be prepared to watch
your dromon, so as to be able to receive the signal for what is
appropriate to do and then, when it has been given, perform
promptly what has been indicated.
42 You must certainly, strate2gos, have your own dromon with
soldiers picked from the entire stratos for size of body and
courage and skill and conspicuous for the rest of their
armament. And the dromon should stand out from all the others
by its size and speed since it is the head of the entire formation.
And you should set up the dromon of your Gloriousness [to be]
of the kind known as pamphylos*.
43 Similarly the other commanders under you, who have some
dromons under them, [must] choose men from these and have
them in their own [ships], so that they also are distinguished
from the rest. And all these, and the remaining [ships] [should]
watch your dromon and organize and arrange themselves by it
------------------------------
31
Cf. Appendix One, §§9.23 & 9.44.
502 APPENDIX TWO
to;n tou' polemou' kairovn, eij mh; a[ra e{terovn ti paravdoxon tw'n
bebouleumevnwn ajnafanh'/ kai; devetai meqovdou eJtevra".
mdV Ei\nai de; shmei'on iJstavmenon ejn tw'/ sw'/ drovmwni ei[te bavndon ei[te
flavmoulon ei[te ti e{teron eij" tovpon uJyhlovn, i{na di aujtou'
shmaivnontov" sou tiv dei' pravttein, eujqevw" ejpilambavnwntai tou'
dovxanto" e[rgou oiJ loipoiv, ei[te sumbavllein eij" povl emon crhv,
ei[te ajnacwrei'n ajpo; polevmou, ei[te ejx elivssein eij" kuvklwsin
kata; tw'n polemivwn, ei[te eij" bohqeivan kataponoumevnou mevrou"
sundramei'n, ei[te ajrgh'sai th;n ejlasivan, ei[te tacuvteron
ejlauvnein, ei[te e[gkrumma devon genevsqai, ei[te ajpo; ejgkruvmmato"
ejxelqei'n h] a[lla tina; kaq e{kasta ajpo; shmeivwn tou' sou'
drovmwno" a{panta uJpodevcesqai ajforw'nta" o{pw" dei' poiei'n.
meV Ouj ga;r duvnataiv ti" ejn toiouvtw/ kairw'/ ajpo; fwnh'" h] boukivnou
paraggevllein ta; devonta diav te to;n qovrubon kai; to;n tavracon
kai; to;n th'" qalavssh" h\con kai; to;n a[llon ktuvpon th'" te
sugkrouvsew"32 kai; kwphlasiva" tw'n dromwvnwn kai; pollw'/ plevon
th'" boh'" tw'n polemouvntwn.
mıV To; de; shmei'on uJposhmainevtw h] ojrqo;n iJstavmenon h] ejpi; dexia;
klinovmenon h] ejp ajristera; kai; ejpi; dexia; metaferovmenon pavl in
h] ejp ajristera; h] tinasssovmenon h] uJyouvmenon h] kamhlouvmenon
h] pantelw'" ejpairovmenon h] metatiqevmenon h] dia; th'" ejn aujtw'/
kefalh'" a[llote a[llw" fainomevnh" ajllassovmenon h] dia;
schmavtwn h] dia; crwmavtwn oi|ovn pote toi'" palaioi'" ejgivneto. 33
mzV En ga;r polevmou kairw'/ shmei'on ei\con th'" sumbolh'" ai{ronte"
eij" u{yo" th;n legomevnhn foinikivda: h\n de; to; legovmenon kame-
lauvkion ejpi; kontarivou uJyouvmenon, mevl an th;n crovan kai; a[lla
tina; kata; to;n o{moion trovpon uJpodeiknuvmena. Asfalevsteron de;
tavca dia; th'" sh'" ceiro;" ta; shmei'a uJpodeicqhvsetai.
mhV Kai; ou{tw" e[stw soi hJ ejnevrgeia, w\ strathgev, tw'n toiouvtwn
shmeivwn gegumnasmevnh w{ste pavnta" tou;" uJpo; se; a[rconta" o{soi
dromwvnwn hJgou'ntai e[cein th;n pei'ran ajsfalh' tw'n toiouvtwn
uJpodeigmavtwn kai; dia; tiv givnetai e{kaston kai; povte kai; pw'", kai;
------------------------------
32
Cf. Thucydides, Peloponnesian war, VII.70.6 (vol. 4, p. 140): “kai; to;n ktuvpon
mevgan ajpo; pollw'n new'n xumpiptousw'n e[kplhxivn te a{ma kai; ajpostevrhsin th'" ajkoh'" w|n
oiJ keleustai; fqevggointo parevcein.”.
33
In MS. A “oi|ovn pote toi'" palaioi'" ejgivneto.” is part of §47.
during the time of engagement, in case anything should happen

contrary to what had been planned and should require different
conduct.
44 [There must] be a signal placed on your dromon, either a
banner or a standard or something else in a high place, so that
when you signal what should be done, the rest can immediately
understand the action decided on, whether they should join in
the engagement, or withdraw from the engagement, or begin an
encirclement of the enemy, or hasten to the assistance of a
section in difficulties, or slow the rowing down, or speed up the
advance, or [whether] they should set up an ambush or come
out of ambush, or anything else; they should receive each and
every [command] from the signals of your dromon, noticing
how it is necessary to act.
45 For no one can give the necessary [orders] at such a time by
voice or by trumpet, because of the hullabaloo and confusion,
the noise of the sea and the other din from the collisions and
rowing of the dromons and, even more, the shouts of those
fighting.
46 The signal should indicate commands by being held upright, or
being inclined to the right or left, or being shifted to the right
again or to the left, or by being waved, or raised or lowered, or
completely removed, or having its position moved, or being
changed by having its “head” (kephale2*) sometimes made to
look different, or through patterns or colours, as used to be done
in the past.34
47 For in time of engagement they [the ancients] used to have a
signal for attack, raising on high the so-called phoinikis*.35 This
was the so-called kamelaukion*, raised on a pole, black36 in
colour, and some other [objects] displayed in a similar way. It is
very much safer [when] signals will be displayed by your hand.
48 Your technique, strate2gos, in these signals should be well
practised so that all the commanders under you who are in
charge of dromons are very experienced in these signs,37 and
why each is made, and when and how, and should not make
------------------------------
34
Cf. Maurice, Strate2g ikon, VII B.16 (pp. 260-62).
35
36
Note that Dain misread the manuscript at this point, reading “red”, ejr uqro;n
(erythron), for “black”, mevlan (melan). The kamelaukion was black not red, as some
have been misled by Dain to believe.
37
Cf. Appendix One, §8.
504 APPENDIX TWO
mh; diasfavllesqai, i{na peri; tau'ta kalw'" ejggumnasavmenoi e;n

kairw'/ creiva" e{toimoi gevnwntai pro;" to; gnwrivz ein aujta; kai;
pravttein suntovmw" ta; di aujtw'n keleuovmena.
mqV Th;n de; tw'n dromwvnwn paravtaxin ejn kairw'/ prosbolh'", ei[per, wJ"
ei[rhtai, tosauvth pavrestin ajnavgkh h] eujkovl w" th;n nivkhn ejlpiv-
zei", poihvsei" kaqw;" a]n sunivdh/" aJrmovdion ei\nai prov" te to;n
kairo;n kai; to;n tovpon kai; pro;" th;n tw'n polemivwn paraskeuh;n
kai; paravtaxin: ouj ga;r nu'n ejsti levgein ajsfalw'" peri; tw'n tovte
mellovntwn sumbhvsesqai.
nV Pote; me;n mhnoeidw'" oi|on sigmatoeidw'" eij" hJmikuklivou tavxin,
tou;" me;n a[llou" drovmwna" e[nqen kajkei'qen oi|on kevratav tina h]
cei'ra" kai; mavlisata ejn tw'/ a[krw/ proavgonta" tou;"
ijscurotevrou" kai; meivzona": ejn de; tw'/ bavqei tou' hJmikuklivou
oiJoneiv tina kefalh;n th;n sh;n ejndoxovthta w{ste; pavnta
periskopei'n ka; diatavttein, kai; dioikei'n kai; o{pou dei'
bohqeiva" ejpibohqei'n meq w|n a]n bouvl h/ eij:" tou'to aujto;
eujkairouvntwn. To de; sch'ma to; mhnoeide;" ginevsqw w{ste tou;"
ejmpivptonta" polemivou" e[swqen ajpokleivesqai th'"
kuklwvsew".
naV Pote; de; paratavx ei" ijsometwvpou" ta;" nau'" ejp eujq eiva" w{ste
creiva" kalouvsh" ejmpivptein toi'" polemivoi" kata; prwv/ran kai; dia;
tou' puro;" tw'n sifwvnwn katakaivein ta;" ejkeivnwn nau'".
nbV Pote; de; kai; eij" diafovrou" merivzesqai paratavx ei" ei[te duvo h]
trei'" kata; th;n posovthta tw'n uJpo; se; dromwvnwn. Kai; th'" mia'"
paratavxew" sumbalouvsh" hJ a[llh ejmpesei'tai kata; tw'n
polemivwn h[dh ejmpeplegmevnwn h] o[pisqen h] ejk plagivou kai; dia;
th'" bohqeiva" th'" ejpelqouvsh" kat aujtw'n ajpagoreuvsousin oiJ
ejnantivoi to; mavcesqai.
ngV Pote; de; kai; di ejgkruvmmato": ajpoplanwmevnwn ga;r tw;n pole-
mivwn kai; ejmpiptovntwn wJ" pro;" ojlivgou", ajnafane;n ajqrovw" to;
e[gkrumma kai; tavraxan aujtou;" to;n tovnon th'" ejnstavsew" aujtw'n.
ndV “Allote de; di ejl afrw'n kai; tacutavtwn dromwvnwn sumballov-
ntwn aujtoi'" kai; schmatizomevnwn fughvn, ejkeivnwn de; ejn th'/
diwvxei kopoumevnwn kai; biazomevnwn mevn, mh; katalambanovntwn
mistakes. Thus, being well practised in these, in time of need

they are ready to recognize them and do quickly what is ordered
by them.
49 In time of attack, if indeed, as has been said, there is a great
need [for one] or you expect an easy victory, you will organize
the formation of the dromons as you consider is suitable to the
weather and the topography and the preparation and formation
of the enemy. It is not possible now to give precise instructions
for what may happen in the future.
50 Sometimes [you should draw up] a crescent-shaped or sigma-
shaped [i.e., capital sigma “C”] formation in a semi-circle,38
with the rest of the dromons placed on one side and the other
[i.e., of the flagship] like horns or hands and making sure that
the stronger and larger [ships] are placed on the tip. Your
Gloriousness [should be positioned], like a head in the deep of
the semi-circle,39 so that you can observe and control and
oversee everything and where if help is needed provide
assistance with whatever [ships] you wish that are to hand for
this purpose. The crescent arrangement should be such that, as
the enemy attack, they are enclosed within the curve.
51 Sometimes you will form up the ships on an equal front in a
straight [line],40 so that, when the need summons, [you can]
attack the enemy at the prow and burn their ships with fire from
the sipho2nes.
52 Sometimes it [the fleet] should be divided into several
formations,41 either two or three according to the number of
dromons under you. When one formation has attacked, the
other falls on the enemy either at the rear or from the flank
when they are already engaged, and with these reinforcements
attacking them the enemy breaks off fighting.
53 Sometimes [you should] use an ambush. For when the enemy
are deceived and are attacking an apparently small [force], the
sudden and disturbing appearance of the ambushers will take
the heart out of their resistance.
54 On other occasions, when light and very fast dromons have
attacked the enemy and are pretending flight, and they [the
enemy] are wearied by the pursuit and exhausted and are not
------------------------------
38
39
40
Cf. Maurice, Ek tou' Maurikivou, §7 (p. 42); Appendix One, §§9.35-41.
41
506 APPENDIX TWO
de; tou;" feuvgonta", h] kai; tinwn ajpokoptomevnwn ajllhvlwn th'"

suneceiva", e{teroi sou drovmwne" a[kopoi kai; ajnapepaumevnoi
kata; tw'n kekopwmevnwn oJrmhvsante" nikhvsousin aujtouv" h[, eij
kai; ta; dunata; tw'n ejcqrw'n ploi'a parelqei'n ijscuvsa" ti", toi'"
ajsqenestevroi" ejpiteqh'/.
neV Pote; de; sumbalw;n kai; iJkanw'" ejk ceiro;" polemhvsa" tai'"
ejnantivai" nausi; tou' teleivw" kopwqh'nai tou;" enantivou",
ajpoplevzei"42 me;n tou;" drovmwna", eJtevrou" de; pavlin ejpipevmyei"
toi'" polemivoi" ajkopiavtou" toi'" kekopwmevnoi" kai; ejkluqei'sin
ajpo; th'" mavch", kai; ou{tw" th;n kat aujtw'n nivkhn ejrgavsh/: mavl ista
de; tou'to givnetai o{tan perissotevrou" aujto;" e[ch/" drovmwna"
uJpe;r tou;" polemivou".
nıV Pote; de; fugh;n prospoiouvmeno" meta; dromwvnwn tacinw'n pro;"
divwxin ejkkalevsh/ tou'" polemivou" kata; pruvmnan e[cwn aujtouv".
Kakei'noi oJrmhvsante" diwvkein dialuvsousi th;n tavxin aujtw'n. Kai;
ou{tw" suntovmw" ajnqupostrevya" diesparmevnoi" toi'" diwvkousi
mavlista kai; plei'on ejkeivnwn drovmwna" e[cwn ejpevlqh/" aujtoi'"
kata; prwv/ran: kai; h] kaq e{na h] kata; duvo ejpavgwn tou;" sou;"
drovmwna" tw'/ eJni; ploivw/ tw'n polemivwn nikhvsei" aujtouv".
nzV Prosbavllein de; polemivoi" crh; ejn naumaciva/ kai; o{tan tuvch/
aujtou;" nauagh'sai kai; o{tan ajpo; zavlh" diataracqevnte" ajtonhv-
swsin, h] ejn nukti; ejpelqovnta ejmprh'sai ta;" ejkeivnwn nau'", h] ejn
th'/ cevrsw/ ajscoloumevnwn, h] wJ" a]n hJ creiva kalevsh/ kai; aujto;"
ejpinohvsh/" ou{tw" kai; poihvsei" ta;" prosbolav".
nhV Poikivlh" ga;r ou[sh" th'" tw'n ajnqrwvpwn gnwvmh" ajduvnatovn tina
ta; mevllonta ejmpivptein ejn tai'" toiauvtai" paratavxesin h]
proginwvskein h] prolevgein a{panta: dio; oujde; ta;" kat aujtw'n
ajntiparatavx ei" ejn tw'/ parovnti lovgw/ dunatovn meqodeuvein, ajll
eij" th;n qeivan provnoian a{panta tau'ta ajnatiqevnai kai; devesqai
tou' Qeou' i{na ejn toi'" toiouvtoi" ojx evsi kairoi'" duvnataiv ti" kai;
bouleuvesqai kai; dianoei'sqai kai; pravttein ta; devonta.
nqV Polla; de; kai; ejpithdeuvmata toi'" palaioi'" kai; dh; kai; toi'" nevoi"
ejpenohvq h kata; tw'n polemikw'n ploivwn kai; tw'n ejn aujtoi'" pole-
mouvntwn: oi|on tov te skeuastovn pu'r meta; bronth'" kai; kapnou'
propuvrou dia; tw'n sifwvnwn pempovmenon kai; katakaivon aujtav.
------------------------------
42
ajpoplivzei" MS. A. We emend to ajpoplevzei" following Appendix Five, §52.
54 able to overtake [those] in retreat, either when some of them are

cut off from each other in the main [fleet], other fresh and
rested dromons of yours, setting on the exhausted ships, will
defeat them or, if some [dromon] is able to pass the powerful
enemy ships, it should attack the weaker [ones].
55 Sometimes, when you have attacked and fought the opposing
ships sufficiently at close quarters until the enemy are
completely exhausted, you will disengage the dromons and
send in other fresh [ones] against enemies who are weary and
downcast from battle. And thus you will gain the victory over
them. This can best be done when you have more dromons than
the enemy.43
56 Sometimes, when [you] are pretending to flee with fast
dromons you may incite the enemy to pursuit, having them at
your stern. They, having set off in pursuit, will break up their
formation.44 And so, turning round quickly, you should attack
them at the prow as they pursue in disorder, especially having
more dromons than they. And bringing up your dromons, either
singly or in twos, against a single enemy ship, you will defeat
them.
57 It is necessary to attack the enemy in battles at sea, both when
they happen to be shipwrecked and whenever they are
disheartened, having been scattered in a squall.45 Either burn
their ships, attacking by night or when they are occupied on
shore, or you should make an attack when the need arises and
when you can devise [it].
58 Since men’s opinion is varied, it is impossible for anyone either
to foresee or to foretell all that will take place in such
formations. Thus it is not possible in the present discussion to
deal with counter formations against them, but [one must] leave
all this to divine providence and pray God that in such moments
of acute crisis one is able to devise and invent and put into
practice what is required.
59 Many devices have been invented by men of old and especially
in recent times against enemy ships and those fighting in them;
such as the processed fire, which is expelled from sipho2nes with
thunder and propyra*, forefire46 smoke and sets them on fire.
------------------------------
43
44
A standard battle tactic. See Pryor, “Roger of Lauria”, p. 203.
45
Cf. Appendix Two [b], §5.
46
Cf. Appendix Five, §56.
508 APPENDIX TWO
xV. Kai; toxoballivstrai de; e[n te tai'" pruvmnai" kai; tai'" prwv/rai"
kai; kata; tw'n duvo pleurw'n tou' drovmwno" ejkpevmpousai sagivta"
mikra;" ta;" legovmena" muiva". Kai; qhriva e{teroi ejpenovhsan ejn
cuvtrai" kekleismevna kai; kata; tw'n ploivwn tw'n polemivwn
rJiptovmena: oi|on o[fei" kai; ejcivdna" kai; sauvra" kai; skorpivou"
kai; ta; o{moia touvtwn ijobovla: w|n suntribomevnwn ta; qhriva
davknousi kai; sumfqeivrousi dia; tou' ijou' tou;" polemivou" e[swqen
tw'n ploivwn.
xaV Kai; cuvtra" de; a[lla" ajsbevstou plhvrei" w|n rJiptomevnwn kai;
suntribomevnwn oJ th'" ajsbevstou ajtmo;" sumpnivgei kai; skotivz ei
tou;" polemivou" kai; mevga ejmpovdion givnetai.
xbV Kai; trivboloi de; sidhrai' rJiptovmenai ejn toi'" ploivoi" tw'n
polemivwn ouj mikra; luphvsousin aujtou;" kai; ejmpodivsousin pro;"
to;n kata; th;n w{ran ojfeivlonta ajgw'na.
xgV ÔHmei'" de; keleuvomen kai; puro;" skeuastou' gegemis-mevna"
cuvtra" ejpirrivptesqai kat aujtw'n kata; th;n uJpodeicqei'san
mevqodon th'" aujtw'n skeuasiva": w|n suntribo-mevnwn eujkovl w" ta;
ploiva tw'n polemivwn katakahvsetai.
xdV Crhvsasqai de; kai; th'/ a[llh meqovdw/ tw'n dia; ceiro;" ballomevnwn
mikrw'n sifwvnwn o[pisqen tw'n sidhrw'n skoutarivwn para; tw'n
stratiwtw'n kratoumevnwn, a{per ceirosivfwna levgetai, para; th'"
hJmw'n basileiva" a[rti kateskeuasmevna: rJivyousi ga;r kai; aujta;
to; skeuasto;n pu'r kata; tw'n proswvpwn tw'n polemivwn.
xeV Kai; trivboloi de; meivzone" sidhrai' h] ejn sfairivoi" xulivnoi" h|loi
ojxei'" ejmpephgmevnoi, stuppivoi" de; kai; eJtevra/ u{lh/ ejneilhmevnoi47:
‹a}›48 ejmpurisqevnta kai; kata; tw'n polemivwn ballovmena, ei\ta
pivptonta ejn toi'" ploivoi" dia; pollw'n merw'n ejmprhvsousin aujtav.
xıV Alla; eij kai; dia; to; sbevsai oiJ polevmioi th;n aujtw'n flovga
patevsousin aujta; oiJ plei'stoi tou;" povda" plhghvsontai kat
aujth;n th;n sumbolh;n tou' polevmou kai; ouj mikro;n e[stai toi'"
ejnantivoi" ejmpovdion.
xzV Dunato;n de; kai; diav tinwn geranivwn legomevnwn h[ tinwn oJmoivwn
ejpithdeumavtwn gammatoeidw'n49 kuvklw/ peristrefo-mevnwn h]
pivssan uJgra;n pepurwmevnhn h] skeuh;n h[ tina u{l hn eJtevran
ejpicuvsai toi'" polemikoi'" ploivoi" dia; tw'n dromwvnwn desmou-
mevnoi" tou' maggavnou strefomevnou kat aujtw'n.
xhV Dunato;n de; kai; oJlovklhron th;n nau'n ajnatrevyai tw'n pole-
------------------------------
47
ejneilhmevnoi, thus Dain: ejneilhmevna MS. A.
48
Thus Dain, as added to MS. A by Desrousseaux. MS. A does not have this.
49
gammatoeidw'n, thus Dain: gammatoeidw'" MS. A.
60 And there should be toxobalistrai* (bow-ballistae)50 at the

sterns and the prows and along the two sides of the dromon to
shoot the small arrows known as muiai* (flies). And others
have thought of putting poisonous creatures into jars and
throwing [them] into the ships of the enemy; such as snakes,
vipers, lizards, scorpions and other such venomous [creatures].
When [the jars] break, the creatures bite and destroy with their
venom the enemy in the ships.
61 And other jars full of unslaked lime; when thrown and broken,
the fume from the lime chokes and kills the enemy and causes
great confusion.
62 Iron caltrops thrown into the ships of the enemy will cause
them no little harm and will hinder [them] in the struggle in
which they should be engaged at that time.
63 We give instructions [that] jars full of processed fire, made
according to the usual method of construction, are to be thrown
at them. When they break, the ships of the enemy will easily be
set on fire.
64 You should also employ the other method, with small sipho2nes
throwing [i.e., the fire] by hand which are held behind iron
skoutaria* (shields) by the soldiers. These are known as hand-
sipho2nes and were recently invented by our Majesty. They also
throw processed fire into the faces of the enemy.
65 Also larger iron caltrops or sharp nails embedded in wooden
balls and wrapped round with tow and other matter, these when
ignited and thrown against the enemies, and then landing on the
ships, will set them on fire in many parts.
66 If the enemies stamp on the flame to extinguish [it], most of
them will injure their feet during the clash of battle and this will
be no small nuisance for the opposition.
67 It is also possible by means of some [things] called gerania*

(cranes) or some similar contrivances, shaped like a [capital
letter] gamma (i.e., a “G” shape), turning in a circle, to pour
either wet flaming pitch or the processed [fire] or anything else
into the enemy ships when they are coupled to the dromons
when the manganon is turning over them.
68 It is also possible to capsize an entire enemy ship if -- having
------------------------------
50
Cf. Maurice, Ek tou' Maurikivou, §3 (p. 41).
510 APPENDIX TWO
68 mivwn eja;n pleura;n para; pleura;n dhvsa" aujth;n tw'/ drovmwni,

kai; tw'n polemivwn ejpi; e}n mevro", wJ" e[qo" e[cousi, pro;" th;n ejk
ceiro;" mavchn sundramovntwn kai; dokouvntwn ejpakoumbivzein
to; eJautw'n ploi'on tw'/ drovmwni, ejpevlqh/ me;n e{tero" drovmwn
kata; th'" pleura'" th'" ejn th'/ pruvmnh/ th'" polemiva" kai; tauvthn
wjqhvsh/ sfodrw'" th'/ sugkrouvsei, kai; oJ Jme;n drovmwn dunhqh'/
luvsa" eJauto;n tou' desmou' uJpocwrh'sai mikro;n w{ste mh; ei\nai wJ"
ajkouvmbisma th'" polemiva", barhvsh/51 de; oJ e{tero" drovmwn pavsh/
dunavmei, pavntw" ajnatrevyei su;n aujtoi'" toi'" ajndravsi th;n
polemivan nau'n. Dei' de; kanonivsai to;n desmo;n mh; pavntw" kat
ijsovthta gevnesqai, ajlla; mikro;n ajf ei'nai gumnav tina pleura;
kata; pruvmnan th'" polemiva", di w|n ejmpesw;n oJ drovmwn wjqhvsei
pro;" th;n ajnatroph;n tw'n polemivwn th;n nau'n.
xqV Pro;" touvtoi" kai; to; nu'n hJmi'n ejpinohqevn, w{ste ajpo; th'" kavtw tou'
drovmwno" ejlasiva" dia; tw'n ojpw'n h[toi ajpo; truphmavtwn tw'n
kwpivwn ejkferovmena mevnaula katasfavttein tou;" polemivou",
tw'n panuv moi ajnagkaivwn dokei'.
oV Alla; kai; e{teron touvtou ajnagkaiovteron, ei[ ge ceirw'n eujf uw'n
ejpituvch/, to; dia; th'" kavtwqen tou' drovmwno" ejl;asiva" th'/
uJpodeicqeivsh/ meqovdw/ di ojph'" paraskeuavsei plhsqh'nai u{dato"
th;n nau'n tw'n polemivwn.
oaV Eijsi; de; kai; e{tera toi'" ajrcaivoi" ejpinohqevnta ejn tw'/ plwi?mw/
polevmw/ ejpithdeuvmata, kai; e[ti de; ejpinohqh'nai dunavmena, a{per
ejn tw'/ parovnti gravf ein dia; th;n suntomivan ajnoivkeion hJghsavmeqa,
tina; de; kai; ajsuvmfora dia; to; mh; faulivz esqai toi'" polemivoi" kai;
ma'llon ejkeivnou"52 crh'sqai aujtoi'" kaq hJmw'n. Ta; ga;r
strathghvmata a{pax katanohqevnta duvnantai ajntistrathgei'sqai
kai; katameqodeuvesqai para; tw'n polemivwn: ajll e{kaston to;
ejpinohqe;n mevcri th'" pravx ew" e[cein ejn musthrivw/.
obV En de; tw'/ biblivw/ tw'n ajrcaivwn taktikw'n kai; strathghmavtwn
zhtw'n ti" euJrhvsei kai; ta; touvtwn pleivona: ouj ga;r dunatovn, wJ"
ei[rhtai, pro;" e{kasta ta; ejmpivptein mevllonta dia; to; a[peiron
aujtw'n gravfein, ajlla; ta; iJkanav.
ogV Plh;n kefalai'on eijpei'n, e{stwsan oiJ drovmwne" ejxwplismevnoi
teleivw", ajpov te stratiwtw'n ajndreivwn kai; ejk ceiro;" mavcesqai
dunamevnwn kai; tw'/ th'" yuch'" parasthvmati tolmhrw'n kai;
pepaideumevnwn kai; gegumnasmevnwn: ou|toi de; e[stwsan kaqw-
------------------------------
51
barhvsh/, thus Dain: barhvsei MS. A.
52
ejkeivnou", thus Dain: ejkeivnoi" MS. A.
68 coupled it side by side to the dromon, and the enemy rush to

one side, as is their habit, to engage in hand-to-hand-fighting
and expect their own ship to lay against the dromon -- another
dromon were then to run at the side of the enemy vessel
towards the stern and strike it hard as they collide, and if the
one (first) dromon should be able to free itself from the
coupling and back off a little so that it is not laying against the
enemy, and if the other (second) dromon were to weigh down
with all vigour, it will capsize the enemy ship and her crew
completely. You should organize the coupling so that it does
not hold the [enemy] ship evenly but leaves at the enemy ship’s
stern some of the sides a little exposed, where the dromon will
be able to attack and exert pressure to capsize the enemy ship.
69 In addition, [there is] the [technique] we have recently devised
so that when pikes are thrust from the lower bank of the
dromon through the holes or trype2mata* (oarports) of the oars,
they slaughter the enemy; this seems to me especially useful.53
70 But there is another [technique] even more useful than this, if it
falls to experienced hands, [and that is] when the enemies’ ship
is filled with water through a hole made by the dromon’s lower
oar-bank by the usual method.54
71 There are other devices for naval warfare invented by the
ancients and [others] that can still be invented, which we have
considered it inappropriate to describe at present in summary.
And [there are] some which are inadvisable [to mention] since
they may be taken over by the enemy, and indeed they may use
them against us. For once stratagems have been invented, a
counter-stratagem and defence can be devised by the enemy.
Every [scheme] once invented [should be] kept secret until it is
carried out.
72 Anyone who looks into the book on ancient tactics and
strategies will find more on these [matters]. For it is impossible,
as has been said, to write about every future [event] that will
happen, but [this is] sufficient.
73 But to mention the main point, the dromons should be
completely armed, with brave soldiers capable of fighting at
close quarters, and bold in their mental attitude, and trained and
exercized. These should be armed with the weapons with which
------------------------------
53
Cf above p. 405.
54
512 APPENDIX TWO
plismevnoi o{ploi" oJpoivoi" kai; oJ ejn th'/ xhra'/ stratiwvth"

oJplisqh'nai diwvristai, dhlonovti katavfrakto". Kai; ou{tw
pavnte" oiJ th'" a[nw ejl asiva" oJplisqhvsontai.
odV Pro;" de; th;n tw'n ejcqrw'n poiovthta tw'n posovthta tw'n ploivwn,
kai; aujtov", w\ strathgev, diaskeuavsei" tou;" drovmwna" kai; kaqo-
plivsei" o{pw" mh; ejlavttona strato;n e[ch/ oJ hJmevtero" drovmwn tou'
polemivou, o{sti" mavlista eij" i[shn mavchn ejlqei'n eJtoimavzetai dia;
tou' eij" ajllhvlou" desmou', 55 a;ll eij dunato;n kai; pleivona: ajmfo-
tevrwn ga;r ajndreivw" macomevnwn, oiJ pleivone" uJpernikhvsousin.
oeV Ea;n ga;r sunora'/" e[cein tou;" polemivou" ploi'a pleivona strato;n
uJpodecovmena, oujsiwvsei" kai; aujto;" tou;" i[sou"56 drovmwna" ejn
plhvqei. Eklevxh/ de; ajpo; pavntwn tou;" ajrivstou" kai; ejx aujtw'n
ejxoplivsei" th;n ajrkou'san duvnamin dia; dromwvnwn teleivwn kai;
ijscurotavtwn: w{ste eij ou{tw tuvch/ h] tw'n duvo to;n strato;n eij" e{na
ejmbibavsh/" h] ejk pavntwn ejpilevx h/ tou;" ajrivstou", wJ" ei;rhtai: kai;
genhvsontai a[cri kai; diakosivwn stratiwtw'n h] kai; pleivone"
kata; drovmwna eJna, wJ" a]n kai; tw'/ plhvqei kai; tw'/ megevqei tw'n
dromwvnwn kai; th'/ eujtuciva/ tw'n stratiwtw'n ejpikratevstero" tw'n
polemikw'n ploivwn genovmeno", su;n qew'/ th;n kat aujtw'n nivkhn
ajpolavbh/".
oıV Dei' dev se kai; mikrotevrou" ejxoplivzein drovmwna" kai;
ejlafrotevrou" tw'n sunhqw'n, w{ste kai; diwvkonta"
katalambavnein tou;" polemivou" kai; diwkomevnou" mh;
katalambavnesqai kai; touvtou" e[cein ejn kairw'/ th'" aJrmozouvsh"
aujtoi'" creiva", w{ste duvnasqai aujtouv" h] kakovn ti poihvsai tou;"
ejcqrou;" h] mh; paqei'n ti kako;n par aujtw'n.
ozV Mikrou;" de; kai; megavlou" drovmwna" kata; th;n poiovthta tw'n
polemivwn ejq nw'n kataskeuavsei". Ouj ga;r oJ aujtov" ejstin stovlo"
tw'n ploivwn tw'n te Sarakhnw'n kai; tw'n legomevnwn ÔRw'" boreivwn
Skuqw'n. OiJ me;n ga;r Sarakhnoi; koumbarivoi" crw'ntai meivz osi
kai; ajrgotevroi", oiJ de; oi|on ajkativoi" mikroi'" kai; ejlafrotevroi"
------------------------------
55
desmou', thus Dain: devsmou" MS. A.
56
i[sou", thus Dain, following Desrousseaux: sou;" MS. A.
the land soldier is decreed to be armed, that is, [as a]

kataphraktos. All those in the upper oar-bank should be armed
like this.
74 You yourself, strate2gos, will equip the dromons to match the
quality of the enemy and the quantity of [their] ships and will
arm them so that our dromon has a stratos no less than [that of
the dromon] of the enemy, [and], indeed, one that is prepared
for an even fight when coupled to each other, but if possible
greater; for, when both sides fight bravely, the greater will be
victorious.57
75 If you realize that the enemy has ships with a greater stratos,
you yourself will ousia*58 an equivalent number of dromons.
From all these [ships] you will select the best and from these
you will arm an adequate force of effective and very strong
dromons. Thus, if this is [what] happens, you either combine
the crew from two [ships] into one or, as has been said, you
select the best from all [the crews]. There should be up to two
hundred, or more, soldiers in one dromon, so that, being
superior to the enemy ships in both number and size of the
dromons and in the good fortune of the soldiers, with God you
will achieve victory over them.
76 You must also arm dromons [which are] smaller and lighter
than the usual so that, when pursuing they overtake the enemy,
or, when being pursued, they are not overtaken, and [you
should] have these at a time of appropriate need for them, so
that they can either inflict some damage on the enemy or not
suffer damage from them.59
77 You will equip small and large dromons according to the
quality of the enemy nation. For the fleet of ships of the
Saracens is not the same as that of the so-called Russians,
northern Skythians.60 The Saracens use larger and slower
koumbaria,61 while the Skythians use akatia*, which are small,
------------------------------
57
58
Note the verb oujsiwvsei", the verbal action of providing an oujsiva. The verb
oujsiovw thus meant to provide a ship with a crew. See also Appendix Five, §68.
59
60
Skythians, from the ancient people known to the Greeks and Romans. A generic
term used by Byzantines for peoples to the North outside the frontiers of the Empire.
61
“Koumbavrion” was a transliteration into Greek from Arabic. The original Arabic
word was most probably qunba2r (pronounced qumba2r), which was used in documents
of the Cairo Geniza for a large sailing ship. See Goitein, Mediterranean society, pp.
306, 331, 480 n. 6. Christides has suggested that the Arabic may have been marqib
514 APPENDIX TWO
kai; gorgoi'", oiJ Skuvqai: dia; potamw'n ga;r eij" to;n Eu[xeinon
ejmpivmtonte" povnton ouj duvnantai meivzona e[cein ploi'a.
ohV Kai; tau'ta me;n peri; paratavxewn eijrhvsqw. ”Otan de; ajpalla-
gh'nai bouvlh/ th'" mavch", mhnoeidw'", wJ" ei[rhtai, th;n paravtaxin
tw'n dromwvnwn poihvsa" ou{tw" uJpocwrhvsei" dia; to; ajsfale;"
ei\nai to; toiou'ton sch'ma ejn tai'" toiauvtai" kai; proovdoi" kai;
uJpocwrhvsesin, wJ" marturou'si tine" tw'n palaiw'n touvtw/ tw'/
trovpw/ crhsavmenoi.
oqV Meta; de; th;n luvsin tou' polevmou devon se, w\ strathgev, ta; wJ"
eijko;" krathqevnta ajpo; tw'n polemivwn lavfura ejx i[sou
diamerivzein toi'" stratiwvtai" kai; ajristopoiei'n kai;
filofronei'sqai aujtouv", kai; tou;" me;n ajristeuvsanta" kai;
dwrew'n kai; timw'n ajx iw'sai, tou;" dev ajnavxiovn ti stratiwvtou
poihvsanta" ejpitimh'sai deovntw".
pV Givnwske dev, w\ strathgev, o{ti plh'qo" dromwvnwn ajnavndrou"
ejcovntwn stratiwvta" oujde;n ijscuvei, oujd a]n kai; pro;" ojlivgou"
machvswntai tou;" ejnantivou" ajndreivou" kai; eujyuvcou": ou[te ga;r
plh'qo" ajndrw'n kata; ojlivgwn ijscuvsei eij mh; kai; th'/ proqumiva/ kai;
th'/ oJplivsei stratiw'tai ajlhqei'" ajpodeivknuntai. Tiv ga;r oujk
ejrgavsontai deino;n kai; ojlivgoi luvkoi pro;" polla;" ciliavda"
poimnivou…
paV Dio; crhv se sunora'n a{panta meta; ajkribeiva" pavsh" ta; tw'n
ejcqrw'n wJ" diavkeintai kai; ou{tw" thvn te tw'n dromwvnwn
kataskeuh;n kai; th;n tw'n stratiwtw'n o{plisin kai; to; plh'qo"
aujtw'n kai; to; mevgeqo" kai; ta; a[lla ejpithdeuvmata aJrmodivw" kata;
tw'n ejnantivwn paraskeuavz ein.
“Ecein de; kai; mikrou;" kai; tacei'" drovmwna" ouj pro;" povlemon
ejxwplismevnou", ajlla; pro;" ta;" bivgla" kai; ta; mandavta kai; ta;"
a[lla" ajpantwvsa" oJmoivw" creiva": kai; e[ti tav te monhvria
legovmena kai; ta;" galeva", plh;n kai; aujtou;" ejnovplou" dia; ta; wJ"
eijko;" kai; kata; tuvchn ejmpivptonta.
pbV Kai; se; de; aujto;n dia; pavntwn ei\nai dei' spoudai'on kai; gennai'on
kai; ajtavracon kai; ojxu;n ejn tai'" ejnagkaivai" mavlista tw'n
pragmavtwn ejgceirhvsesiv te kai; pravx esin, i{na kai; Qew'/
eujavresto" kai; th'/ hJmetevra ejk Qeou' basileiva/ eu[crhstov" te kai;
dovkimo" ajnafanei;" strathgo;" ajmfotevrwqen kerdhvsh/" ta;" ajx iva"
tw'n povnwn ajmoibav", ejk Qeou' me;n misqou;" ajqanavtou" uJpe;r th'"
aujtou' klhronomiva" ajgwnizovmeno", ejx hJmw'n de; kai; tima;" kai;
------------------------------
kabı3r, “large ship”, but this seems much less likely. See Christides, Conquest of Crete,
p. 66.
lighter, and fast, as they cannot have larger ships when raiding
down rivers to the Black Sea.
78 That is enough about formations. When you wish to disengage
from a battle, having, as has been said, drawn the formation of
the dromons into a crescent, you will withdraw in this way
since this arrangement is safe in advances and retreats of this
kind,62 as some of the ancients indicate [by their] having used
this method.63
79 When the engagement has ended, strate2gos, you should divide
the spoils that have been acquired from the enemy as is usual
equally among the soldiers and praise them and make much of
them, and reward the outstanding soldiers with gifts and
honours, and you should penalize accordingly those whose
behaviour has been unbecoming to a soldier.
80 You should appreciate, strate2gos, that a number of dromons
with cowardly soldiers achieves nothing, not even when
fighting a few opponents [if these are] brave and of good heart;
neither will a number of men achieve anything against a few
unless they prove to be true soldiers in energy and arms. Will
not a few wolves do great damage to many hundreds of
thousands of sheep?
81 Therefore you should observe with great accuracy the enemy’s
situation and then prepare the equipment of the dromons, the
armament of the soldiers, their number, the size [of the ships],
and other needs in a manner appropriate to the opposition.
Equally, [there is] need to have small and fast dromons
[which are] not armed for battle but can be used as scouts, for
messages and other similar purposes. And also what are known
as mone2reis (monoremes) and also galeai, except that they
should be armed against normal eventualities.
82 You should be keen, valiant, calm, and alert throughout

everything, particularly in the inevitable conflicts and periods
of action; thus you may be pleasing to God and serve our
Majesty under God and be a renowned strate2gos and you will
gain due recompense for your labours from both, an eternal
reward from God for your struggles on behalf of his dominion,
and honours and their attendant gifts from myself, when you
------------------------------
62
63
See Thucydides, Peloponnesian war, III.78.3 (vol. 2, p. 136), and cf. VII.70.4
(vol. 4, p. 138).
516 APPENDIX TWO
dwrea;" ta;" proshkouvsa", mh; yeudovmeno" th;n klh'sin, ajll

ajlhqh;" strathgo;" kai; w]n kai; kalouvmeno". Tosau'ta kai; peri
naumaciva" wJ" ejn sunovyei metrivw" eijrhvsqw.
APPENDIX TWO [b]
LEO VI, EK TOU KUROU LEONTOS TOU BASILEWS

Ek tou' Kuvrou Levonto" tou' basilevw"

165 ÔIstorhvsw soi e[ti kai; e{teron nautikou' stovlou strathvghma:
o{tan ga;r eij" tovpou" limevna" mh; e[conta" kai; yammwvdei" th;n
ajpovbasin mevllh/" poihvsasqai, eja;n ou{tw tuvch/ ejn kairw'/
nautikh'" strathgiva", savkkou" pollou;" gemivsa" a[mmou kai; toi'"
scoinivoi" prosdhvsa" ajpo; eJkavstou drovmwno" ejkkremavsei" tou;"
ajrkou'nta" oiJonei; sidhra;" ajgkuvra", kai; ou{tw" to;n legovmenon
pelagolimevna poihvsa" eujkovlw" kata; to;n tovpon nukto;" ejpelqw;n
th;n bebouleumevnhn soi katadromh;n poihvsei".
266 Polemivou pote; nautikou' stovlou meta; oijkeiva" dunavmew"
nautikh'" uJpocwrw'n strathgo;" mhnoeidh' paravtaxin poiouvmeno"
uJpostrefevtw plevwn kata; pruvmnan kai; ou{tw" bouleuevsqw
ajpocwrivz esqai tw'n polemivwn: kai; ga;r ouj feuvgwn, ajlla;
fugomacw'n, eJtoivmou" e{x ei ta;" nau'" kai; pavl in ejpelqei'n toi'"
polemivoi" kata; prwv/ran. Ei[ ge kai; touvtou creiva gevnhtai, ta;"
prwv/ra" e[cein pro;" aujtouv": kai; ga;r oujde; qarrhvsousin ejn tw'/
koilwvmati eijselqei'n th;n kuvklwsin uJforwvmenoi.
------------------------------
64
Edited from a microfilm of folios 331r-v of the tenth-century manuscript Milan,
Biblioteca Ambrosiana, MS. B 119-sup. [gr. 139]. A text was published in Dain,
Naumachica, pp. 35-8.
65
Constitution XX, §196 of Leo VI, Taktika (PG).
We believe that we have translated the Greek accurately here; however, we have
no idea what the emperor supposed this paragraph to mean. He appears to have
thought that if a fleet made landfall on an open beach where there was no harbour into
which to put, that one would then need to construct a “sea-harbour”, pelagolimevn
(pelagolimen), by lowering sandbags to the sea bed to hold the ships in position. Why
one could not use for this purpose the multiple anchors that all medieval ships carried,
escapes us. Then, the emperor appears here not to have appreciated that dromons
could simply be beached, thus negating the need for any such “sea-harbour”;
although, elsewhere he did appreciate that fleets could be beached. In any case, why
constructing such a “sea-harbour” would contribute to making a raid successful
appears to be entirely obscure.
66
have not fallen short of your calling but are a true strate2gos in
both name and deed. Enough has now been said sufficiently on
naval warfare in this brief survey.
APPENDIX TWO [b]
LEO VI, FROM THE LORD LEO, THE EMPEROR,

EDITION AND TRANSLATION
From the Lord Leo, the Emperor

1 I will tell you now of yet another stratagem for a naval fleet.
When you intend to make a landing in sandy places which do
not have harbours, should this happen during your period of
naval command, having filled many sacks with sand and having
tied [them] with ropes, you will hang a sufficient number like
iron anchors from each dromon; and having in this way
constructed what is known as a ‘sea harbour’, attacking the
place by night, you will easily carry out your planned raid.
2 A strate2gos retiring on some occasion before an enemy fleet
with his own naval force, making a crescent formation should
withdraw, sailing by the stern [i.e., backing water], and should
plan to disengage from the enemy in this way; for by not
fleeing but making a fighting retreat, he will have his ships
ready to attack the enemy once again from the prow. If there is
need of this, [you should] have the prows towards them, for
they will not have the courage to enter the hollow, suspecting
encirclement.
518 APPENDIX TWO
367 »Wn e{neka, eij68 tovpwn h] povl ewn ejkpevmpein mevllei" nautiko;n
stovlon, kruvptein se dei' kai; tou;" tovpou" kai; ta;" povl ei" w{ste
mhdevna prognw'nai pou' mevllei givnesqai oJ katavplou". “Entalma
de; gravya" kai; sfragisavmeno" aujto; ajsfalw'" ejpivdo" tw'/
kaqistamevnw/ para; sou' nauavrcw/, i{na kata; to; pevlago" ejxelqw;n
tovte luvsh/ th;n sfragi'da kai; mavqh/ pou' mevllei poreuvesqai:
ou{tw" ga;r poihvsa" lavqh/" tou;" polemivou".
469 Tw'/ de; ajgaqw'/ strathgw'/ crevo" ejsti;n pro;" pa'n e[q no" aJrmozomevnw/
diafovrou" pro;" e{kaston ta;" strathgiva" ejpinoei'sqai. Eij dev
pote kai; nauarciva" kairo;" ejpisth'/, ajkivndunon th;n tou' stovlou
tavxin diafulavx ei e[mpeiro" w]n th'" tou' ajevro" kinhvsew". Kai; ta;"
oujragiva" dev, h[toi tou;" ojpisqofuvlaka", eujtavktw" sunagagei'n,
i{na mh; uJpo; tovpou h] uJpo; zavl h" qalassiva" h] uJpo; polemivwn
ajnagkazovmenoi fqeivrwntai.
570 En toi'" nautikoi'" mavcai sunavptontai h] o{tan oiJ polevmioi
nauaghvswsin, h] o{tan uJpo; ceimw'no" talaipwrhqw'sin.
------------------------------
67
68
eij thus Dain: h] MS. A.
69
Epilogue, §§44, 45 of Leo VI, Taktika (PG).
70
Epilogue, §47 fin. of Leo VI, Taktika (PG).
3 For these reasons,71 if you intend to send a naval fleet from72

places or cities you should conceal both the places and the cities
so that no one can know in advance where the landing is to be.
When you have written your order and sealed it securely, hand
it over to the squadron commander appointed by you so that
when he is on the open sea, he will then break the seal and learn
where he is to make for. By doing this you will escape the
enemy’s notice.
4 The good strate2gos needs to adapt himself to every nation and
to devise different strategies for each Should the weather ever
put a stop to a naval expedition, he will keep the fleet’s
formation intact, since he is experienced in the wind’s
movement. And he should assemble the rear guard, that is, the
last ranks, in good order, so that they are not scattered by the
place or a squall at sea, or by harrassment of the enemy.
5 In naval matters battles are joined either when the enemy have
been shipwrecked or when they are in difficulties in a storm.73
------------------------------
71
“For these reasons” referred to cautious conduct and the need to gather
intelligence when dealing with representatives of foreign or enemy powers. See Leo
VI, Taktika (PG), Constitution XX, §219.
72
The sense of this passage demands “to” rather than “from” at this point, in spite
of the fact that the Greek is quite clear. The edition of the clause in Leo VI, Taktika
(PG) at Constitution XX, §220 is the same, with the minor emendation of h] for eij in
accordance with MS. A and others, which makes no change to the sense. However,
like us, Meursius obviously sensed that an emendation was necessary and in his Latin
translation wrote: “Ad quae loca vel civitates classem emissurus es nemini indicare
oportet, ...”. See Leo VI, Taktika (PG), col. 1075.
73
APPENDIX THREE
ANONYMOUS, NAUMACIKA SUNTACQENTA PARA BASILEIOU

PATRIKIOU KAI PARAKOIMOUMENOU, EDITION AND
TRANSLATION1

------------------------------
1
Edited from a microfilm of folios 339-42 of the unique tenth-century manuscript
Milan, Biblioteca Ambrosiana, MS B 119-sup. [gr. 139], referred to by Dain, and
hereafter here also, as MS. A. A text was published in Dain, Naumachica, pp. 61-8.
Cf. above pp. 183-6
Dain’s edition is most unsatisfactory; see also the comments of Mazzuchi,
“Basilio Parakimomenos”, p. 294, n. 78. His misreadings and omissions have been
corrected tacitly here. However, because Dain’s text has been cited so often in so
many different contexts by maritime historians, for the sake of convenience we have
retained his numbering of the sections and paragraphs.
We have not indicated the presence or absence of apostrophes, iota subscripts, or
the enclitic usage.
As far as maritime historians are concerned, the most serious errors in Dain’s
edition, which have been corrected here, occur in §§2.1, 2.3, 2.5, 2.10, 2.13, 3.1, 4.1
and after 7.5.
522 APPENDIX THREE
Naumacika; suntacqevnta para; Basileivou patrikivou kai;

parakoimoumevnou
Aujsonivwn sofivh" dedidagmevno" e[xoca e[rga
Kai; stivca" hjde; favlagga" ijd ajrrageva" paratavx ei"

ÔOplitw'n prulevwn, kraterw'n pavlin iJppokorustw'n
ÔIstorikai'" selivdessin uJp eujgenevwn basilhvwn,
Deicqei;" Aujsonivwn stratih'" panupevrtato" ajrcov", 2
Cambda'n hJtthvsa" Aravbwn gevno" ejx enarivz ei",

Nhw'n ‹d›3 wjkupovrwn dedahvmenai ei[ pote bouvl h/,4
“Andica naumacivh"5 o[fr eujkleva mhvseai e[rga,
Bivblou th'sde, fevriste, nohvmata pavnta kat ai\san
“Ommasi soi'" skopivaze kai; ejn fresi; bavlleo sh'/sin.

“Enqen dh;, Basivl eie, pevdon Krhvth" ajlapavxei"
Kai; geneh;n ojlevsei" Karchdonivwn megaquvmwn.
Prooivmion
1 Ei[per kai; a[llo ti tw'/ bivw/ lusitelei'n oi|de kai; sunista'n

politeivan kai; meivzona to;n eJl ovmenon tw'n ejcqrw'n ajpodeiknuvnai,
oi\mai mhdeno;" ajpoleivpesqai touvtwn h] ta; deuvtera fevrein to;
peri; naumacivan gumnavzesqai kai; peri; tauvthn ejnascolei'sqai
kai; tauvth/ plevon tw'n a[llwn semnuvnesqai.
2 Kalo;n me;n ga;r kai; to; ejn hjpeivrw/ taktikoi'" ejggumnasavmenon

provteron paratavttesqai kai; katagwnivz esqai tou;" ejcqrou;" kai;
ajfeidw'" touvtou" diwvkein nw'ta didovnta" kai; tou;" prostucovnta"
aiJrei'n kai; tou;" porrwtavtw ejlpivz ein aiJrhvsesqai kai;
katalabei'n: kavllion de; o{sw/ kai; duscerevsteron to; tai'" nausi;n
ejformei'n te kai; naulocei'n, ajnqormei'n te kai; ajnteformei'n kai;
ta;" tw'n ejcqrw'n nh'a" katalambavnein te kai; perikuklou'n.
------------------------------
2
ajrcov", thus Dain: a[rco" MS. A.
3
d was added to MS. A in Brunck, Analecta, vol. 3, p. 277 (no. 896).
4
Bouvlh/, thus Dain: bouvlei MS. A.
5
naumacivh", thus Dain: naumaciva" MS. A.
NAUMACIKA SUNTACQENTA PARA BASILEIOU 523
Naval warfare, commissioned by Basil, the patrikios* and

parakoimo2menos*
Having been instructed in the outstanding works of the wisdom

of the Ausonians 6
and in files and phalanxes and unbroken battle-lines
of close-massed hoplites, indeed of mighty chariot marshalls,
in the pages of history by well-born kings,
having been shown to be the all supreme leader of the
Ausonians’ army,
having defeated Chambdan,7 you despoil the Arab race,
and if ever you wish to learn about swift-moving ships
so that, far from war at sea, you may recall doughty deeds,
gaze with your eyes, best [of men], on all the thoughts
[contained] duly
in this book and cast them within your mind.
Then indeed, Basil, you will sack the plain of Crete
and destroy the race of great-hearted Carthaginians.
Preface
1 If there is anything else that can be beneficial to life and

supports the state and demonstrates that he who choses it is
greater than the enemy, I think [that] to be concerned with
naval warfare and to be preoccupied with this and devoted to it
more than anything else falls behind none of these nor takes
second place.8
2 For it is good that the [force] that has been trained previously in
tactics on land should be drawn up and contend with the enemy
and pursue them relentlessly as they turn their backs and seize
those whom they chance upon and expect to seize and capture
those further off. But how much better, and more difficult, to
attack (ephormein) with ships and naulochein, and anthormein
------------------------------
6
Ausonians: mythological early inhabitants of Ausonia (Italy). In Byzantine
political rhetoric “Ausonian” had reference to the most ancient, Roman layer of
Byzantium’s classical heritage. Here the word simply meant “Byzantine”.
7
Sayf al-Dawla ‘Alı3 I (H4amda2n id emir of Aleppo, 945-67).
8
A syntactically confused sentence whose import is the hardly profound thought
that the study of naval warfare is of great importance.
524 APPENDIX THREE
3 Sumbaivnei ga;r ejn me;n tai'" kat h[peiron paratavx esin pleivona"
katevcesqai tou;" ceivrona", ejlavttona" de; h] kai; pavnu bracei'"
tou;" kreivttona", ejkavstou th;n oijkeivan kai; movnon swthrivan
mhcanwmevnou te kai; ejpithdeuvonto": ejn de; tai'" kata; qavlattan
a{ma tw'/ ajgennei' kai; oJ gennai'o" pesei'tai kai; kataduvsei ejn tw'/
buqw'/ h] kai; uJpofqavseie katalhfqei;" mhde;n plevon tou' ajsqenou'"
ejndeixavmeno".
4 Kai; tau'tav moi divdw" neanieuvesqai oJ strathgikwvtato" su; kai;
katorqwvmasi pa'si kosmouvmeno", oJ krataio;" qeravpwn tou'
krataiou' basilevw" hJmw'n, oJ tou' ajsfalou'" ajsfalh;" uJphrevth"
kai; tou' anjdreivou ajndrei'o", oJ toi'" kat h[peiron ajgwnivsmasi kai;
aujtou;" basileva" eujfravna" eujmenw'" e[conta" kai; pa'n to;
uJphvkoon galhvnh" kai; carmonh'" kaqupodeivx a" mestovn, oJ ta; tw'n
a[llwn aJpavntwn ajndragaqhvmata tw'n te nu'n o[ntwn tw'n te pavl ai
gegenhmevnwn tapeinwvsa" kai; kavtw qevmeno", kai; ta; kata;
qavlassan ei[ pou dehvsoi toi'" kat h[peiron deivx wn parovmoia.
5 Peivqomai gavr, ou{tw" e[cw, kai; ou{tw logivzomai kai; toiauvtai"

ejlpivsin ejpaivromai. Toi'" soi'" ga;r eJkavstote ta; tw'n a[llwn
paratiqei;" kai; parexetavzwn w{sper tino;" e[rga paido;" uJp
ajndro;" teleivou euJrivskw hJttwvmena kai; ajpoleipovmena kai; oi|on
diapaizovmena. 9
6 OiJ me;n ou\n a[lloi pavnte" o{soi megavl a katwvrqwsan to;n a[neton
eujqu;" aiJrou'ntai bivon kai; eujdiavcuton kai; deutevroi"
ajpoknou'sin ejgceirei'n10 katorqwvmasin, i{na mh; pollavki" toi'"
u{steron ajtuchvsante" kai; th;n tw'n protevrwn uJpovlhyin
ajmaurotevran paravscwsin. Aujto;" de; ajei; tw'n prwvtwn ejktelei'"
ta; deuvtera meivzona kai; th'/ touvtwn uJperbolh'/ pavntwn
periginovmeno" mhcana'/ tw'/ megevq ei tw'n deutevrwn ajpokruvptein
ta; prw'ta.
7 Kai; pavscein taujto;n sumbaivnei toi'" eujnoou'siv te kai; dus-

menaivnousin: hJttwvmenoi ga;r ejpivsh" ajgavllontai te kai;
caivrousin, oiJ me;n o{ti th'/ sh'/ promhqeiva/ nikw'si te kai; neni-
khvkasin, oiJ de; o{ti mh; par a[llwn hJtthvq hsan h] para; sou' tou'
pavnta" nikw'nto" eujbouliva/ kai; rJwvmh/: di w|n ajnavgkh touvtou"
------------------------------
9
diapaizovmena, thus Dain: diapezovmena MS. A.
10
ejgceirei'n, thus Dain: ejnceirei'n MS. A.
and antephormein, and to capture the ships of the enemy and

also to surround [them].11
3 For it happens in battle lines on land that larger [forces] can
restrain inferior, and fewer, even if very puny, [can restrain]
stronger, since each man is devising and contriving his safety
alone. But in [those] at sea the noble will fall with the ignoble
and plunge into the deep or may suddenly be made captive
without achieving more than a feeble [man].
4 You permit me to express my youthful exuberance on these

matters,12 you most eminent strate2gos*, adorned with great
endeavours, the valiant attendant of our valiant emperor, the
sure and brave servant of a sure and brave [lord], you who by
conflicts on land have gladdened the emperors themselves,
feeling secure, and have demonstrated that their every subject is
full of peace and joy, you who have humbled and brought low
the valiant deeds of all other [men] both who live now and were
born in former times, you who will show, if ever there should
be need, that [deeds] at sea are equal to those on on land.
5 I obey, for this is my opinion and this is my belief and I am
borne up by hopes of this kind. For every time that I compare
the [deeds] of others to yours and examine [them] together I
find [that it is] as though the actions of a child are outmatched,
outstripped, and outplayed by [those] of a grown man.
6 Indeed all the others who have achieved great things
immediately choose the quiet and comfortable life and decline
to undertake further endeavours less perchance they may often
fail in their subsequent [actions] and render the recollection of
their former [deeds] less glorious. But you always accomplish
subsequent [actions] that are greater than the first and,
becoming superior to all, in your pre-eminence in these
[matters] you contrive to conceal your first [achievements] by
the magnitude of the subsequent ones.
7 However, it happens that both those who favour [you] and those
who bear [you] ill-will suffer the same [thing]. When they are
defeated they rejoice and take pleasure equally, the one group
because they are victorious and remain so through your
forethought, the other because they have been defeated by no
------------------------------
11
On the meaning of these terms according to the Anonymous, and his probable
sources, see below §7.2.
12
526 APPENDIX THREE
uJpeivkein kai; uJpotavttesqai, kai; th'" a[llwn nivkh" aiJrei'sqai

ma'llon th;n ejk sou' prosgenomevnhn h|ttan aujtoi'", meq w|n
hJttw'ntai poihtaiv te kai; rJhvtore", w{sper ejkei'noi toi'" e[rgoi"
ou{tw kai; ou|toi toi'" lovgoi" ajpoleipovmenoi.
8 Pavnte" me;n ou\n dia; tau'ta wJ" ijscuvo" e{kasto" e[tucen e[cwn, th;n
eujfhmivan soi prosavgein kai; to;n ejk lovgwn e[painon
ejgceirou'sin. Kai; ouj diovti mh; ajxivw" prosavgein soi duvnantai,
h[dh kai; tou' panto;" ejl attou'sqai bouvlontai, ajll eujf hmou'nte"
o{son e{kasto" oi|ov" te h\/ to; pa'n prosenhnocevnai soi kai; mhdeno;"
aJmartavnein oi[ontai: fivlon ga;r kai; Qew'/ kai; ajnqrwvpoi" ejsti; to;
eij" duvnamin.
9 »Wn kai; hJmei'" ejsme;n oiJ tai'" sai'" kat ejcqrw'n ajristeivai"
pollavki" ejfhsqevnte" kai; sunecw'" touvtwn katentrufhvsante"
kai; terfqevnte" o{son eijko;" kai; megavlhn soi dia; tau'ta th;n cavrin
ojfeivlonte" o{ti ka]n tauvtai" i[sa toi'" pa'si kekoinwnhvkamen kai;
koinw'" eujfravnqhmen kai; hjgalliasavmeqa.
10 Anq w|n soi thvnde th;n sullogh;n di ejntolh'" sh'" suneilevcamen
ejk pollw'n me;n iJstoriw'n, pollw'n de; strathgikw'n sullexavmenoiv
te kai; ejklexavmenoi, dw'ron soi pavntwn ejrasmiwvteron kai; tw'n
a[llwn, wJ" eijpei'n, poqeinovteron, plh;n o{ti kai; polla; kata; tuvchn
kai; tovl man oujk ajnalovgw" tai'" paraskeuai'" eu{romen
ajpobaivnonta w{sper kai; muriavda" o{ti pleivsta" kaiv tina" tovl ma"
kai; megavla" paraskeua;" kaqairou'nta" tou;" su;n nw'/ kai; meta;
logismou' kinduneuvein ejqevlonta". All ejgceirhtevon h[dh e[rgou
ejcomevnou" kai; mh; toi'" prooimivoi" ejpi; polu; ejmbraduvnonta".
1.13 Peri; tw'n th'" new;" merw'n
1 Kai; prw'ton me;n peri; tw'n th'" new;" lektevon merw'n. Ei[dh14 ga;r
new'n perivergon a]n ei[h levgein ejn tw'/ parovnti: w{sper ga;r
a[nqrwpo" ajnqrwvpou th'/ bracuvthti dienhvnocen kai; tw'/ megevqei h]
------------------------------
13
aV (1) in the margin of MS. A.
14
A Platonic term, referring to the concept of ‘Forms’ or ‘Ideas’, translated here as
“kinds”. The following sentences also use Platonic phraseology.
others than by you, who are superior to all in good counsel and
might. And so, they must yield and submit and prefer the defeat
brought upon them by you to victory over others. And, with
them, poets and rhetoricians are defeated for, while they (the
former) fall short in their deeds, these (the latter) fall short in
their words.
8 All therefore for these reasons, according to the ability each
possesses, attempt to offer you adulation and praise in words.
And it is not because they are not able to approach you
worthily, for already they wish to be diminished in every
respect,15 but in offering you adulation as best each can, they
think that they have offered you everything and fall short in
nothing; for acting within one’s capabilities is acceptable to
both God and men.
9 We are amongst those who have often delighted in your deeds
of valour against the enemy and have endlessly rejoiced in them
and taken all the pleasure that is usual and owe you great
gratitude for this because, even if we have shared in your
exploits equally with everyone else, we rejoice and take delight
in common.
10 In return for this we have, on your instructions, gathered
together this collection, having selected and chosen from many
histories and from many manuals on strategy, a gift16 more
pleasurable than all and more desirable, so to say, than the rest.
However, we have discovered that many [things] come about
by chance and daring and not according to their preparation,
just as those planning hazardous undertakings with care and
forethought bring many hundreds of thousands of ventures and
great schemes to nothing. But we must make the attempt,
having already begun the task, and should not linger too long
over these introductory remarks.
1. Concerning the parts of the ship
1 First we must discuss the parts of the ship. It would be

inappropriate for the present to discuss kinds of ships; for, just
as one man differs from another in smallness and largeness or
------------------------------
15
A confused, and confusing, modesty topos.
16
That is, this text that he presents to Basil.
528 APPENDIX THREE
th'/ ajndreiva/ h] ajnandreiva/ h] tini touvtwn, pavntwn de ei\dov" ejsti

to; aujtov, ka]n oJ me;n tuvch/ polloi'" komw'n katorqwvma-
sin, oJ de; polloi'" ejl attwvmasi, kai; koinw'" me;n o{ te deilo;" o{ te
mh; deilov", o{ te smikro;" o{ te mh; smikrov", o{ te crhsto;" kai; oJ mh;
crhsto;" a[nqrwpo" ei[rhtai, ijdivw" de; a[llo" a[llo ti wjnovmastai
kai; h] di ajreth;n h] dia; kakivan h] dia; mevgeqo" oijkeivan
proshgorivan ejkthvsato,
2 ou{tw dh; kai; ejpi; nhw'n: koinw'/ me;n ojnovmati pa'sai kalou'ntai
nh'e", ijdivw" de; aiJ me;n trihvrei", aiJ de; dihvrei", aiJ de; monhvrei"
kata; to; ajnavlogon th'" eijresiva" kthsavmenai ta; ojnovmata, u{lh17
de; pasw'n ejstin hJ auJthv18 ka]n th'/ kataskeuh'/ polu; diafevrousin,
kai; aiJ me;n meivzosin, aiJ de; meivosin, kai; aiJ me;n pleivosin, aiJ de;
ejlavttosin xuvloi" kataskeuavzontai. Diovper, wJ" ei[rhtai, peri;
tw'n merw'n aujtw'n ei[pwmen ejx w|n th;n gevnesin e[cousi kai; th;n
suvstasin. 19
2.20 Mevrh newv"21
1 Druvocon, 22 trovpi", tropivdia, spei'ra,23 tropoiv. --- Kai; druvocon

me;n su;n polloi'" a[lloi" noeivsqw te kai; legevsqw to; kalouvmenon
para; pa'si koravkion, o} sunevcei pavnta kai; sugkratei' kai; w|/
prosdevdentai kai; oiJonei; ejpereivdontai ta; loipav. Koinw'" me;n
ga;r wjnovmastai druvocon a{pan xuvlon ejpivmhke" sunevcon kai;
oiJonei; proshlou'n24 e[tera braceva te kai; pollav. Noeivtw d a]n
ou{tw kai; to; perivtonon. Shmaivnei de; kai; th;n ojph;n tou' pelevkew"
ejn h|/per oJ steileio;" ejntivqetai, wJ" kai; ”Omhro":
------------------------------
17
An Aristotelian term for primordial matter.
18
auJth; MS. A., with e (= eJauth;) added in a second hand.
19
These are general Platonic terms for ‘being’ and ‘coming to be’.
20
No numbering in MS. A.
21
Even the form of the rest of this treatise closely follows and reflects that of
Pollux. It is a very juvenile exercize.
22
Cf. Pollux, Onomasticon (Bethe), I.85 (vol. 1, p. 27): “druvocon, trovpi",
[trovpide"], tropivdia, stei'ra [tropoiv].”.
23
On the Anonymous’s understandings of spei'ra see below §2.3. However, for
his speira, read steira, “cutwater” on the basis of Pollux, Onomasticon (Bethe), I.85.
24
We translate proshlovw in the sense of “fasten together”, rather than in the
classical sense of “to nail” since a piece of wood cannot “nail” anything.
bravery or cowardice or some other respect but is nevertheless

of the same kind, even if one happens to be adorned with many
achievements and another with many defects, and in general
usage both the craven and the not craven, both the small and the
not small, both the reliable and the not reliable, are called
“man”, while on particular points different men have different
appellations and achieve separate categorization because of
good qualities or bad qualities or size,
2 thus it is with ships. All are known by the general name of
ships, but in more precise usage some [are known as] triremes,
others as biremes, and others as monoremes, taking their names
according to their eiresia* (oarage). All have the same matter,
even if they differ greatly in their method of construction and
some are constructed from larger timbers and others from
smaller [ones], some from more and others from fewer.
Wherefore, as has been said, let us discuss their parts, from
which they derive their creation and their existence.
1. Parts of a ship
1 Stock (dryochos/dryochon*), keel, garboard strakes (tropidia*),

speira*, through beams (tropoi*). --- And you should class and
consider in many respects as a dryochos/dryochon what is
called by everyone the korakion*, which clasps and holds
everything together and to which other things are attached, and,
as it were, supported by it. Generally every long continuous
timber which fastens together many other short timbers is
known as a dryochos/dryochon. You should also class the wale
(peritonon*) in this way. It also means the hole in the axe into
which the haft is placed,25 as Homer [says]:
------------------------------
25
In fact this was not the meaning of dryochos and the Anonymous misunderstood
Homer’s line, which said that the axes were set up in a row, like dryochoi: “He set up
in a row like dryochoi, twelve [axes] in all;”.
530 APPENDIX THREE
”Istasc ejx eih;"26 druovcou" duvo kai; devka pavnta".27

2 Tropivdia de; ta; proshrmosmevna th'/ trovpei, peri; h|" u{steron
ejrou'men tw'/ pleivono" dei'sqai lovgou tauvthn h] ta; loipav.28 Ex
aujth'" ga;r a{pan a[rcetai mevro", kai; tauvthn wJsperei; qemevlion
e[cei.
3 Spei'ra de; para; me;n ajrcitevktosi tw'n stuvlwn oiJ pro;" th'/ bavsei
livqoi.29 Eu{rhtai de; kai; par AiJoleu'si to; a[rmenon, wJ" ejn tw/
Thlou' de; spei'ron kai; ejpivkrion e[mpese povntw/. 30

Shmaivnei de; kai; to; iJmavtion, wJ" ejn tw'/
Spei'ra kavk ajmf w[moisi balwvn: 31
kai; para; tw'/ poihth'/, wJ" ejn tw'/
Ai[ ken a[ter speivrou kei'tai polla; kteativssa".32
4 Tropoi; de; oiJ tropwth're": 33 kai; ”Omhro":

Hrtuvnanto de; kwvpa" tropoi'"34 ejn dermativnoisin.35
Ex ou| kai; tropwvsasqai levgetai to; th;n kwvphn sundh'sai tw'/
tropwth'ri.36
5 Trovpi" de; to; katwvtaton mevro" th'" nho;" kai; oi|on qemevlio".
Tauvth" de; to; me;n e}n mevro" ejx ou| hJ prwv/ra dianivstatai
proemboli;" kalei'tai, to; de; pro;" th;n pruvmnan podovsthma, e[nqa
------------------------------
26
ejxeih;", thus Dain: ejxeivh" MS. A.
27
Homer, Odyssey, 19.574 (vol. 2, p. 277): “i{stasc eJxeivh ", druovcou" w{", dwvdeka
pavnta":”.
28
Cf. Pho2tios, Lexicon (Naber), vol. 2, p. 229: “tropivdia: ta; eij" trovpin new;"
eujqetou'nta xuvla: ...”.
29
Cf. Pollux, Onomasticon (Bethe), I.85 (vol. 1, p. 27): “mevson de; th'"
proembolivdo" kai; tou' ejmbovlou hJ stei'ra kaloumevnh .” [rejected by the Anonymous];
Hesychios, Lexicon (Schmidt), S.1445 (vol. 4, p. 64): “Spei'ra: oiJ pro;" th'/ bavsei livqoi.
kai; th'" new;" skeu'ov" ti. kai; suvstremma ejk scoinivou, h] rJavkh.”. On “stones at the base
of pillars”, see above pp. 197-8.
30
Homer, Odyssey, 5.318 (vol. 1, p. 204): “thlou' de; spei'ron kai; ejp ivkrion e[mpese
povntw/.”.
31
Homer, Odyssey, 4.245 (vol. 1, p. 148): “spei'ra kavk ajmf w[moisi balwvn, oijkh'i
ejoikwv", …”.
32
kteativssa", thus Dain: kteativsa" MS. A.
Homer, Odyssey, 2.102 (vol. 1, p. 52): “ai[ ken a[ter speivrou kei'tai polla;
kteativssa".”. Cf. Hesychios, Lexicon (Schmidt), S,1445 (vol. 4, p. 64): “... kai;
iJmavtia. kai; iJstia. a[lloi ei\do" iJmativou eujmevgeqe" gunaikeivo u”.
33
Cf. Pollux, Onomasticon (Bethe), I.88 (vol. 1, p. 28): “oiJ de; peri; stei'ran
eJkatevrwqen parateinovmenoi tropoi; prw'to" kai; deuvt ero", oJ kai; qalavmio".” [rejected
by the Anonymous].
34
tropoi'" MS. A.: trovpois Dain.
35
Homer, Odyssey, 4.782 (vol. 1, p. 176): “hjrtuvnanto d ejretma; tropoi'" ejn
dermativnoisin,”.
36
Cf. Hesychios, Lexicon (Schmidt), T.1503 (vol. 4, p. 180): “tropwvsasqai: to; th;n
kwvphn pro;" tovn skalmo;n dh'sai tw'/ tropwth'ri. h] dia; mhcanh'" nikh'sai”.
He set up in a row all twelve dryochoi.

2 Tropidia are [what are] attached to the keel, about which we
will speak later since this needs more discussion than the rest.
Every part begins from this and uses this as if it were a
foundation.
3 Speira [is the name given] by architects to the stones at the base
of pillars. It is also found among the Aiolians [as a term for]
sail, as in
The sail (speiron) and yard arm fell far away into the sea.
It also means clothing, as in
putting filthy clothing (speira) around his shoulders;
and by the poet [Homer], as in
if he, though having acquired great wealth, were to lie
without a shroud (speiron).
4 Tropoi* are the oar-grommets (tropo2te2res*), and Homer [says]:
They attached the oars in the leather oar-grommets.
Hence fastening the oar with the oar-grommet is called
“grommeting up”.
5 The keel is the lowest part of the ship and its foundation, as it
were. One part of this, from which the prow extends, is called
proembolis*. That by the stern [is called] podoste2ma*, where a
532 APPENDIX THREE
dh; kai; skhnh; phvgnutai tw'/ strathgw'/ h] trihravrcw/, h[goun

kravbato".37 En oi|" de; oJ kravbatto" ejpereivdetai, trocanth're"
kai; a[flasta, oiJ legovmenoi bovrdwne". 38
6 Th'" de; pruvmnh" ta; mevrh pavlin eJkavtera pevtasoi kai; scista; kai;
ejpwtivde" levgontai, ejn oi|" ejpivkeintai ta; phdavl ia. Kai; to; me;n
a[kron tou' phdalivou h[toi tou' aujcevno" levgetai oi[ax:39 o{pou de; oJ
kubernhvth" ejpiklivnetai a[gklima kalei'tai. 40 To; de; pa'n oi[ax te
kai; phdavlion, to; de; teleutai'on uJperuvption, to; de; loipo;n
aujchvn.41 To; de; mevson th'" pruvmnh" kai; new;" mevcri th'" prwv/ra"
ajsavnidon.42
7 Ta; de; eJkatevrwqen tw'n toivcwn katavstega katavstrwma levgetai

kai; qra'no" kai; sanidwvmata, w|n a[nwqen hJ prwvth eijresiva kai; oiJ
oJpli'tai kai; toxovtai kai; peltastaiv, kavtwqen de; tou'
sanidwvmato" hJ deutevra h{ti" di o{lou ejrevttei, tucovntwn ejpi; tou'
katastrwvmato" a[nwqen polemouvntwn. Kai; oiJ me;n ejpi; tou'
qravnou" kaqhvmenoi qrani'tai levgontai, oiJ de; eij" ta; zuga; zuvgioi:
kai; qalavmioi de; e[stin o{te eij e[cei trei'" eijresiva" hJ nau'". 43
8 Kai; to; me;n e[dafo" aujth'" kuvto" kai; ajmfimhvtrion ojnomavzetai. 44

Plevousa de; ijdiwtikw'" ejkei' dhvpou kai; quriv" ejstin eij" ejkroh;n
tou' u{dato", h{ti" eujdiva"45 kalei'tai. Tauvthn de; dhlonovti th;n
plevousan sunevcousi ta; e{rmata, h[toi aiJ legovmenai e{drai: kai;
ta; ijkriva, a} ejgkoivl ia kalou'ntai.
------------------------------
37
Cf. Pollux, Onomasticon (Bethe), I.89 (vol. 1, p. 29): “ta; de; peri; th;n pruvmnan
prouvconta xuvla peritovnaia kalei'tai. ejkei' pou kai; skhnh; ojnomavzetai to; phgnuvmenon
strathgw'/ h] trihravrcw.”.
38
bovrdwne" MS. A: bavrdwne" Dain. Cf. Hesychios, Lexicon (Schmidt), T.1523
(vol. 4, p. 181): “trocanth're": pro;" ta; phdavlia. kalei'tai th'" pruvmnh" mevro".”.
39
Cf. Pollux, Onomasticon (Bethe), I.89 (vol. 1, p. 29): “to; de; a[kron tou' phdalivou
ªoi[ax: to; de; pa'nº oi[ax te kai; phdavlion ªkalei'taiº.”.
40
Cf. Pollux, Onomasticon (Bethe), I.90 (vol. 1, p. 29): “i{na de; kataklivnetai oJ
kubernhvth", a[gklima kalei'tai.”.
41
Cf. Pollux, Onomasticon (Bethe), I.89-90 (vol. 1, p. 29): “to; de; mevson aujtou'
fqei;r h] rJivza h] uJpovzwma, to; de; teleutai'on pteruvgion, to; de; loipo;n aujchvn.”.
42
Cf. Pollux, Onomasticon (Bethe), I.90 (vol. 1, p. 29): “ªto;º mevson de; th'" pruvmnh"
sanivdion, ou| to; ejnto;" ejnqevmion, to; d ajphrthmevnon aujtw'/ ejp iseivwn.”.
43
Cf. Pollux, Onomasticon (Bethe), I.87 (vol. 1, p. 28): “kai; to; me;n e[dafo" th'"
new;" ... kaloi'to d a]n kai; qavlamo", ou| oiJ qalavmioi ejr evttousi: ta; de; mevsa th' new;" zugav,
ou| oiJ zuvgioi kavqhntai, to; de; peri; to; katavstrwma qra'no", ou| oiJ qrani'tai.”.
44
Cf. Pollux, Onomasticon (Bethe), I.87 (vol. 1, p. 28): “kai; to; me;n e[dafo" th'"
new;" kuvto" kai; gavstra kai; ajmfimhvtrion ojnomavzetai.”.
45
eujdiva", thus Dain, following Hesychios: eujdia;" MS. A.
tent or berth (krabattos*) is fixed for the strate2gos or

trie2rarchos*.46 [Those things] on which the berth is supported
[are] trochante2res* and aphlasta*, the so-called bordo2nes*.
6 The parts on each side of the stern are called “spreaders”
(petasoi*) and “dividers” (schista*) and epo2tides*, on which
the rudders (pe2dalia*) rest. The top of the rudder or shaft
(auche2n*) is called the tiller (oiax*). Where the helmsman
(kyberne2te2s*) leans is called the anklima*. The whole [is
called] both tiller and rudder, the last [part] is the hyperyption*
and the rest is the shaft. The middle of the stern and the ship as
far as the prow is undecked (asanidon*).
7 The covered [parts] on each side of the hulls are called [the]
deck (katastro2ma*) and [the] bench (thranos*) and planking
(sanido2mata*). Above these is the first oar-bank and the
hoplites and archers and peltasts.47 Below the planking is the
second [oar-bank] which rows through everything, when there
is fighting on the deck above. Those sitting on the thranos are
called thranitai*, those on the thwarts (zyga*) [are called]
zygioi*, and sometimes [there are] thalamioi* if the ship has
three oar-banks.
8 And the bottom [of the ship] is named the hold (kytos*) and the
floor (amphime2trion*). And somewhere there, when sailing
idio2tiko2s* there is an opening, which is called a eudias* (bung
hole), for the removal of water. Indeed as [the ship] sails, the
shores (hermata*), that is, what are known as seats (hedrai*),
close this; and [there are] the decking [timbers] (ikria*), which
------------------------------
46
Cf. above pp. 215-16 & n. 156, 269.
47
“Hoplites” and “peltasts” were, of course, classical terms for types of soldiers
and quite irrelevant to tenth-century military practise.
534 APPENDIX THREE
9 Th'" de; pleouvsh" mevson ejpi; th'" trovpio" prosarmovzetai hJ

travpeza, h|" ejnto;" oJ iJsto;" i{statai, h[toi to; katavrtion. Tou' de;
katartivou to; me;n proshlouvmenon th'/ trapevzh/ katwvteron mevro"
ptevrna kalei'tai, ejx ou| kai; to; ejx eptevrnisen, o{tan uJpo; ajnevmou
biazomevnh e[x w th'" trapevzh" ejkbh'/. 48
10 ÔIstodovkh de; kai; keraiva to; keratavrion. ÔIstivon de; to; a[rmenon. 49
Kai; oiJ legovmenoi kaqormei'" ejpi; th'" trovpio" sterew'"
proshvlwntai kata; stoi'con trei'" o[nte", ejf w|n hJ keraiva
katagomevnh ejpivkeitai. Kai; trei'" de; stami'ne"50 h[goun
sthmonavria i{stantai kai; aujta; kata; stoi'con oi|" ejpereivdetai to;
katavstrwma.51
11 Eijsi; de; kaiv tina xuvla diavtona dihvkonta ajpo; tou' eJno;" toivcou th'"
nho;" e{w" tou' eJtevrou, ejf w|n ejpivkeitai: ta; de; tou;" toivcou"
e[xwqen sunevconta perivtona kalou'ntai. 52
12 ÔH de; sani;" di h|" aiJ kw'pai ejx evrcontai qureovn, kai; o{qen me;n
ejkdevdentai skalmov", w|/ de; ejndevdentai tropwthvr.53 To; de; ejpi; tw'n
skalmw'n ejpiskalmiv". Di w|n de; ei[retai hJ kwvph trhvmata.54 To; de;
pro;" aujtw'/ tw'/ skalmw'/ devrma a[skwma, to; par hJmi'n
manikevllion. 55
13 Tauvth" de; a[nwqen th'" eijresiva" perivtonon, ei\ta sani;" eJtevra, hJ

legomevnh pevl a, ei\ta perivtonon, ei\ta pavlin qureovn, e[nqa hJ
a[nwqen eijresiva. “Anwqen de; pavntwn hJ ejphgkeniv", to; a[rti
legovmenon katapathtovn: ejkei'sev pou kai; to; kastevllwma
givnetai, e[nqa ta;" ajspivda" oiJ stratiw'tai kremw'si.
14 Th'" de; prwv/ra" plhsivon eJkatevroi" toi'" mevresi perivboloi ejmpe-

------------------------------
48
Cf. Pollux, Onomasticon (Bethe), I.91 (vol. 1, p. 29): “kai; to; me;n uJpode-covmenon
to;n iJsto;n lhnov" ªkalei'taiº, to de; ejnarmozovmenon aujt w'/ ptervna, ...”.
49
Cf. Dindorf, Scholia Graeca, B.427 (vol. 1, p. 117): “iJstivo n to; a[rmenon, ...”.
50
stami'ne", thus Dain: stamivde" MS. A.
51
Cf. Pollux, Onomasticon (Bethe), I.92 (vol. 1, p. 30): “ta; de; xuvla ejf w|n aiJ
sanivde" ejpivkeintai, kanovnia kai; stami'ne".”.
52
Cf. Pollux, Onomasticon (Bethe), I.92 (vol. 1, p. 30): “to; de; sunevcon a[nwqen
eJkatevrou" tou;" toivcou" peritovnaion kalei'tai.”.
53
Cf. Pollux, Onomasticon (Bethe), I.87 (vol. 1, p. 28): “kai; o{qen me;n aiJ kw'pai
ejkdevdentai ªskalmov", w|/ de; ejkdevdentaiº, tropwthvr, ...”.
54
trhvmata, thus Dain: trivmata MS. A.
55
Cf. Pollux, Onomasticon (Bethe), I.88 (vol. 1, p. 28): “to; d uJpo; to;n skalmo;n
ªejpiskalmiv"º. ... di w|n de; dieivr etai hJ kwvph, trhvmata. to; de; pro;" autw'/ tw'/ skalmw'/ devrma
a[skwma.”.
are called the floor timbers (enkoilia*).

9 When [the ship] is sailing, the mast step (trapeza*), in which
the mast (histos*), that is, the katartion*, is set up, is fixed in
the middle on to the keel. The lower part of the katartion which
is fixed in the mast step is called the heel (pterna*); hence [the
expression] “is unheeled” when it comes out of the mast step
under pressure from the wind.
10 [There is] the mast receiver (histodoke2*) and the yard, the
keratarion*. The sail (histion) [is] the sail (armenon). And what
are known as the kathormeis* are fixed firmly in a row on the
keel, there being three of them, on which the yard (keraia) rests
[when] lowered. And three futtocks (stamines*), or ste2monaria,
on which the deck is supported, are also fixed in a row.
11 There are also some timbers (xyla diatona*) which stretch from
one side of the ship to the other and by which it is braced.
Those [timbers] which enclose the sides on the outside are
called wales.
12 The strake through which the oars come out [is the] thyreon*,
and [that] from which they are hung [is the] thole (skalmos*),
and what they are hung with [is] the oar-grommet. What is on
the tholes [is] the episkalmis*. [The parts] through which the
oar is passed [are] oarports (tre2mata*). The hide over the thole
[is the] asko2ma*, the manikellion* according to us.
13 Above this oar-bank [is] the wale, then another plank, called
pela*, then a wale, and then a thyreon again, where the upper
oar-bank [is]. Above everything [is] the gunwale (epe2nkenis*),
known lately as the katapate2ton*. Somewhere here there is also
the pavesade (kastello2ma*) where the soldiers hang their
shields.
14 Near the prow on both sides periboloi* (catheads) are fixed
536 APPENDIX THREE
14 phgmevnoi i{stantai di w|n aiJ a[gkurai krevmantai, aiJ th;n nau'n
iJstw'si calwvmenai.56 Epi; de; th'" prwv/ra" oJ sivf wn o}" katakovrax 57
levgetai ejnergw'n o{tan w\sin aiJ nh'e" ajntivprw/roi: kai; duvo de;
plavgioi kai; aujtoi; ejnergou'nte" o{tan plavgiw" prosbavlwsi.
15 Ta; de; th'" new;" scoiniva: kavloi, provtonoi, peivsmata, ajpovgaia,

prumnhvsia, kai; e[mboloi, oi} tou;" oi[aka" sunevcousin kai; di w|n
eij" to;n trocanth'ra ajpodesmw'ntai. 58
16 Au|tai me;n aiJ ojnomasivai oijkei'ai celandivou kai; drovmwno": ejk

tw'n aujtw'n ga;r nhi?wn xuvlwn ajmfotevrwn aiJ kataskeuai;
givnontai, eij kai; peri; th;n kaqovlou klh'sin dienhnovcasi: kai; to;
me;n drovmwn wjnovmastai, to; de; celavndion.
3.59 Peri tw'n ojnomasiw'n tw'n nhw'n tw'n pro;" povlemon

ejpithdeivwn
1 Kai; to;n me;n tw'n dromwvnwn60 ajriqmo;n kai; tw'n ejn aujtoi'"
stratiwtw'n ajneivkastovn ejsti kai; a[dhlon diorivsasqai61 o{ti mhde;
ejn toi'" iJstorhvsasi peri; touvtwn palaioi'" eu{romen ejf eJno;"
ajriqmou' di;" th;n aujth;n diafulattomevnhn kai; throumevnhn tavxin
th'" nautikh'" stratia'", ajlla; pro;" to; tw'n ejnantivwn plh'qo" tou;"
i[sou" pollavki" h] kai; pleivona" ejntavttousavn te kai;
ajntexavgousan kata; to; tw'n dromwvnwn kai; tw'n ajndrw'n plh'qo"
kai; mevgeqo".
2 «Hn ga;r kai; meizovnwn plh'qo" dromwvnwn kai; nu'n ei\nai crh; oi|"
ejpistw'sin ta; legovmena pavntw" xulovkastra, ajlla; kai; mevsai
trihvrei" kai; monhvrei" tine;" tacinaiv, leptai; galevai, ai|" ejn tai'"
bivglai"62 crhstevon pro;" ejreqismo;n tw'n ejnantivwn kai; diavlusin
tavxewn kai; o{sa tw'/ th" naumaciva" ei[dei suntelei'n ei[wqen, oi|on
------------------------------
56
Cf. Pollux, Onomasticon (Bethe), I.93 (vol. 1, p. 31): “a[gkurai ajmfivboloi,
ajmfivstomoi, eJterovstomoi: ...”.
57
katakovrax, thus Dain: katakovraka MS. A.
58
Cf. Pollux, Onomasticon (Bethe), I.93 (vol. 1, p. 31): “... iJstov", iJstodovkh,
keraiva, scoiniva, kavloi, provtonoi, kalwv/dia, peivsmata, ajpovgua, ªejpivguaº, prumnhvsia:
...”. Cf. Hesychios, Lexicon (Schmidt), T.1523 (vol. 4, p. 181): “trocanth're": pro;" ta;
phdavlia. kalei'tai th'" pruvmnh" mevro".”.
59
bV (2) in the margin of MS. A.
60
dromwvnwn, thus Dain: dromovnwn MS. A.
61
Cf. Appendix Two [a], §12: “To;n de; tw'n dromwvnwn ajriqmo;n kai; tw'n ejn aujtoi'"
stratiwtw'n ajneikastovn ejstin kai; a[dhlo;n diorivsasqai: ...”.
62
Cf. Appendix Two [a], §10: “Kai; e[ti de; kataskeuavsei" drovmwna" mikrotevrou"
14 from which are suspended the anchors, which halt the ship
when they are let down. On the prow [is] the flame-thrower
(sipho2n*) called the katakorax*, which functions when the
ships are prow to prow. And [there are] also two at the sides,
which function when engaged by the side.
15 The ship’s cordage: brails, forestays, stern cables, mooring
lines, bow mooring lines, stern mooring lines, and emboloi*,
which restrain the tillers and by which these are bound to the
trochante2r.
16 These are the terms appropriate to a chelandion and a dromon.
Both are constructed from the same ships’ timbers, even if they
differ in their overall nomenclature, the one being called
dromo2n and the other chelandion.
3. Concerning terms for the ships that are suitable

for war
1 About the number of dromons and of the soldiers in them it is

impossible and unrealistic to be prescriptive, because not even
in the ancient [writers] who discuss these matters do we find
one single figure kept and observed twice for the battle line
(taxin) for a naval expedition, but [an expedition] usually
marshalling and drawing up against the opposing [force],
[dromons] equal to or greater than [those of the enemy] in
respect of the quantity and size of dromons and men.
2 There used to exist a number of larger dromons and now there
should be [constructed] those on which are placed what are
called wooden castles (xylokastra*), but also [there should be]
middle-sized triremes and some fast monoremes, light galeai,*
which should be used as scouts, to provoke the opposition and
------------------------------
gorgotavtou", oiJonei; galeva" kai; monhvr ei" legomevnou", tacinou;" kai; ejlafrouv", oi|sper
crhvvsh/ ejn te tai'" bivglai" kai; tai'" a[llai" spoudaivai" creivai".”.
538 APPENDIX THREE
ejfolkivde", levmboi, kevl hte", ejpaktrivde", ejpaktrokevlhte", nh'e "

stratiwvtide", oJplitagwgoiv, iJppagwgoiv, forthgoiv. 63
3 Eijsi de; kaiv tina ploi'a krioi; kai; travgoi legovmena, oi|"
crhsamevnou" Lukivou" ajnevgnwmen, wJ" eijkavz ein o{ti toiou'tovn ti
ploi'on kai; oJ tau'ro" h\n oJ th;n Eujrwvphn ajpagagwvn. 64
4.65 Peri; tw'n ojnomasiw'n tw'n ajrcovntwn tou' plwi?mou

strateuvmato"
1 AiJ de; tw'n tou' toiouvtou plwi?mou strateuvmato" hJgemovnwn

ojnomasivai eijsi;n ai{de. Strathgov", oJ pavntwn a[rcwn h[goun
stratiwtw'n kai; tw'n aujtw'n hJgemovnwn: kovmh", 66 oJ ejpi; trisi;n h]
kai; pevnte drovmwsin tetagmevno" hJgemoneuvein. 67
2 ÔEkatontavrch" oJ ejpi; mia'" nho;" eJkato;n ajndrw'n hJgouvmeno" o{sti"
kai; trihvrarco" kevklhtai. “Esti de; oJ legovmeno" kevntarco":
kevntoum ga;r para; ÔRwmaivoi" oJ eJkato;n ajriqmo;" proshgovreutai
kai; kevntarco" oJ eJkato;n ajndrw'n hJgouvmeno". Epi; de; tw'n
qematikw'n dromwvnwn drouggavrioi kai; tourmavrcai uJpo; th;n tou'
strathgou' cei'ra kai; aujtoi; telou'nte". 68
------------------------------
63
This list of terms for both specific types of ships and boats and also the uses to
which ships might be put was fundamentally classical rather than Byzantine;
although, such words as hippago2gos and phorte2gos were still used in the tenth
century. We have given the best understanding of the classical meanings of the terms
now available; however, it should not be thought that the Anonymous realized what
they had once meant. In fact, once again, he simply lifted them from Pollux. Cf.
Pollux, Onomasticon (Bethe), I.82-3 (vol. 1, p. 26): “Peri; new'n kai; nautikw'n
ojnomavtwn. ... ejfolkivde", levmboi, kuvdaroi, gau'loi, kevlhte", kelhvtia, ejpaktrivde",
ejpaktrokevlhte" [barei'"], ... levgoito d a]n tacei'a nau'" ... oJplitagwgov", [stratiw'ti",
strathgiv", iJppagwgov"], fortiv", forthgov", ...”.
64
Cf. Pollux, Onomasticon (Bethe), I.83 (vol. 1, p. 27): “[Luvkia] [[legovmena]]
krioi; kai; travgoi, wJ" eijkavzein o{ti toiou'tovn to ploi'on kai; oJ tau'ro" h\n oJ th;n Eujrwvphn
ajpagagwvn.”.
65
gV (3) in the margin of MS. A.
66
Cf. Appendix Two [a], §25: “... e{na to;n legovmenon kovmhta, o{sti" nauvarcov" te
kai; hJgemw;n tw'n uJp aujt w'n dromwvnwn uJpavrcwn frontivsei prosecevsteron peri; pavntwn
eujkovl w" kai; diatavxei pro;" a{panta.”.
67
Cf. Appendix Two [a], §25: “... ajll ejpisthvsei" aujtoi'" a[rconta" h] kata; pevnte h]
kata; trei'" drovmwna", ...”.
68
Cf. Appendix Two [a], §26: “... ejpi; de; tw'n qematikw'n dromwvnwn kai; drouggavr ioi
ejpisthvsontai kai; tourmavrcai, kai; aujtoi; tw'/ strathgw'/ uJpotaghvsontai kai; toi'" ejkeivnou
paraggevlmasin uJpakouvsousin.”.
break up their battle lines, and all the [ships] that usually form a
part of the pattern of naval warfare, such as ships’ boats (epho-
lkides), light galleys (lemboi), merchant galleys (kele2t es), small
galleys (epaktrides), light kele2tes (epaktrokele2tes), troop trans-
ports, hoplite transports, horse transports, [and] supply vessels.
3 There are also other ships called rams (krioi) and goats (tragoi)
which we have read that the Lycians used, from which [we
may] conjecture that the bull that carried off Europa was a type
of boat.
4. Concerning the terms for the commanders (archontes*) of

the naval force
1 The terms for the officers (he2gemones*) of a naval force of this

sort are as follows. A strate2gos [is] the commander of all, or
rather of the soldiers and their officers; a kome2s* is appointed to
command three or five dromons.
2 A hekatontarche2s* is the officer of a ship of one hundred men,
and is also called a trie2rarchos*. There is also what is called a
kentarchos*, for centum is the Roman word for the number one
hundred and a kentarchos is the officer [in charge] of one
hundred men. In the thematic dromons [there are] droungarioi*
and tourmarchai* [who] themselves serve under the hand of the
strate2gos.
540 APPENDIX THREE
3 Ef eJkavsth" de; tw'n new'n a[rcontev" eijsi;n oi{de: trihvrarco" kai;
penthkovntarco", eJkatovntarco" te kai; nauvarco" kai;
epjistoleuv". Levgetai de; oJ ejpi; tou' stovlou diavdoco" tou'
nauavrcou: nauvarco" d a]n ojnomavzoito oJ par hJmi'n
prwtokavrabo". Prosqetevon de; touvtoi" kai; trihrauvlhn kai;
keleusthvn: e[sti de; oJ me;n boukinavtwr, 69 oJ de; keleusth;" oJ to;
flavmoulon katevcwn.70
5.71 Peri; tw'n stratiwtw'n oJpoivou" dei' ei\nai tou;" ejpi; tou'
katastrwvmato" a} sanidwvmata kalou'ntai
1 Eujyuvcou" kai; rJwmalevou" kai; proquvmou" eij" th;n kata; tw'n

ejnantivwn ejgceivrhsin, ejpei; koinw'" kai; mavcimoi kevklhntai. Ta;
de; tw'n toiouvtwn o{pla ajspivde", qwvrake", knhmi'de", kravnh te
kai; xivf h, dovratav te kai; drevpana kai; ceirovyella.72 Kai; toxeiva
ejsti;n o{te to; pa'n ejk diasthvmato" ejnergou'sa kata; tw'n ejn tai'"
cersi; th;n ijscu;n kai; duvnamin ejcovntwn ejnantivwn oJplitw'n.
2 Pavntw" de; ajndrei'oi e[stwsan kai; tw'n kata; povlemon

ejmpeirovtatoi,73 wJ" a]n ejn touvtoi" oiJ kubernh'tai qarrou'nte"
prosbavllwsi nausi; polemivai" kai; prosdesmw'sin eujkovl w",
diakovptonte" tauvta" kai; kataduvonte".
6.74 Peri; tw'n oJplivsewn eJkavsth" nhov"
1 Pro;" touvtoi" de; kai; aiJ tw'n nhw'n eJkavsth" oJplivsei" e[stwsan
ai{de: devrrei" kai; difqevrai75 kai; sivf wne" kai; o{sa ejn tw'/ peri
------------------------------
69
boukinavt wr, thus Dain: ibukinavtwr MS. A.
70
Cf. Pollux, Onomasticon (Bethe), I.96 (vol. 1, p. 32): “a[llh" de; creiva "
trihvrarco", penthkovntarco", nauvarco", ejp istoleuv": ou{t w ga;r ejkalei'to oJ ejp i; tou'
stovlou diavdoco" tou' nauavrcou. oJ de; stovlo" kaloi't a]n kai; ajpovstolo". [prosqetevon de;
touvtoi"] [kai; trihrauvlhn kai; keleusthvn].”; I.119 (vol. 1, p. 39): “Ta; de; th'"
navumaciva": aiJ me;n fevrousai trihvrei", makra; ploi'a, tacei'ai nh'e", katavfrakta ploi'a,
oiJ de; a[rconte" trihvrarcoi kai; penthkovntarcoi, kai; nauvarcoi kai; ejpistolei'". To; [de;]
pra'gma, [nauarciva], trihrarciva, penthkontarciva. th'" de; tou' nauavrcou new;" [to;]
o[noma, nauarci;", kai; strathgiv".”.
71
dV (4) in the margin of MS. A.
72
Cf. Pollux, Onomasticon (Bethe), I.120 (vol. 1, p. 39): “Ta; de; tw'n ejmpleovntwn
ajspivde", qwvrake", knhmi'de", kravnh, xivfh, dorudrevpana, cei'r e" sidhrai': ...”.
73
74
eV (5) in the margin of MS. A.
75
difqevrai, thus Dain: divfqerai MS. A. Cf. Pollux, Onomasticon (Bethe), I.94
3 In each of the ships, the commanders are as follows: trie2rarchos

and pente2kontarchos*, hekatontarchos and navarchos* and
epistoleus*; he is known in the fleet as the deputy of the
navarchos. A navarchos would be called pro2tokarabos* by us.
To these should be added the trireme’s flute-player
(trie2raule2s*) and the keleuste2s*, the first is the trumpeter and
the keleuste2s is in charge of the standard.
5. Concerning the soldiers who should be on the katastro2ma*

which is called sanido2mata
1 [These should be of] good spirit, sturdy and eager for the
encounter with the opposition since they are generally called
warriors. The weapons of such men [should be] shields, breast
plates, greaves, helmets and swords, spears, rigging cutters and
vambraces. And bows [are needed] sometimes, functioning
over a distance against enemy hoplites who are strong and
powerful in hand-to-hand [fighting].
2 All in all, these should be brave men with considerable
experience in battle, so that the helmsmen can rely on them in
attacks on enemy ships and couple [ships] easily, and break
through [their line] and sink [them].
1. Concerning the armaments of each ship
1 In addition the armaments of each of the ships should be these:

leather hides and screens (diphtherai) and sipho2nes and
------------------------------
(vol. 1, p. 31): “... devrrei", difqevrai. ...”.
542 APPENDIX THREE
merw'n aujth'" perielavbomen.
2 “Eti de; kai; pro;" ta;" ejmbola;" tw'n ejnantivwn buvrsai tauvtai"
proshlouvsqwsan, o{pw" oJ sivdhro" periolisqaivnh/ pro;" to;
ajntivtupon ajntilabh;n oujk e[cwn, wJ" a]n oiJ oJpli'tai toi'" kontoi'"
ajpwqou'nte" ajp ajllhvl wn ta; skavfh diavgwsin76 eij" th;n eJautw'n
swthrivan, eij mh; pro;" cei'ra" qarrou'si parakerdaivnonte".
7.77 Peri; th'" tw'n iJstorikw'n levx ewn diasafhvsew"
1 Epei; de; wJ" ejcrh'n peri; tau'ta dihvlqomen, e[lqwmen dh; kai; ejpi;
tw'n levx ewn tw'n ejn toi'" iJstorikoi'" tetagmevnwn th;n diasavfhsin.
Epitrihravrchma oJ crovno" ejsti;n o{n ti" ejpetrihravrchse,
ejxhvkonto" me;n aujtw'/ tou' kairou', braduvnonto" de; tou'
diadovcou. 78 Sugkekrothmevnon to; kalw'" peplhrwmevnon: to; de; wJ"
eJtevrw" ajpoplhvrwton ajsugkrovthton. 79
2 Naulocei'n to; tou;" polemivou" parafulavttein: ajnqormei'n de; kai;

ajnteformei'n to; ajntikaqesthkevnai pro;" naumacivan ejsti kai;
ejxormei'n to; prosekpleu'saiv pou: periormei'n de; to; proskaqe-
zesqai nhvsw/ poliorkhtikw'". 80
------------------------------
76
Cf. Appendix Two [a], §28. Cf. also Pollux, Onomasticon (Bethe), I.120 (vol. 1,
p. 39): “pro;" de; ta;" ejp ibola;" aujtw'n ajntesofivzonto buvrsa" proshlou'nte" [pro;" ta;
toicivsmata tw'n new'n], o{pw" oJ sivdhro" ojlisqavnh/, pro;" to; ajntivt upon ajntilabh;n oujk
e[cwn. [kontoi'"] ajpewqou'nto kai; dih'gon ajp ajllhvlwn ta; skavfh.”. Pollux’s source was
almost certainly Thucydides. See Thucydides, Peloponnesian war, VII.65.2 (vol. 4, p.
128): “ta;" ga;r prwv/ra" kai; th'" new;" a[nw ejpi; polu; katebuvrswsan, o{pw" a]n
ajpolisqavnoi kai; mh; e[coi ajntilabh;n hJ tw'n sidhrw'n ceirw'n (65.1)] cei;r
ejpiballomevnh .”.
77
ıV (6) in the margin of MS. A.
78
Cf. Pollux, Onomasticon (Bethe), I.123 (vol. 1, p. 40): “tw'/ de; qa'tton poihvsanti
tou'to a\qlon stevfano" h\n. e[sti de; aujtw'/ lovgo" kai; peri; tou' trihrarchvmato":
ejpitrihravrchma dev ejstin oJ] crovno" o{n ti" ejpetrihravrchsen ...”.
79
Cf. Pollux, Onomasticon (Bethe), I.121 (vol. 1, p. 40): “kai; ta; me;n kalw'",
peplhrwmevna, sugkekrothmevna, ta; de; wJ" eJtevrw" ajplhvrwta [kai; hJmiplhvrwta] [kai;
ajsugkrovthta ].”.
80
Cf. Pollux, Onomasticon (Bethe), I.122 (vol. 1, p. 40): “to; de; fulavttein tina;"
ejformei'n kai; naulocei'n, kai; to; ajntikaqesthkevnai pro;" naumacivan ajnqormei'n [kai;
ajnteformei'n, kai; to; prosedreuvein] [prosormei'n], kai; [to;] proekpleu'sai proexormei'n
kai; proormei'n, kai; to; sthvsasqai th;n nau'n prosormivsasqai, kai; to; ejn kuvklw/
periplei'n nh'son kai; proskaqh'sqai nhvsw/ poliorkhtikw'" ajpo; new'n, periormei'n. kai;
periormivzein th;n nau'n peri; to; cw'mav fhsi Dhmosqevnh".”.
everything we mentioned in the [section] concerning parts [of a

ship].
2 Moreover, also let there be hides fastened to the [ships] against
“ramming” by the opposition, so that the iron might glance off
in the opposite direction and does not take hold, [and] so that
the hoplites, fending the vessels off from each other with
spears, achieve their own safety, even if they had no great
expectation of success in hand to hand fighting.
7. Concerning the interpretation of historical vocabulary
1 Since we have investigated these topics as was required, let us

discuss the vocabulary found in the historians. Epitrie2arche2ma
[is] the additional period a man serves as trie2rarchos beyond
the expiry of his office, in the absence of his successor.
Synkekrote2menon (welded together) means fully trained and
conversely asynkrote2ton (not welded together) means poorly
trained.
2 Naulochein [means] to be on the watch for the enemy.
Anthormein and antephormein [mean] to take position for
fighting a naval battle, and exormein [means] to set sail for
somewhere. Periormein [means] to set up a siege round an
island.
544 APPENDIX THREE
3 Kai; to; me;n eij" toujpivsw th;n nau'n ajnakrousasqai to; eij"
ejmbolh;n uJpavgein ejstiv: to; de; eij" th;n pruvmnan krouvsasqai. 81 Kai;
to; me;n eij" toujpivsw th;n nau'n ajnakrousasqai to; eij" ejmbolh;n
uJpavgein ejstiv: to; de; eij" th;n pruvmnan krouvsasqai.82 Kai;
ajnagwgh; me;n ejsti;n oJ ajpo; tou' limevno" ejpi; to;n povnton pro;"
povlemon plou'", kai; ajntanagwgh; kai; ajntepivpleusi": ajnavpeira
de; hJ pro;" naumacivan ajntepistrofhv.
4 Kai; perivplou" mevn ejstin oJ pro;" perikuvklwsin plou'":
paravplou" de; o{tan plagivw" toi'" polemivoi" oiJ e{teroi plhsivstioi
periplevwsi, kai; dievkplou" o{tan ejmbavlwsin aiJ nh'e" kata; to;
mevson tw'n polemivwn kai; pavlin uJpostrevywsi kai; pavlin
ejmbavl wsi, 83 diakovptonte" kai; buqivzonte" ta;" tw'n ejnantivwn. Kai;
uJperkera'sai mevn ejsti to; ta; tw'n ejnantivwn kevrata
perikuklw'sai, w{sper uJperkerasqh'nai to; uJpo; tw'n ejnantivwn
perikuklwqh'nai.
5 Parexeiresiva dev ejsti to; o[pisqen mevro" th'" pruvmnh", e[nqa ta;
paravptera tw'n new'n eijsin, a} ejpwtivde" kevklhntai: levgetai de;
ou{tw" dia; to; parekto;" th'" eijresiva" ei\nai to; phdavlion oiJonei;
eJrevtton kai; ijquvnon th;n nau'n.
6 Tau'ta me;n ou\n iJkanw'" hJmi'n dieivlektai kai; oujde;n parei'tai tw'n
ojfeilomevnwn mnhmoneuqh'nai: metitevon de; h[dh ejpi; ta; ei[dh tw'n
paratavxewn.
8.84 Naumaciva" paravtaxi" kuklikhv
1 Kukliko;n85 kalei'tai to; sch'ma th'" tavxew" o{tan tw'/ mh; didovnai
dievkploun ...86
------------------------------
81
Cf. Pollux, Onomasticon (Bethe), I.125 (vol. 1, p. 41): “kai; to; me;n eij" ejmbolh;n
uJpagagei'n eij" toujpivsw th;n nau'n ajnakrouvsasqai, to; d eij" fugh;n Ôpruvmnan
krouvsasqai: ”. Hesychios, Lexicon (Schmidt), A.4375 (vol. 1, p.175): “ajnavkrousi":
ejn naumaciva/ ejlevgeto ejp i; tou' pruvmnan krouvein”.
82
Cf. Pollux, Onomasticon (Bethe), I.125 (vol. 1, p. 41): “kai; to; me;n eij" ejmbolh;n
uJpagagei'n eij" toujpivsw th;n nau'n ajnakrouvsasqai, to; d eij" fugh;n Ôpruvmnan
krouvsasqai: ”. Hesychios, Lexicon (Schmidt), A.4375 (vol. 1, p.175): “ajnavkrousi":
ejn naumaciva/ ejlevgeto ejp i; tou' pruvmnan krouvein”.
83
Cf. Hude, Scholia, I.49.3 (p. 44): “dievkploi: dievkplou" ejsti; to; ejmbalei'n kai;
pavlin uJpostrevyai kai; au\qi" ejmbalei'n.”.
84
zV (7) in the margin of MS. A.
85
The following incomplete sentence appears to have been based on Thucydides’
account of the first battle of Naupaktos in 430 B.C.E. See Thucydides, Peloponnesian
war, II.83.5 (vol. 1, p. 412): “kai; oiJ me;n Peloponnhvsioi ejtavxanto kuvklon tw'n new'n wJ"
mevgiston oi|oiv t h\san mh; didovnte" dievkploun, ...”.
3 To bring the ship into reverse for “ramming” is anakrousas-

thai;87 [to bring it] by the stern [is] krousasthai. Anago2ge
(bringing up) is the departure from harbour to the open sea for
battle,88 and [so is] antanago2ge2 (bringing up against) and
antepipleusis (sailing against).89 Anapeira is the turning around
for a naval battle.90
4 Periplous (sailing round) is the sailing movement of

encirclement. Paraplous (sailing past) is when each group sails
past the enemy side on with full sails. Diekplous (sailing
through) is when the ships attack through the line of the enemy
and turn around and attack once again, destroying and sinking
the [ships] of the opposition. And hyperkerasai (outflanking) is
the encirclement of the wings of the opposition just as ‘to be
outflanked’ is to be encircled by the enemy.
5 The parexeiresia* (outrigger) [is] the rear part of the stern
where the paraptera of the ships are, which are called epo2tides.
It is called this because the rudder, which, as it were, directs
and guides the ship, is outside the eiresia.* 91
6 We have now said sufficient on this subject and nothing has
been omitted that should have been mentioned. We should now
proceed to the types of battle-lines.
8. The encircling formation of a sea battle
The arrangement of the formation is called kyklikon* (encirc-

ling) when by not making a diekplous (sailing through) ...
------------------------------
86
The text ends at the end of the last line of fol. 342v but appears to have been
continued in subsequent folios which have been lost.
87
The Anonymous’s sources were Pollux and Hesychios but in following them he
got it all wrong since no galley would “go into reverse”, or backwater in order to ram.
The confusion arose from Pollux’s reading of Thucydides’ description of the battle of
Syracuse. See Thucydides, Peloponnesian war, VII.70.4 (vol. 4, p. 138): “..., aiJ me;n
ejmbolai; dia; to; mh; ei\nai ta;" ajnakrouvsei" kai; dievkplou" ojlivgai ejgivgnonto, ...”.
Thucydides wrote that conditions were so crowded in Syracuse harbour that ramming
was impossible because the ships could not backwater to give themselves room or
perform the break through the line, diekplous. There was no room to manœuvre to
ram. However, in reading him, Pollux and Hesychios associated “backing water”
(anakrousis) with “ramming” (embole2) and the Anonymous slavishly followed them.
88
Cf. Pollux, Onomasticon (Bethe), I.102 (vol. 1, p. 34).
89
Cf. Pollux, Onomasticon (Bethe), I.124 (vol. 1, p. 40).
90
Cf. Pollux, Onomasticon (Bethe), I.123-4 (vol. 1, p. 40).
91
APPENDIX FOUR
FLEETS, ARMAMENTS, AND EQUIPMENT FOR DROMONS,

PAMPHYLOI, AND OUSIAKA CHELANDIA ACCORDING TO
THE INVENTORIES FOR THE EXPEDITIONS TO CRETE OF 911
AND 949 IN THE DE CERIMONIIS AULAE BYZANTINAE
ATTRIBUTED TO CONSTANTINE VII PORPHYROGENNETOS1
[a] THE EXPEDITION OF 911
[b] THE EXPEDITION OF 949

discussed elsewhere in the text or appendices, are asterisked the first
time they are used. They may be accessed through the Index.
-----------------------------
1
Texts adapted from Haldon, “Theory and practice”, pp. 203-13, 219-33 with
reference to Constantine VII, De cerimoniis, II.44, II.45 (vol. 1, pp. 651-60, 664-5,
669-77). The translations are our own. The text here includes only those sections of
De cerimoniis, II.44 & II.45 related to the naval forces per se, not the complete text.
We have followed the abbreviations for novmismata/nomismata used by Reiske
and Haldon, two large commas [,,], and that used by Haldon for
miliarhvsia/miliare2sia, . For the abbreviations for 0 miliaresia used by Reiske and
Haldon, we have used zero/m [0/m]. We have expanded their abbreviations for
kente2narion/kente2naria. For the archaic letter koppa, representing the numeral 90, we
have used #.
On the expedition of 911 see Vasiliev, Byzance et les Arabes.Tome I, pp. 208-
16; Haldon, “Theory and practice”, pp. 239-42. For that of 949 see Vasiliev/Canard,
Byzance et les Arabes. Tome II, part 1, pp. 320-41. This should be read, however,
with much caution. Vasiliev/Canard did not appreciate that an ousia was a ship’s
company rather than an actual ship and this effects much of their analysis. See also
Haldon, “Theory and practice”, pp. 334-9.
548 APPENDIX FOUR
[a] The expedition of 9112
ÔH genomevnh ejxovplisi" kai; The fitting out and the cost

e[xodo" kai; to; poso;n th'" rJovga" and the sum of the pay and of
kai; tou' laou' tou' the force sent against the
ajpostalevnto" kata; th'" impious Crete with the
qeolevstou Krhvth" meta; tou' patrikios* and logothete2s tou
patrikivou ÔHmerivou kai; dromou* Himerios in the time
logoqevtou tou' drovmou ejpi; of the Lord Leo, beloved of
Levonto" tou' filocrivstou Christ.
despovtou.
1 To; basilikoplovi>mon ciliavde" The imperial fleet, 12,000;

ibV.ÔRw'" yV. Rho2s, 700.
ejdevxato oJ strathgo;" th'" tw'n The strate2gos of the
Kiburraiwtw'n e[cein Kibyrrhaio2tai undertook to
strato;n ÀecV, kai; diplou'" provide a force of 5,600,
ÀaV: oJmou ÀıcV. and 1,000 reserves; 6,600 in
all.
ejdevxato oJ strathgo;" th'" The strate2gos of Samos
Savmou e[cein strato;n ÀdV, undertook to provide a force
kai; diplou'" ÀaV: oJmou ÀeV. of 4,000, and 1,000
reserves; 5,000 in all.
ejdevxato oJ strathgo;" tou' The strate2gos of Aigaion
Aijgaivou pelavgou" e[cein Pelagos undertook to
strato;n ÀgV, kai; diplou'" ÀaV: provide a force of 3,000,
oJmou ÀdV. oJmou' to; pa'n and 1,000 reserves, 4,000 in
xiliavde" khV kai; tV. all. In all a total of 28
thousand and 300.
Dia; tou' basilikou' ploi?mou Concerning the imperial fleet
2 Drovmwne" xV e[conte" ajna; 60 dromons having 230
ajndrw'n kwphlatw'n slV kai; oarsmen and 70 marines
ajna; polemistw'n oV: oJmou' each; in all 18,000. 40
ciliavde" ihV. pavmfuloi mV ejx pamphyloi*, of which 20
w|n oiJ me;n kV pavmfuloi ajna; pamphyloi [have] 160 men
ajndrw'n rxV, oiJ de; e{teroi kV each [and] the other 20
ajna; ajndrw'n rlV, kai; ÔRw'" yV: [have] 130 men each, and
-----------------------------
2
Haldon, “Theory and practice”, pp. 203, 205, 207, 209, 211, 213.
FLEETS FOR THE CRETAN EXPEDITIONS OF 911 AND 949 549
oJmou' ÀewV. oJmou' to; pa'n 700 Rho2s; in all 5,800. In all
ciliavde" kgV kai; wV [duvo,3 the total 23,800.
Reiske].
Dia; tou' qevmato" tw'n Concerning the thema* of the
Kiburraiwtw'n Kibyrrhaio2tai
3 Drovmwne" ieV e[conte" ajna; 15 dromons each having 230
ajndrw'n kwphlatw'n slV kai; oarsmen and each 70
ajna; polemistw'n oV: oJmou' marines; in all 4,000 and
ciliavde" dV kai; fV. 500.
pavmfuloi iıV e[conte" oiJ me;n ıV 16 pamphyloi, 6 of them each
ajna; ajndrw'n rxV, oiJ de; having 160 men, the other
e{teroi iV ajna; ajndrw'n rlV: 10 each 130 men; in all
oJmou' ciliavde" bV kai; sxV. 2,000 and 260.
oJmou' to; pa'n ciliavde" ıV kai; In all, the total 6,000 and 760.
yxV.
Dia; tou' qevmato" th'" Savmou Concerning the thema of
Samos
4 Drovmwne" iV e[conte" ajna; 10 dromons each having 230
ajna; polemistw'n oV: oJmou' ÀgV. marines; in all 3,000.
pavmfuloi ibV, e[conte" oiJ me;n dV 12 pamphyloi, 4 of them each
ajna; ajndrw'n kwphlatw'n rxV, having 160 oarsmen, the
oiJ de; hV ajna; ajndrw'n rlV: [other] 8 each 130 men; in
oJmou' ÀacpV. all 1,680.
oJmou' to; pa'n dia; tou' qevmato" In all, the total for the thema of
th'" Savmou ÀdcpV. Samos 4,680.
Dia; tou' qevmato" tou' Aijgaivou Concerning the thema of
Pelavgou" Aigaion Pelagos
5 Drovmwne" zV e[conte" ajna; 7 dromons each having 230
ajna; polemistw'n oV: oJmou' marines; in all 2,100.
ÀbrV.
pavmfuloi zV e[conte" oiJ me;n gV 7 pamphyloi, 3 of them each
ajna; ajndrw'n rxV, oiJ de; having 160 men, the other 4
e[teroi dV ajna; ajndrw'n rlV: each 130 men; in all 1,000.
------------------------------
3
A copyist’s error in the manuscript. See Haldon, “Theory and practice”, pp. 247
n. 39.
550 APPENDIX FOUR
oJmou' ÀaV.
oJmou' to; pa'n dia; tou' qevmato" In all, the total for the thema of
th'" Aijgaivou Pelavgou" ÀgrV. Aigaion Pelagos 3,100.
Dia; tou' qevmato" ÔEllavdo" Concerning the thema of
Hellas
6 Drovmwne" iV e[conte" ajna; 10 dromons each having 230
ajna; polemistw'n oV: oJmou' ÀgV. marines; in all 3,000.
Dia; tw'n Mardai>tw'n Concerning the Mardaites
7 Mardai?tai, strato;" su;n The Mardaites, army with
ajrcovntwn, ÀdpzV, kai; kata; officers 4,087, and as an
prosqhvkhn e{teroi Àa: oJmou' auxiliary another 1,000; in
ÀepzV. all 5,087
oJmou to; pa'n diav te tou' In all the total for the dromons
basilikou' ploi?mou, diav te of the imperial fleet and the
tw'n qemavtwn drovmwne" rbV,4 themata 112, 75 pamphyloi,
pavmfuloi oeV, a[ndre" 34,000 oarsmen ‹and 200›
kwphlavtai ciliavde" ldV and 7,340 marines and 700
‹kai; sV› kai; polemistai; Rho2s and 5,087 Mardaites.
ÀztmV kai; ÔRw'" yV kai;
Mardai?tai ÀepzV.
AiJ rJovgai dia; tou' basilikou' The pay for the imperial fleet
ploi?mou
8 Strato;" su;n ajrcovntwn The men of the fleet together
ciliavde" ibV kai; fbV. rJovga with officers 12,000 and
kenthnavria ieV, livtrai # : ,, 502. Pay of 15 kente2naria,
iV. 90 litrai, 10 nomismata.
prosqhvkhn aujtw'n Àa ajna; nom. Their auxiliary of 1,000 each 5
eV, ginovmenon livtrai xqV, nomismata, making 69
nom. lbV. litrai, 32 nomismata.
ÔRw'" yV. rJovga kenthnavrin aV. 700 Rho2s; pay of 1
kente2narion.
oJmou' dia; tou' ploi?mou kai; tw'n In all for the fleet and the Rho2s
ÔRw'" rJovga kenthnavria izV, pay of 17 kente2naria, 59
livtrai nqV ,, mbV litrai, 42 nomismata.
Dia; tou' qevmato" tw'n Concerning the thema of the
-----------------------------
4
Sic MS; recte ribV (112).
Kiburraiwtw'n Kibyrrhaio2tai
9 Strato;" su;n ajrcovntwn ÀıyxV. Men of the fleet together with
rJovga kenthnavria bV, livtrai officers 6,760. Pay 2
kaV ,, mbV su;n toi'" diploi'". kente2naria, 21 litrai, 42
nomismata with the
reserves.
Dia; tou' qevmato" th'" Savmou Concerning the thema of
Samos
10 Strato;" su;n ajrcovntwn ÀdcpV, Men of the fleet together with
kai; ajpo; tw'n diplw'n Àa. rJovga officers 4,680, and 1,000
kenthnavria bV, livtrai aV ,, from the reserves. Pay 2
iaV. kente2naria, 1 litra, 11
nomismata.
Dia; tou' qevmato" tou' Aijgaivou Concerning the thema of
Pelavgou" Aigaion Pelagos
11 Strato;" su;n ajrcovntwn ÀgrV kai; Men of the fleet together with
ajpo; tw'n diplw'n Àa. rJovga officers 3,100, and 1,000
kenthnavrin aV, livtrai ndV ,, from the reserves. Pay 1
gV. kente2narion, 54 litrai, 3
nomismata.
Dia; tw'n Mardai>tw'n th'" Concerning the Mardaites of
duvsew" the West
12 Strato;" su;n ajrcovntwn ÀdpzV. Men of the fleet together with
rJovga kenthnavria dV, livtrai officers 4,087. Pay 4
xıV ,, lbV. kai; hJ prosqhvkh kente2naria, 66 litrai, 32
ajndrw'n Àa ajna; ,, hV, nomismata. And the
ginovmenon kenthnavrin aV, auxiliary of 1,000 men 8
livtrai ia ,, hV. nomismata each, making 1
kente2narion, 11 litrai, 8
nomismata.
oJmou' to; pa'n dia; tw'n In all the total for the pay of
Mardai>tw'n th'" duvsew" the Mardaites of the West 5
rJovga kenthnavria eV, livtrai kente2naria, 77 litrai, 42
ozV ,, mbV.5 nomismata.
kai; ojmou' to; pa'n diav te tou' And in all the total for the pay
basilikou' ploi?mou, tw'n of the imperial fleet, the
-----------------------------
5
Sic MS; recte mV (40).
552 APPENDIX FOUR
ÔRw'", tw'n qematikw'n Rho2s, the thematic fleets,

ploi?mwn kai; tw'n and the Mardaites of the
Mardai>tw'n duvsew" rJovga West, 29 kente2naria, 13
kenthnavria kqV, livtrai igV ,, litrai, 66 nomismata.
xıV.
Dia; tou' procrevou Concerning mobilization pay
13 ... ...
ijstevon, o{ti ejdevxato oJ strath- Note that the strate2gos of the
go;" tw'n Kiburraiwtw'n kai; Kibyrrhaio2tai and the
oJ katepavnw tw'n Mardai>tw'n katepano2 of the Mardaites
Attaliva", i{na oJ me;n of Antalya undertook that
strathgo;" eujtrepivsh/ celav- the strate2gos would prepare
ndia duvo ajpo; tw'n oujsiw'n two chelandia from the
tw'n pourmarcw'n, oJ de; kate- ousiai* of the tourmarchai,
pavnw tw'n Mardai>tw'n the katepano2 of the
eujtrepivsh/ galeva", kai; dia; Mardaites would prepare
Martivou mhno;" ajposteiv- galeai*, and during the
lwsi aujta eij" Surivan, i{na month of March would
peri; pavntwn tw'n ejkei'se despatch them to Syria, so
meletwmevnwn kai; prattomev- that they might bring back a
nwn ejnevgkwsin ajpovkrisin report and a true account
kai; ajlhqe;" manda'ton. regarding everything
prepared and done there.
Peri; tw'n ojf eilovntwn Concerning what should have
eJtoimasqh'nai eij" Qra/khsivou", been prepared in Thrake2sion,
h[goun that is to say
14 tw'n kV ciliavdwn tou' kriqarivou of the 20,000 [modioi] of
kai; peri; tw'n mV ciliavdwn barley, and concerning the
tou' te sivtou kai; tou' 40,000 of wheat and biscuit,
paxamativou kai; ajreurivou and flour and concerning
kai; peri; tou' oi[nou tw'n lV the 30,000 of wine and
ciliavdwn kai; peri; tw'n concerning the 10,000
sfaktw'n tw'n iV ciliavdwn. [animals] for slaughter.
15 kai; peri; tou' eJtoimasqh'nai and concerning the preparation
linavrion lovgw/ tw'n proof 10,000 [measures of] flax
puvrwn kai kalafathvsew" for the propyra* and the
ciliavda" iV, i{na e[ch/ eij" ta; caulking, let them be held
Fuvgela, kai; karfiva ciliav- at Phygela, and 6,000 nails
da" ıV lovgw/ th'" hJl wvsew" for the nailing of the
tw'n dromwvnwn. ejdevx ato peri; dromons. The pro2tonotarios

touvtwn oJ prwtonotavrio" tw'n of the Thrake2sion undertook
Qra/khsivwn. ejdevxato kai; oJ these items. And the
Limnogavlakto", i{na sunv- [official] of Limnogalaktos
dravmh/ aujto;n eij" to;n oi\non. likewise undertook to assist
him with the wine.
16 peri; tou' eJtoimasqh'nai Concerning preparing a nail 5-
karfivon pentadaktulai'on fingers [long] for the fabric
lovgw/ th'" strwvsew" tw'n of the dromons, as regards
dromwnivwn, eij" ta;" skavl a" the gangways and as
kai; eij" ta;" pavqna" ciliavda" regards the mangers,
lV, kai; katevlqwsin eij" ta; 30,000, and they [i.e., the
Fuvgela. ejdevxato de; peri; nails] should “go” [i.e., be
touvtou oJ strathgo;" th'" sent] down to Phygela. The
Savmou tou' lambavnein strate2gos of Samos
e[xodon para; tou' undertook to obtain the
prwtonotarivou. expenses concerning this
from the pro2tonotarios.
17 peri; tw'n sandalivwn tw'n Concerning the sandalia made
kamoqevntwn lovgw/ tw'n dro- for the dromo2nia: for the
mwnivwn eij" to;n trevconta, courier, let him be
i{na ajpostalh'/ th'" eJtaireiva" despatched from the
meta; keleuvseu" pro;" to;n hetaireia with an order for
katepavnw, kai; dwvsh/ aujto;n the katepano2, who should
prwtokagkellavrion kai; give him a pro2tokankel-
pa'san sundromhvn, kai; kra- larios and full support, and
thvsh tou;" Korfitiavnou" let him then hold the
ÔHrakleiva", kai; ejpavrh/ nauv- Korphitianoi of He2rakleia
ta" uJpe;r eJkavstou sandav- and take four sailors for
liou dV. ajposteivlh/ de; aujta; each sandalion. He should
dia; suntomiva" dia; tou' send them off without delay
prwtokagkellarivou. i{na de; through the pro2tokankel-
e[ch/ e{kaston sandavlion to; larios. Each sandalion
katavrtion aujtou' kai; to; should have a mast and a
keratavrion kai; ajna; kwpivwn yard and each 4 oars and the
dV kai; to; parakwvpion. steering oar.
554 APPENDIX FOUR
[b] The expedition of 949 6
[b].I
ÔH kata; th'" nhvsou Krhvth" The expedition which took

genomevnh ejkstrateiva kai; place against the island of
ejxovplisi" tw'n te ploi?mwn kai; Crete and the arming of both
kaballarikw'n ejpi; Kwnstan- the ships and the cavalry,
tivnou kai; ÔRwmanou' tw'n under Constantine and
Porfurogennh;twn ejn Cristw'/ Ro2manos,7 the emperors born
pistw'n basilevwn eij" in the purple, faithful in
ijndiktivona zV. Christ, in indiction seven.
1 To; basiliko;n plovi>mon oujsivai The imperial fleet, 150 ousiai,

rnV, ejx w|n pavmfuloi ıV kai; of which 6 [were] hand-
oiJ ajrtivw" kataskeuasqevn- picked (pamphyloi)* and 2
te" bV. recently mobilised.
oujsiaka; celavndia rV. ejx aujtw'n 100 ousiaka chelandia,* of
tw'n rV oujsivwn [Rousivwn, which 100 ousiai, 78 ousiai
Reiske] e[n te Durracivw/ kai; in Dyrrachion and Dalma-
ejn Dalmativa/ oujsivai zV, ejn tia, 3 ousiai in Calabria, 3
Kalabriva/ oujsivai gV, meta; ousiai with the ostiarios and
tou' ojstiarivou Stefavnou kai; nipsistiarios Stephen for
niyistiarivou eij" th;n JIspa- service in Spain.
nivan douliva oujsivai gV.
eij" fuvlaxin th'" qeofulavktou As defence for the God-
povlew" pavmfulo" aV kai; guarded city, one pamphy-
oujsivai kdV. los and 24 ousiai.
2 ta; mevllonta taxeideu'sai ejn The [ships] intended to
Krhvth/ campaign in Crete
pavmfuloi zV, oujsiaka; celav- 7 pamphyloi, 33 ousiaka che-
ndia lgV, oJmou' celavndia mV. landia, 40 chelandia in all.
-----------------------------
6
Haldon, “Theory and practice”, pp. 219, 221.
7
The future Ro2manos II, son of Constantine VII, born 939 and crowned co-
emperor with his father on 6 April 945.
8
Treadgold suggested emendation of zV to kV . See Treadgold, “Army”, p. 146. We
reject this because it was based on an attempt to make the figures add up rather than
on a different reading of the manuscript. Haldon also rejects it but suggests the
attractive emendation of “ejx aujtw'n tw'n JRousivwn ...” to “ejx aujtw'n tw'n rV oujsivwn ...” on
the grounds that the Rho2s are always referrred to elsewhere as ÔRw'" not ÔRousivwn. See
“Theory and practice”, p. 219. Since we are not convinced that the figures were ever
actually intended to add up, and since Haldon’s emendation makes them come close
to doing so in any case, we prefer his emendation to Treadgold’s.
drovmwne" kV ajna; oujsiw'n bV. 20 dromons, each of two

oujsivai mV. ousiai. 40 ousiai.
3 oiJ ÔRw'" a[ndre" fpdV kai; paidiva The Rho2s, 584 men and
ta; poiou'nta taxeideu'sai servants going to campaign,
a[ndre" meV, oJmou' ÔRw'" ckqV. 45 men: 629 Rho2s in all.
4 oiJ Toulmavtzoi a[ndre" txhV, oiJ The Toulmatzoi, 368 men; the
aijcmavl wtoi a[ndre" yV [Ày, prisoners, 700 men.
Reiske]9
5 ejavqhsan eij" fuvl axin th'" For the guard of the City, the
povlew" oiJ strathgoi; ‹tw'n strate2goi of Aigaion
ploi>moqemavtwn: oJ strath- Pelagos with six chelandia
go;"› tou' Aijgaivou pelavgou" pamphyla,* each of 120
meta; celandivwn pamfuvl wn men and 4 ousiaka chelan-
ıV ajna; ajndrw'n rkV kai; dia, each of 108 men.
celandivwn oujsiakw'n dV ajna;
ajndrw'n rhV.
6 kateleivfqh de; kai; miva oujsiva Also left behind was one ousia
eij" to; kovyai th;n th'" ojgdovh" to cut the wood for the
ijnd. [iktivono"] xulhvn. eighth indiction.
7 oJ strathgo;" th'" Savmou meta; The strate2gos of Samos with 6
celandivwn pamfuvlwn ıV ajna; chelandia pamphyla, each
ajndrw'n rnV kai; celandivwn of 150 men, and 6 chela-
oujsiakw'n ıV ajna; ajndrw'n rhv. ndia ousiaka, each of 108
men.
8 ajpestavlhsan de; meta; tou' There were sent away to
prwtospaqarivou Iwavnnou Africa with the pro2tospa-
kai; ajshkrhvth" ejn Afrikh'/ tharios and ase2kre2tis John,
celavndia gV kai; drovmone" dV 3 chelandia and 4 dromons,
ajna; ajndrw'n skV. each of 220 men.
9 oJ strathgo;" tw'n Kiburrai- The strate2gos of the Kibyrr-
wtw'n meta; celandivwn pam- haio2tai with six chelandia
fuvlwn ıV ajna; ajndrw'n rnV kai; pamphyla, each of 150 men
celandivwn oujsiakw'n ıV ajna; and 6 chelandia ousiaka,
ajndrw'n riV: each of 110 men.10
10 kateleivfqh de; kai; eij" fuvla- There were also left behind as
-----------------------------
9
A printing error for yV 700.
10
Treadgold emended “riV ” [110] to “rhV” [108]. See Treadgold, “Army”, p. 146.
However, this was not based on a re-reading of the manuscript but rather on analogy
to the figures elsewhere. We see no necessity to make the emendation because of the
close approximation of the figures in any case and because of the analogy with the
110 men of the ousiai for the dromons.
556 APPENDIX FOUR
xin tou' qevmato" pavmfuloi a guard for the theme 2

bV, oujsiaka; dV. Kateleivfqh pamphyloi, 4 ousiaka
de; kai; eij" to; kovyai th;n th'" [chelandia]. There were
ojgdovh" ijndiktivono" xulh'n also left behind 2 ousiai to
oujsivai bV. cut wood for the eighth
indiction.
11 kateleivfqh de; kai; eij" fuvl axin There were also left behind in
tou' kurou' Stefavnou tou' Rhodes to guard11 the lord
gunaikadelfou' tou' basi- Stephen, the brother in law
levw" ejn ÔRovdw/ oujsiva aV kai; of the emperor, one ousia
dromovnwn dV ajna; ajndrw'n and 4 dromons, each of 220
skV. men.
12 galevai th'" Attaliva" ieV. ejx 15 galeai of Antalya. Of these
aujtw'n kateleivfqh eij" fuvla- 6 galeai were left behind as
xin tou' qevmato" galevai ıV. a guard for the theme.
13 galevai th'" Antioceiva" bV. Two galeai of Antioch. These
Kateleivfqhsan kai; au|tai were also left behind as a
eij" fuvlaxin tou' aujtou' guard for the same theme.
qevmato".
14 galevai th'" Karpavqou. Kate- Galeai of Karpathos. There
leivfqhsan eij" fuvl axin th'" was left behind one galea as
nhvsou Karpavqou galeva aV. a guard for the island of
Karpathos.
15 ajpo; tou' qevmato" Peloponnhv- From the theme of Peloponne-
sou oJ tourmavrch" th'" para- sos the tourmarche2s* of the
livou meta; celandivwn dV. coast with four chelandia.
………. ……….
[b].II12
“Estin hJ ejxovplisi" drovmono" aV The arming of one dromon is
-----------------------------
11
To watch him rather than to protect him. He was a prisoner. This was Stephen
Lekape2nos, son of emperor Ro2manos I Lekape2nos and brother in law of Constantine
VII through the marriage of the latter to his sister Helena in 919. On 20 December 944
Stephen and his brother Constantine deposed their father in order to prevent the
accession of Constantine VII, to whom Ro2manos had given precedence over them in
his will of 943. When Constantine VII seized the throne outright on 27 January 945,
Stephen and Constantine Lekape2nos were sent into prison in exile. The former was
sent first to Pro2te2 in the Princes’ islands in the Sea of Marmara, then to Proikonne2sos,
then to Rhodes, and later to Mityle2ne2. He died in 967.
12
Haldon, “Theory and practice”, p. 225.
1 klibavnia oV. 70 klibania* (lamellar

corselets)
2 lwrivkia yila; lovgw/ tw'n 12 lo2rikia* (light hauberks) for
prwtokaravbwn kai; sifw- the helmsmen and the
narivwn kai; prorevwn ibV, operators of the sipho2nes*
and the bow hands*
3 e{tera lwrivkia koina; iV, 10 other standard corselets
4 kasivdia pV, 80 helmets
5 aujtoprovswpa iV, 10 helmets with visors
6 ceirovyella zugai; hV, 8 pairs of vambraces
7 spaqiva rV, 100 swords
8 skoutavria rJapta; oV, 70 sewn shields
9 skoutavria Ludiavtika lV, 30 “Lydian” shields
10 kontavria meta; tribellivwn pV, 80 trident pikes (corseques)
11 logcodrevpana kV, 20 longchodrepana* (lance-
sickles) (rigging cutters)
12 menauvl ia rV, 100 pikes
13 rJiktavria rV, 100 javelins
14 toxareva" ÔRwmaiva" su;n kovrdwn 50 “Roman” bows with double
diplw'n nV, strings13
15 nauvkla" meta; ceirotoxo- 20 navklai* with cheirotoxo-
bolivstrwn kai; covrdwn bolistrai* (hand-spanned
metaxotw'n kV, crossbows) and silk strings14
16 sagivta" ciliavda" iV, 10,000 arrows
17 mu'a" sV, 200 “mice/flies” (quarrels)15
18 tribovlia ciliavda" iV, 10,000 caltrops
19 ajgrivfou" meta; ajlusidivwn dV, 4 grapnels with chains
20 ejpilwvrika nV, 50 surcoats
21 kamelauvkia nV, 50 kamelaukia*
22 oJ drovmwn ojfeivlei e[cein a[ndra" The dromon should have 300
tV, oiJ me;n slV plovi>moi men, of these 230 men of
kwphlavtai h[toi kai; the ship [should be]
polemistaiv, kai; oiJ e{teroi oV oarsmen and also marines,
a[ndre" polemistai; ajpo; tw'n and the other 70 men
kaballarikw'n qemavtwn kai; marines from the cavalry
ajpo; tw'n ejqnikw'n. themata and the barbarians.
-----------------------------
13
See Haldon, “Theory and practice”, p. 271.
14
15
See Haldon, “Theory and practice”, p. 273 & n. 111.
558 APPENDIX FOUR
dia; tw'n e}x pamfuvlwn for the six pamphyloi
23 dovrka" o{sa" oJdhghvsei oJ Qeo;" as many hide shields as God

to;n basileva to;n a{gion: may guide the holy emperor
[to provide]
24 klibavnia ajna; xV, 60 lamellar corselets each
25 kasivdia ajna; xV, 60 helmets each
26 lwrivkia ajna; iV. 10 hauberks each
dia; tw'n oujsiakw'n celandivwn for the ousiaka chelandia
27 klibavnia ajna; iV, 10 lamellar corselets each

28 kasivdia ajna; iV, 10 helmets each
29 lwrivkia yila; bV, kai; koina; hV. 2 light hauberks and 8 standard
ones
………. ……….
[b].III16
Dia; tw'n ojfeilovntwn Concerning the equipment

frontisqh'nai ajpo; tou' which should have been
sekrevtou tou' eijdikou' eij" provided by the Department
ejxovplisin tw'n kV dromonivwn of the Eidikon for the arming
of 20 dromons
1 molivbin [Molivbion, Reiske] 5 sheets of lead each for the

lovgw/ tw'n kolumbwmavtwn ka(o)lymbomatoi,* total of
[kalumbomavtwn, Reiske] 100 sheets, weighing 3,000
ajna; cartw'n eV. oJmou' cavrtai litrai
rV, staqmivon [sti, Reiske]
livtrai Àg.
2 bursavria lovgw/ tw'n aujtw'n 20 hides for the same
kolumbwmavtwn [kalubomav- kolymbomatoi 17
twn, Reiske] kV,
3 pevtala megavla [lovgw/] portw'n 300 large petala of portai18
tV,
-----------------------------
16
17
18
Elsewhere, the inventories use a similar phrase in specifying various items of
gear needed for siege engines. See Haldon, “Theory and practice”, pp. 225, 227 and
4 tribovlia ciliavde" fV, 500,000 caltrops19

5 pelevkia sV, 200 double-bladed battle axes
6 tzikouvria fV, 500 single-bladed battle axes
7 kevntoukla kata; perivsseian sV, 200 extra [lengths of] felt
8 ajjrmenovpoula kata; perivsseian 100 extra armenopoula*
rV, (small sails)
9 cavl kwma ajrgo;n livtrai sV, 200 litrai of unworked bronze
10 kassivteron livtrai sV, 200 litrai of tin
11 molivbin [molivbion, Reiske] 200 litrai of unworked lead
ajrgo;n livtrai sV,
12 khri;n [khrivon, Reiske] livtrai 100 litrai of wax
rV,
13 skafivdia sV, 200 spades
14 kaldavria sV, 200 tubs
15 kapouvlia ciliavda" bV, 2,000 levers 20
16 ptuavria ciliavda aV, 1,000 spades
17 kovrda" metaxwta;" paceva" 5 heavy silk-spartum bow
spartivna" eV, strings21
18 kai; eij" ta;" mikra;" toxobo- and another 5 spartum [bow
livstra" spartivna" eV, strings] for the small bow-
ballistae*
19 bursavria lovgw/ tw'n celandivwn 100 hides for the chelandia
rV,
20 kavdou" rV, 100 kadoi* (amphorae)
21 ajtevgia kilikevi>na kata; dromwv- 10 “Cilician” [goat’s hair]
nin [dromovnion, Reiske] iV, covers for each dromon, in
oJmou' sV, total 200
22 sfendovnai petzevi>nai kdV. 24 leather slings
------------------------------
cf. Constantine VII, De cerimoniis, II.45 (vol. 1, pp. 670, 671): “..., kai; hJ touvt wn
ejxovplisi". ... pevtala portw'n lovgw/ ejnduvsew" tw'n diafovrwn trocilivwn, ...”, “..., pevtala
portw'n eij" ta; trocivlia sV , ...”. Here the petala of the portai clearly had something to
do with trochilia, the sheaves of blocks, or the blocks themselves.
A block has various parts - a “shell” (the casing) - a “sheave” (the pulley) - a
“pin” (the spindle on which the sheave turns) - a “swallow” (the hole in which the
sheave is set) - a “strop” (a rope around the casing to hold the casing together) - and
an “eye” (a ring formed by running the strop around a ring to allow the block to be
fastened to something else). Porta was not a classical Greek word but Latin
porta/portus could have the sense of an “opening” or a “hole”. Petaleion/petalon
could mean a “leaf”, hence something flat and round, a plate, or a covering over
something. Our best guess at the meaning of petala of portai is therefore the casings
of the swallows of blocks. Cf. Haldon, “Theory and practice”, p. 276.
19
20
Cf. Haldon, “Theory and practice”, p. 276.
21
560 APPENDIX FOUR
23 Ijstevon, o{ti hJ e[xodo" tw'n Note that the expense of the

ajrmevnwn kai; tw'n difqerivwn sails and the leather
ojfeivl ei ejx evrcesqai ajpo to; [screens] should come from
eijdikovn. the Eidikon.
[b].IV22
Dia; tw'n ojfeilovntwn Concerning the equipment

frontisqh'nai ajpo; tou' which should have been
sekrevtou tou' basilikou' provided by the Department
bestiarivou eij" ejxovplisin tw'n of the Vestiarion basilikon for
kV dromonivwn. the arming of 20 dromons
1 Sifwvnia ajna; gV, oJmou' xV, 3 sipho2nia each, in total 6023

2 kai; kata; perivsseian gonavtia and 40 extra gonatia akontia*
ajkovntia meta; boukolivwn mV, for the boukolia* [of the
sipho2nia]24
3 a[rmena kV, 20 sails
4 difqevria xV, 60 leather [screens]
5 parapelevkia kV, 20 anti-axe [screens]
6 kastelwvmata [kastelovmata, kastelo2mata* (pavesades)
Reiske] kata; tuvpon, according to the norm
7 manikevl ia ajna; nV, oJmou' Àa su;n 50 manikelia* (oar sleeves)
tw'n gonativwn aujtw'n, each, in total 1,000, to-
gether with their gonatia*
8 kwpiva ajna; rkV, oJmou' ÀbuV, 120 oars each, in total 2,400
9 calkivsia kV meta; kai; ta; loipa; 20 chalkisia* (block masts),
mavggana, together with the rest of the
mangana* (blocks)25
10 yelliva kV, 20 psellia* (parrels)
11 maxilavria mV, 40 fenders?26
-----------------------------
22
Haldon, “Theory and practice”, p. 227. Most of the items in this list are repeated
in the inventories, together with others, in another list of additional items provided
from the treasury of the Vestiarion basilikon to the droungarios tou ploimou*. See
Part F below.
23
24
25
26
Maxilavrion was not a classical Greek term. Haldon suggests that its meaning
was related to that of Modern Greek maxilavri (maxilari) a pillow or cushion and a
“fender” in a nautical context. See Haldon, “Theory and practice”, p. 281. The
reservation that we have about this is that two fenders per dromon seems very few.
12 perovnia kV, kataprovswpa su;n 20 peronia* (spurs), the kata-

tw'n katakoravkwn aujtw'n, proso2pa* (face things) [of
the bows] together with
their katakorakes* (coup-
lings)27
13 sivdhra bolistika; rkV, 120 side2ra bolistika* (anchors)
14 sidhrobovl ia rkV, 120 side2robolia* (anchor
chains)
15 ajnagokatavgonta su;n tw'n 20 anagokatagonta* (wind-
iJmantarivwn aujtw'n kV, lasses) with their cables28
16 peripetovmena ajna; kdV, oJmou' 24 peripetomena* 29 each, in
upV, total 480
17 filourevai ajna; ibV, oJmou' smV, 12 Linden cables (cables made
from the inner bark of the
Linden tree) each, in total
240
18 seivsta" sV, 200 crowbars
19 tzovkou" uV, 400 sledge-hammers
20 ajxjinoruvgia uV, 400 mattocks
21 perovna" kata; perivsseian sV 200 extra belaying pins
22 karfi;n aJrpavgivn [karfivon ajr- 3,000 hooked/barbed spikes
pavgion, Reiske] ciliavda" gV,
23 gulariko;n30 karfi;n [karfivon, 3,000 screw spikes
Reiske] Àg,
24 tetradaktuliai'on karfi;n 6,000 “four-finger-long”
[karfivon, Reiske] Àı, spikes
25 kai; th'" parhlwvsew" ciliavde" and 12,000 for fastening
ibV,
26 sivdhron ajrgo;n livtrai Àg, 3,000 litrai of unworked iron
27 katziva pV. 80 braziers
-----------------------------
27
Haldon, “Theory and practice”, pp. 227, n. 83, 281-3, points out that this entry
can be read in two ways. As it stands with the punctuation of the manuscript, it can be
read as two separate items: “20 peronia. The kataproso2pa together with their
katakorakes.” If the full stop is removed, however, the text can be read as a single
item. We prefer this because otherwise the specification “kataprovswpa su;n tw'n
katakoravkwn aujtw'n” would have to stand as the only item in the entire list not given
a numerical value and because we have made sense of what kataproso2pa and
katakorakes probably meant in the context of peronia. See above pp. 208-9.
28
29
30
Cf. G.13: gurariko;n.
562 APPENDIX FOUR
[b].V31
ÔUpe;r ejxoplivsew" tetrarevwn On the arming of four

[tetraraivwn, Reiske] dV, tetrareai, four lambdareai,
labdarevwn dV, magganikw'n dV. four manganika32
1 krikevllia lV, 30 rings

2 pavgouroi ieV, 15 clamps 33
3 yelliva lV, 30 shackles/parrels?
4 kai; lovgw/ tw'n megavl wn and for the large bow-
toxobolivstrwn, ballistae,
5 krioi; eij" ta;" celwvna" ieV, 15 rams for the tortoises,
6 daktuvl ioi ieV, 15 bolts
7 bareva" megavla" kV, 20 large weights
8 kai; mikrotevra" lV, and 30 smaller [ones]
9 kai; lovgw/ tw'n megavlwn toxobo- and iron/chains? for the large
livstrwn sivdhra kata; tuvpon, bow-ballistae according to
the norm
10 pivssa livtrai ciliavde" iV, 10,000 litrai of pitch
11 uJgropivssin magarika; 300 round pots of liquid pitch
strogguvla tV,
12 kedreva magarika; nV, 50 pots of pine distillate34
13 linavrin [linavrion, Reiske] 8,000 litrai of linen
livtrai ciliavde" hV,
14 kanavbin [kanavbion, Reiske] 2,000 litrai of hemp
ciliavde" bV,
------------------------------
31
Haldon, “Theory and practice”, p. 229. This inventory is a confused, composite
one which incorporates items both for engines and also for ships. At least by item 10
the inventory has moved to items for ships and it is entirely possible that in fact the
entire inventory should in fact belong to the previous inventory for the 20 dromons.
32
The latest discussion of these engines is in Haldon, “Theory and practice”, pp.
273-5, who argues that tetrareva were catapults powered by men hauling on ropes
(such as are frequently depicted in medieval manuscripts of chronicles of the
Crusades), that labdareva were either another form of stone throwers or a three-legged
anti-personnel device set up around encampments to hinder enemy attacks, and
magganikav, some sort of large bow-ballistae using blocks and tackles to draw them.
33
The coincidence between items E.2, 3, 5 and items G.7, 8, 9 below again casts
doubt on the authenticity of this inventory.
34
According to Pliny, pix and cedrium were distillates from the wood of the pitch
pine: taeda. See Pliny, Natural history, XVI.21-2 [52-3] (vol. 4, pp. 420-23): “Pix
liquida in Europa e taeda coquitur, navalibus muniendis multosque alios ad usus.
lignum eius concisum furnis undique igni extra circumdato fervet. primus sudor aquae
fluit canali; hoc in Syria cedrium vocatur, cui tanta vis est ut in Aegypto corpora
hominum defunctorum perfusa eo serventur. sequens liquor crassior iam picem fundit;
...”. See also Meiggs, Trees and timber, pp. 410-16, 467-71.
15 sandavlou" kV, 20 ship’s boats35

16 sfendovnai sidhrai' ibV, 12 iron slings
17 sidhrobolistika; kata; perivs- 50 extra iron anchors
seian nV,
18 sidhrovbola nV, 50 anchor chains
19 filourevai rV, 100 Linden cables
20 peripetovmena rV, 100 peripetomena
21 spartivna" rV, 100 spartinai* (spartum
cables)
22 leptavria sV, 200 leptaria*
23 tetravkoula eij" ta; sifwvnia rV, 100 tetrakoula* for the
sipho2nia36
24 linavrion eij" tou;" sfovggou" nV 50 [400] linaria (some things
[uV , Reiske], made of flax) for the
sphongoi* (sponges)
25 skalodevmata uV. 400 skalodemata* (mooring
cables)37
26 oiJ ‹ojktw;› [penthvkonta, 24 sipho2nia for the 8 [50]
Reiske], pavmfuloi sifwvnia pamphyloi
kdV:
27 ta; mV oujsiaka; sifwvnia pV. 80 sipho2nia for 40 ousiaka
[ships]
28 karfi;n [karfivon, Reiske], 6,000 deck nails
stegadero;n ciliavde" ıV.
[b].VI38
Ta; ajpo; tou' sekrevtou tou' What was expended from the
eijdikou' ejxodiasqevnta uJpe;r Department of the Eidikon for
tou' taxeidivou th'" Krhvth" the expedition to Crete
-----------------------------
35
By this point the inventory has certainly changed from items for engines to items
for ships. Sandaloi were small boats, as in sandavlion, savndali", savndalo". See Jal,
Glossaire nautique, p. 1315. This would make no sense in the context of engines,
unless the word had an unknown meaning with relation to them. The equation
between the number of sandaloi, 20, and the 20 dromons in the previous inventory
can hardly be coincidental.
36
37
38
Haldon, “Theory and practice”, pp. 229, 231, 233. We have omitted some items
in this inventory between items 1-2 and 21-2 which appear not to have been related to
the naval forces per se.
564 APPENDIX FOUR
1 Ejphvrqh ajpo; to;n zugo;n lovgw/ 24 litrai in gold was taken

ajgora'" diafovrwn eijdw'n tou' from the mint for the
taxeidivou th'" Krhvth" su;n purchase of various items
ejxoplivsei tw'n qV ÔRousikw'n for the expedition to Crete,
karabivwn kai; tw'n bV together with the equipping
monerivwn tw'n aijcmalwvtwn of the 9 karabia of the Rho2s
su;n tw'n oJrisqevntwn ajpo; and the 2 mone2r eis of the
keleuvsew" doqh'nai ta; kata; prisoners, together with the
perivsseian diafovrwn eijdw'n various extra supplies and
eij" to;n qeovswston remaining items ordered by
basiliko;n stovlon tou' command to be given to the
basilikou' ploi?mou kai; God-preserved imperial
loipw'n crusou' [cV, Reiske] fleet of the imperial navy,
livtrai kdV, ta; kai; ejxodias- and which was disbursed as
qevnta ou{tw". follows.
………. ……….
2 ejdovq h uJpe;r ajgora'" tw'n panivwn There were provided for the
tw'n rJasiakw'n lovgw/ poihv- purchase of bolts of cloth
sew" ajrmevnwn enjneva [enjeva, for the manufacture of 9
Reiske], ajna; phcw'n lV tw'n qV sails, each of 30 pecheis, for
karabivwn tw'n ÔRw'", the 9 karabia* of the Rho2s,
3 kai; eJtevrwn ajrmevnwn bV ajna; and another two sails, each of
phcw'n khV tw'n bV monerivwn 28 pecheis, for the two
tw'n aijcmalwvtwn su;n tw'n moneria* of the prisoners,
doqevntwn panivwn rJasikw'n together with extra bolts of
kata; perivsseian tou;" cloth provided for the said
aujtou;" ÔRw'"39 Rho2s,
4 uJpe;r panivwn dia; tw'n for the bolts for them [i.e., the
ajmfotevrwn ‹nomivsmata› 40 ships] all, 1154 ‹nomis-
ÀarndV, mata›,41
5 ta; kai; ajgorasqevnta ajpo; tou;" and which were purchased 42
ajbbavda" eij" ta; ejnoikhka; from the monks as rent for 43
-----------------------------
39
Reiske and Haldon have a full stop here but it does not appear in the manuscript
and we have omitted it since the whole sentence flows on to 1154 [nomismata].
40
Additions supplied by Haldon are between arrows.
41
See Haldon, “Theory and practice”, p. 284. The figure of 1154 nomismata is
impossibly high. Haldon suggests that the copyist of the manuscript mistook the
abbreviation for nomismata for an a (the numerical figure for 1,000), thus
inadvertently making “154 nominsmata” into “1154 ‹nomismata›”. “Theory and
practice”, p. 229, n. 92.
42
“ajgorasqevnta ajpo; tou;" ajbbavda"”, which should mean “purchased from the
monks”, here clearly had the meaning of “paid by the monks”.
th'" monh'" tou' kurou' the monastery of the lord

ÔRwmanou' kai; ajpo; tou;" Ro2manos, and from the
ajnagnafarivou" tou' fovrou drapers on the market, ‹and›
‹kai;› ajpo; diafovrwn proswv- from various people, ‹for›
pwn ‹dia;› tw'n triw'n, nom. tkV the three [groups], 320
kai; dV [ d, Reiske], ta; kai; nom[ismata], and 4 miliare-
ajgorasqevnta katenwvpion sia, and that purchased
tou' sakellarivou kai; tou' under the supervision of the
bestiarivou. [Departments of the] Sakel-
larios and the Vestiarion.
6 ejdovq hsan uJpe;r misqou' tw'n There were provided for the
ajrmenoravf wn tw'n kamovn- payment of sailmakers who
twn ta; aujta; a[rmena su;n made the said sails, together
ajgora'" nhvmato" ,, lgV. with the purchase of thread,
33 nomismata.
7 ejdovq h uJpe;r ajgora'" There were provided for the
xulokeraivwn lovgw/ tw'n purchase of wooden yards
aujtw'n iaV karabivwn ,, ‹...› for the said 11 karabia, ?
nomismata
8 ejdovq h uJpe;r ajgora'" scoinivwn There were provided for the
lovgw/ kruptw'n ejpikhvrwn kai; purchase of cordage for the
podiodrovmwn tw'n aujtw'n iaV concealed waxed (or perish-
ajrmevnwn ,, gV. able) [boltropes] and sheets
of the said 11 sails, three
nomismata.44
9 ejdovq h uJpe;r ajgora'" difqerivwn There were provided for the
idV, Ú xbV, wJ" tw'n ejtevr wn iqV purchase of 14 leather
difqerivwn doqevntwn ajpo; screens: 62 [nomismata],
tw'n ajpokeimevnwn eij" to; the other 19 leather screens
eijdikovn. having been provided from
those stored in the Eidikon.
10 ejdovq h uJpe;r ajleifh'" tw'n aujtw'n There were provided for the
difqerivwn ,, zV 0/m [,, z 0/m, greasing of the said leather
------------------------------
43
“eij" ta; ejnoikhka;” possibly means “drawn on the rents of”. Cf. Haldon, “Theory
and practice”, p. 228.
44
Cf. Haldon, “Theory and practice”, p. 284. The meaning of krupta; ejpivkhra and
podiovdromo" is very unclear. Krupta; implies something concealed, ejp ivkhra
something either waxed or perishable. What they mean together is anybody’s guess.
Bolt ropes for the edges of the sails seems most probable. Waxed ropes used for the
bolt ropes of sails might be concealed, but they would not be especially perishable.
Podiovdromo" we have translated as “sheets” on the analogy to classical Greek pou'"
and medieval Latin and Italian poggia/pozia for a sheet of a sail. However, what the
qualification drovmo" attached to podiov- here was meant to signify escapes us.
566 APPENDIX FOUR
Reiske]. screens, seven nomismata, 0

[miliaresia].
11 ejdovq h uJpe;r ajgora'" maggavnwn There were provided for the
lovgw/ tw'n aujtw'n iaV purchase of blocks for the
karabivwn ,, qV, ‹› ıV [,, q said 11 karabia, 9 nomis-
ıV, Reiske]. mata, 6 [miliaresia].
12 ejdovq h uJpe;r ajgora'" kwpivwn There were provided for the
tpeV ,, eV. purchase of 385 oars, 5
nomismata.
13 ejdovq h uJpe;r nauphghvsew" tw'n There were provided for the
aujtw'n iaV karabivwn tw'n te construction of the said 11
scistw'n kai; tw'n peribovl wn karabia, both their schistai*
aujtw'n kai; loipw'n ,, iaV. and their periboloi,* and
everything else, 11 nomis-
mata.
14 ejdovq h uJpe;r kalafathvsew" tw'n There were provided for the
aujtw'n iaV karabivwn ,, lgV. caulking* of the said 11
karabia, 33 nomismata.
15 ejdovq h uJpe;r ajgora'" eJtevrwn There were provided for the
panivwn rJasiakw'n rV tw'n purchase of another 100
doqevntwn kata; perivsseian bolts of cloth extra to that
eij" to;n aujto;n stovlon ,, khV, provided for the said fleet,
ibV. 28 nomismata, 12 miliare-
sia.
16 ejdovq h uJpe;r ajgora'" bursa- There were provided for the
rivwn boeivwn rkbV ,, phV 0/m. purchase of 122 ox hides,
88 nomismata, 0 [miliare-
sia].
17 ejdovq h uJpe;r ajgora'" kentouv- There were provided for the
klwn sV tw'n doqevntwn kata; purchase of 200 [lengths of]
perivsseian oJmoivw" eij" to; felt, extra to those provided
basiliko;n plovi>mon ,, khV for the imperial ships, 28
0/m. nomismata, 0 [miliaresia].
18 ejdovq h uJpe;r ajgora'" kilikivwn Àa There were provided for the
tw'n ıV ,, rxıV, ıV. purchase of 1,000 “Cili-
cians” (goats’ hair covers),
each at 6 [miliaresia], 166
nomismata, 6 miliaresia.
19 ejdovq h uJpe;r ajgora'" scoinivwn There were provided for the
leptarivwn kai; cartarivwn purchase of cordage, lepta-
kai; loiph'" ejxovdou ,, iıV ria,* and chartaria45 and the

ıV, rest of the outlay, 16 nomis-
mata, 6 miliaresia,
20 lovgw/ poih'sai ajtegivwn [There were provided] for the
kilikivnwn rV ,, rpg 0/m. making of 100 “Cilician”
(goats’ hair) covers, 183
nomismata, 0 [miliaresia].
21 ejdovq h uJpe;r ajgora'" kassi- There were provided for the
tevrou [kasitevrou, Reiske] purchase of 200 litrai of tin,
livtrai sV ,, ldV. 34 nomismata.
………. ……….
22 ejdovq h uJpe;r aJgora'" eJtevrou There were provided for the
kassitevrou [kasitevrou, purchase of another 200
Reiske] livtrai sV ta; doqevnta litrai of tin, which was
Micah;l cuth'/ lovgw/ kata- provided to the caster
kollhvsew" diafovrwn e[rgwn Michael for the brazing
tw'n sifounivwn tou' basili- together of various parts of
kou' ploi?mou ,, l, 0/m. the siphons for the imperial
fleet, 30 nomismata, 0
[miliaresia].
23 ejdovq h uJpe;r aJgora'" khrivou There were provided for the
livtrai rV ,, eV. purchase of 100 litrai of
wax, 5 nomismata.
24 ejdovq h uJpe;r aJgora'" molibivou There were provided for the
ajrgou' livtrai sV ,, dV. purchase of 200 litrai of
unworked lead, 4 nomis-
mata.
25 ejdovq h uJpe;r aJgora'" calkwv- There were provided for the
mato" diafovrou tw'/ doqevnti purchase of various [items]
lovgw/ th'" uJpourgiva" tou' of bronze which was given
drouggarivou tou' ploi?mou for the service of the
uJpe;r kakabivwn megavl wn bV, droungarios of the fleet, for
kai; eJtevrwn kakabivwn two large cauldrons, and
mesaivwn bV, kai; cutroka- two other medium cauld-
kabivwn ganwtw'n dV, kai; rons, and four tinned pot-
koukoumivwn megavl wn bV, kai; cauldrons, and two large
tiganivwn megavl wn bV, kettles, and two large frying
calkostamnivou ganwtou' pans, one tinned bronze urn,
-----------------------------
45
What chartaria mean in this context is obscure. Haldon, “Theory and practice”,
p. 231, n. 102 suggests sheets or strips of anything, such as lead or leather.
568 APPENDIX FOUR
eJno;", flaskivwn ganwtw'n bV, two tinned flasks, two sets

cerniboxevstwn bV ,, kdV. of pitcher and basin, 24
nomismata.
26 oJmou' to; pa'n e[xodo" crusou' [cV , In all, the total outlay [was] 21
Reiske] livtrai kaV ,, nzV, litrai of gold, 57 nomis-
miliarhvsia ıV. mata, 6 miliaresia in gold.
27 ejdovq hsan para; tou' eijdikou' Silk and spartum bowstrings
kovrdai metaxwtai; spar- were provided from the
tivnai. ejdovqhsan e{terai Eidikon. Other small silk
kovrdai metaxwtai; mikrai; bowstrings were provided
lovgw/ tw'n toxobolivstrwn. for the bow-ballistae.
28 ejdovq h oJmoivw" para; tou' eijdi- Likewise there were provided
kou' lovgw/ tw'n kolumbwmav- from the Eidikon for the
twn [kalubomavtwn, Reiske] kolymbo2matoi of the
tw'n celandivwn tou' basili- chelandia of the imperial
kou' ploi?mou ajna; cartivwn eV ships, 5 sheets of lead each,
molivbin [molivbion, Reiske] 100 sheets, that is in weight
cartiva rV, staqmivon [sti, 3,000 litrai.
Reiske] livtrai Àg [gV,
Reiske].
29 ejdovq h to;n drouggavrion tou' There were provided to the
ploi?mou ajpo; tou' kavtw droungarios of the fleet
ajrmamevntou 46 [katepavnw from the lower armoury
tou' a[rmato", Reiske]
30 spaqiva Àg, 3,000 swords
31 skoutavria Àg, 3,000 shields
32 kontavria Àg, 3,000 pikes
33 sagivta" ciliavde" smV, 240,000 arrows
34 eJtevra" sagivta" lovgw/ tw'n another 4,000 arrows, [i.e.,]
toxobolivstrwn muva" ciliav- “mice/flies”, for the bow-
de" dV. ballistae.
[b].VII47
Ta; ajpo; tou' sekrevtou tou'' What was provided as

bestiarivou doqevnta tw'/ additional [things] from the
drouggarivw/ tw'n ploi?mwn kata; Department of the Vestiarion
-----------------------------
46
Emendation suggested by Oikonomides, “To; kavtw ajrmamevnton”.
47
perivsseian uJpe;r tou' to the droungarios to2n

taxeidivou th'" Krhvth". ploimo2n for the expedition to
Crete
1 seivsta" rnV, 150 crowbars

2 perovna" tw'n celandivwn rlV, 130 peronai for the chelandia
3 sfendovbola sidhra' ibV, 12 iron slings
4 tzovkou" smV, 240 sledge hammers
5 tzapiva tV, 300 mattocks
6 perovnia sidhra' megavl a tou' 12 large iron peronia for the
xulokavstrou ibV, xylokastron*
7 pavgouroi sidhroi' ieV, 15 iron “crabs” (clamps?)
8 yelliva sidhra' lV, 30 iron hoops (shackles?)
9 davktuloi ieV, 15 “fingers” (bolts?)
10 katziva hV, 8 braziers
11 bareva" iV, 10 weights48
12 ceirobarevai kdV, 24 hand weights
13 karfi;n [karfivon, Reiske] 4,000 screw spikes
gurariko;n49 ciliavde" dV,
14 karfi;n aJrpavgin [karfivon aJr- 2,000 hooked/barbed spikes
pavgion, Reiske] ciliavde" bV,
15 tetradaktulai'on Àe, 5,000 [spikes] “four-finger-
long”
16 th'" parhlwvsew" Àh, 8,000 [spikes] for fastening
17 stegadero;n Àı, 6,000 deck nails
18 ajkovntia calka' ieV, 15 bronze [headed] akon-
tia*(boathooks)50
19 ganwta; calka' lV, 30 tinned bronze [some-
things]51
-----------------------------
48
49
Cf. D.23: gulariko;n.
50
In classical Greek a kontvo " could be a gaff or boathook. See Jal, Glossaire
nautique, p. 890. Obviously one would not make the entire implement from bronze
but might well make the head from bronze rather than iron in order to resist corrosion.
Cf. above p. 404.
51
This and the following three items look suspiciously like cooking equipment.
The tinned bronze [somethings] may well have been tinned cooking pots. Cooking in
untinned bronze or copper pots is not a good idea; even though many bronze cooking
pots have been found in Byzantine contexts, for example in the seventh-century Yassı
Ada shipwreck. See Bass, et al., Yassi Ada volume I, 269-73. Haldon, “Theory and
practice”, p. 284 suggests that drakton was derived from dravx or dravgma, for a
handful, and that a paradraktion would be a cup or vessel of some kind. He suggests
that here the word may have referred to some part of a block and tackle system and
points out that the verb dravw/dravssomai is used in the Strate2gikon of Kekaumenos for
570 APPENDIX FOUR
20 paradravktia calka' lV, 30 bronze paradraktia52

21 boutiva calka' lV, 30 bronze boutia*
22 ajnagontiteva calka' ieV, 15 bronze anagontitea*53
23 scoiniva sidhrovbola xV, 60 schoinia side2robola (iron
anchor chains)
24 peripetovmena rmV, 140 peripetomena
25 spartivna" rmV, 140 spartinai (spartum cables)
26 leptavria skV, 220 leptaria
27 skalodevmata tV, 300 skalodemata (mooring
cables)
28 koubavria rV, 100 koubaria*54
29 kanavbi [kanavbion, Reiske] 2,000 litrai of hemp
livtrai Àb,
30 kai; ajnti; linarivou ciliavdwn zV, and instead of 7,000 [litrai] of
,, pV, linen, 80 nomismata
31 kai; uJpe;r ajgora'" pivssh" and for the purchase of 10,000
ciliavde" iV, kai; uJgropissiv- [litrai] of pitch, and
ou ciliavde" tV, kai; kedreva" 300,000 [litrai] of liquid
ciliavde" rV, ,, kgV, pitch, and 100,000 [litrai]
of cedar resin, 23
nomismata
32 sivdhron ajrgo;n livtrai Àg, 3,000 litrai of unworked iron
33 krioi; sidhroi' megavloi iV. 10 large iron “rams”
------------------------------
siege hooks, aJrpagaiv (harpagai), lowered from walls to seize the rams of attackers.
Against this we consider that they may have been bronze tripods or something similar
with hooks for suspending pots over fires. The bronze boutia were almost certainly
bronze buckets or tubs with handles for carrying or hanging.
52
53
54
APPENDIX FIVE
NIKEfiPHOROS OURANOS, PERI THALASSOMACIAS,


discussed elsewhere in the text or appendices, are asterisked the first
time they are used. They may be accessed through the Index.
------------------------------
1
Edited from a microfilm of the fourteenth-century manuscript Munich,
Bayerische Staatsbibliothek, Cod. Monac. 452, folios 82r-89v, referred to by Dain and
hereafter here also as MS. N. Dain had wished to publish his edition from this
manuscript but his transcripts were lost in the War and he was compelled to use those
he had made from the sixteenth-century manuscript, Florence, Biblioteca Medicea
Laurenziana, MS. Laurentianus LVII-31, referred to by Dain and hereafter here also
as MS. l. MS. l was copied from MS. N at Corfu by Antonios Eparchos in 1564. We
have compared the Laurentian manuscript to the Munich one but have noted readings
from the former only where they effect the understanding of the text.
The text in MS. N (Cod. Monac. 452) is not rubricated. It has no titles and lacks
the initial letters of paragraphs, which were no doubt intended to be added in red later.
It also has no paragraph numbers in the manuscript. A heading has been added in a
later hand: Peri; tou' ginomevnou eij" th;n qalavssan stovlou (Concerning an expedition
taking place at sea). However, for the sake of convenience we have retained Dain’s
invented title, Peri; qalassomaciva", and his paragraph numbering.
Phonetic orthographic variants in the manuscript have not been noticed; for
example, the oblique cases of dromwn- are frequently spelled drwmwn-.
572 APPENDIX FIVE
Peri; Qalassomaciva"
1 Af h|” de; ei[pomen tau'ta ajparti; i{na diataxwvmeqav soi kai;
peri; th'" ginomevnh" eij" th;n qavlassan mavch" dia; tw'n
dromwvnwn: plh;n eij" me;n ta; polla;2 taktika oujde;n eu{romen
peri; aujth'", ejx w|n de; e[gnwmen hJmei'" skorpista; w|de kai; ejkei'
kai; eij" o{sa ejl avbomen ojlivghn pei'ran. Ex w|n ajnemavqomen para;
tw'n ploi?mwn strathgw'n th'" basileiva" hJmw'n --- a[lla me;n ga;r
ejpoivhsan ejkei'noi o[pisqen, a[lla de; e[paqon u{steron --- ejk
touvtwn hJmei'" ajnalexavmenoi ojlivga o{son ajformh;n dou'nai toi'"
mevllousi mavcesqai kai; eij" th;n qavlassan dia; tw'n dromwvnwn
i{na dioriswvmeqa ejn ojlivgoi" lovgoi”.
2 Prw'ton me;n, strathge;, tou' stovlou ojf eivl ei” e[cein pei’ran kai;
ginwvskein tw'n ajevrwn kai; ajnevmwn ta" kinhvsei": i{na de;
proskoph'/" kai; proginwvskh/" aujta" ajpo; tw'n fainomevnwn
ajstevrwn kai; ajpo; tw'n shmadivwn tw'n ginomevnwn eij" ta; a{stra kai;
eij" to;n h{l ion kai; eij" th;n selhvnhn. ÔArmovz ei de; ginwvskein kai;
se; ta;" ejnallaga;" th;n kairw'n: ajpo; ga;r tou' e[cein pei'ran eij"
aujta; fulavttesqai e[cei" ajsfalh" kai; ajkivnduno" ajpo; tw'n
ceimwvnwn th'" qalavssh".
3 ÔArmovzei kamwqh'nai kai; drovmwna" eij" povl emon tw'n polemivwn
ajrkou'nta" to;n dia; qalavssh": plh;n w{" ejstin hJ katavstasi" tou'
stovlou tw'n polemivwn, ou{tw" i{na poihvsh/" kai; su; th;n
kataskeuh;n tw'n sw'n dromwvnwn dunath;n eij" pavnta pro" to;
ajntimavcesqai. ÔH de; kataskeuh; tw'n dromwvnwn mhvte polu; e[stw
pacei'a, i{na mh; gevnwntai ajrgoi; eij" ta;" ejl asiva", mhvte pavl in
kata; polu; lepthv, i{na mh; uJpavrch/ ajduvnato" kai; saqra; kai;
paraluqh'/ tacevw" uJpo; tw'n kumavtwn kai; uJpo; th'" sugkrouvsew"
tw'n polemivwn, ajlla; suvmmetron ejcevtw th;n kataskeuh;n oJ
drovmwn, i{na kai; ejlaunovmeno" mh; uJpavrch/ ajrgo;" polu;, kai;
kludwnizovmeno" uJpo; tw'n kumavtwn i{na mh; paraluvhtai, h]
sugkrouovmeno" para; tw’n ejcqrw'n i{na euJrivskhtai par aujtou;"
ijscurovtero".
4 Ecevtwsan de; oiJ drovmwne" ajnelliph’ kai; dipla' pavnta ta; pro"
o{plisin aujtw'n, oi|on aujcevnia, kwpiva, skarmouv", tropwth'ra"
kai; ta; a[rmena; de; aujtw'n kai; keratavria kai; katavrtia kai; a[lla
o{sa hJ nautikh; tevcnh ajpaitei'. Ecevtw de; oJ drovmwn kai; ejk peris-
------------------------------
2
Polla; MS. N: palaia; MS. l.
NIKEfiPHOROS OURANOS, PERI QALASSOMACIAS 573
On fighting at sea
1 (= Leo VI, §1) After speaking of these points let me now

discuss with you the warfare that takes place at sea with
dromons. We have found nothing on this except in many
tactical [manuals], from which I have gathered [some
information] scattered here and there, and from events in which
we acquired a little experience. On some matters we learned
from the naval strate2goi* of our empire3 some things [that] they
had done previously [and] others they suffered later. Having
selected a few points from these that can give an introduction to
those intending to fight also at sea with dromons, let me set
[these] out in succinct words.
2 (= Leo VI, §2) First, strate2gos, you should have experience of
the fleet and should know the movements of the airs and winds.
You should look out for and anticipate these from the stars that
appear and from the signs that happen in the stars, the sun, and
the moon. It is appropriate for you to have knowledge of the
changes of the seasons, for from experience in these you may
be preserved safe and sound from storms at sea.
3 (= Leo VI, §§3, 4) It is appropriate that dromons should be

built [that are] adequate for fighting the enemy at sea. However,
you should make the equipment of your dromons correspond to
the condition of the fleet of the enemy [and] able to withstand
them in all respects. The construction of the dromons should be
neither too heavy, or they will be sluggish when under way, nor
on the other hand too light, or they will be weak and unsound
and quickly broken up by the waves or the attacks of the
enemy. Let the dromon have suitable construction, so that it is
not sluggish when sailing and is not broken up by waves in a
gale and, when struck by the enemy, proves stronger than them.
4 (= Leo VI, §5) The dromons should have a complete supply in

duplicate of their tackle, such as rudders (auche2n es*), oars,
tholes (skarmoi*), oar-grommets (tropo2te2res*), and their sails
and yards and masts and everything the nautical art demands.
------------------------------
3
Note that even though Nike2phoros Ouranos was only a magistros himself and
not, of course, emperor, he preserved Leo VI’s syntax here and elsewhere, and wrote
as though he were the emperor.
574 APPENDIX FIVE
sou' xuvl a tina; kavtw eij" to;n pavton kai; sanivda" kai; stuppei'a
kai; pivssan kai; uJgrovpisson: kai; nauphgo;n e{na ejk tw'n ejlatw'n
meta; pavntwn tw'n ejrgaleivwn aujtou’ oi|on skepavrnou, trupavnou,
privono" kai; ei[ ti o{moion.
5 Ecevtw de; oJ drovmwn to;n sivfwna e[mprosqen eij" th;n prwv/ran
kaloevnduton4, wJ" e[cei hJ sunhvq eia, i{na di aujtou' ajpoluvh/ to;;
skeuasto;n pu'r kata; tw'n ejcqrw'n. “Anwqen de; tou' toiouvtou
sivf wno" ejcevtw wJ" pavton ajpo; sanivdwn periteteicismevnon5
gurovqen6 meta; sanivdwn, pro" to; i{stasqai eij" aujto; a[ndra"
polemista;" oi{tine" i{{na mavcwntai pro;" tou;" ejpercomevnou” ajpo;
th'" prwv/ra" polemivou" h] kai; bevl h o{sa a]n qevl wsi kai;
ejpinohvswsin i{na rJivptwsin ajp ejkei' oujk eij" th;n prwv/ran kai; eij"
th;n pruvmnan tou' polemikou', ajlla; kai; eij" o{lon to; polemikovn.
6 Alla; kai; ta; xulovkastra periteteicismevna uJpo; sanivdwn i{na
sthvkwsin eij" tou;" megavlou" drovmwna" pro;" to; mevson tou'
katartivou pro;" to; sthvkein a[ndra" eij" aujta; kai; rJivptein mevson
eij" to; polemiko;n h] livqou" megavlou" mulikou;" h] sivdhra bareva,
oi|on maziva wJ" xifavria, i{na di aujtw'n suntrivywsi to;n polemiko;n
drovmwna h] tou;" o[nta" eij" aujto;n i{na kataklavswsin ejpavnw
aujtou' pivptonta 7: oiJ de; iJstavmenoi eij" ta; xulovkastra ojfeivlousin
ejpicevein tiv eij" to; polemiko;n to; dunavmenon ejmprh'sai aujto; h]
foneu'sai tou;" o[nta" eij" aujtov. Ei|" de; e{kasto" ejk tw'n dromwvnwn
e[stw makrov", suvmmetro" kai; ejcevtw ta" duvo ejlasiva" th;n a[nw
kai; th;n kavtw.
7 Dia; de; eJkavsth" ejlasiva" ejcevtw zugou;" to; ojligovteron ei[kosi
pevnte eij" ou}" kavqhntai oiJ ejlavtai, pro;" to; ei\nai zugou;"8 a[nw
me;n ei[kosi pevnte, kavtw de; oJmoivw" ei[kosi pevnte, oJmou'
penthvkonta zugouv". {Ina de; kaqevzwntai eij" e{na e{kaston zugo;n
ejlavtai duvo, ei|" me;n dexiav, ei|" de; ajristerav, pro;" to; ei\nai o{lou"
tou;" ejlavta", tou;" a[nw kai; tou;" kavtw, eJkatovn: oiJ de; aujtoi; i{{na
w\si kai; stratiw'tai. [Exw de;; touvtwn i{na uJpavrch/ oJ kevntarco" tou'
drovmwno" kai; oJ kratw'n to; flavmoulon kai; oiJ duvo
prwtokavraboi, kai; a[llo" o{sti" aJrmovz ei eij" uJphresivan tou'
kentavrcou. OiJ de; prw/rai'oi ejl avtai duvo oiJ o[nte" eij" th;n a[kran,
oJ me;n ei|" e[stw sifwnavtwr, oJ de; a[llo" i{na bavllh/ ta; sivdhra eij"
------------------------------
4
kaloevnduton MSS N & l. Dain emended to calkw'/ e[nduton on the basis of Leo
VI, Naumachika Leontos Basileo2s (Appendix Two [a]), §6, which is reasonable.
5
periteteicismevnon, thus Dain: periteteicismevnwn MSS N & l.
6
gurovqen, thus Dain: gurw'q en MSS N & l.
7
pivptonta, thus Dain: pivptonta" MSS N & l.
8
zugou;", thus Dain: zuga; MSS N & l.
The dromon should have some extra timbers below the deck
(patos)9, and planks, tow, pitch and liquid pitch. And one of the
oarsmen [should be] a shipwright with all the tools, such as an
adze, an auger, a saw, and such like.
5 (= Leo VI, §6) The dromon should have a sipho2n* (flame-
thrower) in front at the prow, bound well [emend to “in
bronze”] as is the custom, so that processed fire can be thrown
through it against the enemy. Above this sipho2n there should be
a kind of floor of planks fortified all around with planks, so that
marines can stand on it to fight the enemy attacking from the
prow, or so that they can throw whatever weapons they want
and can devise from there, not at the prow and stern of the
enemy but at the whole enemy [ship].
6 (= Leo VI, §7) Moreover, they should set up xylokastra*
(wooden castles), fortified with planks, on the large dromons
towards the middle of the mast, so that men can stand on them
and throw into the middle of the enemy [ship] great mill stones
or heavy iron [weights], like sword-shaped blooms, so that they
can smash the enemy dromon with these or crush those on
board it as [the weights] fall on to it. Those standing on the
wooden castles should also pour onto the enemy ship a
substance that can set it on fire or kill those on it. Each of the
dromons should be of a suitable length and should have two
elasiai* (oar-banks), one above and one below.
7 (= Leo VI, §8) For each oar-bank there should be at least
twenty-five zygoi* (thwarts) on which the oarsmen sit, so that
there are 25 thwarts above and similarly 25 below, making a
total of fifty thwarts. Two oarsmen should sit on each thwart,
one on the right and one on the left, so that in all, with those
above and those below, there should be one hundred oarsmen;
and these should also be soldiers. Apart from these, there
should be the kentarchos* (“captain”) of the dromon and the
one who keeps the standard and two pro2tokaraboi*
(helmsmen), and whoever else is suitable to serve the
kentarchos. Of the two oarsmen at the prow who are at the end,
------------------------------
9
Patos was not a technical term for a deck. However, its sense of “something
trodden upon” seems to imply the deck here.
576 APPENDIX FIVE
th;n qavlassan: oJ de; prw/reu;"10 i{na kaqevzhtai ejpavnw th'" prwv/ra"

ejxwplismevno" meta; tw'n ajrmavtwn aujtou'. ÔO de; kravbato" tou'
kentavrcou ojf eivl ei givgnevsqai ejpi; th'" pruvmnh": a{ma me;n i{na
uJpavrch/ para; mivan ajf wrismevno" eij" aujto;n oJ a[rcwn, a{ma de; i{na
fulavtthtai eij" kairo;n sumbolh'" polevmou ajpo; tw'n rJiptomevnwn
belw'n para; tw'n polemivwn: ajpo; de; tou' toiouvtou krabavtou
blevpei pavnta oJ a[rcwn kai; pro;" th;n creivan keleuvei to;n
drovmwna.
8 Ofeivlousi de; givnesqai kai; a[lloi drovmwne" megalwvteroi

cwrou'nte" ajpo; diakosivwn a]ndrw'n h] plevon touvtwn h] ojl igwvteron
pro;" th;n creivan th;n ajpaitou'san tovte eij" to;n kairo;n kata; tw'n
polemivwn: kai; oiJ me;n penthvkonta i{na ejnergw'sin eij" th;n kavtw
ejlasivan, oiJ de; eJkato;n penthvkonta i{na sthvkwsin o{loi a[nw
ejxwplismevnoi meta; tw'n ajrmavtwn kai; polemw'si pro;" tou;"
ejcqrouv".
9 ”Ina de; poihvsh/" kai; makrotevrou" drovmwna" wJ" galeva" kai;
monhvria11 gorga; kai; ejl affa; pro;" to; e[cein aujta; eij" ta;" bivgla"
kai; eij" ta;" a[lla" spoudaiva" creiva".
10 ”Ina de; poihvsh/" kai; a[lla ploi'a fortika; kai; a[lla ploi'a pavl in
eij" a{per a]n w\si ta; iJppavria, ta; legovmena iJppagwgav, a{per
ojfeivlousin ei\nai eij" to;n stovlon wJ" tou'ldon kai; bastavzein kai;
ta; pravgmata kai; ta;" creiva" tw'n stratiwtw'n dia; to; mh;
barei'sqai eij" aujta12 tou;" drovmwna", ejx airevtw" eij" kairo;n
polevmou, o{tan e[cwsin oiJ drovmwne" creivan ojlivghn trofh'" h]
a[rmavtwn h] a[llwn tinw'n, i{na ajnalambavnwtai ta;" dioikhvsei"
aujtw'n ajpo; tw'n forthgw'n kai; tw'n loipw'n ploivwn w|n ei[pomen. To;
de; povsoi drovmwne" gevnwntai kai; povsoi stratiw'tai i{na w\sin eij"
sujtou;" ouj dunavmeqa oJrivsai, ajll wJ" e[cei kai; ajpaitei' hJ creiva
pro;" to;n kairo;n kai; th;n duvnamin tw'n polemivwn ou{tw" i{na
poihvsh/" kai; to; plh'qo" tw'n dromwvnwn. Kai; pavlin pro;" to;
mevgeqo" tw'n dromwvnwn i{na poihvsh/" ajriqmo;n tou' laou' tou'
ojfeivlonto" ei\nai eij" aujtouv" kai; th;n aJrmovzousan aujtw'n
polemikh;n ejxovplisin.
11 Ta; de; fortika; kai; iJppagwga; ploi'a ejcevtwsan tou;" ajrkou'nta"
eij" aujta; nauvta" e[conta" th;n ejxovplisin aujtw'n, oi|on toxavria kai;
sagivta" kai; rJiptavria kai; a[llo ei[ ti e[cei creiw'de" eij" to;n povl e-
------------------------------
10
prw/reu;", thus Dain: prwrai'o" MS. N, prorai'o" MS. l.
11
monhvria, thus Dain: monevria; MSS N & l.
12
aujta;, thus Dain: aujta;" MSS. N & l.
one should be the sipho2nato2r (operator of the flame thrower),

and the other should throw the “irons” (the anchors) into the
sea. The bowman should be stationed above the prow and
equipped with his weapons. The krabatos* (berth) of the
kentarchos should be at the stern, both so that the archo2n*
(commander) should be set apart in it, and also so that he is
protected in time of attack from the missiles thrown by the
enemy. For the commander can see everything from this berth
and give orders for the dromon as necessary.
8 (= Leo VI, §9) Other larger dromons should be prepared with
space for two hundred men, perhaps more or fewer according to
the need demanded by the moment against the enemy. Fifty
should operate from the lower oar-bank and one hundred and
fifty should all be placed above, fully armed with their
weaponry, and should fight the foe.
9 (= Leo VI, §10) You should also build other longer [emend to
“shorter”]13 dromons, such as galeai* and fast, light mone2ria
(monoremes), so as to have them for sentinels and other
essential tasks.
10 (= Leo VI, §§11, 12) You should build other ploia phortika
(supply ships) and others again on which horses can be loaded,
called horse transports. These should be in the fleet as a kind of
baggage-train and should carry equipment and the necessities of
the soldiers so that the dromons are not burdened with them;
especially in time of battle, when the dromons’ needs for food,
arms, and other things are small, they should undertake the
distribution of these from the supply [ships] and the other ships
that I have mentioned. We cannot be prescriptive about how
many dromons should be built nor how many soldiers should be
in them, but you should build the number of dromons as the
situation requires according to the demands of the moment and
the enemy force. Once again, according to the size of the
dromons you should supply the number of the force that should
be in them and an appropriate warlike armament for them.
11 (= Leo VI, §13) The supply ships and horse transports should
have on board sufficient nautai* (sailors) with their armament,
such as bows and arrows and javelins and anything else neces-
------------------------------
13
Makrotevrou" means “longer”. Here it is most probably a mistake for
mikrotevrou", meaning “shorter”.
578 APPENDIX FIVE
mon dia; ta;" ajnagkaiva" peristavsei". Ecevtwsan de; oiJ toiou'toi

nau'tai kai; a[rmata ejk perissou': pollavki" ga;r leivpousin
a[rmata eij" tou;" stratiwvta" kai; ajnalambavnontai ejx aujtw'n.
Alla; kai; aujta; ta; ploi'a aJrmovz ei i{na e[cwsi mavggana kai;
a[rmata o{sa eijsi pro;" creivan mhv pote kai; sumbh'/ i{na leivywsi
katadapanwvmena eij" tou;" polevmou".
12 Cwri;" de; tw'n stratiwtw'n h[goun tw'n a[nw ejl atw'n o{soi a]n w\sin
ajpo; tou' kentavrcou kai; e{{w" tou' ejscavtou ojf eivlousin ei\nai
katavfraktoi, a[rmata e[conte" skoutavria, mevnaula, toxavria,
sagivta" ejk perissou', spaqiva, rJiptavr ia, lwrivkia, klibavnia
e[conta e[mprosqen pevtala eij tavca kai; o[pisqen oujk e[cousi,
kassivda", ceirovyella, kai; ejx airevtw" oiJ ajgwnizovmenoi kai;
polemou'nte" eij" cei'ra" ejn th'/ sumbolh'/ th'" mavch". OiJ de; loipoi;
pavnte" oiJ mh; e[conte" lwrivkia h] klibavnia, i{na forw'si ta;
legovmena neurikav, a{per givnontai ajpo; diplw'n kendouvklwn. Kai;
i{na sthvkwsin ou|toi o[pisqen tw'n a[llwn dia; to; skevpesqai uJp
aujtw'n i{na toxeuvwsin. Ecevtwsan de; eij" tou;" drovmwna" kai;
livqou" kai; toxavria14 polla; dunavmena rJivptesqai pollav:
rJivptousi ga;r aujta; kata; tw'n polemivwn ajpo; ceirw'n kai; ouj
blavptousin aujtou;" ojl igwvtera para; ta; a[lla a[rmata: a[rmata
gavr eijsi kai; oiJ livqoi eujkovl w" euJriskovmena kai; ajnelliph'.
13 Plh;n mh; rJivptwsi tou;" livqou" movnon w{ste katadapanhqh'nai th;n
duvnamin aujtw'n kai; ajpostaqh'nai h] kai; aujta; ta; rJiptovmena
plhrw'sai, i{na mh; poihvswsi suvskouta oiJ ejcqroi; kai; devx wntai
ta; rJiptovmena, kai; ajf ou| plhrwqw'sin ejx aivf nh" a[rxwntai
polemei'n meta; tw'n spaqivwn kai; tw'n menauvlwn kai; euJrhqw'sin
oiJ me;n ejcqroi; ajpo; ajnapauvsew" kai; gevnwntai ijscurovteroi, oiJ de;
hJmevteroi ajpo; kovpou kai; kataponhqw'sin eujkovlw".
14 Poiou'si de; tou'to oiJ Sarakhnoi; kai; uJpomevnousi prw'ton th;n
bivan tou' polevmou, kai; o{te i[dwsi tou;" hJmetevrou" ajpostaqevnta"
kai; plhrwvsanta" ta; rJiptovmena par aujtw'n oi|on sagivta" kai;
livqou" h] a[lla tinav, tovte ajnaphdw'si kai; a{ma me;n ejkfobou'si
tou;" hJmetevrou", a{ma de; kai; polemou'sin ajpo; ceiro;" meta;
spaqivwn kai; menauvl wn ijscurw'" kai; meta; dunavmew".
15 Dio; prevpei meta; skopou' poiei'n th;n sumbolh;n th;n tou' polevmou,
i{na ma'llon oiJ polevmioi pavqwsin o{sa eijsi; pro;" blavbhn, oujc oiJ
hjmevteroi: aJrmovz ei ga;r tou;" hJmetevrou" fulavttein th;n ijdivan
------------------------------
14
toxavria, thus MSS N & l. However, obviously, one does not throw bows. There
are, apparently, manuscript errors here. Leo VI had “kovclaka"”, pebbles. We suggest
that toxaria should possibly have been riptaria, missiles.
sary for battle in difficult situations. These sailors should also

have extra weapons. For often soldiers run short of weapons
and they can draw on these. It is also appropriate for these ships
to have mangana*15 and other arms as needed, to prevent these
running short when used up in battle.
12 (= Leo VI, §14) Apart from the soldiers or upper oarsmen, [all
others] however many there might be, from the kentarchos
down to the last [man] should be kataphraktoi* --- having [as]
weapons shields, pikes, bows, extra arrows, swords, javelins,
corselets, lamellar cuirasses with plates in front even if they
have none behind, helmets, [and] vambraces --- especially those
engaged in hand-to-hand fighting in the front line of attack in
battle. All the rest who do not have corselets or lamellar
cuirasses should wear what are called neurika, which are made
from double layers of felt. These should stand behind the others
to be protected by them as they use their bows. There should
also be on the dromons stones and many toxaria (bows) able to
be thrown far. For they can throw these at the enemy by hand
and these arms do no less harm than others, for stones are arms
that are easily obtained and abundant.
13 (= Leo VI, §15) However, they should not just throw the stones
in such a way that their energy is expended and they break off,
or use up the missiles, in case the enemy links shields and
absorbs the missiles and, when these are used up, then suddenly
begins to fight with swords and pikes, and since the enemy are
not weary they are stronger while our [men] are also easily
worn down because of their exertions.
14 (= Leo VI, §16) The Saracens do this and at first they endure
the impetuosity of the battle and when they see that our men are
breaking off and have used up the missiles [they have] with
them, such as arrows and stones or other such things, then they
rush out and both terrify our men and also fight hand-to-hand
with swords and pikes strongly and with vigour.
15 (= Leo VI, §17) So you should make your attack on the enemy
with forethought, so that it is rather the enemy who suffer harm
[and] not our [men]. It is appropriate that our [men] preserve
------------------------------
15
Cf. §61 below.
580 APPENDIX FIVE
duvnamin kai; ta;" boula;" aujtw'n ajp ajrch'" e{w" tevlou" th'" mavch"
kai; metrei'n kai; tw'n polemivwn th;n duvnamin kai; th;n proqumivan
kai; ou{tw" poiei'n to;n povlemon.
16 Frovntison de;, strathgev, kai; peri; th'" deouvsh" tw'n stratiwtw'n
dapavnh" pro;" to; e[cein aujtou;" ta; ajnagkai'a i{na mh; gevnhtai
lei'yi" touvtwn eij" aujtou;" kai; h] stasiavswsin h] ajdikw'si kai;
turannw'si tou;" o[nta" eij" th;n hJmetevran cwvran ajnagkazovmenoi
dia; th;n lei'yin tw'n creiwdw'n. All, eij dunatovn ejstin, a[pelqe ejn
tavcei eij" th;n polemivan gh'n kai; ejx aujth'" ejcevtw ta;" creiva" oJ
stratov".
17 Paravggeilon de; kai; toi'" a[rcousin i{na mh; ajdikw'siv tina ejk tw'n
uJpo; cei'ra aujtoi'" stratiwtw'n h] dw'rovn ti par aujtw'n
lambavnwsin16 h] ta;" legomevna" sunhqeiva": th;n ga;r ejndoxovthtav
sou oi[damen wJ" oujde; ejqumhqh'nai duvnasai toiou'tovn ti, ejpeidh;
oujde; aJrmovz ei soi dw'ron oiJondhvpote ajpo; mikrou' h] megavlou ejk
tw'n uJpo; cei'ra soi lambavnein to; suvnolon.
18 Tou;" de; stratiwvta" ejpilevgou ajndreivou" kai; crhsivmou",
ejxairevtw" tou;" eij" th;n a[nw ejlasivan o[nta", oi{tine" kai;
polemou'sin ajpo; ceiro;" pro;" tou;" polemivou". ‘An d eu{rh/" tina;"
ejk tw'n stratiwtw'n ajnavndrou", ajpovlue aujtou;" eij" th;n kavtw
ejlasivan, kai; a]n lavbh/ h] ajpoqavnh/ ti" ejk tw'n stratiwtw'n, i{na
ajnaplhrwvsh/" ejk tw'n kavtw to;n ejkeivnou tovpon.
19 ÔArmovzei gavr i{na ginwvskh/" kai; eJno;" eJkavstou ejk tw'n uJpo; se;
stratiwtw'n th;n e{xin kai; th;n ajndreivan kai; th;n ejpithdeiovthta,
w{sper oiJ kunhgoi; ginwvskousin eJno;" eJkavstou skulivou ta;"
ejpithdeiovthta" kai; e[cousin aujta; e{toima eij" o} qevl ousin.
20 Ou{tw" i{na poihvsh/" kaqw;" pavnta ginwvskei" o{ti ajrkou'si pro;" to;
taxivdion o{per e[cei", oi|on tou;" drovmwna" kai; tou;" ejn aujtoi'"
stratiwvta" kai; ta; a[rmata kai; ta;" trofa;" kai; th;n a[llhn
ajposkeuh;n tou' stovlou h{ti" ojfeivlei ei\nai eij" a[lla ploi'a, wJ"
ajnwtevrw ei[pomen: h}n kai; poih'sai e[cei" eij" tou'l don kai;
ajfei'nai aujth;n eij" ajsfalei'" tovpou" o{tan ejlpivzh/" povlemon.
21 ‘An de; gevnhtai creiva, i{na e[ch/" kai; iJppavria eij" ta; iJppagwga;
ploi'a pro;" to; e[cein se kai; kaballarivou" eij" th;n cwvran tw'n
polemivwn kai; aJplw'" i{na pavnta teleiwvsh/" kai; ou{tw" i{na
peripathvsh/" wJ" aJrmovzei.
22 Kai; prw'ton mevn, pri;n ajpokinhvsh/", i{na leitourghqw'si pavnta ta;;
flavmoula tw'n dromovnwn, eij gevnhtai para; tw'n iJerevwn eujch; pro;"
to;n Qeo;n uJpe;r kateuodwvsew" tou' stratou' kata; tw'n polemivwn.
------------------------------
16
lambavnwsin, thus Dain: lambanei'n MSS N & l.
their own energy and intentions from the beginning to the end
of the battle and measure the energy and eagerness of the
enemy and organize the battle accordingly.
16 (= Leo VI, §18) Take consideration, strate2gos, for the essential
supplies of the soldiers, for them to have what is necessary, and
so that a lack of these things does not arise and they either rebel
or oppress and mistreat those in our territory, compelled
through lack of necessities. But, if possible advance quickly
into enemy land and let the stratos* satisfy its needs there.
17 (= Leo VI, §19) Instruct your commanders that they are not to
wrong any of the soldiers under them or to accept any gift from
them or what is known as the customary perquisites. I know
that your Gloriousness has not been able to consider any such
[thing], since it is not in your character to acccept any gift
whatsoever from [anyone] great or small under your command.
18 (= Leo VI, §20) Choose courageous and reliable soldiers,
especially those on the upper oar-bank who fight the enemy
hand to hand. If you find [that] any of the soldiers are cowardly,
dismiss them to the lower oar-bank, and if any of the soldiers
should be captured or die, you should fill his place from those
below.
19 (= Leo VI, §21) It is appropriate for you to know the attitude
and bravery and capability of each soldier under you, as
huntsmen know the capabilities of each single dog and have
them ready for their requirements.
20 (= Leo VI, §22) You should arrange everything as you know is
sufficient for the campaign you have, such as the dromons and
the soldiers in them and arms and food and the remaining
equipment for the fleet, which should be in other ships, as we
said above; this you should organize as a baggage-train and
leave in safe places when you anticipate fighting.
21 (= Leo VI, §23) If the need arises, you should also have horses
on the horse-transport ships so that you have cavalry in enemy
territory, and [to put it] simply you should arrange everything
and thus advance appropriately.
22 (= Leo VI, §24) First, before you move off, the standards of the
dromons should be blessed, [preferably] with a prayer to God
from the priests for the successful venture of the army against
582 APPENDIX FIVE
Ei\ta i{na dialalhvsh/" pro;" to;n o{lon to;n lao;n kai; pro;" tou;"
a[rconta" pavl in ijdivw" ta; aJrmovzonta pro;" to;n kairo;n kai; ou{tw"
i{na proqumopoihvsh/" to;n strato;n kai; ajpokinhvsh/" o{tan pneuvsh/
ejpithvdeio" a[nemo" kai; oujk ejnantivo".
23 Plh;n mh; peripatw'sin wJ" fqavsousin oiJ drovmwne", ajlla; sth'son
eij" aujtou;" a[rconta" h] kata; pevnte h] kata; trei'", e{na to;n
legovmenon kovmhta, o{sti" e[stw ajrchgo;" tw'n dromwvnwn w|n e[cei"
ajpodou'nai aujtw'/, i{na frontivsh/ ejpimelw'" peri; pavntwn kai;
diatavxh/ pro;" a{panta.
24 OiJ de; toiou'toi a[rconte" tw'n dromwvnwn ojfeivlousin ei\nai uJpo; se;
kai; devcesqai para;; sou' ta; paraggevl mata kai; lalei'n aujta;
pavlin eij" tou;" uJpoceirivou" aujtw'n. Kai; tau'ta me;n i{na givnwntai
ejpi; tou' basilikou' ploi?mou: ejpi; de; tw'n qematikw'n kai; ÔRwmaivwn
[ÔRwmai>kw'n]17 i{na w\si drouggavrioi kai; tourmavrcai, kai; i{na
uJpotavsswntai kai; aujtoi; tw'/ strathgw'/ kai; poiw'si ta;
paraggellovmena par aujtou'.
25 Oujk ajgnow' de; o{ti kata; th;n oJmoivwsin tou' basilikou' ploi?mou
kai; oiJ tw'n pleustikw'n qemavtwn strathgoi; drouggavrioi
ejlevgonto to; palaio;n kai; oiJ o[nte" uJp aujtou;" ejlevgonto kovmhte"
kai; kevntarcoi movnon: ajlla; nu'n to; drouggaravton eJno;" eJkavstou
eij" th;n strathgivda ajnevbh kai; kalei'tai hJ kefalh; strathgo;" kai;
kratei' ajxiwvmata kai; oiJ baqmoi; merivzontai eij" ta;" strathgika;"
tavxei".
26 ”Ina de; gumnavsh/" kai; tou;" ploi?mou" stratiwvta" kai; aujtou;"
tou;" drovmwna" kata; a[llo kai; a[llo sch'ma : kai; a[llote me;n i{na
poihvsh/" th;n gumnasivan kata; e{na e{kaston a[ndra, a[llote de; kai;
kata; perissotevrou", kai; i{na devcwntai katevnanti ajllhvlwn
meta; spaqivwn kai; skoutarivwn. Kai; aujtou;" de; tou;" drovmwna"
ou{tw gumnavsh/" i{na ejpevrcwntai kat ajllhvlwn wJ" ejpi;
paratavxew": kai; a[llote me;n i{na desmw'sin, a[llote; de; i{na
ajpoluvwsin kai; i{na poiw'si kai; aujtoi; kat a[llo kai; a[llo sch'ma
wJ" dh'qen kata; ajllhvl wn th;n sumbolh;n tou' polevmou: kai; i{na
meta; tw'n kontarivwn prowqw'si ta; ploi'a tw'n polemivwn pro;" to;
mh; plhsiavsai kai; dh'sai aujtouv": ouj gavr ejsti pavntote crhvsimon
i{na oiJ polemou'nte" desmw'sin ajllhvlou" meta; sidhrw'n
kamakivwn: givnontai ga;r ejk touvtou kivndunoi pollavki" ou{" ouj
------------------------------
17
This emendation was suggested to us by John Haldon on the grounds that the
distinction made here was between the traditional “Roman” themata and the new
“Armenian” themata of Nike2phoros’s own age such as Lycia, Cilicia, and Northern
Syria.
the enemy. Then you should address the entire force, and the
commanders especially, with suitable words for the occasion,
and so you should inspire the stratos* and move off when a
favourable, and not adverse, wind blows.
23 (= Leo VI, §25) However, the dromons should not advance in
the order in which they arrive, but put commanders in them [in
charge] of every five or three, a so-called kome2s*, who should
be leader of the dromons you have handed over to him, so that
he can have special responsibility in all matters and make all
arrangements.
24 (= Leo VI, §26) These commanders of the dromons should be
under you and receive their instructions from you and report
them in turn to those under their command. This should be the
system in the imperial fleet. In the thematic and Roman [fleets]
there should be droungarioi* and tourmarchai* and they
should be subordinate to the strate2gos and carry out his
instructions.
25 (= Leo VI, §27) I am not unaware that by analogy with the
imperial fleet the strate2goi of the naval themes were formerly
called droungarioi and those under them were called kome2t es
and kentarchoi only. But now [the position of] droungarios of
each [theme] has risen to that of strate2gos and the head [of the
naval theme] is called strate2gos and holds axio2mata,18 and the
positions are classed in the ranks of strate2gos.
26 (= Leo VI, §28) You should exercize the naval soldiers and the
dromons in one manner or another. Sometimes you should hold
the exercize as each individual man and sometimes in larger
groups so that they engage each other with swords and shields.
And you should exercize the dromons in such a way that they
attack each other as if in formation. Sometimes they should
couple and sometimes uncouple and in one manner or another
they should practice the clash of battle against each other. They
should push the ships of the enemy away with poles to prevent
their coming close and coupling. For it is not always advanta-
geous for those fighting to couple themselves together with
side2rai kamakes* (iron rods), for dangers often develop from
this which no one can escape.
------------------------------
18
Axio2mata: “Dignities”. Axio2mata pertained to those official positions for which
the emperor conferred the insignia of office. See Oikonomides, Listes de préséance,
pp. 281-90.
584 APPENDIX FIVE
duvnataiv ti" fugei'n.19

27 Guvmnaze de; aujtou;" kai; kat a[llon trovpon pro;" ta;" ejpinoiva" a}"
e[cei" noh'sai20 o{ti ejndevcontai gevnesqai kata; tw'n polemivwn, i{na
ejk touvtou suneqivzwntai pro;" tou;" ktuvpou" kai; ta;" krauga;" kai;
th;n a[llhn kivnhsin tou' polevmou kai; mh; taravsswntai wJ"
ajguvmnastoi kai; ejxaivvfnh" ajnelpivstw" ejrcovmenoi eij" aujtav.
28 Ou{tw guvmnason kai; sunevqison aujtouv": ei\ta ojfeivlousi plevein

meta; suntavx ew" sunhgmevnoi tosou'ton o{son mh; ejmpodivz ein
ajllhvloi" eij" ta;" ejl asiva" kai; eij" ta;" bivgla" tw'n ajnevmwn ta;"
ginomevna" eij" th;n qavl assan, ajll wJ" paratagh; polla;
gumnasqei'sa, ou{tw" i{na plevwsin. ”Otan de; suskalw'sai pro;" to;
ajplhkeu'sai ojf eivlousi, poiei'n to; skavl wma meta; tavxew" kai;
fevrein tou;" drovmwna" ejnovrdina h] eij" limevna h] eij" a[llon tovpon
ejpithvdeion eij" to; skalw'sai i{na kai; zavl h a[n sumbh'/ mh; lavbwsi.
29 ÔArmovzei de; proginwvskein ajpo; tw'n shmadivwn tw'n ginomevnwn

tw'n ajstevrwn, tou' hJl ivou kai; th'" selhvnh" to; poi'o" a[nemo" mevllei
pneu'sai eij" to;n kairo;n ejkei'non. Kai; pro;" to;n a[nemon ejklevgou
kai; to;n tovpon tou' skalwvmato": kai; ei[per ouj katepeivgei se
ajnavgkh mh; ajpokinhvsh/" eij" ajnepithvdeion plou'n, ajll o{te ejstin
a[nemo" ejpithvdeio" kai; galhvnh kai; o{tan e[ch/" bebaivan ejlpivda
swthriva": aJrmovz ei ga;r uJfora'sqai kai; ta; shmavdia tw'n a[stevrwn
a{per parathrou'sin oiJ plovi>moi kai; a[lla o{sa eijsi; sumfevronta,
kai; ou{tw" i{na pleuvsh/".
30 Eij" de; ta; a[plhkta a]n me;n skalwvsh/" eij" th;n hJmetevran cwvran
kai; oujk e[ch/" fovbon ajpo; tw'n polemivwn, skavl wson meta; tavxew"
kai; mhdevna tw'n ejntopivwn blavpth/" h] ajdikh'/", mhde; aJrpavz h/ kakov",
mhde; poih'/ ejrhvmwsin eij" aujtouv".
31 “An de; plhsiavz h/" eij" th;n polemivan gh'n h] a]n ejl pivzh/" o{ti ejgguv"
eijsiv pou oiJ polemivoi, prevpei pavntw" i{na kai; bivgla" e[ch/" ajpo;
makrovq en kai; eij" th;n qavl assan ajsfavleian pollh;n kai; uJpavrch/"
a[grupno" wJ" eij" paravtaxin, ejpeidh; pollaiv eijsin aiJ ejpiboulai;
tw'n polemivwn. Kai; ga;r h] eij" th;n gh'n euJrhvsousi21 kai; polemh-
sousiv soi, a]n tuvch/ de; i{na ejmpurivswsi kai; to;n stovlon kai; i{na
fanw'si pro;" th;n qavlassan kai; poihvswsi pro;" se; sumbolh;n
------------------------------
19
fugei'n, thus Dain and MS l: fugh'n MS. N.
20
noh'sai, thus Dain: nohvsein MSS N & l.
21
euJrhvsousi, thus Dain, who deleted “a]n”, which immediately follows in MSS N
& l.
27 (= Leo VI, §29) Exercize them in other ways according to the

techniques you have perceived that they can expect from the
enemy, so that in this way they are accustomed to the blows,
cries and general commotion of war and will not be confused
through being untrained should they encounter these things
suddenly and unexpectedly.
28 (= Leo VI, §30) Exercize them and accustom them. Then they
should sail in formation, a sufficient distance apart to prevent
their colliding with each other when rowing and on the watch
for the winds that happen at sea. Moreover, they should proceed
according to the formation which has been often exercized.
When they have to put in at an aple2kton (naval station) they
should make the skalo2ma (landing) in formation and bring the
dromons up in a regular manner either into harbour, or another
place suitable for landing, so that they will not be damaged
should a squall arise.
29 (= Leo VI, §31) It is appropriate to anticipate from the signs
given by the stars, sun and moon what wind is likely to blow at
that season. And choose your landing-place according to the
wind. If there is no urgent need, do not move off on an
inauspicious voyage, but [only move off] when there is a
favourable wind and calm [sea] and when you have sure
expectation of safety. It is appropriate to take into account the
stars’ signs that seamen observe and all other relevant matters
and then set sail.
30 (= Leo VI, §32) If you land in aple2kta in our own territory and
you have no fear from the enemy, land in good order and do not
harm any of the local inhabitants nor should any evil [man] do
them wrong or seize or lay their land waste.
31 (= Leo VI, §33) But if you approach enemy land or you expect
the enemy to be near you, you should always have scouts some
way off and great security at sea and be alert in the formation,
since the devices of the enemy are many. For either they will
find [you] on land and attack you, if they are able to set fire to
the fleet, and should they appear at sea they will attack you
fiercely either by night or by day. And if you find yourself
586 APPENDIX FIVE
polevmou h] th;n nuvkta h] th;n hJmevran. Kai; a]n euJrevqh/" su; ajnevtoi-
mo" kai; ejkei'noi e{toimoi pavntw", i{na se nikhvswsin, a]n de; eu{rw-
si se e{toimon, i[sw" i{na gevnhtai a[prakto" hJ ejpiboulh; aujtw'n.
32 Epei; de; summevtrw" peri; touvtwn dietaxavmeqa, ei[pwmen a[rti ejn
suntovmw/ kai; pw'" i{na paratavxh/" kai; tw'" a[ra poihvsh/" ta;"
sumbola;" eij" tou;" polemivou" kaqw;" dietaxavmeqa kai; eij" ta;"
ginomevna" eij" th;n xhra;n sumbola;" tw'n polemivwn.
33 ”Otan de; ejlpivzh/" kairo;n polevmou, strathgev, suvntaxon tou;"
stratiwvta" kai; cwvrison kata; tavx ei" aujtw'n kai; ajnavgnwqi
aujtoi'" ta; stratiwtika; ejpitivmia, a{per ei[pomen eij" th;n
stratiwtikh;n gumnasivan th'" xhra'" kai; proqumopoivhson kai;
ejndunavmwson aujtou;" ajpo; lovgou kai; parainevsew" kai; dievqison
eij" to;n povl emon, i{na to; me;n dia; to;n fovbon tw'n ejpithdeivwn, to; de;
dia; th;n sh;n paraivnesin gevnwntai ajndrei'oi kai; tolmhroi; kai; eij"
tou;" mevllonta" kinduvnou" tou' polevmou ajgwnivzwntai ajpo;
ceirov".
34 ÔArmovzei de; i{na di ejpidromh'" kai; dia; a[llwn tinw'n
ejpithdeumavtwn kai; strathghmavtwn poih'/" tevcna" kata; tw'n
polemivwn kai; h]22 meq o{lou 23 tou' uJpo; se; laou' kai; stovlou, h]
meta; mevrou" tino;" ejx aujtou': cwri;" ga;r ajnavgkh" megavlh"
katepeigouvsh" oujk ofeivlei" poiei'n dhmovsion povlemon: poll;a;
ga;r sumbaivnousin ejnantiwvmata kai; polla; givnontai eij" to;n
polevmon a} provteron oujk a[n ti" h[lpise.
35 Dia; tou'to pavntote ojfeivlei" fulavttesqai kai; mh; poiei'n
parataga;"24 ejxairevtw" eij" th;n qavl assan e[nqa desmou'ntai met
ajllhvlwn oiJ drovmwne" kai; givnetai mavch ajpo; ceirw'n, h}n ouj
duvnatai ti" fugei'n oujde; euJrei'n to; sumfevron aujtou'.
36 Kai; tau'ta me;n ojf eivl ei" fulavttesqai ei[per ouj qarrei'" eij" to;
plh'qo" tw'n dromwvnwn kai; eij" th;n ajndreivan kai; ejxovplisin kai;
proqumivan tw'n stratiwtw'n wJ" i{na nikhvsh/" tou;" polemivou".
37 Ouj ga;r ejk tou' e[cein se ploi'a polla; kai; megavl a givnetai hJ nivkh
------------------------------
22
“... h] meq o{lou tou' uJpo; se; laou' kai; stovlou, h] ...”, as emended by us by
comparison to Leo VI, Naumachika Leontos Basileo2s, §36 because “... mh; meq o{lou
tou' uJpo; se; laou' kai; stovlou, mhde; ...” as in MSS N & l does not make sense.
23
meq o{lou, thus Dain: meta; o[lou MSS N & l.
24
This does not make sense. Parataghv is an unexpected form. The sense required
by comparison to Leo VI, Naumachika Leontos Basileo2s, §37 is “attack” and the
forms required should be either paravtaxi" or paratavgmata. Obviously, making
attacks was of paramount importance, as Nike2phoros himself emphasized in many
places. It appears that either Nike2phoros himself or, more probably, someone else
involved in the manuscript transmission process, extrapolated from Leo VI’s warning
against over confidence in attack in §37 to a general veto on attacks.
unprepared and they are quite prepared, they will defeat you;
but if they find you prepared, perhaps their devices will achieve
nothing.
32 (= Leo VI, §34) Since we have discussed these [matters]
adequately, let us now say briefly how you should organize a
formation and make attacks on the enemy, as we have indicated
also for attacks that take place on dry [land].
33 (= Leo VI, §35) When you anticipate a period of fighting,
strate2gos, draw up the soldiers and divide them into their
formations and read them the military code of penalities which
we discussed in the course of land-based military training; and
encourage and hearten them with a speech and exhortation and
accustom them to war, so that partly through fear of the
regulations and partly through your exhortation, they become
brave and daring and fight hand-to-hand in the coming dangers
of engagement.
34 (= Leo VI, §36) It is appropriate that, through incursions and
other practices and stratagems, you should contrive ruses
against the enemy, and not with the whole of the force and the
fleet under you nor with a part of it. For without some urgent
and compelling reason you should not begin a general
engagement. Many obstacles arise and many things happen in
war which no one would previously have anticipated.
35 (= Leo VI, §37) Therefore you should always be on guard and
should not make attacks, especially at sea where the dromons
are coupled to each other and hand-to-hand fighting takes place,
which no one can avoid or find [any] benefit from.
36 (= Leo VI, §38) And you should take these precautions if you
are not confident in the number of the dromons, the bravery,
armament, and enthusiasm of the soldiers to enable you to
defeat the enemy.
37 (= Leo VI, §39) Victory in war does not depend on your pos-
588 APPENDIX FIVE
tou' polevmou, ajll ejk tou' e[cein eij" aujta; polemista;" tolmhrou;"
kai; proquvmou" kata; tw'n polemivwn kai; pro; pavntwn ejk tou' e[cein
se th;n tou' qeou' bohvqeian kai; sunevrgeian kai; ajpo; tou' e[cein
aujtou;" kaqaro;n bivon kai; fulavttein dikaiosuvnhn kai; pro;" tou;"
hJmetevrou" kai; eij" tou;" polemivou", eij de; oujde;n poiou'sin eij"
tou;" aijcmalwvtou" h] aijscro;n h] rJuparovn, h] o{sa eijsi;n eij"
aijscuvnhn aujtw'n, h] ei[p er ouj deiknuvousin eij" aujtou;" wJmovthta
kai; ajphvneian, kai; oujk ajdikei'" e[qno" tuvcon h] a[llou" tina;" mh;
ajdikouvmeno" par aujtw'n: tou;" ga;r ajdikou'nta" prevpei
ajmuvnesqai, meta; th'" tou' Qeou' bohqeiva".
38 Eij de; ajpaitei' i{na gevnhtai dhmovsio" povl emo", paravtaxon tou;"
drovmwna" eij" diavfora kai; poikivla schvmata kaqw;" oJ kairo;" kai;
oJ tovpo" ajpaitei'. Plh;n a]n qarrh'/" i[na nikhvsh/" tou;" polemivou"
kai; dia; tou'to poiei'" dhmovsion povl emon, mh; poihvsh/" th;n mavchn
plhsivon th'" hJmetevra" gh'": ejkei' ga;r a]n i[dwsin ajnavgkhn oiJ
stratiw'tai ejlpivzousin i{na kataxulwvswsi kai; swqw'sin: ajlla;
ma'llon plhsivon th'" gh'". Ekei' ga;r a]n i[dwsin ajnavgkhn oiJ
stratiw'tai kai; ejmpevswsin eij" deilivan, ejlpivzousin i{na swqw'si
dia; th'" gh'" kai; rJivptousi gorgo;n ta; a[rmata kai;; oujde;n
protimw'ntai para; th;n fugh;n oi} ei;" kairo;n paratagh'"
prokrivnousi to; zh'n para; to; fugei'n.
39 Pro; de; th'" hJmevra" tou' polevmou aJrmovzei bouleuvesqai meta; tw'n
uJpo; se; ajrcovntwn tiv prevpei i{na poihvsh/", kai; o{per ajpo; koinh'"
boulh'" fanh'/ crhvsimon tou'to i{na poihvsh/". Kai; paravggeilon toi'"
a[rcousi tw'n dromwvnwn i{na w\sin e{toimoi pro;" to; teleiw'sai ta;
bouleuqevnta ka]n a[ra kai; a[llo ti tuvcon ejnantivon ajpanth'/sh ejk
th'" ejpidromh'" tw'n ejnantivwn. Plh;n kai; tovte o{tan ajpanthvsh/ to;
ejnantivon kai; ouj poihvsh/" ta; bouleuqevnta, aJrmovz ei pavntw"
eJtoivmou" a{panta" ei\nai kai; blevpein eij" to;n so;n drovmwna pro;"
to; labei'n ejx aujtou' shmavdion tiv a[ra ojf eivlousi poih'sai: ajf
o{tou de; i[dwsi to; toiou'ton shmavdion, i{na poiw'si kai; ejkei'noi
suntovmw" o{per a]n deivxh/" aujtoi'".
40 Pavntw" ga;r aJrmovzei soi, strathgev, ejpilevxasqai ejx o{lou tou'

stratou' stratiwvta" kai; megalwtevrou" kai; ajndreiotevrou",
e[conta" kai; ajreth;n perissotevran kai; ejx ovplisin meivz ona, kai;
e[cein aujtou;" eij" to;n drovmwna to;n so;n kai; poih'sai aujto;n
pavmfulon. Kai; aujto;" de; oJ drovmwn oJ so;" ojf eivlei kai;
megalwvtero" ei\nai para; tou;" a[llou" pavnta" kai; wJ" kefalh; th'"
paratavxew" o{lh" faivnesqai.
41 ÔOmoivw" de; aJrmovzei kai; tou;" a[llou" tou;" uJpo; se; a[rconta" o{soi
sessing many large ships but on their having bold fighting men
enthusiastic against the enemy, and above all from your having
God’s help and support and from their living pure lives and
preserving justice towards our [subjects] and the enemy, if they
do nothing disgraceful or foul to the prisoners, or what is a
disgrace to them, or if they do not treat them roughly or
harshly, and if you do not injure any people, or any one else,
when you are not injured by them. Wrong-doers should be dealt
with through God’s assistance.
38 (= Leo VI, §40) If a general battle is required, form up the

dromons in a variety of different ways, as the season and
topography require. However, if you are confident that you will
defeat the enemy and you are starting a general battle for this
reason, do not set up the battle near our land for, if they see the
need the soldiers would expect to abandon ship there and find
refuge, but rather near [enemy] land. For, if the soldiers see the
need and succumb to cowardice, they, who in a time of conflict
prefer life to flight, would hope to find refuge on land there and
would abandon their weaponry quickly and put nothing before
flight.
39 (= Leo VI, §41) Before the day of engagement it is appropriate
to discuss with the commanders under you what you should do
and you should act on what appears to the general consensus to
be useful. Issue instructions to the commanders of the dromons
to be ready to carry out what has been planned unless indeed a
contrary decision emerges after an incursion of the opposition.
However, when a contrary decision is made and you do not
carry out what has been discussed, then it is nevertheless
appropriate for all to be prepared and to watch your dromon to
receive from it the signal for what they should do. When they
see this signal, they should do promptly whatever you have
indicated.
40 (= Leo VI, §42) It is certainly appropriate, strate2gos, for you to
select from the whole stratos* larger and braver soldiers with
superior skills and better armament, and to have them on your
dromon and to make it pamphylos*. And your dromon should
be larger than all the others and should be obvious as the head
of the entire formation.
41 (= Leo VI, §43) Similarly, it is appropriate that each of the

590 APPENDIX FIVE
e[cousin uJpoceirivou" auJtw'n25 drovmwna", i{na ei|" e{kasto" a[rcwn

ejpilevxhtai ejk tw'n uJpoceirivwn aujtou' dromwvnwn a[ndra" kai;
pamfuleuvsh/26 to;n i[dion drovmwna kai; pamfuleuvsh/ aujto;n kai;
katasthvsh/. Plh;n kai; aujtoi; oiJ a[rconte" kai; oiJ loipoi; pavnte"
ojfeivlousi blevpein pro;" to;n drovmwna to;n so;n kai; par aujtou'
i{na kanonivz wntai kai; panqavnwsi tiv ojfeivlousi poih'sai eij" to;n
kairo;n tou'' polevmou, a]n a[ra sumbh'/ ti ajnevl piston kai; ejnantivon
pro;" a} e[cei" bouleuvsasqai: kai; prevpei i{na gevnhtai pro;"
ejkei'no pavlin a[llh mevqodo".
42 “Estw de; shmavdion iJstavmenon eij" to;n so;n drovmwna, ei[te
flavmoulon, ei[te bavndon, ei[te a[llo ti, kai; sthkevtw eij" tovpon
uJyhlovn, i{na deiknuvh/" di aujtou' tiv prevpei poiei'n kai; parauta;
poiw'si tou'to oiJ loipoi; drovmwne" ei[te sumbolh;n povlemou, ei[te
ajnacwvrhsin ajpo; polevmou, ei[te i{na gurisqw'si meta; suntavx ew"
eij" to; kuklw'sai tou;" polemivou", ei[te i{na sunacqw'sin eij"
bohvq eian mevrou" tino;" kataponoumevnou, ei[te i{na ajrghvswsi th;n
ejlasivan, ei[te tacuvteron ejlauvnwsin, ei[te e[gkrumma aJrmovzei
genevsqai, ei[te ajpo; ejgkruvmmato" i{na ejxevlqwsin h] a[lla tinav,
i{na manqavnh/ oJ stovlo" o{la ajpo; shmadivwn tou' sou' drovmwno" kai;
ginwvskwsi pw'" ojfeivlousi poiei'n.
43 Ouj ga;r duvnataiv ti" ajpo; fwnh'" h] boukivnou paraggevllein ta;
devonta diav to;n qovrubon kai; th;n tavrach;n kai; to;n h\con th'"
qalavssh" kai; dia; to;n a[llon ktuvpon th'" sugkrouvsew" kai; th'"
ejlasiva" tw'n dromwvnwn kai; to; plevon dia; th;n kraugh;n tw'n
polemouvntwn.
44 To; de; shmei'on o{te qevl ei dei'xaiv ti, h] ojrqo;n iJstavmenon i{na
deiknuvh/ aujtov, h] ejpi; ta; dexia; klinovmenon h] ejpi; ta; ajristerav, h]
tinassovmenon, h] uJyouvmenon, h] camhlouvmenon, h] pantelw'"
ejpairovmenon, h] metatiqevmenon, h] i{na ejnallavsshtai, hJ kefalh;
tou' toiouvtou shmeivou kai faivnhtai a[llh/ kai; a[llh/, h] ajpo; tou'
schvmato", h] ajpo; th'" crova", h[goun i{na ejnallavsshtai h] to; sch'ma
h] th;n croia;n aujth'" kaqw;" ejpoivoun oiJ palaioiv.
45 Eij" ga;r to;n kairo;n tou' polevmou ei\con shmavdion th'" sumbolh'"
kamelauvkion27 mau'ron uJyouvmenon ejpi; kontarivou: ei\con de; kai;
a[lla shmei'a tina uJpodeiknuvmena aujtoi'" oJmoivw". Asfalevs-
teron dev ejstin i{na su; aujto;" meta; ceiro;" sou deiknuvh/" ta;
shmavdia.
------------------------------
25
auJtw'n, thus Dain: aujtw'n MSS N & l.
26
pamfuleuvsh/, thus Dain: pamfileuvsh/ MSS N & l, both here and in the next
clause.
27
kamelauvkion, thus Dain: kamalauvkion MS. N, kamaulauvkion MS. l.
commanders under you, who have dromons subordinate to

them, should choose men from the dromons subordinate to
them and make their own dromon pamphylos* and keep it
pamphylos and maintain it. However, these commanders and all
the others should watch your dromon and regulate themselves
by it and learn what they have to do in time of engagement, in
case anything unexpected happens and contrary to what you
have planned. In these circumstances a different conduct should
be developed.
42 (= Leo VI, §44) Let there be a signal placed on your dromon, a
standard or a banner or something else, and put it in a high
place so that you can use it to show what needs to be done and
the rest of the dromons can immediately do it, an attack in
engagement or withdrawal from engagement, or for them to
curve in formation to encircle the enemy, or to go to the
assistance of a section in difficulties, or to slow the rowing
down, or speed up the advance, or whether it is appropriate to
set up an ambush or come out of an ambush, or anything else,
so that the fleet can receive all [commands] from signals from
your dromon and know what they ought to do.
43 (= Leo VI, §45) For no one can give the necessary [orders] by
voice or by trumpet because of the hullabaloo and confusion
and the noise of the sea and the other din from the collisions
and the rowing of the dromons and, even more, the shouts of
those fighting.
44 (= Leo VI, §46) When the signal is to convey a message, it
should do so either standing upright or inclining to the right or
to the left, or being waved, or raised or lowered, or completely
removed, or altering its position, or it should be changed, and
the signal’s head should look different, either in pattern or
colour, or it should be changed in pattern or colour as the
ancients used to do.
45 (= Leo VI, §47) For in time of engagement they used to have a
signal for attack a black kamelaukion* raised on a pole; they
had some other signals displayed in a similar way. It is safer for
you to show the signals with your own hand.
592 APPENDIX FIVE
46 Poivhson dev, strathgev, pavnta" uJpo; se; a[rconta" o{soi kratou'si

drovmwna" i{na e[cwsin ajsfalh' th;n pei'ran tw'n toiouvtwn
shmadivwn kai; guvmnason aujtou;" eij" tau'ta i{na ginwvskwsi to;
shmavdion e}n e{{kaston dia; tiv givnetai kai; pw'" kai; mhv pw"
sfavllwntai, ajlla; ma'llon i{na gumnasqw'si kalw'" eij" aujta; kai;
poiw'si suntovmw" ta; keleuovmena.
47 Eij dev ejsti tosauvth ajnavgkh h] ejlpivzei" o{ti eujkovlw" nikhvsei" kai;
mevllei" poih'sai dhmovsion povl emon, poivhson th;n paratagh;n tw'n
dromwvnwn kaqw;" ginwvskei" o{ti e[ni aJrmovdion kai; pro;" to;n
kairo;n kai; pro;" to;n tovpon kai; pro;" th;n eJtoimasivan tw'n
polemivwn kai; pro;" th;n paratagh;n aujtw'n: ouj ga;r dunavmeqa a[rti
levgein met ajsfaleiva" peri; tw'n mellovntwn sumbaivnein.
48 “Allote me;n i{na poihvsh/" wJ" si'gma th;n paratagh;n kai; tou;" me;n
a[llou" drovmwna" i{na poihvsh/" e[nqen kakei'qen oi|on kevrata h]
kei'ra" h] ta;" a[kra" i{na sthvsh/" ejx airevtw" tou;" ijscurotevrou"
kai; megalwtevrou": eij" de; to; bavqo" h[goun eij" th;n mevshn
uJpavrch/" su; pro;" to; periblevpein ejkei'qen pavnta kai;; diatavttein
kai; dioikei'n kai; bohqei'n e[nqa ejsti; creiva bohqeiva": e[ce de; kai;
eujkaivrou" tina;" pro;" to; ejpibohqei'n meta; sou' o{te ejsti; creiva
bohqeiva": to; de; sch'ma tou'to th'" paratagh'" to; o]n wJ" si'gma dia;
tou'to givnetai dia; to; ajpokleivesqai e[sw tou' kuvklou th'"
paratagh'" tou;" ejpercomevnou" polemivou".
49 “Allote de; pavlin i{na paratavxh/" tou;" drovmwna" ojrqa; pro;" to;
e[cein i[sa ta; mevtwpa aujtw'n, i{na o{tan gevnhtai creiva
ejpipivptwsin eij" ta;" prwv/ra" tw'n polemikw'n kai; dia; tw'n
sifwvnwn tou' puro;" katakaivwsin aujtav".
50 “Allote de; i{na cwrivsh/" tou;" drovmwna" kai; eij" diafovrou"
paratagav", ei[te eij" duvo, ei[te eij" triva, pro;" to; plh'qo" tw'n
dromwvnwn w|n e[cei". Kai; o{tan poihvsh/ sumbolh;n hJ miva parataghv/,
i{na ejpipevsh/ kai; hJ a[llh h] o[pisqen h] ejk plagivou kata; tw'n
polemivwn wJ" e[ti eijsi;n ejmpeplegmevnoi kai; blevponte" o{ti ejph'lqe
bohqeiva kat aujtw'n parauta; ejxatonou'sin.
51 “Allote de; kai; di ejgkruvmmato" polevmhson aujtoi'": plh;n
plavnhson touvtou" prw'ton di ojlivgwn tinw'n, kai; o{tan ejpipevswsi
kat aujtw'n, tovte i{na fanh'/ to; e[gkrumma kai; paratavxh/ kai;
ejkluvsh/ aujtouv".
52 “Allote de; ajpovluson tou;" drovmwna" ejl afrou;" kai; gorgou;" i{na
poihvswsi pro;" ejkeivnou" sumbolh;n polevmou pro;" ta; polemika;
kai; i{na polemhvsh/" aujta; ajpo; ceirw'n e{w" ou\ kopwqw'si teleivw"
46 (= Leo VI, §48) Ensure, strate2gos, that all the commanders

under you who have control of dromons are very experienced in
these signals, and [you should] practice them in these so that
they know why each signal is made, and how, and they make
no mistakes, but rather are well practised in these matters and
do [what is] ordered promptly.
47 (= Leo VI, §49) If there is a great need or you expect that you
will be easily victorious, and you intend to make a general
attack, organize the formation of the dromons as you know is
suitable to the weather, the topography, the readiness of the
enemy and their formation. We cannot now speak precisely
about what might happen.
48 (= Leo VI, §50) On other [occasions] you should make a
formation like a [capital letter] sigma (i.e., a “Ç” shape) and
you should put the rest of the dromons on one side and the other
like horns or hands and you should make sure that you position
the stronger and larger on the tips. You yourself should be at
the deepest [point], that is, in the middle, to be able to see
everything from there and control and oversee and assist where
there is need of assistance. Have by you at hand some [ships] to
go off to help where there is need of assistance. This sigma-
shaped formation is used for this purpose, to enclose the
enemy’s attacking formation within the circle.
49 (= Leo VI, §51) On other [occasions] again you should form up
the dromons in a straight [line] to have an equal front, so that
when the need arises, they can attack the enemy [ships] at the
prow and burn them with fire from the sipho2nes.
50 (= Leo VI, §52) On other [occasions] you should divide the
dromons into several formations, either into two or into three,
according to the number of dromons you have. When one
formation has made an attack, the other should fall on the
enemy either at the rear or the flank when they are still
engaged, and when they see that reinforcements are attacking
them, they are immediately disheartened.
51 (= Leo VI, §53) On other [occasions] fight them with an
ambush. However, first deceive them with a small [force] and,
when they attack these, then the ambushers should appear and
confuse and scatter them.
52 (= Leo VI, §§54, 55) On other occasions you should send out
light and fast dromons to make a warlike attack on them [the
enemy] against the enemy [ships], and so that you fight them at
594 APPENDIX FIVE
oiJ polevmioi. Kai; tovte i{na ajpoplevxh/" tou;" polemou'nta" met

aujtw'n drovmwna" kata; tw'n polemivwn wJ" e[ti eijsi;n ajpo; kovpou kai;
ejxeluvqhsan ajpo; th'" mavch" kai; ou{tw" i{na poihvsh/" th;n kat
aujtw'n nivkhn. Tou'to de; givnetai ejx airevtw" o{tan perissotevrou"
drovmwna" e[cei" para; tou;" polemivou".
53 “Allote de; schmativsqhti o{ti feuvgei": plh;n e[ce eij" th;n toiauvthn
fugh;n drovmwna" tacei'" kai; poivhson tou;" polemivou" i{na
diwvkwsin ejxopivsw. Kai; o{tan i[dh/" o{ti diwvkonte" parevl usan th;n
duvnamin aujtw'n, tovte uJpovstreyon suntovmw" wJ" e[ti eijsi;n
ejskorpismevna ta; polemika; kai; ejpivpeson aujtoi'" ajpo; tw'n
e[mprosqen: plh;n e[ce par ejkeivnou" drovmwna" perissotevrou"
i{na h] ei|" kata; eJno;" h] duvo ejk tw'n sw'n dromwvnwn ejpevrcwntai
kata; eJno;" polemikou', kai; ou{tw nikhvsei" aujtou;".
54 ÔArmovzei de; poiei'n sumbolh;n pro;" polemivou" eij" qavlassan kai;
o{tan sumbh'/ kindineu'sai aujtou;" eij" th;n qavl assan kai; o{tan ajpo;
zavlh" ejxatonhvswsi taracqevnte" h] i{na ejpevlqh/" eij" nuvkta kai;
ejmprh'sh/" ta; ploi'a aujtw'n h] wJ" e[ti perispw'ntai ejkei'noi eij" th;n
xhra;n h] wJ" ajpaitei' hJ creiva kai; dunhvqh/" ejpinohvsai kai; suv,
ou[tw" i{na poihvsh/" ta;" sumbola;" aujtw'n.
55 OiJ ga;r a[nqrwpoi pollavki" e[cousi gnwvma" kai; ajduvnatovn ejsti
proginwvskein tina; h] prolevgein o{la ta; mevllonta sumbaivnein eij"
ta;" toiauvta" paratagav": kai; dia; tou'to ouj dunavmeqa eijpei'n
ajrtivw" o{la" ta;" kat aujtw'n ajntiparatavxei", ajlla; ma'llon
ojfeivlomen ajnatiqevnai tau'ta pavnta th'/ pronoiva/ tou' Qeou' kai;
parakalei'n aujto;n i{na eij" tou;" toiouvtou" katepeivgonta"
kairou;" duvnhtaiv ti" kai; bouleuvesqai kai; ejnqumei'sqai kai; ta;
aJrmovzonta poiei'n.
56 Polla; de; ejpithdeuvmata kai; oiJ palaioi; kai; oiJ nevoi ejpoivhsan
kata; tw'n polemikw'n ploivwn kai; tw'n polemouvntwn eij" aujtav:
oi|on to; skeuasto;n pu'r, h[goun to; lamprovn, meta; bronth'" kai;
kapnou' tw'n propuvrwn pempovmenon dia; tw'n sifwvnwn kai;
katakai'on autav.
close quarters until the enemy are completely exhausted.28 Then

you should disengage the dromons fighting with them, against
the enemy, when they are still exhausted and downcast from the
battle, and thus you may achieve victory over them.29 This is
best done when you have more dromons than the enemy.
53 (= Leo VI, §56) On other occasions, pretend that you are
fleeing; however, for this kind of flight have fast dromons and
make the enemy follow behind. When you see that they have
broken up their force in the pursuit, then immediately turn
while the enemy are still scattered and attack them from the
front. However, you should have more dromons than them, so
that either one of your dromons attacks one enemy ship or two
attack one, and thus you will defeat them.
54 (= Leo VI, §57) It is appropriate to make an attack on the
enemy at sea, both when they happen to be in danger at sea and
when they are disheartened after being scattered in a squall;
either you should attack and burn their ships at night or when
they are still dispersed on dry land, or you should make attacks
on them as need arises and as you can devise.
55 (= Leo VI, §58) Men often have opinions and it is impossible to
foresee everything or to foretell all that is likely to happen in
these formations. Because of this we cannot discuss precisely
all counter formations against them [the enemy] but we ought
rather to leave all this to the providence of God and pray to Him
that in such moments of urgent crisis one will be able to devise,
invent, and act upon appropriate [measures].
56 (= Leo VI, §59) Men of old and of recent times have invented
many devices against enemy ships and those fighting in them;
such as processed, that is brilliant, fire, which is expelled from
sipho2nes with thunder and smoke from the propyra and sets
them alight.30
------------------------------
28
Nike2phoros breaks off Leo VI’s §54 here, in the process clearly changing Leo’s
meaning. It is almost as though either the feigned retreat referred to by Leo, which
was a standard naval manœuvre in the Middle Ages, was not used in the Byzantine
navy in the age of Nike2phoros, or, and more probably, someone else involved in the
manuscript transmission process was unfamiliar with the stratagem.
The changes to Leo’s §§54-55 made by Nike2phoros in this §52 are the most
radical in the whole constitution.
29
This makes no sense without the mention of sending in a second squadron, here
omitted from the paraphrase of Leo VI, Naumachika Leontos Basileo2s, §55.
30
Although the vocabulary is the same as that of Leo VI here, the meaning
syntactically is different. Propyra has become a nominal rather than adjectival form.
596 APPENDIX FIVE
57 Kai; toxobalivstra"31 kai; eij" ta;" pruvmna" tou' drovmwno" kai; ta;"
prwv/ra" kai; eij" ta; duvo pleura;" aujtou' rJiptouvsa" sagivta"
mikra;" ta;" legomevna" muiva". “Alloi de; kai; qhriva ejpenovhsan
ajpokekleismevna eij" tzukavlia kai; rJiptovmena e[swqen tw'n
polemikw'n ploivwn: oi|on o[f ei" kai; ejcivdna" kai; sauvra" kai;
skorpivou" kai; ta; a[lla o{sa e[cousin ijovn: klw'ntai ga;r ta;
tzukavlia kai; ejkbaivnousi ta; qhriva kai; davknousi kai; foneuvousi
dia; tou' ijou' tou;" e[swqen tw'n ploivwn.
58 Epenovhsan de; kai; e{tera tzukavl ia gevmonta ajsbevstou kai;
rJiptomevnwn tw'n tzukalivwn kai; klwmevnwn oJ ajtmo;" tou' ajsbevstou;
skotivz ei kai; sumpnivgei tou;" polemivou" kai; givnetai mevga
ejmpovdion eij" aujtouv".
59 Kai; trivbolia de; sidhra' rJiptovmena eij" ta; ploi'a tw'n polemivwn
oujk ojlivga lupou'sin aujtou;" kai; ejmpodivz ousin eij" to;n ginovmenon
povlemon pro;" th;n w|ran.
60 ÔHmei'" de; tzukavl ia keleuvomen gevmonta puro;" skeuastou' i{na
rJivptwntai e[swqen tw'n polemikw'n ploivwn: klwmevnwn ga;r tw'n
tzukalivwn, eujkovlw" katakaivontai ta; ploi'a. Krateivtwsan de;
o[pisqen tw'n sidhrw'n skoutarivwn ceirosivf wna a{per ejpoivhsen
a[rti hJ basileiva mou, i{na kai; aujta; ajpoluvswsi to; skeuasto;n
pu'r eij" ta; provswpa tw'n polemivwn. Kai; trivbovlia de; megavl a
sidhra' h] hJlavria ojxeva ejmpephgmevna eij" xuvla strovggula wJ"
pw'ma kai; ejntetuligmevna wJ" stuppei'a kai; eij" skeuh;n kai; eij"
navfqan kai; eij" ta; loipa; ta; kaivonta rJiptovmena kata; tw'n
polemivwn ajpo; pollw'n merw'n kai; pivptonta eij" ta; ploi'a aujtw'n
ejmprhvsousin aujtav. “An de; katapathvswsin oiJ polevmioi th;n
flovga aujtw'n dia; to; sbevsai aujthvn, kah'nai e[cousi oiJ povde"
aujtw'n eij" aujth;n th;n sumbolh;n tou' polevmou kai; ouj mikro;n e[cei
genevsqai ejmpovdion eij" tou;" polemivou".
61 Dunatovn de; ejsti kai; to; dia; geranivwn h] a[llwn ejpithdeumavtwn
oJmoivwn: i{na w\sin wJ" gavmma kai; strevfwntai guvrwqen ejpicu'sai
eij" ta; polemika; ploi'a kai; uJgrovpisson brasto;n h] skeuh;n h]
a[llhn tina; u{lhn: plh;n i{na desmhvswsi prw'ton ta; polemika; oiJ
drovmwne" kai; tovte i{na strevf wntai to; mavgganon kai; ejpicevh/
a{per ei[pomen.
62 Dunato;n dev ejsti kai; to; periegei'rai oJlovklhron to;n polemikovn,

a]n dhvsh/" aujto; eij" to;n drovmwna pleura;n para; pleura;n kai;
------------------------------
31
toxobalivstra", thus Dain: toxobolivstra" MSS N & l.
57 (= Leo VI, §60) There should be toxobalistrai* (bow-ballistae)

at the stern of the dromon and the prow and along the two sides
to shoot the small arrows known as muiai* (flies). Others have
thought of putting poisonous creatures into pots and throwing
[them] into the enemy ships; such as snakes, vipers, lizards,
scorpions and other venomous [creatures]; for the pots break
and the creatures come out and bite and kill with their venom
those in the ships.
58 (= Leo VI, §61) They have also devised other pots full of
unslaked lime and when these pots are thrown and broken, the
fume from the lime kills and chokes the enemy and causes them
great confusion.
59 (= Leo VI, §62) Iron caltrops thrown into the ships of the
enemy cause them no little harm and hinder them in the fighting
taking place at that time.
60 (= Leo VI, §§63-66) We give instructions that pots full of
processed fire should be thrown into the enemy ships; when the
pots break, the ships are easily burned. Hand-sipho2nes, which
my Majesty (i.e., Leo VI) recently made, should be held behind
iron shields so that they too throw processed fire into the faces
of the enemy. Large iron caltrops or sharp nails embedded in
round [pieces of] wood like a disk, and wrapped round with tow
[soaked] in the processed [fire] and naphtha and other
combustibles, and thrown at the enemy from many directions
and landing on their ships, will set them ablaze. If the enemies
stamp on the flame to extinguish it, their feet will be burned
during this attack and this is no small inconvenience for the
enemies.
61 (= Leo VI, §67) There can also be [made] a [device] by means

of gerania* (cranes) or other similar contrivances. These should
be shaped like a [capital letter] gamma (i.e., a “G” shape) and
should turn around to pour burning wet pitch or the processed
[fire] or anything else onto the enemy ship. However, the
dromons should first couple to the enemy [ships] and then they
should turn the manganon and it should pour the [substances]
we have mentioned.
62 (= Leo VI, §68) It is possible to overturn an entire enemy [ship]
if you couple it to the dromon side by side, and the enemy rush
598 APPENDIX FIVE
sunacqw'sin oiJ polevmioi eij" to; mevro" ejkei'no to; o]n eij" to;n
drovmwna wJ" e[cousin e[qo" pro;" to; poih'sai ajpo; ceiro;" mavchn
ejlpivzonte" o{ti ejpakoumbivz ei to; ploi'on aujtw'n eij" to;n drovmwna:
ei\ta i{na ejpevlqh/ a[llo" drovmwn kata; th'" pleura'" tou' polemikou'
th'" ou[sh" eij" th;n pruvmnan kai; i{na sugkrouvsh/ kai; prwvsh/
ijscura; to; toiou'ton polemikovn: kai; oJ me;n prw'to" drovmwn oJ
dhvsa" to; polemiko;n i{na dunhqh'/ ajpoluvsein aujto;n ejk tou' desmou'
kai; uJpocwrh'sai ojlivgon pro;" [to;]32 mh; e[cein eij" aujto;
ajkouvmbisma to; polemikovn: oJ de; a[llo" drovmwn i{na barhvsh/ o{son
duvnatai kai; a]n gevnhtai ou{tw", i{na periegeivrh/ to;33 polemiko;n
meta; tw'n o[ntwn ajndrw'n eij" aujtov. Prevpei de; i{na mh; dhvsh/" o{lon
to; polemikovn, ajll ojlivgon ti, i{na ajfhvsh/" pleura; gumna; eij" th;n
pruvmnan tou' polemikou', eij" a} i{na sugkrouvsh/ oJ drovmwn pro;" to;
periegei'rai to; polemiko;n meta; tw'n polemivwn.
63 Anagkai'on de; faivnetai moi kai; o{per ejpenohvsamen hJmei'" i{na
ajpo; th'" kavtw ejl asiva" tou' drovmwno" dia; tw'n truphmavtwn tw'n
kwpivwn ejkbaivnonta ta; mevnaula sfavzwsi tou;" polemivou".
64 “Esti; de; kai; a[llo ajnagkaivoteron a]n eu{rh/ cei'ra" ejpidexivou",

to; ejkbavllein ajpo; tw'n eijrhmevnwn truphmavtwn tw'n kwpivwn th'"
kavtw ejl;asiva" mevnaula kai; truph'sai to; polemiko;n w{ste
eijselqei'n u{dwr kai; gemivsai aujtov. Epenovhsan de; kai; a[lla tina;
ejpithdeuvmata oiJ ajrcai'oi eij" to;n povl emon th'" qalavssh". Eijsi; de;
kai; a[lla dunavmena ejpinohqh'nai, ajll ejq evlonte" th;n suntomivan
oujc ijscuvomen gravfein aujtav: eijsi; gavr tina ejx aujtw'n ajsuvmfora
pro;" to; gravfesqai i{na mh; givnwntai fanera; eij" tou;" polemivoi",
i{na mh; ma'llon ejkei'noi poiw'sin aujta; kaq hJmw'n. Ta; ga;r
strathghvmata a]n a{pax katanohvswsin oiJ polevmioi, duvnantai
ajntistrathgei'n kai; ejpithdeuvein aujtav: dia; tou'to prevpei
e{kaston o{per a]n ejpinohvsh/ e[cein aujto; ejn musthrivw/, e{w" a]n
poihvsh/ aujtov.
65 Eij" de; to; biblivon tw'n ajrcaivwn taktikw'n kai; strathghmavtwn
ejpeunw'n euJrhvsei" kai; perissovtera: ouj gavr ejstin, wJ" ei[rhtai, 34
dunato;n gravf ein pro;" pavnta ta; mevllonta givnesqai dia; to; ei\nai
a[peira.
66 Plhvn, i{na ei[pw to; kuriwvteron, e{stwsan oiJ drovmwne"
ejxwplismevnoi teleivw" ajpo; stratiwtw'n ajndreivwn kai; dunamev-
nwn ajpo; ceiro;" polemei'n kai; tolmhrw'n kai; gegumnasmevnwn:
------------------------------
32
to; added by Dain. MSS N & l do not have this.
33
to;, thus Dain: to;n MSS N & l.
34
ei[rhtai, thus Dain and MS. l: ei[rh MS. N.
to that part which is against the dromon, as is their cusom, to

engage in hand-to-hand fighting, hoping that their ship will lay
against the dromon. Then another dromon should run at the
enemy [ship]’s side at the stern and should strike and push the
enemy [ship] severely. And the first dromon, the one coupled to
the enemy [ship], should be able to free itself from the coupling
and back off a little so that the enemy [ship] does not lay
against it. The other dromon should weigh down as much as it
can and if it does this, it should up end the enemy [ship] with
the men in it. You should not couple the whole enemy [ship]
but only a little, so that at the enemy’s stern you leave the sides
bare, where the dromon can strike in order to overturn the
enemy [ship] with the enemy [crew].
63 (= Leo VI, §69) [A technique] which we have devised seems to

me useful: when pikes, coming out through the trype2mata*
(oarports) for oars on the lower bank on a dromon, slaughter the
enemy.
64 (= Leo VI, §§70, 71) There is something even more useful if
experienced hands are available, [and that is] to thrust pikes out
from the above-mentioned oar-ports of the lower oar-bank and
make a breach in the enemy [ship] so that water enters and
floods it. The ancients invented other devices for naval warfare
and others can also be invented, but since I wish to give a
summary, I cannot describe them. It is also inappropriate for
some to be described, to prevent their becoming known to the
enemy and indeed their using them against us. For if the enemy
once get information about a stratagem, they can work out a
counter-stratagem and put it into practice. And so every
[scheme] that is invented [should be] kept secret until it can be
carried out.
65 (= Leo VI, §72) You will find more [information] when you
look in the book of ancient tactics and strategies. For it is
impossible, as has been said, to write about everything that will
happen since these are infinite.
66 (= Leo VI, §73) But, let me mention the more important
[point]. The dromons should be completely armed with brave
soldiers capable of fighting at close quarters, and bold and
600 APPENDIX FIVE
ejcevtwsan de; a[rmata kai; ejxovplisin oi{an wJrivsamen i{na e[ch/ oJ

stratiwvth" oJ w]n eij" th;n xhra;n oJ katavfrakto": kai; ou{tw"
e[stwsan ejx wplismevnoi pavnte" oiJ th'" a[nw ejlasiva".
67 Pro;" de; th;n poiovthta tw'n ejcqrw'n kai; pro;" to; plh'qo" tw'n ploiv-
wn aujtw'n poivhson kai; suv, strathgev, drovmwna" kai; ejxovplison
aujtou;" pro;" to; mh; e[cein to;n drovmwna hJmw'n ojligovteron strato;n
para; to; polemiko;n ploi'on, o{sti" ejx airevtw" eJtoimavzetai dh'sai
meta; tou' polemikou' kai; polemh'sai: a;ll ei[ ejsti dunato;n i{na
e[ch/ kai; perissovteron strato;n oJ hJmevtero" drovmwn: polemouvn-
twn ga;r ajndreivw" kai; tw'n duvo oiJ perissovteroi nikhvsousin a[n.
68 Eij de; sunora'/" o{ti e[cousin oiJ polevmioi ploi'a e[conta
perissovteron stratovn, bavle kai; su; plh'qo" perissovteron eij"
tou;" drovmwna" kai; oujsivwson aujtouv". Plh;n e[klexai ajpo; pavntwn
tou;" kreivttona" a[ndra" kai; ejx aujtw'n ejxovplison th;n ajrkou'san
duvnamin eij" drovmwna" teleivou" kai; ijscurotavtou": kai; h] tw'n
duvo dromwvnwn to;n strato;n, a]n tuvch/, i{na bavlh/" eij" to;n e{na
drovmwna, h] ejk pavntwn i{na ejpilevxh/" tou;" kreivttona", wJ"
ei[rhtai, kai; gevnwntai a[cri diakosivwn stratiwtw'n h] kai;
perissovteroi kaq e{na35 e{kaston drovmwna, i{na kai; ajpo; tou'
plhvqou" kai; ajpo; tou' megevqou" tw'n dromwvnwn kai; ajpo; th'"
ajndreiva" tw'n stratiwtw'n nikhvsh/" su;n Qew'/ ta; polemika; ploi'a.
69 ”Ina dev ejxoplivsh/" kai; mikrotevrou" drovmwna" kai; ejl afrotevrou"
para; tou;" a[llou" ou}" e[comen sunhqeiva/, 36 i{na kai; eja;n diwvkwsi
tou;" polemivou", fqavswsin aujtav, kai; eja;n diwvkwntai para; tw'n
polemivwn, mh; fqavnwntai par aujtw'n, kai; touvtou" i{na e[ch/" eij"
kairo;n creiva" pro;" to; duvnasqai aujtou;" h] poih'saiv ti kako;n
tou;" ejcqrou;" h] mh; paqei'n ti kako;n par aujtw'n.
70 Mikrou;" de; kai; megavlou" drovmwna" poih'son pro;" th;n poiovthta
tw'n polemouvntwn soi ejq nw'n. Ouj ga;r to;n aujto;n e[cousi stovlon
oiJ Sarakhnoiv. e[cousi de; kampavria37 megalwvtera kai;
ajrgovtera: oiJ de; ÔRw'soi ajkavtia mikrovtera kai; ejlafrovtera kai;
gorgav: diabaivnousi ga;r potamouv" kai; ou{tw" ejmbaivnousin eij"
to;n Eu[xeinon Povnton: kai; dia; tou'to ouj duvnantai e[cein
megalwvtera ploi'a.
------------------------------
35
kaq e{na, thus Dain: kata; e[na MSS N & l.
36
sunhqeiva/, thus Dain, following Desrousseaux: sunhvqeian MSS N & l.
37
kampavria MS. N, kampavdia MS. l, kai; mpavdia Dain. The koumbavria of Leo
VI, Naumachika Leontos Basileo2s, §77 became kampavria in MS. N, which was
misread as kampavdia by Antonios Eparchos in MS. l, and this was then guessed at as
kai; mpavdia by Dain. Something of an object lesson in manuscript transmission
processes!
exercized. They should have weaponry and armament such as

we have decreed the soldier who is on land, the kataphraktos,
should have. All those in the upper oar-bank should be armed
like this.
67 (= Leo VI, §74) You, strate2gos, should build dromons to match
the quality of the enemy and the number of their ships and arm
them so that our dromon has a stratos no fewer than the enemy
ship, [and] one that is especially prepared to couple with the
enemy ship and fight; but if it is possible our dromon should
have a larger crew, for when two crews fight bravely, the larger
will win.
68 (= Leo VI, §75) If you realize that the enemy has ships with a
larger stratos, put a larger number into the dromons and
ousia*38 them. However, choose the stronger men from the
whole [force] and from these arm a sufficient force of effective
and very strong dromons. Either, if this is [what] happens, put
the crew from two dromons into one dromon, or choose the best
from all, as has been said, and there should be up to two
hundred soldiers or more on each dromon, so that through the
number and size of the dromons and the bravery of the soldiers
you may, with God, defeat the enemy ships.
69 (= Leo VI, §76) You should arm dromons [which are] smaller
and lighter than those we usually have, so that if they pursue the
enemy, they can catch up with them, and if they are pursued by
the enemy, they are not caught up with by them. You should
have these for a time of need, so that they can either inflict
some damage on the enemy, or avoid damage from them.
70 (= Leo VI, §77) Build small and large dromons according to
the quality of the peoples warring against you. For the Saracens
do not have the same fleet; they have larger and slower
kamparia*. The Russians have smaller, lighter, and fast akatia
for they cross rivers and thus come down into the Black Sea
and so cannot use larger ships.
------------------------------
38
See Appendix Two [a], §75 and n. 58.
602 APPENDIX FIVE
71 Kai; tau'ta me;n ei[pomen peri; tw'n paratagw'n. ”Otan de; qevlh/"
cwrisqh'nai ajpo; th'" mavch", poivhson wJ" si'gma th;n paratagh;n
tw'n dromwvnwn kai; ou{tw" uJpocwvrhson ejxopivsw, oi|on hJ prwv/ra
i{na blevph/ pro;" tou;" polemivou" kai; hJ pruvmna i{na uJpavrch/ ojpivsw:
ajsfale;" gavr eJsti tou'to to; sch'ma th'" paratagh'" kai; o{tan
uJpavgh/" pro;" tou;" polemivou" kai; o{tan uJpocwrh'/" ejx aujtw'n, wJ"
marturou'si tine" tw'n palaiw'n poihvsante" aujtoiv. Kai; ga;r o{tan
uJpocwrh'/" ejx aujtw'n ouj feuvgei", ajlla; fugomacei'" kai; e[cei" kai;
tou;" drovmwna" eJtoivmou" pro;" to; ejpelqei'n pavlin kat aujtw'n a]n
gevnhtai creiva ejk tou' e[cein se ta;" prwv/ra" pro;" aujtouv": kai; oiJ
polevmioi pavl in ouj qarrou'sin ejmbh'nai polla; eij" to; kuvklwma
th'" paratagh'" sou, uJforwvmenoi i{na mh; kuklwqw'sin.
72 Af ou| de; luqh'/ oJ povl emo", aJrmovzei sev, strathgev, ta;
krathqevnta ajpo; tw'n polemivwn diamerivz ein ejpivsh" eij" tou;"
stratiwvta" kai; poiei'n trapevz a" kai; kalei'n kai; filofro-
nei'sqai: kai; oiJ me;n ajndragaqhvsante" i{na lavbwsi kai; dwrea;"
kai; timav", oiJ de; poihvsantev" ti ajnavx ion stratiwvtou i{na
ejpitimhqw'si aJrmozovntw".
73 Givnwske dev, strathgev, a]n e[ch/" plh'qo" dromwvnwn kai; e[peita

e[ch/" kai; ajnavndrou" stratiwvta", [o{ti]39 oujde;n ijscuvei oujd a]n
e[ch/"40 povlemon pro;" ojl ivgou" ejcqrou;" kai; w\sin ejkei'noi
ajndrei'oi kai; tolmhroiv: oujde; ga;r polloi; a[ndre" ijscuvousiv ti
kata; ojlivgwn ajndrw'n, eij ou[k eijsi kai; ajpo; th'" prqumiva" kai; ajpo;
th'" ejxoplivsew" stratiw'tai ajlhqei'". Tiv ga;r kako;n ouj mh;
poihvsousi kai; ojlivgoi luvkoi polla;" ciliavda" probavtwn;
74 Dio; prevpei i{na blevph/" met ajkribeiva" pollh'" pavnta ta; tw'n
ejcqrw'n o{pw" eijsi; kai; ou{tw" i{na poihvsh/" kai; tw;n dromwvnwn th;n
kataskeuh;n kai; to; plh'qo" aujtw'n kai; to; mevgeqo" kai; th;n
ejxovplisin tw'n stratiwtw'n kai; ta; a[lla ejpithdeuvmata aJrmodivw"
kai; kata; tw'n ejcqrw'n. “Ece de; kai; mikrou;" kai; taceiva"
drovmwna", ouj pro;" povl emon ejx wplismevnou", ajlla; pro;" ta;"
bivgla" kai; ta; mandavta kai; ta;" a[lla" ta;" ajpaitouvsa" oJmoivw"
creiva". “Ece de; kai; tav monhvria41 kai; ta;" galeva" kai; aujta
ejxwplismevna meta; ajrmavtwn dia; ta; pollavki" sumbaivnonta.
75 Su; de; ojfeivl ei" eij" pavnta ei\nai spoudai'o" kai; ajndrei'o" kai;
ajtavraco" kai; tacu;" eij" ta;" ajnagkaiva" ejxairevtw" tw'n
pragmavtwn ejgceirhvsei" kai; pravx ei" i{na kai; tw'/ Qew'/ ajrevsh/" kai;
------------------------------
39
Thus Dain. MSS N & l do not have this.
40
e[ch/", thus Dain, following Desrousseaux: e[ch MSS N & l.
41
monhvria, thus Dain: monevria MSS N & l.
71 (= Leo VI, §78) That is enough about formations. When you

wish to withdraw from a battle, make the formation of the
dromons into a [capital letter] sigma (i.e., “Ç” shaped), and thus
withdraw backwards so that the prow faces the enemy and the
stern remains to the rear.42 This type of formation is safe both
when you approach the enemy and when you withdraw from
them, as some of the ancients have indicated, having done
[that].43 For when you withdraw from them you are not in flight
but are avoiding battle, and you keep the dromons ready to
attack again, if need arises, by your having the prows towards
them. And the enemy in turn do not have the courage to enter
your curved formation for fear of encirclement.44
72 (= Leo VI, §79) When the engagement has ended, it is
appropriate, strate2gos, for you to divide then amongst the
soldiers what has been captured from the enemy, and to hold
banquets and invite [the men] and make much of them. Those
who have acted bravely should receive rewards and honours
and those whose behaviour has been unbecoming to a soldier
should be penalized accordingly.
73 (= Leo VI, §80) You should realize, strate2gos, that if you have
a number of dromons but then you have cowardly soldiers, that
it is no use even if you are fighting against a few enemies but
ones who are brave and bold. For not even many men can
achieve anything against a few unless they are true soldiers
both in energy and arms. For will not a few wolves do terrible
things to many hundreds of thousands of sheep?
74 (= Leo VI, §81) Therefore you should look with great accuracy
at the whole [situation] of the enemy, and then organize the
equipment of the dromons, and their number and size, and the
armament of the soldiers and other needs in a manner
appropriate to the enemy. Have small fast dromons which are
not armed for battle but as scouts, for messages and other
similar necessary purposes. You should also have mone2ria
(monoremes) and galeai, both armed with weaponry against
many eventualities.
75 (= Leo VI, §82) You must be keen, brave, calm, and vigilant
through everything, especially in the inevitable conflicts and
periods of action, so that you may both please God and appear a
------------------------------
42
43
The following sentences were added by Nike2phoros Ouranos to Leo VI’s text.
44
604 APPENDIX FIVE
th'/ ejx aujtou' basileiva/ hJmw'n dovkimo" fanh'/" strathgo;" kai; ejx
ajmfotevrwn kerdhvsh/" ajxiva" ajmoibav", ajpo; Qeou' me;n misqou;"
ajqanavtou" wJ" ajgwnizovmeno" uJpe;r th'" klhronomiva" aujtou', ejx
hJmw'n de; timav" te kai; dwrea;" ta;" prepouvsa", ouj yeudovmeno" to;
o{noma tou' strathgou', ajlla; stathgo;" th'/ ajlhqeiva/ kai; w]n kai;
legovmeno". Tosau'ta peri; qalassomaciva" ejn suntovmw/
eijrhvkamen.
notable strate2gos in the service of my Majesty under God and

receive worthy recompense from both: an eternal reward from
God for your struggles on behalf of his dominion, and honours
and gifts from myself since you have not fallen short of the
name of strate2gos but are a strate2gos in truth both in word and
deed. We have said enough in brief about warfare at sea.
APPENDIX SIX
GREEK FIRE
This Appendix is not an attempt to solve the vexed problems of either

the composition of the raw material of Greek Fire or of the delivery
systems for it. We do not pretend to have considered in toto the
scholarship on Greek Fire, the literature on which is voluminous and
has a long history.1 Here we have merely collected those observations
on Greek Fire which we have made in the course of our research into
the Byzantine navy and which may be useful to others or which
inform our discussion elsewhere. The entire debate about Greek Fire
has been bedevilled by the fact that the term became used widely in
both Arabic and Latin for almost any combustible, irrespective of the
delivery system. Pots full of combustibles known as Greek Fire,
hurled by catapults, were utilized by both Muslims and Latins, as well
as by Byzantines. However, here we are concerned only with that type
of combustible which was either shot by sipho2nes or which the
Byzantine sources suggest was the same material, even if hurled or
poured.
Although the relevant section of his chronicle is confused
chronologically, Theophane2s the Confessor, followed by Constantine
VII in the De administrando imperio, ascribed the invention of Greek
Fire and the projection of it through flame-throwers, sipho2nes or
sipho2nia, to an artificer from Heliopolis of Syria by the name of
Kallinikos during the first Muslim siege of Constantinople. He first
wrote that in A.M. 6164 (September 671-August 672) the emperor
Constantine IV stationed “large biremes carrying fire-cauldrons and
dromons carrying sipho2nes” in the small harbour of Caesarius on the
south side of Constantinople in preparation to defend it against the
Muslim fleets en route to assault the city. According to Theophane2s ,
the Muslim fleets did not arrive until A.M. 6165 (spring-summer of
673). He then described the siege of the city from April to September
------------------------------
1
See in particular Partington, Greek Fire and gunpowder; Ellis Davidson, “Secret
weapon”; Haldon and Byrne, “Greek Fire”; Haldon, “‘Greek Fire’ revisited”;
Christides, “New light”; Korres, «ÔUgro;n pu'r»; Pasch, “Fuoco greco”; Russo, “Fuoco
marino”.
We are extremely grateful to John Haldon for his many discussions with, and
communications to, us concerning Greek Fire and his practical experiments to build a
sipho2n weapon, and for a copy of his “‘Greek Fire’ revisited” before its publication.
608 APPENDIX SIX
673, the withdrawal of the Muslim fleet to Kyzikos to winter over 673-
4 and its return in the following spring of 674, the siege lasting in this
way for seven years according to him. Only at the end of his entry for
A.M. 6165, in a passage identified as being derived from a lost Syriac
chronicle, did Theophane2s write that: “At that time Kallinikos, an
artificer from Heliopolis of Syria, having taken refuge with the
Romans [and] having prepared sea fire, ignited the ships of the Arabs
and burned them with their crews. In this way the Romans came back
in victory and acquired the sea fire.”2 The later part of the entry for
A.M. 6165 is in fact chronologically generic and the fact that
Theophane2s wrote that the Byzantines had dromons carrying sipho2nes
in A.M. 6164 does not negate his own ascription of the development
of Greek Fire to Kallinikos. He referred to it as “sea fire”, pu'r
qalavssion (pyr thalassion), or “wet fire”, pu'r u{gron (pyr hygron), and
said that Kallinikos had “prepared” or “processed”, kataskeuavsa"
(kataskevasas), it.
Writing around the middle of the twelfth century George Kedre2nos
uniquely reported that Kallinikos was from Heliopolis of Egypt rather
than Syria and that from him were descended the family of
“Lampros”, “Brilliance”, who still manufactured the fuel in his own
day.3 “Brilliant” was one of the adjectival terms commonly used for
the fire. However, there is no corroborating evidence for Kedre2nos’s
story and the idea that the secret of the fuel had remained confined to
the members of one private and obscure family and had been handed
down within it from generation to generation for six centuries is not
credible.
Theophane2s also wrote that in 713, when preparing against the
coming Muslim assault on Constantinople, Anastasios II built fire-
carrying die2reis, amongst other ships. And, during the actual assault,
Leo III had fire-carrying sipho2nes made and mounted on dromons and
die2reis which he sent against the Muslim fleets. In 726 the fleets of
Hellas and the Cyclades revolted against Leo III because of his
persecution of iconophiles but were defeated by the imperial fleet
using “artificial” fire. In 743 the usurping emperor Artabasdos, sent
out “fire-carrying die2reis” against the fleet of the Kibyrrhaio2tai
------------------------------
2
Theophane2s, Chronographia, A.M. 6164-5 (vol. 1, pp. 353-4), esp. A.M. 6165
(vol. 1, p. 354): “tovte Kallivniko" ajrcitevktwn ajpo; ÔHlioupovlew" Suriva" prosfugw;n
toi'" ÔRwmaivo i" pu'r qalavssion kataskeuavsa" ta; tw'n Aravbwn skavfh ejnevprhse kai;
suvmyuca katevkausen. Kai; ou{tw" oiJ ÔRwmai'o i meta; nivkh" uJp evstreyan kai; to;
qalavssion pu'r eu|ron.”. Cf. Constantine VII, De administrando imperio, §48 (p. 226).
3
George Kedre2nos, Synopsis historio2n, vol. 1, p. 765.
GREEK FIRE 609
supporting Constantine V approaching Constantinople but the

Kibyrrhaio2tai repulsed them, suggesting that they also had Greek Fire.
Finally, In 812 the Bulgar Khan, Krum, captured the town and fortress
of Develtos on the Black Sea coast and found in it 36 bronze, calkou'"
(chalkous), sipho2nes and a quantity of the fuel. However, there is no
evidence that the Bulgarians ever used the Greek Fire themselves and,
apparently, obtaining possession of the fuel and the delivery
mechanism did not in itself reveal the secret of the weapon system.4
In the tenth century the De administrando imperio repeated
Theophane2s’ account, referring to the fuel as “wet” fire, as did the
Vita Basilii of the Theophane2s continuatus. Genesios called it fire “for
war”, polemiko;n pu'r (polemikon pyr).5 Leo VI referred to it as
“processed” fire, pu'r ejskeuasmevnon (pyr eskevasmenon), and
Nike2phoros Ouranos followed him; although, he also called it
“brilliant” fire, pu'r lamprovn (pyr lampron). Both said that as well as
being projected through sipho2nes, it could be hurled in pots, kuvtrai
(kytrai) or tzukavlia (tzykalia), or poured from what may have been
cauldrons operated by mangana, or have tow wrapped around
caltrops, trivboloi (triboloi), soaked in it.6 The treatise known as the
Sylloge2 taktiko2n, which has been dated to the early tenth century, also
referred to it as “wet fire” or “brilliant” fire and described the
projection devices as swivels, strepta; (strepta): “Useful are what are
known as strepta which send by a device the wet fire, which is also
known as brilliant [fire] by many.” The Sylloge2 taktiko2n once had a
chapter 70 entitled, “How the fire that is called wet can be put out and
how, when it is thrown at wood or walls, it does not affect them”;
however, this chapter has been lost.7
The Muslims also acquired possession of Byzantine fire-ships and
the question arises as to whether they did in fact acquire the secret of
the siphone2s and their fuel. Ibn al-Athı3r, writing of a naval expedition
------------------------------
4
Theophane2s, Chronographia, A.M. 6209, 6218, 6235, 6305 (vol. 1, pp. 397,
405, 419, 499).
5
Constantine VII, De administrando imperio, §48, ll. 28-30 (p. 226): “ Istevo n,
o{ti ejp i; Kwnstantivnou, uiJou' Kwnstantivnou, tou' kai; Pwgwnavtou kaloumevnou,
Kallivnikov" ti" ajpo; ÔHlioupovlew" ÔRwmaivoi" prosfugwvn, to; dia; tw'n sifwvnwn
ejkferovmenon pu'r uJgro;n kateskeuvasen, ...”; Theophane2s continuatus, V.59 (p. 298):
“... tou' pneuvmato" tw'/ uJgrw'/ puri; katenevprhsan.”; Genesios, Basileiai, B.2, B.5, D.34
(pp. 24, 27, 85).
6
Appendix Two [a], §§6, 59, 63 (kytrai), 64, 67 (manganon); Appendix Five,
§§5, 56, 60 (tzykalia), 61 (manganon). On the mangana see above pp. 378-9.
7
Sylloge2 taktiko 2n, 53.8 (pp. 102-3): “... lusitelei' ta; strepta; kalouvmena ta; dia;
mhcanh'" to; uJgro;n pevmponta dhladh; pu'r, o} dh; kai; lampro;n para; toi'" polloi'"
ojnomavzetai, kai; ta; legovmena ceirosivf wna, a{per nu'n hJ basileiva hJmw'n ejpenovhse ...”.
See also the rubric for ch. 70 at p. 15.
610 APPENDIX SIX
commanded by the Aghlabid emir Abu2 ’l-Aghlab Ibra2hı3m ibn ‘Abd

Alla2h against Pantelleria in 827, wrote that the Muslims captured a
Byzantine fire ship of the h5arra2qa type.8 Other Arabic authors called
ships or weapons for launching fire h5arra2qa2t or naffa2t6a2t and
specialists in fire-launching h5arra2qu2n or naffa2t6u3n.9
The Latin treatise attributed to Marcus Graecus entitled Liber
ignium ad comburendos hostes survives in several manuscripts, the
oldest of which is apparently the late thirteenth- or early fourteenth-
century Paris, Bibliothèque Nationale, MS. Lat. 7156. It contains
much information about what it calls Greek Fire. Who Marcus
Graecus was is unknown, although Muslim alchemical writers and
their Latin translators knew a text or texts attributed to a certain
Marcouh or Marcouneh, entitled King of Egypt. The title was only an
honorific but the treatise attributed to this author by Muslim
alchemists was associated with a Greek tradition. The translation into
Latin was made in the late twelfth or thirteenth centuries. This treatise
was concerned with combustibles of many types which supposedly
could not be extinguished by water, rather than with the Byzantine
weapon per se. It represents a much wider alchemical tradition on
combustibles and is of little use. There is no mention of sipho2nes in
it.10
The otherwise unknown Muslim author Murd4a2 ibn ‘Alı3 ibn Murd5a2
al-T4artu2sı3, who wrote a treatise on armaments for S5ala2h5 al-Dı3n,
described the composition of a naphtha-based fire fuel which would
float on water:
Manufacture of a naphtha which runs on water and is good for burning

ships. Pitch, one part; mineral sulphur, that is to say naphtha, one part;
resin, the same; sandarak, the same; pure and clear dolphin’s fat, the same;
grease of kidneys of goat, the same; yellow sulphur, the same. Grind that
which should be ground. Put the pitch on the fire in the cauldron for a
while and, when the pitch boils, the sandarak should be added and beaten
until it is mixed. That finished, mineral sulphur should be added, which
has been covered in old oil, and take off [the fire]. When you need it, take
it and boil it until you know that it has reached the point to ignite as fire,
and send it on the water towards the desired ships. It will cause a great
------------------------------
8
Ibn al-Athı3r, Al-Ka2m il (Fagnan), p. 192.
9
See Canard, “Textes”.
10
The treatise was edited by Berthelot from the oldest manuscript, Paris,
Bibliothèque Nationale, MS. Lat. 7156, with reference to three other manuscripts also.
See Marcus Graecus, Liber ignium, pp. 89-94 and 108.
GREEK FIRE 611
11
conflagration and it runs on water and cannot be extinguished.
It is extremely improbable that the Byzantine Greek Fire fuel had

anything in common with such a mixture, or indeed with any of the
many other similar such mixtures to be found in Muslim sources.
A dromon’s primary sipho2n was undoubtedly the one at the prow,
above which, according to Leo VI and Nike2phoros Ouranos, there was
a fortified foredeck.12 In the anonymous Arabic translation of sections
of the Naumachika Leontos Basileo2s inserted by Ibn Mankalı3 into his
Al-adilla and Al-ah5ka2m, Leo’s §6 was translated as: “At the bow of
each ship there should be tubes (ana2bı3b) from which they throw fire.
They [the tubes] are called in the old Greek (Ru2mı3) tongue sifuna, and
above the aforementioned tubes should be covered planks, the latter in
turn protected from above by other planks”. The translator apparently
envisaged more than one such sifo2n at the prow. The translator also
preserved the reference to the fortified foredeck above the sifuna. The
text of §51 of the Naumachika Leontos Basileo2s, was also reproduced
quite closely: “On other occasions, the arranging of your ships should
be in a straight line, so that if time allows, you can ram your enemy’s
ships with the bow of your ships, and shoot fire at them”. Where Leo
VI said that the sipho2n was bound in copper, the Arabic translator
omitted this. However, interestingly, in the Al-adilla, the sipho2nator
was said to have had his own “elite squad” of men, as, no doubt, he
would have had.13 The translator used the Arabic ana2bı3b, “tubes” for
the flame-thrower. He knew Greek well but had no idea about the
Byzantine weapon because sivf wn did, of course, have the primary
sense of a tube or pipe, as well as of a force-pump. He simply equated
it to the types of combustibles known to Muslims and translated the
Greek literally.
There is no doubt that the Muslims possessed combustibles for use
in war both at sea and on land. They could certainly hurl such
combustibles in pottery “grenades” by hand or with catapults.14 The
only question is whether they also possessed the secret of the sipho2nes
and could project fire in the manner of a flame-thrower. The fact that
the translator of Leo VI could translate the Greek sivfwn accurately by
------------------------------
11
Al-T4artu2sı3, “Traité”, pp. 123 [Arabic] and 146 [French translation]; here
translated from the Arabic by Ahmad Shboul.
12
13
Appendix Eight [a], pp. 241-2, 243, 247; [b], pp. 21, 123 The texts vary slightly.
14
See Christides, Conquest of Crete, pp. 63-6; idem, “Transmission”, pp. 91-5;
idem, “Parallel naval guides”, pp. 62-4; idem, “New light”, pp. 4-25; Haldane, “Fire-
ship of al-Sa2lih”.
612 APPENDIX SIX
the Arabic unbu2b for a tube does not really prove anything more than
that he knew Greek well. In fact, among all the Arabic references to
“Greek Fire” combustibles collected by De Goeje and Canard, there is
only one description of fire-throwing devices, h5arra2qa2t al-naft’, by a
poet included in the Nishwa2r of Abu2 ‘Alı3 al-Muh5assin ibn ‘Alı3, al-
Tanu2khı3 (Syria and Iraq, 941-94 C.E.) which does suggest that
Muslims may have had such a device: “And lo [there is] something
yellow (of brass or bronze) in whose mouth is mucus of the same
colour which whenever she [the yellow object] vomits forth, then it
[the mucus] plays with the wind and floats like a mirage. ... She spits
out lightning flashes, between two nights, from entrails up through the
mouth of a snake where you can see no teeth. She plunges into the
tumult naked, to make it more frightening, and if she were asked she
would not recognise fear or safety”.15
According to John Kaminiate2s, the Muslims under Leo of Tripoli
assaulting Thessalonike2 in 904 used fire emitted from sipho2nes in the
hands of men stationed on bridges running from the mastheads of the
ships,16 and it is probable that Muslims did in fact acquire the secret of
the weapon system, although hard evidence is extremely elusive.17 It is
clear from Joinville’s description that the Greek Fire used by the
Egyptians against the Crusaders at Damietta in 1249 was in
earthenware pots hurled by catapult.18
Was the secret acquired by the Latin West? Geoffrey Malaterra,
reported that in 1081 the Norman fleet off Dyrrachion was confronted
and defeated by the Venetians, who “... skilfully blowing the fire,
which they call Greek and is not extinguished by water, from hidden
passages of tubes beneath the waves, cunningly burned between those
same waves of the sparkling sea-top a certain ship of ours [of the
Normans] which they call a cattus”.19 Malaterra clearly did not
------------------------------
15
Canard, “Textes”; De Goeje, “Observations”. Canard’s work built upon De
Goeje’s. Our translation is suggested by Michael Carter from Al-Tanu2khı3, Nishwa2r,
vol. 2, p. 303. It varies considerably from those of Margoliouth and Canard.
16
John Kaminiate2s, De expugnatione Thessalonicae, 34.7 (p. 32): “..., pu'r te dia;
tw'n sifwvnwn tw'/ ajevri fushvsante", ...”.
17
Examination of the sources cited by those who claim the Muslims did have the
secret of the weapon system, Canard, Christides, Eickhoff, Haldane, Vasiliev, and
others, reveals a lack of hard evidence for the sipho2n system. There is plenty of
evidence for fireships, combustibles, earthenware grenades, and fire-arrows, but not
for the sipho2n system.
18
Joinville, Vie de saint Louis, §206 (pp. 100-101).
19
Geoffrey Malaterra, De rebus gestis, III.26 (p. 73): “Sed illi artificiose ignem,
quem graecum appellant, qui nec acqua extinguitur, occultis fistularum meatibus sub
undis perflantes, quandem navem de nostris, quam cattum nominant, dolose inter
ipsas liquidi aequoris undas comburunt.”.
GREEK FIRE 613
understand how the Greek Fire weapon worked, but he did believe
that the Venetians had access to it. This is the only mention known to
us in Western sources of a weapon resembling the sipho2nes of tenth-
century dromons being used by anyone other than Byzantines.
Similarly, no depiction of any weapon resembling a sipho2n is known
to us from illustrated Western manuscripts, with the exception of the
single illustration in the Sicilian manuscript of the Synopsis historio2n
of John Skylitze2s. Although drawn in Sicily, this illustration was
probably based on an earlier Byzantine one in the original manuscript
from which it was copied.20
Figure 57
Dromon using Greek Fire in the Synopsis historio2n of John Skylitze2s
(Madrid, Biblioteca Nacional, vitr. 26-2, fol. 34v), ca 1160.
However, we draw attention to a hitherto almost unnoticed mention of

what may have been sipho2nes of the Byzantine type used in the fleets
of the Angevin Kingdom of Sicily in the 1270s. Angevin galleys used
ampule, bottles or jars, filled with ignis, “fire”, variously described as
sulphureus (sulphurous), silvestris (lustrous, silvery), or Grecus
(Greek).21 These were presumably hurled by hand or by catapults. The
correlation between the Latin silvestris and the Greek lampron used in
Byzantine sources is striking and it is difficult to avoid the conclusion
------------------------------
20
See Appendix Seven; here esp. Table 9, Byzantine Four; Table 10, no. 10.
21
Filangieri, Registri, vol. 12, p. 223; vol. 13, pp. 104-5; vol. 17, pp. 147-8; vol.
23, p. 289.
614 APPENDIX SIX
that the fuel was the same. Moreover, one order of 1275 from the
royal chancery specified that galleys should be armed with “roccette
ad ignem proiciendum XXV”; that is, with “25 ‘rockets’ for throwing
fire”.22 These sound so much like the sipho2nes of the Byzantines that it
is difficult to believe that the reference was to anything else. However,
we know of no collateral evidence for the use of anything like a
sipho2n or roccetta for projecting fire actually being used by Angevin
fleets. The chronicle record of the operations of Angevin fleets has no
mention of the use of such devices.
The actual composition of the fuel is an unknown quantity.
Understandably, Byzantine sources did not reveal the secrets of the
fuel and its method of projection.23 There is, however, a very curious
description of what appears to be Greek Fire and sipho2nes on folio
157r-v of a manuscript at Wolfenbüttel.24 Most of this manuscript
contains a text of the De compendiosa doctrina of the early fourth-
century author Nonius Marcellus, a glossary of Latin terms profusely
illustrated with quotations from late Republican Latin authors and
with frequent sprinklings of Greek phrases. The folio in question here
is in fact the last two pages of the manuscript and is written in a script
of the ninth century. There are also some jottings in a fourteenth-
century hand. These pages contain miscellaneous bits and pieces and
the Greek Fire text is sandwiched between a Greek alphabet and list of
diphthongs and vowels on the one hand and an excerpt from St
Augustine on the other. It may have been intended to refer to how one
might make a nice fire for a recitation of the canticle of the three boys
in the fiery furnace (Daniel, III.52-88). It reads:
The material of the fire of the three boys: naptha, tow, pitch, a fire arrow.
Naptha [is] a species of balsam originating in Babilonia [Egypt] in humid
places, which colloquially we call marisci [recte, marismi; i.e.,
maremmas], and it seems to swim there upon the water like fat. Also,
there are two kinds of balsam. One originating from Mount Sinai, exuding
from rock, whence “rock of oil” [i.e., petroleum]; the other [originating
from] twigs which mixed together produce an inextinguishable fire. For
when the Saracens proceed in war to a naval battle, having built a furnace
------------------------------
22
Filangieri, Registri, vol. 13, p. 105. First noticed by Pryor in “Galleys of Charles
I of Anjou”, pp. 78-9 and Table One.
23
Constantine VII, De administrando imperio, §13, ll. 73-103 (pp. 68-70).
24
Wolfenbüttel, Herzog August Bibliothek, Cod. Guelf., 96 Gud. lat., fol. 157r-v.
We are indebted to the Head of the Department of Manuscripts at Wolfenbüttel, Dr
Helmar Härtel, and his staff, for their efforts in tracing this text from the inaccurate
reference given in Forbes, More studies, p. 83, and for providing us with a photocopy
of the manuscript. The text was transcribed in Bischoff “Anecdota Carolina”, pp. 6-7.
GREEK FIRE 615
right at the front of the ship, they [the Saracens?] set on it a copper vessel
full of these things, having put fire underneath. And one of them, having
made a bronze tube similar to that which the rustics call a squitiatoria,
“squirt”, with which boys play, they spray [it] at the enemy.25
Figure 58
Wolfenbüttel, Herzog August Bibliothek, Cod. Guelf., 96 Gud. lat.,
fol. 157r-v.
This is an interesting text both because it is dated so early

chronologically and also because it is the only text known which
------------------------------
25
The Latin text reads as follows. The transcription and suggested emendations are
Bischoff’s with one addition by us from the manuscript given in square brackets:
“Materia ignis trium puerorum. Napta. Stupa pix. malleolis. Napth genus balsami
nascens in Babilonia in humentibus locis, quos vulgo mariscos (korr. aus marismos)
appellamus, et quasi saginum ibi [super] aquam videtur natare. Sunt etiam duo genera
balsami: unum nascens in monte Sina sudans ex petra, unde petra olei; alterum ex
virgultis que simul mixta procreant ignem inextinguibilem. Nam pergentibus
Saracenis ad bellum navali (nova mit einkorr. li Hs.) certamine, in prima fronte navis
facta fornace illi insidunt vas eneum his plenum subposito igne, et unus eorum fistula
facta aerea ad similitudinem quam rustici squitiatoriam vocant, qua ludunt pueri, in
hostem spargunt.”.
616 APPENDIX SIX
discusses the composition and method of projection of what appears to

have been Greek Fire. Although it must be taken with many grains of
salt, its identification of the fuel, naphtha, as a form of balsam being
obviously incorrect, and the latter in itself being misunderstood,
nevertheless its references to a furnace, a copper vessel, and a squirt
are obviously important, as is its identification of the fuel as petroleum
oil. It is difficult to believe anything other than that the author had
seen Greek Fire weapons in action but how and why such a
description of them ended up in such an odd Western manuscript in a
recipe for the “fire of the three boys” must make the text very
problematical.
In a passage of Yngvars Saga Ví∂förla which appears to be based
on actual experience,26 the hero Yngvarr and his companions coming
down the Russian rivers encountered heathen “pirates” (illger∂amenn)
who used fire weapons. Covering their ships with reeds to disguise
them as islands, they attacked Yngvarr’s ships. However, Yngvarr
managed to defeat these heathen “pirates” and their fire weapons by
shooting a flaming arrow lit from a tinder box into the mouth of “the
tube jutting from the furnace”.
... But when the Vikings found how tough the opposition was, then [the
pirates] began blowing with smiths’ bellows at a furnace in which there
was fire and there came from it a great din. There stood there also a brass
[or bronze] tube and from it flew much fire against one ship, and it burned
up in a short time so that all of it became white ashes. ... But that arrow
flew from a bow with the fire into the tube which came out of the furnace
and the fire was turned on the heathens themselves and the island burned
up in a short blink of an eye.27
There was obviously a good deal of fiction involved in this account;
------------------------------
26
Yngvarr Eymundsson was a historical personage, recorded as having died in
1041, who led a host to the East some time before that. Many runic inscriptions
survive recording names of men who sailed with him. He was killed in the East in
“Særkland”. As it survives, the saga was probably written early in the thirteenth
century and was based on a now-lost Latin work which may have amounted to a life
of Ingvarr by the monk Oddr Snorrason, who belonged to the Benedictine monastery
of Thingeyrar in Iceland. Oddr’s work was based on both oral narrative and written
sources.
27
Yngvars Saga, §6 (p. 441): “En er víkingar fundu, at fast var fyrir, tá tóku teir
at blása smi∂belgjum at ofni teim, sem eldr var í, ok var∂ af tví mikill gny)r. Tar stó∂
ok ein eirtrumba, ok ór henni fló eldr mikill á eitt skipit, ok brann tat á lítilli stundu,
svá at allt var∂ at fölska. ... En sú ör fló af boganum me∂ eldinn í trumbuna, tá er stó∂
ór ofninum, ok sny)st eldrinn á sjálfa hei∂ingja, ok brann á litlu augbrag∂i eyin me∂
öllu saman, mönnum ok skipum.”
GREEK FIRE 617
however, the mention of a furnace, the brass or bronze tube, the great
din, and the emission of fire can leave little doubt that the origin of the
story lay in someone’s experience with Greek Fire.
One other property of Greek Fire deserves attention. The Liber
ignium ad comburendos hostes attributed to Marcus Graecus repeated
an ubiquitous specification that all “inextinguishable fire” could in
fact be extinguished by strong vinegar, old urine, sand, or by felt
soaked in vinegar three times and dried out after each soaking. Many
different testimonies are clear that although it could not be put out
with water, in fact it floated and burned on water, it could be
extinguished with vinegar, presumably wine vinegar, Greek o[xo"
(oxos), Latin acetum, or urine, as well as sand.28 Hides soaked in
vinegar were resistant to it. Sand would obviously smother the fire and
extinguish it by depriving it of oxygen. However, why either vinegar
or urine may have been effective, when water was not, is unknown. It
could be dismissed as an old wives’ tale were it not that these two
chemicals alone were specified as being effective in so many different
works. It could be that authors simply accepted what others wrote and
so the specification was passed from hand to hand over the centuries;
however, just possibly, there may have been some chemical reaction
which produced carbon dioxide or nitrogen to smother the fire.
Leo VI wrote that there should be one sipho2n at the prow below the
fortified foredeck and that others which were to be used from behind
iron shields, skoutaria side2ra, presumably along the sides when
engaged broadside, were hand-held ceirosivfwne" (cheirosipho2nes),
which he himself had invented. Nike2phoros Ouranos repeated Leo’s
reference to cheirosipho2nes, but it is very apparent that the Arabic
translator of Leo included by Ibn Mankalı3 in his Al-ah5ka2m was
completely bemused by the weapon.29
In what may possibly be the earliest reference to such
cheirosiphones after Leo VI the author of the treatise on defence
against sieges known as the De obsidione toleranda wrote that if the
enemy built siege engines the defending commander should prepare
pine torches, tow and pitch, and cheirosipho2nes to burn them.30 This is
a text which may possibly have pre-dated the encyclopaedic works
associated with Constantine VII in the mid tenth century. It appears to
------------------------------
28
See Marcus Graecus, Liber ignium, p. 108.
29
Appendix Two [a], §§6, 64; Appendix Five, §§5, 60; Appendix Eight [b], p. 124.
30
De obsidione toleranda, §113 (pp. 188-9): “eij de; kai; mhcana;" oiJ ejcqroi;
kateskeuvasan, proeutrepivzein da'ida" kai; stupei'on kai; pivssan kai; ceirosivfwna, kai,
…”.
618 APPENDIX SIX
have been contemporary with the Sylloge2 taktiko2n.

The Anonymous also wrote that there was one sipho2n at the prow
and another two which could be used along the sides when engaged
broadside; although he did not say whether the latter were hand-held
or mounted.31 Liudprand of Cremona wrote that when fifteen derelict
chelandia were armed to meet the Rho2s invasion of 941, they were
armed with devices to throw the fire not only at the prow and on both
sides but also at the stern.32 Two inventories for the Cretan expedition
of 949 also specified three sipho2nia per dromon, without being any
more specific than that, and 80 sipho2nia for 40 ousiaka [ships].33
These latter were probably chelandia and it would seem that they only
carried two sipho2nia each. It also specified 24 sipho2nia for 50
pamphyloi, but this was almost certainly a copyist’s error for the eight
pamphyloi, said elsewhere to have sailed with the expedition, which
would thus have been armed with three sipho2nia each, just like the
dromons.34 The treatise known as the Praecepta militaria, dated to
soon after 963, also referred to cheirosipho2nes and described the fire
as “glutinous”, pu'r kollutikovn (pyr kollytikon), as well as
“prepared”.35
Hand-held weapons for Greek Fire, referred to by various terms,
are mentioned in the context of land warfare in so many sources that
whatever the fuel was, it must have been capable of being projected in
this way as well. Leo VI recommended destroying an enemy’s siege
towers with fire-throwers, pyrobola, and stone-throwers. Nike2phoros
Ouranos expanded the reference to strepta with fire, sipho2nes,
cheirosipho2nes, and manganika. In the Sylloge2 taktiko2n the
corresponding reference was to strepta “which shoot clearly by
machine the liquid fire that is also called brilliant by the many, and the
so called cheirosipho2nes which our majesty have now devised”.
Nike2phoros Pho2kas also specified the use of strepta with brilliant
------------------------------
31
See Appendix Three, §2.14: “Epi; de; th'" prwv/ra" oJ sivfwn o}" katakovrax levgetai
ejnergw'n o{tan w\sin aiJ nh'e" ajntivprw/roi: kai; duvo de; plavgioi kai; aujtoi; ejnergou'nte" o{tan
plavgiw" prosbavllwsi.”.
32
Liudprand of Cremona, Antapodosis, V.15 (p. 138).
33
Appendix Four [b], §§IV.1, V.27 [= Haldon, “Theory and practice”, pp. 227,
229; Constantine VII, De cerimoniis, II.44 (vol. 1, pp. 672, 673)].
34
See Appendix Four [b], §V.26 [= Haldon, “Theory and practice”, p. 229;
35
Nike2phoros Pho2kas, Praecepta militaria, p. 5: “Dei' de; to;n ajrchgo;n tou' stratou'
e[cein kai; ceiromavggana mikra;, hjlakavtia triva kai; strepto;n meta; lamprou' kai;
ceirosivfouna, i{na, ka]n i[sw" kai; oiJ ejcqroi; th'/ oJmoiva/ kai; i[sh/ paratavxei crhvswntai, diav
te tw'n ceiromaggavnwn diav te tou' skeuastou' kai; kollutikou' puro;" ejpikratevsteroi
givnontai tw'n uJpenantivwn kai; paraluvswsin aujtouv".”.
GREEK FIRE 619
Figure 59
Soldier using a hand-held flame thrower in a treatise on poliorcetics attributed
to He2ro2n of Byzantium (Rome, Biblioteca Apostolica Vaticana, MS. Vat. Gr.
1605, fol. 36r), eleventh century.
[fire] and cheirosipho2nes by field armies.36

The sipho2nes used by men stationed on flying bridges running from
the mastheads of ships reported by Kaminiate2s must also have been
------------------------------
36
Leo VI, Taktika (PG), XV.51 (coll. 899-900): “Pro" de; tou;" ejpagomevnou"
puvrgou" purobovla ei[dh kai; petrobovloi, …”; Nike2phoros Ouranos, Taktika, coll.
1348-9: “Pro;" de; tou;" prosferomevnou" puvrgou" ei;" to; tei'co", i{na w\si strepta; meta;
lamprou' kai; sufwvnia, kai; ceirosuvfwna, kai; magganikav.”; Sylloge2 taktiko2n, 53.8 (pp.
102-3): “Pro;" mevntoi tou;" dia; kulivndrwn toi'" teivcesi prosagomevno" xulivnou"
puvrgou", ou}" oiJ taktikoi; movsuna" ojnomavzousi, lusitelei' ta; strepta; kalouvmena ta; dia;
mhcanh'" to; uJgro;n pevmponta dhladh; pu'r, o} dh; kai; lampro;n para; toi'" polloi'"
ojnomavzetai, kai; ta; legovmena ceirosivfwna, a{per nu'n hJ basileiva hJmw'n ejp enovhse, kai;
puvrgoi pro;" touvtoi" oiJ katevnanti aujtw'n uJyouvmenoi livqoi" h] plivnqoi" h] xuvloi", kai; u{lh
pantoiva ejn tw'/ metaxu; tovpw/ sumforhqei'sa kai; meta; mikro;n exafqei'sa.”; Nike2phoros
Pho2kas, Praecepta militaria, I.15 (p. 20): “dei' de; to;n ajrchgo;n tou' stratou' e[cein kai;
ceiromavggana mikrav, hjlakavtia triva kai; strepto;n meta; lamprou' kai; ceirosivfouna,
i{na, ka]n i[sw" kai; oiJ ejcqroi; th'/ oJmoiva/ kai; i[sh/ paratavxei crhvsontai, diav te tw'n
ceiromaggavnwn diav te tou' skeuastou' kai; kullhtikou' puro;" ejpikatevsteroi givnwntai
‹aiJ paratavxei" hJmw'n› tw'n uJpenantivwn kai; paraluvswsin auJtouv".”.
620 APPENDIX SIX
hand-held.37 The words which he used, “tw'/ ajevri” (“with air”), suggest
that a blast of air of some kind may have been used to project the fire
from the sipho2n. That the Byzantines did possess the technology to
project Greek Fire through cheirosipho2nes is proven by an illustration
in the eleventh-century manuscript Rome, Biblioteca Apostolica
Vaticana, MS. Vat. Gr. 1605 of the Parangelmata poliorke2tika
attributed to He2ro2n of Byzantium. At folio 36r a soldier is depicted on
a flying bridge attacking the top of the walls of a town with a hand-
held flame thrower, described in the text as a “swivelling, fire-
throwing, hand held [implement]”.38 It illustrates a passage taken from
the Syntaxis Me2chanike2 of Philo2n of Byzantium, dated to the late
third-century B.C.E., which referred to the use of “fire-throwing,
hand-held [implements]”. However, the treatise on poliorcetics
attributed to He2ro2n was first compiled in the reign of Constantine VII
and some of its illustrations, even though based on antique models,
may be assumed to represent the technology of that period.39
Whatever the various technologies for projecting Greek Fire may
have been, and there appear to have been more than one, they were not
confined to apparatus fixed on ships. When the fleet of amı3r Ya2zama2n
al-Kha2dim of Tarsos attacked Euripos sometime after 883, according
to the Theophane2s continuatus the strate2gos of Hellas, Oiniate2s,
destroyed the Muslim ships from the walls of the town with “wet
fire”.40
It is possible that the third figure from the left in the illustration of
naval warfare in the Marciana manuscript of the Kyne2g etika of Pseudo
Oppian holds a cheirosipho2n. Whereas the flute or trumpet players at
the sterns of the galleys are clearly blowing into their instruments, this
other figure is not. He appears to be holding over his shoulder a tube
of some sort bound with bands, which would accord with the
presumed construction of such a weapon. [See Figure 26] It is quite
similar in construction to the sipho2n at the bow of the Byzantine
galley in the Madrid manuscript of the Synopsis historio2n of John
Skylitze2s.
It might be asked why the foredeck was constructed above the
------------------------------
37
See above, n. 16.
38
He2ro2n, Parangelmata poliorke2tika, §49 (pp. 98-9) and fig. 22: “… streptw'n
ejgceiridivwn purobovlwn …”. See also Schneider, “Byzantinische Feuerwaffe”.
39
See Dain, “Stratégistes”, pp. 358 (where the manuscript is wrongly identified as
Vaticanus Gr. 1614) and 388.
40
Theophane2s continuatus, V.59 (p. 298). John Skylitze2s changed the report,
making Oiniate2s dispel the Muslim fleet with his own trie2reis. See John Skylitze2s,
Synopsis historiarum, Basivleio" oJ Makedw'n.29 (p. 151).
GREEK FIRE 621
sipho2n, or alternatively why the sipho2n was located below the

foredeck rather than on it? The answer may have been that because the
heat generated by the flame burning oxygen would tend to make the
tongue of the flame curve upwards as it heated the air, it may have
been necessary to locate the source of the flame as low to the water as
possible in order for the tongue to make contact with an enemy ship,
rather than passing right over it. Medieval galleys of all kinds always
rode as low in the water as safety and other considerations allowed, in
order to maximize the mechanical advantage of the oars. A tongue of
flame whose end had curved upwards for more than three or so metres
would pass right over a medieval galley, even if generated virtually at
the waterline. However, Haldon’s recent experiments suggest that this
may not have been the case. In the machine which he constructed,
because the fuel was incompletely vapourised as it left the nozzle, the
jet of flame in fact curved downwards. [See Figure 61] This may help
to explain how Anna Komne2n e2, or her sources Landulf or Tatikios,
describing the alleged defeat of a Pisan fleet during the First Crusade,
wrote that the Pisans were terrified because they were not familiar
with flames which instead of rising were directed wherever the
sipho2nator wanted, often downwards or sideways.41
Bearing in mind that the deck of a dromon at the prow cannot have
been more than approximately 1.5 metres above the calm water line
and that moderate breezes of Beaufort Scale Four, 11-16 knots, raise
waves with crests up to around 0.8 metres above that, Liudprand of
Cremona’s comment that calm winds and seas were necessary if the
sipho2nes were not to become a danger to their own ships becomes
comprehensible.42 And, obviously, any wind would have to be astern.
If an enemy fleet managed to gain the weather guage with the wind
behind it, the sipho2nes would have become useless. When Oiniate2s
used Greek Fire against the fleet of Ya2zama2n al-Kha2dim at Euripos,
the descriptions of both the Theophane2s continuatus and John
Skylitze2s suggest that he did so only when the wind turned
favourable.43 It would also have been highly desirable, and probably
------------------------------
41
Anna Komne2ne2, Alexiade, XI.x.4 (vol. 3, p. 44): “... (oujde; ga;r ejqavde" h\san
toiouuvtwn skeuw'n h] puro;" a[nw me;n fuvsei th;n fora;n e[conto", pempomevnou d ejf a}
bouvletai oJ pevmpwn katav te to; prane;" pollavki" kai; ejf eJkavt era), … ”.
42
Liudprand of Cremona, Antapodosis, V.15 (p. 138): “Denique miserator et
misericors Dominus, qui se colentes, se adorantes, se deprecantes non solum
protegere, sed et victoria voluit honorare, ventis tunc placidum reddidit mare; secus
enim ob ignis emissionem Grecis esset incommodum.”.
43
Theophane2s continuatus, V.59 (p. 298); John Skylitze2s, Synopsis historio2n,
Basivleio" oJ Makedw'n.29 (p. 151).
622 APPENDIX SIX
absolutely necessary, for the sipho2nes to be able to be turned and

aimed in various directions. Fixed weapons would have been almost
unusable in the varying conditions encountered at sea. That is no
doubt why the word strepta, “swivels” was used for the weapon, both
by land and at sea. Liudprand of Cremona also wrote that in 941 the
Byzantine fleet dispelling the Rho2s assault on Constantinople “threw
the fire all around”.44
What exactly sipho2nes were made of is unclear. Theophane2s the
Confessor wrote that the sipho2nes that the Bulgarian Khan Krum
captured at Develtos in 812 were made of bronze.45 Both Leo VI and
Nike2phoros Ouranos said that they were “bound” in bronze, and Anna
Komne2ne2 or her sources said that those which Alexios I had made had
mouths in the form of the heads of lions and other animals made of
iron or bronze:
..., on the prow of each ship he [Alexios I] had fixed the heads of lions and
other land animals in bronze or iron, with their mouths open, surrounding
them with gold to make the mere appearance terrifying. He prepared the
fire that was to be emitted against the enemy to come out through their
mouths so that the lions and other animals appeared to be belching out
fire.46
One of the inventories for the Cretan expedition of 949 said that 30
nomismata were spent on providing 200 litrai of tin to a metal worker
named Michael for soldering, or brazing, together, various parts of the
sipho2nia.47 This makes sense. A tin-based solder would almost
certainly be used for fusing together sections of sipho2nes if they were
were made of bronze.
The word sivfwn, and its Latin transliteration sifo/sipho, could mean
a variety of things in classical Greek and Latin: tube, pipe, siphon for
drawing liquids, water spout. However, the one which is relevant to
------------------------------
44
Liudprand of Cremona, Antapodosis, V.15 (p. 138): “... ignem circumcirca
proiciunt.”.
45
Theophane2s, Chronographia, A.M. 6305 (vol. 1, p. 499): “... ejn oi|" kai; sivfwna"
calkou;" eu|ron lıV, kai; tou' di aujtw'n ejkpempomevnou uJgrou' puro;" oujk ojlivgon, ...”.
46
Anna Komne2ne2, Alexiade, XI.x.2 (vol. 3, p. 42): “... ejn eJkavsth/ prwvra/ tw'n
ploivwn dia; calkw'n kai; sidhvr wn leovntwn kai; ajlloivwn cersaivwn zwv/wn kefala;" meta;
stomavtwn ajnew/gmevnwn kataskeuavsa", crusw'/ te peristeivla" aujta; wJ" ejk movnh" qeva"
fobero;n faivnesqai, to; dia; tw'n streptw'n kata; tw'n polemivwn mevllon ajfivesqai pu'r dia;
tw'n stomavtwn aujt w'n pareskeuvase dii>evnai, w{ste dokei'n tou;" levo nta" kai; ta\lla tw'n
toiouvtwn zwv/wn tou'to ejxereuvgesqai.”.
47
See Appendix Four [b], §VI.22 [= Haldon, “Theory and practice”, p. 231;
Constantine VII, De cerimoniis, II.44 (vol. 1, pp. 675-6)]. On the construction of
sipho2nes, see now Haldon “ ‘Greek Fire’ revisited”.
GREEK FIRE 623
discussion of Greek Fire is that of a force pump. The three most

detailed classical discussions of force pumps were by Ktesibios of
Alexandria (fl. ca 270 B.C.E.), Philo2n of Byzantium (fl. ca 240
B.C.E.), and He2ro2n of Alexandria (fl. ca 50 C.E.). Vitruvius attributed
the invention of the force pump to Ktesibios.48 A different version of
one appears in a text related to the Arabic version of the Pneumatika,
book five of his Mechanike2 syvntaxis, of Philo2n and He2ro2n described
in detail the construction of one used for a fire engine and called a
sifo2n.49 Such force pumps were also mentioned by other classical
authors and in all but two or three cases among seventeen catalogued
literary references the word used for them was either sifo2n or
sipo/sifo/sipho.50 He2ro2n of Alexandria’s pump consisted of two
cylinders set into a sump with pistons operated in them in tandem via
a pivoted connecting rod. Water was sucked into the cylinders through
valves in their bases by the up stroke of the pistons and was then
expelled into a horizontal connecting tube through other valves by the
down stroke. The pistons forced the water out through the connecting
tube into a mouth tube set into it, the latter being fitted with joints
allowing the nozzle to be swivelled in any direction, both laterally and
vertically. The whole machine was made of bronze. Surviving
manuscripts of Philo2n of Byzantium and He2ro2n of Alexandria contain
drawings of such force pumps.51
At least 21 actual examples survived in whole or in part into
modern times. Nine of these, dating from the first to third centuries,
were made of bronze, four of them having double cylinders and 5
having single cylinders, the latter all coming from the shipwrecks
Dramont D and La Tradelière. These were most probably linked in
tandem and were bilge pumps. However, other locations in which the
pumps have been found, including mines, suggest that they served a
variety of purposes. The most complete and spectacular still surviving
example is the pump from the mine at Sotiel Coronada near Valverde,
province of Huelva, Spain. Surviving pumps vary greatly in size from
the smallest, which had an estimated discharge capacity of 0.236
litres, to the largest, which had a discharge capacity of 3.4 litres. The
Sotiel Coronada pump has a cylinder bore diameter of 7.5 centimetres
------------------------------
48
Vitruvius, De architectura, X.vii.1-3 (p. 239).
49
Philo2n of Byzantium, Me2chanike2 syntaxis, pp. 192-4; He2ro2n of Alexandria,
Pneumatika, 1.28 (pp. 130-37).
50
See Oleson, Water-lifting devices, p. 20 for a list of references. All the texts are
cited in Greek or Latin and translated. Force pumps are discussed at pp. 300-25.
51
Schiøler, “Piston pumps”, p. 19; Oleson, Water-lifting devices, figs 13, 14, 27,
28.
624 APPENDIX SIX
and height of 27 centimetres, with a cylinder displacement of 0.795

litres and a discharge capacity of 1.60 litres. The diameter of flow at
the nozzle is narrowed to a mere 8 milimetres so that the velocity of
flow is increased 25 times.52 There appears to have been no attempt to
estimate how far such a pump could eject its spurt of water; however,
since Isidore of Seville wrote they were used to clean the ceilings of
buildings, the water was obviously capable of being expelled with
considerable force for quite a distance.53
Ktesibios’s work is known only through Vitruvius. The Pneumatika
of Philo2n of Byzantium is known only through a ninth-century Arabic
translation. The oldest surviving manuscript of the Pneumatika of
He2ro2n of Alexandria is the thirteenth-century manuscript Venice,
Biblioteca Marciana, MS. Gr. 516. Whether the descriptions of force
pumps by such authors were known in Byzantium is a moot point;
however, Hero2n of Byzantium certainly knew sipho2nes as fire-engines
in the tenth century.54
There are simply too many parallels between the names and
physical attributes of these Greco-Roman force pumps and the
Byzantine Greek-Fire sipho2nes for it to be mere coincidence. Surely
what Kallinikos did was to adapt the idea of the force pump to
projection of some form of processed petroleum naphtha.
Immediately after mentioning the sipho2nia, the inventory for the
Cretan expedition of 949 said that there should be 40 extra gonavtia
ajkovntia (gonatia akontia), literally “hinged or jointed poles or pikes”,
for the boukovl ia (boukolia), often interpreted as bucklers or shields.55
The context of this specification means that it must have had
something to do with the sipho2nia. Gonatia akontia must have had the
sense of something jointed and pointed. Assuming that the only
sipho2nia requiring some kind of side protection were those at the bow,
then the 40 gonatia akontia would correspond to two boukolia or
guards either side of the 20 sipho2nia for the 20 dromons. They may
have been hinges by which the boukolia were swung into place when
going into battle. The boukolia themselves may have been heat shields
------------------------------
52
Schiøler, “Piston pumps”; Rouanet, “Quatre pompes”; Oleson, Water-lifting
devices, pp. 192-5, 198-9, 206-7, 219, 268-9, 313-16, 321.
53
Isidore of Seville, Etymologiae, XX.vi.9: “Sifon vas appellatum quod aquas
sufflando fundat; utuntur enim hos [in] oriente. Nam ubi senserint domum ardere,
currunt cum sifonibus plenis aquis et extingunt incendia, sed et camaras expressis ad
superiora aquis emundant.”.
54
He2ro2n, Parangelmata poliorke2tika, §39, ll. 24-8 (pp. 84-5).
55
Constantine VII, De cerimoniis, II.44 (vol. 1, p. 672)]. See also Constantine VII, De
cerimoniis, vol. 2, p. 794 and now Haldon, “‘Greek Fire’ revisited”.
GREEK FIRE 625
Figure 60
The Sotiel Coronada

Roman force pump from
Valverde, Huelva, Spain,
probably dated to the first
century C.E., Madrid,
Museo Arqueologico
Nacional.
In Haldon’s experiment the heat generated by the device was so great

that the sipho2nator needed a shield between himself and the nozzle.56
It is noticeable that the illustration of the Greek Fire sipho2n on folio
34v of the Madrid Skylitze2s manuscript and also that on folio 23r of
the Marciana manuscript of the Kyne2getika of Pseudo-Oppian, if the
latter is indeed of a sipho2n, both show the weapon as a flared tube.
Force-pump nozzles must have been within such tubes, which would
have operated as heat shields. It is certainly possible that such shields
may have been for this purpose and it is to be noted that Leo VI and
Nike2phoros Ouranos both wrote that the cheirosipho2nes were also
------------------------------
56
See Haldon, “ ‘Greek Fire’ revisited”.
626 APPENDIX SIX
operated from behind iron shields.57 Shields were not normally made
of iron and it must have been the heat generated which made this
necessary.
The inventory for the Cretan expedition of 949 also said that there
should be 100 tetravkoula (tetrakoula) for the sipho2nia,58 the meaning
of which is unknown. Reiske suggested emendation of tetravkoula to
tetravkwla, something four-legged, and that they were carriages for
the sipho2nia performing the same function as the gun-carriages of
later times, which is a possibility. Haldon suggests four-legged grates
or bases, perhaps for the hearth on which a brazier rested, again a
possibility.59
Immediately after this item the inventory also specified that there
should be 400 linaria (some things made of flax) for the “sponges”,
sfovggoi (sphongoi).60 This phrase appears to be corrupt. Reiske
suggested emendation to “400 [pounds] of flax/linen for making
sponges”.61 But what was the sense of “sponge” here? Obviously, if
flax or linen was involved, the “sponges” had nothing to do with
natural sponges. Perhaps they were flax or linen mops for cleaning the
barrels of the sipho2nia. In the inventories for the expedition of 911 to
Crete, there was also a specification for 10,000 [units] of linaria for
caulking and for the provpura (propyra), “fore-fires”.62 “Fore-fires”
sounds suspiciously like something required to prepare the sipho2nia
for “firing”, perhaps wicks of match. If the fuel was forced through
the nozzle of a force pump in a fine spray, then a burning wick of
match underneath the nozzle may have been what ignited it. The word
provpuron is otherwise almost unknown in Byzantine Greek except
that Leo VI wrote of the processed fire that it was expelled from
sipho2nes with thunder and “forefire” smoke. This appears to make
little sense and it is surely no accident that Nike2phoros Ouranos
emended the emperor’s syntax to convert the adjective provpuro" to
the noun provpuron, changing the syntax to mean expelled from
------------------------------
57
58
59
See Haldon, “Theory and practice”, p. 283; idem, “‘Greek Fire’ revisited”;
Constantine VII, De cerimoniis, vol. 2, pp. 795-6.
60
61
See Constantine VII, De cerimoniis, II.45 (vol. 1, p. 673) [the Latin translation].
62
See Haldon, “Theory and practice”, p. 211: “... kai; peri; tou' eJtoimasqh'nai
linavrion lovgw/ tw'n propuvrwn kai; kalafathvsew" ciliavda" iV, ...”; Constantine VII, De
cerimoniis, II.44 (vol. 1, p. 658); Appendix Two [a], §59.
GREEK FIRE 627
sipho2nes with thunder and smoke from the “forefires”.63 The Greek
prefix prov- could have the sense, amongst many others, of something
in front of something else, either chronologically or spatially. What
comes before fire chronologically is smoke. But Leo VI can hardly
have meant “smoking smoke”. The way in which Nike2phoros Ouranos
changed his syntax suggests that the magistros knew that propyra
were a physical part of the sipho2nes but that the emperor did not.
Haldon argues that the 10,000 units of flax must have been for flax
fibre fires to heat the oil on the grounds that such large amounts would
not have been needed for ignition wicks. That is certainly true, but it
overlooks the other use specified for the flax, caulking. With a fleet
the size of that of 911 a huge amount of caulking material would have
been needed. Moreover, if one wanted a slow match-like source of
heating which was safe at sea, why would one have used flax rather
than charcoal, which was used universally for such purposes? Against
that, however, charcoal would not have produced the large amounts of
smoke apparently associated with the weapon and, certainly, Haldon
has made a weapon work using a fire of flax fibre.64 The question
remains open, as does that of whether the fuel was actually heated
aboard ship or not.
Haldon is surely correct to point to the fact that whatever the raw
material of Greek Fire was, whether petroleum oil or something else,
it was processed in some way before use. Leo VI and Nike2phoros
Ouranos are very clear that it was the same processed fire fuel which
was used for the cheirosipho2nes and that it could also be hurled in
pots, or poured from what may have been cauldrons hung from cranes,
or have tow-wrapped caltrops soaked in it. However, Haldon’s
reconstruction of a sipho2n as a complex arrangement of an oil
container, bellows and hearth to heat the oil, pump, and tube and
nozzle, could have been applicable only to a heavy weapon fixed in
place; such as a main sipho2n at the bow of a dromon. He concludes
that the cheirosipho2nes were different and that they merely squirted
unignited raw material.65 They undoubtedly were different in some
ways since, at least as claimed, they were not developed until over two
centuries after Kallinikos invented the original sipho2nes. However, the
illustration to folio 36r of He2ro2n’s Parangelmata poliorke2tika [see
Fig. 59] shows that they did throw the flame itself, not only unignited
fuel. Moreover, Haldon’s original arrangement of a tank of oil under
------------------------------
63
64
See Haldon, “‘Greek Fire’ revisited”; Haldon and Byrne, “Greek Fire”, p. 94.
65
Haldon and Byrne, “Greek Fire”, p. 97, n. 19; Haldon, “‘Greek Fire’ revisited”.
628 APPENDIX SIX
pressure over a burning hearth would have been highly dangerous, yet
there are no recorded instances of fire ships blowing up. In his second
experiment he abandoned this approach for this reason and used a
force pump between the tank of heated oil and the nozzle to put the
fuel under pressure.66
We are convinced that the essential mechanism of sipho2nes was an
adaptation of a Greco-Roman force pump.67 The fuel itself was
undoubtedly processed from petroleum obtained at various times from
wells in the regions of Tmutorakan, Tziliapert, Erzurum, and Zichia
on the north coast of the Black Sea, in Georgia, in Eastern Turkey, and
on the east coast of the Black Sea respectively. That much at least was
revealed by the De administrando imperio. Even today, petroleum
from these regions seeps to the surface through clay sediments and
reaches the surface as very light crude.68 What was then done to it to
“process” it is a matter of debate. Haldon and Byrne considered the
idea of distilling it but rejected this on the grounds that it would have
been too dangerous. Since then Haldon has continued to use raw
petroleum, although he has added around three kilogrammes of pine
resin per 45 litres of fuel to make the fuel burn longer, be more
adhesive, and burn at a higher temperature. By heating the fuel/resin
mixture for his weapon, he has produced a very fluid liquid which
burns very readily. When his weapon is used without heating the
fuel/resin mixture, it does still ignite at the nozzle but only partially.
Heating produces a longer range, up to 15 metres, and more fierce
heat.69
Whether or not the fuel was indeed actually heated aboard ship is
another issue. In fact, the only evidence that the fuel was heated in
some sort of container on a hearth comes from two highly
questionable Western sources, the obscure text of the Wolfenbüttel
manuscript and Yngvars Saga Ví∂förla. There is nothing in the
------------------------------
66
Haldon, “Theory and Practice”, pp. 278-80; Haldon, “‘Greek Fire’ revisited”.
67
Against all the evidence in the sources, logical argument, and experimentation,
Korres has continued to maintain that force pumps could not possibly have been the
main mechanism of sipho2nes and has continued to insist that Greek Fire was projected
by catapults only, his main argument being that force pumps could not possibly eject
the fuel for a sufficient distance. See Korres, «ÔUgro;n pu'r»; idem, “Greek Fire”.
However, if the fuel was distilled to a consistency no more viscous than water, since
we know that force pumps were used to clean the ceilings of ancient temples, there is
absoltuely no reason why they could not have projected a tongue of flame for a
sufficient distance. Whatever that may have been is arguable.
68
Constantine VII, De administrando imperio, §53, ll. 493-511 (pp. 284-6). See
also Haldon and Byrne, “Greek Fire”, p. 92 & n. 4; Haldon, “‘Greek Fire’ revisited”.
69
Haldon and Byrne, “Greek Fire”, p. 92; Haldon, “‘Greek Fire’ revisited”.
Additional information in a personal communication from John Haldon to John Pryor.
GREEK FIRE 629
Byzantine or Arabic sources to suggest heating of the fuel aboard ship.

If it were not for these two Western texts, would heating aboard ship
be even considered?
Figure 61
“Greek” or liquid fire siphon built by Colin Hewes and Andrew Lacey under
the direction of John Haldon.
© John Haldon
The question of how the Byzantines may have “prepared” or

“processed” the fuel is related to this. Virtually the one constant in the
Byzantine sources is that the fuel was prepared or processed in some
way. Refining petroleum, leaving aside the modern cracking process,
is fundamentally a simple process of distilling the oil by heating it
until various fractions are given off at various temperatures and then
recondensed. The lighter the fraction, the lower the temperature
necessary. The lightest liquid fraction, gasoline, will separate out at
temperatures between 38˚ and 204˚ Celsius, temperatures which
Byzantine technology was certainly capable of achieving. To make the
point, the melting point of copper is 1083˚ Celsius and Romans and
Byzantines could certainly cast copper. The temperatures necessary
would not have posed any problem and distillation techniques were
well known in the Greco-Roman world.70 Dioskorides discussed the
distillation of pine pitch and there is no reason why petroleum could
------------------------------
70
See Forbes, Art of distillation, pp. 13-28.
630 APPENDIX SIX
not have been distilled either.71 The light crude seeping to the surface
in the regions enumerated in the De administrando imperio would
have been relatively easy to distill if the potential dangers from
volatile gasses could be overcome. Distillation of petroleum became
widespread from the seventeenth to the nineteenth centuries,72 and,
although there is no evidence that they actually did so, there seems no
reason prima facie as to why the Byzantines would not have been able
to do so and to produce a light paraffin or kerosene which could have
been used in a force pump without the need to heat the fuel. What
Kallinikos may have achieved was to distill petroleum to produce a
light paraffin or kerosene. This appeals as an appropriate
understanding of what the process of “preparing” the fuel may have
been.
After the tenth century, Greek fire and fire-bearing ships continued
to be mentioned by various authors; for example, Michael Psellos,
who described in classicizing language the remnants of the imperial
fleet which scattered the last Rho2s attack on Constantinople in 1043 as
being composed of trie2reis and fire-carrying ships, purfovroi nh'e"
(pyrphoroi ne2es), and who described their use of Greek Fire, with
which the Rho2s ships were destroyed.73 John Kinnamos and Nike2tas
Cho2niate2s also referred to the continuing use of Greek Fire in the
twelfth century. Kinnamos wrote that the Salju2qid sulta2n ‘Izz al-Dı3n
Qı£lı£j Arslan II was treated to a demonstration of it when he visited
Constantinople in 1162. Fire ships were prepared against the Normans
of Sicily in 1147 and pursued a Venetian ship fleeing Constantinople
at the time of the arrest of the Venetians in the Empire in 1171. Liquid
fire was also used against the fleet of the rebel strate2gos Alexios
Branas in 1187.74 However, very interestingly, when Cho2niate2s
described the preparations of Alexios III Angelos to resist the
imminent arrival of the Fourth Crusade, he made no mention of Greek
Fire. He said that the emperor failed to construct warships,
polemisthvriai nh'e" (polemiste2riai ne2es), and only at the last moment
supposedly repaired what rotting little skiffs, skafivdia (skaphidia), he
could find at Constantinople. Cho2niate2s’s account should be read with
some skepticism since he was seeking to explain why civilization as
------------------------------
71
Dioskoride2s, De materia medica (Wellmann), I.72 (vol. 1, p. 71).
72
See Forbes, Studies; idem, More studies, passim.
73
Michael Psellos, Chronographia, XCIII (vol. 2, p. 10).
74
John Kinnamos, Historiae, V.3, VI.10 (pp. 207, 283); Nike2tas Cho2niate2s,
Historia, Basivl eiva Manouh;l tou' Komnhnou' BV (p. 77); Basivl eiva Manouh;l tou'
Komnhnou' EV (p.172); Basvleiva Isaakivo u tou' Aggelou' AV (p. 381).
GREEK FIRE 631
he knew it had been destroyed by the Crusaders. His pejorative

language may well have exaggerated. Nevertheless, it is clear that by
1203 the negligent rule of the Angeloi emperors had allowed the great
navy created by the first three Komne2noi emperors to decay and that
no effective opposition to the Venetian battle fleet could be mounted.
According to Nike2tas, the Venetians covered their galleys with ox
hides as protection against fire, almost as though they expected to
have to counter Greek Fire; however, neither he nor any Latin
chronicler mentioned it actually being used against them. Some naval
resistance was mounted in the Golden Horn by a few Byzantine
“trie2reis” but these were either destroyed or driven ashore and
abandoned.75 It is striking that whereas in 1043 the remnants of a
similarly decayed Byzantine navy had been able to scatter the Rho2s
attack on Constantinople with Greek Fire, it was apparently not used
in 1203. The implication is clear. At Constantinople in 1203 the
Byzantine galleys were not equipped with Greek Fire.
What happened to the sipho2nes for Greek Fire? Why did they
apparently fall out of use? Was it simply the case that the Byzantines
lost access to the sources of their fuel?76 The Erzerum region was
certainly part of the Empire until the eleventh century but Tmutorakan
was under Khazar and then Rho2s rule from the eighth to eleventh
centuries and Zichia was probably never under Byzantine rule. It is
true that the sources from which the fuel was obtained would all have
been lost to the Empire by the end of the eleventh century. But, with
the exception of those around Erzurum, they had not been within the
frontiers of the Empire even in the age of Constantine VII. The
Byzantines certainly had a presence and influence in the Tmutorakan
and Zichia regions but would nevertheless have had to have obtained
petroleum from them by trade or through relationships with client
states. These regions were later conquered by the Mongols, who
encouraged free trade, yet this does not appear to have enabled
supplies to flow again.
------------------------------
75
Nike2tas Cho2niate2s, Historia, pp. 540, 541, 544.
76
See Haldon, “ ‘Greek Fire’ revisited”.
APPENDIX SEVEN
THE GALLEYS OF THE MANUSCRIPT, MADRID, BIBLIOTECA

NATIONAL, VITR. 26-2, OF JOHN SKYLITZE›S’ SYNOPSIS
HISTORIO›N AND ITS DATING AND ART STYLES
The renowned Madrid manuscript of the Synopsis historio2n of John

Skylitze2s (Biblioteca Nacional, Madrid, vitr. 26-2), which is the only
surviving illustrated manuscript of any Byzantine historian, has been
the object of much attention, particularly from art historians concerned
with its 574 illustrations and from palaeographers concerned with its
dating. Nigel Wilson dated the manuscript to the mid twelfth century,
assigned to it a provenance in the Norman Kingdom of Sicily,
probably the royal court at Palermo, and argued convincingly that the
manuscript and its illustrations were, at least initially, copied from a
de luxe illustrated Byzantine manuscript which may have come from
imperial circles in Constantinople and have been brought back to
Sicily by the embassy of Henricus Aristippus to Constantinople in
1158.1 The complete series of its illustrations was published by
Estopañan and the contributions of various artists to the miniature
series were discussed at length by Grabar and Manoussacas. A
facsimile of the entire manuscript was produced in 2000 and most
recently Tsamakda has subjected the codex to palaeographical,
historical, and art-historical analysis, including good quality
reproductions of the complete corpus of illustrations. She has
concluded that the manuscript was produced in the scriptorium of the
Basilian Greek monastery of San Salvatore in Messina some time in
the third quarter of the twelfth century.2
------------------------------
1
See Wilson, “Madrid Skylitzes”.
2
EstopanNan, Skyllitzes Matritensis; Grabar and Manoussacas, L’illustration;
Tsamakda, Ioannes Skylitzes. Grabar and Manoussacas reproduced only around half
the illustrations. For further discussion of the scholarship on the codex see Tsamakda,
Ioannes Skylitzes, pp. 1-21.
Tsamakda’s reassignment of the manuscript to the monastery of San Salvatore
raises more questions than it answers. From where would such a monastery have
obtained the archetype to copy? Why would a Basilian monastery produce such a
deluxe manuscript and why would it employ a miscellany of artists rather than its own
monks. The artists painting in Western styles are very unlikely to have been Basilian
monks. If it was produced by the monastery for either the royal court or for some
Greek patrician, why did it remain in the monastery’s library after completion?
634 APPENDIX SEVEN
Table 8: Artists of the manuscript Madrid, Biblioteca Nacional, Vitr. 26-2 of
the Synopsis historio2n of John Skylitze2s according to (1) Grabar and
Manoussacas and (2) Tsamakda
Quires Folios Artists Notes
Grabar and Tsamakda

Manoussacas
1 9-163 Aa
9r-16v A1
2 17-24 Ab
17r-24v A2
3 25-32 Aa
25r-32v A1
4 32-40 Aa
33r-40v A1
5 41-48 Bc
41r-48v A1
6 49-56 Bc
49r-56v A1
7 57-63 Ab?
57r-63v A2
8 64-71 Ab
64r-71v A2
9 72-79 Bc
72r-79v A1
10 80-87 Bc
80r(a) B1
80r(b)-87v A1
11 88-95 No miniatures
12 96-102 Cd
96r-102v B1
13 103-110 Cd
103r-110v B1
14 111-118 Cd or Ce Cd or perhaps a different artist Ce
111r-118v B1
15 119-126 Ce or C? Ce if Cd was the artist of quire 14. C?
if Ce was the artist of quire 14
119r-126v B2
16 quire missing
17 127-134 Cd
127r-134v B1
18 135-142 Cd
135r-142v B1
19 143 C?? Another artist responsible for all
144-5 Cd except folios 144-5, for which Cd was
146-50 C?? responsible.
143r-v B3
144r-145r B4
145v-150v B3
20 151-156 Cf
151r-156v B3
------------------------------
3
The Synopsis historio2n begins on fol. 9r.
GALLEYS OF THE MADRID SKYLITZEfiS 635
(Table 8 continued)
21 157-164 Dg
157r-164v B5
22 165-172 Dg
165r-172v B5
23 173-178 Dg
173r-178v B5
24 179-186 Dg
179r-186v B5
25 187-94 No miniatures
26 195-202 Cf
195r-v B2
196r-200r B5
200v-201r B2
201v B5
202r-v B2
27 203-210 Cd or Ce B1
203r-210v
28 211-218 Cd or Ce
211r-218v B1
29 219-226 Dg
219r-226v B5
30 227-234 Ab
227r-234v A1
An analysis of the styles of depiction of galleys in the manuscript has

not previously been attempted; although, Babuin has made a selective
study of some of the more important illustrations and has made some
interesting observations. This is curious because the galley
illustrations differ so markedly, and some can so clearly be assigned
to different artists, that they are in fact extremely good evidence for
the various arguments. Although Tsamakda devoted a long chapter to
the iconography and its sources, discussing in particular combat and
battle scenes, she paid no attention to the ships at all, with the single
exception of the imperial galley using Greek Fire against Thomas the
Slav on fol. 34v.4
The illustration series for the manuscript was executed by a series
of artists working in different styles. On various grounds, Grabar and
Manoussacas identified four different styles of art in the miniatures
and at least seven different artists. For the most part, they deduced that
an individual artist was responsible for the illustrations in any one
------------------------------
4
See Babuin, “Illuminations”; Tsamakda, Ioannes Skylitzes, p. 314. Babuin
misunderstands many of the illustrations she discusses.
636 APPENDIX SEVEN
Table 9: The galleys of the manuscript Madrid, Biblioteca Nacional, Vitr. 26-2 of
the Synopsis historio2n of John Skylitze2s classified by artistic style.
Galley Folios Description

Group
Byzantine 14v, 15r, 26r, These galleys are drawn in a rounded “banana boat”
One 31v, 32r, style, with only one bank of oars, with no spurs or
33v, 35v, ornaments at the prow, but with duplex stern ornaments.
38r, 38v, 39r Two (14v, 15r) have single masts with beaked mastheads
39v, 40v, and obvious lateen sails. In these same two cases the stern
227r ornaments have been applied mistakenly to the bow and
in one other (38v) some oarsmen are mistakenly facing
the bows. In a number of cases (14v, 31v, 33v, 35v, 38r,
38v, 39v), some of the oarsmen wear lamellar cuirasses.
The galleys on fol. 38v clearly have two quarter rudders.
The ship on fol. 15r appears to have an inscription on the
upper strake at the stern. These galleys, and also those of
groups Byzantine Two and Byzantine Three, bear many
similarities to others found in Byzantine manuscripts from
Mt Athos.5 One, the well-known depiction of the arrival
of Thomas the Slav at Abydos (fol. 31v) shows both a
forecastle at the bow of his galley and also horses on one
of the accompanying galleys. Why these galleys and those
of Groups Two and Four do not include such fundamental
characteristics of Byzantine galleys as spurs is difficult to
comprehend.
Byzantine 20v, 21r There are only two examples in this group. These
Two galleys are very similar to those of Byzantine One except
that they have a higher, more recurved prow and a triplex
stern ornament, in one of them made out to be the bow.
The colouring and style is so distinctive that it would be
difficult to believe that anyone other than a unique artist
drew these two galleys. The galley on fol. 21r appears to
show both of the two helmsmen.
Byzantine 29v There is only one illustration in this group. The galley
Three is also drawn in a rounded “banana boat” style and with
only one bank of oars, but has have duplex ornaments at
both bow and stern. It has a pronounced spur. Here again
the oarsmen have been mistakenly drawn facing the bow,
probably because the object of the illustration, Thomas
the Slav, is also depicted at the bow. This is also the first
illustration in the manuscript to show a galley flying a
standard, composed of a head and three streamers.
Byzantine 34v This illustration is also unique. It is the famous
Four illustration of an imperial galley attacking one in the fleet
of the rebel Thomas the Slav in 821 with Greek Fire. The
galleys are drawn in a rounded “banana boat” style and
with only one bank of oars, but they have no spurs and no
stern or bow ornaments. The imperial galley has a mast
with a beaked masthead and a lateen sail and two shields
hanging from the gunwale.
------------------------------
5
See Pelekanides, OiJ Qhsauroiv, figs 55 (p. 61), 79 (p. 69), 299 and 300 (p. 175).
(Table 9 continued)
Byzantine 41r, 44r There are only two illustrations in this group, both in
Five quire 5. The galleys have extremely high, markedly
recurving duplex stern ornaments but low bows devoid of
ornamentation. There is a similarity in the low bows to
those of the dromon in the manuscript of the Sermons of
St Gregory of Nazianzos, Mount Athos, Pantelee2mon,
Cod. 6 [see Fig. 47] and in both cases spurs may have
been intended. The galleys of fol. 41r both have two
quarter rudders. The right-hand galley of fol. 44r
mistakenly has the stern ornaments at the bow.
Western 110v, 111v There are only two illustrations in this group, at the end
Six of quire 13 and the beginning of quire 14. These galleys
are depicted in a completely new style. They are very long
and flat, show only one row of oar ports, have duplex
stern ornaments, and have pronounced spurs but no stem
posts. On fol. 111v the four galleys representing the fleet
of Leo of Tripoli attacking Thessalonike2 have lavishly
decorated sterns, probably intended to suggest Muslim
ships, and two quarter rudders. In one of these the galley
has three oars at the stern rowed in a second file from
above the gunwale in addition to the file rowing through
oarports. In terms of the manuscript, this is the first
depiction of the new Western bireme galeae of the late
eleventh and early twelfth centuries. Two of the galleys
on fol. 110v have recurved stemposts like those of Group
Seven.
Western 110v, 123v, This is a large group of illustrations. These galleys are
Seven 124r, 129v, very similar to those of group Six but in addition they
130r, 132v, have prominent recurved stem posts. Some (fol. 130r,
134v, 138v, 234v, 138v, 146v) have two quarter rudders. All have
140r, 146v, pronounced spurs. In two cases (130r, 132v) the oarsmen
212r face forward. There is no doubt that this group also
represents the new Western bireme galea of the late
eleventh and early twelfth centuries.
Western 145r There is only the one unique illustration in this group
Eight on fol. 145r. The galleys in it are slightly curved in a
“banana boat” style similar to those of Groups One, Two
and Three, although much less markedly so. They have
duplex stern ornaments, recurved stemposts, prominent
spurs, and two files of oars, one rowed from above the
gunwale and the other through oar ports below it. They
are the best representations of Western bireme galeae
before the Peter of Eboli illustration [see Figure 54].
Western 146v There is only the one illustration in this style on folio
Nine 146v. It is similar to galleys in the style of group Seven
but two of the galleys have two files of oars using the
same oarage system as in the one illustration on fol. 111v
of Group Six and Group Eight.
Western 146v, 147r, One of the galleys on fol. 146v is of a different style
Ten 147v (twice), also found on subsequent folios of quire 19. These are
149v drawn in a heavy “banana boat” style devoid of either
stern or bow ornaments but with recurved sternposts and
638 APPENDIX SEVEN
(Table 9 continued)
stemposts. They seem to represent a reversion by an artist

to a rather crude version of an earlier Byzantine style
Muslim 157r, 159r, These galleys are flat, with pronounced spurs and with
Eleven 167v, 168v, identical fan-shaped ornaments at the bow and the stern.
219v, 222r, Their style is extremely distinctive and was probably
224r, 225v, Muslim inspired.6 Many of the illustrations are extremely
226r, 226v minimalist. Fol. 168v shows the only galley in the
Western or Muslim styles to carry a mast and sail, clearly
lateen. One of the galleys on fol. 226r clearly shows two
quarter rudders and two helmsmen.
quire; although, some artists were responsible for the illustrations of

several quires and a few quires seemed to them to have been
illustrated by more than one artist. They concluded that the first two
styles (A and B), found in the first ten quires of the manuscript, folios
9-87, were based on Byzantine styles of the eleventh or early twelfth
centuries, whereas the second two (C and D), found in the remaining
quires 12 to 29, folios 92-226, reflected Sicilian or South Italian styles
of the twelfth or thirteenth centuries and some Muslim influences.
Quires 11 and 25 were not illustrated. Spaces for illustrations were
left in the text of the manuscript in these quires but were never filled
in. Three artists worked in the two Byzantine styles A and B and
Grabar and Manoussacas numbered them Aa, Ab, and Bc. At least
four artists worked in the two Western and Muslim styles and they
numbered them Cd, Ce, Cf, and Dg. Beyond this, Grabar and
Manoussacas had difficulties with some quires and some folios where
it appeared to them that various artists may have collaborated or
where the identification of a particular artist was uncertain. Tsamakda
has returned to the analysis of the artists and their art styles, reaching
conclusions which are somewhat different to those of Grabar and
Manoussacas but without providing any reasoning for her differing
opinions.7 Whereas Grabar and Manoussacas believed that they could
distinguish three Byzantine-style artists and at least four Western-style
artists, Tsamakda distinguishes only two Byzantine style artists but
five Western-style ones, numbering the artists A1-2 and B1-5. Both of
their conclusions are summarized in Table 8.
------------------------------
6
They have similarities to depictions of Muslim galleys from Egypt and Iraq. See
Christides, “Naval history”, figs 6 and 9; idem, “Dha2t as5-S4awa2rı3”, fig. 2.
7
Grabar and Manoussacas, L’illustration, pp. 169-95; Tsamakda, Ioannes
Skylitzes, pp. 373-8.
Turning to the 48 illustrations of ships in the manuscript, almost all

of them clearly depict galleys rather than sailing ships. There are only
two which depict sailing ships. One represents the famous large
sailing ship owned by his wife Theodo2ra, to which the emperor
Theophilos took such objection, and the other is apparently a small
boat.8 They were drawn by Bc and either Cd or Ce (Grabar and
Manoussacas) and A1 and B1 (Tsamakda). Although they have been
included in Table 10 below, they have been excluded from the
analysis of the various styles of galley depictions given here. As far as
the illustrations of galleys are concerned, there is no evidence that any
of the artists depicted anything other than what were either copies of
illustrations of galleys as they found them in the original Byzantine
manuscript or else depictions of contemporary galleys as they knew
them. As Wilson has written:
... it must be said that the absence of an illustrated copy to work from
would have forced the illuminators to apply their inventive energies in the
styles to which they were accustomed. The alternative is to assume that
the model was a book from Constantinople. In that case the assumption
must be that it was illustrated in a more or less coherent, at any rate fully
Byzantine style, but that some of the Sicilian illuminators could not or
would not copy the model very closely. There is no harm in crediting
them with a desire to show a degree of independence or originality.9
Medieval artists working under commission normally followed

programmatic models as laid down by their employers.10 Therefore, if
the original Byzantine manuscript of Skylitze2s was illustrated but
some of the artists of the extant manuscript apparently emulated the
styles of illustration as found in the original manuscript whereas
others did not do so and drew illustrations in other styles, there can be
only one of two conclusions. Either the artists who did not follow the
Byzantine styles were allowed to deviate from the original by their
employers or else the illustrations of those parts of the manuscript that
they illustrated had been lost by the time that they were working, or
had never existed, and they and their employers had no style to
emulate. The original manuscript from Constantinople may have been
only partially illustrated.
------------------------------
8
See Table 10, nos 19, 40.
9
Wilson, “Madrid Skylitzes”, p. 216.
10
Which in itself raises another question not considered by Tsamakda. If the
manuscript was commissioned, why did the commissioner not require the execution
of the miniatures in any one particular style?
640 APPENDIX SEVEN
Table 10: Comparison of styles of depictions of galleys in the manuscript Madrid,
Biblioteca Nacional, Vitr. 26-2 of the Synopsis historio2n of John Skylitze2s with the
general styles and artists as identified by (1) Grabar and Manoussacas and
(2) Tsamakda
Pryor and Jeffreys11 Grabar and Manoussacas Tsamakda
Folio Number Galley group Figure (Plate) Artist Figure Artist
14v 1 One - Aa 16 A1
15r 2 One - Aa 17 A1
20v 3 Two - Ab 33 A2
21r 4 Two - Ab 34 A2
26r 5 One - Aa 49 A1
29v 6 Three 15 Aa 57 A1
31v 7 One 20 Aa 61 A1
32r 8 One - Aa 63 A1
33v 9 One - Aa 67 A1
34v 10 Four 24 (pl. VI) Aa 70 A1
35v 11 One - Aa 73 A1
38r 12 One pl. VII Aa 79 A1
38v 13 One 29 (pl. VIII) Aa 81 A1
39r 14 One - Aa 82 A1
39v 15 One - Aa 83 A1
40v 16 One - Aa 86 A1
41r 17 Five 30 Bc 88 A1
44r 18 Five - Bc 98 A1
44r 19 Sailing ship - excluded
110v 20 Six 129 Cd 251 B1
111v 21 Six 132 Ce 254 B1
123v 22 Seven - Ce 287 B2
124r 23 Seven 144 & 145 Ce 289 B2
129v 24 Seven 156 Cd 309 B1
130r 25 Seven 157 [pl. XXVIII] Cd 310 B1
132v 26 Seven 161 Cd 317 B1
134v 27 Seven? - Cd 322 B1
138v 28 Seven - Cd 334 B1
140r 29 Seven 172 Cd 339 B1
145r 30 Eight 183 Cd 357 B4
146v 31 Nine & Ten - C? 363 B3
147r 32 Ten - C? 364 B3
147v 33 Ten - C? 366 B3
147v 34 Ten - C? 367 B3
149v 35 Ten & Seven - C? 375 B3
157r 36 Eleven 202 Dg 395 B5
159r 37 Eleven 207 Dg 401 B5
167v 38 Eleven 214 Dg 421 B5
168v 39 Eleven 215 [pl. XXXII] Dg 423 B5
208v 40 small boat - excluded
212r 41 Seven 249 Cd or Ce 501 B1
219v 42 Eleven - Dg 521 B5
222r 43 Eleven - Dg 526 B5
------------------------------
11
Our references are to the manuscript and to the reproduction John Skylitze2s,
Suvnoyi" iJstoriw'n.
(Table 10 continued)
224r 44 Eleven 266 Dg 530 B5

225v 45 Eleven - Dg 535 B5
226r 46 Eleven - Dg 536 B5
226v 47 Eleven - Dg 537 B5
227r 48 One - Ab 538 A1
It should be noted that there is one very long section of the manuscript
between quire 5, folio 44r and quire 13, folio 110v which has no
illustrations of ships at all. Another, somewhat shorter, section
between quire 22, folio 168v and quire 27, folio 208v also has no
depictions of ships. These lacunae may have been the result of nothing
more than happenstance or there may have been something more to it.
Certainly there are plenty of matters naval and maritime in the
sections of the Synopsis historio2n in question which might have been
illustrated.
Our analysis of the illustrations of galleys in the manuscript has led
us to classify them in eleven groups as in Table 10. Comparison of the
styles of galley depictions with the general analysis of the art styles
and individual artists by Grabar and Manoussacas on the one hand and
Tsamakda on the other has a number of instructive implications.
Leaving aside consideration of all other aspects of the artistic styles,
and accepting the obvious probabilities that individual artists may well
either have been working to instructions, or may have been copying
different styles from the original Byzantine manuscript, or may have
drawn galleys in more than one style on their own volition, our
correlations are as in Table 11.
Analysis of the styles of the depications of galleys suggests that
some modifications to the conclusions of Grabar and Manoussacas
[G-B] on the one hand and Tsamakda [T] on the other should be
considered. The preliminary conclusions to be drawn are that in the
Byzantine style the galleys in the style of Group One were drawn only
by artist G-BAa/TA1 except possibly for the stray illustration at folio
227r, which Grabar and Manoussacas attribute to Ab but Tsamakda to
Al. Almost certainly Tsamakda is correct. G-BAa/TA1 also drew the
two unique galleys in the styles of groups Three and Four as well.
This was Artist One. The two galley illustrations of Group Two are
unique to artist G-BAb/TA2. This was Artist Two. G-BBc/TA1 also
642 APPENDIX SEVEN
Table 11: Galley Group artists correlated to (1) Grabar and Manoussacas and
(2) Tsamakda
Galley Quires Folios Artists Artists Artists

Group according to according to according to
Grabar and Tsamakda Pryor and
Manoussacas Jeffreys
Byzantine 1 14v, 15r Aa A1 One

One 3 31v Aa A1
4 32r, 33v, 35v, 38r, 38v, Aa A1
39r, 39v, 40v
30 227r Ab A1
Byzantine 2 20v, 21r Ab A2 Two
Two
Byzantine 3 29v Aa A1 One
Three
Byzantine 4 34v Aa A1 One
Four
Byzantine 5 41r, 44r Bc A1 Three
Five
Western 13 110v Cd B1 Four
Six 14 111v Cd or Ce B1 Four
Western 13 110v Cd B1 Four
Seven 15 123v, 124r Ce or C B2 Five
17 129v, 130r, 132v, 134v Cd B1 Four
18 138v, 140r Cd B1 Four
19 146v C?? B3 Six
28 212r Cd or Ce B1 Four
Western 19 145r Cd B4 Seven
Eight
Western 19 146v C?? B3 Six
Nine
Western 19 146v, 147r, 147v C?? B3 Six
Ten (twice), 149v
Muslim 21 157r, 159r Dg B5 Eight
Eleven 22 167v, 168v Dg B5
29 219v, 222r, 224r, 225v, Dg B5
226r, 226v
supposedly drew the two galleys in the distinctive style of Group Five
as well and here it is much more likely that Grabar and Manoussacas
were correct and that this was a different artist rather than G-
BAa/TA1. We have called him Artist Three.
Of the galleys drawn in Western and Muslim styles, G-BCd/TB1
definitely drew galleys in the styles of both groups Six and Seven
because both are found in the illustration on fol. 110v. We have called
him Artist Four. Grabar and Manoussacas were unsure whether the
illustration on fol. 111v was by another artist but this appears unlikely.
Both Grabar and Manoussacas and also Tsamakda distinguish a

separate artist G-BCe (or C)/TB2 for the two illustrations in the style
of Group Seven on folios 123v and 124r of quire 15. From the style of
the galleys alone there is no real reason to do so; however, there is a
distinctive quality to the brushwork in this quire which suggests that it
was the case. We have called him Artist Five. The stray illustration at
folio 212r of galleys in the style of Group Seven attributed by Grabar
and Manoussacas to G-BCd or G-BCe and by Tsamakda to TB1 could
be attributed to either Four or Five on the style of the galleys alone;
however, the brushwork suggests Four rather than Five.
Quire 19 is the most complex of the manuscript from an artistic
point of view and both Grabar and Manoussacas and also Tsamakda
distinguished two different artists. Although the style of three of the
galleys in the illustration on folio 146v is the same as that of Group
Seven, the drawing style is different to that of artists Four and Five
and this illustration, which includes Group Nine, together with those
of Group Ten, should be attributed to a different Artist Six.
The unique illustration in the style of Group Eight is attributed by
Grabar and Manoussacas to G-BCd but they are almost certainly
incorrect and Tsamakda correct in attributing it to a unique artist TB4.
The illustration is so different in style to the others of Artist Four that
it seems impossible that it could be by him. We have called this one
Artist Seven.
The galleys in the Muslim style of Group Eleven were attributed to
a new artist by both Grabar and Manoussacas (G-BDg) and by
Tsamakda (TB5). Both are undoubtedly correct. We have called him
Artist Eight.
We conclude that there were three artists working in the Byzantine
style, four in the Western style, and one in the Muslim style. Artist
One drew galleys in three distinctive styles, suggesting that he was
faithfully copying illustrations which were already in different styles.
If it were not for other evidence suggesting that the quires in question
were illustrated by different artists, he might well be considered to
have been capable of producing groups Two and Five as well;
however, the other evidence is clear that they were produced by two
other artists, Two and Three. Artists Four, Five, and Six all drew at
least some of their galleys in very similar styles based on Group
Seven, suggesting that they came from the same social milieu. Artists
Seven and Eight appear to have come from different milieux.
We emphasize that these suggested modifications to the
conclusions of Grabar and Manoussacas and also Tsamakda apply
644 APPENDIX SEVEN
explicitly to the illustrations of galleys only. Whether they have any

implications for wider analysis of the art styles and artists of the rest
of the illustrations of the manuscript is for others to decide.
The three important illustrations of bireme galleys discussed in
Chapter Five, here numbers 21, 30, and 31, also confirm Wilson’s and
Tsamakda’s dating of the manuscript to the second half of the twelfth
century at the latest. First, they clearly show the same bireme oarage
system as that of the illustration at folio 119r of the Berne manuscript
of Peter of Eboli’s De rebus Siculis carmen,12 but in less detail. Since
number 30 above is in the same artistic style as that of the Peter of
Eboli illustration, it can reasonably be presumed to have come from
the same atelier as the latter, but almost certainly predated it.
Secondly, the styles of galley depictions of groups Six-Nine are
extremely similar to those of the manuscript of the Annales Ianuenses
of Genoa accompanying the entries for 1170, 1175, and 1191.13
Although no files of oars are shown in the Genoese miniatures, it may
be presumed that they also depicted the new bireme oarage system of
the Skylitze2s manuscript in which both files of oars were rowed from
above deck but only one through oar ports. Thus the evidence of the
illustrations of galleys in the Skylitze2s manuscript confirms the dating
of the manuscript to between ca 1160 and 1200.
------------------------------
12
See above pp. 426-9 & Figs 51-4.
13
See above pp. 424-6 & Fig. 50.
BIBLIOGRAPHY AND ABBREVIATIONS
(a) Manuscripts consulted in microfilm or photocopy
Florence, Biblioteca Medicea Laurenziana, MS. Laurentianus LVII-31.

Milan, Biblioteca Ambrosiana, MS. B 119-sup [Gr. 139].
Munich, Bayerische Staatsbibliothek, Cod. Monac. Gr. 452.
Rome, Biblioteca Apostolica Vaticana, MS. Vat. Gr. 1605.
Wolfenbüttel, Herzog August Bibliothek, Cod. Guelf., 96 Gud. lat.
(b) Manuscripts and books edited and translated in the Appendices
Appendix One Syrianos Magistros, Naumacivai Surianou' Magivstrou, Milan,

Biblioteca Ambrosiana, MS. B 119-sup [Gr. 139], fols
333r-338v.
Appendix Two Leo VI, Naumacika; Levonto" Basilevw": Milan, Biblioteca
Ambrosiana, MS. B 119-sup [Gr. 139], fols 323r-331v.
Appendix Three Anonymous, Naumacika; suntacqevnta para; Basileivo u
patrikivou kai; parakoimoumevnou: Milan, Biblioteca
Ambrosiana, MS. B 119-sup [Gr. 139], fols 339-42.
Appendix Four Constantine VII Porphyrogenne2tos, De cerimoniis aulae
Byzantinae: Leipzig, Universitätsbibliothek, MS. Rep. I.17
[Bibliothek Urb. 28], fols 220r-227r.
Appendix Five Nike2phoros Ouranos, Peri; qalassomaciva": Munich,
Bayerische Staatsbibliothek, Cod. Monac. Gr. 452, fols
82r-89v.
Appendix Eight Ibn Mankalı3, Muh5ammad, Al-Adilla al-Rasmiyyah fı3 ’l-Ta‘a3bı 3
al-H4arbiyya, ed. M. S. Khat6t6a2b (Bagdhad, 1988).
---------------, Al-Ahka2m al-Mulu2kiyya wa2 ’l-dawa2bit al-
Na2musiyya fı3 fan al-kita2l fı3 al-bahr m‘a2 dirasah ‘an fan
al-kita2l al-bahari fı3 ‘asr sla2tin al-Mama2lyk, ed. ‘Abd al-
‘Aa2ziz Mahmou2d ‘Abd al-Da2im (Ph.D.; Cairo, 1974).
(c) Printed Primary Sources
Abel, Orphica Abel, E., ed., Orphica (Leipzig, 1885).

Abu2 ’l-Fı3d a, Mukhtas5ar Abu2 ’l-Fı3da, Isma2‘ı3l ibn (al-Afd5al) ‘Alı3, Mukhtas5ar ta’rikh
al bashar, selections Fr. trans. W. Mac Guckin de Slane
as Résumé de l’histoire des Croisades tiré des Annales
d'Abou ’l-Fedâ, in RHCHOr, vol. 1, 1-165.
Abu2 ’l-Mah5a2sin, Al- Abu2 ’l-Mah5a2sin Jama2l al-Dı3n Yu2suf ibn Taghribı3rdi, Al-
Nuju2m Nuju2m al-Za2hira fı3 Mulu2k Mis5r wa ’l Qa2hira, selections
Fr. trans C. A. C. Barbier de Meynard as Extraits du
Nodjoûm ez-zahire, in RHCHOr, vol. 3, 477-509.
Abu2 Sha2m a, Kita2b al- Abu2 Sha2m a, Shiha2b al-Dı3n Abu2 ’l-Qa2sim ‘Abd al-Rah5m a2n,
rawd5atayn Kita2b al-rawd5atayn fı3 Akhba2r al-Dawlatayn, selections
trans. C. A. C. Barbier de Meynard as Le livre des deux
jardins: histoire des deux règnes, celui de Nour ed-Dı3n
et celui de Salah ed-Dı3n, in RHCHOr, vol. 4, 3-522;
vol. 5, 1-206.
668 BIBLIOGRAPHY AND ABBREVIATIONS
Achilles Tatios, Achilles Tatios, Tw'n kata; Leuvkipphn kai; Kleitofw'nta, in
Leukippe2 and S. Gaselee, trans., Achilles Tatius, rev. ed. (London,
Kleitopho2n 1969).
Achmet, Oneirocri- Achmet, Achmetis Oneirocriticon, ed. F. Drexl (Leipzig,
ticon 1925). Trans. S. E. Oberhelman, The Oneirocriticon of
Achmet: a medieval Greek and Arabic treatise on the
interpretation of dreams (Lubbock, 1991).
Aeneas the Tactician, Aeneas the Tactician, Peri; tou' pw'" crh; poliorkoumevnou"
Poliorke2tikon ajntevcein, in The Illinois Greek Club, ed., Aeneas
Tacticus, Asclepiodotus, Onesander (London, 1923).
Æthelweard, Chronicle Æthelweard, The chronicle of Æthelweard, ed. A.
Campbell (London, 1962).
Agapios, Kita2b al- Agapios of Hierapolis, Kita2b al-‘Unwa2n, ed. & trans. A. A.
‘Unwa2n Vasiliev, in PO, V.4 (pp. 557-692), VII.4 (pp. 457-591),
VIII.3 (pp. 397-550).
Agathias, Historiae Agathias of Myrina, Agathiae Myrinaei historiarum libri
quinque, ed. R. Keydell (Berlin, 1967).
Agnellus, Liber ponti- Agnellus, Liber pontificalis ecclesiae Ravennatis, ed. O.
ficalis Holder Egger, in MGHScriptRerLang, 265-391.
Ailian the Tactician, Aijlianou' taktikh; qewriva, in H. Köchly and W. Rüstow,
Taktike2 theo2ria Griechische Kriegschriftsteller, 3 vols (Leipzig, 1853-
55); vol. 2, 199-471.
Akhba2r Majmu2‘a Anonymous, Akhba2r Majmu2‘a, ed. E. Lafuente y
Alcántara, Ajbar Machmuâ. (Coleccion dè tradiciones.)
Crónica anónima del siglo XI (1867; rpt, Madrid,
1984).
Akropolite2s, George, Akropolite2s, George, Georgii Acropolitae opera, ed. A.
Opera Heisenberg, 2 vols (Leipzig, 1903).
Al-‘Ayni, ‘Iqd al- Al-‘Ayni, Abu2 Muh5ammad Mah5mud ibn Ah5mad ibn Mu2sa2
Juma2n Badr al-Dı3n, ‘Iqd al-Juma2n fı3 ta’rı3kh ahl al-Zama2n,
fragment ed. and trans. E. Fagnan, in Giuffrida, Scritti
Michele Amari, vol. 2, 86-7.
Al-Bakrı3, Kita2b al- Al-Bakrı3, Abu2 ‘Ubayd ‘Abd Alla2h ibn ‘Abd al-‘Aziz, Kita 2b
mughrib al-mughrib fı3 dhikr bila2d Ifrı3qiya wa’l-Maghrib [part of
his al-Masa2lik wa’l-Mama2lik] ed. and trans. W. Mac
Guckin de Slane, Description de l’Afrique
septentrionale, rev. ed. (Paris, 1965).
Al-Bala2dhurı3, Kita2b Al-Bala2dhurı3, Abu2 ’l-‘Abba2s Ah5m ad ibn Yah5ya2 ibn Ja2bir,
Futu2h5 al-Bulda2n Kita2b Futu2h5 al-Bulda2n, vol. 1 trans. P. K. Hitti, vol. 2
trans. F. C. Murgotten, The origins of the Islamic state
(N.Y., 1916, 1924).
Al-Baydhaq, Histoire Al-Baydhaq, Abu2 Bakr ibn ‘Alı3 al-S4inha2jı3, [Arabic title of
des Almohades work unknown], trans. E. Lévi-Provençal as L’histoire
des Almohades d’Abu2 Bakr b. ‘Alı3 as5-S4anha2g°ı3,
surnommé al-Baidak5, in his Documents inédits
d’histoire Almohade: fragments manuscrits du “legajo”
1919 du fonds Arabe de l’Escurial (Paris, 1928), 75-
224.
Albert of Aachen, Albert of Aachen, Historia Hierosolymitana, in RHCHOcc,
Historia Hierosoly- vol. 4, 265-713.
mitana
Aliscans Aliscans, ed. C. Regnier (Paris, 1990).
Al-Kindı3, Governors Al-Kindı3, Abu2 ‘Umar Muh5ammad ibn Yu2suf al-Tujı3bı3, The
and judges governors and judges of Egypt, ed. R. Guest (Leiden,
1912).
Al-Ma2liki, Muh@tas5ir Al-Ma2liki, Abu2 Bakr ‘Abd Alla2h ibn Abı3 ‘Abd Alla2h
BIBLIOGRAPHY AND ABBREVIATIONS 669
Kita2b Riya2d5 Muh5ammad, Muh@tas5ir Kita2b Riya2d5 al nufu2s fı3 t6abaqa2t
‘ulama2’ Qayrawa2n wa-’Ifrı3qiya, trans. Amari in BA-S,
vol. 1, 294-324.
Al-Maqqarı3, Nafh5 al-t5ı3b Al-Maqqarı3, Shiha2b al-Dı3n Abu2 ’l-‘Abba2s Ah5m ad, Nafh5 al-
t5ı3b min ghus5n al-Andalus al-rat5ı3b wa-dhikr wazı3riha2
Lisa2n al-Dı3n ibn al-Khat5ı3b, trans. De Gayangos as
Nafhu-t-tíb min ghosni-l-Andalusi-r-rattíb wa táríkh
Lísanu-d-Dín ibni-l-Khattíb, in his Mohammedan
dynasties, vol. 1, 17-548, vol. 2, 1-544.
Al-Maqrı3zı3, Al-Muqaffa2 Al-Maqrı3zı3, Taqı3 al-Dı3n Abu2 ’l-‘Abba2s Ah5m ad ibn ‘Alı3,
Al-Muqaffa2, selection Fr. trans. E. Fagnan as
“Traduction de la biographie continue dans le
«Mok5affa» de Makrizi, in Giuffrida, Scritti Michele
Amari, vol. 2, 35-85.
Al-Maqrı3zı3, Sulu2k ---------------, Al-Sulu2k li-ma‘rifat al-mulu 2k, part. Fr. trans. E.
(Blochet) Blochet, “Histoire d’Égypte de Makrizi”, ROL, 6
(1898), 435-89; 8 (1900-1901), 165-212, 501-53; 9
(1902), 6-163, 466-530; 10 (1903-4), 248-371; 11
(1905-8), 192-239. [Cited for information not contained
in Broadhurst’s translation only.]
Al-Maqrı3zı3, Sulu2k ---------------, Al-Sulu2k li-ma‘rifat al-mulu2k, part. Eng. trans. R.
(Broadhurst) J. C. Broadhurst, A history of the Ayyu2bid sultans of
Egypt (Boston, 1980).
Al-Marra2kushı3, Al- Al-Marra2kushı3, ‘Abd al-Wa2h5id ibn ‘Alı3 al-Tamı3m ı3, Al-
Mu‘jib Mu‘jib fi Talkhı3s5 Akhbar al-Maghrib, trans. E. Fagnan,
Histoire des Almohades d’Abd el-Wâh’id Merrâkechi
(Algiers, 1893).
Al-Mas‘u2dı3, Muru2j Al-Mas‘u2dı3, Abu2 ’l H4asan ‘Alı3 ibn al-H4u sayn, Muru2j al-
Dhahab wa Ma‘a2din al-Jawhar, ed. and trans. C.
Barbier de Meynard, Macoudi: Les prairies d’or, 9 vols
(Paris, 1861-77).
Al-Mas‘u2dı3, Al-Tanbı3h ---------------, Kita2b al-Tanbı3h wa’l Ishra2f, précis by A.
Silvestre de Sacy, Le livre de l’indication et de
l’admonition ou l’indicateur et le moniteur in al-
Mas‘u2dı3, Muru2j, vol. 9, 301-62.
Al-Muqaddası3, Ah5san Al-Muqaddası3 (Al-Maqdisı3), Muh5ammad ibn Ah5mad,
Ah5san al-Taqa2sı3m fı3 Ma‘rifat al-Aqa2lı3m , 3rd ed., in
BGA, vol. 3.
Al-Nuwayrı3, Al-Ilma2m Al-Nuwayrı3, Muh5ammad ibn al-Qa2sim al-Iskandara2ni,
[attributed to] Kita2b al-Ilma2m fı3ma2 jarat bihi ’l-ah5ka2m
al-maqd5ı3ya fı3 wa2qi‘at al-Iskandariyya, selection It.
trans. Amari as Anonymous, Kita2b ’al ’ima2m, ecc., in
BA-S, Appendice, 26-41.
Al-Nuwayrı3, Niha2yat Al-Nuwayrı3, Shiha2b al-Dı3n Ah5m ad ibn ‘Abd al-Wahha2b,
‘al-Arab (Amari) Niha2yat al-arab fı3 funu2n al-adab, selections It. trans.
Amari as Dal Nihâyat ‘al ‘arib ecc., in BA-S, vol. 2,
110-60.
Al-Nuwayrı3, Niha2yat ---------------, Niha2yat al-arab fı3 funu2n al-adab, selections trans.
‘al-Arab (Caussin) J. J. A. Caussin in J. H. von Riedesel, Voyages en Sicile,
dans la Grande Grèce at au Levant (Paris, 1802), 400-
48.
Al-Nuwayri, Niha2yat ---------------, Niha2yat al-arab fı3 funu2n al-adab, selections trans.
‘al-Arab (De Slane) Mac Guckin de Slane in Ibn Khaldu2n, ‘Ibar (De Slane),
vol. 1, 313-447; vol. 2, 483-93.
Al-Qurt6ubı3, Kita2b al- Al-Qurt6ubı3, Abu2 Ja‘far ibn ‘Abd al-H4aqq al-Khazrajı3,
iktifa2 Kitab al-iktifa2’ fı3 akhbar al-khulafa2’, trans. De
Gayangos as Kita2bu-l-iktifa2 fı3 akhba2ri-l-kholafa2, in his
Mohammedan dynasties, vol. 1, Appendix D (pp. xlii-l).
Al-T4abarı3, Ta’rı3kh (De Al-T4abarı3, Abu2 Ja‘far Muh5ammad ibn Jarı3r, Ta’rı3kh al-
Goeje) rusul wa’l- mulu2k (History of the prophets and kings),
ed. M. J. de Goeje, Annales quos scripsit, 15 vols
(1879; rpt, Leiden, 1964-65).
Al-T4abarı3, Ta’rı3kh ---------------, Ta’rı3kh al-rusul wa’l-mulu2k (History of the
(Yar-Shater) prophets and kings), trans. E. Yar-Shater et al., The
History of al- T4abarı3, 38 vols (Albany, 1985 -); vol. 15
(The crisis of the early Caliphate), trans. R. S.
Humphreys; vol. 18 (Between civil wars: the Caliphate
of Mu‘a2wiyah), trans. M. G. Morony; vol. 22 (The
Marwa2nid restoration), trans E. K. Rowson; vol. 23
(The zenith of the Marwa2n id house), trans. M. Hinds;
vol. 24 (The Empire in transition), trans. D. S. Powers;
vol. 25 (The end of expansion), trans. K. Y.
Blankinship; vol. 26 (The waning of the Umayyad
Caliphate), trans. C. Hillenbrand; vol. 29 (Al-Mans5u 2r
and al-Mahdı3), trans. H. Kennedy; vol. 30 (The
‘Abba2sid Caliphate in equilibrium), trans. C. E.
Bosworth; vol. 32 (The reunification of the ‘Abba2sid
Caliphate), trans. C. E. Bosworth; vol. 34 (Incipient
decline), trans. J. L. Kraemer; vol. 37 (The ‘Abba2sid
recovery), trans. P. M. Fields; vol. 38 (The return of the
Caliphate to Bagdhad), trans. F. Rosenthal.
Al-Tanu2khı3, Nishwa2r Al-Tanu2khı3, al-Muh5assin ibn ‘Alı3, Abu2 ‘Alı3, Nishwa2r al-
muh5a2d5ara wa-akhba2r al-mudha2kara, ed A. Schalchy
[al-Sha2ljı3], The Table-Talk of a Mesopotamian judge, 8
vols (Beirut, 1971-3).
Al-T4artu2sı3, “Traité” Cahen, C., “Un traité d’armurerie composé pour Saladin”,
BEO, 12 (1947-8), 103[1]-146[44].
Al-Tija2nı3, Rih5la Al-Tija2nı3, Abu2 Muh5ammad ‘Abd Alla2h ibn Muh5ammad,
Rih5la, part. trans. A. Rousseau as “Voyage du Scheikh
et-Tidjani dans la Régence de Tunis pendant les années
706, 707 et 708 de l’Hégire (1306-1309)”, Journal
Asiatique, ser. 4, 20 (1852), 57-208; ser. 5, 1, tome 1
(1853), 101-68, 354-425.
Al-Ya‘qu2bı3, Al-Bulda2n Al-Ya‘q5u2b ı3, Ah5m ad ibn Abı3 Ya‘qu2b ibn Ja‘far ibn Wahb
ibn Wa2d5ih5, Kita2b al-Bulda2n, trans. G. Wiet, Les pays
(Cairo, 1937).
Amatus of Monte Cas- Amatus of Monte Cassino, L’ystoire de li Normant et la
sino, L’ystoire de li chronique de Robert Viscart, ed. J. J. Champollion-
Normant Figeac (Paris, 1835).
Ambroise, L’estoire de Ambroise, L’estoire de la guerre sainte: histoire en vers de
la guerre sainte la troisième croisade (1190-1192) par Ambroise, ed. G.
Paris (Paris, 1897).
Anastasius Bibliothe- Anastasius Bibliothecarius, Chronographia tripertita, in
carius, Chronogra- Theophane2s, Chronographia, vol. 2, 33-346.
phia
Anastasius Bibliothe- ---------------, Anastasii Bibliothecarii sedis apostolicae
carius, Collectanea Collectanea ad Joannem Diaconum, in PL, vol. 129,
coll. 557-742.
Andrew of Bergamo, Andrew of Bergamo, Historia, ed. G. Waitz, in
Historia MGHScriptRerLang, 220-30.
Anglo-Saxon chronicle The Anglo-Saxon chronicle, ed. & trans. D. Whitelock
(London, 1961).
Annales Barenses Anonymous, Annales Barenses, ed. G. Pertz, in MGHSS,
vol. 5, 51-6.
Annales Beneventani Anonymous, Annales Beneventani, ed. G. Pertz, in
MGHSS, vol. 3, 173-85.
Annales Bertiniani Prudentius of Troyes, Annalium Bertinianorum pars
secunda inde ab anno 835 usque ad annum 861;
Hincmar of Reims, Annalium Bertinianorum pars tertia
inde ab anno 861 ad annum 882; ed. G. Pertz, in
MGHSS, vol. 1, 429-54, 455-515.
Annales Ceccanenses Annales Ceccanenses, ed. G. H. Pertz in MGHSS, vol. 19,
275-302.
Annales regni Franco- Anonymous, Annales regni Francorum inde ab a. 741
rum usque ad a. 829 qui dicuntur Annales Laurissenses
maiores et Einhardi, rev. ed., ed. F. Kurze, in
MGHScriptRerGerm, recusi vol. 6, (1895; rpt, Hanover,
1950). Trans. B. W. Scholz and B. Rogers, Carolingian
chronicles (Ann Arbor, 1972), 37-125.
Anonymous, African Anonymous, The African war, in A. G. Way, trans.,
war Caesar: Alexandrian, African and Spanish wars
(London, 1955).
Anonymous, Breve Anonymous, Breve chronicon de rebus siculis a Roberti
chronicon Guiscardi temporibus inde ad annum 1250, ..., in J.-L.-
A. Huillard Bréholles, Historia diplomatica Friderici
secundi, 6 vols (1852-61; rpt, Turin, 1963), vol. 1, part
2, 887-908.
Anonymous, Hypothe- Anonymous, ÔUpoqevsei" ejk tw'n strathgikw'n pravxewn, in
seis Polyainos, Strate2ge2mata, 427-504.
Anonymous, Anonymous, Menologii Graecorum pars secunda, in PG,
Menologion vol. 117, coll. 185-331.
Anonymous chronicle of Anonymous, Ignoti civis Barensis sive Lupi protospatae
Bari chronicon, in RISS, vol. 5, 147-56.
Anonymous Rhetor, Anonymous Rhetor, “Tou' aujtou' ejpitavfio" lovgo" eij" to;n
Laudatio funebris ajoivdimon basileva ku'r Manouh;l, to;n ... pepoivto”, in
Manuelis impera- Regel, Fontes, vol. 1, fasc. 1, No XII (pp. 191-228).
toris
Apollo2nios Sophista, Apollo2nios Sophista, Apollonii Sophistae lexicon
Lexicon Homericum, ed. I. Bekker (Berlin, 1833).
Appian, Civil wars Appian, The civil wars, in H. White, trans., Appian’s
Roman history, 4 vols (London, 1912-13), vols. 3-4.
Appian, Mithridatic ---------------, The Mithridatic wars, in H. White, trans.,
wars Appian’s Roman history, 4 vols (London, 1912-13), vol.
2.
Aristotle, Me2chanika Aristotle, Me2chanika, in W. S. Hett, trans., Aristotle: Minor
works (London, 1936).
Arrian, Anabasis Arrian, Anabasis of Alexander, in P. A. Brunt, trans.,
Arrian, 2 vols (London, 1976-83).
Asklepio2dotos, Techne2 Asclepio2dotos, Tevcnh taktikhv, in The Illinois Greek Club,
taktike2 trans., Aeneas Tacticus, Asclepiodotus, Onesander
(London, 1923).
Asser, Life of King Asser, Asser’s Life of King Alfred together with the annals
Alfred of St Neots erroneously ascribed to Asser, ed. W. H.
Stevenson (Oxford, 1904).
Athe2n aios Me2chanikos, Athe2n aios Me2chanikos, Peri me2chane2mato2n, in Wescher,
Peri me2chane2mato2n, Poliorcetique, 3-40.
Athe2n aios of Naukra- Athe2n aios of Naukratis, Athenaeus: the Deipnosophists,
tis, Deipnosophis- trans. C. B. Gulick, 7 vols (London, 1927-41).
tae (Gulick)
Athe2n aios of Naukra- ---------------, Athenaei Naucratitae dipnosophistarum libri XV,
tis, Deipnosophis- ed. G. Kaibel, 3 vols (Leipzig, 1887-90).
tae (Kaibel)
Attaleiate2s, Michael, Attaleiate2s, Michael, Historia, eds W. Brunet de Presle and
Historia I. Bekker [CSHB, 34] (Bonn, 1853). Fr. trans. chs 1-34,
H. Grégoire, “Michel Attaliatès Histoire”, Byzantion, 28
(1958), 325-62. Fr. trans. pp. 151-66 in E. Janssens, “La
bataille de Mantzikert (1071) selon Michel Attaliate”,
Annuaire de l’Institut de philologie et d’histoire
orientales et slaves, 20 (1968-72), 291-304.
Auxilius, In defensio- Auxilii in defensionem sacrae ordinationis Papae Formosi,
nem in E. Dümmler, ed., Auxilius und Vulgarius. Quellen
und Forschungen zur Geschichte des Papsttums im
Anfange des zehnten Jahrhunderts (Leipzig, 1866), 59-
64.
Becker, “Arabische Becker, C. H., “Arabische Papyri des Aphroditofundes”,
Papyri” Zeitschrift für Assyriologie und verwandte Gebiete, 20
(1907), 68-104.
Becker, “Papyrusstu- ---------------, “Papyrusstudien”, Zeitschrift für Assyrio-logie
dien” und verwandte Gebiete, 22 (1909), 137-154.
Bede, Historia Bede, the Venerable, Historia ecclesiastica gentis
ecclesiastica Anglorum, trans. J. E. King, 2 vols (London, 1930).
Bekker, Scholia Bekker, I., ed., Scholia in Homeri Iliadem (Berlin, 1825).
Belgrano, Documenti Belgrano, L. T., ed., Documenti inediti riguardanti le due
inediti crociate di San Ludovico IX (Geneva, 1859).
Bell, Greek papyri. IV Bell, H. I., ed., Greek papyri in the British Museum.
Catalogue with texts. Vol. IV: the Aphrodito papyri
(London, 1910).
Bell, Greek papyri. V ---------------, Greek papyri in the British Museum. Catalogue
with texts. Vol. V (London, 1917).
Bernold of St Blasien, Bernold of St Blasien, Chronicon, ed. G. H. Pertz in
Chronicon MGHSS, vol. 5, 385-467.
Bevers saga Bevers saga, in B. Vilhjálmsson, ed., Riddarasögur, vol. 1
(Reykjavík, 1961), 283-398.
Blancandin Sweetser, F. P., ed., Blancandin et l’orgueilleuse d’amour:
Roman d'aventure du XIIIe siècle (Geneva and Paris,
1964).
Blockley, Fragmentary Blockley, R. C., The fragmentary classicising histori-ans of
classicising the Later Roman Empire: Eunapius, Olympio-dorus,
historians Priscus and Malchus. Vol. 2: text, translation and
historiographical notes (Liverpool, 1983).
Boeve de Haumtone Der Anglonormannische Boeve de Haumtone, ed. A.
Stimming (Halle, 1899).
Brunck, Analecta Brunck, R. F. P., Analecta veterum poetarum Graeco-rum,
3 vols (Strasbourg, 1772-6).
Bryennios, Nike2phoros, Bryennios, Nike2phoros, Hyle historias, ed. and Fr. trans., P.
Hyle historias Gautier, Nicéphore Bryennios histoire, (Brussels, 1975).
Bueve de Hantone Der festländische Bueve de Hantone. Fassung I, ed. A.
Stimming (Dresden, 1911).
Caesar, Civil wars Julius Caesar, The Civil wars, trans A. G. Peskett
(Cambridge, Mass., 1979).
Caesar, Gallic war ---------------, The Gallic war, trans. H. J. Edwards (Cambridge,
Mass., 1979).
Caffaro, Annales Caffaro, Annales Januenses, in L. T. Belgrano and C.
Januenses Imperiale, eds, Annali genovesi di Caffaro e de’ suoi
continuatori dal MXCIX al MCCXCIII, 5 vols [FStI]
(Rome and Genoa, 1890-1929), vol. 1, 5-75.
Caffaro, De liberatione ---------------, De liberatione civitatum Orientis liber, in
Belgrano and Imperiale, eds, Annali genovesi, vol. 1,
95-124.
Cassiodorus, Variae Cassiodorus, Variarum libri XII, ed. A. J. Fridh, in CCSL,
vol. 96.
Champollion-Figeac, Champollion-Figeac, J. J., Documents historiques inédits
Documents histor- tiré des collections manuscrites de la Bibliothèque
iques Royale et des archives ou des biblio-thèques des
départements, 4 vols (Paris, 1841-8).
Chanson de Guillaume Recherches sur la Chanson de Guillaume: études
accompagnées d’une édition, ed. J. Wathelet-Willem
(Paris, 1975).
Chanson de Roland Les textes de la Chanson de Roland, ed. R. Mortier, 8 vols
(Paris, 1940-44).
Charroi de Nimes Le Charroi de Nimes: Chanson de geste, eds R. Van Deyck
et al. (Saint-Aquilin-de-Pacy, 1970).
Chevalier au Cygne The Old French Crusade cycle. Volume II: Le Chevalier au
Cygne and La Fin d’Elias, ed. J. A. Nelson (Alabama,
1985).
Cho2niate2s, Michael, Ta Cho2niate2s, Michael, Micah;l Akominavtou tou' Cwniavtou ta;
So2zomena Sw/zovmena , ed. S. Lampros (1879-80; rpt, Athens, 1968).
Cho2niate2s, Nike2tas, Cho2niate2s, Nike2tas, Nicetae Choniatae historia, ed. I. A.
Historia van Dieten [CFHB, 11] (Berlin, 1975).
Chronica monasterii Hoffmann, H., ed., Die Chronik von Montecassino, in
Casinensis MGHSS, vol. 34.
Chronicle of Albelda Anonymous, Chronicon Albeldense, in PL, vol. 129, coll.
1123-46.
Chronicle of Alfonso III Die Chronik Alfons’ III: Untersuchung und kritische
Edition der vier Redaktionen, ed. J. Prelog (Frankfurt,
1980). Trans. Wolf, Conquerors and chroniclers, 159-
77.
Chronicle of Cambridge Anonymous, La cronaca Siculo-Saracena di Cam-bridge,
ed. G. Cozza-Luzi (Palermo, 1890).
Chronicle of Moissac Chronicle of Moissac (Chronicon Moissiacense), ed. G.
Pertz, in MGHSS, vol. 1, 280-313.
Chronicle of the Morea The chronicle of the Morea, To; croniko;n tou' Morevw", ed. J.
Schmitt (London, 1904).
Chronicle of the Tocco Anonymous, Cronaca dei Tocco di Cefalonia di anonimo,
ed. G. Schiro (Rome, 1975).
Chronicle of 741 Anonymous, Chronicle of 741 [Chronica Byzantia-
Arabica], ed. I. Gil, Corpus scriptorum Muzarabic-
orum, 2 vols (Madrid, 1973), vol. 1, 7-14.
Chronicle of 754 Anonymous, Chronicle of 754 (Isidore of Seville,
Continuatio Hispana a. DCCLIV), ed. T. Mommsen, in
MGHAA, vol. 11, 323-69. Trans. Wolf, Conquer-ors
and chroniclers, 111-58.
Chronicon Casinense Anonymous, Chronicon Casinense, ed. G. Pertz, in
MGHSS, vol. 3, 222-230.
Chronicon Paschale Chronicon Paschale, ed. L. Dindorf, 2 vols [CSHB, 9]
(Bonn, 1832). Trans. M. and M. Whitby, Chronicon
Paschale 284-628 AD (Liverpool, 1989).
Chronicon Salernita- Anonymous, Chronicon Salernitanum: a critical edition
num with studies on literary and historical sources and on
language, ed. U. Westerbergh (Stockholm, 1956).
Claudian, De bello Claudius Claudianus, De bello Gildonico, in MGHAA, vol.
Gildonico 10, 54-73.
Claudian, De consulatu ---------------, De consulatu Stilichonis, in MGHAA, vol. 10,
Stilichonis 189-233.
Claudian, De quarto ---------------, Panegyricus de quarto consulatu Honorii
consulatu Honorii Augusti, in MGHAA, vol. 10, 150-74.
Cligés Chrétien de Troyes, Les romans de Chrétien de Troyes. II:
Cligés, ed. A. Micha (Paris, 1968).
Codex Carolinus Codex Carolinus, ed. W. Gundlach, in MGHEp, vol. 3,
469-657.
Constantine VII, De Constantine VII Porphyrogenne2tos, Constantini Por-
cerimoniis phyrogeniti imperatoris de cerimoniis aulae Byzan-
tinae libri duo, ed. I. Reiske, 2 vols [CSHB, 5] (Bonn,
1829).
Constantine VII, De ---------------, De administrando imperio, vol. 1, ed. G.
administrando Moravcsik, trans. R. J. H. Jenkins, (Budapest, 1949); R.
imperio H. Jenkins, Volume II: commentary (London, 1962).
Constantine VII, De ---------------, Peri; tw'n qemavtwn, ed. A. Pertusi, Costantino
thematibus Porfirogenito de thematibus (Vatican City, 1952).
Constantine VII, Ex- ---------------, Excerpta historica iussu imperatoris Constantini
cerpta historica Porphyrogeniti confecta, 4 vols, ed. U. P. Boissevain, et
al. (Berlin, 1903-6).
Constantine VII, Prae- ---------------, Praecepta imperatori Romano bellum cogitanti ...
cepta observanda, in Constantine VII, De cerimoniis, vol. 1,
444-508.
Constantine VII, Three ---------------, Constantine Porphyrogenitus: three treatises on
treatises imperial military expeditions, ed. and trans. J. F. Haldon
(Vienna, 1990).
Consularia Italica Consularia Italica, ed. T. Mommsen, in MGHAA, vol. 9,
274-339.
Continuatio Sigeberti Gemblacensis Continuatio Praemonstratensis, ed.
Praemonstratensis D. L. C. Bethmann, in MGHSS, vol. 6, 447-56.
Continuation de La continuation de Guillaume de Tyr (1184-1197), ed. M.
Guillaume de Tyr R. Morgan (Paris, 1982). Trans. P. Edbury, “The Old
French continuation of William of Tyre, 1184-97”, in
his The conquest of Jerusalem and the Third Crusade:
sources in translation (Aldershot, 1998), 11-145.
Couronnement de Louis Le couronnement de Louis: Chanson de geste, ed. E.
Langlois (Paris, 1887).
Crisafulli and Nesbitt, Crisafulli, V. S. and J. W. Nesbitt, trans., The miracles of St
Miracles of St Artemios: a collection of miracle stories by an
Artemios anonymous author of seventh-century Byzantium
(Leiden, 1997).
Cronaca Capuana Anonymous, La cronaca della dinastia Capuana, ed. N.
Cilento, in Italia meridionale longobarda (Milan and
Naples, 1966), 122-34.
Dain, Naumachica Dain, A., ed., Naumachica partim adhuc inedita (Paris,
1943).
Dandolo, Andrea, Dandolo, Andrea, Chronica per extensum descripta, aa.
Chronica 46-1280 d. C., ed. E. Pastorello, in RISSNS, tome 12,
parte 1.
Daphnopate2s, Theo- Daphnopate2s, Theodore, Théodore Daphnopatès corres-
dore Correspond- pondance, eds and trans J. Darrouzès and L. G.
ance Westerink (Paris, 1978).
De Boislisle, “Projet de De Boislisle, A., “Projet de Croisade du premièr Duc de
Croisade” Bourbon (1316-33)”, Annuaire-bulletin de la Société de
l’histoire de France, 9 (1872), 230-6, 246-55.
De Chaumont, On De Chaumont, F. S. B., On ventilation and cubic space
ventilation (Edinburgh, 1867).
De Gayangos, Moham- De Gayangos, P., The history of the Mohammedan
medan dynasties dynasties in Spain; extracted from the Nafh5 al-t6ib min
ghus5n al-Andalus al-rat6ı3b by Ahmed ibn Mohammed al-
Makkarí, a native of Telemsán, 2 vols (1840-43; rpt,
London, 1964).
De Mas Latrie, Traités De Mas Latrie, L., Traités de paix et de commerce et
de paix documents divers concernant les relations des
Chrétiens avec les Arabes de l’Afrique septentrionale
au moyen-âge, 2 vols (1866; rpt, N.Y., n.d.).
Dennis, Three treatises Dennis, G. T., ed., Three Byzantine military treatises
(Washington, 1985).
Denys of Tell-Mah5ré, Denys of Tell-Mah5ré, Chronique de Denys de Tell-Mah5ré.
Chronique Quatrième partie, ed. and trans. J.-B. Chabot (Paris,
1895).
De obsidione toleranda Sullivan, D., “A Byzantine instructional manual on siege
defense: the De obsidione toleranda. Introduction,
English translation and annotations (with a reprint of
the Greek text edited by Hilda van den Berg)”, in J. W.
Nesbitt, ed., Byzantine authors: literary activities and
preoccupations. Texts and translations dedicated to the
memory of Nicolas Oikonomides (Leiden and Boston,
2003), 139-277. Original ed., De obsidione toleranda,
ed. H. van den Berg (Leiden, 1947).
De Smedt, “Acta De Smedt, C., et al., eds, “Acta graeca SS. Davidis,
graeca” Symeonis et Georgii Mitylenae in insula Lesbo”,
AnalBoll, 18 (1899), 209-259.
Dindorf, Scholia Dindorf, G., ed., Scholia Graeca in Homeri Odysseam
Graeca (1855; rpt, Amsterdam, 1962).
Dio Cassius, Roman Dio Cassius, Roman history, in E. Cary, trans., Dio’s
history Roman history, 9 vols (London, 1914-27).
Diodo2rus Siculus, Bib- Diodo2rus Siculus, Bibliothe2ke2 historike2, in C. H. Old-
liothe2ke2 historike2 father, et al., trans., Diodorus of Sicily, 12 vols
(London, 1933-67).
Dioskoride2s, De Dioskoride2s, Pedanius, Pedanii Dioscuridis Anazarbaei de
materia medica materia medica libri VIII ... nunc inter thesauros
(Pierpont Morgan) Pierpont Morgan bibliothecae asservatus (Paris, 1935).
Dioskoride2s, De ---------------, Codex Vindobonensis med. Gr. I der
materia medica Österreichischen Nationalbibliothek, 2 vols (Graz,
(Vienna) 1965-70).
Dioskoride2s, De ---------------, Pedanii Dioscuridis Anazarbaei de materia
materia medica medica libri quinque, ed. M. Wellmann, 3 vols (Berlin,
(Wellmann) 1907-14).
Doanidu, “ÔH paraiv- Doanidu, S., “ÔH paraivthsi" Nikolavou tou' Mouzavlwno"
thsi" Nikolavou tou' ajpo; th'" ajrciepiskoph'" Kuvprou. jAnevkdoton
Mouzavlwno"” ajpologhtiko;n poivh ma”, Hellenica, 7 (1934), 109-150.
Du Chesne, Historiae Du Chesne, C., Historiae Francorum scriptores, 5 vols
Francorum (Paris, 1636-49).
scriptores
Ekkehard of Aura, Ekkehard of Aura, Hierosolymita, in RHCHOcc, vol. 5, 1-
Hierosolymita 40.
Epistolae Austrasicae Epistolae Austrasicae, ed. W. Gundlach, in MGHEp, vol.
3, 110-53.
Epistolae ad res Orien- Epistolae ad res Orientales spectantes, eds E. Caspar and
tales spectantes G. Laehr, in MGHEp, vol. 7, 371-84.
Eracles L’estoire de Eracles empereur et la conqueste de la terre
d’Outremer, in RHCHOcc, vols 1-2.
Erbse, Scholia graeca Erbse, H., Scholia graeca in Homeri Iliadem, 7 vols
(Berlin, 1969-88).
Erchempert, Historia Erchempert, Historia Langobardorum Beneventanorum,
Langobardorum ed. G. Waitz, in MGHScriptRerLang, 231-64.
Ernoul Ernoul, Chronique d’Ernoul et de Bernard le Trésorier, ed.
L. de Mas Latrie (1871; rpt, Famagusta, 1974).
Eunapios, History Eunapios, E nova editione historiae Eunapii Sardiani qua
(Bekker) Dexippum continuavit Excerpta de legationibus gentium
ad Romanos, in I. Bekker and B. G. Niebuhr, eds,
Dexippi, Eunapii, Petri Patricii, Prisci, Malchi,
Menandri historiarum quae supersunt [CSHB, 7]
(Bonn, 1829), 41-118.
Eunapios, History ---------------, Fragmenta, in Blockley, Fragmentary
(Blockley) classicising historians, vol. 2, 6-150
Eunapios, History ---------------, Eunapii Fragmenta, in Eunapii Sardiani vitas
(Boissonade) sophistarum et fragmenta historiarum, ed. J. F.
Boissonade (Amsterdam, 1821), vol. 1, 452-545.
Euripides, Fabulae Euripides, Euripidis fabulae, ed. J. Diggle, 3 vols (Oxford,
1984-94).
Eustathios of Thessa- Eustathios of Thessalonike2, Eustathios of Thessaloniki: the
lonike2, Capture of capture of Thessaloniki, text of S. Kyriakidis (Palermo,
Thessaloniki 1961), trans. J. R. Melville Jones (Canberra, 1988).
Eustathios of Thessa- ---------------, Eustathii archiepiscopi Thessalonicensis
lonike2, Commenta- commentarii ad Homeri Odysseam, ed. J. G. Stallbaum,
rii ad Homeri Odys- 2 vols (Leipzig, 1825-6).
seam
Eustathios of Thessa- ---------------, “Tou' aujtou' lovgo" eij" to;n
aujtokravtora ku'r
lonike2, “Oratio ad Manouh;l to;n Komnhnovn, o{te h\n e[ti th'" aJgiwtavth" tw'n
Manuelem Muvrwn ejkklhsiva" uJpoyhvfio"”, in Regel, Fontes, vol. 1,
imperatorem [2]” fasc. 1, No III (pp. 24-57).
Eustathios of Thessa- ---------------, “Lovgo" rJhtoriko;" tou' aJgiwtavtou mhtropolivtou
lonike2, “Oratio ad Qessalonivkh" ku'r Eujstaqivo u, tou' kata; Flw'ron, ei;"
Manuelem to;n aujtokravtora ku'r Manouh;l to;n Komnhnovn”, in
imperatorem [4]” Regel, Fontes, vol. 1, fasc. 1, No VI (pp. 92-125).
Ex miraculis Sancti Ex miraculis Sancti Agrippini, ed. G. Waitz, in
Agrippini MGHScriptRerLang, 463-5.
Fath5 al-Andalus Anonymous, Fath5 al-Andalus, ed. and trans. J. de
Gonzalez, as Fatho-l-Andaluçi. Historia de la conquista
de España: codice Arábigo del siglo XII (Argel, 1889).
Ferretto, Codice diplo- Ferretto, A., ed., Codice diplomatico delle relazioni fra la
matico Liguria, la Toscana e la Lunigiana ai tempi di Dante
(1265-1321), parte prima, in Atti della Società Ligure di
storia patria, 31, fasc. 1 (1901).
Filangieri, Registri Filangieri, R., ed., I registri della cancelleria angioina, vols
1-33 (Naples, 1950-81).
Fin d’Elias The Old French Crusade cycle. Volume II: Le Chevalier au
Cygne and La Fin d’Elias, ed. J. A. Nelson (Alabama,
1985).
Fragmentum historiae Fragmentum Historiae monasterii-novi Pictaviensis
monasterii-novi auctore Martino monacho ejusdem monasterii, in E.
Pictaviensis Martene and U. Durand, Thesaurus novus
Anecdotorum, 5 vols (1717; rpt, Farnborough, 1968-9),
tome 3, coll. 1209-1220.
Fredegar, Continua- Fredegar, The fourth book of the chronicle of Fredegar with
tions its continuations, ed. and trans. J. M. Wallace-Hadrill
(London, 1960).
Fulcher of Chartres, Fulcher of Chartres, Fulcheri Carnotensis Historia
Historia Hierosoly- Hierosolymitana (1095-1127), ed. H. Hagenmeyer
mitana (Heidelberg, 1913). Trans. H. S. Fink and F. R. Ryan, A
history of the expedition to Jerusalem (1095-1127)
(N.Y., 1973).
Galen, De usu partium Galen, Galhnou' periv creiva" movriwn izV/Galeni de usu
partium libri XVII, ed. G. Helmreich, 2 vols (Leipzig,
1907, 1909). Trans. M. T. May, Galen on the usefulness
of the parts of the body: Peri; creiva" morivwn/de usu
partium (Ithaca, 1968).
Genesios, Basileiai Genesios, Basileivwn tovmoi D, in Iosephi Genesii regum
libri quattuor, ed. A. Lesmueller-Werner and I. Thurn
[CFHB, 14] (Berlin and N.Y., 1978). Trans. A.
Kaldellis, Genesios: on the reigns of the emperors
(Canberra, 1998).
Genoa, Codice diplo- Genoa, Codice diplomatico della repubblica di Genova, ed.
matico C. Imperiale di Sant’ Angelo, 3 vols [FStI, nos. 77, 79,
89] (Rome, 1936-42).
George Hamartolos, George Hamartolos, Croniko;n suvntomon, ed. E. de Muralto,
Chronikon syntomon in PG, vol. 110, coll. 41-1286.
George of Pisidia, George of Pisidia, Eij" th;n genomevnhn e[fodon tw'n barbavrwn
Bellum Avaricum kai; eij" th;n aujt w'n ajstociva n/Bellum Avaricum, in A.
Pertusi, ed., Giorgio di Pisidia poemi. I: Panegirici
epici (Ettal, 1959), 176-224.
Gesta abbatum Fonta- Anonymous, Gesta abbatum Fontanellensium, ed. G. Pertz,
nellensium in MGHSS, vol. 2, 270-301.
Gesta episcoporum Anonymous, Gesta episcoporum Neapolitanorum. Pars I,
Neapolitanorum ed. G. Waitz, in MGHScriptRerLang, 398-424.
Gesta triumphalia per Gesta triumphalia per Pisanos facta de captione
Pisanos facta Hierusalem et civitatis Maioricarum et aliarum
civitatum et de triumpho habito contra Ianuenses, in
Bernardo Maragone, Annales Pisani, 87-96.
Glaber, Ralph, Glaber, Rodolfus, Historiarum libri quinque, ed. and trans.
Historiae J. France (Oxford, 1989).
Goetz, Glossarii Latini Goetz, G., ed., Corpus glossariorum Latinorum, 7 vols
(Leipzig and Berlin, 1888-1923).
Greek New Testament The interlinear Greek-English New Testament. The Nestle
Greek text with a literal English translation, ed. A.
Marshall (London, 1958).
Greene, Scholia Plato- Greene, G. C., ed., Scholia Platonica (Haverford, 1938).
nica
Gregory I, Epistulae Gregory I, Pope, S. Gregorii Magni registrum epistula-
rum, ed. D. Norberg, 2 vols, in CCSL, vols 140, 140A.
Gregory of Nazianzos, Gregory of Nazianzos, Tou' ejn aJgivoi" patro;" hJmw'n Grh-
Logoi gorivou tou' Qeolovgou ajrciepiskovpou Kwnstanti-
noupovlew" lovgoi, in PG, vol. 36, coll. 213-38.
Gregory of Tours, Gregory of Tours, Historiarum libri X, ed. B. Krusch, 2nd
Historiae ed., in MGHScriptRerMer, vol. 1, part 1, 2 fasc. Trans
L. Thorpe, Gregory of Tours, The history of the Franks
(Harmondsworth, 1974).
Grenfell, Oxyrhynchus Grenfell, B. P., et al., eds, The Oxyrhynchus papyri. Part
papyri XVI (London, 1924).
Grettis saga Grettis saga Ásmundarsonar, in Gu∂∂ ni Jónsson, ed.,
Íslenzk Fornrit, vol. 7 (1936; rpt, Reykjavík, 1964), 1-
290.
Grumel, “Lettre du Pape Grumel, V., “La lettre du Pape Étienne V à l’empereur
Étienne V” Basile Ier”, REB, 11 (1953), 129-55.
Hagenmeyer, Epistulae Hagenmeyer, H., ed., Epistulae et chartae ad historiam
primi belli sacri spectantes quae supersunt aevo
aequales ac genuinae: Die Kreuzzugsbriefe aus den
Jahren 1088-1100 (1901; rpt, Hildesheim, 1973).
Haldon, “Theory and Haldon, J., “Theory and practice in tenth-century military
practice” administration: chapters II, 44 and 45 of the Book of
Ceremonies”, TM, 13 (2000), 201-352.
Heliodo2ros, Aithiopika Heliodo2ros of Emesa, Les Éthiopiques (Théagène et
Chariclée), ed. R. M. Rattenbury and T. W. Lumb,
trans. J. Maillon, 3 vols (Paris, 1935-43).
Herodotos, Histories Herodotos, Histories, in A. D. Godley, trans., Herodotus, 4
vols (London, 1920-25).
He2ro2n, Parangelmata Anonymous, Paraggevlmata poliorkhtikav, ed. and trans.
poliorke2tika D. Sullivan, Siegecraft: two tenth-century instructional
manuals by “Heron of Byzantium” (Washington, 2000),
26-113.
He2ro2n of Alexandria, He2ro2n of Alexandria, ÔHrw'no" Alexandrevw" pneumatikw'n
Pneumatika AV BV, ed. W. Schmidt, Heronis Alexandrini opera quae
supersunt omnia, vol. 1 (Leipzig, 1899), 1-333.
Hesiod Hesiod: the Homeric hymns and Homerica, trans. H. G.
Evelyn-White, (Cambridge, Mass., 1959).
Hesychios, Lexicon Hesychios of Alexandria, Hesychii Alexandrini lexicon, ed.
(Latte) K. Latte (Copenhagen, 1953-).
Hesychios, Lexicon ---------------, Hesychii Alexandrini lexicon, ed. M. Schmidt, 4
(Schmidt) vols (Jena, 1858-62).
Historia ducum Historia ducum Veneticorum, ed. H. Simonsfeld, in MGH
Veneticorum SS, vol. 14, 72-89.
Historia miscella Historia miscella ab incerto auctore consarcinata,
complectans Eutropii Historiam, quam Paulus
Diaconus ... ad tempora Justiniani deduxit, ed L.
Muratori, in PL, vol. 95, 739-1144.
Historia pseudo- Anonymous, Historia pseudoIsidoriana cod. Parisini 6113,
Isidoriana ed. T. Mommsen, in MGHAA, vol. 11, 378-88.
Homer, Iliad Homer, The Iliad, trans. A. T. Murray, 2 vols (London,
1924-25).
Homer, Odyssey ---------------, The Odyssey, trans. A. T. Murray, rev. G. E.
Dimock (Harvard, 1995).
Hude, Scholia Hude, C., ed., Scholia in Thucydidem ad optimos codices
collata (Leipzig, 1927).
Huygens, Peregrinati- Huygens, R. B. C., ed., Peregrinationes tres: Saewulf, John
ones tres of Würzburg, Theodoricus, in CCContMed, vol. 139.
Hydatius, Chronicle Hydatius, The Chronicle of Hydatius and the Consularia
Constantinopolitana: two contemporary accounts of the
final years of the Roman Empire, ed. & trans. R. W.
Burgess (Oxford, 1993).
Ibn ‘Abd al-H4akam, Ibn ‘Abd al-H4akam, ‘Abd al-Rah5m a2n, Futu2h5 Ifrı3qiya wa’l-
Afrique Andalus, ed. and trans. A. Gateau, Conquête de
l’Afrique du Nord et de l’Espagne (Algiers, 1947).
Ibn ‘Abd al-H4akam, ---------------, Ibn Abd-el-Hakem’s History of the conquest of
Spain Spain, ed. and trans. J. H. Jones (Göttingen, 1858).
Ibn abı3 Zar‘, Rawd5 al- Ibn abı3 Zar‘, Abu2 ‘l-‘Abba2s Ah5m ad al Fa2sı3, Kita2b al-Anı3s
Qirt6a2s al-mut6rib bi-rawd5 al-qirt6a2s fı3 akhbar mulu2k al-Maghrib
wa-ta’rı3kh madı3nat Fa2s, trans. A. Baumier, Roudh el-
Kartas: histoire des souverains du Maghreb (Espagne
et Maroc) et annales de la ville de Fès (Paris, 1860).
Ibn al-‘Adı3m , Zubdat Ibn al-‘Adı3m , Kama2l al-Dı3n Abu2 ’l-Qa2sim ‘Umar, Zubdat
al-h5alab (RHCHOr) al-h5alab fı3 ta’rı3kh H4alab, extracts trans. C. A. C.
Barbier de Meynard, in RHCHOr, vol. 3, 571-690.
Ibn al-‘Adı3m , Zubdat ---------------, Zubdat al-h5alab fı3 ta’rı3kh H4alab, extracts trans.
al-h5alab (Blochet) E. Blochet, in ROL, 3 (1895), 509-65, 4 (1896), 145-
320, 5 (1897), 37-107, 6 (1898), 1-49.
Ibn al-Athı3r, Al-Ka2m il Ibn al-Athı3r, ‘Izz al-Dı3n Abu2 ’l-H4asan ‘Alı3, Al-Ka2mil fı3 ’l-
(Amari) ta’rı3kh, selections ed. and trans. Amari in BA-S, vol. 1,
353-507.
Ibn al-Athı3r, Al-Ka2m il ---------------, Al-Ka2mil fı3 ’l-ta’rı3kh, selections trans. E. Fagnan,
(Fagnan) Annales du Maghreb et de l’Espagne (Algiers, 1898).
Ibn al-Athı3r, Al-Ka2m il ---------------, Al-Ka2m il fı3 ’l-ta’rı3kh, selections trans. as Extrait
(RHCHOr) de la chronique intitulée Kamel-Altevarykh par Ibn-
Alatyr, in RHCHor, vol. 1, 189-744; vol. 2. part 1, 1-
180.
Ibn al-Athı3r, Al-Ka2m il ---------------, Al-Ka2mil fı3 ’l-ta’rı3kh, ed. C. J. Tornberg, 13 vols,
(Tornberg) (1851-76; rpt, Beirut, 1965-66).
Ibn al-Athı3r, Al-Ka2m il ---------------, Al-Ka 2mil fı3 ’l-ta’rı3kh, ed. Tornberg, selections
(Vasiliev) trans. in Vasiliev, Byzance et les Arabes. Tome I, 349-
69.
Ibn al-Athı3r, Al-Ka2m il ---------------, Al-Ka 2mil fi ’l-ta’rı3kh, ed. Tornberg, selections
(Vasiliev/Canard) trans. in Vasiliev/Canard, Byzance et les Arabes. Tome
II, part 2, 128-62.
Ibn al-Khat6ı3b, A‘ma2l al- Ibn al-Khat6ı3b, Abu2 ‘Abd Alla2h Muh5ammad, A‘ma2l al-
a‘lam a‘lam fi-man bu2yi‘a k5abl al-ih5tila2m min mulu2k al-Islam,
part. trans. H. H. Abul-Wahab, “Contribution à
l’histoire de l’Afrique du Nord et de la Sicile: extrait du
‘A‘ma2l al-a‘la2m’ ‘Ibn al-H@at6îb”, in Giuffrida, Scritti
Michele Amari, vol. 2, 427-94.
Ibn H4awqal, S5u2rat al- Ibn H4awqal, Abu2 ’l-Qa2sim ibn ‘Alı3 al-Nas5ı3b, Kita2b S5u 2rat
Ard5 al-Ard5, 3rd ed., ed. J. H. Kramers in BGA, vol. 2.
Ibn ‘Idha2rı3, Al-baya2n Ibn ‘Idha2rı3, al-Marra2kushı3, Al-baya2n al-mughrib fı3 akhba2r
al-mughrib al-Maghrib, trans. E. Fagnan, Histoire de l’Afrique et
de l’Espagne intitulée al-Bayano’l-Moghrib, 2 vols
(Algiers, 1901-4).
Ibn Iya2s, Histoire des Ibn Iya2s, Abu2 ’l-Baraka2t Muh5ammad ibn Ah5m ad, Bada2’i‘
Mamlouks al-zuhu2r fı3 waqa2’i‘ al-duhu2r, trans. G. Wiet, Histoire
des Mamlouks Circassiens. Tome II (872-906), (Cairo,
1945).
Ibn Khaldun, ‘Ibar Ibn Khaldu2n, Walı3 al-Dı3n ‘Abd al-Rah5ma2n ibn
(Amari) Muh5ammad, Kita2b al-‘Ibar wa-dı3wa2n al-mubtada’ wa
’l-khabar, selections trans. Amari in BA-S, vol. 2, 163-
243
Ibn Khaldu2n, ‘Ibar ---------------, Kita2b al-‘Ibar wa-dı3wa2n al-mubtada’ wa ’l-
(Bu2la2q) khabar, 7 vols (Bu2laq, 1867-8).
Ibn Khaldu2n, ‘Ibar (De ---------------, Kita2b al-‘Ibar wa-dı3wa2n al-mubtada’ wa ’l-
Slane) khabar, selections trans. W. Mac Guckin de Slane,
Histoire des Berbères et des dynasties musulmanes de
l’Afrique septentrionale, new ed., 4 vols (1925, 1956;
rpt, Paris, 1978).
Ibn Khaldu2n, ‘Ibar (Des ---------------, Kita2b al-‘Ibar wa-dı3wa2n al-mubtada’ wa ’l-
Vergers) khabar, selections trans. A. N. des Vergers, Histoire de
l’Afrique sous la dynastie des Aghlabites (184/800-
297/909) et de la Sicile sous la domination Musulmane
(Paris, 1841).
Ibn Khaldu2n, Muqqa- ---------------, The Muqqadimah: an introduction to history, 3
dimah vols, trans. F. Rosenthal (Princeton, 1958).
Ibn Muyassar, Akhba2r Ibn Muyassar, Ta2j al-Dı3n Muh5ammad ibn Yu2suf, Akhba2r
Mis5r Mis5r, selection Fr. trans. C A. C. Barbier de Meynard in
RHCHOr, vol. 3, 461-73.
Ibn al-Qala2nisı3, Dhayl Ibn al-Qala2nisı3, Abu2 Ya‘la2 H4amza ibn Asad, Dhayl ta’rı3kh
ta’rı3kh Dimashq Dimashq, extracts trans. H. A. R. Gibb as The
Damascus chronicle of the Crusades extracted and
translated from the chronicle of Ibn al-Qala2nisı3
(London, 1932)
Ibn Sabba2t6, Dı3wa2n S4ilat Ibn Sabba2t6, Abu2 ‘Abd Alla2h Muh5ammad ibn Abı3 ’l-H4asan
al-simt6 ‘Alı3 ibn al Mis5rı3, al-Tu2zarı3, Dı3wa2n S4ilat al-simt6,
selections trans Amari in BA-S, vol. 1, 345-52.
Ibn Sa2h5ib, Al-Mann bi Ibn Sa2h5ib al-S4ala2t, Abu2 Marwa2n ‘Abd al-Malik, Al-Mann
’l-ima2ma bi ’l-ima2ma ‘ala ’l-mustad5‘afı3n bi-an ja‘alahum Alla2h
al-a’imma wa-ja‘alahum al-wa2rithin, trans. A. Huici
Miranda as Al-Mann bil-ima2ma (Valencia, 1969).
Ibn Shadda2d, Al-Nawa- Ibn Shadda2d, Baha2’ al-Dı3n Abu2 ’l-Mah5a2sin, Al-Nawa2dir al-
2dir al-sult6a2niyya sult6a2niyya wa ’l-mah5a2sin al-Yu2sufiyya, trans. D. S.
Richards, The rare and excellent history of Saladin or
al-Nawa2dir al-sult6a2niyya wa’l-Mah5a2sin al-Yu2sufiyya by
Baha2’ al-Dı3n Ibn Shadda2d (Aldershot, 2001).
‘Ima2d al-Dı3n, Al-fath5 ‘Ima2d al-Dı3n Muh5ammad ibn Muh5ammad al-Ka2tib al-
al-qussı3 Is5faha2nı3, Al-fath5 al-qussı3 fı3 ’l-fath5 al-qudsı3, trans. H.
Massé, Conquête de la Syrie et de la Palestine par
Saladin (al-Fath5 al-qussî fî l-fath5 al-qudsî) [Documents
relatifs a l’histoire des Croisades, 10] (Paris, 1972).
Isidore of Seville, Isidore of Seville, Isidori Hispalensis episcopi
Etymologiae etymologiarum sive originum libri XX, ed. W. M.
Lindsay, 2 vols (Oxford, 1911).
Isidore of Seville, ---------------, Historia Gothorum Wandalorum Sueborum, ed.
Historia T. Mommsen, in MGHAA, vol. 11, 267-303. Partial
trans. Wolf, Conquerors and chroniclers, 81-110.
Itinerarium peregrino- Anonymous, Das Itinerarium peregrinorum: eine
rum (Mayer) zeitgenössische englische Chronik zum dritten Kreuzzug
in ursprünglicher Gestalt, ed. H. E. Mayer (Stuttgart,
1962).
Itinerarium peregrino- Chronicles and memorials of the reign of Richard I. Vol. I:
rum (Stubbs) Itinerarium peregrinorum et gesta regis Ricardi, ed. W.
Stubbs, Rolls Series, tome 38, vol. 1. Trans. J. J.
Nicholson, Chronicle of the Third Crusade: a
translation of the Itinerarium Peregrinorum et Gesta
regis Ricardi (Aldershot, 1997).
James of Vitry, Historia Jacobus de Vitriaco, Historia Hierosolimitana, in J.
Hieroso-limitana Bongars, ed., Gesta Dei per Francos, 2 vols (Hanau,
1611), vol. 1, 1047-1145.
Jerome, St, Epistolae Jerome, St, Sancti Eusebii Hieronymi Stridonensis
presbyteri epistolae, in PL, vol. 22, 325-1224.
John VIII, Register John VIII, Pope, Das Register Papst Johannes VIII, 872-
882, ed. D. Lohrmann (Tubingen, 1968).
John VIII, Registrum ---------------, Registrum Iohannis VIII Papae, ed. E. Caspar, in
MGHEp, vol. 7, 1-272.
John VIII, Fragmenta ---------------, Fragmenta registri Iohannis VIII Papae ind. VI-
IX (A. 872-876), ed. E. Caspar, in MGHEp, vol. 7, 273-
312.
John Lydos, On powers John Lydus, Ioannes Lydus on powers or the magistracies
of the Roman state, ed. and trans. A. C. Bandy
(Philadelphia, 1983).
John of Naples, Gesta John of Naples, Gesta episcoporum Neapolitanorum pars
altera auctore Iohanne Diacono, in MGH-
ScriptRerLang, 424-35.
John the Deacon, Cro- John the Deacon, Cronaca Veneziana, ed. G. Monticolo, in
naca Veneziana Chronache Veneziane antichissime [FStI, 9], (Rome,
1890), 57-171.
Joinville, Vie de saint Jean de Joinville, Vie de saint Louis, ed. J. Monfrin (Paris,
Louis 1995).
Jordanes, Romana Jordanes, De summa temporum vel origine actibusque
gentis Romanorum, ed. T. Mommsen, in Iordanis
Romana et Getica, MGHAA, vol. 5, part 1, 1-52.
Jordanes, Getica ---------------, De origine actibusque Getarum, ed. T.
Mommsen, in Iordanis Romana et Getica, MGHAA,
vol. 5, part 1, 53-138. Trans. C. C. Mierow, The Gothic
history of Jordanes, 2nd ed. (Princeton, 1915).
Julian [emperor], Julian [emperor], Orations, in W. C. Wright, trans., The
Orations works of the emperor Julian, 3 vols (London, 1913-23).
Kaminiate2s, John, De Kaminiate2s, John Ioannis Caminiatae de expugnatione
expugnatione Thessalonicae, ed. G. Böhlig (Berlin, 1973). Trans. D.
Thessalonicae Frendo and A. Fotiou, The capture of Thessaloniki
(Perth, 2000).
Kedre2nos, George Kedre2nos, George, Suvnoyi" iJstoriw'n [Compendium histo-
Synopsis historio2n riarum], ed. I Bekker, Georgius Cedrenus Ioannis
Scylitzae ope, 2 vols [CSHB, 23] (Bonn, 1838-9).
Kekaumenos, Strate2g i- Kekaumenos, Strategikon, ed. B. Wassiliewsky and V.
kon (Wassiliewsky) Jernstedt, Cecaumeni strategicon et incerti scriptoris de
officiis regiis libellus (St Petersburg, 1896).
Kekaumenos, Strate2g i- Kekaumenos, Strategikon, ed. and It. Trans., M. D.
kon (Spadaro) Spadaro, Raccomandazioni e consigli di un galantuomo
(Strathgikovn) (Alessandria, 1998).
Khludov psalter [Schepkina], å. Ç. [V.],

[Miniatures of the Khludov
psalter] (Moscow, 1977).
Kiessling, Sammelbuch Kiessling, E., Sammelbuch Griechischer Urkunden aus
Ägypten, vol. 8 (Wiesbaden, 1965).
Kinnamos, John, Kinnamos, John, ÔIstoriw'n bibliva ıV, ed. A. Meineke,
Historiae Ioannis Cinnami epitome rerum ab Ioanne et Alexio
Comnenus gestarum [CSHB, 15] (Bonn, 1836). Trans.
C. M. Brand, Deeds of John and Manuel Comnenus
(N.Y., 1976).
Kita2b al-‘uyu2n Anonymous, Kita2b al-‘uyu2n wa’l-h5ada2’iq fı3 ah@ba2r al-
h5aqa2’iq, trans. Vasiliev/Canard, Byzance et les Arabes.
Tome II, part 2, 220-25.
Kolybas, Sergios, Kolybas, Sergios, “Lovgo" tou' megalopravttonto" prwtonw-
“Oratio ad Isaacium taivou basilikou' grammatikou' ku'r Sergivo u tou' Koluba'
Angelum imperato- ei;" to;n aujtokravtora ku'r Isaavkion to;n “Aggelon”, in
rem” Regel, Fontes, vol. 1, fasc. 2, No XVI (pp. 280-91).
Komne2ne2, Anna Alexi- Komne2ne2, Anna, Alexiade, ed. and trans. B. Leib, 3 vols
ade (Paris, 1937-45).
Lambros, “Triva keivme- Lambros, S., “Triva keivmena sumbavllonta eij" th;n iJstorivan
na sumbavllonta” tou' nautikou' para; toi'" Buzantivnoi"”, Nevo"
ÔEllhnomnhvmwn, 9 (1912), 162-77.
Leo III, Epistolae Leo III, Pope, Epistolae X, ed. K. Hampe, in MGHEp, vol
5, 85-104.
Leo VI, Taktika (Vari) Leo VI, Leonis imperatoris tactica, ed. R. Vari, 2 vols
(Budapest, 1917-22).
Leo VI, Taktika (PG) ---------------, Leonis imperatoris tactica, ed. J. Meursius, in
PG, vol. 107, 669-1094.
Leo Grammatikos, Leo Grammatikos, Leonis Grammatici chronographia, ed.
Chronographia I. Bekker [CSHB, 31] (Bonn, 1842).
Leo Marsicanus, Leo Marsicanus [Leo of Ostia], Chronica monasterii
Chronica Casinensis, ed. W. Wattenbach, in MGHSS, vol. 7, 551-
727.
Leo the Deacon, Leo the Deacon, Leonis diaconi Caloensis historiae libri
Historiae decem, ed. C. B. Hase [CSHB, 3] (Bonn, 1828).
Liber Maiolichinus Liber Maiolichinus de gestis Pisanorum illustribus, ed. C.
Calisse [FstI, 29], (Rome, 1904).
Liber Pontificalis Le Liber Pontificalis, ed. L. Duchesne, 2nd ed., 3 vols
(Paris, 1955-7).
Liudprand of Cremona, Liudprand of Cremona, Antapodosis, in Die Werke
Antapodosis Liudprands von Cremona, ed. J. Becker, 3rd ed., in
MGHScriptRerGerm, separatim editi, vol. 41, 1-158.
Trans. F. A. Wright, The works of Liudprand of
Cremona: Antapodosis, Liber de rebus gestis Ottonis,
Relatio de legatione Constantinopolitana (N.Y., 1930),
25-212.
Liudprand of Cremona, ---------------, Liudprandi relatio de legatione Constantino-
Relatio politana, ed. J. Becker, 3rd ed., in MGHScriptRerGerm,
separatim editi, vol. 41, 175-212. Trans. F. A. Wright,
The works of Liudprand of Cremona: Antapodosis,
Liber de rebus gestis Ottonis, Relatio de legatione
Constantinopolitana (N.Y., 1930), 235-77.
Livy, Ab urbe condita Livy, From the foundation of the city, in B. O. Foster, et al.,
trans., Livy, 14 vols (London, 1919-59).
Longus, Daphnis and Longus, Daphnis and Chloe, trans G. Thornley, rev. J. M.
Chloe Edmonds (London, 1916).
Louis II, Epistola Ludovici II. imperatoris epistola ad Basilium I.
imperatorem Constantinopolitanum missa, ed. W.
Henze, in MGHEp, vol. 7, 385-94.
Lucan, Civil war Lucan, The civil war, trans. J. D. Duff, Lucan (London,
1928).
Lucian Lucian, Lucian, ed. and trans. A. M. Harmon, 8 vols
(London, 1915).
Lupus Protospatharios, Lupus Protospatharios, Annales, ed. G. Pertz, in MGHSS,
Annales vol. 5, 51-63.
Mai, Homerus et Mai, A., Homeri et Virgili picturae antiquae (Rome, 1835).
Virgilius
Mai, Iliad ---------------, Iliadis fragmenta antiquissima cum picturis item
scholia vetera ad Odysseam (Milan, 1819).
Mai, Picturae ---------------, ed., Picturae antiquissimae bellum Iliacum
repraesentantes (Milan, 1819).
Malalas, John Malalas, John, Chronographia, ed. J. Thurn (Berlin, 2000).
Chronographia Trans. E. Jeffreys et al., The chronicle of John Malalas:
a translation (Melbourne, 1986).
Malaterra, Geoffrey, De Malaterra, Gaufredus, De rebus gestis Rogerii Calabriae et
rebus gestis Siciliae comitis et Roberti Guiscardi ducis fratris eius,
ed. E. Pontieri, in RISSNS, tome 5, pt. 1.
Malchos, Byzantiaka Malchos of Philadelphia, Byzantiaka, ed. and trans.,
Blockley, Fragmentary classicising historians, vol. 2,
402-70.
Mansi, Concilia Mansi, J. D., Sacrorum conciliorum nova et amplissima
collectio, tome XVI (Venice, 1771).
Maragone, Bernardo Maragone, Bernardo, Annales Pisani, ed. M. L. Gentile in
Annales Pisani RISSNS, tome 6, pt. 2, 1-74.
Marcellinus Comes, Marcellinus Comes, Chronicon, ed. T. Mommsen (Berlin,
Chronicon 1894). Trans. B. Croke, The chronicle of Marcellinus
(Sydney, 1995).
Marcus Graecus, Liber Marcus Graecus, Liber ignium ad comburendos hostes, in
ignium M. Berthelot, La chimie au moyen-âge. Tome I: Essai
sur la transmission de la science antique au moyen-âge
(Paris, 1893), 89-135.
Marius of Avenches, Marius of Avenches, Marii episcopi Aventicensis chronica,
Chronica ed. T. Mommsen, in MGHAA, vol. 11, 232-9.
Marsden, Technical Marsden, E. W., Greek and Roman artillery: technical
treatises treatises (Oxford, 1971).
Maurice, Ek tou Maurice, Ek tou' Maurikivou pw'" dei' diapleei'n tou;"
Maurikiou potamou;" kai; ta;" diabavsei" aujtw'n poiei'sqai ejcqrw'n
ajntikaqistamevnwn, in Dain, Naumachica, 39-42.
Maurice, Strate2gikon Maurice, Das Strategikon des Maurikios, eds G. T. Dennis
and E. Gamillscheg (Vienna, 1981). Trans. G. T.
Dennis, Maurice’s Strategikon: handbook of Byzantine
military strategy (Philadelphia, 1984).
Michael Rhetor, “Ora- Michael Rhetor, “Tou' aujtou' eij" to;n aujtokravtora ku'r
tio ad Manuelem Manouh;l to;n Komnhnovn, o{te h| prwtevkdiko"”, in Regel,
imperatorem [2]” Fontes, vol. 1, fasc. 1, No IX (pp. 152-65).
Michael the Syrian, Michael the Syrian, Chronique de Michel le Syrien,
Chronique Patriarche Jacobite d’Antioche (1166-1199), ed. and
trans. J.-B. Chabot, 4 vols (Paris, 1899-1910).
Miklosich and Müller, Miklosich, F. and J. Müller, Acta et diplomata Graeca
Acta et diplomata medii aevi, 6 vols (1860-90; rpt, Aalen, 1968).
Minieri-Riccio, Saggi Minieri-Riccio, C., Saggi di codice diplomatico formato
sulle antiche scritture dell’Archivio di Stato di Napoli.
Volume primo: che principia dal 15 agosto dell’ anno
964 e termina nel 27 ottobre 1285 (Naples, 1878).
Miracles of Saint Les plus anciens recueils des miracles de Saint Demetrius
Demetrius et la pénétration des Slaves dans les Balkans, ed. and
trans. P. Lemerle, 2 vols (Paris, 1979).
Morgan, Continuation Morgan, M. R. ed., La continuation de Guillaume de Tyr
(1184-1197) (Paris, 1982).
Moschos, John, Moschos, John, Tou' makarivou Iwavnnou tou' Eujkrata'
Spiritual meadow bivblo", hJ legomevnh Leimwvn, in PG, vol. 87, part III, coll.
2843-3116. Trans. J. Wortley, The spiritual meadow
(Pratum spirituale) (Kalamazoo, 1992).
Munitz, Letter of the Munitz, J. A., et al., eds, The letter of the three Patriarchs
three Patriarchs to emperor Theophilus and related texts (Camberley,
1997).
Neapolitanorum victo- Neapolitanorum victoria facta, ed. G. Waitz, in
ria facta MGHScriptRerLang, 465-6
Nicholas I, Epistolae Nicholas I, Pope, Nicolai I Papae epistolae, ed. E. Perels,
in MGHEp, vol. 6, 257-690.
Nicholas I, Letters Nicholas I, Patriarch, Nicholas I Patriarch of
Constantinople: letters, eds R. J. H. Jenkins and L. G.
Westerink [CFHB, 6] (Washington, 1973).
Nike2phoros I, Historia Nike2phoros I, Patriarch, ÔIstoriva suvntomo", trans. C.
syntomos Mango, Short history (Washington, 1990).
Nike2tas David Paphla- Nike2tas David Paphlagon, Vita S. Ignatii archiepiscopi
gon, Vita S. Ignatii Constantinopolitani, in PG, 105, coll. 487-574.
Nike2tas Magistros, Vita Nike2tas Magistros, Vita S. Theoctistae in Acta Sanctorum
S. Theoktistae Novembris, vol. 4 (Brussels, 1925), 224-33. Trans. A.
C. Hero, “The life of our blessed mother Theoktiste of
Lesbos who practiced asceticism and died on the island
named Paros”, in A.-M. Talbot, ed., Holy women of
Byzantium: ten saints’ lives in English translation
(Washington, 1996), 95-116.
Nonius Marcellus, De Nonius Marcellus, De compendiosa doctrina, ed. W. M.
compendiosa Lindsay, 3 vols (Leipzig, 1903).
doctrina
Odo of Deuil, De profe- Odo of Deuil, De profectione Ludovici VII in orientem, ed.
ctione Ludovici VII and trans. V. G. Berry (N.Y., 1948).
Ogier de Danemarche La chevalerie d’Ogier de Danemarche, ed. M. Eusebi
(Milan, 1963).
Oikonomides, Listes de Oikonomides, N., ed., Les listes de préséance byzantines
préséance des IXe et Xe siècles (Paris, 1972).
Oliver of Paderborn, Oliverus Scholasticus (Oliver of Paderborn), Historia
Historia Damiatina Damiatina, in H. Hoogeweg, ed., Die Schriften des
Kölner Domscholasters, späteren Bischofs von
Paderborn und Kardinal-Bischofs von S. Sabina
Oliverus [Bibliothek des litterarischen Vereins in
Stuttgart, 202] (Stuttgart, 1894), 159-280.
Olympio2doros, Books of Olympio2doros of Thebes, Books of history, ed. and trans. in
history Blockley, Fragmentary classicising historians, 152-
208.
Orkneyinga Saga Orkneyinga Saga, ed. F. Gu∂mundsson, in Íslenzk fornrit,
35 vols (Reykjavík, 1968-82), vol. 34, 1-300.
Orosius, Historiae Paulus Orosius, Pauli Orosii historiarum adversum
adversum paganos paganos libri VII, ed. K. Zangermeister (Leipzig, 1889).
Trans. R. J. Deferrari, Orosius, The seven books of
history against the pagans (Washington, 1964).
Ottobono Scriba, Ottobono Scriba, Annales Ianuenses, in L. T. Belgrano and
Annales Ianuenses C. Imperiale, eds, Annali genovesi di Caffaro e de’ suoi
continuatori dal MXCIX al MCCXCIII, 5 vols [FStI]
(Rome and Genoa, 1890-1929), vol. 2, 3-66.
Ouranos, Nike2phoros, Ouranos, Nike2phoros, Ek tw'n taktikw'n Nikhfovrou
Ek to2n taktiko2n Oujranou' kefavlaia riqV-rkgV, in Dain, Naumachika, 89-
104.
Ouranos, Nike2phoros, ---------------, Taktika, ed. J. Meursius, Constantini
Taktika Porphyrogennetae imp. Liber Tacticus terra marique
pugnantium ordinationem continens, in Ioannes Lamius
(G. Lami), ed., Ioannis Meursi opera omnia, 12 vols
(Florence, 1741-63), vol. 6, coll. 1209-1418. Chapters
56-65 trans. McGeer, Dragon’s teeth, 89-163.
Parastaseis syntomoi Parastavsei" suvntomoi cronikaiv, ed. T. Preger in
chronikai Scriptores originum Constantinopolitanarum, 2 vols
(1901-7; rpt, N.Y., 1975), vol. 1, 19-73. Trans. A.
Cameron and J. Herrin, Constantinople in the early
eighth century: the Parastaseis syntomoi chronikai
(Leiden, 1984).
Patria Ko2n stantinou- Pavtria Kwnstantinoupolevw" III, ed. T. Preger in
poleo2s III Scriptores originum Constantinopolitanarum, 2 vols
(1901-7; rpt, N.Y., 1975), vol. 1, 214-83.
Paul the Deacon, Paul the Deacon, Pauli historia Langobardorum, ed. L.
Historia Lango- Bethmann and G. Waitz, in MGHScriptRerLang, 45-
bardorum 187.
Peeters, “Vie grecque” Peeters, P., ed., “Une vie grecque du Pape S. Martin I”, in
AnalBoll, 51 (1933), 225-262.
Pernoud, Marseille Pernoud, R., ed., Les statuts municipaux de Marseille
(Monaco and Paris, 1949).
Pero Tafur, Andanças e Pero Tafur, Andanças e viajes por diversas partes del
viajes mundo avidos, ed. G. Bellini (Rome, 1986).
Peter the Deacon Peter the Deacon, Continuatio of the Chronica monasterii
Casinensis, ed. W. Wattenbach, in MGHSS, vol. 7, 727-
844.
Peter of Eboli, De rebus Peter of Eboli, De rebus Siculis carmen, ed. E. Rota, in
Siculis RISSNS, tome 31, pt. 1.
Philo2n of Byzantium, Philo2n of Byzantium, Me2chanike2 syntaxis, ed. B. Carra de
Me2chanike2 syntaxis Vaux, Le livre des appareils pneumatiques et des
machines hydrauliques par Philon de Byzance (Paris,
1902).
Philostratos, Eijkovne" A Philostratus the elder, Imagines, in A. Fairbanks, ed.,
Philostratus Imagines, Callistratus Descriptions
(London, 1931).
Pho2kas, Nike2phoros, Pho2kas, Nike2phoros, Praecepta militaria [Strathgikh;
Praecepta militaria “Ekqesi" kai; Suvntaxi" Nikhfovrou despovtou], ed. J.
Kulakovskij, “Nikephori praecepta militaria ex codice
Mosquensi”, Mémoires de l’Académie impériale des
sciences de St.-Pétersbourg, VIII sér. Classe historico-
philologique, No 9 (St Petersburg, 1908), 1-21. Trans.
McGeer, Dragon’s teeth, 12-59.
Pho2tios, Homilies Photios, The homilies of Photius, Patriarch of
Constantinople, trans. C. Mango (Cambridge, Mass.,
1958).
Pho2tios, Lexicon ---------------, Photii Patriarchae lexicon, ed. S. A. Naber, 2
(Naber) vols (Leiden, 1864-65).
Pho2tios, Lexicon ---------------, Photii Patriarchae lexicon, ed. C. Theodoridis,
(Theodoridis) vol. 1 (Berlin, 1982).
Plato, Laws Plato, Laws, trans. R. G. Bury, 2 vols (London, 1961).
Plato, Timaeus ---------------, Timaeus, in R. G. Bury, trans. Plato, 12 vols,
(London, 1929), vol. 9.
Pliny, Natural history Gaius Plinius Secundus, Historia naturalis, in H. Rackham,
trans., Pliny, Natural History, 10 vols (London, 1938-
62).
Plummer and Earle, Plummer, C., and J. Earle, eds, Two of the Saxon chronicles
Saxon chronicles parallel (Oxford, 1892).
Plutarch, Dion Plutarch, Dion, in B. Perrin, trans., Plutarch’s lives, 11 vols
(London, 1914-26), vol. 6, 1-123.
Plutarch, Moralia ---------------, Plutarch's Moralia, trans E. I. Minar, et. al., 15
vols (Cambridge, Mass., 1961).
Pollux, Onomasticon Pollux, Julius, Iulii Pollucis Onomasticon, ed. G. Dindorf,
(Dindorf) 5 vols (Leipzig, 1824).
Pollux, Onomasticon ---------------, Pollucis onomasticon, ed. H. Bethe, 3 vols
(Bethe) (Leipzig, 1900-37).
Polyainos, Strate2ge2mata Polyianos, Polyaeni Strategematon libri VIII, ed. I. Melber,
(1887; rpt, Stuttgart, 1970).
Polybios, Histories Polybios, The Histories, in W. R. Paton, trans., Polybius:
the histories, 6 vols (London, 1922-27).
Priskos, History Priskos, History, ed. and trans. in Blockley, Fragmentary
classicising historians, 222-376.
Prodromos, Manga- Theodori Prodromi de Manganis, ed. S. Bernardinello
neios, Poems (Padua, 1972).
Prodromos, Theodore, Prodromos, Theodore, Theodoros Prodromos historische
Historical poems Gedichte, ed. W. Hörandner [Wiener Byzantinische
Studien, 11] (Vienna, 1974).
Prodromos, Theodore, ---------------, Theodori Prodromi de Rhodanthes et Dosiclis
Rhodanthe and amoribus libri IX, ed. M. Marcovich (Stuttgart, 1992).
Dosikles
Prokopios, History of Prokopios of Caesarea, History of the wars, in Procopius, 7
the wars vols, ed. and trans. H. B. Dewing (London, 1914-40).
Prokopios, Opera ---------------, Procopii Caesariensis opera omnia, ed. J. Haury,
omnia 4 vols (Leipzig, 1962-4).
Prosper, Epitoma Prosper Tiro (of Aquitaine), ed. T. Mommsen, Prosperi
chronicon Tironis epitoma chronicon, in MGHAA, vol. 9, 385-
499.
Psellos, Michael, Psellos, Michael, Chronographie ou histoire d’un siècle de
Chronographia Byzance (976-1077), ed. and trans. E. Renauld, 2 vols
(Paris, 1926).
Pseudo al-Wa2qidı3, Pseudo al Wa2qidı3, Abu2 ‘Abd Alla2h Muh5ammad ibn ‘Umar,
Futu2h5 al-S!a2m wa- Futu2h5 al-S!a2m wa-Mis5r, selections trans. Amari in BA-
Mis5r S, vol. 1, 329-38.
Pseudo Ibn Qutayba, Pseudo Ibn Qutaybah, Ah5a 2dı3th al-Ima2mah wa-l-Siya2sah,
Ah5a2dı3th al-Ima2ma trans. in De Gayangos, Mohammedan dynasties, vol. 1,
Appendix E, (pp. l-xc).
Pseudo-Kodinos, Traité Pseudo-Kodinos, Traité des offices, ed. J. Verpeaux (Paris,
des offices 1976).
Pseudo Symeon magis- Sumewvn magivstrou kai; logoqevtou cronografiva, in
tros, Chronographia Theophanes continuatus, vol. 2, 603-760.
Rabe, Scholia in Rabe, H., ed., Scholia in Lucianum (1906; rpt, Stuttgart,
Lucianum 1971).
Ralph of Caen, Gesta Ralph of Caen, Gesta Tancredi in expeditione
Tancredi Hierosolymitana, in RHCHOcc, vol. 3, 587-716. Trans.
B. S. and D. S. Bachrach, The Gesta Tancredi of Ralph
of Caen: a history of the Normans on the First Crusade
(Aldershot, 2005).
Raymond of Aguilers, Raymond of Aguilers, Historia Francorum qui ceperunt
Historia Francorum Iherusalem, in RHC HOcc, vol. 3, 231-309. Trans. J. H.
and L. L. Hill, Raymond d’Aguilers: Historia
Francorum qui ceperunt Iherusalem (Philadelphia,
1968).
Rea, Oxyrhynchus Rea, J. R., ed., The Oxyrhynchus papyri. Vol. LV (London,
papyri 1988).
Regel, Fontes Regel, V. E., and N. I. Novosadskij, eds, Fontes rerum
Byzantinarum: rhetorum saeculi XII orationes politicae,
2 vols (1891-1917; rpt, Leipzig, 1982).
Rhetorica militaris Köchly, H. (A.), ed., Anonymi byzantini rhetorica militaris
(Leipzig, 1856).
Robert of Clari, Robert of Clari, La conquête de Constantinople par Robert
Conquête de de Clari, trans. A. Micha, texte de P. Lauer [1924]
Constantinople (Paris, 1991).
Roman d’Auberon Le Roman d’Auberon, ed. J. Subrenat (Paris and Geneva,
1973).
Roman de Rou Le Roman de Rou de Wace, ed. A. J. Holden (Paris, 1973).
Roman de Troie Benoit de Sainte-Maure, Le Roman de Troie, ed. L.
Constans, 6 vols (Paris, 1904-12).
Romuald of Salerno, Romuald of Salerno, Chronicon [A.M. 130 – A.C. 1178],
Chronicon ed. C. A. Garufi in RISSNS, tome 7, parte 1, 1-297.
Russian primary The Russian Primary Chronicle: Laurentian text, eds and
chronicle trans, S. H. Cross and O. P. Sherbowitz-Wetzor, rev. ed.
(Cambridge, Mass., 1953).
Rutherford, Scholia Rutherford, W. G., ed., Scholia Aristophanica: being such
Aristophanica comments adscript to the text of Aristophanes as have
been preserved in the Codex Ravennas, 2 vols (London,
1896).
Saewulf Saewulf, in R. B. C. Huygens, ed., Peregrinationes tres:
Saewulf, John of Würzburg, Theodericus, in
CCContMed, vol. 139, 59-77.
Sanudo Torsello, Marino Sanudo Torsello, Liber secretorum fidelium crucis
Secreta fidelium super Terrae Sanctae recuperatione et conservatione
crucis (1611; rpt, Jerusalem, 1972).
Sa2wı3ris, History of the Sa2wı3ris ibn al-Muqaffa‘, Siyar al-bay‘a al-muqaddasa, ed.
Patriarchs (Evetts) and trans. B. Evetts as History of the Patriarchs of the
Coptic Church of Alexandria, 4 vols , in PO, I.2, I.4,
V.1, X.5.
Sa2wı3ris, History of the Sa2wı3ris ibn al-Muqaffa‘, Siyar al-bay‘a al-muqaddasa,
Patriarchs trans. O. H. E. KHS-Burmester, Y. ‘Abd al-Ması3h5, A.
(Burmester) S. Atiya, and A. Khater as History of the Patriarchs of
the Egyptian Church: known as the History of the Holy
Church [Publications de la Société d’archéologie Copte.
Textes et documents, X-XII] vol. 2, parts 1-3; vol. 3,
parts 1 & 2 paginated continuously (Cairo, 1943-1970).
Schiaparelli, Schiaparelli, C., Vocabulista in Arabico (Florence, 1871).
Vocabulista
Schreiner, Schreiner, P., Die byzantinischen Kleinchroniken, 3 vols
Kleinchroniken [CFHB, 12/1-3] (Vienna, 1975-9).
Sebeos, Armenian Sebeos, The Armenian history attributed to Sebeos, trans.
history R. W. Thomson, commentary J. Howard-Johnston, 2
vols (Liverpool, 1999).
Septuagint Septuagint, ed. A. Rahlfs, Septuaginta: id est Vetus
Testamentum Graece iuxta LXX interpretes, 9th ed., 2
vols (Stuttgart, 1935). Trans., The Septuagint version of
the Old Testament, with an English translation
(London, 1956).
Sextus Empiricus Sextus Empiricus, Sextus Empiricus, trans. R. G. Bury, 4
vols (London, 1971).
Sibt6 ibn al-Jawzı3, Sibt6 ibn al-Jauzı3, Shams al-Dı3n Abu2 ’l-Muz5affar Yu2suf ibn
Mir’a2t al-zama2n Qı£zughlu, Mir’a2t al-zama2n fı3 ta’rı3kh al-Aiya2m,
selections trans C. A. C. Barbier de Meynard as Extraits
du Mirât ez-zèmân, in RHCHOr, vol. 3, 512-70.
Sidonius Apollinaris, Sidonius Apollinaris, Carmina, ed. C. Luetjohann, in
Carmina MGHAA, vol. 8, 173-264.
Simokatte2s, Theophy- Simokatte2s, Theophylaktos, Theophylacti Simocattae
laktos, Historiae historiae, ed. P. Wirth (Stuttgart, 1972).
Skoutario2te2s, Theo- Skoutario2te2s, Theodore, Synopsis Chronike2, ed.
dore, Synopsis Heisenberg, as Theodori Scutariotae additamenta ad
Chronike2 Georgii Acropolitae Historiam, in George Akropolite2s,
(Heisenberg) Opera, vol. 1, 275-302.
Skoutario2te2s, Theo- ---------------, Synopsis Chronike2, ed. K. N. Sathas as
dore, Synopsis Anonymous, Suvnoysi" cronikhv, in his Mesaiwnikh;
Chronike2 (Sathas) Biblioqhvkh/Bibliotheca Graeca medii aevi, 7 vols
(1872-94; rpt, Hildesheim, 1972), vol. 7, 1-610.
Skylitze2s, John, Syno- Skylitze2s, John, Ioannis Scylitzae synopsis historiarum, ed.
psis historio2n I. Thurn [CFHB, 5] (Berlin, 1973).
Skylitze2s, John, Suvno- ---------------, Iwavnnou Skulivt zh Suvnoyi" iJstoriw'n/ Joannis
yi" iJstoriw'n Scylitzae Synopsis historiarum: Codex Matritensis
Graecus vitr. 26-2 [facsimile edition] (Athens, 2000).
Skylitze2s, John, conti- Tsolake2, E. Th., ed, JH Sunevceia th'" cronografiva" tou'
nuatus jIwavnnou Skulivtsh [Etaireia Makedonikw'n spoudw'n,
105] (Thessalonike2, 1968).
Snorri Sturluson, Snorri Sturluson, Heimskringla sí∂ara bindi, ed. B. S.
Heimskringla Kristjánsdóttir et al. (Reykjavík, 1991).
Souda Adler, A., ed., Suidae lexicon, 5 vols (Leipzig, 1928-38).
Souter, Novum Souter, A., ed., Novum Testamentum Graece, (Oxford,
Testamentum 1910).
Stephen V, Epistolae Stephen V, Pope, Stephani Papae V epistolae, diplomata et
privilegia, ed. J. D. Mansi, in PL, vol. 129, 785-822.
Sylloge2 taktiko2n Dain, A., ed., Sylloge tacticorum quae olim «inedita Leonis
tactica» dicebatur (Paris, 1938).
Symeon Logothete2s, Symeon Logothete2s, Cronografiva, in Leo Grammaticus,
Chronographia Chronographia, 3-331.
Syngraphe2 Anonymous, Suggrafh; cronografiva ta; kata; Levonta uiJo;n
chronographia Bavrda tou' Armenivo u perievcousa, in Leo Grammaticus,
Chronographia, 333-62.
Syrianos Magistros, Syrianos Magistros, Naumacivai Surianou' Magivstrou, in
Naumacivai Dain, Naumachica, 43-55.
Syrianos Magistros, Anonymous, Peri; Strathgiva", in Dennis, Three treatises,
Peri strate2gike2s 1-135
Tacitus, Annals Tacitus, The annals of Tacitus, trans. J. Jackson, in Tacitus,
3 vols (Cambridge, Mass., 1992).
Tafel and Thomas, Tafel, G. L., and G. M. Thomas, eds, Urkunden zur älteren
Urkunden Handels- und Staatsgeschichte der Republik Venedig, 3
vols (1857; rpt, Amsterdam, 1964).
Theodore of Studios, Theodore of Stoudios, Theodori Studitae epistulae, ed. G.
Epistulae Fatouros (Berlin, 1992).
Theodosios the Monk, Theodosios the Monk, La espugnazione di Siracusa nell’
Espugnazione di 880: testo greco della lettera del monaco Teodosio, ed.
Siracusa C. O. Zuretta, in Giuffrida, Scritti Michele Amari, vol.
1, 165-73. Latin version, Epistola ad Leonem
archidiaconum, in PG, vol. 135, n. 83 to coll. 51-60.
Theophane2s, Theophane2s the Confessor, Theophanis chronographia, ed.
Chronographia C. de Boor, 2 vols (Leipzig, 1883). Vol. 1 trans. C.
Mango and R. Scott, The chronicle of Theophanes
Confessor: Byzantine and Near Eastern history AD
284-813 (Oxford, 1997).
Theophane2s continuatus Theophanes continuatus, Ioannes Cameniata, Symeon
magister, Georgius monachus, ed. I. Bekker, 3 vols
[CSHB, 22] (Bonn, 1838), vol. 1.
Thietmar of Merse- Thietmar of Merseburg, Chronicon, ed. R. Holtzmann, 2nd
burg, Chronicon ed., in MGHScriptRerGerm, vol. 9.
Thucydides, Thucydides, History of the Peloponnesian war, in C.
Peloponnesian war Forster Smith, trans., Thucydides, 4 vols (London,
1919-23).
Tjader, Tjäder, J.-A., Die nichtliterarischen lateinischen papyri
Nichtliterarischen Italiens aus der Zeit 445-700, 2 vols (Lund, 1955,
lateinischen papyri 1982).
Tornikios, George, Tornikios, George, “Lovgo" tou' logiwtavtou maivstoro" tw'n
“Oratio ad Isaacium rJhtovrwn ku'r Gewrgivo u tou' Tornivkh eij" to;n
Angelum imperato- aujtokravtora ku'r Isaavkion to;n “Aggelon ...”, in Regel,
rem” Fontes, vol. 1, fasc. 2, No XV (pp. 254-80).
Translatio Sancti Monachi anonymi Littorensis, Historia de translatione
Nicolai sanctorum Magni Nicolai, terra marique miraculis
gloriosi, ejusdem avunculi, alterius Nicolai,
Theodorique, martyris pretiosi de civitate Mirae in
monasterium S. Nicolai de Littore Venetiarum, in
RHCHOcc, vol. 5, 253-92.
Translatio Sancti Translatio Sancti Severini auct. Iohanne Diacono, ed. G.
Severini Waitz, in MGHScriptRerLang, 452-9.
Turner, Oxyrhynchus Turner, E. G., et al., eds, The Oxyrhynchus papyri. Part
papyri XXVII (London, 1962).
Tzetze2s, John, Epistu- Tzetze2s, John, Epistulae, ed. P. A. M. Leone (Leipzig,
lae 1972).
Vegetius, Epitoma Flavius Vegetius Renatus, Epitoma rei militaris, ed. C.
Lang (Stuttgart, 1872). Trans. N. P. Milner, Vegetius:
Epitome of military science (Liverpool, 1993).
Vegetius, Mulomedicina Vegetius Renatus, Publius, P. Vegeti Renati digestorum
artis mulomedicinae libri, ed. E. Lommatzsch (Leipzig,
1903).
Victor of Tunnuna, Victor of Tunnuna, Victoris Tonnennensis episcopi
Chronica chronica, ed. T. Mommsen, in MGHAA, vol. 11, 184-
206.
Victor Vitensis, Victor Vitensis, Historia persecutionis Africanae
Historia provinciae temporibus Geiserici et Hunirici regum
Wandalorum, ed. K. Halm, in MGHAA, vol. 3, part 1,
1-58. Trans. J. Moorhead, Victor of Vita: History of the
Vandal persecution (Liverpool. 1992).
Vita Antonii junioris Anonymous, Bivo" kai; politevia tou' oJsivo u Antwnivou tou'
Nevou (Life and conduct of the holy Antony the
Younger), ed. A. Papadopoulos-Kerameus, in Sylloge2
Palaistine2s kai Syriake2s hagiologias [= Pravoslavnj
Palestinskij sbornik, 19.3] (St Petersburg, 1907), 186-
216.
Vita Caesarii Vitae Caesarii episcopi Arelatensis libri duo, ed. B.
Arelatensis Krusch, in MGHScriptRerMer, vol. 3, 431-501.
Vita di Sant’ Elia il Anonymous, Vita di Sant’Elia il giovane, ed. and trans. G.
giovane Rossi Taibbi (Palermo, 1962).
Vita Hludowici Vita Hludowici imperatoris, ed. G. Pertz, in MGHSS, vol.
2, 604-48. Trans. A. Cabaniss, Son of Charlemagne: a
contemporary life of Louis the Pious (Syracuse, 1961).
Vita S. Gregorii tou Dvornik, F., ed., La vie de Saint Grégoire le Décapolite et
Dekapolitou les Slaves Macédoniens au IXe siècle [Travaux publiés
par l’Institut d’études Slaves, V] (Paris, 1926).
Vita S. Ignatii Nike2tas David Paphlagon, Vita S. Ignatii archiepiscopi
Constantinopolitani, in PG, vol. 105, coll. 487-574.
Vita S. Nikolai Anonymous, The Life of St Nicholas of Sion, ed. and trans.
I. and N. P. S!evc°enko (Brookline, 1984).
Vita S. Nili Anonymous, Vita S. Nili junioris, in PG, vol. 120, 15-166.
Vita S. Theodori Oikonomides, N., “ÔO bivo" tou' aJgivou Qeodwvrou Kuqhvrwn
(10o" aij.) (12 Mai?ou --- BHG3, ajr. 2430)”, Trivton
Paniovnion Sunevdrion, Praktikav, vol. 1 (Athens, 1967),
264-91; rpt. in his Byzantium from the ninth century to
the Fourth Crusade (Aldershot, 1992), No VII.
Vita Theodorae Vita sanctae Theodorae imperatricis, in W. Regel, ed.,
Analecta Byzantino-Russica (1891-8; rpt, N.Y., n.d.), 1-
19.
Vita Theophanis Methodios, Patriarch of Constantinople, “Bivo" tou' oJsivou
patro;" hJmw'n kai; ojmologhtou' Qeofavnou": poivhma
Meqodivou patriavrcou Kwnstantinoupovlew"”, ed. V. V.
Latyshev, Zapiski Rossiiskoi Akademii Nauk po
istoricheskomu-filologicheskomu otdeleniyu, 8 ser., 13.4
(1918), 1-40.
Vitruvius, De Vitruvius, Vitruvii de architectura libri decem, ed. F.
architectura Krohn (Leipzig, 1912). Trans. F. Granger, Vitruvius on
architecture, 2 vols (London, 1934); M. H. Morgan,
The ten books on architecture (NY, 1960).
Wendel, Scholia Wendel, C., ed., Scholia in Apollonium Rhodium vetera
(Berlin, 1958).
Wescher, Poliorcetique Wescher, C., ed., Poliorcétique des grecs. Traités
théoriques - récits historiques (Paris, 1867).
Wessely, Studien Wessely, C., Studien zur Palaeographie und Papyruskunde.
III & VIII: Griechische Papyrusurkunden kleineren
Formats (Leipzig, 1904-8).
William of Apulia, William of Apulia, Gesta Roberti Wiscardi, ed. R.
Gesta Wilmans, in MGHSS, vol. 9, 239-98.
William of Tyre, William of Tyre, Willelmi Tyrensis archiepiscopi
Chronicon chronicon, ed. R. B. C. Huygens, 2 vols paginated
continuously, in CCContMed, vols 63 and 63A.
Wolf, Conquerors and Wolf, K. B., Conquerors and chroniclers of early medieval
chroniclers Spain (Liverpool, 1990).
Xenopho2n, Anabasis Xenopho2n, Anabasis, in C. L. Brownson, trans., Xenophon,
7 vols, (London, 1918-25), vol. 3 (rpt, London, 1980).
Xenopho2n, Hellenika ---------------, Hellenika, in C. L. Brownson, trans., Xenophon, 7
vols, (London, 1918-25), vols 1-2 (rpt, London, 1985-
86).
Yah5ya2 ibn Sa‘ı3d, Yah5ya2 ibn Sa‘ı3d , al-Ant6a2k i, Histoire de Yah5ya2 ibn Sa‘ı3d
Histoire d’Antioche continuateur de Sa‘ı3d -ibn-Bit6rı3q , ed. and
trans. I Kratchkovsky and A. Vasiliev, 2 vols, in PO,
XVIII.5, XXIII.3.
Yngvars Saga Yngvars Saga Ví∂förla, in G. Jónsson, ed., Fornaldar Sögur
Nor∂urlanda, vol. 2 (Reykjavik, 1959), 423-59.
Zo2naras, John, Epitome2 Zo2naras, John, Ejpitomh; iJstoriw'n/Epitome historiarum, ed.
historio2n L. Dindorf, 6 vols (Leipzig 1868-75).
Zo2simos, Historia nova Zosimi comitis et exadvocati fisci historia nova, ed. L.
Mendelssohn (Leipzig, 1887). Trans. R. T. Ridley,
Zosimus, New history (Sydney, 1982).
(d) Secondary Works: Reference Works
Bosworth, Islamic Bosworth, C. E., The Islamic dynasties: a chronological

dynasties and genealogical handbook (Edinburgh, 1967).
Doursther, Diction- Doursther, H., Dictionnaire universel des poids et mesures
naire universel anciens et modernes (Brussels, 1840).
Du Cange, Glossarium Du Cange, C. du Fresne, seigneur, Glossarium ad
scriptores mediae et infimae Graecitatis (1687; rpt,
Paris, 1943).
EI.2 The encyclopedia of Islam. New edition, ed. H. A. R. Gibb,
et al., 11 vols plus indices and supplements (Leiden and
London, 1960-2002).
Glassé, Concise encyc- Glassé, C., The concise encyclopedia of Islam (San
lopedia of Islam Francisco, 1989).
Graesse-Benedict- Graesse, J. G. T., Orbis Latinus: Lexikon lateinischer
Plechl, Orbis geographischer Namen des Mittelalters und der
Latinus Neuzeit, 4th ed., ed. H. Plechl (Braunschweig, 1972).
Jal, Glossaire nautique Jal, A., Glossaire nautique: répertoire polyglotte de termes
de marine anciens et modernes (Paris, 1848).
Kemp, Ships and the Kemp, P., ed., The Oxford companion to ships and the sea
sea (Oxford, 1976).
Liddell and Scott, Liddell, H. G., and R. Scott, A Greek-English lexicon, 9th
Greek-English ed. with revised supplement (Oxford, 1996).
lexicon
Martini, Metrologia Martini, L., Manuale di metrologia ossia misure, pesi e
monete (Turin, 1883).
Niermeyer, Lexicon Niermeyer, J. F., Mediae latinitatis lexicon minus (Leiden,
1954-76).
Oxford classical Oxford classical dictionary, The, 2nd ed., ed. N. G. L.
dictionary Hammond and H. H. Scullard (Oxford, 1970).
ODB Oxford dictionary of Byzantium, The, ed. A. P. Kazdhan, 3
vols. (N.Y. and Oxford, 1991).
Setton, HC Setton, K. M., ed., A history of the Crusades, 6 vols
(Madison, 1969-1989).
Smith, Greek and Smith, W., ed., Dictionary of Greek and Roman geography,
Roman geography 2 vols (London, 1854-7).
Sophocles, Greek Sophocles, E. A., Greek lexicon of the Roman and
lexicon Byzantine periods (from B.C. 146 to A.D. 1100 (N.Y.,
1887).
USNO, Sun and moon United States Naval Observatory, Complete sun and moon
data for one day, @ http:/aa.usno.navy.mil/ data/docs/
RS_OneDay.html
(e) Secondary Works: Books, articles, chapters
Abulafia, “Ancona” Abulafia, D., “Ancona, Byzantium and the Adriatic, 1155-
1173”, Papers of the British School at Rome, 52 (1984),
195-216. Rpt in his Italy, Sicily and the Mediterranean,
1100-1400 (London , 1987), N o IX.
Abulafia, “Norman ---------------, “The Norman kingdom of Africa and the Norman
kingdom of Africa” expeditions to Majorca and the Muslim
Mediterranean”, in R. Allen Brown, ed., Anglo-Norman
studies VII: proceedings of the Battle Conference, 1984
(Woodbridge, 1985), 26-49. Rpt in his Italy, Sicily and
the Mediterranean, 1100-1400 (London, 1987), No XII.
Abulafia, “Pisan ---------------, “The Pisan Bacini and the medieval
Bacini” Mediterranean economy: a historian’s viewpoint”, in C.
Malone and S. Stoddart, eds, Papers in Italian
archaeology IV: the Cambridge Conference, part IV,
classical and medieval archaeology [B.A.R.
International series, 246] (Oxford, 1985), 287-302. Rpt
in his Italy, Sicily and the Mediterranean, 1100-1400
(London, 1987), No XIII.
Abu Lughod, Before Abu Lughod, J., Before European hegemony: the world
European hegemony system A.D. 1250-1350 (N.Y., 1989).
Ahmad, Islamic Sicily Ahmad, A., A history of Islamic Sicily (Edinburgh, 1975).
Ahrweiler, Byzance et Ahrweiler, H., Byzance et la mer: la marine de guerre, la
la mer politique, et les institutions maritimes de Byzance aux
VIIe-XVe siècles (Paris, 1966).
Alertz, “Naval Alertz, U., “The naval architecture and oar systems of
architecture” medieval and later galleys”, in Morrison, Age of the
galley, 142-62.
Alexandres, ÔH Alexandres, K. A., ÔH qalassiva duvnami" eij" th;n iJstoriva n
qalassiva duvnami" th'" buzantinh'" aujtokratoriva" (Athens, 1956).
Anderson, Oared Anderson, R. C., Oared fighting ships from classical times
fighting ships to the coming of steam (Kings Langley, 1962).
Andriotes, Buzantinav Andriotes, N., et al., eds, Buzantinav kai Arabikav
kai Arabikav istio- istiofovra ploiva (7o"-13o" ai.)/Byzantine and Arab
fovra ploiva sailing ships (7th-13th cent.) (Athens, 2001).
Antoniadis-Bibicou, Antoniadis-Bibicou, “A propos de la première mention
“Stratège des d’un «Stratège des Caravisiens»”, Byzantinoslavica, 27
Caravisiens” (1966), 71-91.
Antoniadis-Bibicou, ---------------, Études d’histoire maritime de Byzance: à propos
“Thème des du “Thème des Caravisiens” (Paris, 1966).
Caravisiens”
Ashburner, “Byzantine Ashburner, W., “The Byzantine mutiny act”, JHS, 46
mutiny act” (1926), 80-100.
Ashtor and Kedar, “Una Ashtor, E., and B. Z. Kedar, “Una guerra fra Genova e i
guerra” Mamluchi negli anni 1380”, Archivio storico italiano,
133 (1975), 3-44.
Babuin, “Illuminations” Babuin, A., “Illuminations of nautical subject in Skylitzes
Matritensis: a preliminary report”, Graeco-Arabica, 7-8
(1999-2000), 17-30.
Babuin, “remarks on ---------------, “Some remarks on Arab ships in Byzantine
Arab ships” iconography”, in Yousef al-Hijji, Arab seafaring, 25-
40.
Babuin, “Standards” ---------------, “Standards and insignia of Byzantium”,
Byzantion, 71 (2001), 5-59.
Baldwin, “Peri Strate2- Baldwin, B., “On the date of the anonymous PERI
gike2s” STRATHGIKHS”, BZ, 81 (1988), 290-93.
Bandinelli, Hellenistic - Bandinelli, R. B., Hellenistic-Byzantine miniatures of the
Byzantine minia- Iliad (Ilias Ambrosiana) (Olten, 1955).
tures
Basch, “Galley in Basch, L., “A galley in Istanbul: the kadı3rga”, MM, 60
Istanbul” (1974), 133-5.
Basch, “Navires et Basch, L., “Navires et bateaux coptes: état des questions en
bateaux coptes” 1991”, Graeco-Arabica, 5 (1993), 23-62.
Basch, “Note sur le ---------------, “Note sur le calfatage: la chose et le mot”,
calfatage” Archaeonautica, 6 (1986), 187-98.
Bass, History of Bass, G. F., ed., A history of seafaring based on
seafaring underwater archaeology (London, 1972).
Bass and Van Door- ---------------, and F. Van Doorninck, jr, eds, Yassi Ada. Volume
ninck, Yassi Ada. I: a seventh-century Byzantine shipwreck (College
Volume I Station, 1982).
Bass et al., Serçe ---------------, S. D. Matthews, J. R. Steffy, and F. van
Limani Doorninck, jr, eds, Serçe Limani: an eleventh-century
shipwreck. Volume I: the ship and its anchorage, crew,
and passengers (College Station, 2004).
Beckwith, Early medi- Beckwith, J., Early medieval art (N.Y., 1965).
eval art
Beihammer, “Zypern” Beihammer, A., “Zypern und die byzantinisch-arabische
Seepolitik vom 8 bis zum beginn des 10 Jahrhunderts”,
in Yousef al-Hijji, Arab seafaring, 41-61.
Bel, Benou Gha2nya Bel, A., Les Benou Gha2nya [Publications de l’École des
Lettres d’Alger, 27] (Paris, 1903).
Bevere, “Ordigni ed Bevere, R., “Ordigni ed utensili per l’esercizio di arti, ed
utensili” industrie, mezzi di trasporto, ed armi in uso nelle
provincie napolitane dal XII al XVI secolo”, Archivio
storico per le provincie napoletane, 22 (1897), 702-38.
Bischoff, “Anecdota Bischoff, B., “Anecdota Carolina”, in W. Stach and H.
Carolina” Walther, Studien zur lateinischen Dichtung des
Mittelalters: Ehrengabe für Karl Strecker zum 4.
September 1931 (Schriftenreihe der historischen
Vierteljahrschrift, Heft 1) (Dresden, 1931), 1-11.
Blackman, “Hull Blackman, D. J., “Further early evidence of hull
sheathing” sheathing”, IJNA, 1 (1972), 117-119.
Blavatsky and Peters, Blavatsky, V. D., and B. G. Peters, “A wreck of the late 4th
“Donuzlav wreck” to early 3rd century BC near Donuzlav”, trans. D. J.
Blackman, IJNA, 2 (1973), 25-31.
Bondioli, et al., “Oar Bondioli, M., R. Burlet, and A. Zysberg, “Oar mechanics
mechanics” and oar power in medieval and later galleys”, in
Morrison, Age of the galley, 172-205.
Bonino, Archeologia Bonino, M., Archeologia e tradizione navale tra la
Romagna e il Po (Ravenna, 1978).
Bonino, “Rams” ---------------, “Rams of Byzantine dromons”, MM, 81 (1995),
80-82.
Bosworth, “Arab Bosworth, C. E., “Arab attacks on Rhodes in the pre-
attacks on Rhodes” Ottoman period”, in Yousef al-Hijji, Arab seafaring,
63-74
Bouras, “Basil Bouras, L., “O Basivleio" Lekaphnov" paraggeliodovth"
Lekapenos” e[rgwn tevcnh"”, in Constantine VII Porphyrogenitus and
his age: Second international Byzantine conference,
Delphi, 22-26 July 1987 (Athens, 1989), 397-434.
Bovini, “Felice Kibel” Bovini, G., “Principali restauri compiuti nel secolo scorso
da Felice Kibel nei mosaici di S. Apollinare Nuovo di
Ravenna”, Corsi di cultura sull’arte ravennate e
bizantina, 13 (1966), 83-104.
Bragadin, “Navi” Bragadin, M. A., “Le navi, loro strutture e attrezzature
nell’alto medioevo”, in La navigazione mediterranea
nell’alto medioevo (Settimane di studio del Centro
Italiano di studi sull’alto medioevo, XXV), 2 vols
(Spoleto, 1978), vol. 1, 389-407.
Brett, “Armies of Brett, M., “The armies of Ifrı3q iya. 1052-1160”, Cahiers de
Ifrı3qiya” Tunisie, 48 (1977), 107-25; rpt in his Ibn Khaldun and
the medieval Maghrib (Aldershot, 1999), No XII.
Brokkaar, “Basil Brokkaar, W. G., “Basil Lacapenus. Byzantium in the 10th
Lacapenus” century”, in W. F. Bakker, et al., eds, Studia Byzantina
et neohellenica neerlandica (Leiden, 1972), 199-234.
Brooks, “Relations Brooks, E. W., “The relations between the empire and
between the empire Egypt from a new Arabic source”, BZ, 22 (1913), 381-
and Egypt” 91.
Brubaker, Vision and Brubaker, L., Vision and meaning in ninth-century
meaning Byzantium: image as exegesis in the homilies of
Gregory of Nazianzus (Cambridge, 1999).
Bryer, “Shipping” Bryer, A., “Shipping in the Empire of Trebizond”, M.M.,
52 (1966), 3-12.
Burlet, et al., “Comment Burlet, R., J. Carrière, A. Zysberg, “Mais comment
pouvait-on ramer” pouvait-on ramer sur les galère du Roi-Soleil”, Histoire
& Mesure, 1 (1986), 147-208.
Bury, “Ceremonial Bury, J. B., “The ceremonial book of Constantine Por-
book” phyrogennetos”, EHR, 22 (1907), 209-27, 417-39.
Canard, “Textes” Canard, M., “Textes relatifs à l’emploi du feu grégeois
chez les Arabes”, Bulletin des études Arabes, 6.26
(1946), 3-7.
Carile and Cosentino, Carile, A., and S. Cosentino, Storia della marineria
Marineria bizantina bizantina (Bologna, 2004).
Carre and Jézégou, Carre, M.-B., and M.-P. Jézégou, “Pompes à chapelet sur
“Pompes à chapelet” des navires de l’antiquité et du début du moyen-âge”,
Archaeonautica, 4 (l984), 115-43.
Casson, “Origin of the Casson, L., “The origin of the lateen”, American Neptune,
lateen” 31 (1971), 49-51.
Casson, Ships and ---------------, Ships and seamanship in the ancient world
seamanship (Princeton, 1971).
Casson, Athlit ram ---------------, and J. R. Steffy, eds, The Athlit ram (College
Station, 1991).
Catalogue général Catalogue général des manuscrits des bibliothèques
publiques des départements. Tome IV: Arras ---
Avranches --- Boulogne (Paris, 1872).
Chapman, Insects Chapman, R. F., The insects: structure and function
(London, 1969).
Cheira, Lutte Cheira, M. A., La lutte entre Arabes et Byzantines: la
conquête et l’organisation des frontières aux VIIe et
VIIIe siècles (Alexandria, 1947).
Chevedden, “Artillery” Chevedden, P. E., “Artillery in late antiquity: prelude to the
Middle Ages”, in I. A. Corfis and M. Wolfe, eds, The
medieval city under siege (Woodbridge, 1995), 131-73.
Christides, “Arab- Christides, V., “Arab-Byzantine struggle in the sea: naval
Byzantine struggle” tactics (7th-11th C. A.D.): theory and practice”, in
Yousef al-Hijji, Arab seafaring, 87-106.
Christides, “Byzantine ---------------, “Byzantine dromon and Arab s5h5ı3nı3: the
dromon and Arab development of the average Byzantine and Arab
s5h5ı3nı3” warships and the problem of the number and function
of the oarsmen”, in Tzalas, Tropis III, 111-122.
Christides, ---------------, “Once again Caminiates’ ‘Capture of
“Caminiates” Thessaloniki’ ”, BZ, 74 (1980), 7-10.
Christides, Conquest of ---------------, The conquest of Crete by the Arabs (ca. 824): a
Crete turning point in the struggle between Byzantium and
Islam (Athens, 1984).
Christides, “Dha2t as5- ---------------, “The naval engagement of Dha2t as5-S4awa2rı3 A.H.
S4awa2rı3” 34 / A.D. 655-656: a classical example of naval warfare
incompetence”, Byzantina, 13 (1985), 1329-1345.
Christides, “Ibn al- ---------------, “Ibn al-Manqalı3 (Manglı3) and Leo VI: new
Manqalı3” evidence on Arabo-Byzantine ship construction and
naval warfare”, Byzantinoslavica, 56 (1995), 83-96.
Christides, ---------------, “Introduction, reconstructing medieval Byzantine
“Introduction” and Arab ships: remarks and conclusions” in Andriotes,
Buzantinav kai; Arabikav istiofovra ploiva , 27-33.
Christides, “Military ---------------, “Military intelligence in Arabo-Byzantine naval
intelligence” warfare”, in Oikonomides, Byzantium at War, 269-81.
Christides, “Naval ---------------, “Naval history and naval technology in medieval
history” times: the need for interdisciplinary studies”,
Byzantion, 58 (1988), 309-32.
Christides, “New light” ---------------, “New light on navigation and naval warfare in
the eastern Mediterranean, the Red Sea and the Indian
Ocean (6th-14th centuries A.D.)”, Nubica, III/1 (1994),
3-42.
Christides, “Parallel ---------------, “Two parallel naval guides of the tenth century:
naval guides” Qudama’s document and Leo VI’s Naumachica: a
study on Byzantine and Moslem naval preparedness”,
Graeco-Arabica, 1 (1982), 51-103.
Christides, “Raids of the ---------------, “The raids of the Moslems of Crete in the Aegean
Moslems” Sea: piracy and conquest”, Byzantion, 51 (1981), 76-
111.
Christides, Sailing ships ---------------, ed., Sailing ships of the Mediterranean Sea and
the Arabian Gulf: Volume II. Navigation in the Red
Sea, the Arabian Gulf and the Indian Ocean (Athens,
2000).
Christides, ---------------, “The transmission of Chinese maritime
“Transmission” technology by the Arabs to Europe”, American
Neptune, 52 (1992), 38-45.
Christidou, Arab-Byza- Christidou, S., A. Apostolopoulos, and V. Christides, eds,
ntine navigation Treasures of Arab-Byzantine navigation/Qhsauroiv th"
Arabo-Buzantinh" nausiploia" (Athens, 2004).
Chryssos, Griechen- Chryssos, E., et al., eds, Griechenland und das Meer:
land und das Meer Beiträge eines Symposions in Frankfurt im Dezember
1996 [Peleus: Studien zur Archäologie und Geschichte
Griechenlands und Zyperns, Bd 4] (Mannheim and
Möhnesee, 1999).
Ciampini, Vetera Ciampini, G. G., Vetera monimenta in quibus praecipuè
monimenta musiva opera sacrarum … illustrantur, 2 parts (Rome,
1699) [non vidimus].
Clover, “Count Gaïnas” Clover, F. M., “Count Gaïnas and Count Sebastian”,
American journal of ancient history”, 4 (1979), 65-76.
Coates, “Development” Coates, J. F., “Development of the design” in Shaw,
Trireme project, 71-4.
Coates, “Naval ---------------, “The naval architecture of European oared
architecture” ships”, The Royal Institution of naval architects, Spring
Meetings 1993, Paper N o 9.
Coates, “Naval ---------------, “The naval architecture and oar systems of
architecture and oar ancient galleys”, in Morrison, ed., Age of the galley,
systems” 127-141.
Coates, “Oar” ---------------, “Should the oar be rigged aft or forward of the
tholepin”, in Shaw, Trireme project, 48-9.
Coates, “Reconstruct- ---------------, “The reconstruction”, in Morrison and Coates,
ion” Trireme reconstructed, 17-25.
Coates, “Spanish ---------------, “The strength and behaviour of tension-
windlasses” tourniquets, or Spanish windlasses of natural fibre
ropes”, IJNA, 16 (1987), 207-11.
Coates and Morrison, ---------------, and J. Morrison, “The sea trials of the
“Sea trials” reconstructed Athenian trireme Olympias: a reply to
Lucien Basch”, MM, 79 (1993), 131-41.
Coates and Shaw, ---------------, and (J.) T. Shaw, “Speculations on fitting
“Speculations” hypozomata”, in Shaw, Trireme project, 82-6.
Coates, et al., Trireme ---------------, S. K. Platis, and J. T. Shaw, The trireme trials
trials 1988: report on the Anglo-Hellenic sea trials of
Olympias (Oxford, 1990).
Codera, “Mochéhid” Codera, F., “Mochéhid, conquistador de Cerdeña”, in
Giuffrida, Scritti Michele Amari, vol. 2, 115-33.
Collins, Arab conquest Collins, R., The Arab conquest of Spain 710-797 (Oxford,
1989).
Cosentino, “Epitafio Cosentino, S., “Un epitafi sardo del secolo VI concernante
sardo” probabilmente un dromonarius”, in Carile and
Cosentino, Marineria bizantina”, 185-207.
Cosentino, “Flotte ---------------, “La flotte byzantine face à l’expansion
byzantine” musulmane. Aspects d’histoire institutionellle et
sociale (VIIe-Xe siècles), BF, 28 (2004), 3-20.
Cosentino, “Syrianos’s ---------------, “The Syrianos’s «Strategikon»: a ninth-century
«Strategikon»” source?”, Bizantinistika: rivista di studi Bizantini e
Slavi, 2nd ser., 2 (2000), 243-80.
Cotterell and Kammin- Cotterell, B. and J. Kamminga, Mechanics of pre-industrial
ga, Mechanics of technology: an introduction to the mechanics of ancient
pre-industrial and traditional material culture (Cambridge, 1990).
technology
Courtois, “Politiques Courtois, C., “Les politiques navales de l’Empire romain”,
navales” Revue historique, 64 (1939), 17-47, 225-59.
Courtois, Vandales Courtois, C., Les Vandales et l’Afrique (Paris, 1955).
Cowdrey, “Mahdia Cowdrey, H. E. J., “The Mahdia campaign of 1087”, EHR,
campaign” 92 (1977), 1-29.
Croke, Count Croke, B., Count Marcellinus and his chronicle (Oxford,
Marcellinus 2001).
Dain, Elien le tacticien Dain, A., Histoire du texte d’Élien le tacticien des origines
à la fin du moyen-âge (Paris, 1946).
Dain, Nicephore Oura- ---------------, La “Tactique” de Nicéphore Ouranos (Paris,
nos 1937).
Dain, “Stratégistes” ---------------, “Les stratégistes Byzantins” (completed by J.-A.
de Foucault), TM, 2 (1967), 317-92.
Davis, Medieval war- Davis, R. H. C., The medieval warhorse: origin,
horse development and redevelopment (London, 1989).
De Chaumont, “On De Chaumont, F., “On ventilation and cubic space”,
ventilation” Edinburgh medical journal, 12.11 (May, 1867), 1024-
34.
De Goeje, De Goeje, M. J., “Quelques observations sur le feu
“Observations” Grégois”, in Homenaje á D. Francisco Codera en su
jubilación del profesorada (Zaragoza, 1904), 93-8.
De Hamel, Illuminated De Hamel, C., A history of illuminated manuscripts, 2nd
manuscripts ed. (London, 1994).
Delogu, Longobardi e Delogu, P., A. Guillou, and G. Ortalli, Longobardi e
Bizantini. Bizantini (Turin, 1980).
Denham, Aegean Denham, H. M., The Aegean: a sea-guide to its coast and
islands, 5th ed. (London, 1983).
Dennis, “Byzantine Dennis, G. T., “Byzantine battle flags”, BF, 8 (1982), 51-
battle flags” 59.
Dennis, “Flies, mice, ---------------, “Flies, mice, and the Byzantine crossbow”,
and the Byzantine BMGS, 7 (1981), 1-5.
crossbow”
Der Nersessian, Der Nersessian, S., L’illustration des psautiers Grecs du
L’illustration moyen-âge. II: Londres, Add. 19.352 (Paris, 1970).
Di Stefano, “Antichi Di Stefano, G., “Antichi relitti nella baia di Camarina”, in
relitti” P. S. Gianfrotti, ed., IV rassegna di archeologia
subacquea. IV premio Franco Papò. Atti (Messina,
1991), 127-34.
Dionisotti, “Greek Dionisotti, A. C., “Greek grammars and dictionaries in
grammars and Carolingian Europe”, in M. W. Herren, ed., The sacred
dictionaries” nectar of the Greeks: the study of Greek in the West in
the early Middle Ages (London, 1988), 1-56.
Dohrn, Ficoronische Dohrn, T., Die Ficoronische ciste in der Villa Giulia in
ciste Rom (Berlin, 1972).
Dolley, “Warships” Dolley, R. H., “The warships of the later Roman Empire”,
JRS, 38 (1948), 47-53.
Dolley, “Meteorology” ---------------, “Meteorology in the Byzantine navy”, MM, 37
(1951), 5-16.
Dolley, “Naval tactics” ---------------, “Naval tactics in the heyday of the Byzantine
thalassocracy”, Atti dell’ VIII congresso di studi
bizantini, I (Rome, 1953), 324-39.
Dolley, “Rig” ---------------, “The rig of early medieval warships”, MM, 35
(1949), 51-5.
Dotson, “Galley design” Dotson, J. E., “Merchant and naval influences on galley
design at Venice and Genoa in the fourteenth century”,
in C. L. Symonds, ed., New aspects of naval history:
selected papers presented at the Fourth naval history
Symposium, United States naval Academy (Annapolis,
1981), 20-32.
Dotson, “Economics ---------------, “Economics and logistics of galley warfare”, in
and logistics” Morrison, Age of the galley, 217-23.
Dotson, “Simone ---------------, “The voyage of Simone Leccavello: a Genoese
Leccavello” naval expedition of 1351”, in Saggi e documenti VI
(Genoa, 1985), 267-82.
Dufrenne, L’illustra- Dufrenne, S., L’illustration des psautiers Grecs du moyen-
tion âge. I: Pantocrator 61, Paris Grec 20, British Musuem
40731 (Paris, 1966).
Dufrenne and Villain- ---------------, and C. Villain-Gandossi, “Bateaux figurés dans
Gandossi, “Bateaux des oeuvres Carolingiennes”, Archaeonautica, 4
figurés” (1984), 243-60.
Dunbabin, Mosaics of Dunbabin, K. M. D., The mosaics of Roman North Africa:
Roman North Africa studies in iconography and patronage (Oxford, 1978).
Dunlop and Williams, Dunlop, R. H., and D. J. Williams, Veterinary medicine: an
Veterinary medi- illustrated history (St Louis, 1996).
cine
Dvornik, Intelligence Dvornik, F., Origins of intelligence services: the ancient
services Near East, Persia, Greece, Rome, Byzantium, the Arab
Muslim empires, the Mongol empire, China, Muscovy
(New Brunswick, 1974).
Edlin, What wood is Edlin, H. L., What wood is that: a manual of wood
that identification (London, 1969).
Ehrenkreutz, “Place of Ehrenkreutz, A., “The place of Saladin in the naval history
Saladin” of the Mediterranean Sea in the Middle Ages”, Journal
of the American Oriental Society, 75 (1955), 100-16.
Eickhoff, “Galeeren- Eickhoff, E., “Galeerenkriege im Mittelmeer (7.-11.
kriege” Jahrhundert): technologische Traditionen”, in
Ordinamenti militari in Occidente nell’alto medioevo
[Settimane di studio del Centro italiano di studi
sull’alto medioevo, XV], 2 vols paginated continuously
(Spoleto, 1968), vol. 2, 979-1007.
Eickhoff, Seekrieg und ---------------, Seekrieg und Seepolitik zwischen Islam und
Seepolitik Abendland: das Mittelmeer unter byzantinischer und
arabischer Hegemonie (650-1040) (Berlin, 1966).
Eisenberg, “Metallur- Eisenberg, S., “Metallurgical analysis of the ram”, in
gical analysis” Casson, Athlit ram, 40-50.
Ellis Davidson, “Secret Ellis Davidson, H. R., “The secret weapon of Byzantium”,
weapon” BZ, 66 (1973), 61-74.
Estopañan, Skyllitzes Estopañan, S. C., ed., Skyllitzes Matritensis. Tomo I:
Matritensis reproducciones y miniaturas (Barcelona and Madrid,
1965).
Face, “Caffaro” Face, R. D., “Secular history in twelfth-century Italy:
Caffaro of Genoa”, JMH, 6 (1980), 169-84.
Fahmy, Muslim naval Fahmy, A. M., Muslim naval organisation in the Eastern
organisation Mediterranean from the seventh to the tenth century
A.D., 2nd ed. (Cairo, 1966).
Fantar, Mosaique en Fantar, M. H., ed., La mosaïque en Tunisie (Paris and
Tunisie Tunis, 1994).
Farello, “Niceforo Farello, F. A., “Niceforo Foca e la riconquista di Creta”,
Foca” Medioevo greco. Rivista di storia e folologia bizantina,
1 (2001), 141-60.
Featherstone, “Further Featherstone, M., “Further remarks on the De cerimoniis”,
remarks” BZ, 97 (2004), 113-21.
Featherstone, “Prelim- ---------------, “Preliminary remarks on the Leipzig manuscript
inary remarks” of De cerimoniis”, BZ, 95 (2002), 457-79.
Fine, Early medieval Fine, J. V. A., jr, The early medieval Balkans: a critical
Balkans survey from the sixth to the late twelfth century (Ann
Arbor, 1983).
Fiori, Costantino Fiori, F., Costantino hypatos e doux di Sardegna (Bologna,
hypatos 2001).
Fitzgerald, “The ship” Fitzgerald, M. A., “Chapter VI: the ship”, in J. P. Oleson,
ed., The harbours of Caesarea Maritima: results of the
Caesarea ancient harbour excavation project 1980-85.
Volume II: the finds and the ship (Oxford, 1994), 163-
223.
Foerster, “Warships of Foerster, F., “The warships of the kings of Aragón and
Aragón” their fighting tactics during the 13th and 14th centuries
AD”, IJNA, 16 (1987), 19-29.
Foerster-Laures, “Bilge --------------- (-Laures), F., “The problem of the bilge and the
and pump” pump in antiquity”, in Tzalas, Tropis I, 91-6.
Forbes, Art of Forbes, R. J., Short history of the art of distillation from
distillation the beginnings up to the death of Cellier Blumenthal
(Leiden, 1948).
Forbes, More studies ---------------, More studies in early petroleum history 1860-
1880 (Leiden, 1959).
Forbes, Studies ---------------, Studies in early petroleum history (Leiden, 1958).
Foucher, Navires et Foucher, L., Navires et barques figurés sur des mosaiques
barques découvertes à Sousse et aux environs (Institut national
d’archéologie et arts, Musée Alaoui. Notes et
documents, XV) (Tunis, 1957).
Frendo, “St Demetrius” Frendo, D. C., “The miracles of St Demetrius and the
capture of Thessaloniki”, Byzantinoslavica, 58 (1997),
205-24.
García Moreno, “Byz- García Moreno, L. A., “The creation of Byzantium’s
antium’s Spanish Spanish province. Causes and propaganda”, Byzantion,
province” 66 (1996), 101-119.
Garlan, Poliorcetique Garlan, Y., Recherches de poliorcétique grec (Paris, 1974).
grecque
Gaudefroy- Gaudefroy-Demombynes, M., “Une lettre de Saladin au
Demombynes, calife Almohade”, in Mélanges René Basset: études
“Lettre de Saladin” nord-africaines et orientales publiés par l’Institut des
hautes-études Marocaines, 2 vols (Paris, 1925), vol. 2,
279-304.
Gay, L’Italie méridion- Gay, J., L’Italie méridionale et l’empire byzantin depuis
ale l’avènement de Basile Ier jusqu’a la prise de Bari par
les Normands (867-1071) (Paris, 1904).
Giuffrida, Scritti Giuffrida, G., int., Scritti per il centenario della nascità di
Michele Amari Michele Amari [Documenti per servire alla storia di
Sicilia pubblicati a cura della Società Siciliana per la
storia patria. Diplomatica – serie IV – cronache e scritti
vari], 2 vols (1910; rpt, Palermo, 1990).
Goitein, Mediterranean Goitein, S. D., A Mediterranean society: the Jewish
society communities of the Arab world as portrayed in the
documents of the Cairo Geniza. Vol. I: economic
foundations (Berkeley and Los Angeles, 1967).
Gorini, Mirnik, and Gorini, G., I. Mirnik, and E. Chino, “I falsi di Meneghetti”,
Chino, “Falsi di Bollettino del Museo Civico di Padova, 80 (1991), 321-
Meneghetti” 57.
Goudas, “Mesaiwnika; Goudas, M., “Mesaiwnika; karavgmata ploivwn ejpi; tou'
karavgmata” Qhseivo u”, Byzantis, 2 (1910), 329-57.
Grabar and Manoussa- Grabar, A., and M. Manoussacas, L’illustration du
cas, L’illustration manuscrit de Skylitzès de la Bibliothèque Nationale de
Madrid (Venice, 1979).
Gransden, Historical Gransden, A., Historical writing in England c. 550 to c.
writing 1307 (London, 1974).
Great Britain, Black Sea Great Britain, Admiralty, Black Sea Pilot, 11th ed.
Pilot (London, 1969).
Great Britain, Mediter- ---------------, Mediterranean Pilot. Vol. IV, 9th ed. (London,
ranean Pilot 1968); Vol. V, 6th ed. (London, 1976).
Guilland, “Drongaire” Guilland, R., “Le Drongaire de la flotte, le Grand
Drongaire de la flotte, le Duc de la flotte, le Megaduc”,
in his Recherches sur les institutions Byzantines, 2 vols
(Berlin and Amsterdam, 1967), vol. 1, 535-62
[originally published in BZ, 44 (1951), 212-40].
Guilmartin, Gunpow- Guilmartin, J. F., Gunpowder and galleys: changing
der and galleys technology and Mediterranean warfare at sea in the
sixteenth century (Cambridge, 1974).
Haldane, “Fire-ship of Haldane, D., “The fire-ship of al-Sa2lih Ayyu2b and Muslim
al-Sa2lih” use of ‘Greek Fire’ ”, in D. J. Kagay and L. J. Andrew
Villalon, eds, The circle of war in the Middle Ages:
essays on medieval military and naval history
(Woodbridge, 1999), 139-44.
Haldon, “Expedition- Haldon, J., “The organisation and support of an
ary force” expeditionary force: manpower and logistics in the
Middle Byzantine period”, in Oikonomides, Byzantium
at War, 111-51.
Haldon, “ ‘Greek Fire’ ---------------, “‘Greek Fire’ revisited: recent and current
revisited” research”, in E. Jeffreys, ed., Byzantine style, religion,
and civilization: in honour of Sir Steven Runciman
(Cambridge, forthcoming in 2006).
Haldon and Byrne, ---------------, and M. Byrne, “A possible solution to the
“Greek Fire” problem of Greek Fire”, BZ, 70 (1977), 91-9.
Halm, Empire of the Halm, H., The empire of the Mahdi: the rise of the
Mahdi Fatimids (Leiden, 1996).
Hansen, Athenian Hansen, M. H., The Athenian democracy in the age of
democracy Demosthenes: structure, principles, and ideology
(Oxford, 1991).
Harris, “Frederic af Harris, D. G., “Admiral Frederic af Chapman’s auxiliary
Chapman” vessels for the Swedish inshore fleet”, MM, 75 (1989),
211-29.
Harris, “Bessarion on Harris, J., “Bessarion on shipbuilding”, Byzantinoslavica,
shipbuilding” 55 (1994), 291-303.
Hayes, Horses on board Hayes, M. H., Horses on board ship: a guide to their
ship management (London, 1902).
Héliot, Manuscrits Héliot, P., Les manuscrits illustrés de la Bibliothèque de
illustrés Boulogne (Paris, 1934).
Hesseling, Mots Hesseling, D. C., Les mots maritimes empruntés par le
maritimes Grec aux langues romanes (Amsterdam, 1903).
Hewitt, Organization of Hewitt, H. J., The organization of war under Edward III
war 1338-62 (Manchester, 1966).
Hocker, “1995 field Hocker, F., “The Byzantine shipwreck at Bozburun,
season” Turkey: the 1995 field season”, INA [Institute of
nautical archaeology] quarterly, 22.4 (1995), 3-8.
Hocker, “1997 field ---------------, “The Byzantine shipwreck at Bozburun, Turkey:
season” the 1997 field season”, INA [Institute of nautical
archaeology] quarterly, 25.2 (1998), 12-17.
Hocker, “Final cam- ---------------, “Bozburun Byzantine shipwreck excavation: the
paign” final campaign”, INA [Institute of nautical
archaeology] quarterly, 25.4 (1998), 3-13.
Hocker, “Galleys and ---------------, “Late Roman, Byzantine, and Islamic galleys and
fleets” fleets”, in Morrison, Age of the galley, 86-100.
Hocker, “Lead hull ---------------, “Lead hull sheathing in antiquity”, in Tzalas,
sheathing” Tropis III, 197-206.
Höckmann, Antike Höckmann, O., Antike Seefahrt (Munich, 1985).
Seefahrt
Höckmann, “Liburni- ---------------, “The Liburnian: some observations and insights”,
an” IJNA, 26 (1997), 192-216.
Hodges and White- Hodges, R., and D. Whitehouse, Mohammed, Charlemagne
house, Mohammed and the origins of Europe: archaeology and the
Pirenne thesis (Ithaca, 1983).
Howard-Johnston, Howard-Johnston, J., “Anna Komnene and the Alexiad”, in
“Anna Komnene” M. Mullett and Dion Smythe, eds, Alexios I Komnenos.
I: Papers (Belfast, 1996), 260-302.
Hunger, Hochsprach- Hunger, D., Die hochsprachliche profane litteratur der
liche profane Byzantiner, 2 vols (Munich, 1978).
litteratur
Huxley, “Porphyro- Huxley, G., “A Porphyrogenitan Portulan”, Greek, Roman,
genitan Portulan” and Byzantine Studies, 17 (1976), 295-300.
Hyland, Equus Hyland, A., Equus: the horse in the Roman world (London,
1990).
Hyland, Medieval ---------------, The medieval warhorse from Byzantium to the
Warhorse Crusades (Stroud, 1994).
Idris, Berbérie orientale Idris, H. R., La Berbérie orientale sous les Zı3rı3d es Xe-XIIe
siècles, 2 vols paginated continuously (Paris, 1962).
Jal, Archéologie navale Jal, A., Archéologie navale (Paris, 1840).
Jenkins, “Cyprus” Jenkins, R. J. H., “Cyprus between Byzantium and Islam,
A.D. 688-965”, in G. E. Mylonas and D. Raymond,
eds, Studies presented to David Moore Robinson (Saint
Louis, 1953), vol. 2, 1006-14.
Johnson and West, Johnson, A. C., and L. C. West, Byzantine Egypt:
Byzantine Egypt economic studies (Princeton, 1949).
Joranson, “Spurious Joranson, E., “The spurious letter of Emperor Alexius to
letter” the Count of Flanders”, American historical review, 55
(1950), 3-43.
Kahane, “Byzantino- Kahane, H. and R., “Byzantinoromanica”, in Polychronion
romanica” [Festschrift Fr. Dölger] (Heidelberg, 1966), 304-17; rpt
in their Graeca et Romanica, 451-64.
Kahane, “Eléments ---------------, “Les éléments byzantins dans les langues
byzantins” Romanes”, Cahiers Ferdinand de Saussure, 23 (1966),
67-73; rpt in their Graeca et Romanica, 465-71.
Kahane, Graeca et ---------------, Graeca et Romanica scripta selecta. Volume one:
Romanica Romance and Mediterranean lexicology (Amsterdam,
1979).
Kahane, “Massaliotica” ---------------, “Massaliotica”, Romanica [Gerhard Rohlfs
Testimonial] (Halle, 1958), 231-42; rpt in their Graeca
et Romanica, 319-330.
Kahane, “Two nautical ---------------, “Two nautical terms of Greek origin: typhoon
terms” and galley”, in Etymologica (Tübingen, 1958), 417-39;
rpt in their Graeca et Romanica, 295-317.
Kahane and Pietran- ---------------, and A. Pietrangeli, “Cultural criteria for Western
geli, “Cultural borrowings from Byzantine Greek”, in Homenaje a
criteria” Antonio Tovar (Madrid, 1972), 205-29; rpt in their
Graeca et Romanica, 509-33.
Kahane and Tietze, ---------------, and A. Tietze, The Lingua Franca in the Levant:
Lingua Franca Turkish nautical terms of Italian and Greek origin
(1958; rpt., Istanbul, 1988).
Kalligas, Monemvasia Kalligas, H., Byzantine Monemvasia. The sources
(Monemvasia, 1990).
Karlin-Hayter, “Military Karlin-Hayter, P., “ ‘When military affairs were in Leo’s
affairs” hands’: a note on Byzantine foreign policy, (886-912)”,
Traditio, 23 (1967), 15-40.
Kaster, Guardians of Kaster, R. A., Guardians of language: the grammarian and
language society in late antiquity (Berkeley, 1988).
Kazhdan, “Some Kazhdan, A., “Some questions addressed to the scholars
questions” who believe in the authenticity of Kaminiates’ ‘Capture
of Thessalonica’ ”, BZ, 71 (1978), 301-14.
Kennedy, Early Kennedy, H., The early Abbasid Caliphate: a political
Abbasid Caliphate. history (London and Sydney, 1981).
Khoury, “Leo Tripol- Khoury Odetallah, R., “Leo Tripolites-Ghula2m Zura2fa and
ites-Ghula2m Zura2fa” the sack of Thessaloniki in 904”, Byzantinoslavica, 56
(1995), 97-102.
Kienast, Kienast, D., Untersuchungen zu den Kriegsflotten der
Untersuchungen Römischen Kaiserzeit (Bonn, 1966).
Kilby, Cooper Kilby, K., The cooper and his trade (Fresno, 1971).
Koder, “Thalassokratia Koder, J., “Aspekte der Thalassokratia der Byzantiner in
der Byzantiner” der Ägäis”, in Chryssos, Griechenland und das Meer,
101-9.
Kolias, “Byzantini-sche Kolias, T. G., “Die byzantinische Kriegsmarine. Ihre
Kriegsmarine” Bedeutung im Verteidigungssystem von Byzanz”, in
Chryssos, Griechenland und das Meer, 133-9.
Kolias, Byzantinische ---------------, Byzantinische Waffen: ein Beitrag zur
Waffen Byzantinischen Waffenkunde von den Anfängen bis zur
Lateinischen Eroberung (Vienna, 1988).
Kolias, “Kamelaukion” ---------------, “Kamelaukion”, Jahrbuch der Österreichischen
Byzantinistik, 32 (1982) [XVI. Internationaler
Byzantinistenkongress, Wien, 4.-9. Oktober 1981.
Akten. II.Teil. 3. Teilband], 493-502.
Kolias, “Taktika of Leo ---------------, “The Taktika of Leo VI the Wise and the Arabs”,
VI” Graeco-Arabica, 3 (1984), 129-35.
Korres, “Greek Fire” Korres, T. K., “ ‘Greek Fire’: problems concerning the use
of the ‘secret’ weapon of the Byzantine navy”, in
Arcaiva Ellhnikhv tecnologiva. IO¡ Dieqne"V Sunevdrio.
Praktikav (Athens, 1997), 533-43.
Korres, «ÔUgro;n pu'r» ---------------, «ÔUgro;n pu'r»: VEna ovplo th" buzantinhv" nautikhv"
taktikhv" (Thessalonike2, 1995).
Koukoulès, “Nautiko;" Koukoulès, P., “ÔO nautiko;" bivo"”, in his Buzantinw'n bivo "
bivo"” kai; politismov", V (Athens, 1952), 344-86.
Koutrakou, “Diploma- Koutrakou, N., “Diplomacy and espionage: their role in
cy and espionage” Byzantine foreign relations, 8th-10th centuries”,
Graeco-Arabica, 6 (1995), 125-44.
Koutrakou, “Spies of ---------------, “ ‘Spies of towns’. Some remarks on espionage
towns” in the context of Arab-Byzantine relations (VIIth-Xth
centuries)”, in V. Christides and T. Papadopoullos, eds,
Proceedings of the Sixth International Congress of
Graeco-Oriental and African studies, Nicosia 30 April
– 5 May 1966 [Graeco-Arabica, 7-8 (1999-2000)]
(Nicosia, 2000), 243-66.
Kreutz, Before the Kreutz, B., Before the Normans: Southern Italy in the ninth
Normans and tenth centuries (Philadelphia, 1991).
Kubiak, “Byzantine Kubiak, W. B., “The Byzantine attack on Damietta in 853
attack” and the Egyptian navy in the 9th century”, Byzantion,
40 (1970), 145-66.
Kyrris, “Cyprus” Kyrris, C. P., “The nature of the Arab-Byzantine relations
in Cyprus from the middle of the 7th to the middle of
the 10th century A.D.”, Graeco-Arabica, 3 (1984), 148-
75.
Lagardère, Almoravides Lagardère, V., Les Almoravides jusqu’au règne de Yu2suf b.
Ta2s°fı3n (1039-1106) (Paris, 1989).
Lagardère, Djihad ---------------, Les Almoravides: le Djihad Andalou (1106-1143)
Andalou (Paris and Montréal, 1998).
Law, Grammar and Law, V., Grammar and grammarians in the early Middle
grammarians Ages (London, 1997).
Lendle, “Das Lendle, O., “Das Karchesion: Gerät am Masttopp --
Karchesion” Trinkegefäss -- Drehkipp-Gelenk”, Acta classic
Universitatis Scientiarum Debreceniensis, 32 (1997
[for 1996]), 85-117.
Lepper and Frere, Lepper, F., and S. Frere, Trajan’s column: a new edition of
Trajan’s column the Cichorius plates (Gloucester and Wolfboro, 1988).
Lev, “Fa2t6imid navy” Lev, Y., “The Fa2t6imid navy, Byzantium and the
Mediterranean sea 909-1036 C.E. / 297-427 A.H.”,
Byzantion, 54 (1984), 200-52.
Lev, State and society ---------------, State and society in Fatimid Egypt (Leiden,
1991).
Lev, Saladin in Egypt ---------------, Saladin in Egypt (Leiden, 1999).
Levi della Vida, Levi della Vida, G., “A papyrus reference to the Damietta
“Damietta raid” raid of 853 A.D.”, Byzantion, 17 (1944-5), 212-21.
Lévi-Provençal, L’Es- Lévi-Provençal, E., Histoire de l’Espagne musulmane, 3
pagne musulmane vols (Paris, 1950-53).
Lewis, Naval power and Lewis, A. R., Naval power and trade in the Mediterranean
trade A.D. 500-1100 (Princeton, 1951).
Lewis, “Balkan ---------------, “The economic and social development of the
peninsula” Balkan peninsula during Comneni times, A.D. 1081-
1185”, Actes du II Congrès international des études du
sud-est européen (Athens, 1972), 407-15.
Lewis and Runyan, ---------------, and T. J. Runyan, European naval and maritime
Naval and maritime history, 300-1500 (Bloomington, 1985).
history
Lindsay, Early Lindsay, W. M., Studies in early mediaeval Latin
mediaeval Latin glossaries, ed. M. Lapidge (Aldershot, 1996).
glossaries
Livadas, “Medieval Livadas, G. K., “Some questions of medieval nautical
nautical technolo- technology in Kameniates’ ‘Sack of Thessaloniki’ (904
gy” AD)’, in Tzalas, Tropis III, 283-7.
Loud, Robert Guiscard Loud, G., The age of Robert Guiscard: Southern Italy and
the Norman conquest (Harlow, 2000).
Macrides, “Pen and the Macrides, R., “The pen and the sword: who wrote the
sword” Alexiad”, in T. Gouma-Peterson, ed., Anna Komnene
and her times (N.Y. and London, 2000), 63-81.
Madden, Enrico Madden, T. F., Enrico Dandolo and the rise of Venice
Dandolo (Baltimore and London, 2003).
Magdalino, Manuel I Magdalino, P., The empire of Manuel I Komnenos, 1143-
Komnenos 1180 (Cambridge, 1993).
Magdalino, “Non-juri- ---------------, “The non-juridical legislation of Leo VI”, in S.
dical legislation” Troianos, ed., Analecta Atheniensia ad ius Byzantinum
spectantia. I (Athens, 1997), 169-82.
Mahjoubi, “Nouvelle Mahjoubi, M. A., “Découverte d’une nouvelle mosaïque de
mosaique” chasse à Carthage”, Comptes rendus des séances de
l’Académie des Inscriptions et Belles lettres, (1967),
264-77.
Makris, “Griechischer Makris, G., “Ein griechischer lingua franca-text”, Studi e
lingua franca” testi, 331 (1988), 187-221.
Makrypoulias, “Byz- Makrypoulias, C. G., “Byzantine expeditions against the
antine expeditions” emirate of Crete c. 825-949”, Graeco-Arabica, 7-8
(1999-2000), 347-62.
Makrypoulias, “Muslim ---------------, “Muslim ships through Byzantine eyes”, in
ships” Yousef al-Hijji, Arab seafaring, 179-90.
Makrypoulias, “Navy” ---------------, “The navy in the works of Constantine
Porphyrogenitus”, Graeco-Arabica, 6 (1995), 152-71.
Malamut, Iles de Malamut, E., Les îles de l’Empire byzantin VIIIe-XIIe
l’Empire byzantin siècles, 2 vols (Paris, 1988).
Malamut, “Les ---------------, “Les insulaires des 10e-12e siècles: marins ou
insulaires” soldats? La démobilisation de la marine insulaire du
10e au 12e s.”, Jahrbuch der österreichischen
Byzantinistik, 32 (1982), 63-72.
Manfroni, Marina Manfroni, C., Storia della marina italiana dalle invasioni
italiana. I barbariche al Trattato di Ninfeo (anni di C. 400-1261)
(1899; rpt, Milan, 1970).
Martin, Art and Martin, L. R., The art and archaeology of Venetian ships
archaeology and boats (College Station and London, 2001).
Martin, Transport of Martin, E. E., The transport of horses by sea (Calcutta,
horses 1901).
Mazzuchi, “Basilio Mazzuchi, C., “Dagli anni di Basilio Parakimomenos (Cod.
Parakimomenos” Ambros. B 119 sup.)”, Aevum, 52 (1978), 267-318.
McGeer, Dragon’s teeth McGeer, E., Sowing the dragon’s teeth: Byzantine warfare
in the tenth century (Washington, 1995).
McGeer, “Tradition and ---------------, “Tradition and reality in the Taktika of
reality” Nikephoros Ouranos”, DOP, 45 (1991), 129-40.
Meiggs, Trees and Meiggs, R., Trees and timber in the ancient Mediterranean
timber world (Oxford, 1982).
Meinardus, “Medieval Meinardus, O., “Medieval navigation according to
navigation” akidographemata in Byzantine churches and monas-
teries”, Deltivon th'" Cristianikh'" Arcaiologikh'"
ÔEtaireiva ", per. DV, 6 (1970-72), 29-52.
Mellows “Observat- Mellows, J. S., “Observations relating to sicknesses and
ions” casualties that may be expected to occur among the
horses of the 1st Regiment of Dragoon Guards,
commanded by Colonel the honourable George
Cathcart, on the approaching voyage from Canada to
England”, The veterinary record and transactions of
the Veterinary Medical Association, 4 (1848), 101-7.
Merzagora, “Navigazi- Merzagora, M., “La navigazione in Egitto nell’età greco-
one in Egitto” romana”, Aegyptus: rivista italiana di Egittologia e di
Papirologia, 10 (1929), 105-48.
Montfaucon, Palaeo- Montfaucon, B., Palaeographia Graeca (Paris, 1708).
graphia Graeca
Morrison, Age of the Morrison, J. (S.), ed., The age of the galley: Mediterranean
galley oared vessels since pre-classical times (London, 1995).
Morrison, Greek and ---------------, Greek and Roman oared warships (Oxford,
Roman oared 1996).
warships
Morrison, “Lessons” ---------------, “Lessons from the trials of Olympias”, in Tzalas,
Tropis III, 321-5.
Morrison, “Triereis” ---------------, “Triereis: the evidence from antiquity”, in Shaw,
Trireme project, 11-20.
Morrison, “Trireme” ---------------, “The trireme”, in Morrison, Age of the galley, 49-
65.
Morrison and Coates, ---------------, and J. F. Coates, eds, An Athenian trireme
Trireme recon- reconstructed: the British sea trials of Olympias, 1987
structed (Oxford, 1989).
Morrison, et al., ---------------, J. F. Coates, and N. B. Rankov, eds, The
Athenian trireme Athenian trireme: the history and reconstruction of an
ancient Greek warship, 2nd ed. (Cambridge, 2000).
Mott, Development of Mott, L. V., The development of the rudder: a
the rudder technological tale (College Station and London, 1997).
Moutsos, “Greek Moutsos, D., “Greek CELANDION and Latin
CELANDION” CELUNDRIA”, Byzantion, 62 (1992), 402-13.
Mueller, “Venetian Mueller, R., “Venetian ships and shipbuilders before the
ships” millenium: Jal’s chelandia or the fortunes of a fake”, in
M. Arnoux and P. Monnet, eds, Le technicien dans la
cité en Europe occidentale 1250-1650 (Rome, 2004),
61-76.
Murray, “Athlit ship” Murray, W. M., “Classification of the Athlit ship: a
preliminary report”, in Casson, Athlit ram, 72-5.
Musca, Emirato di Bari Musca, G., L’emirato di Bari 847-871 (Bari, 1978).
Musso, “Armamento” Musso, G. G., “Armamento e navigazione a Genova tra il
Tre e il Quattrocento”, in Guerra e commercio
nell’evoluzione della marina genovese tra XV e XVIII
secolo (Genoa, 1973), vol. 2, 6-77.
Nadel and Bussolari, Nadel, E. R., and S. R. Bussolari, “The Daedalus project:
“Daedalus project” physiological problems and solutions”, American
scientist, 76.4 (July-August, 1988), 351-60.
Odetallah, “S4ala2h al- Odetallah, R. K., “S4ala2h al-Dı3n and the sea. The case of
Dı3n and the sea” ‘Akka2”, in Yousef al-Hijji, Arab seafaring, 207-15.
Oikonomides, Oikonomides, N., ed., Tov empovlemo Buzavntio (9o” - 12o”
Byzantium at war ai.)/Byzantium at war (9th-12th c.) (Athens, 1997).
Oikonomides, “To; kavtw ---------------, “To; kavtw ajrmamevnton”, Archeion Pontou, 26
ajrmamevnton” (1964), 193-6.
Oleson, Water-lifting Oleson, J. P., Greek and Roman mechanical water-lifting
devices devices: the history of a technology (Toronto, 1984).
Omont, Miniatures Omont, H., Miniatures des plus anciens manuscrits grecs
de la Bibliothèque Nationale du VIe au XIV siècles
(Paris, 1929).
O’Sullivan, “Arab O’Sullivan, S., “Sebeos’ account of an Arab attack on
attack” Constantinople in 654”, BMGS, 28 (2004), 67-88.
Pace, Mosaici Pace, B., I mosaici di Piazza Armerina (Rome, 1955).
Parker, Ancient Parker, A. J., Ancient shipwrecks of the Mediterranean &
shipwecks the Roman provinces [British Archaeological Reports.
International series, 580] (Oxford, 1982).
Partington, Greek Fire Partington, J. R., A history of Greek Fire and gunpowder
and gunpowder (Cambridge, 1960).
Pasch, “Fuoco greco” Pasch, G., “Il fuoco greco”, Archeologia medievale, 25
(1998), 359-68.
Peacock and Williams, Peacock, D. P. S., and D. F. Williams, Amphorae and the
Amphorae and the Roman economy: an introductory guide (London,
Roman economy 1986).
Pelekanides, OiJ Pelekanides, S. M., et al., OiJ Qhsauroiv tou' aJgivo u o[rou".
Qhsauroiv Seira; AV: Eijkonografhmevna ceirovgrafa parastavsei” -
ejpivtitla - ajrcika; gravmmata. Tovmo" BV (Athens, 1975).
Pertusi, “Ordinamenti Petrusi, A., “Ordinamenti militari, guerre in Occidente e
militari” teorie dei Bizantini (secc. VI-X)”, in Ordinamenti
militari in Occidente nell’alto medioevo [Settimane di
studio del Centro italiano di studi sull’alto medioevo,
XV], 2 vols paginated continuously (Spoleto, 1968),
vol. 2, 631-700.
Platis, “Greek crew Platis, S., “The Greek crew trials with Olympias in 1988”,
trials” in Tzalas, Tropis III, 335-45.
Polemis, The Doukai Polemis, D. I., The Doukai: a contribution to Byzantine
prosopography (London, 1968).
Polunin, Trees and Polunin, O., Trees and bushes of Europe (London, 1976).
bushes
Pryor, “Byzantium and Pryor, J. H., “Byzantium and the Sea: Byzantine fleets and
the Sea” the history of the Empire in the age of the Macedonian
emperors, c. 900-1025 CE”, in J. B. Hattendorf and R.
W. Unger, eds, War at sea in the Middle Ages and the
Renaissance (Woodbridge, 2003), 83-104.
Pryor, “Crusade of ---------------, “The Crusade of Emperor Frederick II, 1220-29:
Frederick II” the implications of the maritime evidence”, American
Neptune, 52 (1992), 113-32.
Pryor, “Eracles” ---------------, “The Eracles and William of Tyre: an interim
report”, in B. Z. Kedar, ed., The Horns of H4at6t6in
(Jerusalem and London, 1992), 270-93.
Pryor, “From dromo2n to ---------------, “From dromo2n to galea: Mediterranean bireme
galea” galleys AD 500-1300”, in Morrison, Age of the galley,
101-116.
Pryor, “Galleys of ---------------, “The galleys of Charles I of Anjou, King of
Charles I of Anjou” Sicily: ca. 1269-84”, Studies in medieval and
Renaissance history, 14 [old series, 24] (1993), 33-103.
Pryor, “Geographical ---------------, “The geographical conditions of galley
conditions” navigation in the Mediterranean”, in Morrison, Age of
the galley, 206-16.
Pryor, Geography, ---------------, Geography, technology, and war: studies in the
technology, and war maritime history of the Mediterranean, 649-1571
(Cambridge, 1988).
Pryor, “Mediterranean ---------------, “The Mediterranean round ship”, in R. W. Unger,
round ship” ed., Cogs, caravels and galleons (London, 1994), 59-
76.
Pryor, “Modelling ---------------, “Introduction: modelling Bohemond’s march to
Bohemond’s march Thessalonike2”, in idem, ed., Logistics of warfare in the
to Thessalonike2” age of the Crusades (Aldershot, 2005), 1-24.
Pryor, “Naval architec- ---------------, “The naval architecture of Crusader transport
ture” ships: a reconstruction of some archetypes for round-
hulled sailing ships”, MM, 70 (1984), 171-220, 275-92,
363-86.
Pryor, “Naval archi- ---------------, “The naval architecture of Crusader transport
tecture revisited” ships and horse transports revisited”, MM, 76 (1990),
255-73.
Pryor, “Roger of ---------------, “The naval battles of Roger of Lauria”, JMH, 9
Lauria” (1983), 179-216.
Pryor, “Rutilius ---------------, “The voyage of Rutilius Namatianus: from Rome
Namatianus” to Gaul in 417 CE”, MHR, 4 (1989), 271-80.
Pryor, “Stadiodromi- ---------------, “The Stadiodromikovn of the De Cerimoniis of
kovn” Constantine VII, Byzantine warships, and the Cretan
expedition of 949”, in J. Chrysostomides, C. Dendrinos,
and J. Harris, eds, The Greek islands and the sea:
Proceedings of the First International Colloquium held
at the Hellenic Institute, Royal Holloway, University of
London, 21-22 September 2001 (Camberley, 2004), 77-
109.
Pryor, “Transportation ---------------, “Transportation of horses by sea during the era of
of horses by sea” the Crusades: eighth century to 1285 A.D.”, MM, 68
(1982), 9-27, 103-125.
Pryor, “Types of ships” ---------------, “Types of ships and their performance
capabilities”, in R. Macrides, ed., Travel in the
Byzantine world: Papers from the Thirty-fourth Spring
Symposium of Byzantine Studies, Birmingham, April
2000 (Aldershot, 2002), 33-58.
Pryor, “Voyages of ---------------, “The voyages of Saewulf”, in Huygens,
Saewulf” Peregrinationes tres, 34-57.
Pryor, “Venetian fleet” ---------------, “The Venetian fleet for the Fourth Crusade and
the diversion of the Crusade to Constantinople”, in The
experience of Crusading. Volume one: Western
approaches, eds M. Bull and N. Housley (Cambridge,
2003), 103-23
Pryor, “Water, water ---------------, “‘Water, water everywhere, Nor any drop to
everywhere” drink’. Water supplies for the fleets of the First
Crusade”, in M. Balard, et al., eds, Dei gesta per
Francos: études sur les croisades dédiées à Jean
Richard (Aldershot, 2001), 21-8.
Purpura, “Relitto biz- Purpura, G., “Il relitto bizantino di Cefalù”, Sicilia
antino di Cefalu” archeologica, 16 (1983), 93-104.
Quand voguaient les Quand voguaient les galères: exposition du 4 octobre 1990
galères au 6 janvier 1991, Musée de la Marine Paris (Paris,
1990).
Rackham, Ancient Rackham, O., Ancient woodland: its history, vegetation
woodland and uses in England (London, 1980).
Rankov, “Reconstruct- Rankov, B., “Reconstructing the past: the operation of the
ing the past” trireme reconstruction Olympias in the light of
historical sources”, MM, 80 (1994), 131-46.
Rankov, “Rowing ---------------, “Rowing Olympias; a matter of skill”, in Shaw,
Olympias” Trireme project, 50-57.
Reddé, Mare nostrum Reddé, M., Mare nostrum: les infrastructures, le dispositif
et l’histoire de la marine miitaire sous l’Empire romain
[Bibliothèque des Écoles françaises d’Athènes et de
Rome, 260] (Rome, 1986).
Richardson, New Richardson, L., jr, A new topographical dictionary of
topographical ancient Rome (Baltimore and London, 1992).
dictionary
Rose, “Islam versus Rose, S., “Islam versus Christendom: the naval dimension,
Christendom” 1000-1600”, The Journal of military history, 63 (1999),
561-78.
Rosenberger, “Contrôle Rosenberger, B., “Le contrôle du détroit de Gibraltar aux
du détroit” XIIe-XIIIe siècles”, in M. Hamman, ed., L’Occident
musulman et l’Occident chrétien au moyen-âge (Rabat,
1995), 15-42.
Rosenthal, Vergilius Rosenthal, E., The illuminations of the Vergilius Romanus
Romanus (Cod. Vat. Lat. 3867): a stylistic and iconographical
analysis (Zürich, 1972).
Rosse, Hollinshead’s Rosse, C., and P. Gaddum-Rosse, Hollinshead’s textbook
anatomy of anatomy, 5th ed. (Philadelphia, 1997).
Rouanet, “Quatre Rouanet, G., “Étude de quatre pompes à eau romaines
pompes” provenant de l’épave Dramont D”, Cahiers
d’archéologie subaquatique, 3 (1974), 49-79.
Runciman, First Runciman, S., A history of the first Bulgarian empire
Bulgarian empire (London, 1930).
Runciman, Romanus ---------------, The emperor Romanus Lecapenus and his reign:
Lecapenus a study of tenth-century Byzantium (Cambridge, 1929).
Russo, “Fuoco marino” Russo, F., “Fuoco marino. Tra leggende e storia”, Rivista
maritima, 137 (2004), 55-68
Santamaria, “L’épave Santamaria, C., “L’épave Dramont «E» à Saint Raphaël
Dramont” (Ve siècle après J.C.)”, Archaeonautica, 13 (1995),
whole issue.
Santamaria-Arandez, Santamaria-Arandez, A., “La reconquista de las vias
“Reconquista” maritimas”, Anuario de estudios medievales, 10 (1980),
41-134.
Savvides, “Secular Savvides, G. C., “The secular prosopography of the
prosopography” Byzantine maritime theme of the Carabisians/
Cibyrraeots”, Byzantinoslavica, 59 (1998), 24-45.
Schneider, “Byzanti- Schneider, R., “Eine byzantinische Feuerwaffe”, Zeitschrift
nische Feuerwaffe” für historische Waffenkunde, 5 (1909-11), 83-6.
Schiøler, “Piston Schiøler, T, “Bronze Roman piston pumps”, History of
pumps” technology, 5 (1980), 17-38.
Senac, Musulmans et Senac, P., Musulmans et Sarrasins dans le sud de la Gaule
Sarrasins du VIIIe au XIe siècle (Paris, 1980).
Serre, Marines de Serre, P., Les marines de guerre de l’antiquité et du
guerre moyen-âge, 2 vols (Paris, 1885, 1891).
S!evc°enko, “Constan- S!evc°enko, I., “Re-reading Constantine Porphyrogenitus”,
tine Porphyrogeni- in J. Shepard and S. Franklin, eds, Byzantine
tus” diplomacy: papers from the twenty-fourth spring
symposium of Byzantine studies, Cambridge, March
1990 (Aldershot, 1992), 167-95.
Shaw, “Oar mechanics” Shaw, (J.) T., “Oar mechanics and oar power in ancient
galleys”, in Morrison, Age of the galley, 163-171.
Shaw, “Meshing” ---------------, “The meshing of oars in Olympias and in a
‘stretched’ design”, in Shaw, Trireme project, 75-7.
Shaw, “Rowing astern” ---------------, “Rowing astern”, in Shaw, Trireme project, 69-
70.
Shaw, “Steering to ram” ---------------, “Steering to ram: the diekplous and periplous”, in
Shaw, Trireme project, 99-104.
Shaw, Trireme project ---------------, ed., The trireme project: operational experience
1987-90, lessons learnt (Oxford, 1993).
Shaw, “Voyage and ---------------, “The voyage and speed trials of Olympias in
speed trials” 1990”, in Shaw, Trireme project, 39-44.
Shrader, “Handlist” Shrader, C. R., “A handlist of extant manuscripts
containing the De re militari of Flavius Vegetius
Renatus”, Scriptorium, 33 (1979), 280-305.
Shirley, Transport of Shirley, A., Remarks on the transport of cavalry and
cavalry artillery with hints for the management of horses,
before, during, and after a long sea voyage (London,
1854)
Sleeswyk and Meijer, Sleeswyk, A. W., and F. Meijer, “The water supply of the
“Water supply of the Argo and other oared ships”, MM, 84 (1998), 131-38.
Argo”
Smith, Early history Smith, F. H., The early history of veterinary literature and
its British development, 4 vols (1916-33; rpt, London,
1976).
Smith, Manual of ---------------, A manual of veterinary hygiene, 3rd ed. rev.
veterinary hygiene (London, 1905).
Steffy, “Kyrenia ship” Steffy, J. R., “The Kyrenia ship: an interim report on its
hull construction”, American journal of archaeology,
89 (1985), 71-101.
Steffy, “Medieval cargo ---------------, “The medieval cargo ship: evidence from nautical
ship” archaeology”, in C. L. Symonds, ed., New aspects of
naval history (Annapolis, 1981), 13-19.
Steffy, “Ram and bow ---------------, “The ram and bow timbers: a structural
timbers” interpretation”, in Casson, Athlit ram, 6-39.
Steffy, “Shell to ---------------, “The Mediterranean shell to skeleton transition: a
skeleton” northwest European parallel?”, in R. Reinders and P.
Kees, eds, Carvel construction technique (Oxford,
1991), 1-9.
Steffy, Wooden ship ---------------, Wooden ship building and the interpretation of
building shipwrecks (College Station, 1994).
Stephenson, Basil the Stephenson, P., The legend of Basil the Bulgar-slayer
Bulgar-slayer (Cambridge, 2003).
Stevenson, Miniature Stevenson, T. B., Miniature decoration in the Vatican
decoration Virgil: a study in late antique iconography (Tübingen,
1983).
Stratos, “Naval Stratos, A. N., “The naval engagement at Phoenix”, in A.
engagement at E. Laiou-Thomadakis, ed., Charanis studies: essays in
Phoenix” honour of Peter Charanis (New Brunswick, 1980),
229-47.
T4a2h a, Muslim conquest T4a2h a, A. D., The Muslim conquest and settlement of North
Africa and Spain (London, 1989).
Talbi, L’émirat Talbi, M., L’émirat Aghlabide 184-296/800-909: histoire
Aghlabide politique (Paris, 1966).
Thompson, Goths in Thompson, E. A., The Goths in Spain (Oxford, 1969).
Spain
Tougher, “Imperial Tougher, S., “The imperial throught-world of Leo VI: the
thought-world” non-campaigning emperor of the ninth century”, in L.
Brubaker, ed., Byzantium in the ninth century: dead or
alive. Papers from the thirtieth Spring symposium of
Byzantine studies, Birmingham, March 1996
(Aldershot, 1998), 51-60.
Tougher, Leo VI ---------------, The reign of Leo VI (886-912): politics and
people (Leiden, 1997).
Treadgold, “Army” Treadgold, W., “The army in the works of Constantine
Porphyrogenitus”, RSBN, n.s. 29 (1992), 77-162.
Treadgold, Byzantium ---------------, Byzantium and its army 284-1081 (Stanford,
and its army 1995).
Treadgold, Byzantine ---------------, The Byzantine revival 780-842 (Stanford, 1988).
revival
Trombley, “Mediterra- Trombley, F. R., “Mediterranean sea culture between
nean sea culture” Byzantium and Islam c. 600-850 A.D.”, in The dark
centuries of Byzantium (7th–9th c.)/Oi skoteinoi
Aiwne" tou Buzantiou (7o"-9o" ai.) [National Hellenic
research foundation, Institute for Byzantine research,
International Symposium 9] (Athens, 2001), 133-69.
Trombley, “War, ---------------, “War, society and popular religion in Byzantine
society and popular A natolia (6th-13th centuries”, in Byzantine Asia Minor
religion” (6th-12th cent.)/H Buzantinhv Mikrav Asiva (6o”-12o” ai.)
(Athens, 1998), 97-139.
Tsamakda, Ioannes Tsamakda, V., The illustrated chronicle of Ioannes
Skylitzes Skylitzes in Madrid (Leiden, 2002)
Tsougarakis, Byzantine Tsougarakis, D., Byzantine Crete from the 5th century to
Crete the Venetian conquest (Athens, 1988).
Tu™ma, “Puzzle of a Tu™ma, O., “The puzzle of a decline and a rise: the
decline and a rise” Byzantines and the Italians on the Sea”,
Byzantinoslavica, 53 (1992), 53-7.
Tzalas, Tropis I Tzalas, H., ed., Tropis I: First international symposium on
ship construction in antiquity, Piraeus, 30 August - 1
September 1985. Proceedings (Piraeus, 1989).
Tzalas, Tropis III ---------------, Tropis III: Third international symposium on ship
construction in antiquity, Athens 1989 (Athens, 1995).
Van Doorninck, Van Doorninck, F. H., jr, “Did tenth-century dromons have
“Waterline ram” a waterline ram? Another look at Leo Tactica,
XIX.69”, MM, 79 (1993), 387-92.
Varaldo, “Inventario” Varaldo, C., “Inventario ed armamento di una flotta di
galee a Savona nel 1476”, Atti e memorie della Società
savonese di storia patria, n.s., 14 (1980), 85-96.
Vasiliev, Byzance et les Vasiliev, A. A, Byzance et les Arabes, 2 vols in 4 parts
Arabes, (Brussels, 1935-68). Tome I: la dynastie d’Amorium
(820-867), ed. H. Grégoire and M. Canard (Brussels,
1935). Tome II, part 1: Les relations politiques de
Byzance et des Arabes à l'époque de la dynastie
Macédonienne première période: de 867 à 959
(Brussels, 1968). Tome II, part 2: La dynastie
Macédonienne (867-959): extraits des sources Arabes
(Brussels, 1950) [Volume two was “mis au jour”,
updated, by Marius Canard to 1968 and is here referred
to as Vasiliev/Canard].
Vasiliev, Russian attack ---------------, The Russian attack on Constantinople in 860
on Constantinople (Cambridge, Mass., 1951).
Viereck, Römische Viereck, H. D. L., Die römische Flotte (Herford, 1975).
Flotte
Von Falkenhausen, Von Falkenhausen, V., “Between two empires: Byzantine
“Between two Italy in the reign of Basil II”, in P. Magdalino, ed.,
empires” Byzantium in the year 1000 (Leiden and Boston, 2003),
135-59.
Waley, “Combined Waley, D., “Combined operations in Sicily, A.D. 1060-
operations” 78”, Papers of the British School at Rome, 22 [n.s., 9]
(1954), 118-25.
Wallace, “Amphora Wallace, M. B., “Progress in amphora capacities
capacities” measurement”, in J.-Y. Empereur and Y. Garlan, eds,
Recherches sur les amphores grecques (Paris, 1986),
87-94.
Wallace Matheson, Wallace Matheson, P. M., and M. B. Wallace, “Some
“Rhodian amphora Rhodian amphora capacities”, Hesperia, 51 (1982),
capacities” 293-320.
Weitzmann, Byzantine Weitzmann, K., Byzantine book illumination and ivories
book illumination (London, 1980).
Weitzmann, Sacra ---------------, The miniatures of the Sacra Parallela: Parisinus
Parallela Graecus 923 (Princeton, 1979).
Weitzmann, Studies ---------------, Studies in classical and Byzantine manuscript
illumination, ed. H. L. Kessler (Chicago and London,
1971).
Whittle, “Carabisiani” Whittle, J., “The ‘theme’ of Carabisiani”, in Christidou,
Arab-Byzantine navigation, 139-45.
Wilson, “Madrid Wilson, N. G., “The Madrid Skylitzes”, Scrittura e civiltà,
Skylitzes” 2 (1978), 209-19.
Wilson, Scholars ---------------, Scholars of Byzantium (London, 1983).
Wolfram, Goths Wolfram, H., History of the Goths, rev. ed., trans. T. J.
Dunlap (Berkeley, 1988).
Wolseley, Soldier’s Wolseley, G. J. [General Viscount], The soldier’s pocket-
pocket-book book for field service, 5th ed. (London, 1886).
Woody, “Sagena Woody, K. M., “Sagena piscatoris: Peter Damiani and the
piscatoris” Papal election decree of 1059”, Viator, 1 (1970), 33-54.
Wright, Codicological Wright, D. H., Codicological notes on the Vergilius
notes Romanus (Vat. lat. 3867) (Città del Vaticano, 1992).
Yannopoulos, “Cibyrra” Yannopoulos, P. A., “Cibyrra et Cibyrréotes”, Byzantion,
61 (1991), 520-29.
Yousef al-Hijji, Arab Yousef al-Hijji, Y., and V. Christides, eds, Aspects of Arab
seafaring seafaring: an attempt to fill in the gaps of maritime
history (Atens 2002).
Zacos, Byzantine lead Zacos, G. and A. Veglery, Byzantine lead seals, 3 parts
seals plus plates volume (Basel, 1972).
Zorzi, Biblioteca Zorzi, M., ed., Biblioteca Marciana · Venezia (Florence,
Marciana 1988).
Zuckerman, “Military Zuckerman, C., “The military compendium of Syrianus
compendium” Magister”, Jahrbuch der Österreichischen
Byzantinistik, 40 (1990), 209-24.
INDICES
The Tables of Rulers, Gazeteer, Glossaries, and Appendices One to Five and Eight are
not indexed except where there is relevant discussion in the notes.
Items which are included in the Gazeteer are asterisked (*) and those in the Glossary
of Greek, Arabic and Latin terms are marked with an obelus (†).
Abbreviations
‘Ab. ‘Abba2sid Lat. Latin (i.e., West European)

am. amı3r/amı3rate† mag. magistros†
Byz. Byzantine mag. mil. magister militum†
C. Count/County mts mountains
Cal. Caliph P. Prince/Principality
D. Duke/Duchy pat. patrikios†
droung. droungarios† prov. province
Emp. Emperor R. Rome/Roman
Fa2t6. Fa2t5imid strat. strate2gos†
gov. governor sult6. sult6a2n†
Gr. Greek th. thema/themata†
hist. history/historian Um. Umayyad
im. ima2m† Ts. Tsar
K. King/Kingdom W. West/Western
Kh. Khan
[a] General Index
‘Abba2sids, in general, 33-5, 40-41, 45, ‘Abd al-Malik ibn Abu2 ‘Amı3r
50-51, 60, 62, 72, 94 Muh5ammad al-Mans5u2r, 76, 88
------, naval forces and expeditions, 41, ‘Abd al-Malik ibn Mughı3th, Um. am.,
45, 51, 62-3, 400 42
‘Abd al-‘Azı3z, gov. Egypt, 28 ‘Abd al-Mu’min, Almohad Cal., 97-8,
‘Abd al-‘Azı3z ibn Mu2sa2, gov. al- 104, 414
Andalus*, 30 ‘Abd al-Rah5m a2n al-Gha2fiqı3, gov. al-
‘Abd Alla2h ibn al-Mu‘izz, 90 Andalus*, 30
‘Abd Alla2h ibn al-Zubayr, 27 ‘Abd al-Rah5m a2n ibn H4abı3b, Um. gov.
‘Abd Alla2h ibn Iba2d5, 35 Ifrı3qiya*, 45
‘Abd Alla2h II ibn Ibra2hı3m , Aghlabid ‘Abd al-Rah5ma2n ibn Mu‘a2wiya, al-
am., 68 Da2khil, Um. am. al-Andalus*, 33-
‘Abd Alla2h ibn Ibra2hı3m ibn Ah5mad, 4, 42
am. Sicily, 65 ‘Abd al-Rah5m a2n II, Um. am. al-
‘Abd Alla2h ibn Ish5a2q ibn Ja2m ı3, Andalus*, 42-3
admiral of Seville, 98 ‘Abd al-Rah5m a2n III, Um. am., then
‘Abd Alla2h ibn Mu2sa2, 33 Cal., al-Andalus*, 68-70, 76
‘Abd Alla2h ibn Rustam, 35 Abolita glosses, Rome, Biblioteca
‘Abd Alla2h ibn Sa‘d ibn Abı3 Sarh5, Apostolica Vaticana, MS. Vat. Lat.
gov. Egypt, 24-5 3321, 244 n. 263
‘Abd Alla2h ibn Ya2sı3n, Almoravid Abu2 ‘Abd Alla2h al-Shı3‘ı3, Fa2t6. da2‘ı3†,
missionary, 94 35, 40, 51
‘Abd al-Malik, Um. Cal., 27 Abu2 ‘Abd Alla2h Muh5ammad ibn
712 INDICES
Maymu2n, Almoravid admiral, 95- 88, 101, 149
6, 104 ------, Norman Sicilian raids into, 106,
Abu2 Ah5mad Ja‘far, Fa2t6. h5a2jib, 70 112-13, 118, 121
Abu2 ‘Amı3r Muh5ammad al-Mans5u 2r, Aeneas, ship of, 194-5 & n. 88
Um. h5a2jib al-Andalus*, 76, 88 Aeneas the Tactician, Poliorke2tikon,
Abu2 H4afs5 ‘Umar ibn ‘Iflsa2, Andalusi 176
corsair, 46 Aeolian islands, 16
Abu2-H4afs, Sicilian rebel, 90 Æthelweard, 411-12 & n. 21
Abu2 ’l-‘Abba2s, son of ‘Abd Alla2h II Africa*, R. & Byz. prov., 7 & n. 3, 8-
ibn Ibra2hı3m, 68 10, 12, 26-8 105, 126-7, 164, 259-
Abu2 ’l-Aghlab Ibra2hı3m ibn ‘Abd 60, 282, 325-6, 331, 384, 449
Alla2h, Aghlabid am., 609-10 Aga2dir, 27
Abu2 ’l-Misk Ka2fu2r, Ikhshı3did regent, Agapios, Byz. hist., 26 n. 38
50, 72-3 Agathias, Byz. hist., 197 n. 99, 219 n.
Abu2 ’l-Qa2sim, am. Sicily, 75, 168 n. 167, 279
24 Agay, shipwreck, 147
Abu2 Marwa2n ‘Ubayd Alla2h, Um am., Aghlabids, in general, 35, 40, 45, 51,
42 452
Abu2 Sa‘ı3d, Almohad gov. Ceuta, 98 ------, naval expeditions, 40, 45, 48-9,
Abu2 Ya‘qu2b Yu2suf I, Almohad Cal., 64-6, 68, 386, 392
98 ------, conquest of Sicily, 40, 45, 48, 64-
Abu2 Yu2suf Ya‘qu2b al-Mans5u 2r, 5, 70
Almohad Cal., 97 Agila, K. of Visigoths, 12
Abu2 Zar‘a T4arı3f ibn Ma2lik al-Mu‘a2firı3, Agnellus of Ravenna, Lat. hist., 13,
29 170
Abundantius, praetorian prefect†, 13- Agrippa, Marcus Vipsanius, 231-2
14, 124 n. 6, 129-30 Ah5mad al-Akhal, am. Sicily, 90
Abydos*, 62, 88, 121-2, 236, 264 n. Ah5mad al-S4iqillı3, Almohad admiral,
335, 265, 308, 335-6 & n. 513, 98
373, 636 Ah5mad ibn T4u2lu2n, gov., then am.
Achelo2on*, battle of (917), 67 Egypt, 50, 62
Achilles Tatios, Leukippe2 and Ah5mad ibn Ya‘la2, Andalusi Um. am.,
Kleitopho2n, 254 69
Achmet, son of Seire2m, Ah5mad ibn Ziya2d at Alla2h, am. Sicily,
Oneirokritikon, 171 68
Acre, 112, 115, 117-18 Ahrweiler, H., 111 n. 197, 256
------, siege of (1189-91), 118-20, 417- Aigaion Pelagos*, naval th., 47, 77,
18, 431-3 88, 256, 259, 266-7, 372 & n. 587,
Actium*, battle of (31 B.C.E.), 7, 125, 391
232 Aigina*, 47
Acts of the Apostles, 226 Ailian the Tactician, Taktike2 theo2ria,
Adalperga, Duchess Benevento, 44 177 & n. 11
Adana, 386 Aistulf, K. of Lombards, 24, 44
Adrian, pat., 65 Aizo, 42
Adrianople, 67 Akropolite2s, George, Chronike2 syn-
Adriatic Sea, 7, 13, 15-18, 32, 64, 89, graphe2, 418 & n. 53
99, 389, 408 Al-‘Abba2s ibn ‘Abd al-Mut6t6alib, 34
------, Muslim raids into, 64 Al-‘Abba2s ibn al-Fadl ibn Ya‘qu2b,
------, Venice and, 45, 67-8, 99 gov. Syria, 414
Aegates* islands, battle of (241 Al-Afd5al, Fa2t6. am. al-juyu2sh†, 88, 108
B.C.E.), 231, 351 Al-Andalus*, 50, 88, 97-8, 102-3, 256-
Aegean Sea, 46-7, 63, 76, 101, 116, 7, 259 (see also mulu2k al-t6awa2’if†)
121-2, 334 n. 507, 335-7 & n. 514, ------, Muslim conquest of, 29-30
341, 360, 374, 389 (see also ------, Umayyad am., Cal. of, 33-5, 40-
Aigaion Pelagos) 42, 68-71, 76, 452
------, Muslim raids into, 26, 47, 61-2, ------------, naval forces and expeditions,
INDICES 713
42-3, 68-70, 76 Aliulfus, saio†, 14, 124 n. 6
Al-Ant6a2ki, Yah5y a2 ibn Sa‘ı3d, Melchite Al-Kahı3n a, Berber queen/soothsayer,
hist., 414 28
Al-‘Azı3z, Fa2t6. Cal., 75 Al-Kala, 70
Alarcos*, battle of (1195), 97 Al-Ku2fa*, 34
Alaric I, K. of Visigoths, 8, 12 Al-Mahdiyya*, 51, 68, 70-71, 98, 103-
Alaric II, K. of Visigoths, 12 4
Alberic, C. Tusculum, 189 n. 64 ------, campaign of 1087 against, 93, 103
Alboin, K. of Lombards, 19 Al-Ma’mu2n, ‘Ab. Cal., 34, 46
Alcacer do Sal, 76 Al-Mans5u2r, ‘Ab. Cal., 34
Aleppo, 73 Al-Mans5u2r bi ’lla2h , Fa2t6. Cal., 71
Alexander, Byz. Emp., 67 Al-Mans5u2r/Almanzor, see Abu2 ‘Amı3r
Alexandres, K. A., 208, 229 n. 26, 239 Muh5ammad al-Mans5u2r
n. 245, 253 n. 298, 256, 258-9 n. Al-Maqrı3zı3, Muslim hist., 7 n. 1, 117
319, 286 n. 412, 312 Al-Mas‘u2dı3, Muslim hist. and
Alexandria, 24, 75, 102, 107, 109, geographer, 72
115, 117, 333-4 & n. 507, 342 n. Almeria, 35, 69-70, 76, 97-8 (see also
530 Pechina*)
Alexios, strat., 62 & n. 100 Almohads, in general, 96-8, 117
Al-Farama2’*, 47 ------, naval forces and expeditions, 97-
Al-Fayyu2m*, 165 8, 452
Alfonso I, K. Aragon, 95 Almoravids, in general, 89, 94-7
Alfonso VI, K. León/Castile, 88, 95 ------, and Jews and Christians, 95
Alfonso VIII, K. León/Castile, 97 ------, naval forces and expeditions, 95-
Al-Fust6a2t6*, 25, 109 6, 104, 452
Algeciras, 29, 43 Al-Mu‘izz, Fa2t6. Cal., 51, 70-72, 75
Al-H4akam I, Um. am. al-Andalus*, 42 Al-Muktafı3, ‘Ab. Cal., 50
Al-H4akam II, Um. Cal. al-Andalus*, Al-Muqaddası3, Ah5san, 190
76 Al-Muqtadir, ‘Ab. Cal., 51, 63-4 & n.
Al-H4a2k im, Fa2t6. Cal., 87 106
Al-H4asan, Zı3rid am., 104 Al-Muqtadir ibn Hu2d, malik†
Al-H4asan al-S4ams5a2m al-Dawla, am. Zaragoza, 89
Sicily, 90 Al-Mura2b it6u2n , see Almoravids
Al-H4asan ibn Ah5mad, Fa2t6. am. Sicily, Al-Mustans5ir, Fa2t6. Cal., 51
70, Al-Mu‘tamid, ‘Ab. Cal., 35
Al-H4asan ibn ‘Alı3, 40 Al-Mu‘tas5ı3m, ‘Ab. Cal., 35
Al-H4asan ibn ‘Alı3 al-Kalbı3, Fa2t6. am. Al-Muwah5h5idu2n, see Almohads
Sicily, 70, 72, 75 Alp-Arslan, ‘Ad5ud al-Dawla, Salju2q id
Alhucemas Bay, Morocco, 41 sult6., 94
Al-H4u rr ibn ‘Abd al-Rah5m a2n al- Al-Qa2hira, 51 (see also Cairo)
Thaqafı3, gov. al-Andalus*, 30 Al-Qa2’im,‘Ab. Cal., 94
‘Alı3, Zı3rid am., 103-4 Al-Qayrawa2n*, 28, 35, 51, 117
‘Alı3 ibn Abı3 ’l-H4usayn al-Kalbı3, am. Al-Rawd5a* island, Nile river, 25
Sicily, 71 Al-Saffa2h5, Abu2 ’l-‘Abba2s, ‘Ab. Cal.,
‘Alı3 ibn Abı3 T4a2lib, Cal., 34, 50 34
‘Alı3 ibn Ghaniya, am. Balearics, 98 Al-Samh5 ibn Malik al-Khawla2nı3, gov.
‘Alı3 ibn Ifisa2 ibn Maymu2n, Almoravid al-Andalus*, 30
admiral, 97 Al-T4abarı3, Muslim hist., 33, 63, 169,
‘Alı3 ibn Muja2hid, am. Balearics, 89 399 n. 670
‘Alı3 ibn Ni‘ma ibn al-H4awwa2s, am. Al-Tanu2khı3, Abu2 ‘Alı3 al-Muh5assin ibn
Enna, Sicily, 90, 93 ‘Alı3, Nishwa2r, 612
‘Ali ibn Yu2suf, Almoravid am. al- Al-T4artu2sı3, Murd4a2 ibn ‘Alı3 ibn Murd5a2,
Muslimı3n†, 95-6 610-11
‘Alı3, son of Zı3rid am. Tamı3m , 93 Al-Walı3d I, Um. Cal., 31, 33
‘Alı3ds, 40 (see also Shı3‘a) Al-Walı3d II, Um. Cal., 33
Aliscans, 413 Al-Z4a2hir, Fa2t6. Cal., 87
714 INDICES
Amalafrida, 14 5, 190, 192-3 & n. 86, 195-203 (&
Amalaric, K. of Visigoths, 12 nn. 96, 107, 117), 209 n. 138, 210,
Amalasuntha, Ostrogothic queen 215 n. 156, 216-25, 227-30 (& nn.
regent, 105, 326, 329 206, 207), 234, 240, 244, 248-52 &
Amalfi, 65, 75, 102-3, 166 n. 291, 253, 255, 266-70, 272-9 &
Amalric I, K. Jerusalem, 109, 114-15, n. 388, 280-82 & n. 400, 284, 325,
416 382-3, 403-4, 617-18
Ambroise, L’estoire de la guerre Antai, 19
sainte, 203 n. 124, 318-19, 331, Antalya, 32, 62, 64, 112, 190-91, 353,
432-3 372 (see also Kibyrrhaio2tai*)
‘Amma2r, brother of al-H4asan ibn ‘Alı3 ------, Gulf of, 41, 191
al Kalbı3, 72 ------------, battle of (790), 41, 385
Amplonianum primum glossary, Antenori, Obelerio degli, Doge
Erfurt, Wissenschaftliche Venice, 45
Allgemeinbibliothek, Amplon. Fol. Anthemios, R. Emp. (467-72), 124 n.
42, 244 n. 263 7
Ampurias*, C. of, 43 Antioch, 73, 110, 112, 118
’Amr ibn al-’As, gov. Egypt, 27 ------, P. of, 110, 112
Anastasios I, Byz. Emp., 13 Antioch on Cragus*, 191
Anastasios II, Byz. Emp., 31-2, 169- Anti-Photian collection, 169 n. 27
70, 608 Antonius, Marcus, 125, 231-2
Anastasius Bibliothecarius, 168 & n. Antony the Younger, St, Life of, 307-
25, 170, 172-3, 445 8, 385-6, 450
Anatolia, 31, 94, 112, 215 n. 155 (see Aphrodite2** papyri, 150 (see also
also Asia Minor) Egypt, papyri)
Anatolikon*, th., 46, 371, 391 Apion, 185
‘Anbasa ibn Suh5aym al-Kalbı3, gov. al- Apollodo2ros of Athens, 216-17 & n.
Andalus*, 29-30 160
Ancona, 17-18, 48, 113-14, 116, 420 Apollo2nios of Rhodes, Argonautika,
& n. 56 and scholia on, 194 n. 87, 197 n.
Anderson, R. C., 407 n. 1 98, 199 n. 108, 217, 226, 250-51
Andros, 111 nn. 290, 291
Angeloi emperors, 121-2, 410, 452, Apollo2nios Sophista, 185, 197 n. 96,
630-31 216-17 n. 160, 250 n. 287
Angelos, Alexios III, Byz. Emp., 122, Apostype2s, Leo, pat., commander in
630 Italy, 66
Angelos, Constantine, Byz. admiral Appenine mts, 69
(title unknown), 114 Appian, R. hist., 231, 242
Angelos, Isaac II, Byz. Emp., 121 Apsimaros, droung.† of Kibyrrhaio2-
Angevin registers, see Sicily, tai*, see Tiberios III
Kingdom of Apulia, 19, 24, 49, 65, 70, 75, 77, 91-
“Angili” of “Brittia”, 131 3, 100 (see also Italy)
Anglo-Saxons, 411-12 Aquileia*, 8, 19
Anglo-Saxon chronicle, 412 n. 21 Aquitaine, 43,
Anna, sister of Basil II, 87 ------, D. of, 68
Anonymous, Hypotheseis ek to2n ------, K. of, 68
strate2giko2n praxeo2n, 178 Aquitania Secunda*, R. prov., 12
Anonymous, Naumachika Arabs, in general, 34-5, 40
syntachthenta para Basileiou ------, in al-Andalus*, 69, 76, 88
patrikiou kai parakoimoumenou, ------, in Ifrı3qiya*, 35
183, 186, 269, 445 Arab tribes of Iraq, 34
------, sources, 185 Arados*, 393
Anonymous, the, author of Aragon, 42
Naumachika syntachthenta para ------, K. of, 96
Basileiou patrikiou kai Ardabourios, mag. mil., 8
parakoimoumenou, 134 n. 35, 184- Arduin, Milanese mercenary, 91
INDICES 715
Arethas of Caesarea*, 185 & n. 49, 291, 366
186, 228 n. 201 Athe2n aios Me2chanikos, Peri me2chane2-
Arginousai*, 340-41 mato2n, 177, 242 n. 257, 379
Argyros, Ro2m anos III, Byz. Emp., 77 Athens/Athenians, 185, 219, 290, 305,
Argyros, Eustathios, droung. tou 340 & n. 525, 351, 382, 392 n.
ploimou†, 62, 399 634, 401
Argyros, Marianos, pat., 72 ------, Assembly, 340 n. 525
Argyrus, son of Melo, katepano2†, 91-2 ------, trie2reis, 186, 201 n. 115
Arichis, D. Benevento, 44 Athlit ram, 145; Fig. 10
Aristarchos of Samothrace, 185 Atlantic Ocean, 27, 68, 71, 95
Aristippus, Henricus, 633 Atlas mts, 96
Aristophanes, Acharnenses, scholia Attaleiate2s, Michael, Byz. hist., 408
on, 276-7 n. 382, 279 n. 389 Attalus, usurper, 8
------, Peace, scholia on, 220 n. 172 Attila the Hun, 8
------, Thesmophoriazusae, scholia on, Augustine, St, 614
194 n. 87 Authari, K. of Lombards, 163
Aristotle (pseudo), Me2chanika, 226-7 Autun, 30
Arithmos†, regiment of the, 257 Avars, 19, 133
Arles, 43, 153 Avlona*, 110-11, 113
Armenia, 89 Axouch, John, megas domestikos†,
Armeniakon*, th., 391 113-14
Arsinoe*, 165 Axouch, Alexios, son of John, 114
Arsuf*, 332 Ayas*, battle of (1294), 231
Artabane2s, strat., 17-18 Ayvalik, 341
Artabasdos, Byz. Emp., 608 Ayyu2b, son of Zı3rid am. Tamı3m , 93
Artemios, St, Miracles of, 365 & n. Ayyu2bids, in general, 107-8, 118
575 ------, naval forces and expeditions, 107-
Artemios, pro2tase2kre2tis†, Byz. Emp., 8, 117-20
see Anastasios II
Asad ibn al-Fura2t, Aghlabid Babuin, A., 635 & n. 4
commander, 48 Badareno river, 170
Ascalon, 108, 109 & n. 192, 118, 332 Ba2dis*, 98
------, battle of (1099), 107 Badr al-Jama2lı3, Fa2t6. am. al-juyu2sh†, 87
Ascoli, 91 Baghdad, 34-5, 40, 94
Ashtor, E., 359 Balard, M., 360 n. 555
Asia Minor, 46, 86, 101-2, 105, 110- Balbillus, Claudius, praefectus Aegy-
11, 265, 327, 331, 329, 371, 389 pti†, 342 n. 530
(see also Anatolia) Baldwin I, K. Jerusalem, 107, 331
As5ı3la*, 43 Balearic islands, in general, 9, 11, 13,
Askle2piade2s of Myrlea, 250 n. 291 43, 96, 102
Asklepio2dotos, Techne2 taktike2, 177 ------, Muslim attacks on/conquest of,
Asmundson, Thorstein, 414 28, 33, 42-3
Aspar, son of Ardabourios, mag. mil., ------, conquest by Muja2h id al-
8 Muwaffaq, 88-9
Asser, Life of king Alfred, 424 n. 1 ------, under banu2 Gha22niya, 96, 98
Astakos*, Gulf of, battle in (1185), Baligant, am., 413, 424 n. 1
118, 121 Balkans, 67, 70, 100-11
Asti, 69 Balletto, L., 203 n. 124
Astley, P., 317 n. 468 banu2 Gha22niya, see Balearics
Asturian mts, 29 banu2 Maymu2n, Almoravid admirals,
Asturias, K. of the, 30, 42 95-7
Athanagild, Visigothic noble, 12 Barbara, St, church, St Luke of Stiris,
Athanasius II, bishop-duke Naples, 65 239 n. 247
Athe2n agoras of Syracuse, 305 Barcelona, in general, 42, 89
Athe2n aios of Naukratis, Deipnosoph- ------, Frankish March of, 42
iste2s, 197 n. 98, 198, 250-51 & n. Bardane2s, strat., see Philippikos
716 INDICES
Bardas, Caesar†, 47, 265, 327 Bevere, R., 318 n. 471
Bari, 114, 168 Bevers Saga, 414 (see also Bueve de
------, am. of, 48-9, 64-5 Hantone)
------, capital of Byz. th. of Longo- Bija2ya*, 51, 96, 98, 104
bardia*, 65, 74, 91-4, 99 Bilbays*, 120
Barqa*, 27-8 Bithynia*, R. prov., canal of, 102
Basch, L., 152 n. 77 Black Sea, 7, 76-7, 389, 609
Basegio, Marco, Venetian admiral, Blancandin et l’orguielleuse d’amour,
231 413
Basil I, Byz. Emp., 41, 47, 49-50, 62 Blatado2n, monastery, Thessalonike2,
& n. 100, 64-6, 164 n. 7, 166, 168, 239 n. 246, 419; Fig. 48
270 Blockley, R. C., 123 n. 2
Basil II, Byz. Emp., 73-4, 77, 87, 182 Bodrum museum, 362 n. 564
n. 35, 183, 414, 452 Boer War, 317 n. 468
------, Me2nologion of, 400 n. 675 Bohemond of Taranto, P. Antioch, 99,
Basil the parakoimo2m enos†, see 110-11
Lekape2nos, Basil Boio2anne22s, Basil, katepano2† of
Basilicata, 70 Longo-bardia*, 77
Basques and region, 29, 42 Bona*, 45, 90
Battle of the Masts/Dha2t al-S4awa2rı3, Bondioli, M., 434
(655), 25, 385, 390, 399, 451 Boniface, comes Africae†, 8
Beaufort, Francis, British rear admiral, Boniface, Carolingian gov. Corsica, 45
Beaufort Scale, 263, 335-6, 343, Bonifacio, 43
353 Boris I, Kh. of Bulgars, 51
Bebrycians, spring of the, 362 Boris II, Ts. of Bulgars, 73
Bede, the Venerable, Historia eccle- Bosporos, 112, 144, 240, 254, 331 &
siastica gentis Anglorum, 412 n. 21 n. 502, 340
Bedouin, 51 (see also Banu2-Hila2l, Botaneiate2s, Nike2phoros III, Byz.
Banu2-Sulaym) Emp., 86, 99
Beirut, 73, 106-8, 117, 119, 358-9, Boulgarophygon*, battle of (896), 66
418 Bouthro2ton*, 394
Belgrade, 133 Boutoumite2s, Manuel, 111
Belgrano, L. T., 424-5 Bozburun, shipwreck, 147
Belisarios, mag. mil., 10-17, 105, 125- Branas, Alexios, strat., 630
6, 131 & n. 29, 133, 153, 325-6, Branas, De2m e2trios, Byz. general, 112
329, 449 Branimir, P. Croatia, 67
Benevento, 66, 74 Brasidas, Spartan admiral, 219
------, bishop of, 166, 168 Brescia, 18
------, Lombard D./P. of, 19, 44, 49, 68, Brindisi, 48, 114
75, 92 (see also Capua-Benevento) Bringas, Joseph, parakoimo2menos†,
Benoit de Sainte-Maure, Roman de 354
Troie, 413 “Bristol” Psalter, London, British
Berbers, in general, 12, 27-8, 30, 35, Library, MS. 40731, 245
88 (see also al-Kahı3na, Kusayla/ broumalia†, festival of, 258
Kası3la ibn Lamzan) Bruzzano, 70
------, Awraba, 27 Bryennios, Nike2phoros, rebel, 86
------, in al-Andalus*, 76, 88 Bryennios, Nike2phoros, Caesar†, Byz.
------, Keta2m a, 35, 40, 51 hist., Hyle historias, 100, 110 n.
------, Mas5mu2d a, 96 196, 408-10
------, S4anha2ja, 51, 94 ------, and Alexiad of Anna Komne2n e2,
------, Zenata, 35, 40 110 n. 196, 408-10 & n. 14, 419 n.
Bernard, C. Barcelona, 42 55
Bernold of St Blasien, Lat. chronicler, Bueve de Hantone/Boeve de Haum-
101 n. 179 tone, 413
Bessarion, Cardinal, 191 Bulgarian empire, First, 41, 45-6, 51,
Bethe, H., 228 60, 66-7, 73-4, 87
INDICES 717
Bulgars, 19, 167, 271 n. 364, 307, 408, ------, naval forces and expeditions, 24-6
609 Camargue, the, 43
Burchard, Carolingian C. of the Stable, Camarina*, horse transport shipwreck,
42 312
Burgundy, D. of, 68 Camerino, 68
------, K. of, 68 Campania, 9, 49, 64-5, 72
Bury, J. B., 224 n. 186 Çanakkale, 336 n. 513
Busta Gallorum*, battle of (552), 18 Canard, M., 612
Bu2yids, 94 Çandarli, Gulf of, 341
Byrides, harbour in Thrace, 392 & n. Candiano, Pietro, supposed Doge
636 Venice, 237-8 & n. 238
Byzacena*, R. prov., 8 Candiano, Pietro I, Doge Venice, 67,
Byzantion*, 339-40 237 n. 238
Byzantium/Byzantines, in general, 14- Candiano, Pietro II, Doge Venice, 237
15, 24-5, 87, 89, 99 n. 238
------, in south Italy, 18-20, 25-6, 43-4 & Candiano, Pietro III, Doge Venice, 67,
n. 77, 49, 66, 69, 71-2, 75, 77, 91- 237 n. 238
2, 94 Candiano, Pietro IV, Doge Venice,
------, intelligence systems, 393 & n. 237 n. 238
642, 394, 451 Cannae*, battle of (1018), 77, 91
------, naval forces/populations, in Cape Bon, battle off (468), 9, 451
general, 27, 32, 76-7, 86-7, 89-90, Capitanata*, 77, 91-2
100-101, 109-13, 116, 120-22, Capua-Benevento, Lombard P., 69,
370, 384-93, 403, 418-19, 444, 451 71-2, 74-5, 91-2
(see also Aigaion Pelagos, Carcassonne, 30
basilikon plo2imon, Karabisianoi, Carmen in victoriam Pisanorum, 103
Kibyrrhaio2tai, Samos) n. 182
------------, mercenaries in, 110, 121-2 Carolingians, 42-4, 452
------------, naval/military services, 89-90 ------, naval expeditions, 42-3, 45
------, naval expeditions (from Justinian Cartagena, 8-9, 12
I), 10-12, 14-18, 24-5, 27-8, 31, Cartenna*, 35
33, 44-9, 61-2, 64, 66, 68-9, 72-3, Carter, Michael, 612 n. 15
75-6, 77/86, 90, 100, 105, 109-14, Carthage*/Carthaginians 8-9, 14, 18,
115-18, 120, 125-6, 166, 260, 323, 28, 45, 193-4 n. 86, 231, 325, 384,
325-6, 383-5, 415, 452 (see also 388 n. 624, 400
Crete) ------, Dermech mosaic, 221-2 n. 177,
309-10; Fig. 35
Cadiz, 95, 97 ------, fishing boat mosaic, 245
Caesar, Julius, 331 C!aslav Klonimirovic°, P. Serbia, 67
------, Gallic War, 146, 230 n. 208 Casoni, G., 237
Caesarius of Arles, St, Life of, 153 Caspian Sea, 94
Caffaro di Caschifellone, Genoese Cassiodorus, 13-14, 124 & n. 6, 125 n.
consul, 96, 105, 424-6 8, 129-30
Cagliari, 18 Casson, L., 159, 226, 227 n. 198, 229,
------, church of St Saturninus, 124-5 230 n. 208
Cairo, 51, 75, 109, 115, 120 Castile, K. of, 96
------, genı3za† of, 102, 363 Castronuovo*, 93
Calabria, 14, 17, 19, 24, 72, 75, 91-2, Catti, G., 237
122, 256-7, 259 (see also Italy) Catulus, Gaius Lutatius, R. admiral,
------, Aghlabid attacks on, 48, 65, 68 351
------, Byz. strat. of, 68 Cefalù, shipwreck, 151 n.76
------, Sicilian attacks on, 70, 71 Cerami*, battle of (1063), 93
------, Almoravid attack on, 95, 104 Cetara*, 65
Calatrava*, 96 Cetina river, 67
Caliphate, Muslim (see also Ceuta, 13, 51, 69-71, 76, 95, 97-8
Umayyads, ‘Abba2sids) Chalke2do2n*, 331 & n. 502
718 INDICES
Chandax*, 47, 266, 333, 371 Clover, F. M., 123 n. 2
Chanson de Guillaume, 413 Clovis, K. of Franks, 12
Chanson de Roland, 413 & n. 25, 424 Coates, J., 143 n. 57, 201 n. 115, 263
n. 1 n. 334, 280 n. 397, 338 & n. 521,
Charlemagne, Carolingian K., W. 355-6, 432 n. 14, 434 n. 18
Emp., 42, 44-5 ------, S. K. Platis & J. T. Shaw, 338 n.
Charles Martel, Merovingian Mayor of 521
the Palace, 30, 43 Comacchio, 64
Charles I of Anjou, K. Sicily, galleys Conrad II, W. Emp., 91
of, 138, 203, 217 n. 162, 275, 312- Conrad III, K. of the Germans, 112
14, 322, 430, 434 n. 19, 436 n. 22; Constans II, Byz. Emp., 25-6, 43, 385,
Fig. 56 (see also Sicily, Angevin 390, 399
K. of) Constantine I, R. Emp. (306-37), 7 &
Charles II of Anjou, K. Sicily, 318 n. n. 3, 130, 132
470 Constantine IV, Byz. Emp., 25, 169
Charles the Bald, K. Aquitaine, W. Constantine V, Byz. Emp., 166-7, 307,
Emp., 154 408, 609
Charroi de Nimes, 413 Constantine VI, Byz. Emp., 44
Chelidonia*, Cape, battle/storm off Constantine VII, Byz. Emp., 62, 71-2
(842), 47, 385 & n. 128, 150, 183-4, 186-7, 188 n.
Cherso2n*, 31, 167, 197 n. 99, 279, 353 62, 191, 242 n. 257, 554 n. 7, 556
------, th., 391 n. 11, 617, 631
Chiessi, shipwreck, 366 ------, De administrando imperio, 164 n.
Childebert II, K. of Franks, 163 7, 186, 188, 256-7, 270-71 & n.
Chios, 88, 111, 116, 264 n. 335, 265, 364, 281 n. 400, 324 n. 483, 607,
336, 340-41, 373, 376, 452 609, 630
------, battle north of (912), 63, 385-6 ------, De cerimoniis, 63, 150, 187-8,
Cho2niate2s, Michael, Ta So2zomena, 419 258, 391 (see also Crete)
n. 54 ------------, Leipzig, Univ./Urb., MS. 28
Cho2niate2s, Nike2tas, Byz. hist., 111 & [Rep.i.17], 187
n. 198, 112-17, 121-2, 410-11 & n. ------, Excerpta, 241-2
17, 416-18, 630-31 ------, Praecepta historica, 186
Chouroup, Byz. naval commander, ------, Praecepta imperatori Romano
113 bellum cogitanti ... observanda,
Chrétien de Troyes, Cligés, 413 178, 186, 188, 191 n. 78, 215 n.
Christiana*, Ta, 264 n. 335, 265 155, 367 n. 579
Christides, V., 187 n. 57, 207 n. 131, ------, treatise on weather and seasonal
236, 239 & n. 246, 407 n. 1 navigation for, 191 & n. 78, 392
Christodoulos, Sicilian admiral, 104 Constantine IX, Byz. Emp., 87-9, 91-2
Chronicle of Alfonso III, 26 Constantine, Pope, 43
Chronicle of the Morea, 420 Constantine Gongyle2s, he2ge2tor
Chronicle of the Tocco, 420 naumachias, 71 & n. 128, 187
Chronicon Salernitanum, 168 n. 25 Constantinople, 8, 11, 17, 19, 24-6, 28,
Chrysopolis*, 331 n. 502 32, 43, 47, 66-8, 73-4, 76-7, 90-92,
Chuka Channel, 146 94, 99-101, 105, 109-10, 112-17,
Cilicia*, 26, 31, 50, 60, 73, 112, 452 120, 133, 135, 140, 150, 157, 163-
------, Muslim naval forces, 62-4, 386, 4, 170, 172, 175, 185, 188 n. 62,
403, 452 216, 225, 241-3, 256, 259, 264-6,
cista Ficoronica, 362 & n. 562, 368; 308, 311-12, 326-7, 331, 333-4 (&
Fig. 43 nn. 507, 510), 335-7, 353-4, 358,
Civitate*, battle of (1053), 92 371-4, 390, 393, 410, 416-17, 630-
Civitavecchia, 43 31, 633, 639 (see also Golden
Classe*, port of Ravenna, 15, 19, 24 Horn)
Cleopatra, 403 n. 689 ------, Bulgarian attacks on, 45-6, 66-7
Clermont, Council of (1095), 101-2 & ------, Great Palace, Chalke2 entrance
n. 179 vestibule, 32
INDICES 719
------, Magyar attack on, 70 303, 305-6, 308-9, 327, 329, 331.
------, Rho2s† attacks on, 60, 66-7, 72, 354, 370-72, 390, 396, 408, 449-
87, 189, 384-5, 452, 618, 622, 630- 50, 452
31 ------, De cerimoniis inventories for
------, Muslim attacks on, 26-7, 31-2, 47, ------------, (911), 150, 186-7, 190-92,
149, 169-70, 225, 242, 383, 385, 260-61, 286, 305-6, 309, 363, 407,
451, 607-8 446
Constantius, R. Emp. (337-61), 7 ------------, (949), 150, 186-7, 189, 191,
Constantius, mag. mil., 8 208-14, 221, 230, 244, 246 & n.
Contarini, Domenico, Doge Venice, 271, 247-8, 255-7, 259-62, 264,
99 266-9, 275, 280-81, 284-6, 305,
Cordoba, 40, 43, 68, 71 361-3, 371-2, 379-80, 395, 398,
Corfu, 18, 99, 106, 111-13, 332-3 (see 403-4, 407, 446, 618, 622, 624,
also Korkyra) 626
------, battle of (1084), 100, 410 & n. 14 Croats/Croatia, 19, 49, 67, 99 (see also
------, Strait, 392 n. 634 Branimir, Tomislav)
Corinth/Corinthians Crotone, 16-18, 326
------, ancient, 219, 223, 392 n. 634 Crusader states, 107-8, 117-20, 453
------, Gulf of, 67 Crusades, 305, 319 n. 476
------------, battle of (879), 61, 378 n. 592, ------, First, 94, 101-3, 105, 109-10, 112,
385, 392 409, 621
------, isthmus of, 392 ------, of Bohemond of Taranto, 110
------, medieval, 107, 410 ------, of Frederick II, 106
------------, ship on plate, 245 ------, of Louis IX, 314, 327
corsairs, in general, 45, 88, 111 n. 198 ------, of C. Louis of Clermont, 328
------, Andalusi, 43, 46, 452 ------, Second, 106, 112-13
------, Balearic, 89, 96, 98 ------, Third, 106, 118-20, 283, 319, 417,
------, Cilician, 62-4, 72 431-3 (see also Acre, siege of)
------, Cretan, 47, 61, 64, 190 ------, Fourth, 120, 122, 243, 311, 417-
------, Fa2t6imid, 108 18, 630
------, Neretljani, 67 ------, Sixth, 328, 612
------, of Fraxinetum*, 69 ------, Venetian (of 1122-3), 106, 108,
------, Sicilian, 64, 70 111, 332-3
------, Vandal, 12 ------, Northern fleets, 97, 106, 1 19
------, Western, 121 Cutty Sark, clipper ship, 342 n. 530
------, Zı3rid, 90 Cyclades islands, 47, 88, 111
Corsica, in general, 11, 32 ------, fleet of, 32, 608
------, Gothic attack on, 18, Cyprus, 47, 63-4, 109-10, 112, 120,
------, Muslim attacks on, 40, 42-3, 45 319, 331, 334, 358, 385-6, 400,
------, Vandal attack on, 9 418
Cosentino, S., 180 ------, battles off (747 & 965), 32-3, 73,
Cosenza, 68 385
Cotterell, B. & J. Kamminga, 338 n. ------, Byzantine reconquests, 62 & n.
521 100, 73, 452
Couronnement de Louis, Le, 413 ------, covenantry status of, 25, 33, 42-3,
Covadonga*, battle of (718/22), 29 63 (see also ‘ahd†)
Crete, 27-8, 63-4, 106, 109, 118, 232, ------, Muslim attacks on, 63, 393
264, 325, 333, 354-5, 389 ------, Muslim subjection of, 25-7, 33,
------, Muslim attacks on, 25-7, 33, 42 41-2
------, Muslim conquest of, 46-8, 452 Cyril Greek-Latin glosses, British
------, Muslims of, 47, 61, 63-4, 184, Library, MS. Harley 5792, 220 n.
190, 378 n. 592, 385, 390 n. 629, 172, 220 n. 172, 221, 251 n. 294,
392 254 n. 302, 272 n. 370, 367 n. 578
------, Byzantine expeditions against,
46-7, 63 & n. 105, 71-2, 105, 181, Dacia*, R. prov., 19, 136-7
184, 188, 190-91, 257, 259, 262-9, Daedalus Project, 355, 357 & n. 547
720 INDICES
Dain, A., 175, 180-83, 218 n. 163, Dio Cassius, R. hist., 231-2, 331
249, 455 n. 1, 483 n. 1, 521 n. 1, Diodo2ros Siculus, Gr. hist., 358
571 n. 1 et passim in the notes to Diogene2s, Ro2m anos IV, Byz. Emp.,
Appendices 1-3, 5 89-90, 94
Dalasse2nos, Constantine, admiral, 101 Diogene2s, kome2s†, 66, 166
Dalmatia, 61, 67, 256-7, 259 Dioskoride2s, Pedanius, De materia
------, th., 391 medica, 150, 629-30; Fig. 11
Dalmatian coast, 67-8 Dnepr river, 60, 73
Damascus, 26, 50, 73, 112 Dolley, R. H., 204, 229 n. 206, 231 n.
Damianos, am. Tarsos, 63-4 213, 234-5, 238 & n. 242, 239-40,
Damietta, 109, 115, 117, 612 241 n. 252, 292 n. 424, 394-5, 405
------, Byz. attacks on, 33, 47, 115, 117. n. 697
169 D’Oria, Conrad, Genoese admiral, 236
Dandolo, Andrea, Doge Venice, Lat. n. 237
Hist., 113 Dorostolon*, 73
Dandolo, Enrico, Doge Venice, 122 Dositheus, Hermeneumata Monacen-
Danes, 412 sia, Munich, Bayerische Staatsbib-
Danube, river, 19, 89, 133, 384 liothek, Cod. Monac. Lat. 13002,
Daphnopate2s, Theodore, eparchos† of 135 n. 38, 251 n. 293
the city, Byz. hist., 71 n. 128, 188 ------, Hermeneumata Vaticana, Rome,
& n. 62, 189, 354 Biblioteca Apostolica Vaticana,
Dardanelles, 8, 47, 62, 116, 118, 121, MS. Vat. Lat. 6925, 251 n. 294
334 n. 507, 335, 340-41 Dotson, J., 435 n. 21
------, battle of (324), 7, 130, 132 Doukaina, Eire2n e2, 99
Darna*, 28 Doukas, Constantine, 99
David, Symeon, and George, bishops Doukas, Constantine, at Ancona, 420
of Mityle2ne2, 172 & n. 56
“Day of the Apostle Thomas”, battle Doukas, John, megas doux†, 101, 109
on the, 63, 385 Doukas, Michael VII, Byz. Emp., 94,
De Boislisle, A., 328 n. 495 99
De cerimoniis, see Constantine VII Drako2n*, river, 18
De Chaumont, F., 330 & n. 500, 356 Dramont D, shipwreck, 623
De Goeje, M. J., 612 Dramont E, shipwreck, 147
Del Giudice, G., 314 n. 457 Dubrovnik, 49, 64 (see also Ragusa*)
De2m e2trias*, 394 Du Cange, C., 281-2 n. 400
De2m e2trios I Poliorke2te2s, later K. Duklja*, 67
Macedonia, 358 Dvornik, F., 389 n. 628
De2m e2trios, strat., 15 Dyrrachion*, 13-14, 16, 67, 99-100,
De2m e2trios of Thessalonike2, St, 110-11, 256-7, 259, 612
miracles of, 163 ------, th., 67
De2mokleitos, 399
Denia, 98 Ebro river, 43, 96
------, and the Balearics, taifa mamlaka† ------, battle of (217 B.C.E.), 388 n. 624
of, 88-9 Echimos, John, see Antony the
Dennis, G. T., 381 Younger, St
De obsidione toleranda, 241-2, 617 Edremit, 31
Desiderius, K. of Lombards, pretender, ------, Gulf of, 341
44, 166 Eg°ribucak Point, 341
Develtos*, 609, 622 Egypt, 7 n. 3, 24, 27-8, 40, 47, 50-51,
Dha2t al-Himmah, 41 74-5, 88, 117-20, 158-9, 342, 358,
Dhu2 ’l-Nu2nid taifa mamlaka† of 363, 612 (see also Ayyubids,
Toledo, 88 Fa2t6imids, Ikhshı3dids, Mamlu2ks,
Dia, 264 n. 335, 265, 373 T4u2lu2nids)
Didymos of Alexandria, 216 ------, Byzantine attacks on, 24, 33, 47,
Digene2s Akritas, 41 72, 75, 77, 109, 112, 115, 117, 415
Dindorf, G., 228 n. 201 ------, Crusader & Frankish attacks on,
INDICES 721
109, 114-15, 117, 120, 415, 612 ------, in Ifrı3qiya*, 51, 68-71, 74-5
------, naval forces and expeditions, 26, ------, in Sicily, 70
31-3, 47, 62, 73, 87, 107-9, 117- ------, naval forces, 51, 68-9, 75, 87-8,
20, 164-5, 384-6, 393, 431 107-9
------, papyri, 123 n. 1, 128, 150, 164-5 ------, naval expeditions, 51, 68-70, 72,
(see also Aphrodite2* , Oxyrhyn- 74-5, 88, 107-9, 358
chus*) Fez, 40
------, Sicilian Norman attacks on, 107, Ficoroni, F., 362 n. 562
114, 117 Filangieri, R., 436, n. 22
Eickhoff, E., 229 n. 206, 234 & n. 228, Fin d’Elias, La, 413
235, 238, 256, 261 n. 327 Fink, H. S. & F. R. Ryan, 332 n. 504
Eidikon†, Department of the, 210, Flavius Basiliskos, mag. mil., 9, 127,
246-8, 361-4 451
Eire2ne2, Byz. Empress, 44 Flavius Eutropius, 178 n. 15
Ekkehard of Aura, Lat. hist., 101 & n. Florence, Biblioteca Medicea
179 Laurenziana, MS. Laurentianus
Eknomos*, battle of (256 B.C.E.), 400 LV-4, 176-80
Elba, 366 Florence, Biblioteca Riccardiana, Cod.
El Cid, see Rodrigo Díaz de Vivar 217, Arabic-Latin dictionary, 154
Elias the Younger, St, Life of, 394 n. 82
Ellis Davidson, H. R., 383 n. 613 Foniás, Cape, 376
Elvira*, 69 Fortore river, 92
Engelberga, Carolingian Empress, 166 Fragmentum historiae monasterii-novi
English Channel, 146 Pictaviensis, 102 n. 179
Enna, 48, 64, 90, 93, 414 France/the French, 101, 103, 166
Eparchos, Antonius, 183 Franks, in general, 24, 30, 42-3 (see
Epidauros* (of Argolis), 352 also Carolingians, Merovingians)
Epidauros* (of Illyricum), 14 ------, in Italy, 41, 44 & n. 77, 163-4
Epiros*, 9, 394 ------, in K. of Jerusalem, 107-9, 331-2
Eracles, Old French trans. Of William ------, in Languedoc and Provence, 19,
of Tyre, 286, 417 30
Ernoul, continuator of William of ------, in Spain, 42
Tyre, 418 Fravitta, mag. mil., 8
Erzurum, 628, 631 Fraxinetum*, 69, 72
Escorial Taktikon, 77, 391 Frederick I, Barbarossa, W. Emp.,
Esphigmenou, Mount Athos, Cod. 14, 116-17
Me2nologion, 245 Friuli, margraves of, 69-70
Estopañan, S. C., 633 frontiers, Byzantine/Muslim, 31, 33,
Eunapios of Sardis, R. hist, 123, 127 41, 60
Euphe2mios, tourmarche2s† in Sicily, Fulcher of Chartres, Lat. Hist., 108,
48 331-3, 358
Euripides, Helen, 226
Euripos*, 62, 116, 620-21 Gabes, 104
Eustathios of Thessalonike2, 282, 411 Gaeta, 49, 65, 68, 75
Eustathios, strat. Of Calabria, 68, 270- Gafforio, Genoese corsair, 122
71 & n. 363 Gainas, mag. mil., 8
Excerpta historica, see Constantine Gaiseric, K. of Vandals, 8-10, 12, 124
VII n. 7
Galata*, tower of, 122
Famagusta, 358-9 Galatia*, 31
Farwald II, D. Spoleto, 24 Galatision, church, Athens, 239 n. 247
Fasti Vindobonenses priores, 13 Galen, 224 & n. 186
Fa2t6ima, daughter of Muh5ammad, 51 Galerius, Gaius, praefectus Aegypti†,
Fa2t6imids, in general, 35, 40, 50-51, 68, 342 n. 530
107 Galicia, 12, 68-9
------, in Egypt, 51, 75-6, 87-8 Garigliano river, Muslim corsair nest,
722 INDICES
65, 68 Great Zab, river, battle of (750), 34
Gascony, D. of, 68 Greece, western, battle off (880), 66,
Gaudiosus, dromonarius, 124-5 & n. 8 385, 392
Gaza, 358 Gregory I, the Great, Pope, 164
Gelimer, K. of Vandals, 11, 126 Gregory III, Pope, 32
Genesios, Byz. hist., 609 Gregory, Byz. strat. Bari, 65, 168
Genoa/Genoese, 97 Gregory, spatharios†, 66, 166
------, Annales Ianuenses, 407 n. 1 Gregory of Dekapolis, St, Life of, 394
------------, (Paris, Bibliothèque Nationale, Gregory of Nazianzos, St, Sermons of
MS. Suppl. Lat. 773), 424-31, 644; (Paris, Bibliothèque Nationale,
Fig. 50 MS. Gr. 510), 157-8, 245; Fig. 15
------, aqueduct on docks, 373 Gregory of Nazianzos, St, Sermons of
------, sacked by Fa2t6imids, 70 (Mount Athos, Pantelee2mon, Cod.
------, naval forces and expeditions, 88- 6), 245, 300, 637; Fig. 47
9, 95-6, 102-6, 110-11, 119, 122, Grettis saga Ásmundarsonar, 414
231, 314, 358-9 Grimoald, D. Benevento, 44
------, quarter in Constantinople, 115 Grumel, V., 168 n. 27
George of Antioch, Sicilian admiral, Guadalete river, battle of (711), 29
103-4, 106, 113 Guadalquivir river, 43
George Hamartolos, Byz. hist., 271 n. Guadiana river, 96
364, 419 n. 55 Guaimar IV, P. Salerno, 91, 92
George, imperial legate in S Italy, 166 Gula2m Zura2fa, see Leo of Tripoli
Georgius, Byz. “stratigus” in Italy, Gundobad, K. of Burgundians, 153
168 Guy of Lusignan, K. Jerusalem, 118
Gepids, 19
Gerace, 71 H4abla, freedman of Abu2 ‘Iqa2l al-
Germanicus Caesar, Aratea, 221-2 n. Aghlab, 48
177; Fig. 21 Hadrian I, Pope, 44
Germanos, Byz. general, 17 Hadrumetum*, 27, 137 & n. 48, 250;
Germany, 101, 413 Figs 3, 24 (see also Su2sa)
Gerona, 42 Hagenmeyer, H., 332 nn. 504-5
Gesalec, Visigothic claimant, 14 Haifa, 119
Gha2nim ibn Muh5ammad ibn Marda- Haldon, J., 214-5 & n. 154, 256, 263
nı3sh, Almohad admiral, 98 n. 334, 327 n. 493, 365, 367 n. 580,
Ghassa2nı3†, Arabs, 69 371, 372 n. 587, 380 n. 599, 381,
Ghuzz, see Turks 554 n. 8, 561 n. 27, 562 n. 32, 564
Gibraltar, 12 n. 41, 569 n. 51, 582 n. 17
------, Straits of, 13, 29, 68, 97 ------, De cerimoniis, II.44-5, 187, 372
Gildo, comes Africae†, 7 n. 587
Gisolf I, P. Salerno, 74 ------, Greek Fire, 607 n. 1, 621, 625-8
Glossae Aynardi (Metz, Bibliothèque & n. 69; Fig. 61
Publique, Cod. Metensis 500), 244 ------, & M. Byrne, 628
n. 263 H4amda2nids, 50
Glossae Nonii (Leiden, Bibliothek der H4amma2d ids, Algerian Zı2rı3ds, 51, 96,
Rijksuniversiteit, MS. BPL 67F), 104
272 n. 370 H4ammu2d ids of Malaga, 88
Golden Horn, 31, 121-2, 167, 240, harbour chains (Acre, Golden Horn,
403, 631 Palermo), 31, 93, 103, 119, 122
------, chain of, 31 Harmathous*, 341
Goths, of Crimea, 7 Harris, J., 152 n. 78
Grabar, A. and M. Manoussacas, 633 Hartel, Helmar, 614 n. 24
& n. 2, 635, 638-9, 641-3 Ha2ru2n al-Rashı3d, ‘Ab. Cal., 34, 40-42,
Grado*, 64 H4asan al-Kalbı3, 72
Granada, taifa mamlaka† of, 95 Hasdrubal, 388 n. 624
Grand Congloué, shipwreck, 146 n. H4assa22n ibn al-Nu‘ma2n al-Ghassa2nı3, 28
63, 148 n. 69 H4at6t6ı3n , Horns of*, battle of (1187),
INDICES 723
118 Hocker, F. M., 147 n. 65, 256
Hauteville, family (see also Robert Holy Apostles, church, Thessalonike2,
Guiscard, Roger I) 239 n. 247
------, Drogo, Norman C., 91-2 Homer, Iliad, and scholia on, 185-6,
------, Humphrey, Norman C., 91 217, 250 & nn. 287, 289
------, William, Norman C., 91 ------, Odyssey, and scholia on, 185-6,
H4aydara2n*, battle of (1052), 51 194, 196, 197-8 (& nn. 99, 102),
Hayes, M. H., captain, 317 n. 467, 319 199 n. 107, 203, 216, 249, 251, 282
n. 473, 320 n. 477, 321 n. 479, 324 Honorius I, R. Emp. (395-423), 8
n. 482, 327 n. 493 Howard-Johnston, J., 110 n. 196, 410
Helena, wife of Constantine VII, 556 n. 14, 419 n. 55
n. 11 Hugh, C. St Pol, 243
Heliodo2ros of Emesa, romance author, Hugh of Arles, K. Italy, 72
131 Hugh of Provence, K. Italy, 324
Heliopolis* of Egypt, 608 Hunayn*, 98
Heliopolis* of Syria, 607-8 Hungary/Hungarians, 99
Hellas*, th., fleet of, 31-2, 62, 608, Huns, 197 n. 99, 279
620 H4usa2m al-Dı3n Lu’lu al-Mas‘u2dı3,
Henry III, W. Emp., 92 h5a2jib†, Ayyu2bid admiral, 118-19
Hephaisteion (St George), church, H4usayn ibn ‘Alı3, Shı3‘a† im., 34
Athens, 239 n. 247 Huxley, G., 264 n. 335
Heraclianus, comes Africae†, 8 Hyland, A., 320 n. 478, 322 n. 480,
He2rakleia*, 133, 264 n. 335, 265, 325, 327 n. 493
335 & n. 512, 373
He2rakleia Pontike2* , 339-40 Iba2d5ı3yya, 35 (see also Khawa2rij†)
He2rakleios I, Byz. Emp., 24, 282 Iberia*, th. and army of, 89
He2rakleios of Edessa, mag. mil., 9 Ibn abı3 Zar‘, Muslim hist, 98
Herodotos, 133, 186, 382 Ibn al-Athı3r, Muslim hist., 7 n. 1, 48
He2ro2n of Alexandria, Pneumatika, (& nn. 87, 88), 68, 90, 98, 414,
623-4 609-10
He2ro2n (of Byzantium), Parangelmata Ibn al-Qala2nisı3, Muslim hist. 108, 109
poliorke2tika, 177, 242 & n. 257, n. 192
363 & n. 567, 380, 620, 624, 627; Ibn H4awqal, S5u2rat al-Ard 5, 169, 190
Figs 23, 59 Ibn ‘Idha2rı3, Muslim hist., 7 n. 1, 95, 98
Hesiod, 201 Ibn Khaldu2n, Muslim hist., 7 n. 1, 41,
Hesseling, D. C., 228 43 n. 72, 48 & n. 88, 69, 95, 98,
Hesychios of Alexandria, Lexicon, 103
131, 185, 196, 197 n. 96, 198, 199 Ibn Mankalı3, Muh5ammad, Al-ah5ka2m
n. 107, 200-202, 216, 218-21, 224- al mulu2kı3yya, 161 n. 92, 192, 207,
7 & n. 199, 234, 240, 244, 250 & 255 n. 306, 272 n. 367, 395 n. 652,
n. 287, 279, 282-3 404 nn. 693 & 696, 405 nn. 697-8,
Hexamilite2s, Basil, strat. Kibyrrha- 611
io2tai*, 72, 385, 399-400, 452 ------, Al-adilla al-rasmiyya, 235-6, 611
Hierax*, 65 & n. 112 Ibn Mardanı3sh, Abu2 ‘Abd Alla2h
Hiero2 II of Syracuse, great ship of, Muh5ammad ibn Sa‘d, taifa malik†
198-9 & n. 104, 366 Valencia, 97
H4ija2z, 40, 50-51 Ibn Ta2fratust, qa2’id†, Almoravid naval
Hilderic, K. of Vandals, 14 commander, 96
Himerios, pat. and logothete2s tou Ibra2hı3m I ibn al-Aghlab, Aghlabid
dromou†, 63 & n. 103, 186, 333, am., 45
385-7 & n. 620, 452 Ibra2hı3m II, Aghlabid am., 65
hippodrome, 257 Iceland, 413
Hisha2m I, Um. am. al-Andalus*, 42 Iconoclasm, 32, 46, 170-71 & n. 34,
Hisha2m II, Um. Cal. al-Andalus*, 76 385-6
Historia ducum Veneticorum, Lat. Idrı3sids, 35, 40-41, 51, 76
hist., 113 Ifrı3qiya*, 27-8, 33, 35, 40, 45, 51, 68,
724 INDICES
70, 72, 75, 94, 97-8, 103-4, 106,
117, 414, 452 Jabal T4a2riq, 29 (see also Gibraltar)
Ignatios, Patriarch, 169 n. 27, 172 Jaffa, 107-8
Igor, P. Kiev, 384 Jal, A., 196 n. 93, 201 n. 115, 209 n.
Ikhshı3dids, 50, 72 136, 210 n. 144, 217 & n. 162,
------, naval expeditions, 73 228, 236-8, 244, 367 n. 579
Ile d’Or, Provence, 147 Jaroslav I, P. Kiev, 87, 384
Ilias Ambrosiana (Milan, Biblioteca Jason and the Argonauts, 362; Fig. 43
Ambrosiana, Cod. Ambros. F. 205 Jawhar al-S4aqlabı3, Fa2t6. general, 51,
Inf.), 135-6 & n. 40, 139-40, 142, 70-71
152, 155-7, 282, 299; Figs 6, 7, 13, Jeffreys, M., 312 n. 456
14 Jenkins, R. J. H., 256
ima2m†, of Iba2d5ı3yya, 35 Jerba, 104
ima2ms†, Shı3‘a†, 34, 50 Jerome, St, 227
Indoulph, 17-18 Jerusalem, 91, 107, 118, 332
Innocentius, praetorian prefect† of ------, K. of, 107, 111, 115, 118, 331-2,
Africa*, 164 358
intelligence systems, 181, 393 & n. Jews, 70, 95
642, 394 John I Tzimiske2s, strat. Anatolikon*,
Ionian islands, 105-6, 110 Byz. Emp., 73-4, 77, 183-4
Ionian Sea, 61, 65 & n. 113, 112 John III Doukas Vatatze2s, Emp.
Ios, 264 n. 335, 265, 331, 333, 371, Nicaea*, 418 n. 53
373, 376 John VIII, Pope, 65, 166, 168
Iraq, 34-5 John XI, Pope, 189
Irmingar, C. Ampurias*, 43 John of Cappadocia, praetorian
Ischia, 45 prefect†, 11
Isidore of Seville, St, Lat. hist. & John of Damascus, St, see Sacra
encyclopedist, 13, 126 & n. 14, Parallela
126, 128, 134-5, 166, 244 n. 263, John, Byz. general, nephew of
272 n. 370, 280 n. 394, 411, 412 n. Vitalian, son-in-law of Germanos,
21, 415 n. 43, 624 16-18, 451
Isma2‘ı3l, Shı3‘a† im., 50 John, logothete2s†, 44
Istanbul, 334 n. 510, 335 n. 512, 340 John, pro2tospatharios† and ase2kre2tis†,
n. 523 (see also Constantinople) 259
------, Turkish Naval Museum, 216 n. John, pat., 28, 325, 384
156 John, Byz. general in Libya, 18
Italy, in general, 14-19, 41, 43-5, 70, John, D. Naples, 71
102, 112-14, 452 (see also Apulia, John, abbot of St John “ad Titum”, 170
Calabria) John Lydos, 124 & n. 7
------, Byzantine, 68, 189, 424, 452 John of Poutze2, collector of revenues,
------, Byzantine expeditions to (934-5), 111 & n. 198
inventories for, 189, 324, 446 John the Deacon, Lat. hist., 168
------, east coast, 68 Joinville, Jean de, 612
------, west coast, 48-9, 64-5, 68, 70, 72 Julian, C., gov. Ceuta, 29
------, K. of, 44, 68-70 Justin I, Byz. Emp., 167
------, Muslim attacks on, 40, 48-9, 64-6, Justinian I, Byz. Emp., 12, 14-18, 124,
68, 70-72, 75 178, 326
------, southern, 40, 44, 48-9, 64-6, 71-2, Justinian II, Byz. Emp., 31, 167, 192
74-5, 77/86, 90-93, 424 Justinian, son of Germanos, Byz.
Itinerarium peregrinorum, 144 n. 59, general, 17-18
283, 417-18, 431-3
Ivrea*, margraves of, 69-70 Kahane, H. & R., 253 n. 298
Izmir, 26, 101 ------, and A. Tietze, 218 n. 163, 224 n.
------, Gulf of, 341 184
‘Izz al-Dı3n Qı£lı£j Arslan II, Salju2q id Kakorizos, koubikoularios†, 393
sult6., 630 Kalbı3te am. of Sicily, see Sicily
INDICES 725
Kallinikos, 26 n. 38, 383, 607-8, 624, 110 & n. 196, 111, 114, 381, 408-
627, 630 10, 419, 621-2
Kallixeinos of Rhodes, 197 n. 98 Komne2ne2, Theodo2ra, wife of megas
Kaminiate2s, John, De expugnatione doux† Isaakios, 407
Thessalonicae, 187 & n. 57, 220 n. Komne2nos, Alexios I, Byz. Emp., 87,
172, 225, 232-3, 240-42, 379, 612, 94, 99-102 & n. 180, 109-11, 266,
619 334, 452
Kantakouze2nos, Byz. general, 110 Komne2nos, Alexios III, Emp.
Karabisianoi†, fleet of, 25, 32, 267 Trebizond, 420
Kardam, Kh. of Bulgars, 45 Komne2nos, Alexios, pro2tosebastos†,
Kardia*, battle off (873), 61 121
Karpathos, 47, 191, 372, 389 Komne2nos, Andronikos, rebel, Byz.
Kasion*, 358 Emp., 121
376-8; Fig. 46 Komne2nos, Isaac I, Byz. Emp., 89
Katakylas, Leo, mag., 178 Komne2nos, Isaac, Emp. Cyprus, 118
Katasyrtai*, battle of (917), 67 Komne2nos, John II, Byz. Emp., 111
Kedar, B., 359 Komne2nos, Manuel I, Byz. Emp., 112-
Kedre2nos, George, Byz. hist., 408, 608 16, 120-21, 411, 415-16, 419 n. 54
Kefalle2nia, 17, 65, 88, 100 Komne2nos Bryennios, Alexios, megas
Kekaumenos, Strate2gikon of, 86, 89, doux†, 114
381, 394, 408, 569 n. 51 Ko2nstantianos, commander of imperial
Kellia*, Egypt, monastic complex, grooms, 14-15
158-9, 245; Fig. 17 Kontomyte2s, Constantine, strat.
Ke2poi*, aple2kton†, 47, 265, 327 Thrake2sio2n*, 47, 385
Khafa2ja ibn Sufya2n , Aghlabid gov. Kontostephanos, Andronikos, megas
Sicily, 64 doux†, 115-16
Kha2lid ibn Kaysa2n, 33 Kontostephanos, Isaac, megas doux†,
Khawa2rij†, 35, 40 (see also Iba2d5ı3yya) 111, 419 n. 55
Khazars, 31, 631 Kontostephanos, Stephen, megas
Khazdan, A., 187 n. 57 doux†, 113
Khludov Psalter, Moscow, Historical Korkyra/Korkyraioi*, 392 n. 634, 401
Museum, MS. 129 D, 157-9 & n. Koro2ne2, 334 n. 507
89, 245; Fig. 16 Korres, T. K., 628 n. 67
Khoury Odetallah, R., 187 n. 57 Kos, 47, 336
Khura2sa2n*, 35, 40 Kosmas, 32
Khusraw II, K. of Persia, 24 Koukoulès, P., 215 n. 155
Kibel, F., 154 n. 83 Krateros, strat. Kibyrrhaio2tai*, 46,
Kibyrrhaio2tai*, naval th. and fleet of, 327
28, 32-3, 46-7, 62, 64, 77, 88, 189, Krum, Kh. of Bulgars, 41, 45, 609,
191, 256, 259, 266-7, 307, 353, 622
371-2, 384-6, 390- 91 Ktesibios of Alexandria, 623-4
------, naval operations, 32-3, 41, 42, 46, Kumans, see Turks
62, 72, 88, 327, 400, 452, 608-9 Kusayla/Kası3la ibn Lamzan, Berber
Kiev, 73, 87 chief, 27
Kinnamos, John, Byz. hist., 113, 115- Kutrigurs, 19
17, 416-17 n. 45, 630 Kvarner, Gulf of, battle of (842), 49
Klaudianos, Byz. commander at Kyrenia, shipwreck, 145, 148 n. 69,
Salo2nes*, 17 195 n. 90, 364
Klazomenite2s, Katakalo2n, strat. Kythe2ra, 47, 191, 396
Ragusa*, 408 Kyzikos*, 27, 101 n. 179, 242, 608
Kleidion*, battle of (1014), 74
Kleoxenos, 399 Labienus, Titus, 331
Köbenhaven continuations of Prosper, Lamos* river, 60
13 Lampedusa, 45
Kolias, T. G., 381 Lampros family, 608
Komne2ne2, Anna, Byz. hist., Alexiad, Lampsakos*, 167, 190
726 INDICES
Landolfus Sagax, Lat. hist, 170 karabos†, 270 & n. 360
Landulf, Lat. admiral in Byz. service, Leo the Deacon, Byz. hist., 187-8,
110, 409, 621 308-9
Languedoc, 13, 30 Leontios, strat. Hellas*, Byz. Emp.,
Las Hormigas*, battle of (1285), 392 31, 325
Las Navas de Tolosa*, battle of Leovigild, K. of Visigoths, 12
(1212), 97 Lepanto, battle of (1571), 395 n. 649
Latakia, 33, 106, 109 Lesbos, 111, 116, 341
Lateran Council, Third, 416 Les livres des histoires du commence-
La Tradelière, shipwreck, 623 ment du monde (Paris, Biblio-
Laureate2*, 17 thèque National, Ms. Fr. 301), 312;
Lawson, P., 317 n. 468 Fig. 37
Lebounion*, Mt, battle of (1091), 100- Letter of the three Patriarchs (Mt
101 Athos, Iviron, Codex 381), 420-21
Leccavello, Simone, galley of, 359-60 Levant, 50, 63, 102, 105-7, 118, 386
Lech, battle of the (955), 70 Levantine waters, 76-7, 112, 114, 453
Lekape2nos, Basil, pat., parakoimo2- (see also Syria/Palestine, coasts of)
menos†, 77, 134, 175, 183-7, 240, Lévi-Provençal, E., 42 n. 71
447 Liber Maiolichinus de gestis Pisa-
Lekape2nos, Constantine, 189, 556 n. norum illustribus, 412
11 Liberios, pat., 17
Lekape2nos, Ro2manos I, pro2tokarabos, Liburni, Illyrian people, 125
droung. tou ploimou†, Caesar†, Libya, 128
Byz. Emp., 64, 67, 72, 167, 183, Licinius, R. Emp. (308-25), 7 & n. 3,
188 n. 62, 189 & n. 64, 190-91, 130
271 & n. 364, 390, 556 n. 11 Liguria, 43
Lekape2nos, Stephen, 259-60, 556 n. 11 Lilybaion*, 11, 351
Lemnos, battle off (923), 64, 385 Lisbon, 43, 76, 96, 98
Leo I, R. Emp. (457-74), 9, 124 n. 7 Liudprand, bishop of Cremona, 74, 76,
Leo III, strat. Anatolikon*, Byz. Emp., 150, 189, 271, 618, 621-2
31-2, 169-70, 608 Liutprand, K. of Lombards, 24, 32
Leo V, Byz. Emp., 32 Livadas, G. K., 233 n. 220, 241 n. 253
Leo VI, Byz. Emp., 47, 66-7, 133, Lombards, in general, 19, 25, 41, 163
151-2, 159-61, 164 n. 7, 182 n. 35, ------, K. of, 24, 32, 43-5 & n. 77
257, 266, 270, 391 ------, in southern Italy, 19, 24, 44, 49-
------, Ek tou kyrou Leontos tou Basil- 50, 65, 74, 77, 91-2, 323 (see also
eo2s, 181, 395 Benevento, Capua, Salerno,
------, Taktika, 66, 175-6, 181-2 & n. 35, Spoleto)
367 n. 580, 386-7, 395 Lombard (and Pisan and Genoese)
------, Naumachika Leontos Basileo2s, naval expedition, supposed, 111
134, 152, 160-61, 175-6, 180-83, London, 342 n. 530
185, 190-92, 202-3 & n. 121, 204, Longobardia*, th., 65, 69, 77, 166,
206-7, 210-12, 215 n. 155, 220 n. 168
172, 221, 229, 231-2, 234-6, 238, Lothar I, W. Emp., 48-9
246, 252, 254-5, 258, 260-61, 266- Louis I, K. Aquitaine, 42
70, 273, 275, 278 & n. 388, 283-5, Louis II, K. Italy, W. Emp., 49, 64-5,
304-5, 354, 361, 378, 381-3 & n. 168 & n. 25
612, 387-8 & n. 624, 391-405, 423, Louis VII, K. France, 106, 113
445, 448, 451, 609, 611, 617-18, Louis IX, K. France, 314, 327-8
622, 625-7 Louis, C. Clermont, 32
------------, sources, 176-80 Lucian, Lexiphane2s, scholia on, 212 &
Leo I, Pope, 154 n. 148
Leo IX, Pope, 92 ------, Zeus trago2dos, scholia on, 222,
Leo of Tripoli, 62-4, 232-3, 240, 385, 253
399, 452, 612, 637 Luni, 24, 32, 43, 45
Leo the Armenian, pro2telate2s†, pro2to- Lycia*, 25-6, 31
INDICES 727
------, battles off (956 & 1099), 72, 110, Marozia, Senatrix, 189 n. 64
385 Marrakesh, 94, 98
Marsala, 104
Maeander* river, 47, 265, 327 Marseilles, 43, 69
Maghrib, 27-8, 30, 34-5, 40, 45, 51, ------, horse transports, 314, 319-20, 323
69-71, 97-8, 102, 117, 363, 452 ------, Informationes pro passagio
Maghribin sailors, 117 transmarino (Paris, Bibliothèque
Magnaura†, 258 Nationale, MS. Fr. 12814), 327-8
Magnentius, usurper, 7 & n. 495, 356-7
Magyars, 70 Martin I, Pope, 216
Mahdia, 103 n. 183 (see also al- Martin, E. E., veterinary lieutenant,
Mahdiyya) 316 n. 463, 320 n. 477, 321-2 n.
Mai, Angelo, Cardinal, 140, 155 479, 324 n. 482, 327 n. 493
Maio of Bari, amı3r of amı3rs, 114 n. Maruffo, Niccolò, 359
206 Marwa2n II al-H4ima2r, Um. Cal., 34
Majorian, R. Emp. (457-61), 9 Mascezel, 7
Makroio2anne2s, Byz. naval com- Ması3la, 51
mander, 71 Maslama ibn ‘Abd al-Malik, 31-2,
Makrypoulias, C. G., 191, 256-7, 285- 169-70, 242
7, 301-2 Matthew Paris, Lat. hist., 424 n. 1
Malaga, 12, 88 Mauretania Caesariensis*, R. prov., 8,
------, graffito, 163 n. 1 11
Malake2nos, pat., 71 Mauretania Sitifensis*, R. prov., 8, 11
Malalas, John, Byz. hist., 125, 127, Maure2x, Michael, Byz. naval
403 n. 689 commander, 100
Malaterra, Geoffrey, Lat. hist., 412, Maurice, Byz. Emp., 133, 333-4 & n.
612-13 507
Malchos of Philadelphia, R. hist., 13 Maurice, Ek tou Maurikiou, 315, 394-
Malea, Cape, 65 n. 113, 110, 113, 116, 5 & n. 652, 397, 401
452 ------, Po2s dei diapleein potamous ...,
Malı3kı3 jurists, 95 179 & n. 22, 180, 232
Malta, 17, 40, 64, 226 ------, Strate2gikon, 163, 177 n. 11, 179
mama2lı3k (mamlu2ks)†, 76, 88 & n. 20, 183, 273, 282-3, 315, 379,
------, of Egypt, 358 394, 395 n. 652, 396-7
Mamora*, 98 Maurokatakalo2n, Nicholas, Byz.
Mams*, battle of the (688), 27-8 admiral, 409, 419
Mane2s, strat. Kibyrrhaio2tai*, 32 Maurozome2s, Theodore, 416
Mango, C., 149 n. 70 mawa2lı3†, 34
Maniake2s, George, strate2gos Maximinos, praetorian prefect†, 15
autokrato2r, 77/86, 91 Maximus, usurper, R. Emp. (383-8), 7
Mantzikert*, battle of (1081), 94 Mazara, 48, 93
Manuel, pat., 74 McCormick, M., 164 n. 7
Marcellinus, comes†, 9 Megara, Christ the Saviour, temple,
Marcellinus Comes, Rom./Byz. hist., 239 n. 246
124 & n. 4 Meinardus, O., 239 n. 237
Marcus Graecus, Liber ignium ad Melfi, 92
comburendos hostes, 610, 617 Melilla, 69-70
Mardaites† of Antalya, 190-92, 384 Melito2n, tourmarche2s†, 191, 396
Margaritus of Brindisi, Sicilian Mellows, J. S., veterinary surgeon, 317
admiral, 118 n. 468
Marianus (Marinus), Valerius, senator, Melo, Apulian rebel, 77, 91
342 n. 530 Meneghetti, Alvise, Venetian jeweller,
Maritime Alps, 40 antique dealer, forger, 237-8
Maritsa river, 100 Merovingians, 43-4
Marmara, Sea of, 7, 26, 100, 121-2, Mesardonite2s, Basil, katepano2† of
215 n. 155, 335 Longobardia*, 77
728 INDICES
Mesopotamia*, th. and army, 73, 89 40
Messina, 48 & n. 87, 92, 319 Muh5ammad II al-Mu‘tamid, ‘Abba2d id
------, Straits of, 92, 342 n. 530 malik† Seville, 95
------------, battles in (888, 901, 965), 66, Muh5ammad al-Na2s5ir, Almohad Cal.,
74, 385 97
Metho2n e22, 66, 354, 386 Muh5ammad ibn2 ‘Abd Alla2h ibn
Michael I, Byz. Emp., 46 Tu2mart, Almohad Mahdı3†, 96
Michael II, Byz. Emp., 32, 46, 48, 386 Muh5ammad ibn al-Ash‘a2th ibn al-
Michael III, Byz. Emp., 47, 50, 166 ‘Uqba al-Khuza2‘ı3, ‘Ab. gov.
Michael VIII Palaiologos, Byz. Emp., Ifrı3qiya*, 35
420 Muh4ammad ibn ‘Alı3 ibn Yu2suf al-
Michael, topote2re2te2s†, 271 n. 364 Masu2fi ibn Gha22niya, Almoravid
Michael the Rhetor, 411 gov. Balearics, 96
Michael the Syrian, Jacobite hist., 26 Muh5ammad ibn Ibra2hı3m ibn al-
n. 38 Thumna, am. Syracuse, 90
Michael, a metalworker, 622 Muh5ammad ibn Khafa2ja, gov. Sicily,
Michiel, Domenico, Doge Venice, 108 64
Michiel, Vitale, Doge Venice, 116 Muh5ammad ibn T4ughj al-Ikhshı3d, gov.
Milan, Biblioteca Ambrosiana, MS. B Egypt, 50
119-sup. [Gr. 139], 175, 178-80, Muja2h id al-Muwaffaq, am. Denia and
183-4, 217 Balearics, 88-9, 102
Milazzo, battle off (888), 66, 385 muja2hidu2n†, 41
Mindaros, Spartan admiral, 340 mulu2k al-t6awa2’if†, taifa “kings”, 88-9,
Minieri Riccio, C., 314 n. 457 94-5
Mithridate2s VI Eupator, K. Pontos, Munich, Bayerische Staatsbibliothek,
242, 401 Cod. Monac. Gr. 452, 182-3
Mityle2n e2, 264 n. 335, 265, 308, 334, Mu2sa2 ibn Nus5ayr, gov. Ifrı3qiya*, 28-9,
336, 340 & n. 525, 373, 556 n. 11 33
(see also Lesbos) Mylai*, battle of (36 B.C.E.), 231
------, David, Symeon, and George, Sts, Myra*, 88
bishops of, 172, 334 Myriokephalon*, battle of (1176), 117
“Milk Mountain”, battle of (552), 18
Monemvasia, 65 n. 112, 114 Nahr al-Ibra2hı3m*, 359
Mongols, 631 Nahray ben Nissı3m, 363
Monte Calascibetta, fortress, 93 Naku2r*, city state, 40-41, 43
Montecassino, abbey of, 65 Naples, 15-16, 42, 49, 65, 72, 75, 166,
Montfaucon, B., 420 n. 56 168
Morea*, 420 ------, battle off (ca 880), 66
Morocco, 94 ------, Gulf of, battle of (1284), 391 n.
Morrison, J., 340-41 632, 400
Moschio2n, 198, 366 Narbonensis Prima*, R. prov., 12
Moschos, John, Spiritual meadow, 358 Narbonne, 30, 42
Mott, L., 163 n. 1, 220 n. 171, 222 & Narse2s, Byz. general, 17-18
n. 177 Nasar, pat., droung. tou ploimou†, 65-
Mouikouron*, 17 6, 270 & n. 360, 385-6, 392, 452
Mouzalo2n, Nicholas, archbishop Na2s5ir al-Dawla Ba2dı3s, Zı3rı3d am., 51
Cyprus, patriarch Constantinople, Naupaktos, 88
334-5 ------, first battle of (430 B.C.E.), 351,
Mu‘a2wiya ibn Abı3 Sufya2n, gov. Syria, 382, 544 n. 85
Cal., 24-5, 26 & n. 38, 27, 393 ------, second battle of (413 B.C.E.),
Mu‘a2wiya ibn H4udayj al-Saku2nı3, 27 219, 222
Mueller, R., 238 & n. 240 naval expeditions, wrecked, 32, 72, 75,
Muh5ammad, the Prophet, 34, 50 391-2 & n. 634
Muh5ammad I, Um. am. Al-Andalus*, Naxos, 47, 264 n. 335, 265-6, 331,
68 333, 371, 373
Muh5ammad al-Muntas5ir, Idrı3sid Cal., Neilos of Rossano, St, Life of, 189,
INDICES 729
386 d’, 413
Nelson, Horatio, admiral, 334 n. 507 Oiniate2s, strat. Hellas*, 62, 620 & n.
Neretljani/Arentanoi, 67, 99 40, 621
Neretva river, 67 Oinoussai ship model exhibition at
Nicaea*, 31, 94, 410 Eighth Congress on Graeco-
------, Empire of, 418 Oriental and African Studies, 236
------, Second Council of, 32 & n. 234, 239
Nicholas I Mystikos, Patriarch of Old French/Anglo-Norman literature,
Constantinople, 63-4 (& nn. 103, 413-14
106), 67, 167, 190 “old Michael”, pro2telate2s†, pro2to-
Nicholas I, Pope, 166 karabos†, 270-71
Nicholas II, Pope, 92 Old Norse literature, 414
Nicholas of Sion, St, Life of, 211 Oleg, Rho2s† P. Novgorod and Kiev?,
Nicotera, 95, 104 66-7, 384
Niebuhr, B. G., 123 n. 2 Olympias, daughter of Robert
Niger river, 94 Guiscard, 99
Nikander, Theriaca and Alexiphar- Olympias, reconstructed Greek trie2re2s,
maca, 154 n. 85 201 n. 115, 233 & n. 221, 263 n.
Nikanor, St, monastery, Zavorda, MS 334, 279 n. 392, 292, 315 n. 461,
95, Pho2tios, Lexicon, 240 n. 249 335 n. 511
Nike2phoros I, Byz. Emp., 45 ------, crews, 206-7 n. 130, 274, 351-2,
Nike2phoros, pat., 27 356-7, 358 & n. 549
Nike2phoros Magistros, doux†, 189 ------, horizon, 389 n. 626,
Nike2tas David Paphlagon, Byz. writer, ------, oars/oarage system, 278-9 n. 388,
169 n. 27, 172, 224 n. 184 280-81, 286, 287 & n. 415, 289-96,
Nike2tas Magistros, Life of St Theok- 298 & n. 426, 299, 432, 434-5 & n.
tiste2, 333 19
Nikome2deia*, Gulf of, 102 ------, performance in waves, 337, 351-3
Nikopolis*, 9 ------, speed/manœuvrability, 143, 338-
------, th., 394 9, 342, 344-5
Nile river, 25 ------, stability, 206-7 n. 130
Nîmes, 30, 43 Olympias Mark II, reconstructed
Nineveh*, 24 Greek trie2re2s, 290-92, 294, 296,
Nishapur, 94 299, 302, 432 n. 14, 438-9
Nonius Marcellus, De compendiosa Olympic games, rowing, 131 n. 26
doctrina, 272 n. 370, 614 Omurtag, Kh. of Bulgars, 41, 46
Normans, (see also Sicily, Norman K. Onasandros, Strate2gikos, 177 n. 11
of) O›oryphas, probably droung. tou
------, in Byzantine service, 89 ploimou†, 46
------, in Ifrı3qiya*, 97 O›oryphas, Nike2tas, droung. tou
------, in southern Italy and Sicily, 45, ploimou†, 49, 61, 64, 378 n. 592,
77, 90-93, 103, 284, 424 385, 392
------, invade Balkans, 99-102, 612 Ophryneion*, 264 n. 335, 265
Norse, 43, 76, 414 Opsikion*, th., 31
Norway, 413 Oran, 98
Noto, 93 Oreste2s, Byz. commander in Italy, 77
Novem Populi*, R. prov., 12 Oria, 70
Numidia*, R. prov., 8 Orosius, R. hist., 8
Orseolo, Pietro II, Doge Venice, 68
Oberhelman, S. E., 171-2 n. 37 Orfewv” Argonautikav, 226
Octavian, 125 Osimo, 15, 18
Oddr Snorrason, 616 n. 26 Ostia*, port of Rome, 9, 16
Odo of Deuil, Lat. hist., 106 Ostrogoths, in Italy and Balkans, 14-
Odovacer, K. in Italy, 13 18, 24, 450-51
Odysseus, 194, 442 n. 27 Otranto, 16-17, 111, 166, 326, 394
Ogier de Danemarche, La chevalerie ------, Straits of, 389
730 INDICES
Otto I, W. Emp., 70, 74, 76 Pavia, 19, 44, 163
Otto II, W. Emp., 74-5, 168 n. 24 Pechenegs, Turkish? people, 73, 89,
Otto III, W. Emp., 75 100-102
Ouranos, Nike2phoros, mag., ho krato2n Pechina,* 35, 69 (see also Almeria)
te2s Anatole2s Pe2ge2*, battle of (922), 67
------, Ek to2n taktiko2n (Oxford, Bodleian Pelagos, shipwreck, 147
Library, MS. Baroccianus Graecus Pelagius/Pelayo, K. Asturias, 29-30
131), 183, 315, 360 & n. 558, 387, Peloponne2sos/Peloponne2sian, 45, 48,
396, 398 & n. 664 65, 305, 340, 382, 401
------, Peri thalassomachias, 181-2, 202 ------, th., 324 n. 483, 372
n. 121, 204, 206, 210-12, 215 n. Pepin I, K. Italy, 42, 45
155, 220-21 & n. 172, 229 & n. Pepin III/I, Merovingian Mayor of the
207, 231-2, 234, 238, 252 n. 295, Palace, 43; Carolingian K., 44, 166
254-5, 258, 261, 266-70, 273, 275, Perboundos, K. of Rynchinoi, 163
278 & n. 388, 284-5, 304-5, 361, Perikle2s, 305
378, 381, 383, 387-8, 393, 396-9, Persia/Persians, 19, 24, 393
403-5, 445, 448, 573 n. 3, 609, Peter, Kh. of Bulgars, 67
611, 618, 622, 625-7 Peter the Hermit, 109
------, Taktika, 180, 181 n. 29, 182 & n. Peter of Eboli, De rebus Siculis
35, 271 n. 364, 398 n. 667 carmen, 230-31, 283 n. 404, 407 n.
Outhwaite, B., 363 n. 565 1, 429-31, 442, 452, 637, 644; Fig.
Outremer*, 106, 114, 117, 120, 286, 54
412, 414-15 Petreius, Marcus, 331
------, coasts of, see Syria/Palestine Peukia*, Ta, 264 n. 335
Oxyrhynchus* papyri, 128 (see also Philippikos, Byz. Emp. (Bardane2s,
Egypt, papyri) strat.), 31
Philokale2s, Eumathios, Byz. gov.
Palermo, 48, 66, 72, 93, 103 Cyprus, 109-10
Palestine, 108 (see also Philokte2te2s, 131 n. 29
Syria/Palestine) Philo2n of Byzantium, Me2chanike2
Pamphylia*, 191-2 syntaxis, 176-7, 620, 623-4
Pamplona, 43 Philostratos the elder, 131
Panaghia islet, 116 Philotheos, Kle2torologion of, 271 n.
Pandolf I, P. Capua-Benevento, 74 364, 391
Pandolf IV, P. Capua-Benevento, 91-2 Philoxenos Latin-Greek glosses (Paris,
Pannonia*, R. prov., 19 Bibliothèque Nationale, MS. Lat.
Pantelleria, 90, 104, 610 7651), 194-5 n. 87, 213 n. 149, 217
Papacy, see Rome n. 161, 251 n. 294, 272 n. 370
Paphlagonia*, 100 Phoenicia, 7 n. 3
Paphos, 335-6 & nn. 510, 516 Phoinikous*, 25, 336, 385, 390, 451
Parastaseis syntomoi chronikai, 164 Pho2kas, Byz. Emp., 333-4 n. 507
Parekbolai, 381 Pho2kas, Bardas, 73, 386
Paris, 30 Pho2kas, Nike2phoros, domestikos to2n
Paris, Bibliothèque Nationale, MS. Gr. scholo2n†, 66
1564, Me2naion, 420 & n. 56 Pho2kas, Nike2phoros, domestikos to2n
Paros, 333 scholo2n†, Byz. Emp., 72-4, 76,
Partecipazio, Urso I, Doge Venice, 49 105, 182 n. 35, 183-4, 187-8, 265,
Paschal II, Pope, 110-11 268 n. 351, 308-9, 327, 354, 374,
Passero, Cape, 64 396, 408, 618
Patara*, 336 & n. 516 ------, Praecepta militaria, 182 n. 35,
Patria Ko2n stantinoupoleo2s, 167 271 n. 364, 387 n. 622, 398 n. 677
Patti, 95 Phormio2n, Athenian admiral, 351, 382
Paul, St, 226 Pho2teinos, strat. Anatolikon*, 46
Paul I, Pope, 166 Pho2tios, Patriarch, 169 n. 27
Paulos, exarchos† of Ravenna, 171 n. ------, Lexicon, 131, 194 n. 87, 196-7,
34 200, 202, 216, 221, 227 n. 199,
INDICES 731
234, 240 & n. 249, 383 (see also Prodromos, Manganeios, 312 & n. 456
Nikanor) Prodromos, Theodore, Rhodanthe and
Phournoi islets, 264 n. 335, 265 Dosikles, 283, 410
Phygela*, aple2kton†, 105, 265, 306-7, Proikonne2sos*, 47, 264 n. 335, 265,
327, 371, 373, 376 335, 373, 556 n. 11
Piacenza, Council of (March 1095), Prokopios of Caesarea, Byz. hist., 11,
101-2 & n. 179 16, 18, 125-30, 131 & n. 29, 132-4,
Piazza Armerina, Sicily, mosaics 137 153, 163, 173, 194-5 & n. 88, 229,
n. 48, 221-2 n. 177, 309-10; Fig. 283, 325-6, 393
36 Pro2tevon, John, proto2spatharios†,
Piccono, Pietro, 358-9 strat. Peloponne2sos, 324 n. 483
Picenum*, 17, 326 Provence, 24, 43, 50,
Pilatos, John, ase2kre2tis†, 71 ------, C. of, 68, 69
Piraeus, port of Athens, 340 & n. 525 ------, K. of, 68
Pisa/Pisans, in general, 97, 121 Pryor, J. H., 245-6, 401 n. 680, 430
------, naval forces and expeditions, 45, nn. 7-8, 434 n. 19
88-90, 95-6, 102-3, 105, 110-11, Psellos, Michael, Byz. hist., 86, 630
118-19, 409, 621 Pserimo, 146
Plane (Marseilles), shipwreck, 147 Pseudo John of Damascus, letter
Platis, S., commander, 352 attributed to, 170, 420-21
Plato, 186 Pseudo Kodinos, Traité de offices, 420
------, Laws, 196, 233 Pseudo Oppian, Kyne2getika (Venice,
------, Republic, 233 Biblioteca Marciana, MS. Gr. 479
------, Timaeus, scholia on, 194 n. 87 [coll. 881]), 273-4, 282, 620, 625;
Pliny, Natural History, 342 & n. 530, Fig. 26
562 n. 34 Pseudo Symeon magistros, Byz. hist.,
Plutarch, Table-Talk, 201 408
Po, river, 15, 164 Pula, 16
Podaro2n, pro2telate2s†, pro2tokarabos†, Punic Wars, 193-4 n. 86
270 & n. 360 Purpura, G., 152 n. 76
Pollux, Julius, Onomasticon, 185 & n. Pylai*, 215 n. 155
49, 186, 192-3, 195-6 & n. 93, Pylos* battle of (425 B.C.E.), 219
197-9, 200 & n. 111, 201-2, 210 n. Pyrenees mts, 8, 12, 29-30, 42
142, 215 n. 156, 216, 220 & n. Pythagorion (see Kastron Samos)
171, 221, 227-9, 232-4, 240, 244,
250-51 & n. 291, 253, 268-70, 272, Qal‘at Banı3 H4amma2d, 51
274, 276-9, 282, 284, 383 Qaramat6a2, Shı3‘a† sectarians, 50, 75
Polyainos, Strate2ge2mata, 178 Qı£lı£j Arslan I, sult6. of Ru2m*, 101
Polybios, R. hist., 231, 241-2, 351, Qutalmïsh, Salju2q id chieftain, 94
388 n. 624, 399
Ponza, 45 Ra2fi‘ ibn Maggan ibn Ka2m il, gov.
------, battle of (1300), 236 n. 237 Gabes, 103-4
Poros island, 351-2, 357 n. 546 Ra2ghib, client of al-Muwaffaq, brother
Portugal/Portuguese, K. of, 96, 98 of ‘Ab. Cal. Al-Mu‘tamid, 62
Portus*, 16 Ragusa*, th., 408
Port-Vendres A, shipwreck, 147 Rainulf I, C. Aversa, 91
Pozzuoli, 342 n. 530 Rainulf II, C. Aversa, 92
Praecepta imperatori Romano bellum Rametta, 74
cogitanti … observanda, see Ramon Berenguer III, C. Barcelona,
Constantine VII 96
Praecepta militaria, 618 Rankov, B., 358 n. 549
Praeneste*, 362 n. 562 Raoul, Manuel, 407
Preslav, the Little*, 73 Ravenna, 8, 13, 15-17, 19, 24, 123-4
Prevesa, battle of (1538), 395 n. 649 & n. 3, 170-71
Princes’ islands, 556 n. 11 ------, exarchate† of Italy, 19, 32, 163-4
Priskos, strat., 133 ------, papyri, 123-4 & n. 3
732 INDICES
------, St Apollinare Nuovo, ship Rögnvald, Jarl of the Orkneys, 414
mosaic, 154 & n. 83 Roman d’Auberon, 413
Raymond, P. Antioch, 112 Roman de Rou of Wace, 413
Reconquista, 30, 88, 96-7 Roman Vergil, (Rome, Biblioteca
red galley of Provence, 436 n. 22, 439 Apostolica Vaticana, MS. Vat. Lat.
Reggio, 17, 66, 68 3867), 135-6, 138, 152, 154-5,
Reiske, I, 214-15 (& nn. 153, 154), 299; Fig. 4
280 n. 396, 362-3 n. 565, 364-5, Romania*, 86, 121
380, 626 Ro2manos I Lekape2nos, Byz. Emp., see
Rhade2nos, John, pat., 64 Lekape2nos
Rhaidestos*, 335 & nn. 512, 513 Ro2manos II, Byz. Emp., 72, 178, 188
Rhetorica militaris, see Syrianos n. 62, 354, 554 n. 7
Magistros Ro2manos III Argyros, Byz. Emp., see
Rhodes/Rhodians, in general, 47, 334- Argyros
5 & n. 507, 389, 401, 556 n. 11 Ro2manos IV Diogene2s, Byz. Emp., see
------, Byzantines and, 31, 72, 259-60 Diogene2s
------, Crusader fleets and, 105 Ro2manos, exarchos† of Italy, 163
------, Muslim attacks on and occupation Rome/Romans, ancient, 7-8, 12, 27,
of, 25-7, 42, 40, 125, 127-30, 134-9, 152, 159,
------, prevailing winds, 336-7 & nn. 161, 186, 193-4 n. 86, 203, 215,
514, 516 219, 222, 225, 231-2, 241, 245,
Rhoiteion*, 340-42 250, 388 n. 624
Rhône river, 30, 43 ------, Byzantine, 9, 15-16, 18-19, 24-6,
Rho2s†, 74, 207, 631 164, 166, 168, 187-8, 195 n. 88,
------, attacks on Constantinople, 60, 66- 326
7, 72, 86-7, 144, 189, 207, 384-5, ------, medieval, 43-4, 48, 65-6, 74, 308
452, 618, 622, 630-31 n. 448 (see also Western Roman
------, in Cretan fleet of 949, karabia of, Empire)
189, 210, 214, 221, 246, 248, 554 ------------, Papacy, 19, 32, 43-4, 65-9, 92,
n. 8 166, 189 n. 64, 318 n. 470
Riba2t6 al-Fath (Rabat), 98 Romuald of Salerno, Lat. hist., 114
Ricca, E., 314 n. 457 Roncesvalles, battle of (778), 42
Richard I, Cœur de Lion, K. England, Rosetta, battle off (919), 51
119, 319, 331, 418 Rossano, 17, 326, 386
Richard de Templo, Itinerarium Rothari, K. of Lombards, 24
peregrinorum et gesta regis Rotrud, daughter of Charlemagne, 44
Ricardi, 203 n. 124, 319 n. 474 Roussillon, 147
Ricimer, mag. mil., 9 Ruffus, Jordanus, Medicina equorum,
Rif, 98 318 n. 472
Rimini, 13, 15 Ru2m*, 102-3
Robert Guiscard, Norman C./D. Runciman, S., 324 n. 483
Apulia and Calabria, 87, 91-3, 99- Russian primary chronicle, 67 n. 116
100, 410 n. 14 Rustamids, 35, 40-41, 51
Robert, D. Normandy, 109
Robert I, “the Frisian”, C. Flanders, S4a2bir, Fa2t5. Slavic am., 70
102 & n. 180 Sacra Parallela of St John of Damas-
Robert II, C. Flanders, 109 cus, (Paris, Bibliothèque Nationale,
Robert of Clari, Lat. hist., 243, 311 & MS. Gr. 923), 140, 142, 158, 239
n. 453, 331 & n. 502 n. 244, 245, 299; Fig. 8
Roderick, K. of Visigoths, 29 Saepinum*, 65
Rodrigo Díaz de Vivar, el Cid, 88 Saewulf, pilgrim, 336 n. 516, 412
Roger I, Great C. Sicily, 90, 92-3, 103 S4affa2rids, 35
Roger II, K. Sicily, 98, 103-4, 106, Sagrajas*, battle of (1086), 95
113 S4ala2h5 al-Dı3n, Ayyu2bid sult6., 97, 117-
Roger of Lauria, admiral, 236 n. 237, 20, 417-18, 610
391 n. 362, 400 Salamiella, Juan Alvares, Libro de
INDICES 733
Menescalcia de albeiteria et fisica 51, 90
de las bestias, 318 n. 472 sharı3‘a†, 95
Salé, 98 Shaw, (J.) T., 434
Salerno, 64, 70, 92 Shı3‘a†, 34-5, 40, 50-51
------, Gulf of, 65 ships as siege engines see bridges,
------, Lombard D./P. of, 44, 49, 68-9, flying
74-5, 91 shipwrecks, 69, 146, 150. The large
S4a2lih5 ibn Mans5u2r, 41 number of wrecks discussed at
Salju2q ids, 94 148-9, n. 69 are not indexed
------, of Ru2m*, 101-2, 117 separately. See also Agay,
Salo2nes*, 8, 14, 16-18 Bozburun, Cefalù, Dramont D &
Sa2m a2nids, 35 E, Grand Congloué, Kyrenia, La
Samarra, 35 Tradelière Pelagos, Plane, Port-
Samos, 25, 88, 111, 264 n. 335, 336, Vendres A, Serçe Limani, St
373, 376, 419 n. 54 Gervais B, Yassi Ada; naval
------, naval th., 47, 77, 88, 256, 259, expeditions, wrecked.
266-7, 306, 372, 376, 391 Shirley, A., lt colonel, 319 n. 473,
Samosata*, 184 321-2 n. 479, 327 n. 493
Samothrace, Victory of, monument, “short chronicles”, Byz., 420
378 n. 387, 442 n. 27 Sicily, 8-9, 14-18, 25, 28, 45, 50, 105,
Samuel, Ts. of Bulgars, 73-4, 182 n. 170-71, 305, 326, 331
35 ------, Byz. attempts at reconquest, 66,
San Salvatore, monastery, Messina, 74, 77/86, 271 n. 363
633 & n. 2 ------, Muslim attacks on, 26, 28, 33
Santarem, 98 ------, Muslim conquest of, 40, 45, 48 &
Santiago de Compostela, 76 n. 88, 64-5, 452
Sanudo Torsello, Marino, 442-3 ------------, Aghlabid period, 51, 64-6, 390
Sardinia, in general, 9, 11, 18, 45, 103 n. 629
------, Muslim attacks on, 28, 33, 40, 42- ------------, Fa2t5imid period, 70, 390 n. 629
3, 45, 48, 88, 102 ------------, Kalbı3te period, 70-72, 74-5,
Saronic Gulf, 352 390 n. 629
Sava river, 19 ------------------, naval expeditions, 75
Savona, 360 ------, Angevin K. of
Sayf al-Dawla ‘Alı3 I, H4amda2nid am. ------------, Angevin registers, 138, 217 n.
Aleppo (945-67), 184, 523 n. 7 162, 218 & n. 163, 314 n. 457, 318
Sayf al-Dı3n al-‘A›dil, Ayyu2bid admiral, n. 471, 430, 452
sult6., 120 ------------------, Formularium curie Caroli
Scheffer, J., 228 Secundi regis Sicilie, 318 n. 470
scholia, 187 ------------, galleys of, 138, 203, 217 n.
Se2lymbria*, 133 162, 218, 230, 233, 243-4, 248,
Senegal river, 94 260-61, 275, 287, 289-92, 304,
Senogallia*, battle of (551), 17-18, 312-13, 359, 430, 434 n. 19, 435-7
134, 451 & n. 22, 438, 442-3, 452, 613-14
Sens, 30 ------------, taride, horse transports of,
Septimania*, 30 312-14 & n. 457, 320, 322
Septuagint, 228-9 ------------------, paliolus, false floor, 322 n.
Serbs/Serboi/Serbia, 67, 408 (see also 479
Vlastimir, C! aslav Klonimirovic°) ------------------, slings, 317-18
Serçe Limani, shipwreck, 147, 211 ------, Norman conquest, 90-93, 103
Sergius II, Pope, 308 n. 448 ------, Norman K. of, 97, 103, 114, 633
Sergius IV, D. Naples, 91 ------------, Byzantine alliance, 121
Serre, P., 123 n. 3, 292 n. 424 ------------, conquest of Ifrı3qiya*, 104, 106
Seville, 29, 43, 76, 95, 97 ------------, naval forces and expeditions,
Sextus Empiricus, Against the 98-9, 104, 106-7, 111-13, 114 & n.
professors, 224 206, 117-18, 121, 312, 410-11,
Sharaf al-Dawla al-Mu‘izz, Zı33rı3d am., 452, 630
734 INDICES
Sidon, 106-7, 120 64
Sigeion*, 341 Stephen, brother of Maio of Bari, 114
Sigismund, K. of Burgundians, 153 & n. 206
Silves, 76 Stephen, ostiarios† and nipsistiarios†,
Simokatte2s, Theophylaktos, Byz. hist., 256-7, 259
133, 173, 333-4 & n. 507 St Gall, abbey, Switzerland, 69
Sisebut, K. of Visigoths, 13 ------, Stiftsbibliothek MS. 912, Latin
Skakki, Erling, 414 glosses, 166
Skle2ros, Bardas, 73, 77, 386 St Gervais B, shipwreck, 147
Skoutariote2s, Theodore, Byz. hist., 101 St Piero a Grado, church, Pisa, 102
n. 179, 116 Stilicho, mag. mil., 7, 8
Skylitze2s, John, Byz. hist., Synopsis Stilo*, Punta di, battle of (981), 75
historio2n, 308, 408, 620 n. 40, 621, ------, battle off (880), 66, 270 n. 360,
641 385
------, Madrid, Biblioteca Nacional, vitr. storms, 15-17, 27, 32, 72-3, 90, 95,
26-2, MS. of, 144, 207-8 & n. 132, 104, 110, 113, 115, 391-2
236, 255 & n. 307, 282, 300, 308, St Peter’s, 48, 74 (see also Rome)
426-31, 613, 620, 625, 633-44; Strobilos*, island fortress, 64
Figs 9, 33, 51-3, 57 Stryphnos, Michael, megas doux†, 121
------------, galleys of, 636-8; Table 9 St Symeon*, 105
Skyros, 116 St Tropez, 69
Sleeswyk, A. W. & F. Meijer, 362, Sturluson, Snorri, Saga inga konungs
368 og brå∂ra hans, 414
Smaragdus, pat., exarchos† of Italy, Styppeio2te2s, Kesta, domestikos to2n
164 scholo2n†, 62
Smith, F. H., colonel, later general, Suevi, 12
320 n. 477, 321-2 n. 479, 324 n. Sulayma2n, Um. Cal., 169
482, 327 n. 493, 330 & n. 500, 356 Sulayma2n ibn Qutalmïsh, Salju2q id
Smyrna, see Izmir chieftain, 94
Sofia, 17 Sunnı3/sunna†, 35, 40
Sophon*, lake, 102 Su2sa, 70 (see also Hadrumetum*)
Sotiel Coronada, force pump, 623-4; Svjatoslav, P. Kiev, 73, 77
Fig. 60 Sybota*, battle of (433 B.C.E.), 392 n.
Souda, lexicon, 123 n. 2, 131, 194 n. 634
87, 201-2, 203-4 n. 125, 216, 219, Sydney (Australia), 342 n. 530
221-2, 234, 240, 250 Sylloge2 taktiko2n , 609, 618
Spain, 8, 12, 28, 102, 256-7 (see also Symeon, Ts. of Bulgars, 60, 66-7
Reconquista) Symeon Logothete2s, Byz. hist., 167,
------, Visigothic, 26 188, 271 n. 364
------, Muslim invasion of, 29-30, (see Symeon of Mytile2n e2, St, 334
also al-Andalus) Syracuse/Syracusans,
spies, see intelligence systems ------, ancient, 222, 241, 305, 366, 545
------, of towns (kataskopoi to 2n poleo2n), n. 87
394 ------, medieval, 15, 17, 25, 48, 64-5, 90,
Spoleto, Lombard D./P., 19, 24, 65, 93, 95, 385, 414
68-70 ------, modern, 334 n. 507,
Sporadhes islands, 336 Syria, 73, 87, 89, 182 n. 35, 190-91,
Stadiodromikon (of 949), 264-6, 354, 393
373, 376 ------, naval forces, 62, 64, 75, 386, 393
Steffy, J. R., 145 Syria/Palestine, 24, 33, 50-51, 69, 393
Steiriones, John, Calabrian corsair, ------, coasts of, 24, 75, 107, 109, 112
122 (see also Levantine waters)
Stenon*, fleet of the, 271 n. 364 Syrianos Magistros
Stephen II, Pope, 44 ------, Naumachiai Syrianou Magistrou,
Stephen VI, Pope, 66, 168 & n. 27 178, 179-80 & n. 22, 181 & n. 29,
Stephen, droung. of Kibyrrhaio2tai*, 183, 360 & n. 558, 387-8 & n. 624,
INDICES 735
389, 395-6, 398-9, 401-2 24 n. 32, 149-50 & n. 70, 166-7,
------, Peri strate2gike2s, 178 169-70, 172-3, 188, 225, 242, 282,
------, Rhetorica militaris, 180 307, 322, 329-30, 333 n. 507, 371,
393, 399 n. 670, 400, 408, 445,
T4abarqa*, 70 607-9, 622
Tafur, Pero, 153-4 n. 82 Theophane2s continuatus, Byz. hist., 71
Tagus river, 43 n. 128, 133, 186, 188 & n. 62, 190,
Ta2hart*, 35 283, 308, 354, 378 n. 592, 386,
T4a2hirids, 35 393, 609, 620-21
Tahu2da*, battle of (681), 27 Theophano2, niece of John Tzimiske2s,
Taifa, “Party” kings, see mulu2k al- 74
t6awa2’if Theophilos, Byz. Emp., 32, 170, 385-
Taktika, see also Escorial Taktikon 6, 639
------, Taktikon Benes°evic° (ca 934-44), Theophilos, bishop of Alexandria, 227
391 Theophilos, strat. Kibyrrhaio2tai*, 41,
------, Taktikon Uspenskij (ca 842-3), 385, 400 & n. 675
390 Theophylaktos, tourmarche2s†, 66, 166
Tamı3m, Zı3rid am., 93, 103 The2ra, 149 n. 70, 333, 355, 373;
Tancred, C. Lecce, Sicilian admiral, The2ra-The2rasia, 264 n. 335, 265
118 Thermopylae*, 18
Tangier, 27, 29, 71, 76, 98 Thessalonike2, 7 n. 3, 63 & n. 103, 111-
Taormina, 64-5, 68, 74, 93 12, 118, 163, 225, 240, 379, 399,
Taranto, 16, 48-9, 66, 91, 326 612, 637
------, battle off (867), 49 Theudis, Ostrogothic gov., 12-13
T4a2riq ibn Ziya2d, gov. Tangier, 29 Thietmar of Merseburg, Lat. hist., 75,
Tarsos, 51, 61-2, 72-3, 232, 385-6, 215 n. 155, 260
400, 452, 620 Thomas the Slav, revolt of, 46, 236,
Tatikios, Byz. general, 110, 409, 621 308, 384-6, 392, 403, 635-6
Tauros mts, 41 (see also frontiers) Thousand and One Nights, 41
Teias, K. of Ostrogoths, 18 Thrace, 8, 31, 86, 101, 325
Telerig, Kh. of Bulgars, 45 Thrake2sio2n*, th., 47, 371, 391
Tenedos*, 264 n. 335, 265, 336, 373 Thrasamund, K. of Vandals, 14
Teruel, Spain, 138; Fig. 5 Thucydides, Gr. hist., Peloponnesian
Tervel, Kh. of Bulgars, 31, 32 war, 131 & n. 29, 133, 185-6, 216,
Thasos, battle off (839), 47, 385 219, 305, 338-42 (& nn. 525, 526),
Thebes*, 107, 410 351, 382-3, 401
Themetra*, ship mosaics, 137 & n. 48, ------, scholia on, 209 n. 137, 218-19,
250 n. 288, 251; Fig. 3 222-3, 272, 276-7 n. 382, 382
Theodamı3r, Visigothic D., 30 Thugga*, 137 n. 48
Theodo2ra, Byz. Empress, 32, 172, 334, Tiber river, 16
639 Tiberios III (Apsimaros), Byz. Emp.,
Theodore, anchorite, 358 28, 31, 167-8, 390
Theodore of Kythera, St, Life of, 191 timber supplies, 117
Theodore of Stoudios, St, 167 Tingitania*, R. prov., 8, 11
Theodore Psalter (London, British Tinnis*, 33, 107, 114, 117
Library, MS. Add. 19.352), 245 Tisza river, 19
Theodoric I, the Great, K. of Tjäder, J.-O., 123 n. 3
Ostrogoths, 12-14, 124 & n. 6, 125 Tlemcen, am. of, 96
n. 8, 129-30, 170 Tmutorakan*, 628, 631
Theodosios III, Byz. Emp., 31, 170 T4oghrïl I, Rukn al-Dunya2 wa ’l Dı3n,
Theodosius I, R. Emp. (379-95), 7 Salju2q id sult6., 94
Theodosius II, R. Emp. (408-50), 8-9 Toledo, 29, 88, 95
Theoktistos, mag. and logothete2s tou Tolo, 352
dromou†, 46-7 Tomislav, P., then K., Croatia, 67
Theophane2s, pat., 72 Torlonia relief, ship, 225
Theophane2s the Confessor, Byz. hist., Tortosa (Lebanon), 42, 96
736 INDICES
Tortosa (Spain), 118 Urbikios, 179
Totila, K. of Ostrogths, 15-18, 326 Urmia, lake, 89
Toulouse, 12, 30, Utica*, 45
------, C. of, 68
Tours, battle of (732), 30 Valdabron, am., 413
Tower of the Flies, 119 Valence, 43
Tradonico, Pietro, Doge Venice, 67 Valentinian II, R. Emp. (375-92), 7
Trajan’s column, 136-9, 251, 428 n. 6; Valentinian III, R. Emp. (425-55), 8
Fig. 3 Valentinos, Byz. commander, 16
Trajan’s Gates*, pass of, 73 Valerian, Byz. general, 16
Trani, 91 Valia, K. of Visigoths, 11
Trapani, 93 Van Doorninck, F., jr, 204 & n. 127,
Treadgold, W., 192 n. 81, 256, 389 n. 206
628, 390 n. 629, 554 n. 8 Vandals, 8-9, 12-14, 24, 124 n. 7, 451
Tricamaron*, battle of (533), 326 ------, Byz. conquest of, 10-11, 105,
Trieste, Gulf of, 64 126, 132, 325-6, 449-50
Tripoli (of Libya), 9, 27, 104 ------, naval forces and expeditions, 9,
Tripoli (of Lebanon), 63, 73, 75, 106- 132
7, 109, 111, 118, 385, 414 Varangians, 77
Tripolitana*, R. prov., 8 Vari, R., 177 n. 11
Trojan War, 131 n. 29 Vasiliev, A. A., 386-7 n. 620
Tsamakda, V., 633 & n. 2, 635, 638-9 Vatican Vergil (Rome, Biblioteca
& n. 10, 641, 643-4 Apostolica Vaticana, MS. Vat. Lat.
Tselevinia Strait, 352 3225), 136, 154; Fig. 1
T4u2lu2nids, 50, 62 (see also Ah5mad ibn Vegetius, Flavius V. Renatus, Epitoma
T4u2lu2n) rei militaris, 128, 177-8 & n. 15,
------, fleet of, 50 230 & n. 208, 230-2, 316-17 n.
Tunis, 28, 327 467, 329 n. 498, 396, 431, 433
Turin, C. of, 69 Vegetius, Publius, Mulomedicina,
Turks, 35, 87, 89-90, 94, 101-2 (see 316-17 n. 467
also Pechenegs, Salju2qids) Veneti, ships of, 146, 230 n. 208
------, in Byzantine service, 89, 114 Venice/Venetians, in general, 19, 32,
------, Kumans, Qipc°aq Turks, 100, 102 45, 67-8, 99, 113, 121, 273, 327,
------, Oghuz†, 94 420
Tuscany, margraves of, 68-70 ------, and Neretljani, 67
Tyre, 75, 108-9, 117-18 ------, Byzantine alliance, 121
Tyrrhenian Sea, 24, 102, 166 ------, imprisonment of by Manuel
Tzachas/Çaka, Turkish am. Smyrna, Komne2nos, 115-16, 630
101, 102, 109 ------, privileges in Byz. emp., 87
Tziliapert*, 628 ------, naval forces and expeditions, 48-
9, 67-8, 87, 99-100, 105-6, 108
‘Ubayd Alla2h al-Mahdı3, 40, 50 111, 115-17, 119-22, 168, 231,
Ukraine, 60 237, 243, 286, 305, 332-3, 385,
‘Umar ibn ‘Abd al-‘Azı3z, Um. Cal., 409-10, 417-18, 452, 612-13, 630-
29, 32 31
Umayyads, in general, 33-4, 45 ------, Biblioteca Marciana, MS. Gr. 335
------, naval forces and expeditions, 26- [coll. 645], 191
8, 31-3, 45, 383, 385, 393-4, 607-8 ------, San Marco, Pala d’Oro, 245
Umayyads, of al-Andalus*, 33, 35, 40- Venosa, 91
42 & n. 71, 68, 71, 76, 451 Verona, 18
------, naval forces and expeditions, 42- Versinikia*, battle of (813), 45-6
3, 68-70, 76 (see also Pechina) Ververouda, 352
‘Uqba ibn al-H4ajja2j al-Salu2lı3, gov. al- Vestiarion basilikon†, Department of
Andalus*, 29-30 the, 208-10, 212, 214, 244, 246,
‘Uqba ibn Na2fi‘ al-Fihrı3, 27 362
Urban II, Pope, 101 & n. 179, 102 Vesuvius, mt, 18, 65
INDICES 737
Vienne, 69 Xenopho2n, Anabasis, 339-40 & n. 522
Vigilius, Pope, 16 ------, Hellenika, 382
Vijosë, river, 113
Visigoths, in Balkans, 8 Yah5ya2 IV, Idrı3sid Cal., 40
------, at Narbonne, 8 Yah5ya2 ibn al-‘Azı3z, H4amma2did am.
------, in Spain, 12-13, 19, 24, 29-30 Bija2ya*, 104
Vita Basilii, see Theophane2s Yah5ya2 ibn Ibra2hı3m , S4anha2ja Berber
continuatus chieftain, 94
Vitalian, Byz. usurper, 16 Yah5ya2, Zı3rid am., 103
Vitruvius, 226, 623-4 Yaqz5a2n ibn Muh5ammad, Rustamid
Vivian, abbot of St Martin of Tours, im., 35
Bible commissioned by, 154 Yarmu2k , battle of the (636), 24
Vladimir I, P. Kiev, 60, 87 Yassı Ada, 4th-century shipwreck,
Vladimir, P. Serbia, 67 146, 148 n. 69, 152
Volubilis*, 40 ------, 7th-century shipwreck, 147, 211,
Von Sind, J. B, L’art du manège pris 362, 569 n. 51
dans vrais principles, suivi d’une Ya2zama2n al-Kha2dim, am. Tarsos, 62,
nouvelle méthode pour l’embou- 620-21
chure des chevaux, 318 n. 472 Yazı3d I, Um. Cal., 27
Vouillé, battle of (507), 12-13 Yazı3d III, Um. Cal., 33
Vsevolod, son of Jaroslav, P. Kiev, 87 Yngvars Saga Ví∂förla/Yngvarr
Vvilia, C. of the Patrimony, 13-14, Eymun-dsson, 616-17 & n. 26, 628
124 n. 6 Yu2suf Buluggı3n ibn Zı3rı3, Fa2t6 gov.
Ifrı3qiya*, 51
Wadı3 Abu2 Raqra2q, 98 Yu2suf ibn Tashufı3n, Almoravid Amı3r
Wamba, K. of Visigoths, 26 al-Muslimı3n†, 94-5, 104
watch towers, 24
West, Latin, 95-6, 99, 100-122, 134, Zachary, Pope, 32, 44
275, 380, 444 Zakynthos, 9, 65
Western Roman Empire, 44, 70, 74, 91 Zanj, 35
William “the Iron Arm”, see Zara, 99
Hauteville, William of Zaragoza, 30, 42,
William I, K. Sicily, 98 ------, taifa mamlaka†, 89, 94
William II, K. Sicily, 118 Zeno, R. Emp. (474-91), 9
William of Tyre, Lat. hist., 108-109 & Zenta* and Stamnos*, Serbian topar-
n. 192, 115, 117-18, 286-7 & n. che2s† of, 408
414-15 & n. 43, 416-18 Zichia*, 628, 631
Wilson, N., 429, 633, 639, 644 Zı3rı3ds, 51, 90, 93, 103-4
Witigis, K. of Visigoths, 14, 15 ------, naval forces and expeditions, 90,
Wolfenbüttel, Herzog August 93, 103-4
Bibliothek, Cod. Guelf., 96 Gud. Ziya2dat Alla2h I, Aghlabid am., 40, 48
lat., 614-15, 628 Ziya2dat Alla2h III, Aghlabid am., 40
“Wolf King”, see Ibn Mardanı3sh Zo2naras, John, Byz. hist., 408
Wolseley, G. J., general viscount, 327 Zo2simos, Byz. hist., 7 n. 3, 130
n. 493 Zuckerman, C., 180
Zuhayr ibn Qays al-Balawı3, 27
738 INDICES
[b] Technical Index
In this Index, where ship types used by one culture are referred to by names used in
another, they have been given in single quotation marks to indicate that it is a non-
indigenous attribution.
A first construction of this section of the Index by alphabetical entry alone quickly
became unusable because of the complex and multifarious nature of the book. Cross
references became legion and many entries would have had no meaning to general
readers. It became necessary to group many items under general headings and these
have been indicated in bold: arms/armaments, chelandia, crews, dromons, galleys,
Greek Fire, horse transports, hulls, oars, ships. These headings have sub-headings,
and sub-sub headings.
‘abı3d†, slave soldiers, 87 bow-ballistae, 380-81, 448

‘ahd†, covenant of peace, 25, 43 ------------------, chordai metaxotai, silk
akontia chalka, 404 & n. 693 bowstrings, 380
akritai†, 41 ------------------, navklai, 380
Al-H4ajj†, 94 ------------, toxa/toxaria, bows, 382, 448
alysis, chain, 31 (see also Golden ------------, toxareai Rho2maiai, “Roman”
Horn) bows, 285-6
amı3r al-juyu2sh†, 87 ------------, toxobalistrai, bow-ballistae,
amı3r al-Muslimı3n†, 94-5 247, 379 & n. 594, 381, 403, 448,
anchoring/mooring, 373-8, 449 562 n. 32
antle2te2rion/sentinaculum bilge pump, ------------------, trochilia, “blocks” for, 379
367 & n. 578 ------------, tzangra, Latin crossbow, 380-
aple2kta†, 105, 265, 327, 329, 331, 81
373, 376-7, 394-5 (see also ------, cranes, gerania, 378-9 & n. 392,
Ke2poi*, Phygela*) 404, 448
arms/armaments, 16, 230, 264, 285- ------------, karchesion, 379
6, 359, 378-82, 443 (see also Greek ------------, ke2lo2n (eia), 379 n. 592
Fire) ------------, manganon, 378-9
------------, katergon, 419 ------, missiles, 129, 384, 402-3, 410,
------, armour, 255 & n. 307 450
------------, cheiropsella, vambraces, 382 ------------, caltrops, 380, 402-3
------------, epilo 2rika, surcoats, 286 ------------, quicklime/unslaked lime, 380,
------------, karche2sion, 379 403
------------, kassides, helmets, 382 ------------, mues/muiai, “mice/flies”,
------------, klibania, lamellar cuirasses, quarrels, 379, 381 & n. 604, 403
285, 381, 636 ------------, reptiles, pots of, 380, 403
------------, kne2mides, greaves, 382 ------------, rhiktaria/rhiptaria, javelins,
------------, lorikia, mail corselets, 381 285, 382, 403, 448
------------, neurika, felt jackets, 381 ------------, sagitai, arrows, 382, 403
------------, skoutaria, shields, 382, 403 ------, rigging cutter, drepane2/drepa-
------------, skoutaria Lydiatika, “Lydian” non, 230, 382
shields, 285 ------------, longchodrepana, rigging
------------, skoutaria rhapta, sewn cutters, 230
shields, 285 ------, swords and pikes
------, bows & ballistae ------------, kontaria meta tribellio2n, tri-
------------, ballistae, 379 & n. 596 dent pikes/corseques, 285
------------, bows, 16, 403 ------------, menaulia, pikes, 285, 382, 405
------------, catapults, 402 ------------, spathia, swords, 285, 382, 448
------------, cheirotoxobolistrai, hand-held
INDICES 739
bacini, of Pisa, 102 ------------, shalandiyya2t†, 98, 169, 414
baggage train, see touldos ------, Western
barile, of Naples, 369 ------------, calant/chalant, 413
basilikon plo2imon, imperial fleet based ------------, chelandre/salandriae, et var.
at Constantinople, 32, 46, 61, 72-3, 168 (nn. 25, 27), 190, 215 n. 155,
77, 114, 116, 164 & n. 7, 261, 266, 237-8 & Fig. 22, 260, 305, 310,
270 n. 360 417
bourgesioi†, 115 ------------, lectus, commander’s berth,
bridges 215 n. 155
------, assault/boarding ------------, sandanum†, transport galley,
------------, corvus, 193-4 n. 86 105
------------, epibathrai, 193-4 n. 86, 242 n. chrysobull†, 87
257 comites, see kome2tes†
------, flying, 64, 177, 225, 240-41 & n. corvus, see bridges
242-3, 619-20; Fig. 23 crews, Byzantine, 66, 121, 124, 131-
------------, epibate2ria, 242 & n. 257 4, 254-76, 306 n. 438, 351, 384,
------, landing bridges/gangways/ramps, 418, 450 (see also strate2goi†,
306-12, 314, 326, 415 droungarios tou ploimou†)
------------, klimakes, 308-9 ------, fleet command, 266-9
------------, pons, 307 ------------, arche2gos, leader/squadron
------------, scala/skala, 265, 306-7, 309 commander, 268, 397
bungs/bung holes, 200-202 ------------, archontes, squadron com-
manders, 397
caulkers/caulking, see hulls, caulking ------------, droungarioi†, 267-9, 450
cavalry ------------, he2gemo2n, officer/squadron
------, Byzantine, 16-17, 265, 271 n. commander, 268, 397
364, 305-31 ------------, kome2tes†, 268-9, 397, 450
------------, hippeis, 325 ------------, topote2re2tai, port admirals, 271
------, Greek, hippeis, 305 & n. 364
------, Muslim, 43 ------------, tourmarchai†, 267-9, 450
------, Ostrogothic, 17 ------------------, to2n ploimo2n, 271 n. 364
------, Sicilian, 104 ------, ship’s command, 268-74
ceol, keel, 412 & n. 21 ------------, kentarchoi, commanders of
chelandia ships, 215, 268-9, 274-6, 381, 450
------, Byzantine, 48, 65-6, 72, 75, 89, ------------, kyberne2te2s/gubernator, helms-
166-70, 183 & n. 37, 186, 188-91, man, 220, 275
209, 233, 236-8, 246, 255, 258, ------------, naupe2gos, shipwright, 152
283, 290, 325, 355, 357, 364, 372- ------------, navarchos, see pro2tokarabos
3, 384-5, 395-6, 408, 413-14, 426, ------------, pro2reus, bowman, 271-2, 275,
445, 447, 618 (see also spur) 398, 450
------------, as horse transports, 167, 307, ------------, pro2telatai, “first oarsmen”,
309-10, 312, 315, 317, 320, 322-5, 270 & n. 361, 271
328-31, 371, 417, 420, 449-450 ------------, pro2tokaraboi, helmsmen, 270-
------------, celandria, 170 71, 393, 398, 450
------------, decks, 233 ------------, sipho2nator, 271-2, 275, 398,
------------, die2reis†, 191 450, 611, 621
------------, etymology of, 167 & n. 18 ------------, trie2rarchos, 215 n. 156
------------, kaly(m)bomatoi, sumps/water ------, ships’ crews, 254-66, 274-
tanks?, 363-7 ------------, archers, 144, 232, 236, 381
------------, ousiaka, 189, 256-7, 259-60, ------------, as supernumeraries, 262-6,
372 370-71
------------, pamphyla, 189, 256-7, 259-60, ------------, auteretai, 126 n. 13, 131-2 (&
372 (see also galleys, Byzantine, nn. 27, 29), 133
pamphyloi) ------------, chelandarioi, 167
------------, ‘shalandiyya2t’†, 47, 71, 169, ------------, elatai, oarsmen, 258, 274
414 ------------, ibykinato2r/boukinato2r, trum-
------, Muslim peter, 273
740 INDICES
------------, kataphraktoi, 381 ------------, marines & oarsmen, 14, 124 n.
------------, ko2pe2latai, oarsmen, 128-34, 6
181, 192, 232, 255, 260-64, 270- ------, Western, 351
72, 274, 285-6, 288-90, 294, 296, ------------, ciurma/zurma, et var., galley’s
298-302, 381 crew, 272-3
------------, marines, 131-3, 144, 181, 192 ------------, comiti, galley commanders,
& n. 81, 203, 208, 255, 261-2, 283, 275-6
285, 381-2, 402, 450-51 ------------, marines, 431, 433
------------------, polemistai, 192 n. 81, 261, ------------, nauclerii, helmsmen, 275
274 ------------, oarsmen, 105, 314, 431
------------------, stratio2tai, 232, 255, 274, ------------, of galleys of the K. of Sicily,
325, 381 260-61, 275
------------, nautai, sailors, crews of ------------, weight of, 171 n. 36
sailing ships, 274-5, 305
------------, of dromons, 17, 126, 131-4, da2‘ı3†, 40, 51
152, 181, 188, 191-2 (& nn. 81, dekarche2s, 268 n. 351
82), 232 doukade2s†, 90
------------, ousia, ship’s company, 255-8 dromons, Byzantine (see also
& n. 319, 259-64, 272-3, 357, 372- armaments, Greek Fire, signals,
3, 450, 547 n. 1 spur)
------------------, to ousia, 258 ------, dromo2nes (specifically, dromon in
------------------, arming of, 132, 381-2, 636 general is not indexed), 13, 123 &
------------------, pamphyla, hand-picked, n. 1, 124-8, 130-31, 133-4, 157,
259 160-61, 164 n. 9, 166-7, 169-73,
------------------, weight of, 262-4 & n. 334, 183 & n. 37, 188-90, 192, 407-9,
359-60 411-15, 417-18, 421, 423, 426,
------------, ousioo2, to crew, 258 445-7
------------, periousia, ship’s company, ------------, as biremes, 130-31, 133, 172-
258 3, 192, 447
------------, stratos, crew, 254, 258 ------------, as horse transports, 306-7,
------------, tsourma/tzourma, galley’s 309-10, 312, 315, 317, 320, 322-5
crew, 272-3 ------------, as monoremes, 127-31, 133,
------, Greek/Roman, 351 172-3, 190, 192, 447-8
------------, epistoleus, vice-admiral, 274 ------------, as quadriremes, 262
------------, hekatontarche2s, commander of ------------, as transports, 415-18, 447
a ship of 100 men, 268, 274 ------------, as triremes, 172-3, 192
------------, hortator/pausarius, master of ------------, ballast, 338
oars, 272 ------------, commands and orders, 273-4
------------, keleuste2s, master of oars, 272, ------------, dromades, 123 & n. 2
274 ------------, dromonarii, “men of the
------------, kyberne2te2s/gubernator, helms- dromons”, 124-5 & n. 8
man, 227, 270, 272 ------------, dromo2n basilikos/dromo2nion
------------, navarchos, admiral, fleet/ basilikon, imperial dromon, 164 &
squadron commander, 268-70, 274, n. 7, 167, 188, 258, 270, 408, 410
397 ------------, dromones, 13, 124 & n. 16
------------, pente2kontarchos, purser, 274 ------------, etymology, 125 & n. 12, 127-
------------, thalamios/thalamite2s, oars- 8, 130-31, 143
man, 276-7, 289 ------------, olkadas tachynautousas, 133
------------, thranite2s, oarsman, 276-7, 289 ------------, performance capabilities, 333-
------------, toxotai, archers, 408 78, 449
------------, trie2rarchos, commander of a ------------------, range of, 370-71
trie2re2s, 268-9 ------------------, speed/manœuvrability,
------------, trie2raule2s, flautist, 273-4 126-7 & n. 15, 130-31, 139, 143,
------------, zygios/zygite2s, oarsman, 276-7 160-61, 264, 333-5, 338-53, 449
------, Muslim, 25, 31, 119 ------------------, under sail, 336-8
------------, marines, 117 ------------, ventilation, 331, 355-6, 450
------, Ostrogothic ------------, xylokastra, wooden castles,
INDICES 741
229-30 (& nn. 207, 210), 232, 234- n. 424, 322, 373, 448
7, 404, 448 ------------------, size, 127, 130-33, 260-62
------, anchors/anchoring, mooring ------------------, tonnage, 304, 359-60
------------, anagokatagonta, windlasses, ------------, dryochon, stocks, 192-5 (& nn.
213-14 86, 87), 200, 529 n. 25
------------, anagontitea, ?, 213-14 ------------, enkoilia, floor timbers, 152,
------------, himantaria, “lifts”, anchoring 196, 202-3 & n. 121, 448
systems, 213 ------------, epenkenis, see katapete2ton
------------, koubaria, reels for mooring ------------, eudia(io)s, bung hole, 200-
cables?, 214 & n. 154, 362-3 n. 202, 448
565 ------------, flotation levels, 263-4
------------, leptaria, anchor buoys and ------------, garboard strakes, 196-7
lines?, 214 & n. 153 ------------, hull design, 127, 131, 143,
------------, periboloi, catheads, 210, 213, 290-304, 423, 446
221, 448 ------------------, flaring of, 292, 296, 298-
------------, peripetomena, windlass bars?, 300, 304
214 ------------, interscalmium, 287-9, 291,
------------, philoureai, lower anchor 296, 300; Fig. 28
cables, 212-13 & n. 149 ------------, kastello 2ma, pavesade, 282-3,
------------, side2ra (bolistika), anchors, 285, 404, 448
210-12, 272 ------------, katapate2ton, gunwale, 282,
------------, side2robola/side2robolia, schoi- 447-8
nia side2robola, anchor chains, 212 ------------, keelson, 196 & n. 93
------------, skalodemata, mooring cables, ------------, korakion, 193-4 & n. 86
214-15 & n. 154 ------------, kytos, hold, 227
------------, spartinai, upper anchor cables, ------------, orophe2, deck, 126, 129, 446
213 ------------, pela, strake, 281 n. 400
------, bows, pro 2ra, prow, 127, 135-40, ------------, peritona, wales, 195, 199, 200
143, 203-14 & n. 113, 281, 448
------------, gonatia, 280-81 ------------, pitch/liquid pitch for 152
------------, kataproso2pa, face of the bows, ------------, planks, extra, 152
209, 561 n. 27 ------------, scuppers, 201-2
------------, proembolis, 197-8 ------------, skoutaria, shields, 404
------------, pseudopation, foredeck, 203, ------------, stamides/stamines/ste2monaria,
402, 448, 611, 617, 620-21 upper futtocks of frames, 199, 448
------------, speira/steira, 192, 196-9, 528 ------------, stro2sis, fabric, 306
n. 23 ------------, thyreon, strake with oar ports,
------------, stemposts, 139-40, 142-3 278, 280-81
------, equipment ------------, tropidia, garboard strakes?,
------------, boukinon, trumpet, 273-4 192, 195-7
------------, kadoi, water amphorae, 214-15 ------------, tropis, keel, 143, 192, 195,
n. 154 198, 248
------------, kalymbomatoi, sumps/water ------------, tropis sterea, supposed
tanks?, 363-7 stiffener keel, 248-9
------------, kamelaukion, see signals ------------, xyla diatona, cross-beams,
------------, pathnai, mangers, 306 199-200
------------, phlamoulon, see signals ------------, zo2ste2r, wale, 410-11
------------, skalai, gangways, 306-7 ------, oars, oarage, oar benches, 127-
------, hull 34, 173, 181, 260-64, 283-304,
------------, amphime2trion, floor, 227 343, 432, 448 (see also oars,
------------, decks, 126-30, 132, 227-34, oarage systems, seated stroke)
423 ------------, askomata, see manikellia
------------, depth in hold, 292 ------------, elasia, oar bank, 254, 448, 450
------------, dimensions ------------, manikellia, oar sleeves, 279,
------------------, beam, 207, 244, 248, 292 280 & n. 395, 281, 284-5, 299,
& n. 424, 448 338, 447-8
------------------, length, 244, 248, 291-2 & ------------, pella, oars, 281 n. 400
742 INDICES
------------, schoinia, oar-grommets, 252, 224-5 & n. 184, 253
278 n. 388 ------, stern, prymne, poop, 215-27
------------, skalmoi, tholes, 278-9 (& nn. ------------, anklima, helmsman’s station,
387, 388), 287-8 & n. 417, 448 220
------------, thranoi, benches, 276-7 ------------, aphlaston, see stern ornament
------------, tre2mata/trimata/trype2mata, oar ------------, bordo2n es, unknown, 216-18,
ports, 278 & n. 387, 281, 294, 405, 447
448 ------------, krab(b)at(t)os/krebate2/graba-
------------, tropoi/tropo2te2res, oar grom- tus, commander’s berth, 215-16 &
mets, 192, 198-9, 278-9 (& nn. n. 155, 217, 269, 365, 448
388, 389), 448 (see also schoinia) ------------, parexeiresia, 218-19
------------, zygoi, thwarts, 276-8, 448 ------------, podostema/podostama, stern-
------, rigging: masts, sails, yards, 238- post, 198 & n. 101
54, 448 ------------, ske2ne, poop berth, 215-16 & n.
------------, armena, sails, 127-8, 142, 156, 448
153-9, 246 & n. 271, 247-9, 251-2 ------------, stern ornament, 216-18, 448,
------------, armenopoula, storm sails?, 636-8
247 dromons, Muslim
------------, chalkisia, blockmast, 244-6 & ------, dromonarion/dromu2n 164 & n. 9,
n. 264, 448 169
------------, histodokai, mast crutches, 248- dromons, Ostrogothic
52 ------, dromones, 13-14, 129-30
------------, karya, block sheaves/pulleys?, dromons, Western
252-3 & n. 298 ------, dromo/dromundus, et var., 411-13
------------, katartia, masts, 142, 229-31, ------, dromones, 166
238-51, esp. 249, 448-9 ------, dromont/dromund, et var., 413-
------------------, beaked mastheads, 245-6, 14, 418
636 ------, drómundr, 414
------------------, dimensions, 243-4, 248 droungarios tou ploimou/droungarios
------------------, mangana, blocks, 244-6 to2n ploimo2n†, 46, 49, 62, 64-5, 67,
------------, kathormeis, yard crutches, 77, 90, 210, 258, 266, 270 & n.
248-9, 251-2, 447-8 360, 271 n. 364, 362, 390-91, 450
------------, keraiai/kerataria, yards, 234-
5, 245-9, 251, 448 fire ships, 9 (see also galleys
------------, psellia, parrels?, 246-7, 448 [Muslim], Greek Fire
------------, pterna, heel of mast, 248, 250- flags, see signals/signalling
51 & n. 291
------------, trapeza, mast step, 248, 250- galleys, Byzantine, 13, 65, 72, 75,
51 & nn. 289, 291 113, 115, 117, 121-2, 126-30, 133-
------, rudders 5, 139-45, 152, 155-7, 160-61,
------------, auche2n, rudder/rudder shaft, 410-11 (see also chelandia,
220 & n. 172 dromons)
------------, emboloi (as rudder tackles), ------, agrarion rousion, imperial
224 n. 183, 225, 253 crimson barge, 164 n. 7, 270 n. 360
------------, epo2tides/paraptera 218-24 ------, akatos†, 258, 286
------------, hyperyption, rudder blade, 220 ------, die2reis†, 65, 169-70, 172-3, 191,
& n. 171 232, 409-10, 608
------------, oiakes, tillers, 134 n. 35, 220 ------, dromades nee2s, 409
& n. 172, 224-5, 253 ------, epaktrokele2tes, 113
------------, pe2dalia, quarter rudders, 207- ------, ‘galea’, 415-18
8, 220-27 (& nn. 172, 176, 177, ------, galeai, 190-91, 259-60, 283-4,
197), 448, 637 372, 396, 423-4, 426, 448, 452
------------, petasoi, rudder housings?, ------, ‘galee’, naves longe rostrate,
220-21 & n. 176, 448 115, 117, 415
------------, schista, rudder through ------, graffiti of, 239 (& nn. 246, 247),
beams?, 220-21, 448 407
------------, trochante2r/trechante2ri, 216, ------, karaboi/karabia†, 188-9, 210,
INDICES 743
214, 221, 246, 248, 258, 408, 420 ------, forecastles, 136
------------, sails, 246 & n. 272, 248 ------, histodokai, mast crutches, 234-5
------, katergon/katirgon, 418-21, 444, & n. 228, 248-9, 251-2
447; Figs 48, 49 ------, interscalmia, 287 & n. 415, 291,
------------, exkoussaton, 419 & n. 55 294, 296, 300
------, mone2reis/mone2ria†, 190, 246, ------, katastro 2mata, half decks, 128-9,
248, 283, 396 404
------, myoparo2nes†, 113, 283 ------, kele2s, 167
------, ne2es makrai/ne2es de makrai, 408, ------, liburnae, 8, 123 & n. 2, 125 & n.
416-17 11, 127-30, 161, 186, 197 n. 99,
------, ne2es polemiste2riai, 630 219, 221-2 n. 177, 230 & n. 208,
------, ne2es pyrphoroi, 629 274, 342, 423, 428 n. 6, 430, 433,
------, ne2es tachynautousai, 65, 133 446; Figs 1, 2
------, on seals, 407 ------------, decks, 231-2
------, ousia/ousiakos (as ship type), ------------, half decks, 127, 129
255-8 (but see crews) ------------, oars, 428 n. 6, 433
------, pamphyloi, 189, 191-2 (& nn. 81, ------------, square sails, 153
82), 256-60, 372, 618 (see also ------------, yards, 230
chelandia pamphyla) ------, longae naves, 411
------, pente2konteroi†, 113, 133, 283 ------, oars, 198 n. 102
------, ploia kastello2m ena, 282 ------, parexeiresia, outrigger, 218-19 &
------, ploia makra†, 14, 18, 115, 118, n. 167, 291-2
121 ------, pe2dalia, quarter rudders, 226
------, pyrphoroi, 113 ------, pente2konteroi†, 190, 226
------, sage2nai†, 163, 258 ------, phoinikis, battle flag, 397-8
------, scout ships, 284, 387, 389, 396, ------, proembolis, 197-8 & n. 99
398 ------, rudder mounts, 221-2 (& nn. 176,
------, skaphidia, skiffs, 122, 630 177)
------, trie2reis†, 65, 113, 116, 121, 167, ------, scaphae, scout ships, 396
169-70, 172-3, 188, 308-9, 408-11, ------, skalmoi, tholes, 279, 282
416-17, 445, 620 n. 40, 630-31 ------, stability, 206-7 n. 130
------, zo2ste2r, wale, 410-11 ------, steira, part of stempost, 196 n. 93,
galleys, Greek/Roman, 7-8 & n. 3, 197-8 & n. 99, 199
125, 127-30, 136-40, 146 (see also ------, stemposts, 136, 138-9, 143, 197
ram) ------, stolos/perikephalaia, head of the
------, akatos/akation†, 164 prow, 197
------, akrostolion, head of the prow, ------, tetre2re2s†, “four”, 145
197-8 nn. 98, 99, 217 ------, thalamia, oarport, 280
------, aphlaston/aplustre, stern orna- ------, thalamios, 199
ment, 216-17 & n. 161 ------, thranoi, benches, 277
------, asko2mata, oar sleeves, 279 & n. ------, triacontoroi/triak(ch)onte2re2s†, 7
389, 280, 299 & n. 3, 123 & n. 2, 130 & n. 25,
------, aulos, flute, 274 132
------, bucina, trumpet, 274 ------, trie2reis†, 7 & n. 3, 128, 130, 136,
------, castles, 231-2 153, 210, 218-19, 273-4, 277, 291-
------, celeusma, command setting the 2, 294, 336-7, 383, 404, 430
stroke of the oars, 272 ------------, as horse transports, 315 n. 461
------, columbarium, oarport, 280 ------------, ballast, 338 n. 520
------, dimensions, 291-2 ------------, hypozo2mata, 233-4 & n. 221
------, eiresia, oar bank, 218, 254-5 ------------, speed under oars, 338-42 & n.
------, epe2n kenis, “gunwale”, 282 525; Fig. 41
------, episkalmis, stringer for tholes, ------------, water supplies, 361, 368
279-80 & n. 392, 282 ------, tropoi, through beams, 198-9
------, epitonos, backstay, 218, 221-2 n. ------, wales, 146, 197
177 ------, wrecks, 392 & n. 634
------, epo2tides, cheek timbers, 210, 219, ------, zeuglai/funiculi, rudder tackles,
222-3 226-7
744 INDICES
Galleys, Muslim, 96, 98, 103-4, 107, ------------, outrigger, 435, 438
109, 117, 119-20, 163 n. 1 (see ------------, with two oar banks?, 424-6,
also chelandia, dromons) 430-31
------, aghriba† [ghura2b ], 98, 103-4 ------------, galee/galie, 423-4 n. 1
------, akation/akatenarion†, 164 & n. 9 ------, French, 17th-century, 357
------, al-‘(m?)aun, p. 164 n. 8 ------, gati, see catti
------, decks, 234 ------, golafri/garabi†, 412
------, ‘die2reis’†, 165 & n. 11 ------, gumbariae†, 67
------------, ‘kastellatoi’, 165 & n. 11 ------, interscalmia, 289, 435, 437 n. 22,
------, ‘galeae’, 108-9, 190 442
------, h5arra2qa2t†, fire ships, 64, 117, 610 ------, ‘katerga’, 420
------, kadirga, 216 n. 156, 420 ------, lintres, vessels, 412
------, ‘karabo2s/karabion’†, 164-5, 188- ------, manichilium, oar sleeve, 280 n.
9, 270 395
------------, ‘kastellatoi’, 165 & n. 11 ------, masts, 243-4
------, ‘koumb(p)aria’†, 62, 513 n. 51 ------, Renaissance, 434
------, ‘myoparo2n es’†, 190 ------, rudders, 637
------, naffa2t6a2t, fire ships, 610 ------, sagene†, 412
------, ‘pente2konteroi’†, 190 ------, ‘shawa2nı3’†, 103-4, 107
------, qa2dis, 164 n. 8 ------, shields, 283, 431
------, qit6‘a†, 286-7 n. 414 ------, squarciavele, rigging cutters, 230
------, rudders, 638 ------, stemposts, 637
------, sails, lateen, 638 ------, telaro, rowing platform, 439 & n.
------, ‘sa(k)tourai’†, 190 23, 442-3, 453
------, shawa2nı3†, 96, 98, 103, 107, 117, ------, temones, quarter rudders, 225
119, 414 ------, ‘trie2reis’, 409-11
------, ‘skaphe2’†, 61 ------, triremes, 442-3 & n. 29
------, ‘trie2reis’†, 307 ------, wales, 411 n. 18
------, ‘usha2rı3, transport galley, 258-9 n. ------, water supplies, 359
319, 310-11 n. 448, 311-12 n. 453 geniza†, of Cairo, 102, 363
galleys, Ostrogothic, 14, 124 n. 6 grappling irons, 193-4 n. 86
------, ‘ploia makra’†, 14, 17-18 Greek Fire
galleys, Western, 105-7, 116, 118, ------, Byzantine, 26 & n. 38, 27, 31-2,
122, 138, 230-31, 239, 314 (see 46, 61-2, 72-3, 86, 110, 144-5, 169,
also Sicily, Angevin K. of; spur) 189, 208, 239, 378-80, 383-5, 392,
------, ‘aghriba’, 104 402, 404, 408, 448, 450-51, 607-31
------, banda, 217-18 & n. 162 (see also Kallinikos, Lampros)
------, barce, 412 ------------, cheirosipho2nes, hand sipho2-
------, bellicosae naves, 48 nes, 275, 617-20, 625-7
------, callati/collativi/collaturi, tie tack- ------------------, skoutaria side2ra, iron
les, 230 n. 208 shields for, 617, 625-6
------, carabi/currabii, 170-71, 412 ------------, composition of fuel, 614-17,
------, catti†, 286-7 & n. 414, 412, 612 624, 628
------, columbaria, oarports, 280, 426, ------------, extinguishing of/resistance to,
442 617
------, decks, 233-4 ------------, gerania, cranes, 378-9 & n.
------, deck beams, 289 392, 404, 609, 627
------, dimensions, 243-4, 302 ------------, hurled by catapult, 607, 611-
------, ‘eiresia’, file of oarsmen, 411 & 12
n. 17 ------------, in pots, 607, 609, 611-12, 627
------, galeae(galee), 96, 105-6, 109, ------------, manganika, 618
116-18, 122, 171 n. 37, 284, 290- ------------, pyr eskevasmenon, processed
91, 300, 305, 312, 316 n. 467, 342, fire, 378, 402, 608-9, 618, 627-8
355, 368, 407 n. 1, 415-17, 420, ------------------, distillation, 628-30
423-44, 447, 452-3, 637 ------------, pyr hygron, wet fire, 608-9
------------, apostis, outrigger “gunwale”, ------------, pyr kollytikon, glutinous fire,
430, 443 618
INDICES 745
------------, pyr lampron, brilliant fire, ------, weight of, 324-5
609, 613 horse transports
------------, pyr polemikon, fire for war, ------, British, 315-17 (& nn. 463, 467),
609 320 & n. 477, 321 n. 479, 323-4 & n.
------------, pyr thalassion, sea fire, 608 482, 327 n. 493, 636
------------, pyrobola, fire throwers, 618 ------------, foot-boards & shingle, 321-2
------------, sipho2nator, 611, 621 n. 479
------------, sipho2nes/sipho2nia, 169, 203, ------------, slings, 317-18 & n. 468
209 n. 138, 257, 275, 378, 384, ------, Byzantine, 115, 167, 265, 304-
402, 448, 608-12 & n. 17, 613-14, 33, 415-17 & n. 45 (see also
617-22, 624-8 & n. 67, 631 bridges, chelandia)
------------------, boukolia, heat shields, 624- ------------, hippago2ga/hippago2goi†, 113,
6 274, 395, 416-17 n. 45, 538 n. 63
------------------, force pumps, 622-6, 628-9 ------------, ‘naves maiores ad deportan-
------------------, furnace, 615-17 dos equos deputate’, 115
------------------, gonatia akontia, hinged/ ------------, ne2es hippago2gai/hippago2goi,
jointed poles/pikes, 624 305
------------------, hearth, 627-8 ------------, pathne2, manger, 265, 306-7,
------------------, linaria/sphongoi, 626 316
------------------, of brass or bronze, 615-17, ------------, ploia hippago2ga, 305
622 ------------, taretes, 420, 447
------------------, propyra, fore-fires, match- ------------, thyrides, stern ports, 312, 326,
es?, 626-7 415, 450
------------------, tetrakoula, 626 ------------, ventilation, 330-31, 450
------------, squitiatoria, squirt, 615-16 ------, Greek/Roman, 305, 309-10, 315
------------, strepta, swivels, 609, 618, 622 n. 461
------------, triboloi, caltrops, 609, 627 ------------, phatne2, manger, 307, 316
------, Muslim, 609-12 & n. 17 ------, Muslim,
------------, ana2bı3b, tubes, 611-12 ----------, t6ara2’id†, 98, 117, 308 & n. 448,
------------, h5a rra2q /h5arra2qu2n, 610-11 415, 417
------------, naffa2t6/naffa2t5u3n, 610 ------, Western, 107, 312
------------, naphtha, 616 ------------, bote, water butts, 328
------, Western, 612-14 ------------, chelandre, 305, 310-11
------------, ignis silvestris, 613-14 ------------, huissier/oxerius/uscerius et
------------, roccette ad ignem proicie- var., 258-9 n. 319, 311, 417-18
ndum, 614 ------------, lanzones/cyntae, slings, 317-
19 & nn. 470-73
h5a2jib†, 70, 76, 118 ------------, paliolus, false floor, 321-2 n.
hekakontarchos, 268 n. 351 479
horizon, see visibility ------------, pontes, ramps, 307, 310, 314,
horses, barley for, 305-6 (& nn. 438, 417
439) ------------, ostia, stern ports, 310-14, 417
------, kele2s, “courser”, 167, 325 ------------, ‘tara 2’id’†, 107
------, seasickness, 315-16 & n. 463 ------------, taride†, 305, 310-14 & n. 457,
------, transportation of by sea, 315-16 & 316 n. 467, 328-9 & n. 496, 417,
n. 463, 318-23, 328-9 n. 496, 330, 420
371, 449 ------------------, cat[h]ena mortua, manger,
------------, air/oxygen requirements, 330, 314, 316-17 & n. 467
450 hulls, 131, 137, 143, 145-52 (see also
------------, azoturia, 330 dromons)
------------, laminitis, 330-31 ------, caulking, 147-52 & n. 69
------------, to Holy Land, 106, 319, 331-3 ------------, calafata, caulker, 150
------------, Western, 318-20 ------------, kalaphate2s/kalaphate2seo2s,
------, war/cavalry horses, 319 & n. 476, caulker/caulking, 150
322 & n. 480 ------------, tow, oakum, 147, 152
------, water requirements, 327-9 & n. ------, floor timbers, enkoilia, 196-7
495, 332, 371, 449-50 ------, frames/ribs, 145, 195, 199
746 INDICES
------, keel, 149 n. 70, 195-6 & n. 90 muku2s† [maks], non Qur’a2n ic taxes,
------, keelson, 196 & n. 93 95
------, lead/copper sheathing, 147-9 & n.
69, 152, 364 naukle2ros, ship master, 358
------, pitch, 149 & nn. 69 & 70, 562 n. navigation, coastal, 105, 341, 354
34
------, shell construction, 145-52 (& nn. oars (see also crews, Byzantine, ships’
60, 69), 195, 446 crews; dromons)
------, skeleton construction, 147, 151 n. ------, gearing, 181 n. 30, 290 & n. 421,
76, 152 n. 78, 195, 446 292-3, 435
------, waterproofing coatings, 148-9 & ------, grommets, see dromons,
n. 69 Byzantine, oars, tropoi/tropo2te2res
------, in general, 285-304
indictions, 125 n. 8 ------, slippage, 440 n. 26
iron, casting of, 204 n. 126 ------, spare, 285
------, thalamian, 289-91, 293-4, 296,
jiha2d†, 41, 96 298-9, 351, 356, 432 n. 14
jizya†, tribute tax, 30 ------, thranite, 289, 291, 296, 298 n.
jund†, 76 426, 299, 432 n. 14
------, weight in hand, 290 & n. 421
kadoi, see water/watering ------, with multiple oarsmen, 286, 300-
kamelaukion, see signals 302, 415
katepano†, 77, 90, 92, 190 ------, zygian, 291, 298-9, 351
katergokistai, 419 n. 54 ------, oarage systems, 133-4, 181 n. 30,
kleisourarchai†, 391 261-4, 284-304, 343-4
kochlias, Archimedes screw pump, ------------, alla sensile, 284, 305, 315,
367 426-44, 452; Figs 51-6
ku2ra†, 43 ------------, a scaloccio, 434 & n. 16
------------, ordines, 415-16, 431
limbers/limber holes, 366 ------------, seated stroke, 284, 287-8, 295-
logothete2s to2n agelo2n†, 306 6, 343, 430, 433
logothete2s tou dromou†, 306 ------------, stand-and-sit stroke, 284, 430
& n. 8, 433-6, 443, 453
magistri militum†, 7-9 ------, oar benches, 276-8, 286-7, 290,
mama2lı2k (mamlu2ks)†, 76, 88 431-3
masts/mastheads (see also dromons; ------------, canting of, 290, 439-42
bridges, flying) ------------, footrests, 433
------, Greek, 229 ------, oarsmen, (see also crews,
------------, akateios/akatia, 240, 253 Byzantine, ships’ crews)
------------, dolo2n, 240 ------------, mechanical advantage, 435
------------, epidromos(n), 240 ------------, physiology, 355-7
------, lowering of, 218, 231 ------------, thalamian, 357 & n. 546, 358
------, Muslim, 64, 234-5, 240 & n. 549
------, Roman, 155 & n. 87, 221-2 n. ------------, water supplies, 368-73, 449
177, 230 n. 208
------, Western, 218, 231 paria†, 88, 95
------, with lateen rig, 230 n. 208, 235, patrikioi†, 17, 26 n. 38, 27-8, 63-6,
236 71-2, 134, 183, 186, 270-71, 325,
megas domestikos†, 113 384, 414
megas doux†, 101, 111, 113-15, 121 pelagolimen, “sea harbour”, 516 n. 65
meltemi, 336-7, 341, 371, 374 pente2kontarchos, 268 n. 351
merarchoi/moirarchoi, divisional phylira (Lat. tilia), linden tree, 212-13
army or fleet commanders, 396-7 & n. 149
mere2, squadrons of a fleet, 397 pithos, see water/watering
praetorian prefect†, 11, 13, 15, 164
prokoursatores, cavalry scouts, 271 n.
364
INDICES 747
pro2tospatharios te2s phiale2s†, 271 dicuntur’, 115
provisions/provisioning, 105, 115, ------------, ne2es, generic ships, 127, 134,
264-6, 306 n. 438, 443 153
------, areurion, flour, 306 n. 438 ------------, phortago2goi†, transport ships,
------, barley for horses, 306 (& nn. 438, 113
439) ------------, phorte2goi/phortika†, transport
------, paxamation, biscuit, 306 n. 438 ships, 274, 305z, 395, 538 n. 63
------, sitos, wheat, 306 n. 438 ------------, ploia†, generic ships, 127,
pumps, 366-7 134, 163, 334
------------, porthmeia, transports, 309
qa2’id†, 96 ------------, skeve2† , barytera, vessels of
quartarolo, of Genoa, 369 burden, 163
------------, skevophora†, supply ships,
ram, waterline/ramming, 127, 131, 274, 305
134-40, 143-6, 152, 197, 203-4, ------, Greek
206-8, 291, 383-4, 423, 446, 448, ------------, amphime2trion, floor, 227 & n.
450 199
------, embolas, eperchomenous, “attack- ------------, ankoina, halyard
ing”, 134 ------------, apogaia/protonoi, offshore
------, embolon/embolos, embole2/emba- mooring lines, 253-4
llein, 134 & n. 35, 135 & n. 38, ------------, chalinos, parrel, 247
145-6, 197, 203-4 n. 125, 206, 545 ------------, cheimaros, bung, 201
n. 87 ------------, che2niskos, stern ornament, 222
------, rostrum/rostrate, 135 & n. 38, ------------, enkoilia, floor timbers, 196,
145-6, 203 n. 124, 415 200, 202
------, synkrousis/synkrouein, collision/ ------------, enthemion, 227-8 & n. 204
collide, 206 ------------, episeio2n, 228-9
rhogai, annual cash salaries, 391 ------------, epitonos, backstay, 218
rig/rigging, see sails ------------, ergate2s, windlass/capstan, 213
rudders, sternpost, 224 n. 184 & n. 151
------------, eudiaios, bung hole, 201 & n.
sailing season, 329 & n. 498 115
sails ------------, hedrai, seats, 200, 202
------, flax for, 342 n. 530 ------------, hermata, shores, 200, 202
------, lateen, 127-8, 153 & n. 82, 154 n. ------------, histia, sails, 248-9, 251
82, 155-9 & n. 90, 163 n. 1, 218 & ------------, histos, mast, 248, 250
n. 164, 230 n. 208, 238-9, 241 n. ------------, hypomochlion, 226
252, 423, 446, 448, 636 ------------, ikria, deck, 128, 196, 200,
------------, tacking with, 135 202-3
------------, ties and tie tackles, 230 n. 208 ------------, kaloi, brails of sails, 253-4
------, square, 127-8, 153, 154 (& nn. 83, ------------, kanonia, deck beams, 199
85), 155-7, 221-2 n. 177, 423 ------------, karche2sion, mast head, 244
------------, slings, 203 n. 208, 241 n. 253 ------------, mesodme2/le2nos, mast step,
------------, himantes, lifts, 213, 230 n. 208 250-51 & n. 291
saio†, 14 ------------, oiakes/oie2kes, rudders, 226-7
scout ships (see galleys, Byzantine, & n. 196
Greek/Roman) ------------, pareia, cathead?, 210 n. 144
sea breezes (see winds, coastal) ------------, pe2dalia, quarter rudders, 233-4
sekreton†, 90 ------------, peismata, mooring lines, 253
sharı3‘a†, 95 ------------, periago2geus, windlass/capstan,
ships (in general) 213
------, Anglo-Saxon ------------, peritonaia, 200 & n. 111
------------, scip, 412 & n. 21 ------------, phalke2s, deutera tropis, second
------, Byzantine keel/keelson, 196 n. 93
------------, cheimarrous, bung hole, 201 ------------, protonoi, forestays, 218, 253-4
------------, ‘naves’†, 117 ------------, prymne2sia, stern mororing
------------, ‘naves maxime que dromones lines, 253
748 INDICES
------------, sanis/sanidion/sanido2ma, ------------, carchesium, mast head, 244
deck, 128-9, 227-8; sanides. Planks ------------, cymbae, light boats/skiffs, 171
of a deck, 199 ------------, dromundus, 412
------------, scuppers, 201-2 ------------, ‘la2t6ana’ [pl. ‘lawa2t6in’], 154 n.
------------, side2robolia, anchors, 212 n. 82
148 ------------, latenae, 153
------------, stamines, upper futtocks of ------------, ‘mara2kib’†, 103, 107
frames, 199 ------------, naves†, 103, 106-7, 109, 116-
------------, stege2/stegos, deck, 128 17
------------, stropheion, windlass/ capstan, ------------, prodanus/prodano/peronus,
213 from protonos, mast-lowering
------------, tropis/tropideion, keel, 196 cable, 218 & n. 164
------------, tropos/zygon, cross beam, 200 ------------, stemonaria, upper futtocks,
------------, zo2ste2r, wale, 200 & n. 113 199
------, Modern Greek ------------, ‘sufun’†, 107
------------, despentsa, storeroom, 228 ------------, tertiarola, storm sails, 247
------------, ergate2s, capstan, 213 n. 151 ------------, trozze/trocte, et var., parrels,
------------, kaponi, cathead, 210 n. 144 247
------------, kassaron, poop, 228 ships’ boats, 132 & n. 30, 254, 270
------------, koubari, spool/reel, 214-15 n. signals/signalling, 272, 389, 392, 396-
154 9
------------, koumpania, storeroom, 228 ------, banda, signal/squadron flags,
------------, mpouka, bung, 201 n. 115 396-7
------------, trotsa, parrel, 247 ------, kamelaukion, signal flag, 396-9,
------, Muslim 503 n. 36
------------, al-s5a2rı3, mast, 235-6 ------------, kephale2, head, 396-9
------------, but6sa†, 119 ------, mirrors, 396
------------, ‘kate2nai’†, 169 ------, phlamoulon, ship’s standard/ tail
------------, mara2kib†, 43, 47, 68-70, 76, or streamer of signal flag, 272,
98, 169, 414 397-8
------------, qawa2rib†, 98 ------, smoke, 396
------------, quarter rudders, 242 spur/spurs, in general, 127, 131, 134-
------------, qunbar/‘koumbarion’ 513 n. 46, 152, 163 n. 1, 384, 423, 446,
51 450, 636-8
------------, sufun†, 164 n. 8 ------, chelandia, 209
------------------, ‘naves’†, 109 ------------, peronai, 209
------, Roman ------, dromons, 203-10 (& nn. 124,
------------, chalatorii, halyards, 230 & n. 136), 448
208 ------------, couplings, 138, 203
------------, constratum, deck, 129 ------------------, katakorakes, 208-9 & n.
------------, falx, sickle, rigging cutter, 230 138
------------, forus, deck, 129 ------------, peronia, 208-9
------------, funes, halyards, 230 n. 208 ------, Western galleys, 203, 424, 426,
------------, pons, deck, 129 430
------------, stega, deck, 129 ------------, calcar/speronus 209, 144
------------, tecta, deck, 129 ------------, naves rostratae, galleys with
------, Western spurs, 332, 415
------------, arbor de medio, 231 strate2goi†, various, 15, 31, 62, 77, 90,
------------, barca, 154 n. 82 133, 176, 180-81, 192 n. 82, 211,
------------, bardone/bradone, backstay, 215 n. 156, 258-9, 266-9, 271 (nn.
218 & n. 163 263, 264), 272, 354, 360, 386-7,
------------, ‘but6sa2t’†, 117, 120 391, 393, 395, 405, 450
------------, calcet/cholzexe, et var., block- ------, of Aigaion Pelagos*, 77, 256,
mast, 244 (& nn. 263, 264) 259, 266-7, 372 & n. 587, 391
------------, capone, cat tackle, 210 n. 144 ------, of Anatolikon*, 31, 46, 184, 391
------------, car, lower spar of a yard, 253 ------, of Armeniakon*, 391
n. 298 ------, of Calabria, 68, 271 & n. 363
INDICES 749
------, of Cherso2n*, 391 ------, paraplous, sailing past, 382, 383
------, of Dalmatia, 391 ------, periplous, sailing around, 382
------, of Hellas*, 31, 62 ------, shipwreck, 391-2
------, of the Kibyrrhaio2tai*, 32, 41, 46, ------, storms, 391-2
72, 77, 88, 189, 191, 256, 259, ------, sun, at the back/in the eyes, 391
266-7, 307, 353, 385, 390-91, 399- tagmata, squadrons of a fleet, 397
400 tarsianatus, keepers of arsenals, 318
------, of Longobardia*, 65, 166, 168 themata†, and armies, 77, 89-90, 391,
------, of Nikopolis*, 394 423
------, of Peloponne2sos, 324 n. 483 ------, naval, 32, 46-7, 64, 77, 88, 90,
------, of Ragusa*, 408 189, 191, 259, 260 & n. 323, 261,
------, of Samos, 77, 256, 259, 266-7, 266-7, 286, 307-8, 327, 371, 376,
306, 376, 391 384, 391 (see also Aigaion
------, of Thrake2sio2n*, 47, 391, 396 Pelagos*, Kibyrrhaio2tai*, Samos,
strate2goi plo2imoi, naval strate2goi†, themata)
386 ------, squadrons of a fleet, 267-8, 395,
strate2gos autokrato2r, 77 397
sult6a2n†, 94 ------, various, 25, 31, 46, 62 & n. 100,
sumps, 366-7 67, 69, 324 n. 483, 371, 394
touldos/touldon, baggage train, 305,
tactics, 8, 66, 130, 382-406 (see also 315, 395
Greek Fire) tourma, 267
------, ambushes, 388, 391, 397
------, battle lines, 208 visibility, 388-9
------, caution, 387 & n. 622
------, control of the land, 390 water/watering, 105-6, 264-6, 333,
------, diekplous, sailing through, 382, 352, 354-6 (& n. 544), 357-78,
545 n. 87 390, 443, 449
------, desmos/desmein, couple/to ------, amphorae/kadoi/pithos, 361-2 &
couple, 403-4 n. 564, 368-71, 373, 377
------------, akontia/kontaria, poles pre- ------, aqueducts, 373, 377
venting coupling, 404 ------, barrels, 362-3 & nn. 565-6, 363,
------------, kamakes side2rai, grappling 367-71, 373, 377
rods, 404 ------, boutia, 362-3
------, engaging off enemy coasts, 388 & ------, buckets, 368
n. 624 ------, ka(o)ly(m)bomatoi, water tanks?,
------, feigned flight, 400 363-7
------, formations, 208, 392, 395, 399- ------, water skin, byrsa/phlaske2/asko-
402, 450-51 daula, et var., 367 & n. 580
------------, counter formations, 399-400 ------, weight of, 359-60
------------, crescent-moon, 400-402, 451 ------, wells, 368-9, 373, 377
------, grappling, 208, 451 water resistance, 143 & n. 57
------------, links/iron rods, 208 weather, 191 & n. 78, 389, 392-3
------, hulls, holing of, 405-6 winds (see also meltemi)
------, hyperkerasai, to outflank, 382-3 ------, coastal, 359
------, kyklikon, encircling, 382 ------, contrary, 65 & n. 113, 343-4
------, menaula, pikes, use of, 405 ------, prevailing, 191 n. 78, 264, 333,
------, missiles, 208, 236, 384, 402-3, 335-6 (& nn. 512-14, 516), 340
450-51
------, night attacks, 391-2
ACKNOWLEDGMENTS
Figure 1 Liburnae in the Vatican Vergil (Rome, Biblioteca

Apostolica Vaticana, MS. Vat. Lat. 3225, fol. 43v);
by permission of the Biblioteca Apostolica Vaticana.
Figure 2 Liburnae on Trajan’s column, ca 114 C.E.; by
permission of Professor Sheppard Frere.
Figure 3 Galley on a mosaic from the baths at Themetra, near
Hadrumetum, Tunisia, ca 200-220 C.E.
Figure 4 Dromons in the Roman Vergil (Rome, Biblioteca
Apostolica Vaticana, MS. Vat. Lat. 3867, fol. 77r);
Figure 5 Catalan galley on a painted beam from a church near
Teruel (Barcelona, Museu Nacional d’Art de
Catalunya, Ref. 15839); by permission of the Museu
Nacional d’Art de Catalunya. ©MNAC - Museu
Nacional d’Art de Catalunya, Barcelona, 2006.
Photographers: Calveras/Mérida/Sagristà.
Figure 6 Dromon in the Ilias Ambrosiana (Milan, Biblioteca
Ambrosiana, Cod. Ambros. F. 205 Inf., min. VIII);
by permission of the Biblioteca Ambrosiana, Milan.
Figure 7 Dromons in the Ilias Ambrosiana (Milan, Biblioteca
Ambrosiana, Cod. Ambros. F. 205 Inf., min.
XXXVIII); by permission of the Biblioteca
Ambrosiana, Milan,
Figure 8 Two-masted, lateen-rigged dromons in a manuscript of
the Sacra Parallela attributed to St John of
Damascus (Paris, Bibliothèque Nationale, MS. Gr.
923, fol. 207r); by permission of the Bibliothèque
Nationale de France.
Figure 9 Dromons rolling over Russian ships with their spurs in
the Synopsis historio2n of John Skylitze2s (Madrid,
Biblioteca Nacional, vitr. 26-2, fol. 130r); by
permission of the Biblioteca Nacional de España,
Madrid.
Figure 10 The Athlit ram, third-second centuries B.C.E.;
photograph courtesy of the photographer, Professor
William M. Murray.
Figure 11 Caulkers at work extracting old pitch from the hull of a
752 ACKNOWLEDGMENTS
ship in a manuscript of the De materia medica of

Dioskoride2s (N.Y., Pierpont Morgan Library, Cod.
652, fol. 240r); by permission of the Morgan
Library, New York.
Figure 12 Square sail on a galley in a manuscript of the Bible
commissioned by abbot Vivian of St Martin of Tours
(Paris, Bibliothèque Nationale, MS. Lat. 1, fol. 3v);
by permission of the Bibliothèque Nationale de
France.
Figure 13 Lateen sails on dromons in the Ilias Ambrosiana
(Milan, Cod. Ambros. F. 205 Inf., min. XXVII),
early sixth century; after A. Mai, Homeri et Virgili
picturae antiquae (Rome, 1835).
Figure 14 Lateen? sails on dromons in the Ilias Ambrosiana
(Milan, Cod. Ambros. F. 205 Inf., min. XXVII); by
permission of the Biblioteca Ambrosiana, Milan.
Figure 15 Lateen-rigged ship in a manuscript of the Sermons of St
Gregory of Nazianzos (Paris, Bibliothèque
Nationale, MS. Gr. 510, fol. 367v); by permission of
the Bibliothèque Nationale de France.
Figure 16 Lateen-rigged ship in a manuscript of the Psalms, the
Khludov Psalter (Moscow, Historical Museum, MS.
129 D, fol. 88r).
Figure 17 Two-masted, lateen-rigged ship in a painting from
Kellia, Egypt, from a drawing in L. Basch, “Navires
et bateaux coptes: état des questions en 1991”,
Graeco-Arabica, 5 (1993), Fig. 25 (p. 53); by
permission of Professor V. Christides, editor,
Graeco-Arabica.
Figure 18 Galley on a lustre-ware bowl from al-Fayyu2m, Egypt
(Cairo, Museum of Islamic Art, Inv. No. 7900).
Figure 21 Through-hull rudder mounts on a galley representing
the constellation Argo in a manuscript of the Aratea
attributed to Germanicus Caesar (Boulogne-sur-Mer,
Bibliothèque Municipale, MS. 188, fol. 78.); by
permission of the Bibliothèque Municipale,
Boulogne-sur-Mer.
Figure 22 Chelandium on a medal forged by Alvise Meneghetti
(1691-1768) attributed to a Venetian Doge Pietro
Candiano as published by Augustin Jal; after A. Jal,
Glossaire nautique: répertoire polyglotte de termes
ACKNOWLEDGMENTS 753
de marine anciens et modernes (Paris, 1848), p. 465.

Figure 23 Ship borne siege engines in a treatise on poliorcetics
attributed to He2ro2n of Byzantium (Rome, Biblioteca
Apostolica Vaticana, MS. Vat. Gr. 1605, fol. 40r);
Figure 24 Mosaic of a galley with a lowered mast from a
sepulchre at Hadrumetum, Tunisia, third century.
Figure 26 Dromons in the Kyne2getika of Pseudo-Oppian (Venice,
Biblioteca Marciana, MS. Gr. 479 [coll. 881], fol.
23r); by permission of the Biblioteca Nazionale
Marciana.
Figure 34 The fleet of Thomas the Slav advancing on Abydos and
carrying horses aboard a galley in the Synopsis
historio2n of John Skylitze2s (Madrid, Biblioteca
Nacional, vitr. 26-2, fol. 31v); by permission of the
Biblioteca Nacional de España.
Figure 36 Mosaic of a galley from Piazza Armerina, Sicily, early
fourth century; photograph, John H. Pryor.
Figure 37 Horses unloaded from ports at the sterns of galleys in a
manuscript of Les livres des histoires du
commencement du monde (Paris, Bibliothèque
National, Ms. Fr. 301, fol. 58v); by permission of the
Bibliothèque Nationale de France.
Figure 43 The cista Ficoronica: Jason and the Argonauts watering
at the spring of the Bebrycians; by permission of
Dietrich Reimer Verlag, Gebr. Mann Verlag,
Deutscher Verlag für Kunstwissenschaft.
Figure 46 The harbour of Kastron Samos, adapted from Great
Britain, Admiralty, Hydrographic Office, Chart No
1568 of 1967.
Figure 47 Dromon in a manuscript of the Sermons of St Gregory
of Nazianzos (Mount Athos, Pantelee2mon, Cod. 6;
fol. 138r), twelfth century.
Figure 48 Graffito of a katergon? From the monastery of the
Blatado2n at Thessalonike2; after O. Meinardus,
“Medieval navigation according to akidographemata
in Byzantine churches and monasteries”, Deltivon th'"
Cristianikh'" Arcaiologikh'" ÔEtaireiva", per. DV, 6
(1970-72), 29-52.
Figure 49 Graffito of a katergon? from Hagia Sophia, Trebizond;
after A. Bryer, “Shipping in the empire of
754 ACKNOWLEDGMENTS
Trebizond”, Mariner’s Mirror, 52 (1966), 3-12; fig.

6.
Figure 50 Galleys in the Annales Ianuenses of Genoa (Paris,
Bibliothèque Nationale, MS. Suppl. Lat. 773); after
L. T. Belgrano and C. Imperiale, eds, Annali
genovesi di Caffaro e de’ suoi continuatori dal
MXCIX al MCCXCIII, 5 vols [FStI] (Rome and
Genoa, 1890-1929).
Figures 51-3 Bireme galleys in the Synopsis historio2n of John
Skylitze2s (Madrid, Biblioteca Nacional, vitr. 26-2,
fols 111v, 145r, 146v); by permission of the
Biblioteca Nacional de España, Madrid.
Figure 54 Sicilian galley in a manuscript of the De rebus Siculis
carmen of Peter of Eboli (Bern, Burgerbibliothek,
Cod. 120, fol. 119r); by permission of the
Burgerbibliothek, Bern.
Figure 57 Dromon using Greek Fire in the Synopsis historio2n of
John Skylitze2s (Madrid, Biblioteca Nacional, vitr.
26-2, fol. 34v); by permission of the Biblioteca
Nacional de España, Madrid.
Figure 58 Wolfenbüttel, Herzog August Bibliothek, Cod. Guelf.,
96 Gud. lat., fol. 157r-v; by permission of the
Herzog August Bibliothek, Wolfenbüttel.
Figure 59 Soldier using a hand-held flame thrower in a treatise on
poliorcetics attributed to He2ro2n of Byzantium
(Rome, Biblioteca Apostolica Vaticana, MS. Vat.
Gr. 1605, fol. 36r); by permission of the Biblioteca
Apostolica Vaticana.
Figure 60 The Sotiel Coronada Roman force pump from
Valverde, Huelva, Spain, probably dated to the first
century C.E. (Madrid, Museo Arqueologico
Nacional); by permission of the Museo
Arqueologico Nacional.
Figure 61 “Greek” or liquid fire sipho2n built by Colin Hewes and
Andrew Lacey under the direction of John Haldon;
by permission of J. Haldon, C. Hewes, and A. Lacey.
THE
MEDIEVAL MEDITERRANEAN
PEOPLES, ECONOMIES AND CULTURES, 400-1500
Editors: Hugh Kennedy, Paul Magdalino (St. Andrews), David Abulafia

(Cambridge), Benjamin Arbel (Tel Aviv), Mark Meyerson (Toronto),
Larry J. Simon (Western Michigan University).
This series provides a forum for the publication of scholarly work relating to the
interactions of peoples and cultures in the Mediterranean basin and the Black Sea
area and is intended for readers with interest in late antiquity, the Middle Ages
(Italy, Spain, the Latin East), Byzantium, Islam, the Balkans and the Black Sea.
Manuscripts (in English, German and French) should be 60,000 to 120,000 words
in length and may include illustrations. The editors would be particularly
interested to receive proposals for monograph studies; studies with texts; editions
with parallel translations of texts or collections of documents; or translations
provided with full annotation.
1. Shatzmiller, M. (ed.). Crusaders and Muslims in Twelfth-Century Syria. 1993.

ISBN 90 04 09777 5
2. Tsougarakis, D. The Life of Leontios, Patriarch of Jerusalem. Text, Translation,
Commentary. 1993. ISBN 90 04 09827 5
3. Takayama, H. The Administration of the Norman Kingdom of Sicily. 1993.
ISBN 90 04 09865 8
4. Simon, L.J. (ed.). Iberia and the Mediterranean World of the Middle Ages. Studies in
Honor of Robert I. Burns S.J. Vol. 1. Proceedings from Kalamazoo. 1995.
ISBN 90 04 10168 3
5. Stöckly, D. Le système de l’Incanto des galées du marché à Venise (fin XIII e- milieu XV e
siècle. 1995. 90 04 10002 4.
6. Estow, C. Pedro the Cruel of Castile, 1350-1369. 1995. ISBN 90 04 10094 6
7. Stalls, W.C. Possessing the Land. Aragon’s Expansion into Islam’s Ebro Frontier
under Alfonso the Battler, 1104-1134. 1995. ISBN 90 04 10367 8
8. Chevedden, P.E., D.J. Kagay & P.G. Padilla (eds.). Iberia and the Mediterranean
World of the Middle Ages. Essays in Honor of Robert I. Burns S.J. Vol. 2.
Proceedings from ‘Spain and the Western Mediterranean’, a Colloquium
Sponsored by The Center for Medieval and Renaissance Studies, University of
California, Los Angeles, October 26-27, 1992. 1996. ISBN 90 04 10573 5
9. Lev, Y. (ed.). War and Society in the Eastern Mediterranean, 7th-15th Centuries. 1997.
ISBN 90 04 10032 6
10. Ciggaar, K.N. Western Travellers to Constantinople. The West and Byzantium,
962-1204: Cultural and Political Relations. 1996. ISBN 90 04 10637 5
11. Skinner, P. Health and Medicine in Early Medieval Southern Italy. 1997.
ISBN 90 04 10394 5
12. Parry, K. Depicting the Word. Byzantine Iconophile Thought of the Eighth and
Ninth Centuries. 1996. ISBN 90 04 10502 6
13. Crisafulli, V.S. & J.W. Nesbitt. The Miracles of St. Artemios. A Collection of
Miracle Stories by an Anonymous Author of Seventh-Century Byzantium.
1997. ISBN 90 04 10574 3
14. Antonopoulou, T. The Homilies of the Emperor Leo VI. 1997.
ISBN 90 04 10814 9
15. Tougher, S. The Reign of Leo VI (886-912). Politics and People. 1997.
ISBN 90 04 10811 4
16. O’Callaghan, J.F. Alfonso X and the Cantigas de Santa Maria. A Poetic Bio-
graphy. 1998. ISBN 90 04 11023 2
17. Gilmour-Bryson, A. The Trial of the Templars in Cyprus. A Complete English
Edition. 1998. ISBN 90 04 10080 6
18. Reyerson, K. & J. Drendel (eds.). Urban and Rural Communities in Medieval
France. Provence and Languedoc, 1000-1500. 1998. ISBN 90 04 10850 5
19. Kagay, D.J. & T.M. Vann (eds.). On the Social Origins of Medieval Institutions.
Essays in Honor of Joseph F. O’Callaghan. 1998. ISBN 90 04 11096 8
20. Ferreiro, A. (ed.). The Visigoths. Studies in Culture and Society. 1999.
ISBN 90 04 11206 5
21. Lev, Y. Saladin in Egypt. 1999. ISBN 90 04 11221 9
22. Burns, R.I., SJ, P.E. Chevedden & M. de Epalza. Negotiating Cultures. Bilingual
Surrender Treaties in Muslim-Crusader Spain under James the Conqueror.
1999. ISBN 90 04 11230 8
23. Webster, J.R. Carmel in Medieval Catalonia. 1999. ISBN 90 04 11435 1
24. Bareket, E. Fustat on the Nile. The Jewish Elite in Medieval Egypt. 1999.
ISBN 90 04 11439 4
25. Daileader, P. True Citizens. Violence, Memory, and Identity in the Med-ieval
Community of Perpignan, 1162-1397. 2000. ISBN 90 04 11571 4
26. Hames, H.J. The Art of Conversion. Christianity and Kabbalah in the Thir-
teenth Century. 2000. ISBN 90 04 11715 6
27. Fryde, E. The Early Palaeologan Renaissance (1261-c. 1360). 2000.
ISBN 90 04 11714 8
28. Smith, J.M.H. Early Medieval Rome and the Christian West. Essays in Honour of
Donald A. Bullough. 2000. ISBN 90 04 11716 4
29. Andrea, A.J. Contemporary Sources for the Fourth Crusade. With Contributions by
Brett E. Whalen. 2000. ISBN 90 04 11740 7
30. Brett, M. The Rise of the Fatimids. The World of the Mediterranean and the
Middle East in the Fourth Century of the Hijra, Tenth Century A.D. 2001.
ISBN 90 04 11741 5
31. Kennedy, H. (ed.). The Historiography of Mamluk Egypt (c. 950-1800). 2001.
ISBN 90 04 11794 6
32. Orvietani Busch, S. Medieval Mediterranean Ports. The Catalan and Tuscan
Coasts, 1100 to 1235. 2001. ISBN 90 04 12069 6
33. NecipoÅlu, N. Byzantine Constantinople. Monuments, Topography and
Everyday Life. 2001. ISBN 90 04 11625 7
34. Stewart, A.D. The Armenian Kingdom and the Mamluks. War and Diplomacy
during the Reigns of Het#um II (1289-1307). 2001. ISBN 90 04 12292 3
35. Peltomaa, L.M. The Image of the Virgin Mary in the Akathistor Hymn. 2001.
ISBN 90 04 12088 2
36. Mavroudi, M. A Byzantine Book on Dream Interpretation. The Oneirocriticon of
Achmet and Its Arabic Sources. 2002. ISBN 90 04 12079 3
37. Reyerson, K.L. The Art of the Deal. Intermediaries of Trade in Medieval
Montpellier. 2002. ISBN 90 04 12129 3
38. Loud, G.A. & A. Metcalfe (eds.). The Society of Norman Italy. 2002.
ISBN 90 04 12541 8
39. Lev, Y. (ed.). Towns and Material Culture in the Medieval Middle East. 2002.
ISBN 90 04 12543 4
40. Pastor, R., E. Pascua, A. Rodríguez-López & P. Sánchez-León. Beyond the
Market. Transactions, Property and Social Networks in Monastic Galicia
1200-1300. 2002. ISBN 90 04 11953 1
41. Parani, M.G. Reconstructing the Reality of Images. Byzantine Material Culture
and Religious Iconography 11th-15th Centuries. 2003.
ISBN 90 04 12462 4
42. Holmes, C. & J. Waring (eds.). Literacy, Education and Manuscript Transmission in
Byzantium and beyond. 2002. ISBN 90 04 12096 3
43. VanLandingham, M. Transforming the State. King, Court and Political Culture
in the Realms of Aragon (1213-1387). 2002. ISBN 90 04 12743 7
44. O’Connor, I.A. A Forgotten Community. The Mudejar Aljama of Xàtiva,
1240-1327. 2003. ISBN 90 04 12846 8
45. Magdalino, P. Byzantium in the year 1000. 2003. ISBN 90 04 12097 1
46. Zeldes, N. “The Former Jews of this Kingdom”. Sicilian Converts after the
Expulsion, 1492-1516. 2003. ISBN 90 04 12898 0
47. Moore, J.C. Pope Innocent III (1160/61–1216). To Root Up and to Plant.
2003. ISBN 90 04 12925 1
48. Kelly, S. The New Solomon. Robert of Naples (1309-1343) and Fourteenth-
Century Kingship. 2003. ISBN 90 04 12945 6
49. Nesbitt, J.W. Byzantine Authors: Literary Activities and Preoccupations. Texts
and Translations dedicated to the Memory of Nicolas Oikonomides. 2003.
ISBN 90 04 12975 8
50. Edbury, P.W. (ed.). John of Ibelin. Le Livre des Assises. 2003.
ISBN 90 04 13179 5
51. Winter, M. & A. Levanoni (eds.). The Mamluks in Egyptian and Syrian Po-
litics and Society. 2004. ISBN 90 04 13286 4
52. Hames, H.J. (ed.). Jews, Muslims and Christians in and around the Crown of
Aragon. Essays in honour of Professor Elena Lourie. 2004.
ISBN 90 04 12951 0
53. Alfonso, I., H. Kennedy, & J. Escalona (eds.). Building Legitimacy.
Political Discourses and Forms of Legitimacy in Medieval Societies. 2004.
ISBN 90 04 13305 4
54. Hartnup, K. ‘On the Beliefs of the Greeks’. Leo Allatios and Popular Ortho-
doxy. 2004. ISBN 90 04 13180 9
55. Hamilton, B., J. Hamilton & S. Hamilton. Hugh Eteriano: Contra Patare-
nos. 2004. ISBN 90 04 14000 X
56. Andrews, F., C. Egger & C.M. Rousseau (eds.). Pope, Church and City.
Essays in Honour of Brenda M. Bolton. 2004. ISBN 90 04 14019 0
57. Hunt, H. Joy-bearing Grief. Tears of Contrition in the Writings of the Early
Syrian and Byzantine Fathers. 2004. ISBN 90 04 14123 5
58. Nicolaou-Konnari, A. & C. Schabel (eds.). Cyprus. Society and Culture
1191-1374. 2005. ISBN 90 04 14767 5
59. Drews, W. The Unknown Neighbour. The Jew in the Thought of Isidore of
Seville. 2006. ISBN 90 04 14964 3
60. Böninger, L. Die deutsche Einwanderung nach Florenz im Spätmittelalter. 2006.
ISBN 90 04 15047 1.
61. Barber, C. & D. Jenkins (eds.). Reading Michael Psellos. 2006.
ISBN 90 04 15180 X.
62. Pryor, J.H. & E.M. Jeffreys. The Age of the DROMWN . The Byzantine
Navy ca 500-1204. 2006. ISBN-13: 978-90-04-15197-0,
ISBN-10: 90-04-15197-4

Byzantine Navy 500-1204

Uploaded by

Copyright:

Available Formats

Byzantine Navy 500-1204

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Byzantine Navy 500-1204

Uploaded by

Copyright:

Available Formats

THE AGE OF THE DROMWN

Hugh Kennedy (St. Andrews)

This book is printed on acid-free paper.

Library of Congress Cataloging-in-Publication Data

© Copyright 2006 by Koninklijke Brill NV, Leiden, The Netherlands

Authorization to photocopy items for internal or personal

printed in the netherlands

Preface and appreciations………………………………...…….. xi

List of figures, maps, and tables ……………………………...... xiii

Note on orthography, translation, citation, and dating ………… xix

Note on metrology and hours of daylight ……………………… xxiii

Selective gazeteer of historical place names …………………... xxv

Selective glossary of Greek, Latin, and Arabic terminology ….. xxxix

Glossary of English nautical terminology ……………………... xlix

Note on citation of Greek and Latin glossaries ………………... lix

Abbreviations ………………………………………………….. lxiii

Maps …………………………………………………………… lxv

Chapter One: the operational context ………………………….. 7

Chapter Two: the origins of the Dromon …………………….... 123

Chapter Three: from the sixth to ninth centuries ………………. 163

Chapter Four: the Dromon in the age of the Macedonian

Chapter Five: The demise of the Dromon ……………………... 407

Chapter Six: The triumph of the Galea ………………………... 423

Conclusion ……………………………………………………... 445

Appendix One: Syrianos Magistros, Naumacivai Surianou'

Appendix Two: Leo VI, Naumacika; Levonto" Basilevw", edition

Appendix Three: Anonymous, Naumacika; suntacqevnta para;

Appendix Four: fleets, armaments, and equipment for dromons,

Appendix Five: Nike2phoros Ouranos, Peri; qalassomaciva",

Appendix Six: Greek Fire ……………………………………... 607

Appendix Seven: The galleys of the manuscript, Madrid,

Appendix Eight: Muh5ammad ibn Mankalı3, [a] Al-adilla al-

Bibliography and abbreviations …………….....………………. 667

Indices …………………………………………………………. 711

Acknowledgments …………………………………………....... 751

We have shared a happy collaboration in this book. Our friendship

Frontispiece: Reconstruction of a bireme dromon of the era of

Figure 12: Square sail on a galley in a manuscript of the

century ……………………………………….... 249

early fourth century …………………….……... 311

Skylitze2s (Madrid, Biblioteca Nacional, vitr.

Map 1: The Atlantic and western Mediterranean:

Map 7: The southern Aegean and Lycia ………………... lxxi

Table 1: Rulers of the first period, ca 400-560 ……………. 10

proper nouns except where there is a common English equivalent, thus

In translating Greek, Latin, and Arabic texts we have have kept in

Citation of Primary Sources

In terms of the practical limitations of medieval shipbuilding and also

British Imperial Metric

[b] Hours of Daylight

For hours of daylight we have utilized the United States Naval

Ar. Arabic lit. literally

Abydos “Abudo", et var. (cl. & Byz. Gr.), Abydus, et var.

var. (med. Lat.), al-Farama2’ (Ar.): town, al-

Armeniakon Armeniakovn: original thema† of Asia Minor,

Bu2na, et var. (Ar.), Bona, et var. (med. Lat.):

the road on the W coast of the Black Sea, 20

Ifrı3qiya Ifrı3qiya (Ar.): Arabic name for the R.-Byz. prov. of

coast of Catalonia, NE of Palamos, Spain.

Pannonia Pannoniva (cl. & Byz. Gr.), Pannonia (cl. Lat.),