Madgearu, Alexandru - Byzantine Military Organization On The Danube, 10th-12th Centuries

Byzantine Military Organization on the Danube,
10th–12th Centuries
East Central and Eastern
Europe in the Middle Ages,
450–1450
General Editor
Florin Curta
VOLUME 22
The titles published in this series are listed at brill.com/ecee

Byzantine Military Organization
on the Danube,
10th–12th Centuries
By
Alexandru Madgearu
LEIDEN • BOSTON
2013
Cover illustration: Salterio di Basilio II, Marc. Gr. Z. 17 (=421), f. IIIr. (Basil II Bulgaroctonus,
miniature). Su concessione del Ministero per i Beni e le Attività Culturali—Biblioteca Nazionale
Marciana.
Library of Congress Cataloging-in-Publication Data
Madgearu, Alexandru.
[Organizarea militara bizantina la Dunare in secolele X–XII. English]
Byzantine military organization on the Danube, 10th–12th centuries / by Alexandru Madgearu.
pages cm. — (East Central and Eastern Europe in the Middle Ages, 450–1450 ; Volume 22)
Includes bibliographical references and index.
ISBN 978-90-04-21243-5 (hardback : alk. paper) — ISBN 978-90-04-25249-3 (e-book) 1. Danube
River Region—History, Military. 2. Byzantine Empire—History, Military—527–1081. 3. Byzantine
Empire—History, Military—1081–1453. I. Title.
DR49.26.M3313 2013
355.309495—dc23
2013015322
This publication has been typeset in the multilingual “Brill” typeface. With over 5,100 characters
covering Latin, IPA, Greek, and Cyrillic, this typeface is especially suitable for use in the humanities.
For more information, please see www.brill.com/brill-typeface.
ISSN 1872-8103
ISBN 978-90-04-21243-5 (hardback)
ISBN 978-90-04-25249-3 (e-book)
Copyright 2013 by Koninklijke Brill NV, Leiden, The Netherlands.

Koninklijke Brill NV incorporates the imprints Brill, Global Oriental, Hotei Publishing,
IDC Publishers and Martinus Nijhoff Publishers.
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 Koninklijke Brill NV
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.
This book is printed on acid-free paper.

Contents
List of Illustrations ........................................................................................... vii

List of Abbreviations ....................................................................................... ix
Introduction ...................................................................................................... 1
I. The Recovery of the Danubian Frontier ........................................... 7
II. The Military Organization of the Danube Region ......................... 59

1 The Theme of Dristra (later, Paradunavon) ............................. 59
2 The strategoi as City Commanders in the Theme
Dristra/Paradunavon ....................................................................... 88
3 The Theme of Sirmium and the New Bulgarian Theme
on the Middle Danube .................................................................... 95
III. The Evolution and Function of the Danube Frontier of

Byzantium (1000–1204) ........................................................................... 101
1 The Fortifications .............................................................................. 101
2 The Danube Frontier in the 11th Century ................................. 115
3 The Danube Frontier in the 12th Century ................................. 144
Conclusion ......................................................................................................... 167
Bibliography ...................................................................................................... 173
Index of People ................................................................................................. 199
Index of Geographical Names ...................................................................... 204
Index of Sources ............................................................................................... 210
Index of Modern Authors .............................................................................. 211
List of Illustrations
1. The dikes between Cernavoda and Constanţa (after Barnea,

Ştefănescu 1971, 168, map IV) ............................................................... 25
2. The military organization between 971 and 986 ........................... 44
3. The Danubian region in 1000–1004 .................................................... 52
4. The Danubian region in 1018/1019 ...................................................... 58
5. The fort and settlement of Nufăru (after Damian et alii
2012, 195) ..................................................................................................... 103
6. The fort of Dinogetia (after Barnea 1980, Plate XV) ..................... 105
7. The group of fortifications Păcuiul lui Soare-Dervent-Oltina
(after Diaconu, Vâlceanu 1972, 10, fig. 1) ........................................... 109
8. The fortifications affected by the attacks of 1032–1035 ............... 119
9. The fortifications affected by the attack of 1036 ............................ 120
12. The fortifications affected by the rebellion of 1072–1091 ............ 133
13. The fortifications affected by the attack of 1122 ............................. 146
14. The fort of Braničevo (after Popović 1991, 171) ............................... 156
15. The fort of Belgrade (after Popović 1991, 172) ................................. 156
list of ABBREVIATIONS
AAnt Acta Antiqua Academiae Scientiarum Hungaricae,

Budapest
AB Archaeologia Bulgarica, Sofia
AEMA Archivum Eurasiae Medii Aevi, Wiesbaden
AIIAI Anuarul Institutului de Istorie şi Arheologie “A. D.
Xenopol”, Iaşi
AM Arheologia Moldovei, Iaşi
AMN Acta Musei Napocensis, Cluj-Napoca
AMV Acta Musei Varnensis, Varna
Arh Arheologija, Sofia
B Byzantion. Revue internationale des études byzan-
tines, Bruxelles
BB Byzantinobulgarica, Sofia
BBA Berliner Byzantinische Arbeiten, Berlin
BF Byzantinische Forschungen, Amsterdam
BHR Bulgarian Historical Review, Sofia
BM Bulgaria mediaevalis, ed. V. Gjuzelev, V. Tăpkova-
Zaimova, K. Nenov, Sofia
BMI Buletinul Monumentelor Istorice, Bucureşti
BMGS Byzantine and Modern Greek Studies, Birmingham
BNJ Byzantinisch-Neugriechische Jahrbücher, Athens
BS Byzantine Studies, Pittsburgh
BSH Bulletin de la Section Historique de l’Académie Rou-
maine, Bucarest
BSNR Buletinul Societăţii Numismatice Române, Bucureşti
Borders Borders, Barriers, and Ethnogenesis. Frontiers in Late
Antiquity and the Middle Ages (Centre for Medieval
Studies. University of York. Studies in the Early Mid-
dle Ages, 12), edited by Florin Curta, Turnhout, 2005
Byzantium and Byzantium and East Central Europe. Papers of a
East Central Symposium, Cracow, 24–26 September 2000, edited by
Europe G. Prinzing, M. Salamon, P. Stephenson, Cracow, 2001
Byzantium at War Byzantium at War (9th–12th c.). National Hellenic
Research Foundation, Institute of Byzantine
Research (International Symposium, 4), edited by
K. Tsiknakis, Athens, 1997
x list of abbreviations
Byzanz und Byzanz und seine Nachbarn, hrsg. von A. Hohlweg (Südost-
Nachbarn Europa Jahrbuch, 26), München, 1996
Byzsl Byzantinoslavica, Prague
BZ Byzantinische Zeitschrift, München
CCDJ Cultură şi civilizaţie la Dunărea de Jos, Călăraşi
CFHB Corpus Fontium Historiae Byzantinae
CIEB XII Actes du XIIe Congrès International d’Études Byzantines
(Ochride, 10–16 septembre 1961), 2 vol., Belgrade, 1964
CIEB XIV Actes du XIVe Congrès International des Études Byzantines,
3 vol., Bucurarest, 1974–1976
CIEB XV Actes du XVe Congrès International d’Études Byzantines
(Athènes, Septembre 1976), Athènes, 1980
CIEB XX XXe Congrès International des Études Byzantines. Pré-actes,
3 vol., Paris, 2001
CN Cercetări numismatice, Bucureşti
Cronica 1999 Cronica cercetărilor arheologice din România. Campania
1999, Deva, 2000
Cronica 2000 Cronica cercetărilor arheologice. Campania 2000, Suceava,
2001
Cronica 2001 Cronica cercetărilor arheologice. Campania 2001, Buziaş,
2002
2002, Sf. Gheorghe, 2003
2003, Cluj-Napoca, 2004
2004, Mangalia, 2005
2005, Constanţa, 2006
2006, Tulcea, 2007
2007, Iaşi, 2008
2008, Târgovişte, 2009
2009, Suceava, 2010
2010, Sibiu, 2011
list of abbreviations xi
Cronica 2011 Cronica cercetărilor arheologice din România. Campa-

nia 2011, Târgu Mureş, 2012
DAW Österreichische Akademie der Wissenschaften,
Philosophisch-Historische Klasse, Denkschriften
Dobrudža Dobrudža, Varna
Dobrudža. Études Dobrudža. Études ethno-culturelles, Sofia, 1987
DOP Dumbarton Oaks Papers, Washington DC
EAIVR Enciclopedia arheologiei şi istoriei vechi a României,
3 vol., Bucureşti, 1994–2000
EB Études Balkaniques, Sofia
EBP Études byzantines et post-byzantines, Bucureşti, Iaşi
ECEEMA East Central and Eastern Europe in the Middle Ages,
450–1450, General Editor F. Curta
Il Mar Nero Il Mar Nero. Annali di archeologia e storia, Roma-
Paris
INMV Izvestija na Narodnija Muzej (previously Izvestija na
Varnenskoto Arheologičesko Družestvo), Varna
Istro-Pontica Istro-Pontica. Muzeul tulcean la a 50-a aniversare,
1950–2000. Omagiu lui Simion Gavrilă la 45 de ani de
activitate, 1955–2000, Tulcea, 2000
JÖB Jahrbuch der Österreichischen Byzantinistik, Wien
MCA Materiale şi cercetări arheologice, Bucureşti
MGH Monumenta Germaniae Historica
MN Muzeul Naţional, Bucureşti
Peace and War Peace and War in Byzantium. Essays in Honor of
George T. Dennis S.J., edited by Timothy S. Miller and
John Nesbitt, Washington DC, 1995
Polychronion Polychronion. Festschrift Franz Dölger zum 75.
Geburstag, hrsg. von P. Wirth, Heidelberg, 1966
PP Pliska-Preslav, Šumen
Prinos Prinos lui Petre Diaconu la 80 de ani. Volum îngrijit
de Ionel Cândea, Valeriu Sârbu, Marian Neagu, Brăila,
2004
Proceedings Proceedings of the International Symposium dedicated
Haralanov to the Centennial of the Dr. Vassil Haralanov, held in
Shumen in September the 13th–15th 2007, coord. Ivan
Jordanov, Shumen, 2008
RdI Revista de istorie, Bucureşti
REB Revue des Études Byzantines, Paris
xii list of abbreviations
RESEE Revue des Études Sud-Est Européennes, Bucarest

RHSEE Revue Historique du Sud-Est Européen, Bucarest
RI Revista Istorică, Bucureşti
RIR Revista Istorică Română, Bucureşti
RRH Revue Roumaine d’Histoire, Bucarest
SB Studia Balcanica, Sofia
SBS Studies in Byzantine Sigillography, Washington DC
SCIV(A) Studii şi cercetări de istorie veche (şi arheologie),
Bucureşti
SCN Studii şi cercetări de numismatică, Bucureşti
SEER The Slavonic and East European Review, London
SMIM Studii şi materiale de istorie medie, Bucureşti
SMMIM Studii şi materiale de muzeografie şi istorie militară,
Bucureşti
SOF Südost-Forschungen, München
SRH Scriptores Rerum Hungaricarum tempore ducum
regumque stirpis Arpadianae gestarum. Edendo operi
praefuit Emericus Szentpétery, 2 vol., Budapest, 1937–
1938
Studii Studii. Revistă de istorie, Bucureşti
TM Travaux et Mémoires, Centre de Recherche d‘Histoire
et Civilisation Byzantines, Paris
Vyzantina Vyzantina. Epistemonikon organon Kentrou Vyzan-
tinon Ereunon Philosophikes Scholes Aristoteleiou
Panepistemiou, Thessaloniki
Von der Scythia Von der Scythia zur Dobrudža, herausgegeben von Ch.
Choliolčev, R. Pillinger, R. Harreither (Miscellanea
Bulgarica, 11), Wien, 1997
ZRVI Zbornik Radova Vizantoloshkog Instituta, Belgrade
Introduction
The first monograph offering a comprehensive survey of the Byzantine

military organization in Bulgaria and Dobrudja was written by Nicolae
Bănescu in 1946,1 as final result of his studies published between 1922 and
1945 (for instance: Bănescu 1922; Un duc byzantin du XIe siècle: Katakalon
Kékauménos, BSH, 11, 1924, 25–36; Bănescu 1927; La question de Paristrion,
ou conclusion d’un long débat, B, 8, 1933, 277–308; Bizanţul şi romanitatea
de la Dunărea de Jos, Bucureşti, 1938). Besides literary sources (almost the
same as those available and known today), his work was based on a small
number of lead seals, and on very scarce archaeological data from the
fortifications in Capidava and Garvăn, where excavations had started few
years before his book was published (see G. Florescu, Fouilles et recherches
archéologiques à Calachioi (Capidava?) en 1924 et 1926, Dacia, 3–4, 1927–
1932, 483–515; Idem, Capidava în epoca migraţiilor, RIR, 16, 1946, 325–343;
Gh. Ştefan, Dinogetia, I. Risultati della prima campagna di scavi, Dacia,
7–8, 1937–1940 (1941), 401–425).
The book published in 1946 was the first one exclusively dedicated to
the military and administrative organization of the provinces Paradu-
navon and Bulgaria. Before Bănescu, Nicolae Iorga,2 Gheorghe I. Brătianu,3
Dragutin Anastasijević,4 Petăr Mutafčiev,5 Constantin Necşulescu6 have
dealt with this topic only tangentially. The first seals of military and civil-
ian dignitaries were published in the 1930s. Many came from Silistra, a city
within the borders of Romania at that time (see N. Bănescu, Les sceaux
byzantins trouvés à Silistrie, B, 7, 1932, 1, 321–331; N. Bănescu, P. Papahagi,
Plombs byzantins découverts à Silistrie, B, 10, 1935, 2, 601–606). Gheorghe I.
Brătianu’s outstanding monograph dedicated to the Black Sea also con-
tributed to the study of the Byzantine period in the Danubian region.
However, although written between 1946 and 1947, the book was pub-
lished only in 1969, as the author was sent to prison by the Communist
1 Bănescu 1946.
2 Iorga 1920, 33–46.
3 Brătianu 1935, 9–96; Brătianu 1942.
4 Anastasijević 1930, 20–36.
5 Mutafčiev 1932.
6 Necşulescu 1937, 122–151; Necşulescu 1939, 185–206.
2 introduction
regime, and the manuscript was taken to France by Vitalien Laurent, and
published there a cultural association of the Romanian diaspora).7
The development of the archaeological researches in Romania and Bul-
garia after 1948 led to the identification of other Byzantine fortifications,
in addition to those sites on which excavations continued: Capidava,
Garvăn, Isaccea, Păcuiul lui Soare (in the Romanian part of Dobrudja),8
Pliska, Preslav, Silistra and Varna (in Bulgaria). The next work building
upon of N. Bănescu’s was Ion Barnea’s contribution to the third volume of
the series “Din istoria Dobrogei”, published in 1971.9 Barnea studied mili-
tary organization on the basis of archaeological finds and of the then most
recently discovered seals. He also analyzed the list of offices published
by Nicolas Oikonomides in 1965. The monograph included the results of
archaeological excavations carried out by I. Barnea himself together with
Gheorghe Ştefan and Bucur Mitrea at Garvăn and Isaccea, and also took
into account the research of the younger generation of archaeologists
(Petre Diaconu and Radu Florescu) at Păcuiul lui Soare and Capidava. The
archaeological monograph of Dinogetia-Garvăn published in 1967 brought
a decisive contribution to drawing a clear chronology of the period.10 A
brief synthesis of the then current research on the Byzantine frontier in
the 10th–12th centuries was presented at the 13th International Congress of
Byzantine Studies (London, 1966).11 Several studies appeared in the 1960s
on the military history and historical geography of the Danubian region
in the Byzantine period, by such scholars as Alexandru Bolşacov-Ghimpu,12
Constantin Cihodaru,13 Petre Diaconu,14 Alexander Kuzev,15 Petre Ş.
Năsturel,16 Eugen Stănescu,17 Andrew Urbansky,18 Tadeusz Wasilewski.19
On the occasion of the 14th International Congress of Byzantine Stud-
ies in Bucharest (1971), a remarkable catalogue was published under the
7 Brătianu 1999.
8 The excavation reports were published each year in SCIV, and then in MCA.
9 Barnea, Ştefănescu 1971, 71–335.
10 Ştefan, Barnea 1967.
11 Condurachi, Barnea, Diaconu 1967.
12 Bolşacov-Ghimpu 1967; Bolşacov-Ghimpu 1973.
13 Cihodaru 1963; Cihodaru 1965; Cihodaru 1968.
14 Diaconu 1962; Diaconu 1965 a; Diaconu 1965 b; Diaconu 1966.
15 Kuzev 1966, 23–50; Kuzev 1967, 41–70; Kuzev 1968, 27–55; Kuzev 1969, 137–157.
16 Năsturel 1965; Năsturel 1966, 382–387; Năsturel 1969.
17 Stănescu 1966; Stănescu 1968 a; Stănescu 1968 b; Stănescu 1968 c; Stănescu 1970;
Stănescu 1971.
18 Urbansky 1968.
19 Wasilewski 1964.
introduction 3
title “Cultura bizantină în România” [Byzantine Culture in Romania], and

together with it a number of papers presented at the congress, which were
dedicated to the history of the Byzantine military organization of the Dan-
ube region.20 Petre Diaconu’s monograph on the Pechenegs, which had
been published the previous year, also included some discussion of the
military organization of Paradunavon.21 He continued his studies in the
same vein with his book on the Cumans.22
A new generation of historians approached the topic in the 1970s,
using either literary (Ivan Božilov,23 Jonathan Shepard,24 Nicolae Şerban
Tanaşoca,25 Vasilka Tăpkova-Zaimova),26 or archaeological sources, par-
ticularly the ever-growing data resulting from excavations in Păcuiul lui
Soare, Capidava, Garvăn, Isaccea, as well as from newly opened sites
in Nufăru, Turcoaia, Tulcea, Krivina, Odărci, Vetren, Braničevo (Stefka
Angelova, Silvia Baraschi, Victor H. Baumann, Ljudmila Dončeva-Petkova,
Gheorghe Mănucu-Adameşteanu, Ernest Oberländer-Târnoveanu, Marko
Popović, Ioan Vasiliu). In Serbia, the building of the power plants in the
Iron Gates region offered the opportunity for many rescue excavations,
which brought out some results relevant to the history of the 10th–12th
centuries. A small sector of the Danube frontier was thus studied in an
almost exhaustive manner, although only on the southern bank of the
river (no similar research was carried out on the northern, Romanian
bank). Finally, excavations in Sirmium (Sremska Mitrovica), Belgrade and
Braničevo concerned also the Byzantine fortifications in the area.27
A significant progress came with the discovery of many lead seals con-
cerning the military organization, which were later published by Georgi
Atanasov, Ion Barnea, Ivan Jordanov, Ljubomir Maksimović, Gheorghe
Mănucu-Adameşteanu, Ernest Oberländer-Târnoveanu, and Werner
Seibt. However, by far the most important discovery was that of an entire
archive of hundreds of seals in Preslav, the result of Ivan Jordanov’s inten-
sive work on that site. This was a material of unique value not just for the
region, but for the Byzantine Empire as a whole, since nowhere else had
20 Stănescu 1974; Lewis 1975; Wasilewski 1975; Barnea 1975 b, 503–508; Tăpkova-Zaimova
1975 b, 615–619.
21 Diaconu 1970.
22 Diaconu 1978.
23 Božilov 1973 b, 111–122; Božilov 1976.
24 Shepard 1975; Shepard 1979.
25 Tanaşoca 1973.
26 Tăpkova-Zaimova 1973.
27 For all these sites, see the works quoted in chapter III.
4 introduction
so many seals been found in a precisely dated context. The numismatic

researches of Gabriel Custurea, I. Jordanov, Gh. Mănucu-Adameşteanu
and E. Oberländer-Târnoveanu were also important contribution to the
refinement of the chronology of the Danube frontier.
After 1990 new archaeological excavations were carried out in Piatra
Frecăţei, Hârşova, Preslav, Pliska, Skala, Rujno, and field work continued
at Păcuiul lui Soare, Capidava, Garvăn, Nufăru, and Isaccea. Those excava-
tions were conducted by G. Atanasov, Alexandru Barnea, V. H. Baumann,
Costel Chiriac, Oana Damian, I. Jordanov, Gh. Mănucu-Adameşteanu,
Ioan C. Opriş, Dorel Paraschiv, and Valeri Yotov.
The rapidly growing amount of information and the general development
of studies on Byzantine military history made possible a number of syn-
thetic studies, which clarified various aspects of the military organization
of the Danube provinces, or of the frontier history. The studies published
by I. Barnea,28 P. Diaconu,29 I. Jordanov,30 Gh. Mănucu-Adameşteanu,31
N. Oikonomides32 and T. Wasilewski,33 brought important contributions
to the chronology of the Byzantine military and administrative organi-
zation, as well as of the Pecheneg and Cuman invasions. Moreover, the
studies of the Hungarian medievalist Ferenc Makk concerning the Byzan-
tine-Hungarian relations during the 11th and 12th centuries dealt also with
the Danube frontier, as the military confrontations between Byzantium
and Hungary involved the middle course of the river.34 Other significant
contributions to the study of the Danube region may be found in stud-
ies of the general military history of the Byzantine Empire published by
John Haldon,35 Hans-Joachim Kühn,36 Jonathan Shepard37 and Warren
Treadgold.38 The British Byzantinist Paul Stephenson wrote several stud-
ies, later incorporated into a monograph about the Balkan provinces of
the Byzantine Empire in the 10th–12th centuries, with many innovative
ideas on the evolution of the Danube frontier.39 Ion Bica’s book on the
28 Barnea 1993 b; Barnea 1997; Barnea 2001.

29 Diaconu 1986; Diaconu 1987 b; Diaconu 1988 b.
30 Jordanov 1987 a; Jordanov 1993; Jordanov 1995; Jordanov 2003 b; Jordanov 2011 a.
31 Mănucu-Adameşteanu 1998; Mănucu-Adameşteanu 2001 a.
32 Oikonomides 1998.
33 Wasilewski 1995.
34 Makk 1989; Makk 1990; Makk 1994.
35 Haldon 1999.
36 Kühn 1991.
37 Shepard 1999.
38 Treadgold 1995.
39 Stephenson 2000; Stephenson 2003.
introduction 5
province Paradunavon is hardly accessible, as it was published only in a

small number of copies.40 Florin Curta’s chapters on the Byzantine Dan-
ube provinces and the nomad warriors in the history of early medieval
South-Eastern Europe should also be mentioned for giving more attention
to the archaeological and numismatic data than Stephenson’s book.41
The recent acceleration over the last few decades of the development of
Byzantine Studies, especially in the fields of archaeology and sigillography
had a major contribution to a better understanding of the military orga-
nization in the Danube region in the 10th–12th centuries. A new synthesis
is now possible and necessary.42 The present book, an updated version
of the original manuscript published in Romanian in 2007 by “Cetatea de
Scaun” publishing house, is the final result of several of studies dedicated
to the beginnings of the Danube provinces, the evolution of the military
organization, and different chronological aspects concerning this topic.43
Some of my initial opinions have changed after learning about the discov-
ery of new seals or the refinement of the analysis of the existing data. The
research for this book owes much to the Fulbright post-doctoral grant I
received from Ohio State University (October 2002–March 2003), and to
further studies in the libraries of Rome and the Dumbarton Oaks Center of
Byzantine Studies in Washington DC, as well as in Bulgaria. Many thanks
are due to those who have provided me with books and studies which
were instrumental for the writing of this work: Viorel Achim, Georgi Ata-
nasov, Jean-Claude Cheynet, Florin Curta, Gabriel Custurea, Anton Cuşa,
the late Petre Diaconu, Stela Doncheva, Sergiu Iosipescu, Ivan Jordanov,
Gheorghe Mănucu-Adameşteanu, Ernest Oberländer-Târnoveanu, Paul
Stephenson, Victor Spinei, and Valery Yotov.
40 Bica 2003.
41 Curta 2006, 180–365.
42 An overview of the recent research in Tăpkova-Zaimova, Stoimenov 2004, 341–348.
43 Madgearu 1998; Madgearu 1999 a; Madgearu 1999 b; Madgearu 1999–2000; Madgearu
2003 a; Madgearu 2003 b; Madgearu 2005.
Chapter One
The Recovery of the Danubian Frontier
After the Avar and Slavic attacks1 that caused the abandonment of the
Danubian frontier in the first decades of the 7th century, and after the
establishment of Bulgaria in 680, the Byzantine navy continued to exert
in the intervening period some degree of control over the river, engaging
in operations against Bulgaria. On the Danubian sector close to the sea, it
is possible that Noviodunum (Isaccea) possibly remained a harbor for the
Byzantine ships throughout the entire 7th century,2 and a stopping point
on the way to Durostorum (today Silistra), as did Carsium (Hârşova) for
a shorter period.3 The city of Durostorum was involved in the war of Sep-
tember 680 against the Bulgar ruler Asparukh who occupied the so-called
Onglos, as a lead seal issued by Constantine IV between 679 and 680 is
showing (the seal belonged to a message sent to a high Byzantine digni-
tary, most probably military, from Durostorum).4 Closer to the mouths of
the Danube, at Nufăru, the lead seal of a dignitary named Kyriakos (dated
to 696–697) is a valuable proof that the small fortification that existed
there in the 4th–7th centuries was still under Byzantine domination some
years after the arrival of the Bulgars.5 Unfortunately, no extensive excava-
tions were possible within the modern village, although the site appears
to have grown in importance after the 10th century.
Not too far from Nufăru and Isaccea, the earthworks enclosing an area
near the village of Niculiţel are still a conundrum. Their identification
with Onglos is not suitable, because it is quite clear from the accounts
of both Theophanes Confessor and Nikephoros that that well defended
place, surrounded by rivers and marshes, was located north, not south
1 Dimitrov 1997, 26–34; Madgearu 1997, 315–324; Curta 2006, 66–69; Madgearu 2007,
265–266.
2 There are many 7th century seals (Barnea 1997, 354), but very few coins issued after
Phokas (Oberländer-Târnoveanu 1996, 104; Iacob 2000, 493).
3 For coin circulation at Durostorum: Oberländer-Târnoveanu 1996, 103–106. For Carsium,
a coin from 629–630 attests the preservation of the contacts with the empire (Custurea 1986,
277, nr. 6). Haldon 1999, 74 admitted that some Danubian fortresses survived in the 7th cen-
tury because they were supplied by the fleet.
4 Barnea 1981, 625–628; Jordanov 2009, 82–83, nr 51.
5 Barnea 2001, 107–108, nr. 5.
8 chapter one
of the Danube. Moreover, both authors write of Onglos as a place, not a

region (Theophanes even said that the Onglos was appropriate to accom-
modate the small number of people that remained after the wars).6 The
“steep rocks” mentioned by Nikephoros seem to be his own invention,
because Theophanes, who was better informed, did not record them.7
Nikephoros may have had in mind the legendary Riphaei Mountains.
Some unidentified “fortifications” (οχύρομα) were included in the Onglos.
Between Onglos and the Danube was a small marshy zone. The rivers that
bordered the Onglos are not named, but they cannot be either the Dnieper
or the Dniester, because the Bulgars are said to have crossed them before
settling in Onglos. The name Onglos comes from the Türkic word agul /
aul, which means “court”, “enclosure”.8 Many historians identified the
Onglos with the entire southern part of Moldavia, bordered by the rivers
Siret or Prut, and the Dniester, as well as by the Vadul lui Isac-Tatarbunar
earthen dike,9 but the best identification points to a small area between
the mouths of the rivers Siret and Prut, confined by the earthen dike built
there in the early 2nd century between modern villages of Şerbeşti and
Tuluceşti, whose purpose was to provide an extra-defense for the Roman
camp at Barboşi.10
The theory presented above was rejected by Rašo Rašev, who believed
that before 680 Asparukh had conquered a large region bordered by the
dike in southern Bessarabia, the Şerbeşti-Tuluceşti dike, and the so-called
Small Earthen Dike in Dobrudja. The center of this region, the first Bulgar
state, was the ringwork at Niculiţel—the actual Onglos.11 It is true that,
according to the Bulgarian Vision of Prophet Isaiah written in the second
half of the 12th century, Ispor (Asparukh) had built a great fortification
or dike from the Danube to the sea,12 but this particular text cannot be
6 Theophanes, AM 6171 (transl. Mango, Scott, 498–499); Nikephoros, 35, 36 (ed. Mango,
88–91). The story is repeated by Zonaras, XIV, 21.10–20 (ed. Büttner-Wobst, 227).
7 Hălcescu 1989, 341.
8 Fehér 1931, 24; Božilov 1975, 31. The etymology from the Latin angulus, supported by
some historians, is wrong, because Nikephoros specifically mentions that Onglos was called
so in the language of the Bulgars.
9 Among them: Fehér 1931, 9–12; Bănescu 1948, 6; Gjuzelev 1984, 35–36; Hălcescu 1989,
339–351; Curta 2006, 79; Sophoulis 2011, 109.
10 Božilov 1975, 33–36; Toynbee 1973, 440, 452; Sophoulis 2011, 109. For the Roman origin
of these walls see Napoli 1997, 359–361.
11 Rašev 1982, 76–79; Rašev 1987, 49–51; Rašev 1997, 49–54; Rašev 2004, 277–282. This is
a developed form of the theory first advanced by Škorpil 1918, 145–152, according to whom
Onglos referred to the earthworks as Niculițel.
12 Fehér 1931, 16; Gjuzelev 1984, 33–34; Squatriti 2005, 59–60, 63, 70; Petkov 2008, 195. For
the value of this text, see Dimitrov 1993, 97–109 (only the data about the 11th century are
somehow reliable, see Stanev 2012, 21–25).
the recovery of the danubian frontier 9
trusted at all in matters pertaining to the early history of Bulgaria, except

in what concerns the memory of its half-legendary characters preserved
at the beginning of the second millennium. Regarding Asparukh as their
founding hero, medieval Bulgarians attributed to him such impressive
earthworks, much like Emperor Trajan was credited with the building of
other earthen dike, few of which have actually been built under his reign.
It is therefore a gross mistake to take this text at face value in support of
archaeological interpretations concerning the 7th century Bulgars.
The Small Earthen Dike between Cochirleni and Constanţa is 61 km
long, with a ditch to the south. There are no associated forts. Many hypoth-
eses have been advanced concerning the date of that rampart, which is
the oldest of all three dikes running across the central part of Dobrudja
(the Small Dike is overlapped by the other two—the Great and the Stone
Dike—near Constanţa). Because the Great Earthen Dike seems to be a
Late Roman construction (see below), the Small Dike must therefore be
of pre-Roman origin. Its dating to the 7th century has been advanced by
those who believed the Great Earthen Dike to be a 9th-century building.13
There is in fact no clear explanation for when and why the Small Dike was
built in the first place. Most recently, Pavel Georgiev advanced the idea
that the Small and the Great dikes were built at the same time, with the
Small one operating as a subsidiary line of defense behind the Great Dike.14
If so, it remains why there are no forts associated with it. As for dike in
southern Bessarabia, between the village of Vadul lui Isac and Tatarbunar,
which has its ditch to the north, its dating to the 7th century is based on
a number of refuse pits and huts associated with 6th–7th century hand-
made pottery, which were found on the southern side of the rampart at
Kubei (Bolgrad district).15 Nonetheless, it’s the Roman origin of the dike
is clear, since 4th century coins were found in the filling of the ditch.16
Therefore, both earthen dikes believed by Rašev to have been the north-
ern limit of Onglos are of Roman origin. Rašev’s ideas have been accepted
and reproduced by D. I. Dimitrov and Uwe Fiedler.17
The earthworks in Niculițel consist of a circular rampart, 27 km long,
which encloses an area of 48.3 km2; four earthen forts; and two median
13 Barnea, Ştefănescu 1971, 97–98, 110, 116–117; Diaconu 1972 a, 377–378; Papuc 1992, 327;
Mănucu-Adameşteanu 2001 a, 77; Bogdan-Cătăniciu 2006, 411 (for the pre-Roman origin);
Sophoulis 2011, 101.
14 Georgiev 2010, 413–422.
15 Curta 2006, 80; Fiedler 2008, 164; Sophoulis 2011, 102.
16 Napoli 1997, 60, 104, 373–378.
17 Dimitrov 1985, 119; Fiedler 1986, 461; Fiedler 2008, 152–153.
10 chapter one
ramparts linking the outer precinct to the forts in the interior.18 The
shape of this fortification system strongly suggests its Türkic origin, and
the very scarce archaeological evidence shows that the ramparts could
be dated between the 5th and the 7th centuries.19 However, there is no
way to credit Asparukh’s Bulgars with that the building of this system of
circular fortifications, as other nomadic population, such as the Avars,
may also come into question. After all, an updated version of the Geogra-
phy of Pappus of Alexandria, in Armenian translation claims that “there
are two mountains [in Thrace] and one river, the Danube, which has six
tributaries and which forms a lake, and an island called Peuce. On this
island lives Asparukh, son of Kubrat, a fugitive from the Khazars from
the mountains of the Bulgars, who expelled the Avar nation and settled
there”.20 The island of Peuce was not a real island, but the northern part of
Dobrudja, near the Danube Delta (called Peuce in Antiquity). The alliance
of this Avar group with the empire is documented by finds of gold and
silver coins received as subsidies during the period between Heraclius and
Constantine IV, when the Byzantine navy continued to exert its control
along the Danube up to Durostorum. The gold and silver coins found in
the northern Dobrudja were payments for an ally, which could well be the
Avar group, split from the central power after 626, when the failed siege
of Constantinople caused centrifugal movements of the subjected popula-
tions. Such coins have been found at:
Agighiol—1 silver 668–669;

Galaţi (hoard)—3 silver of 615–638, 4 silver of 647–659, 5 silver of 668–681;
Isaccea—1 gold of 659–664;
Istria—1 gold of 674–681;
Lunca (Ceamurlia commune)—1 gold of 674–680;
Niculiţel—1 silver of 674–681;
Piua Petrii (hoard)—1 silver of 641–668, 2 silver of 668–685;
18 See the descriptions in Škorpil 1918, 119–134 and Barnea, Ștefănescu 1971, 117–118.
19 Diaconu 1972 b, 316–318. After the field surveys during the Bulgarian occupation of
Dobrudja in 1917, none of which produced any results (Škorpil 1918, 113–141), the excavations
carried out 1953–1954 produced 6th- to 7th-century handmade pottery from the filling of
the rampart.
20 Ananias, 48. The author of this geographical treatise was long believed to have been
Moses Chorenatzi (Beševliev 1981, 173–174), but the latter lived in the 5th century. The
real author was Ananias of Širak (590–670). The interpolation about Asparuch belongs
to the abridged version, written in the 9th century. See Spinei 2009, 80–81 for the late
interpolations.
Tulcea—1 gold of 641–646;

Valea Teilor—2 silver of 641–646 and 668–673 (perhaps from a hoard).21
The hoard from Galaţi (which belongs to the same group) shows a long
accumulation from Heraclius to Constantine IV. The stipends were most
likely paid over a long period of time between Heraclius (most probably
after 626) and Constantine IV. Heraclius is in fact known for having estab-
lished an alliance with Kubrat at the time of the latter’s conflict with the
Avars.22 The fortifications of Niculiţel could have well been built by this
group, for the Byzantines, during a longer period than the short presence
of Asparukh near the Danube Delta.
The Bulgar state established in 680 in the former Roman province
Moesia challenged the Byzantine domination over the western coast of
the Black Sea, dangerously close to the imperial capital. The security of the
Black Sea region required a policy of containment of Bulgaria or even the
destruction of the Bulgarian state which was growing fast. Upon his return
from the exile in Crimea, and with military assistance from of the Bulgar
khan Tervel (701–718), Emperor Justinian II gave the Bulgars the region of
Zagora between Stara Planina and Burgas in 705. The treaty of 717 estab-
lished the new frontier from Mileone (Jabalkovo, Haskovo region) to Cape
Emona; Anchialos (Pomorie), Mesembria (Nesebăr) and Develtos (Burgas)
remained in the Byzantine Empire. Zagora was recovered after the victory
at Markellai (Krumovo, 7 km west of Karnobat) obtained by the Byzantine
army in 756, or according to other opinions, in 760 or 761.23
In the course of the following Byzantine-Bulgar conflicts, naval opera-
tions in the Black Sea and on the Danube were a key component of the
Byzantine strategy, which was based on attacking the enemy on two fronts.
The fact that Bulgaria did not become a naval power (although some
boats are said to have gone up the Drava River against the Franks in 827)24
was a great advantage for the Byzantine offensive policy. This strategy of
encirclement was applied by Constantine V, during the wars with Bulgaria
21 Oberländer-Târnoveanu 1996, 104–105; Iacob 2000, 485–498; Madgearu 2007, 270–271;
Custurea, Talmaţchi 2011, 314.
22 Beševliev 1981, 511–516.
23 Theophanes, AM 6251 (transl. Mango, Scott, 596); Nikephoros, 73 (ed. Mango, 145);
Beševliev 1971, 6–9; Browning 1975, 48; Beševliev 1981, 209–210; Gagova 1986, 67–69; Dim-
itrov 1992 a, 36–37; Sophoulis 2011, 90–93.
24 Annales Regni Francorum, a. 827 (ed. Kurze, 25–26); Beševliev 1981, 285, 470–471; Egg-
ers 1995, 62; Curta 2006, 158–159; Sophoulis 2011, 123, 297.
12 chapter one
which started after the decisive victories against the Arabs (as a rule, the
military effort was directed at Bulgaria when the other enemy, the Caliph-
ate, was in crisis or defeated).25 The main goal of Constantine V’s cam-
paigns was to stop the Bulgar expansion into Thrace, the most important
Byzantine agricultural region of that period.26 Durostorum was reached
again by the Byzantine navy during these wars fought by Constantine V
against Bulgaria. In 756, a fleet of 500 boats moved up the Danube, while
the land army commanded by the emperor marched toward the Stara
Planina range, obtaining the victory at Markellai. In 763, another fleet of
800 chelandia with 9,600 horsemen on board entered the Danube. Mean-
while, the land army advanced to Anchialos, where it won another vic-
tory on June 30, 763. In 767, the defeated Bulgars are said to have taken
refuge in the swamps downstream from Durostorum. The large number
of boats involved in those operations indicates the existence of some sup-
ply points along the river. However, all those attempts at restoring the
Byzantine control over the Danube ended with the Bulgar victories of 792
and 796.27
The operations along the Danube are also documented by several lead
seals dated to the 8th century. One of them belongs to a strategos called
Phokas, which suggests the existence in Durostorum of an official receiv-
ing messages from a theme commander. This man was perhaps the chief
of a turma (one of three subdivisions of a theme).28 A pier found at Duro-
storum, radiocarbon-dated to 780 AD ± 200 may have been built during
those naval campaigns that reached that city, even if some Bulgarian
archaeologists suppose that the pier was used for trade only.29 The large
margin of error involved in the radiocarbon dating makes it impossible to
decide on the chronology and historical significance of the pier.
Even if scattered, those bits of informaton suggest a continuous Byzantine
concern with recuperating the positions lost along the Danube, including
Durostorum, within the context of the 8th-century conflicts with Bulgaria.
The presence of the imperial navy on the river represented a real threat
25 Howard-Johnston 2006, 344.

26 Shepard 1995, 232–233.
27 Theophanes, AM 6254, 6265 (transl. Mango, Scott, 599, 617); Nikephoros, 73, 76, 79
(ed. Mango, 144/145, 148/149, 152/153); Zonaras, XV, 8.4–19 (ed. Büttner-Wobst, 279–281);
Beševliev 1971, 5–17; Beševliev 1981, 209, 213–234; Gagova 1986, 67–69; Gjuzelev 1992, 24–25,
30; Busetto 1992, 323–330; Dimitrov 1992 a, 36–37; Curta 2006, 85–88; Sophoulis 2011, 93–95,
166–172.
28 Barnea, Ştefănescu 1971, 15; Damian 2004, 290–291.
29 Angelova, Koleva 2004, 22.
to Bulgaria. During the subsequent centuries, the navy appears to have

cooperated with allies the Byzantine emperors summoned from the
steppe lands to the north of the river. However, the rise of Bulgaria as a
major power in the region, ca. 800, interrupted these attempts at restor-
ing the Byzantine naval domination on the Danube. The Bulgars secured
Durostorum with a restored precinct and a new gate, and a residence was
built there for khan Omurtag (815–831).30 After the failure of Emperor
Nikephoros I’s 811 campaign against Bulgaria, Krum (803–814) launched
a counteroffensive in 812, which resulted in the conquest of Mesembria,
Anchialos, Develtos, Sozopolis and Agathopolis. The attacks continued in
Thrace in 813–815, until the decisive victory of the new emperor Leo V
(813–820) at Mesembria in 814. The peace treaty of 816 divided the western
coast of the Black Sea between Byzantium and Bulgaria. The Byzantines
recuperated the port cities of Sozopolis and Agathopolis. Develtos appears
to have been shared between the two powers along the river Sredetska.
Mesembria and Anchialos were left to Bulgaria. The new frontier was
established on the line Develtos—Agathonike (Orjakhovo)—Constantia
(Simeonovgrad), up to the Hebros River (Maritsa). The access between
Develtos and Maritsa was blocked by an earthen dike now known as
Erkesija, which is 137 km long.31
To the north, the expansion of Bulgaria resulted in the occupation of
parts of Walachia and Moldavia during the reign of Krum, shortly after the
collapse of the Avar khanate. In 813, thousands of prisoners of war taken
from Adrianople and Macedonia were moved to the so-called “Bulgaria
beyond the Danube.”32 Much has been written on the exact location of this
territory, but the right solution may only be provided by the archaeological
evidence. Ninth-century century artifacts pertaining to an urban culture
of Byzantine origin (clay water-pipes, bricks, and a specific kind of pottery
mostly associated with Byzantine towns) have been found especially to
the west and north-west from the city of Olteniţa (Chirnogi, Căscioarele,
30 Angelova, Koleva 2004, 21–22; Atanasov 2012.

31 Theophanes, AM 6303–6305 (transl. Mango, Scott, 672–675, 679, 682–686); Skylitzes,
Leo V the Armenian, 1 (ed. Thurn, 13; trad. Flusin, 15; transl. Wortley, 5); Zonaras, XV, 17.15–19;
19.13–19 (ed. Büttner-Wobst, 315–319, 321–322); Browning 1975, 49–50; Gagova 1986, 70–72;
Browning 1986, 29; Dimitrov 1992 b, 43–45; Sophoulis 2011, 184–286. For the Erkesija wall, see
Beševliev 1981, 476–477; Rašev 2005, 52; Curta 2006, 149–156; Fiedler 2008, 167.
32 Leo Grammaticus, 231–232; Scriptor Incertus, 345–346; George Monachos, 817–818;
Symeon, Chronikon, 131.9–10 (ed. Walgren, 235–236); Bănescu 1927, 14–15; Bănescu 1948, 6–7;
Brezeanu 1984, 121–122; Teodor 1987, 2–3; Damian 2003, 483–484; Spinei 2009, 56–57; Teodor
2011, 95–96; Sophoulis 2011, 102, 256–257.
14 chapter one
Greaca, Radovanu, Curcani, Mironeşti), as well in other places in Wala-

chia. They must be associated with the Byzantines forcefully moved north
of the Danube, on the road to the salt mines of the present-day counties
of Prahova and Buzău. Similar artifacts have been found farther to the
northeast, near the mouth of the river Siret (Dragosloveni, Budeşti, Câmp-
ineanca, Gugeşti, Şendreni), the terminal of another salt road. The brick
fortress at Slon (Prahova County) was also built for the Bulgars by the
same Byzantines resettled beyond the Danube. This fortress was located
in the Prahova salt area and its purpose was to defend the route to Tran-
sylvania across the Carpathian Mountains, through the Tabla Buţii pass.33
The sudden appearance of 9th-century settlements south of the Roman
dike between Vadul lui Isac and Tatarbunar, in southern Bessarabia, has
also been linked to the same Byzantine prisoners.34
The territories north of the Danube in present-day Walachia and
southern Moldavia appear to have occupied primarily because of the salt
resources (vital to any medieval society), in addition to their strategic
position as a buffer against the Khazars, then against the Magyars, and
the Byzantine outpost at the mouths of the Danube (after 863). Byzantine
sources show the lands north of the Danube to have been under their own
commanders, and not under a direct or even very strict Bulgar control.
Cordyles, the chief of the Byzantine prisoners, escaped in 836 and asked
for assistance from emperor Theophilos (829–842). A fleet was sent in 837
or 838 to the Danube to rescue the former prisoners. The Bulgars sum-
moned the Magyar warriors from Levedia (the region between the Don
and the Bug to which they had moved after 830) to push back the Byzan-
tine expedition.35 The Magyars were at that time the new nomads in the
area, who were usually hostile to Bulgaria. A campaign of the Bulgar army,
which had reached the Dnieper River a few years before those events,
may have well been directed against them.36 To stop Magyar attacks, the
Bulgars used as a defense line the Great Earthen Dike across Dobrudja,
between Cochirleni on the Danube and Constanţa. This 54 km-long dike
has ditches on both sides, in addition to 35 large and 28 small forts along
33 Teodor 1987, 9–12; Comşa 1982–1984, 39–44; Damian 2003, 485–491; Ciupercă 2010,
279–289; Teodor 2011, 96–98.
34 Postică, Hâncu, Tentiuc 1999, 288.
35 Leo Grammaticus, 232; George Monachos, 818; Symeon, Chronikon, 131.11–12 (ed. Wal-
gren, 236–237); Beševliev 1981, 354; Brezeanu 1984, 121–129; Božilov, Dimitrov 1985, 54–56;
Treadgold 1988, 291; Kristó 1996, 15, 107–111; Zuckerman 1997 a, 55–59; Mladjov 1998, 87–90;
Spinei 2006, 58, 64; Spinei 2009, 58.
36 Beševliev 1981, 283; Spinei 2006, 66; Sophoulis 2011, 292–293.
the inner defence line. The Great Dike was most likely built in the 4th–
5th centuries by the Late Roman army, but its repair in the 9th century is
proved by finds of pottery from an excavation in Palas, near Constanţa.37
The increasing turmoil in the region between the Don and the Volga
after 830, caused by the migrations of the Pechenegs and the Magyars,
re-activated the Byzantine-Khazarian alliance which made possible the
reestablishment of a true imperial administration in Crimea, in the form
of the new theme of Cherson, created in 841 by Emperor Theophilos. Its
2,000 soldiers were commanded by the same Petronas, who had been sent
to Sarkel in 839 to build a stronghold for the Khazars. The initial plan
was to occupy a larger area, called the theme of Klimata. The name of
the province was changed after some decades, becoming the theme of
Cherson.38
In 860, Constantinople was attacked by a fleet of 200 ships carrying
8,000 Rus’ warriors, who ravaged the suburbs and carried off a large
amount of booty.39 This event demonstrated that the Khazar allies of the
Byzantine Empire could not be trusted as a real shield against invasions
from the north. Defence against future attacks required the strengthen-
ing of the Byzantine positions in the north and the establishment of new
and effective alliances, such as that made in the next year with the Mag-
yars, by the future Apostles of the Slavs. Saints Constantine (Cyril) and
Methodius, who were sent to Khazaria by Emperor Michael III (842–867),
but encountered a party of Magyars on their way. That the Magyars were
located between the Don and the Bug made them excellent allies against
Rus’ attacks. They were also in a good position to attack Bulgaria and the
Frankish Empire, if needed. For more than three decades, the Magyars
37 The description of the Great Earthen Wall at Barnea, Ştefănescu 1971, 99–100. For the
excavation at Palas: Panaitescu 1978 a, 241–245. For the Late Roman origin: Comşa 1951, 233–
235; Barnea, Ştefănescu 1971, 115; Petre 1973, 31; Bogdan-Cătăniciu 1996, 202–207; Napoli 1997,
102–104, 341–347, 354–355; Georgiev 2005, 23–40; Bogdan-Cătăniciu 2006, 412–418. For the
supposed construction in the 9th century: Diaconu 1973–1975, 199–209; Rašev 1987, 52–53;
Mănucu-Adameşteanu 2001 a, 78–79; Rašev 2005, 53; Fiedler 2008, 168.
38 Skylitzes, Theophilus, 22 (ed. Thurn, 73; transl. Flusin, 66–67; transl. Wortley, 75);
Wozniak 1975, 56–62; Noonan 1992, 114; Whittow 1996, 233–235; Zuckerman 1997, 51–58
(who has demonstrated that the threat was represented by the Magyars, and that the right
date for Sarkel building is 839, not 833). The name Klimata concerns the small zones around
the fortresses from the mountain region of the peninsula (Zuckerman 1997 a, 67). The name
change reflected the abandonment of the region outside the town (Zuckerman 1997 a,
67–73).
39 Skylitzes, Michael III and Theodora, 18 (ed. Thurn, 107; transl. Flusin, 94; transl. Wort-
ley, 107); Russian Primary Chronicle, 60; Franklin, Shepard 1996, 50–55; Whittow 1996, 239.
16 chapter one
therefore replaced the Khazars as the guarantors of the Byzantine security

in the lands north of the Black Sea. Unlike the Khazars, the Magyars had
no interest in Crimea, a good portion of which was thus turned into the
theme of Cherson established in 841.40
A good defense against the maritime attacks of the Rus’ also required
the control of certain strategic points on the western coast of the Black
Sea, but this region was partially lost after the treaty of 816, which gave
to Bulgaria the ports of Anchialos and Mesembria, as well as a part of
Develtos.41 Furthermore, Bulgaria received another, 25-mile wide strip of
land at the beginning of the reign of Boris (852–889), as Empress Theodora
(842–856) wanted to establish good relations with the Bulgar ruler during
the concomitant Byzantine campaigns in the East. After 860, therefore
it became imperative for the Byzantines to recover as much as possible
of the Black Sea coast, which would otherwise be available to northern
invaders if Bulgaria turned hostile. The Rus’ danger was thus a strong
reason for an offensive Byzantine policy against Bulgaria. Another reason,
no less important, was the alliance forged by the Bulgar khan Boris with
the Frankish Empire in 862. This alliance changed the balance of power in
South-Eastern Europe. The Byzantine emperor Michael III reacted imme-
diately by attacking Bulgaria in 863. Boris was defeated and converted
to Christianity under the name of Michael. Following the war, Emperor
Michael III recuperated Develtos, Anchialos, and Mesembria.42 Their
recovery made possible the building of a strong defensive system directed
both against Bulgaria and the Rus’, a system that represented a first attempt
at the rebuilding of Byzantine hegemony on the western Black Sea coast,
one century before the reconquest of the Danube region.
If the northern element of the defensive system directed against the
Rus’ was the Cherson theme, its southernmost element was Mesembria,
a port now turned into the headquarters of a kleisura. The kleisurai were
initially military structures located on the frontier, especially around
mountain passes or other points of access to and from the empire. The
first kleisurai were established in the East in the 7th–8th centuries, and
40 For the alliance with the Magyars, see Nikolov 1997, 79–92; Zuckerman 1997 a, 51–74.
For the theme see Zuckerman 1997 b, 210–222.
41 Browning 1975, 50–51; Beševliev 1981, 190–206; Gjuzelev 1992, 25; Dimitrov 1992 b,
43–45.
42 Skylitzes, Michael III and Theodora, 7 (ed. Thurn, 90–91; transl. Flusin, 80–81; transl.
Wortley, 90–92); Zonaras, XVI, 2.1–15 (ed. Büttner-Wobst, 387–389); Browning 1975, 54–55;
Gjuzelev 1978, 52; Browning 1986, 30; Gjuzelev 1992, 25; Gagova 1986, 73; Fine 1991, 113–130;
Dimitrov 1992 a, 38–40; Dimitrov 1992 b, 47; Curta 2006, 166–168.
they were later transformed into small themes in the 9th–10th centuries.
The commanders of kleisurai (kleisurarches) were under the direct com-
mand of the emperor, but had a rank lower than that of the strategoi of
themes.43 In the Balkans, the first kleisura was organized on the Strymon
valley in Macedonia, in 688.44 In Mesembria, the lead seals of several klei-
surarches are now dated with certainty by their archaeological contexts
between 864 and 917. Another kleisura was created in Sozopolis around
840.45 Ivan Jordanov first believed that those lead seals should be dated to
the third quarter of the 11th century, for he assumed that the kleisura func-
tioned in parallel to the katepanate of Mesembria.46 We shall see that that
katepanate was indeed organized during that period as a shield against
the rebels who had taken power in Paradunavon in 1073. Two katepanoi
of Mesembria are known, Simeon and Valatzertes (see chapter II.1). No
mountain pass existed at Mesembria, but the word kleisura received the
meaning of any small frontier area organized for defense. Its mission was
to protect the naval base at Mesembria, which was a staging post on the
route to the Danube and the Crimea.
Between the Crimea and Mesembria, the Danube Delta was another
region of strategic importance, because it was on the route taken by Rus’
boats, as indicated in the De Administrando Imperio.47 It is thus not sur-
prising that there is some evidence for a form of Byzantine military orga-
nization in this area in the 9th century. The existence of a Danube Delta
theme called Lykostomion was inferred from the dedication of Photios’
Lexicon to a certain Thomas, protospatharios and archon of Lykostomion.
The Lexicon was written around 880–890. Hélène Ahrweiler suggested that
the 9th-century Lykostomion was the same as that recorded in the Danube
Delta region in the 14th century.48 This identification was then endorsed by
many other historians, especially from Romania,49 but rejected by Vasilka
Tăpkova-Zaimova, who maintained that the 9th-century Lykostomion
43 Ahrweiler 1960, 81–82; Oikonomides 1972, 342; Ahrweiler 1974 a, 216–218; Ferluga 1976,
71–85; Haldon 1999, 79, 114.
44 Stavridou-Zafraka 2000, 128–129; Krsmanović 2008, 129.
45 Jordanov 2003 a, 119–120, nr. 47.2, 159–160, nr. 74.1; Jordanov, Žekova 2007, 118, nr. 308;
Jordanov 2009, 454–455, nr. 1340–1345.
46 Jordanov 1993, 140, nr. 280, 281.
47 Constantine Porphyrogenitus, IX, 91–102 (ed. Moravcsik, 62/63).
48 Ahrweiler 1966, 89–90.
49 For instance: P. Ş. Năsturel, book review of Ahrweiler 1966, RESEE, 4, 1966, 3–4, 649–651;
Barnea, Ştefănescu 1971, 12; Florescu, Ciobanu 1972, 381–382; Iliescu 1978, 234; Diaconu 1981
a, 218; Diaconu 1994 a, 368; Barnea 1997, 355; Barnea 2000, 296–297; Mănucu-Adameşteanu
1998–2003, 163; Damian 2004, 286.
18 chapter one
was probably a place in Thessaly, Epirus, or Acarnania. Tăpkova-Zaimova

noted that no other sources mention the Lykostomion-on-the-Danube at
such an early date.50 Likewise, Ahrweiler’s identification of the so-called
archon Boulgarias mentioned on some seals with a Byzantine commander
of a Bulgarian maritime theme, although accepted by Ion Barnea,51 Petre
Diaconu52 and Robert Browning,53 has eventually proved to be wrong.
It is now clear that the title in question applied to the Bulgarian ruler
Boris-Michael.54
Under those circumstances, the location of the Lykostomion theme
somewhere in the Danube region (at Periprava, as Octavian Iliescu thought,
or in some other place) is in need of supplementary proof. No archaeolog-
ical remains of a 9th-century fortress have yet been found in the Vâlcov-
Periprava area, despite several field surveys, although potsherds from
Letea, near Periprava, attest to a 9th- to 11th-century settlement.55 Yet, the
possibility of a Lykostomion theme in the Lower Danube area results also
from the inscription on a lead seal found at Isaccea in 1993, which belongs
to the stratilates John Chaldos, the strategos of Chaldia, a theme located
on the southern shore of the Black Sea. That commander took part in a
plot against Michael III in 867.56 The seal thus shows the existence of
the Byzantine authority on the site of ancient Noviodunum around 867.
This further suggests the existence of a theme in the area, possibly that
of Lykostomion. In addition, a solidus struck for Basil I and Constantine
(868–879) was discovered in unknown circumstances near the village of
Suvorovo (Ismail, Odessa region, Ukraine) in 1997.57 The place is signifi-
cant, because it is located near the Vadul lui Isac-Tatarbunar earthen dike
mentioned above, and next to a presumed fortress of that same period.58
Since gold coins from this period are very rare in the lands north of the
river Danube, its presence at a small distance from the supposed Lykos-
tomion theme is really significant.
The contacts between the strategos of Chaldia and a commander at
Noviodunum attested by that lead seal could have concerned security
50 Tăpkova-Zaimova 1970, 82–86.

51 Barnea, Ştefănescu 1971, 12; Barnea 2000, 296.
52 Diaconu 1981 a, 218.
53 Browning 1975, 137.
54 Tăpkova-Zaimova 1970, 83. For these seals, see Jordanov 2001, 31–34.
55 Iliescu 1978, 234; Iliescu 1994, 244–245; Mănucu-Adameşteanu 2001 a, 67–68.
56 Barnea 1993 a, 55–56; Barnea 1997, 355.
57 Russev, Fokeev 2001–2002, 183–185.
58 Postică, Hâncu, Tentiuc 1999, 289.
problems in the Black Sea region, because the mission of the Chaldia
theme was the defense of the southern coast against the attacks of the
Rus’. The provinces of Chaldia and Paphlagonia were created in 820 after a
Rus’ maritime invasion that affected the city of Amastris.59 When the Rus’
launched another, much more destructive attack against Constantinople
in 860, the solution was to recover the ports previously taken by Bulgaria,
but also the mouths of the Danube, in order to revamp the defense system
in the Black Sea region In this way, the advanced position at Cherson was
inserted into a wider security system, completed after 863 by the creation
of the Lykostomion theme, located between Cherson and the western
coast ports. The defensive system created after 863 in the western Black
Sea area ensured the security of Constantinople against the Rus’ and Bul-
garia, while the alliance with the Magyars secured the stability in the north.
Therefore, the establishment of a theme in the area of the mouths of the
Danube was required by the threats to the security in the Black Sea area
after 860. No such military reasons could be invoked for any other place
called Lykostomion in the 9th-century Byzantium. This substantiates the
hypothesis put forward by H. Ahrweiler, as Vasilka Tăpkova-Zaimova has
not explained which strategic task could a theme of Lykostomion serve in
Thessaly, the Tempe valley, or in Epirus. None of the places mentioned in
her study had any military importance in the 9th century.
Ion Barnea claimed that Lykostomion was the headquarters of the
Byzantine navy in the Black Sea,60 but this could hardly have been the
case, since the naval base of Cherson in Crimea was of far greater impor-
tance. The theme of Lykostomion included the maritime sector of the
Danube, but not the area north of the Great Earthen Dike, for that military
structure had been established at the height of Bulgar power. It is more
probable that a buffer zone existed between the Bulgar and the Byzantine
territories in Dobrudja.61 During this period, the Byzantine navy was no
longer in a position to control Durostorum as it had been at some points
in the 7th and 8th centuries; during the ninth century, there is clear evi-
dence that the town was in Bulgar hands. The base at Lykostomion was
then a staging post on the coastal route to Cherson. There is no proof
that this base was used by the Byzantine navy for the repatriation of the
59 Treadgold 1988–1989, 140–143; Whittow 1996, 254–255.

60 Barnea 1993 b, 584–585.
61 Harhoiu 2000, 353.
20 chapter one
prisoners escaping from the lands north of the river Danube.62 It came
into existence only after 863.
On the basis of the alleged existence in Capidava of a ninth-century
Byzantine fortress, Radu Florescu once believed that the Lykostomion
theme came into being at some point during the last third of the 8th cen-
tury, shortly after the successful campaigns led by Emperor Constantine
V against Bulgaria. In his view, the Byzantine domination extended as
far south as the Small Earthen Dike between Cochirleni and Constanţa,
which he believed to have marked the Byzantine-Bulgar frontier.63 If so,
the territorial expansion of Bulgaria during the reign of Krum left no room
for any Byzantine territory extending as far south. While it is theoretically
possible that the theme was established first during the reign of Constan-
tine V,64 its operation must have been interrupted between 812 and 863.
On the other hand, there is no proof that the Small Earthen Dike was
erected in the late eighth century.
The theme of Lykostomion was involved in the Byzantine-Bulgarian
war, which began in 894, when the Bulgarian ruler Symeon (893–927)
retook Mesembria and Anchialos. Emperor Leo VI (886–912) ordered a
counteroffensive in southern Bulgaria combined with a naval operation
on the Danube. He summoned the Magyar allies, for his own forces were
apparently not sufficient (the Byzantines were engaged at that same time
in war with the Arabs in Anatolia). The action followed a defensive plan
based on the theme of Lykostomion and on cooperation with the Magyars,
who were at that moment the masters of the region known as Atelkuz
(between the Dnieper or the Bug and the Danube). The Magyars had been
expelled by the Pechenegs from Levedia, and they had moved to Atelkuz
ca. 889. Now close to the Danube, they were useful allies in a war against
Bulgaria. The Magyars ruled by Arpad and Kusan crossed the Danube on
boats of the Byzantine navy, and then ravaged northern Bulgaria, winning
two or three battles. A war on two fronts was too much for Symeon (who
apparently put up some fierce resistance in Dorostolon).65 Realizing that
he could rely on nomads, much like Leo VI, he appealed for help from the
Pechenegs, who promptly invaded the land of their enemies, the Magyars.
62 As advocated by Diaconu 1981 a, 218 and Custurea 1986, 276.

63 Florescu, Ciobanu 1972, 381–387; Florescu 1986, 171–177.
64 In such a case, the seals of eighth-century military officials found in Durostorum could
be ascribed to officers of that theme.
65 The ancient, Latin name of Durostorum is rendered Dorystolon or Dorostolon in 9th–
10th centuries sources written in Greek.
Arpad was thus forced to leave Bulgaria, and the Magyars soon after that
left for Pannonia. With Pecheneg assistance, Symeon was able to restore
his power on the Danube. The war of 894–896 ended with a major Bulgar-
ian victory at Bulgarophygon (southern Thrace), in the spring of 896. The
theme of Lykostomion must have ceased to exist after that defeat, which
drastically changed the balance of power in the western Black Sea area.66
The sudden increase of the Pecheneg power required a change of the
Byzantine policies in the northern Black Sea region, for the Pechenegs
were a real threat to Cherson. Like the Magyars, they seem to have occu-
pied the interior of Crimea.67 One obvious solution was to turn them
into imperial allies. In 917, the strategos of Cherson, John Bogas, in fact
requested their military assistance in yet another war against Symeon. He
convinced the Pechenegs to march toward the Danube, while the Byzan-
tine fleet was sent to transport them over to the southern bank. The dis-
pute between Bogas and the commander of the imperial navy Romanos
Lekapenos (the future emperor) prevented that action, and the Byzan-
tine army remained without the allied Pecheneg forces. That was a great
advantage for Symeon, who directed his troops to the southern conflict
area, where he won the crushing victory of Acheloos (near Anchialos),
on August 20th, 917 against the army led by general Leo Phokas.68 Some
historians believe that the Bulgarians were allied with the Magyars in 917,69
but the source on which such an interpretation is based (Al-Masudi)
refers most probably to the Magyar and Pecheneg attacks of 934.70 By
the Byzantine-Bulgarian treaty of 927, Anchialos entered thus again under
Bulgarian rule, together with the ports of Mesembria, Develtos, Sozopolis,
66 Leo, Taktika, 18. 40 (ed. Dennis, 452/453); Annales Regni Francorum, a. 896 (ed. Kurze,
129–130); Symeon, Chronikon, 133. 17–21 (ed. Walgren, 276–277); Theophanes Continuatus,
358–359; Constantine Porphyrogenitus, XL, 8–20 (ed. Moravcsik, 176/177); Skylitzes, Leo VI
the Wise, 12, 14 (ed. Thurn, 176–178; transl. Flusin, 148–150; transl. Wortley, 170–172); Zonaras,
XVI, 12.18–29 (ed. Büttner-Wobst, 442–444); Gjuzelev 1978, 52; Božilov 1980, 73–81; Wozniak
1984, 303; Dimitrov 1986, 67–68; Gjuzelev 1996, 228; Kristó 1996, 182–190; Tougher 1997,
176–280; Howard-Johnston 2000, 342–354; Spinei 2006, 73; Spinei 2009, 62; Todorov 2010,
318–319. For the migration of Hungarians to Atelkuz and its location, see Oikonomides 1965,
69–72; Božilov, Dimitrov 1985, 57–68; Malamut 1995, 110; Zuckerman 1997 a, 61–66; Spinei
2006, 62–63; Spinei 2009, 63–64.
67 Obolensky 1979, 129.
68 Nicholas I, Letters, 60–63; Theophanes Continuatus, ed. Bekker, 387–390; Skylitzes,
Constantinos VII (913–959), 8 (ed. Thurn, 203–204; transl. Flusin, 172; transl. Wortley, 198–
199); Zonaras, XVI, 17.1–4 (ed. Büttner-Wobst, 464–465); Wozniak 1984, 304–306; Dimitrov
1986, 75–77; Haldon 2001, 87–88; Todorov 2010, 323–324.
69 Božilov 1973 a, 14; Dimitrov 1986, 75–76; Mladjov 1998, 120; Makk 1999, 12.
70 Diaconu 1970, 18–19; Spinei 2006, 168.
22 chapter one
and Agathopolis. The Byzantine control on the western Black Sea coast
and the security of Constantinople against Rus’ attacks were thus seri-
ously endangered.71
Because the Pechenegs were quickly growing into the most important
power in the steppe lands north of the Black Sea, it was necessary to stop
them by supporting their rivals. This was the second solution found for the
Pecheneg problem. The Rus’ were a suitable partner, because they were
at the same time enemies of the Khazars and of the Pechenegs. During
the first half of the 10th century, the Byzantine Empire broke the alliance
with the Khazars, who were no longer of any use, because their power
had declined. On the other hand, the new missionary policy started after
860 opposed an alliance with a Judaic state. It is true that the same Rus’
were potential enemies of the Byzantine Empire, but they were not able
to launch offensives toward the Black Sea if they were hold back by the
Pechenegs who controlled the Dnieper cataracts. For that reason, it was
imperative to manage the hostility between the Rus’ and the Pechenegs
by means of a balance policy.72 The joint raid of the Magyars and Pech-
enegs in 934 into Bulgaria and the Byzantine Empire73 only showed how
serious the danger was, and that required using the Rus’ to put the breaks
on the Pechenegs, even if the two populations were traditional enemies.
In 940, the Rus’ prince Oleg started a war against the Khazars at the
request of Emperor Roman I Lekapenos (920–944), who wished to pun-
ish the Khazars for their alleged persecution of Christians, but a Khazar
chief persuaded Oleg to turn against the emperor and to attack Constanti-
nople. The invasion took place on June 11th, 941, under the common com-
mand of Oleg and Igor. A fleet composed of 1000 small boats besieged
the capital and destroyed the suburbs much like in 860, but it was finally
defeated by means of the “Greek Fire”.74 The Byzantine diplomacy was
at an impasse, but the idea of using the Rus’ as an imperial instrument
in the northern Black Sea area was not abandoned. The Rus’ launched
71 Diaconu 1970, 14–15; Gjuzelev 1981, 17; Wozniak 1984, 304–306; Fine 1991, 149–150, 161;
Malamut 1995, 107–109.
72 Toynbee 1973, 458–460; Wozniak 1984, 301–315; Huxley 1984, 84–87.
73 Theophanes Continuatus, 422–423; Masoudi, II, 59–64; Skylitzes, Romanos Lekapenos,
29 (ed. Thurn, 228; transl. Flusin, 192; transl. Wortley, 220); Moravcsik 1970, 55–56; Diaconu
1970, 17–19; Oikonomides 1973, 1–3; Makk 1999, 12; Stephenson 2000, 40; Curta 2006, 188;
Spinei 2006, 109, 168; Spinei 2009, 92.
74 Russian Primary Chronicle, 72; Ostrogorsky 1956, 303; Noonan 1992, 115–116; Zucker-
man 1995, 256–257, 264–265; Whittow 1996, 257; Franklin, Shepard 1996, 113–117; Noonan
1998–1999, 210–211.
in 943 another attack against the Khazar fortresses Sudak and Tmutoro-
kan, and then against the Arab territories in the Caucasus. Tmutorokan
was taken. This gave the Rus’ the opportunity to gain a foothold on one
side of the Kerč Strait. Some have argued that Tmutorokan had already
been taken in 913, during a Rus’ campaign to the Caspian Sea,75 but more
recent studies have demonstrated that Rus’ conquest of the town cannot
be dated before 943. In 944, the Rus’ planned a new raid into Bulgaria
and Byzantium, this time in alliance with the Pechenegs, but the invasion
was eventually stalled by Byzantine bribes for Prince Igor. Moreover, Igor
was thus convinced to enter a treaty with the Byzantine Empire, in 944 or
945. The Byzantine enovys met Igor somewhere on the Danube frontier of
Bulgaria. The prince promised to defend the theme of Cherson theme and
the isthmus of Crimea against the “Black Bulgars,” which were apparently
coming from the lands north of the Sea of Azov, as well as against the
Pechenegs, and to stay away from the mouth of the Dnieper. At the same
time, the treaty allowed the “Black Bulgars” to attack the Khazar territo-
ries in Crimea, because that apparently served the Byzantine interests.76
The Rus’ attack of 941 made necessary a drastic overhaul of the orga-
nization of the navy. Emperor Roman I Lekapenos, a former naval officer
(drongarios), took measures to improve the Black Sea defense.77 Accord-
ing to some, those were the circumstances under which the staging post at
Tomis (mentioned in sources by its new name Constantia or Constanteia)
was restored. The surge in finds of Byzantine coins from Constanţa, all of
which may be dated to this period, has already been associated with the
revival of that town (another port of call was perhaps at Mangalia, the
old Greek and Roman town of Callatis).78 Such Byzantine staging ports
had both commercial and military functions.79 Taking into account the
peaceful Bulgarian-Byzantine relations established after 927, the settle-
ment of some Byzantine staging posts in Dobrudja could not be excluded,
but the scarce archaeological and numismatic evidence cannot support
that hypothesis.
According to De Administrando Imperio, in the mid-tenth century Bul-
garia had control of Dobrudja up to the mouths of the Danube, which
75 Stokes 1960, 507–514, Soloviev 1960, 574.

76 Russian Primary Chronicle, 73–77; Diaconu 1970, 20; Obolensky 1979, 130; Wozniak
1984, 307; Huxley 1984, 85–86; Whittow 1996, 257; Spinei 2006, 170, 174; Spinei 2009, 92–93;
Gordiyenko 2012, 165.
77 Ahrweiler 1966, 106.
78 Mănucu-Adameşteanu 1991 a, 304–308; Mănucu-Adameşteanu 1995–1996, 289–290.
79 Ahrweiler 1974 b, 163–164.
24 chapter one
means that Constantia was within Bulgarian territory. The Rus’ who
sailed along the coast “come to the Selinas, to the so-called branch of the
Danube River. And until they are past the river Selinas, the Pechenegs
keep pace with them. And if it happens that the sea casts a ‘monoxylon’
on shore, they all put in to land, in order to present a united opposition
to the Pechenegs. But after the Selinas they fear nobody, but, entering
the territory of Bulgaria, they come to the mouth of the Danube. From
the Danube they proceed to the Konopas, and from the Konopas to Con-
stantia, and from Constantia to the river of Varna, and from Varna they
come to the river Ditzina, all of which are Bulgarian territory. From the
Ditzina they reach the district of Mesembria”.80 Given that the treaty of
927 restored to Bulgaria all the ports between Mesembria and Agathopo-
lis, it is unlikely that the Empire could have maintained two small ports
of call in Dobrudja, and the island of Lykostomion. While the number of
Byzantine coins in Dobrudja increased after the middle of the 10th cen-
tury (according to finds not only from Mangalia and Constanţa, but also
from Isaccea, Hârşova, Capidava and other points),81 that is not enough
to prove a restoration of the Byzantine power in northern Dobrudja. The
increasing number of coin finds is simply a reflection of the accelerated
commercial exchanges between Bulgaria and the Byzantine Empire.
During the period when Bulgaria had control over the entire region
now known as Dobrudja, the so-called “Stone Dike” was erected, which is
the most recent among the three linear fortifications across that land from
Cernavoda to Constanţa (the Stone Dike overlaps the Small as well as the
Great Earthen Dikes) (see Fig. 1). The Stone Dike is an earthen rampart,
59 km long and 1,7–2,2 m thick, surmounted by a wall built of limestone
blocks and mortar, with a ditch to the north. Twenty-six forts with an area
between 2,5 and 10,8 hectares were erected along the dike. Pottery of the
Dridu type, which is typical for the 10th century was found in some of those
forts. Moreover, a solidus struck in 945–959 is known from fort XIV.82 An
excavation made along the dike at Medgidia identified a fireplace with the
same type of pottery, which was used when the foundation of the wall was
built (the stone traces of soot from the fireplace).83 The limestone blocks
80 Constantine Porphyrogenitus, IX, 91–102 (ed. Moravcsik, 62/63).

81 Mănucu-Adameşteanu 1996, 280–285.
82 Barnea, Ştefănescu 1971, 100–104, 111–114; Petre 1973, 27–31; Papuc 1992, 328; Bogdan-
Cătăniciu 1996, 202–207; Mănucu-Adameşteanu 2001 a, 80–84; Bogdan-Cătăniciu 2006,
413–417.
83 Papuc 2000, 58 (for some unknown reason, Papuc dates the fireplace to the 11th
century).

the recovery of the danubian frontier
1. The dikes between Cernavoda and Constanţa (after Barnea, Ştefănescu 1971, 168, map IV)
25
26 chapter one
carved in smaller dimensions than in the Roman period were extracted

from quarries like that in Medgidia.84 Petre Diaconu in fact believed that
the wall had been erected by the Bulgarians between 986 and 1000 as a
defense line against the Byzantine-controlled area in northern Dobrudja.85
According to other opinions, the dike was erected by the Byzantines dur-
ing the reign of Basil II.86
There is an incontrovertible proof that the Stone Dike was built in
the 10th century, namely the Cyrillic inscription discovered in 1950 in
fort VIII near the village of Mircea Vodă (Constanţa County). The inscrip-
tion, although damaged, reads, translated by the Slavicist Damian Bogdan
as: “. . . against the Greeks/ in the year 6451/ at the time of Demetrius/
župan . . .”87 Later, it became clear that the stone from Mircea Vodă has
two inscriptions, carved at two different moments. The first one includes
two lines from the end of an inscription which mentions the year 6451
(943 AD), while the last two lines referring to a župan named Demetrius
were carved with less precision at a later date.88 Župan Demetrius has
therefore nothing to do with the events of 943, whatever those may have
been. Some believed that in that year the Pechenegs invaded Bulgaria,
despite the failure of the campaign planned by Igor.89 However, no source
mentions any Pecheneg raid either in 943 or in 944.
The region behind the dike was a march-like district of medieval Bul-
garia. It appears that the rock monastery built in the chalk quarry in
Murfatlar-Basarabi was there at the time the wall was built, for the stones
were extracted from that same quarry.90 One of the Cyrillic inscriptions in
84 Panaitescu 1978 b, 247–251.

85 Diaconu 1962, 1215–1236; Diaconu 1965 b, 189–199, 383–394; Diaconu 1972 a, 375–378.
86 Cihodaru 1963, 1128–1129; Cihodaru 1965, 267; Salamon 1971, 495; Barnea 1975, 96;
Mănucu-Adameşteanu 2001 a, 84.
87 Barnea, Ştefănescu 1971, 64, 69; Božilov 1973 a, 12–13; Petkov 2008, 38. The inscription
was first published by Comşa 1951, 237 and Bogdan, Comşa, Panaitescu 1951, 122–128. Its
authenticity was questioned by Nandriş 1960, 530–534, who noted that the language pres-
ents old Russian and not old Bulgarian features, as one could expect: “Whereas the philolog-
ical arguments plead, in Bogdan’s view, for a Russian version of the text, palaeographically
the inscription is South Slavonic”. A refugee from Communist Romania, Nandriş lived at
that time in London and was vehemently anti-Communist. His intention seems to have
been to cast doubts about what he viewed as a forgery designed to prove the earliest Russian
presence in Romania. His views were not adopted by anyone else, and the inscription is still
viewed as one of the oldest Slavonic monuments.
88 Bogdan 1978, 151.
89 Božilov 1973 a, 23; Stănescu 1974, 396; Diaconu 1981 a, 219; Petkov 2008, 38.
90 Curta 1999, 129–149; Madgearu 1999–2000, 20; Curta 2006, 229–232. The most recent
study on the rock monastery in Damian, Samson, Vasile 2009, 117–158.
that monastery is dated to the mid-10th century on paleographic grounds,

and it is also known that the monastery must have been in existence
during the Prince Svyatoslav’s campaign to Bulgaria (968), because the
monks carved in the church B3 some figures of Rus’ (Varangian) warriors
and even Svyatoslav’s tamgha sign.91 Another inscription from church B4
reads ZHUPAN I IMAET GEORGE ONC TEBE ESTEK KRAIN I REZHETE,
which could be translated as “the župan and guardian George promises
you eighty pieces of gold and so must it be!” This reading has been cor-
rected by Kazimir Popkonstantinov, who believes that the word župan
should be replaced with Tupai (the name of the man who built a church
consecrated to Saint George).92 If so, there is no other mention of a župan
in central Dobrudja except that from Mircea Vodă.
The limestone block with the inscription was not found in situ, but had
apparently fallen next to the precinct of fort VIII. The block was appar-
ently extracted from the wall (as the late Eugen Comşa, the author of the
discovery, once told me); it was therefore not part of the foundation.93 It
is obvious that the block was once in the wall, but its exact position is
unknown. On the other hand, it is sure that the stone was taken from
another construction, because mortar remains may still be observed on
the block (it is important to note on the other hand that the wall was
not built with mortar). This construction was most probably a monument
erected to commemorate a battle (several Bulgar inscriptions attest to
that practice).94 Because the block belonged to the elevated part of the
wall, the added words “Župan Demetrius” were visible, and they could be
linked with a building activity done under the leadership of Demetrius,
when the wall was built or when it was repaired, sometime after 943. It
is more probable that the block with the inscription was inserted during
a repaired wall at the time when the dike served as protection against
the Byzantine forces in northern Dobrudja, that is between 986 and 1000
(see below).95 However, taking into account the huge amount of work
required for the building of the wall a fundamentis, it seems more prob-
able that the insertion of the block was done at some point during the
91 Popkonstantinov 1987, 120–125 (for the inscription); Pintescu 1999, 75; Agrigoroaei
2006, 25–49; Fetisov 2007, 299–314; Spinei 2009, 54.
92 Popkonstantinov 1987, 128–132.
93 As maintained by Barnea, Ştefănescu 1971, 64, 69, 114, and by other authors who repro-
duced their views.
94 Bogdan, Comşa, Panaitescu 1951, 129.
95 The idea that a repair made after 986 could be proved on the basis of the Mircea Vodă
inscription was first put forward by Florescu, Ciobanu 1972, 388–389.
28 chapter one
reign of Symeon or that of Peter.96 Paolo Squatriti has in fact advanced the
idea that the dike was only a “pharaonic work” intended to commemorate
the power and the prestige of one of the Bulgarian rulers, without any
real military purpose.97 However, the addition of 26 large forts to the dike
strongly suggests that it did indeed have a military function in the 10th
century. This linear fortification must have been conceived as a protection
against Pecheneg attacks directed at Preslav and other Bulgarian settle-
ments, mostly because it was provided with forts for soldiers.
Not surprisingly, the region next to the Stone Dike was well developed
and populated, as a recent archaeological excavation has demonstrated.
A large 10th-century settlement with stone houses was identified dur-
ing the rescue excavations made in May 2011 near the village Valul lui
Traian, not far from Constanţa. A cemetery of more than 300 graves was
also found near the northern side of the Small Earthen Dike (some of the
graves cut through the rampart). From the area of the same settlement a
small part of a coin hoard with 50 nomismata was recovered in 1935. The
eight known pieces have been struck for Constantine VII and Romanos II
between 945 and 959. Another small hoard of only 9 nomismata issued
for Constantine VII and Romanos II and another for Nikephoros II was
found in the surroundings. Both hoards were hidden during the events of
969–971. Other nomismata issued between 945 and 959 are known from
various points along the Stone Dike or just behind it, at Oltina, Rasova,
Urluia, Basarabi (Murfatlar), and Constanţa.98
After the treaty with Igor, the Byzantine Empire tried to maintain a bal-
ance between the Rus’ and the Pechenegs, in order to contain the power
of both. This reversal of alliances in the Black Sea region led to a joint
action against Bulgaria in 966, in addition to a Magyar raid against the
Empire. Emperor Nikephoros II Phokas (963–969) stopped the payment of
the usual tribute for Bulgaria, which had been established by the treaty of
927, and prepared for a large-scale campaign, which had to be cancelled at
the last moment because of a rebellion in Antioch.99 Instead, the emperor
96 Rašev 1987, 55–56; Rašev 2005, 57; Curta 2006, 230.

97 Squatriti 2005, 86–90.
98 Barnea, Ştefănescu 1971, 23; Mănucu-Adameşteanu 1996, 281; Poenaru-Bordea,
Ocheşeanu, Popeea 2004, 131–133; Custurea, Matei 2007, 105–113; Custurea, Talmaţchi 2011,
292, 374, 376; Cliante et alii 2012, 294–295.
99 Leo the Deacon, IV. 5 (ed. Talbot, Sullivan, 109–111); Skylitzes, Nikephoros II Phokas, 20
(ed. Thurn, 277; transl. Flusin, 232; transl. Wortley, 265); Zonaras, XVI, 27.13–15 (ed. Büttner-
Wobst, 513); Ostrogorsky 1956, 315–317; Stokes 1962, 50–54, 466–467; Gjuzelev 1981, 17; Fine
1991, 181–182; Busetto 1996, 11; Whittow 1996, 294, 326; Treadgold 1997, 502; Spinei 2006, 112.
chose to punish Bulgaria by proxies, namely his northern allies, the Rus’.
It was a better option to strike this country from the north, because an
offensive from the south required access to the mountain passes across
Stara Planina, which were too dangerous for the Byzantine army. The
emperor was certainly aware of the catastrophe of 811, when a Byzantine
army was ambushed and slaughtered in one of those passes, in which the
usual combat formation was impossible to deploy.100
Following the order of Nikephoros Phokas, the strategos of Cherson
Kalokyros convinced Prince Svyatoslav of Kiev (945–971) to attack Bul-
garia, but Kalokyros acted by his own treacherous interest, using the alli-
ance with the Rus’ warriors for his own plan to seize the imperial throne.
He suggested to Svyatoslav that in return, he would keep Bulgaria, in addi-
tion to a generous payment of 1,500 pounds of gold pounds (around 500
kg). In the summer of 968, the Kievan prince launched his campaign with
an army said to be as large as 40,000 men, with which he reached the
mouths of the Danube and then Dorostolon. The Bulgarian army, of about
30,000 men tried to defend the city, but was easily defeated, and Svyato-
slav conquered several fortresses in eastern Bulgaria, including the capi-
tal at Preslav. Some Bulgarian aristocrats went to the Rus’ side, choosing
that over the Byzantine rule. Thus, the rising Rus’ maritime power gained
control of the Lower Danube region, with the consent and support of the
Byzantine emperor. This appears to have been a major strategic mistake,
for the Rus’ gained control not only over the river, but also over a country
of great strategic and economic importance. Svyatoslav was fully aware of
the real meaning of his achievement. According to the Russian Primary
Chronicle, he declared that “I do not care to remain in Kiev, but should
prefer to live in Pereiaslavetz on the Danube, since that is the centre of
my realm, where all riches are concentrated; gold, silks, wine, and vari-
ous fruits from Greece, silver and horses from Hungary and Bohemia, and
from Rus’ furs, wax, honey, and slaves”.101 Indeed, precisely at this time
the Danube turned again into an important axis of trade route between
Central Europe and the East, at the same time as the revival of urban life
in Central Europe and the sedentization of the Magyar nomads. Jewish
merchants are known to have traveled from Spain to Khazaria by several
routes, one of which followed the Danube, while Kievan traders are said
to have already reached the German towns on the Upper Danube River in
100 Marinow 2011, 453.

101 Russian Primary Chronicle, 85–86.
30 chapter one
the mid-9th century. In that respect, gaining control of the Lower Danube
was both a military and an economic target for the Rus’.102
Confronted with such a critical situation, Nikephoros Phokas decided to
force the return to Rus’ of at least some of Svyatoslav’s forces by asking the
Pechenegs to attack Kiev. He acted like Symeon during the Magyar inva-
sion in Bulgaria in 894–895, but with less success, since the Rus’ were a
redoubtable force, difficult to stop by means of a simple Pecheneg raid. In
fact, the Rus’ prince showed no desire to abandon his expansionist plans.
Svyatoslav returned to Preslav from Kiev in August 969, subduing the new
Bulgarian emperor Boris II (969–971), who had recovered his capital a few
months before that with Byzantine assistance. In the second campaign,
Svyatoslav was wise enough to ally himself with the Pechenegs and the
Magyars. From Preslav, he attacked the Byzantine territory in coopera-
tion with those Bulgarians who had accepted his rule. That shows how
ineffective the diplomacy of Nikephoros Phokas has become. The inter-
nal conflicts within the empire, in which Kalokyros had intended to get
involved, have by now undermined any possible success in the military
field. Nonetheless, because the Byzantine general Bardas Skleros obtained
a decisive victory against the Rus’ at Arkadiopolis (today Lule Burgas), 140
km to the west from Constantinople, in the spring or summer of 970. The
Byzantine army of no more than 12,000 men, was most certainly outnum-
bered by the Rus’ coalition with as many as 30,000 men, but managed to
obtain the victory by means of moving an army corps to the flanks, thus
encircling the enemy by surprise.103
At that moment, Nikephoros Phokas had already been overthrown
by John Tzimiskes (969–976), a very good general, who began his reign
with a truce with the Rus’, as he was challenged from a different corner
by Bardas Phokas’s rebellion. The truce, on the other hand, left Svyato-
slav unprepared for what Tzimiskes had in plan for him Tzimiskes in
102 Leo the Deacon, IV. 6; V. 1–3 (ed. Talbot, Sullivan, 111–112, 128–132); Skylitzes, Nike-
phoros II Phocas, 20; John Tzimiskes, 5 (ed. Thurn, 277, 288–291; transl. Flusin, 232–233, 242;
transl. Wortley, 265, 275–276); Zonaras, XVI, 27.16–18; XVII, 1.17–20 (ed. Büttner-Wobst, 523–
525); Stokes 1962, 467–476; Lewis 1975, 363–364; Wozniak 1984, 309–310, Fine 1991, 182–183;
Barnea 1993 b, 586; Franklin, Shepard 1996, 88–89, 145–146; Busetto 1996, 12–14; Whittow
1996, 260–261; Stephenson 2000, 48–51; Curta 2006, 238. The location of Pereiaslavetz will
be discussed in chapter II. 2.
103 Leo the Deacon, VI. 8, 10–13 (ed. Talbot, Sullivan, 153, 155–161); Skylitzes, John Tzi-
miskes, 5–6 (ed. Thurn, 288–291; transl. Flusin, 242–245; transl. Wortley, 275–279); Russian
Primary Chronicle, 87–88; Stokes 1962, 480–493; Zonaras, XVII, 1.21–26 (ed. Büttner-Wobst,
525); Wozniak 1984, 310–311; Fine 1991, 183–186; Hanak 1995, 138–151; Franklin, Shepard 1996,
145–147; Busetto 1996, 15–18; Treadgold 1997, 504, 508; Krsmanović 2008, 35, 131.
the spring of 971. As soon as he finished dealing with Bardas Phokas, the
Byzantine army moved rapidly through eastern Bulgaria, taking back the
ports that had been lost in 917. According to Leo the Deacon, the Byzan-
tine forces included 13,000 pedestrians and 15,000 cavalrymen. The army
was commanded by the emperor himself, assisted by the stratilates Bardas
Skleros and the stratopedarches Peter Phokas (who commanded troops
from Thrace and Macedonia). The capital Preslav was taken from the Rus’
and the Bulgarians on April 4th. The city was renamed Ioannoupolis after
the emperor. The victory at Preslav was only the first step in a large offen-
sive against the Rus’, for the main body of the army headed by Svyatoslav
was at that time concentrated in Dorostolon. For Svyatoslav, that city was
not only a place of refuge, but also a position that allowed him to main-
tain control over the Danube.104 He ignored the threat from the Byzantine
navy led by drongarios Leo, which had already been sent to Dorostolon
in late March, with the specific mission to block the fords by which the
Rus’ could withdraw to the north. A seal of that commander of the fleet
was in fact found in Preslav,105 a clear indication of pre-planned, joint
operations. Meanwhile, the Byzantine land army, marching from Pre-
slav toward Dorostolon, took several other fortresses, such as Pliska and
Dineia.106 A number of coin hoards were hidden during this campaign,
such as that found in Razgrad.107 Having reached Dorostolon on April
23rd, the land army laid the siege. Initially, Svyatoslav appears to have
effectively defended the city and even to have launched small attacks
against the Byzantine troops. However, when the Byzantine fleet showed
up on April 25th, the Rus’ suddenly found that their retreat routes had
been blocked. Moreover, the Byzantines could now use the “Greek fire”
from the ships against those on the ramparts.108
The Byzantine army focused on the western gate of Dorostolon (the
troops in that area were under the command of Bardas Skleros), while
the eastern gate was under the attack of Peter Phokas. The chronicle of
John Skylitzes mentions that a delegation came to the emperor during
the siege: “A delegation now came to him from Constanteia and the other
104 Iorga 1998, 170–171.

105 Jordanov 2009, 378–379, nr. 1049.
106 Dineia has been identified with modern Vojvoda by Beševliev 1962, 72.
107 Dzanev 2007.
108 Leo the Deacon, VIII, 1–10; IX, 1–2, 6 (ed. Talbot, Sullivan, 176–188, 192–193); Skylitzes,
John Tzimiskes, 9–11 (ed. Thurn, 295–300; transl. Flusin, 247–251; transl. Wortley, 281–286);
Zonaras, XVII, 2.16–37 (ed. Büttner-Wobst, 528–530); Franklin, Shepard 1996, 149; Busetto
1996, 19–26; Haldon 2001, 99–101.
32 chapter one
fortresses established beyond the Danube. They sought an amnesty for

their misdeeds [in return for] handing over themselves and the strong-
holds. He received them kindly, dispatching officers to take charge of the
fortresses and with sufficient troops to secure them.”109 That information
is missing in the minutely detailed report of Leo the Deacon, no doubt
because Skylitzes culled it from a now lost source. Unlike Leo the Dea-
con who framed the narrative in comparison to battles of Antiquity, Sky-
litzes paid more attention to the details.110 In that respect, even though he
wrote at a later date, Skylitzes is a more trustworthy source than Leo the
Deacon in certain respects. To be sure, he does not explain of what exactly
have been guilty the garrisons of those fortresses. Asking for forgiveness
from the emperor would make sense only if they had been either Byz-
antines, or allies disobeying orders or betraying oaths. Since it is hard to
imagine the emperor being so clement, had those soldiers been his own,
it is more likely that they were Bulgarians. Tzimiskes wished to be seen as
a liberating Bulgaria from the Rus’ occupation. He recognized Boris II as
emperor (the Bulgarian ruler was captured by Svyatoslav and freed when
Preslav had been conquered). In the mean time, some Bulgarians had
chosen Svyatoslav’s side. Such opportunists have by now started to leave
the Rus’ as the victorious Byzantine army was approaching.111 That may
be the context of Skylitzes’ story of forts surrendering to Tzimiskes. The
emperor had no reason to reject the offer of the Bulgarians who controlled
those forts, but could not trust them either. For this reason he set new
garrisons in those forts, as if following the advice of the military treatise of
Syrianos Magister, according to which “we must not entrust the safety of
these forts or assign to their garrisons men who have once been captured
by the enemy.”112
If the unnamed fortifications were located on the northern bank of the
Danube, that would apply to Constanteia as well, which is absurd. This
Constanteia is most certainly the same as the port mentioned by Constan-
tine Porphyrogenitus, the ancient city of Tomis and the present-day city of
Constanţa. As a consequence, some have advanced the idea that the pas-
sage contains two different sentences, which have been collapsed—one
109 Skylitzes, John Tzimiskes, 12 (ed. Thurn, 301; transl. Flusin, 252; transl. Wortley, 287).
See also Zonaras, XVII, 2.40–42 (ed. Büttner-Wobst, 530).
110 McGrath 1995, 157–158.
111 Stokes 1962, 479, 486; Oikonomides 1965, 63.
112 Peri strateias, 30/31. For the date and authorship of this treatise, see Rance 2007–2008,
701–737.
about Constanteia and the other about the northern fortifications.113 There
is however no evidence to support such an interpretation. Much more
likely is the reading of the passage as “from Constanteia, and from other
forts, [located] beyond the Danube.” In other words, only the un-named
forts were beyond the river. Some have proposed to locate those forts in
northern Walachia (at Slon), or even in the Banat or in Transylvania,114 as
no forts are known to have existed on the northern bank of the Danube
in the vicinity of Dorostolon. Since they took for granted the location of
those forts in the lands north of the river Danube, others advanced solu-
tions that ignore the political and military framework of the events. For
example, Mihai Sâmpetru believed that the forts in question must have
been located to the west from the mouth of the Argeş river, only because
most 9th-century building materials believed to be Byzantine have been
found in that area.115 As we have seen, those archaeological remains are
more likely to be associated with the Byzantine prisoners forcefully moved
there by the Bulgars in 813. Since they are known to have returned to the
empire in 838, their settlements could not have possibly been the same as
fortresses said to been in operation in 971.
The forts must have been in the vicinity of Dorostolon, and if so, their
mission was probably to prevent Pecheneg attacks. It is impossible that
John Tzimiskes would have sent his soldiers at a great distance, at a time
when he needed them for the siege of Dorostolon.116 One could go as far
as to interpret this as a tactical move in relation to the siege, for it would
otherwise make no sense to give up on those troops at the very moment
the emperor most needed them. Therefore, one would have to admit that
the forts mentioned by Skylitzes could not have been too far from Doros-
tolon. Control over them meant a serious blow to the enemy’s forces and,
perhaps, an attempt to cut any possibility of withdrawal. It is obvious that
Constanteia (Constanţa) served as anchorage for the Rus’, in case they
needed to withdraw by sea. Another way to withdraw was by the Danube,
either on water or along the valley, by land. Given that the Byzantine navy
was already at Dorostolon, withdrawing by water on monoxyles was not
an option. Meeting in council with his leaders, Svyatoslav was aware that
113 Stănescu 1974, 398; Mănucu-Adameşteanu 1991 a, 317.

114 Stănescu 1974, 398–399; Diaconu 1969 b, 395–396; Diaconu 1987 a, 217–218; Diaconu,
Vâlceanu 1972, 15; Bica 2003, 45–46 (Spinei 2009, 100 is quite skeptical about such a remote
location).
115 Sâmpetru 1974, 256.
116 Skylitzes, John Tzimiskes, 12 (ed. Thurn, 301; transl. Flusin, 252; transl. Wortley, 287).
34 chapter one
the “Greek Fire” would prevent any Rus’ attempt to flee in canoes on the
Danube.117 The only hope the Rus’ could entertain was to withdraw by
land to Constanteia (and then to go by sea), or to march along the Dan-
ube, the latter being obviously a longer route. That was without any doubt
the reason for which Tzimiskes gladly accepted the offer to extend his
control over Constanteia and other forts, the position of which could have
blocked the access of the Rus to the Danube route. This is in fact another
argument in favor of locating those forts in the vicinity of Dorostolon.118
The chronicle of Yahya-ibn-Said from Antioch (which was compiled in
1013) mentions that, after Svyatoslav’s capitulation, Tzimiskes took the city
(Thaisaira, an Arabicized version of Dristra, an alternative, perhaps native
of Dorostolon) “and the neighboring forts previously conquered by the
Russians.”119 The same appears in Skylitzes: “Once the Russians had sailed
away, the emperor turned his attention to the fortresses and cities along
the banks of the river and then he returned to Roman territory.”120 The
Arab chronicler continues: “the emperor returned to Constantinople after
he appointed commanders for these forts”, which must be different from
those which had surrendered before.121 The two sources are therefore in
agreement: some of the forts controlled by the Rus’ surrendered willingly,
while others were taken by the Byzantine army at the end of the military
operations. At any rate, they were all located within the main theater of
operations, either on the right, or on the left bank of the Danube. Petre
Diaconu believed that the forts taken by the Byzantines after the war were
all on the right bank, since in various parts of Skylitzes’ chronicle, the
word ὂχϑας (plural of ὂχϑη = “bank”) is used for a single bank.122 None-
theless, nothing indicates that Skylitzes needed to be so specific in his
terminology at this point, or anywhere else in his work.
If the fortress from the Păcuiul lui Soare island was built before the
971 war, then it could well be one of those Bulgarian strongholds in the
vicinity of Dorostolon, especially since it was initially on the left bank of
the river (the present branch of the Danube known as Ostrov was the
117 Leo the Deacon, IX. 7 (ed. Talbot, Sullivan, 195).

118 Andronic 1969, 207–215; Barnea, Ştefănescu 1971, 73–74; Božilov 1973 b, 111–122; Frank-
lin, Shepard 1996, 149–150; Busetto 1996, 9–32; Madgearu 1999–2000, 7–14.
119 Yahya-ibn-Said, 833. The reference given by Oikonomides 1965, 63, footnote 25 and
uncritically reproduced after him by Barnea, Ştefănescu 1971, 74 is wrong.
120 Skylitzes, John Tzimiskes, 18 (ed. Thurn, 310; transl. Flusin, 258; transl. Wortley,
293–294).
121 Not the same, as Božilov 2008, 93 thinks.
122 Diaconu 1969 a, 43–44.
main stream of the river at that time). The landscape around Păcuiul lui
Soare has changed considerably since the Middle Ages, and it is quite pos-
sible that some other forts existed on the left bank, which have mean-
while been flooded.123 However, the excavator of Păcuiul lui Soare, Petre
Diaconu, has demonstrated that the first phase of the fort is to be dated
just after the siege of Dorostolon in 971.124 The problem of this fort’s chro-
nology will be discussed in further detail in chapter III. 1.
If there were any Bulgarian forts on the northern bank of the Danube
in the vicinity of Dorostolon, then those forts must have been part of the
defense system built against the Magyars and the Pechenegs, to which
belonged a number of fortifications on the southern bank, such Oltina,
Rasova and Cochirleni-“Cetatea Pătulului”—all dated between the 9th
and the 10th century on the basis of the fine Gray Ware of the Dridu B
type found inside each one of them. The fort at Cochirleni, located at 100
m from the western end of the Great Earthen Dike, has ramparts made
of stones and mortar, enclosing an area of 3 ha. The fort near Rasova was
built in the same technique, with a 4 m wide rampart. No excavations
have been carried so far in any of them.125 The excavations in Oltina (a
fortress located 35 km downstream from Dorostolon, near Satu Nou, at a
place called “Capul Dealului”) suggest that that fort, which was erected on
a 12 ha-large promontory defended by a single earthen rampart, was built
in the 9th century, as demonstrated by a coin struck for Emperor Theo-
philos (830–842) and found on the site.126 Everything points therefore to
a system of forts defending the axis of the Danube downstream (i.e., to
the east and northeast) from Dorostolon, on the southern bank, since the
9th century. Those are most likely the forts mentioned by Skylitzes and
Yahya-ibn-Said.
The last attempt to break the siege of Dorostolon was decided on July
20th, 971 by Svyatoslav in a council with his chieftains, but ultimately
failed. Leo the Deacon mentions that the council in question was called
κομέντον in the language of the Rus’.127 This, however, is a word of Latin
origin, from which the Romanian word for “word” (cuvânt) derives. Niko-
laos Oikonomides remarked that Leo the Deacon had found the word
123 Vâlceanu 1967, 593–615.

124 See especially Diaconu 1969 b, 395–400; Diaconu 1988 a, 181–183. For other opinions
on the chronology of the naval base in Păcuiul lui Soare, see chapter III. 1.
125 Diaconu 1973–1975, 200–203; Florescu 1986, 173; P. Diaconu, Cochirleni, in EAIVR, I,
324.
126 Custurea 2000–2001, 589–590; Chiriac, Custurea 2002, 222.
127 Leo the Deacon, IX. 7 (ed. Talbot, Sullivan, 194).
36 chapter one
in a military report going back to information collected from spies who

spoke Romanian, and was thus misled into believing that the word was
Rus’. This in fact is a valuable piece of information about the presence of
Romanians (or at least speakers of Romanian) in the area, and of their
use for collecting intelligence for the Byzantine army. Incidentally, the
same word (κομέντoν) is used by Skylitzes in reference to a council of the
Pecheneg chieftains during the events of 1048–1049. Some believe that
the word is Pecheneg.128 Pavel Georgiev attempted to to reject Oikono-
mides’ interpretation by arguing that the word was borrowed by the Rus’
from the Bulgarians, who in turn took it from the Greek language (in
which the word must have been komventon).129 While this theory seems
to hold water at the first glance, Georgiev ignores the fact that the word
in question had already disappeared from the Greek language, at the time
the Bulgarians were supposed to have borrowed it. That, in fact, was the
reason for which Oikonomides proposed its transmission via Romanian.
A military treatise composed under Emperor Basil II insists that “actual
spies, however, are the most useful. They go into the enemy’s country and
can find out exactly what is going on there and report it all back to those
who sent them. The domestic and the generals along the border should
be sure to have spies not only among the Bulgarians but also among all
the other neighboring peoples, for example, in Patzinakia, in Turkey [the
land of the Hungarians], in Russia, so that none of their plans will not be
known to us.”130
After much fighting, overwhelmed by both land and naval forces that
had blocked the city, Svyatoslav finally surrendered on July 24th, 971, and
agreed to enter a peace treaty with John Tzimiskes. The Rus’ promise to
withdraw from Bulgaria under the supervision of the Byzantine navy.131
Following his victory, John Tzimiskes sent envoys to the Pechenegs, asking
for an alliance to prevent any further Rus’ attacks on the newly conquered
Bulgaria. Svyatoslav was in fact killed by the Pechenegs in an ambush near
128 Oikonomides 1987, 187–190.

129 Georgiev 1988, 87–92.
130 Anonymous Book on Tactics, 292/293.
131 Leo the Deacon, IX. 8–11 (ed. Talbot, Sullivan, 196–200); Skylitzes, John Tzimiskes,
14–18 (ed. Thurn, 304–310; transl. Flusin, 254–258; transl. Wortley, 289–293); Russian Pri-
mary Chronicle, 89–90; Zonaras, XVII, 3.19–22 (ed. Büttner-Wobst, 535); Barnea, Ştefănescu
1971, 73; Fine 1991, 187; Franklin, Shepard 1996, 149–150; Busetto 1996, 27–28; Stephenson
2000, 53; Haldon 2001, 101–104.
the Dnieper cataracts on his way back to Kiev.132 Meanwhile, Boris II was
taken to Constantinople where the emperor “took off the royal insignia
(they were a tiara with purple border, studded with gold and pearls, and
a purple robe and scarlet boots), and honored him [Boris II] with the
rank of magistros.”133 This was the final act of the long battle for restor-
ing the Byzantine power over what Leo the Deacon called Mysia, the for-
merly Roman province which had falled under the barbarian Bulgars after
Emperor Constantine IV’s failed expedition against them. The immedi-
ate consequence of the 971 war was the disappearance of the Bulgarian
state. The main chronicler of the events, Leo the Deacon, in fact treated
Tzimiskes’s campaign as a reintegration of “Mysia” into the empire (he
must have used that archaic name for the land in order to emphasize that
that was a Roman province recovered from the barbarians).134 Indeed, the
territories gained by the Byzantine Empire were quite extensive. Besides
the region between the Danube and the Stara Planina, the part of Thrace
previously within Bulgarian borders was now included into the duchy of
Adrianople. The western boundary of the newly conquered lands along
the Danube was probably the valley of the Iskar river.135 Most troops were
in fact stationed near the Danube and in northeastern Bulgaria, since that
was the direction from which further Rus’ attacks were expected.136
The military organization established by John Tzimiskes is reflected in
a list of offices known as the Taktikon Scorialensis, which was dated to 975,
published, and studied by Nikolaos Oikonomides.137 That text includes
data on otherwise unknown military and administrative units, the exis-
tence of some of which has been later confirmed by means of lead seal
finds. The theme of Ioannoupolis was created in northeastern Bulgaria,
and placed under the command of a strategos of Thrace and Ioannoup-
olis residing in Preslav, the name of which, as we have seen, had been
changed into Ioannoupolis. Another strategos, according to the Taktikon
Scorialensis, resided in Dristra.138 This is in fact the the first source to
132 Skylitzes, John Tzimiskes, 18 (ed. Thurn, 310; transl. Flusin, 259; transl. Wortley, 294);
Russian Primary Chronicle, 90; Zonaras, XVII, 3.22–23; 4.5 (ed. Büttner-Wobst, 535, 536);
Tăpkova-Zaimova 1973, 91; Pritsak 1975, 232; Stephenson 2000, 53; Malamut 1995, 116; Spinei
2006, 176.
133 Skylitzes, John Tzimiskes, 18 (ed. Thurn, 310; transl. Flusin, 258–259; transl. Wortley,
294). See also Leo the Deacon, IX. 12 (ed. Talbot, Sullivan, 201).
134 Brezeanu 1991, 112.
135 Tăpkova-Zaimova 1975 a, 113–115.
136 Haldon 1999, 64.
137 Oikonomides 1965, 57–60; Oikonomides 1972, 255–277.
138 Oikonomides 1972, 264/265, 268/269.
38 chapter one
mention the new name of the city previously known as Dorostolon. The
name Dristra is most likely modeled after the Bulgarian name Drăstăr,139
which suggests that the Byzantine administration was making efforts to
win over its new subjects. However, the Bulgarian form of the name had
been familiar to the Byzantines since the late 9th century. In the so-called
Chronicle of Monemvasia, a confused tradition about the Avars maintains
that they were received by Justinian in “Dorostolon, called today Dris-
tra” (Δωροστόλῳ, τῇ νῦν καλουμένῃ Δρίστρᾳ).140 The Bulgarian name of the
town, Drăstăr, seems to have its origins in a person’s name—Dristros, a
Bulgarian count known from the reign of Symeon.141
In his account of the battle of Dorostolon, Leo the Deacon claims that
Emperor John Tzimiskes changed the name of the city into Theodoroupo-
lis to honor a miracle by Saint Theodore Stratilates, which had helped the
Byzantine army during the siege.142 That information does not appear in
other sources and it is even contradicted by the mention of the name Dris-
tra in the Taktikon Scorialensis and in the inscriptions of several lead seals
dated to that same period. Nonetheless, the name change from Dorostolon
to Theodoroupolis has by now been accepted by some historians.143 Four
seals from Preslav belonging to a man named Sisinios, protospatharios and
katepano of Theodoroupolis, confirmed that opinion.144 Ivan Jordanov ini-
tially believed that no less than three themes have been established in
971: Dorostolon, Thrace and Ioannoupolis, and Theodoroupolis. Accord-
ing to him, the latter was located north of the Danube and in the north-
ern part of Dobrudja.145 Later, Jordanov changed his position and claimed
that Theodoroupolis was in fact not Dorostolon, but Presthlavitza or Little
Preslav146 (the location of which will be discussed in chapter II. 2). Petre
Diaconu, following Nikolaos Oikonomides, was of a different opinion:
Sisinios was the commander of a tagma recruited from Euchaita, a city
in Paphlagonia (now Avhat, in Turkey) which was famous for the relics
of Saint Theodore Stratilates. Moreover, the metropolitan of Euchaita was
139 Mutafčiev 1932, 193–194.

140 Kislinger 2001, 199.
141 Mutafčiev 1932, 195; Beševliev 1963, 218, nr. 46.
142 Leo the Deacon, IX, 9, 12 (ed. Talbot, Sullivan, 197, 200).
143 Bănescu 1946, 68; Stănescu 1974, 397; Kuzev 1979, 34; Kuzev, Gjuzelev 1981, 184;
Wasilewski 1995, 190, 200; Soustal 1997, 120.
144 Jordanov 1982 a, 20–21; Jordanov 1993, 124–125, nr. 228–231; Jordanov 2003 a, 88–89,
nr. 33.1; Krsmanović 2008, 139; Jordanov 2009, 426, nr. 1208–1211.
145 Jordanov 1982 a, 12–23.
146 Jordanov 1987 b, 200–201; Jordanov, Tăpkova-Zaimova 1988, 120; Jordanov 1993, 124–
125, nr. 228–231; Tăpkova-Zaimova 1993, 96. See also Stephenson 2000, 53.
at Dorostolon with the Byzantine army. Thus, Leo the Deacon must have
mistaken Dorostolon for the city of Euchaita, about which Skylitzes knew
that it had received the new name Theodoroupolis in honor of the victory
of 971, attributed to the assistance of the military saint.147
Oikonomides offered a third solution: Dorostolon was indeed called
Theodoroupolis, but only for a short time. Sisinios must have been the
first governor of that province, and he had the rank of katepano. Dristra
was the new name introduced for the city and for the theme at some
point before 975 (the date of the Taktikon Scorialensis), probably at the
same time as Euchaita was renamed Theodoroupolis in honor of the local
saint, who had contributed to the victory of Dorostolon.148 After a while,
the rank of the commander residing in Dorostolon/Dristra was lowered
from katepano to strategos. That was the situation captured in the Tak-
tikon Scorialiensis, ca. 975.
Oikonomides placed in the same region of the Lower Danube another
province mentioned in Taktikon Scorialensis—“Western Mesopotamia”,
ruled by a katepano who had a strategos under his orders. The katepano
of Mesopotamia was not called “from the west”, but the existence of a
duke of Mesopotamia (an equivalent function)149 led Oikonomides to the
conclusion associate that the katepano was of Western Mesopotamia.150
One should note the high rank reserved for the katepano of Western
Mesopotamia in this list of offices, namely immediately after the dukes
of the eastern provinces.151 The dukes or katepanoi recorded in the list
were commanders of provinces—Antioch (duke), Mesopotamia (duke),
Chaldia (duke), Western Mesopotamia (katepano), Italy (katepano), Thes-
saloniki (duke), Adrianople (duke). Oikonomides believed that West-
ern Mesopotamia was the same as Atelkuz (a Hungarian word with the
same meaning, namely “between rivers”).152 This, however, would imply
that John Tzimiskes took over southern Moldavia as well. Oikonomides’
interpretation was adopted by others,153 but there is no proof for such an
147 Oikonomides 1986 a, 327–335; Diaconu 1986, 170–171; Diaconu 1987 c, 483–484.

148 Oikonomides 1998, 586–587. Jordanov 2003 a, 88–89 agreed with this.
149 For the synonymy see Cheynet 1985, 181; Kühn 1991, 158–168; Krsmanović 2008, 78.
Only Wasilewski 1975, 644 claimed that the dukes were commanders of mercenaries,
while katepanoi were commanders of stratiotai, the soldiers who owed military service in
exchange for land property.
150 Oikonomides 1965, 57–60, 73–74; Oikonomides 1972, 262/263, 268/269.
151 Kühn 1991, 221.
152 Oikonomides 1965, 68–72.
153 Andronic 1969, 211; Božilov 1976, 19–32; Diaconu 1970, 24–25; Diaconu 1987 a, 218–
219; Makk 2002, 273–281; Holmes 2005, 398; Božilov 2008, 94; Krsmanović 2008, 133. Some
40 chapter one
extension of Byzantine power north of the river Danube in 971. Tzimiskes

had neither the necessary forces, nor indeed the need to occupy such a
distant territory, the security of which was better left to the allied Pech-
enegs, who were the enemies of the Rus’. He definitely had more important
problems in the East, where the war with the Arabs was resumed in 972.154
True, copper coins (folles) struck for John Tzimiskes and Basil II have been
found in Walachia, but such finds could hardly indicate a military occupa-
tion, as similar coins are also known from northern Moldavia, the Banat
and Hungary—all regions that were most certainly not occupied by the
Byzantines.155 Those coins are rather an indication of trade relations. On
the other hand, it is quite obvious that the single, most important goal of
the 971 campaign was the defeat of Svyatoslav. The occupation of a larger
area north of the river Danube, or even of bridgeheads, was pointless, for
it could have meant only a waste of precious military resources.
Oikonomides returned to the problem of the Western Mesopotamia
katepanate in one of his latest studies, without changing his position on
the location of that administrative unit in the region of the Danube Delta,
or on the left bank of the river. He also maintained that that province
did not survive the Bulgarian offensive of 986.156 The discovery in Preslav
of the seals of one Damian Dobromir, patrikios and anthypatos, duke of
Thrace and Mesopotamia, changed the terms of the equation.157 The uni-
fied command over the provinces Thrace and Mesopotamia concerned in
fact troops under the command of Damian Dobromir.158 That person is
the same as Dobromir, the commander of the Bulgarian fortress of Berhoia
(Veria) in Macedonia, who went to the Byzantine side in 1001, after which
he received the title of anthypatos (according to an interpolation into the
chronicle of Skylitzes, he was married to a niece of Emperor Samuel).159
researchers accepted only the existence of the strategy of Western Mesopotamia, the
katepanate being located in the east (Tăpkova-Zaimova 1973, 95–96; Wasilewski 1975, 641;
Tăpkova-Zaimova 1993, 96; Wasilewski 1995, 198–200; Madgearu 1999 a, 421–422; Madgearu
1999–2000, 5–7, 16; Stephenson 2000, 56). I have given up that interpretation as it was not
possible to have both a duke and a katepano at the same time within the same eastern
Mesopotamia.
154 Ostrogorsky 1956, 321; Treadgold 1997, 511.
155 See the gazetteer in Custurea 2000, 185–199.
156 Oikonomides 1998, 584–585.
157 Jordanov 1984, 99–105; Jordanov, Tăpkova-Zaimova 1988, 120; Jordanov 1993, 127–128,
nr. 237–238; Jordanov 2003 a, 98–99; Jordanov 2006, 130–131, nr. 168–169; Seibt 2008, 104;
Jordanov 2009, 435–436, nr. 1248–1249.
158 Oikonomides 1998, 583–584.
159 Skylitzes, Basil II and Constantine, 27 (ed. Thurn, 344; transl. Flusin, 287; transl. Wort-
ley, 326); Zonaras, XVII, 8.10 (ed. Büttner-Wobst, 559).
It is even possible that Damian Dobromir was of Vlach origin, since it is

quite unlikely for a Bulgarian to have received that office in the midst of
the conflict with Bulgaria.160 Members of the Bulgarian aristocracy who
submitted to the Byzantine authority were bestowed titles and office away
from the Balkans, in the distant provinces of the East. Oikonomides, who
believed Damian Dobromir to have been a Bulgarian, rejected the idea
that he could have been appointed a duke at the time when the Bulgar-
ians were still fighting the Byzantine army. Instead, according to Oikono-
mides, Damian Dobromir was invested with that office soon after 971.
Oikonomides does not seem to have realized that this Damian Dobromir
was the same person as that mentioned by Skylitzes for the year 1001.
I first advanced that identification in 1999, and Stepan Nikolov reached
the same conclusion independently.161 In short, this would make Damian
Dobromir the commander of the joint forces from the themes Thrace and
Mesopotamia, which were sent to the Danube in 1000 or shortly after that.
The office of duke was equivalent to katepano. Another officer from that
province was Leo, imperial spatharocandidate and turmarches of Mesopo-
tamia, known from three lead seals discovered in Silistra and dated to the
10th century.162 Given the place of their discovery, Leo was probably the
governor of western Mesopotamia.
The location of Western Mesopotamia in the northern part of Dobrudja,
between the Danube, the Black Sea and the Carasu valley, is now accepted
by most scholars,163 who agree that the name referred to that position
“between rivers.” According to another opinion, Western Mesopotamia
included the strategies of Dristra and Ioannoupolis, which were separated
after 975.164 However, Werner Seibt has brought to attention two lead seals
160 As I sustained in Madgearu 1999–2000, 18.

161 Madgearu 1999 a, 422; Madgearu 1999–2000, 18; Nikolov 2001, 146. See also Curta 2006,
244. Jordanov 1984, Oikonomides 1998, 585, Stephenson 2000, 67, 77 and Strässle 2006, 161,
328 did not notice the identity between Dobromir from Veria and Damian Dobromir. More
recently, Krsmanović 2008, 137–138, 144, while citing my study, advanced the idea that the
Bulgarian nobleman was appointed duke of Western Mesopotamia soon after 971, after
which he joined Samuel, before turning again to the Byzantine side in 1001.
162 Jordanov 2003 a, 125–126, nr. 48.4; Jordanov 2006–2007, 521; Jordanov, Žekova
2007, 119, nr. 311; Jordanov 2008, 42–43, nr. 9–10; Seibt 2008, 104; Jordanov 2009, 458–459,
nr. 1353–1355.
163 Andronic 1969, 211; Tăpkova-Zaimova 1973, 95–96; Stănescu 1974, 399; Shepard 1985,
253; Diaconu 1986, 167–173; Kühn 1991, 221–222; Tăpkova-Zaimova 1993, 96; Madgearu 1999–
2000, 5–12; Mănucu-Adameşteanu 2001 a, 92; Tăpkova-Zaimova, Stoimenov 2004, 342–345;
Jordanov 2005, 263–272; Mărculeţ 2006, 295–319; Strässle 2006, 222; Yotov 2007, 322; Božilov
2008, 94; Spinei 2009, 101.
164 Mărculeţ 2005, 28–35; Mărculeţ 2006, 297–298, 316.
42 chapter one
of strategoi of Mesopotamia, which were in Izvorovo and Simeonovgrad

(southern Bulgaria),165 and dated to the 9th century, before the recovery of
eastern Mesopotamia, the province between the Tigris and the Euphrates.
According to Seibt, a new province of Mesopotamia was established in the
early 9th century somewhere between Adrianople and Thessaloniki, in an
area to which settlers from Mesopotamia have been brought.166 If Seibt is
right, then also Damian Dobromir should also be placed in Macedonia,
near Thrace, the other theme which provided troops under his command.
A similar point of view had been advanced some time ago by Alexandru
Bolşacov-Ghimpu, but his idea remained unknown to Seibt.167
In my opinion, the 9th-century seals from southern Bulgaria are an
incontrovertible evidence of a theme named Mesopotamia established in
southern Thrace at a time when the Byzantine-Bulgar frontier ran along the
line Develtos—Agathonike (Orjakhovo)—Constantia (Simeonovgrad)—
Hebros (Maritsa) river, as established by the treaty of 816. This area had
a great importance for the forward defense of Constantinople.168 The
absence of any information for the subsequent period strongly suggests
that the theme had a very short life. On the other hand, locating the
katepanate of Western Mesopotamia in southern Thrace soon after 971
made no sense whatsoever from a strategic point of view, since nothing
threatened that region after the disappearance of Bulgaria. By contrast,
such a military and administrative unit was imperiously necessary near
the Danube frontier, which was threatened by the Rus’, or at least per-
ceived to be so. For this reason, I am inclined to believe that Western Mes-
opotamia mentioned in the Taktikon Scorialensis was in fact in northern
Dobrudja, its southern border being marked by the Stone Dike. Damian
Dobromir and Leo were commanders in that theme, and not of the theme
in southern Thrace, which had by then disappeared. They both resided
probably in Noviodunum-Isaccea, the most important center of northern
Dobrudja, where several seals dated after 971 have been found. As for the
residence of the strategos under the orders of the katepano of Western
Mesopotamia, he must have been only the commander of a fortress from
that province.169 Constanteia seems to be the best candidate for that posi-
tion, because the metropolitanate of Tomis was revived exactly at that
165 Jordanov 2003 a, 124, 125, nr. 48. 1, 3.

166 Seibt 2008, 103–108.
167 Bolşacov-Ghimpu 1973, 559.
168 Beševliev 1981, 190–206; Gagova 1986, 70–72; Dimitrov 1992 b, 43–45.
169 Stephenson 2000, 56; Krsmanović 2010, 608.
the same time.170 The primary goal of the katepanate of Western Meso-
potamia was the defense against future Rus’ maritime attacks. Therefore,
the first structure created in 971 by John Tzimiskes in the Danube region
consisted of the katepanate of Western Mesopotamia (centered perhaps
at Noviodunum), the katepanate of Dristra (later turned into a strategy),
and the strategy of Ioannoupolis-Preslav (see Fig. 2).
After 975, the area dominated by the Byzantine Empire along the Dan-
ube extended farther to the west, as a strategy of Morava was established,
which is documented by the lead seal of Adralestos Diogenes, imperial
protospatharios and strategos of Morava. Morava, in this case, is the town
Moravon (modern Dubravica) located at the mouth of river by that same
name. Because this man was active during the reign of John Tzimiskes, it
could be inferred that he was appointed shortly after 976. The strategy of
Morava is not mentioned in the Taktikon Scorialensis, written before 975.171
It is not excluded that the advance toward the Morava was made from the
south, namely from Dyrrachion—and important position of the empire
in the western Balkans until 997. One indication of a southern direc-
tion is the ephemeral existence of a katepanate of Ras (near Novi Pazar),
which is documented by the seal of John, protospatharios and katepano
of Ras, dated to that same period. According to a tradition preserved in
the so-called Chronicle of the Priest of Duklja, the Byzantine army occupied
Raška shortly after the conquest of Bulgaria, but the local župan regained
his independence after Tzimiskes’ death.172 Because the authenticity of
that source was recently contested with strong arguments,173 a Byzantine
conquest of Ras in the aftermath of the conquest of Bulgaria remains to
be proven. If the information in the chronicle is to be trusted at all, the
katepanate of Ras must have existed for a short while after 971. The strat-
egy of Morava belonged to that katepanate which had no common border
with the theme of Dristra.
170 Madgearu 1999–2000, 16; Madgearu 2001, 75–76. Tăpkova-Zaimova 1973, 96 supposed

that the residence of the strategos was at Păcuiul lui Soare, but this interpretation could
not be admitted if the katepanate included only the northern part of present Romanian
Dobrudja. Kostova 2008 a, 217 considers that the residence of the strategy was either at
Constanţa, either at Presthlavitza.
171 Nesbitt, Oikonomides 1991, 195–196; Pirivatrić 1997, 173–201; Oikonomides 1998, 589;
Krsmanović 2008, 135, 141. For the archaeological discoveries see chapter II. 3.
172 Ferluga 1980, 437–438; Nesbitt, Oikonomides 1991, 100–101; Krsmanović 2008, 82, 135,
141, 143.
173 Bujan 2008, 5–38.
44
chapter one
2. The military organization between 971 and 986

Soon after the conquest of 971, another strategy was established at

Anchialos (the lead seal of its strategos Anastasios may be dated in the
last decades of the 10th century).174 In Mesembria, the old kleisura was
revived, which had been created after 864, and then lost to Bulgaria in 917.
The analysis of the iconography of the seals of two kleisurarchoi, Alexios
and Pankratios, discovered at Preslav, indicates a date in the last decades
in the 10th century. A same dating may be accepted for could the seal of
the strategos Nikolaos of Mesembria.175 The strategoi of Anchialos and the
kleisurarchoi of Mesembria were perhaps under the orders of the duke
of Adrianople. Much like in the mid-9th century, the Rus’ threat was the
primary reason for those new developments in the defense on the western
coast of the Black Sea. The strategic purpose of the new Danube theme
was to prevent or to deter future Rus’ attacks. The same mission had
Crimea, where a new office, the strategos of the Cimmerian Bosphorus
(Kerč Straits) was established soon after 971. It appears that that port city
had been conquered by Svyatoslav from the Khazars in 965, and then by
Byzantines after the victory of 971.176 The need for a better defense against
the Rus’ led to the creation in the same period of yet another office, the
strategos of Pontos Euxinos.177
A major change took place in Byzantine-Rus’ political and military rela-
tions during the reign of Basil II (976–1025). The emperor sought military
support from the new prince Vladimir (980–1015) during the civil war he
waged against the general Bardas Phokas. After his baptism, Vladimir mar-
ried Anna, Basil II’s sister. The Rus’ ruler sent in the spring of 988 6,000
elite Varangian warriors who had a decisive contribution in the victori-
ous battles of Chrysopolis and Abydos. In 989, the Rus’ attacked Cherson,
as that city was under the control of a commander allied with the rebel
Bardas Phokas. They came down on the Don River and then by the Black
Sea, and laid a siege of the city which lasted several months. Encircled
by the Rus’ army and betrayed from the inside, Cherson surrendered to
the Rus’ in 990, but Vladimir gave it back to his brother-in-law as a mar-
riage gift.178 Thus the Byzantine-Rus’ relations entered a relatively long
174 Nesbitt, Oikonomides 1991, 168–169, nr. 73.1.

175 Oikonomides 2002, 7. The seals are published by Jordanov 2003 a, 119–120, 47.1, 47.4;
Jordanov 2009, 454–455, nr. 1340, 1343–1345.
176 Oikonomides 1972, 269, 363.
177 Oikonomides 1972, 266/267, 358; Kostova 2006, 582. The seal of Michael, strategos of
Pontos Euxinos dated to the mid-11th century, in Jordanov 2009, 425, nr. 1207.
178 Poppe 1976, 195–244; Franklin, Shepard 1996, 162–163; Holmes 2005, 460, 510–511;
Krsmanović 2008, 51.
46 chapter one
period of peace, interrupted in 1043 by a renewed Rus’ attack (discussed

in chapter III. 2).
For the moment, however, the danger from the north has been removed,
and even the Pechenegs were at peace. The situation on the Danube fron-
tier was again destabilized, however, by the revived Bulgarian state. The
four sons (comitopuloi) of the Bulgarian count Nikola from Macedonia—
Samuel, Aaron, David and Moses—started in 976 a rebellion against the
Byzantines. Some historians date the revolt of the Comitopouloi at an
earlier date (969) and regard it as a centrifugal movement against the
Bulgarian emperor Boris II, in the circumstances surrounding the Rus’-
Byzantine war. True, both Skylitzes and Zonaras placed the apostasias of
the four brothers at the beginning of Boris II’s reign. The same scholars
also maintain that the presence of Bulgarian envoys at the court of Otto I
(936–973) in Quedlinburg in the spring of 973 suggests that some kind of
Bulgarian state had survived the Byzantine onslaught, and that state must
have by then been restricted to Macedonia.179 However, it has been noted
that only the German emperor’s vassals were present in Quedlinburg,
which would imply a more westerly location for the lands from which
the Bulgarian envoys had come, perhaps around Sirmium, a city which
was conquered by the Byzantine army only at the end of the conflict with
Bulgaria, in 1018.180 Indeed, Bulgarian Macedonia was not occupied before
that by the Byzantine forces, and the collapse of the Bulgarian state in
968–969 must have offered a good opportunity for centrifugal actions in
that remote area, against the central power of Preslav.181 The problem
remains open.
Samuel became the leader of the resistance against the Byzantines. He
established a state, the military power of which was based on that part of
Macedonia which remained un-occupied by John Tzimiskes. From that
region, he launched a series of attacks on the Byzantines, taking advantage
of the civil war between Emperor Basil II and the general Bardas Skleros
(976–979). Between 976 and 986, the Bulgarians in Macedonia obtained
several major victories against the Byzantine army, the most important
being that of Serdica (Sofia), on August 17th, 986. Samuel became emperor
179 Skylitzes, Basil II and Constantine, chapter 13, 5; chapter 16, 11 (ed. Thurn, 255–256,
328; transl. Flusin, 216, 275; transl. Wortley, 246, 312); Zonaras, XVI, 23.32–34; XVII, 6. 2–5 (ed.
Büttner-Wobst, 495, 547); Stokes 1962, 483; Antoljak 1972, 379–384; Döpmann 1983, 47, 50.
Contra: Anastasijević 1930, 20–36; Stănescu 1968 a, 409–412; Ferluga 1976, 345–354.
180 Ljubinković 1973, 949–950; Popović 1978, 36–39.
181 Fine 1991, 188–189; Nikolov 2001, 142.
of Bulgaria in 997 because Romanos, the brother of the former emperor

Boris II was ineligible, since he had been castrated.182 It was in fact a
new state developed around Ochrid, which now became the see of the
Bulgarian Patriarch. It is worth insisting that this was a Bulgarian, not a
Macedonian state, as some historians from the Republic of Macedonia
wrongly assume.183
Over the first ten years, Samuel fought the Byzantines mostly in the
central, western and southern parts of Bulgaria. The Danube region does
not seem to have been a matter of concern for him at that time, because,
being a good strategist, he realized that victory depended on the con-
centration of forces in naturally defended regions. In his work of 982,
an anonymous Persian geographer made a distinction between Burjan,
a province with a district (shahr) called Thrace, paying land-taxes to the
king of Rum, and Bulghari, a people living on the mountains to the north-
west of the empire, perpetually at war with the Byzantines.184 The latter
were most likely Samuel’s subjects, as Samuel had not yet taken control
of the lowlands. He did launch attacks to the north in the direction of the
Danube region, but only after his power was firmly settled in Macedonia
and in the mountainous part of Bulgaria. In the course of several offen-
sives, Samuel managed to conquer a large part of the Byzantine territories
in the northern and central Balkans, taking advantage of the internal con-
flict within the empire, as well as of the Byzantine wars with the Arabs.
The expansion of Samuel’s state towards the Danube began in 986, after
the conquest of Serdica and the subsequent battle of the Trojan Gorge,
where the Byzantine army led by general Stephen Kontostephanos, the
domestikos of the western scholai (commander of the army in the Balkan
Peninsula) suffered a serious defeat on August 17th, 986.185 Because there
no seal among those found in the administrative building in Preslav (see
next chapter) could be dated between 986 and 1000, it has been suggested
that the Byzantine rule over that city was interrupted during that period.
This means that Preslav was conquered by Samuel in 986, most probably
182 Fine 1991, 192–195; Stephenson 2000, 59–63.

183 For the propaganda value of this interpretation linked directly to the historical legiti-
mization of the present state, a successor of former Yugoslavia, see Fine 1991, 191; Madgearu
2008, 181–185.
184 Hudud al-’Alam, 42 (ed. Minorsky, 157–158).
185 Skylitzes, Basil II and Constantine, 12 (ed. Thurn, 330–331; transl. Flusin, 276–277;
transl. Wortley, 313–314); Leo the Deacon, X. 8 (ed. Talbot, Sullivan, 213–215); Tăpkova-
Zaimova 1973, 100; Fine 1991, 192; Holmes 2005, 490–492; Strässle 2006, 274; Yotov 2008 a,
350; Krsmanović 2008, 50, 147.
48 chapter one
soon after the victory in the Trojan mountain pass. During this campaign
two messages from Stephen Kontostephanos were sent at Preslav. Kon-
tostephanos’ predecessor for the years 976–985, Leo Melissenos, was also
in correspondence with the strategy of Preslav.186 It has been suggested
that Preslav remained under Byzantine control even after 986 because
there were too many strategoi of Ioannoupolis for the years 971–986,187
but nothing substantiates that suggestion. There is no reason to question
the possibility of five strategoi in succession over a 16-year long period
(for which see chapter II). Although he conquered the old capital, Samuel
continued to rule from Ochrid, which had a a better defense.
The Byzantine push into the Middle Danube region through the cre-
ation of a strategy of Morava must have caused a rapprochement between
the Hungarian duke Geza and Samuel, which took the form of a dynastic
alliance. In or around 995, Geza offered one of his daughters in marriage
to Gabriel, Samuel’s son and heir apparent (according to Skylitzes, Gabri-
el’s mother was a woman from Larissa). The marriage with the Hungar-
ian princess was however abandoned around 997, when Gabriel took as
wife a woman from Larissa, named Irene.188 At that moment, the strategy
of Morava had disappeared, most probably because of Samuel’s northern
offensive of 986. In that same year, some troops were transferred from
the Danube region to other provinces, which were under greater threats.
At Dinogetia, the barracks of the soldiers in the garrison installed there
soon after 971 were abandoned after a very short time, being replaced by
civilian dwellings.189 Northern Dobrudja (the katepanate of Western Mes-
opotamia) remained under Byzantine control after 986. A seal found in
Isaccea shows that the port was in operation after 986. The seal belonged
to David Kouropalates,190 the Armenian prince of Tayk who received the
186 Jordanov, Tăpkova-Zaimova 1988, 121; Diaconu 1989, 11–14; Tăpkova-Zaimova 1993,

97; Jordanov 1993, 88–90, nr. 159–162; Jordanov 2006, 287–289, nr. 459–461; Strässle 2006,
406–407; Jordanov, Žekova 2007, 103–104, nr. 266, 267; Krsmanović 2008, 57–59; Jordanov
2009, 361–362, nr. 992–996. Leo Melissenos is not mentioned in the list of Kühn 1991, 153.
187 Frankopan 2001, 87–88.
ley, 332; Prokić 1906, 31, 36, 43, 52; Moravcsik 1970, 62; Fine 1991, 195–196; Makk 1994 a, 27;
Strässle 2006, 158, 333; Révész 2009, 83. Gabriel was also called Radomir or Romanos. Accord-
ing to Iorga 1937, 10 and Risos 1990, 206–207 (accepted by Strässle 2006), the name Romanos
indicates that his mother from Larissa was of Vlach origin. This, in turn, would supposedly
be the first attestation of the ethnic name Romanians gave to themselves.
189 Diaconu 1969 a, 48–49; Tăpkova-Zaimova 1973, 101.
190 Barnea 1993 a, 56–57.
title of kouropalates in 990 and who died in 1000.191 Since he was a high
official in the Byzantine army, his message was obviously addressed to a
local commander. A coin hoard of 49 nomismata in which the latest coins
have been struck between 977 and 989 was buried in Greci, 17 km south of
Garvăn, most likely in the circumstances of the war against Samuel.192
The Stone Dike became again useful for Bulgaria as the northern part of
Dobrudja remained under Byzantine control. The defense required repa-
rations and a military presence along that frontier. It appears that that the
mission entrusted to župan Demetrius at some point between 986 and
1000, as attested by the inscription from Mircea Vodă. The situation in
Dristra after 986 is far from clear. Although the Stone Dike most certainly
was the frontier of Bulgaria after 986, some historians believe that Dristra
remained under Byzantine control, because the historical narrative about
the offensive of 1000 makes no mention of that city.193 The current state
of research does not allow any conclusive answer to this question, but it
is more likely that Dristra was taken over by the Bulgarians in 986 or soon
after that. The city was too important to be left aside after the victorious
campaign of Samuel. A proof of the Bulgarian rule over the city is the exis-
tence of a workshop producing cast Bulgarian imitatations of Byzantine
coins, namely folles of types A1 and A2.194
An older generation of historians brought the period of Bulgarian revival
in association with a controversial source. In 1819, the French Byzantinist
of German origin Charles B. Hase published three notes in the appendix of
his edition of the Leo the Deacon’s Chronicle, which dealt with the life of
a Byzantine commander of a fort called Klimata, supposedly threatened
by un-named northern barbarians. The commander is said to have been
compelled to seek assistance against those enemies from a powerful ruler
called basileos, who lived somewhere in the north. Following Hase, histo-
rians called this text “The Note of the Gothic Toparch”, “The Note of the
Greek Toparch”, or “Hase’s Anonymous.” Nobody had seen the manuscript
after which Hase published his edition. The very confusing information in
191 Grousset 1947, 506, 512, 530; Grünbart 1998, 29.

192 Barnea, Ştefănescu 1971, 23; Mănucu-Adameşteanu 1996, 281; Poenaru-Bordea,
Ocheşeanu, Popeea 2004, 130–134; Mănucu-Adameşteanu 2010 d, 444; Custurea, Talmaţchi
2011, 211–212.
193 Bănescu 1946, 47–48; Diaconu 1969 b, 396; Salamon 1971, 494–496; Stănescu 1974, 401;
Mănucu-Adameşteanu 2001 a, 97; Bica 2003, 59; Mărculeţ 2005–2006, 311; Krsmanović 2008,
148, 195.
194 Oberländer-Târnoveanu 1983, 262; Mănucu-Adameşteanu, Poll 2006, 446–447; Yotov
2008 a, 351.
50 chapter one
the text encouraged speculations and many offered different interpreta-

tions. While to some, the region ruled by the toparch must have been
in Crimea, others believed the events in question to have taken place in
Dobrudja, either in 992, or in 1051, the latter year being established on
the basis of some astronomical observation mentioned in the text.195 Still
others went as far as to look for the residence of the toparch in Capidava,196
Lykostomion,197 or Vicina.198
All discussion on this matter abruptly ended when Ihor Ševčenko dem-
onstrated that the three fragments published by Hase were his own com-
position, in other words a forgery.199 Hase had been paid by the Russian
chancellor Nikolai Rumiantsev to find and edit ancient sources concern-
ing the old history of Russia. Growing greedy, Hase made up a text, which
could be extremely useful in the context of early nineteenth-century con-
cerns with legitimizing the Russian occupation of Crimea and the Lower
Danube region (this was in fact just a few years after the occupation of
the Romanian region of Bessarabia in 1812). Indeed, most historians writ-
ing before Ševčenko’s revelations believed the northern protector of the
toparch to have been some Rus’ ruler (perhaps Vladimir). We cannot pres-
ent here the entire demonstration made in a 70 pages study, but we still
mention some of the most significant facts. The most important clue is
the name Klimata, which in the text appears to be the name of a city. Dur-
ing Hase’s lifetime, it was commonly believed that that name, as found
in Constantine Porphyrogenitus’ work, applied to a fortress (and not to a
theme or a region), which Hase himself further identified with Balaklava.
As we have seen, Klimata was not a city, but a small theme in Crimea,
which existed only for a short time between 841 and ca. 870.200 More-
over, the name Maurokastron, which is mentioned in the text, could not
have been in existence in the 10th century. Hase was apparently not aware
that ancient Tyras (Cetatea Albă, now Bilhorod Dnistrovs’kyi in Ukraine)
was not a city in the 10th century any more, and that the twin towns of
Maurokastron and Asprokastron appeared only during the second half of
195 Diaconu 1962; Cihodaru 1963; Diaconu 1965 b; Cihodaru 1965; Diaconu 1968, 357–369;
Bolşacov-Ghimpu 1972, 104–116; Mănucu-Adameşteanu 2001 a, 42–43; Damian 2004, 299;
Mărculeţ 2006, 304, 307–312.
196 Diaconu 1962, 1228.
197 Mărculeţ 2005, 35–36, 41–42; Mărculeţ 2005–2006, 310.
198 Bolşacov-Ghimpu 1972, 111.
199 Ševčenko 1971, 115–188.
200 See footnote 38 and Sokolova 1993, 99.
the 13th century.201 Contrary to what historians commonly believed dur-

ing Hase’s lifetime, there was no other Maurokastron in Crimea in the
10th century. Even more significant is the mismatch between the use in
the text of the expression “Hellenic way of life” and the real significance
of the word “Hellenes” for the Byzantines: before the 13th century, “Hel-
lenes” referred to “heathens” or “followers of the ancient philosophy.” A
tenth-century author could not have possibly used the adjective “Hellenic”
in an ethnic sense (“Greek”). Finally, no Greek author would have ever
called the Rus’ prince basileos. Ševčenko’s study irritated many,202 but no
serious arguments have since been put forward against his claim that the
text was Hase’s forgery, and no explanations have been so far provided for
the anachronisms to which he pointed. The so-called “Note of the Greek
Toparch” is therefore of no real use for the subject of this book.203
The survival of the Byzantine rule in the northern Dobrudja after
986 enabled the future recovery of the entire Danube region, as soon as
circumstances turned favorable. Basil II’s victories against the Arabs in
995–999 were followed by his alliance with the Fatimid Caliph al-Hakim
(1001) and the transfer of the army to Europe, in order to continue the
war against the Bulgarians. The first major Byzantine victory against Sam-
uel was in 997 on the river Spercheios, when the duke of Thessaloniki
Nikephoros Ouranos ambushed the Bulgarian army returning from a raid
deep into Greece. The offensive of 1000 in eastern Bulgaria was led by the
generals Theodorokanos (duke of Adrianople) and Nikephoros Xiphias
(strategos of Philippopolis). The cities of Pliska, Preslav, and Little Preslav
were quickly taken by this army moving from the theme of Thrace.204 The
territory previously occupied by John Tzimiskes in 971 was thus recov-
ered, but the theme of Ioannoupolis was not recreated, as its territory
was now included into the theme of Dristra (the name and evolution of
this theme will be discussed in the next chapter), while the katepanate of
Western Mesopotamia continued to exist for a while, as Damian Dobro-
mir is known to have beeen its commander until after 1002 (he entered in
the Byzantine service in 1001) (see Fig. 3).
201 For the history of this medieval city see now Rădvan 2010, 473–484.
202 Božilov 1978, 245–259; Diaconu 1981 b, 1111–1133.
203 For the forgery see now Medvedev 2000.
204 Skylitzes, Basil II and Constantine, 26 (ed. Thurn, 343–344; transl. Flusin, 287; transl.
Wortley, 326); Zonaras, XVII, 8.3–10 (ed. Büttner-Wobst, 559); Fine 1991, 197; Whittow 1996,
381; Haldon 2001, 106–107; Stephenson 2003, 111–112; Holmes 2005, 410, 413–414, 495; Yotov
2008 a, 349; Krsmanović 2008, 53, 149, 161. For the seals of Nikephoros Ouranos found in
Preslav and Silistra, see Jordanov 2009, 362–363, nr. 997–999.
52
chapter one
3. The Danubian region in 1000–1004

A new campaign led by Basil II in 1002 targeted the western part of the
Danube region, the area around Vidin, an important city conquered after
a difficult siege, which lasted from March to December. The navy made
again use of the “Greek Fire,” much like at Dristra in 971, but the besieged
were able to extinguish it. The strength of the precinct wall made the con-
quest difficult, and it seems that only the depletion of resources available
to people in Vidin made possible Basil II’s victory.205 As Nicolae Iorga has
suggested, Vidin was perhaps the center of an autonomous polity within
Samuel’s state.206 That city and its hinterland were now to the theme of
Dristra. The empire has become a neighbor of Hungary.
Two coalitions emerged in the Danube region in 1002. The aggressors
were Byzantium and Hungary, both powers aiming to take control over
the area. On the defending side there were Bulgaria, and a duchy that
had just rebelled against the Hungarian king. The information about that
duchy in the region now known as the Banat is included in the Legenda
Major Sancti Gerardi,207 the vita of the first bishop of Morisena-Cenad.
Morisena was the residence of a duke called Ahtum, who rose against the
King Stephen I of Hungary. Two dates have been advanced for the rebel-
lion and the subsequent war: 1003–1004, and sometimes between 1028 and
1034. The second interpretation208 supports the later chronology because
Ahtum was allied with the “Greeks”, as it is written in the source. In other
words, the war against Ahtum was waged at a moment of decline for the
Byzantine power in the region, namely after Basil II’s death in 1025. The
relation recorded in the tenth chapter in the surviving text of the Leg-
enda Major Sancti Gerardi contains several confusions and anachronisms,
which make difficult any attempt to establish a clear chronology and the
context of the events. For instance, Ahtum is said to have owned the for-
tresses of Severin and Vidin. However, this is more likely the situation of
the 13th and 14th centuries, when both fortresses belonged to Hungary,
and not that of the early 11th century. No fort existed in (Turnu) Severin
during the latter period. On the other hand, because the Byzantine army
had conquered Vidin in 1002, it is unlikely that Ahtum possessed it after
that. The most important objection concerns the international relations.
ley, 328); Zonaras, XVII, 8.13–14 (ed. Büttner-Wobst, 560); Kuzev 1968, 37–44, 55; Kuzev,
Gjuzelev 1981, 103–105; Stephenson 2000, 65; Nikolov 2001, 151; Holmes 2005, 496; Strässle
2006, 168, 275.
206 Iorga 1937, 14.
207 SRH, II, 480–560.
208 Bálint 1991, 116–117.
54 chapter one
What reasons would the Byzantines have to support an enemy of their

ally, King Stephen I of Hungary, who had fought against Bulgaria together
with Basil II in 1002, as well as later?209 The real enemy of the Byzantines
in the Danube region after 1017 were the Pechenegs, not the Hungarians
(see below), and the Pechenegs were also the traditional enemies of the
Hungarians. Until at least 1028, Byzantium had no interest to support an
enemy of Hungary. The same alliance is unlikely for 1002, when Stephen I
is known to have provided military assistance to Basil II. No military con-
flict is mentioned between Hungary and the Byzantine Empire during the
period between 1002 and 1038 (the latter year being the latest date so far
advanced for Ahtum’s rebellion and his war with Stephen). The only solu-
tion is that before 1003, Ahtum was in fact Samuel’s ally. Some have there-
fore maintained that the “Greeks” mentioned in the text were Bulgarians,
whose name had been replaced since in the late 11th century (when the
text was written) Bulgaria was a Byzantine territory.210 The name “Greeks”
is therefore interpreted as a generic designation for eastern monks, regard-
less their ethnicity. It follows that the date of the war between Ahtum and
Stephen I should be placed around 1002.
Ahtum is said to have been baptized in Vidin before the Byzantine
conquest of that town, at a time when he was allied with Samuel and
Vidin was under the jurisdiction of the Bulgarian patriarchate in Ochrid.
Samuel’s 997 campaign into Dalmatia and present-day Serbia was the
opportunity for a rebellion against the Hungarian duke Vajk (the future
king Stephen I). In that year, Gabriel abandoned his Hungarian wife,
which brought Bulgaria in conflict with the Hungarians. In 1002, Samuel
was therefore in conflict not only with Basil II, but also with Stephen I.
This strongly suggests that Ahtum was Samuel’s ally.211 Stephen I’s goal in
attacking Ahtum was to take over the control over the salt trade along the
Mureş valley. In brief, the conflict between Stephen I and Ahtum must be
regarded as a side development of the larger confrontation between Basil
II and Stephen I, on one hand, and Samuel, on the other hand, for the
domination of the Danube region between Vidin and Braničevo.
According to some recent interpretations, Basil II entered a peace
treaty with Samuel in 1005, whereby he recognized the Bulgarian state,
209 For the participation of Stephen I, see Györffy 1964, 149–154; Kosztolnyik 2002, 33.
210 Fehér 1921, 152–155; Györffy 1964, 149.
211 The same opinion in Strässle 2006, 155, 333. Makk 1994 a, 27–29 accepted that Samuel
was in conflict with Stephen I, but thought that Ahtum was his vassal, fighthing against
Samuel.
but only in Macedonia. At that moment, the emperor’s intention was not
to eliminate Bulgaria, for he was satisfied with the recovery of the region
between the Lower Danube and the Stara Planina, as well as of the city of
Dyrrachion on the Adriatic coast, a position of considerable strategic and
economic importance. The future of Bulgaria was that of a client state.212
However, in 1014 Basil II changed his mind and decided to occupy all of
Bulgaria. He launched a campaign aimed at the conquest of Ochrid, the
Bulgarian capital. The decisive victory took place in the Kimbalongon
mountain pass on the Strymon River, on July 20th, 1014. The Hungarian
king Stephen I took part in that campaign as Basil’s ally.213 The follow-
ing Bulgarian emperors Gabriel (1014–1015) and John Vladislav (1015–1018)
continued the fight, but without much success. The last counteroffensive
was organized in 1017 by John Vladislav in cooperation with Krakras, a
local ruler from Pernik. The campaign ended in failure when Tzotzikios,
the strategos of the theme of Dristra, learned about their plans. The Bul-
garians tried to recruit the Pechenegs on their side, but the nomads did
not budge.214 This suggests that in 1017 the Pechenegs or at least some of
them, were again hostile to the Byzantine Empire, and that the army of
the theme of Dristra was able to prevent their invasion.
The remaining part of Macedonia, including Ochrid, the last capital of
Bulgaria, was occupied in 1018. The theme of Bulgaria was organized in the
same year and included the central region of the Balkans conquered from
Bulgaria after 1014, but without the region by the Danube, which remained
attached to the province of Dristra. The residence of the duke of Bulgaria
was in Skopion (Skopje), and the first to occupy that position was David
Arianites, former duke of Thessaloniki.215 This book is not dealing with
212 Stephenson 2000, 69–70; Stephenson 2003, 117–120.

transl. Wortley, 331–332); Zonaras, XVII, 9.2–10 (ed. Büttner-Wobst, 563–564); Györffy 1964,
151; Stănescu 1968 a, 413; Fine 1991, 198; Makk 1994 a, 30; Whittow 1996, 387–388; Stephen-
son 2000, 71–74; Stephenson 2003, 117–119; Strässle 2006, 333. Kimbalongon is the Grecized
version of the Vlach name Câmpulung (“long field”). According to Strässle 2006, 182–184,
Kimbalongon is the Rupel Pass.
214 Skylitzes, Basil II and Constantine, 36–38, 40–41 (ed. Thurn, 350–359; transl. Flusin,
293–299; transl. Wortley, 332–339); Zonaras, XVII, 8.10–26 (ed. Büttner-Wobst, 564–567);
Bănescu 1946, 70; Barnea, Ştefănescu 1971, 93; Jordanov 1995, 210; Holmes 2005, 415, 418;
Strässle 2006, 334, 410; Božilov 2008, 95. Skylitzes gives the name Tzitzikios, but the real
name was Tzotzikios, identified with Pherses Tzotzikios, anthypatos, patrikios and strategos
of Cappadocia (McGeer, Nesbitt, Oikonomides 2001, 120–121, nr. 43.13) or, better, with his
nephew (Stepanenko 2009, 237, 245).
ley, 339); Bănescu 1946, 118–120; Stephenson 2000, 72–74; Holmes 2005, 419–422; Borisov
2007, 71; Krsmanović 2008, 55–56; Jordanov 2011 c, 179–180.
56 chapter one
the history of that theme, except in relation to the changes in its organiza-
tion taking place after 1071, when a part of the Middle Danube region was
attached to the theme of Bulgaria.
The region along the Danube to the west from Vidin made up a dif-
ferent province, created after 1019, when the Byzantine army moved
into the north-western parts of Bulgaria, all the way to Sirmium (Srem-
ska Mitrovica), a city said to have been under a Bulgarian “strategos.”
The name Sermon given for this commander by Skylitzes appears to be
a confusion with the name of the city.216 Gold imitations of Byzantine
nomismata, struck by a certain Sermon stratilates (the latter word being
the Greek term for “general”) and said to have been found somewhere in
the Danube region have been attributed to that Bulgarian commander.217
They are nothing else but forgeries dated ca. 1870.218 Following some strat-
agem, the commander of Sirmium was killed by the general Constantine
Diogenes. The theme of Sirmium was thus created in the north-western
territory, perhaps on the basis of the former theme of Morava. The new
province included the cities of Belgrade and Braničevo, which, like Sir-
mium, had been part of Bulgaria since the early 9th century. The first
commander of the theme of Sirmium was duke Constantine Diogenes,
who later moved to the theme of Bulgaria.219 The theme of Serbia, which
is sometimes mistaken for the theme of Sirmium,220 was actually in what
is now Montenegro. That theme was administrated by local Serbian rul-
ers, like Ljudevit, who is mentioned in 1039 with the title of strategos of
Zachlumia and Serbia.221
It appears that the theme of Sirmium had a bridgehead on the left bank
of the Danube, at Kuvin, the purpose of which was the defense of the
transl. Wortley, 345); Bănescu 1946, 26–27, 120–123, 135 Stephenson 2000, 66; Nikolov 2001,
149–150.
217 Schlumberger 1925, II, 417–418, 420 (the image of the coin); Iorga 1937, 9.
218 Metcalf 1979, 54; Oberländer-Târnoveanu 2005, 188, 190; Jordanov 2011 d, 135–139. The
forgeries may be attributed to a Bulgarian or a Serbian person, someone well acquainted
with the medieval history of the region, for Sermon appears to have been taken as a symbol
of the unity of the southern Slavs (Bulgarians and Serbs), during the period just before the
liberation of Bulgaria from the Ottomans (1878) and while Serbia existed as an autonomous
kingdom supported by Russia.
219 Wasilewski 1964, 473–474; Kühn 1991, 233–235; Bálint 1991, 104; Stephenson 2000, 66,
74, 124; Holmes 2005, 423, 425; Strässle 2006, 406; Krsmanović 2008, 198–199; Krsmanović
2009, 76.
220 Laurent 1957, 185–195; Wasilewski 1964, 465–472.
221 Maksimović 1996, 54–56; Stephenson 2000, 66, 74, 123, 126–129; Stephenson 2003,
122–124.
ford at Braničevo. I have demonstrated elsewhere that Kuvin is the same

as Dibiskos mentioned in 1019 and 1020 as a parish of the bishopric of
Braničevo.222 The existence of a parish integrated into a Byzantine eccle-
siastical structure implies that the Byzantine administration extended on
the northern bank of the Danube, the strategic value of which would be
obvious in the subsequent conflicts with Hungary. There is, however, no
proof that the theme of Sirmium extended any farther into the Banat.
In 1019, therefore, two Byzantine provinces existed along the Danube—
Dristra and Sirmium (see Fig. 4). The first one was extended along the
lower course of the river, all the way to its mouths. The boundary between
the themes of Dristra and Sirmium must have been somewhere in the
Iron Gates region around Vidin. It is not known for how long after 1002
did the katepanate of Western Mesopotamia continue to exist in northern
Dobrudja, but the absence of later testimonies suggests that it was soon
after that incorporated into the theme of Dristra.
222 Madgearu 2001, 80–84.

58
chapter one
4. The Danubian region in 1018/1019

Chapter Two
The Military Organization of the Danube Region
1. The Theme of Dristra (later, Paradunavon)
More than 700 lead seals dated from 971 to 986, 1000 to 1050, and 1060 to
1088 have been found in a single building in Preslav. Most of them were
attached to messages dispatched to military commanders and officials in
that town, while others accompanied copies of letters sent from Preslav,
which were preserved in the archive stored in that building. The number
of lead seals from Preslav is so large, that it may be used to clarify the
prosopography and administrative evolution of the territories conquered
by the Byzantine Empire in Bulgaria. The Preslav archive includes lead
seals of the following strategoi of Ioannoupolis that could be dated from
971 to 986, the latter being the date on which the city was conquered by
the revived Bulgarian state of Samuel:
– Leo Sarakinopoulos (Ioannoupolis and Dorostolon);

– Theophanes (Thrace and Ioannoupolis);
– Staurakios (Thrace and Ioannoupolis);
– Peter (Ioannoupolis);
– Katakalon (Ioannoupolis).1
All of them had the rank of basilikos protospatharios. Besides the chief
commanders of the theme, we know the name of an officer, Adrian, who
also had the rank of basilikos protospatharios, but served as a simple tur-
march of Preslav. His seal is dated after the moment the city reverted to
its old name.2 A deputy of the commander (ek prosopou) of the theme
1 Jordanov 1982 b, 35–39; Jordanov 1987 a, 89–92; Jordanov 1993, 19, 128–137; Oikonomides
1998, 583–584; Stephenson 2000, 56; Frankopan 2001, 75–97; Jordanov 2003 a, 100–102, nr. 35
B. 15–18, 105, nr. 38. 1, 106, nr. 38. 2, A. 3; Božilov 2008, 93, 95; Yotov 2008 a, 348; Krsmanović
2008, 138; Jordanov 2009, 437–438, nr. 1261–1266, 443–444, nr. 1283–1290; Jordanov 2011 a, 81,
nr. 4–23; Jordanov 2011 b, 201. Stoimenov 1996, 84 believes that the residence of strategos
Katakalon was Helioupolis (Emesa in Syria), which received its name in 975.
2 Nesbitt, Oikonomides 1991, 153–154, nr. 69.2.
60 chapter two
Thrace and Ioannoupolis, named Adralestos Spanopoulos, is known from

a seal from Preslav.3
Nikephoros Xiphias, who appears on two Preslav seals, was the strat-
egos of Thrace and Ioannoupolis, but not before 986, since he must have
been too young at that time. As Valery Yotov has noted, his title of strat-
egos title should be linked with the war of 1000,4 but unlike Yotov, I see
Nikephoros Xiphias only as the commander of the theme of Thrace, who
moved with his army to Ioannoupolis. The bone of contention here is the
date at which Ioannoupolis got back its initial name, Preslav.
The first strategos of Dristra appointed after katepano Sisinios was Leo
Sarakinopoulos. He appears on many of his seals with the rank of impe-
rial protospatharios and the title of strategos of Dorostolon, but most such
seals are from an earlier period of his career. He succeeded Nikephoros
Xiphias as strategos of Thrace, Ioannoupolis and Dorostolon, most prob-
ably in 972. After a while, his authority appears to have been limited to
Ioannoupolis and Dorostolon alone. The change reflected the departure
to Thrace of the troops displaced from that theme. Their commander
remained in Ioannoupolis.5 Leo Sarakinopoulos was sometimes mistaken
for another Leo. Vitalien Laurent has demonstrated in 1967 that the initial
reading of a fragmentary seal found near Călăraşi,6 Λέοντ[ι] στρατ(ηγῷ)
[Ἰ]στριην(ῷ), was wrong. Laurent’s alternative reading, Λέοντ[ι] (βασιλικῷ)
στράτορι καὶ τριβυνῷ, has meanwhile accepted as much better, even if
some still persist in the old error.7 The seal is nevertheless important for
the military history of the area around Dristra because, being dated to the
10th century, it could be related to John Tzimiskes’ campaign. The strator
(shield-bearer) was a member of the imperial escort, a kind of bodyguard,
3 Jordanov 1993, 144, nr. 290; Jordanov 2003 a, 102, nr. 35 B 19; Jordanov 2009, 439,
nr. 1269.
4 Jordanov 1993, 131–132, nr. 250; Jordanov 2003 a, 101, nr. 35B18; Jordanov 2006, 310,
nr. 526; Yotov 2008 a, 351–352; Jordanov 2009, 438–439, nr. 1267–1268.
5 Jordanov 1982 a, 12–23; Diaconu 1986, 173–177; Jordanov 1987 a, 91; Jordanov 1993, 136–
137, nr. 259–277, 232, nr. 271 a; Frankopan 2001, 88; Jordanov 2002, 82, nr. 3, 85, nr. 12; Jordanov
2006, 360–362, nr. 604–630; Yotov 2008 a, 346–347; Božilov 2008, 93; Jordanov 2009, 414–415,
nr. 1164–1165, 436–437, nr. 1250–1260, 445, nr. 12929–1310; Jordanov 2011 a, 81, nr. 2–23.
6 Mititelu, Barnea 1966, 46–48; Barnea, Ştefănescu 1971, 75.
7 Tăpkova-Zaimova 1973, 91–92; Jordanov 2009, 259, nr. 709. For persistence in the old
error, see Mărculeţ 2005, 27–28; Mărculeţ 2005–2006, 306; Mărculeţ 2008, 177–181; Mărculeţ
2010 a, 203–207. Moreover, Mărculeţ reproduces Ion Barnea’s wrong idea about the owner
of the seal being an “Istrian strategos” and identifies him with the strategos Leo Chalkotubes
mentioned in 1036 (for whom, see below).
the military organization of the danube region 61
and the presence of such a person in the area is obviously linked to that
of the emperor.
The list of strategoi succeeding Leo Sarakinopoulos in office is as
follows:
– Arkadios (protospatharios and strategos of Dorostolon);8

– Marianos (protospatharios and strategos of Dorostolon);9
– Peter (protospatharios and strategos of Dristra);10
– Basilakes (protospatharios and strategos of Dristra);11
– David (protospatharios and strategos of Thracia and Dristra); certainly
dated to the period after the reconquest, as he commanded troops
moving with him from Thrace to Dristra;12
– Theodor (primikerios and strategos of Dristra);13
– Tzotzikios (patrikios and strategos of Dorostolon in 1017).14
Because the name Dorostolon was more similar to the ancient Durosto-
rum, it appears to have been first used in the official title, before Dristra
was adopted instead. As already mentioned, the new name first appears
in 975 in the Taktikon Scorialensis. If Arkadios and Marianos are the only
commanders of Dorostolon, then they must have followed Leo Saraki-
nopoulos, and each other, in rapid succession before 975. There are so
far no other known commanders of Dorostolon. Tzotzikios’s title, which
8 Jordanov 1993, 118, nr. 217; Jordanov 2002, 82, nr. 4; Jordanov 2003 a, 65, nr. 23.2; Ivanov
2008, 138, nr. 2; Božilov 2008, 93; Jordanov 2009, 415, nr. 1166, 499, nr. 1510–1516; Jordanov
2011 b, 202.
9 Jordanov 2011 a, 81, nr. 1.
10 Jordanov 1993, 119, nr. 218–219; Jordanov 2002, 83, nr. 7; Jordanov 2003 a, 66, nr. 23.5;
Božilov 2008, 93; Yotov 2008 a, 348; Jordanov 2009, 415–416, nr. 1170–1171; Jordanov 2011 a, 82,
nr. 30–31. Possibly the same with the strategos of Ioannoupolis mentioned above.
11 Seibt 1995, 224 (who read Basileios); Jordanov 2002, 82–83, nr. 5; Jordanov 2003 a, 65,
nr. 23.3, Jordanov 2009, 415, nr. 1167; Jordanov 2011 a, 82, nr. 25–26.
12 Stănescu 1968 b, 44; Diaconu, Vâlceanu 1972, 18; Jordanov, Tăpkova-Zaimova 1988, 120,
122; Jordanov 1992 a, 286, nr. 13; Tăpkova-Zaimova 1993, 97–98; Jordanov 2000, 138, nr. 27;
Jordanov 2002, 85, nr. 13; Jordanov 2003 a, 102, nr. 35 C 20; Božilov 2008, 95; Krsmanović
2008, 138–139, 195 (dated to the beginning of the 11th century); Jordanov 2009, 439–440,
nr. 1270–1271.
13 Bănescu 1927, 17; Bănescu 1946, 52, 69–70; Stănescu 1968 b, 42–43; Barnea, Ştefănescu
1971, 89; Tăpkova-Zaimova 1973, 103; Jordanov, Tăpkova-Zaimova 1988, 122; Jordanov 1992
a, 288–289, nr. 19; Jordanov 2000, 138, nr. 31; Jordanov 2002, 83, nr. 6; Jordanov 2003 a, 66,
nr. 23.4; Božilov 2008, 95; Jordanov 2009, 416, nr. 1168, 1169; Jordanov 2011 a, 82, nr. 27–29.
A signet ring of the same commander was published by Markov 1998, 63–66.
ley, 337); Bănescu 1946, 70; Stănescu 1968 b, 43; Barnea, Ştefănescu 1971, 93; Jordanov 1995,
210; Božilov 2008, 95.
62 chapter two
also refers to Dorostolon, not to Dristra, is most likely an archaic form

deliberately employed by Skylitzes. In conclusion, Sisinios, Leo Sarakino-
poulos, Arkadios and Marianos were all commanders of Theodoroupolis/
Dorostolon/Dristra between 971 and 974, one for each year. All of them
were commanders of the province the center of which was the city of
Dorostolon/Dristra, not commanders of the city. At that time, themes
were still under the command of strategoi, but the transition to an army
based on professional, paid soldiers (tagmata) would involve the replace-
ment of those commanders with dukes or katepanoi.15 It is not easy to
explain why all four commanders stayed in office for such short periods
of time. Catherine Holmes observes that the “joint command and rapid
turnover have striking parallels with initial periods of rule on newly con-
quered frontiers in the east.”16 Perhaps the leadership in that newly con-
quered province was considered the most difficult position in the chain
of command in the first years after the conquest. That no seals of civilian
officials from the years between 971 and 986 have been found in the Pre-
slav archive suggests that no civilian administration was introduced dur-
ing the military occupation.17
In the Danube region, only one of these initial katepanates was pre-
served: Western Mesopotamia. The other one, Dristra, was turned into a
strategy a few years later, but still under Tzimiskes’s reign. This change in
status implies a reduction of troops coupled with the decrease in impor-
tance in comparison with other frontier regions. The alliance with the
Pechenegs across the Danube must have been regarded as sufficient, while
troops were badly needed in the East, where the war against the Arabs
had resumed. The strategy of Preslav was sufficient for defense in depth
and for monitoring one of the main routes to Constantinople. The troops
detached from the Thracian province remained at Preslav most probably
until 986, under the command of the strategos of Ioannoupolis (Preslav).
The title of strategos of Thrace and Ioannoupolis does not imply the col-
lapse of those provinces into a single administrative unit.18 It means only
that the army brought from Thrace was put under the leadership of the
strategos based in Ioannoupolis-Preslav. The presence of a strong military
15 Cheynet 1985, 186–187, 193.

16 Holmes 2005, 400.
17 Oikonomides 1998, 588–589.
18 As maintained, for instance, by Diaconu 1986, 177; Madgearu 1999–2000, 16–17; Stephen-
son 2000, 56; Holmes 2005, 400; Yotov 2008 a, 345, 348; Krsmanović 2008, 133; Krsmanović
2010, 610.
force in the new theme shows that that was still an occupation regime,
not a civilian administration.19
The final victory against Bulgaria in 1018 brought a new military and
administrative organization of the Balkan Peninsula, primarily through
the creation of the themes of Bulgaria, Sirmium and Serbia. Follow-
ing Nicolae Bănescu20 and several other historians,21 I have elsewhere
advocated the idea that the province of Paradunavon may be dated to
1018–1020, at the same time as the themes of Bulgaria and Sirmium.22 By
contrast, Ivan Jordanov,23 Paul Stephenson,24 Boris Borisov,25 Ioan Bica26
and Bojana Krsmanović27 believe that the absence of seals of the katepano
of Paradunavon dated before the 1050s can only mean that the theme of
Paradunavon theme had not yet been established. According to them,
the theme called Dristra must have remained in existence until the Pech-
eneg troubles of 1045 caused a re-shuffle of the organization of the Lower
Danubian region. Tadeusz Wasilewski advanced a third interpretation,28
according to which the theme or duchy of Paradunavon appeared in 1027,
as a result of the Pecheneg raid of that same year. The discovery of a seal
of Leo Drymis, anthypatos, patrikios and katepano of Dristra has put an
end to this controversy. The title of katepano proves that Leo was the
commander of the theme, and not of the city of Dristra.29 It is known
that he was vestis and katepano of Bulgaria sometime between 1055 and
1065 (Werner Seibt has established a new chronology for his seals).30 He
19 Whittow 1996, 296; Oikonomides 1998, 588–589.

20 Bănescu 1946, 54–60.
21 Barnea, Ştefănescu 1971, 76, 93–95; Tăpkova-Zaimova 1973, 103, 112; Kuzev 1979, 37; Jor-
danov, Tăpkova-Zaimova 1988, 122; Kühn 1991, 223–224; Tăpkova-Zaimova 1993, 98; Mărculeţ
2005, 57–59; Strässle 2006, 460. Diaconu 1992 b, 327 accepted Bănescu’s theory on the basis
of a wrong identification of Michael, vestarches and katepano of Paradunavon with a ruler of
the Paristrian cities in 1045–1047 (see below). The date accepted by Diaconu was regarded as
sufficient proof that Bănescu had been right about the katepano of Paradunavon appearing
soon after the 1020s. I have also followed Diaconu in this respect in my previous studies, but
this opionion is now obsolete.
22 Madgearu 1999 a, 422; Madgearu 1999–2000, 20; Madgearu 2001, 76.
23 Jordanov 1987 b, 202; Jordanov 2002, 81; Jordanov 2003 a, 136; Jordanov 2003 b, 73.
24 Stephenson 2000, 78, 94; Stephenson 2003, 115.
25 Borisov 2007, 71.
26 Bica 2003, 98–107.
27 Krsmanović 2008, 194–195.
28 Wasilewski 1975, 642, 645.
29 Jordanov 2002, 80–81, nr. 1; Jordanov 2003 a, 62; Jordanov 2006, 147, nr. 209.2; Jordanov
2011 a, 82, nr. 34.
30 Laurent 1969, 146; Seibt 1995, 227; Jordanov 2011 c, 168–169, 182. Previously, it was con-
sidered that Leo Drymis was duke of the Bulgarian theme by the end of the 12th century
(Bănescu 1946, 151–152).
64 chapter two
must therefore have been katepano of Dristra before that, which implies
that the official name Dristra for the Danube province remained in use
until the mid-11th century. This is now a final conclusion.
The first known commander of the theme of Dristra, whose term may
be associated with precisely dated events (the Rus’ attack of 1043, for
which see chapter III. 2), is vestis Katakalon Kekaumenos, whom Skylitzes
describes as “ruler (archon) of the Istrian cities and lands” (ἄρχων ὢν τῶν
παρὰ τὸν Ἴστρον πόλεων καὶ χωρίων). This is in fact a well known char-
acter of Byzantine history. In April 1042, Katakalon Kekaumenos was in
Constantinople, while at some point during the last months of 1045 he
was appointed duke of Iberia. This must therefore have become katepano
between 1042/1043 and 1045.31 Two seals of Κατακαλὼν Καμεν, ἀνϑύπατος
καὶ κατεπάνο were found in Pliska.32 They show that the office of katepano
has been introduced to the Danube province of Dristra in the 1040s, most
probably as a response to the serious Pecheneg invasion of 1036. A katepano
or a duke was the commander of tagmata, a kind of professional soldiers
dispatched from the central army. In older themes, commanders contin-
ued to be called strategoi until the first two or three decades of the 11th
century, but provinces created under Basil II had dukes or katepanoi as
commanders, no doubt because troops under their command were units
of the central army (tagmata). The first dukes are mentioned in the west-
ern provinces of the Empire, in Thessaloniki and Adrianople, and then in
Western Mesopotamia. After 976, dukes became involved primarily in the
war against Samuel, and new and better troops stationed in Thessaloniki
and Adrianople were put under the command of dukes, who replaced the
strategoi. Thessaloniki and Adrianople became the most important mili-
tary bases of the Balkan region under Basil II.33
The next katepano of the Danube province was patrikios Michael, son
of Anastasios. According to Skylitzes, he was the “ruler of the Paristrian
cities” (ἄρχων τῶν παριστρίων πόλεων). Michael is mentioned two times
in relation to the Pecheneg crisis of the 1040s, which will be discussed
31 Skylitzes, Constantine Monomachos, 6 (ed. Thurn, 433; transl. Flusin, 360; transl. Wort-
ley, 407); Bănescu 1946, 74–78; Barnea, Ştefănescu 1971, 127; Wasilewski 1975, 643; Jordanov
1995, 210–211; Mănucu-Adameşteanu 2001 a, 177; Jordanov 2003 b, 63–64; Bica 2003, 98–99;
Božilov 2008, 96; Jordanov 2011 a, 113. For his career, see Savvides 1986–1987, 22–23.
32 Jordanov 1992 a, 290–291, nr. 24; Jordanov 2000, 139, nr. 35; Jordanov 2006, 174–176,
nr. 257–258; Jordanov 2009, 488, nr. 1465–1466.
33 Cheynet 1985, 186; Holmes 2005, 403–412; Strässle 2006, 216; Krsmanović 2008, 77–79,
143, 148–159; Krsmanović 2010, 609–610, 623–625, 631.
in the following chapter.34 The chronology of the events was confused

before Jacques Lefort’s conclusive study, dedicated to the oration of John
Mauropous, the metropolitan of Euchaita. Lefort showed that the inva-
sion of the Pechenegs under their chief named Tyrach, which took place
while Michael was still in office, must be dated between December 1046
and January 1047.35 Meanwhile, Alexander Kazhdan argued that the other
Pecheneg group led by Kegen had crossed the Danube in 1045, when
Michael was already a commander there.36 This implies that he was in
office between 1045 and 1047, after Katakalon Kekaumenos. He is most
likely to have lost his position after the invasion of Tyrach, since he was
incapable of dealing with the Pecheneg crisis and securing the defense of
the province. Skylitzes makes it very clear that the second Pecheneg inva-
sion was made possible by the frontier not being guarded, while according
to Kekaumenos (the Byzantine writer and aristocrat from Larissa, not the
military commander!) the Pechenegs took advantage of the incompetence
of the local commanders (akritai).37 At any rate, Michael was replaced by
one of two subsequent katepanoi of Dristra, either Leo Drymis or Con-
stantine (. . .)polites. The latter’s truncated name is known from a seal,
on which he appears with the title patrikios and katepano of Dristra. The
seal may be dated between 1030 and 1040,38 but also after 1047, the year
of Michael’s removal. In that case, Leo Drymis must have succeeded the
katepano with truncated name, sometime in the early 1050s. Michael, on
the other hand, resurfaced as magistros in 1057, and was involved in the
plot that led Isaac I Comnenos to imperial power.39
34 Skylitzes, Constantine Monomachos, 16–17 (ed. Thurn, 455–459; transl. Flusin, 377–380;
transl. Wortley, 427–430); Bănescu 1946, 74–78; Barnea, Ştefănescu 1971, 129–130; Wasilewski
1975, 643–644; Mănucu-Adameşteanu 2001 a, 178; Jordanov 2003 b, 64–65; Bica 2003, 99–100;
Božilov 2008, 96.
35 Lefort 1976, 273–284.
36 Kazhdan 1977, 65–77. Shepard 1975, 71–75 had dated the invasion of Tyrach to the
winter 1048–1049, but it was only Lefort’s demonstration that settled the issue.
37 Skylitzes, Constantine Monomachos, 17 (ed. Thurn, 458; transl. Flusin, 379; transl. Wort-
ley, 429–430); Kekaumenos, 47 (ed. Spadaro, 82/83).
38 Nesbitt, Oikonomides 1991, 150, nr. 65.1; Jordanov 2002, 81–82, nr. 2; Jordanov 2003 a,
62; Božilov 2008, 96. Only the letters . . . ist . . . remained from the name of the province, and
this cannot be “Paristrion” as suggested in the Dumbarton Oaks Catalogue, because all other
seals indicate “Paradunavon.” Mănucu-Adameşteanu 2001 a, 183 dated him without any
arguments to the third quarter of the 11th century. I have previously (Madgearu 1999 a, 422)
advanced the idea that this katepano was in office in 1010ss or 1020s, as I believed the theme
of Paradunavon to have come into being immediately after 1018. My (wrong) conclusion was
reproduced by Mărculeţ 2005, 60 and Meško 2006, 131.
39 Skylitzes, Michael the Old, 12 (ed. Thurn, 498; transl. Flusin, 409; transl. Wortley, 463).
66 chapter two
The following katepano was no other than Romanos Diogenes, who

later became emperor between 1068 and 1071. According to Michael Attal-
iates, as he was still duke of Serdica in 1067, Romanos summoned the
people whom the Byzantine author calls Sauromatai to assist him against
Emperor Constantine X Dukas (1059–1067). The Sauromatai apparently
knew him since he had fought against them, as a ruler of the Paristrian cit-
ies (ἄρχων τῶν περὶ τὸν Ἴστρον πόλεων). Magister Nikephoros Botaneiates
had saved Romanos Diogenes in the war against the Sauromatai. Much
of the discussion surrounding this story revolves around the identity of
the Sauromatai and the chronology of the events. Although Attaliates
employs the archaism Sauromatai mostly in reference to Hungarians, in
this case he seems to have had the Pechenegs in mind. The general con-
text of events fits the year 1053, when Nikephoros Botaneiates is known
to have assisted Romanos during the battle near Preslav, in which Basil
Monachos, the duke of Bulgaria, was killed. Troops from the Danube
theme took part in that battle.40 It is unlikely that the events mentioned
took place in 1059, since the campaign of that year was under the com-
mand of Isaac I Comnenos, who defetead the Pechenegs.41 Romanos Dio-
genes must have been katepano of Dristra in 1053, but the length of his
term remains unknown. A seal found at Vetren has him with the titles
vestarches and katepano, but without the name of the province.42 Another
seal from the same boulloterion was later found in Silistra.43 That seal must
be dated after Romanos Diogenes served as commander of Dristra, for he
received the rank of vestarches in 1067, as a reward for a victory against
the Pechenegs obtained at the time he was duke of Serdica.44 There is a
strong possibility that the letter sent to Vetren on the Danube concerned
40 Attaliates, ed. Bekker, 37–43, 97 (ed. Pérez Martín, 29–33, 73–74); Skylitzes, Constan-
tine Monomachos, 17 (ed. Thurn, 458; transl. Flusin, 379–380; transl. Wortley, 430–431); Keka-
umenos, 67 (ed. Spadaro, 100/101); Bănescu 1946, 83; Stănescu 1966, 55; Diaconu 1970, 75,
93; Mănucu-Adameşteanu 2001 a, 178; Bica 2003, 139–141; Stoimenov 2008, 177–180. In 1059,
Nikephoros Botaneiates was duke of Edessa and Antioch (Cheynet 1983, 461; McGeer, Nes-
bitt, Oikonomides 2005, 24; Karagiorgou 2008, 108–111, 128).
41 Attaliates, ed. Bekker, 67 (ed. Pérez Martín, 51–52); Skylitzes Continuatus, 107;
Zonaras, XVIII, 6.1–5 (ed. Büttner-Wobst, 671); Matthew of Edessa, II, 5 (ed. Dostourian,
90–91); Michael the Syrian, XV.2 (ed. Chabot, 165) (the attack of Isaac against Partiqayê, the
Pechenegs); Diaconu 1970, 76–77; Spinei 2006, 197; Spinei 2009, 112.
42 Atanasov, Jordanov 1994, 37–40, nr. 116; Jordanov 1995, 212; Jordanov 2003 b, 65; Jor-
danov 2006, 128–130, nr. 167; Krsmanović 2008, 195; Božilov 2008, 96 (dated around 1065);
Jordanov 2009, 494, nr. 1490.
43 Jordanov, Žekova 2007, 125, nr.326; Jordanov 2009, 494, nr. 1491.
44 Zonaras, XVIII, 10.12 (ed. Büttner-Wobst, 684); Dančeva-Vassileva 2004, 24. For the
titles of Romanos Diogenes, see Jordanov 2003 b, 65.
military operations against the Pechenegs north of the Danube who con-
trolled at that time the southern bank of the river as well.
In my opinion, Romanos Diogenes’s successor was Demetrios Katakalon,
the first commander to be called katepano of Paradunavon. Five seals
(among which two from Silistra and one from Pliska) have him with the
title anthypatos, patrikios and katepano of Paradunavon.45 A chronologi-
cal point of reference for his career is the destruction of Krivina, a set-
tlement on which another one of his seals was found, which bears the
titles patrikios and strategos. The Krivina seal must obviously be dated
before Demetrios Katakalon became katepano of Paradunavon. Another
seal mentioning him as strategos, not katepano was found in Silistra.46 As
patrikios and strategos, Demetrios Katakalon must have been the com-
mander of an unknown city in the region. According to the numismatic
evidence, the settlement in Krivina was destroyed during the Pecheneg
invasion of 1047. Petre Diaconu therefore believed that Demetrios Kataka-
lon had been appointed strategos in the late 1030s.47 He then became the
commander of Paradunavon at some point after 1050 or even after 1055,
since is known to have become proedros at a later time, after obtaining
the command of Paradunavon. The rank of proedros was bestowed upon
provincial dukes particularly during the reign of Constantine X.48 On the
other hand, according to three seals from Silistra, Demetrios Katakalon
received two other titles while being katepano of Paradunavon, namely
anthypatos and vestis.49 This can only mean that he remained in office
for quite some time. This strongly suggests that his term post-dated that
of Romanos Diogenes, and that it therefore started after 1055. The time
span between the moment in which he was appointed strategos (around
1040) and the moment in which he is known to have been proedros and
katepano (after 1055) is consistent with the cursus of the average Byzan-
tine commander.50
45 Bănescu 1946, 95–97 (Silistra and Constantinople); Nesbitt, Oikonomides 1991,

153, nr. 67.1 (unknown location); Jordanov 2003 a, 136–137, nr. 58. 1; Jordanov 2006, 198–
199, nr. 302–303; Ivanov 2008, 140, nr. 6 (Pliska); Jordanov 2009, 463–464, nr. 1373–1375,
487, nr. 1462.
46 Jordanov 1981, 92–95; Jordanov 2003 a, 137; Jordanov 2006, 198, nr. 300–301; Jordanov
2009, 502–503, nr. 1530, 1531.
47 Diaconu 1992 a, 179–181.
48 Laurent 1969, 145; Diaconu 1992 a, 181.
49 Jordanov 2006, 198–199, nr. 302–304.
50 Bănescu 1946, 95–97 (accepted by Jordanov 1981, 92–94; Jordanov, Tăpkova-Zaimova
1988, 123; Jordanov 2003 a, 136–137, nr. 58.1; Jordanov 2003 b, 68–69). Stănescu 1968 b,
48 proposed the end of the 11th century. Mănucu-Adameşteanu 1995, 350 (the same in
68 chapter two
After Demetrios Katakalon, Theodore Pegonites was invested with the

office of katepano of Paradunavon. He also had the ranks of patrikios and
anthypatos, according to his seal found in Izvorovo (near Silistra), which
is dated between 1050 and 1060. A seal with the same legend has recently
appeared in an auction (its findspot is therefore unknown). In another
seal from the 1040s, Theodore Pegonites is mentioned only as strategos,
and sometimes before 1065 he appears to have gained the title of duke
of Edessa.51
With all those commanders, the name of Paradunavon attached to
their titles indicates a new organization of the Danube region. The theme
of Dristra was thus replaced by the theme of Paradunavon, no doubt cov-
ering much of the same area. The change was introduced under Isaac I
Comnenos (1057–1059), mostly likely in 1059, during the war against Hun-
gary and the Pechenegs, which will be discussed below. The implication
of this dating is that both theories advanced earlier are wrong. Bănescu’s
idea (which I have also endorsed) places the change in 1018, at the same
time as the creation of the themes of Bulgaria and Sirmium. Wasilewski’s
idea, later developed by Paul Stephenson, links the transformation of the
military organization at the Danube to the new policy implemented dur-
ing the Pecheneg crisis, which started in 1045. The analysis of the seals of
commanders shows, however, that the name Paradunavon was not used
as an official title before the late 1050s.52 The first katepano of Paradu-
navon was Demetrios Katakalon, in function in 1059.
The name Paristrion which appears in the written sources was never
used officially, as it is absent from the nomenclature of the seals.53 The
official name Paradunavon means “near the Danube.” The river, in this
case, is not the traditional (and by this time quite archaic) Istros or the
equally pretentious Δάνουβις (a name of Latin origin employed by Anna
Comnena and Zonaras), but the Dunav, a name of Bulgarian origin. The
Byzantine authorities most obviously borrowed the name from the popu-
lation living by the river, and preferred it to old provincial name derived
Mănucu-Adameşteanu 2001 a, 181–182) thought that Demetrios Katakalon was duke in

ca. 1070. Božilov 2008, 97 even proposed the 1080s.
51 Matthew of Edessa, II, 27 (ed. Dostourian, 108); McGeer, Nesbitt, Oikonomides 2001,
163, nr. 73.2; Jordanov 2003 a, 137, nr. 58.2; Jordanov 2003 b, 69–70; Jordanov 2006, 343,
nr. 573; Wassiliou-Seibt 2008, 134; Wassiliou-Seibt 2009, 307–309, nr. 5 a–c; Jordanov 2009,
464, nr. 1376; Jordanov 2011 a, 84, nr. 51–51a.
52 The same opinion in Jordanov 2003 b, 73; Jordanov 2011 a, 85–86.
53 Stănescu 1968 b, 42–64 considered that the province had three successive names: Dris-
tra, Paradunavon, Paristrion. This point of view is now obsolete.
from Dristra. As Eugen Stănescu noted, it must have been a concession

made to the local population, and it is quite possible that the new provin-
cial name is at the origin of the Bulgarian name Podunavia, which appears
in 14th-century Walachian charters.54
In his account of the Oghuz invasion of 1065,55 Michael Attaliates men-
tions the names of two “rulers of the Danubian cities” (ἐπάρχοντος κατὰ
τὸν Ἴστρον πόλεων), Basil Apokapes and Nikephoros Botaneiates. The con-
tinuator of Skylitzes and Zonaras56 reproduce the same information. Basil
Apokapes was taken prisoner by the Oghuz, but managed to escape with
the assistance of a barbarian. Because the existence of two commanders
for the same province at once is impossible, Tadeusz Wasilewski57 and
Petre Diaconu58 have proposed that in 1065 the theme of Paradunavon
had been divided into two smaller provinces—Western and Eastern
Paristrion. Wasilewski dated the change to 1059, under Isaac I Comnenos.
Diaconu embraced the idea, and suggested that Basil Apokapes had been
entrusted with the command of the Western Paristrion. Nicolae Bănescu
had a different opinion on this matter. According to him, in 1065 Nike-
phoros Botaneiates was duke of Bulgaria in 1065,59 but he failed to produce
any evidence to support his argument. On the contrary, it is likely that
that office was at the time occupied by vestarches Andronikos Philokales,
who is attested in 1066.60 Bănescu also tried to prove that Basil Apokapes
was duke of Paradunavon in 1059, and that he had remained in that office
until 1065, when he was captured.61 Diaconu, based on the order in which
the two dukes are mentioned by the continuator of Skylitzes, believed
that Basil Apokapes was the commander of the “Bulgarian” troops, which
are mentioned first. Those “Bulgarians” were soldiers from the province
54 Stănescu 1968 b, 56–57. The name was most obviously still in use in the 14th century
in Bulgaria. Emperor Ivan Alexander is said to have been ruler of Podunavie in a document
dated to 1337 (Petkov 2008, 468).
55 Some studies mention the year 1064, but the right date results from the chronicle of
Matthew of Edessa (see below).
56 Attaliates, ed. Bekker, 83 (ed. Pérez Martín, 63); Skylitzes Continuatus, 113–114; Zonaras,
XVIII, 9.2 (ed. Büttner-Wobst, 678).
57 Wasilewski 1964, 479–480 (Basil Apokapes in west, Botaneiates in east). The theory
was reproduced with some changes in Wasilewski 1995, 198.
58 Diaconu 1970, 82–99.
59 Bănescu 1946, 32–34, 89–90, 142–143. See also Cheynet 1983, 461–462; Kühn 1991, 230;
Grünbart 1998, 38; Karagiorgou 2008, 128–129; Jordanov 2011 c, 182.
60 Kekaumenos, 181 (ed. Spadaro, 218/219); Bănescu 1946, 144; Kühn 1991, 230; Jordanov
1993, 160–161, nr. 323; Jordanov 2009, 483, nr. 1444; Valeriev 2010, 428; Jordanov 2011 c, 157–
158, 182.
61 Bănescu 1946, 84–90.
70 chapter two
based in Serdica (Sofia), to which Diaconu referred as Western Paristrion.

The hypothesis is not very convincing, and even Diaconu admitted that
the solution to the problem may be different. That the theme of Para-
dunavon was split into two smaller provinces has also been maintained
by Vasilka Tăpkova-Zaimova, who thought that Nikephoros Botaneiates
ruled the western part of the theme, while Basil Apokapes was the com-
mander of the eastern part.62
Other sources concerning Basil Apokapes substantiate an alternative
interpretation.63 He appears, for example, in the will of an aristocrat from
Cappadocia named Eustathios Boilas.64 A man called Βασιλείου μαγίστρου
τοῦ παραδούναβι is also mentioned among the various provincial com-
manders at the end of a will drafted for Boilas by the monk Theodoulos
on April 4th, 1059. He appears with his full name (Basil Apokapes) among
the executors of the will. He was born in Tayk, much like Boilas and the
local bishop (the other executors). Because only dukes of several eastern
provinces were mentioned in Theodoulos’ note, the presence of a com-
mander of Paradunavon in the East for this occasion is hard to imagine.
No explanation was offered in this respect either by Bănescu or by other
scholars who followed him.65
The reading Βασιλείου μαγίστρου τοῦ παραδούναβι was established in 1910
by the first editor of the will, Spyros Lambros, and was accepted as such by
Bănescu.66 However, the reading is far from clear, as the lines are placed
on both sides of the cross of Saint John Climacus, in the eschatocol of the
document, and could thus be read in two different ways. Vasil N. Zlatarski,67
followed by Rats M. Bartikian,68 regarded the words τοῦ παραδούναβι as
linked to the name of Aaron, the duke of Mesopotamia, and the son of the
last Bulgarian Emperor, John Vladislav.69 Paradoúnabi was in this case a
nickname pertaining to the man’s origin from a certain region (“from the
62 Tăpkova-Zaimova 1973, 106–109.

63 The first form of this argument was presented in Madgearu 1998. The study of other
works not available to me at the time, and consultation of studies published after 1998 have
led to a refined version of the same ideas presented in Madgearu 2005.
64 The source was translated and commented by Vryonis 1957, 263–277. Another major
contribution in this respect is Lemerle 1977, 15–63.
65 Barnea, Ştefănescu 1971, 133; Lemerle 1977, 40–42; Barnea 1986, 271; Savvides 1991–
1992, 98–103; Grünbart 1998, 37–40; Mohov 1999, 158–168; Stephenson 2000, 94; Mănucu-
Adameşteanu 2001 a, 179–180.
66 Bănescu 1946, 84–89; Bănescu 1963, 155–158.
67 Zlatarski 1929, 22–34.
68 Bartikian 1967, 315–319.
69 About Aaron, see Jordanov 1990, 107–108; Jordanov, Žekova 2007, 65, nr. 140.
province near the Danube”). If so, then Basil Apokapes was simply duke
of Iberia, like John Monasteriotou, who is mentioned after him. The exis-
tence of two dukes in Iberia means that one of them, Basil Apokapes, was
in fact the ruler of the Armenian-Georgian district of Tayk.70
In conclusion, the final part of the will reads κατεπάνω Ἐδέσες
Ἰωάννου τοῦ Δουκήτζη, δουκῶντος Ἀδριανοῦ Ἀντιοχείας, Ἀαρὼν προέδρου
καὶ αυτοδέλφου τῆς αυγούστης Μεσοποταμίας τοῦ παραδούναβι, Βασιλείου
μαγίστρου, Ἰωάννου Μοναστιριότου Ἠβηρίας, πανκρατίου Βαασπρακανίας καὶ
Ἰωάννου κοροπαλάτου καὶ δομεστίκου τῶν σχολῶν αυτοδέλφου τοῦ βασιλέως
Κομνηνοῦ. John Dukitzes, Aaron, John Monasteriotes and John Comne-
nos are all well-known men. Adrian was identified with a member of the
Dalassenos family; Pankratios (the duke of Vaspurakan) is the same with
the Armenian prince Bagrat Vakhac’i.71 Magister Basil has no relation with
the name Paradunavon. On the other hand, the form τοῦ παραδούναβι may
simply be a mistake if it is taken as the title of a provincial commander. In
that case, it should read τοῦ παραδούναβου. At any rate, the phrase must be
understood as “the man from the Danubian region,” and applies to Aaron,
and not to Basil.
Ion Barnea72 defended Bănescu’s idea with new arguments concerning
the relationship between Basil Apokapes and Paradunavon, when pub-
lishing a seal found during archaeological excavations in Nufăru.73 Three
other, identical seals have been discovered in Silistra, Bradvari, and an
unknown location in north-eastern Bulgaria, while a fourth one is known
from the Zacos collection.74 All of them have the legend “Basil Apokapes,
magistros, vestis and dux” (μαγίστρῳ βέστῃ καὶ δουκὶ). The complete text
of the legend was established by M. Grünbart after the piece from the
Zacos collection (the single one preserved entirely).75 None of those seals
mentions the name of the province, for which Basil Apokapes served
as duke.
Because several seals of Basil Apokapes have been discovered within the
territory of Paradunavon, Barnea believed he had been the commander of
70 So Bartikian. Yuzbashian 1973–1974, 164–165 rejects the idea of Apokapes being duke
of Iberia, but failed to offer an identification for Tayk.
71 Lemerle 1977, 40–42. For Pankratios see also Yuzbashian 1973–1974, 152.
72 Barnea, Ştefănescu 1971, 133.
73 Barnea 1986, 271; Barnea 1987 a, 194; Barnea 1987 b, 84–85, nr. 7.
74 Jordanov 1986, 123–124; Grünbart 1998, 37–38; Jordanov 2003 b, 66; Jordanov 2006,
56–57, nr. 43–45; Jordanov 2009, 485, nr. 1449–1451.
75 Grünbart 1998, 37–38.
72 chapter two
that province.76 The same line of reasoning was followed in a much cited
study of the diffusion of seals.77 In fact, all that those seals from Paradu-
navon show is that Basil Apokapes had for a while some office imply-
ing correspondence between him and officers in that province. The seals
themselves do not prove in any way that he was the commander of the
province, since it is quite possible that he was the commander of an army
corps moved to Paradunavon from elsewhere.
Basil Apokapes is also mentioned in the chronicle of Matthew of Edessa
as having been at the head of an army, which Emperor Constantine X
sent to the Danube against the Oghuz. The Armenian chronicle covers
the period between 952 and 1136, the year of its author’s death. Although
primarily dealing with events in Armenia, it often offers useful details
about developments elsewhere in the Empire, on the basis of informa-
tion from other, now lost sources. Edouard Dulaurier’s old translation was
used by most historians who wrote on Basil Apokapes. The more recent
English translation is based on a diferent, and apparently better manu-
script. According to that manuscript, “in the year 514 of the Armenian
era [1065–1066], during the reign of the Roman emperor Dukas, a great
war broke out in the West caused by the nation of the Uzes. The emperor
Dukas collected troops from all the Greeks and from the forces of Arme-
nia. He appointed the illustrious Roman magnate Basil, the son of Abukab,
as commander of these forces. Basil, advancing with many troops, came
and descended upon the great river called the Danube. Here on the banks
of the river a violent battle took place between the Romans and the Uzes,
and there was heavy slaughter on both sides (. . .). The enemy captured
Basil, the Roman general, and led him into captivity to their country (. . .).
For many years Basil remained captive in the country of the Uzes (. . .).
After a while one of the infidel troops contemplated freeing Basil, and the
general, in turn, promised to give him many things, including a position
of high rank from the emperor. A few days later this man, with the help
of some of his friends, snatched Basil and immediately brought him to
the emperor Dukas. Thus there was much rejoicing among all the Greeks,
and the emperor gave many gifts to those who had brought Basil. After
this Basil came to his father Abukab in Edessa.”78 Matthew of Edessa
76 Barnea 1987 a, 194. The same opinion (Basil Apokapes as duke of Paradunavon between
1059 and 1065) at Mărculeţ 2005, 65–68; Mărculeţ 2008, 189–196; Božilov 2008, 96; Mărculeţ
2009, 163–177; Jordanov 2011 a, 114–115.
77 Cheynet, Morrisson 1990, 118.
78 Matthew of Edessa, II, 24 (ed. Dostourian, 105); Yuzbashian 1973–1974, 146–147.
mentions another transfer of troops from the East to the Danube, when
in 1040 Emperor Michael IV moved troops from the themes of Sebasteia,
Taron and Vaspurakan against the Bulgarian rebel Peter Delian.79 So, this
was not something unusual.
To judge from the information culled from the chronicle of Matthew
of Edessa, Basil Apokapes came to the Danube in 1065 as duke of an
army gathered from several provinces, including Armenia. It is quite clear
that he was not appointed commander over any province in the region
to which the army had been sent.80 This explains quite neatly why the
seals of Basil Apokapes have no mention of province in relation to his title
of duke. He was neither a deputy, nor the commander of Paradunavon,
but the general of the troops summoned for assistance. Like Michael
Attaliates and the continuator of Skylitzes, Matthew of Edessa knew that
Basil Apokapes had escaped from the captivity, but added that he later
returned to his father Aboukab (Michael Apokapes) in Edessa. He gives
further details about Basil’s life, including the fact that he died Edessa in
1083 or 1084.81 For this Armenian historian from Edessa, Basil Apokapes
was a great local hero, the member of an illustrious family. Even though
he wrote at a great distance from the Danube, he was well informed about
his hero’s life. There is no reason to doubt his testimony, and completely
pointless to rely instead on the chronicle of Smbat Sparapet,82 who pro-
vides no additional details of any relevance, since his is just a summary
of Matthew’s account.
Michael Attaliates and Skylitzes wrote about Basil Apokapes battling
the Seljuk Turks as commander of Mantzikert (a fortress in the theme
of Vaspurakan, which was the province created after the occupation of
the Armenian kingdom with the same name in 1022).83 In 1054 or 1055 he
was the strategos of that city (Attaliates calls him a hegemon), but not the
katepano of the province, as attested by a seal from the Zacos collection,
79 Matthew of Edessa, I, 80 (ed. Dostourian, 68).

80 As noted Guilland 1971, 1–2, Basil Apokapes “fut chargé de défendre avec le magistros
Nicéphor Botaniate les places fortes du Danube”. This seems to be the best interpretation
of the source.
81 Other seals show that later in his life Basil Apokapes was duke of Edessa: proedros, pro-
toproedros, protonobelissimos and finally sebastos: Grünbart 1998, 38–40; McGeer, Nesbitt,
Oikonomides 2001, 163, nr. 73.1; Cheynet 2001 a, 59–66, nr. 32, 33. His brother Pharasmanes
is also known. He was strategos of Strumitza in the 1040s and vestarches in 1059 (Grünbart
1998, 31; Jordanov 2006–2007, 522–524; Jordanov 2008, 45–49, nr. 16).
82 Mărculeţ 2008, 193–194; Mărculeţ 2009, 169.
83 Yuzbashian 1973–1974, 148.
74 chapter two
on which he appears with the titles of protospatharios and strategos.84

Seven other seals belong to a later period of his life, when already had the
titles of vestarches and katepano of Vaspurakan (Βαασπουρακανίας). Two
of those seals are from Pliska, one from Malka Popina (west of Silistra),
one from an unknown location in north-eastern Bulgaria, one from the
collection in the Hermitage Museum, and two from the Zacos collection.85
M. Grünbart (based on some stylistic features of the seals) has dated all
of them to the 1030s, and has attributed them to another Basil Apokapes.86
However this cannot be so, because the title of vestarches was most typi-
cal for military commanders between 1040 and 1080.87 A better solution is
to have Basil Apokapes promoted from vestis to vestarches at the time of
his appointment as katepano of Vaspurakan. Because four seals recording
the title of vestarches were found in Paradunavon, the promotion must
have taken place while Apokapes was already in the Danube region, but
still a katepano of Vaspurakan. It is theoretically possible that the letters
to which those seals were attached had been sent from Vaspurakan to
Dristra and Pliska, but it is more likely that Apokapes was already in Para-
dunavon with his army, when sending those letters out.
The order in which Basil Apokapes’ titles appear on seals is the following:
patrikios, magistros, vestis, vestarches, proedros, etc. In 1054 or 1055, he was
only patrikios, but he later became magistros and dux. He may have been
promoted dux of Tayk as a reward for his victory at Mantzikert. In this case,
he became magistros and duke of Tayk in 1055 or shortly after that.
Basil Apokapes’ cursus may therefore be reconstructed as following:88
in 1054 or 1055, he was patrikios and strategos of Mantzikert;

after 1055 he was appointed
– magistros and duke of Tayk, and was still in that office in 1059 when he
is mentioned in the will of Eustathios Boilas;
84 Attaliates, ed. Bekker, 46 (ed. Pérez Martín, 35); Skylitzes, Constantine Monomachos, 19
(ed. Thurn, 462–464; transl. Flusin, 382–383; transl. Wortley, 432); Bănescu 1946, 85; Grous-
set 1947, 597–598, 600; Grünbart 1998, 33.
85 Jordanov 1986, 126; Jordanov 1992 a, 283; Grünbart 1998, 35–36; Jordanov 2000, 138;
Mănucu-Adameşteanu, Jordanov 2002, 125; Jordanov 2003 a, 43, nr. 14.1; Jordanov 2006, 56,
nr. 40–42; Jordanov, Žekova 2007, 108, nr. 278; Jordanov 2009, 407–408, nr. 1130–1133.
86 Grünbart 1998, 36.
87 Seibt 1978, 226; Cheynet 1983, 471. For instance: Iberia: Michael Iasites (1044–1045),
Ioannes Monasteriotes (1059); Bulgaria: Michael Saronites (1073–1074); Dyrrachion: Michael
Maurik (1067). See Yuzbashian 1973–1974, 164; Kühn 1991, 189, 226, 231, 238.
88 For his career see also Savvides 1991–1992, 98–103; Grünbart 1998, 37–40.
in 1065, he was
– vestarches and katepano of Vaspurakan, as well as magistros, vestis and
duke of an unspecified province;
he became
– proedros, protonobelissimos, and duke of Edessa after 1077;
finally, he was sebastos and duke of Edessa between 1081 and 1083.
According to Alexios Savvides, Basil Apokapes had been transferred from

Armenia because he had become dangerously popular there, and Emperor
Constantine X was afraid of a mutiny in that part of the Empire.89 This
is of course possible, but Savvides is wrong when writing that Apokapes
came to Paradunavon in 1059. Constantine X became emperor in 1059,
but when Theodoulos wrote his will Isaac I Comnenos was still on the
throne, as mentioned in the eschatocol.90 If Apokapes came to Paradu-
navon in 1059 or earlier, the emperor sending him there must have been
Isaac. But Matthew of Edessa makes it clear that Apokapes was sent to
the Danube frontier in the circumstances surrounding the invasion of the
Oghuz. In this case, the emperor in question was Constantine X. In 1065,
relations between the Byzantine administration and the Armenian aristo-
crats from Cappadocia were tense, because the emperor has begun per-
secuting those who refused to embrace Chalcedonian orthodoxy.91 Given
his military prestige, Basil Apokapes was a potential leader of an Arme-
nian rebellion. This may indeed explain why troops from distant Armenia
were dispatched to the Danube, precisely at a moment when the Seljuk
threat required their presence in the East. The emperor’s decision proved
to be a major mistake of strategy, as the eastern frontier was deprived of
key forces, while the war on the Danube was a complete failure. In any
case, Basil Apokapes was not duke of Paradunavon in 1065. He was sent
to the Danube to assist the commander of the province, Nikephoros Bota-
neiates. Basil Apokapes was only the duke of an army corps (tagma) sent
as reinforcement to the province Paradunavon. His title of duke appears
therefore on his seals without the name of any province.92
89 Savvides 1991–1992, 100–101.

90 Lemerle 1977, 40.
91 Dédéyan 1975, 105–109.
92 Similar conclusions in Jordanov 1986, 123–127; Stepanenko 1994, 27–34; Šandrovskaja
1994, 158–160; Jordanov 2003 a, 43–44; Jordanov 2006, 57–59.
76 chapter two
A seal discovered at Constantinople belongs to a certain Βασιλεὶου ἐκ

προσώπου τοῦ παραδούναβου. Ion Barnea thought that dignitary to be Basil
Apokapes, who had been sent to Paradunavon as a deputy of the com-
mander of the theme.93 The title ek prosopou refers indeed to a deputy, but
nothing indicates this to be Basil Apokapes. Vasile Mărculeţ also thought
that Basil Apokapes had been sent as a deputy for an unknown duke of
Paradunavon in 1059, being transferred from Vaspurakan,94 but appar-
ently did not know that Barnea’s identification of Basil ek prosopou with
Basil Apokapes was just a supposition.95 The seal from Constantinople is
important for this book only because it mentions yet another rank in the
organization of the province. It does not belong to Basil Apokapes.
The account of the two dukes in Paradunavon in Michael Attaliates, the
continuator of Skylitzes, and Zonaras may thus be understood only as a
reference to Nikephoros Botaneiates as duke the theme of Paradunavon
in 1065. Between October 1061 and February 1063, he had been duke of
Thessaloniki, and after Paradunavon, he became duke of Cyprus, between
1065 and 1067. He must therefore have been duke of Paradunavon between
1063 and 1065.96 If, on the other hand, there was only one duke of Para-
dunavon in 1065, the theory about the division of the province in 1059 is
false. The only change taking place in that year consisted in the intro-
duction of the official name Paradunavon for the province extending to
west all the way to Vidin. Some have viewed Romanos Diogenes’s 1067
title of duke of Serdica as proof for the split of Paradunavon, since he
was the supposed commander of the Western Paradunavon. After being
appointed duke of Serdica in 1066 by Emperor Constantine X, Romanos
Diogenes won a victory against the Pechenegs.97 He was not duke of the
Bulgarian theme, as that office was occupied between 1066 and 1067 by
Andronikos Philokales. The office of duke of Serdica does not concern
the command over that city. Apparently, Serdica was in 1067 the center
93 Zacos 1984, 424, nr. 956; Barnea 1986, 270, nr. 3.

94 Mărculeţ 2008, 194–195; Mărculeţ 2009, 171.
95 Božilov 2008, 97 does not identify him with Basil Apokapes.
96 For these stages of the career of Nikephoros Botaneiates, see: Zacos, Veglery 1972,
1464–1465; Cheynet 1983, 461–462; Cheynet 1985, 187, 192; Oikonomides, N. 1986 b, 89–90,
nr. 91, 92; Kühn 1991, 178, 212–213, 241; Karagiorgou 2008, 111–116, 128–129; Valeriev 2010,
425–427.
97 Zonaras, XVIII, 10.12 (ed. Büttner-Wobst, 684). See also Tăpkova-Zaimova 1973, 107
who argued that Romanos Diogenes was the duke of a theme based in Serdica. Dančeva-
Vassileva 2004, 24 mentioned the appointment of Romanos Diogenes as duke of Serdica in
order to battle the Pechenegs who had settled in that area. He was “strategos of the army of
the city and of the region”.
of a new military and administrative unit,98 but there is no proof that

that unit extended to the north as far as the Danube. On the contrary, the
relationship between Romanos Diogenes and Nikulitzas, who was strat-
egos of Larissa at the same time, strongly suggests that the new province
extended to the south.99 It is possible that parts of the theme of Bulgaria
were re-shuffled in 1059 to form a separate province based on Serdica,
as a measure against Hungary. The appointment of a new duke certainly
meant a larger number of troops in the region.
In conclusion, Isaac I Comnenos did not divide the theme of Paradu-
navon. He only detached a part of the Bulgarian theme to create the theme
of Serdica along the Morava valley, between Niš and Philippopolis.
Following Nikephoros Botaneiates, a new katepano of Paradunavon
was appointed at some point after 1065—Michael, who had the title of
vestarches. The first known seal of a man with that title, which is now in
the Zacos collection, was ascribed either to Michael, son of Anastasios,
who ruled the Danube region in 1045–1047,100 or to the general Michael
Dokeianos, who was killed by the Pechenegs in 1050.101 Recently, however,
two more seals have been published, one discovered during excavations
in Pliska, the other from a private collection from Varna.102 The title of
vestarches does not fit in the cursus of Michael son of Anastasios, who is
mentioned by Skylitzes as patrikios in 1047 and as magistros in 1057.103 This
was therefore a different person, somebody who occupied the office at a
time when katepanoi were typically given the rank of vestarches. Nothing
is known about the duration of his term, and it is even sure whether or
not he came immediately after Nikephoros Botaneiates.104
Another commander was katepano Symeon, a vestis. Three seals with
that title were found in Constantinople, Garvăn and Preslav, respectively.
This katepano of Paradunavon must be the same as Symeon, vestarches and
katepano of Adrianople and Mesembria, mentioned on seals dated to the
98 Diaconu 1970, 95–97.

99 Kekaumenos, 183 (ed. Spadaro, 220/221); Diaconu 1970, 96. Şesan 1978, 49 believed
the theme of Serdica to have extended as far as Vidin in the early 12th century, given that its
duke fought against the Cumans in that region in 1114. However, it is more probable that that
region was under the duke of Niš, an interpretation to which I will return below.
100 Zacos 1984, 300, nr. 602; Barnea 1986, 269–270, nr. 2; Diaconu 1992 b, 326–327;
Madgearu 1999 a, 424.
101 Stephenson 2000, 94; Stephenson 2003, 115. Accepted by Mărculeţ 2008, 184–188, who
considers that Michael the son of Anastasios was followed by Michael Dokeianos.
102 Ivanov 2008, 140–141, nr. 7; Jordanov 2009, 465, nr. 1378–1379.
103 Jordanov 1993, 143; Jordanov 2003 b, 64–65, 71; Božilov 2008, 96 (dated after 1072).
104 Mănucu-Adameşteanu 2001 a, 182 places his term between 1060 and 1070.
78 chapter two
1060s or 1070s.105 The katepanate of Mesembria was established as a shield

against the rebels from Paradunavon, namely after 1072 (see below). Symeon
became vestarches during or after his term in Paradunavon. The conclusion
is that Symeon could be placed as katepano in Paradunavon between the
years 1065 and 1072, before the appointment of katepano Nestor.106 Nicolae
Bănescu thought Symeon to have been katepano during the third decade of
the 11th century, and treated him as Emperor Romanos III’s brother-in-law.107
Later archaeological finds gave the impression that he was right about the
chronology. The seal from Garvăn was found in the filling of grave 46 from
the cemetery around the chapel. Ion Barnea dated the grave between the
middle of the 11th century and the middle of the 12th century.108 In this
view, the seal came from the underlying layers excavated for the grave pit.
However, the chronology of the cemetery in question is different from that
initially advanced by Ion Barnea and Eugen Comşa. The graves are most
certainly not of the same date with the adjacent chapel, which was most
likely erected just after 971. In fact the cemetery belongs to the last phase
of occupation on the site, which is dated to the 13th–14th centuries.109 The
archaeological context of the Garvăn seal is therefore in contradiction with
Bănescu’s idea and dating.
Nine identical seals of Constantine Theodorokanos were found on dif-
ferent sites in the southeastern part of Paradunavon (Preslav, Silistra,
Păcuiul lui Soare and other points). The rank and office of Constantine
Theodorokanos are not specified, but he was identified with a proedros
mentioned in the sources as having been arrested in 1077 or 1078 by the
usurper Nikephoros Bryennios. Petre Diaconu believed Theodorokanos to
have been the commander of Paradunavon, and explained that the men-
tion of his office was not needed as locals knew him well.110 If indeed he
was katepano of Paradunavon, he must have been in office between 1065
and 1072, but other three commanders are mentioned at that same time.
105 Ştefan, Barnea 1967, 332–334; Barnea 1980, 271–273; Jordanov 1993, 110, nr. 195–196, 144,
nr. 289; Jordanov, Tăpkova-Zaimova 1988, 124; Stephenson 2000, 94; Mănucu-Adameşteanu
2001 a, 182; Jordanov 2003 a, 137–138, nr. 58. 3; Jordanov 2003 b, 70; Jordanov 2009, 401–402,
nr. 1109–1110, 464–465, nr. 1377.
106 The same date at Božilov 2008, 96.
107 Bănescu 1946, 70–71. Followed by Stănescu 1968 b, 46; Barnea, Ştefănescu 1971, 93–94;
Mărculeţ 2005, 60.
108 Ştefan, Barnea 1967, 332–334.
109 See Diaconu 1969 a, 49 (the chapel); Diaconu 1978, 129 (the cemetery). The necropolis
is published by Ştefan, Barnea 1967, 367–373.
110 Diaconu 1992 c, 359–361; Jordanov 1993, 193–194, nr. 402–403; Mănucu-Adameşteanu
2001 a, 183; Jordanov 2006, 155–156, nr. 229–236.
Because the seals of Constantine Theodorokanos have been found within

the area most affected by the rebellion against Michael VII, his office of
commander of Paradunavon must pre-date that event. However, there is
no proof that he truly had that position, and only the discovery of new
seals specifying his title could clarify this problem.
The rebellion mentioned above was the result of the financial poli-
cies introduced by Nikephoritzes, Emperor Michael VII Dukas’ parakoi-
monenon (minister), who established a state monopoly over the wheat
trade in the capital, only to see prices going up. As his goald was appar-
ently to stock food in the provinces, the people from the cities of Para-
dunavon rose up in rebellion. They were also affected by Nikephoritzes’
decision to cease any cash payments to them, which had until then been
crucial in securing the peace in the region. Paul Stephenson has noted
that Nikephoritzes’s ultimate goal may have been to reform the economic
relations with Paradunavon by means of a strict control of the local mar-
kets, similar to that in the capital, and of a substitution of stipends from
the central government with local resources.111 This, on the other hand,
rhymes well with the notion of the security of Paradunavon preserved by
means of gifts and bribes to the Pechenegs, as opposed to costly military
operations. On the other hand, the continuator of Skylitzes explains that
the stratiotai were unhappy with their exclusion from the administration
of the theme (διοίκησις). This may refer to a shift of power in favor of civil-
ian aristocrats in towns, the same class that had lost previous financial
privileges. In 1072, the citizens of Dristra accepted as their ruler a Pech-
eneg chief named Tatrys or Tatós (called ἐξάρχον by Attaliates), who had
inherited the autonomy granted by Emperor Constantine IX in 1053 to the
Pechenegs (see chapter III. 2). Attaliates mentions that the inhabitants of
Dristra gave to the Pecheneg ruler the control of the frontier (τὴν ἐξουσίαν
τῆς ἂκρας).112 Nicolae Iorga and Nicolae Bănescu strove to demonstrate
that those events represented the first Romanian state in history,113 while
other historians excluded any participation of the Romanian population.114
111 Stephenson 2000, 99–100.

112 Skylitzes Continuatus, 166; Attaliates, ed. Bekker, 204–205 (ed. Pérez Martín, 150);
Zonaras, XVIII, 7.3–4 (ed. Büttner-Wobst, 713). The word ἂκρα has two meanings: “hill-
top fortress” and “border, frontier” (from the second one derives ὰκρίτες, frontier guard).
Although Anna Comnena, VII, 3.3 (transl. Sewter, 193) wrote about two akropoleis in Dristra,
the context of the sentence requires this translation made by Pérez Martín.
113 Iorga 1920, 33–46; Bănescu 1922, 287–310; Iorga 1937, 82–90.
114 For instance: Mutafčiev 1932, 236–259; Gyóni 1943–1944, 83–188; Tăpkova-Zaimova
1980, 332–335.
80 chapter two
The involvement of the Romanian population in the rebellion led by

Tatós results from a remark of Michael the Syrian, Patriarch of Antioch
(1126–1199), who in his chronicle mentions that Alexios I Comnenos, at
the beginning of his reign, “by his wisdom has freed their cities [of the
Greeks] from the Franks, Cumans, Serbs and Vlachs (Balakaye)”. As Victor
Spinei has demonstrated, Balakaye are without any doubt the Romanians,
and the context concerns the conflict against the Pechenegs from Para-
dunavon, two decades after the beginning of the rebellion.115 This means
that the rebellion of 1072 was a multiethnic affair involving the Pechenegs,
but also Greeks, Bulgarians and Romanians. Contrary to Iorga’s beliefs, the
name Tatós has no relation with the Romanian Tatu, and is most obviously
a Turkic name, with multiple analogies in the Turkic languages.116 Tátos’s
other name was Chalis, a word quite similar to Chalisi, the Turkic Muslim
group mentioned by John Kinnamos as having been settled between the
Danube and the Sava.117 The rebellion of 1072 was therefore no “national”
movement, but instead was no different from other centrifugal develop-
ments taking place in the 11th century in the Balkans.118 Different now
is the strong military position of the Pechenegs, who were thus able to
hijack the movement for their own interests.119
Emperor Michael VII Dukas sent to Dristra a man from his entourage,
the vestarches Nestor, and appointed him as the new katepano of Para-
dunavon, with the mission to restore the imperial authority. Nestor was
escorted by some citizens from Dristra who promised they would help
him regain control of the city. Those people must have been members of
a faction in the city which had remained loyal to the emperor. According
to Attaliates, Nestor was a former doulos of the emperor’s father, and of
Illyrian descent.120 The title doulos is confirmed by four seals, two of are
from Silistra and another from Bolyarovo (Elkhovo region).121 Both have
the legend τῷ ἀν(ϑρώπ)ῳ ἄνακτο(ς) Δούκα. The meaning is “the man of the
115 Michael the Syrian, XV.12 (ed. Chabot, 204); Spinei 2009, 119.
116 Bogrea 1971, 49–50; Drăganu 1933, 573–575; Necşulescu 1937, 135–141, 148–150. More-
over, that Tátos was a Pecheneg is explicitly mentioned by both Attaliates and Zonaras.
117 Kinnamos, III, 8; V, 16 (transl. Brand, 86, 186); Bogrea 1971, 35; Stephenson 2000, 192,
225.
118 Madgearu 2006, 213–221.
119 Necşulescu 1937, 122–151; Diaconu 1970, 100–116; Tanaşoca 1973, 64; Malamut 1995,
130–131; Spinei 2009, 119.
120 Skylitzes Continuatus, 166; Attaliates, ed. Bekker, 205 (ed. Pérez Martín, 151); Zonaras,
XVIII, 7.3–4 (ed. Büttner-Wobst, 713).
121 Oikonomides 1986 b, 93–94, nr. 95; Jordanov 1992 b, 238–239, nr. 14–15; Jordanov 2006,
306–308, nr. 521–522; Jordanov 2009, 652, nr. 2005–2007.
master Dukas” (ἄναξ is “lord”, and it was used in that sense to refer to an
emperor, a king, but also to the master of a house). The word ἄνϑροπος
denotes the condition of doulos of the emperor. It is important to note
that the same phrase (ἄνϑροπος αυτοῦ) was employed by Anna Comnena
to refer to the Bohemond of Taranto as her father, Emperor Alexios’
liegeman.122 On the other hand, according to the same Attaliates, Nestor
was οἰκειοτάτως. He therefore belonged to the group of oikeioi, the entou-
rage of the emperor, oikeios being identical with doulos. An oikeiotatos
was on the highest position in the oikos (the imperial court), almost at
the same level as the emperor’s kinsmen.123 Because the archaic name
“Illyrian” could refer only to Serbs or to Vlachs,124 Nestor must have been
a formerly local ruler from Dioclea or Macedonia, who had offered his
territory to Constantine X, or an aristocrat who had entered the impe-
rial service. According to the continuator of Skylitzes, Nikephoritzes had
previously confiscated his wealth. Revenge must therefore have been on
his mind, when, according to Attaliates, he joined the rebels, because of
being of the same origin (τῷ ὁμοτίμῳ τοῦ γένους). Some have interpreted
this passage as proof that Nestor was Bulgarian, under the assumption
that Dristra was inhabited only by Bulgarians.125 However, the Byzantine
author and other like him in Constantinople regarded people both from
the Danube regions and from Illyria as not quite Byzantine, either “half-”
or even “under-Byzantines,” or “inside foreigners.”126 For Attaliates, there-
fore, Nestor shared the same inferior condition with the inhabitants of the
city, who like him were on the periphery of the civilized world, although
not necessarily “Illyrian” stricto sensu.
The discovery of two seals of Nestor in Silistra shows that prior to his
joining the rebels, he had been in contact with them or with someone in
the city. Shortly after his appointment, he must have been based elsewhere
in Paradunavon, for he would not othwerise have used the phrase “man of
the emperor” on his seal. In other words, it took some time between his
122 Ferluga 1976, 407.

123 Ferluga 1976, 242; Cheynet 1990, 287–301; Stephenson 2000, 123–129.
124 For the idea that Nestor was Vlach, see Diaconu 1970, 103–104; Angold 1984, 98. The
Bulgarian origin of Nestor, which was advocated by Zlatarski, Mutafčiev, and other Bulgar-
ian historians, is contradicted by the fact that the archaic word employed for Bulgarians
in the Byzantine sources was “Mysian,” not “Illyrian.” In Attaliates’ work, Illyrikon refers to
Macedonia. Therefore, as Petre Diaconu pointed out, the use of the archaic word Illyrian in
reference to a Vlach is not impossible.
125 Tăpkova-Zaimova 1974, 672–673; Tăpkova-Zaimova 1980, 334.
126 Malamut 1995, 131.
82 chapter two
correspondence sealed with seals mentioning that phrase and his defec-
tion to the rebels’s side. Where was Nestor, when writing his letters? He
may have been in Isaccea or some other city in Paradunavon not affected,
at least initially, by the rebellion of 1072. Upon his arrival to Dristra, Nestor
realized that he could not appease the rebels. In control of the situation was
the Pecheneg chieftain Tatós. Under such circumstances, Nestor had no
choice but to put himself at the head of the rebellion against the emperor
and his personal enemy, Nikephoritzes. It is possible that he established
contact with the Bulgarian and Serbian rebels from the theme of Bulgaria,
who are also mentioned in 1072.127 Nestor appears to have shared power
with Tatós and led a Pecheneg army against Constantinople in 1075. There
were not enough troops in the capital to march against the Pechenegs, but
the march was unexpectedly interrupted when Nestor learned Tatós had
sent men to kill him. It is not altogether clear whether this is what hap-
pened or simply a rumor, possibly spread by someone in the service of the
emperor. In any case, Nestor and his men returned to Paradunavon (with
plenty of prisoners and booty from Macedonia and Thrace), and nothing
is known about him after that.128
For some years, the northern part of Dobrudja remained under the
authority of the empire. A provincial mint, in which coins were cast,
instead of being struck, operated in Isaccea until 1080. Cast coins do not
commonly appear in southern Dobrudja. During Michael VII’s reign, the
mint produced only issues with his name, which implies that Isaccea was
not in the hands of the rebels. On the other hand, continuous relations
with the capital result from finds of coins struck in the Constantinopolitan
mint on behalf of Nikephoros III Botaneiates (1078–1081), which appear in
greater numbers in northern Dobrudja than locally produced, cast coins.129
If the office of katepano of Paradunavon was still in existence between 1072
and 1081, then his residence could have been only in Isaccea. However, no
commander is so far known from any seal of a duke or a katepano, which
could be dated between 1072 and 1081. On the other hand, the authority
of whoever was running the northern part of Dobrudja still loyal to the
127 Sacerdoţeanu 1939–1940, 89–91; Ferluga 1976, 81–84; Fine 1991, 213–214; Stephenson
2000, 141–143.
128 Skylitzes Continuatus, 166; Attaliates, ed. Bekker, 208–209 (ed. Pérez Martín, 152–154);
Zonaras, XVIII, 7.4–6 (ed. Büttner-Wobst, 713). For the events from Paradunavon in 1072–
1091, see Gyóni 1944, 83–188; Stănescu 1966, 56–65; Tanaşoca 1973, 61–82; Malamut 1995,
129–141; Stephenson 2000, 98–102; Madgearu 2003 a, 49–56; Spinei 2006, 199–204.
129 Oberländer-Târnoveanu 1983, 261–270; Mănucu-Adameşteanu 1997, 119–149; Mănucu-
Adameşteanu 2001 a, 137–147; Mănucu-Adameşteanu, Poll 2006, 443–444; Curta 2006, 299.
central administration must have been quite limited, because it is certain

that after 1078, the rebellion spread to some northern cities as well. A
seal of Nikephoros Basilakios, duke of Dyrrachion, was found in Nufăru.130
He rose in rebellion against Nikephoros III Botaneiates in that same year
(1078) and even called the Pechenegs for assistance.131 The fortress in
Nufăru must have been a gateway to the Pecheneg lands, and it certainly
was one of the most important centers in northern Paradunavon.
A new katepanate was created to the south from the territory con-
trolled by the rebels, in Mesembria. Two individuals are mentioned as
commander of the new administrative unit: Symeon (the former katepano
of Paradunavon) and Valatzertes (Valtzar, the son of the Pecheneg chief
Kegen, who was in Byzantine military service, much like his father).132 Six
ek prosopou (deputy commanders) of the katepanate of Mesembria are
also known: Andronikos, Theodore, John, Leo, Kyriakos and Himerios
Solomon.133 Meanwhile, the military command of northern Paradunavon
after 1087 was in the hands of Gregory Mavrokatakalon. His seals found
in Silistra (4 specimens), Preslav (4 specimens), Isaccea (2 specimens),
Vetren and Oltina (one specimen each) describe him as patrikios, anthy-
patos and katepano, without any specific province. This may imply that
he was regarded as a commander of the entire theme of Paradunavon.134
Gregory Mavrokatakalon twice appears in Anna Comnena’s biography of
her father as Alexios’ adviser during the campaign against the Pechenegs
in the summer of 1087, which will be discussed in chapter III. 2.135 The
title of anthypatos and the office of katepano are undoubtedly indication
130 Barnea 1993 a, 61–65, nr. 9.

131 Attaliates, ed. Bekker, 298 (ed. Pérez Martín, 213); Zonaras, XVIII, 19.17 (ed. Büttner-
Wobst, 723); Malamut 1995, 133.
132 Skylitzes, Constantine Monomachos, 21 (ed. Thurn, 465; transl. Flusin, 385; transl.
Wortley, 435).
133 Seibt 1978, 299; Jordanov 1990, 110; Nesbitt, Oikonomides 1991, 173; Jordanov 1993, 193;
Seibt 2004, 258–259; Jordanov 2003 a, 40, nr. 10.3, 121, 122–123, nr. 47. 5; Jordanov 2009, 405,
nr. 1124, 1125, 45, nr. 1346. Treadgold 1995, 37, note 51 has maintained without any reason that
the katepanate came into existence soon after 1000.
134 I first advanced this interpretation in Madgearu 1999 a, 430, before the discovery
of the seal of Oltina, published by Chiriac 2001, 113–121 (who endorsed my point of view).
Mănucu-Adameşteanu 2001 a, 184; Mănucu-Adameşteanu 2001 c, 110 and Mărculeţ 2005, 74
rejected the idea. The seals from Bulgaria are published by Jordanov 1981, 94–97; Jordanov
1993, 162–163, nr. 328–330; Jordanov 2006, 280–282, nr. 442–451; Jordanov, Žekova 2007, 124,
nr. 323, Jordanov 2009, 486–487, nr. 1454–1461, 501–502, nr.1527–1527A, and those from Isac-
cea by Barnea 1990, 323, nr. 10; Barnea 1997, 357. Another seal from an unknown location is
in Paul Stephenon’s personal collection.
135 Anna Comnena, VII, 2.3; 3.4 (transl. Sewter, 219, 223); Barnea, Ştefănescu 1971, 148–149;
Jordanov 1981, 96–97.
84 chapter two
that he was the commander of a province. Ivan Jordanov believed how-

ever that Mavrokatakalon had no such position, as no province name is
mentioned.136 Nonetheless, there are examples of dukes of specific prov-
inces, whose seals do not mention the names of those provinces, i.e.,
the seal of the katepano of Bulgaria, Andronikos Philokales, which was
published by Ivan Jordanov himself.137 Another seal of Gregory Mavroka-
takalon as patrikios and strategos was found in Preslav, but it is certainly
earlier than 1081 and has no relation with the command of Paradunavon.138
Two other seals from Isaccea without titles are also dated too before
Gregory Mavrokatakalon’s promotion as katepano. They are still impor-
tant because they indicate the continuity of the military organization in
Isaccea between 1078 and 1081, when Gregory Mavrokatakalon was on the
staff of Alexios I Comnenos, who was at that time the commander of the
western army.139 Gregory Mavrokatakalon was appointed katepano of Par-
adunavon after the offensive of 1087, in which he served on the emperor’s
side. Knowing that Isaccea was under imperial control between 1087 and
1091, given that a seal has been found there of Adrian—Emperor Alexios’s
brother and the new commander of the western army140—it is possible
that Gregory Mavrokatakalon was a katepano residing in Isaccea. The
messages sent to Dristra and Oltina were perhaps intended to convince
the rebels to take Emperor Alexios’s side, after the failed siege of Dristra
(see chapter III. 2).
The last known duke of the Paradunavon theme was the proedros
Leo Nikerites. He had already been appointed to that office in May 1091,
when Alexios I Comnenos sent somebody in exile in that province.141 The
beginning and the end of his term are unknown, but he may have been
demoted after the Cuman invasion of 1095.142 Two seals of Leo Nikerites
have been found in Isaccea and Păcuiul lui Soare.143
136 Jordanov 1981, 97.

137 Jordanov 1993, 160–161, nr. 323.
138 Jordanov 1982 b, 96; Jordanov 1993, 169, nr. 354; Jordanov 2006, 280, nr. 441.
139 Barnea 1997, 357.
140 Barnea 1986, 273; Barnea 1997, 357. The message attested by this seal was most likely
related to the operations launched by Alexios I for the recovery of Paradunavon.
141 Anna Comnena, VIII, 9.7 (transl. Sewter, 268).
142 For this attack, see chapter III. 2.
143 Bănescu 1946, 93–95; Barnea, Ştefănescu 1971, 153; Jordanov 1995, 213–214; Mănucu-
Adameşteanu 2001 a, 184–185; Jordanov 2003 b, 67–68; Curta 2006, 302. The seals are pub-
lished by Barnea 1986, 272–273, nr. 5 and Diaconu 1992 a, 181–182. Other seals of Leon
Nikerites as vestarches and protoproedros link him to Bulgaria, but not to Paradunavon, in
No duke or katepano of Paradunavon is known for the period after the

Cuman attack of 1095, which of course does not necessarily mean that the
province disappeared. The office of katepano, however, was most likely
abolished as part of the military reform introduced by Alexios I Comne-
nos, whose purpose was a more effective defense with fewer forces. The
alternative was a shorter and stronger line of defense, placed on the Stara
Planina range, instead of the Danube. The territory between the Danube
and the new line of defense remained under Byzantine administration
until the Vlach-Bulgarian uprising of 1185–1186, but had no troops under
the command of any duke or katepano.144
This re-shuffling of the defense mechanism in the region revived the
importance of two ports, Mesembria and Anchialos, which were located
next to the easternmost segment of the Stara Planina range. Besides the
katepanate of Mesembria created during the secession of Paradunavon, a
newly established military structure was also meant support the defense
line on the Stara Planina range—the theme of Anchialos. That new theme
was detached from Thrace by Alexios I Comnenos in 1087, and included
the port towns of Anchialos, Mesembria, Sozopolis, and Develtos.145 The
rationale for the new theme seems to have been the same strategic concept
as that applied by John Tzimiskes to the situation following the conquest
of 971, when a strategy of Anchialos was created to enforce the defense
of the western Black Sea coast against possible seaborne attacks by the
Rus’. In 1087, there was of course another enemy, but the defense against
the Pechenegs coming from north also required a naval force capable of
moving on the Danube. It is no accident that within that same year, a fleet
commanded by George Euphorbenos was sent to the Danube against the
Pechenegs (see chapter III. 2). If the office of duke of Paradunavon was
indeed abolished after 1095, then that would imply that no tagmata were
stationed any more between the Danube and the Stara Planina Moun-
tains, and that troops still garrisoned in some of the forts on the river were
Jordanov 2006, 308–309. nr. 523–524; Jordanov, Žekova 2007, 64–65, nr. 138–139; Jordanov
2009, 163–164, nr. 400–401. Another one at Seibt 1995, 225–226.
144 Stănescu 1966, 65 believed that the absence of any mention in the sources indicates
that the Paristrian theme was not restored, or that it existed only for a brief time under
Alexios I. The establishment of another kind of military structure, the kleisura, should not
in principle be excluded, but there is no evidence for that so far. The seal of a kleisurarch
named Kalokyres Sarakinopoulos does not mention Paradunavon, as wrongly stated by
Meško 2006, 130 (for this seal found in Pliska, see Jordanov 2006, 364, nr. 639; Jordanov 2009,
520, nr. 1602).
145 Ahrweiler 1966, 188; Nesbitt, Oikonomides 1991, 169; Kühn 1991, 168.
86 chapter two
under the command of the duke of Anchialos. At any rate, the Danube
province was not abandoned, only defended by fewer forces stationed in
fewer towns.
At the end of this survey seeking to reconstruct the list of commanders
of Dristra and Paradunavon, it is perhaps important to summarize the
results as following:
Katepanate of Western Mesopotamia (971–after 1002):

Damian Dobromir, patrikios and anthypatos, duke of Thrace and Mesopo-
tamia (around 1002)
Theme of Ioannoupolis (Preslav)—between 971 and 986:
Peter, strategos of Ioannoupolis, basilikos protospatharios (971–?)
Leo Sarakinopoulos, strategos of Thrace and Ioannoupolis, basilikos pro-
tospatharios (?–975)
Theophanes, strategos of Thrace and Ioannoupolis, basilikos protospath-
arios (975–986)
Staurakios, strategos of Thrace and Ioannoupolis, basilikos protospatharios
(975–986)
Katepanate of Theodoroupolis (971):
Sisinios, katepano, protospatharios (971)
Theme of Dorostolon (972–975):
Leo Sarakinopoulos, strategos of Dorostolon, protospatharios
Arkadios, strategos of Dorostolon, protospatharios
Marianos, strategos of Dorostolon, protospatharios
Theme of Dristra (975–986, 1000–1018?):
Peter, strategos of Dristras, protospatharios
Basilakes, strategos of Dristras, protospatharios
David, strategos of Dristras, protospatharios
Theodore, strategos of Dristras, primikerios
Tzotzikios, strategos of Dorostolon, patrikios
Theme of Dristra or of the Paristrian cities (1018?–1059):
Katakalon Kekaumenos, archon ton para ton Istron poleon (katepano in
two seals), anthypatos, next vestis (1042–1045)
Michael, son of Anastasios, archon ton Paristrion poleon, patrikios
(1045–1047)
Constantine (. . .)polites, katepano of Dristra, patrikios (after 1047)
Leo Drymis, katepano of Dristra, anthypatos, patrikios (around 1050)
Romanos Diogenes, archon ton peri ton Istron poleon, patrikios (before and
after 1053)
Theme of Paradunavon (1059–1095?):

Demetrios Katakalon, katepano of Paradunavon, anthypatos, patrikios
(circa 1055–1062)
Theodore Pegonites, katepano of Paradunavon, patrikios, anthypatos,
vestis (circa 1055–1062)
Nikephoros Botaneiates, archon ton kata ton Istron poleon, magistros
(1062–1065)
Michael, katepano of Paradunavon, vestarches (inbetween 1065–1072)
Symeon, katepano of Paradunavon, vestis (inbetween 1065–1072)
Nestor, katepano of Paradunavon, vestarches (1072)
Gregory Mavrokatakalon, katepano of Paradunavon, patrikios, anthypatos
(after 1087)
Leo Nikerites, duke of Paradunavon, proedros (1091–1095?).146
It is clear from this list that the first commander with the title of katepano
(of the theme of Dristra) appears in 1042. Contemporary seals, however,
have only the title, without the name of the province. By contrast, the
title of archon appears only in the written sources, namely in the chron-
icles of Skylitzes and Attaliates. This was a title commonly used for rul-
ers of autonomous regions on the periphery of the Empire,147 although
in the case of the Danube province such an autonomy is out of question.
It may be that the chroniclers thought of that province as one inhabited
by “mixobarbarians.” In fact, Michael Attaliates employed the word mixo-
barbaroi in reference to the rebels of 1072, a usage interpreted as illus-
trating his scornful attitude towards that remote region on the frontier.
Some, however, have taken the phrase to imply the mixed ethnicities of
the population in the region,148 but a careful study of Attaliates’ vocabu-
lary suggests otherwise. In Classical Antiquity, mixobarbaroi was used
to describe the intermediary status of people “caught” between civiliza-
tion and barbarism.149 For educated authors such as Michael Attaliates
146 The discovery of new seals and the publication of new studies dedicated to the mili-
tary prosopography of the region considerably altered changed the list I have first proposed
in Madgearu 1999 a, 423–431, which was also accepted in general lines by Meško 2006,
131–133.
147 Ahrweiler 1966, 56–60.
148 Attaliates, ed. Bekker, 205 (ed. Pérez Martín, 150); Bănescu 1946, 101; Diaconu 1970, 101;
Bonarek 2007, 196–200.
149 Stănescu 1965, 45–49; Stănescu 1974, 404; Tanaşoca 1973, 66–74; Tăpkova-Zaimova
1975 b, 615–616; Malamut 1995, 131.
88 chapter two
or Anna Comnena (who also employed the phrase in reference to several

individuals), the Danube rebels were indeed semi-barbarians, in sharp
opposition to the order specific for true civilization, irrespective of their
ethnicity. In this context, even the use of the word archon may be seen
as a narrative strategy designed to shed a certain, negative light upon the
peripheral province. This was of course literary licence, and had nothing
to do with the official usage. In other words, the use of archon cannot be
interpreted as evidence that the province was an “archontate.”150
2. The Strategoi as City Commanders in the Theme

of Dristra/Paradunavon
The army of a theme was made up of units called turma, each garrisoned
in a different city. A seal found in Istanbul of John, turmarch of Paradu-
navon, was dated on iconographical grounds to the 1060s.151 Two seals
(one of them from Sofia) attest the existence of a turmach called Andron-
ikos in Arachilava. That town has been identified with Orjakhovo (known
in Antiquity as Appiaria).152 However, doubts about that identification
has also been raised when it turned out that the medieval fortress on that
site cannot be dated earlier than the 13th century.153
The garrisons of the most important cities were under the command
of strategoi.154 The title was initially reserved for commanders of themes,
but its meaning began to change in the early 11th century. The new strat-
egoi of cities are attested in Paradunavon by seals and a few written
sources. For example, Skylitzes mentions the capture of five strategoi—
John Dermokaites, Bardas Petzes, Leo Chalkotubes, Constantine Pterotos
and Michael Strabotrichares—by the Pechenegs in 1036,155 but there is
no mention of the cities in which they served as commanders. It is worth
observing though that the capture of no less than five strategoi at the
same time suggest a quite developed military organization in the cities
along the Danube.
150 Mărculeţ 2011, 55–68.

151 Zacos 1984, 275, nr. 530; Barnea 1986, 269; Jordanov 2003 b, 71.
152 Seibt 2004, 255–256.
153 Kuzev 1968, 29–34; Kuzev, Gjuzelev 1981, 120–124.
154 Ahrweiler 1960, 40–41, 46–50, 90; Krsmanović 2008, 76–77.
155 Skylitzes, Michael IV the Paphlagonian, 10 (ed. Thurn, 399; transl. Flusin, 330–331;
transl. Wortley, 376). For the attack, see Bănescu 1946, 73–74; Diaconu 1977 a, 1897; Malamut
1995, 118–119. A seal of a Strabotricharites found near Preslav may be that of Michael Stra-
botrichares (Jordanov 2006, 382, nr. 671).
Preslav was no more the center of a theme after the reconquest of 1000.
In that city resided a simple strategos (instead of a duke). The seals of three
strategoi of Preslav have been found in the archive discovered on the site
(the name of which was rendered as Περσθλαβα and Πραισθλαβας). First,
Constantine Karantinos was the brother-in-law of Emperor Romanos III
(1028–1034). Since it is known that he was duke of Antioch after 1030, he
must have been appointed before that to the otherwise inferior office in
Preslav. The term of the second strategos, Andronikos Dukas, could be
placed in the 1030s, while that of the third strategos, John, could be dated
only generally within the first half of the 11th century. He is known from
seven seals found in Preslav and another from Silistra. All those strategoi
were also protospatharioi.156
Pliska, the first capital of Bulgaria, was also the residence of a strategos
until its destruction dated to the 1060s. The archaeological excavations
carried out until 1999 have discovered 568 coins, the latest of which have
been struck for Emperor Constantine X Dukas. A few specimens from
1075–1080 and 1092–1118 are not sufficient for supporting the idea that
the town was rebuild after its destruction most likely during the invasion
of the Oghuz in 1065 (see next chapter).157 Before that, that a strategos
resided in Pliska results from the seal of Philotheus Frangopoulos, pro-
tospatharios and strategos.158
Another strategos resided in Varna, but only for a shorter while, namely
during the early conflicts in the 1050s with the Pechenegs in Paradunavon,
which required an increased protection of the coastline. Only one strategos
is known from Varna, namely Asoteos, patrikios and anthypatos, whose
seal (and term) is dated to ca. 1064.159 A strategos was in residence in
Vetren, a fortress where the seals of three military commanders have been
found: Tyrach, protospatharios and eparchos:160 the katepano Romanos
156 Jordanov 1987 a, 93, 95; Jordanov 1993, 146–150, nr. 291–303; Jordanov 2000, 138, nr. 19;
Jordanov 2003 a, 149–150, 152, nr. 63.2, 3, 6; Jordanov 2006, 137, nr. 192–195, 186, nr. 273–278;
Jordanov, Žekova 2007, 121–122, nr. 316, 317; Božilov 2008, 95; Yotov 2008 a, 347; Jordanov
2009, 469–472, nr. 1393–1403A, 1407–1413.
157 Jordanov 2000, 165.
158 Jordanov 2003 a, 143, nr. 62. 1; Jordanov, Žekova 2007, 120, nr. 312; Jordanov 2009, 467,
nr. 1384.
159 Seibt 2004, 254; Jordanov 2003 a, 45; Kostova 2006, 589; Kostova 2008 a, 214–215; Jor-
danov 2009, 408, nr. 1134.
160 Atanasov, Jordanov 1994, 41; Spinei 2006, 191. This is in fact the Pecheneg chieftain,
who had entered the Byzantine military service upon his surrender (see chapter III. 2).
90 chapter two
Diogenes;161 and Gregory Mavrokatakalon.162 By the same token, it would

also be possible to claim that there were strategoi in both Garvăn and
Krivina, given that both sites produced seals of katepanoi of Paradunavon
(Symeon and Demetrios Katakalon, respectively). Similarly, because of
several seals of commanders of other themes or of other officials have
been found in Nufăru, the existence of an 11th-century strategos in that
town is quite possible.163 Finally, Isaccea was definitely one of the most
important centers in Paradunavon, and must have had a strategos. The
city was the temporary headquarters of the theme during the secession of
Dristra in 1072–1091, and the discovery of many seals of high officials and
even of those of the emperors Alexios I and Isaac II Angelos164 suggests
that a high-rank military commander resided in that city.
One of the most controversial issues regarding the local military organi-
zation in the theme of Paradunavon is the identification of Presthlavitza,
a town known to have had a strategos from several seals discovered in
Bulgaria or preserved now in various collections:
– Aetios, protospatharios and strategos of Presthlavitza (eight seals found

in Preslav, one from Silistra, one from Topolovgrad-Melnitsa, one from
the Sliven region, another one in the Dumbarton Oaks collection);165
– Leo Pegonites, protospatharios and strategos of Presthlavitza (eight seals
from Preslav, one from Pliska, two from Silistra, one in the Dumbarton
Oaks collection, and another in the Hermitage collection, all dated by
the middle of the 11th century. The term of this strategos may have
coincided with that of Theodore Pegonites as katepano of Paradunavon.
Leo may have been Theodore’s father);166
161 Atanasov, Jordanov 1994, 37.

162 Jordanov 2006, 281.
163 Duke Basil Apokapes (Barnea 1986, 270–271; Barnea 1987 a, 194); Basilakes, duke of
Dyrrachion (Barnea 1993 a, 61–65); Constantine Anemas, spatharos and inspector of the
Armeniakon theme (Barnea 1986, 272); George Spanopoulos, vestarches (Barnea 1997 b,
97–98); Constantine Kokkinobepheos, protospatharios and krites (Barnea 1983, 268–269);
Clement, vestarches (Barnea 1996–1997, 190).
164 Barnea 1997, 355–359.
165 Diaconu 1987 b, 279; Nesbitt, Oikonomides 1991, 153, nr. 69.1; Jordanov 1993, 151–152,
nr. 304–307; Frankopan 2001, 79; Jordanov 2003 a, 148–149, nr. 63.1; Jordanov, Žekova 2007,
121, nr. 315; Božilov 2008, 95; Jordanov 2009, 467–469, nr. 1385–1392, 497, nr. 1495–1502.
Possible the same as Aetios, imperial protospatharios and strategos, without a city name
(Jordanov 1993, 165–166, nr. 334–340; Jordanov, Žekova 2007, 126, nr. 327).
166 Bănescu 1946, 41; Šandrovskaja 1982, 167; Oikonomides 1983, 1–2; Nesbitt, Oikono-
mides 1991, 179, nr. 78.4; Jordanov 1993, 153–154, nr. 309–310; Jordanov 2003 a, 152–153, nr. 63.7
(who thought he was the son of Theodore Pegonites); Jordanov 2006, 344–346, nr. 575–584;
– John Malesis, patrikios and strategos of Presthlavitza (one seal from

Dervent, one from Preslav, one from Silistra);167
– John Maleas, protospatharios and strategos of Presthlavitza, different
from the previous individual, according to the last studies on his two
seals found int Preslav;168
– Melias, protospatharios and strategos of Presthlavitza (three seals from
Preslav);169
– Omalis, protospatharios and strategos of Presthlavitza (one seal from
Preslav);170
Presthlavitza is also mentioned on seals of three kommerkiarioi (custom

officials)—Sergios, John, and Eustratios Romanos.171
There are two conflicting points of view regarding the location of
Presthlavitza. One of them sees that as a place different from Preslav, and
interprets the suffix -itza as a diminutive, thus equating Presthlavitza with
Little Preslav, a fortress mentioned only once in the account of the cam-
paign of 1000.172 According to Petre Diaconu, since such pairs of names
were given to neighboring places, Little Preslav must be somewhere in the
vicinity of Preslav.173 This is indeed correct, for the fortress is mentioned
in a description of a trip to Preslav, but the location is still unknown.
According to the second opinion, Presthlavitza is the Grecized version of
Jordanov, Žekova 2007, 122, nr. 318; Wassiliou-Seibt 2008, 134–135; Božilov 2008, 95; Wassil-
iou-Seibt 2009, 306–307, nr. 4; Jordanov 2009, 472, nr. 1414–1419, 510–511, nr. 1560–1565. He
was not a strategos of a small province called Preslav detached from the theme of Dristra, as
wrongly maintained by Mărculeţ 2005, 50; Mărculeţ 2005–2006, 312.
167 Oikonomides 1983, 2; Diaconu 1987 b, 279; Jordanov 1992 b, 232, nr. 5; Jordanov 1993,
171–172, nr. 359; Diaconu 1994 b, 355–356; Jordanov 2003 a, 151, nr. 63.5; Jordanov 2006, 268–
269, nr. 412–413; Božilov 2008, 95; Kostova 2008 a, 215; Jordanov 2009, 470–471, nr. 1406, 506,
nr. 1546. Mănucu-Adameşteanu 2001 a, 176 was wrong when making John Malesis the com-
mander of the province in 1030–1040. As strategos, in that period he could have been only a
city commander. Meško 2006, 132 has him also as a commander of Paradunavon.
168 Jordanov 1993, 152–153, nr. 308, 232, nr. 308 a; Jordanov 2003 a, 151, nr. 63.4; Jordanov
2006, 267–268, nr. 267; Jordanov 2009, 470, nr. 1404–1405.
169 Jordanov 1993, 154, nr. 311; Jordanov 2003 a, 154, nr. 63.9; Božilov 2008, 95; Jordanov
2009, 473, nr. 1421–1422; Jordanov 2011 b, 202.
170 Jordanov 1993, 154–155, nr. 312; Jordanov 2003 a, 154, nr. 63.8; Božilov 2008, 95; Jordanov
2009, 473, nr. 1420.
171 Oikonomides 1983, 2–4; Nesbitt, Oikonomides 1991, 178–179, nr. 78.1–3; Jordanov
1993, 150.
172 Šandrovskaja 1982, 168; Oikonomides 1983, 4–9; Nesbitt, Oikonomides 1991, 178–179;
Tăpkova-Zaimova 1993, 97; Wasilewski 1995, 199–200.
173 Năsturel 1965, 30–36 has supported its location in Păcuiul lui Soare. Stănescu 1970, 124
agrees that Little Preslav must have been somewhere in the vicinity of Preslav.
92 chapter two
the name of the capital of capital of Bulgaria, now given a feminine name.174
In other words, Presthlavitza is in fact Preslav. Scholars on both sides seem
to agree that Presthlavitza is the same as Pereiaslavetz mentioned in the
Russian Primary Chronicle as the town in which Svyatoslav wanted to set-
tle upon conquering the Danube region in 968 (see chapter I). Because the
Danube appears in that account in relation to Pereiaslavetz, some have
argued that its location must be sought on the southern bank of the river,
possibly at Nufăru (in northern Dobrudja, on the Saint George branch of
the Danube), the old name of which until the early 20th century was Pris-
lav or Prislava. Moreover, on 14th- to 16th-century maps, that same place
appears as Proslavitza.175 Those rejecting the identification of Presth-
lavitza with Preslav believe that Prislav/Nufăru was indeed Pereiaslavetz
or Presthlavitza.176 The discovery in Nufăru of a Byzantine fort dated to the
10th–11th centuries has only give more fuel to that intepretation. Its main
opponent, Petre Diaconu, believed that the phrase “na Dunaj” in the text
of the Russian Primary Chronicle should not be taken at face value, since,
according to him the Danube was only regarded as a landmark for a larger
area through which the river flows. The comparison between Byzantine
and Rus’ sources led Diaconu to the conclusion that Svyatoslav’s residence
was in Preslav. That the Rus’ chronicler chose Pereiaslavetz as a name for
what was otherwise Preslav may have been the result of confusion with
the Rus’ town Pereiaslavl’ (so Diaconu). No other Preslav is mentioned
in the Russian Primary Chronicle, which means that Pereiaslavetz must be
the name of the Bulgarian capital.177 Petre Năsturel also noted that it was
quite normal for Svyatoslav to chose as his residence, even if temporary,
the very capital city of the Bulgarian ruler whom he had just defeated, and
that, on the other hand, a residence on the southern bank of the Danube
would have exposed to Pecheneg attacks.178 Diaconu was therefore right
174 Diaconu 1987 b, 279–293. See also Jordanov 1983, 105.

175 Grămadă 1930, 241–242; Cihodaru 1978–1979, 286–288; Todorova 1986, 64; Baraschi
1991, 53–65.
176 Brătianu 1942, 149; Cihodaru 1968, 225–227; Cihodaru 1978–1979, 286–288; Oikono-
mides 1983, 7–9; Nesbitt, Oikonomides 1991, 178; Perkhavko 1994, 278–290; Wasilewski 1995,
200; Hanak 1995, 142–144; Franklin, Shepard 1996, 147; Soustal 1997, 122; Pintescu 1999, 72;
Stephenson 2000, 56–57; Meško 2006, 131; Strässle 2006, 110, 169; Yotov 2007, 323. Different
points of view at Bromberg 1937, 459 (Hârşova) Stokes 1962, 474–475 (Garvăn-Dinogetia).
Busetto 1996, 13 believes that Presthlavitza was Dristra, but that is impossible.
177 Diaconu 1965 a, 37–43; Diaconu 1987 b, 284–286.
178 Năsturel 1965, 25.
in identifying Pereiaslavetz with Preslav, the former being the name given
by the Rus’ chronicler to the Bulgarian capital.179
The archaeological excavations in Nufăru did not offer any support to
the idea that that was Svyatoslav’s residence, given that the ramparts of the
fort were erected on top of an earlier, non-fortified settlement, which has
been dated before 971. By the time Svyatoslav was in Bulgaria, therefore,
this was just a village, not an important town. Gold coins struck between
945 and 959 for Constantine VII and Romanos II were found in second-
ary positions, namely in sunken-floored buildings dated after 971.180 Those
coins have been in circulation for a few decades before their deposition.
If Pereiaslavetz was the Rus’ name for Preslav, then where was Presth-
lavitza, the residence of several strategoi mentioned in the 11th century?
In other words, was Pereiaslavetz the same as Presthlavitza? There are
again three possibilities: Preslav, Little Preslav, or some other city which
was not in the hinterland of Preslav. Petre Diaconu firmly believed that
Presthlavitza was Preslav, as he saw the former as the feminine form of
the latter name. In his opinion, which I have shared in some previous
studies,181 the strategoi of Presthlavitza were commanders of Preslav when
the city was no more the residence of a theme. I initially thought that
kommerkiarioi would not appear in Preslav when that town was on the
border of the empire, during the secession of Paradunavon in 1072–1091.
Instead, their presence makes more sense somewhere on the Danube, in
a point where trade was sufficiently active to be taxed, namely in a city
which was also the residence of a strategos. Trans-Danubian trade implies
the use of a ford. Besides, that office cannot be associated with anything
close to Dristra, which had its own kommerkiarioi.182
Rossina Kostova’s idea that Presthlavitza was in Isaccea (the ancient
Noviodunum)183 is consistent with those conditions, but there is another
solution which is perhaps even better, given the survival of the name:
Nufăru. During the 11th century, that was definitely an important fort and
179 Atanasov 1994, 116 rejected the identification of Pereiaslavetz with Nufăru.

180 Damian 1995, 217; Damian, Andonie, Vasile 2000, 67; Damian et alii 2003 b, 73.
181 Madgearu 1999 a, 429; Madgearu 2003 a, 53.
182 Three kommerkiarioi of Dristra are known: Andronikos (protospatharios), Dionysios
(spatharocandidatos) and John Spondyles. See Seibt 1978, 306, nr. 9; Nesbitt Oikonomides
1991, 151, nr. 65.2; Jordanov 2002, 84, nr. 8–11; Jordanov 2003 a, 66–67, nr. 23.6–8; Jordanov,
Žekova 2007, 110–111, nr. 285, 286; Jordanov 2009, 417, nr. 1172–1174; Jordanov 2011 a, 84,
nr. 35–38.
183 Kostova 2008 a, Table XXIII.
94 chapter two
possibly the residence of a strategos. It was also far enough from Dristra,
and placed right next to a key ford across the Danube. Al-Idrisi, who com-
piled his geographical work in 1154 mentions a city named Barisklafa near
a river and a swamp, at a distance of four days to the east from Daristar
(Dristra), through the wilderness. Al-Idrisi’s editor, Konrad Miller, has
proposed that the itinerary Daristar-Barisklafa-Disina-Akli followed the
Danube to the Delta, for he believed Akli to have been Kilia, which how-
ever did not exist in the 12th century. Nor does the description of Akli,
apparently located in a fertile land, south of the mountains, fits with Kilia
in the Danube Delta. Barisklafa and Disina were therefore not along the
Danube. From Daristar, the route indicated by al-Idrisi went to the south-
east, across the southern part of Dobrudja. Barisklafa has been viewed
as a misunderstood (or mispronounced) form of Presthlavitza. Petre Dia-
conu has tried to locate it in Pliska,184 but this is not possible, since Pliska
was not in existence in the mid-twelfth century any more, having been
destroyed by the Oghuz in 1065. Barisklafa could well be Little Preslav,
given that because Preslav itself appears as Migali Berisklafa in al-Idrisi’s
work. There are indeed marshy zones near Preslav, and the river in ques-
tion may be the Tiča. The location of Barisklafa and Migali Berisklafa on
al-Idrisi’s map, in the interior and next to a river, matches that identifica-
tion. However, Barisklafa cannot be the Presthlavitza mentioned on seals
of kommerkiarioi, because no custom points are known to have existed
in the interior. Moreover, this location was too close to Preslav, which
already had a strategos. Al-Idrisi does not therefore mention Presthlavitza,
the residence of strategoi and kommerkiarioi (present-day Nufăru), for it
has nothing to say about the northern part of Dobrudja.185
In conclusion, several cities were ruled by strategoi in the Dristra/
Paradunavon province: Preslav, Pliska, Varna, Isaccea, Presthlavitza
(Nufăru), Garvăn, Krivina, and Vetren. It is possible that all major forts
and towns had their own strategoi.
184 Al-Idrisi (ed. Jaubert, 386; ed. Miller, 129); Tomaschek 1886, 301–302, 311; Grămadă
1930, 242; Brătianu 1942, 147–148; Diaconu 1965, 50; Năsturel 1965, 27; Diaconu 1968, 359–361;
Diaconu 1976 a, 430–431; Baraschi 1981, 317–318.
185 The identifications of several names from this al-Idrisi’s work with places in north-
ern Dobrudja such as Halmyris or Argamum (Brătescu 1920, 23–31; Barnea, Ştefănescu 1971,
163–166; Cihodaru 1978–1979, 289) are simply wrong. Neither Halmyris (Murighiol) nor
Argamum (Jurilovca) were inhabited any more in the 12th century.
3. The Theme of Sirmium and the New Bulgarian Theme

on the Middle Danube
The Byzantine presence in the Middle Danube region became dangerous

for the new Hungarian kingdom seeking expansion to the south, in the
direction of the Danube, trade along which had been revived in the early
11th century. During his first reign (1038–1041), Peter Orseolo, king of Hun-
gary, may have offered support to the rebels of the Bulgarian nobleman
Peter Delian, who claimed to be the son of the former emperor Gabriel
Radomir and of his Hungarian wife, the sister of King Stephen I. The Bul-
garian rebellion has started as a protest against the removal of the fiscal
and religious privileges established by Basil II in 1018. By 1040, the rebels
were already in control of important cities such as Niš and Skopje, and
were moving on Thessaloniki. No Byzantine army was able to stop them,
but a supposedly heir of the Bulgarian dynasty appeared in the person
of Alusian, the son of the last emperor John Vladislav, to challenge Peter
Delian. After joining the rebels, Alusian blinded Peter Delian and swiftly
made peace with Emperor Michael IV.186
Hostilities with Hungary opened in 1059, when Belgrade was briefly
taken by Hungarian troops, before Isaac I Comnenos’ offensive from Ser-
dica in that same year. The Hungarian attack was justified by the need to
punish the Byzantines who had encouraged Pechenegs from the themes
of Dristra and Bulgaria to raid southern Hungary.187 Another Hungarian
attack took place in 1071, this time to punish the Pechenegs themselves,
who had invaded the region around Sirmium. The Pechenegs were appar-
ently encouraged to do so by the Byzantine commander of Belgrade,
dux Nicota (Niketas). This duke was most likely the commander of the
Sirmium province, and not just the strategos of Belgrade.188 Despite the
concentration of Byzantine and Pecheneg forces, Sirmium, Niš and Bel-
grade were taken by the Hungarians, the latter town after a siege that
lasted three months.189 It is important to note that on both occasions the
186 Prokić 1906, 49–50; Ferluga 1976, 341, 385; Iljovski 1991, 98–99; Fine 1991, 204; Makk
1999, 36, 47; Curta 2006, 283–284; Madgearu 2008, 66; Révész 2009, 84–88.
187 Attaliates, ed. Bekker, 66–67 (ed. Pérez Martín, 51–52); Wasilewski 1964, 478–479;
Moravcsik 1970, 62–63; Makk 1990, 17; Shepard 1999, 67.
188 As considered Bănescu 1946, 36–37.
189 Kinnamos, V. 8 (transl. Brand, 171); SRH, I, 373, 374, 377; Chalandon 1912, 54; Wasilewski
1964, 480–481; Moravcsik 1970, 64–65; Makk 1990, 17–18; Popović 1991, 170, 173; Kühn 1991,
235; Makk 1999, 64, 66; Shepard 1999, 69; Kosztolnyik 2002, 383; Spinei 2006, 187; Stojkovski
2009, 383–387.
96 chapter two
Hungarian attacks were in retaliation for Pecheneg raids from the Byzan-
tine Empire.
The victory of the Hungarian king Salomon (1063–1074) was made pos-
sible by the difficult situation in the Empire in 1071, the year of the major
defeat of the Byzantine army at Mantzikert at the hands of the Seljuk
Turks. Following that, in 1072 the rebellion of Paradunavon started, which
led to the secession of a large part of the province, and in that same year
began another rebellion is mentioned in the theme of Bulgaria. The Bul-
garian rebels were led by George Vojtech from Skopje, who bestowed
the title of emperor on Constantine Bodin, the son of the prince Mihailo
Vojslav of Dioclea (1046–1081). Moreover, George took the name Peter in
reference both to the 10th-century Bulgarian emperor Peter and to the
rebel Peter Delian. The rebels advanced to Niš and Ochrid, but they were
defeated after a few months.190 It is very likely that the rebels had Hun-
garian support, much like in 1127 and 1149, when Serbs rebelled against
Byzantium at the time of the Byzantine-Hungarian wars.191
The theme of Sirmium theme ceased to exist in the aftermath of the
Hungarian attack of 1071. Niketas was its last commander. The region
to the east and south from Belgrade remained under Byzantine control,
but within the theme of Bulgaria. This results from the analysis of the
seal of Nikephoros Batatzes, duke of Bulgaria, which was discovered in
Moroviskos,192 a town which had previously been within the theme of Sir-
mium (see below). It is known that this dignitary was in office after 1075.193
The message sent to Moroviskos was addressed to a strategos under the
orders of the duke of Bulgaria.
Peaceful relations between the Byzantine Empire and Hungary were
established in 1075 by Michael VII and Géza I (1074–1077), and sealed
by means of the latter’s marriage with Synadene, the emperor’s cousin
of the emperor.194 The empire was looking for allies in the aftermath of
the catastrophe at Mantzikert and the numerous mutinies in the Balkans.
190 Sacerdoţeanu 1939–1940, 89–91; Ferluga 1976, 81–84; Fine 1991, 213–214; Stephenson
2000, 141–143.
191 However, Shepard 1999, 70 does not exclude that the possibility of the attack against
the theme Sirmium taking place heme happened before the rebellion in the Bulgarian
theme.
192 Maksimović, Popović 1993, 127–128.
193 Laurent 1969, 144–147; Kühn 1991, 232; Jordanov 2009, 410, nr. 1146; Jordanov 2011 c,
172–174, 182.
194 Moravcsik 1970, 65–69; Shepard 1999, 72–74; Stephenson 2000, 188–189; Cheynet
2002, 7–10.
Under the circumstances, it was important to turn the northern neighbor

into an ally.195 Géza I’s favorable attitude towards Byzantium helped the
case. The emperor conceded to Géza the peripheral region of Sirmium,
in exchange for security for Belgrade down the stream. The region given
to Hungary between Sava and Danube was known as Sirmia or Franko-
chorion.196 Since Roman antiquity, the city of Sirmium (modern Sremska
Mitrovica) had been an important economic, military and religious cen-
ter, revived after 1018. Its abandonment meant a decreasing Byzantine
presence in the Middle Danube region, and allowed Hungary to obtain
a valuable strategic position, which may be used as a launchpad for any
future wars with Byzantium.
Božidar Ferjančić and Ferenc Makk believed that the Byzantines
recovered Sirmium in 1075. Their main argument was the 1980 discov-
ery in Mačvanska Mitrovica of a seal of Alexios I Comnenos, who was
then megas domestikos (commander of the western army). The seal could
therefore be dated between the late 1078 and April 1081.197 But the gen-
eral situation in the Empire at that time was so difficult that it is hard
to imagine the recovery of a remote town such as Sirmium was a real
concern for the imperial administration. A different explanation was later
offered for the presence of Alexios’ seal in Mačvanska Mitrovica: the mes-
sage had been sent to the Byzantine bishop who continued to reside in
Sirmium, and more exactly in Mačvanska Mitrovica, on the southern bank
of the Sava, after the Hungarian conquest of 1071.198 However, it has not
yet been noted that this seal is one of many attached to messages the
megas domestikos Alexios sent out in 1080 during his war against the Nor-
mans led by Robert Guiscard. Such messages called for the assistance of
the commanders residing in Preslav, Beroe, Tărnovo, Zlati Voivoda and
Melnitsa.199 Does that then mean that Alexios was calling for help from a
commander in Sirmium? Since there is no other proof that Sirmium was
in Byzantine hands after 1071, the message may have been sent to the
Hungarian king Ladislas I (1077–1095).
F. Makk has also maintained that the region around Belgrade was occu-
pied again by the Hungarians in 1091, during Ladislas I’s campaign into
Croatia. Makk believed this region to have been the Messia mentioned in
195 Cheynet 2002, 7.
196 Stephenson 2000, 141, 191–193; Cheynet 2002, 11.
197 Ferjančić 1982, 47–52; Makk 1989, 125; Makk 1990, 18; Maksimović, Popović 1990, 216.
198 Stephenson 2000, 191.
199 Frankopan 2011, 99–106.
98 chapter two
Ladislas I’s title as rendered by a charter of 1091, but Messia was in fact
Bosnia, a land which was indeed conquered by Ladislas I in the course of
his war against Croatia. This was in fact the interpretation favored by the
Hungarian historian György Györffy, to whom Makk also referred, even
though he tried to identify Messia with the formerly Roman province of
Moesia Prima, an identification for which there is however no evidence.200
There is in fact no proof that any specific knowledge of the location of the
formerly Roman province had survived until the 9th–13th centuries. For
instance, both in Bishop Pilgrim of Passau’s forgeries of 971–991 and in the
chronicle of Simon of Keza, the name Moesia is applied to Moravia.201
Everything, therefore, point to the conclusion that Sirmium remained
in Hungarian hands after 1071, while Belgrade was under Byzantine rule.
When the crusaders showed up in 1096, they are said to have entered the
Byzantine territory in Belgrade. In fact, according to Albrecht of Aachen,
there was a military commander of the theme of Bulgaria residing in
that city: duce, Nichita nomine, principe Bulgarorum et praeside civitatis
Belegravae. The same man appears in the chronicle of William of Tyre,
as Bulgarorum dux.202 He was a namesake of the 1071 commander of
the theme of Sirmium in 1071 and of the protoproedros Niketas Karykes
or Karikes, who was a duke of the theme of Bulgaria. Following Robert
Guilland, some have mistaken Niketas Karykes for Leo Nikerites, the last
known commander of Paradunavon,203 even after the correct reading of
their respective seals was published by Günter Prinzing.204 Only Ivan Jor-
danov expressed doubts about the identification of duke Niketas men-
tioned in the western sources with Niketas Karykes, on the grounds that
the title of princeps could not have been the equivalent of protoproedros
and that the Niketas of the western chroniclers is not named Karykes as
well.205 Such arguments are not sufficient, however, for rejecting the pos-
sibility of duke Niketas being Niketas Karykes. The latter is known to have
been the commander of the Bulgarian theme and of the city of Belgrade
at the same time, which suggests a change of organization taking place
at that time, no doubt in order to improve the defense. During the last
200 Makk 1994 a, 65–67; Makk 1999, 79–84.

201 Eggers 1995, 390–391.
202 Albertus Aquensis, I, 7–9, 13 (ed. Edgington, 12/15, 16/17, 18/19, 28/29); William of Tyre,
I, 18, 19 (ed. Babcock, Krey, I, 98, 100); Runciman 1949, 211–213; Popović 1991, 170.
203 Mănucu-Adameşteanu 2001 a, 185.
204 Bănescu 1946, 149; Kühn 1991, 231; Prinzing 1995, 220–224.
205 For the seals see Jordanov 2003 a, 50–51, nr. 19.2; Jordanov 2006, 191–192, nr. 288; Jor-
danov 2009, 410, nr. 1145; Jordanov 2010, 184–185; Jordanov 2011 c, 170–172, 183.
decades of the 11th century, a duke of Skopje (Skopion) is mentioned, first

in 1078 (Alexander Kabasilas) and then around 1100 (John Comnenos).206
This suggest that the large theme of Bulgaria was divided into two smaller
provinces, one under a duke residing in Skopje, the other, which was still
called Bulgaria, under a duke residing in Belgrade. A similar action had
taken place in 1059, when the duchy of Serdica was detached from Bul-
garia. The northern part of the theme of Bulgaria was more exposed to
Hungarian attacks, and it may be on that basis that Pecheneg groups were
moved to Belgrade and its hinterland in the late 11th century. The Pech-
enegs were in military service under the command of the duke and they
had small boats (naviculas) to monitor traffic on the Sava and the Danube
river, as indeed they did with the crusaders of 1096.207
At some point before 1114, Alexios I Comnenos brought another change
to the administrative organization in the region, as he moved the head-
quarters of the Bulgarian theme from Belgrade to Niš (Nisos, the ancient
Naissos). That the province extended at that time as far north as Braničevo
and Belgrade is mentioned by Anna Comnena.208 The retreat inside of the
province, into the mountain region may reflect the same strategic con-
cept which had been applied to the eastern parts of the Balkans, when
the emperor moved the defense line on the Stara Planina range of moun-
tains. A duke of Niš under Emperor Alexios I, named Nikephoros Deka-
nos, is attested by three seals.209 Tadeusz Wasilewski believed that the
duchy of Niš came into being after 1071, but at that time, as well as in
1096, the residence of the commander of the theme of Bulgaria was in
Belgrade, not in Niš. It is in the former, not in the latter, that duke Nike-
tas is said to have taken refuge when the crusaders of Peter the Hermit
crossed the river. At that moment, Niš belonged to the southern Bulgarian
theme, for the crusaders led by Walter the Penniless are said to have were
encountered there another dux et princeps Bulgarorum.210 This strongly
suggests that the transfer of the theme’s headquarters from Belgrade to
Niš took place sometime after 1096. The subsequent history of the duchy
of Niš remains unclear until two other dukes are mentioned, one in 1147
206 Bănescu 1946, 149; Kühn 1991, 231–233; Nesbitt, Oikonomides 1991, 98, nr. 30.
207 Albertus Aquensis, I, 8 (ed. Edgington, 18/19); Marjanović-Vujović 1974, 183–188.
208 Anna Comnena, XIV, 8.1 (transl. Sewter, 462); Wasilewski 1964, 481; Stephenson 2000,
152; Jordanov 2010, 177.
209 Jordanov 2003 a, 131–132, nr. 55.1; Jordanov 2006, 125–136, nr. 163–164; Jordanov 2009,
461–462, nr. 1367–1367 A; Jordanov 2010, 179–180.
210 Albertus Aquensis, I, 6 (ed. Edgington, 10/11).
100 chapter two
(Michael Branas)211 and the other in 1153 (the future emperor Andron-
ikos Comnenos).212 Both commanders appear in the sources at the time
of the military confrontation with Hungary. However, because it is certain
that Braničevo was under Byzantine administration in 1127,213 it may have
belonged to the duchy of Niš.
Before the headquarters of the Bulgarian theme were established in
Belgrade, that city had its own strategos, who is mentioned in 1026, in the
account of the pilgrimage of St. Simon of Trier returning from the Monas-
tery Saint Catherine in Sinai. When he got to Belgrade, he was prevented
by the princeps civitatis to enter Hungary (pervenientes itaque usque ad
civitatem Bellegradam, quae est in confinio Bulgariorum atque Ungari-
orum, a civitatis infelicissimo principe prohibitus est nobiscum transire).214
The princeps was most likely the strategos of the city. Another strategos
resided in Braničevo, a town which emerged in the 9th century on the site
of the ancient city of Viminacium, at the confluence between the Mlava
and the Danube. Braničevo grew quickly in the 11th century, and became
the main center of the defensive system in the region.215 Several seals sug-
gest the existence of a strategos residing in Moroviskos (Moravon), a for-
tified settlement at the mouth of Morava (today Dubravica), which was
occupied continuously between the 10th and the 12th century.216 The forts
of Moroviskos and Braničevo were meant to defend the Morava valley,
the axis of the main road to Thessaloniki and Constantinople. Finally, the
existence of a strategos in Niš is proved by the seal of the protospatharios
Nikephoros Lalakon, which is dated after the mid-11th century.217
In comparison with the katepanate of Paradunavon, the region to the
west from Vidin appears as less important for the Byzantine strategy, at
least until Hungary began to expand in Belgrade-Braničevo area. This area
will become a sensitive issue in the 12th century and later, when the place
of the Byzantine Empire in confrontations with Hungary will be taken by
the Vlach-Bulgarian Empire.
211 Kinnamos, II. 13 (transl. Brand, 60); Bănescu 1946, 160; Stephenson 2000, 259; Jordanov
2010, 186.
212 Niketas Choniates, Manuel Comnenos, 3 (ed. Dieten, 101; transl. Magoulias, 58);
Bănescu 1946, 43, 161; Urbansky 1968, 80–81; Stephenson 2000, 233–234.
213 See chapter III, footnote 218.
214 Eberwinus, 210; Wasilewski 1964, 478; Bálint 1991, 104; Stephenson 2000, 124 (who sup-
posed that he was a “local magnate” who ruled the city); Holmes 2005, 425. For the date, see
Klein 2005, 187.
215 Popović, Ivanišević 1988, 125–179; Maksimović, Popović 1990, 222; Milošević 1991,
187–195.
216 Maksimović, Popović 1993, 127–129.
217 Nesbitt, Oikonomides 1991, 100 (nr. 32.2); Stephenson 2000, 124; Jordanov 2010, 178.
Chapter Three
The evolution and function of the Danube frontier

of Byzantium (1000–1204)
1. The Fortifications
After the war of 968–971, the strategic target of Emperor John Tzimiskes’
military policies on the Danube was to prevent any future attacks of the
Rus’ by means of both diplomacy and fortifications. The first diplomatic
action in that direction was the alliance with the Pechenegs, right after
Svyatoslav’s capitulation,1 even though relations with the Pechenegs
would develop in a way different from that that John Tzimiskes had envis-
aged. Those nomadic warriors would in fact turn into the next threat to
the security of the frontier. For the moment, however, and for the next
fifty years or so, the Pechenegs in the area north of the Black Sea and in
Moldavia acted as shield against the Rus’.
On the other hand, the building activity along the Danube consisted of
the restoration of several old Roman forts and the erection of new ones
at strategic points. The forts along the Danube were initially only for gar-
risons of soldiers, but they gradually grew into larger settlements, inhab-
ited by civilians, as well as the military. Houses began to be built next to
the ramparts, as in Garvăn and Nufăru.2 When possible, the Byzantine
builders reused the remains of the old Roman structures, especially the
ramparts. The new walls were made of ashlar with a core of stones mixed
with mortar. Their thickness varies between 2.5 and 3.5 meters. Tenth-
to eleventh-century forts are smaller in area than the Roman forts they
commonly overlap, which speaks volumes about the smaller number of
soldiers in their garrisons.
Silver coins (miliaresia worth 1/12 of a gold coin) struck for John Tzi-
miskes and found in (6 specimens), Păcuiul lui Soare (2 specimens),
Dervent (one specimen), Oltina (3 specimens), Constanţa (2 specimens),
Vetren (one specimen), and Valul lui Traian (one specimen) betray the
1 Shepard 1985, 253 remarked that “the Byzantines intended to seal the Danube frontier
by means of a considerable military force” against future Rus’ attacks.
2 Barnea 1971, 354 (Garvăn); Baraschi, Moghior 1983, 137 (Nufăru).
102 chapter three
presence of the military (even though Valul lui Traian is a non-fortified

settlement), as such coins were distributed as salary for the soldiers.3
There are many more such coins in forts than in cities in the interior, most
likely because they were used to pay the army during or shortly after the
campaign of 971.4 A lead sheet was found in Silistra, which was used for
the production of silver coins, no doubt indicating the presence in Dristra
of the mobile mint accompanying the army.5 The cluster of coin finds in
the hinterland of Dristra mirrors the military operations discussed in the
first chapter of this book. As a matter of fact, the presence of the Byzan-
tine coins in the aftermath of the Byzantine conquest of the northern and
northeastern Balkans is an indication of the presence of the military, not
of active trade relations.6
Archaeology is now in a position to provide sufficient data for estab-
lishing a rough chronology of the defense system on the Danube frontier.
Soon after the conquest the reconstruction of the Roman forts and the
building of new ones began. The first Roman fortress to be reoccupied
after 971 was that Mahmudia (known as Salsovia in the Roman period).
Coins and other artifacts recovered from field surveys or known as stray
finds point to the reuse of the formerly Roman fort (area: 150 × 120 m)
since at least the reign of Basil II. A miliaresion struck for that emperor
may indicate the existence of a garrison, but it is not certain that the coin
was actually found in Mahmudia. Other bronze coins are dated up to the
reign of Romanos IV (1067–1071). Nothing is known about a possible res-
toration of the old precinct (the results of the excavations from 2004 are
limited to Late Roman finds).7
The fortress of Nufăru (fig. 5) was built immediately after 971, as indi-
cated by the analysis of coin finds,8 on the site of the Late Roman fort
that operated there until the early 7th century. The Byzantine rampart
3 Damian 1995, 218; Custurea 2000, 136, 137, 148, 152, 157; Mănucu-Adameşteanu 2001 a,
35; Custurea, Matei 2002–2003, 433–438; Poenaru-Bordea, Ocheşeanu, Popeea 2004, 133,
134; Custurea, Talmaţchi 2011, 292.
4 Metcalf 1976, 96.
5 Damian 1995, 219.
6 Metcalf 1976, 89–97.
7 Oberländer-Târnoveanu 1980 b, 66–70; Vasiliu, Mănucu-Adameşteanu 1984, 150 şi nota
56; Mănucu-Adameşteanu 1992, 399–400; Damian 1995, 220; Custurea 2000, 143; Mănucu-
Adameşteanu 2001 a, 65–66, 414 (tabel 16); Oberländer-Târnoveanu 2003, 386; Topoleanu
et alii 2005, 215–216; Mănucu-Adameşteanu 2010 a, 419–449; Mănucu-Adameşteanu 2010 e,
225–226.
8 Mănucu-Adameşteanu 1998, 80; Custurea 2000, 146–147; Mănucu-Adameşteanu 2001 a,
63–64; Damian, Andonie, Vasile 2003, 239.
the danube frontier of byzantium (1000–1204) 103
5. The fort and settlement of Nufăru (after Damian et alii 2012, 195)
is 2.6–2.7 m thick and encloses an area of about 6 hectares. Five towers

have been so far identified on the eastern, northern and western sides.
Three of them are separated from the rampart (on the eastern side, at the
north-western corner, and on the northern side), while two other towers
are embedded into the the rampart (on the western side).9 The fortress
had a harbor from the very beginning. At the site, which to this day is used
as anchorage, there was a 14.2 m-long jetty—a wall attached to the rock
and made of large stones bonded with mortar. The jetty was 2 m thick
and must have been built in the late 10th century, at the same time as the
first rampart. The connection with the fortress was made by an entrance
through the northern side of the rampart.10 Assuming that Presthlavitza
was at Nufăru, the harbor must have also serves as a customs point with
an office for kommerkiarioi. The presence of the Varangians is indicated by
the wooden buildings and a timber pavement, 5.7 m long and 1.2 m wide,
9 Damian, Andonie, Damian 1994, 168–170; Mănucu-Adameşteanu 2001 a, 61–62;

Damian et alii 2003, 72–75; Damian 2005, 177–180; Damian, Vasile et alii 2010, 31–33.
10 Damian, Andonie, Vasile 2000, 67; Damian, Andonie, Vasile 2002, 217; Damian et alii
2003, 74; Damian et alii 2007; Damian et alii 2012; Damian, Andonie, Vasile 2003, 213–216;
Mănucu-Adameşteanu, Poll 2006, 436; Damian, Vasile et alii 2010, 33; Damian, Mărgineanu-
Cârstoiu et alii 2011, 91.
104 chapter three
which was uncovered in 2002 and 2003. It remains unclear when the
Varangians may have come to Nufăru, either before or after the establish-
ment of the Byzantine fortress.11 The large number of coin finds12 and of
other artifacts, particularly imports and luxury goods indicate a flourish-
ing life in this fortified settlement during the 11th century, when the occu-
pied area was extended up to 800 m farther on the bank of the Danube.
The suburb was however abandoned and turned into a cemetery after the
Pecheneg attack of 1122.13
At Tulcea, the old Roman fort of Aegyssus located on the hill known as
“Hora” was reoccupied during the Byzantine period. The Roman polygo-
nal precinct was restored (a segment of the new wall was found in 1993).
Houses and huts were discovered inside the fortified area. At first, archae-
ologists believed that the fort has been established under John Tzimiskes,
but the analysis of the coin finds suggests a date after 1000 for the foun-
dation of the fort (anonymous folles of the A1 type, which are dated to
the reign of John Tzimiskes, have not been found on the site). The fort
remained in use until the reign of Constantine IX (1042–1055). In other
words, the Tulcea fort (whatever its Byzantine name) was abandoned
after the Pecheneg attack of 1047.14
The Roman ramparts in Isaccea (ancient Noviodunum) were most cer-
tainly restored soon after 971.15 The seal of a commander named Niketas
Hagiozacharites may be dated to this early period, since he is to have been
captured by the rebel general Bardas Phokas in March 979.16 At any rate,
the seal shows the presence of the military in Isaccea before 979, which
indirectly suggests that the fort may have been restored for the occasion.
As mentioned in the first chapter, it is also certain that Isaccea was used
by the Byzantine army after 990. The coin finds also point to the occu-
pation beginning during the reign of John Tzimiskes.17 On the northern
side of the fort along the Danube (excavated for 250 m), were six towers
11 Damian, Andonie, Vasile 2003, 214–215; Damian, Andonie, Vasile 2004, 219; Yotov
2007, 323; Damian, Vasile 2011, 275–290.
12 Mănucu-Adameşteanu 1991 b, 497–554; Mănucu-Adameşteanu 2001 a, 413 (table 15).
13 Damian, Andonie, Vasile 2003, 243.
14 Vasiliu, Mănucu-Adameşteanu 1984, 149; Mănucu-Adameşteanu 1995 b, 363; Cus-
turea 2000, 155; Mănucu-Adameşteanu 2001 a, 58–59; Oberländer-Târnoveanu 2003, 389;
Damian 2005, 147; Mănucu-Adameşteanu 2010 a, 295–418; Mănucu-Adameşteanu 2010 c,
223, 228–230.
15 Mănucu-Adameşteanu 2001 a, 55–59.
16 Barnea 1987 b, 81.
17 Custurea 2000, 141–142.
6. The fort of Dinogetia (after Barnea 1980, Plate XV)
of rectangular plan (each was 9.5 × 10 m). On the eastern side of the late
Roman rampart a new 3 m thick curtain was built in the Byzantine period,
at a distance of 12 m from the 6th century wall (which was 285 m long), in
the interior. The so-called Large Tower (9 × 25 m) was in fact built in the
early 4th century on the southern side of the fort, and then restored in the
11th century, before being abandoned in that same century. The occupied
area extended outside the ramparts during that century, but that part of
the settlement was abandoned after the Cuman attack of 1095, or after
that of 1122, and was turned into a cemetery during the 12th century.18
The 4th- to 6th-century fort located on an island at Garvăn (ancient
Dinogetia) was restored and a garrison of stratiotai was established inside
it, their houses being dug into the leveled debris from the Late Roman
period (fig. 6). The polygonal fort has an area of 1.2 hectares with a 2.8–3 m
thick rampart and 14 towers (11 along the walls and 3 in the corners). The
main gate on the southern side is 2.5 m large. Two other entrances exist
18 Baumann, Mănucu-Adameşteanu 2001, 219; Mănucu-Adameşteanu 2001 a, 57;

Baumann et alii 2002, 158; Baumann et alii 2004, 147; Damian 2005, 173–174; Mănucu-
Adameşteanu 2009, 625, 628–629.
106 chapter three
on the western and northern sides, respectively.19 The first occupation

phase of the Byzantine site is coin-dated to the reign of Emperor John
Tzimiskes.20 Moreover, ten gold coins (tetartera) struck for Basil II and
found on the site have led some scholars to the conclusion that a provin-
cial mint for the army may have operated in Garvăn.21 As in Nufăru and
Isaccea, the occupied area extended beyond the walls, but was abruptly
interrupted by the Pecheneg invasions ofthe 1080s. During the 12th cen-
tury, the area outside the rampart became a cemetery.22
At Măcin (ancient Arrubium), the Byzantine occupation is documented
primarily by pottery finds typical for the 10th and 11th centuries, as well as
by 25 folles dated without interruption between 976 and 1081. A silver coin
(stamenon) struck for Emperor Alexios I Comnenos in 1092–1118 is also
known from this site. The scarce archaeological evidence available so far
is not sufficient for deciding whether the walls of the early Roman camp
(220 × 170 m) or those of the late Roman fort (45 × 75 m) were restored
after 971.23
At Turcoaia (ancient Troesmis), only the western fort was occupied
in the early 11th century. The eastern fort was not restored, but the area
between the two forts was apparently occupied. That area shrank after
the Pecheneg invasion of 1036 which has been also blamed for a layer of
destruction by fire inside the western fort. That fort seems to have been
abandoned during the secession of 1072–1091, and then reoccupied after a
while until the Pecheneg and Cuman invasion of 1122. The series of coin
finds covers with some interruptions the whole period between Basil II
and Manuel I Comnenos.24
At Piatra Frecăţei (Ostrov commune, Tulcea county), archaeologists
have established that Beroe, the old Roman fort, and its suburb area
19 Diaconu 1969 a, 44–49; Barnea, Ştefănescu 1971, 77; Barnea 1971, 353–354; Barnea
1980, 243–244; Damian 2005, 165–167.
20 Mănucu-Adameşteanu 2001 a, 52–53; Custurea 2000, 137.
21 Metcalf 1979, 53–54; Mănucu-Adameşteanu, Poll 1999, 345; Mănucu-Adameşteanu
2001 d, 50.
22 Barnea 1973, 292–293, 298–301; Barnea 1980, 259; Barnea et alii 2004, 126–127.
23 Condurachi, Barnea, Diaconu 1967, 184; I. Barnea, P. Diaconu, Arrubium, EAIVR, I, 1994,
120–121; Mănucu-Adameşteanu 1992, 400; Custurea 2000, 144–145; Mănucu-Adameşteanu
2001 a, 51; Oberländer-Târnoveanu 2003, 386–387, 394; Damian 2005, 148–149; Mănucu-
Adameşteanu 2010 a, 267–294; Mănucu-Adameşteanu 2010 f, 237–242.
24 Mănucu-Adameşteanu 1980, 230–234; Oberländer-Târnoveanu 1980 a, 267–269,
274–278; Custurea 2000, 141; Mănucu-Adameşteanu 2001 a, 49–50; Oberländer-Târnove-
anu 2003, 389, 395; Damian 2005, 163–165; Mănucu-Adameşteanu 2010 a, 181–265; Mănucu-
Adameşteanu 2010 d, 439–469.
were again occupied after 971. In a sunken-floored building dug into the
ruins of the former rampart a coin was found, which had been struck for
Emperor Michael IV. This has been interpreted to mean that a part of the
old ramparts, which had not been repaired after 971, was still used for the
protection of the inhabitants, but only until the destruction caused by
the Pecheneg invasions of 1032–1036. The occupation continued after that
without a repaired precinct, perhaps until the Cuman attack of 1122.25
At Hârşova (ancient Carsium), the 1.5 hectare promontory is surrounded
by three concentrical precincts, which were used not only by the Byzan-
tines, but also by the Ottomans until their final dismantlement in 1829. The
southern side of the fort was protected by a natural cliff. According to the
results of the latest excavations on the western side, the outside rampart
(I), which 1.5 m thick and measures 80 m to the north, 76 m to the east,
and 86 m to the west, is the most recent, and may be dated to the Byzan-
tine period (and not to Roman period, as scholars previously believed). Its
foundation trench cuts through a layer with 9th–10th-century pottery. The
actual Roman precinct is the innermost (III), but it too was used in the
10th–12th centuries. Outside the walls, a civilian settlement existed in the
11th century. Coin finds from Hârşova are dated between the reigns of
John Tzimiskes and Alexios I Comnenos, the latest being a stamenon from
the first series dated after 1092. It is important to note that silver coins
struck for all Comnenian emperors in the 12th century have been found
on the site, a very unusual situation for Dobrudja, which may point to
the special significance of this fort.26 The Romanian name appears to be
related to the ancient name Carsium through a Slavic intermediary form,
which must have been in use during the Byzantine period. The absence
of any coins struck for John II Comnenos may indicate that the fort was
destroyed in 1122.
Capidava was also on a promontory, which is now to the south from
the village of Topalu. The medieval fort occupied the same site as the
Roman camp. Radu Florescu believed the first occupation phase to be
dated to the 9th century by means of pottery remains. The earliest fort
was defended by a rampart made of remains of Roman buildings, and by
25 Custurea 2000, 147–148; Mănucu-Adameşteanu 2001 a, 47–48; Barnea et alii 2002, 255;
Stănică 2004, 357, 365; Damian 2005, 149; Barnea et alii 2009, 161; Mănucu-Adameşteanu
2010 a, 267–294; Mănucu-Adameşteanu 2010 b, 94–95, 103–104.
26 Condurachi, Barnea, Diaconu 1967, 184–185; Barnea, Ştefănescu 1971, 80; Panait et
alii 1995–1996, 122–127; Custurea 2000, 140; Mănucu-Adameşteanu 2001 a, 45–47, 403 (table
5); Oberländer-Târnoveanu 2003, 386, 394; Damian 2005, 160–163; Covacef, Nicolae 2005,
169–171; Covacef, Nicolae 2007, 178–181; Nicolae et alii 2008, 319–323.
108 chapter three
a ditch.27 Petre Diaconu rejected that interpretation, and argued that the
first occupation phase cannot be dated before 971.28 To be sure, despite
intensive archaeological research on the site, so far only two coins dated
before 971 are known.29 Radu Florescu’s chronology is not supported by
any shred of evidence. The Byzantine fort used a part of the Late Roman
precinct, which is 2.75 m thick, surmounted by a new, 2 m-thick wall on
the northern, eastern and southern sides, made of spolia from the Roman
ruins. The series of coin finds has a gap between 986 and 1000. The most
recent excavations show the existence of another precinct on the eastern
side. Capidava was abandoned after a Pecheneg invasion in 1047. It is not
clear if the Roman harbor was also used during the Byzantine period.30
Given its incorporation into the precinct of a military facility, no sys-
tematic excavations have ever been carried out in Hinog (ancient Axio-
polis) after those of Grigore Tocilescu (1895–1896) and Carl Schuhhardt
(1916–1917). Three consecutive fortifications were built on raised ground
in front of the Hinog island, about 3 km south of Cernavoda, in front of the
Hinog Island. The southern fortress was used in the Roman and Byzantine
periods. According to the available data, it is a quadrangular fortification
with sides measuring 161 m (to the north), 210 m (to the east), 200 m (to
the south), and 250 m (to the west), respectively. Taking into account that
it is strategically located next to one of the most important fords across the
Danube, Axiopolis may have already been occupied in 971. An anonymous
follis of type A1 found on the site may support that supposition. All other
coins are from the subsequent reigns between Basil II and Romanos IV,
but there is also one struck for Emperor Alexios I after 1092.31 Since Axio-
polis was the see of a bishop in the 11th century,32 this may have been
quite an important site.
At Oltina, the systematic excavations carried on the site at “Capul Dea-
lului” since 2001 have established that the fortress was used during the 11th
27 Florescu 1967, 259–268; Florescu 1986, 175–176; Opriş 2004, 69. Accepted by Fiedler
2008, 199.
28 Diaconu 1969 a, 46–48.
29 Gândilă 2007, 608, 615–616 (one follis from Leo VI, 886–912 and another one from
Romanos I, 931–934).
30 Florescu, Covacef 1988–1989, 204–244; A. Barnea, I. Barnea, Capidava, EAIVR, I, 1994,
249; Custurea 1995–1996, 301–307; Custurea 2000, 134; Mănucu-Adameşteanu 2001 a, 43–44,
402 (table 4); Damian 2005, 158–159; Pinter, Ţiplic, Urduzia, 2008, 85–86; Pinter, Urduzia,
2009, 89–90.
31 Barnea 1960, 69–78; I. Barnea, Axiopolis, EAIVR, I, 1994, 146; Mănucu-Adameşteanu
2001 a, 40–42, 160; Damian 2005, 148.
32 Popescu 1994, 421–438.
7. The group of fortifications Păcuiul lui Soare-Dervent-Oltina (after Diaconu,

Vâlceanu 1972, 10, fig. 1)
century, but the earthen rampart on the western side was not replaced
with a stone wall after 971. The rampart was uncovered along a span of 410
m. The gate was on the south-western side. The coins discovered during
the excavations indicate its occupation between the reigns of John Tzi-
miskes and Nikephoros III Botaneiates (1078–1081). The number of coins
visibly decreased after Constantine X. The seal of Gregory Mavrokataka-
lon is a valuable proof that the fort remained in existence until the last
two decades of the 11th century.33
In Dobrudja, another important ford across the Danube is in Der-
vent (the Turkish name actually means “ford, passage”) (see Fig. 7). On
a 40m-high promontory, a new fort was built next to the ruins of the old
Roman one. The Byzantine fort has a 1.85 m-thick rampart, but its size
remains unknown, for only the eastern side has been uncovered, and it is
120 m long. That the fort was built during the reign of John Tzimiskes is
indicated by the silver coin already mentioned. The other coins found on
the site are dated between the reign of Tzimiskes and that of Michael IV,
which means that the final destruction may be attributed to the Pecheneg
33 Custurea 2000, 151; Custurea 2000–2001, 583–594; Mănucu-Adameşteanu 2001 a,

39–40; Chiriac et alii 2003, 220; Chiriac et alii 2004, 222–224; Custurea 2006, 415–421; Cus-
turea 2009, 612–621.
110 chapter three
invasions of 1035–1036. The coins struck for Constantine IX, Constantine X,

and Alexios I recorded for this site in older studies are in fact from Păcuiul
lui Soare. The exact date of the other destruction by fire, which has been
observed in the site’s stratigraphy remains unknown, and it may be the
result either of some catastrophic event during peacetime, or of Samuel’s
campaign of 986. Outside the fortress, the place was transformed into
a cemetery for the population from the neighbor settlement Păcuiul lui
Soare.34
The island Păcuiul lui Soare near Dervent has the most spectacular
Byzantine fortress so far known in Dobrudja. It most certainly operated as
a naval base between 971 and at least 986. At the time of its building, the
fort was in fact on the left bank of the river, i.e., a bridgehead. The largest
part of the precinct (which was probably trapezoidal) is now sunken, but
the total area could be estimated at around 4 or 5 hectares. The 6 m-thick
walls were built on oak beams in order to cope with the instability of the
soil. They were made of large ashlars bonded with water-resistant mortar.
The north-eastern wall is still standing on a 42 m-long segment, while the
south-eastern one runs for 240 m. The main gate was on the north-eastern
side—a rectangular tower (10.5 × 14.7 m), reaching 8.5 m to the outside.
Another gate may have existed on the southern side. The military harbor
was on the south-eastern side and consisted of a 24 m-long platform, with
a 4 m-large entry and two large, rectangular towers. The southern tower
is 9.3 m long and 5.6 m wide. The initial goal of this naval base was to
prevent any Rus’ attacks directed at Dristra, which is 18 km up the river.
The naval base at Păcuiul lui Soare was integrated into the same system
of fortifications as the forts in Oltina and Dervent (perhaps also as those
from Rasova and Cochirleni, which are likely to have survived after 971).
The fortress was abandoned in 986. After 1000, it became a fortified settle-
ment without any military function.35
As for Nufăru, Ernest Oberländer-Târnoveanu denied that the fortress
in Păcuiul lui Soare was built during the reign of John Tzimiskes, since
34 Diaconu, Anghelescu 1968, 348–349; Diaconu 1970, 44; Diaconu 1977 b, 62; P. Dia-
conu, Dervent, EAIVR, II, 1996, 50; Custurea 2000, 137; Mănucu-Adameşteanu 2001 a, 37–39,
111, 400 (table 2); Mănucu-Adameşteanu 2001 b, 93, 400 (table 2); Mănucu-Adameşteanu
2003, 303, 306; Damian 2005, 146–147; Damian, Vasile 2010, 338–339.
35 Diaconu 1966, 367–370; Condurachi, Barnea, Diaconu 1967, 190; Diaconu 1969 b;
Diaconu 1976 a, 409–447; Diaconu, Vâlceanu 1972, 23–25, 52–54; Diaconu, Damian, Vasile
2001, 170–172; Diaconu, Damian, Mărgineanu-Cârstoiu 2004, 226–227; Diaconu, Damian,
Mărgineanu-Cârstoiu 2005, 252; Damian 2005, 150–153; Damian, Bănăseanu 2006, 249;
Damian, Ene 2011, 95; Damian, Olteanu 2012, 94.
only a few folles of type A2 (dated 976–1020) are known from the site.36
However, the stratigraphical evidence from the site is indisputable. Mor-
evoer, it has been recently demonstrated that about a quarter of the folles
of type A2 are earlier variants, which may be dated before 1000, in which
case they actually support the idea of the naval base being in existence
between 971 and 986. The site has actually produced five coins struck for
Emperor John Tzimiskes.37 Another group of scholars believe that the
fortress was in existence already before 971. According to such opinions,
Păcuiul lui Soare was Mundraga, to which Emperor Symeon fled when
attacked by the Magyars in 895.38 However, it is more likely that Mun-
draga was present-day Tutrakan.39 According to a third theory, Păcuiul lui
Soare was the so-called “Omurtag’s palace from the Danube” mentioned
in a ninth-century, Bulgar inscription from Tărnovo,40 but it is now clear
that that palace was in fact in Dorostolon/Silistra.41 The ramparts made
of large ashlars without a filling, but set on timber beams reminds one
of the walls from Pliska and Preslav, and are substantially different from
those of the Byzantine forts located north of Axiopolis and built after 971.
Radu Florescu saw that difference as a key argument in favor of the idea
that Păcuiul lui Soare was in fact a Bulgarian fort erected during the reign
of Symeon.42 But the building technique in question is not Bulgarian, and
even in Pliska and Preslav its application must be attributed to Byzan-
tine masons or to Bulgarian masons working after Byzantine models. On
the other hand, there are practical reasons for its application in Păcuiul
lui Soare, particularly because of the instability of the soil.43 Only arti-
facts securely dated to the 9th century—either coins, or pottery—such as
36 Oberländer-Târnoveanu 1998, 76.

37 Mănucu-Adameşteanu 2001 a, 33–36, 399 (table 1); Mănucu-Adameşteanu 1998–2003,
nr. 146–151, 164–165.
38 Kuzev 1972, 418.
39 Tsankova-Petkova 1975, 35–40. Another opinion: Beševliev 1985, 17–21 (one of the
akropoleis of Dristra).
40 Mutafčiev 1932, 192; Kuzev 1969, 138; Kuzev, Gjuzelev 1981, 196–200; Ovčarov 1987,
57–68; Gjuzelev 1991, 86. For the inscription see Beševliev 1963, 247–260; Petkov 2008, 11.
The Bulgarian origin of this fortress was also admitted by Beševliev 1981, 469; Soustal 1997,
119; Curta 2006, 241.
41 Beševliev 1962, 71 and Petkov 2008, 11 observed that the distance between Pliska and
that palace recorded in the inscription is the same as that between Pliska and Dristra.
See also Fiedler 2008, 193. The recent study of Atanasov 2012 brought final archaeological
evidence in this respect.
42 Florescu, Ciobanu 1972, 390, 394–395; Florescu 1986, 175. A similar opinion in Comşa
1983, 102.
43 See the remarks of Diaconu 1988 a, 181–183 about Ovčarov 1987.
112 chapter three
found in Oltina, Rasova and Cochirleni-Cetatea Pătulului, may solve this

problem. So far, however, no such evidence has emerged.
Dristra, the headquarters of the province, was incorporated into the
Bulgar state at an early date, but it is possible that it was recovered for
short periods by the Byzantines during the campaigns of Constantine V,
when the Byzantine navy moved on the Danube. The development of the
city as a Bulgar center began in the early 9th century under Krum, as
indicated by an intensive building activity. At that time, a new wall was
built, which enclosed an area of 0.45 ha on the bank of the Danube. This
was also the wall in use during the Byzantine period, along with the old
Roman precinct. Recent archaeological excavations have identified the
pier. Its timber beams have been radiocarbon-dated to 780± 200, a date
the margin of error of which is far too large for deciding whether the pier
is a Byzantine or a Bulgarian construction. At any rate, Dristra was the see
of a Bulgarian bishop.44 During the 11th century, a new fort was built in
the southern part of the city. After 1088, the area within the walls shrank,
with previously occupied sectors turning into burial grounds.45
To the west from Dristra, the archaeological evidence of Byzantine forts
is rather meager. The fort in Vetren (20 km west of Silistra) was erected
on the site of the Roman camp of Tegulicium, which had meanwhile been
occupied by a ninth-century rural settlement. Several seals are known from
Vetren, among which the most important is that of Tyrach (protospath-
arios and eparchos) and that of katepano Romanos Diogenes. The coins
finds cover the entire period between John Tzimiskes to Alexios I, with
the latest coins being three tetartera struck after 1092. A special find is a
hoard of 12 nomisma histamenon.46 The imperial tent of Alexios I, which
was used during the siege of Dristra in 1087 (see chapter III. 2), is known to
have been sent to Vetren (Vetrinon).47 A (yet unpublished) seal of Alexios I
is known from Spanţov, on the opposite (northern) bank of the Danube.48
However, according to the information in the archive of the Institute of
Archaeology in Bucharest, where the seal is now preserved, the actual
find spot is in the village of Stancea, which is very close to the fortress of
44 Soustal 1997, 119–123; Angelova, Koleva 2004, 21–28.

45 The information about two fortresses called with the unfit word akropoleis was trans-
mitted by Anna Comnena, VII, 3.3 (transl. Sewter, 193). For the fortifications in the 11th–
12th centuries, see Kuzev 1969, 138–139; Angelova 1993, 52–57.
46 Atanasov, Jordanov 1994, 42–45; Oberländer-Târnoveanu 2003, 390, 395; Custurea,
Talmaţchi 2011, 378.
47 Anna Comnena, VII, 3.6 (transl. Sewter, 194); Stanev 2012, 21–25.
48 Condurachi, Barnea, Diaconu 1967, 193; Diaconu 1978, 55.
Vetrinon. It is possible that the seal has been attached to a message sent
to a military unit which had crossed the Danube against the Pechenegs.
At a distance of only 5 km to the west, on the southern bank, at Popina,
there was another fort with a stone rampart enclosing an area of 140 ×
150 m. The fort may have been built in the 9th century, but coin finds
clearly indicate its use during the Byzantine period until the 1080ss. Not
far from it, at Gradiščeto, there was another fort with an earthen rampart,
which is also dated to the 11th century. A seal of Basil Apokapes is known
from Popina.49 The re-occupation of the 9th- to 10th-century stronghold
in Krivina (on the site of the Roman fort Iatrus) may be coin-dated to the
1030s, that is to the reign of Michael IV (1034–1041). There are so far no
anonymous folles of type A2, which are known to have been struck under
Basil II and Constantine VIII. The fort in Krivina appears to have remained
in use until the Pecheneg invasion of 1047.50 Scarce information is avail-
able for the forts in Tutrakan, Ruse, Svištov, Gigen, Nikopol, and Vidin.
Vidin and Svištov are actually mentioned by al-Idrisi in 1154 as Bidenu and
Bestcastro (Suvestcastro), respectively.51
The segment of the Danube between Braničevo and Sirmium was the
most exposed to the Hungarian attacks, which started in 1059 and were
directed at the valley of the Morava. The restoration of the defense system
in this region was therefore the result of the measures taken by Isaac I
Comnenos. The Iron Gates area was part of the theme of Sirmium estab-
lished in 1018, but was threatened by attacks from Hungary, as long as
the Hungarians did not yet control the valley of the Cerna river vis-à-vis
the mouth of the Timok river. Hungarians came to control the valley of the
Cerna only after 1127. Nonetheless, the Roman watch-tower in Veliki
Gradac (ancient Taliata) was restored from ground up in the 11th century,52
But no other formerly Roman forts in the Iron Gates area are known to
have been integrated into the defense system, even if some of them appear
to have occupied in the 11th century as well.
In conclusion, it appears that the main reason for restoring the defense
system on the Danube frontier was the protection of Dristra and of the
main fords at Nufăru, Isaccea, Garvăn, and Dervent. Troops garrisoned in
49 Mănucu-Adameşteanu, Jordanov 2002, 125–126; Oberländer-Târnoveanu 2003, 388.

50 Diaconu 1988 b, 202–203; Schönert-Geiss 1991, 239; Diaconu 1992 a, 180.
51 Al-Idrisi (ed. Jaubert, 386; ed. Miller, 129); Tomaschek 1886, 299, 300; Kuzev 1966,
23–50; Kuzev 1967, 41–70; Kuzev 1968, 37–49; Cihodaru 1968, 224; Kuzev, Gjuzelev 1981,
98–115, 149–156.
52 Popović 1991, 175.
114 chapter three
the theme of Dristra, later Paradunavon, were expected to prevent and

repel attacks from the lands north of the Danube. In most cases, the Byz-
antines simply rebuild—sometimes on a smaller scale—the old Roman
forts. The restoration skipped a number of old Roman forts along the
Danube, no doubt because of lack of sufficient soldiers for the garrisons.
Other formerly Roman sites were reoccupied, but not refortified, which
suggests that they had no military function.
Only one bridgehead on the left (northern) bank of the Danube is so far
known for sure, namely the fort at Păcuiul lui Soare, but there may have
been others as well. For example, the name of the site at Grădiştea, 3 km
to the southwest from Călăraşi, derives from the Slavonic word for “for-
tress.” A fortress may have existed there before its complete destruction
by the Danube, as suggested by two folles of types A1 and A2, respective,
which were found there during field surveys.53 Other crossing points may
have been located at Borcea-Pietroiu (in front of Cochirleni), Borduşani
(in front of Capidava) and Şendreni (in front of Garvăn).54 It was also
been surmised that the polygonal tower from Giurgiu was built after 971,
and that it had the same function as the naval base in Păcuiul lui Soare,
given that it was made in the same technique (big ashlars with bands of
brick). This polygonal tower lies under the north-western tower of the
later fortification, which was built there in the late 14th century. However,
the problem is that there are no archaeological finds from this site which
could be dated to the 10th–12th centuries. An 11th-century reconstruction
was also surmised for the fortress in Celei, the formerly Roman bridgehead
at Sucidava, given that 11th-century coins have been found on the site.
Coin finds have also been used as evidence for bridgeheads at Piua Petrii,
Zimnicea and Turnu Severin, all of which have been regarded as possible
anchorages for the Byzantine navy.55 This cannot of course be excluded,
but no archaeological evidence exists so far to support that idea.
After 1000, the Bulgarian fortifications in the interior, such as Odărci,
Skala or Tsar Asen were also integrated into the Byzantine defense sys-
tem, for they now monitored the access to Preslav and further south, to
Constantinople.
53 Diaconu 1970, 38; Ioniţă 2005, 133.

54 Diaconu 1970, 38.
55 Ioniţă 2005, 51–52, 121, 132.
2. The Danube Frontier in the 11th Century
Jonathan Shepard has proposed the idea that after 1000, Basil II aban-
doned the strategy based on a very strong frontier along the Lower Dan-
ube, as the Rus’ were no more a threat. The emperor left instead only few
forces, in order to move resources to other Balkan regions, which were
still under Bulgarian occupation. According to Shepard, this was a kind
of “policy of a minimal military commitment.”56 This is in fact consistent
with the abandonment of the naval defense, which is indicated among
other things by the transformation, after 1000, of the fort in Păcuiul lui
Soare into a fortified settlement without naval function.57 No permanent
naval force capable of preventing barbarian raids seems to have been in
use throughout the existence of the Dristra/Paradunavon theme. When
needed, ships could be dispatched from the naval base in Mesembria, and
afer 1078, in Anchialos.58 On the basis of a passage in the Life of Saint
Cyril the Phileote, who was a sailor on the Danube between 1042 and 1045,59
Vasilka Tăpkova-Zaimova has maintained that a navy must have been in
existence on the Danube at that time. However, the passage in question
appears to refer to a civilian fleet engaged in trade, and not to the navy.60
The “policy of a minimal military commitment” is in stark contrast with
the heavy investments in the Danube fleet during the Early and Late
Roman periods.
The decline of the Byzantine naval power is in fact highlighted by the
events of July 1043. The Rus’ prince Vladimir of Novgorod, the son of the
Kievan prince Yaroslav, launched a maritime expedition against Con-
stantinople, probably in connection with and to the assistance of George
Maniakes’ rebellion. After his victory against the Pechenegs in 1036,61
Vladimir was eager to get involved in the Byzantine affairs. When 10,000
or perhaps as many as 20,000 Rus’ warriors arrived on small boats from the
Dniepr into the Black Sea, approaching Constantinople, nobody expected
a naval attack against the capital, which does not appear to have been
defended any more by a naval force. A fleet had to be quickly improvised,
and some of the enemy boats were destroyed by means of the “Greek
56 Shepard 1985, 254–259. Similar opinions at Haldon 1999, 64; Stephenson 2003, 114.
57 Diaconu 1966, 369.
58 Ahrweiler 1966, 167; Gjuzelev 1978, 52–53; Gjuzelev 1981, 18.
59 Tăpkova-Zaimova 1980, 330.
60 Barnea 1993 b, 589–590; Stephenson 2000, 84, 96.
61 Curta 2006, 302–303; Spinei 2006, 181; Spinei 2009, 107.
116 chapter three
Fire.” If even the naval defense of Constantinople was neglected, what

could be expected from the periphery? It is no surprise therefore that the
defense of the Danube province was left to the land troops. Indeed, when
they returned by land, the Rus’ were attacked at Varna by troops of the
Dristra theme under the command of katepano Katakalon Kekaumenos.
They were also repelled by the same general when they landed for sup-
plies, perhaps at Presthlavitza-Nufăru, easily accessible by boat from Seli-
nas (Sulina) through the Saint George arm of the Danube Delta.62
The troubles in the Danube region appear to have been underesti-
mated, mostly in what concerns the Pechenegs, and this would have seri-
ous consequences for the security of the frontier. After a failed 1017 by
Tzotzikios, the strategos of Dristra, the Pechenegs returned to the Danube
region in 1027, when reached Niš in the theme of Bulgaria, which appears
to have been left deserted for some time after that attack. It is very prob-
able that the same invasion affected Hungary, where a Pecheneg attack
was recorded for 1028. For a while, scholars viewed this attack as targeting
exclusively the western part of the Danube region (the themes of Sirmium
and Bulgaria), given that no destruction dated around 1027 has been iden-
tified in the forts of Dobrudja.63 A new examination of the archaeologi-
cal and numismatic evidence, led Gheorghe Mănucu-Adameşteanu to the
conclusion that Garvăn and Capidava must have also been attacked in
1027 (coins from Basil II and Constantine VIII were recovered from burned
layers).64 Both forts were repaired after destruction. At Garvăn, the south-
ern gate was doubled on the outside. Other repairs dated to the same
period have been observed at the towers 11 and 13.65 A small hoard of folles
from Isaccea was hidden during the attack of 1027.66
62 Skylitzes, Constantine Monomachos, 6 (ed. Thurn, 430–433; transl. Flusin, 357–360;

transl. Wortley, 407); Attaliates, ed. Bekker, 19–20 (ed. Pérez Martín, 16–17); Russian Pri-
mary Chronicle, 142; Zonaras, XVII, 24 (ed. Büttner-Wobst, 632–633); Bănescu 1946, 74; Ahr-
weiler 1966, 128–129; Barnea, Ştefănescu 1971, 127; Shepard 1984, 147–212; Angold 1984, 12–14;
Franklin, Shepard 1996, 216–217; Kostova 2006, 583–589; Spinei 2009, 108.
63 Skylitzes, Constantine VIII, 2 (ed. Thurn, 373; transl. Flusin, 309; transl. Wortley, 352);
Zonaras, XVII, 10.2 (ed. Büttner-Wobst, 571); SRH, II, 109–111 (Chronicon Henrici de Mügeln,
3); Diaconu 1970, 40–42; Barnea, Ştefănescu 1971, 123; Spinei 2009, 107.
64 Mănucu-Adameşteanu 2001 b, 88–91; Mănucu-Adameşteanu 2003, 303–304; Mănucu-
Adameşteanu, Poll 2006, 437–438. For Garvăn: Barnea 1971, 355; Barnea 1973, 305 (older
studies linked the destruction level with the attack of 1036).
65 Mănucu-Adameşteanu 2001 a, 52.
66 Mănucu-Adameşteanu, Poll 2006, 435–437, 451–455.
The Pecheneg raids of 1032–1036 were disastrous for the entire Danube
region. The attacks took advantage of the inclement weather (the horse-
men crossed the frozen Danube during one of these inroads, in 1035), and
were repeated at short intervals (three raids occurred in 1036 alone).67
Destruction layers that could be dated to this period on the basis of the
coin finds are documented in Tulcea,68 Isaccea,69 Garvăn,70 Turcoaia,71
Capidava,72 Oltina,73 Dervent,74 Popina-Gradiščeto,75 and Gigen.76 Inland
fortresses were also sacked: Constanţa,77 Tsar Asen,78 Rujno, Okorš,
Kladenci,79 Skala,80 Odărci,81 Šumen,82 Kavarna, and Balčik.83 Even Dristra
was affected by the invasions of 1032–1036.84
The precise chronology of events depends on the dates for the anony-
mous folles of types B and C, which have been found in the destruction
layers. According to Cécile Morrisson,85 the B-type folles were struck
between 1028 and 1034, and the C type between 1034 and 1041, while Philip
Grierson86 proposed the years 1035–1042 and 1042–1050, respectively.
67 Skylitzes, Romanos III Argyros, 10; Michael IV Paphlagonianus, 9, 10 (ed. Thurn, 385,
397, 399; transl. Flusin, 319, 328, 330–331; transl. Wortley, 364, 374–376); Glykas, 584 (attack
dated in 1032 or 1033); Zonaras, XVII, 12.9; 14.26,30 (ed. Büttner-Wobst, 579, 589, 590); Dia-
conu 1970, 43–49; Malamut 1995, 118; Stephenson 2000, 81; Curta 2006, 293–294; Spinei
2006, 187; Spinei 2009, 107.
68 Mănucu-Adameşteanu 2001 a, 113; Mănucu-Adameşteanu 2010 c, 230.
69 Baumann, Mănucu-Adameşteanu 2001, 110.
70 Barnea 1971, 355; Barnea 1973, 308; Barnea 1980, 245; Mănucu-Adameşteanu 2001 a,
111–112; Mănucu-Adameşteanu 2001 b, 94–95.
71 Mănucu-Adameşteanu 2001 a, 113, 128; Mănucu-Adameşteanu 2010 d, 440.
72 Mănucu-Adameşteanu 2001 a, 110; Mănucu-Adameşteanu 2001 b, 93. The old opinion
sustained the final destruction in 1036 (Diaconu 1970, 44; Barnea, Ştefănescu 1971, 124; Flo-
rescu, Covacef 1988–1989, 244), but later numismatic discoveries have shown that it was
peopled until the Pecheneg invasion of 1047: Custurea 1995–1996, 301.
73 Mănucu-Adameşteanu 2001 a, 39; Custurea 2000–2001, 590; Custurea 2009, 614.
74 Mănucu-Adameşteanu 2001 a, 111; Mănucu-Adameşteanu 2003, 305; Damian, Vasile
2010, 339.
75 Mănucu-Adameşteanu, Jordanov 2002, 126.
76 Mănucu-Adameşteanu, Jordanov 2002, 130; Borisov 2007, 75.
77 Mănucu-Adameşteanu 1991, 323.
78 Dimova 1993, 65, 73; Atanasov 2003, 291.
79 Atanasov 1991, 84, 88; Atanasov 2003, 291; Mănucu-Adameşteanu, Poll 2006, 439.
80 Yotov 1995, 182; Yotov, Atanasov 1998, 198; Atanasov 2003, 291.
81 Mănucu-Adameşteanu, Poll 2006, 438; Dončeva-Petkova 2007, 644.
82 Žekova 2005, 169–170.
83 Mănucu-Adameşteanu, Jordanov 2002, 130–131.
84 Angelova 1987, 94 (the destruction is attested by an anonymous B type coin found
in a burned level).
85 Morrisson 1970, 586–600.
86 Grierson 1973, 634.
118 chapter three
Grierson’s chronology was accepted by some,87 but Gheorghe Mănucu-

Adameşteanu has demonstrated that the date 1050–1060, which Grierson
advanced for the D type does not match what it is known about the his-
tory of the Danube region in those years. As a consequence, the chronol-
ogy of the B and C types must also be wrong.88 Furthermore, the forts
with destruction layers dated by means of B-type coins, or with no C-type
coins whatsoever, must have been the victims of the attacks of 1032–1035
(see Fig. 8): Dristra, Oltina, Popina-Gradiščeto, Vetren, Gigen, Skala, Tsar
Asen, Okorš, Rujno, Kladenci, Kavarna, and Balčik. At some point during
that same interval, a coin hoard was buried in Središte, which included
34 miliaresia and 27 folles, the last coins being B-type specimens. The
conclusion is that the attacks of 1032–1035 were directed mostly at the
region near Dristra and from there to Preslav, while the invasions of 1036
struck especially northern Dobrudja (see Fig. 9).89 For this reason it is
more probable that the fort at Dervent was attacked in 1035, and not in
1036. The abandonment of that fortress left undefended one of the most
important fords, which was only at a small distance from Dristra. Among
forts attacked in 1032–1036, those from Gigen, Skala, Rujno, Tsar Asen,
Odărci, Kladenci, and Okorš were also abandoned. Furthermore forts in
the interior were not restored at all.
The Pechenegs were bent on sacking Dristra and its hinterland. Next
they moved for more plunder to northern Dobrudja, which had until
then been spared by their depredations. All raids taking place between
1032 and 1036 had serious demographic consequences. Many settlements
in northern Bulgaria were destroyed during the first half of the 11th cen-
tury, most probably between 1032 and 1036. For instance, in the district of
Silistra, out of eight forts and 72 settlements, only three survived, while in
the district of Dobrič only 11 out of 121.90
The disaster of 1032–1036 may be blamed on Emperor Romanus III’s
failed policies. His grand strategy was not based upon a lucid analysis of
the reality, being driven only by the emperor’s quest for glory for himself
and for his soldiers.91 Michael Psellos wrote in this respect that: “setting
his heart on military glory, he prepared for war against the barbarians,
87 For instance Atanasov 2003, 289–298.

88 Mănucu-Adameşteanu 2001 b, 107–108; Mănucu-Adameşteanu 2003, 307–308.
89 Atanasov 2003, 291–295; Mănucu-Adameşteanu 2003, 301–308. For the coin hoard of
Središte see Parušev 1993, 164; Custurea, Talmaţchi 2011, 350–351.
90 Borisov 2007, 74–75.
91 Shepard 2002, 77.

the danube frontier of byzantium (1000–1204)
8. The fortifications affected by the attacks of 1032–1035

119
120
chapter three
9. The fortifications affected by the attack of 1036

east and west. Victory over the western barbarians, however easy, seemed
no great triumph, but an attack on the barbarians of the east, he thought,
would win him fame.”92 The neglect of the western part of the empire, to
which fewer troops were allocated, was simply an invitation for trouble
from the Pechenegs.
Following the death of Basil II in 1025, the implementation of peacetime
policies caused the emperors Constantine VIII (1025–1028), Romanos III
(1028–1034) and Michael IV (1034–1041) to cut funds for the army, even
though the state budget increased considerably as a consequence of the
expansion of the urban economy and successive tax raises. Basil II had
spent huge resources for an army of more than 250,000 men93 which had
been permanently on campaign for almost half a century. After 1025, how-
ever, there were no more permanent troops in the forts.94 On the other
hand, the protection granted to estates owned by stratiotai was removed,
although those small landowners had been an essential component of
the army during the 9th–10th centuries. The measure, on the other hand,
encouraged the commutation of the military obligations into cash pay-
ments, as funds were now needed for civilian expenses (some of them
completely useless, such gifts for favorites), and not for the building of
a well maintained military force. The new professional army came into
being only during the reign of Constantine IX (1042–1055), but it became
efficient only during the reign of Alexios I Comnenos (1081–1118).95
Constantine IX also introduced to all border provinces the dukes or
the katepanoi, the office of commanders of major units of the Byzantine
army.96 The theme of Dristra was no exception, as it received its first
katepano in the person of Katakalon Kekaumenos, in 1042–1045, at the
beginning of Constantine IX’s reign. Appointing dukes or katepanoi to
the themes actually meant the transfer of tagmata (army corps of profes-
sional military) to those provinces, which now relied on those military
units for their defense, instead of locally recruited troops.97 A duke was
in fact the commander of the tagma garrisoned within a given territory.
In other words, the presence of a duke or a katepano (the two terms were
often used interchangeably) implies the existence of professional troops.
92 Michael Psellos, III, 7 (transl. Sewter, 42).

93 Treadgold 1995, 285.
94 Cheynet 2005, 109.
95 Ostrogorsky 1956, 346–347, 354–355; Oikonomides 1976, 141–147; Lemerle 1977, 251–
312; Angold 1984, 62–65; Cheynet 1991, 64–73.
96 Cheynet 1985, 193.
97 Cheynet 1985, 194.
122 chapter three
From an archaeological point of view, this is confirmed by the sudden

appearance in a particular region of a great number of gold and silver
coins, which were sent from the center as payments for the troops. This
is in fact the reason for which, unlike other provinces in the interior, such
coins appear in forts on the Danube in surprisingly larger numbers than
the folles.98 The total number of troops stationed in the theme of Dristra
may be estimated at 5,000.99
Without a fleet, devoid even of an effective fortification system, and
with fewer soldiers, the Lower Danube region thus required a different
kind of defense after the disastrous invasions of 1032–1036. The new policy
was based on the idea of maintaining peaceful relations with the Pech-
enegs to whom payments were not made and with whom free trade was
encouraged. The two offices of kommerkiarioi in the Danube region—
Presthlavitza and Dristra—were established to monitor the trade with
the Pechenegs in Moldavia and Walachia. It is important to note that two
such offices existed in a relatively small area. By comparison, there were
four offices of kommerkiarioi on the western and northern coasts of the
Black Sea, at Constantinople, Develtos, Mesembria and Cherson.100 Paul
Stephenson has noted that this change in policy toward the Pechenegs
must have been initiated by John Orphanotrophous, Michael IV’s minister.101
The new policy was implemented through a peace treaty with the Pech-
enegs which was concluded in 1036.102
The new policy proved to be successful over the following decade,
and was most likely responsible for the relative prosperity of the cities in
Paradunavon. The Pechenegs lived in peace with the Byzantines when
Katakalon Kekaumenos was appointed katepano of Dristra (later on, a
Pecheneg called Koulinos would remember that at the time those people—
Pechenegs and Byzantines—intermingled).103 Nevertheless, the policy
failed when large number of Pechenegs moved into the empire, first as
allies, then as enemies. The conflict between a commander named Kegen
and the paramount chief named Tyrach spilled into the Empire, when the
latter sought refuge in the theme of Dristra in 1045, together with 20,000
followers. The Pecheneg refugees first moved into a swampy area, most
98 Frankopan 1997, 30–39.

99 Treadgold 1995, 84.
100 Antoniadis-Bibicou 1963, 208–209.
101 Stephenson 2000, 80–83, 114. See also Angold 1984, 1–11; Haldon 1999, 91.
102 Diaconu 1970, 51–55; Malamut 1995, 119.
103 Skylitzes, Constantine Monomachos, 22 (ed. Thurn, 469; transl. Flusin, 387; transl.
Wortley, 438).
likely Balta Ialomiţei or Borcea (itself a name of Turkic origin), not far
from Dristra. A linguistic study has shown to this day Balta Ialomiţei is
the area with the highest concentration of place names and river names
of Turkic origin (Pecheneg or Cuman) in the entire Walachia. Besides Der-
vent, no less than three other such names exist in the immediate vicinity
of Păcuiul lui Soare—Bugeac, Canlia and Galiţa.104
Kegen was eager to put his men to the service of the Byzantine Empire,
and the katepano Michael decided to send the Pecheneg chief to Constan-
tinople, to the emperor. In the capital, Kegen was baptized with Emperor
Constantine IX as sponsor at the baptismal font, and was given a new
name (John) and the title of patrikios. He became an ally of the empire
(symmachos). Kegen’s Pechenegs were all baptized in the waters of the
Danube, received land and three unidentified fortifications.105 The pur-
pose of this colonization was twofold: to stop further Pecheneg attacks
and to use them against their rivals beyond the Danube.106 In that respect,
the Pechenegs became a kind of stratiotai who owed military service in
exchange for the use of that land. The seals with the inscription Ιοάννες
μαγὶστρος καὶ ἄρχον Πατζινακὶας (one found at Silistra, another one pre-
served in a museum in München) belonged to Kegen.107 They point to
a later moment in the Pecheneg chieftain’s career, for the title of magis-
tros was higher than that of patrikios. The term Patzinakia indicates the
appearance of an autonomous Pecheneg territory, located somewhere
in the Danube region. This territory remained under the control of the
empire, because the title of archon was given only to rulers of autono-
mous regions on the periphery. As a matter of fact, Kegen’s seals are the
first official source using the title of archon for the Danube region.
Two large coin hoards of miliaresia discovered in Gigen may represent
the booty gathered by Pecheneg warriors. One of them includes 710 coins
(Basil II and Constantine VIII: 113; Romanos III: 1; Constantine IX: 596),
while the other has 23 coins struck for Constantine IX. Both hoards were
104 Iorga 1937, 42; Conea, Donat 1958, 158–159; Diaconu 1977 b, 62–63.
105 Skylitzes, Constantine Monomachos, 16 (ed. Thurn, 456–457; transl. Flusin, 378;
transl. Wortley, 427–428); Attaliates, ed. Bekker, 30–31 (ed. Pérez Martín, 24); Zonaras,
XVII, 26.1–9 (ed. Büttner-Wobst, 641–642); Necşulescu 1937, 125–127; Stănescu 1966, 51;
Diaconu 1970, 51–61; Barnea, Ştefănescu 1971, 126; Angold 1984, 15; Malamut 1995, 119–123;
Stephenson 2000, 90–91; Mănucu-Adameşteanu 2001 b, 98–99; Krumova 2005, 210–212;
Curta 2006, 296; Spinei 2006, 188, 190; Spinei 2009, 108–109.
106 Diaconu 1970, 57–61; Barnea, Ştefănescu 1971, 126.
107 Jordanov 1992 c, 79–82; Jordanov 2003 a, 138–142, nr. 59. 1; Jordanov 2006, 201–206,
nr. 307; Jordanov 2009, 465–466, nr. 1380; Spinei 2006, 191; Spinei 2009, 109.
124 chapter three
buried at a time when the fort, which was destroyed in 1032–1035, had
already been abandoned.108 The most significant archaeological evidence
for the beginning of the sedentization of the Pechenegs in this area comes
from a cemetery excavated in Odărci (535 graves). Inhumations began
there during the reign of Constantine IX, first on the site of fort destroyed
in 1032–1036. Many graves present clear Pecheneg features such as horse
bones or leaf-shaped pendants; trepanation was observed on 53 skulls,
and the individuals in questions must have been Pechenegs. Some graves
are certainly Christian, as Kegen’s men were baptized in the Danube.109
Another cemetery is known from Pliska, which included 40 graves dug
into the ruins of a church destroyed during the Pecheneg attack of 1036.
A group of three graves is also known from the ruins of the palatial
compound in Pliska. Artifacts associated with those graves are typically
Pecheneg, much like those in another cemetery from Preslav (20 graves
with horse gear and belt fittings).110 However, inhumation in cemeteries
is unknown among the Pechenegs living outside the empire. Once they
settled in the Byzantine province and were converted to Christianity, they
apparently adapted to the burial customs of the native population, and it
is probable that mixed marriages occurred as well. The next generation
descended from those Pechenegs, who were settled in 1046, would par-
ticipate in the rebellion of 1072. Pendants of Pecheneg origin were found
at Dristra and in forts such as Păcuiul lui Soare, Garvăn, Isaccea, Nufăru,
Mahmudia, and even at Varna. Other such objects are known from graves
in the countryside (Istria, Târguşor, Valea Dacilor, Vălnari). The memory
of this population was preserved by two place names, Pecineaga and Pece-
neaga, in the Constanţa and Tulcea counties, respectively.111
Settling barbarians (Carpi, Goths) on the Roman soil in the Danube
region was a practice known since Late Antiquity. In fact, there are strik-
ing parallels between the Pecheneg settlement and that of the Goths
in 376.112 Like the Gothic foederati, the Pechenegs quickly turned into
enemies and caused much trouble over the following decade. Relying on
those unreliable barbarians for the defense of the theme of Dristra must
have been a desperate measure to cope with the lack of local resources.
108 Metcalf 1979, 55; Penčev 1998, 76–95; Mănucu-Adameşteanu, Jordanov 2002, 130;
Oberländer-Târnoveanu 2003, 392.
109 Dončeva-Petkova 2007, 644–658.
110 Dončeva-Petkova 2003, 244–258; Michailova 2003, 259–266; Krumova 2005, 215–216;
Schmitt 2006, 482; Dončeva-Petkova 2007, 657.
111 Madgearu 2003 a, 52–55; Spinei 2006, 200.
112 Tăpkova-Zaimova 1975 b, 617–618; Schmitt 2006, 477.
The Pechenegs were definitely regarded as important in the defense of

the province, since they were given three fortresses and an autonomous
territory. Assigning of fortresses to autonomous allies had already been
used at that time by Emperor Constantine IX. He had applied the same
measures in the theme of Armeniakon (northern Asia Minor), where sev-
eral estates and fortifications were granted to the Norman mercenaries,
who were regarded as symmachoi like the Pechenegs, but who would later
rebel against the imperial authorities.113 Grants of forts became so com-
mon that Michael VII tried to slow down the process by prohibiting the
transfer for those forts through inheritance and by forcing those to whom
they had been granted to accept financial responsibility for the repairs. All
this clearly shows that the imperial power was in no position to maintain
the entire system of fortifications, and preferred instead to grant them to
various warlords who had the capability and the interest to take care of
them, even though, at least theoretically, the emperor was still the owner
of those forts.114 This “leasing” of the frontier defense began with the Pech-
enegs on the Danube.
Following the rise of Patzinakia, the Danube ceased to be a clear-cut
frontier between Byzantium and the barbarians. The Danube lands now
turned into a transition zone, a periphery in which the population and
the mode of living were half barbarian.115 John Haldon has argued that
in the region of this permeable frontier the Byzantine government intro-
duced a deliberate policy of depopulation.116 However, the archaeological
evidence shows the opposite. True, a certain degree of depopulation may
be surmised, but only for the 12th century, when some settlements were
abandoned. However, even in the 12th century, the city of Dristra contin-
ued to prosper, at least according to al-Idrisi’s testimony (to whom the
city was known as Daristar).117
113 Magdalino 1997, 27–29.

114 Oikonomides 1966, 413–417.
115 Independent from Haldon 1999, 63–64, I wrote of a “permeable frontier” in Madgearu
1999b, and my idea was then picked up by Stephenson 2003, 128.
116 Haldon 1999, 64.
117 Al-Idrisi (ed. Jaubert, 386; ed. Miller, 129); Tomaschek 1886, 300–301; Barnea,
Ştefănescu 1971, 164–165; Baraschi 1981, 316; Shepard 1979, 233–237. Chiriac 1993, 447–454
believed that Linokastro, another fort mentioned by al-Idrisi may have located north of
Hârşova. The name means “wool town” and may be compared with the local placename
Cetatea Oii (Sheep Fort). Chiriac appears to have built his theory on the wrong inter-
pretation advanced by Brătescu 1920, 29, who had located Linokastro in Dobrudja, even
though it is quite clear from the text that the town in question was near Rosokastron (now
126 chapter three
After becoming an ally of the emperor, Kegen decided to use this advan-
tage to take revenge on Tyrach, and he launched an attack in 1047 across
the Danube against the Pechenegs living north of the Danube. This action
only caused the migration of more Pechenegs into the empire. Another
factor contribution to this population movement was the pressure of the
Uzes from the east, who in turn had been driven away by the Cumans. The
Pechenegs crossing the Danube in great numbers could not be stopped by
a fleet of 100 triremes sent from Constantinople, since, again, the river had
frozen. According to Skylitzes, the Danube had no defense whatsoever.
Sources indicate that as many as 800,000 Pechenegs made the crossing,
but this is evidently an exaggeration. Faced with this crisis, the katepano
Michael asked for reinforcements from Constantinople. They came under
the command of Constantine Arianites and Basil Monachos (the dukes of
Adrianople and Bulgaria, respectively), but the Pechenegs were eventu-
ally forced to withdraw because of a disease to which many of them fell
victims. The survivors were settled in the region between Niš and Sofia.
Tyrach and some of his men were baptized, receiving offices in the Byzan-
tine army, like Kegen. Those measures were intended to pacify the region
and to turn the Pechenegs into a reliable and sedentary population.118 The
seal found in Vetren confirms the integration of Tyrach into the Byzan-
tine military structures.119 However, the policy ultimately failed, because
Tyrach’s Pechenegs rebelled when they were sent against the Seljuks in
1049, and in the process created havoc in Thrace and Macedonia for the
subsequent years until 1053. They established their base of operations in
a region rich in grazing fields, woods and water called Hekaton Bounoi,
which was located north and east of Preslav. Sent for negotiations, Kegen
was killed by the Pechenegs. After the battle at Preslav, in which the duke
of Bulgaria Basil Monachos was killed, Constantine IX concluded another
peace for thirty years peace, which stipulated payments of stipends to the
Pechenegs.120
Rusokastro, 25 km to the west from Burgas)(Al-Idrisi, ed. Jaubert, 388; ed. Miller, 132). See
Tomaschek 1886, 315–316 and Grămadă 1930, 215.
118 Skylitzes, Constantine Monomachos, 17 (ed. Thurn, 458–459, 465–473; transl. Flusin,
379–380; transl. Wortley, 429–430); Attaliates, ed. Bekker, 30–35 (ed. Pérez Martín, 24–27);
Zonaras, XVII, 26.10–22 (ed. Büttner-Wobst, 642–644); Necşulescu 1937, 127–128; Bănescu
1946, 128–129; Stănescu 1966, 52; Diaconu 1970, 62–65; Barnea, Ştefănescu 1971, 127–129; Ste-
phenson 2000, 90–91; Krumova 2005, 211–212; Kostova 2006, 589–590; Curta 2006, 296, 306;
Spinei 2006, 190–192; Schmitt 2006, 479–480; Krsmanović 2009, 77; Spinei 2009, 108–110.
119 Atanasov, Jordanov 1994, 41; Spinei 2006, 191.
120 Skylitzes, Constantine Monomachos, 21–25, 28 (ed. Thurn, 465–475; transl. Flusin,
384–392; transl. Wortley, 434–443); Attaliates, ed. Bekker, 37–43 (ed. Pérez Martín, 28–33);
Tyrach’s attack of 1047 (see Fig. 10) is responsible for destruction in the
Danube region and for the abandonment of Capidava (where the most
recent coins are those of Constantine IX)121 and Preslav (where the impe-
rial administration was restored in the 1060s, as seals from the administra-
tive building demonstrate).122 Destruction layers dated to that same time
have been found in Garvăn, where anonymous folles of type D appear
in several sunken-floored buildings that were apparently burned down.123
The same circumstances may have been responsible for the abandonment
of the fort in Tulcea.124 One of the largest coin hoards found in Paradu-
navon and dated to this period, which consists of 106 gold specimens (the
most recent from 1042/1055), was buried in Garvăn, most likely during this
attack.125 Other coin hoards dated to this period have been discovered in
Păcuiul lui Soare and Popina.126 On the former site, the gate was blocked
by means of a wall made of recycled materials.127 The fort at Krivina was
destroyed during those Pecheneg invasions and was abandoned until the
early 13th century.128 Both Varna and Mesembria suffered as well.129 Fol-
lowing the raids, a part of the population from the lowlands moved to
higher elevation on the southern slopes of the Stara Planina Mountains.
The archaeological excavations revealed a cluster of settlements in that
area, all dated to mid-11th century (one of them, Djadovo, was completely
excavated).130
The permeability of the Danube frontier allowed the settlement in 1059
of another Pecheneg group, which was fleeing the Uzes. According to
Kekaumenos, 64, 67 (ed. Spadaro, 96/97–98/99, 100/101); Zonaras, XVII, 26.23–24 (ed.
Büttner-Wobst, 644); Matthew of Edessa, I, 95 (ed. Dostourian, 80); Necşulescu 1937, 129;
Diaconu 1970, 62–65, 73–76; Barnea, Ştefănescu 1971, 127–128; Kazhdan 1977, 65–77 (who
clarified the date at which Kegen’s Pechenegs were settled); Angold 1984, 15–17; Malamut
1995, 119–128; Madgearu 1999 a, 435–436; Stephenson 2000, 91–92; Mănucu-Adameşteanu
2001 b, 100–105; Curta 2006, 297; Spinei 2006, 194–197; Schmitt 2006, 484–485; Jordanov
2009, 393. For Hekaton Bounoi see: Grămadă 1925–1926, 88–89; Diaconu 1970, 66–69, 73–76;
Madgearu 2003 a, 51–52; Schmitt 2006, 482.
121 Custurea 1995–1996, 301, 307; Mănucu-Adameşteanu 2001 a, 110; Curta 2006, 296.
122 Jordanov 1982 b, 41; Jordanov 1987 a, 95.
124 Oberländer-Târnoveanu 2003, 389; Mănucu-Adameşteanu 2001 a, 271, 411 (tabel 13),
412 (tabel 14); Mănucu-Adameşteanu 2010 c, 229–230.
125 Barnea 1971, 356; Metcalf 1979, 75; Barnea 1980, 274; Vîlcu 2008, 87–96 (who indicated
the right date of the coins, against the opinion of Mănucu-Adameşteanu 2001 a, 119, 127).
126 Mănucu-Adameşteanu 2001 a, 119; Mănucu-Adameşteanu, Jordanov 2002, 129.
127 Condurachi, Barnea, Diaconu 1967, 191; Mănucu-Adameşteanu 2001 a, 118.
128 Diaconu 1992 a, 180; Curta 2006, 297.
130 Borisov 2007, 76–78.
128
chapter three

Michael Psellos, “by their plundering and ravaging [the Uzes] compelled
them [i.e., the Pechenegs] to abandon their own homes and seek new
ones. So, at a time when the Ister was frozen over, they crossed as though
on dry land and emigrated from the Trans-Danubian territories to our
province. The whole nation was transported, bag and baggage, over our
borders, incapable of living at peace themselves, and bound to spread con-
sternation among their former neighbours.” The Pechenegs seem to have
taken advantage of an attack, which Andrew I, King of Hungary (1046–
1061) launched on the theme of Bulgaria. The attack stopped after negotia-
tions opened with the emperor in Serdica. Meanwhile, Isaac I Comnenos
defeated the Pechenegs (September 1059). In the course of those events, a
Pecheneg chief named Selté took a fortress, which Petre Diaconu identi-
fied with Loveč. However, if we are to take the account of the continuator
of Skylitzes at face value, the fort in question must have been somewhere
on the Danube. Petre Ş. Năsturel proposed the fort at Turcoaia, which fits
the description given in the source—a “steep rock” next to “swamps of
the Ister.” However, other sites such as Măcin or Hârşova may be equally
be regarded as good matches. At any rate, it is important to note that the
Pechenegs apparently had no problems crossing the river and that one of
their chieftains was able to take a fort on the frontier.131
When the Uzes crossed the Danube in 1065, they were met by the joint
forces of Nikephoros Botaneiates and Basil Apokapes. According to Attali-
ates and Michael Glykas, 600,000 Uzes crossed the Danube, but this is
undoubtedly an exaggeration.132 The figure of 60,000 given by Zonaras is
probably closer to reality. Another detail from Glykas’s account deserves
more attention: in the invasion took part the “most noble of the Pech-
enegs” (τῶν Πατζιναϰῶν οἱ ευγενέστεροι). This seems to imply that at the
head of the coalition of 1065 were some Pecheneg chiefs.133 The invasion
of 1065 created more havoc in Paradunavon. All sources agreed on the
large number of invaders, against whom it was necessary to summon an
army from the East, under the command of Basil Apokapes. Because even
131 Attaliates, ed. Bekker, 67 (ed. Pérez Martín, 51–52); Skylitzes Continuatus, 106–107;
Glykas, 602; Michael Psellos, VII, 67 (transl. Sewter, 241–242); Anna Comnena, III, 8.6
(transl. Sewter, 122–123); Zonaras, XVIII, 6.1–5 (ed. Büttner-Wobst, 671); Matthew of Edessa,
II, 5 (ed. Dostourian, 90–91). For the attack see: Iorga 1937, 81; Necşulescu 1939, 185; Gyóni
1943–1944, 88–92; Stănescu 1966, 53; Diaconu 1970, 76–78; Malamut 1995, 128; Curta 2006,
298; Spinei 2009, 110, 112.
132 Necşulescu 1939, 195; Malamut 1995, 129; Spinei 2009, 197.
133 Glykas, 605; Attaliates, ed. Bekker, 83 (ed. Pérez Martín, 63).
130 chapter three
with that, it was apparently not possible to contain the invasion, nego-
tiations were initiated to convince the Uzes to return to the lands north
of the Danube. They were convinced to do so only when pestilence and
starvation killed many of them.134 Again, the archaeological excavations
have confirmed the data from the written sources (see Fig. 11). Destruc-
tions that may be dated in 1065 are documented in Garvăn (where two
coin hoards were buried, with a total of 113 nomismata and 4 miliaresia)135
and Oltina.136 We know that the fort in Nufăru put up some resistance, for
the local commander received a message from Basil Apokapes during the
operations. Southern Paradunavon was also affected. The series of coin
finds in Pliska stops at this particular moment (see chapter II), and coin
hoards were buried under the same circumstances in Păcuiul lui Soare
and 23 August (Constanţa county).137 That three seals of Basil Apokapes
alone were found within a relatively short distance from each other in
134 Attaliates, ed. Bekker, 83–85 (ed. Pérez Martín, 63–64); Skylitzes Continuatus, 114–
115; Zonaras, XVIII, 9.1–9 (ed. Büttner-Wobst, 678–679); Necşulescu 1939, 193–196; Stănescu
1966, 54; Barnea, Ştefănescu 1971, 133–135; Malamut 1995, 129; Spinei 2006, 286–287; Spinei
2009, 114.
135 Ştefan, Barnea 1967, 29; Diaconu 1970, 79–80; Barnea 1971, 356; Barnea 1973, 292,
305, 315; Barnea 1980, 245; Mănucu-Adameşteanu 2001 a, 119–120, 127, 128, 213–216; Vîlcu
2008, 89.
136 Custurea 2000–2001, 589, 593.
137 Diaconu 1976 b, 235–239; Mănucu-Adameşteanu 2001 a, 128.
Silistra, Bradvari, and Popina, is remarkable. Those messages show that

troops were sent to the west from, perhaps because the duke expected
an attack from the valley of the Mostiştea, at a point of crossing in use
from the Antiquity and throughout the Middle Ages.138 Popina is in fact
located just 25 km to the west from Dristra, Bradvari at 23 km south-west
of Dristra, on the way to Pliska. The Uzes may have also used the fords at
Dervent, Garvăn, and perhaps also Isaccea.
The weakness of the Byzantine domination on the Lower Danube
results from the secession of the Danube province for about two decades,
between 1072 and 1091. While several centrifugal movements and rebel-
lions are known to have taken place in the Empire at that time, in this
case the Pechenegs played a key role both at the beginning and in the
course of the mutiny, which started as a reaction to the lack of protection
from the central government and deteriorated into complete secession.139
During the secession, the Byzantine frontier, if one can still speak of
such a thing, was limited to some points in northern Dobrudja, foremost
among them being Isaccea, the new residence of Paradunavon after the
occupation of Dristra by Tatós. When Nikephoros III Botaneiates became
emperor in 1078, since he had previously been a katepano of Paradunavon,
the relations between the native inhabitants and the Pechenegs began
to change. Attaliates reports that a delegation of “Scythians” (an archaic
name no doubt referring to people from the lands once within the Roman
province of Scythia Minor) came to the emperor to ask for forgiveness and
to promise that they would never again attack the empire together with
the Pechenegs. This delegation may be dated to the fall of 1080.140 Attali-
ates’ account shows that the local population broke the alliance with the
Pechenegs, because the initial reason for the rebellion had disappeared
after the dethronement of Michael VII. This change of attitude may have
also been influenced by the fact that the new emperor was a former duke
in Paradunavon, whom many in the region must have known personally.
Archaeologically, this period is characterized by the interruption of
the coin series in Garvăn, Turcoaia, and Hârşova—all three sites without
any finds of anonymous folles of the K type, which are dated between
138 Developing an idea from Iorga 1937, 90, Iosipescu 1994, 257, 259 suggested that the
so-called Bordoni, a warrior group attested in 1187, were located on the Mostiştea valley,
because their name derived from the Slavic brod (“ford”) that could be linked with to Slavic
most (“bridge”).
139 Stănescu 1966, 60–61; Tanaşoca 1973, 80.
140 Attaliates, ed. Bekker, 302–303 (ed. Pérez Martín, 216); Skylitzes Continuatus, 185;
Stănescu 1966, 59–60; Meško 2011, 137.
132 chapter three
1085 and 1092. Turcoaia was reoccupied shortly after that, as indicated
by coins struck from 1093–1094 onwards. In Garvăn, the settlement out-
side the walls was abandoned in the 1080s. A coin hoard of 15 nomismata
struck for Emperors Romanos IV and Michael VII was found inside the
fort.141 The fort at Mahmudia may have also been abandoned after 1072,
since the most recent coins from that site are issues of Romanus IV.142 The
defense system in northern Dobrudja was therefore seriously damaged,
with only Isaccea and Nufăru remaining in operation (see fig. 12). Farther
to the south, the fort at Popina was destroyed around 1080,143 while that in
Oltina continued in use, as indicated by a seal of Gregory Mavrokatakalon
dated between 1087 and 1091. The fort in Păcuiul lui Soare was apparently
spared, as no destruction layer dated to 1070–1080 has been archaeologi-
cally identified. Nonetheless, the general anxiety in the area is reflected in
the burial of a small hoard 33 coins, which was found in a sunken-floored
building.144 The lack of any traces of destruction may be explained as indi-
cating that together with Dristra, Păcuiul lui Soare had fallen earlier on
into the hands of the rebels.
The destruction may also be attributed to the invasion of those whom
Anna Comnena calls “Scythic people”—either Cumans, or another Pech-
eneg group. The newcomers crossed the Danube and immediately opened
negotiations with Tatós and two other chiefs, named Sesthlav and Satza,
who, according to Anna Comnena, controlled “Vitzina and the remaining
areas.”145 It is not altogether clear whether both Sesthlav and Satza were
rulers of Vitzina, or only one of them. While Sesthlav is definitely a Slavic
name (perhaps Časlav), Satza is of Turkic origin. A Cuman prince called
Sakz’ is mentioned in the Russian Primary Chronicle, as having been killed
in battle with Vladimir Monomakh at Bela Vezha in 1093. The name may
141 Barnea 1971, 356; Metcalf 1979, 75; Barnea 1980, 245; Mănucu-Adameşteanu 1995,
354–356; Mănucu-Adameşteanu 2001 a, 134–136, 216–218; Mănucu-Adameşteanu 2010 a,
194; Mănucu-Adameşteanu 2010 d, 446.
142 Mănucu-Adameşteanu 1992, 400; Oberländer-Târnoveanu 1998, 80; Mănucu-
Adameşteanu 2001 a, 128–129; Mănucu-Adameşteanu 2010 e, 226, 228.
144 Conovici, Lungu 1980, 397–402; Custurea, Talmaţchi 2011, 308. Other coin hoards
were hidden at Belene, Silistra, Giurgengik and Plopeni (Constanţa county): Metcalf 1979,
75–76; Custurea, Talmaţchi 2011, 53, 313, 330–331.
145 Anna Comnena, VI, 14.1 (trad. Sewter, 212): a “Scythian tribe, having suffered inces-
sant pillaging at the hands of Sarmatians, left home and came down to the Danube”;
Brătianu 1935, 16; Necşulescu 1937, 139–144; Gyóni 1944, 87–100; Diaconu 1970, 112–115;
Barnea, Ştefănescu 1971, 137–146; Cihodaru 1978–1979, 285; Malamut 1995, 130.

the danube frontier of byzantium (1000–1204)
12. The fortifications affected by the rebellion of 1072–1091

133
134 chapter three
thus have been in use among members of the Pecheneg or Cuman aristoc-
racy, and it is quite possible that Satza mentioned by Anna Comnena is
one and the same person as that later be killed by Vladimir Monomakh.146
The place name Isaccea has in fact the same origin, and some have even
suggested that the name derives from Satza.147 The Tatar name Saqčy first
appears on coins struck from 1286 onwards in the name of the Mongol
ruler Nogai (who had his residence in Isaccea between 1296 and 1301), and
later on coins of the Genoese colony established there. During the first
decades of the 14th century, the name Sakdji was mentioned in the his-
tory of Baybars al-Mansuri (Zubdat al-fikra) and in Abulfida’s Geography
of 1325.148 In the 18th century, Romanians employed the name Sacce,149 its
present form Isaccea having developed as the result of a confusion with
the personal name Isac. It therefore seems likely that the city took a name
of Turkic origin under the Mongols. In other words, whatever the origin
and meaning of the name Saqčy from which Sacce/Isaccea derives, there
is absolutely no reason to believe that it derived from the name of an
obscure chief mentioned only by Anna Comnena. Another Vitzina, “a river
which flows down from the neighboring hills,” is mentioned in Anna Com-
nena’s Alexiad somewhere north of the pass of Sidera (Riš), and south of
Pliska. That river has been identified with Kamčija, also called Tiča, which
appears in Constantine Porphyrogenitus as Ditzina (the Bulgarian form
was most likely Dičina). The same name (Disina or Desina) was later given
by al-Idrisi to a city at a distance of four days to the east from Barisklafa
(Little Preslav).150 This is most likely Venzina, placed between Kamčija
and Varna, a city destroyed by the Ottoman Turks in 1388, as mentioned
in the chronicle of Mehmed Nešri.151 In Anna Comnena’s account, Vitzina
appears as almost of the same importance as Dristra, and, moreover, the
residence of a rebel. As a consequence, some have proposed this to be the
146 Russian Primary Chronicle, 213. For the fights with the Polovtsy (Cumans) see
Franklin, Shepard 1996, 266, 272.
147 Brătescu 1920, 31; Brătianu 1935, 24–26; Cihodaru 1978–1979, 295; Brezeanu 2002, 38
(who considered that Satza reflects a Greek form of a Romanian placename Saca); Ciocîl-
tan 2011, 415.
148 Brătianu 1935, 39, 45, 70; Baraschi 1981, 336; Oberländer-Târnoveanu 1995–1996,
200–211; Vásáry 2005, 90, 91, 97; Spinei 2009, 32.
149 Giurescu 1971, 258.
150 Constantine Porphyrogenitus, IX, 101 (ed. Moravcsik, 62/63); Anna Comnena, VII,
3.1 (transl. Sewter, 222); Al-Idrisi (ed. Jaubert, 386, 397; ed. Miller, 123, 129, 132); Tomaschek
1886, 311–312; Brătescu 1920, 9, 30; Brătianu 1935, 12–18; Brătianu 1942, 146–147; Gyóni 1952,
505–512; Beševliev 1962, 69; Cihodaru 1978–1979, 283; Beševliev 1985, 21–22.
151 Brătianu 1935, 83–84; Cihodaru 1978–1979, 284–285.
first mention of the famous and enigmatic city of Vicina or Vecina, known
to have been in existence and to have prospered during the 13th and 14th
centuries on an island somewhere on the Danube. This Vicina has noth-
ing to do with the place mentioned by al-Idrisi near the Kamčija river, but
could indeed be the same as that mentioned by Anna Comnena.152
The location of Vicina was and remains one of the most vexed ques-
tions of the Romanian historiography. This is not the place for an exten-
sive presentation of the problem, but some considerations are necessary
to clarify what could be said about the 11th-century Vitzina, which appears
in Anna Comnena’s Alexiad. In the late 13th century, the city of Vicina was
already so important to become the residence of a metropolitan bishop,
first mentioned in 1285 (Vitzina in the Greek spelling). Already in 1296, the
portolan chart entitled Il compasso da navigare calls the Danube flume de
Vicina. The city was recovered by the Byzantines shortly after 1261, and
become a Genoese colony, but fell under Mongol rule around 1338. The
subsequent decline coincided in time with the rise of Chilia and Brăila as
rival trading centers.153 It is quite possible, although not certain, that the
fortress mentioned by Anna Comnena is the same as the famous medieval
city. In that case, it remains to be explained why Vicina was not men-
tioned in any of the rather abundant sources between the 1080s and 1281
(the year of its first mentioned in the notarial documents of the Genoese
colony in Pera). The only exception may be an epigram celebrating the
exploits of the general John Dukas in the war of 1167 against Hungary
(see below). The text preserved in the manuscript Marcianus Graecus 524
mentions that Dukas crossed the Danube at Vidin, which is in contradic-
tion with the account of the same expedition in John Kinnamos, accord-
ing to whom the army commanded by John Dukas entered Hungary via
Moldavia. Some have therefore proposed that the author of the epigram
mistook Vidin for Vicin, that is Vicina.154 However, it is equally possible
that Kinnamos was wrong.155
152 Still, some historians considered that Desina from the work of Al-Idrisi and Vic-
ina from Dobrudja were the same place (Bromberg 1937, 177; Stănescu 1970, 124; Barnea,
Ştefănescu 1971, 164; Kuzev 1981, 272; Cihodaru 1978–1979, 284–289; Todorova 1984, 431;
Atanasov 1994, 112–113; Malamut 1995, 135).
153 For the history of Vicina see, for instance: Brătianu 1935, 9–96; Brătianu 1942, 144–
154; Cihodaru 1978–1979, 295–302; Todorova 1984, 429–441; Năsturel 1987, 145–171; Iliescu
1994, 229–236.
154 Diaconu 1978, 103; Baraschi 1981, 317; Cândea 1999, 154 (identified with Măcin).
155 Năsturel 1969, 180–181.
136 chapter three
The name Vicina or Vecina derives from the Romanian or the Italian
(in any case, Romance) word for “neighbor,” which suggests that the city
developed in close proximity of an older one. If the fort mentioned by
Anna Comnena was really the same as the city first mentioned in 1281,
then the name must have been of Romanian, and not Italian origin. The
14th-century sources about Vicina are so vague that its location was much
disputed. Leaving aside such impossible identifications as Măcin (because
of a supposed name similarity, but without any support in the archaeolog-
ical evidence),156 or Nufăru157 (which was in fact Presthlavitza), two major
theories may be distinguished. According to one of them, Vicina was on
an island near Isaccea or Isaccea itself (some have even maintained that
that city had two names).158 Aerial photography revealed a site of urban
character (the street grid is still visible) about 1 km south from the ancient
Noviodunum in Isaccea, which may be for Vicina.159 Hydrological stud-
ies also revealed the existence of an old, now sunken island farther to
the west, next to the mouth of the Prut river.160 A different theory has
been put forward by Petre Diaconu, who identified Vicina with Păcuiul lui
Soare.161 A key argument in this theory is the distance between Vicina and
the mouths of the Danube, which is given in the portolan chart of 1296
as 200 Genoese miles. That roughly corresponds to the distance between
Păcuiul lui Soare and the Danube Delta, but there are other problems with
this theory. As Petre Năsturel has noted, a notarial act issued in Chilia on
October 18, 1360 mentions a ship departing from Chilia to Vicina and then
back to the Black Sea, which implies that those three points of destina-
tion were not too far from each other, given that sailing upstream in the
winter was very difficult, if not impossible. This further means that Vicina
must have been somewhere on the maritime segment of the Danube, i.e.,
between the Danube Delta and the present-day city of Galaţi. The main
commodity exported in Vicina was wheat, which required land trans-
portation, too difficult through the Bărăgan steppe in eastern Wallachia,
but much easier along the valleys of the Moldavian rivers Siret and Prut.
Moreover, establishing a new metropolitan bishopric in close proximity to
156 Tomaschek 1886, 30; Brătescu 1920, 30; Bromberg 1938, 19–29; Gyóni 1943–1944, 20;
Ciocîltan 2011.
157 Atanasov 1994, 114–128.
158 Grămadă 1924, 458; Giurescu 1971, 258; Vásáry 2005, 161.
159 Rada et alii 1988, 203–204.
160 Botzan 1992, 61–73. A similar point of view was sustained by Cihodaru 1978–1979,
294–295 (between Isaccea and Somova).
161 See especially Diaconu 1976 a.
Dristra made no sense whatsoever. When Vicina began to decline, its posi-
tion was quickly taken by Brăila, the terminal of the trade route connect-
ing Transylvania to the Danube across the Carpathian Mountains during
the second half of the 14th century. In conclusion, the best solution is to
look for Vicina on a now disappeared island close to Garvăn, or to the
south from the old town in Isaccea.
The restoration of imperial authority in Paradunavon was the result
of the victories Alexios I Comnenos obtained against the Pechenegs. After
the conclusion of the general crisis caused by internal rebellions and the
Norman invasion, the emperor turned to the Pecheneg problem and the
secession of Paradunavon. The general Gregory Pakourianos (the com-
mander of the western army) won a victory against the Pechenegs in the
early months of 1083 at an unknown location (there is a lacuna in Anna
Comnena’s text at this point),162 but the Pechenegs mauled another army
sent against them, and the generals Gregory Pakourianos and Nicholas
Branas were killed in battle in the pass of Veliatova. The rebels were
then stopped near Philippopolis (Plovdiv) by another army led by Con-
stantine Humbertopoulos, the commander of an elite troop of Flemish
mercenaries.163 The Pechenegs continued to raid the lands south of Stara
Planina Mountains, sometimes in cooperation with other rebels, not just
with Nikephorus Basilakios. In 1086, they forged an alliance with Traulos,
the chief of the rebel Paulicians in Thrace, while in 1087 they raided
together with the former Hungarian king Salomon and the Cumans.
The crisis in Paradunavon did not end until 1091 mainly because of the
Pechenegs, and it is possible that the natives were more inclined during
those years to side with the central power against the Pechenegs. An even
greater threat came in the spring of 1087, when under the command of
Tzelgu, a large number of Pechenegs crossed the Danube together with
a group of Cumans and with the Hungarian followers of the former king
Salomon, who had meanwhile been dethroned by Ladislas I (1077–1095).
The coalition ravaged the entire region between the Danube and the
Stara Planina Mountains, before entering Thrace. After that, they with-
drew to southern Dobrudja, Tatós’s power center of Tatós.164 This time,
162 Frankopan 1996, 278–281; Spinei 2006, 202–203; Meško 2011, 142.

163 Anna Comnena, VI, 14, 3–4 (transl. Sewter, 213); Diaconu 1970, 116; Barnea, Ştefănescu
1971, 146–147; Birkenmeier 2002, 71; Stephenson 2000, 101; Curta 2006, 290, 300.
164 Anna Comnena, VII, 1.1 (transl. Sewter, 217–218); SRH, I, 409; Iorga 1937, 40, 86;
Necşulescu 1937, 144–145; Stănescu 1966, 60–63; Diaconu 1970, 117; Diaconu 1978, 35–36;
Birkenmeier 2002, 71; Stephenson 2000, 102; Mănucu-Adameşteanu 2001 a, 149; Kosztolnyik
2002, 385; Curta 2006, 300; Spinei 2006, 203–204, 367.
138 chapter three
the pursuing Byzantine army commanded by Nicholas Mavrokatakalon

was successful. Tzelgu and Salomon were killed, and the army of Alexios
I counter-attacked in the summer and fall of 1087. The strategy employed
in this campaign seems to have been the same joint naval and land oper-
ations that Leo VI applied in 895 and John Tzimiskes in 971. The field
army advanced through Thrace, reached Pliska and then Dristra. At the
same time, the navy commanded by George Euphorbenos was sent to the
Danube, in an attempt to repeat the successful operation of 971.165 It is
possible that the message sent by Adrian Comnenos (mentioned in the
previous chapter) concerned the operations of the navy, given that Isac-
cea was a major anchorage.
The itinerary of the army is marked by imperial seals and gold coin
hoards found at Golyam Izvor, Melnitsa, Zlati Voyvoda, Preslav, Kirkovo,
Drandar, Vodno, Gurgendzhik, Iširkovo, and Păcuiul lui Soare.166 During
the siege of Dristra, three hoards of folles, one of 3,226 specimens, were
buried there.167 In the first phase of the battle, the Byzantine army was
able to break through the outer precinct, but the citadel resisted. The siege
ultimately failed because Tatós was able to launch a night attack against
the imperial camp, which had been set at 24 stadia (about 4.5 km) from
the city, on the site of an old Roman vicus. Alexios I’s attempt to reset the
siege ended in disaster. In the end, the victory was secured by a Cuman
force summoned to Dristra from the lands north of the Danube by Tatós.
The Byzantine lines were broken and quickly withdrew.168 However, when
the Cumans and the Pechenegs began to quarrel over the booty, the latter
were overwhelmed, and feld to a place called Ozolimne. In a note to her
main narrative, Anna Comnena explains that “the lake now called by us
Ozolimne is in diameter and circumference very big, in surface area not
inferior to any other lake described by geographers. It lies beyond the
‘Hundred Hills’ and into it flow very great and noble rivers; many ships
and large transport vessels sail on its waters, from which one can deduce
165 Anna Comnena, VII, 2–7 (transl. Sewter, 220–222); Necşulescu 1937, 145; Diaconu
1970, 117; Malamut 1995, 136–137; Mănucu-Adameşteanu 2001 a, 150; Kostova 2006, 590–593;
Spinei 2006, 205; Meško 2011, 134–135, 143. He was not the “governor of Paradunavon”, as
believes Mărculeţ 2005, 94.
166 Yotov 2008 b, 172–176; Yotov 2008 c, 262; Jordanov 2009, 101–103; Valeriev 2010,
653–657.
167 Custurea, Talmaţchi 2011, 339, 341–342.
168 Anna Comnena, VII, 3–4 (transl. Sewter, 223–229); Barnea, Ştefănescu 1971, 148–149;
Diaconu 1970, 117–118; Malamut 1995, 137–138; Birkenmeier 2002, 72–74; Curta 2006, 290,
300–301; Spinei 2006, 205–206; Yotov 2008 c, 257–263.
how deep the lake is. It has been called Ozolimne, not because it emits
an evil or unpleasant odor, but because an army of Huns once visited the
lake and the vernacular word for Huns is ‘Ouzi’. The Huns bivouacked
by the banks of this lake and the name Ouzolimne was given it, with the
addition of the vowel ‘u’. No congregation of Huns in that area has ever
been mentioned by the ancient historians, but in Alexius’ reign there was
a general migration there from all directions—hence the name.” Although
this description makes it very clear that Ozolimne was navigable, Petre
Diaconu tried to demonstrate that this was a swampy area near Preslav,
for which he took Hekaton Bounoi as a reference point. Most other his-
torians believe that Ozolimne is in fact Balta Ialomiţei, the marshy area
in which both Pechenegs and Uzes stayed when arriving at the Danube.
Needless to say, the etymology advanced by Anna Comnena is wrong, as
there are several other marshes called Ozolimne, a word which means
“bad-smelling lake.”169
After Alexios I’s campaign failed, the Pechenegs took the initiative
and for four years raided deeply into Thrace, all the way to the walls of
Constantinople (1088–1091). The situation became desperate for Emperor
Alexios, when the Pechenegs established contacts with Tzachas, the Seljuk
emir who had already occupied a part of Asia Minor and had attacked a
number of islands in the Aegean Sea. With a fleet built with the assis-
tance of a Greek renegade, Tzachas intended to besiege Constantinople,
and his cooperation with the Pechenegs is strikingly similar to the joint
Avar-Persian attack on that city in 626. Alexios I succeeded to use the ten-
sions between Cumans and Pechenegs to attract the latter onto his side.
With their assistance, he then inflicted a major defeat upon the Pech-
enegs at Lebounion, in Thrace, on the Maritsa valley (April 29th, 1091).
The involvement of the Cumans was a key to success, because their light
cavalry encircled the Pechenegs at a crucial moment in the battle.170 The
duke of Paradunavon (either Gregory Mavrokatakalon or Leo Nikerites)
set up contacts with the Cumans north of the Danube. In the aftermath of
Lebounion, the imperial authority was reestablished over the Danube.
169 Anna Comnena, VII, 5 (transl. Sewter, 229); Grămadă 1925–1926, 85–97; Iorga 1937,
69; Conea, Necşulescu 1937, 146; Donat 1958, 159; Diaconu 1970, 121–129; Spinei 2006, 287.
170 Anna Comnena, VII, 6–11; VIII, 1–6 (transl. Sewter, 230–260); Zonaras, XVIII, 23.1–6
(ed. Büttner-Wobst, 740); Matthew of Edessa, II, 90 (ed. Dostourian, 155); Ostrogorsky 1956,
380–381; Ahrweiler 1966, 184–185; Diaconu 1970, 117–120, 130–132; Angold 1984, 110–111; Mala-
mut 1995, 135–142; Treadgold 1997, 617; Birkenmeier 2002, 70–77; Curta 2006, 301; Spinei
2006, 208–209, 367–368; Spinei 2009, 120.
140 chapter three
The Byzantine army at Lebounion also included a tagma of 5,000

mountain people whom Anna Comnena calls Areimanioi.171 Some have
proposed that those were Vlachs, who call themselves Armâni.172 But
Ἀρειμανής appears in classical texts, such as Strabo, with the meaning of
“full of warlike frenzy” (possessed by the god Ares), as correctly translated
in the English edition of the Alexiad.173 Anna was very fond of such clas-
sical references, by which she wanted to show her sophistication. This,
however, does not exclude the possibility of those warriors being Vlachs,
as they are said to have come from the mountains. It is known that on
the eve of Lebounion, the emperor recruited heavily from among Bulgar-
ians and Vlachs in southern Thrace, an area under the command of Nike-
phoros Melissenos174 A seal was found in Isaccea of one George, “strategos
of the Vlachs.”175 Ion Barnea believed this man to be the same as George
Dekanos who, being under suspicion of conspiracy, was sent in 1091 to Leo
Nikerites, the duke of Paradunavon, “apparently to assist him in protect-
ing the Danube area, but in fact he was sent so that Nikerites might keep
an eye on him.”176 Be as it may, the first mention of a “strategos of the
Vlachs” proves that a military unit—a tagma—existed, which was made
up of Vlachs. In the late 11th century, the title of strategos was bestowed
mostly on commanders of cities, but could also be used in a more general
sense as “military commander, general.” Military units made up of Pech-
enegs settled within the empire or Norman mercenaries became particu-
larly important during the 11th century.177 This Vlach unit could have been
recruited from the Stara Planina Mountains, which had a concentration
of Romanian population, the same area in which the Vlach brothers Peter
and Asan would start their rebellion in the late twelfth century.
It is worth mentioning that Alexius I’s war with the Pechenegs has
nothing to do with the story inserted in the 13th-century “Chronicle of
Laon” published by Krijnie Ciggaar. According to that source, a group
of English knights in the Byzantine army was sent to the northern coast of
the Black Sea to fight against the heathens, most likely in 1083. This means
171 Anna Comnena, VIII, 5.1 (transl. Sewter, 257).

172 Iorga 1937, 88; Birkenmeier 2002, 76.
173 Gyóni 1951, 251–252.
174 Anna Comnena, VIII, 3.4 (transl. Sewter, 251); Gyóni 1951, 241–243, 249–251; Vásáry
2005, 20.
175 Barnea 2001, 103–104.
176 Anna Comnena, VIII, 9.6–7 (transl. Sewter, 268). A seal of George Dekanos was
found in Sozopol (Jordanov 2006, 122–123, nr. 161).
177 Ahrweiler 1960, 33–35, 90; Treadgold 1995, 116–117.
that they were already in Constantinople during the Pecheneg siege of

1075. The text mistakenly took Emperor Michael VII for the better-known
Alexios I.178 The knights are said to have liberated the region, which now
became a “New England,” and to have built several forts there. An Icelan-
dic saga based on English sources even gives the (English) of those forts.
Even though, as early as 1974, Jonathan Shepard has demonstrated on the
basis fo place names of English origin preserved on later maps that this
episode concerns an area to the east from the Crimea,179 Răzvan Theodo-
rescu still advanced the idea that the region in question, which is called
Domapia in the chronicle, was in fact Dobrudja or north-eastern Bulgaria,
an idea later shared by Krijnie Ciggaar.180 The details about the length of
the barbarian rule in Domapia do not really fit with the known history of
Paradunavon, in which the Pecheneg effectively ruled only between 1072
and 1091. Moreover, not all Pechenegs were pagan throughout that period.
Judging from the chronicle’s account, contact between Constantinople
and Domapia was possible only by sea. Finally, it is possible to derive the
name Domapia from the name of the river Don. Thus, the area settled by
the Englishmen was Zichia on the north-eastern coast of the Black Sea,
where 14th- to 16th-century portolan charts placed cities with such names
as Londina and Susaco (from Saxon). A speech of Manuel Straboromanos
commemorates in a few words the restoration of the Byzantine rule over
the Cimmerian Bosporus under Emperor Alexios I Comnenos, and a lead
seal of an archon of Zichia has been dated to the same period.181 The Eng-
lish knights requested priests of the Latin rite from Hungary. This was
seen as a proof that Domapia was next to Hungary, but in reality that
country was the closest Catholic state to Crimea, and not just to the Dan-
ube area. Based on the supposed location of Domapia near the Danube,
Ferenc Makk even developed the theory of a crusader state established
by English knights, but then used by the Hungarian king Ladislas I (1077–
1095) as a launchpad for his expansionist policy into the Balkans.182 It was
the same Jonathan Shepard who noted that no evidence exists for such
a policy.183 In conclusion, the so-called “New England” has nothing to do
with the history of Paradunavon.184
178 Ciggaar 1974, 301–342.

179 Shepard 1974, 18–39.
180 Theodorescu 1981, 637–646.
181 Bănescu 1941, 59, 73–74; Shepard 1974, 22–23, 25, 28.
182 Makk 1994 a, 59–67; Makk 1999, 79–84.
183 Shepard 1999, 68.
184 For a detailed discussion of this issue, see Madgearu 2003 b, 137–144.
142 chapter three
Shortly after the restoration of the Byzantine authority over Paradu-

navon, the emperor was confronted with another problem on the Dan-
ube frontier, namely the Cuman attack of 1095. A Cuman chieftain named
Tugorkan or Togortak offered his support to an impostor pretending to be
Nikephoros Diogenes, Emperor Romanos IV’s son, for whom he was now
trying to obtain the power in Constantinople. An army was sent to Isaccea
in order to stop the Cumans, who had occupied Paristrion.185 Proof for
that is the seal of Katakalon Tarchaneiotes, who is known to have partici-
pated in this campaign against the Cumans.186 The troops were paid with
coins struck in the Isaccea mint, which had produced silver specimens
on behalf of Emperor Alexios I through the monetary reform of 1092. The
mint then continued to operate throughout Alexios’ entire reign.187
Scholars used to believe that northern Dobrudja was spared by the
Cuman invasion, as Tugorkan may have crossed the Danube at Dervent.
To be sure, the fort at Păcuiul lui Soare, which is the closest to the Der-
vent ford, was destroyed and abandoned throughout the 12th century.188
The attack then moved to the south, as indicated by the large hoard from
Kalipetrovo.189 New studies have however shown that Isaccea was also
destroyed in the late 11th century, while Păcuiul lui Soare may have been
attacked, but was certainly not destroyed in 1095. Instead, its annihilation
came only with the last Pecheneg invasion into the Empire, that of 1122
(see below). It is therefore likely that the Cumans of 1095 crossed the Dan-
ube at Isaccea. The settlement outside the walls was abandoned on that
occasion, and became a cemetery in the 12th–13th centuries.190 Another
Cuman group may have crossed the Danube at Axiopolis, given that that
fort is supposed to have been destroyed in 1095. The supposition is based
on a text mentioning the transfer of the bishop of Axiopolis, who had
remained without see. The general context of those events fits with the
year 1095, and consequently with the Cuman attack.191
185 Anna Comnena, X, 2 (transl. Sewter, 297–302); Zonaras, XVIII, 23.26 (ed. Büttner-
Wobst, 744); Russian Primary Chronicle, 180; Diaconu 1978, 41–51; Mănucu-Adameşteanu
2001 c, 110–111 (who established the right date of the attack, 1095, not 1094); Spinei 2006,
374–375; Spinei 2009, 121–122. About this pretender, see Frankopan 2005, 147–165.
186 Barnea 1997, 358. Three other seals of Katakalon Tarchaneiotes have been found in
Dristra and in two unknown places in north-eastern Bulgaria (Jordanov 2006, 398–399,
nr. 701–703).
188 Diaconu 1978, 41–55; Mănucu-Adameşteanu 1995, 355–359.
189 Oberländer-Târnoveanu 1992, 42–43; Custurea, Talmaţchi 2011, 246–247.
190 Mănucu-Adameşteanu 2001 c, 111–115.
191 Popescu 1994, 421–438.
The Cumans that reached the lands south of the Stara Planina Moun-
tains were beaten by a Byzantine army from Anchialos. In that fort, a
leader of the Vlachs came to the emperor, named Pudilos (which may
render the Romanian name Budilă or Bădilă). Pudilos brought news about
another group of Cumans who had meanwhile crossed the Danube.192
This implies that some Vlachs were used by the Byzantines at the time to
watch over the mountain passes or as scouts.193 Romanians are known to
have acted as spies for the Byzantine army on other occasions as well, for
example in the war of 971 against Svyatoslav and against the Pechenegs
in 1049. Pudilos/Bădilă was a local ruler from the Danube lands,194 but he
may have also had an office in the Byzantine army. Other Vlachs from the
Stara Planina Mountains were meanwhile hostile to the Byzantines, and
told the Cumans about routes to bypass the defended mountain passes.195
It is interesting to note in this respect what a military manual composed
by Emperor Basil II himself has to say about such situations: “The men
we call guides are not simply men who know the roads, for the lowliest
peasant can do that, but men who, in addition to knowing the roads, are
able to conduct the army through the mountain passes.”196 It is, however,
a gross exaggeration to treat the Romanians and the Cumans of 1095 as an
anti-Byzantine coalition, perhaps a rehearsal for the events of 1185–1186.197
In the late eleventh century In 1095, the Vlachs had no reason to rise in
rebellion against the state, and some of them were even recruited for the
Byzantine army, while other were opportunists who preferred to take
advantage of the Cuman attack.
That the Cumans were capable to bypass the defended mountain passes
had major consequences for the future strategy of the Byzantine army.
This seems to have shifted the emphasis from defending the line of the
river Danube to defending the points of access across the Stara Planina
Mountains. The attack of 1095 has demonstrated the key role mountain
passes had in the defense of all routes leading to Constantinople, by far
great than the fords across the Danube. For that reason, it was necessary
to obtain and to maintain the loyalty of the mountain population. The
192 Anna Comnena, X, 2.6 (transl. Sewter, 298); Giurescu 1931, 118; Brătianu 1935, 26;
Iorga 1937, 91; Brătescu 1920, 13; Gyóni 1952, 502–503; Barnea, Ştefănescu 1971, 155; Vásáry
2005, 21; Spinei 2006, 375–376.
193 Stănescu 1989, 28. Spinei 2006, 375.
194 Diaconu 1977 a, 1898; Diaconu 1978, 57.
195 Anna Comnena, X, 3.8 (transl. Sewter, 299); Gyóni 1952, 496–501.
196 Anonymous Book on Tactics, 290/291; Dagron, Mihăescu 1986, 251.
197 Vásáry 2005, 21.
144 chapter three
Byzantine army already had the experience of recruiting from the moun-
tain population of Albanians in Epirus, who then joined the rebellion of
Nikephoros Basilakios in 1078. However, during Bohemond of Taranto’s
invasion of 1108, the Albanians from the mountains blocked the narrow
passes along Via Egnatia with palisades and timber barriers.198 The use
of highlanders for military operations involving the defense of mountain
passes increased the military role of the Romanians in the Stara Planina
Mountains during the 12th century. When the administration tried to
extract revenue from those free men, loyalty made room for hostility, and
the same mountains turned into bases of operations for rebels. The Bul-
garian and the Vlach forces gathered by the Vlach brothers Peter and Asan
made great use of their advantages in the mountains when confronting
Byzantine armies sent against them, and were thus capable of obtaining
remarkable victories, such as that following the ambush of Trjavna in the
spring of 1190.199
3. The Danube Frontier in the 12th Century
A peaceful period followed the Cuman invasion of 1095. The following

raid in 1114, reached only the western part of Paradunavon and the theme
of Bulgaria. The Cumans crossed the Danube at Vidin, most probably
through the ford known to this day as Vadul Cumanilor (“the Ford of the
Cumans”). Emperor Alexios I led the expedition against them in per-
son, which crossed over onto the northern bank, within Little Walachia
(Oltenia), a region with a cluster of place names of Turkic (Pecheneg or
Cuman) origin.200 A much greater impact on Paradunavon had the Pech-
eneg attack of 1122, the consequences of which may be comparable to
those of 1033–1036. After their defeat in 1121 at the hands of Grand Prince
of Kiev, Vladimir II Monomakh (1113–1125), the Pechenegs from the steppe
lands north of the Black Sea moved to the Danube area. This population
movement was not unlike that of the Uzes of 1065. According to John
Kinnamos, the Pechenegs came to the Byzantine lands with their families.
Michael the Syrian and a anonymous source commemorating the victory
198 Ducellier 1968, 357–358, 364–368.

199 Niketas Choniates, Isaac Angelos, III. 3 (ed. Van Dieten, 429–430; transl. Magoulias,
236); Akropolites, 133; Tsankova-Petkova 1981, 182.
200 Anna Comnena, XIV, 9.1 (transl. Sewter, 467); Iorga 1937, 69; Conea, Donat 1958, 154;
Diaconu 1978, 59–61; Mănucu-Adameşteanu 2001 c, 114; Mărculeţ 2005, 101; Curta 2006, 312;
Spinei 2006, 386–388; Spinei 2009, 124.
later obtained against them insist that there also Cumans among them.
The large number of enemies seems to have prompted the emperor to
adopt a cautious attitude and to try to bribe some members out of the
coalition. He then focused on those nomads who had crossed the Danube
and have reached Thrace, where the Byzantine army won a major victory
at Beroe in April 1122.201
The 1122 invasion of the Pechenegs and the Cumans is responsible for
the destruction of the forts in Păcuiul lui Soare, Turcoaia, Garvăn, and
Nufăru. Isaccea suffered as well (see Fig. 13). Coin series on all those
sites show gaps between the later part of Alexius I’s reign and the early
regnal years of John II Comnenos.202 At Garvăn, the attack ended the
phase, which archaeologists call “of above-ground dwellings,” and the site
remained unoccupied until the late 12th century, when the new phase
begins, which is called “of the burned dwellings.”203 The seal of Michael
“archbishop of Russia” found in Garvăn in the filling of sunken-floored
building 168—which belongs to the phase “of the burned dwellings”—was
attributed to a metropolitan bishop of Kiev whose term began in 1130 and
ended in 1145.204 Petre Diaconu has however argued that he was in fact a
bishop of Iur’ev, who between 1072 and 1073 was then the deputy of the
metropolitan bishop.205 The position in which the seal was discovered
suggests that it actually came from the layer into which the building was
dug. Moreover, the absence from this site of reliquary crosses of Kievan
type, which appear Dobrudja in the second half of the 12th century is a
strong argument in favor of the idea that the site was abandoned during
that period.206 The fort in Nufăru was partially restored under the early
regnal years of Manuel I Comnenos. After the invasion of 1122, the occupied
area shrank behind a new rampart (Z 2, identified at a point known as
201 Kinnamos, I. 3 (transl. Brand, 16); Niketas Choniates, John II Comnenos, 4 (ed. Van
Dieten, 13–15; transl. Magoulias, 10–11); Michael the Syrian, XV.12 (ed. Chabot, 206); Cha-
landon 1912, 48–50; Barnea, Ştefănescu 1971, 156; Diaconu 1978, 62–71; Angold 1984, 153;
Birkenmeier 2002, 90; Spinei 2006, 216–217; Meško 2007, 3–26; Spinei 2009, 125–126; Ivanov,
Lubotsky 2010, 595–603.
202 Diaconu 1978, 62–71; Oberländer-Târnoveanu 1998, 81; Mănucu-Adameşteanu
2001 a, 159, 192–194; Mănucu-Adameşteanu 2001 c, 113–114; Curta 200, 312, 314; Mănucu-
Adameşteanu 2010 a, 194; Mănucu-Adameşteanu 2010 d, 446.
203 Diaconu 1975, 387–394; Diaconu 1978, 62–65, 71, 120–129; Mănucu-Adameşteanu
2001 a, 193.
204 Ştefan, Barnea 1967, 335–336; Barnea 1980, 282–283.
205 Diaconu 1992 a, 183–185.
206 Mănucu-Adameşteanu 1987, 289.
146
chapter three

“Dispensar”), while the area outside it became a cemetery.207 Very few

coins have been found which could be dated after 1122, in sharp contrast
to the situation under Alexios I.208 The site continued to be occupied, but
this decline in the number of coin finds suggests rapid economic decline,
as well as the absence of a garrison. This further suggests that in the
aftermath of the 1122 attack, the most important settlement in the mari-
time segment of the Danube was Isaccea. The fort in Nufăru was severely
damaged and ceased to be part of the defense system, while Garvăn was
entirely abandoned. Judging from the existing evidence, Isaccea remained
the only outpost in a land of deserted or impoverished settlements.209 The
large number of 12th-century coins found in Isaccea210 reflects the pros-
perity of that city, especially during the reign of Manuel Comnenos, while
the existence of an episcopal see may be surmised on the basis of the
equally large number of engolpia (pectoral crosses).211 Dristra remained
the main city in Paradunavon. Its description by al-Idrisi (“a city with
large streets, many bazaars and rich resources”) is one of a prosperous
urban center. The city was also the residence of a metropolitan bishop,
Leo Charsianites, who was in office between 1145 and 1170.212
The changing balance of power in the northwestern Balkans led to
frequent and intense confrontations between the Byzantine Empire and
Hungary in the Middle Danube region. After the disappearance of the
theme of Sirmium in 1071, the territory under Byzantine in that region,
which was annexed to the theme of Bulgaria, was much diminished. Its
westernmost point was the city of Belgrade, still in Byzantine hands in
1096. Its strategic importance was linked to its position at the confluence
of the Tisza and Danube rivers. The control over this city by any power
located in the south could effectively open up the way to Central Europe
along the two rivers. Much later, this is exactly what happened after
the Ottoman conquest of Belgrade (1521). For the Byzantines, however,
Belgrade was simply a launchpad for campaigns against Hungary led by
Emperors John II and Manuel.
In those campaigns, the Byzantine army attacked along the roads cross-
ing the center of the Balkan Peninsula and reaching Vidin, Braničevo, and
207 Mănucu-Adameşteanu 1998, 80; Damian, Andonie, Vasile 2001, 165; Damian,

Andonie, Vasile 2002, 217.
208 Mănucu-Adameşteanu 2001 a, 413 (table 15).
209 The same conclusion at Curta 2006, 319–320.
210 Oberländer-Târnoveanu 1992, 44–51, 58–59.
211 Mănucu-Adameşteanu 1987, 285–292.
212 Barnea, Ştefănescu 1971, 163; Shepard 1979, 233.
148 chapter three
Belgrade in the north. The navy was also used in those operations, which
were made possible by the existence of several anchorages, such as those
of Isaccea, Axiopolis (at least until 1148), and Dristra.
That the region between Braničevo and Belgrade is called “Paristrian”
in the main accounts of those campaigns, which may be found in John
Kinnamos and Niketas Choniates,213 shows that the name previously
employed metaphorically for the lower course of the Danube was now
extended to its middle course. Such a change may reflect a shift in stra-
tegic interests, but not the creation of a province named Paristrion. The
name Paristria appears in the letter of Emperor Isaac II Angelos (1185–1195)
to Pope Celestin III (1191–1198), which was drafted by the logothetes Dem-
etrios Tornikes in 1193,214 and clearly refers to the Middle Danube region
in the context of the Hungarian attack on Serbia in the winter of 1192/1193.
The emperor complains about Christians attacking Christians: “This pas-
sion which overcame almost all those who are called Christians started in
Germany and Sicily and extended even to the Ocean (. . .) and now this
passion, after crossing the Ocean and the Rhine, flows into Paristria.” He
then describes the situation in Serbia. While this is solid evidence that the
name “Paristrion” was now applied to the region between Belgrade and
Braničevo, there is no mention of a province by that name anywhere in
the northwestern Balkans.
The expansion of the Hungarian kingdom to the southwest must be
seen in the context of the occupation of Croatia in 1091. The Byzantine
reaction to those events was a campaign aiming at restricting the Hun-
garian push to Dalmatia, which was still in Byzantine hands.215 Know-
ing that a Pecheneg raid devastated Hungary in that same year, it is not
impossible that in the aftermath of Lebounion, the Byzantines may have
used the Pechenegs to create a diversion. With imperial authority over
Paradunavon restored, contact with the Pechenegs north of the Danube
was made possible through such centers as Isaccea. It is likewise possible
that the invaders came from Paradunavon. With the power of Hungary
on the rise, Alexios I tried to prevent a coalition between Hungary and
his arch-enemy, Bohemond of Taranto, who was at the time contemplat-
ing a matrimonial alliance with the Hungarian king Coloman. In 1105,
Alexios’ son John married Princess Piroska (Irene), the daughter of the
213 Kinnamos, III, 11 (transl. Brand, 94); Niketas Choniates, Manuel Comnenos, IV, 1 (ed.
Van Dieten, 127; transl. Magoulias, 72).
214 Demetrios Tornikes, 342/343; Mănucu-Adameşteanu 2001 a, 206.
215 Urbansky 1968, 28–29; Makk 1989, 10; Makk 1990, 18; Fine 1991, 284; Makk 1994 a, 65.
former king Ladislas I. The emperor also allowed Hungary to take Dalma-
tia, a province which was almost independent, and in which Hungary had
already occupied important cities, such as Zara and Spalato.216 He then
enjoined Venice to the alliance against Bohemond’s invasion of Epirus in
1107. In 1115, Venice attacked the Hungarian domain in Dalmatia, in retali-
ation for what was perceived as interference in the Venitian economic
hinterland.217
As for the peaceful Byzantine-Hungarian relations, they were abruptly
interrupted in 1127, when one of the claimants to the Hungarian crown,
Prince Almos, sought refuge and assistance in the empire, as Emperor
John II Comnenos (1118–1143) had married his (Almos’s) cousin Piroska,
now renamed Irene. In retaliation, the Hungarian king Stephen II (1116–
1131) took Braničevo and then advanced in the direction of Niš, Sofia, and
Philippopolis. John II Comnenos’s counter-attack in the spring of 1128
recovered all losses, including Braničevo, and also conquered the region
between the Sava and the Danube, which was abandoned in 1071 (Franko-
chorion). The navy equipped with “Greek Fire” assisted the land troops,
which marched through Sofia to Braničevo and Belgrade. The ships were
then used for the crossing onto the left bank of the Danube at Zemun,
the fort in front of Belgrade. Haram (Nova Palanka), at the mouth of river
Nera, in front of Braničevo, was the other bridgehead taken on this occa-
sion. That the Byzantine navy was capable of moving beyond the Iron
Gates indicates the existence of some anchorages along the Danube. The
involvement of the navy in such a remote segment of the Danube was
a novelty for the Byzantine military history, which must have involved
a serious logistic effort. All previous operations had concerned only the
segment between the Danube Delta and Dristra. This naval campaign was
in fact decisive for the victory, the most important achievement of which
was the reconquest of Braničevo. The king of Hungary resumed the war in
1129 with the assistance of his Czech vassals, and retook Braničevo. Dur-
ing this campaign, Hungary was also assisted by the great zhupan Uroš I
(circa 1125–circa 1145), the ruler of the Serbs in Raška, who had risen in
rebellion against the Byzantines. His daughter Helen later married in 1130
the future Hungarian king Béla II (1131–1141). The Serbian defection posed
a serious threat to the security of the road linking Belgrade to Niš and
216 Urbansky 1968, 32–34, 37–38; Makk 1989, 14; Fine 1991, 234, 285; Stephenson 2000,
180–181, 197–199; Rostkowski 2001, 162–163.
217 Chalandon 1912, 55–56; Makk 1989, 15; Stephenson 2000, 203; Rostkowski 2001, 163.
150 chapter three
Sofia. Another Hungarian attack in 1129 caused a reaction from John II,
who crushed the Serbian rebels and retook Braničevo, forcing Hungary
to recognize the city as Byzantine domain through the peace treaty con-
cluded in October 1129.218 The war of 1127–1129 had a defensive character
for the Byzantine Empire, because the target was to preserve the control
over the strategic position of Braničevo. The initiative belonged to Hun-
gary, and John II had no intention to occupy the region of Sirmium. He
preferred to concentrate forces in the East, a much more important the-
ater of war.219
Manuel I Comnenos’ strategic goal during his long reign (1143–1180) was
to restore the western frontiers to their line established by Basil II. The
confrontation with Hungary was only a part of this ambitious plan, the
inadequacy of which was shown by the final failure in the wars against
the German Empire and the Seljuk Turks.220 Unlike John II, Manuel waged
an offensive war against Hungary, for his intention seems to have been to
transform Hungary into a buffer state between the Byzantine and Ger-
man empires, at a time when, after 1156, the deterioration of his relations
with Frederick I Barbarossa (1152–1190) had led to what Paul Magdalino
has called “cold war.221 Manuel led 13 campaigns between 1149 and 1167,
the result of which was to put an end to the Hungarian expansion in the
direction of the Adriatic Sea and of Serbia.
A new Cuman raid affected Dobrudja in 1148, before hostilities started
with Hungary. At that time, the Byzantine army was involved in the war
against Roger II of Sicily (1130–1154), who in 1147, taking advantage of
the Second Crusade, had Corfu and a number of forts in Peloponnesus.
Manuel appealed to Venice. He also intended to send an expedition to
recover Corfu and to reestablish the Byzantine rule over southern Italy.222
The campaign was put on hold, however, when news arrived about
Cumans having ravaged the entire region between the Danube and the
218 Kinnamos, I, 4 (transl. Brand, 17–19); Niketas Choniates, John II Comnenos, 5 (ed. Van
Dieten, 17–18; transl. Magoulias, 11–12); SRH, I, 439–443; Chalandon 1912, 58–61; Urbansky
1968, 45–46; Angold 1984, 154, 174; Kosztolnyik 1987, 88–90; Makk 1989, 24–27; Fine 1991,
234–235; Stephenson 2000, 207–209; Rostkowski 2001, 166–169; Birkenmeier 2002, 90–91;
Curta 2006, 328–329; Kostova 2008 b, 270.
219 Makk 1999, 109–110; Stephenson 2000, 206–210.
220 Ostrogorsky 1956, 402–414; Angold 1984, 161–184; Treadgold 1997, 638–650; Stephen-
son 2000, 211–274.
221 Magdalino 1993, 62; Stephenson 2000, 239–274. See also Ferluga 1976, 193–213; Fine
1991, 236–242.
222 Chalandon 1912, 318–323; Urbansky 1968, 62–63; Makk 1989, 43–45; Magdalino 1993,
46–53; Stephenson 2000, 223.
Stara Planina Mountains. The emperor decided to go in person against

those enemies. At the head of the land army, he reached the Danube at
an unknown location. While waiting for the arrival of the navy sent from
Constantinople, he decided to raid the territories on the northern side of
the river, which he crossed on fishing boats collected from the neighbor-
hood. A man who did not know who he was, complained to him about
the emperor’s lack of concern for the population in the Danube region.
Manuel crossed the Danube with 500 soldiers and made a surprise attack
on a Cuman campsite.223 To judge from Kinnamos’ account, this action
was meant to serve as propaganda means for distracting attention from
the lack of any defense of the Lower Danube, in sharp contrast to the
situation on the border with Hungary. The Cumans were regarded as less
dangerous than the Arpadian kingdom, and because of that there is very
little information about what was going on in Dobrudja during the 12th
century. This one-sided strategy in fact encourage the rise of the Cuman
power in the second half of the 12th century, when the Cumans became
a regional power, a “steppe empire”, which will dominate the area until
the Mongol invasions. The Cuman invasion of 1148 was followed by other
attacks perfectly timed to coincide with the Byzantine involvement in
the war with Hungary, which raises the possibility of a Hungarian-Cuman
cooperation. For instance, in 1154, the Cumans plundered the Byzantine
cities along the Danube. A commander called Kalamanos (who, in fact,
was Boris, the illegitimate son of the Hungarian king Coloman) was sent
with some troops against the Cumans, but he was easily defeated because,
in Niketas Choniates’s words, “he conducted the campaign ineptly.” Choni-
ates’ account suggests that there were no permanent forces garrisoned in
the forts along the Danube. A third Cuman raid is mentioned by Kinna-
mos for 1161.224
The study of coin hoards buried during the 1148 attack reveals that
northern Dobrudja was the most affected area.225 The identification of
the place where Manuel crossed the Danube has long been a matter of
dispute. The fisherman who complained to the emperor, without knowing
223 Kinnamos, III, 3 (transl. Brand, 76–77); Niketas Choniates, Manuel Comnenos, II, 2
(ed. Van Dieten, 78; transl. Magoulias, 46); Urbansky 1968, 63; Barnea, Ştefănescu 1971,
157–160; Diaconu 1978, 78–80; Spinei 2006, 388.
224 Kinnamos, IV, 24 (transl. Brand, 153); Niketas Choniates, Manuel Comnenos, II. 7
(ed. Van Dieten, 93; transl. Magoulias, 54); Chalandon 1912, 323–325; Urbansky 1968, 81;
Diaconu 1978, 89–90; Makk 1989, 148; Mănucu-Adameşteanu 2001 a, 197–198; Spinei 2006,
389; Spinei 2009, 129–130.
225 Oberländer-Târnoveanu 1992, 44, 56; Mănucu-Adameşteanu 2001 a, 196–199.
152 chapter three
who he was, also revealed to him that a fort named Demnitzikos had been
destroyed by the Cumans and left deserted. Kinnamos also mentioned
a name, Tenouormon, which refers to a place in the region north of the
Danube which Manuel’s men raided. The fort of Demnitzikos was first
identified with Dinogetia (Garvăn), where archaeological excavations had
revealed a layer of destruction dated to the mid-12th century,226 but that
destruction seems to be linked to the Pecheneg attack of 1122 (see above).227
Others have proposed Zimnicea, mainly on the basis of name similarity,
but no 12th-century is known from that city. Still others advanced the idea
that Demnitzikos was Svištov, and that the name had later been moved
north of the Danube.228 This identification could not make use of a further
identification of Tenouormon with Teleorman (now a county just north of
the Danube), as Petre Ş. Năsturel has demonstrated that the name Ten-
ouormon (which means “wild forest”) was most likely the Cuman name for
the entire forested area in Walachia and southern Moldavia. This area is
in effect called “the large forest” in medieval Romanian charters. The pro-
gressive shrinking of that area during the late Middle Ages and the early
modern period made it possible to restrict its use the present county of
Teleorman.229 In a later study, Petre Ş. Năsturel re-analyzed the passage in
Kinnamos and noted that the expression “a city worth of its name” (πόλιν
λόγου ἀξίαν) is a slightly veiled allusion to Axiopolis. His conclusion was
that Manuel crossed the Danube at Cernavoda, in which case the small
fort (phrourion) Demnitzikos could only be the fort located at the end of
the Stone Dike, now called “Cetatea Pătulului.” The two streams which
Manuel’s soldiers are said to have crossed would then be the main branch
and the Borcea branch of the Danube.230 This seems to be the best inter-
pretation of the passage. Axiopolis remained a Byzantine fort until the
attack of 1148, but no archaeological confirmation of this date is so far
available. However, it is quite possible that Axiopolis was one of the few
forts left, to the existence of which the Cuman attack of 1148 put an end.
Much more important than the location of Demnitzikos is the con-
clusion one can draw about the general situation of the Lower Danube
226 Bolşacov-Ghimpu 1967, 543–549 (accepted by Barnea, Ştefănescu 1971, 158–160 and

Barnea 1971, 355).
227 Diaconu 1978, 62–71.
228 Diaconu 1978, 86–88. The identification with Zimnicea was also proposed by Iorga
1937, 70, 205; Bănescu 1927, 20; Bănescu 1946, 104; Kostova 2008 b, 275.
229 Năsturel 1981, 81–91.
230 Năsturel 2004, 521–529. Spinei 2009, 129 quotes this study, still endorsing the previ-
ous opinion of Năsturel.
region in the mid-12th century, namely that it now took a backseat to the
main action taking place in the northwestern Balkans. This disregard for
the Lower Danube frontier allowed the Cuman power to rise north of the
Danube, especially after the Cumans established an alliance in 1159 with
Ivan Rostislavich and got involved in the fratricide strife in Halych. Dan-
ube Cumania emerged as a new power between the Byzantine Empire,
Hungary and the Rus’ principalities. The “steppe empire” created by the
Cumans (called Qipchak) extended from Bărăgan (the steppe in eastern
Wallachia) to the Volga, and was made up of several power centers.231
Cumania blocked the Rus’ expansion toward the Danube, but did not
prevent commercial relations between Halych and the Byzantine cities
on the Danube. Halychian merchants traveled south along the Dniester
and the Prut rivers, and Rus’ imports (primarily pectoral crosses and vari-
ous garments) began to appear in Moldavia and in settlements from Para-
dunavon, such as Isaccea, Garvăn, Dristra, and Păcuiul lui Soare.232 To
the south, Halych reached as far as north-eastern Moldavia, on the upper
course of rivers Prut and Dniester. The theory of a complete incorporation
of Moldavia within Halych, which was put forward by the Russian and
Soviet historians is based on an erroneous interpretation of the presence
of the Rus’ traders, and on the story of the so-called Berladniks. Prince
Ivan Rostislavich gathered the Cumans in 1159 in order to prey on Rus’
traders and their boats on the Danube. As Victor Spinei has demonstrated,
the Berladniks who were led by that prince in the fight against Iaroslav
Osmomysl have no relation with the Moldavian town Bârlad, a settlement
which appeared much later. The city of Berlad mentioned in the Rus’
chronicles was in Volyhnia.233 Cumania, not Halych, was the dominant
power in the region north of the Lower Danube in the mid-12th century.
The princes of Suzdal’ Vasilko (son of Yuri Dolgoruki) and Vladimir, who
are said to have settled down in 1165 somewhere on the Danube, were
most likely in control of a number of forts within the Byzantine territory.
That appanage could hardly be regarded as an extension of the Rus’ prin-
cipality to the Danube.234
231 Diaconu 1978, 95; Shepard 1979, 207; Pritsak 1982, 367; Perkhavko 1996, 18; Vásáry
2005, 7, 32; Spinei 2006, 393; Spinei 2009, 131.
232 Popescu 1992, 145–150.
233 Cihodaru 1979, 179; Boldur 1988, 228–231; Purici 1997, 189–202; Curta 2006, 315–316;
Spinei 2009, 135–137.
234 Kinnamos, V, 12 (transl. Brand, 178); Bănescu 1927, 21; Brătianu 1935, 29; Frances
1959, 57; Diaconu 1977 a, 1897; Diaconu 1978, 98; Tăpkova-Zaimova 1980, 334.
154 chapter three
The conflict with Hungary during Manuel I Comnenos’s reign was ini-
tiated by Hungary. In 1149, Géza II gave his support to another Serbian
rebellion in Raška against the Byzantines. The Serbs entered a dynastic
alliance with Hungary, as Jelena, the daughter of the great zhupan Uroš I
(circa 1125–circa 1145) married Béla II (1131–1141). On the other hand, the
great zhupan Uroš II (circa 1145–circa 1160) concluded an anti-Byzantine
agreement with Roger II of Sicily. The war in Serbia delayed the Byzantine
response to the Hungarian aggression, because Manuel had to use a part
of his army to quell the rebellion of the Serbs. By 1150, however, the Byz-
antine army was ready to march against Hungary, and managed to defeat
the Hungarians before they joined the Serbian rebels. The Hungarians had
attacked, as usual, along the valley of the river Morava in the direction of
Niš. After the decisive battle of Tara (north of Niš) in September 1150, the
great zhupan was forced to abandon his alliance with Géza II and Roger II.
The pacification of the Serbian lands was a sine qua non of the campaign
against Hungary. In 1166, Manuel put all of Raška under the rule of his
faithful ally, the great zhupan Tihomir.235
It was easier for the Byzantines to deal with Hungary and Raška than
with the Normans in Sicily.236 Manuel therefore adopted an offensive
strategy in the Middle Danube region. In October or November 1150, at
the time the largest part of the army of the Hungarian king Géza II was
on campaign in Halych, the Byzantine navy was again sent from Con-
stantinople to the Danube, at the same time as the land troops marched
to Belgrade and crossed the river Sava. An army corps under the com-
mand of the emperor occupied the region between the Danube and the
Sava, including the city of Sirmium. Another corps led by general Theo-
dore Batatzes besieged the fort at Zemun, and a third expedition reached
the Timiş valley in present-day Banat. The Byzantine troops were quickly
withdrawn from the Danube-Sava region as soon as the Hungarian army
returning from Halych counter-attacked, and the results of the 1150
campaign were indecisive. On the other hand, Géza II’s attacks on the
Byzantine territories (April 1151) had no concrete results.237
235 Kinnamos, III, 6–9 (transl. Brand, 83–90); Niketas Choniates, Manuel Comnenos, II.
7 (ed. Van Dieten, 93; transl. Magoulias, 54); Chalandon 1912, 384–391, 401; Ostrogorsky
1956, 412; Urbansky 1968, 48–49, 52, 71–75; Moravcsik 1970, 80; Kosztolnyik 1987, 146–147;
Makk 1989, 48–51; Fine 1991, 236–244; Magdalino 1993, 54–55; Stephenson 2000, 225–226,
267; Curta 2006, 329–330.
236 Magdalino 1993, 55.
237 Kinnamos, III, 10, 11 (transl. Brand, 90–93); Niketas Choniates, Manuel Comnenos,
II. 7 (ed. Van Dieten, 92–93; transl. Magoulias, 54); Chalandon 1912, 402–408; Urbansky
1968, 72–77; Moravcsik 1970, 81; Kosztolnyik 1987, 147–148; Makk 1989, 52–56; Magdalino
After 1150, the emperor appointed his cousin Andronikos as duke of Niš
in 1153. He had under command the entire territory from Niš to Belgrade.
That does not seem to have been a wise decision, as Andronikos had his
mind on the imperial throne, and opened negotiations with Géza II, to
whom he promised the duchy in exchange for the the Hungarian king’s
support. Soon after that, Andronikos was arrested (1154), but Géza, who
does not seem to have been aware of his ally’s fate, made a bold attack on
Braničevo and Belgrade together with the Cumans, who were then raid-
ing in the Danube region. In 1155, Manuel decided to retaliate, fearing an
alliance between Géza II and intended to make an alliance with Emperor
Frederick Barbarossa. The navy was again sent to the Danube, but the
Hungarian king quickly asked for peace and returned all prisoners taken
in the previous year.238
The events of 1155 led Manuel to a re-shuffle of the defense policy on
the Danube. After 1155, the system of fortifications on the river bank and
in the interior was reinforced. Manuel was fully aware of the strategic value
that such forts as Belgrade and Braničevo had in the war against Hungary.
The reconstruction work archaeologically documented on both sites may
be dated to the mid-12th century. At Braničevo, a new fortification was
built in Veliki Grad after 1156 next to the ramparts of the 6th-century walls
of Viminacium, themselves reused after 1018 (fig. 14). The area of the city
was thus doubled, with the purpose to accommodate the troops brought
from the south for the operations in the region of Braničevo. In Belgrade,
a new fort of 60 × 136 m was built inside the area of the 6th-century city
of Singidunum. This fort, which was separated from the rest of the settle-
ment by a ditch became the residence of the commander (fig. 15).239
Manuel also entered an alliance with Halych against Hungary. This
alliance was also directed against the Cumans, who were at that time in
conflict with the Rus’240 The new alliance allowed for the concentration
of the war efforts in the western part of the Danube frontier.
1993, 54–55; Stephenson 2000, 230–231 (according to his own chronology); Kostova 2008
b, 271–272.
238 Kinnamos, III, 17, 19 (transl. Brand, 100, 103–105); Niketas Choniates, Manuel Comne-
nos, 3 (ed. Van Dieten, 101–103; transl. Magoulias, 58–59); Chalandon 1912, 409–414; Bănescu
1946, 43, 161; Urbansky 1968, 80–84; Şesan 1978, 50; Kosztolnyik 1987, 148–150; Makk 1989,
60–62; Magdalino 1993, 56–58; Stephenson 2000, 233–234; Curta 2006, 330–331; Kostova
2008 b, 272; Stojkovski 2009, 389–390.
239 Popović, Ivanišević 1988, 178; Maksimović, Popović 1990, 217–219; Popović 1991, 169–
185; Milošević 1991, 187–195; Maksimović, Popović 1993, 129–133; Stephenson 2000, 241–245;
Kalić 2003, 91–96; Curta 2006, 331–332.
240 Wasilewski 1964, 481–482; Moravcsik 1970, 78–92; Diaconu 1978, 94–95; Angold 1984,
176.
156 chapter three
14. The fort of Braničevo (after Popović 1991, 171)
15. The fort of Belgrade (after Popović 1991, 172)

In addition, Manuel gave his support to claimants to the Hungarian

crown, all of whom were expected to follow a pro-Byzantine policy. First
Stephen, Géza II’s brother, then Béla, King Stephen III’s elder brother ful-
filled, if only briefly, Manuel’s ambition to turn Hungary into a client state.
A marriage was planned between Béla and Manuel’s daughter, Maria. Ste-
phen III quickly offered Sirmium and Dalmatia, the terms of the treaty
he concluded with Manuel in 1164. The recovery of Sirmium led to the
organization of a new duchy centered upon that city, but smaller than
the earlier one. Even so, Manuel continued to support a third claimant
to the Hungarian throne, a man named Stephen who had occupied the
fortress Zemun in 1164. Betrayed in his expectations, Stephen III attacked
the Byzantine Empire in 1166, and following a victory over duke Michael
Gabras, Sirmium was again occupied, if only for a while, by the Hungar-
ians. In Dalmatia, the Hungarians meanwhile won another victory near
Split against the duke of that province, Nikephoros Chaluphes. This time,
the emperor’s reaction was quick and effective. An army under the com-
mand of Andronikos Kontostephanos crushed the Hungarian at Zemun
(July 8th, 1167).241 Meanwhile, the general Leo Batatzes attacked Hungary
through one of the Carpathian passes (most probably Oituz).242 According
to John Kinnamos, this expedition went “by way of the regions near the
so-called Euxine, whence no one had ever assailed them.” Once in Tran-
sylvania, the Byzantine army ”after they had passed through some weari-
some and rugged regions and had gone through a land entirely bereft of
men, they burst into Hungary; encountering many extremely populous
villages, they collected a great quantity of booty and slew many men, but
took captive many more.” Finally, a third army under the command of
John Dukas made a similar maneuver, entering Hungary from the, some-
where near the border with Halych.243
241 Kinnamos, V, 5–6, 8, 10, 14–16 (transl. Brand, 160–165, 167–171, 174, 179–186); Nike-
tas Choniates, Manuel Comnenos, 4 (ed. Van Dieten, 127–128, 132–135; transl. Magoulias,
72–73, 75–77); Chalandon 1912, 470–487; Wasilewski 1964, 481; Urbansky 1968, 85, 93–103;
Moravcsik 1970, 82–88; Kosztolnyik 1987, 185, 187; Makk 1989, 85–92, 99–100; Magdalino
1993, 79–83; Stephenson 2000, 247–260; Curta 2006, 332–333; Kostova 2008 b, 272–273.
242 Cihodaru 1979, 171 suggested that the army marched through the Bugeac (southern
Bessarabia), which would mean that it crossed the Danube at Isaccea.
243 Kinnamos, VI. 3 (transl. Brand, 195–196); Niketas Choniates, Manuel Comnenos, V, 1
(ed. Van Dieten, 151–157; transl. Magoulias, 86–89); Frances 1959, 58–59; Urbansky 1968,
104–106; Moravcsik 1970, 84–85; Stănescu 1974, 407; Diaconu 1978, 102–103; Kosztolnyik
1987, 187, 189, 191; Makk 1989, 100–101; Stephenson 2000, 260–261; Haldon 2001, 138–139;
Birkenmeier 2002, 119–120; Curta 2006, 333; Kostova 2008 b, 273–274. Some have seen this
158 chapter three
Leo Batatzes’ army also included ”a large group of Vlachs, who are said
to be formerly colonists from the people of Italy.” The word ὃμιλος used by
Kinnamos means “mob” and concerns a unit of irregulars, recruited most
probably from Dobrudja, the region from which the attack was launched.244
There is, however, no indication of a military organization of the local
Romanian population in Dobrudja. On the contrary, the impression one
gets is that the Byzantines had to improvise, as they recruited unskilled
men (Romanians in this case) to complement the apparently small
number of professional soldiers taking part in the expedition. There is no
proof that those Vlachs were from the Stara Planina Mountains or that
they were organized in some kind of unit specializing in fighting in the
mountains.245
The victory of 1167 was the final moment of the Byzantine-Hungarian
wars under Manuel I. The Byzantine control over the disputed territories
in Dalmatia, Croatia and in the Middle Danubian region was sanctioned
by the peace treaty of July 1167. The same victory curbed any Cuman
attacks for a while, as it demonstrated the capability of the Byzantine
army to wage war on a large scale in the regions under its control.246 The
hegemony of the Byzantine Empire was restored in the western Balkans
and on the Danube, but not for too long.
After Manuel’s death in 1180, Prince Béla, crowned king of Hungary by
Manuel after Stephen III’s death (with the name Bela III, 1172–1196) took
the opportunity to recover Dalmatia and Sirmium. Involved in conflicts at
the Byzantine court, he attacked the empire under the pretext of defend-
ing the interests of Alexios II Comnenos against the usurper Andronikos.
He therefore annexed a large region extending from Sirmium to Sofia
after two campaigns carried out in 1182–1183 in cooperation with the great
zhupan of Serbia, Stephen Nemanja. Andronikos Comnenos’ usurpation
was also used as a pretext by the king of Sicily, William II (1166–1189) for
launching an attack on the Byzantine Empire. That maneuver was meant
to give support to a claimant to the Byzantine throne calling himself
Alexios II (the real Alexios II had in fact been killed by Andronikos). On
June 24th 1185, the Norman army conquered Dyrrachion, at the western
expedition as aimed to counteract the Hungarian expansion into the regions south and
south-east of the Carpathian Mountains (Cândea 1999, 153–155).
244 Stănescu 1971, 589; Stănescu 1974, 407–408; Năsturel 1969, 178–184; Curta 2006, 316;
Kostova 2008 b, 274.
245 Năsturel 1969, 179.
246 Năsturel 1969, 183–184.
terminal of Via Egnatia, connecting Thessaloniki to Constantinople. The

Normans then sacked Thessaloniki on August 24th, 1185, but were stopped
at Dimitritsi in southern Thrace on November 7th, 1185.247
The new emperor Isaac II Angelos made peace with Béla. The Hun-
garian king offered him his daughter Margaret (Mary) in marriage, with
the region Belgrade-Niš (which had been conquered from the Byzantine
Empire in 1182–1183) as dowry. In exchange for that, Hungary kept Dal-
matia and Sirmium.248 To cover the wedding expenses, Isaac II imposed a
surtax, which was the immediate cause of the rebellion of the Vlachs and
Bulgarians in the Stara Planina Mountains. As Niketas Choniates wrote,
when the emperor Isaac II was encamped at Kypsella in Thrace [today
Ypsala], two leaders of the Romanian (Vlach) population from the Hae-
mus Mountains requested to be recruited in the army and be awarded
by imperial rescript a certain estate (χωρίον) situated in the vicinity of
Mount Haemus, which would provide them with a little revenue.249 The
two leaders’ names were Peter and Asan. Peter’s real name appears to
have been Theodore, but he adopted Peter in reference either to the rebel
Peter Deljan or to the 10th-century emperor of Bulgaria. That the rebellion
began in October 1185 has been demonstrated by Günter Prinzing.250
This book is not concerned with the state created by the Assenids.
Instead, the following considerations focus only on events that caused
the end of the Byzantine military organization along the Danube. In fact,
one of the reasons the rebels were so successful is that some of them had
military experience. The appearance under Alexios I Comnenos of Vlach
military units recruited from the mountain regions may have spurred
the rise of a military elite. The forts in the Stara Planina mentioned by
Niketas Choniates were most likely centers of that elite’s regional power.251
The military skills of those who rebelled in 1185 were the unforeseen con-
sequence of the Comnenian emperor’s policy of local recruitment. The
reforms introduced by Alexios I Comnenos increased the military capabil-
ity of the empire through support granted to the pronoiars. Pronoia was
a lifetime possession bestowed to aristocrats who were thus obliged to
247 Ostrogorsky 1956, 422–426; Urbansky 1968, 109–111, 122; Moravcsik 1970, 90–92; Kosz-
tolnyik 1987, 207–208; Makk 1989, 115–117; Fine 1994, 6–7; Treadgold 1997, 652–654; Ste-
phenson 2000, 281–283.
248 Guilland 1964, 125; Moravcsik 1970, 93; Kosztolnyik 1987, 209; Makk 1989, 120; Fine
1994, 9–10; Treadgold 1997, 657; Stephenson 2000, 283–284; Vásáry 2005, 14.
249 Niketas Choniates, Isaac Angelos, I. 5 (ed. Van Dieten, 369; transl. Magoulias, 204).
250 Prinzing 1999–2000, 257–265.
251 Iosipescu 1994, 252.
160 chapter three
ensure recruits from their estates. In the long term, the pronoia is respon-
sible for the centrifugal trends in Byzantine politics, especially after it
became hereditary (1258), but in the late 11th and early 12th century, the
reorganization of the professional army on that basis was a success story.
The small farming property of the stratiotai survived, but the develop-
ment of a feudal army not unlike those of Western Europe is obvious. The
recruitment on the basis of local (and therefore often ethnic) criteria was
encouraged by the rise of the pronoiars, and that contributed to a rapid
decline of units of foreign mercenaries. Several groups of Pecheneg and
Cumans were settled in the European provinces of the Empire, former
prisoners of war thus becoming stratiotai. The military profession was
again attractive to peasants, even if they happened to be were tenants of
the pronoiars. According to Niketas Choniates, “everyone wanted to enlist
in the army (. . .) while some ran away from their charge of grooming
horses and others, washing away the mud from brick making and wiping
off the soot from working the forge, presented themselves to the recruit-
ing officers. After handing over a Persian horse or paying down a few gold
coins, they were enrolled in the military registers without due examina-
tion and immediately were provided with imperial letters awarding them
parcels of dewy land, wheat bearing fields, and Roman tributaries to serve
them as slaves.”252
The request of the two Vlach brothers Asan and Theodore (Peter) was
rejected. Furthermore they were apparently humiliated in front of the
emperor, perhaps with his approval. This fact sparked the rebellion which
soon found supporters among the cattle and sheep owners affected by the
tax increases. From the very beginning, therefore, this was a rebellion of
mountaineers with considerable military skills. In his account of Isaac II
Angelos’ campaign against the rebels (October 1187), Niketas Choniates
criticizes the emperor for not dispatching soldiers to the forts, most likely
those in the Stara Planina Mountains and on the Danube.253 Instead,
according to Choniates, the rebels put those forts to good use and to their
advantage.254 The development of the region in which those mountain
forts were located is betrayed by the surge in coin finds of the second
half of the 12th century, which have been discovered in the area between
252 Niketas Choniates, Manuel Comnenos, VII (ed. Van Dieten, 209; transl. Magoulias,
118–119); Ostrogorsky 1956, 392–393; Ostrogorsky 1970, 41–54.
253 Niketas Choniates, Isaac Angelos, II. 1 (ed. Van Dieten, 394; transl. Magoulias, 217).
For the campaign, see Tsankova-Petkova 1978, 116–118.
254 Niketas Choniates, Isaac Angelos, I. 4 (ed. Van Dieten, 369; transl. Magoulias, 204).
Tărnovo and Šumen.255 The rebellion started in Tărnovo, a town which,

according to Niketas Choniates, was “the best fortified and most excellent
of all the cities along the Haemus, encompassed by mighty walls, divided
by a river stream, and built on a ridge of the mountain.” Tărnovo became
a real capital, a fortified town in which buildings symbolizing political
and religious power—the imperial palace, churches, later the patriarchal
church and palace—were later erected on the hill now called Tsarevets
(the palace of the tsars, churches, and later on the patriarchal church and
palace).256 From this initial center, the movement spread to the entire
territory of the formerly Bulgarian state, at the same time gaining a new
political goal—complete independence from Byzantium, as symbolized
by the use of the title of emperor (tsar). The events of 1185 eventually led
to the rise of a state which for all practical purposes was regarded as the
heir of early medieval Bulgaria.
The population involved in this movement was both Romanian (Vlach)
and Bulgarian. Nationalistic concerns have long been on the mind of Bul-
garian historians rejecting the participation of the Vlachs and the Roma-
nian (Vlach) origin of the movement’s leaders, although both Byzantine
and western sources are quite clear in that respect.257 Some have gone as
far as to claim that the Assenids descended from a community of Turkic
Bulgars who had resisted Slavicization until the 12th century.258 Others
have recently claimed a Cuman origin for the Assenids on the basis of
nothing else but the supposedly Turkic origin of the name Asan.259 Lit-
tle attention has been paid to the fact that Cuman names were given to
people who were not viewed as, nor claimed to be Cumans, such as many
Romanians in medieval Walachia and Moldavia.260 To be sure, a mixed,
Vlach-Cuman origin of Peter and Asan is not impossible. However, the
name Asan (if it is Turkic at all) was given to the son of a high-status
man, who would use a foreign name for his child to distinguish him (and
himself) from the other Vlachs. Ever since the invasion of 1122, which had
255 Oberländer-Târnoveanu 2003, 360.

256 Niketas Choniates, Alexios Angelos, I. 5 (ed. Van Dieten, 470; transl. Magoulias, 258);
Akropolites, 137; Gjuzelev 1991, 95–98; Dochev 2002, 673–674; Žekova 2004, 344–349; Curta
2006, 320–321, 360.
257 For the history of the historiographic dispute between Romanian and Bulgarian his-
torians surrounding the origin of the Assenids and the (ethnic) character of the state, see
Tanaşoca 1989, 153–181; Lazăr 2005, 3–31.
258 Dimitrov 1993, 103.
259 Vásáry 2005, 38–41, who also quoted the previous studies.
260 See Spinei 2009, 311–330.
162 chapter three
reached Thrace, Cumans lived in close contact with the Vlachs, for exam-
ple, in the region of Moglena.261 It is important to note that during that
invasion, warriors had brought their families with them, and it is reason-
able to believe that at least some Cuman women were taken captive after
the victory of Beroe.
The role of the Vlachs in the rebellion and the rise of the Assenid state
is clearly spelled out by various sources, such as Niketas Choniates, Geof-
froy of Villehardouin and Robert de Clari. The Byzantine historian even
mentioned the language of the Vlachs which was apparently spoken by
a priest whom the rebels had captured,262 while the French chroniclers
wrote about Blaquie and Blacs. Of crucial significance in this respect is
that Pope Innocentius III (1198–1216) knew very well that the people of
Kaloyannes, the third, younger brother who took over the power in 1197,
were of Roman origin (audito quod de nobili Urbis Romae prosapia pro-
genitores tui originem traxerint).263 A particularly interesting testimony
about the ethnic character of the state established by the Peter and Asan
brothers is that of the Heimskringla, written by Snorri Sturlusson in Ice-
land around 1230. The region in Thrace in which the Byzantine-Pecheneg
war of 1091 took place is called Blökumannaland (“the land of the Vlachs”)
which is evidently an anachronism, but reflects the situation at the time
Snorri wrote Heimskringla.264 Those testimonies thus represent a body
of uncontrovertible evidence about the participation of the Vlachs in
the rise of what is now commonly referred to as the “Second Bulgarian
Kingdom.”265
The rebels led by Peter and Asan occupied the mountain passes in
the eastern range of the Stara Planina Mountains, but were pushed back
by the Byzantine army (April 1186). At this point, the intervention of
the Cumans was decisive in the ultimate victory of the rebels. From this
261 Anastasijević, Ostrogorsky 1951, 22–29.

262 Niketas Choniates, Alexios Angelos, I. 5 (ed. Dieten, 468; transl. Magoulias, 257).
263 Wolff 1949, 190–192, 201–202; Brătianu 1980, 75–78; Tanaşoca 1981, 582; Vásáry 2005,
26–27.
264 Pintescu 1999, 41–61; Spinei 2006, 143–1144. Others have dated the events to which
Heimskringla refers to 1122, but that does not change the issue of the (ethnic) anachronism
(Diaconu 1978, 72–77; Bløndal 1981, 148–151; Meško 2007, 9–13). A third opinion linking the
events to the 1114 expedition north of the river Danube (in which the Vlachs in question
were north, not south of the Danube) cannot be accepted, because no major battle is
known to have been fought there to match the account in the Heimskringla. See Gyóni
1956, 303–311; Horedt 1969, 180–181; Mărculeţ 2005, 101–103; Mărculeţ 2010 b, 585–594.
265 And has been recognized as such by several historians: Wolff 1949, 180–181; Angold
1984, 273; Cheynet 1990, 450; Stephenson 2000, 288–294; Fine 1994, pp. 12–13.
moment onwards, the Cumans would get involved in several campaigns

until 1206.266 Emperor Isaac II’s expedition of October 1187 was a disaster,
despite the victory his troops obtained with much difficulty at Lardeas on
October 11th, 1187. The following campaign interrupted by winter, led to
another failure, as after a long siege in the spring of 1188, Lovitzos (Loveč)
could not be taken. Through the peace concluded in the summer of 1188,
the rebels obtained the entire region north of a line linking Plovdiv, Stara
Zagora, and Akhtopol. According to Niketas Choniates, the rebels ”were
not content merely to preserve their own possessions and to assume con-
trol of the government of Mysia; they also were compelled to wreak havoc
against the Roman territories and unite the political power of Mysia and
Bulgaria into one empire as of old.”267
After 1188, what had started as a domestic rebellion turned into a conflict
between two states. Peter had claimed the imperial title since the begin-
ning of the rebellion, and he even wanted to be recognized as emperor
of the Greeks by Emperor Frederick Barbarossa when the imperial army
crossed the Balkans in the direction of Constantinople at the beginning
of the Third Crusade. The Serbs of Stephen Nemanja and the Vlachs and
Bulgarians led by Peter offered their alliance against Isaac II, but the Ger-
man emperor chose peace with the Byzantine Empire. To judge by con-
temporary accounts, the forces commanded by Peter consisted of 40,000
Vlach and Cuman archers.268
Isaac II resumed hostilities in 1190, but the Vlach-Bulgarian army
inflicted a serious defeat on the Byzantine army in the pass of Trjavna.
Over the next five years, the military confrontations between the Byz-
antines, one one hand, and the Vlachs and the Bulgarians on the other
hand, took place mostly south of the Stara Planina Mountains, in Thrace.
In 1195, the Assenids conquered a region in the valley of the river Strymon
(Struma).269 To the north, their state seems to have extended all the way
to the Danube since the beginning of the rebellion. In a letter to his son
Henry, Emperor Frederick Barbarossa wrote that “Kalopeter the Vlach and
266 Diaconu 1978, 114–119, 130–133; Vásáry 2005, 17, 21, 42; Lazăr 2006, 17.
267 Niketas Choniates, Isaac Angelos, I. 5; II. 1 (ed. Van Dieten, 372–374, 394–399; transl.
Magoulias, 206, 217–219); Tsankova-Petkova 1978, 106–119; Tsankova-Petkova 1980, 61–63;
Stephenson 2000, 290–292; Vásáry 2005, 43–44; Lazăr 2010, 12–13.
268 Ansbertus, 58 (transl. Loud, 84); Wolff 1949, 184; Fine 1994, 24; Iosipescu 1994, 265–
268; Stephenson 2000, 294–300; Vásáry 2005, 44.
269 Niketas Choniates, Isaac Angelos, III. 3 (ed. Van Dieten, 429–430; transl. Magoulias,
236–237); Akropolites, 133; Wolff 1949, 184–188; Kosztolnyik 1987, 217; Stephenson 2000,
300–306; Vásáry 2005, 45–47; Lazăr 2006, 18–19.
164 chapter three
his brother Assan with the Vlachs subject to them were exercising tyran-
nical rule over much of Bulgaria, and especially in the region where the
Danube flows into the sea”.270 This means that in 1189 Dobrudja or a good
part of it was already under the control of the rebels. Gheorghe Mănucu-
Adameşteanu is therefore wrong when arguing that before 1204 the Ass-
enid state did not include Dobrudja.271 Knowing that the rebels threatened
early on Thrace, the province closest to the capital, most resources were
concentrated for its defense. Maintaining large numbers on the Danube
frontier, which by now was accessible only the sea, would have been a
useless effort. As in the case of the rebellion of 1072, the ultimate bastion
of Byzantine authority in the Danube region was Isaccea. In 1190, Isaac II
Angelos decided to move again against Bulgaria. The land army besieged
Tărnovo, but without any success, for the Cumans arrived to attack the
besieging army. The Byzantines were defeated after they withdrew in the
mountains, in that already mentioned ambush of the Trjavna Pass. This
campaign is covered by both Niketas Choniates and George Akropolites.
Choniates, in a speech addressed to the emperor in 1190, refers to fighting
taking place in the “Danube islands”: “(Istros) is after many years struck
by Romaic [Byzantine] oars.”272 This, as Genoveva Tsankova-Petkova has
demonstrated, is only wishful thinking. It appears that the idea of send-
ing the navy to the Danube in order to hinder the crossing of the river
by the Cumans was never put to practice, because of unknown reasons.
The speech included congratulations made before the launching of the
campaign.273 Be as it may, the aforementioned plan may be associated
with messages the emperor sent to Isaccea, as illustrated by two of his
seals found on the site.274
A direct consequence of the defeat inflicted upon the Byzantine army
in the Trjavna Pass was the abandonment of Varna, the most important
port on the maritime way to the Danube (it was recovered only in 1193).
Mesembria was also taken at about the same time.275 Varna was again
retaken by Bulgarians on March 24th 1201, after a three day siege. This
270 Ansbertus, 33 (transl. Loud, 64): in Bulgariae maxima parte ac versus Danubium,

quousque mare influat, quidam Kalopetrus Flachus ac frater eius Assanius cum subditis Fla-
chis tyranizabat; Brătescu 1919, 22.
272 Niketas Choniates, Orationes, A (ed. Van Dieten, 3–4).
273 Tsankova-Petkova 1981, 181–185. Gjuzelev 1986, 205 admitted that the expedition to
the Danube was real.
274 Barnea 1971, 361; Barnea 1997, 359; Mănucu-Adameşteanu 2001 a, 206.
275 Wolff 1949, 184–188; Guilland 1964, 135; Gjuzelev 1986, 205–208.
was the last point the Empire still controlled north of the Stara Planina
Mountains. Constanteia was also taken in 1202. The Bulgarian occupation
of those port cities made it impossible for the Byzantine navy to get to
the mouths of the Danube.276 As a matter of fact, no other attempt to
attack Bulgaria from the Danube is known after 1190. The Danube region
was lost, although it was partially recovered for a brief while by Emperor
Michael VIII Paleologos (1259–1282). In a speech Manuel Holobolos wrote
for the occasion, Michael is praised for having forged in 1273 an alliance
against Bulgaria with the Mongol warlord Nogai against Bulgaria, as “the
many Paristrian islands” returned to Byzantine control.277 Holobolos’ fig-
ure of speech does not refer to some small islands in the Danube Delta,
but to northern Dobrudja, previously known as Western Mesopotamia.
During the third quarter of the fourteenth century, this territory was a
Mongol-Byzantine condominium centered upon Isaccea.278
The western part of the Danube region was also lost shortly after the
rebellion in Bulgaria. In 1189, when the crusaders entered the Balkans on
their way to Constantinople, there was still a duke in Braničevo (dux de
Brandicz). According to Ansbertus’ Historia de expeditione Friderici imper-
atoris, the troops under the command of the duke intercepted the crusad-
ers moving to Niš. 279 Stephen Nemanja and Peter, the leader of the Vlachs
and of the Bulgarians, offered their support to Frederick I Barbarossa for a
common war against Isaac II, but the German emperor preferred to main-
tain good relations with Byzantium.280 Soon after the incidents of 1189,
there is no mention of Byzantine authority in the region. In 1190, in coop-
eration with Peter, Stephen Nemanja, who had already started the expan-
sion of Serbia in the central part of the Balkans, took over the region of
Braničevo-Niš from the Byzantines. Isaac II responded with an expedition
against Serbia in 1191. After a victory won somewhere in the valley of the
Morava river, a peace was concluded, which returned Niš, Braničevo and
Belgrade to the empire.281 King Béla III of Hungary took advantage of the
276 Gjuzelev 1978, 53; Gjuzelev 1981, 19; Gjuzelev 1986, 209; Stephenson 2000, 310.
277 Laurent 1945, 188; Stănescu 1974, 412; Năsturel 2003, 351.
278 Laurent 1945, 184–198; Bănescu 1946, 113–115; Oberländer-Târnoveanu 1995–1996,
208–209.
279 Ansbertus, 27–28 (transl. Loud, 60); Stephenson 2000, 295.
280 Ansbertus, 58 (transl. Loud, 84–85); Wolff 1949, 184; Guilland 1964, 132–133; Iosipescu
1994, 265–268; Stephenson 2000, 294–300; Curta 2006, 361.
281 Niketas Choniates, Isaac Angelos, III (ed. Van Dieten, p. 434; trad. Magoulias, pp.
238–239); Niketas Choniates, Orationes, D (ed. Van Dieten, 27–33); Guilland 1964, 135–136;
Makk 1989, 123; Fine 1994, 25–26; Stephenson 2000, 301; Vásáry 2005, 45–46; Curta 2006,
335.
166 chapter three
defeat inflicted upon Nemanja to attack Serbia in the winter of 1192–1193.

The Hungarians occupied an area south of Braničevo, from which they
eventually withdrew at the request of Pope Celestin III. The Byzantines
remained in control of the region between Niš, Braničevo and Belgrade
until 1199, when it fell to the Bulgarians. The privilege granted to Venice
in November 1198 mentions provincia Nisi and Uranisoue (Braničevo)282
as under Byzantine control. However, in 1202 Braničevo must have been
under Bulgarian rule, for the local bishop was a suffragan of the Bulgarian
patriarch. The Bulgarians may have taken Braničevo in 1199, knowing that
Kaloyannes campaigned in the area in that year, as mentioned in his cor-
respondence with Pope Innocent III.283
Thus ended the dispute between Hungary and the Byzantine Empire
for domination in the Middle Danube region. The geopolitical situation
of the Danube area entered a new phase marked by the confrontation
between Bulgaria, Hungary, and the Cumans.
282 Tafel, Urkunden I, 261.

283 Curta 2006, 382.
Conclusion
John Tzimiskes’ conquest of the Lower Danube region was the unex-
pected result of the war against Svyatoslav, who, taking the chance offered
by Nikephoros Phokas’ strategic mistake, had attempted to establish his
domination over that region. “Unexpected” in this context refers to the
lack of any evidence that Nikephoros Phokas’s goal was either to destroy or
to occupy Bulgaria. His intention in 968, when he summoned Svyatoslav’s
assistance, was to keep Bulgaria under Byzantine hegemony, as a buffer
state against attacks from the north. His intention was however made
irrelevant by Svyatoslav’s plans to ensconce himself in the region by the
Black Sea and the river Danube. The military operations carried out on
the Lower Danube by Byzantium’s land and naval forces in 971 were only
a reaction to the plans of the Rus’ prince. The unfolding of the political
events had left a power vacuum in Bulgaria, which John Tzimiskes’s con-
quest now attempted to solve. The restoration of the Byzantine domina-
tion in Bulgaria was initially conceived as a means to prevent future Rus’
attacks. That was the primary purpose of dispatching troops from Thrace
in the new theme of Ioannoupolis, and that was the reason for which no
such forces were stationed in western or southern Bulgaria, where the
Byzantines wages a comparatively longer and much more difficult war
against the rebels of 976 led by Samuel. The occupation of those regions
was not regarded as necessary, or even possible, for no threat from the
Rus’ or any other enemy was expected there. The same strategic goal
explains why a katepanate called Western Mesopotamia was established
south of the Danube Delta, in the region commonly crossed by the Rus’
on their way to Constantinople.
This neglect of the other parts of Bulgaria turned out to be another
strategic mistake, with very serious consequences. Once the rebels of 976
revived the Bulgarian state, they effectively became a mortal threat to all
Byzantine possessions in the Balkans. The was against Samuel led to a
new organization of the Danube region, after the territories lost in 986
were recovered and new ones conquered, all the way to Sirmium. This
organization was implemented in three stages (1000, 1002, 1018–1019) and
consisted of an extension of the theme of Dristra to the west, in addition
to the creation of another theme of Sirmium. In the southern parts of
the former Bulgarian state the themes of Bulgaria and Serbia were estab-
lished. The occupation was based on garrisons in forts, some of which
168 conclusion
were restored Roman forts, others built anew, especially next to the fords
across the Danube. There were altogether fewer fortifications than those
in existence between the 2nd and the 7th century, an indication that the
military forces available were not as numerous as in Antiquity.
Since the Rus’ had ceased to be a threat to the Danube frontier since
the age of Basil II, there was no need of a navy any longer. The new ene-
mies, Pechenegs and Hungarians, had no naval forces. The defense system
developed by Basil II was therefore based only on monitoring the crossing
points around the main fords, which mounted warriors could use to cross
the Danube. This implied a small number of forts manned by stratiotai.
This strategy failed in the long run, because forts around the fords or other
points along the Danube could not effectively prevent the Pecheneg, Uze
and Cuman raids, many of which took place in the winter, when the entire
river was frozen. The alternative solution adopted by Michael IV in 1036,
and developed by Constantine IX in 1045, was to appease the Pechenegs,
first by means of bribes, next by allowing them to settle in the lands south
of the Danube, as frontier guards. Such a policy proved effective for as
long as the forts opened for contacts with the barbarians remained under
the control of the central administration, i.e. until the secession of 1072.
That secession marked the turning point in the history of the region, as
the Pechenegs previously established in the empire took over the forts,
including Dristra, soon to be joined by the Pechenegs who had remained
in the lands north of the river.
The numerous raids and invasions across the Danube led to the adop-
tion of a third solution, namely moving the defense line from the Danube
to the Stara Planina range. According to Paul Stephenson, this solution
was adopted in the 12th century because of an ever diminishing number
of troops available for the defense of the Danube.1 Under the new circum-
stances, the province of Paradunavon became a half-open space. Only a
few points on the river (Isaccea, Hârşova, Axiopolis and Dristra) remained
under direct Byzantine control. Those were outposts needed for the moni-
toring of the fords, as well as anchorages for the navy based in Anchialos.
Nonetheless, it is a mistake to draw the conclusion that the Byzantines
had by then been completely lost the region along the Danube.2 The lands
between the Danube and the Stara Planina remained a periphery, but a
periphery of Byzantium.
1 Stephenson 2000, 103–105.

2 Tăpkova-Zaimova 1980, 336.
conclusion 169
It is quite possible that the imperial authority at Dristra was exer-

cised through some some local leaders—either Pechenegs or Cumans—
who, although subjects to the emperors, were for all practical purposes
autonomous.3 This was after all the status granted earlier to Kegen for
Patzinakia. An agreement with Pecheneg chieftains in Paradunavon could
have been reached in the aftermath of the victory of 1091. That is why
Gheorghe I. Brătianu has long suggested that the appearance in the late
eleventh century of Turkic autonomous polities in Dobrudja, all of which
were politically dependent upon the empire.4 Under such circumstances,
the presence of tagmata commanded by a duke was no more necessary
for the defense of that remote province. The lack of any information about
commanders in the cities on the Danube cities is in itself significant. The
region continued to be part of the empire, but the Byzantine defense line
was moved to the south, on the Stara Planina Mountains.
The transformation of the region between the Danube and the Stara
Planina Mountains into a half-open space (ἐρημία) may have been inspired
by the experience on the eastern frontier, where Arab and especially Seljuk
attacks have rendered obsolete the network of fortifications. The solution
adopted in the East was to turn the borderlands into a “no man’s land,”
with no population whatsoever, in order to create difficult conditions for
any invasion which would suffer because of lack of supplies.5 After 1081,
the eastern frontier was de facto outside imperial control of the empire,
with only a few isolated forts left as residences for the dukes.6
The invasions of 1027, 1032–1036, 1047 and 1065 show that how ineffec-
tive the blockade of the fords across Danube really was when confronted
with large numbers of nomadic warriors looking for booty in cities and
the countryside. Moving the defense line on the mountain range provided
the opportunity to fight the nomads in an environment with which they
were not familiar and in which ambushes could easily be set up in moun-
tain passes, which were altogether much easier to defend than the line
of the Danube. Much like in the east, the region deserted in the north,
between the mountains and the Danube, was meant to operate as a “no
man’s land.”7 William of Tyre noted that Byzantines had depopulated the
3 Shepard 1979, 238.

4 Brătianu 1999, 234.
5 Ahrweiler 1974, 219–221. For the features of the eastern frontier, see Holmes 2002,
83–104.
6 Cheynet 2001 b, 65.
7 Magdalino 1993, 134.
170 conclusion
former ancient provinces of Dacia Ripensis and Dacia Mediterranea for

military purposes: in utraque videlicet Dacia, etiam hodie non permittunt
habitatores introire. Likewise, Epirus, a region that could be crossed in
four days, was left deserta et invia et alimentis carentia.8 Although this
account refers to the First Crusade, it is likely that the situation described
is that of William’s lifetime, for it does not appear in earlier sources. In
other words, William must have experienced the situation himself, when
in 1169 he traveled from Constantinople to Rome. This, then, was the
strategy applied by Manuel I, although William of Tyre’s authorial finger
appears to point to Basil II. Modern historians emphasize the deliberate
nature of this policy aimed at creating buffer zones next to the frontiers.9
If areas next to the frontier were depopulated, then there was no booty—
goods or prisoners—to attract the Pechenegs or the Cumans. There is no
evidence that the nomads ever intended, of their own will, to settle in the
lands south of the Danube. All known cases of settlement were the result
of Byzantine decisions and took place under Byzantine supervision.
The events of 1095 have demonstrated the advantage of placing the
defense line on the Stara Planina Mountains. This new direction taken
by the defense system under Alexios I is strikingly similar to the decision
taken in 275 by Emperor Aurelian, who abandoned Dacia north of the
Danube in favor of a better defense on the line of the river. In both cases,
this strategy of retrenchment was a response to increasing pressure on the
earlier line of defense and to diminishing resources. The only difference
is that unlike Aurelian, Alexios I did not entirely abandon the position to
the north of the new line of defense (the province of Paradunavon), but
turned into a half-open space in front of the theme of Anchialos.
The depopulation of areas between the Danube and the Stara Planina
Mountains probably began shortly after the Pecheneg crisis of 1047–1053,
and was definitely well on its way during the secession of 1072–1091.
When the former Hungarian king Salomon withdrew to the Danube after
being defeated by Nicholas Mavrokatakalon in 1087, he had to go through
“deserted places” (errantes itaque ferebantur per inania), before reach-
ing a castrum desertum et vacuum, which was located somewhere on the
bank of the river.10 The archaeological excavations in northern Bulgaria
have revealed clusters of settlements in the highlands, along the Danube,
8 William of Tyre, II, 4 (ed. Babcock, Krey, I, 122).

9 Hendy 1985, 39.
10 SRH, I, 410.
conclusion 171
and on the Black Sea coast. There are, however, no settlements at all in
the entire region between the Danube and the Stara Planina Mountains,
which could be dated to the 11th and 12th centuries.11 During the former
century, there is a visible increase of population in all forts along the Dan-
ube, which in most cases witnessed the growth of a civilian settlement
outside the fort’s walls (Garvăn, Isaccea, Nufăru, Tulcea, Hârşova). By the
12th century, however, there is evidence of a serious demographic decline,
with most extramural settlements being abandoned, while in other cases
the occupation ceased inside the fort as well.
The demographic decline in the Danube forts must be regarded in con-
nection with the withdrawal of the defense line to the Stara Planina Moun-
tains. This made most forts along the Danube obsolete, and any efforts to
repair or rebuild them utterly unnecessary. The forts had failed in their
mission to prevent invasions, and no troops left behind would have been
able to win against the Pechenegs and the Cumans. Military successes, if
any, obtained by the Byzantines against the raiders were always caused
by their ability to move behind the enemy lines or to attack the Pech-
enegs or the Cumans when exhausted. This was no doubt the conclusion
Emperor Alexios I reached when deciding to change the general strategy
in the region.
The transformation of the region between the Lower Danube and the
Stara Planina Mountains into a buffer zone coincided in time with the
increasing military importance of the region on the border with Hungary.
Resources began to concentrate in that region under John II and espe-
cially Manuel I, as Hungary was increasingly perceived as capable to block
access to the road leading to Thessaloniki and Constantinople along the
Morava valley. This is the ultimate reason for the growth of both Belgrade
and Braničevo during the reign of Manuel, when both forts were com-
pletely rebuilt. The Serbs were an additional threat, as they often rebelled
with Hungarian assistance. The strategy implemented by the Comnenian
emperors, therefore, was meant to increase the security in the central parts
of the Balkans, at the expense of the Lower Danube. In the end, Manuel
was successful against Hungary by means of both war and a dynastic alli-
ance. But the ultimate (albeit unintended) consequence of the marriage
between his daughter and the future Hungarian king was that Béla III got
involved in the conflict caused by the usurpation of Andronikos, which
11 Borisov 2007, 74–78.

172 conclusion
reopened the Hungarian-Byzantine dispute over the region of Belgrade

and Niš region, a dispute Hungary later took with Bulgaria as well.
The collapse of the Byzantine military organization on the Danube was
the direct consequence of the Vlach-Bulgarian rebellion of 1185. The rebels
were assisted by Cumans, who brought in valuable military experience
and power. A century after the secession of Paradunavon, primarily
because of economic reasons, the region along the Danube was finally
completely lost by the Byzantines, who were no more capable to match
the military power of the rebels and prevent them from taking over key
strategic points in the mountains (passes and forts), which had until then
secured the defense of the European parts of the Empire.
BIBLIOGRAPHY
Primary Sources
Akropolites: George Akropolites: The History, Introduction, translation and commentary

by Ruth Macrides, Oxford, 2007.
Al-Idrisi: Géographie d’Edrisi traduite de l’Arabe en français d’après deux manuscrits de la
Bibliothèque du roi et accompagnée de notes par P. Amedée Jaubert, vol. II, Paris, 1840
(Recueil de voyages et de mémoires publié par la Société de Géographie, 6); Map-
pae Arabicae. Arabische Welt- und Länderkarten des 9.–13. Jahrhunderts in arabischer
Urschrift, lateinischer Transkription und Übertragung in neuzeitliche Kartenskizzen. Mit
einleitenden Texten herausgegeben von Konrad Miller, Band II. Die Länder Europas und
Afrikas im Bilde der Araber, Stuttgart, 1927.
Albertus Aquensis: Albert of Aachen, Historia Ierosolimitana. History of the Journey to Jeru-
salem, edited and translated by Susan B. Edgington, Oxford, 2007.
Ananias: The geography of Ananias of Širak: Ašxarhac’oyc’, the long and the short recen-
sions. Introduction, translation and commentary by Robert H. Hewsen, Wiesbaden, 1992
(Beihefte zum Tübinger Atlas des Vorderen Orients, 77).
Anna Comnena: Anna Comnena, The Alexiad. Translated by E. R. A. Sewter, London,
2003.
Annales Regni Francorum: Annales Fuldenses sive Annales Regni Francorum Orientalis ab
Einhardo, Ruodolfo, Meginhardo Fuldensibus Seligenstadi, Fuldae, Mogontiaci Conscripti
cum Continuationibus Ratisbonensi et Altahensibus, post editionem G. H. Pertzii recog-
novit Fridericus Kurze, MGH, Scriptores Rerum Germanicarum in usum scholarum,
vol. 7, Hannovra, 1891.
Ansbertus: Ansbertus, Historia de expeditione Friderici imperatoris, in MGH. Scriptores
Rerum Germanicarum, Nova Series, tomus V, Berlin, 1928, 1–115; G. A. Loud, The Cru-
sade of Frederick Barbarossa. The History of the Expedition of the Emperor Frederick and
Related Texts, Farnham (Ashgate), 2010 (Crusade Texts in Translation, 19).
Anonymous Book on Tactics: Anonymous Book on Tactics, in Three Byzantine Military Trea-
tises, Text, translation, and notes by G. T. Dennis, Washington DC, 1985 (CFHB, Series
Washingtoniensis, XXV) (Dumbarton Oaks Texts, 9), 241–335.
Attaliates: Michaelis Attaliotae Historia, opus a Wladimiro Bruneto de Presle inventum,
descriptum, correctum recognovit Immanuel Bekkerus, Bonn, 1853; Miguel Ataliates,
Historia, introduccion, edicion, traduccion y comentario de Immaculada Pérez Martín,
Madrid, 2002 (Nueva Roma, 15).
Constantine Porphyrogenitus: Constantine Porphyrogenitus, De Administrando Imperio,
vol. I (Greek text edited by G. Moravcsik, English translation by R. J. H. Jenkins), Wash-
ington, 1967 (Dumbarton Oaks) (CFHB, Series Washingtoniensis, I) (Dumbarton Oaks
Texts, 1); vol. II. Commentary (edited by R. J. H. Jenkins), London, 1962.
Demetrios Tornikes: Georges et Dèmètrios Tornikès, Lettres et discours. Introduction,
texte, analyses, traduction et notes par J. Darrouzès, Paris, 1970.
Eberwinus: Miracula Sancti Symeonis auctore Eberwino, in MGH, Scriptores, VIII, ed.
G. H. Pertz, Hannoverae, 1848, 209–211.
George Monachos: Georgius Monachos, Bioi ton neon basileon, in Theophanes Continuatus,
Ioannes Cameniata, Symeon Magister, Georgius Monachus, ex recognitione Immanuelis
Bekkeri, Bonn, 1838, 761–924.
Glykas: Michaelis Glycae Annales, recognovit Immanuel Bekkerus, Bonn, 1836 (Corpus
Scriptorum Historiae Byzantinae).
174 bibliography
Hudud al-’Alam: Hudud al-’Alam, The Regions of the World. A Persian Geography, 372 A.H.–
982 A.D., translated and explained by V. Minorsky, Oxford, 1937.
Kekaumenos: Cecaumeno, Raccomandazioni e consigli di un galantuomo (Strategikon).
Testo critico, traduzione e note di Maria Dora Spadaro (Hellenica. Testi e Strumenti di
letteratura greca antica, medievale e umanistica. Collana diretta di Enrico V. Maltese, 2),
Alexandria, 1998.
Kinnamos: John Kinnamos, Deeds of John and Manuel Comnenus, translated by Charles M.
Brand, New York, 1976.
Leo Grammaticus: Leonis Grammatici Chronographia, ex recognitione Immanuelis Bekkeri,
Bonn, 1842, 1–331.
Leo, Taktika: The Taktika of Leo VI. Text, Translation and Commentary by George T. Dennis,
Washington, 2010 (CFHB, Series Washingtoniensis, XLIX) (Dumbarton Oaks Texts, 12).
Leo the Deacon: The History of Leo the Deacon: Byzantine Military Expansion in the Tenth
Century, Introduction, translation, and annotations by A. M. Talbot and D. Sullivan, with
the assistance of G. T. Dennis and S. McGrath, Washington DC, 2005.
Masoudi: Maçoudi, Les prairies d’or. Texte et traduction par C. Barbier de Meynard et
Pavet de Courteille, 9 vol., Paris, 1861–1877.
Matthew of Edessa: Armenia and the Crusades, Tenth to Twelfth Centuries. The Chronicle
of Matthew of Edessa, translated from the Original Armenian with a Commentary and
Introduction by A. E. Dostourian, Lanham, 1993.
Michael Glykas: Michaelis Glycae Annales, recognovit Immanuel Bekkerus, Bonn, 1836.
Michael the Syrian: Chronique de Michel le Syrien, Patriarche Jacobite d’Antioche (1166–1199),
éditée pour la première fois et traduite en français par Jean-Baptiste Chabot, vol. III,
Paris, 1905.
Michael Psellos: The Chronographia of Michael Psellus. Translated from the Greek by
E. R. A. Sewter, London, 1953.
Niketas Choniates: Nicetae Choniates Historiae, recensuit J. V. Van Dieten (CFHB. Series
Berolinensis, XI/1–2), Berlin, New York, 1975; O city of Byzantium, Annals of Niketas
Choniates. Translated by Harry J. Magoulias, Detroit, 1984.
Niketas Choniates, Orationes: Nicetae Choniates Orationes et epistulae, recensuit J. V. Van
Dieten (CFHB. Series Berolinensis, III), Berlin, New York, 1972.
Nikephoros: Nikephoros Patriarch of Constantinople Short History. Text, Translation, and
Commentary by Cyril Mango, Washington, 1990 (CFHB, Series Washingtoniensis, XIII).
Nicholas I, Letters: Nicholas I, Patriarch of Constantinople, Letters. Greek text and Eng-
lish translation by R. J. H. Jenkins and L. G. Westerink, Washington, 1973 (CFHB, Series
Washingtoniensis VI).
Peri strateias: Peri strateias, in Three Byzantine Military Treatises, Text, translation, and
notes by G. T. Dennis, Washington DC, 1985 (CFHB, Series Washingtoniensis, XXV)
(Dumbarton Oaks Texts, 9), 1–137.
Russian Primary Chronicle: S. Hazzard Cross, O. Sherbowitz-Wetzer, The Russian Primary
Chronicle. Laurentian Text, Cambridge (Mass), 1973.
Scriptor Incertus: Scriptor Incertus, Historia de Leone Bardae Armenii filio, in Leonis Gram-
matici Chronographia, ex recognitione Immanuelis Bekkeri, Bonn, 1842, 335–362.
Skylitzes: Ioannes Scylitzes, Synopsis historiarum, recensuit I. Thurn (CFHB. Series Bero-
linensis, V), Berlin, New York, 1973; Jean Skylitzès, Histoire des empereurs de Byzance,
traduction française par Bernard Flusin, notes par Jean-Claude Cheynet, Paris, 2003
(Réalités byzantines, 8); John Skylitzes, A Synopsis of Byzantine History, 811–1057, Trans-
lation and Notes by John Wortley, Cambridge, 2010.
Skylitzes Continuatus: He synécheia tes Chronographias tou Ioannou Skylitse (Ioannes
Scylitzes Continuatus), ed. E. Th. Tsolakis, Thessaloniki, 1968.
Symeon, Chronikon: Symeonis Magistri et Logothetae Chronicon recensuit Stephanus
Wahlgren, Berlin, New York, 2006 (CFHB. Series Berolinensis, XLIV/1).
Tafel, Urkunden I: G. L. Fr. Tafel, G. M. Thomas, Urkunden zur älteren Handels- und Staats-
geschichte der Republik Venedig, mit besonderer Beziehung auf Byzanz und die Levante
(Fontes Rerum Austriacarum. Diplomataria et acta, XII), I (814–1205), Wien, 1856.
bibliography 175
Theophanes: Chronicle of Theophanes Confessor. Byzantine and Near Eastern History,

AD 284–813, translated with introduction and commentary by Cyril Mango and Roger
Scott, Oxford, 1997.
Theophanes Continuatus, in Theophanes Continuatus, Ioannes Cameniata, Symeon Mag-
ister, Georgius Monachus, ex recognitione Immanuelis Bekkeri, Bonn, 1838, 3–481.
William of Tyre: Historia rerum in partibus transmarinis gestarum (William Archbishop of
Tyre. A History of Deeds Done Beyond the Sea. Translated and annotated by E. A. Babcock
and A. C. Krey. 2 vols. New York, 1941).
Yahya-ibn-Said: Histoire de Yahya-ibn-Said d’Antioche, ed. par I. Kratchkovsky, A. Vasiliev,
în Patrologia Orientalis, vol. 18, Paris, 1924, 347–520.
Zonaras: Ioannes Zonaras, Epitomae Historiarum, vol. III, ed. Theodorus Büttner-Wobst,
Bonn, 1897.
Secondary Sources
Agrigoroaei, V. 2006: Vikingi sau ruşi. Noi cercetări asupra complexului de la Basarabi-
Murfatlar, Apulum, 43, 2, 25–49.
Ahrweiler-Glykatzi, H. 1960: Recherches sur l’administration de l’Empire Byzantin aux
IXe–XIe siècles, Bulletin de Correspondance Hellénique, 84, 1–111.
Ahrweiler, H. 1966: Byzance et la mer. La marine de guerre, la politique et les institutions
maritimes de Byzance aux VIIe–XVe siècles, Paris.
—— 1974a: La frontière et les frontières de Byzance, CIEB XIV, vol. I, 209–230.
—— 1974b: H. Ahrweiler, L’escale dans le monde byzantin, Recueils de la Société Jean
Bodin, Bruxelles, 32 (Idem, Byzance: les pays et les territoires, Variorum, London, 1976,
VI), 161–178.
Anastasijević, D. 1930: L’hypothèse de la Bulgarie Occidentale, in L’art byzantin chez les
Slaves. Les Balkans. Premier recueil dédié à la mémoire de Théodore Uspenskij, I, Paris,
20–36.
Anastasijević, D., Ostrogorsky, G. 1951: Les Koumanes pronoïaires, Annuaire de l’Institut de
Philologie et d’Histoire Orientales et Slaves, 11 (Mélanges Henri Grégoire, III), 19–29.
Andronic, A. 1969: Români, bulgari, ruşi şi bizantini la Dunărea de Jos la sfârşitul secolului
al X-lea, Memoria Antiquitatis, 1, 207–215.
Angelova, S. 1987: Sur la caractéristique de la céramique du Haut Moyen Âge, provenant de
Drăstăr (Silistra), in Dobrudža. Études, 93–114.
—— 1993: Kăm topografijata na srednovekovnija Drăstăr prez XII v. (New Aspects of the
topography of the mediaeval town of Drustur), in Prinosi kăm bălgarskata arheologija, II,
ed. D. Ovčarov, Sofia, 52–57.
Angelova, S., Koleva, R. 2004: Der frühmittelalterliche Drăstăr (Silistra), in G. Fusek (ed.),
Zborník na počest Dariny Bialekovej, Nitra, 19–30.
Angold, M. 1984: The Byzantine Empire, 1025–1204. A Political History, London.
Antoljak, S. 1972: Die Wahrheit über den Aufstand der Comitopulen, în Actes du IIe Congrès
International des Études du Sud-Est Européen, 2, Athènes, 379–384.
Antoniadis-Bibicou, H. 1963: Recherches sur les douanes à Byzance. L’“octave”, le “kom-
merkion” et les commerciaires, Paris.
Atanasov, G. 1991: Khristijanskij pametniki ot rannosrednovekovnata krepost do selo Rujno,
Dulovsko, (Christliche Denkmäler aus der frühmittelalterlichen Burg in der Nähe von
Rujno—Region Dulovo), Dobrudža, 8, 28–50.
—— 1994: La Vicina médiévale et la forteresse de Nufăru, EB, 30, 1, 109–128.
—— 2003: De nouveau pour la date initiale de folles byzantines anonymes classe “B”, AMV, 2,
289–298.
—— 2012: Drăstăr (Silistra) i dunavskata rezidentsija na bălgarskite khanove prez părvata
polovina na IX v. (Drăstăr (Silistra) and the Danube residence of the Bulgar khans during
the first half of the 9th c.), Arh, 53, 1, 28–45.
Atanasov, G., Jordanov, I. 1994: Srednovekovnjat Vetren na Dunav, Shumen.
176 bibliography
Bálint, C. 1991: Südungarn im 10. Jahrhundert, Budapest.

Baraschi, S. 1981: Izvoare scrise privind aşezările dobrogene de pe malul Dunării în secolele
XI–XIV, RdI, 34, 2, 311–345.
—— 1989: Sur la topographie ponto-danubienne au Moyen Âge I. Proslaviţa, RRH, 28, 1–2,
53–68.
—— 1991: Unele probleme despre Proslaviţa, Peuce, 10, 1991, 1, 399–409; 2, 371–375.
Baraschi, S., Moghior, N. 1979: Sondajele de la Nufăru-Ilganii de Jos ( jud. Tulcea), SMMIM, 12,
186–191.
—— 1983: Săpăturile de salvare de la Nufăru (1981), SMMIM, 16, 134–139.
—— 1984–1985: Date noi referitoare la fortificaţiile medieval-timpurii din complexul arheo-
logic de la Nufăru-Ilgani ( judeţul Tulcea), SMMIM, 17–18, 144–151.
Barnea, A. et alii 2002: Barnea, A., Paraschiv, D., Stănică, A. Ostrov, jud. Tulcea, in Cronica
2001, 225–226.
—— 2004: Barnea, A., Olariu, C., Ionescu, M.S., Garvăn, com. Jijila, jud. Tulcea (Dinogetia),
in Cronica 2003, 126–127.
—— 2009: Barnea, A., Paraschiv, D., Iacob, M., Stanc, S., Ostrov, com. Ostrov, jud. Tulcea
[Beroe]. Punct: Piatra Frecăţei, in Cronica 2008, 161–162.
Barnea, I. 1960: Date noi despre Axiopolis, SCIV, 11, 1, 69–78.
—— 1971: Dinogetia et Noviodunum, deux villes byzantines du Bas-Danube, RESEE, 9, 3,
343–362.
—— 1973: Noi descoperiri din epoca feudalismului timpuriu la Dinogetia-Garvăn, jud. Tulcea
(1963–1968), MCA, 10, 291–331.
—— 1975a: Murfatlar şi Niculiţel, SCIVA, 26, 1, 95–99.
—— 1980: Dinogetia—ville byzantine du Bas Danube, Vyzantina, 10, 237–287 + Tables VII–
XXVII.
—— 1981: Sceau de Constantin IV empereur de Byzance, trouvé à Durostorum, RRH, 20, 4,
625–628.
—— 1983: Sigilii bizantine inedite din Dobrogea, Pontica, 16, 263–272.
—— 1986: Noi contribuţii la istoria themei Paristrion (Paradunavon), SCIVA, 37, 3, 268–274.
—— 1987a: Din nou despre Vasile Apokapes, SCIVA, 38, 2, 194.
—— 1987b: Sceaux byzantins de Dobroudja, SBS, 1, 77–88.
—— 1990: Sigilii bizantine inedite din Dobrogea (III), Pontica, 23, 315–334.
—— 1993a: Sceaux byzantins inédits de Dobroudja, SBS, 3, 55–67.
—— 1993b: Le Danube, voie de communication byzantine, in Praktika tou B’ diethnous Sym-
posiou he epikoinonia sto Byzantio. Epimeleia N. G. Moschonas, Athens, 577–595.
—— 1996–1997: Sigilii bizantine din Dobrogea, BSNR, 90–91, nr. 144–145, 187–194.
—— 1997a: Noviodunum în lumina sigiliilor bizantine, SCIVA, 48, 4, 353–360.
—— 1997b: Sceaux byzantins inédits de Dobroudja, EBP, 3, 93–98.
—— 2000: Regiunea Dunării de Jos în lumina sigiliilor bizantine, in Faţetele istoriei. Existenţe,
identităţi, dinamici. Omagiu Ştefan Ştefănescu, Bucureşti, 293–307.
—— 2001: Sigilii bizantine din nordul Dobrogei, in Simpozion de numismatică dedicat
împlinirii a patru secole de la prima unire a românilor sub Mihai Voievod Viteazul,
Chişinău, 28–30 mai 2000, Bucureşti, 103–108.
Barnea, I., Ştefănescu, Şt. 1971: Din istoria Dobrogei, vol. III. Bizantini, români şi bulgari la
Dunărea de Jos, Bucureşti.
Bartikian, R. M. 1967: Kăm kogo se otnasija prozvišteto PARADOYNABIS v pripiskata na
monaha Teodul ot 1059 godina? (Pour qui le surnom PARADOYNABIS de la correspon-
dence du moine Théodul de 1059 se rapporte-t-il?), Izvestija na Bălgarskoto Istoricesko
Druzestvo, 25, 315–319.
Baumann, V. H., Mănucu-Adameşteanu, Gh. 2001: Isaccea, in Cronica 2000, 108–110.
Baumann, V. H. et alii 2002: Baumann, V. H., Stănică, A., Mănucu-Adameşteanu, Gh. 2002:
Isaccea, jud. Tulcea (Noviodunum). Punct Cetate, in Cronica 2001, 157–159.
—— 2004: Baumann, V., Stănică, A., Mănucu-Adameşteanu, Gh. 2004: Isaccea, in Cronica
2003, 146–149.
bibliography 177
Bănescu, N. 1922: Les premiers témoignages byzantins sur les Roumains de Bas-Danube,
BNJ, 3, 287–310.
—— 1927: La domination byzantine sur les régions du Bas-Danube, BSH, 13, 10–22.
—— 1941: La domination byzantine à Matracha (Tmutorokan), en Zichie, en Khazarie et en
‘Russie’ à l’époque des Comnènes, BSH, 22, 2, 57–77.
—— 1946: Les duchés byzantins de Paristrion (Paradounavon) et de Bulgarie, Bucarest.
—— 1948: Les frontières de l’ancien État bulgare, in Mémorial Louis Petit. Mélanges d’histoire
et d’archéologie byzantines (Archives de l’Orient Chrétien, I), Bucarest, 4–14.
—— 1963: À propos de Basile Apokapes, duc de Paradounavis (â Paristrion). La notice du
moine Théodule (1059), RESEE, 1, 1–2, 155–158.
Beševliev, V. 1962: Zur Geographie Nordost-Bulgariens in der Spätantike und im Mittelalter,
Linguistique Balkanique, Sofia, 4, 57–80.
—— 1963: Die protobulgarischen Inschriften (BBA, 23), Berlin.
—— 1971: Die Feldzüge des Kaisers Konstantin V gegen die Bulgaren, EB, 7, 3, 5–17.
—— 1981: Die protobulgarische Periode der bulgarischen Geschichte, Amsterdam.
—— 1985: Mundraga i Tiča-Vičina (Mundraga und Tiča-Vičina), INMV, 21 (36), 17–24.
Bica, I. 2003: Thema Paristrion (Paradounavon) în istoriografia bizantină şi română,
Piteşti.
Birkenmeier, J. W. 2002: The Development of the Komnenian Army: 1081–1180 (History of
Warfare, vol. 5), Leiden.
Bløndal, S. 1981: The Varangians of Byzantium. An Aspect of Byzantine Military History,
Cambridge.
Bogdan, D. P. 1978: Paleografia româno-slavă. Tratat şi album, Bucureşti.
Bogdan, D. P., Comşa, E., Panaitescu, P. P. 1951: Inscripţia slavă din Dobrogea din anul 943,
Studii, 4, 3, 122–134 [the authors are not mentioned, the paper being ascribed to a “team
from the Historical Institute”, because the signature of Panaitescu was forbidden by
political reasons in the first years of the Communist regime].
Bogdan-Cătăniciu, I. 1996: I Valli di Traiano nella Dobrugia. Considerazioni sulle fotografie
aeree, in Omaggio a Dinu Adameşteanu, a cura di Marius Porumb, Cluj-Napoca, 201–226.
—— 2006: Cercetări aerofotografice pe valurile din Dobrogea. Aspecte demografice, in
Orgame/Argamum. Supplementa I. À la recherche d’une colonie. Actes du Colloque Interna-
tional 40 ans de recherche archéologique à Orgame/Argamum: Bucarest-Tulcea-Jurilovca,
3–5 octobre 2005; textes reunis par M. Mănucu-Adameşteanu, Bucureşti, 407–427.
Bogrea, V. 1971: Pagini istorico-filologice. Ediţie îngrijitǎ, studiu introductiv şi indice de
M. Borcilǎ şi I. Mǎrii, Cluj.
Boldur, A. 1988: Teritoriul Moldovei faţă de principatele de Kiev şi Halici, AM, 12, 225–232.
Bolşacov-Ghimpu, A. A. 1967: La localisation de la cité byzantine de Demnitzikos, RESEE, 5,
3–4, 543–549.
—— 1972: O ştire bizantină din Dobrogea despre un voievodat creştin de la nordul Dunării,
la sfârşitul secolului al X-lea, Glasul Bisericii, 31, 1–2, 104–116.
—— 1973: Localisations de sites d’époque romaine et byzantine dans la zone du Bas-Danube,
RESEE, 11, 3, 553–561.
Bonarek, J. 2007: Le Bas Danube dans le seconde moitié du XI-ème siècle: nouveaux états ou
nouveaux peuples?, in Byzantium, New Peoples, New Powers: The Byzantino-Slav Con-
tact Zone, From The Ninth To The Fifteenth Century, ed. M. Kaimakamova, M. Salamon,
M. Smorąg-Różycka (Byzantina et Slavica Cracoviensia, V), Cracow, 193–200.
Borisov, B. 2007: Demographic and Ethnic Changes during X–XII Century in Bulgaria, AB,
11, 2, 71–84.
Botzan, M. 1992: Pour localiser Vicina: histoire et milieu géographique, RESEE, 30, 1–2,
61–73.
Božilov, I. 1973a: L’inscription du Jupan Dimitre de l’an 943 (théories et faits), Études Histo-
riques, Sofia, 6, 11–28.
—— 1973b: Kăm tălkuvaneto na dve izvestija na I. Skilitza za gradovete na dolnia Dunav v
kraja na X v (Au sujet de l’interpretation de deux passages chez I. Skylitzès concernant les
villes du Bas Danube vers la fin du Xe siècle), INMV, 9 (24), 111–122.
178 bibliography
—— 1975: Kăm istoričeskata geografija na severozapadnoto Černomorije (Au sujet de la

géographie historique de la région nord-ouest de la Mer Noire), INMV, 11 (26), 27–36.
—— 1976: Kăm istoriceskata geografija na severozapadnoto Cernomorije, Mesopotamia
-Mesopotamia tes Dyseos (Au sujet de la géographie historique de la région du Bas Dan-
ube: Mesopotamia -Mesopotamia tes Dyseos), INMV, 12 (27), 19–32.
—— 1978: Hase’s Anonym and Ihor Sevcenko’s Hypothesis, BB, 5, 245–259.
—— 1980: À propos des rapports bulgaro-byzantines sous le tzar Symeon (893–912), BB, 6,
73–81.
—— 2008: L’administration byzantine en Bulgarie (1018–1186): le cas de Paristrion-
Paradounavon (Paradounavis), in Byzantium and the Bulgarians (1018–1185), ed. by
K. Nikolaou, K. Tsiknakis (National Hellenic Research Foundation, Institute of Byzan-
tine Research, International Symposium 18), Athens, 91–97.
Božilov, I., Dimitrov, H. 1985: About the Historical Geography of the Northern Black Sea
Coast, BHR, 13, 4, 50–68.
Brătescu, C. 1919: Nume vechi ale Dobrogei: Vlahia lui Asan, Vlahia Albă (1186–sec. XIII),
Arhiva Dobrogei, 2, 1, 18–31.
—— 1920: Dobrogea în sec. XII: Bergean, Paristrion, Analele Dobrogei, 1, 1, 3–38.
Brătianu, Gh. I. 1935: Recherches sur Vicina et Cetatea Albă. Contributions à l’histoire de la
domination byzantine et tatare et du commerce génois sur le littoral roumain de la Mer
Noire, Bucarest.
—— 1942: Vicina II. Nouvelles recherches sur l’histoire et la toponymie médiévale du littoral
roumain de la Mer Noire. À propos des “Miscellanies” de M. J. Bromberg, RHSEE, 19, 1,
133–175.
—— 1980: Tradiţia istorică despre întemeierea statelor româneşti, Bucureşti.
—— 1999: Marea Neagră. De la origini până la cucerirea otomană, ed. V. Spinei, Iaşi.
Brezeanu, S. 1984: “La Bulgarie d’au dela de l’Ister” à la lumière des sources écrites médiévales,
EB, 20, 4, 121–135.
—— 1991: “Mésiens” chez Nicétas Choniate. Terminologie archaïsante et réalité ethnique
médiévale, EBP, 2, 105–114.
—— 2002: Toponymy and Ethnic Realities at the Lower Danube in the 10th Century. “The
Deserted Cities” in the Constantine Porphyrogenitus’ De Administrando Imperio, Annu-
ario. Istituto Romeno di Cultura e Ricerca Umanistica, Venezia, 4, 19–46.
Bromberg, J. 1937: Toponymical and Historical Miscellanies on Medieval Dobrudja, Bessara-
bia and Moldo-Wallachia (I-II), B, 12, 1, 151–180; 2, 449–475.
—— 1938: Toponymical and Historical Miscellanies on Medieval Dobrudja, Bessarabia and
Moldo-Wallachia (III), B, 13, 1, 9–71.
Browning, R. 1975: Byzantium and Bulgaria. A Comparative Study across the Early Medieval
Frontier, London.
—— 1986: Byzantine Foreign Policy and the Bulgarian state, seventh to tenth century, in
J. H. Betts, J. T. Hooker, J. R. Green (ed.), Studies in Honour of T. B. L. Webster, I, Bristol,
23–32.
Bujan, S. 2008: La chronique du prêtre de Dioclée. Un faux document historique, REB, 66,
5–38.
Busetto, R. 1992: La flotta bizantina nelle campagne contro i Bulgari condotte dall’imperatore
Constantino V, RRH, 31, 3–4, 323–330.
—— 1996: Giovanni Tzimisce e Svjatoslav di Kiev. Le operazioni militari bizantine nei Balcani
(969–971), AMN, 33, II, 9–32.
Cândea, I. 1999: Expediţia militară a Bizanţului din 1166 şi geneza culoarului Dunăre-Curbura
Carpaţilor, Angustia, Sfântu Gheorghe, 4, 153–155.
Chalandon, F. 1912: Les Comnène. Études sur l’Empire Byzantin aux XIe et XIIe siècles. II/1, 2.
Jean II Comnène (1118–1143) et Manuel I Comnène (1143–1180), New York.
Cheynet, J. C. 1983: Devaluation des dignités et devaluation monetaire dans la seconde moitié
du XIe siècle, B, 53, 2, 453–477.
bibliography 179
—— 1985: Du stratège de thème au duc: chronologie de l‘évolution au cours du XIe siècle,

TM, 9, 181–194.
—— 1990: Pouvoir et contestations à Byzance (963–1210), Paris.
—— 1991: La politique militaire byzantine de Basile II à Alexis Comnène, ZRVI, 29–30,
61–73.
—— 2001a: Sceaux de la collection Zacos (Bibliothèque nationale de France) se rapportant
aux provinces orientales de l’empire byzantin, Paris.
—— 2001b: La conception militaire de la frontière orientale (IXe–XIIIe siècles), in A. East-
mond (ed.), Eastern Approaches to Byzantium. Papers from the Thirty-third Spring Sym-
posium of Byzantine Studies, University of Warwick, Aldershot, 57–69.
—— 2002: L’Empire byzantin et la Hongrie dans la seconde moitié du XIe siècle, Acta Histo-
riae Artium Academiae Scientiarum Hungaricae, 43, 5–13.
—— 2005: Le gouvernement des marges de l‘Empire byzantin, in Le Pouvoir au Moyen Âge.
Idéologies, pratiques, représentations (Séminaire de l‘équipe de recherches Sociétés, idéolo-
gies et croyances au Moyen âge), ed. C. Carozzi, H. Taviani-Carozzi, Aix-en-Provence,
103–121.
Cheynet, J. C., Morrisson, C. 1990: Lieux de trouvaille et circulation des sceaux, SBS, 2, 105–136.
Chiriac, C. 1993: Despre “Linokastro” (Cetatea Lânii) din “Geografia” lui Edrisi, Analele
Brăilei, 1, 1, 447–454.
—— 2001: Un nouveau sceau de Grégorios Mavrokatakalon découvert à Oltina (départ. de
Constanţa), EBP, 4, 113–121.
Chiriac, C., Custurea, G. 2002: Oltina, com. Oltina, jud. Constanţa, in Cronica 2001, 222.
Chiriac, C. et alii 2003: Chiriac, C., Custurea, G., Matei, I., Oltina, com. Oltina, jud. Constanţa.
Punct: Capul Dealului, in Cronica 2002, 220.
—— 2004: Chiriac, C., Aparaschivei, D., Custurea, G., Oltina, com. Oltina, jud. Constanţa,
in Cronica 2003, 222–224.
Cigaar, K. 1974: L’émigration anglaise à Byzance après 1066. Un nouveau texte en latin sur les
Varangues à Constantinople, REB, 32, 301–342.
Cihodaru, C. 1963: Precizări necesare în legătură cu datarea valului de piatră din Dobrogea
şi însemnările toparhului bizantin, Studii, 16, 5, 1123–1135.
—— 1965: Alte precizări în legătură cu valul de piatră din Dobrogea şi cu însemnările topar-
hului bizantin, AIIAI, 2, 261–280.
—— 1968: Litoralul de apus la Mării Negre şi cursul inferior al Dunării în cartografia
medievală (sec. XII–XIII), Studii, 21, 2, 217–229.
—— 1978–1979: Observaţii în legătură cu porturile de pe cursul inferior al Dunării în secolele
XI–XIV, Cercetări Istorice, Iaşi, serie nouă, 9–10, 281–303.
—— 1979: Observaţii cu privire la procesul de formare şi de consolidare a statului feudal
Moldova în sec. XI–XIV (I), AIIAI, 16, 167–186.
Ciocîltan, V. 2011: Argumente în favoarea identităţii Vicina-Măcin, RI, serie nouă, 22, 5–6,
411–430.
Ciupercă, B. 2010: Some observations from the point of view of the Slon fortifications about
power centres in the Lower Danube between the eighth and tenth centuries, in Potestas et
communitas. Interdisziplinäre Beiträge zu Wesen und Darstellung von Herrschaftsverhält-
nissen im Mittelalter östlich der Elbe, hg. von A. Paroń, S. Rossignol, B. S. Szmoniewski,
G. Vercamer, Wrocław, 275–289.
Cliante, T. et alii, 2012: Cliante, T., Talmaţchi, C., Talmaţchi, G., Lungu, L., Valu lui Traian,
com. Valu lui Traian, jud. Constanţa (Autostrada Medgidia-Constanța). Punct: km.
197+700–198+700, in Cronica 2012, 294–295.
Comşa, E. 1951: Cercetări şi observaţii în legătură cu valurile din Dobrogea, SCIV, 2, 2, 233–
238.
Comşa, M., 1982–1984: Un drum care lega ţinutul Vrancei de Dunăre şi existenţa unui cnezat
pe valea Putnei în secolele IX–X, Vrancea. Studii şi comunicări, Muzeul Focşani, 5–7,
39–44.
180 bibliography
—— 1983: Drumuri comerciale între Carpaţi şi Dunăre în sec. IX–X, MN, 7, 101–107.
Condurachi, E., Barnea, I., Diaconu, P. 1967: Nouvelles recherches sur le Limes byzantin du
Bas-Danube aux Xe–XIe siècles, in Proceedings of the XIIIth International Congress of Byz-
antine Studies (Oxford, 5–10 September 1966), London-Oxford, 179–193.
Conea, I., Donat, I. 1958: Contribution à l’étude de la toponymie pétchénègue-comane de la
Plaine Roumaine du Bas-Danube, in Contributions onomastiques publiées à l’occasion du
VIe Congrès international des Sciences Onomastiques à Munich du 24 au 28 Août 1958,
Bucarest, 139–169.
Conovici, N., Lungu, R. 1980: Un nou tezaur de monede bizantine descoperit la Păcuiul lui
Soare, SCIVA, 31, 3, 397–402.
Covacef, Z., Nicolae, C. 2005: Hârşova, jud. Constanţa [Carsium]. Punct: Cetate, in Cronica
2004, 169–171.
—— 2007: Hârşova, jud Constanţa [Carsium]. Punct Cetate, in Cronica 2006, 177–181.
Curta, F. 1999: The cave and the dyke: a rock monastery on the tenth-century frontier of Bul-
garia, Studia Monastica, Barcelona, 41, 1, 129–149.
—— 2006: Southeastern Europe in the Middle Ages, 500–1250, Cambridge.
Custurea, G. 1986: Unele aspecte privind penetraţia monedei bizantine în Dobrogea în secolele
VII–X, Pontica, 19, 273–277.
—— 1995–1996: Catalogul monedelor bizantine anonime descoperite la Capidava, Pontica,
28–29, 301–307.
—— 2000: Circulaţia monedei bizantine în Dobrogea (sec. IX–XI), Constanţa.
—— 2000–2001: Descoperiri arheologice şi numismatice de pe raza localităţii Satu Nou (com.
Oltina, jud. Constanţa), Pontica, 33–34, 583–594.
—— 2006: Noi descoperiri monetare bizantine din aşezarea de la Capul Dealului—Oltina,
Pontica, 39, 415–421.
—— 2009: Monede bizantine descoperite în aşezarea de la Oltina-“Capu Dealului”, Pontica,
42, 612–621.
—— Matei, I. 2002–2003: Un tezaur monetar bizantin descoperit în Dobrogea, Pontica,
35–36, 433–438.
Custurea, G., Matei, I. 2007: A coin hoard found at Valul lui Traian (10th Century), in
G. Custurea, M. Dima, G. Talmaţchi, A. M. Velter, Coin Hoards of Dobrudja, I, Constanţa,
105–113.
Custurea, G., Talmaţchi, G. 2011: Repertoriul tezaurelor monetare din Dobrogea, Constanţa.
Dagron, G., Mihăescu, H. 1986: Le traité sur la guérille (De velitatione) de l’empereur Nicé-
phore Phocas (963–969), Paris.
Damian, O. 1995: Noi descoperiri monetare bizantine în Dobrogea, CCDJ, 13–14, 217–226.
—— 2003: Considérations sur la citadelle en brique de Slon-Prahova, Studia Antiqua et
Archaeologica, Iaşi, 9, 483–496.
—— 2004: Despre prezenţa politică bizantină la Dunărea de Jos în secolele VII–X, in Prinos,
283–318.
—— 2005: Repere arheologice privind oraşele medievale ale diferitelor zone. Dunărea de Jos,
Historia Urbana, 13, 1–2, 141–183.
Damian, O. et alii 2003: Damian, O., Andonie, C., Damian, P., Vasile, M., Elemente de
fortificaţie din epocile romano-bizantină şi bizantină descoperite la Nufăru, jud. Tulcea,
Buletinul Muzeului Militar Naţional, serie nouă, 1, 2, 67–76.
—— 2007: Damian, O., Vasile, M., Stănică, A., Bănăseanu, A., Samson, A., Cercetări arheo-
logice preventive la Nufăru, jud. Tulcea, MCA, serie nouă, 3, 107–152.
—— 2012: Damian, O., Vasile, M., Samson, A., Ene, D., Stănică, A., Cercetări arheologice
preventive la Nufăru, jud. Tulcea (II), MCA, serie nouă, 8, 177–207.
Damian, O., Andonie, C., Damian, P. 1994: Date despre incinta vestică a cetăţii bizantine de
la Proslaviţa-Nufăru, Istros, 7, 167–182.
Damian, O., Andonie, C., Vasile, M. 2000: Nufăru, in Cronica 1999, 67–68.
—— 2001: Nufăru, in Cronica 2000, 164–165.
bibliography 181
—— 2003a: Nufăru, in Cronica 2002, 213–216.

—— 2003c: Cetatea bizantină de la Nufăru. Despre problemele unui sit suprapus de o aşezare
contemporană, Peuce, SN, 1 (14), 237–266.
Damian, O., Bănăseanu, A. 2006: Ostrov, com. Ostrov, jud. Constanţa. Punct: Păcuiul lui
Soare, in Cronica 2005, 249–250.
Damian, O., Ene, D. 2011: Ostrov, com. Ostrov, jud. Constanţa. Punct: Păcuiul lui Soare, in
Cronica 2010, 94–95.
Damian, O., Mărgineanu-Cârstoiu, M. et alii 2011: Damian, O., Mărgineanu-Cârstoiu, M.,
Dumitraşcu, V., Vasile, M., Nufăru, jud. Tulcea. Punct: Trecere bac, in Cronica 2010,
90–91.
Damian, O., Olteanu, R. 2012: Ostrov, com. Ostrov, jud. Constanţa. Punct: Păcuiul lui Soare,
in Cronica 2011, 94–95.
Damian, O., Samson, A., Vasile, M. 2009: Complexul rupestru de la Murfatlar-Basarabi la
jumătate de secol de la descoperire. Consideraţii arheologice, MCA, serie nouă, 5, 117–
158.
Damian, O., Vasile, G. 2010: Vestigii arheologice descoperite pe dealul Dervent ( jud.
Constanţa), in Între stepă şi imperiu. Studii în onoarea lui Radu Harhoiu. Archäologis-
che Studien für Radu Harhoiu zum 65. Geburtstag. Volum îngrijit de Andrei Măgureanu,
Erwin Gáll, Bucureşti, 337–384.
—— et alii 2010: Damian, O., Vasile, M., Stănică, A., Samson, A., Ene, D.: Date noi despre
cetatea bizantină de la Nufăru, jud. Tulcea, Buletinul Muzeului Militar Naţional “Regele
Ferdinand I”, serie nouă, 7–8, 28–35.
—— 2011: Varangians nearby the Danube Delta. About an Archaeological Discovery in the
Byzantine Fortress at Nufăru, Tulcea County, SCIVA, 62, 3–4, 275–290.
Dančeva-Vassileva, A. 2004: Sredetz (Serdica) dans l’histoire politique de l’Empire Byzantin
(1018–1195), BHR, 32, 1–2, 13–40.
Dédéyan, G. 1975: L’immigration arménienne en Cappadoce au XIe siècle, B, 45, 1, 41–117.
Diaconu, P. 1962: Despre datarea valului de piatră din Dobrogea şi localizarea evenimentelor
din nota toparhului grec, Studii, 15, 5, 1215–1236.
—— 1965a: Autour de la localisation de la Petite Preslav, RESEE, 3, 1–2, 37–56.
—— 1965b: Din nou despre valul de piatră din Dobrogea şi nota toparhului grec (I–II), SCIV,
16, 1, 189–199; 2, 383–394.
—— 1966: Quelques problèmes relatifs à la forteresse byzantine de Păcuiul lui Soare, d’après
les dernières fouilles archéologiques, Dacia, NS, 10, 365–371.
—— 1968: “Alte precizări în legătură cu valul de piatră din Dobrogea şi însemnările topar-
hului bizantin”, SCIV, 19, 2, 357–369.
—— 1969a: Une information de Skylitzes-Cédrénos à la lumière de l’archéologie, RESEE, 7,
1, 43–49.
—— 1969b: Rolul cetăţii din insula Păcuiul lui Soare în cadrul situaţiei politice a Dobrogei la
sfârşitul secolului X, Pontica, 2, 395–400.
—— 1970: Les Petchénègues au Bas-Danube, Bucureşti.
—— 1972a: Câteva consideraţii în legătură cu valurile din Dobrogea, Pontica, 5, 373–380.
—— 1972b: Despre datarea “circumvalaţiei” şi a “bisericii treflate” de la Niculiţel, SCIV, 23,
2, 307–319.
—— 1973–1975: Date noi privind “Valul mare de pământ” din Dobrogea, Peuce, 4, 199–209.
—— 1975: Despre datarea nivelului “locuinţelor incendiate” de la Dinogetia-Garvăn ( jud.
Tulcea), SCIVA, 26, 3, 387–394.
—— 1976a: Păcuiul lui Soare-Vicina, Vyzantina, 8, 409–447 + Tableaux 47–58.
—— 1976b: Diaconu, Un tezaur de monede bizantine din sec. al XI-lea descoperit la Păcuiul
lui Soare, MN, 3, 235–239.
—— 1977a: Istoria Dobrogei în unele lucrări străine recente (II), RdI, 30, 10, 1893–1900.
—— 1977b: Câteva consideraţii pe marginea unor toponimice din dreapta Dunării de Jos
(zona Ostrov-Cochirleni), Pontica, 10, 59–63.
182 bibliography
—— 1978: Les Coumans au Bas-Danube aux XIe et XIIe siècles, Bucarest.

—— 1981a: La Dobroudja et Byzance à l’époque de la genèse du peuple roumain (VIIe–Xe
siècles), Pontica, 14, 217–220.
—— 1981b: Realităţi politice la Dunărea de Jos: români, bizantini, bulgari, pecenegi, RdI, 34,
6, 1111–1133.
—— 1984: Pătrunderea pecenegilor în Câmpia Română şi argumentul numismatic, SCIVA,
35, 1, 68–73.
—— 1986: Despre organizarea administrativ-militară a regiunii Dunării de Jos în vremea lui
Ioan Tzimiskes (Contribuţii pe marginea articolului: “Sigiliile lui Leon Sarakinopulos de la
Preslavul Mare”), SCIVA, 37, 2, 167–178.
—— 1987a: Câmpia Română şi Bizanţul în sec. X (note preliminarii), Istros, 5, 217–220.
—— 1987b: De nouveau à propos de Presthlavitza, SOF, 46, 279–293.
—— 1987c: Unde se afla oraşul Theodoroupolis?, AIIAI, 24, 1, 479–484.
—— 1988a: Sur l’histoire de la Dobroudja au Moyen Age, Dacia, NS, 32, 175–193.
—— 1988b: Sur les problèmes archéologiques de la cité de Iatrus, Dacia, NS, 32, 195–204.
—— 1989: Vers quelle date Léon Mélissénos pouvait-il être domestikos des Scholes occiden-
tales?, RESEE, 27, 1–2, 11–14.
—— 1992a: Contribution à la connaissance de l’histoire de la Dobroudja en base de quelques
sceaux, Dacia, NS, 36, 179–185.
—— 1992b: Note de lector (I). 4. Un sigiliu al lui Mihail, fiul lui Anastasie, conducătorul
Paristrionului?, Istros, 6, 326–327.
—— 1992c: Un alt sigiliu al lui Constantin Theodorokanos, Pontica, 25, 359–361.
—— 1994a: Sur la présence des Byzantins au Bas-Danube (IXe–XIVe siècles), RESEE, 32, 3–4,
367–373.
—— 1994b: Câteva consideraţii privind sigiliile strategului Ioan Malesis, Istros, 7, 355–356.
Diaconu, P., Anghelescu, N. 1968: Urme de locuire în colţul de sud-vest al Dobrogei, Revista
Muzeelor, 5, 4, 348–351.
Diaconu, P., Damian, O., Mărgineanu-Cârstoiu, M. 2004: Ostrov, com. Ostrov, jud. Constanţa.
Punct Păcuiul lui Soare, in Cronica 2003, 226–227.
—— 2005: Ostrov, com. Ostrov, jud. Constanţa. Punct: Păcuiul lui Soare, in Cronica 2004,
251–252.
Diaconu, P., Damian, O., Vasile, M. 2001: Ostrov, com. Ostrov, jud. Constanţa, in Cronica
2000, 170–172.
Diaconu, P., Vâlceanu, D. 1972: Păcuiul lui Soare, I. Cetatea bizantină, Bucureşti.
Dimitrov, D. I. 1985: La culture matérielle sur la rive septentrionale gauche du Bas-Danube
aux VIe–Xe siècles, EB, 21, 1, 114–132.
Dimitrov, H. 1986: Bulgaria and the Magyars at the Beginning of the 10th Century, EB, 22,
2, 61–77.
—— 1992a: Die frühmittelalterliche Stadt Develtos zwischen Byzanz und Bulgarien von
Achten bis ins Zehnte Jahrhundert, in Die Schwarzmeerküste in der Spätantike und im
frühes Mittelalter, herausgegeben von R. Pillinger, A. Pülz, H. Vetters (Schriften der
Balkan-Komission. Antiquarische Abteilung, 11), Wien, 35–45.
—— 1992b: Bulgaria against the Byzantine Empire for the Mastery of the Black Sea Coast in
the First Half of the Ninth Century, in Bulgaria Pontica Medii Aevi, III, Sofia, 43–49.
—— 1997: Die Awareneinfälle in die Provinz Scythia Minor (584–626 n.Chr.), in Von der
Scythia, 26–34.
Dimitrov, S. 1993: The Bulgarian Apocryphal Chronicle and Bulgarian Ethnic History, EB,
29, 4, 97–109.
Dimova, V. 1993: Rannosrednovekovnata krepost do selo Tsar Asen, Silistrenko (La forteresse
moyenâgeuse près du village de Tzar Assain, région de Silistra), Dobrudža, 10, 54–75.
Dochev, K. 2002: Turnovo, Sixth-Fourteenth Centuries, in The Economic History of Byzantium
from the Seventh through the Fifteenth Century, edited by A. E. Laiou, Washington DC,
vol. II, 673–678.
Dončeva-Petkova, L. 2003: Pliska i pečenezite (Pliska and the Pechenegs), pp. 9, 244–258.
bibliography 183
—— 2007: Zur ethnischen Zugehörigkeit einiger Nekropolen des 11. Jahrhunderts in Bulgarien,
in Post-Roman Towns, Trade and Settlement in Europe and Byzantium, vol. 2. Byzantium,
Pliska and the Balkans, edited by Joachim Henning (Millennium-Studien zu Kultur und
Geschichte des ersten Jahrtausends n. Chr., Volume 5/2), Berlin-New York, 643–660.
Döpmann, H. 1983: Wechselbeziehungen zwischen Otto I. und Bulgaren auf dem Hintergrund
der deutsch-byzantinischen Beziehungen, in J. Dummer, J. Irmscher (ed.), Byzanz in der
europäischen Staatenwelt. Eine Aufsatzsammlung (BBA, 49), Berlin, 47–51.
Drăganu, N. 1933: Românii în veacurile IX–XIV pe baza toponimiei şi a onomasticei,
Bucureşti.
Ducellier, A. 1968: L‘Arbanon et les Albanais au XIe siècle, TM, 3, 353–368.
Dzanev, G. 2007: Kolektivna nahodka na solidi ot Х vek v Аbritus (А Collective Find of Solids
from the 10th Century in Аbritus), in Izledvanija po bălgarska srednovekovna arheologija.
Sbornik v čest na prof. Rašo Rašev. Otgovoren redaktor P. Georgiev, Veliko Tărnovo, 375–
384.
Eggers, M. 1995: Das ‚Grossmährische Reich‘. Realität oder Fiktion? Eine Neuinterpretation
der Quellen zur Geschichte des mittleren Donauraumes im 9 Jahrhundert, Stuttgart.
Fehér, G. 1921: Bulgarisch-ungarische Beziehungen in dem V.-XI. Jahrhunderten, Budapest.
—— 1931: Les monuments de la culture protobulgare et leurs relations hongroises, Budapest
(Archaeologia Hungarica, 7).
Ferjančić, B. 1982: Vizantijski pečat iz Sirmijuma (A Byzantine Seal from Sirmium), ZRVI,
21, 47–52.
Ferluga, J. 1976: Byzantium on the Balkans. Studies on the Byzantine Administration and the
Southern Slavs from the VIIth to the XIIth Centuries, Amsterdam.
—— 1980: Die Chronik des Priesters von Diokleia als Quelle für die byzantinische Geschichte,
Vyzantina, 10, 429–460.
Fetisov, A. 2007: The “Rurikid Sign” from the B3 Church at Basarabi-Murfatlar, Apulum, 44,
299–314.
Fiedler, U. 1986: Zur Datierung der Langwälle an der mittleren und unteren Donau, Archäol-
ogisches Korrespondenzblatt, 16, 4, 457–465.
—— 2008: Bulgars in the Lower Danube region. A survey of the archaeological evidence and
of the state of current research, in The Other Europe in the Middle Ages. Avars, Bulgars,
Khazars and Cumans, ed. by F. Curta, Leiden, Boston (ECEEMA, 2), 151–236.
Fine, J. V. A. Jr. 1991: The Early Medieval Balkans. A Critical Survey from the Sixth to the late
Twelfth Century, Ann Arbor.
—— 1994: The Late Medieval Balkans: a Critical Survey from the Late Twelfth Century to the
Ottoman Conquest, Ann Arbor.
Florescu, R. 1967: Date noi de la Capidava. În legătură cu cultura materială a zonei Dunării
de Jos în perioada anterioară campaniilor lui Ioan Tzimisces, Apulum, 6, 259–268.
—— 1986: Limesul dunărean bizantin în vremea dinastiilor isauriană şi macedoneană, Pon-
tica, 19, 171–177.
Florescu, R., Ciobanu, R. Şt. 1972: Problema stăpânirii bizantine în nordul Dobrogei în sec.
IX–XI, Pontica, 5, 381–400.
Florescu, R., Covacef, Z. 1988–1989: Stratigrafia Capidavei romane târzii şi feudale timpurii,
Pontica, 21–22, 197–244.
Frances, E. 1959: Les relations russo-byzantines au XIIe siècle et la domination de Galicie au
Bas-Danube, Byzsl, 20, 1, 50–62.
Franklin, S., Shepard, J., 1996: The Emergence of Rus. 750–1200, London, New York.
Frankopan, P. 1996: A Victory of Gregory Pakourianos against the Pechenegs, Byzsl, 57, 2,
278–281.
—— 1997: The numismatic evidence for the Danube region, 971–1092, BMGS, 21, 30–39.
—— 2001: The workings of the byzantine provincial administration in the 10th–12th centuries:
the example of Preslav, B, 71, 1, 73–97.
—— 2005: Unravelling the Alexiad: Who was ‘Devgenevich’ of the Russian Primary Chronicle
and ‘Pseudo-Diogenes’ of the Greek sources?, BMGS, 29, 2, 147–165.
184 bibliography
—— 2011: 11th-century military seals from Bulgaria: some suggestions, BM, 2, (Studies in
honour of Professor Vassil Gjuzelev), 99–106.
Gagova, K. 1986: Bulgarian-Byzantine Border in Thrace from the 7th to the 10th Century (Bul-
garia to the South of the Haemus), BHR, 14, 1, 66–77.
Gândilă, A. 2007: Greek Imperial, Roman, Byzantine and Ottoman Coin Finds from Capidava
(1966–2006) in the Collection of the National History Museum of Romania, Pontica, 40,
607–623.
Georgiev, P. 1988: Encore sur l’origine du mot KOMENTON, EB, 24, 2, 87–92.
—— 2005: Golemijat zemlen val v Dobrudža—istoričeski beležki (La grande vallum a
Dobroudja—notes historiques), AMV, 3/1, 23–40.
—— 2010: Malkijat zemlen val v Dobrudja—săštnost i datirovka (The Little Dyke in Dobrud-
zha—Character and Date), in Velikotărnovskijat universitet “Sv. Sv. Kiril i Metodij” i
bălgarskata arheologija, I, Veliko Tărnovo, 413–422.
Giurescu, C. C. 1931: Despre Vlahia Asăneştilor, Lucrările Institutului de Geografie al
Universităţii din Cluj, 4, 109–124.
—— 1971: Localizarea Vicinei şi importanţa acestui oraş pentru spaţiul carpato-dunărean,
Peuce, 2, 257–260.
Gjuzelev, V. 1978: Die mittelalterliche Stadt Mesembria (Nesebăr) im 6.–15. Jh, BHR, 6, 1,
50–59.
—— 1981: Il Mar Nero e il suo litorale nella storia del medioevo bulgaro, BB, 7, 11–24.
—— 1984, Chan Asparuch und die Gründung des bulgarischen Reiches, Mitteilungen des
Bulgarischen Forschungsinstitutes in Österreich, 6, 2, 25–46.
—— 1986: Bulgarien und Byzanz im Streit um die Schwarzmeergebiete 1185–1204, JÖB, 36,
207–217.
—— 1991: Hauptstädte, Residenzen und Hofkultur im mittelalterlichen Bulgarien, 7.–14. Jh.
(Vom Nomadencampus bis zum Zarenhof ), EB, 27, 2, 82–105.
—— 1992: Anchialos zwischen der Spätantike und dem frühen Mittelalter, in Die
Schwarzmeerküste in der Spätantike und im frühes Mittelalter, herausgegeben von
R. Pillinger, A. Pülz, H. Vetters (Schriften der Balkan-Komission. Antiquarische Abtei-
lung, 11), Wien, 23–33.
—— 1996: Byzanz und Bulgarien, Rivalität und Koexistenz, in Byzanz und Nachbarn,
219–234.
Gordiyenko, D. 2012: The Byzantine-Bulgarian Confrontation in the first Half of the 10th Cen-
tury and Kyivan Rus’, Byzsl, 70, 1–2, 156–166.
Grămadă, N. 1924: Vicina. Izvoare cartografice. Originea numelui. Identificarea oraşului,
Codrul Cosminului. Buletinul Institutului de Istorie şi Limbă, Cernăuţi, 1, 437–459.
—— 1925–1926: Ozolimna, Codrul Cosminului, 2–3, 85–97.
—— 1930: La Scizia Minore nelle carte nautiche del medio evo. Contribuzione alla topografia
storica della Dobrogea, Ephemeris Dacoromana, 4, 212–256.
Grierson, Ph. 1973: Catalogue of the Byzantine Coins in the Dumbarton Oaks Collection and
in the Whittemore Collection, III/2. Basil I to Nikephorus III (867–1081), Washington DC.
Grousset, R. 1947: Histoire de l‘Arménie des origines à 1071, Paris.
Grünbart, M. 1998: Die Familie Apokapes im Lichte neuer Quellen, SBS, 5, 29–41.
Guilland, R. 1964: Byzance et les Balkans, sous le règne d’Isaac II Ange (1185–1195), in CIEB
XII, vol. 2, 125–137.
—— 1971: Patrices du règne de Constantin IX Monomaque, ZRVI, 13, 1–25.
Gyóni, M. 1943–1944: Zur Frage der rumänischen Staatsbildungen im XI. Jahrundert in Paris-
trion (Archaisierende Volksnamen und ethnische Wirklichkeit in der “Alexias” von Anna
Komnene), Archivum Europae Centro-Orientalis, Budapest, 9–10, 83–188.
—— 1951: Le nom de “Blachoi” dans l’Alexiade d’Anne Comnène, BZ, 44, 2, 241–252.
—— 1952: La première mention historique des Vlaques des Monts Balkans, AAnt, 1, 3–4,
495–515.
—— 1956: Les variantes d’une type de légende byzantine dans la littérature ancienne islan-
daise, AAnt, 4, 1956, 1–4, 293–313.
bibliography 185
Györffy, G. 1964: Zur Geschichte der Eroberung Ochrids durch Basileios II, in CIEB XII, vol. 2,
149–154.
Haldon, J. 1999: Warfare, State and Society in the Byzantine World, 565–1204, London.
—— 2001: The Byzantine wars: battles and campaigns of the Byzantine era, Stroud.
Hanak, W. K. 1995: The Infamous Svjatoslav: Master of Duplicity in War and Peace?, in Peace
and War, 138–151.
Harhoiu, R. 2000: Observaţii generale asupra evoluţiei cultural-istorice a Dobrogei în secolele
VII–X, in Istro-Pontica, 349–354.
Hălcescu, C. 1989: Din nou despre Onglos, SCIVA, 40, 4, 339–350.
Hendy, M. 1985: Studies in the Byzantine Monetary Economy, c. 300–1450, Cambridge.
Holmes, C. 2002: Byzantium’s Eastern Frontier in the Tenth and Eleventh Centuries, in
D. Abulafia, N. Berend (ed.), Medieval Frontiers: Concepts and Practices, Aldershot, 83–104.
—— 2005: Basil II and the Governance of Empire (976–1025), Oxford.
Horedt, K. 1969: Blökumanaland şi Blakumen, AM, 6, 179–185.
Howard-Johnston, J. D. 2000: Byzantium, Bulgaria and the Peoples of Ukraine in the 890s, in
Materials in Archaeology, History and Etnography of Tauria, Simferopol, 7, 342–356.
—— 2006: A short piece of narrative history: war and diplomacy in the Balkans, winter
921/2–spring 924, in Byzantine Style, Religion and Civilization. In Honour of Sir Steven
Runciman, Cambridge, 340–360.
Huxley, G. 1984: Steppe-Peoples in Konstantinos Porphyrogennetos, JÖB, 34, 77–89.
Iacob, M. 2000: Aspecte privind circulaţia monetară pe teritoriul României în a doua parte a
sec. VII p. Chr., in Istro-Pontica, 485–498.
Iliescu, O. 1978: À la recherche de Kilia byzantine, RESEE, 16, 2, 229–238.
—— 1994: Nouvelles contributions à la géographie historique de la Mer Noire, Il Mar Nero, 1,
229–259.
Iljovski, R. 1991: Vizantijsko-ungarsko soiuz vo početokot na XI vek protiv Samuil i negovite
naslednitzi. Obid za hronološko opredeluvanje (L’alliance byzantino-hongroise au début
du XIe siècle contre Samuel et ses successeurs. Essai de détermination chronologique),
ZRVI, 29–30, 75–99.
Ioniţă, A. 2005: Spaţiul dintre Carpaţii Meridionali şi Dunărea Inferioară în secolele XI–XIII,
Bucureşti.
Iorga, N. 1920: Les premières cristallisations d’État des Roumains, BSH, 5–8, 1, 33–46.
—— 1937: Histoire des Roumains et de la romanité orientale, Vol. III: Les fondateurs d’état,
Bucarest.
—— 1998: Chestiunea Dunării (istorie a Europei răsăritene în legătură cu această chestie),
ed. V. Spinei, Iaşi.
Iosipescu, S. 1994: Românii şi cea de-a treia cruciadă, RI, serie nouă, 5, 3–4, 249–272.
Ivanov, S. 2008: Novootkriti molibdobuli ot razkopkite na dvorčovnija Ploštad v Pliska (Newly
Found Lead Seals from the Excavations of the Palace Square in Pliska), in Proceedings
Haralanov, 137–146.
Ivanov, S. A., Lubotsky, A. 2010: An Alanic Marginal Note and the Exact Date of John II’s
Battle with the Pechenegs, BZ, 103, 2, 595–603.
Jordanov, I. 1981: Sceaux de deux notables byzantins de la fin du XIe s., EB, 17, 3, 92–97.
—— 1982a: Pečati na Leon Sarakinopul ot Velikij Preslav, Arh, 24, 1982, 1, 12–23.
—— 1982b: Établissement administratif byzantin à Preslav aux Xe–XIe s., JÖB, 32, 2, 35–44
(XVI Internationalen Byzantinistenkongress. Akten, II/2).
—— 1984: Molivdovulj na Damian Dobromir—duk na Trakija i Mesopotamija (Bleisiegel des
Damjan Dobromir—duk von Thrakien und Messopotamien), INMV, 20 (35), 99–105.
—— 1986: Molybdobulles nouvellement découverts de Basile Apokapes, EB, 22, 1, 123–127.
—— 1987a: La stratégie de Preslav aux Xe–XIe siècles selon les données de la sigillographie,
SBS, 1, 89–96.
—— 1987b: Dobrudža (491–1092)—selon les données de la numismatique et la sphragistique,
in Dobrudža. Études, 182–207.
—— 1990: Le molybdobulle du vestès Jean Aaron, Palaeobulgarica, 14, 1, 106–110.
186 bibliography
—— 1992a: Vizantijski olovni pečati ot Pliska, PP, 5, 281–301.

—— 1992b: Neizdadeni vizantijski olovni pečati ot Silistra (IV) Nichtpublizierte byzantinische
Bleisiegel aus Silistra (IV), INMV, 28 (43), 227–245.
—— 1992c: Sceau d’archonte de PATZINAKIA du XIe siècle, EB, 28, 2, 79–82.
—— 1993: Pečatite ot strategijata v Preslav (971–1088), Sofia.
—— 1995: The Byzantine administration in Dobrudja (10th–12th century) according to
sphragistic data, Dobrudža, 12, 204–223.
—— 2000: Monetite i pečatite ot Pliska (1899–1999), PP, 8, 135–167.
—— 2001: Korpus na pečatite na srednovekovna Bălgarija, Sofia.
—— 2002: Vizantijski pečati s imeto na Silistra (Dorostolon/Dristras), Dobrudža, 20,
80–90.
—— 2003a: Corpus of the Byzantine Seals from Bulgaria, vol. 1: Byzantine Seals with Geo-
graphical Names, Sofia.
—— 2003b: The Katepanikion of Paradunavon according to the sphragistic data, SBS, 8,
63–74.
—— 2005: Otnovo za pečatite na Damijan Dobromir—duk na Trakija i Mesopotamija (Once
again about stamps of Damyan Dobromir—a duke of Thrace and Mesopotamia), AMV,
3/1, 263–272.
—— 2006: Corpus of the Byzantine Seals from Bulgaria, vol. 2: Byzantine Seals with Family
Names, Sofia.
—— 2006–2007: Byzantine Seals containing Place Names from Dr. Vasil Haralanov Collec-
tion, CN, 12–13, 519–525.
—— 2008: Pečatite ot kolektsijata na dr. Vasil Haralanov (The Seals in the Collection of dr.
Vasil Haralanov), in Proceedings Haralanov, 35–53.
—— 2009: Corpus of Byzantine Seals from Bulgaria, Volume 3: part 1–2, Sofia.
—— 2010: Nish XI–XII C., According Byzantine Sigillography (SIC !), in Niš & Byzantium. VIII
Symposium, Niš 3–5 June 2009, ed. Miša Rakocija, Niš, 175–188.
—— 2011a: Srednovekovnijat Drăstăr spored dannite na sfragistikata (VІ–ХІІ v.), in Antičnaia
drevnost’ i srednie veka. Văpusk 40. Sbornik naučiuh trudov, Ekaterinburg, 79–120.
—— 2011b: І. Jordanov, Corpus of Byzantine Seals From Bulgaria. Volume 1–3, Sofia, 2003,
2006, 2009. Addenda et Corrigenda, Numizmatika, sfragistika i epigrafika, 7, Sofia, 189–228.
—— 2011c: Pečati na tema Bălgarija, Numizmatika, sfragistika i epigrafika, 7, Sofia, 157–187.
—— 2011d: Le monnayage du dernier défenseur de Sirmium en 1018 (Une énigme de l’histoire
et de la numismatique bulgare au Moyen Age), BHR, 39, 1–2, 135–139.
Jordanov, I., Tăpkova-Zaimova, V. 1988: Quelques nouvelles données sur l’administration
byzantine au Bas-Danube ( fin du Xe–XIe s.), in Géographie historique du monde mediter-
ranéen (sous la direction de H. Ahrweiler) (Byzantina Sorbonensia, 7), Paris, 119–126.
Jordanov, I., Žekova, Ž. 2007: Catalogue of Medieval Seals at the Regional Historical Museum
of Shumen, Shumen.
Kalić, J. 2003: Beograd u XII veku. Tărdžava—grad- polis (Belgrade au XIIe siècle. Forteresse—
ville- polis), ZRVI, 40, 91–96.
Karagiorgou, O. 2008: On the way to the throne: the career of Nikephoros III Botaneiates
before 1078, in Hypermachos. Studien zu Byzantinistik, Armenologie und Georgistik. Fest-
schrift für Werner Seibt zum 65. Geburtstag, hrsg. von Ch. Stavrakos, A. K. Wassiliou,
M. K. Krikorian,Wiesbaden, 105–132.
Kazhdan, A. P. 1977: Once More About the “Alleged” Russo-Byzantine Treaty (ca. 1047) and
the Pecheneg Crossing of the Danube, JÖB, 26, 65–77.
Kislinger, E. 2001: Regionalgeschichte als Quellenproblem. Die Chronik von Monembasia und
das sizilianische Demenna. Eine historisch-topographische Studie (DAW, 294), Wien.
Klein, W. 2005: St. Symeon Pentaglottos of Trier (d. 1035). A New Greek Saint of Mount Sinai in
the Context of Arabic Christian Studies, Parole de l’Orient. Revue semestrielle des études
syriaques et arabes chrétiennes, 30, 173–198.
Kostova, R. 2006: ‘Bypassing Anchialos’: The West Black Sea coast in naval campaigns, 11th
to 12th c. (I), in Tangra. Sbornik v čest na 70-godišnijata na akad. Vasil Gjuzelev, Sofia,
579–596.
bibliography 187
—— 2008a: Space and Power on the West Black Sea Coast late 10th to 12th C.: the Sigillo-
graphic Evidence, in Proceedings Haralanov, 207–223.
—— 2008b: The Lower Danube in the Byzantine Naval Campaigns in the 12th C., CCDJ, 24,
269–281.
Kosztolnyik, Z. 1987: From Coloman the Learned to Béla III (1095–1196). Hungarian Domestic
Policies and their Impact upon Foreign Affairs (East European Monographs, 220), Boulder.
—— 2002: Hungary under the Early Arpads, 890s to 1063 (East European Monographs, 605),
Boulder.
Kristó, G. 1996: Hungarian History in the 9th Century, Szeged.
Krsmanović, B. 2008: The Byzantine Province in Change (on the Threshold between the 10th
and the 11th Century), Belgrade, Athens (Serbian Academy of Sciences and Arts, The
National Hellenic Research Foundation).
—— 2009: Das Problem der sogenannten Zusammengesetzten Bezirke auf dem Balkan im 11.
Jahrhundert—Zwei Fallbeispiele, ZRVI, 46, 65–87.
—— 2010: Beobachtungen zum Taktikon Escorialense, BZ, 103, 2, 605–637.
Krumova, T. 2005: Pecheneg Chieftains in the Byzantine Administration in the Theme of
Paristrion in the Eleventh Century, Annual of Medieval Studies at Central European Uni-
versity, Budapest, 11, 207–221.
Kühn, H. J. 1991: Die Byzantinische Armee im 10. und 11. Jahrhundert. Studien zur Organisa-
tion der Tagmata, Wien.
Kuzev, A. 1966: Prinosi kăm istoriata na srednovekovnite kreposti po Dolnija Dunav I. Tutra-
kan i Ruse (Contribution à l‘histoire des forteresses médiévales le long du Bas-Danube, I.
Tutrakan et Roussé), INMV, 2 (17), 23–50.
—— 1967: Prinosi kăm istoriata na srednovekovnite kreposti po Dolnija Dunav II. Pirgos,
Novgrad, Svištov i Nikopol (Contribution à l’histoire des forteresses médiévales sur le Bas-
Danube, II. Pirgos, Novgrad, Svichtov et Nicopole), INMV, 3 (18), 41–70.
—— 1968: Prinosi kăm istoriata na srednovekovnite kreposti po Dolnija Dunav III. Gigen,
Orjakhovo, Lom, Vidin i Florentin (Contribution à l’histoire des forteresses médiévales sur
le Bas-Danube. III. Guiguène, Oriahovo, Lom, Vidine et Florentine), INMV, 4 (19), 27–55.
—— 1969: Prinosi kăm istoriata na srednovekovnite kreposti po Dolnija Dunav IV. Silistra i
Hărsovo (Contribution à l’histoire des forteresses médiévales sur le Bas-Danube, IV. Silistra
et Harsovo), INMV, 5 (20), 137–157.
—— 1972: Die Festungen an der unteren Donau im XII–XIV Jh., in Actes du IIe Congrès Inter-
national des Études du Sud-Est Européen (Athènes, 7–13 mai 1970), 2, Athènes, 417–423.
—— 1979: Strategičeskata rolja na krepostite po Dolnja Dunav prez IX–XII v. (Die strategis-
che Bedeutung der Festungen an der unteren Donau während des 9. bis 12. Jahrhunderts),
INMV, 15 (30), 25–41.
—— 1981: Mittelalterliche Städte an der Westküste des Schwarzen Meeres nördlich des Bal-
kangebirges, BB, 7, 271–275.
Kuzev, A., Gjuzelev, V. 1981: Bălgarski srednovekovni gradove i kreposti, vol. I. Gradove i
kreposti po Dunav i Černo More, Varna.
Laurent, V. 1945: La domination byzantine aux bouches du Danube sous Michel VIII Paléo-
logue, RHSEE, 22, 184–198.
—— 1957: Le théme byzantin de Serbie au XIe siècle, REB, 15, 185–195.
—— 1969: Deux nouveaux gouverneurs de la Bulgarie byzantine: le proèdre Nicéphore
Batatzès et le protoproèdre Grégoire, RESEE, 7, 1, 143–150.
Lazăr, G. 2005: Observaţii cu privire la societatea Ţaratului Vlaho-Bulgar al Asăneştilor în
lumina bibliografiei selective, MN, 17, 3–31.
—— 2006: Les rapports féudo-vassaliques et les alliances militaires dans l’Empire des Asé-
nides (1185–1241), MN, 18, 15–32.
—— 2010: Scurtă privire asupra strategiilor şi tacticilor militare ale fraţilor Asăneşti (1185–
1207), MN, 22, 5–20.
Lefort, J. 1976: Rhétorique et politique. Trois discours de Jean Mauropous en 1047, TM, 6,
265–303.
188 bibliography
Lemerle, P. 1977: Cinq études sur le XIe siècle byzantin, Paris.

Lewis, A. R. 1975: The Danube Route and Byzantium, 802–1185, in CIEB XIV, II, 359–368.
Ljubinković, R. 1973: L’Illyricum et la question romaine à la fin du Xe siècle et au début du XIe
siècle. Autour de l’Eglise autocéphale de l’État de Samuel, Italia Sacra, Padova, 22 (= La
Chiesa greca in Italia dall’VII al XVI secolo, III), 927–969.
Madgearu, A. 1997: The Downfall of the Lower Danubian Late Roman Frontier, RRH, 36, 3–4,
315–336.
—— 1998: Unele observaţii asupra istoriei themei Paradunavon (despre Vasile Apokapes),
Pontica, 31, 239–244.
—— 1999a: The Military Organization of Paradunavon, Byzsl, 60, 2, 421–446.
—— 1999b: Dunărea în epoca bizantină (secolele X–XII): o frontieră permeabilă, RI, serie
nouă, 10, 1–2, 41–55.
—— 1999–2000: The Restoration of the Byzantine Rule on the Danube, RESEE, 37–38, 1–2,
5–23.
—— 2001: The Church Organization at the Lower Danube, between 971 and 1020, EBP, IV,
Iaşi, 71–85.
—— 2003a: The Periphery against the Centre: the Case of Paradunavon, ZRVI, 40, 49–56.
—— 2003b: Despre “Noua Anglie” de la Marea Neagră (secolul al XI-lea), RI, serie nouă, 14,
5–6, 137–144.
—— 2005: Ruled indeed Basil Apokapes the Paradunavon Theme?, AMV, 3/1, 273–282.
—— 2006: Centrifugal movements in the Balkans in the 11th century, EBP, 5, 213–221.
—— 2007: The Avars and Dobrudja, in Problemi na prabălgarskata istorija i kultura, 4/2,
Sofia, 263–278.
—— 2008: The Wars of the Balkan Peninsula. Their Medieval Origins, Lanham.
Magdalino, P. 1993: The Empire of Manuel I Komnenos, 1143–1180, Cambridge.
—— 1997: The Byzantine Army and the Land: from Stratiotikon Ktema to Military Pronoia,
in Byzantium at War, 15–36.
Makk, F. 1989: The Árpáds and the Comneni. Political Relations between Hungary and Byz-
antium in the 12th Century, Budapest.
—— 1990: Les relations hungaro-byzantines aux Xe–XIIe siècles, in Études historiques hon-
groises 1990, publiées à l’occasion du XVIIe Congrès International des Sciences Historiques
par le Comité National des Historiens Hongrois, 4, Budapest, 11–25.
—— 1994a: Relations hungaro-bulgares au temps de prince Géza et du roi Étienne Ier, in
Hungaro-Bulgarica, V. Szegedi Bolgarisztika, Szeged, 25–33.
—— 1994b: Saint Ladislas et les Balkans, in Acta Universitatis Szegediensis de Attila Józ-
sef Nominatae. Opuscula Byzantina, 9 (Byzance et ses voisins. Mélanges à la mémoire de
Gyula Moravcsik à l’occasion du centième anniversaire de sa naissance), Szeged, 59–67.
—— 1999: Ungarische Aussenpolitik (896–1196), Herne.
—— 2002: Etelköz et la Mésopotamie de l’Occident, in K. Bibok, I. Ferincz, M. Kocsis (ed.),
Cirill és Metód példáját követve . . .: Tanulmányok H. Tóth Imre 70. születésnapjára, Szeged,
273–281.
Maksimović, L. 1996: Verwaltungsstrukturen in Byzanz und in den Balkanländern, in Byzanz
und Nachbarn, 47–63.
Maksimović, L., Popović, M. 1990: Les sceaux byzantins de la région danubienne en Serbie,
SBS, 2, 213–234.
—— 1993: Les sceaux byzantins de la région danubienne en Serbie. II. La collection du Musée
National de Belgrade, SBS, 3, 113–142.
Malamut, E. 1995: L‘image byzantine des Petchénègues, BZ, 88, 1, 105–147.
Malingoudis, Ph. 1980: Die Nachrichten des Nicetas Choniates über die Entstehung des
zweiten Bulgarischen Staates, Vyzantina, 10, 51–147.
Marinow, K. 2011: Hemos comme barriere militaire. L‘analyse des ecrits historiques de Leon
le Diacre et de Jean Skylitzes au sujet de la campagne de guerre des empereurs byzantins
Nicephore II Phocas en 967 et de Jean I Tzymisces en 971, BM, 2 (Studies in honour of
Professor Vassil Gjuzelev), 443–466.
bibliography 189
Markov, N. 1998: Novootkrit prăsten-pečat na vizantijski sanovnik (Anneau sigillaire nouvel-

lement découvert d‘un dignitaire byzantin), Arh, 39, 3–4, 63–66.
Marjanović-Vujović, G. 1974: Archeological Proving the Presence of the Pechenegs in Beograd
Town, Balcanoslavica, Belgrade, 3, 183–188.
Mănucu-Adameşteanu, G. 1980: Urme de locuire din perioada feudal-timpurie la Troesmis,
Peuce, 8, 230–247.
—— 1987: Noi descoperiri arheologice la Isaccea, SCIVA, 38, 3, 285–292.
—— 1991a: Tomis-Constantia-Constanţa, Pontica, 24, 299–327.
—— 1991b: Circulaţia monetară la Nufăru în secolele X–XIV, Peuce, 10, 1, 497–554.
—— 1992: Cronica descoperirilor monetare din nordul Dobrogei (IV), Pontica, 25, 399–417.
—— 1995a: Aspecte ale politicii împăratului Alexios I Comnenul la Dunărea de Jos în lumina
ultimelor descoperiri sfragistice şi numismatice, RI, serie nouă, 6, 3–4, 345–366.
—— 1995b: Raport final asupra locuirii medievale timpurii de la Aegyssus-Tulcea (secolele
X–XV), Peuce, 11, 363–371.
—— 1996: La diffusion de la monnaie byzantine en Dobroudja aux IXe–Xe siècles, RESEE,
34, 3–4, 275–287.
—— 1997: Un atelier monetar dobrogean din secolul al XI-lea, SCN, 12, 119–149.
—— 1998: Nufăru (jud. Tulcea)—oraş bizantin fortificat, la Dunărea de Jos (sec. X–XIII),
Arheologia Medievală, 2, 79–86.
—— 1998–2003: Monnaies byzantines de la collection du Musée du Bas Danube de Călăraşi,
découvertes à Păcuiul lui Soare (Xe–XIIIe siècles), BSNR, 92–97, nr. 146–151, 161–173.
—— 2001a: Istoria Dobrogei în perioada 969–1204. Contribuţii arheologice şi numismatice,
Bucureşti.
—— 2001b: Les invasions des Petchénègues au Bas Danube (1027–1048), EBP, 4, Iaşi, 87–112.
—— 2001c: Din nou despre atacul cumanilor din anul 1095 şi încetarea locuirii de la Păcuiul
lui Soare, in Simpozion de numismatică dedicat împlinirii a patru secole de la prima unire
a românilor sub Mihai Voievod Viteazul, Chişinău, 28–30 mai 2000, Bucureşti, 109–120.
—— 2001d: Monede de aur de la împăratul Vasile al II-lea (976–1025) descoperite pe terito-
riul Dobrogei, CCDJ, 18, 50–52.
—— 2003: Les invasions des Petchénègues au Bas Danube, AMV, 2, 299–311.
—— 2009: Monede bizantine din secolele X–XII, descoperite la Isaccea-Vicina, în campania
arheologică din anul 2003, Pontica, 42, 621–651.
—— 2010a: Monede bizantine descoperite în Dobrogea, vol. I. Monede bizantine descoperite
în nordul Dobrogei, secolele X–XIV. Beroe-Piatra Frecăţei, Troesmis-Turcoaia, Arrubium-
Măcin, Aegyssus-Tulcea, Salsovia-Mahmudia, Bucureşti.
—— 2010b: Contribuţii la cunoaşterea locuirii medio-bizantine de la Beroe-Piatra Frecăţei
(sec. X–XII), SCIVA, 61, 1–2, 89–115.
—— 2010c: Locuirea de epocă medio-bizantină pe teritoriul oraşului Tulcea (sec. X–XIII),
SCIVA, 61, 3–4, 219–250.
—— 2010d: Comuna Turcoaia, punct Igliţa, cetăţile Troesmis Est şi Troesmis Vest. Consideraţii
privind locuirea medio—bizantină din secolele X–XIII, Pontica, 43, 435–469.
—— 2010e: Contribuţii la cunoaşterea locuirii medio-bizantine de la Mahmudia—cetatea
Salsovia (secolele X–XI), Peuce, SN, 8 (21), 223–236.
—— 2010f: Contribuţii la cunoaşterea locuirii medio-bizantine de la Măcin—cetatea Arru-
bium (secolele X–XII), Peuce, SN, 8 (21), 237–250.
Mănucu-Adameşteanu, G., Jordanov, S. 2002: Monede bizantine din secolele X–XI desco-
perite pe teritoriul localităţii Garvăn, ţinutul Silistra şi invazia pecenegilor din 1046/1047,
in Simpozion de numismatică organizat în memoria martirilor căzuţi la Valea Albă la
împlinirea a 525 de ani (1476–2001), Chişinău, 13–15 mai 2001, Bucureşti, 125–138.
Mănucu-Adameşteanu, G., Poll, I. 1999: O monedă de aur de la Constantin X descoperită în
Dobrogea, Pontica, 32, 343–346.
—— 2006: Un tezaur de folles anonimi din clasa A2–A3, descoperit la Noviodunum—Vicina
(?)—Isaccea, Pontica, 39, 435–461.
190 bibliography
Mărculeţ, V. 2005: Stăpânirea bizantină la Dunărea de Jos în secolele X–XII. Aspecte din
istoria Themei Mesopotamia Apusului, a Strategatului de Dristra-Dorostolon şi a Themei
Paristrion-Paradunavon, Mediaş.
—— 2005–2006: Din nou despre organizarea teritoriilor bizantine de la Dunărea de Jos:
strategatul de Dristra-Dorostolon, Peuce, SN, 3–4 (16–17), 305–316.
—— 2006: Noi consideraţii asupra organizării, funcţionării şi rolului themei Mesopotamia
Apusului, 971–c. 1000, Pontica, 39, 295–319.
—— 2008: Asupra funcţionării unor guvernatori ai themei Paristrion-Paradunavon din seco-
lul al XI-lea, in Românii în Europa medievală (între Orientul bizantin şi Occidentul latin).
Studii în onoarea profesorului Victor Spinei, volum îngrijit de D. Ţeicu, I. Cândea, Brăila,
175–198.
—— 2009: Un problème controversé de l’histoire de la thème Paristrion-Paradounavon au
XIe siècle: le gouvernement de Basileios Apokapes, Istros, 15, 163–177.
—— 2010a: Un probable gouverneur du thème Paristrion-Paradounavon de la première par-
tie du XIe siècle, Orientalia Christiana Periodica, Roma, 76, 1, 201–208.
—— 2010b: Considerations Regarding the Area where the Byzantine Campaign against the
Cumans took place in 1114 and the localization of Blökumannaland, in Tracii şi vecinii lor în
Antichitate / The Thracians and Their Neighbours in Antiquity. Studia in Honorem Valerii
Sîrbu, editor I. Cândea, Brăila, 585–594.
—— 2011: “The archontate of the Paristrion cities” or “of the cities from the Istros”—a Thema
with a single name in the Byzantine literary sources, Istros, 17, 55–68.
McGeer, E., Nesbitt, J., Oikonomides, N. 2001: Catalogue of Byzantine seals at Dumbarton
Oaks and in the Fogg Museum of Art, vol. IV. The East, Washington DC.
—— 2005: Catalogue of Byzantine seals at Dumbarton Oaks and in the Fogg Museum of Art,
vol. V. The East (continued), Constantinople and Environs, Unknown Locations, Addenda,
Uncertain Readings, Washington DC.
McGrath, S. 1995: The Battles of Dorostolon (971): Rhetoric and Reality, in Peace and War,
152–164.
Medvedev, I. 2000: Historia enos plastografismatos: ho legomenos Toparcha Gothicus, Vyz-
antina, 21, 237–249.
Meško, M. 2006: Vývin obranného systému Byzantskej ríše v 11. storoči—príklad témy
Paradounavon (Entwicklung des Verteidigungssystems des Byzantinischen Reiches im 11.
Jahrhundert—Beispiel des Thema Paradounavon, Byzantinoslovaca, 1, Bratislava, 128–143.
—— 2007: Bitka pri Beroé 1122: posledný boj pečenehov (Die Schlacht bei Berrhoia im Jahre
1122: der letzte Kampf der Petschenegen), Vojenská História. Časopis pre Vojenskú Históriu
Múzejníctvo a archívnictvo, Bratislava, 11, 4, 3–26.
—— 2011: Notes sur la chronologie de la guerre des Byzantins contre les Petchénègues (1083–
1091), Byzsl, 69, 134–148.
Metcalf, D. M. 1976: Coinage and Coin Finds Associated with a Military Presence in the
Medieval Balkans, in V. Kondić (ed.), Kovanje i kovnice antičkog i srednjovekovnog novca
(Frappe et ateliers monetaires dans l’antiquité et moyen âge), Belgrade, 89–97.
—— 1979: Coinage in South-Eastern Europe (820–1396), London.
Michailova, T. 2003: Kăsnonomadski grobove v dvorčovnija tsentăr na Pliska (Late Nomadic
Graves in the Palace Center of Pliska), PP, 9, 259–266.
Milošević, G. 1991: Medieval Settlement in the Suburbium of Branicevo, Starinar, NS, 42,
187–195.
Mladjov, I. 1998: Trans-Danubian Bulgaria: Reality and Fiction, BS, New Series, 3, 85–128.
Mohov, A. S. 1999: K administrativnoi strukture vizantijskoi imperii na Dunae v period voin’i
s uzami (1064–1065 gg.) Zur Verwaltungsstruktur des byzantinischen Reiches an der Donau
zur Zeit des Krieges mit den Usen (1064–1065), in Antičnaia drevnost‘ i srednie veka. Văpusk
30. Sbornik naučiuh trudov, Ekaterinburg, 158–168.
Moravcsik, G. 1970: Byzantium and the Magyars, Budapesta.
Morrisson, C. 1970: Catalogue des monnaies byzantines de la Bibliothèque Nationale, II. De
Philippicus à Alexis III (717–1204), Paris.
bibliography 191
Mutafčiev, P. 1932: Bulgares et Roumains dans l’histoire des pays danubiens, Sofia.
Nadriş, G. 1960: A Spurious Slavonic Inscription from the Danube Canal, SEER, 38, 91,
530–534.
Napoli, J. 1997: Recherches sur les fortifications linéaires romaines, Roma.
Năsturel, P. Ş. 1965: Peut-on localiser la Petite Preslav à Păcuiul lui Soare? Commentaire à
Anne Comnène, Alexiade, VII. 3, RESEE, 3, 1–2, 17–36.
—— 1966: Note sur la géographie historique de la Dobroudja chez Constantin Porphyrogé-
nète, in Polychronion, 382–387.
—— 1969: Valaques, Coumans et byzantins sous le règne de Manuel Comnène, Vyzantina, 1,
167–186.
—— 1981: À propos de Tenou Ormon (Teleorman) de Kinnamos, in Geographica Byzantina,
sous la direction d’Hélène Ahrweiler (Byzantina Sorbonensia, 3), Paris, 81–91.
—— 1987: Mais où donc localiser Vicina?, BF, 12, 145–171.
—— 2003: Ce temei se poate pune pe mărturia lui Holobolos asupra stăpânirii Bizanţului
la gurile Dunării?, in Studia historica et theologica. Omagiu profesorului Emilian Popescu
(ed. C. C. Petolescu, T. Teoteoi. A. Gabor), Iaşi, 351–354.
—— 2004: Axiopolis sous les Comnènes. Une relecture de Kinnamos (III. 3), in Prinos, 521–
534.
Necşulescu, C. 1937: Ipoteza formaţiunilor politice române la Dunăre în sec. XI, RIR, 7, 1–2,
122–151.
—— 1939: Năvălirea uzilor prin Ţările Române în Imperiul Bizantin, RIR, 9, 185–206.
Nesbitt, J., Oikonomides, N. 1991: Catalogue of Byzantine seals at Dumbarton Oaks and in
the Fogg Museum of Art, vol. I. Italy, North of the Balkans, North of the Black Sea, Wash-
ington DC.
Nesbitt, J., Morrisson, C. 2009: Catalogue of Byzantine seals at Dumbarton Oaks and in the
Fogg Museum of Art, vol. VI. Emperors, Patriarchs of Constantinople, Addenda, Wash-
ington DC.
Nicolae, C. et alii 2008: Nicolae, C. Bănoiu, D., Vlad, N., Aspecte noi privind topografia cetăţii
de la Hârşova (jud. Constanţa), Pontica, 41, 313–343.
Nikolov, S. 1997: The Magyar Connection or Constantine and Methodius in the Steppes,
BMGS, 21, 79–92.
Nikolov, G. 2001: The Bulgarian aristocracy in the war against the Byzantine Empire (971–
1019), in Byzantium and East Central Europe, 141–158.
Noonan, Th. S. 1992: Byzantium and the Khazars: a special relationship?, in Byzantine Diplo-
macy. Papers from the Twenty-fourth Spring Symposium of Byzantine Studies, Cambridge,
March 1990, ed. by J. Shepard, S. Franklin, Aldershot, 109–132.
—— 1998–1999: The Khazar-Byzantine World of the Crimea in the Early Middle Ages: The
Religious Dimension, AEMA, 10, 207–230.
Oberländer-Târnoveanu, E. 1980a: Monede antice şi bizantine descoperite la Troesmis,
Peuce, 8, 248–280.
—— 1980b: Staţiuni antice pe raza comunei Mahmudia ( jud. Tulcea), Peuce, 8, 55–76.
—— 1983: Un atelier monétaire inconnu de la deuxième moitié du XIe siècle dans le thème de
Paristrion, RESEE, 21, 3, 261–270.
—— 1992: Numismatic and Historical Remarks on the Byzantine Coin Hoards from the 12th
Century at the Lower Danube, RESEE, 30, 1–2, 41–60.
—— 1995–1996: Byzantino-Tartarica. Le monnayage dans la zone des bouches du Danube à
la fin du XIIIe siècle et au commencement du XIVe siècle, Il Mar Nero. Annali di archeolo-
gia e storia, Roma-Paris, 2, 191–214.
—— 1996: Monnaies byzantines des VIIe–Xe siècles decouvertes à Silistra dans la collection
de l’académicien Pericle Papahagi, conservée au Cabinet des Médailles du Musée National
d’Histoire de Roumanie, CN, 7, 97–127.
—— 1998: Monede antice şi bizantine descoperite la Nufăru ( jud. Tulcea) păstrate în colecţia
Muzeului Militar Naţional, BSNR, 88–89, nr. 142–143, 71–106.
—— 2003: La monnaie dans l’espace rural byzantin des Balkans orientaux. Un essai de syn-
thèse au commencement du XXIe siècle, Peuce, SN, 1 (14), 341–412.
192 bibliography
—— 2005: Notes on the Beginnings of the Bulgarian Medieval Coinage, AMV, 3/1, 183–214.
Obolenski, D. 1979: The Crimea and the North before 1204, Archeion Pontou, 35, 123–133.
Oikonomides, N. 1965: Recherches sur l’histoire du Bas-Danube aux Xe–XIe siècles: la Méso-
potamie de l’Occident, RESEE, 3, 1–2, 57–79.
—— 1966: The donations of castles in the last quarter of the eleventh century, in Polychron-
ion, 413–417.
—— 1972: Les listes de préséance byzantines des IXe et Xe siècles, Paris.
—— 1973: Vardariotes—W.l.nd.r—V.n.nd.r, Hongrois installés dans la vallée du Vardar en
934, SOF, 32, 1–8.
—— 1976: L’évolution de l’organisation administrative de l’empire byzantin au XIe siècle
(1025–1118), TM, 6, 125–152.
—— 1983: Presthlavitza, the little Preslav, SOF, 42, 1–9.
—— 1986a: Le dédoublement de Saint Théodore et les villes d’Euchaïta et d’Euchaneia, Ana-
lecta Bollandiana, 104, 327–335.
—— 1986b: A Collection of Dated Byzantine Seals, Washington DC.
—— 1987: Des Valaques au service de Byzance? À propos de l’utilisation du mot komenton
aux Xe et XIe siècles, RESEE, 25, 2, 187–190.
—— 1998: À propos de la première occupation byzantine de la Bulgarie (971–c. 986), in Eupsy-
chia. Mélanges offerts à Hélène Ahrweiler (Byzantina Sorbonensia, 16), II, Paris, 581–589.
—— 2002: Problems of Chronology and the Seals of Preslav, SBS, 7, 1–9.
Opriş, I. 2004: Capidava. Sector VI, in Cronica 2003, 68–69.
Ostrogorsky, G. 1956: Histoire de l’État byzantin, Paris.
—— 1970: Die Pronoia unter den Komnenen, ZRVI, 12, 41–54.
Ovčarov, D. 1987: La forteresse protobulgare sur l’île danubienne Pacuiul lui Soare, in
Dobrudža. Études, 57–68.
Panait P. et alii 1995–1996: Panait P. I., Rădulescu, A., Ştefănescu, A., Flaut, D., Cercetările
arheologice de la cetatea Hârşova. Campania 1995, Pontica, 28–29, 121–134.
Panaitescu, A. 1978a: Contribuţii la cunoaşterea valului mare de pământ din Dobrogea, Pon-
tica, 11, 241–245.
Panaitescu, A. 1978b: O carieră de sec. X e.n. la Medgidia, Pontica, 11, 247–251.
Papuc, G. 1992: Despre valurile transdobrogene, Pontica, 25, 323–329.
—— 2000: Medgidia, in Cronica 1999, 58.
Parušev, V. 1993: Nepublikuvani srednovekovni moneti ot Iužna Dobrudža (VIII–XIV v.)
Unpublished medieval coins from South Dobroudja (8th–14th Centuries), Dobrudža, 10,
145–167.
Penčev, V. 1998: Kolektivna nahodka s vizantijski miliarentsii ot X–XI vek, namerena kraij
selo Gigen, Nikopolsko (A Hoard of Byzantine Miliarensia from X–XI Century, Found in
the Neighborhood of the Village of Gigen, the Nikopol Region), Numizmatika i sfragistika,
Sofia, 5, 1, 76–95.
Perkhavko, V. B. 1994: Gde je nakhodilsija dunaijskiji grad Perejaslavetz, Byzsl, 55, 2, 278–
290.
—— 1996: Russia’s Trade with Byzantium’s Danubian Provinces in the Eleventh-Twelfth Cen-
turies, în Acts. XVIIIth International Congress of Byzantine Studies. Selected papers. Main
and communications (Moscow, 1991), II, Shepherdstown, 13–18.
Petkov, K. 2008: The Voices of Medieval Bulgaria, Seventh–Fifteenth Century. The Records of
a Bygone Culture, Leiden, Boston (ECEEMA, 5).
Petre, A. 1973: Date noi în legătură cu valurile antice de apărare din Dobrogea, BMI, 42, 2,
27–31.
Pinter, Z. K., Ţiplic, I. M., Urduzia, C. 2008: Capidava. Sectorul VIII (extra muros), in Cronica
2007, 85–86.
Pinter, Z., Urduzia, C. 2009: Capidava. Sectorul VIII (extra muros), in Cronica 2008, 89–90.
Pintescu, F. 1999: Vlachs and Scandinavians in the Early Middle Ages, Oslo (Centre for
Viking and Medieval Studies).
bibliography 193
Pirivatrić, S. 1997: Vizantijske tema Morava i “Moravje” Konstantina VII Porfirogeneta (Le
Théme byzantin de Morava et la “Moravie” de Constantin VII Porphyrogénète), ZRVI, 36,
173–201.
Poenaru-Bordea, G., Ocheşeanu, R., Popeea, A. 2004: Monnaies byzantines du Musée de
Constanţa (Roumanie) (Collection Moneta, 42), Wetteren.
Popa, R. 1984: Knaufkrone eines wikingerzeitlichen Prachtschwertes von Păcuiul lui Soare,
Germania, 62, 2, 425–431.
Popescu, E. 1992: Quelques données sur les relations de l’Église de Dobroudja et Moldavie
avec l’Église russe aux XIIe–XIVe siècle, in A. E. Tachiaos (ed.), The Legacy of Saints Cyril
and Methodius to Kiev and Moscow. Proceedings of the International Congress on the Mil-
lenium of the Conversion of Rus’ to Christianity (Thessaloniki, 26–28 November 1988), Thes-
saloniki, 143–155.
—— 1994: Notes on the History of Dobroudja in the 11th Century: the Bishopric of Axiopolis, in
Idem, Christianitas Daco-Romana. Florilegium studiorum, Bucureşti, 421–438.
Popkonstantinov, K. 1987: Les inscriptions du monastère rupestre près du village Murfatlar
(Basarabi). État, théories et faits, in Dobrudža. Études, 115–145.
Popović, V. 1978: Episkopiska sedišta u Srbiji od IX do XI veko (Les évêchés médiévaux sur
le territoire de la Serbie (IXe–XIe siècles), GMGB: Godišnjak Muzeja Grada Beograda, 25,
33–40.
Popović, M. 1991: Les forteresses du système defensif byzantin en Serbie au XIe–XIIe siècle,
Starinar, NS, 42, 169–185.
Popović, M., Ivanišević, V. 1988: Grad Braničevo u srednjem veku (Braničevo, cité médiévale),
Starinar, NS, 39, 125–179.
Poppe, A. 1976: The Political Background to the Baptism of Rus’. Byzantine-Russian Relations
between 986–989, DOP, 30, 195–244.
Postică, G., Hâncu, I., Tentiuc, I. 1999: Aşezarea din secolele IX-XI de la Giurgiuleşti şi unele
consideraţii privind siturile medievale timpurii din zona lacurilor dunărene, in Studia in
honorem Ion Niculiţă. Omagiu cu prilejul împlinirii a 60 de ani, Chişinău, 280–298.
Prinzing, G. 1995: Zu Odessos/Varna (im 6. Jh.), Belgrad (1096) und Braničevo (um 1163).
Klärung dreier Fragen aus Epigraphik, Prosopographie und Sphragistik, Byzsl, 56, 1, 219–
225.
—— 1999–2000: Demetrios-Kirche und Aseniden. Zur chronologischen Präzisierung der
Frühphase des Aseniden-Aufstandes, ZRVI, 38, 257–265.
Pritsak, O. 1975: The Pečenegs. A Case of Social and Economic Transformation, AEMA, 1,
211–235.
—— 1982: The Polovcians and Rus’, AEMA, 2, 321–380.
Prokić, B. 1906: Die Zusätze in der Handschrift des Johannes Skylitzes: codex Vindobonensis
hist. graec. LXXIV. Ein Beitrag zur Geschichte des sogenannten Westbulgarischen Reiches,
München.
Purici, Şt. 1997: Problema hotarelor meridionale ale statului halician, in Spaţiul nord-est car-
patic în mileniul întunecat, coordonator V. Spinei, Iaşi, 189–202.
Rada, M. et alii 1988: M. Rada, N. Cochină, G. Corcodel, M. Iuga, Aşezări antice dobrogene
identificate cu ajutorul fotografiilor aeriene, Thraco-Dacica, 9, 197–205.
Rance, Ph. 2007–2008: The date of the military compendium of Syrianus Magister (Formerly
the Sixth-Century Anonymus Byzantinus), BZ, 100, 2, 701–737.
Rašev, R. 1982: L’Onglos—témoignages écrits et faits archéologiques, BHR, 10, 1, 68– 79.
—— 1987: Les vallums de Dobrudza dans le developpement de la fortification ancienne bul-
gare, in Dobrudža. Études, 48–56.
—— 1997: La plus ancienne periode de l’état bulgare, in Von der Scythia, 49–54.
—— 2004: Quelques remarques sur l’Onglos, in Prinos, 277–282.
—— 2005: Remarks on the Archaeological Evidence of Forts and Fortified Settlements in
Tenth-Century Bulgaria, in Borders, 51–58.
Rădvan, L. 2010: At Europe’s Borders: Medieval Towns in the Romanian Principalities, Leiden,
Boston (ECEEMA, 7).
194 bibliography
Révész, É. 2009: Petăr Deljan, ein bulgarisch-ungarischer Dux, in Bălgarskijat ezik i litera-
tura na krăstopătja na kulturite. Mezhdunarodna konfererentzija (Szeged, 21–21 mai 2007),
Szeged, 82–90.
Risos, A. 1990: The Vlachs of Larissa in the 10th Century, Byzsl, 51, 2, 202–207.
Rostkowski, G. 2001: Hungary between Byzantium, Central and Eastern Europe (ca. 1118–1135),
in Byzantium and East Central Europe, 160–177.
Runciman, S. 1949: The first crusader’s journey across the Balkan peninsula, B, 19, 207–221.
Russev, N. D., Fokeev, M. M. 2001–2002: Vizantijskiji solid iz rayona Dunayskoy delt
(A Byzantine Solid Found in the Danube Delta), Stratum plus. Cultural Anthropology and
Archaeology, Chişinău, 6 (Omens of Civilizations), 183–185.
Sacerdoţeanu, A. 1939–1940: Mouvements politiques et sociaux de la Péninsule Balkanique
dans la seconde moitié du XIe siècle, Balcania, 2–3, 83–106.
Salamon, M. 1971: Some Notes on an Inscription from Medieval Silistra (c. 976), RESEE, 9,
3, 494–496.
Šandrovskaja, V. S. 1982: Die Bedeutung der Bleisiegel für das Studium einiger Aspekte
der byzantinischen Geschichte, JÖB, 32, 2 (XVI Internationalen Byzantinistenkongress.
Akten, II/2), 165–173.
Savvides, A. G. C. 1986–1987: The Byzantine Family of Kekaumenos (Cecaumenos) (Late
10th–early 12th Century), Díptycha. Hetaireias Vyzantinon kai Metavyzantinon Meleton,
4, 12–27.
—— 1991–1992: On the Armenian-Georgian-Byzantine family of Apocapes (Abukab) in the
11th Century, Díptycha. Hetaireias Vyzantinon kai Metavyzantinon Meleton, 5, 96–104.
Sâmpetru, M. 1974: La région du Bas-Danube au Xe siècle de notre ére, Dacia, NS, 18, 239–
264.
Schlumberger, G. 1925: L’épopée byzantine à la fin du Xe siècle, Paris.
Schmitt, O. 2006: Die Petschenegen auf dem Balkan von 1046 bis 1072, in Pontos Euxeinos.
Beiträge zur Archäologie und Geschichte des antiken Schwarzmeer- und Balkanraumes
(Schriften des Zentrums für Archäologie und Kulturgeschichte des Schwarzmeerraumes,
10), Langenweissbach, 473–490.
Schönert-Geiss, E. 1991: Die Fundmünzen, in Iatrus-Krivina. Spätantike Befestigung und
frühmittelalterliche Siedlung an der unteren Donau, IV, Berlin, 213–240.
Seibt, W. 1978: Die byzantinischen Bleisiegel in Österreich. I. Kaiserhof (DAW, 324—Veröffen-
tlichungen der Kommission für Byzantinistik, II/1), Wien.
—— 1995: Sigillographische Beiträge zur bulgarischen Geschichte, Dobrudža, 12, 224–232.
—— 2003: Probleme der historischen Geographie Bulgariens im späteren 10. und 11.
Jahrhundert—ein sigillographischer Beitrag, AMV, 2, 253–263.
—— 2008: “Mesopotamia des Westens”—ist est im 9. und 10. Jahrhundert südlich der Marica
suchen?, AMV, 7/2, 100–113.
Ševčenko, I. 1971: The Date and Author of the So-called Fragments of Toparcha Gothicus,
DOP, 25, 115–188.
Shepard, J. 1974: Another New England? Anglo-Saxon Settlement on the Black Sea, BS, 1, 1,
18–39.
—— 1975: John Mauropous, Leo Tornicius and an Alleged Russian Army: The Chronology of
the Pecheneg Crisis of 1048–1049, JÖB, 24, 61–89.
—— 1979: Tzetzes’ letters to Leo at Dristra, BF, 6, 191–239.
—— 1984: Why did the Russians attack Byzantium in 1043?, BNJ, 23, 147–212.
—— 1985: Information, disinformation and delay in Byzantine diplomacy, BF, 10, 233–293.
—— 1995: Slavs and Bulgars, in The New Cambridge Medieval History, vol. II, c.700–c.900,
edited by R. McKitterick, Cambridge, 228–248.
—— 1999: Byzantium and the Steppe Nomads: the Hungarian Dimension, in Byzanz und
Ostmitteleuropa 950–1453. Beiträge zu einer Table-Ronde des XIX International Congress of
Byzantine Studies, Copenhagen 1996, herausgegeben von G. Prinzing, M. Salamon (Main-
zer Veröffentlichungen zur Byzantinistik, 3), Wiesbaden, 55–83.
—— 2000: Messages, ordres et ambassades: diplomatie centrale et frontalière à Byzance
(IXème–XIème siècles), in Voyages et voyageurs à Byzance et en Occident du VIème au XIème
bibliography 195
siècle. Actes du colloque international organisé par la Section d‘Histoire de l‘Université

Libre de Bruxelles, Genève, 375–396.
—— 2002: Emperors and Expansionism: From Rome to Middle Byzantium, in Medieval Fron-
tiers: Concepts and Practices, ed. by D. Abulafia, N. Berend, Aldershot, 55–82.
Sokolova, I. 1993: Les sceaux byzantines de Cherson, SBS, 3, 99–111.
Sophoulis, P. 2011: Byzantium and Bulgaria, 775–831, Leiden, Boston (ECEEMA, 16).
Soustal, P. 1997: Dorostolon-Silistra. Die Donaustadt im Lichte neuerer Forschung, in Von
der Scythia, 115–126.
Spinei, V. 2006: The Great Migrations in the East and South East of Europe from the Ninth
to the Thirteenth Century. I. Hungarians, Pechenegs and Uzes; II. Cumans and Mongols,
Amsterdam.
—— 2009: The Romanians and the Turkic Nomads North of the Danube Delta from the Tenth
to the Mid-Thirteenth Century, Leiden, Boston (ECEEMA, 6).
Squatriti, P. 2005: Moving Earth and Making Difference: Dikes and Frontiers in Early Medi-
eval Bulgaria, in Borders, 59–90.
Stanev, K. 2012: Svedenie v starobălgarski apokrif za Vetren i razpoloženija tam vizantijskii
oboz prez 1087 g., Mediaevalia. Sofijskij universitet “Sv. Kliment Okhridski”, 2, 5, 21–25.
Stavridou-Zafraka, A. 2000: The development of the theme organisation in Macedonia, in
Byzantine Macedonia. Identity, Image and History. Papers from the Melbourne Confer-
ence, July 1995, ed. by J. Burke and R. Scott (Byzantina Australiensia, 13), Melbourne,
128–138.
Stănescu, E. 1965: Les “Mixobarbares” du Bas-Danube au XIe siècle (Quelques problèmes de
la terminologie des textes), Nouvelles Études d’Histoire, 3, 45–53.
—— 1966: La crise du Bas-Danube byzantin au cours de la séconde moitié du XIe siècle,
ZRVI, 9, 49–73.
—— 1968a: Byzantinovlachica. I. Les Vlaques à la fin du Xe siècle-début du XIe siècle et
la restauration de la domination byzantine dans la Péninsule Balkanique, RESEE, 6, 3,
407–438.
—— 1968b: Beiträge zur Paristrion-Frage. Die Bennenungen der unteren Donaugebieten,
JÖB, 17, 41–64.
—— 1970: Beiträge zur Paristrion-Frage. Betrachtungen über das Stadtwesen des Gebietes
zwischen Donau und Meer in den X.–XII. Jahrh., East European Quarterly, 4, 2, 119–140.
—— 1971: Les “Blachoi” de Kinnamos et Choniatès et la présence militaire byzantine au Nord
du Danube sous les Comnènes, RESEE, 9, 3, 585–593.
—— 1974: Byzance et les Pays Roumains aux IXe–XVe siècles, in CIEB XIV, vol. I, 393–431.
—— 1989: Premisele răscoalei Asăneştilor. Lumea românească sud-dunăreană în veacurile
X–XII, in Răscoala şi statul Asăneştilor, Bucureşti, 11–36.
Stănică, A. 2004: Câteva aspecte privind aşezarea medio-bizantină de la Ostrov-Piatra
Frecăţei (Beroe), jud. Tulcea, in Prinos, 357–373.
Stepanenko, V. P. 1994: K cursus honorum Vasila, syva Apuhabu, in Rossijskoe vizantinove-
denia: Itogi i perspektivi, Moscow, 27–34.
—— 2009: Fers Tzotzikiji, antipat patrikiji i strateg Kappadokii (1000–1022) (Pherses
Tzotzikios, Anthypatos, Patrikios and Strategos of Cappadocia), in Numismatica, sphrag-
istica and epigraphica, 5 (Studia in honorem professoris Ivan Jordanov), Sofia, 235–245.
Stephenson, P. 2000: Byzantium’s Balkan Frontier: A Political Study of the Northern Balkans,
900–1204, Cambridge.
—— 2003: The Balkan Frontier in the Year 1000, in Byzantium in the Year 1000, edited by
P. Magdalino (The Medieval Mediterranean: Peoples, Economies and Cultures, 400–
1500, vol. 45), Leiden, 109–133.
Stoimenov, D. 1996: Za molibdobulite na Katakalon, “Strateg na Ioanopol”, Arh, 38, 2–3, 84.
—— 2008: Vizantijiski molibdobuli ot muzeja na Makedonija—Skopie (Molybdobulls from
the Macedonian Museum—Scopje), AMV 7/2, 177–185.
Stojkovski, B. 2009: Niš u vizantijsko-ugarskim odnosima u XI i XII veku (Niš in Byzantine-
Hungarian Relations in XI and XII century), in Niš & Byzantium. VII Symposium, Niš 3–5
June 2008, ed. M. Rakocija, Niš, 383–394.
196 bibliography
Stokes, A. D. 1962: The Background and Chronology of the Balkan Campaigns of Svyato-
slav Igorevich; The Balkan Campaigns of Svyatoslav Igorevich, SEER, 40, 92, 44–57; 95,
466–496.
Strässle, P. M. 2006: Krieg und Kriegführung in Byzanz: die Kriege Kaiser Basileios II. gegen
die Bulgaren (976–1019), Köln.
Şesan, M. 1978: Les thémes byzantins à l‘époque des Comnènes et des Anges (1081–1204),
RESEE, 16, 1, 45–55.
Ştefan, G., Barnea, I. 1967: Dinogetia I. Aşezarea feudală timpurie de la Bisericuţa-Garvăn,
Bucureşti.
Šandrovskaja, V. S. 1994: K istoriji fem’i Vaspurakan, Vizantijskij Vremennik, 55 (80), 1,
152–161.
Škorpil, K. 1918: Anciens monuments bulgares dans la Dobroudja, in La Dobroudja. Géogra-
phie, histoire, ethnographie, importance économique et politique, Sofia, 109–152.
Tanaşoca, N. Ş. 1973: Les Mixobarbares et les formations politiques paristriennes du XIe siècle,
RRH, 12, 1, 61–82.
—— 1981: De la Vlachie des Assénides au second Empire bulgare, RESEE, 19, 3, 581–594.
—— 1989: O problemă controversată de istorie balcanică: participarea românilor la restau-
rarea ţaratului bulgar, in Răscoala şi statul Asăneştilor, Bucureşti, 153–181.
Tăpkova-Zaimova, V. 1970: Quelques observations sur la domination byzantine aux bouches
du Danube—Le sort de Lykostomion et de quelques autres villes côtières, SB, 1, 79–86.
—— 1973: L’administration byzantine au Bas-Danube ( fin du Xe–XIe s.). Tentative d’une mise
au point, EB, 9, 3, 90–112.
—— 1974: Quelques particularités dans l’organisation militaire des régions de Bas-Danube
et la politique byzantine aux XIe–XIIe siècles, in Études de civilisation médiévale (IXe–XIIe
siècles). Mélanges offerts à E-R. Labande, Poitiers, 667–674.
—— 1975a: Les frontières occidentales des territoires conquis par Tzimiscès, SB, 10, 113–118.
—— 1975b: Les Mixobarbaroi et la situation politique et éthnique au Bas-Danube pendant la
seconde moitié du XIe s., in CIEB XIV, II, 615–619.
—— 1980: La population du Bas-Danube et le pouvoir byzantin (XIe–XIIe s.), in CIEB XV,
vol. IV, 331–339.
—— 1993: L’administration byzantine au Bas-Danube ( fin du Xe–XIe siècle), Byzsl, 54, 1,
95–101.
Tăpkova-Zaimova, V., Stoimenov, D. 2004: Otnovo za Dolni Dunav (kraj na X–XI v.). Isto-
riografski i izvorovădski problemi, in Černo More meždu iztoka i zapada. Reka Dunav most
meždu narodi i kulturi. IX Pontijski četenja, Varna, 16–17 mai 2003, Varna, 341–348.
Teodor, D. 1987: Quelques aspects concernant les relations entre Roumains, Byzantins et Bul-
gares aux IXe–Xe siècles n.e., AIIAI, 24, 2, 1–16.
—— 2011: Nouvelles considérations sur la domination de la Bulgarie au Nord du Danube
Inférieur, EBP, VI, Bucarest, 95–102.
Theodorescu, R. 1981: Marginalia to 11th Century. Anglo-Saxons in the Pontic Area, RRH, 20,
4, 637–646.
Todorov, B. 2010: The value of empire: tenth-century Bulgaria between Magyars, Pechenegs
and Byzantium, Journal of Medieval History, 36, 4, 312–326.
Todorova, E. 1984: Gli insediamenti genovesi alle foci del Danubio: Vicina, Chilia, Licostomo,
in Genova e la Bulgaria nel Medioevo, Genova, 427–459.
—— 1986: Au sujet de la géographie historique du Delta danubien, EB, 22, 1, 57–67.
Topoleanu, F. et alii 2005: Topoleanu, F., Haynes, J., Diaconescu, A., Bogdan, D., Mahmudia,
com. Mahmudia, jud. Tulcea [Salsovia]. Punct: La Cetate, in Cronica 2004, 215–216.
Tomaschek, W. 1886: Zur Kunde der Hämus-Halbinsel, II. Die Handelswege im 12. Jahrhun-
dert nach den Erkundigungen des Arabers Idrisi, Sitzungsberichte der philosophisch-
historische Classe der kaiserliche Akademie der Wissenschaften, Wien, 113, 285–373.
Tougher, S. 1997: The reign of Leo VI (886–912). Politics and people, Leiden, New York,
Köln.
Treadgold, W. 1988: The Byzantine Revival, 780–842, Stanford.
bibliography 197
Toynbee, A. 1973: Constantine Porphyrogenitus and his World, London.

Treadgold, W. 1988–1989: Three Byzantine Provinces and the First Byzantine Contacts with
the Rus’, Harvard Ukrainian Studies, 12–13, 132–144.
—— 1995: Byzantium and its Army. 284–1081, Stanford.
—— 1997: A History of the Byzantine State and Society, Stanford.
Tsankova-Petkova, G. 1975: À propos de l’histoire de la forteresse bulgare Moundraga, Ural-
Altaische Jahrbücher. Internationale Zeitschrift für Nord-Eurasien, 47, 35–40.
—— 1978: La libération de la Bulgarie de la domination byzantine, BB, 5, 95–121.
—— 1980: Les forces centrifuges et centripètes à Byzance du début du régne d’Isaak Ange
(L’insurection des Asénides et la révolte d’Alexis Branas), in CIEB XV, vol. IV, 55–64.
—— 1981: Au sujet de la campagne d’Isaac Ange contre la capitale bulgare (1190) BB, 7,
181–185.
Urbansky, A. B. 1968: Byzantium and the Danube frontier. A Study of the Relations between
Byzantium, Hungary and the Balkans during the period of the Comneni, New York.
Valeriev, I. 2009: The Treasure of Golden Byzantine Coins from Professor Ishirkovo, Silistra
Region, Pontica, 42, 653–661.
—— 2010: Unpublished Seal of Nikephoros Botaneiates as a Proedros and Doux of Antioch,
Pontica, 43, 425–433.
Vásáry, I. 2005: Cumans and Tatars: Oriental Military in the Pre-Ottoman Balkans, 1183–1365,
Cambridge.
Vasiliu, I., Mănucu-Adameşteanu, Gh. 1984: Consideraţii finale asupra locuirii feudal-
timpurie (sec. X–XI) de la Aegyssus-Tulcea (Campaniile 1959–1980), Peuce, 9, 143–155,
541–563.
Vâlceanu, D. 1967: Debarcaderul şi problema poziţiei geografice a cetăţii bizantine de la
Păcuiul lui Soare, SCIV, 18, 4, 593–615.
Vîlcu, A. 2008: Monnaies byzantines d’or du tresor decouvert à Dinogetia en 1939, CCDJ, 26,
87–96.
Vryonis, S. 1957: The Will of a Provincial Magnate, Eustathius Boilas (1059), DOP, 11, 263–277.
Wasilewski, T. 1964: Le thème byzantin de Sirmium-Serbie au XIe et XIIe siècles, ZRVI, 8, 2,
465–482.
—— 1975: Le katepanikion et le duché de Paristrion au XIe siècle, in CIEB XIV, vol. II, 641–
645.
—— 1995: L’administration byzantine dans la vallée du Bas-Danube au Xe et XIe siècles selon
la sigillographie, Dobrudža, 12, 190–203.
Wassiliou-Seibt, A. K. 2008: Gelesenes und Geschriebenes. Zur Interpretation byzantinis-
cher Bleisiegel aus Bulgarien. Der Familienname Pegonites, in Proceedings Haralanov,
133–136.
—— 2009: Der Familienname Pegonites auf byzantinischen Siegeln und in anderen schriftli-
chen Quellen, in Realia Byzantina, herausgegeben von S. Kotzabassi und G. Mavromatis
(Byzantinisches Archiv, Band 22), Berlin, New York, 303–319.
Whittow, M. 1996: The Making of Byzantium, 600–1025, Berkeley.
Wolff, R. L. 1949: The “Second Bulgarian Empire”. Its Origin and History to 1204, Speculum,
24, 2, 167–206.
Wozniak, F. E. 1975: Byzantine Policy on the Black Sea or Russian Steppe in the late 830s’,
BSEB, 2, 1, 56–62.
—— 1984: Byzantium, the Pechenegs and the Rus’: The Limitations of a Great Power’s Influ-
ence on its Clients in the 10th Century Eurasian Steppe, AEMA, 4, 299–316.
Yotov, V. 1995: Monnaies et sceaux de la forteresse “Skala” du Haut Moyen Âge, la Dobroudja
du Sud (Rapport préalable), Dobrudža, 12, 180–185.
—— 2003: The Vikings on the Balkans, Varna.
—— 2007: The Vikings in the Balkans (10th to 11th Centuries). Strategic and Tactical Changes.
New Archaeological Data on Weaponry, Archaeologia Baltica, Klaipeda, 8, 321–327.
—— 2008a: Le thème byzantin, la Thrace, Iôannoupolis et la question des terres bulgares du
nord-est à la fin du Xe—premières années du XIe siècles, Pontica, 41, 345–353.
198 bibliography
—— 2008b: Nov pečat na Aleksiji I Komnin (New Seal of Alexios I Comnenos), AMV, 7/2,
172–176.
—— 2008c: The Dristra Battle (1087), CCDJ, 24, 257–267.
Yotov, V., Atanasov, G. 1998: Skala. Krepost ot X–XI vek do s. Kladenci, Tervelsko, Sofia.
Yuzbashian, K. N. 1973–1974: L’administration byzantine en Arménie aux Xe–XIe siècles,
Revue des Études Arméniennes, NS, 10, 139–183.
Zacos, G. 1984: Byzantine Lead Seals, compiled and edited by J. W. Nesbitt, Berna.
Zacos, G., Veglery, A. 1972: Byzantine Lead Seals, vol. I, Basel.
Žekova, Ž. 2004: Za Preslav sled văstanieto na Asenevci (Preslav after the Asenid Uprising),
Preslav. Sbornik, 6, Sofia, 344–349.
—— 2005: Evoluţia circulaţiei monetare în Šumenul medieval şi regiunea învecinată în perio-
ada 971–1204, CN, 9–11, 167–179.
Zlatarski, V. N. 1929: Edna datirana pripiska na greki ot srjednata na XI. vek, Byzsl, 1,
22–34.
Zuckerman, C. 1995: On the Date of the Khazar’s Conversion to Judaism and the Chronology
of the Kings of the Rus Oleg and Igor. A Study of the Anonymous Khazar Letter from the
Genizah of Cairo, REB, 53, 237–270.
—— 1997a: Les Hongrois au pays de Lebedia: Une nouvelle puissance aux confins de Byzance
et de la Khazarie ca. 836–889, in Byzantium at War, 51–74.
—— 1997b: Two notes on the early history of the thema of Cherson, BMGS, 21, 210–222.
—— 2006: Byzantium’s Pontic Policy in the “Notitiae episcopatuum”, in C. Zuckerman (ed.),
La Crimée entre Byzance et le Khaganat khazar, Paris, 201–230.
Index OF People
Aaron (Bulgarian chief ), 46 Basil Apokapes (duke), 69–76, 90, 113, 129,
Aaron (duke of Mesopotamia), 70, 71 130
Adralestos Diogenes (strategos), 43 Basil I (emperor), 18
Adralestos Spanopoulos (ek prosopou), 60 Basil II (emperor), 26, 36, 40, 45, 46, 51,
Adrian Comnenos (general), 84, 138 53–55, 64, 95, 102, 106, 108, 113, 115, 116,
Adrian Dalassenos (duke of Antioch), 71 121, 123, 143, 150, 168, 170
Adrian (turmarch), 59 Basil Monachos (duke), 66, 126
Aetios (strategos), 90 Basil (ek prosopou of Paradunavon), 76
Ahtum (duke), 53, 54 Basilakes (strategos, Dristra), 61, 86
Albanians, 144 Basilakes (duke, Dyrrachion), 90
Alexander Kabasilas (duke), 99 Béla II (Hungarian king), 149, 154
Alexios I Comnenos (emperor), 80, 81, Béla III (Hungarian king), 158, 159, 165, 171
83–85, 90, 97, 99, 106–108, 110, 112, 121, Berladniks, 153
137–139, 141, 142, 144, 147, 148, 159, 170, 171 Blacs, 162
Alexios II Comnenos (emperor), 158 Bohemond of Taranto (prince), 81, 144,
Alexios (kleisurarchos), 45 148, 149
al-Hakim (Caliph), 51 Bordoni, 131
Almos (prince), 149 Boris I (Michael) (Bulgarian king), 16, 18
Alusian, 95 Boris II (Bulgarian emperor), 30, 32, 37,
Anastasios (strategos of Anchialos), 45 46, 47
Andrew I (king of Hungary), 129 Boris or Kalamanos (son of the Hungarian
Andronikos Comnenos (emperor), 100, king Coloman), 151
155, 158 Budilă (Pudilos) (Vlach chief), 143
Andronikos Dukas (strategos, Preslav), 89 Bulgars, 7–14, 16, 19, 23, 27, 33, 37, 111, 112,
Andronikos Kontostephanos (general), 157 161
Andronikos Philokales (duke of Bulgaria), Bulgarians, 9, 18–24, 26–41, 46–49, 51,
69, 76, 84 54–56, 59, 68–70, 73, 80–82, 85, 92, 95,
Andronikos (turmarch in Arachilava), 88 96, 100, 111, 112, 114, 115, 140, 144, 159,
Andronikos (ek prosopou, Mesembria), 83 161–167
Andronikos (kommerkiarios in Dristra), 93 Bulghari, 47
Anna (sister of Basil II), 45
Arabs, 12, 20, 23, 34, 40, 47, 51, 62, 169 Carpi, 124
Areimanioi, 140 Celestin III (pope), 148, 166
Arkadios (strategos of Dorostolon), 61, 62, Chalis (second name of Tatós), 80
86 Chalisi, 80
Armenians, 10, 48, 71–73, 75 Clement (vestarches), 90
Arpad (Hungarian duke), 20, 21 Coloman (Hungarian king), 148, 151
Asan (Vlach ruler), 140, 144, 159–162 Comitopouloi, 46
Asoteos (strategos, Varna), 89 Constantine (. . .) polites (katepano), 65,
Asparukh (Bulgar ruler), 7–11 86
Aurelian (emperor), 170 Constantine IV (emperor), 7, 10, 11, 37
Avars, 7, 10, 13, 139 Constantine V (emperor), 11, 12, 20, 112
Constantine VII Porphyrogenitus
Bagrat Vakhac’i (Armenian prince), 71 (emperor), 28, 93
Balakaye, 80 Constantine VIII (emperor), 113, 1116, 121,
Bardas Petzes (strategos), 88 123
Bardas Phokas (general), 30, 31, 45, 104 Constantine IX (emperor), 79, 104, 110, 121,
Bardas Skleros (general), 30, 31, 46 123–127, 168
200 index of people
Constantine X Dukas (emperor), 66, 67, 72, Géza I (Hungarian king), 96, 97
75, 76, 81, 89, 109, 110 Géza II (Hungarian king), 154, 155, 157
Constantine Anemas (inspector of the Goths, 124
Armeniakon theme), 90 Greeks, 23, 26, 36, 51, 53, 54, 72, 80, 139,
Constantine Arianites (duke), 126 163
Constantine Bodin (Serbian ruler), 96 Gregory Mavrokatakalon (katepano), 83,
Constantine Diogenes (general), 56 84, 87, 90, 109, 132, 139
Constantine Humbertopoulos Gregory Pakourianos (general), 137
(commander), 137
Constantine Karantinos (strategos, Helen or Jelena (daughter of Uroš I), 149,
Preslav), 89 154
Constantine Kokkinobepheos (krites), 90 Heraclius (emperor), 10, 11
Constantine Pterotos (strategos), 88 Himerios Solomon (ek prosopou of
Constantine Theodorokanos (katepano), Mesembria), 83
78, 79 Hungarians, 4, 21, 36, 39, 48, 53–55, 66,
Cordyles (chief of the Byzantine 95–99, 113, 137, 141, 148–151, 153, 154, 155,
prisoners), 14 157–159, 166, 168, 170–172
Cumans, 3, 4, 77, 80, 84, 85, 105–107, 123, Huns, 139
126, 132, 134, 137–139, 142–145, 150–153,
155, 158, 160–164, 166, 168–172 Iaroslav Osmomysl (prince), 153
Czechs, 149 Igor (prince), 22, 23, 26, 28
Illyrians, 80
Damian Dobromir (duke), 40–42, 51, 86 Ioannes Monasteriotes (vestarches), 74
David (Bulgarian chief ), 46 Isaac I Comnenos (emperor), 65, 66, 68,
David Arianites (duke of Bulgaria), 55 69, 75, 77, 95, 113, 129
David (strategos of Thracia and Dristra), Isaac II Angelos (emperor), 90, 148, 159,
61, 86 160, 163–165
David Kouropalates (Armenian prince), 48 Ivan Alexander (Bulgarian emperor), 69
Demetrios Katakalon (katepano), 67, 68, Ivan Rostislavich (prince), 153
87, 90
Demetrios Tornikes (logothetes), 148 Jews, 29
Demetrius (župan), 26, 27, 49 John I Tzimiskes (emperor), 30, 32–41,
Dionysios (kommerkiarios, Dristra), 93 43, 46, 51, 60, 62, 85, 101, 104, 106, 107,
Dristros (Bulgarian chief ), 38 109–112, 138, 167
John II Comnenos (emperor), 107, 145,
Englishmen, 140 147–150, 161
Eustathios Boilas (aristocrat), 70, 74 John Vladislav (Bulgarian emperor), 55,
Eustratios Romanos (kommerkiarios, 70, 95
Presthlavitza), 91 John Bogas (strategos), 21
Franks, 11, 80 John Chaldos (strategos), 18
Frederick I Barbarossa (German emperor), John Comnenos (kouropalates), 71
150, 155, 165 John Comnenos (duke of Skopion), 99
John Dermokaites (strategos), 88
Gabriel Radomir (Bulgaria emperor), 48, John Dukas (general), 135, 157
54, 55, 95 John Dukitzes (katepano), 71
Genoese, 134, 135 John Maleas (strategos), 91
George Dekanos (kouropalates), 140 John Malesis (strategos), 91
George Euphorbenos (commander of the John Monasteriotes (duke), 71
fleet), 85, 138 John Orphanotrophous (minister), 122
George Maniakes (general), 115 John Spondyles (kommerkiarios, Dristra),
George Spanopoulos (vestarches), 90 93
George Vojtech (Bulgarian chief ), 96 John (katepano of Mesembria), 83
George (strategos of the Vlachs), 140 John (kommerkiarios, Presthlavitza), 91
Géza (Hungarian duke), 48 John (turmarch in Paradunavon), 88
index of people 201
John (strategos, Preslav), 89 Michael Apokapes (Aboukab) (Abukab)

John (katepano of Ras), 43 (Armenian aristocrat), 72, 73
Justinian II (emperor), 11 Michael Branas (duke), 100
Michael Dokeianos (general), 77
Kalokyres Sarakinopoulos (kleisurarchos), Michael Gabras (duke), 157
85 Michael Iasites (vestarches), 74
Kalokyros (strategos), 29, 30 Michael Saronites (vestarches), 74
Kaloyannes (Vlach emperor), 162, 166 Michael Strabotrichares (strategos), 88
Katakalon Kekaumenos (katepano), 64, Mihailo Vojslav (prince of Dioclea), 96
65, 86, 116, 121, 122 Mongols, 134, 135, 152
Katakalon Tarchaneiotes (general), 142 Moses (Bulgarian chief), 46
Kegen (Pecheneg chief ), 65, 83, 122–124, Mysians, 81
126, 127, 169
Khazars, 10, 14–16, 22, 23, 29, 45 Nestor (duke), 78, 80–82, 87
Koulinos (Pecheneg warrior), 122 Nichita (Niketas) (duke), 98, 99
Krakras (Bulgarian chief ), 55 Nicholas Branas (general), 137
Krum (Bulgarian king), 13, 20, 112 Nicholas Mavrokatakalon (general), 138,
Kubrat (Bulgar chief ), 10, 11 170
Kusan (Hungarian duke), 20 Nicota (Niketas) (duke), 95, 96
Kyriakos (dignitary), 7 Nikephoritzes (minister), 79, 81, 82
Kyriakos (katepano of Mesembria), 83 Nikephoros I (emperor), 13
Nikephoros II Phokas (emperor), 28–30,
Ladislas I (Hungarian king), 97, 98, 137, 167
141, 149 Nikephoros III Botaneiates (emperor), 66,
Leo V (emperor), 13 69, 70, 75–77, 82, 83, 87, 109, 129, 131
Leo VI (emperor), 20, 138 Nikephoros Basilakios, 83, 137, 144
Leo Batatzes (general), 157, 158 Nikephoros Batatzes (duke), 96
Leo Chalkotubes (strategos), 60, 88 Nikephoros Bryennios (general), 78
Leo Charsianites (bishop), 147 Nikephoros Chaluphes (duke), 157
Leo Drymis (katepano), 63, 65, 86 Nikephoros Dekanos (duke), 99
Leo Melissenos (general), 48 Nikephoros Diogenes (Son of Roman IV
Leo Nikerites, 84, 87, 98, 139, 140 Diogenes), 142
Leo Pegonites (strategos), 90 Nikephoros Lalakon (strategos), 100
Leo Phokas (general), 21 Nikephoros Melissenos (general), 140
Leo Sarakinopoulos (strategos), 59–62, 86 Nikephoros Ouranos (general), 51
Leo (drongarios), 31 Nikephoros Xiphias (strategos), 51, 60
Leo (turmarches), 41, 42 Niketas Hagiozacharites (commander), 104
Ljudevit (Serbian ruler), 56 Niketas Karykes (Karikes) (duke), 98, 110
Nikola (Bulgarian aristocrat), 46
Magyars, 14–16, 19–22, 28–30, 35, 111 Nikolaos (strategos), 45
Manuel I Comnenos (emperor), 106, 145, Nikulitzas (strategos), 77
147, 150, 154, 158, 170, 171 Nogai (Mongol khan), 134, 152
Margaret (Mary) (Hungarian princess), 159 Normans, 97, 125, 154, 158, 159
Maria (Byzantine princess), 157
Marianos (strategos), 61, 62, 86 Oghuzi, 69, 72, 75, 89, 94
Melias (strategos), 91 Oleg (prince), 22
Michael III (emperor), 15, 16 Omalis (strategos), 91
Michael IV (emperor), 73, 95, 107, 109, 113, Omurtag (Bulgarian king), 13, 111
121, 122, 168 Otto I (emperor), 46
Michael VII Dukas (emperor), 7, 80, 82, 96, Ottomans, 107, 134
125, 121, 122, 141
Michael VIII Paleologos (emperor), 152 Pankratios (kleisourarchos), 45
Michael (archbishop of Russia), 145 Pankratios (duke), 71
Michael (son of Anastasios), 64, 77, 86 Partiqayê, 66
202 index of people
Paulicians, 137 Seljuks, 73, 75, 96, 126, 139, 150, 169
Pechenegs, 3, 4, 15, 20–24, 26, 28, 30, 33, 35, Selté (Pecheneg chief), 129
36, 40, 46, 54, 55, 62–68, 76, 77, 79, 80, Serbs, 56, 80–82, 96, 149, 150, 154, 163, 171
82, 83, 85, 88, 89, 92, 95, 96, 99, 101, 104, Sergios (kommerkiarios), 91
106–109, 113, 115–118, 121–127, 129, 131, 132, Sermon (strategos), 56
134, 137–145, 148, 152, 160, 162, 168–171 Sesthlav (chief), 132
Persians, 139 Simeon (katepano), 17
Peter I (emperor of Bulgaria), 28, 96 Sisinios (katepano), 38, 39, 60, 62, 86
Peter Orseolo (king of Hungary), 95 Slavs, 7, 15, 56
Peter Delian (Bulgarian chief ), 73, 95, 96, Staurakios (strategos), 59, 86
159 Stephen I (king of Hungary), 53–55, 95
Peter Phokas (officer), 31 Stephen II (king of Hungary), 149
Peter the Hermit (Crusader chief ), 99 Stephen III (king of Hungary), 157, 168
Peter (strategos of Ioannoupolis and Stephen Kontostephanos (general), 47, 48
Dristra), 59, 61, 86 Stephen Nemanja (Serbian ruler), 158, 163,
Peter (Theodore) or Kalopeter (ruler of the 165, 166
Vlachs), 140, 144, 159–163, 165 Svyatoslav (prince), 27, 29–36, 40, 45, 92,
Petronas (general), 15 93, 101, 143, 167
Pharasmanes (strategos), 73 Symeon (Bulgarian emperor), 20, 21, 27, 28,
Philotheus Frangopoulos, 89 30, 38, 111
Phokas (strategos), 12 Symeon (katepano), 77, 78, 83, 87, 90
Piroska (Irene) (princess), 148, 149 Synadene (princess), 96
Robert Guiscard (duke), 97 Tatós (Tatrys) (Pecheneg chief), 79, 80, 82,
Roger II (king of Sicily), 150, 154 131, 132, 137, 138
Romanians (Vlachs), 35, 36, 48, 79, 80, 132, Tervel (Bulgar ruler), 11
136, 143, 158, 161 Theodora (empress), 16
Romanos I Lekapenos (emperor), 21–23 Theodore Batatzes (general), 154
Romanos II (emperor), 28, 93 Theodore Pegonites (katepano), 68, 87, 90
Romanos III (emperor), 78, 89, 118, 121, 123 Theodore (katepano of Mesembria), 83
Romanos IV Diogenes (katepano and Theodore (strategos), 61, 86
emperor), 66, 67, 76, 77, 86, 90, 102, 108, Theodorokanos (duke), 51
112, 132, 142 Theodoulos (monk), 70, 75
Romanos (Bulgarian prince), 47 Theophanes (strategos), 59, 86
Romans (Roman Empire, Roman period), Theophilos (emperor), 22, 23, 43
8, 9, 11, 14, 15, 23, 26, 37, 97, 98, 101, 102, Thomas (archon), 17
104–109, 112–115, 124, 121, 138, 160, 162, 163 Tihomir (Serbian ruler), 154
Rumiantsev, N. (Russian statesman), 50 Trajan (emperor), 9
Rus’, 15–17, 19, 22–24, 27–37, 40, 42, 43, 45, Traulos (chief of the Paulicians), 137
46, 50, 51, 64, 85, 92, 93, 101, 110, 115, 116, Tugorkan (Cuman chief), 142
153, 155, 167, 168 Tupai, 27
Tyrach (Pecheneg chief), 65, 89, 112, 122,
Saint Constantine (Cyril), 15 126, 127
Saint George, 27 Tzachas (emir), 139
Saint Methodius, 15 Tzelgu (Pecheneg chief), 137, 138
Saint Simon of Trier, 100 Tzotzikios (strategos), 55, 61, 86, 1116
Saint Theodore Stratilates, 38
Salomon (king of Hungary), 96, 137, 138, Uroš I (Serbian ruler), 149, 154
170 Uroš II (Serbian ruler), 154
Samuel (Bulgarian emperor), 40, 41, 46–49, Uzes, 72, 127, 129, 130, 144
51, 53, 54, 59, 64, 110, 167
Satza (Pecheneg chief ), 132, 134 Vajk (Hungarian duke), 54
Sauromatai, 66 Valatzertes (Valtzar) (katepano), 17, 83
Scythians, 131, 132 Varangians, 27, 45, 103, 104
index of people 203
Vasilko (prince), 153 Walter the Penniless (Crusader chief), 99

Vlachs (Romanians), 41, 48, 55, 80, 81, 85, William II (king of Sicily), 158
100, 140, 143, 158–164
Vladimir I (prince), 45, 50 Yaroslav (prince), 115
Vladimir II Monomakh (prince of Kiev), Yuri Dolgoruki (prince), 153
132, 134, 144
Vladimir (prince of Novgorod), 115
Vladimir (prince of Suzdal’), 153
iNDEX OF Geographical names
Abydos, 45 Beroe (Stara Zagora), 97, 145, 162

Acarnania, 18 Beroe (Piatra Frecăţei), 106
Acheloos, 21 Bessarabia, 8, 9, 14, 50, 157
Adrianople, 13, 37, 39, 42, 45, 51, 64, 77, 126 Bestcastro, 113
Adriatic Sea, 55, 150 Bidenu, 113
Aegean Sea, 139 Black Sea, 1, 11, 13, 16, 18, 19, 21–23, 28, 41,
Aegyssus, 104 45, 85, 101, 115, 122, 136, 140, 141, 144, 167,
Agathonike, 13, 42 171
Agathopolis, 13, 22, 24 Blaquie, 162
Agighiol, 10 Blökumannaland, 162
Akhtopol, 163 Bohemia, 29
Akli, 94 Bolyarovo, 80
Amastris, 19 Borcea, 114, 123, 152
Anatolia, 20 Borduşani, 114
Anchialos (Pomorie), 11–13, 16, 20, 21, 45, Bosnia, 98
85, 86, 115, 143, 168, 170 Bradvari, 71, 131
Antioch, 28, 34, 39, 66, 80, 89 Brăila, 135, 137
Appiaria, 88 Braničevo, 3, 54, 56, 57, 99, 100, 113,
Arachilava, 88 147–150, 155, 156, 165, 166, 171
Argamum, 94 Bucharest (Museum), 112
Argeş, 33 Budeşti, 14
Arkadiopolis, 30 Bug, 14, 15, 20
Armenia, 71–73, 75 Bugeac (southern Bessarabia), 157
Armeniakon, 90, 125 Bugeac (village, Constanţa County), 123
Arrubium, 106 Bulgaria (state), 1, 2, 5, 7, 9, 11–16, 18–24,
Asia Minor, 125, 139 26–32, 36, 37, 41–43, 45–47, 49, 51,
Asprokastron, 50 53–56, 59, 63, 69, 71, 74, 83, 89, 90, 92,
Atelkuz, 20, 21, 39 93, 118, 141, 142, 159, 161, 163–166
Avhat, 38 Bulgaria (Byzantine theme), 55, 56, 63, 66,
Axiopolis, 108, 111, 142, 148, 152, 168 68, 69, 74, 76, 77, 82, 84, 95, 96, 98–100,
Azov Sea, 23 116, 126, 129, 144, 147
Bulgarophygon, 21
Balaklava, 50 Burgas, 11, 126
Balčik, 117, 118 Burjan, 47
Balkans (Balkan Peninsula), 4, 17, 41, 43, Buzău (County), 14
47, 55, 63, 64, 80, 96, 99, 102, 115, 141, 147, Byzantine Empire, Byzantines, passim
148, 153, 158, 163, 165, 167, 171
Balta Ialomiţei, 123, 139 Călăraşi, 60, 114
Banat, 33, 40, 53 Callatis, 23
Bărăgan, 153 Câmpineanca, 14
Barisklafa, 94, 134 Canlia, 123
Bârlad, 153 Capidava, 1–4, 20, 24, 50, 107, 108, 114, 116,
Basarabi (Murfatlar), 28 117, 127
Bela vezha, 132 Cappadocia, 55, 70, 75
Belene, 132 Carasu, 41
Belgrade, 3, 56, 95–100, 147–149, 154–156, Carpathian Mountains, 14, 137, 157, 158
159, 165, 166, 171, 172 Carsium (Hârşova), 7, 107
Berhoia (Veria), 40 Căscioarele, 5
index of geographical names 205
Caucasus, 23 Dnieper, 8, 14, 20, 22, 23, 37

Celei, 114 Dniester, 8, 153
Central Europe, 29, 147 Dobrudja, 1, 2, 8–10, 14, 19, 23, 24, 26, 27,
Cerna (river), 113 38, 41–43, 48–51, 57, 82, 92, 94, 107, 109,
Cernavoda, 24, 25, 108, 152 110, 116, 118, 125, 131, 132, 135, 137, 141, 142,
Cetatea Albă (Bilhorod Dnistrovs’kyi), 50 145, 150, 151, 158, 164, 165, 169
Cetatea Oii, 125 Domapia, 141
Cetatea Pătulului, 152 Don, 14, 15, 45, 141
Chaldia, 18, 19, 39 Dorostolon, 20, 29, 31, 33–35, 38, 39, 59–62,
Cherson, 15, 16, 19, 21, 23, 29, 45, 122 86, 111
Chilia (Kilia), 94, 135, 136 Dragosloveni, 14
Chirnogi, 13 Drandar, 138
Chrysopolis, 45 Drăstăr, 38
Cimmerian Bosphorus, 45, 141 Drava, 11
Cochirleni, 9, 14, 20, 35, 110, 112, 114 Dristra (Silistra), 34, 37–39, 41, 43, 49, 51,
Constanţa (city), 9, 14,15, 20, 23–25, 28, 32, 53, 55, 57–69, 74, 79–82, 84, 86–88,
33, 43, 101 90–95, 102, 110–118, 121–125, 131, 132, 134,
Constanţa (county), 124 137, 138, 142, 147–149, 153, 167–169
Constanteia (Constantia) (Constanţa), 23, Dubravica, 43, 100
24, 31–34, 42, 165 Dumbarton Oaks (Collection), 90
Constantia (Simeonovgrad), 13, 42 Durostorum (Silistra), 7, 10, 12, 13, 19, 20,
Constantinople (Istanbul), 10, 15, 19, 22, 30, 61
34, 37, 42, 62, 64, 67, 76, 77, 81, 82, 100, Dyrrachion, 43, 55, 74, 83, 90, 158
114–116, 122, 123, 126, 139, 141–143, 151, 154,
159, 163, 165, 167, 170, 171 Edessa, 66, 68, 72, 73, 75
Corfu, 150 Emesa, 59
Crimea, 11, 15–17, 19, 21, 23, 45, 50, 51, 141 Emona, 11
Croatia, 97, 98, 148, 158 Epirus, 18, 19, 144, 149, 170
Cumania, 153 Erkesija, 13
Curcani, 14 Euchaita, 38, 39, 65
Cyprus, 76 Euphrates, 42
Europe (part of the Byzantine Empire), 51
Dacia Mediterranea, 170
Dacia Ripensis, 170 Frankish Empire, 15, 16
Dacia, 170 Frankochorion, 97, 149
Dalmatia, 54, 148, 149, 157–159
Danube Delta, 10, 11, 17, 40, 94, 116, 136, Galaţi, 10, 11, 136
149, 165 Galiţa, 123
Danube, passim Garvăn (Dinogetia), 1–4, 49, 77, 78, 90, 92,
Daristar (Dristra), 94 94, 101, 105, 106, 113, 114, 116, 117, 124, 127,
Demnitzikos, 152 130–132, 137, 145, 147, 152, 153, 71
Dervent, 91, 101, 109, 110, 113, 117, 118, 123, Georgia, 71
131, 142 German empire, 29, 148, 150
Desina, 134 Gigen, 113, 117, 118, 123
Develtos, 11, 13, 16, 21, 42, 85, 122 Giurgengik, 132, 138
Dibiskos, 57 Giurgiu, 114
Dimitritsi, 159 Golyam Izvor, 138
Dineia, 31 Gradiščeto, 113, 117, 118
Dinogetia (Garvăn), 1, 2, 48, 92, 105, 106, Grădiştea, 114
152 Greaca, 14
Dioclea, 81, 96 Great Earthen Dike, 9, 14, 15, 19, 24, 35
Disina, 94, 134 Greci, 49
Ditzina (Dičina) (river), 24, 134 Greece, 29, 51
Djadovo, 127 Gugeşti, 14
206 index of geographical names
Haemus, 159, 161 Kuvin, 56

Halmyris, 94 Kypsella, 159
Halych, 153–155, 157
Haram (Nova Palanka), 149 Lardeas, 163
Hârşova, 4, 7, 24, 92, 107, 125, 129, 131,168, Larissa, 48, 65, 77
171 Lebounion, 139, 140, 148
Hebros (Maritsa), 13, 42 Letea, 18
Hekaton Bounoi, 126, 127, 139 Levedia, 14, 20
Helioupolis, 59 Linokastro, 125
Hermitage (museum), 74, 90 Little Preslav, 38, 51, 91, 93, 94, 134
Hinog, 108 Londina, 141
Hungary, 4, 29, 40, 53–55, 57, 68, 77, Lovitzos (Loveč), 129, 163
95–100, 113, 116, 129, 135, 141, 147–151, Lule Burgas, 30
153–155, 157–159, 165, 166 Lunca, 10
Lykostomion, 17–21, 24, 50
Iatrus, 113
Iberia (Caucasian region), 64, 71, 74 Macedonia, 13, 17, 31, 40, 42, 46, 47, 55, 81,
Iceland, 162 82, 126, 145
Illyria, 81 Măcin, 106, 129, 135
Ioannoupolis, 31, 37, 38, 41, 43, 48, 51, Mačvanska Mitrovica, 97
59–62, 86, 167 Mahmudia, 102, 124, 132
Iron Gates, 3, 57, 113, 149 Malka Popina, 74
Isaccea (Noviodunum), 2–4, 7, 10, 18, 24, Mangalia, 23, 24
42, 48, 82–84, 90, 93, 94, 104, 106, 113, 116, Mantzikert, 73, 74, 96
117, 124, 131, 132, 134, 136–138, 140, 142, Maritsa, 13, 42, 139
145, 147, 148, 153, 157, 164, 165, 168, 171 Markellai, 11, 12
Iširkovo, 138 Maurokastron, 50
Iskar, 37 Medgidia, 24, 26
Ismail, 18 Melnitsa, 90, 97, 138
Istria, 10, 124 Mesembria (Nesebăr), 11, 13, 16, 17, 20, 21,
Istros, 68, 164 24, 45, 77, 78, 83, 85, 115, 122, 127, 164
Italy, 39, 150, 158 Mesopotamia (the Eastern province), 39,
Iur’ev, 145 40, 42, 70
Izvorovo, 42, 68 Messia, 97
Migali Berisklafa, 94
Jabalkovo, 11 Mileone, 11
Jurilovca, 94 Mircea Vodă, 26, 27
Mironeşti, 14
Kalipetrovo, 142 Mlava, 100
Kamčija, 134, 135 Moesia Prima, 98
Karnobat, 11 Moesia, 11, 98
Kavarna, 117, 118 Moglena, 162
Kerč Straits, 23, 45 Moldavia, 8, 13, 14, 39, 40, 101, 122, 135, 136,
Khazaria, 15, 29 152, 153, 161
Kiev, 29, 30, 37, 115, 144, 145 Montenegro, 56
Kilia, 94 Morava, 43, 48, 56, 77, 100, 113, 154, 165, 171
Kimbalongon, 55 Moravia, 98
Kirkovo, 138 Moravon, 43
Kladenci, 117, 118 Morisena-Cenad, 53
Klimata, 15, 49, 50 Moroviskos, 96
Konopas, 24 Mostiştea, 131
Krivina, 3, 67, 90, 94, 113, 127 Mundraga, 111
Krumovo, 11 Mureş, 54
Kubei, 9 Murfatlar-Basarabi, 26
Murighiol, 94 Pliska, 2, 4, 31, 51, 64, 67, 74, 77, 85, 89, 90,
Mysia, 37, 163 94, 111, 124, 130, 131, 134, 138
Plopeni, 132
Nera, 149 Podunavia, 69
Niculițel, 7–11 Pontos Euxinos, 45
Nikopol, 113 Popina, 113, 117, 118, 127, 131, 132
Niš (Nissos, Naissos), 77, 95, 96, 99, 100, Prahova (county), 14
116, 126, 149, 154, 155, 159, 165, 166, 172 Preslav, 2–4, 28–32, 37, 38, 40, 43, 45–48,
Novgorod, 115 51, 59, 60, 62, 66, 77, 78, 83, 84, 86,
Novi Pazar, 43 88–94, 97, 111, 114, 118, 124, 126, 127, 138,
Noviodunum (Isaccea), 7, 18, 42, 43, 93, 139
104, 136 Presthlavitza, 38, 43, 90–94, 103, 116, 122,
Nufăru, 3, 4, 7, 71, 83, 90, 92–94, 101–104, 136
106, 110, 113, 116, 124, 130, 132, 136, 145, Prislav (Prislava), 92
147, 171 Proslavitza, 92
Prut, 8, 136, 153
Ochrid, 47, 48, 54, 55, 96
Odărci, 3, 114, 117, 118, 124 Qipchak, 153
Oituz, 157 Quedlinburg, 46
Okorš, 117, 118
Oltenia, 144 Radovanu, 14
Olteniţa, 13 Ras, 43
Oltina, 28, 35, 83, 84, 101, 108–110, 112, 117, Raška, 43, 149, 154
118, 130, 132 Rasova, 28, 35, 110, 112
Onglos, 7–9 Razgrad, 31
Orjakhovo, 13, 42, 88 Rhine, 148
Ostrov (branch of the Danube), 34 Riphaei (mountains), 8
Ozolimne, 138, 139 Riš, 134
Rosokastron, 125
Păcuiul lui Soare, 2–4, 34, 35, 43, 78, 84, 91, Rujno, 4, 117, 118
101, 109–111, 114, 115, 123, 124, 127, 130, 132, Rum, 47
136, 138, 142, 145, 153 Ruse, 113
Palas, 15 Rusokastro, 126
Pannonia, 21 Russia, 50
Paphlagonia, 19, 38
Paradunavon, 1, 3, 5, 17, 59, 63, 65, 67–100, Saca, 134
114, 115, 122, 127, 129–131, 137–141, 144, 147, Sacce, 134
148, 168–170, 172 Saint Catherine in Sinai, 100
Paristrion, 65, 68–70, 86, 142, 148 Saint George (branch of Danube in the
Patzinakia (north of the Danube), 36 Delta), 92, 116
Patzinakia (a land given to the Pechenegs Sakdji, 134
in the empire), 123, 125, 169 Sakz’, 132
Peceneaga, 124 Salsovia, 102
Pecineaga, 124 Saqčy, 134
Peloponnesus, 150 Sarkel, 15
Pera, 135 Satu Nou, 35
Pereiaslavetz, 29, 30, 92, 93 Sava, 80, 97, 99, 149, 154
Periprava, 18 Scythia minor, 131
Pernik, 55 Sebasteia, 73
Peuce, 10 Selinas, 24, 116
Philippopolis (Plovdiv), 51, 77, 137, 149, 163 Şendreni, 14, 114
Piatra Frecăţei, 4, 106–107 Şerbeşti, 8
Pietroiu, 114 Serbia, 3, 54, 56, 63, 148–150, 154, 158, 165,
Piua Petrii, 10, 114 166
208 index of geographical names
Serdica (Sofia), 46, 47, 66, 70, 76, 77, 95, Tatarbunar, 8, 9, 14, 18
99, 129 Tayk, 48, 70, 71, 74
Severin, 53 Tegulicium, 112
Sicily, 148, 150, 154, 158 Teleorman, 152
Sidera, 134 Tempe, 19
Silistra (Durostorum, Dristra), 1, 2, 7, 41, 51, Tenouormon, 152
66–68, 71, 74, 78, 80, 81, 83, 89–91, 102, Theodoroupolis, 38, 39, 62, 86
111, 112, 118, 123, 131, 132 Thessaloniki, 39, 42, 51, 55, 64, 76, 95, 100,
Simeonovgrad, 42 159, 171
Singidunum, 155 Thessaly, 18, 19
Siret, 8, 14, 136 Thrace, 9, 12, 13, 21, 31, 37, 38, 40–42, 47, 51,
Sirmia, 97 59–62, 82, 85, 86, 126, 137–140, 145, 159,
Sirmium (Sremska Mitrovica), 3, 46, 56, 162–164, 167
57, 63, 68, 95–98, 113, 116, 147, 150, 154, Tiča, 94, 134
157–159, 167 Tigris, 42
Skala, 4, 114, 117, 118 Timiş, 154
Skopje (Skopion), 55, 95, 96, 99 Timok, 113
Sliven, 90 Tisza, 147
Slon, 14 Tmutorokan, 23
Small Earthen Dike, 8, 9, 20, 24, 28 Tomis, 23, 32, 42
Sofia (Serdica), 46, 70, 88, 126, 149, 150, 158 Topalu, 107
Somova, 136 Topolovgrad, 90
South-Eastern Europe, 16 Transylvania, 14, 33, 137, 157
Sozopolis, 13, 17, 21, 85, 140 Trjavna, 144, 163, 154
Spain, 29 Troesmis, 106
Spalato, 149 Trojan Gorge, 47, 48
Spercheios, 51 Tsar Asen, 114, 117, 118
Sredetska, 13 Tsarevets, 161
Središte, 118 Tulcea, 3, 11, 104, 117, 127
Stancea, 112 Tulcea (county), 124
Stara Planina, 11, 12, 29, 37, 55, 85, 99, 127, Tuluceşti, 8
137, 140, 143, 144, 151, 158–160, 162, 163, Turcoaia, 3, 106, 117, 129, 121, 122, 145
165, 168–171 Turnu Severin, 53, 114
Stara Zagora (Beroe), 163 Tutrakan, 111, 113
Stone Dike, 9, 24, 26–28, 42, 49, 152 Tyras, 50
Strumitza, 73
Strymon, 17, 55, 163 Ukraine, 50
Sucidava, 114 Urluia, 28
Sudak, 23
Sulina, 116 Vadul Cumanilor, 144
Šumen, 117, 161 Vadul lui Isac, 8, 9, 14, 18
Susaco, 141 Vâlcov, 18
Suvestcastro, 113 Valea Dacilor, 124
Suvorovo, 18 Valea Teilor, 11
Suzdal’, 153 Vălnari, 124
Svištov, 113, Valul lui Traian, 28, 101, 102
Syria, 59 Varna, 2, 24, 77, 89, 94, 116, 124, 127, 134,
164
Tabla Buţii, 14 Vaspurakan, 71, 73–76
Taliata, 113 Veliatova, 137
Tara, 154 Veliki Grad, 155
Târguşor, 124 Veliki Gradac, 113
Tărnovo, 97, 111, 161, 164 Venice, 149, 150, 166
Taron, 73 Venzina, 134
Vetren (Vetrinon), 3, 66, 83, 89, 94, 101, 12, Western Mesopotamia, 39–43, 48, 51, 57,
113, 118, 126 62, 64, 76, 165, 167
Via Egnatia, 144, 159
Vicina, 50, 135–137 Ypsala, 159
Vidin, 53, 54, 56, 57, 76, 77, 100, 113, 135,
144, 147 Zachlumia, 56
Viminacium, 100, 155 Zacos collection, 71, 73, 74, 77
Vitzina, 132, 134, 135 Zagora, 11
Vodno, 138 Zara, 149
Vojvoda, 31 Zemun, 149, 154, 157
Volga, 15, 153 Zichia, 141
Volyhnia, 153 Zimnicea, 114, 152
Zlati Voivoda, 97, 138
Walachia, 13, 14, 33, 40, 69, 122, 123, 136,
152, 153, 161 23 August (Constanţa County), 130
Western Europe, 161
Index of Sources
Abulfida, 134 Leo the Deacon, 28, 30–32, 34–39, 47, 49

Albertus Aquensis (Albrecht of Aachen), Leo VI, Taktika, 21
98, 99 Life of Saint Cyril the Phileote, 115
al-Idrisi, 94, 113, 125, 126, 134, 135, 147
al-Masudi, 21 Manuel Holobolos, 165
Ananias of Širak, 10 Manuel Straboromanos, 141
Anna Comnena, 68, 79, 81, 83, 84, 88, 99, Matthew of Edessa, 66, 68, 69, 72, 73, 75,
112, 129, 132, 134–140, 142–144 127, 129, 139
Annales Regni Francorum, 11, 21 Mehmed Nešri, 134
Anonymous Book on Tactics, 36, 143 Michael Attaliates, 66, 69, 73, 74, 79–83,
Ansbertus, 163–165 87, 95, 116, 123, 126, 129–131
Michael Glykas, 129
Baybars al-Mansuri, 134 Michael Psellos, 118, 121, 129
Michael the Syrian, 66, 80, 144, 145
Chronicle of Monemvasia, 38 Moses Chorenatzi, 10
Chronicle of the Priest of Duklja, 43
Chronicon Henrici de Mügeln, 116 Nicholas I, 21
Constantine VII Porphyrogenitus Nikephoros, 7, 8, 11, 12
(De Administrando Imperio), 17, 21, Niketas Choniates, 100, 144, 145, 148, 150,
23, 24, 32, 50, 134 151, 154, 155, 157, 159–165
Note of the Gothic Toparch, 49, 51
Eberwinus, 100
Pappus of Alexandria, 10
Geoffroy of Villehardouin, 162 Peri strateias, 32
George Monachos, 13, 22 Photios, 17
George Akropolites, 144, 161, 163, 164 Pilgrim of Passau, 98
Robert de Clari, 162
Hudud al-’alam, 47 Russian Primary Chronicle, 15, 22, 23, 29,
30, 36, 37, 92, 116, 132, 134, 142
Il compasso da navigare, 135
Innocentius III, 162, 166 Scriptor incertus, 13
Simon of Keza, 98
John Kinnamos, 80, 95, 100, 135, 144, 145, Skylitzes Continuatus, 66, 69, 73, 76,
148, 150–155, 157, 158 79–82, 129–131
John Mauropous, 65 Smbat Sparapet, 73
John Skylitzes, 13, 15, 16, 21, 22, 28, 30–37, Snorri Sturlusson, 162
39–41, 46–48, 51, 53, 55, 56, 61, 62, Symeon, 13, 14, 21
64–66, 73, 74, 77, 83, 87, 88, 116, 117, 122, Syrianos Magister, 32
123, 126
John Zonaras, 8, 12, 13, 16, 21, 28, 30–32, 36, Taktikon Scorialensis, 37–39, 42, 43, 61
37, 40, 46, 51, 53, 55, 66, 68, 69, 76, 79, Theophanes Confessor, 7, 8, 11–13
80, 82, 83, 116, 117, 123, 126, 127, 129, 130, Theophanes Continuatus, 21, 22
139, 142
Vision of Prophet Isaiah, 8
Kekaumenos, 65, 66, 69, 77, 127
William of Tyre, 98, 169, 170
Legenda Major Sancti Gerardi, 53
Leo Grammaticus, 13, 14 Yahya-ibn-Said, 34, 35
index of Modern authors
Ahrweiler Hélène, 17–19 Iliescu Octavian, 18

Anastasijević Dragutin, 1 Iorga Nicolae, 1, 53, 79, 80
Angelova Stefka, 3
Atanasov Georgi, 3, 4 Jordanov Ivan, 3, 4, 17, 38, 63, 84, 98
Baraschi Silvia, 3 Kazhdan Alexander, 65

Barnea Alexandru, 4 Kostova Rossina, 93
Barnea Ion, 2–4, 18, 19, 71, 76, 78 Krsmanović Bojana, 63
Bartikian Rats, 70 Kühn Hans-Joachim, 4
Baumann Victor Heinrich, 3, 4 Kuzev Alexander, 2
Bănescu Nicolae, 1, 2, 63, 68–71, 78, 79
Bica Ioan, 4, 63 Lambros Spyros, 70
Bogdan Damian, 26 Laurent Vitalien, 2, 60
Bolşacov-Ghimpu Alexandru, 2, 42 Lefort Jacques, 65
Borisov Boris, 63
Božilov Ivan, 3 Magdalino Paul, 150
Brătianu Gheorghe, 1, 169 Makk Ferenc, 4, 141
Browning Robert, 18 Maksimović Ljubomir, 3
Mănucu-Adameşteanu Gheorghe, 3, 4, 116,
Chiriac Costel, 4 118, 164
Ciggaar Krijnie, 140 Mărculeţ Vasile, 76
Cihodaru Constantin, 2 Miller Konrad, 94
Comşa Eugen, 27, 78 Mitrea Bucur, 2
Curta Florin, 5 Morrisson Cécile, 117
Custurea Gabriel, 4 Mutafčiev Petăr, 1
Damian Oana, 4 Năsturel Petre Ş., 2, 129, 136, 152

Diaconu Petre, 4, 5, 18, 26, 34, 35, 38, 67, Necşulescu Constantin 1
69, 70, 78, 91–94, 108, 129, 136, 139, 145 Nikolov Stepan, 41
Dimitrov Dimităr I., 9
Dončeva-Petkova Ljudmila, 3 Oberländer-Târnoveanu Ernest, 3, 4, 110
Dulaurier Edouard, 72 Oikonomides Nicolas, 2, 4, 35–41
Opriş Ioan C., 4
Ferjančić Božidar, 97
Fiedler Uwe, 9 Paraschiv Dorel, 4
Florescu Grigore, 1 Popkonstantinov Kazimir, 27
Florescu Radu, 2, 20, 107, 108, 111 Popović Marko, 3
Prinzing Günter, 98, 159
Georgiev Pavel, 9, 36 Rašev Rašo, 8, 9
Grierson Philip, 117, 118
Grünbart Michael, 71, 74 Sâmpetru Mihai, 33
Guilland Rodolphe, 98 Savvides Alexios, 75
Györffy György, 98 Schuhhardt Carl, 108
Seibt Werner, 3, 41, 42, 63
Haldon John, 4, 125 Ševčenko Ihor, 50, 51
Hase Charles, 49, 50 Shepard Jonathan, 3, 4, 115, 141
Holmes Catherine, 62 Spinei Victor, 80, 153
212 index of modern authors
Squatriti Paolo, 28 Urbansky Andrew, 2

Stănescu Eugen, 2
Stephenson Paul, 4, 5, 63, 68, 79, 122, 168 Vasiliu Ioan, 3
Ştefan Gheorghe, 1, 2
Wasilewski Tadeusz, 2, 4, 63, 68, 69, 99
Tanaşoca Nicolae-Şerban, 3
Tăpkova-Zaimova Vasilka, 3, 17–19, 70, 115 Yotov Valeri, 4, 60
Tocilescu Grigore, 108
Treadgold Warren, 4 Zlatarski Vasili, 70
Tsankova-Petkova Genoveva, 164

Madgearu, Alexandru - Byzantine Military Organization On The Danube, 10th-12th Centuries

Uploaded by

Copyright:

Available Formats

Madgearu, Alexandru - Byzantine Military Organization On The Danube, 10th-12th Centuries

Uploaded by

Document Information

Original Title

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

Madgearu, Alexandru - Byzantine Military Organization On The Danube, 10th-12th Centuries

Uploaded by

Copyright:

Available Formats

Byzantine Military Organization on the Danube,

The titles published in this series are listed at brill.com/ecee

Library of Congress Cataloging-in-Publication Data

Copyright 2013 by Koninklijke Brill NV, Leiden, The Netherlands.

This book is printed on acid-free paper.

List of Illustrations ........................................................................................... vii

I. The Recovery of the Danubian Frontier ........................................... 7

II. The Military Organization of the Danube Region ......................... 59

III. The Evolution and Function of the Danube Frontier of

1. The dikes between Cernavoda and Constanţa (after Barnea,

AAnt Acta Antiqua Academiae Scientiarum Hungaricae,

Cronica 2011 Cronica cercetărilor arheologice din România. Campa-

RESEE Revue des Études Sud-Est Européennes, Bucarest

The first monograph offering a comprehensive survey of the Byzantine

title “Cultura bizantină în România” [Byzantine Culture in Romania], and

so many seals been found in a precisely dated context. The numismatic

28 Barnea 1993 b; Barnea 1997; Barnea 2001.

province Paradunavon is hardly accessible, as it was published only in a

The Recovery of the Danubian Frontier

of the Danube. Moreover, both authors write of Onglos as a place, not a

trusted at all in matters pertaining to the early history of Bulgaria, except

Agighiol—1 silver 668–669;

Tulcea—1 gold of 641–646;

25 Howard-Johnston 2006, 344.

to Bulgaria. During the subsequent centuries, the navy appears to have

30 Angelova, Koleva 2004, 21–22; Atanasov 2012.

Greaca, Radovanu, Curcani, Mironeşti), as well in other places in Wala-

therefore replaced the Khazars as the guarantors of the Byzantine security

was probably a place in Thessaly, Epirus, or Acarnania. Tăpkova-Zaimova

50 Tăpkova-Zaimova 1970, 82–86.

59 Treadgold 1988–1989, 140–143; Whittow 1996, 254–255.

62 As advocated by Diaconu 1981 a, 218 and Custurea 1986, 276.

75 Stokes 1960, 507–514, Soloviev 1960, 574.

80 Constantine Porphyrogenitus, IX, 91–102 (ed. Moravcsik, 62/63).

carved in smaller dimensions than in the Roman period were extracted

84 Panaitescu 1978 b, 247–251.

that monastery is dated to the mid-10th century on paleographic grounds,

96 Rašev 1987, 55–56; Rašev 2005, 57; Curta 2006, 230.

100 Marinow 2011, 453.

104 Iorga 1998, 170–171.

fortresses established beyond the Danube. They sought an amnesty for

113 Stănescu 1974, 398; Mănucu-Adameşteanu 1991 a, 317.

117 Leo the Deacon, IX. 7 (ed. Talbot, Sullivan, 195).

123 Vâlceanu 1967, 593–615.

in a military report going back to information collected from spies who

128 Oikonomides 1987, 187–190.

139 Mutafčiev 1932, 193–194.

147 Oikonomides 1986 a, 327–335; Diaconu 1986, 170–171; Diaconu 1987 c, 483–484.

extension of Byzantine power north of the river Danube in 971. Tzimiskes

It is even possible that Damian Dobromir was of Vlach origin, since it is

160 As I sustained in Madgearu 1999–2000, 18.

of strategoi of Mesopotamia, which were in Izvorovo and Simeonovgrad

165 Jordanov 2003 a, 124, 125, nr. 48. 1, 3.

170 Madgearu 1999–2000, 16; Madgearu 2001, 75–76. Tăpkova-Zaimova 1973, 96 supposed

2. The military organization between 971 and 986

Soon after the conquest of 971, another strategy was established at

174 Nesbitt, Oikonomides 1991, 168–169, nr. 73.1.

period of peace, interrupted in 1043 by a renewed Rus’ attack (discussed

of Bulgaria in 997 because Romanos, the brother of the former emperor

III. The Evolution and Function of the Danube Frontier of

1. The dikes between Cernavoda and Constanţa (after Barnea,

in the following chapter.34 The chronology of the events was confused

– Aetios, protospatharios and strategos of Presthlavitza (eight seals found

– John Malesis, patrikios and strategos of Presthlavitza (one seal from

18 Baumann, Mănucu-Adameşteanu 2001, 219; Mănucu-Adameşteanu 2001 a, 57;