THE ATHENIAN

AGORA
RESULTS OF EXCAVATIONS
CONDUCTED BY
THE AMERICAN SCHOOL OF CLASSICAL STUDIES AT ATHENS

VOLUME XXIV
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

LATE
ANTIQUITY:
A.D. 267-700

BY

ALISON FRANTZ

WITH CONTRIBUTIONS BY

HOMER A. THOMPSON AND JOHN TRAVLOS

THE AMERICAN SCHOOL OF CLASSICAL STUDIES AT ATHENS

PRINCETON, NEW JERSEY
1988

American School of Classical Studies at Athens

is collaborating with JSTOR to digitize, preserve, and extend access to
The Athenian Agora ®
www.jstor.org
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

Roman Lamp. Circa A.D. 600 (L 1494, twice actual size)

"Athena Promachos strides away with grim determination,
turning her back on this whole crew of saints and martyrs."
H. A. Thompson
 PUBLISHED WITH THE ASSISTANCE OF THE GETTY GRANT PROGRAM
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

ALL RIGHTS RESERVED

Library of Congress Cataloging-in-Publication Data

Frantz, Alison.
Late Antiquity, A.D. 267-700.

(The Athenian Agora; v. 24)

Bibliography: p.
Includes index.
1. Athens (Greece)-History. 2. Athens (Greece)-Antiqui-
ties. 3. Athens (Greece)-Antiquities, Byzantine. 4. Greece-
Antiquities. 5. Greece-Antiquities, Byzantine. I. Thompson,
Homer A. II. Travlos, John, 1908- III. American School of
Classical Studies at Athens. IV. Title. V. Series.
DF287.A23A5 vol.24 [DF922] 938'.5 s 88-7525
ISBN 0-87661-224-9 [938'.5]

TYPOGRAPHY BY THE AMERICAN SCHOOL OF CLASSICAL STUDIES PUBLICATIONS OFFICE

C/O INSTITUTE FOR ADVANCED STUDY, PRINCETON, NEW JERSEY
PLATES BY THE MERIDEN-STINEHOUR PRESS, MERIDEN, CONNECTICUT
PRINTED IN THE UNITED STATES OF AMERICA
BY THE TOWN HOUSE PRESS, SPRING VALLEY, NEW YORK
 MEMORIAE
CAROLI RVFI MOREY
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

RERVM ANTIQVARVM AESTIMATORIS

AEVI OBSCVRI ILLVMINATORIS
EXPERTI IN STVDIIS MEDIAEVALIBUS
LITTERARVMQVE HVMANIORVM PROPVGNATORIS
DEVOTE DICATVM
 The confusion of the times, and the scarcityof authentic memorials,
oppose equal difficultiesto the historian, who attemptsto preservea
clear and unbrokenthread of narration. Surroundedwith imperfect
© American School of Classical Studies at Athens

fragments, always concise, often obscure, and sometimes contradic-

For personal use only. License: CC-BY-NC-ND.

tory, he is reduced to collect, to compare, and to conjecture:and

though he ought never to place his conjecturein the rank of facts, yet
the knowledgeof human nature, and of the sure operationof its fierce
and unrestrainedpassions,might, on some occasions,supply the want
of historicalmaterials.
EDWARDGIBBON
 PREFACE
The paucity of literary documentationfor the history of Athens during the six centuriesfollowing
the Antonine period, and the physical destructionof most of the buildings, whether by war, negli-
gence or natural decay, have combinedto wrap these centuries in an obscuritythat is only now
beginning to be penetrated. The decline in the strategic importanceof the Province of Achaea
brought with it a similar decreasein the power of the provinceto exert financial pressure on the
© American School of Classical Studies at Athens

central government,and few historians found it necessaryeven to mention the city of which the
For personal use only. License: CC-BY-NC-ND.

chief reminderswere the ancient works of art that now embellishedConstantine'snew capital.
Few new literary sourceshave come to light since Gibbon led the way in plumbingthe depthsof
the basse epoque, and although modern scholarshipmay attack his historicaljudgments, he alone
has providedthe wealth of leisurely detail, derivedfrom primarysources,which make his work the
delight of the readerand a boon for the scholar.No apologyis made for unquestioningreferencesto
Gibbon'sown accountsof peripheralinterestand for acceptingwith gratitudethe ease with which
he has made it possible to follow back to their sometimesobscure sources statementswhich may
have a directbearing on the matter in hand.
A century after Gibbon the German scholarsGregorovius,Wachsmuth,and finally Judeich col-
lected referencesto and interpretedalmost all known passages relating to Athens, and these testi-
monia have formedthe backboneof all subsequentwritings on the subject,so that until recentlyit
could be said, in Gibbon'swords in anothercontext, "allthat learningcan extractfromthe rubbish
of the dark ages is copiouslystated ... ."
The meager literary framework is now gradually being filled out through the archaeological
discoveriesof the past few years:by the systematicexcavationof the AthenianAgora and the area of
the Dipylon, and by the vigilance of the Greek ArchaeologicalServicein recordingthe remains of
ancient buildings unearthed during modern building operations. The objectiveof this book has
been to correlatethe literary and archaeologicalaspects of the subjectand so to present as full a
picture as possible of Athens in Late Antiquity.
From the end of the 3rd centurythe Agora of classicaltimes no longer had any significanceas a
separate entity within the city; therefore it can now be intelligible only as a part of Athens as a
whole. But, except for the Agora and the areas of the Dipylon and Olympieion, almost all the
ancient city lies under the modern (PI. 1). In the course of new constructionbits and pieces of
ancient buildings, walls, streets, and graves are continuallycoming to light, but only sporadically,
and it is seldompossibleto follow them to their limits. Furthermore,they are usually promptlyand
permanentlycoveredup once they have been recorded.Hope of recoveringthe completeplan of the
city is no less elusive than is the vision that the lost historiesof Dexippus and Eunapiuswill one day
emerge from obscurity. In spite of these limitations it has been considerednecessaryto take into
account the whole of the ancient lower city, however sketchily it must be presented, in the full
knowledge that topographicaltheories may be made obsolete at any moment but in the hope that
future investigatorsmay find here some basis on which to build. In this survey, except in specific
instances, little account has been taken of movable finds of Late Antiquity from the excavation.
Special studies of the Agora series have already been devotedto the coins, pottery, terracottafig-
urines, and lamps (see the Bibliography).
The problemof transliterationfamiliar to all classical scholarsis furthercomplicatedby dealing
with a Greek provincewithin the Roman Empire. At one extreme, names from classicalantiquity
 viii PREFACE

in generalcall for a Greek spelling. At the other,namesof Romanemperorsdemandthe Latin form

even when they are of Greek origin, e.g. Theodosius. In the middlegroundsubjectivityreigns, and
inconsistencyis claimedas a right.
The death of my collaboratorJohn Travlos has deprivedthe book of some detaileddescriptions
and drawings which had been projectedto fill out the architecturalaspectsof this survey. He had,
however, completed all the most important drawings. The most serious lack in the text was the
architecturaldescriptionof the large "Palaceof the Giants"in the middleof the Agora, for which he
had drafted a preliminaryversion. On the basis of this, Homer Thompson, who had followed in
detail the excavationand investigationof the building from start to finish, has kindly providedan
expanded, revised, and updated description,with a new interpretationof its purpose, which ap-
© American School of Classical Studies at Athens

pears as Chapter V. Travlos' contributionsto the text are the short sectionon the water supply in
For personal use only. License: CC-BY-NC-ND.

Chapter III and the full descriptionof the Post-Herulian Wall which appears as an Appendix.
This bare listing in no way doesjustice to his invisible contribution.We exchangedviews over the
years on all aspects of the subject.He was generouswith his knowledgeand keen in his judgment.
The finished productis poorer without his final work, but his influencethroughoutwill be recog-
nized by all who knew him. He was a reveredcolleagueand a faithful friend.
That the excavationof the Agorahas been a corporateundertakingfromthe beginningis evident
from the numberof colleagueswhose help it is a pleasureto acknowledge.Aboveall, thanksare due
to Homer A. Thompson who, as Field Director,championedthe rights of Late Antiquityand from
his commandof the whole span coveredby the excavationscontributedtoward putting this late
period in a new perspective.I am also indebtedto T. Leslie Shear, Jr. for permissionto include
some materialuncoveredsince he assumedthe Field Directorshipin 1968. The fieldnotebookskept
by the scholarssupervisingthe varioussectionsof the excavationshave been the primaryand some-
times the only sourceon which the presentstudy of the Agora materialis based.Opinions and even
words of others have been freely used, and it is hoped that any unconsciousplagiarism will be
accepted as a tribute. The excavatorscharged with the chief areas under discussion were John
McK. Camp II, Margaret Crosby, G. Roger Edwards, Richard H. Howland, James H. Oliver,
Arthur W. Parsons,Henry S. Robinson,RebeccaW. Robinson,Dorothy B. Thompson, Homer A.
Thompson, Eugene Vanderool, Frederick0. Waage, and Rodney S. Young. Grateful acknowl-
edgmentis made to Elizabeth G. Caskey, Fred S. Kleiner,John H. Kroll, D. M. Metcalf, the late
George C. Miles, and MargaretThompson for invaluablenumismatichelp; also to John W. Hayes
for his wisdom on Roman pottery.
Thanks are due to the directorsof the museums of Athens for many facilities:the late George
Sotiriou,Manolis Chatzidakis,and Paul Lazaridesin the ByzantineMuseum;Nicholas Yalourisin
the National ArchaeologicalMuseum; the late John Miliadis and George Dontas in the Acropolis
Museum; Markellos Mitsos and Dina Peppas-Delmouzouin the EpigraphicalMuseum.
In Athens over the years the late Anastasios Orlandos was always generous with his time and
counsel, as were other membersof the Greek ArchaeologicalService:Angeliki Kokkou,Demetrios
Pallas, the late John Papadimitriou,the late Stylianos Pelekanidis,Photios Petsas, Nicholas Pla-
ton, EustathiosStikas, the late John Threpsiades,and the late Nicholas Verdelis.
The substanceof this book has been enrichedby conversationand correspondencewith Javier
Arce, Timothy Barnes, Judith Binder, Graeme Clarke, SlobodanCurciic,John Dillon, Noel Du-
val, Garth Fowden, Frank Gilliam, Andre Grabar, Timothy Gregory, Christian Habicht, Chris-
topher Jones, Ernst Kitzinger, Richard Krautheimer, Paul Lemerle, Arthur H. S. Megaw,
FranSoisPaschoud,Kenneth Setton,Jean-Pierre Sodini, and Sir Ronald Syme. Special thanks are
due to Glen Bowersock,whom I have consultedon many points. He has read my manuscript,made
 PREFACE ix

many constructivesuggestions,and deliveredme from some egregious errors. He is not, however,

responsiblefor any that remain.
Financial support from the American PhilosophicalSociety (1977 and 1984) and the National
Endowment for the Humanities (1976) is gratefully acknowledged.My debt to the Institute for
AdvancedStudy, where much of this text was written, is inestimable,for hospitalityextendingfar
beyond my one year as a member.A generous grant from the J. Paul Getty Trust has met a sub-
stantial part of the costs of publication.
The photographsof Agora material are from the files of the Agora Excavations,the work suc-
cessively of Hermann Wagner, Alison Frantz, James Heyle, Eugene Vanderpool,Jr., Robert K.
Vincent, and Craig Mauzy. Air views were taken by J. W. and E. E. Myers in May, 1975. The
© American School of Classical Studies at Athens

rest, unless otherwise specified, are by the undersigned.Almost all the prints, sometimes from
For personal use only. License: CC-BY-NC-ND.

recalcitrantnegatives,owe much to the skill of Nicholas Restakis,the Agora'sdarkroomtechnician

for many years. For assistanceof other kinds I am indebtedto Margot Camp and Jan Diamant,
secretariesof the Agora Excavation,and, in Princeton,to Enid Bayan for typing and to Marian H.
McAllister and the staff of the School PublicationsOffice for skill and forbearancein seeing the
manuscriptthrough final stages.
All dates, unless otherwise specified,are of the Christianera.

The dedicationof this volume to the memoryof Charles Rufus Morey is an inadequateexpres-
sion of gratitudefor leading me into the then almostunchartedrealm of Late Antiquityand for con-
tinuing help and encouragementthroughthe remainingyears of his life.

1988
 TABLE OF CONTENTS
PREFACE .................... .... .
... .. .. vii
LIST OF ILLUSTRATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . XV
BIBLIOGRAPHY AND ABBREVIATIONS . . . . . . . . . . . . . . . . . . . . . . . xix
© American School of Classical Studies at Athens

I. INVASIONS AND THEIR AFTERMATH

For personal use only. License: CC-BY-NC-ND.

THE PRE-HERULIAN THREAT (A.D. 254-267) . . . . . . . . . . . . .. . . 1

THE HERULIANINVASION . . . . . . . . . . . . . . . . . . . . . . . . 1
ACCOUNTS. .
HISTORICAL . . . . . . . . . . . . . . . . . . .. . . 1
DAMAGE. . . .
PHYSICAL . . . . . . . . . . . . . . . . . . . . . . 3
THE POST-HERULIAN
WALL . . . . . . . . . . . . . . . . . . . . . . 5
II. RECOVERY
POST-HERULIANATHENS (A.D. 267-300) . . . . . . . . . . . . . . . . . . 12
THE AGORA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION
DIOCLETIANTO JULIAN THE APOSTATE(A.D. 284-361) . . . . . . . . . . . . . 16
CONSTANTINE AND ATHENS . . . . . . . . . . . . . . . . . . . . . . . 16
THE REFLORESCENCE OF THE SCHOOLS . . . . . . . . . . . . . . . . . . 17
THE PROGRESS OF CHRISTIANITY . . . . . . . . . . . . . . . . . . . . 18
THE CITY OF ATHENS AND ITS BENEFACTORS . . . . . . . . . . . . . . . . 20
JULIAN AND THE REOPENING OF THE TEMPLES . . . . . . . . . . . . . . . 24
THE BUILDINGS . . . . . . . . . . . .... . . . 24
THE RECONSTRUCTION OF OLD BUILDINGS FOR OFFICIAL USE . . . . . . . . . 24
THE POMPEION AND THE PANATHENAIC WAY . . . . . . . . . . . . . . . 26
W ATERWORKS . . . . . . . . . . . . . . . . . ... . . . . . 29
WATER SUPPLY (by John Travlos) . . . . . . . . . . . . . . . . . . . 29
DRAINAGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
BATHS . . . . . . . .. . .. . . . . . . . . . . . . . . . . . . . 30
LATRINES . . . . . . . . . . . . . .. .. . . . . . . . . . . . . 33
DOMESTIC ARCHITECTURE . . . . . . . . . . . . . . . . . . . . . . . 34
PRIVATE HOUSES . . . . . . . . . . . . . . . . . . . . . . . . . . 34
HOUSES ON THE AREOPAGUS: ARCHITECTURE . . . . . . . . . . . . . . . 37
HOUSES ON THE AREOPAGUS: WATER SUPPLY . . . . . . . . . . . . . . 42
"HOUSE OF PROCLUS" . . . . . . . . . . . . . . . . . . . . . . . . 42
HOUSES ON THE AREOPAGUS: IDENTIFICATION . . . . . . . . . . . . . . 44
HOUSES ON THE AREOPAGUS: CHRONOLOGY . . . . . . . . . . . . . . . 47
ATHENS IN A.D. 396 . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
ALARICIN GREECE(A.D. 396) . . . . . . . . . . . . . . . . . . . . . . . 49
THE THREAT . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
DESTRUCTION IN THE AGORA . . . . . . . . . . . . . . . . . . . . . 53
 xii TABLEOF CONTENTS

IV. PROSPERITY,DECLINE, AND DISASTER (A.D. 396-582)

PROSPERITY(A.D. 396-450) . . . . . . . . . . . . . . . . . . . . . . 57
THE TIMEOFTHENEOPLATONISTS . . . . . . . . . . . . . . . . . . . . 57
REVIVAL AFTER ALARIC . . . . . . . . . . . . . . . . . . . . . . . . 58
STATE OF THE FORTIFICATIONS . . . . . . . . . . . . . . . . . . . . 58
OFPUBLICBUILDINGS. . . . . . . . . . . . . . . .
REHABILITATION 58
NEW CONSTRUCTION . . . . . . . . . . . . . . . . . . . . . . . . 60
BUILDING ACTIVITY OFHERCULIUS . . . . . . . . . . . . . . . . . . . 63
THE RISE OF THE CHRISTIAN CITY . . . . . . . . . . . . . . . . . . . . 68
© American School of Classical Studies at Athens

LEGISLATION FOR THE ESTABLISHMENT OF CHRISTIANITY . . . . . . . . . . 69

For personal use only. License: CC-BY-NC-ND.

CLOSING OFTEMPLES. . . . . . . . . . . . . . . . . . . . . . . . 70
CONVERSION OF SECULAR BUILDINGS . . . . . . . . . . . . . . . . . . 71
FIRSTCHRISTIAN CHURCHES. . . . . . . . . . . . . . . . . . . . . 72
DECLINE(A.D.450-529) . . . . . . . . . . . . . . . . . . . . . . . . . 74
THE REMOVAL OF WORKS OF ART TO CONSTANTINOPLE . . . . . . . . . . . 74
THE OBELISK OFTHEODOSIUS . . . . . . . . . . . . . . . . . . . . . 74
"ABOUT THE ELEPHANTS" . . . . . . . . . . . . . . . . . . . . . . 75
THE ATHENAPROMACHOS . . . . . . . . . . . . . . . . . . . . . . 76
THE VANDALSIN ATHENS . . . . . . . . . . . . . . . . . . . . . . 78
LITERARYEVIDENCE. . . . . . . . . . . . . . . . . . . . . . . . 78
ARCHAEOLOGICALEVIDENCE. . . . . . . . . . . . . . . . . . . . . 78
THE STREET OF THE LIBRARY OF HADRIAN . . . . . . . . . . . . . . . . 79
THE BEGINNING OF INDUSTRIALIZATION . . . . . . . . . . . . . . . . . . 79
BRONZE FOUNDRIES . . . . . . . . . . . . . . . . . . . . . . . . . 80
WATERMILLS. . . . . . . . . . . . . . . . . . . . . . . . . . . 80
. . . . . . . . . . . . . . .
AND DISASTER(A.D. 529-582)
DISINTEGRATION 82
THE REIGNOFJUSTINIAN . . . . . . . . . . . . . . . . . . . . . . . 82
NEW CONSTRUCTION . . . . . . . . . . . . . . . . . . . . . . . . . 83
PRIVATEHOUSES. . . . . . . . . . . . . . . . . . . . . . . . . . 83
SHOPS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
NEW COMMERCIALCONSTRUCTION . . . . . . . . . . . . . . . . . . . 84
DECAY OF EDUCATIONAL INSTITUTIONS . . . . . . . . . . . . . . . . . . 84
CONVERSION OF TEMPLES INTO CHURCHES . . . . . . . . . . . . . . . . 92
THE SLAVIC INVASION . . . . . . . . . . . . . . . . . . . . . . . . . 93

V. THE PALACE OF THE GIANTS (by Homer A. Thompson)

EXPLORATION AND IDENTIFICATION. . . . . . . . . . . . . . . . . . . . . 95
LOCATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
OVERALL PLAN . . . . . . . . . . . . . . . . . . . . . . . . . . 97
STATE OF PRESERVATION . . . . . . . . . . . . . . . . . . . . . . . . 98
CONSTRUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . .. 98
NORTHCOURT... . . . . . . . . . . . . . . . . . . . . . . . . 99
PASSAGE FROM THE NORTH COURT TO THE SOUTH COURT . . . . . . . . . 102 .
SOUTH COURT. . . . . . . . . . . . . . . . . . . . . . . . . . 103
SOUTHEAST COURT . . . . . . . . . . . . . . . . . . . . . . . . . . 104
BATH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
 TABLEOF CONTENTS xiii

MARKS OF INCOMPLETION . . . . . . . . . . . . . . . . ....... 108

STRUCTURES TO THE NORTHWEST OF THE PALACE . . . . . . . . . . . ... 109
PLAN AND PURPOSE OF THE BUILDING . . . . . . . . . . . . . . . . . . . . 110

VI. THE DARKAGES(A.D. 582-700)

RECOVERY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
THE BUILDINGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
INDUSTRIAL ESTABLISHMENTS . . . . . . . . . . . . . . . . ....... 120

EPILOGUE. . . . . . . . . . . . . . . . . . . . . . . . . ........ 123

© American School of Classical Studies at Athens

APPENDIX: WALL(by John Travlos) . . . . . . . . . . ....

THE POST-HERULIAN 125
For personal use only. License: CC-BY-NC-ND.

CONSTRUCTION OF THE WALL . . . . . . . . . . . . . . . . ....... 126

DESCRIPTION OF THE POST-HERULIAN WALL BY SECTIONS. . . . . . . . . ... 127
W EST FLANK . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
NORTH FLANK . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
EAST FLANK . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
THE GATES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
COMMUNICATION BETWEEN THE CITY AND THE ACROPOLIS . . . . . . . . . . 141

INDEXES
I. GENERAL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
II. ANCIENTAUTHORS. . . . . . . . . . . . . . . . . . . . . . . . . . 151
III. INSCRIPTIONS
. . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

PLATES
 LIST OF ILLUSTRATIONS
FRONTISPIECE:Sixth-century lamp, L 1494 (Agora VII, no. 2364)
1. The Athenian Agora and Environs.Air View, 1975
2. Athens, 86 B.C. to A.D. 267. Plan
3. The Athenian Agora. Actual-statePlan
4. Athens, 5th century after Christ. Plan
5. The Acropolisand the Post-Herulian Wall. Plan
© American School of Classical Studies at Athens

6. The Athenian Agora, 5th centuryafter Christ. Plan

7. The Post-Herulian Wall
For personal use only. License: CC-BY-NC-ND.

a. East Face
b. West Face
8. The Post-Herulian Wall, West Face
a. In the Library of Pantainos
b. Column Drums from the Middle Stoa
9. The Post-Herulian Wall, Tower W 3
a. Interior
b. Exterior
10. The Post-Herulian Wall
a. Core of Wall, South Stretch
b. Foundationsof Tower W 5 at South End of Stoa of Attalos
11. The Post-Herulian Wall
a. Tower W 6 before Demolition, 1894
b. Tower W 7 at North End of Stoa of Attalos
12. The Post-Herulian Wall, Tower W 6 Foundationsafter Reconstruction
a. From Southwest
b. From Northwest
13. The Post-Herulian Wall, Tower W 7
a. Foundationsof Piers for SupportingUpper Storey
b. Foundationsof Piers as Excavated
14. The Post-Herulian Wall
a. Tower W 2 from Northeast
b. Christ Church Gate from West
c. Shrine of Hekate (?) from Northwest
d. Pyrgiotissa Gate. Plan
15. a. Building inscription,IG II2, 5200
b. Coins Found in Post-Herulian Wall
c. Epigram Honoring Panathenius,IG II2, 5201
16. The Theater of Dionysos
a. From the Acropolis
b. Phaidros Bema and Orchestra
17. a. Theater of Dionysos, Phaidros Inscription,IG II2, 5021
b. Athenian Agora, Great Drain
18. The Pompeion from the Southeast
19. Baths in the Agora Area. Plans
a. Bath on AreopagusSlope
b. East Bath
c. Southwest Bath, Actual State
d. Southwest Bath, Late Roman Phase
 xvi LIST OF ILLUSTRATIONS

20. Baths in the Agora Area

a. East Bath
b. West Bath
c. Southwest Bath
d. Bath in Palace of the Giants
21. a. House C, 6th-centuryBath
b. Latrine in House A
c. Latrine near Southwest Cornerof Agora
d. Latrine West of Palace of the Giants
e. Latrine at Cornerof Broad Street
f. Latrine in Palace of the Giants
© American School of Classical Studies at Athens

22. a. IndustrialSectionof Agora. Plan

b. House on Lower Slope of Areopagus
For personal use only. License: CC-BY-NC-ND.

23. a-c. Sculpturefrom House on Lower Slope of Areopagus

d. Mosaic in House in SoutheastCornerof Agora
24. a. Garden Court on AreopagusSlope
b. Garden Court in Dougga, Tunisia
25. a. Garden Court on AreopagusSlope. Plan
b. Bronze Disk, B 904. Drawing
c. Relief of Selene, S 857
26. Houses on the Areopagus.Plans
a. House A
b. House B
27. a. AreopagusHouse C. Plan
b. "House of Proclus".Plan
28. AreopagusHouse A
a. From North
b. From Northeast
29. AreopagusHouse B
a. Apsidal Room from North
b. General View looking Southwest
30. AreopagusHouse B
a. Oven
b. Room with Hearth
31. AreopagusHouse C. Air View
32. AreopagusHouse C
a. Triclinium, looking West
b. Entranceto Triclinium
33. AreopagusHouse C
a. Nymphaeum, looking West
b. Fountain House from Nymphaeum
34. AreopagusHouse C, Triclinium, Pool, and Arch to Well
35. AreopagusHouse C, SoutheasternQuarter. Air View
36. a. Fountain House next to AreopagusHouse C
b. Mosaic in "House of Proclus"
37. a. AreopagusHouse D, Mosaic Floor
b. AreopagusHouse C, Main Courtyard
38. Sculpturefrom Wells in the AreopagusHouses
a. Female Bust, S 2437, from Well Q 21:4 in House C
b. Female Bust, S 2435, from Well Q 21:4 in House C
c. Herakles, S 2438, from Well Q 21:4 in House C
d. Hermes, S 1054, from Well P 18:2 in House B
39. Sculpturefrom the AreopagusHouses
a. Nemesis, S 1055, from Well P 18:2 in House B
b. Silenus, S 2363, from House C
 LIST OF ILLUSTRATIONS xvii

c. Philosopher,S 1053, from Well P 18:2 in House B

d. Asklepios, S 1068, from House B
e. Athena, S 2337, from House C
40. Sculpturefrom Wells in AreopagusHouse C
a. Antoninus Pius, S 2436, from Well Q 21:4
b. Helios, S 2356, from Well P 21:2
c. Nike, S 2354, from Well P 21:2
d. BeardedMan, S 2355, from Well P 21:2
41. Objectsfrom the Vicinity of AreopagusHouses B and C
a. Marble Bust, S 1026
b. Marble Bust, S 1044
© American School of Classical Studies at Athens

c. Lamp, L 5324
For personal use only. License: CC-BY-NC-ND.

d, e, g, h. Fragmentsof Window Frames

f. Lamp, L 5323
42. ArchitecturalFragmentsfrom AreopagusHouse C and Vicinity
43. Areopagus House C
a, b. PerspectiveRestorations
c. Mosaic in Triclinium with Later Insertion
d. Ionic Column from Peristyle, A 3866
44. a. "House of Proclus",General View from Southeast
b. Shrine in "House of Proclus"
c. Portraitof Philosopher,AcropolisMuseum 1313
45. a. Epigram Honoring lamblichus, I 3542
b. Herm Head, S 2499, in Pit Showing Signs of Burning
c. Mosaic in ReconstructedMetroon
46. a. Round Building in Northwest Cornerof Agora
b, c. Fragmentsof Statue from Round Building, S 2531, S 2529
d. Monument Base on PanathenaicWay
e. Dedication of Building by Aetius, IG 112,5205
47. Inscriptions
48. a. Building East of Stoa of Attalos
b. Broad Street and Building over Library of Pantainos. Plan
49. Colonnadeof the Broad Street and the Building over the Libraryof Pantainos.Air View
50. a. Colonnadeof Broad Street
b. Building over Library of Pantainos, BasementRoom
51. a. Tetraconch in Library of Hadrian from East
b. Library of Hadrian with Tetraconch.Plan
52. The Palace of the Giants. Air View, May 1975
53. The Palace of the Giants. Actual-statePlan Showing Earlier Structures
54. The Palace of the Giants. RestoredPlan
55. a. The Palace of the Giants. PerspectiveRestoration
b. The Palace of the Giants. Tentative Restorationof North Facade
c. Odeion of Agrippa, North Facade after Antonine Remodeling
56. Area of the Palace of the Giants. View from the North, 1935
57. Area of the Palace of the Giants. View from the North, 1969
58. The Palace of the Giants
a, b. Triton, Front and Side Views
c, d. Giant, Front and Side Views
59. The Palace of the Giants
a. Tile Flooring in Rooms 14, 17, and 18, from Northwest
b. Sherd Bedding for Floor in Room 41, from North
60. The Palace of the Giants, SoutheastCourt Complex
a. Air View, 1975
b. From Southeast
 xviii LIST OF ILLUSTRATIONS

61. The Palace of the Giants

a. Room 40, Vault above Northeast Corner
b. Room 41, Masonry at Northeast Corner
62. The Palace of the Giants
a. Room 39 from West. Traces of Column Bases on Threshold
b. Cistern above West End of South Stoa II (Room 55) from South
63. The Palace of the Giants
a. Fragmentsof Marble RevetmentFound in Room 39
b. Fragmentsof Mosaic Found in Room 39
c. Anta Capital, A 2268, Found in Room 39
64. The Palace of the Giants, Bath soon after Excavation,from the South
© American School of Classical Studies at Athens

65. The Palace of the Giants, Bath

For personal use only. License: CC-BY-NC-ND.

a. Room 43, Remains of Hypocaust

b. Room 42, SemicircularBasin
66. The Palace of the Giants
a. Bath, South Side of Room 42
b. Marble Torso of Roman Imperial Official, S 657
67. a. Dura-Europos:Palace of Dux Ripae, beforeA.D. 256. Plan
b. Piazza Armerina,Plan of Villa, First Half of 4th Century
c. Ravenna, Sant'ApollinareNuovo. Mosaic Showing Facade of Palace, ca. A.D. 490
68. a. Bronze-castingPit near Palace of the Giants (Mattusch M)
b. Bronze-castingPit over Mint (Mattusch N)
69. a. Modern Bronze Foundry in Rome
b. Aqueductand Long Late Roman Wall
70. SoutheastCornerof the Athenian Agora. Actual-statePlan
71. a. Central Mill, Mill Room
b. Water Wheel on Thasos
c. Central Mill, Wheel Race
72. a. Sigma Table, A 3869, from AreopagusHouse C, Final Phase
b. Objectsfrom Final Phase, Palace of the Giants
c. Red Stone Tray, ST 781, from AreopagusHouse C, Final Phase
d. Fragmentof StampedPlate, P 20037, Long Late Roman Wall
e. Lamp, L 5628, from AreopagusHouse C, Final Phase
f. Lamp, L 5630, from AreopagusHouse C, Final Phase
73. a. AreopagusHouse C. Clearing Ashes from Floor of Triclinium
b. Stampson Ampullae from Palace of the Giants
c. Well Head in Courtyardof AreopagusHouse C
d. Late House next to Tholos. Plan
74. a. Destruction Debris over AreopagusHouse C
b. IG II2, 3193, Referringto Fortress
c. Laundry East of PanathenaicWay
75. a. Seventh-centuryGrave in Palace of the Giants
b. Tile Kiln near Southwest Bath
76. a. Marble Base for Olive Press in Palace of the Giants
b. Millstone and Foundationfor Crusher in Agora
c. Oil Press in Metroon
d. Olive Crusherand Press in Olympos, Attica
e. Olive Press in Olympos, Attica
f. Fragmentof Plate from Destructionof House near Tholos
 BIBLIOGRAPHY AND ABBREVIATIONS
For the ancient sourcesthe readeris referredto the Index of AncientAuthors, pp. 151-154 below.

PUBLICATIONS ONTHEATHENIAN AGORA BYTHEAMERICAN SCHOOL OFCLASSICAL STUDIES ATATHENS

AgoraI, etc. = The AthenianAgora,Resultsof the Excavationsconductedby the AmericanSchoolof ClassicalStudiesat
Athens, Princeton
© American School of Classical Studies at Athens

I. PortraitSculpture,by Evelyn B. Harrison, 1953

II. Coinsfrom the Roman throughthe VenetianPeriod, by Margaret Thompson, 1954
For personal use only. License: CC-BY-NC-ND.

III. Literaryand Epigraphical Testimonia,by R. E. Wycherley, 1957

V. Potteryof the Roman Period, Chronology,by Henry S. Robinson, 1959
VI. Terracottasand Plastic Lamps of the Roman Period, by Claireve Grandjouan,1961
VII. Lamps of the Roman Period, First to Seventh CenturyafterChrist,by Judith Perlzweig, 1961
XIV. The Agoraof Athens, by Homer A. Thompson and R. E. Wycherley, 1972
XV. Inscriptions:The Athenian Councillors,by BenjaminD. Meritt and John S. Traill, 1974
XX. The Churchof the Holy Apostles,by Alison Frantz, 1971
Hesperia, Journal of the American School of Classical Studies at Athens, 1, 1932- passim, for annual reports of the
excavationsof the Agora. Individualarticles will be cited in the appropriateplace in the text.
Supplementsto Hesperia relevantto the present volume:
Supplement IV. The Tholosof Athens and its Predecessors,by Homer A. Thompson, Baltimore 1940
SupplementVIII. CommemorativeStudiesin Honor of TheodoreLeslie Shear, Princeton 1949
SupplementXII. The Athenian ConstitutionafterSulla, by Daniel J. Geagan, Princeton 1967
Agora Guide3= The Athenian Agora,a Guide to the Excavationand Museum, 3rd edition, 1976

OTHERWORKS
Asimakopoulou-Atzaka,P., EvvrTay.AaTrwvIaAaLoXpLcrLaVLKSWVTrI71LOTW&vrS1 'EAAXXaos, II, IleXo7rovveros-
ZTepea 'EXXadba,Thessalonica 1987
CAH = CambridgeAncient History
Alan Cameron,PCPS = Alan Cameron, "The Last Days of the Academy at Athens," Proceedings of the
CambridgePhilologicalSociety,n.s. 15, 1969, pp. 7-29
Cameron,Averil, Agathias,Oxford 1970
Cameron,Averil, "Agathiasand the Sassanians,"DOP 23/24, 1969/1970, pp. 67-183
Charanis, "Observations" = P. Charanis, "Observationson the History of Greece in the Early Middle Ages,"
Balkan Studies 11, 1970, pp. 1-34
Charanis, "Significance" = P. Charanis, "The Significanceof Coins as Evidencefor the History of Athens and
Corinth in the Seventhand Eighth Centuries,"Historia 4, 1955, pp. 163-172
CIL = Corpusinscriptionumlatinarum
CSHB = CorpusscriptorumhistoriaeByzantinae, B. G. Niebuhr, ed., Bonn 1828-1897
Day = J. Day, An EconomicHistory of Athensunder Roman Domination,New York, 1942
Dunbabin, Mosaics = K. M. D. Dunbabin, The Mosaicsof Roman North Africa,Oxford 1978
EAA = Enciclopediadell'arteantica, classicae orientale,Rome 1958-
EMME I = EVpernptLov r&v McorawoVLK&Vo MvrlpdelwvT7r9 EAAa)bos,Athens 1927-1929
Frantz, DOP = A. Frantz, "From Paganism to Christianity in the Temples of Athens," DOP 19,
1965, pp. 187-205
Frantz, "Julian" = A. Frantz, "Did Julian the Apostate Rebuild the Parthenon?"AJA 83, 1979,
pp. 395-401
EVLKiO(rvve'XevO&sL = FevLKr]TvveXv p.cX^
lr&Vs k v
r1jA v 'AOfrvaLs'ApxaLoXoydtKjslEramplta
Glucker, Antiochus = J. Glucker, Antiochusand the Late Academy,Hypomnemata56, 1978
Graindor,Archontes = P. Graindor,Chronologiedes archontesathenienssous l'empire, Brussels 1922
Gregorovius, Ferdinand, Geschichteder Stadt Athen im Mittelalter, Stuttgart 1889; also published as 'Ioropla rijsv
7rotXEws'AOrlvv Kararovs
7oearovs alwvas, translatedfrom the German and revisedby S. Lambros,Athens 1904
 xx BIBLIOGRAPHY AND ABBREVIATIONS

Groag, Achaia = E. Groag, Die Reichsbeamtenvon Achaiain spdtrbmischeZeit, Budapest 1946

Groag, Diokletian = E. Groag, Die romischen Reichsbeamtenvon Achaia bis auf Diokletian, Vienna/
Leipzig 1939
Hayes = J. W. Hayes, Late Roman Pottery,London 1972
Hoepfner = W. Hoepfner, Das Pompeion und seine Nachfolgerbauten(KerameikosX), Berlin
1976,pp.176-199
IG = Inscriptionesgraecae
ILS = Inscriptioneslatinae selectae,H. Dessau, ed., Berlin 1892-1916
Jacoby, FGrH = Die Fragmente der griechischen Historiker, F. Jacoby, ed., Berlin/Leiden
1923-1954
Judeich2 = W. Judeich, Topographievon Athen, 2nd ed., Munich 1931
© American School of Classical Studies at Athens

Krautheimer,Architecture = R. K. Krautheimer,Early Christian and Byzantine Architecture,revised edition,

For personal use only. License: CC-BY-NC-ND.

Harmondsworth1975
Mattusch = C. C. Mattusch, "Bronze- and Ironworking in the Area of the Athenian Agora,"
Hesperia 46, 1977, pp. 340-379
McKay, Houses, Villasand = A. G. McKay, Houses, Villasand Palaces in the Roman World,Ithaca 1975
Palaces
Metcalf, "Coastlands" = D. M. Metcalf, "The Aegean Coastlandsunder Threat: Some Coin Hoards fromthe
Reign of Heraclius,"BSA 57, 1962, pp. 14-23
Metcalf, "SlavonicThreat" = D. M. Metcalf, "The SlavonicThreat to Greece,"Hesperia 31, 1962, pp. 134-157
Millar, "Dexippus" = F. Millar, "P. Herennius Dexippus: The Greek World and the Third-Century
Invasions,"JRS 59, 1969, pp. 12-29
Muller, FHG = FragmentahistoricorumGraecorum,C. Muller, ed., Paris 1841-1870
Paschoud,Zosime = F. Paschoud, Zosime. Histoire nouvelle. Text, translation, and commentary,Paris
1971-
Patria = Scriptoresoriginum Constantinopolitanarum, T. Preger, ed., Leipzig 1901
PG = Patrologiaecursuscompletus:seriesgraeca, J.-P. Migne, ed., Paris 1857-1904
Pharr = C. Pharr, trans., The TheodosianCode and Novels, and the SirmondianConstitu-
tions, Princeton 1952
PL = Patrologiaecursuscompletus:series latina, J.-P. Migne, ed., Paris 1884-1904
PLRE = A. H. M. Jones, J. R. Martindale, and J. Morris, The Prosopographyof the Later
Roman Empire, I, A.D. 260-395, Cambridge1971; II, A.D. 395-527, Cambridge
1980
PrincetonEncyclopedia = PrincetonEncyclopediaof ClassicalSites, R. Stillwell et al., edd., Princeton1976
RE = A. F. von Pauly, revised by G. Wissowa, Real-Encyclopiidieder classischenAlter-
tumswissenschaft,Stuttgart1894-1972
Rostovtzeffet al., = M. I. Rostovtzeff et al., Excavations at Dura-Europos,Ninth Season 1935-1936,
Dura-Europos Part III, The Palace of the Dux Ripae and the Dolicheneum,New Haven 1952
Saffrey and Westerink = Proclus: Theologieplatonicienne I, H. D. Saffrey and L. G. Westerink, edd., Paris
1968
SEG = Supplementumepigraphicumgraecum
Spiro, Corpus = M. Spiro, Critical Corpus of the Mosaic Pavements on the Greek Mainland,
Fourth/Sixth Centuries,with ArchitecturalSurveys,New York/London 1978
Swoboda,Paldste3 = K. M. Swoboda, Romische und romanische Paldste, 3rd edition, Vienna/
Cologne/Graz 1969
Thompson, "Odeion" = H. A. Thompson, "The Odeion in the Athenian Agora," Hesperia 19, 1950,
pp. 31-141
Thompson, "Twilight" = H. A. Thompson, "AthenianTwilight: A.D. 267-600," JRS 49, 1959, pp. 61-72
Thompson, "WestSide" = H. A. Thompson, "Buildings on the West Side of the Agora," Hesperia 6, 1937,
pp. 1-226
Travlos, 'Apx'Ef 1973 = J. Travlos, <<'H7rvp'roXrlrs TOV)IlapOev&vos V7Tr riv 'EpovAXovKcal 7 e7'TffTKV7
rov KaTa TroVS pvovs OV OV AV ToKpaTopos IovXtavovw, 'ApX'E 1973,
pp. 218-236
Travlos, PDA = J. Travlos, PictorialDictionaryof Athens, London 1971
Travlos, IOAXEO8bo0LtK7 = J. Travlos, H IHoXeoboLKrj TWv'AOivSOv,
'EfEArL^L Athens 1960
 BIBLIOGRAPHY AND ABBREVIATIONS xxi

Travlos and Frantz, = J. Travlos and A. Frantz, "The Church of St. Dionysios the Areopagiteand the Pal-
"St. Dionysios" lace of the Archbishop of Athens in the 16th Century," Hesperia 34, 1965,
pp. 157-202
Wachsmuth = C. Wachsmuth,Die StadtAthen im AltertumI, Leipzig 1874; II, i, 1890
Ward-Perkins,Architecture = J. B. Ward-Perkins,Roman Imperial Architecture,Penguin Books 1981
Wilson, Piazza Armerina = R. J. A. Wilson, Piazza Armerina,Austin 1983
Young, "IndustrialDistrict" = R. S. Young, "An Industrial District of Ancient Athens," Hesperia 20, 1951,
pp. 135-288
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

ABBREVIATIONSOF PERIODICALS

AA = ArchiologischerAnzeiger
AAA = 'ApxaLoXoyLKa 'AvdAcKraek'AOlv^v
AJA = AmericanJournal of Archaeology
AJP = AmericanJournal of Philology
AM = Mitteilungen des deutschenarchiiologischenInstituts,AthenischeAbteilung
Annuario = Annuariodella ScuolaArcheologicadi Atene
'ApX'Et = 'ApXaLoAoyLK7/
'Eqp?IEp[L
BCH = Bulletin de correspondancehellenique
BSA = Annual of the British Schoolat Athens
BZ = Byzantinische Zeitschrift
CP = ClassicalPhilology
CRAI = Comptesrendusde l'Academiedes inscriptionset belles lettres
AEXAr = 'ApxaoAoyoLKv AzXArTov
DOP = DumbartonOaksPapers
GRBS = Greek,Roman and Byzantine Studies
JdI = Jahrbuchdes deutschenarchdologischenInstituts
JHS = Journal of Hellenic Studies
JRS = Journal of Roman Studies
MonPiot = Monumentset memoirespublies par l'Academiedes inscriptionset belles lettres (FondationPiot)
OJh = Jahresheftedes bsterreichischenarchiologischenInstituts
OpusRom = OpusculaRomana
HIpaKTrKa = rIpaKTLKaTr7Sev 'AO7vaLs'ApxaLoXoyLKs Era&peLa
Proc.Phil.Soc. = Proceedingsof the AmericanPhilologicalSociety
REG = Revue des etudesgrecques
RM = Mitteilungen des deutschenarchdologischenInstituts,R6mischeAbteilung
SymbOsl = SymbolaeOsloenses
ZPE = Zeitschriftfiir Papyrologieund Epigraphik
 I
INVASIONS AND THEIR AFTERMATH
THE PRE-HERULIAN THREAT (A.D.254-267)
The upheavals and chaos that pervadedthe Roman Empire in the 3rd centuryof our era became
acute, as far as Greece was concerned,in 254, when a determinedbut finally unsuccessfulsiege of
Thessalonica was followed by disturbancesall over Greece.1Accordingto Zosimus, Athens was
virtually unprotectedat this time, the city walls having been allowed to fall into neglect since their
© American School of Classical Studies at Athens

destructionby Sulla in 86 B.C.2 But now the threat from the north had becomeso acute that all the
For personal use only. License: CC-BY-NC-ND.

Greeks took thought for their defenses,the Atheniansfor their own city, the Peloponnesiansfor the
Isthmus, and "a commondefensewas undertakenfor the securityof the whole country."3
Zosimus' account may be an oversimplification.Excavationon the Pnyx showed that some re-
pairs were made to the walls there in the late Hellenistic period, and a statementby Cassius Dio
(XLII.I4.I) that Caesar's legate Calenus, who, in 48 B.C., had capturedthe Piraeus because it was
unwalled, was unable to take Athens indicatesthat some part of the fortificationswas in order.4But
in the main Zosimus is probablyto be trusted.The "ValerianWall" (to distinguishit fromthe Post-
Herulian Wall erected later in the century) was solidly built on the foundationsof the Themis-
toclean Wall of which apparentlylittle survivedaboveground (PI. 2).5 The new wall was extended
eastwardto include the new quarterof the city believedto have been foundedby Hadrian,6but the
whole area was far larger than Athens, with its presentresources,could defend,and it was power-
less to resist the attackthat soon followed.

THE HERULIAN INVASION

HISTORICALACCOUNTS

Among the barbariansinvadingthe Europeancontinentduringthe latter part of the 3rd century,

the Heruli made up for an otherwiseundistinguishedrole in historyby leavingan indelibleimpres-
sion on the city of Athens.7Their appearancewas brief but savage. Suddenly emerging from the
1Zosimus, 1.29.2-3; Syncellus, 381 (CSHB, p. 715); Zonaras, xII.23. For a date of 254 for these events, see
Paschoud, Zosime I, pp. 150-151.
2
Plutarch's account of Sulla's bombardmentand his demolition of the walls (Sulla, i4) is amply supported by
archaeologicalevidence(AgoraXIV, p. 23 and passim).
3 This chapter is often taken to mean that the EmperorValerian, in whose reign the activitytook place, personally
saw to the rebuildingof the walls of Athens and the other Greek cities, in contrastto the privateenterprisewhich was
responsiblefor the Athenian Post-Herulian Wall (below). But the text indicatesno more than a general interestin the
defenses of all the provinces of the Empire: HapeXOcwv8' BaXAplavoS KOLVf yvo.t 7'poS rp v rIv o'AwvapX v
ro"jovb]v ErOEiTr0ora 7rpay.Aara ev LaBeiLvaL.... Kat ^'Avao& p,ev TrovTELXovsE7r.EAOvro i/e&L&as&,e~orE
vAAaa TOVOTO blfe'4OELpEV, aLwUevT0os fpoVTLros, 0HeAoTrOVvo Lo0L be rov '
Io'V0.ov KKOLV?bE7rapa 7raor'7s
bLereTXL?ov,
'
fvAaK7lr7s EXXAbos ear'a&-aXeia ris Xcopas eyLvero.
4 H. A.
Thompson and R. L. Scranton,"Stoasand City Walls on the Pnyx," Hesperia 12, 1943, pp. 362-372.
5 Cf. Travlos, PDA, p. 161 with references;also Thompson and Scranton,op. cit., p. 372.
6 R. E. Wycherley, The Stones of Athens, Princeton 1978, p. 23 with references;A. J. Spawforthand S. Walker,

"The World of the Panhellenion,"JRS 75, 1985, p. 93.

7 The
precise origin of the Heruli and their activities both before and after their attack on Athens lie outside the
scope of this book. Some loosenessof terminologyon the part of the historianshas led to confusionover the relationof
"Scythians","Goths",and other tribes and hence to discrepanciesin references.For the subjectin general and also the
complexities involved in the accounts of the invasion of Greece, see P. Damerau, Kaiser Claudius II. Goticus (Klio,
Beiheft n.s. 33, 1934, pp. 62-75; A. Alfoldi, "Notes 1. The Sourcesfor the Gothic Invasionsof the Years 260-270,"
 2 I. INVASIONS AND THEIR AFTERMATH

northeast,they crossedthe Black Sea, ravagedmuch of Asia Minor, and went on to Greece. Their
preciseroute varies in the accountsof differentchroniclers,but it is a matterof historicalrecordthat
they sackedAthens and were finally worsted by a band of 2000 Athenians under the leadershipof
P. Herennius Dexippus. From Attica they passed through at least the northernpart of the Pelo-
ponnese before being driven off and dispersedwestward.8They left no settlers, only a trail of fire
and devastationand, in the pages of the historians,only a vague memory.
Only Dexippus, the historian,eyewitnessand participantin the defense,gives a detailedaccount
of the capture and liberation of Athens,9which elsewhere receives scant attention except for the
(unfortunatelyprobablyapocryphal)episode of the Herulian decisionnot to burn the booksof the
© American School of Classical Studies at Athens

Athenianson the groundthat their pursuit of lettersrenderedthem less likely to make war.10It was
For personal use only. License: CC-BY-NC-ND.

Dexippus himself who led the 2000 Athenians who established themselves in the woods near
Athens to harass and finally repel the invaders.11Their action would explain why the Acropolis
seems to have escapeddamage,or at least in large measure,but they could not preventthe destruc-
tion of a great part of the lower city.12
Few physical traces of Herulian depredationshave been observedin mainland Greece to help
determinemore precisely their route up to the time they reachedAthens. More precision can be
attainedfor its later stages as archaeologicalevidencehas begun to corroboratethe sketchyliterary
accountsthat have survived.In the Peloponnesusa hoard of coins found at Sparta and a similar
discoveryat Corinth, addedto extensive remains of destructionby fire, confirmto some extent the
literary evidence for the devastationof these cities.13Inner city walls built of re-used materials,
CAH XII, 1939, pp. 721-723; idem, Studien zur Geschichte der Weltkrise des 3. Jahrhunderts nach Christus, Darm-
stadt 1967, esp. pp. 320-325; L. Schmidt, Geschichte der deutschen Stiimme bis zum Ausgang der Volkerwanderung:
Die Ostgermanen,Munich 1941 (reprint 1969), chap. VI; J. Straub, Studienzur Historia Augusta,Bern 1952, esp.
chaps. I and II. Two recent articles which came to my attentionwhile this book was in press deal with the origin and
later historyof the Heruli: Alvar Ellegard,"The Ancient Goths and the Conceptsof Tribe and Migration,"Vetenskap.
och Omviirdering(Festschriftfor Curt Weibull), 1986, pp. 49-50 (brief summary),and idem, "Whowere the Eruli?"
Scandia 53, 1987, pp. 5-34.
8 Syncellus (381 [CSHB, p. 717]) takes them across Lake Meotis to Pontus with 500 ships. He then traces their

route via Byzantium, Cyzicus, Lemnos, and Skyros, whence they went on to "burnand devastateAthens, Corinth,
Sparta, Argos and all Achaea."The Historia Augusta (Vita Gallieni, I3), Aurelius Victor (33), and Zonaras (xn.26)
prefer a more northerlycourse by way of Thrace and Macedonia. Zosimus (1.39) limits his referenceto "Greeceand
Athens itself", but recent studies suggest that this is the same event as that in 1.42-43. Cf. Straub (footnote7 above),
pp. 59 ff.; Paschoud,Zosime I, pp. 37, note 67, and 157-159. (But cf. Damerau [footnote7 above, p. 1], pp. 63-66.)
Cedrenus (259 A [CSHB, I, p. 454]) speaks of "the cities and even Athens";Ammianus Marcellinus (xxxi.5.15-17)
mentions "Epirus,Thessaly and all Greece".
9 It may be found in
Jacoby, FGrH 100, 28 and Muller, FHG III, p. 680, frag. 21.
10The first known mentionof the incidentis by the 6th-centurywriter Petrus Patricius(Muller, FHG IV, p. 196, 9,
frag. [1] under "Continuatorof Cassius Dio"), and the tale caught the imagination of other writers, e.g., Cedrenus
(259 A) and Zonaras (xII.26), apparentlyusing the same sourceas Petrus. Is this perhapsa variationof the old theme of
encouragingone's enemies in unwarlike pursuits, as found in Herodotus,I.155, where Croesusurged Cyrus to pardon
the Lydians and to commandthem to teach their sons to play the lyre and sing and dance? (Larissa Bonfantekindly
called my attentionto this parallel.)
11 Dexippus, "inwhom the writing and makingof historymomentarilycoalesced,"cameof a distinguishedAthenian
family, and his role in repulsingthe Heruli is the subjectof a definitivearticleby Fergus Millar (Millar, "Dexippus").
For Dexippus, see also E. Kapetanopoulos,"The Family of Dexippos I Hermeios,"'ApX'E4 1972, pp. 133-172 (fam-
ily connectionsonly).
12 For the view that the
Acropolisalso was taken and the Parthenonbadly damaged,see J. Travlos, 'Apx' E4 1973,
pp. 218-236 and Travlos, PDA, p. 444. For a contraryargument,cf. Frantz, "Julian."
13
Sparta, M. Karamessines-Oikonomidou,Xapo-rTrptovyd 'Avaorrdo'lovK. 'OpAvb8ovIII, Athens 1966, p. 377;
Corinth, 0. Broneer, Corinth, I, iv, The South Stoa and Its Roman Successors,Princeton 1952, p. 134, and for more
 THE HERULIAN INVASION 3

similar to the Post-Herulian Wall at Athens (pp. 5-11 below), have been found at Olympia,
Eleusis, Beroea, and elsewhere.
As long as the only availableevidencewas literarythe damageinflictedby the Heruli was under-
estimatedby scholars.'4 It was only in the courseof the excavationof the Agorathat the impactwas
shown to have been catastrophic.From the beginning of the excavations,it became graduallyap-
parent that a majorconflagrationhad consumeda great part of the lower city, so that A.D. 267, the
year in which the city was sacked,15defines clearly the end of the ancient city and its transitionto
the status of a minor provincialtown,'6 a characterwhich it retainedall throughthe Middle Ages,
with life disruptedto such an extent that the old patterncouldneverbe resumed.17That the damage
© American School of Classical Studies at Athens

was not confinedto the Agora has since been shown by the work in the Kerameikosand at scattered
points elsewhere in the city.18The destructionaroundthe Dipylon is reflectedin the suddencessa-
For personal use only. License: CC-BY-NC-ND.

tion of the lamp industryin the latter part of the 3rd century.

DAMAGE
PHYSICAL
Until 267 the appearance of Athens had probably changed little since its transformationby
Hadrian and the Antonines, when the city reachedits architecturalpeak (Pls. 2, 3). To the great
Classical and Hellenistic monumentshad been added the productsof the Roman period, e.g., the
Tower of the Winds, the Market of Caesar and Augustus, the Odeion of Agrippa, the Temple of
Ares in its new location in the Agora, and later, the buildings of Hadrian and the Antonines,e.g.,
the Libraryof Hadrian, the Odeion of HerodesAtticus, and the Southeastand SouthwestTemples
in the Agora.19In Athe allbuildings,
importAgorat the
buildings,
impexceportiontant with
withof the those on south
side of the square, were still standingand in a good state of preservation.20But in this one raid the
Agora lost, temporarilyat least, all of its monumentalcharacteras a civiccenter,and althoughsome

recent discoveries,C. K. Williams, II and 0. H. Zervos, "Corinth, 1981: East of theheTheater," Hesperia 51, 1982,
pp. 118, 132-134; "Corinth,1982: East of the Theater,"Hesperia 52, 1983, pp. 23-24.
14
E.g., Wachsmuth, I, pp. 707-724; Judeich2,p. 104.
15
Whether the invasion took place during the last months of 267 or the first of 268 is uncertainowing to the con-
fusion of the sources regardingthe actual number of invasions by "Goths"during the decade 260-270. Alfoldi, in the
two referencescited (footnote 7 above, p. 1), gives a summaryof the known facts. Paschoud (Zosime I, pp. 157-159)
opts for a date in 268 but still in the reign of Gallienus. All commentatorsagree that the invasiontook place towardthe
end of Gallienus' reign. It seems certain that the Heruli launchedtheir invasion of Asia Minor in 267 whether or not
they reachedAthens in that year. I have thereforeadopted267 as the startingpoint.
16 Cf. T. L. Shear,Jr., "Athens:From City State to ProvincialTown," Hesperia 50, 1981, pp. 356-377. Technically

speaking, it reached that status under Augustus when Greece became part of the Roman Empire as the Provinceof
Achaea, but the transitionwas almost imperceptiblethrough the Antonineperiod.
17 On this whole
subjectsee Thompson, "Twilight"and Agora XIV, pp. 208-210. Also Agora Guide3,pp. 32-34.
For the individualbuildings, Thompson, "WestSide,"pp. 76, 114, 171, 195; Hesperia, Suppl. IV, pp. 101, 136.
18 For Athens in general, Travlos, PDA, chap. VI. For the Kerameikos,includingthe Dipylon and Pompeion,Agora
VII, passim and K. Kubler, "Die spaten Einbautenim griechischenPompeion,"AM 56, 1931, pp. 80 ff.; also Hoepf-
ner, pp. 174-175.
19 For the new dating of the Agora temples, see W. B. Dinsmoor,Jr., "AnchoringTwo Floating Temples,"Hesperia

51, 1982,pp.410-452.
20 The south side had incurredthe heaviest damagein 86 B.C. Most of the buildingswere destroyed,and for the next
200 years the area was given over to small industrial establishments.In the early 2nd century, however, these were
abolished,and at least two of the majorbuildings,the Law Court on the south side and the East Building,were rebuilt,
and the wall of South Stoa II was repaired to form the southern limit of the Agora and to support the "Antonine"
Aqueduct.AgoraXIV, p. 71, also Hesperia 28,1959, p. 97 and Hesperia 29,1960, p. 362 (Thompson);AgoraGuide3,
p.162.
 4 I. INVASIONS AND THEIR AFTERMATH

of its functionswere restored,it sufferedencroachmentsfrom small industries,21most of these after

a considerableperiod of desolation.22
Of the buildingson the south and east sides of the Agora not one remainedintact, and to most the
damagewas so serious as to make them of little use except as a quarry.23The centerof destruction
was the Odeion, whose huge ceiling beams providedample fuel. The mass of burned debrisover-
lying the site at the time of excavationleft no doubtof the cause of the ruin, and coins sealed under
the debris,runningdown to the reign of Gallienus (A.D. 253-268), providedevidencefor the date.24
After the Odeion the heaviest concentrationof damage was to the buildings of the south and east
sides:the Middle Stoa, the Libraryof Pantainos,and the Stoa of Attalos,which showed the greatest
© American School of Classical Studies at Athens

evidence of burning, particularlyat the south end. The Stoa, although not completelydestroyed,
For personal use only. License: CC-BY-NC-ND.

was renderedunusable as an independentbuilding.25

Along the west side of the Agora the damage was confinedto the southern half, with only the
Metroon and Bouleuteriontotally ruined at this time.26But the Tholos sufferedenough to neces-
sitate surroundingthe lower part of the walls with a heavyring of masonry;it was rebuiltand, with
the Stoa of Zeus, the Temple of Apollo Patrobs,and the Stoa Basileios,surviveduntil the end of the
4th century.27
The absence of debris from the Herulian destructionin the north half of the west side of the
Agora leads to speculationthat the invadersenteredAthens in at least two bands,one stormingthe
Dipylon which providedample evidence of fighting around it. From there it advancedalong the
PanathenaicWay, leaving sporadictracks,until it reachedthe centerof the Agora.28The destruc-
tion on the south side of the Agora and the north slopes of the Areopagus, as well as that of the
Bouleuterionand Metroon and the partial destructionof the Tholos, may have been the work of a
force approachingby a road that led from the Piraeus Gate to the southwestcornerof the Agora.A
third band might have enteredthrough a gate on the Pnyx, the "St. Demetrios Gate",whose foun-
dationswere overlaidby a thick accumulationof Herulian debris.29This would have providedeasy
accessto the south side of the Acropolis,which also sufferedheavily in the raid.
The evidencefor the north side is still incomplete.The porch of the large Roman basilica at the
northeastcorner,formerlycalledthe Northeast Stoa, is known to have been burned.30The building
to which it was attached was completely destroyed,the work begun by the Heruli having been

21
For these industries,see pp. 79-80 below.
22 The
word "Agora"by this time is a misnomer,but the term is convenientand will be retainedas denotinga geo-
graphic area, even thoughmost of its original functionshad ceased.
23AgoraXIV, pp. 208-209.
24
Hesperia 5, 1936, p. 9 (Shear, Sr.); Thompson, "Odeion,"p. 134.
25Hesperia 19, 1950, p. 319 (Thompson).
26 Roof tiles of the Metroon
(identifiedby brickstamps) were found in contextsof the 3rd century,an epistyle block
was incorporatedinto a modernfoundationin the line of the Post-Herulian Wall, and potteryand 13 coins of the 2nd
and 3rd centuries coveredthe floor (Thompson, "WestSide,"p. 195 and excavationnotes). For the destructionof the
whole west side, ibid. (see under the individualbuildings), also for their later history, and below, p. 25. The fill inside
the Bouleuterionhad all been removedby the Greek and German excavatorsin the 19th century,but in the courtyard
in front of the building was a considerabledepositof 3rd-centurypotteryand lamps ("WestSide,"p. 171).
27 Below,
pp. 53-55.
28 A concentrateddeposit of coins, metal objects,and pottery along with much ash in the north end of the Stoa of
Attalos, which later becameTower W 7 in the Post-Herulian Wall, was one of the key pieces of evidencefor Herulian
destructionin the Agora. Of 105 coins found there the latest (46) were from the reign of Gallienus.
29Thompson and Scranton(footnote4 above, p. 1), p. 370.
30Hesperia 20, 1951, pp. 53-56 (Thompson).
 THE HERULIANINVASION 5

completedby pillagers.31At the northwestcornerthe Stoa Basileios, along with its neighboracross
the PanathenaicWay, the Stoa Poikile, evidentlyescapedwith only minor damage,as the attackers
turned southward into the Agora proper. The Poikile, at least, must have remained in a usable
conditionuntil well into the 5th century.32
Outside the Agora, as the PanathenaicWay runs up toward the Acropolis,the buildings border-
ing it (the SoutheastBuilding and the Eleusinion) were all destroyedor damaged,and the devasta-
tion continued along the lower slopes of the Acropolis. The extent of damage to the Library of
Hadrian cannotbe determinedat present.It has long been evidentthat a numberof columnbases of
late workmanship in the east colonnade of the court were replacements necessitated by some
© American School of Classical Studies at Athens

catastrophe.33More recentlywork on the south side of the library has uncoveredthree more bases
of the same series;also, "displacedmaterialfromthe northand east internalperistylewas identified
For personal use only. License: CC-BY-NC-ND.

and replaced.5"34 It may be concludedthat the library was out of commissionfor a considerable
period, probablyuntil the early 5th century.35
THE POST-HERULIAN
WALL
A few buildings had escaped damage almost entirely, and others might have been repaired,but
the necessity of reducing the circuit of the defensesoutweighed other considerations.Accordingly,
the more realisticsolutionwas adoptedof systematicallystrippingthe damagedbuildingsand using
the abundant building material salvaged from them to erect an inner fortificationwall to enclose
only that part of the city lying immediatelyto the north of the Acropolis(Pls. 4, 5, 7:a).36A prece-
dent for using materialfrom earlier buildingsfor the constructionof new defenseshad alreadybeen
set by the Athenians after the Persian sack of 480 B.C.,as told by Thucydides (1.90.3, 93.1-2) and
substantiatedby archaeologicalevidence.Also, in 69 B.C.,after the pirate Athenodoroshad ravaged
the island of Delos, the Roman admiralTriarius erecteda wall arounda much restrictedarea of the
town in order to protect at least the sanctuary from a new attack. He, too, used as construction
material the ruins left by the invader.37The closest analogy in time and place is to be found at
Olympia, where a similar wall was built to protect the Temple of Zeus and a small area to the
south. The wall, originally attributedto the 6th century after Christ, has now been shown to date
from the secondhalf of the 3rd.38
31
Hesperia 40, 1971, pp. 261-265 (Shear, Jr.).
32
Below, pp. 55-56.
33J. Travlos, ?<'Aao-Ka4a e' v Tj BLXSALOflTOy'o? VA8ptavdo?,IIpaKTLKC 1950, p. 55; A. Frantz, "Honors to a
Librarian,"Hesperia 35, 1966, p. 379 and pl. 91.
34 H.
Catling, ArchaeologicalReports (JHS) 1982-1983, pp. 8-9.
35Below, p. 63.
36 Not all the material found its way into the wall. During the rest of the 3rd centurythe Agora servedas a quarryfor

privateindividuals,includingthe squatterswho settledin the ruins, especiallyalong the west side. For the historyof the
excavationof the wall, the scholarshiprelating to it, and a detaileddescription,see Appendix.
37 Phlegon of Tralles (2nd century after Christ) is the source of informationfor the attack and the building of new
fortifications(Photius, Bibl., Cod. 97 [Henry, II, p. 65]). His accountwas confirmedwhen the French excavatorsof
Delos discoveredthe wall itself, composedof much re-used materialincludingarchitecturalmembers,statues,and four
inscriptions(P. Bruneau and J. Ducat, Guide de Delos, 3rd ed., Paris 1983, p. 198).
38 A. Mallwitz, Olympiaund seine Bauten, Munich 1972, pp. 110-112; OlympiaBerichtVI, 1953-1954; 1954-1955

(1958), pp. 5-6. One may ventureto questionthe opinion that the wall was built in anticipation,ratherthan as a result,
of the Herulian attack, out of material taken from buildings deliberatelydemolishedfor the purpose. The apparently
haphazardchoiceof structuresand the piecemealapproach(e.g., only the columnsof the Leonidaion,only the southwall
and the colonnadeof the Bouleuterion)are moreeasily explainedby the chancesof war than by deliberateplanning,and
one would expect the sanctity of the Pelopion to outweigh the meager supply of building materialthat it might afford.
 6 I. INVASIONS AND THEIR AFTERMATH

During the centuryand more since the wall was discoveredit has been known under a varietyof
names, being attributedat one time or another to Valerian (A.D. 253-260), Justinian (A.D. 527-
565), and the Florentine Dukes of Athens, the Acciaiuoli (A.D. 1402-1435). Now investigationof
differentparts of the wall in the area of the Agora has establishedbeyonddoubt that at least part
was underconstructionduringthe last quarterof the 3rd century.One of the strikingfeaturesof the
destructiondebrisleft by the Heruli is the numismaticevidence,with considerablenumbersof coins
leading up to, but sharply cut off with, issues of Gallienus (A.D. 253-268). A similar situation
prevails in the case of the wall, with the exception that here the coins stop abruptly with Probus
(A.D. 276-282). One of the most convincingbits of evidenceis a pocketfulof 16 coins droppedon the
mortarbeddingfor the wall when it was still wet, in a sectionof the west flank on the slope of the
© American School of Classical Studies at Athens

Acropolis (PI. 15:b). The earliest were ten coins of Aurelian (A.D. 270-275), followed by two of
For personal use only. License: CC-BY-NC-ND.

Severina (A.D. 274-275) and one each of Florian (A.D. 276) and Probus. Another coin of Probus
was found in a footing trench in the same part of the wall.39It is apparent,therefore,that the wall
was built as a directresult of the invasionand was probablybegun during Probus'reign. A delay of
ten to fifteen years beforefear was translatedinto action might seem excessive.The necessaryprod
was perhapsfurnishedby anotherthreat,this time from Franksto whom Probushad grantedlands
to settle. Some revoltedand, having an ample supply of ships, spreaddisasterover all Greece.40
The new wall enclosedthree sides of a trapezoidalarea northof the Acropolis,the Acropolisitself
formingthe southernlimit (PI. 5). Only the west flank and a little of the northlie within the area of
the Agora.The irregularshape of the enclosurewas dictatedby the desireto economizein materials
and labor by utilizing, so far as possible,the remainsof earlier structuresand to respectimportant
existing roads.Thus, the west flank of the wall took its line fromthe ruinedbuildingsalong the east
side of the Agora, while the north followeda slightly southerlycoursefromthe northend of the Stoa
of Attalos to follow the line of the ancient east-west street and to make use of some still standing
walls of two buildings, the Libraryof Hadrian and another,possibly the Pantheonor, more prob-
ably, a basilica for meetingsof the PanhellenicUnion.41
The courseof the west flank has been preservedfor its entire length almostwithout interruption.
The south end was solidly built against the rock of the Acropolis. The wall then descendedthe

(For a plan and a photograph of a model, see N. Papachatzes, Ilavo-aviov 'EXAAdosHlEprTfyi^': Meo-o'rVLaKga,
'
HLaKca,Athens 1979, p. 241.) Anotherconsiderationis the close resemblanceof the wall at Olympia to that in Athens,
which can be dated during or very shortly after the reign of Probus (A.D. 276-282), below, p. 6. Both use the same
techniques:two outer faces composedof re-usedarchitecturalmembersseparatedby a thick coreof rubblemadeof field
stones and smaller fragmentsof ancient marbles.The masonryof the outer faces is carefullylaid with some eye to the
whic h might not have been a matterof concernin the face of very real and presentdanger.
estheticeffect,
The widespreadfear of barbarianinvasionis well illustratedby walls of the same naturefoundscatteredthroughthe
whole empire, from Spain to Asia Minor, which are only now beginning to be recorded,and many of which are still
unrecorded.For a series of walls of the middleof the 3rd centuryin Spain and France,see I. A. Richmond,"FiveTown
Walls in Hispania Citerior,"JRS 21, 1931, pp. 86-100. For Anemurium, E. Alfoldi-Rosenbaum,"Matronianus,
Comes Isauriae:An Inscriptionfrom the Sea Wall of Anemurium,"Phoenix 26, 1972, pp. 183-186 and C. P. Jones,
"The Inscriptionfrom the Sea Wall at Anemurium,"Phoenix 26, 1972, pp. 396-399. Also, L. Robert, "Epigrammes
relativesa des gouverneurs,"Hellenica 4,1948, pp. 60 ff. for epigramscommemoratingthe buildingof fortificationsin
the east.
39Hesperia 7, 1938, p. 332 (Shear, Sr.).
40Zosimus, 1.71.2: "Kai payKxwv rw
T /3a-tLAL pora-corV7WVKa KO TVxOVTWV oKi/o-ews
sAolpa rTS awoOr-aOa, 7rAooWv
ev7rop71orara,lrjv E'EXXaba -vverapafev i7rao-av. A single coin of Maximian (A.D. 286-305), found in a footing
trenchof the wall in one of the lower stretches,may lower slightly the date of the wall, or it may indicatenormalprog-
ress in what may have been a time-consumingoperation.
41 Travlos, PDA,
pp. 439-443 (Pantheon);Spawforthand Walker (footnote6 above,p. 1), p. 97 (Panhellenion).
 THE HERULIANINVASION 7

slope, following the east side of the Panathenaic Way, where the builders were able to take ad-
vantage successivelyof the foundationsof the peribolosof the Eleusinion and the stylobatesof the
Southeast Building and the Library of Pantainos. All these buildings were cleared of whatever
remainedof walls and columns, and the fortificationwall was laid on the firm and level surfaceof
their floors.The column drumsthemselvestook their place with the architecturalmembersof other
structuresas materialfor the wall (PI. 8). At the north end of the Libraryof Pantainosthe courseof
the wall changed slightly to utilize the wall of the shop fronts of the Stoa of Attalos by taking an
eastwardjog with the south wall of the Stoa. The northend of the Stoa formedthe northwestcorner
of the enceinte.
© American School of Classical Studies at Athens

The north flank takes off at an acute angle from the northeastcornerof the Stoa of Attalos. Its
For personal use only. License: CC-BY-NC-ND.

line has been establishedat intermittentpoints, includingthe east end where its projectedend forms
a right angle with the northernmostpreservedsection of the east wall, thus determiningthe north-
east cornerof the enceinte.42
The course of the east wall is the most problematicsince most of it lies under the thickly settled
district of the Plaka (PI. 1). Two sections were uncoveredby the Greek ArchaeologicalSociety in
1861 and demolished by the excavatorsin order to retrieve the many inscriptionsand pieces of
sculpturebuilt into them.43The wall undoubtedlyendednear the northeastcornerof the Acropolis.

Towersin the West Flank

The wall on the west side of the enceinte was pierced by three gates44and protectedby seven
rectangulartowers composed,like the curtain,of re-used material.
Tower W 1 stoodalmost equidistantfromthe south end of the wall and the Hypapanti Gate, thus
coveringthe upper slopes of the Acropolis.
Tower W 2 protectedthe Christ Church Gate from the south side (PI. 14:a).
Tower W 3 stood at the southwest corner of the Library of Pantainos and the north end of the
Southeast Building (PI. 9). It protected no gate but instead completely blocked a narrow street
which had led into the Agora from the east between the two buildings. In orderto providedrainage
for the surfacewater that must have previouslyflowedfreely from east to west overthe Panathenaic
Way, a drain was carriedunder the west foundationsof the Southeast Building and thence under
the tower itself to communicatewith the easternbranchof the Great Drain. This tower, like Tow-
ers W 1 and W 2, projectedinto the PanathenaicWay and blockedmost of its width at this point.45
It was demolishedin 1959, leaving enough of its foundationsto indicateits dimensions.46
Towers W 4 and W 5 flankedthe PyrgiotissaGate. W 4 was preservedto a height of 3.45 m. W 5
incorporatedthe anta of the south wall of the Stoa of Attalos along with the first of the front col-
umns and a stretchof the steps (PI. 1O:b).In Turkish times the Churchof the Pyrgiotissa(Virgin of

42 For
details, see Appendix, pp. 136-137.
43See Appendix, pp. 137-138.
44
In later times each of the gates was incorporatedin or overlaidby a church. To avoid confusionwith the other
flanks the gates will be given the additionaldesignationof the names of these churches:South Gate = Hypapanti Gate;
Middle Gate = Christ Church Gate; North Gate = PyrgiotissaGate. The towers will be designatedby flank starting
at the south end of the west flank:W 1, etc.
45 Below, p. 29.
46 This tower was particularly rich in fine early marbles, e.g., the series of Ionic columns with painted capitals

(A 2972 and A 2973) and an over-life-size, flamboyantlydrapedfemale statue, perhapsAphrodite(S 1882). For these
and other architectural members and pieces of sculpture found with them, see Hesperia 29, 1960, pp. 350-359
(Thompson) and E. B. Harrison, "New Sculpturefrom the Athenian Agora, 1959," Hesperia 29, 1960, pp. 373-378.
 8 I. INVASIONS AND THEIR AFTERMATH

the Tower) was built into Tower W 5, the rubble filling, if any, having been removed for the
purpose.47Both the churchand most of the tower were demolishedin 1859.48
Tower W 6 stood in the middle of the Stoa of Attalos, in front of shops X and XII. Plate 11:a
shows the tower as it still stoodjust before its demolition in 1900. In Plate 12:a the foundations
appear after being clearedprior to the reconstructionof the Stoa.
Tower W 7, the largest, at the northwestcornerof the circuit,was built into the northend of the
Stoa of Attalos (PI. 11:b). Four piers were set down fairly symmetricallyinside the outer walls to a
depth of 3.60 to 4.45 m. These presumablysupportedan upper storey and perhaps a platformfor
artillery (PI. 13).
© American School of Classical Studies at Athens

Gates in the WestFlank

For personal use only. License: CC-BY-NC-ND.

The South, or Hypapanti, Gate stoodmidway between Towers W 1 and W 2 to providean outlet
for a narrow road leading from the inner city to the PanathenaicWay. The street and gate appear
to have been in continuoususe until Turkish times, when they were blockedby the Church of the
Hypapanti which was built over them. Door frameswith moldingscharacteristicof the 6th century
were found in the church,indicatingprobablerepairsto the gate at that time.
The Middle, or Christ Church Gate, just below the Eleusinion, was protectedby a single tower
(W 2) and allowed passage of the South Road which from earliest times had skirted the lowest
slopes of the Areopagus(Pl. 14:a,b). By settingthe gate at an angle the builderswere able to respect
the road and conformto its course.The gate had a checkeredhistorywhich was tied in with that of a
small sanctuarytuckedin the angle between Tower W 2 and the gate (PI. 14:c).The sanctuarywas
probablyconstructedin the 2nd century,perhapsas an open-air shrine of Hecate.49It had already
undergoneone remodelingwhen the wall was built. The sanctuarywas spared by the builders of
the wall, but Tower W 2 was permittedto encroachslightly on the southeastcorner.Sometimein
the secondhalf of the 4th centurythe shrine fell into disuse;it was strippedof its sacredtokens,the
walls were demolished, and not long after, terracottawater channels were laid over it. The late
survivalof this pagan sanctuaryaddsanotherdimensionto the pictureof Athens as still a pagan city
well along into the 4th century. This was a much frequentedroute, and to preserve the shrine
required an effort which cannot be observedin any other part of the wall. In general, the ruling
principlewas apparentlyto obtain constructionmaterialat any cost.
The North (Pyrgiotissa) Gate gave access to the road leading past the south end of the Stoa of
Attalos to the Roman Market (PI. 14:d). It had been a main thoroughfaresince trafficbetween the
two Agoras had been facilitated by the demolition of the south stair of the Stoa, ca. A.D. 100. As
befittedits importance,the gate was flankedby the two most massivetowers (W4 and W5) except
for W7 at the corner.

Gates in the North and East Flanks

In these flanks the position of only one gate (Krystalliotissa)can be determinedwith certainty,
not far from the east end of the north wall. In its presentstate, with an elaboratelydecorateddoor
frame, it undoubtedlydates fromthe time of Justinian,50but this is very likely a repairof an earlier
gate in the same place.51
47Below, p. 133.
48 G. A. Sotiriou, EMMEI,
p. 110.
49
Hesperia 39, 1960, p. 333 (Thompson);Agora XIV, p. 169.
50G. A. Sotiriou, EMME I, pp. 28-30; also Appendix, p. 140.
51
For the probablepositionsof the other gates, see Appendix, pp. 138-141.
 THE HERULIAN INVASION 9

Building Inscriptions
An inscription found in the wall near the East (St. Demetrios) Gate52gives the name of the
builder as Illyrius. There can be little doubtthat this was Claudius (Leonticus)Illyrius, Proconsul
of Achaea, who was honored by the city with two statues, the bases of which carry identical in-
scriptions giving the names and offices of himself, his father (Claudius) Teres (formerly called
Tebens),53eponymousarchon and Areopagite,and his grandfatherCn. Claudius Leonticus, Pro-
consul of Achaea and Corrector.54
A correspondingepigram was found in the vicinity of the Pyrgiotissa Gate in the west flank
(PI. 15:a).55Although the name of the builder is missing and the phraseology is different, the
© American School of Classical Studies at Athens

similarity of the subject,in which a comparisonis made between the ease with which the walls of
For personal use only. License: CC-BY-NC-ND.

Thebes were erected by Amphion's lyre and the unaided effort required in the present circum-
stances, makes it likely that the west flank also was built by Illyrius. That being the case, it is
probable that he was responsible for the whole wall and that one of the gates in the north flank
would have borne yet another variation on the same theme.56There were apparentlytwo distichs
side by side on 5200: one complete(a); on a now missing piece (b), only the last word of each line.57
Soon after its discoverythe wall became fixed in the minds of the scholars concernedwith it as
part of Valerian's program for fortifying the cities,58thereby providinga date of 250-260 for the
approximatelimits of the careerof Claudius Illyrius. This belief was bolsteredby the fact that the
man chargedwith setting up the statues, M. Iunius Minucianus, was said by the Suda to have lived
in the time of Gallienus,59and to have been the son of Nicagoras,a sophist,who lived in the time of
Philip the Arab (A.D. 244-249),6?0giving Minucianus a precociousnessof dubiouscredibility.That
it was the date of Minucianus that was in erroris made clear by the activitiesof his son, Nicagoras
(II), a priest of Eleusis who accompaniedConstantine to Egypt in 326, probably in an official

52
IG II2, 5199; see footnote 56 below for the text. The text is said to be from a drawing by Cyriac of Ancona, but
E. W. Bodnar has shown that this is not the case. The stone was rediscoveredin 1861 and another small piece was
found on the same site in 1949 (Cyriacus of Ancona and Athens [Collection Latomus XLIII], Brussels 1960,
pp. 175-177).
53For the correctedspelling, see P. Graindor,"Inscriptionsd'epoqueimperiale,"BCH 38, 1914,
pp. 386-388. The
name is otherwise unknown.
54
IG II2, 3689, 3690. Cf. Groag, Diokletian, cols. 88-89. Claudius Illyrius was also responsiblefor building the
walls of Lapethosin Cyprus.J. and L. Robert ("Bulletinepigraphique,"REG 64,1951, pp. 206-208), commentingon
an inscription published by T. B. Mitford ("Some New Inscriptionsfrom Early Christian Cyprus,"Byzantion 20,
1950, pp. 136-139), revised the date proposedby Mitford, and G. W. Bowersockshowed that the subjectwas not a
senator from Cyprus ("Roman Senators from the Near East: Syria, Judaea, Arabia, Mesopotamia,"Tituli 5, 1982,
pp. 669-670).
55 IG II2, 5200.
56 The text of the two inscriptionsruns as follows:
IG II12,5199: 'AftL(wvuoovcraLs KtOap7Js'o[r]7To-'E<7Tr>O),8T?S rcXca- vvv b' ('' 'as 7raTpLos 'IAAvpLo,
a5vXoyov pofa
,oovav 07rcov. rTc KaL bOKEOVO-t
axK/?,T pjELVv7erpara lravTa TE<X>vas.

IG II2, 5200: a b
ov ra'e OeXof
.eAWrs'ApOLovI' ipa[pe L--
#opp.ly{] TEOVS
o0Vi KVK Tcrelas XELtPOs
A[?pE? 3la]. aperTa[]
57For commentary,see J. and L. Robert, "Bulletinepigraphique,"REG 79, 1966, pp. 741-742.
58Above, p. 6. The opinion is still current.Cf. B. E. Thomasson,LaterculiPraesidumI, Goteborg 1984, col. 196.
59Suda, M 1087. Mommsen was the first to suggest this identification
(quoted in annual report, AeAr 5, 1889,
p. 133).
60 Suda, N
373.
 10 I. INVASIONS AND THEIR AFTERMATH

capacity.61If Nicagoras was from 40 to 60 years old at the time, he would have been born
ca. 265-285, and his father'scareerwould have been from 15 to 35 years behindhim. But with the
wall now firmly establishedin the reign of Probus (A.D. 276-282) or slightly later, and the Suda's
dating set aside, Minucianus' career falls at a reasonable time, ca. 255-295. It also makes un-
necessaryan adjustmentin the date of the archonshipof Claudius Illyrius' father,Teres or Tebens,
given in RE II, col. 598 as ca. 258/9 (between239 and 261). Graindornotedthe unlikelihoodof the
father holding the archonship when his son was already eligible for the proconsulate,and so he
moved Teres back into the second quarter of the century.62The new dating of the wall, however,
allows Teres' archonshipto remain in its allotedplace, and so the officesof fatherand son fall into a
© American School of Classical Studies at Athens

normal chronologicalrelationship.Also, the added 20 to 30 years give a more reasonablespan for

For personal use only. License: CC-BY-NC-ND.

the careersof the three generations.Cn. Claudius Leonticus,the grandfather,is generallyagreedto

have flourishedin the first quarterof the 3rd century.63The old dating would have allowed barely
40-45 years for the careersof grandfather,son, and grandson;the new gives a more realisticperiod
of 65-75 years.
Another inscription relating to the fortifications (PI. 15:c) records repairs to a "dilapidated
tower".64Inscribedon a tabella ansata,to be insertedinto masonry,it reads as follows:
ro<v 'T>>p6vaKoOf.LOVovra 7rvp/yov cre'vosev iroAelzooiv
vvv OVTW
O7rVpyWOfEv Xa~a//rpaX'Lp?7lyefLOvoo
OLKOev' Vopaews Hfava/oq[v]lov eopxov avbp&v

The tower,a bulwarkin warsbutrecentlyin disrepair,hasnowbeen

truly fortifiedas a towerby the glorioushandof Panathenius,out-
standingamongmen,in his wisdomandat his ownexpense.
A precise date of A.D. 267 has been assigned to the inscriptionon the basis of a passage in the
Historia Augustastating that Gallienus charged"the Byzantines"Cleodamusand Athenaeuswith
restoring and fortifying the cities and that a battle was fought around Pontus in which the bar-
barianswere defeatedby the Byzantineleaders.65Panatheniusis interpretedas an enhancementof
the name Athenaeus in recognitionof his deeds. Zonaras, presumablyreferringto the same event
but omitting Gallienus' directive about the fortifications,speaks of "Cleodemosan Athenian"
be
(KAeo677losv 'AOvTaZtos- av?rp) as victorious in a naval engagementwith barbarians(after the
Herulian sack),66but this has been generallyregardedas an erroron the part of Zonaras.67

See 0. Schissel, "Die Familie des Minukianos,"Klio 21,1927, pp. 361-373. See also K. Clinton, The Eleusinian
61

Mysteries (Transactionsof the AmericanPhilosophicalSociety64, 3), Philadelphia 1974, pp. 66-67.

62
Graindor,Archontes,pp. 262-263.
63 Groag,
Diokletian,cols. 88-89.
II2, 5201. First notedby S. Ricci ("Miscellaneaepigrafica,"Monumentiantichi 2, 1893, col. 253) as foundin a
64 IG

house in Hermes Street. The two pivot holes indicate at least two periods of re-use. Another translationof aKKO,Up.o
would be "unadorned",but the readingadoptedhere seems betterto fit the comparisonbetweenthe formerand present
states. For anotherexample in a similar context see IG
II2, 3193 (see below, p. 118).
65Vita Gallieni,I3.6. "Quibus compertis Gallienus Cleodamum et Athenaeum Byzantios instaurandis urbibus
muniendisquepraefecit,pugnatumqueest circa Pontum, et a Byzantiisducibusvicti sunt barbari."
66 XII.26.
A. Alfoldi, who alone raised doubts about two commanders(CAH XII [footnote 7 above, p. 1], p. 722), now
67

restatesthis view more positivelyin "RedeuntSaturniaregna, VII," Chiron9, 1979, p. 594. He still, however,equates
Cleodamuswith the "Panathenius"of the inscription(and also with the Cleodamusto whom, with Porphyry,Longinus
had dedicateda work;for Longinus and the dedication,see Porphyry,Plot. xvII.83). Various aspectsof the inscription,
the
in addition to the identificationof "Panathenius",have caught the attentionof scholars:the rank of the Tyeqcoov,
exact meaning ofev pabiws, etc. Cf. A. von Premerstein, "Zu den Inschriften der Ostgermanen,"Zeitschriftfur
 THE HERULIAN INVASION 11

Accordingto the Historia Augusta the instructionsto fortify the cities and the victory over the
Scythians at sea precededthe attack by the Goths (Herulians) on "Cyzicus,Asia and then all of
Achaea",including Athens. Whether there was time within this period to carry out all the actual
constructionwork on the fortificationsis uncertain.If our inscriptiondoes relate to these eventsand
"Panathenius"is one with Athenaeus, the rebuildingof the tower should have been completednot
long after, while the memoryof "Panathenius"was still green. In any case, the inscriptioncan have
no connectionwith the Post-Herulian Wall, which incorporatedno earlier fortificationsand was
not built until the reign of Probus (A.D. 276-282) or slightly after. The tower itself might have been
originally part of the Themistoclean Wall, repaired in the course of the building of the Wall of
© American School of Classical Studies at Athens

Valerian along the same line, or one built expressly for the Wall of Valerian and damagedduring
the Herulian attack.
For personal use only. License: CC-BY-NC-ND.

A concrete example of the repair of towers at about that time is to be found in the extensive
reconditioningof the fortificationsof the Pnyx, especially one of the towers of the gate at the east
end, probablyin the secondhalf of the 3rd century.68

G. Barbieri,L'albosena-
deutschesAltertum60, 1923,pp. 77 ff.;Groag,Diokletian,pp. 96-97, withotherreferences;
torioda SettimioSeveroa Carino,Rome 1952,AppendixIII, "Praesesed 7ye.cWv", p. 577;L. Robert(footnote57
above, p. 9), p. 741; L. Robert, CRAI 1968, p. 423, note 6; also in J. des Gagniers et al., Laodicee du Lycos, Le
Nymphee, Quebec/Paris 1969, pp. 339-351; E. Manni, TrebellioPolliore.Le Vitedi Valerianoe di Gallieno,Palermo
1969, p. 102, note to Vita Gallieni, 13; T. D. Barnes, "Some Persons in the Historia Augusta,"Phoenix 26, 1972,
p. 148, and The Sourcesof the Historia Augusta(CollectionLatomusCLV), Brussels 1978, pp. 68, 73; L. de Blois, The
PolicyoftheEmperorGallienus,Leiden1976,p. 4. Foran epigrampossiblyrelatingto the samebattle,cf. P. E. Riza-
cel
kos,<?rvOov 'ApX'Ef 1912,pp. 40-43. Cf. also DavidArmstrong,"Gallienusin Athens,
ypap.p.aE7rLrvp.zutov>>,
264,"ZPE 70, 1987,pp. 235-258.
68ThompsonandScranton(footnote4 above,p. 1), pp. 366-372.
 II
RECOVERY
POST-HERULIAN ATHENS (A.D.267-300)
By the middle of the 3rd century, even beforethe Herulian sack, the economicsituation in Athens
had deteriorated.The productionof sculpturedsarcophagi,the most importantexport,had ceased,'
leaving the manufactureof lamps as the only importantsourceof income from abroad.During the
3rd and 4th centuriesthese continuedto be exportedall over Greece and also overseas.2Practically
© American School of Classical Studies at Athens

no pottery, other than wares for strictly utilitarian use, was made in Athens in Late Antiquity.
For personal use only. License: CC-BY-NC-ND.

Athenianpotterswere no match for the artisansof Asia Minor and North Africa,whose fine wares
began to be importedin small quantities in the late 4th centuryand in increasingvolume through
the 5th.3
The economic decline of the 3rd century was accompaniedby a general loosening of political
organizations.All through the 2nd century and the first third of the 3rd, prytany decreesand lists
existed in profusion,ceasing suddenly about 230-240, the date of the latest known example being
231/2.4 The institution of the ephebes shared the same fate except that it lasted right up to the
Herulian sack, the latest known inscriptiondating between 260 and 267.5
The Council of the Areopagus lasted at least until toward the end of the 4th century but with
reduced responsibilities.6The archonship continued until 485, the year of Proclus' death in the
archonshipof Nicagoras the Younger.7
An increase ca. A.D. 270 in the membershipof the Boule from 500 to 750 is unexplained. Day
suggeststhat it might reflecta shortageof membersaffluentenough to fulfill the necessaryfinancial
responsibilities.In the 4th century the membershiprevertedto the pre-Hadrianic300 for equally
mysteriousreasons.8
The PanathenaicFestival was still being held close to the time of the Herulian raid, with P. He-
rennius Dexippus as agonothete;9on its later history,see below, pp. 23-24.
The disaster of 267, although of the first magnitude for Athens, would not in itself have been
enough, without contributingcauses, to smotherthe civilizationthat the city represented.Through
sheer vitality Athens had survivedthe Persian sack of 480/79 B.C. and emergedmore powerful and
magnificentthan before. Recoveryfrom Sulla's siege in 86 B.C.was both slower and less complete,
1 See Day, p. 261. On the export, fromits beginningin the time of Hadrian to its cessationafter 267, cf. A. Giuliano,
II commerciodei sarcofagiattici, Rome 1962.
2
AgoraVII, p. 9; Thompson, "Twilight,"pp. 70-72.
3 For this
subjectin detail, see Hayes.
4 Agora
XV, p. 338, no. 491.
5 For the prytaneis,see Hesperia, Suppl. XII, p. 92; for the ephebes, ibid., p. 1, note 3.
6 Ibid., p. 61; cf. Himerius, Or. vii (vii), addressingthe body as those who "nowjudge for the Atheniansconcerning

freedom."
7 PLRE II, Nicagoras iunior.
8 Day, pp. 277-278; Hesperia, Suppl. XII, p. 91; Millar, "Dexippus,"p. 21; IG II2, 3669 (750 members),4222
(300 members).
9 IG II2, 3198. His holdingof the agonothesiacannotbe said to be in exactly 269/70 (Hesperia, Suppl. XII, p. 133),
a most unlikely time for the celebration.The date comes from IG II2, 3669, which honors Dexippus not for his role in
freeing Athens from the Heruli but for his History, which must postdatethat event. His agonothesiais includedonly as
part of his cursushonorum(as also in IG II2, 3670, on the base of a statue of Dexippus erectedby his sons).
 THE AGORA 13

no doubt because the city had already lost its preeminenceand with it some of its motivation.10By
the 2nd centuryafter Christ the great revivalof building and the other arts came largelythroughthe
energy and generosityof individualbenefactorssuch as Hadrian and HerodesAtticus,and the rela-
tive prosperitythat lasted through the Antonine period had little chanceof survivalin the troubles
and uncertaintiesthat were the general lot of the Empire in the secondhalf of the 3rd century.
After the departureof the Heruli, Athens enjoyedan uneasy peace, under constantthreat of fur-
ther attack,until the last decadeof the 4th century. During the first part of this period, i.e., the last
third of the 3rd century,there seems to have been little activitybeyondbasic measuresfor survival.
No new constructionwas undertakenin the area of the Agora except the fortificationwall (PI. 5),
© American School of Classical Studies at Athens

but destructiondebriswas clearedaway in orderto make some of the buildingshabitable,if only by

squatters.Need for water promptedthe cleaningout of as many wells as possible,while most of the
For personal use only. License: CC-BY-NC-ND.

rest servedas dumping places for unsalvageablematerial.1 It would be surprisingto find the Athe-
nians defyingthe danger,at the same time that it was being recognizedby the buildersof the wall, to
the extent of putting up elaborate houses in the undefendedpart of the city. The Agora proper,
boundedby Kolonos Agoraios, the PanathenaicWay, the South Road, and some as yet undefined
line on the north side, certainlyremainedalmostdesolateuntil well along in the 4th century,and the
deep hollow at the west end of the Middle Stoa providedalmostlimitless space for dumping.
The new circuit could never have accommodatedthe entire population, however depleted by
now; it could only have been intendedas an emergencyline of defense in case of an actual raid. At
other times the inhabitantswent about their business outside, but with a lack of confidencethat is
reflectedin the absenceof substantialbuildings.12 Although the Agora remaineddesolatefor many
years the adjacentareas outside the wall gradually took on new life, especially to the south and
southwest.13

THE AGORA
The appearanceof the area south of the Agora during the period from 267 to the end of the 4th
century can never be quite clear. Seventeen centuries of erosion on the slope and of rebuilding,
sometimesgradual and sometimesinvolvingclearanceof fairly large areas in a concertedprogram,
have removedall but the smallest tracesof buildings, includingevidenceof date. Compensationfor
this lack, however, is afforded by the many wells with their filling: a sure guide to habitation.
Taken as a whole, these show a consistentpatternfor the historyof Late Antiquity. Most had been
sunk at least as early as the 1st centuryafter Christ and showed an unbrokenperiodof use until the
last half of the 3rd century,i.e., 267. A periodof disuse followed,representedby an accumulationof
debris resulting from a gigantic clean-up operation, and then a gradual resumption of use as a
source of water. The pace was acceleratedin the 4th century and continuedwithout interruption
into the 6th; in some cases the period of use ends sharply with the end of the first quarter of the
century,presumablyreflectingthe closingof the schoolsof philosophyin 529; in othersit continued

lo Agora XIV, p. 23.

1 This depressedstate is well illustratedby the immediateand spectaculardecline in the quality of the productsof
the Athenian lamp manufacturersduring the same period and its rapid recoveryat the beginning of the 4th century
(AgoraVII, pp. 20-21; see also Day, chaps. VI and VII).
12 For makeshiftrepairs made to damagedbuildings on the west side of the Agora, see below, Chapter III.
13 The north side, except for sporadicfinds, still lies under the moderncity (PI. 1); the east, of course,is boundedby
the wall. The following is a brief surveyof the developmentof the district.Particularswill be discussedin their proper
chronologicalframework.
 14 II. RECOVERY

until towardthe end of the century,when Slavic incursionsof the 580's broughtthe life of the city to
an abrupt halt for a quarterof a century.14
The south districtfalls into two parts, an easternand a western,almost cut off from each otherby
the protrudingmass of the Areopagus(PI. 4). The western half is the flat area occupiedin classical
times by private houses and industrial establishmentsand was approachedfrom the southwest
cornerof the Agora. 5 Here many of the pre-Herulian houses were made habitableonce more, and
the districtseems to have been fairly populous from the middle of the 4th century,with the lower-
lying parts quite quickly taking on the aspectof a slum.16
The eastern half of the districtformeda rough triangle, containedbetween the South Road, the
© American School of Classical Studies at Athens

Dromos from the Middle (Christ Church) Gate to a point between the South (Hypapanti) Gate
and Tower W 1, and the Areopagusitself. Toward the west the area narrowedas the contoursof
For personal use only. License: CC-BY-NC-ND.

the Areopagusalmost met the South Road at its junctionwith the roadcomingin from the west end
of the Areopagusand leading into the Agora at its southwestcorner.
It is in this easternhalf that the concentrationof wells is heaviest,at least a dozen having been in
use in the 4th century. Most of the houses that they servedhave been swept away, either by rains
washing down the slopes or deliberatelyto make roomfor the more substantialbuildingsof the 4th
and 5th centuries.Some of the 4th-centurysettlement,however,may perhapsbe recognizedin pre-
Herulian houses, rehabilitatedin whole or in part, and there is evidenceto show that there was a
general rebuildingin the area in the secondhalf of the 4th century.17
The reason for the more rapid recoveryon the south side is fairly obvious:there were no major
pre-existingbuildings to clutterthe groundwith indigestibleruins;and the higher terrain,forming
a natural limit to the flat expanse of the Agora proper,was well suited to residentialrequirements.
The topographicaladvantageshad already been exploited in classical times when three broadter-
raceswere laid out along the centralpart of the district.Springson the slopes of the Areopagussent
down a constantflow of groundwater to the wells and fountainsbelow, and even though the Agora
itself was used as a dumping ground until its rehabilitationshortly after 400, the rapid growth of
weeds which accumulatesover debriscomposedof mud brick,ashes, and decayedvegetablematter
would have soon obscured the more unsightly aspects of the devastatedarea. The South Road
formedthe northernboundaryof the new residentialdistrict.Along the north edge the backwall of
South Stoa II, which had been rebuilt in the 2nd century to carry the Antonine aqueduct, served
still furtherto detachthe hillside from the plain by providinga sharp drop of over five meters. 8
The graduallyincreasingpace of recoverythroughthe 4th centuryis borneout by the numismatic
evidence. From the period 269 to 305 a total of 508 coins was found in the Agora proper and its
immediatesurroundings,or an averageof approximately15 per year.19During the reign of Con-
stantine the averagenearly doubled,to ca. 27. In the 59 years from the death of Constantineto the
invasion by Alaric in 396 the yearly average increased fivefold, to 132.20By the end of the 4th
century, when the Visigothic threat had become acute, the settlement outside the Post-Herulian
14 Below, Chapter VI.
15 Young, "IndustrialDistrict,"esp. 284-288 for our period.
16
Below, p. 35.
17
Below, Chapter III.
18
For the later use of the aqueduct,see below, pp. 22, 29.
19These
figures are those published by M. Thompson in Agora II in 1954; the numbers have increased,but little
change will be found in the proportions.
20 These statistics can have
only a very general significance,leaving out of accountvarious specificallynumismatic
factors.The numbers,however,are large enough to indicatea definitetrend.
 THE AGORA 15

Wall was substantialenough to warrant restoringthe Themistocleancircuitwall, which had fallen

into disrepairsince it was rebuilt under Valerian and after the attackof 267.21
With the Herulian destructionthe Agora properlost foreverits characteras a public square.The
only enduring features were the roads leading into it which, although modifiedin some respects,
kept in the main their original courses (PI. 6).22 As in precedingcenturies,the PanathenaicWay,
leading from the Dipylon, was still the main road into Athens.23At the northwest corner of the
Agora it was joined by a narrowerroad coming in from the north and mergingwith the road along
the west side of the Agora; a third street, built in the 2nd century after Christ, led from the Sacred
Gate and paralleledthe PanathenaicWay a little to the south as it approachedthe Agora.24At the
entranceto the Agora the PanathenaicWay took a southeasterlycourse leading diagonally across
© American School of Classical Studies at Athens

the Agora to its southeastcorner.25

For personal use only. License: CC-BY-NC-ND.

The Panathenaic Way and the west road remained in use, with some interruptions,through
Byzantine times; the road from the Sacred Gate and the road from the north had a continuous
history through the 6th century,after which they were coveredwith Byzantinehouses.
About the middle of the 5th centurya new west-east road was built on the line of an old one long
in disuse, startingat the point where the PanathenaicWay enteredthe Agora (PI. 4); this led to the
Library of Hadrian and was to be known as the Street of the Library of Hadrian because its pro-
jected course ends exactly at the porch of the Library. It overlaythe ruins of a large basilica of the
Hadrianic periodwhich was destroyedin the Herulian raid. This road is now indicatedonly by the
foundationsof a stoa, presumablyone of a pair flankingthe street.26
South of the Agora the road leading into the heart of the city from the Piraeus forkedin three
directions (PI. 6): one to the north, to join the West Street, a secondto the east (the South Road),
and a third in a southeasterlydirectionhigher up on the slopes of the Areopagus(the Upper South
Road). Of these three the South Road, running around the lower slope of the Areopagus to the
Middle Gate in the Post-Herulian Wall, was the main artery for this area. Establishedas early as
the Bronze Age, it remainedin use, with graduallyrising levels, over substantiallythe same course
until the 20th century. By the Late Roman period encroachmentsfrom the south had permanently
narrowedit to approximatelyhalf of its original width, or ca. 2.50-3.00 m.27
The East ColonnadedStreet, built in the reign of Trajan to connectthe Greek and Roman mar-
kets, between the Stoa of Attalos and the Gate of Athena Archegetis,enteredthe inner city through
the PyrgiotissaGate (PI. 14:d). It continuedin use at least until the 12th century.

21 Below,p. 83.
22 See Agora XIV,
pp. 17-18, 192-194; E. Vanderpool,"Roadsat the Northwest Corner of the Athenian Agora,"
Hesperia28, 1959,pp. 289 ff.
23 For its course and earlier history,with bibliography,see AgoraXIV, locc. citt.; Vanderpool,op. cit., pp. 291-295.
24
AgoraXIV, p. 108;Thompson,"WestSide,"pp. 5, 219, fig. 126;Hesperia6,1937, p. 339 (Shear,Sr.);Hesperia
42, 1973,p. 370 (Shear,Jr.).
25 For the buildings along the street and the vagariesof its courseover the years, cf. below, pp. 26-29.
26Hesperia40, 1971,pp. 264-265 (Shear, andbelow,p. 79.
Jr.)
27
Hesperia25, 1956,pp. 47-49 and29, 1960,pp. 332-333,340, fig.4 (Thompson).
 III
FROM DIOCLETIAN TO THE VISIGOTHIC INVASION
DIOCLETIAN TO JULIAN THE APOSTATE (A.D.284-363)
CONSTANTINE AND ATHENS

Even though an appreciablemeasureof recoverycan be noticedin Athens from the latter part of
the reign of Diocletian and throughthe first quarterof the 4th century,this is a relativeterm. By the
© American School of Classical Studies at Athens

beginning of the 4th century Thessalonica, an upstart by comparisonwith Athens, had becomea
For personal use only. License: CC-BY-NC-ND.

cosmopolitancity, convenientlysituated on the Via Egnatia, the main land route between the east-
ern and western halves of the Empire.After the transferof the capitalto Constantinoplethe impor-
tance of that city increased,while Athens, out of the mainstreamof Imperial business, became a
provincialcultural centerin the Provinceof Achaeaof which Corinthwas the capital (althoughthe
vicariusresidedin Thessalonica).'
The elevation of the Province of Achaea to senatorial status under Constantine (perhaps even
under Diocletian)2 implies an improvementin the state of Greece as a whole; the impression is
heightenedfor Athens itself by the EmperorJulian's referenceto Constantine'sconcernfor the city
"in word and deed, throughouthis whole life."3As evidenceof Constantine'sinterestJulian cites
the fact that he acceptedthe title of "strategosof Athens" (i.e., c7r rTao'ira),4 in recognitionof
which the Athenianserecteda statue of him "withan inscription."This in turn pleasedConstantine
"morethan the highest honors"and inspired him to make an annual distributionof grain to the
city.5The grant was renewedlater in the centuryby his son ConstansI in the formof severalgrain-
producingislands.6
Constantine,eager to please the pagan aristocrats,showed special favor by enabling Nicagoras,
the Priest of Eleusis, to go to Egypt to visit the Syringes.7 The Athenians especially would have
profitedby and appreciatedConstantine'stolerantattitudetoward those who retainedtheir pagan
IA. Piganiol, L'empirechretien,2nd ed., Paris 1972, p. 14.
2
For the date of the change,see Groag, Achaia,pp. 15-16 and A. H. M. Jones, The LaterRoman Empire, 2nd ed.,
Norman [1964], I, pp. 106-107.
3Julian, Or. i.6. This tributeto the father,insertedin Julian's panegyricof ConstantiusII, was perhapsintroduced
as a welcome digressionfrom the uncongenialtask of praising the son.
4At this time the functionof strategoswas to take chargeof the food supply and the provisionmarket(Philostratus,
VS 1.23 (526). For the prestigeof the officesee P. Graindor,Un milliardaireantique. HerodeAtticuset safamille, Cairo
1930, p. 27.
5The passage is worth quoting in full: KaA*v oaS evrav6a KaLrwv aoLLFv 'AOivv Mv KOvaL as EKEovS
EpyOLs KaL AoyoL TLMiWV TOV7ravra XpovovblereAEL.3aaLXEvs yap wV KOLKVpLOSIravTrwv, -Trparn7yovsEKVIVW ov
va)t~??rOaL K~ ' ' ' ' ' '
) '
I
' '* I'
KaXeL0cOaL KaL TroLavrrT7s ELKOvos rvyxavcov /erT e7rypaluuaTos eyavvvro 7rXEov rI TW v ey,LTrCovrT7uov ' aLWLS.
a?ifue36Levos ofe 7rIT
avr- a,ItEL
T"rV7TroAlV,7rvplv
0J.zEvov~ 5LAg" va6 TOOa
K5LWOL e7aorov pvovs
VPO 'ero'oKarov Tbpavopea x apsov-ovo'-
' Ka'trrta
OaL. ev vTrripxE r T7rdoAL
Ap.Vev a6ovoLs eivat, bXftaLuo
VK8Voz rapa r&v 3cEArtrroov.Cf. Wachsmuth,
I, pp. 710-711.
6 Eunapius (VS, 492) describesin detail the award by Constansat the instanceof Prohaeresiusand its confirmation
by the Prefect of Illyricum, Anatolius; see below, p. 25.
7Millar, "Dexippus,"p. 17. P. Graindorpoints out that Constantine'sinterestin Nicagoraswas not becauseof his
religious preeminence but because the emperor was an "ami des lettres" ("Constantine et le dadouque Nicagoras,"
Byzantion 3, 1926, pp. 209-214). Nicagoras' grandfather had held the chair of sophistic in Athens, and the Syringes,
the so-called "pipes", the Egyptian kings of Thebes, were regarded as places of pious pilgrimage
because they had been visited by Plato. (Cf. Pausanias, 1.42.3.)
 DIOCLETIAN TO JULIAN THE APOSTATE (A.D.284-363) 17

beliefs, as expressedin an edict sent to the provinces.In this he denouncedthe cult of false gods,8but
a thread of tolerance runs through the text, especially strong in chapter 56. Those in error must
enjoy the same peace and quiet as the faithful; no one shall molest either ('Kao-roSOTEp 7 vX
/l3ovAeTaL KarTEXET), TOVTCO
KaTaKEXpi 6ow).9

THE REFLORESCENCE OF THE SCHOOLS

The most enduring of all the pagan institutions in Athens were the schools of philosophy and
rhetoric.These had continuedto flourishup to the time of the Herulian incursion,and it was only a
short time beforethat Longinus and his pupil Porphyrywere upholdingthe Platonictraditionand
© American School of Classical Studies at Athens

marking Plato's birthday with a dinner in Athens at which one of the topics of conversationwas
plagiarism in classical Greek literature.10Fergus Millar has examined all the available evidence,
For personal use only. License: CC-BY-NC-ND.

literary and epigraphical,to determineto what extent the intellectuallife of Athens was disrupted
by the events of 267. Although for the last quarterof the 3rd centuryhe findsthe evidence"marked-
ly thin," he concludesguardedly that the "stabilityfrom which the world of Dexippus came was
not, in the long term, destroyedby the Herulian invasion.""11 Although the thread of the philo-
sophical tradition was not broken, it was neverthelessextremely tenuous for the remainderof the
century.Other factorsalso show that althoughthe raid was devastating,the breakwith the past was
not complete.Some institutions,e.g. the enrollmentof the ephebes, apparentlyended abruptly,but
others survived.The Council of the Areopaguswas functioningas late as the secondhalf of the 4th
century,12and the PanathenaicFestival at least as late. Also, the archonswere taking officethrough
the 5th century.13
Around the turn of the century the schools began again to attractnotice abroad.The Christian
Prohaeresiusapparentlycame to Athens as early as the first decadeof the 4th centuryto studywith
one of the most famous of the late sophists,Julian of Cappadocia,who must thereforehave already
been establishedin the city.14 The fame of the schoolsand more especiallyof the fightingthat went
on continuously between the various factions was sufficient to fire the imagination of Libanius
(born in 314) in his boyhoodand to make him resolveto go to Athens at the earliest possibleoppor-
tunity. (When he achieved his wish in 336 he cautiously kept aloof from the fighting, but he ob-
servedit closely and describedit vividly).15

8
Eusebius, v. Const. 11.48-6o(PG XX, ii, pp. 1026-1034 and I. A. Heikel, Die griechischenchristlichenSchrift-
steller, Leipzig 1902, I, pp. 61-65).
9 For more on Constantine'stolerance,see now R.
Krautheimer,Three ChristianCapitals,Berkeley 1983, chap. I.
10Millar, 16-20. For the annual
"Dexippus,"pp. birthdaycelebration,Marinus, Vita Procli, 23; also Saffrey and
Westerink, p. xix, with references.
11Millar,
"Dexippus,"pp. 16-20. For the known details of events and personalities,both beforeand after the raid,
the reader is referredto Millar's admirableanalysis.
12 IG
II2, 4222. Also, the EmperorJulian addressedhis Letter to the Athenians (A.D. 361) Tr t3ovAj Kat TO? i^8l.
13
Graindor,Archontes.
14
For the probabledates of the arrivalof Julian and Prohaeresius,see Millar, "Dexippus,"pp. 16-20.
15Libanius, Ep., 627. Libanius and
Eunapius are the chief sources for the Athenian schools in the 4th century.
Among the most useful modern works are F. Schemmel, "Die Hochschule von Athen im IV. und V. Jahrhundert,"
Neues Jahrbuchfiir Philosophie22, 1908; J. H. Walden, The Universitiesof Ancient Greece, New York 1909 (with
many translationsfrom Libanius);G. Sievers,Das Lebendes Libanius,Berlin 1868; H. I. Marrou, Histoirede l'iduca-
tion dans l'antiquite,6th ed., Paris 1965; M. L. Clarke, Higher Educationin the Ancient World,London 1971;see also
Libanius' Autobiography(Oration I), The Greek Text, edited, with introduction,translation, and notes, by A. F.
Norman, London/New York 1965; also L. C. Ruggini, "Sofistigreci nell' impero romano,"Athenaeum,n.s. 49, 1971,
pp. 402-425; and Glucker, Antiochus.
 18 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

Most of the 4th-centuryteacherswhose careersare known to us in any detail were sophists;this

seeming preponderance,however, is partly due to the fact that Eunapius, our chief source for the
period,studiedonly rhetoricin Athens, whereas his knowledgeof philosophyand philosopherswas
obtained in Pergamon. The balance has now been redressedin part by H. D. Saffrey and L. G.
Westerink,16who have shown that the Platonictraditionwas being kept alive in Athens, although
its exponents were not in complete harmony;hence the lack of an unbrokenchain of Diadochs in
the 4th century. Not only was Platonic philosophythrivingbut the other schoolsas well.17Names
of a few philosophershave emerged,e.g., Priscus, Celsus, Hilarius, and lamblichus (II).18
Almost all the teachers and students known to have been active in Athens in the 4th and 5th
© American School of Classical Studies at Athens

centuries came from abroad, especially the eastern provinces,thus continuingthe traditionof the
For personal use only. License: CC-BY-NC-ND.

4th and 3rd centuriesB.C., when few of the distinguishedmen of arts and letterswere nativeAtheni-
ans. Prohaeresiuswas an Armenian; Hilarius, a Bithynian who was both a philosopher and a
painter,'9"grewold at Athens,"and died at the handsof Alaric'sforcesin 396.20Himerius, another
Bithynian,came to Athens in the secondquarterof the 4th century,marriedan Athenianaristocrat
said to have been a descendantof Plutarch, the biographer,and acquired citizenship. Himerius
joined the EmperorJulian in the east for a time but returnedto Athens, where he remaineduntil
his death.2'The one native Athenianto achievefame in the schoolswas Plutarch,who foundedthe
Neoplatonic School in the last quarterof the 4th century and was still active as Diadoch when he
died in 432.22Indeed,longevitymay have been a factorin the continuityof the academictraditionin
Athens, for Julian the Cappadocianand his pupil Prohaeresius between them spanned almost
three quartersof a century,and Plutarch'smost distinguishedpupil, Proclus,was Diadoch fromca.
440 to his death in 485.23

THE PROGRESS OF CHRISTIANITY

The gradual ascendancyof Christianityin the Empire as a whole had little effect on Athens in
the first five centuriesof the Christianera.24The evidencefor the Christianpopulationin the 2nd
and 3rd centuriesis slight. Three Bishops of Athens (Narcissus, Quadratus,and Publius) suffered
16 In their invaluable
chapter "L'ecoled'Athenesau IVe siecle,"pp. xxxv-xlviii.
17SaffreyandWesterink,p. xl.
18
Celsus, son of Archetimus,a "high-rankingexponentof philosophicalstudies"(optimatemsapientiae),inducedto
leave Athens to teach at Rome (Symmachus,Relationes,5.1 and R. H. Barrow,Prefectand Emperor:The Relationesof
Symmachus,Oxford 1973, p. 52). lamblichus (II) will reappearas a substantialmaterialbenefactorof Athens as well
as a philosopher(below, p. 51).
19Not a later philosopher of the same name whom Proclus refused to teach on accountof his dissipatedlife: o8b
ov
IXAapLos,uXoAo<f>v fEK8f8Lb&Ttf1VOS9BE Ta%VTroya0 TPWovs ?ovais, OVKCTU'rvIHpOKXov6baaarKaAov. Damascius,
Vita Isidori,266 (Epitoma Photiana [Zintzen, p. 187]).
20
Eunapius, VS, 482. For his place in the philosophicalmilieu of Athens, Saffreyand Westerink,p. xlii.
21
H. Schenkel (RE VIII, cols. 1622-1635, with Suppl. III, cols. 1151 ff.) concludesthat Himerius' life must have
been ca. 300-380.
22
He too may have been descendedfrom the biographer(Millar, "Dexippus,"p. 17). For his immediateancestors
and his descendants,see Saffrey and Westerink,p. xxxv. For the school, below, pp. 57-58.
23 For Plutarch and Proclus, see below,
pp. 43, 63; also, for buildings probablyconnectedwith teaching,pp. 37-48.
24
For the progressionfrom the Edict of Galerius (A.D. 311) grantingtoleranceof Christianityto the final edicts of
Theodosius I of 392 and 395, prohibitingall pagan rites (CTh xvi. o0.12-I 3), see J. Geffcken,Der Ausgangdesgriech-
isch-romischenHeidentums, Heidelberg 1920; K. Baus, Handbookof ChurchHistory, H. Jedin and J. Dolan, edd.
(trans.from 3rd ed., New York 1965), I, chaps. 30, 31; N. H. Baynes, Constantinethe Greatand the ChristianChurch,
2nd ed., London 1972, prefaceby H. Chadwick,with extensivebibliography;T. D. Barnes,"Lactantiusand Constan-
tine,"JRS 63, 1973, pp. 30 ff., also idem, Constantineand Eusebius,Cambridge,Mass./London 1981.
 DIOCLETIAN TO JULIAN THE APOSTATE (A.D.284-363) 19

martyrdomin the 2nd century.25Athens' most famousmartyr,Leonides,becameone under Decius

(A.D. 291). Although he is known as a bishop of Athens, recent investigationshave demotedhim
from that rank.26
In the 2nd centurythe Christian communityof Athens receiveda reprimandfrom the Bishop of
Corinth, Dionysius, in whose jurisdictionthe Churchof Athens lay, for their lack of faith, and their
bishop, Quadratus, was praised for having rekindled it.27Origen in 240 evidently found the at-
mospheresufficientlycongenialto visit Athens for long enough to finish his work on Ezekiel and to
start one on the Song of Songs.28The "churchof God at Athens"was, in his opinion, "gentleand
orderly",as opposedto the popular assemblyat Athens which was "unruly".29
© American School of Classical Studies at Athens

The situationchangedlittle in the 4th century.Athenswas once more a universitytown, and only
For personal use only. License: CC-BY-NC-ND.

the lower classes showed themselves susceptibleto Christianityto any degree. Most Athenians of
the upper classes were connectedin some way with the schools, and this ensured almost universal
adherenceto paganism.30Support was also providedby some high officials, most notably Vettius
Agorius Praetextatus,Proconsulof Achaea in 362/4, a personalfriend of the EmperorJulian and
holder of many priesthoodsin his long career (d. 384) including that of Eleusis and the taurobo-
lium.31It is believedthat he may have introducedthe tauroboliumto Athens, first celebratedthere
by the priest Archeleos, as stated on one of two altars with representationsrelating to the cult of
Cybele and Attis.32The second altar was erectedsome 20 years later, in 386, by Musonius, as the
inscriptionrecords,in the consulshipof Honorius and Euodius.33
Visible evidenceof the tenacity of paganism is the shrine, probablyof Hecate, at a conspicuous
crossroadsin the PanathenaicWay, in the middle of the Agora, which continuedin use until the
late 4th century.34In this connectionit may be noted that Aconia Fabia Paulina, the wife of Prae-
textatus, was a priestessof Hecate on Aegina.35
Throughout the 4th centuryand well into the 5th, Athens was to all intents and purposesa pagan
city. Whatever the nature of the celebrationsand other benefactionsmentioned by the panegy-
rists,36nowhere is there any hint of Christian influence. The pagan characterof activities in the
newly restoredTheater of Dionysos is made clear in the dedicationof the new stage by the archon
25
G. A. Sotiriou, EMME I, p. 21.
26
F. Halkin, Rechercheset documentsd'hagiographiebyzantine (Subsidiahagiographica51, Societe des Bollan-
distes), Brussels 1971, pp. 60-66. Michael Choniatesdevotedan entire orationto the martyrdomof Leonides:Michae-
lis Acominati, Opera, S. Lambros,ed., Athens 1879 (reprint 1968), I, pp. 150-156.
27
Eusebius, HE Iv.23.3.
32. For
For more on this
on
28 Eusebius32
more
period, this period, see Dictionnaire d'histoireet de geographie ecclesiastiques,s.v.
Athenes (Janin); also Reallex.f. Antike und Christentum,s.v. Athen II (Frantz, in press).
29 Contra Celsum 111.30.
30
Frantz, DOP, pp. 187-202; also idem, "PaganPhilosophersin ChristianAthens,"Proc. Phil. Soc. 119, 1975, pp.
29-38.
31 For his career see PLRE
I, Praetextatus 1 and for the list of his priesthoods,his epitaph in Rome, CIL VI, 1779.
Also Groag, Achaia, pp. 45-48; Robert (footnote38 above, p. 6), p. 24.
32
irpTrov EVpo TreXtoufpV?s, IG II2, 4841. The two altars are discussedin detail by J. N. Svoronosin Das Athener
Nationalmuseum,Athens 1908, nos. 1746-1747, pp. 474-484, with illustrations (pl. LXXX). See also M. J. Ver-
maseren, Corpus Cultus Cybelae Attidisque II, Leiden 1982, pp. 116-118, nos. 389, 390. Svoronosnotes that the
taurobolium,which the Greeksregardedas a barbariansuperstitionattachedto the worship of Cybele, had spreadover
the whole Roman world by the 2nd century but had apparentlynever before been recognizedofficially as a separate
cult. For the family of Archeleos,see below, pp. 50-51.
33 IG II2, 4842.
34
Above, p. 8.
35 CIL VI, 1779: "sacrataCereri et Eleusiniis sacrataapud Aeginam Hecatae tauroboliatahierophantria."
36 Below, pp. 20-24.
 20 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

Phaidros.37Apart from the Panathenaiathere is little solid evidencefor games after the middle of
the 3rd century,38but the PanathenaicFestival itself was being held regularlyin the middle of the
4th century,as describedin detail by Himerius, and probablyinto the 5th.39
Pheidias'bronzestatueof Athena Promachoscontinuedto be a landmarkon the Acropolis.It was
mentionedby St. Jerome in 372 and again on the base of a statueof Herculius, Prefectof Illyricum,
ca. 410. It is even possible that it was not removedto Constantinopleuntil ca. 465.40
There was apparentlylittle distinctionbetween pagan and Christianstudentsin the 4th century.
In other parts of the Empire some effort was made to provideChristianeducationto competewith
the schools,41 but in Athens the future Church Fathers, Basil the Great and Gregoryof Nazianzus,
© American School of Classical Studies at Athens

were as assiduousin the study of the classicsas was their fellow student,Julian, the future apostate
For personal use only. License: CC-BY-NC-ND.

emperor. It would be agreeableto be able to include John Chrysostomin this distinguishedcom-

pany.42A circumstantialaccountby George, Bishop of Alexandria,has him finishinghis education
in Athens after learning all he could in Antioch.43But it has been pointedout that strictobservance
of the known chronologywould have him achieve this status at the age of 18.44 The person of
Anthemius, priest of Athena and a leading philosopherof Athens who was confoundedby John's
wisdom in the same account,must probablyalso be consignedto the realm of fiction.
The only known Christian teacher in Athens at this time was Prohaeresius,who was highly
esteemedby pagan and Christianalike.45The fightingwhich figuresso prominentlyin the accounts
of student life seems to have been chiefly between the factions of different teachers or between
studentsand townspeopleratherthan between pagans and Christians.

THE CITY OF ATHENS AND ITS BENEFACTORS

During the 4th centurythree high officialsare namedas benefactorsof Athens:two Proconsulsof
Achaea, Cervonius and Ampelius, and the Emperor Julian. The precise date of Cervonius, the
earliest, is uncertain, but he is believed to have held office in the second half of Constantine's
reign.46Ampelius is next, with his term firmly fixed at 359/60, barely two years before Julian's
accession.47The literary evidencefor all three rests on a dubious type of source:the panegyricor
other encomium. The two proconsuls are praised by Himerius on their departurefrom Greece;
37 Below, p. 25.
38
The latest known agonisticinscriptionis datedca. 253-257 (L. Moretti, Iscrizioniagonistichegreche,Rome 1953,
pp. 263-268, no. 90). Moretti notes, however,that there seem to be no more majorgames after that time but that local
and regional contestscontinuedto be held for some decades.
39Below, pp. 23-24.
40
Below, pp. 76-77.
41 Clarke
(footnote 15 above, p. 17), chap. 5.
42 Alan D. E. Cameron,"lamblichusat Athens,"Athenaeum,n.s. 45, 1967,
p. 153, note 19.
43 Photius, Bibl., Cod. 96, Life of St. John Chrysostom.This authoris not to be confusedwith the 4th-centurybishop
of the same name and the same city. The episcopacyof the author of the life of Chrysostomhas been dated in the first
quarter of the 7th century by P. R. Coleman-Norton("The Vita S. Chrysostomiby Georgius Alexandrinus,"CP 20,
1925, pp. 69-72). The Life itself is based in the main on the biographyby Palladius, a contemporaryand friend of
John. The author's own contributions(of which the Athenian episode seems to be one) prove to be almost entirely
figments of the imagination.They are rejectedby C. Baur, John Chrysostomand his Time, M. Gonzaga, trans., 2nd
ed., Westminster,Maryland 1959, pp. 25-26.
44 A.
Naegele, "Chrysostomosund Libanios,"Chrysostomika,Rome 1908, pp. 93-98.
45 See also above, p. 17.
46 Groag, Achaia,pp. 27-28.
47 Ibid., pp. 42-44.
 DIOCLETIAN TO JULIAN THE APOSTATE (A.D.284-363) 21

Mamertinus, on assuming the consulship, exalts Julian.48 A certain similarity pervades the ac-
counts, all drawn from the general language of panegyric,and so the line between truth and fiction
is not always clearly defined.The formulais standard.In each encomiumthe city, which has fallen
into decay, has been transformedby whicheverindividualis being addressedat the moment.Julian
has brought water to places that were formerly dry and panting with thirst. Cervonius too has
brought water to Athens but he is creditedin more flowery language:the Nymphs have left their
groves to play in the streets of Athens, and the place flourishesas a spring meadow. Under Julian
old festivals are celebratedand new ones are inaugurated.With Cervonius,celebrationsand danc-
ing begin again.49By commandof Julian all the cities of Macedonia, Illyricum, and the Pelopon-
© American School of Classical Studies at Athens

nese are restored.Ampelius has transformedthe whole countrysidefrom Thermopylaeto the tip of
For personal use only. License: CC-BY-NC-ND.

the Peloponneseinto cities. But in additionto loose generalitieshe is creditedwith one episodethat
is refreshinglyconvincingin its modestsimplicityand specificdetail:the renewal of a small district
"in the very heart of Athens, centeredaroundthe little streetof Kollytos."50It may be noted, more-
over, that Ampelius is the only one of the three men who is known for specificgifts elsewhere. For
example, reconstructionof the theater at Sparta and the establishmentof the Amyklaion in the
same city, and possiblyother buildingsin Spartaas well.51On Aegina he apparentlyconstructedan
elaboratenymphaeum,52and in Chalkis he built or restoredsome public building.53
In the case of Mamertinus on Julian, no concreteexamples of public or privateworks are cited,
only general prosperity,improvedwater supply, and celebrations.It is even more significantthat
none of these are placed specificallyin Athens, which is not mentionedbeyondthe fact that it had
fallen into decay, along with other named cities. Otherwise,only tacit inclusionamong all the cities
of the Balkan peninsula. The impressionis left that the newly appointedconsul, who had already
held officesunderJulian,54knew nothing at first hand either of Athens or the rest of Greece,and is
speaking in safe and acceptablegeneralities.55
The list of benefactorsof the city, dependentlargely on chance,is necessarilyincomplete.Three
officialsmight certainlyhave qualified,althoughno recordsurvivesof a practicalinterestin Athens.

48
Cervonius:Himerius, Or. iv (xxxviii); Ampelius:Himerius, Ecl. xxxi; Julian: Claudius Mamertinus, Grat.act.
Iul., 9, etc.; Panegyriqueslatins, E. Galletier, ed., Paris 1955, III, pp. 23-24.
49 He is also honoredas a "friendof the Muses" on a statue base found at Thespiae (SEG XV, 323).
50This contributionof Ampelius will be discussedmore fully below, pp. 35-36.
51 For the
inscriptionrecordingthe rebuildingof the theater,see A. M. Woodward,"Excavationsat Sparta 1924-25:
The Inscriptions,"BSA 26, 1923-25, pp. 225-236, no. 20; for the Amyklaion,ibid., p. 229 and IG V 1, 455.
52 Robert (footnote38 above, p. 6), pp. 5-34.
5 Woodward (footnote 51 above), p. 227; also IG XII 9, 907 and A. Joubin and A. Wilhelm,
"Inscriptionsde
Chalcis,"BCH 16, 1892, pp. 102-106. After his proconsulshiphe went on to a distinguishedcareerin otherhigh offices
(Groag, Achaia, pp. 42-44). For the personal reasonswhy Ampelius lost enduringfame, see Amm. Marc., xxvII.4.3.
54 In commandof the sacred
largesses (Amm. Marc., xxi.8.i) and PraetorianPrefectof Illyricum (xxi.i2.25).
55 Since this is the only passage bearingon Julian's possibleconnectionwith building activitiesin Athens, it is worth
quoting in full: "Ipsae illae bonarum artium magistrae et inventricesAthenae omnem cultum publice privatimque
perdiderant.In miserandamruinam concideratEleusina. Sed universas urbes opere imperatorisrefotas enumerare
perlongumest: scire satis est cunctasMacedoniae,Illyrici, Peloponnesicivitatesunis an binis epistulis maximi impera-
toris repentinam induisse novatis moenibus iuventutem, aquas locis omnibus scatere, quae paulo ante arida et siti
anhelantia visebantur, ea nunc perlui, inundari, madere, fora, deambulacra,gymnasia laetis et gaudentibuspopulis
frequentari,dies festos et celebrarivetereset novosin honoremprincipisconsacrari"(Claudius Mamertinus,Grat.act.
lul., 9.3-4). For an analysis of this speechsee H. Gutzwiller, Die Neujahresrededes KonsulsClaudiusMamertinusvor
dem KaiserJulian (Basler Beitrdgezur Geschichtswissenschaft10, 1942).
 22 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

First, Scylacius, Proconsul of Achaea after 343 and the subject of an encomium by Himerius.56
Before becoming proconsul Scylaciuswas a high official, probablyvicarius, in Asia, and much of
Himerius' oration is devotedto his importantwork in regulatingthe courseof the Maeander near
Miletus. His operationsthere are confirmedexplicitly in an epigramdiscoveredrecentlyin Laodi-
cea,57in which he is celebratedfor having constructedin 12 monthsan imposingbuilding (Oavpia),
probablya nymphaeum.
Although evidencefor concreteworks in Athens is lacking, Scylacius'reputationas an energetic,
wise, and considerateadministrator,togetherwith the floweryimplicationof Atheniandescentthat
runs through Himerius' oration, justifies the assumption that he probably left some tangible, if
© American School of Classical Studies at Athens

unidentified,contributionto Athens. Another possible benefactoris Flavius Hermogenes,Procon-

For personal use only. License: CC-BY-NC-ND.

sul of Achaea, probably in the reign of Constantius II, who at one time devotedhimself to philo-
sophy and, during his proconsulship,repairedthe harborat Corinth.58And finally Vettius Agorius
Praetextatus, Proconsul of Achaea between 362 and 364, another learned man.59He, like Cer-
vonius, was honoredwith a statue at Thespiae.60
That the physical aspect of the city had recoveredto some extent by the second half of the 4th
centuryis shown in the Expositio totiusmundi et gentium, which describesthe "landof Achaea' as
having these cities: Corinth and Athens. "Corinthis very active in commerceand has an outstand-
ing structureof an amphitheatre.Athens has the centersof higher learning and ancient historical
monuments and something worthy of special mention, the Acropolis, where by means of many
standingstatues it is wonderfulto see a so-calledwar of the ancients."61It may be concluded,then,
that from at least as early as the second quarter of the 4th century interest was being shown on a
high level in revivingthe city of Athens. The literaryevidencethat substantialmeasureswere taken
to increasethe water supply or improveits distribution,whetherby Cervonius,Julian, or ordinary
civic enterprise,is corroboratedby the rebuildingof the Antonineaqueductin the southeastcorner
of the Agora62and by the sudden surge in the constructionof new baths and the restorationof old
ones all over the city.63
56Himerius, Or. xxvii (xxv); PLRE, Scylacius 1; RE, s.v. Scylacius 1; Groag, Achaia, p. 34; and especially
to Groag,on thebasisof the titulusto
L. Robert,LaodiceeduLycos(footnote67 above,p. 11),pp. 339-351.According
Himerius' oration: rfjs ers8eL'cos apeo7raylTr?,yEv6olevos,Scylacius might have been named an Areopagite, but
Robert, following H. Schenkel in RE, s.v. Himerius, points out that this applies to Himerius himself rather than to
Scylacius.
57Robert, loc. cit.
58IG IV, 209. His careeris reviewedin Himerius, Or. xiv (XLVIII). See also PLRE, Hermogenes9; Groag, Achaia,
pp. 36-38; and Saffrey and Westerink,pp. xxxix-xl.
59 Above, p. 19. He also was eulogized by Himerius (Or. LI [Colonna]),but unfortunatelyonly the headingsurvives.
60
IGVII, 1855 (SEG XV, 322), also A. Plassart,"Fouillesde Thespies: Inscriptions,"BCH 50,1926, pp. 411 146.
61
Expositio totiusmundi et gentium, 52 (introduction,text, and commentaryby J. Rouge [Sourceschretiennes1241,
Paris 1966). For the date of composition(A.D. 359), see pp. 18-19. For a recentanalysis of this passage (and the trans-
lation used here), J. H. Oliver, "Achaia,Greece, and Laconica,"GRBS 21, 1980, pp. 75-81.
There is general agreementon the emendationarcemfor arcum,i.e., "Acropolis"for "triumphalarch".The "stand-
ing statues"have been taken as those dedicatedby Attalus, referredto in Pausanias, 1.25.2 (cf. Rouge's note to lines
14-16, p. 292). The whole descriptionof Athens is containedin a few lines: "Athenasvero <studia et> historias an-
tiquas et aliquid dignum nominatum,arcem ubi multis statuis mirabileest videre dicendumantiquorumbellum."
62 Agora
XIV, p. 202.
63
Travlos, PDA, pp. 180-181 for the locationof those known up to ca. 1970. For these and other hydraulicinstalla-
tions, as well as the continuingconcernfor all such activitiesinto the 5th century,see below, pp. 29-33. In some cases it
is not certainwhether these date from the 4th or the 5th century.
 DIOCLETIAN TO JULIAN THE APOSTATE (A.D.284-363) 23

Almost certainly, consideringJulian's predilectionfor Athens, he would have offered sums for
general rehabilitation,along with provisionfor other cities, after he came to the throne.64But there
was little time for direct personal interventionin the few months before he left for Antioch to pre-
pare for his Persian campaign, and in what time there was his attentionwas directedto the affairs
of the capital, e.g., the constructionof a new harborwith a stoa leading down to it, "sigma-shaped
rather than straight",and a new library "in which he depositedall his books."65He also made an
attemptto acquire from Alexandriathe obelisk which ConstantiusII had greatlydesired,66but his
efforts were equally unsuccessful, and it remained for Theodosius I to secure it finally in 390.67
During his stay in Antioch and his travels in the vicinity, Julian's sympathy and support were
© American School of Classical Studies at Athens

engaged by what he saw, such as the plight of Nicomedia, still in ruins after the earthquakeof
358.68In Athens, to which he never returnedafter his two months as a student, it is impossibleto
For personal use only. License: CC-BY-NC-ND.

attributea single building or public work to his initiative with any degreeof certainty.69The only
one to have been seriously proposedis the reconstructionof the interior of the Parthenon.70It has
been suggestedthat the new Pompeion, built over the ruins of the old by the Dipylon Gate to pro-
vide a new marshalinggroundfor the PanathenaicProcession,was built by Julian. The suggestion
was made on the basis of suitability, but it cannotbe said that among the emperorsin Late Antiq-
uity Julian alone seems to have given his attentionto Athens.71
Whatever the date of the actual building, the mention of "celebrations"in the encomia of Cer-
vonius and Mamertinus raises the question of the continuanceof the PanathenaicFestival. Was it
interruptedin 267 and, if so, for how long? It was certainlystill-or again-being held in the 4th
century because of Himerius' often quoted descriptionof the course of the procession:"The ship
begins its voyage right from the gate, as from a calm haven. Moving from that point, as over a
waveless sea, it is conveyed through the middle of the Dromos, which, descendingfrom above,

64
I.e., in the Imperial letters referredto by Mamertinus.
65
Zosimus,IIL.11.3. But G. Dagron (Naissanced'une capitale, Paris 1974, p. 90, note 1) inclinesto attributethe stoa
and the library to ContantiusII or even Constantinerather than to Julian.
66
Julian, Ep. 59 (58).
67 See also below, pp. 74-75.
68 Amm.
Marc., xvII.7.2-8; XXII.9.3-5. Cf. his abortiveattempt to restorethe Jewish temple at Jerusalem (G. W.
Bowersock,Julian the Apostate,Cambridge,Mass. 1978, appendix).
69 An
inscription published by G. Sotiriou (<<aAaLa XptLartavLK? BaA-Lct 'IXwl\ovy>>,
'ApX'E4 1919, pp. 29-30)
was later associatedwith Julian by A. Piganiol ("La couronnede Julien Cesar,"Byzantion 13, 1938, p. 248, note 1,
and idem, L'empire chretien (footnote 1 above, p. 16), p. 127, note 6); also P. Leveque ("Observationssur l'icono-
graphie de Julian dit l'apostat,"MonPiot 51, 1960, p. 118, note 2). It reads:
IOYAIANO[I
EEB
?A1 ANEOHK[E]
Other scholarshad either ignored or rejectedthe possibilitythat the inscriptionreferredto the Emperor(Sotiriou,loc.
cit.; P. Graindor,Album d'inscriptionsattiques,Paris 1924, p. 19, and idem, Athenesde Tiberea Trajan, Cairo 1936,
p. 3; also W. Peek, AM 67, 1942 [1951], pp. 41-42, on groundsof letter formsand unsuitability[absenceof titles, etc.]).
Piganiol's restoration(A. KAav&Los beforethe name is unacceptablebecauseit is clear from Sotiriou'sdrawingthat the
surfacesabove and to the left were uninscribed.The inscriptionwas carvedon a poros slab, re-used in its fragmentary
condition in the side of a sarcophagusin a late tomb beside the Basilica on the Ilissos, where it presumablystill rests,
once more underground.
70 For the case for the destructionof the Parthenon and reconstructionunder
Julian, see Travlos, 'ApX'.E 1973,
pp. 218-236. Contra,Frantz, "Julian."
71 The archaeologicalevidencefor the date of constructionof the new Pompeionwill be discussedbelow, pp. 26-28.
 24 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

straightand smooth,dividesthe stoas extendingalong it on either side, in which the Atheniansand

the others buy and sell."72
The literary evidencefor the date of the regenerationof the PanathenaicFestival is vague. Hi-
merius lived for at least 40 years inside the 4th century.There is nothing in the orationto indicate
that the resumptionof the festival was a recent event. The Panathenaia,as the most importantof
Athenian festivals, might be expected to receivethe highest priority in this still essentially pagan
city, and the first of the panegyricsto mention the revival of "celebrations"(Cervonius)might be
more reasonablyassociatedwith the Panathenaiathan the last; it might very well have precededthe
building of the late Pompeionby a few years, whetheror not at the instigationof Cervoniusor some
© American School of Classical Studies at Athens

as yet anonymousdonor, ca. 330. A desire to enhance the appearanceof the main entranceto the
city may also have played an importantpart in putting the attendantbuildingprograminto effect.73
For personal use only. License: CC-BY-NC-ND.

JULIAN AND THE REOPENING OF THE TEMPLES

The widely held belief that Julian reopenedthe Parthenon and the other temples of Athens is
unsubstantiatedand is based only on the assumptionthat they had alreadybeen closed. This is in
itself a dubiousclaim. Only one of the many edicts relating to the impositionof Christianityorders
specificallythe closing of all temples,74and the frequent reiterationof all the provisionsagainst
paganism in general shows how widely they were disregarded.Prompt compliancemight be ex-
pected least of all from the Athenians.75The reopening of the temples received attention from
contemporaryand later historians, but the only temples specifically mentioned in surviving ac-
counts are comparativelyinsignificant examples in the east, which was generally a much more
compliant area in this respect than the west. Both Rome and Athens, which might be expectedto
make a greaterimpact on the writers, are conspicuouslyabsent in the accounts.76
THE BUILDINGS
THE RECONSTRUCTION OF OLD BUILDINGS FOR OFFICIAL USE

In its now diminishedstate of importanceAthens' need for public buildings was limited:offices
for the Proconsul or Praetorian Prefect on their occasional state visits, lawcourts, and places of
assembly, where they or other officialscould addresslarge numbersof people, all find mention in
contemporarywriters. Many of Himerius' orationsare statedin the title to have been delivered"in
the praetorium"(ev rTOw rpaLrcoptto).The Praetoriumin the 4th century,as only a branchoffice,as
it were, might have been any building of suitable size and adequate state of preservation.That
would apply also to the lawcourts.The assemblyplace was a differentmatter.A long traditionhad
associated the meetings of the Assembly with theaterlike structures:first the auditorium on the
Pnyx and later the Theater of Dionysos itself (Pls. 5, 16:a). There is good reasonthereforewhy the
customaryterm for the assembly place in the writings of Eunapius and other 4th-centurywriters
was rTOOarpov. After the Herulian attack the only theater or odeion that was not beyond repair
was the Theater of Dionysos;moreover,this had been more and more commonlyused as the meet-
ing place of the Assembly even as early as the 4th century B.C.,77and so it was natural to follow
72Himerius, Or. in (xLvn).I2; trans. Wycherley,AgoraIII, p. 20, no. 1.
21.
73 See footnote48 above, p.
74 CTh xv. I
0.4. by ConstantiusII and Constans(date variouslygiven as 346, 354, 356).
Issued
75 For the legislation in general, see below, pp. 69-70, and for the effect on Athens in particular,Frantz, DOP.

76 The argumentagainst the reopeningof the temples by Julian is developedmore thoroughlyin Frantz, "Julian."
77Cf. K. Kourouniotisand H. A. Thompson, "The Pnyx in Athens,"Hesperia 1, 1932, pp. 137-138, with refer-
ences;also H. A. Thompson, "The Pnyx in Models,"Hesperia, Suppl. XIX, Studiesin Attic Epigraphy, History and
 THE BUILDINGS 25

tradition by rehabilitatingit for the same purpose. In addition to general repairs, a new stage, or
bema, was constructed,the gift of the archon Phaidros, as recordedin the dedicatoryinscription
carvedon the top step (Pls. 16:b, 17:a).78The salient featureof the bema was the sculpturedfrieze
adorningits front, which was taken from the previous,Hadrianic, phase of the theater.79
The long-accepteddate of ca. 400 for this renovationwas based on no solid ground, the archon
Phaidros being otherwise unknown except for the dedicationof a sundial;80and precise archaeo-
logical evidenceis lacking. For historicalreasonsit seems preferableto assign a terminus ante quem
of 345/6, the year in which the Prefectof Illyricum,Anatolius,visitedAthens and addresseda large
crowd "in the theater",very likely the Theater of Dionysos.81The importanceattachedto the as-
© American School of Classical Studies at Athens

sembly place at all times argues for an earlier rather than a later date for its reconstruction.Con-
For personal use only. License: CC-BY-NC-ND.

sideringthe distinctivelypagan flavorof the sculpturedfrieze of the bema, with its Dionysiac scenes
and crouching sileni, as well as the dedicationto the "loverof mysteries"and the more tolerant
attitude of Constantinethan of his successors,a date in his reign seems preferableto a later time.
For other civic needs after 267, makeshiftquarterswere providedat first by patching up build-
ings that had not been completelydestroyed(Pls. 3, 6). The earliest attemptsin this directionwere
made at least as early as the middle of the 4th century.The Tholos, where traditionwas strongest,
was the first building to be rehabilitated.It had suffered enough to necessitatebuttressingwith a
ring of concrete masonry82but not enough to be demolishedcompletely.The marble floor of the
preceding period and the brick dome were still serviceable,and the interior walls were now em-
bellished with marble slabs. The building went out of use finally at the end of the 4th century,
probablyas a result of the attackby Alaric in 396.83
The Metroon was destroyedor at least badly damagedin 267. That it never regainedits original
function is implied in the EmperorJulian's referenceto the Metroon, "whereall the documentsof
the Athenians used to be kept."84Nevertheless, two rooms were rehabilitatedto a modest degree:
the north room in basilica form, the next to the south as some sort of eating establishment (see
P1. 6). This was achieved with a minimum of effort, the rebuildershaving gained ca. 1.50 m. of
extra height for the walls by lowering the floorby that amount.The chief featureof the roomwas a
pair of broiling pits, ca. 6.70 m. and 7.20 m. long, with a maximumdepthof ca. 0.55 m. and a width
of ca. 0.55 m. Both were filled with loose ash and charcoaland a large numberof chickenbones.
This phase of the Metroon was of short duration, for it was destroyedor damagedat the end of
the 4th century, probablyin Alaric's attackin 396.85

Topographypresentedto Eugene Vanderpool,Princeton 1982, p. 145, for Lycurgus'interestin completingthe theater

as an assembly place (with further references).
78IG II2, 5021.
79 Cf. M. Sturgeon,"The Reliefs on the Theater of Dionysos in Athens,"AJA 81, 1977, pp. 31-53 for a recentstudy
of the frieze.
80 IG II2, 5208.
81
For details of the argument,see A. Frantz, "The Date of the PhaidrosBema in the Theater of Dionysos,"in Hes-
peria, Suppl. XX, Studies in Athenian Topography,Architectureand Sculpturepresented to Homer A. Thompson,
Princeton 1982, pp. 34-39.
82 For the latest
history of the Tholos, see Hesperia, Suppl. IV, pp. 54, 101, 136-137; AgoraXIV, p. 210.
83
Pottery and lamps of the 4th and 5th centurieswere found in the debris over the floor;the latest coins were from
the reign of Arcadius.
84Julian, OratioVIII
(v).I59B.
85
For its later history and the possibility that during the whole of the Post-Herulian periodthe remodeledMetroon
was used as a synagogue, see below, p. 59.
 26 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

It was in either the late 4th or early 5th centurythat the Agora was given a moreformal appear-
ance with a series of statues, of which the bases remain, at the northeastcornerof the Metroon.86

THE POMPEION AND THE PANATHENAIC WAY

Sometime during the 4th century the Panathenaic Way took on a new appearance with the
rebuildingof old and constructionof new stoas borderingit on both sides from the Dipylon to the
point where the street turns into the Agora (P1. 4). A building was erectedover the ruins of the
Pompeion and its successor(Magazinbau), a substantialstructureof the Antonine period, and, on
the same axis, some six metersto the east, was a triple arch or gateway.87Presumablyall these were
© American School of Classical Studies at Athens

connectedwith the revival of the Panathenaia, the building over the Pompeion to provide a new
marshalinggroundfor the processionand the Festtoras a ceremonialstartingplace.
For personal use only. License: CC-BY-NC-ND.

The Late Pompeionconsistedof two stoas borderingan open space, abuttingagainstthe city wall
at the west end, closed at the east. Mos the the constructionmaterial (columns and other archi-
tecturalfragments)was salvagedfrom assortedbuildingswhich had probablysufferedin the Heru-
lian attack.Of the Festtoronly the foundationsfor the four massivepiers remain.88
Onthe
thebasis
basisofof all the evidencethen available Hoepfner concludedthat the building of both the
Late Pompeion and the Festtor took place about the middle of the 4th century. This dating has
recentlybeen called in question;it thereforeseems advisableto re-examinethe evidencein the light
of informationfrom the Agora excavationswhich was not availableto Hoepfner.
There is little concreteinformationfor the date of constructionof the two buildings,but the broad
framework of the history of the area of the Pompeion in Late Roman times is clear and well
documented:
1) In 267 the Magazinbauwas destroyed.Lampmakerssoon took over the whole districtaboveand
aroundthe ruins and continuedto beactive
active throughthe first part of the 4th century.89
2) By the middle of the 4th centurythe districtonce more was servingthe needs of the Panathenaic
Festival, accordingto Himerius' testimony.90
3) In 396 Alaric's forces inflictedheavy damage from the Dipylon allalong the PanathenaicWay
and into the Agora,91after which the lampmakersreturnedand took up their trade once more.
Himerius saw the Panathenaicprocession,but what actual buildingsdid he see? Certaintyexists
only in the case of the long stoa borderingthe PanathenaicWay on the south side just before the
street enters the Agora. Its predecessor,built at the end of the 1st or in the 2nd century, was de-
stroyedin 267, and no rebuildingtook place beforethe middle of the 4th century.92Moreover,this
For these see below, p. 60.
86

For the definitivepublicationof the Pompeion in all periods, see Hoepfner. The reporton the investigationand
87

dating of the latest phase, from which the present accountis largely drawn, will be found on pp. 176-199, with refer-
ences. The two new buildings will be referred to as the "Late Pompeion"(Hoepfner's Hallenstrasse) and Festtor,
retaininghis term for what was evidentlya ceremonialgateway,perhapssimilar to one at the entranceto the stadiumat
Delphi.
88 Some doubt has been raised as to whether the
gate was ever finished, since no pieces of the superstructurewere
discovered, but all the material may have been carriedaway for use in other buildings,with a thoroughnessnot unpar-
alleled elsewhere in Athens.
89 See also AgoraVII, p. 20, for the early 4th-centuryrecoveryof the lamp industry.
90 Above, pp. 23-24.
91 Hoepfner attributedthe damageto the earthquakeof 375 ratherthan to Alaricon the basis of Zosimus'reportthat
Alaric left Athens unharmed.But Athens escapedserious damagein the earthquake,and there is now ample evidence
of destructionin 396 in the Agora (below, p. 52) and in the Kerameikos,where coins going down to the reign of Arca-
dius (A.D. 383-408) attest to destructionat that time (AgoraVII, p. 63).
92 For the
evidence,see below, p. 28. Its counterparton the north side of the street undoubtedlydates fromthis same
time, but it is in too fragmentarya state to permit close dating (below, p. 28).
 THE BUILDINGS 27

is the only stretchof the street that answers to Himerius' descriptionof "dividingthe stoas on either
side".Disturbancescausedby the lampmakersremovedmost of the evidencefor the date of the Late
Pompeion. A 4th-century date for the Festtor seemed to be established by the deep filling in its
foundationtrench,but a few sherdsattributedto the late 4th or 5th centuryin the same filling seem
to underminethe earlier dating and thereforebring both buildings down to the 5th century.93This
evidencecannot be ignored,and so the question must be reopened.
A 5th-centurydate presentsgrave difficulties.Although the Late Pompeioncannotbe datedpre-
cisely by either pottery or coins, a sequence of developmentscan be tracedbetween stages 1 and 3
noted above. The cessationof activityof the lampmakers,so well entrenchedthrough the first half
© American School of Classical Studies at Athens

of the 4th century, implies some reason, such as clearing the ground for new construction.On the
For personal use only. License: CC-BY-NC-ND.

more positive side, the Late Pompeion shows two distinct periods, with a rebuildingalmost from
the ground. A new floor was laid, ca. 0.30 m. higher than the first, and on the stump of the south
wall a new one was superimposed(PI. 18). The new building was evidentlyof flimsier construc-
tion, its wall being only 0.60 m. thick. It extendedsix metersfartherto the east, perhapsto compen-
sate for greater destructionat the west end, close to the city wall. That the destructionwas caused
by Alaric was left in no doubt by the massive debris, all over the area, abandonedby the lamp-
makerswho once more movedin immediatelyafter the disaster.94
The Festtor was evidently destroyedby Alaric (if, indeed, it was ever completed)and never re-
built. Although the evidencefor its constructionis contradictory,its destructionand plunderingcan
be fixed with some certaintyin the first half of the 5th centuryby coins foundin the northernpart of
the remaining foundations.95The walls of the secondphase of the Late Pompeioncarriedright up
to the foundations of the Festtor and consequentlypostdate its destruction.Hoepfner concludes
thereforethat the secondphase dates from the first half of the 5th century.The columnsthat were
found over and aroundthe Late Pompeionwere very likely used in both phases of its construction.
Some additions may have been necessary, which would perhaps account for the presence of two
impost blocks,96imposts not being known in Greece beforethe 5th century.97
The published studies of impost blocks are directedmainly toward the decoratedtype, well in-
tegratedwith the general architecturalscheme,ratherthan the undecorated,purely utilitarianform
designedto effect a transitionfrom the capital to the arch. It might be notedthat Greece,where the
chief sourceof columns in Late Antiquity was the cannibalizationof ancientbuildings,would have
been a fertile ground for e ro the type, since thelate
columns in buildings (as in the
Pompeion) were often of disparateheights, which could be rectifiedby the use of the impost.
To revert to the contradictoryevidence for the constructionof the Festtor, i.e., the 5th-century
sherdsfrom the foundationtrench, some doubtmay be entertainedabout its validity.The rest of the
fill, which was deep, was uniformly of the earlier part of the 4th century. The stratificationwas

93This was pointedout to me by J. P. Binder.

94
Some 16,000 lamps were found on the site. First thought to have been from Alaric's destruction,they are now
attributedto these artisans,accordingto J. P. Binder'sdating of the lamps of this period,revisedfromthat publishedin
Agora VII. (For a note on the redatingsee H. Williams, Kenchreai,V, The Lamps, Leiden 1981, p. 104.)
95 Six were of the period 395-450, one each from the time of Valentinian II (A.D. 388-396) and Honorius
(A.D. 413-423), and one from the Middle Ages.
96 Hoepfner, p. 179, fig. 193 and p. 185, fig. 200.
97 F. W. Deichmann, Studien zur Architektur Konstantinopels im 5. und 6. Jahrhundert nach Christus (Deutsche
Beitrdgezur Altertumswissenschaft4), Baden-Baden1956, pp. 41 ff. Deichmann notes, however,that the subjecthas
not been thoroughlystudiedand that little chronologicalevidencehas been found. He is of the opinionthat the origin of
the type is probablyto be found in Greece.
 28 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

very complicated,and the possibilityof human errorcannotbe excluded.98In view of the activityof
the plunderers,e.g. the coins cited above,some intrusionmight be anotherexplanation.
Closely related to the problem is the history of the Panathenaic Way. The stoa borderingthe
south side of the Way between Dipylon and Agorahad a long history. Built in the secondhalf of the
1st to the early 2nd century, it was in use, with several modifications,until 267.99It was almost
completelydestroyedby the Heruli, who left heavily burntdebrisoverthe floors,followedby an ac-
cumulation of pottery of the first half of the 4th century, with coins of ConstantineII (A.D. 337-
341) and Constantius II (350's). The stoa was refurbishedin a substantialway, evidently in the
middleof the 4th century,including,probably,the large statuebase in the propylonat its south end.
This must have been one of the stoas seen by Himerius, since the site had lain desolateafter the de-
© American School of Classical Studies at Athens

structionof its predecessor.Destructionin 396 is well attestedby layers of ash and debris over the
For personal use only. License: CC-BY-NC-ND.

floors,with coins of the 380's.100This stoa, then, and Phase I of the Late Pompeionmay be consid-
ered contemporary.The Festtormay be includedwith these becauseof the close architecturalrela-
tionship and the fact that it was already destroyedand being plunderedby the time of Phase II.
Might not the plunderers,in fact, have been the buildersof Phase II?
The three buildings, Late Pompeion (Phase I), Festtor, and the stoa on the PanathenaicWay,
form a well-integratedgroup. They can easily be thought of as part of the revivalof Athens during
the reign of Constantineand a little later, which, incidentally,coincideswith the time of the known
benefactorsof the city with their emphasis on "celebrations".'01 The Panathenaia,as the most im-
portantof Athenian festivals,might be expectedto receivethe highest priority.A desire to enhance
the appearanceof the main entranceto the city may also have played a part.
Remains of the stoas on the north side of the street between the Dipylon and Agora have been
uncoveredover a considerablestretch.1'02 They are in too ruinous a state to provideany useful evi-
dence of their date, but they were probably part of the same program.A Late Roman bath (5th
century?)was built overone part, and so here at least there was no rebuilding.The final stretchhad
no stoa on the north side owing to the continuedpresenceof an Early Roman temple and a monu-
mental altar of the classical period, discoveredin the 1980-1982 campaignof the Agora Excava-
tions.103These were destroyed,probablyin 396, and their place was taken by a 5th-centurystoa
which abuttedagainst the west wall of the Stoa Poikile.104
The course of the Panathenaic Way after it entered the Agora was more informal (Pl. 6). Its
eastern edge remainedmore or less stable, along the line later followed by the Long Late Roman
Wall and the mill aqueduct.105But the western edge became quite undefinedin Late Antiquity,
particularlyin the northernsectionwhere the graveledsurfacesufferedcontinuallyfrom the accu-
mulatedsilt broughtby the torrentsrushingdown the slopes as the drainagesystembecameless and
less effective.
Toward the southeastcornerof the Agora the courseof the road was deflectedwestwardto avoid
the towers of the Post-Herulian Wall and the mill aqueduct,revertingto its original line only after

98Dr. Hoepfner, who was presentat the excavation,kindlygave this opinion in answer to an inquiry from me about
the details.
99For its final explorationsee Hesperia 42, 1973, pp. 370-382 (Shear,Jr.), with earlier references.
100For details, see below,
p. 54, and Shear, op. cit., p. 380.
101Above, p. 21.
102 AAA 4, 1971,
Vanderpool (footnote 22 above, p. 15), pp. 289-297; Y. Nikopoulou, <TowoypaLcKa 'AOvwv?v>,
pp. 1-9.
103
Hesperia 53, 1984, pp. 24-37 (Shear, Jr.).
104
Ibid., p. 16, fig. 10.
105Below, pp. 80, 81.
 THE BUILDINGS 29

the latter had turned east to pass through the wall.106Above this point the road followed the old
Dromos, even though Towers W 3, W 2, and W 1 (proceedingsouthward) stood squarely in the
middle (Pls. 5, 9). The difficultywas not serious,however,becausethe groundlevels of the roadand
of the area between its west side and the Antonine aqueduct were the same, and the intervening
space was ample to make up for the loss causedby the towers.
Traffic over the road diminishedgreatlyafter the Slavicinvasionof the late 6th century,when the
stretchon the slopes becamegraduallycoveredover with a layer of silt. During the 7th centurythe
road was used by the inhabitantsof the makeshifthouses built over the ruins on either side:those of
the "schools"on the west and miscellaneousbuildingson the east.107But these buildingswere short-
© American School of Classical Studies at Athens

lived. Most of the houses on the east side of the street were destroyedby fire, those on the west by
neglect. The western area was not rebuilt until Turkish times, but the east becameone of the most
For personal use only. License: CC-BY-NC-ND.

prosperousresidentialdistrictsin the 11th and 12th centuries.

WATERWORKS
WaterSupply
Throughout Late Antiquity, from the 4th century down to the Slavic invasion of the 580's, an
effort was made to keep in working order at least some of the fresh-waterand drainagesystemsof
the classicalperiod. In some cases a thoroughcleaningwas all that was needed;in othersthe Heru-
lian damagewas so severeas to necessitatesubstantialrebuilding,usually, however,permittingre-
use only in a much reduceddegree.
Of the water systems of the classical period only the Antonine aqueductsurvivedthe destruction
of 267. It too was seriouslydamagedbut was put back into serviceon a diminishedscale. In connec-
tion with that aqueduct an elaboratesystem of reservoirs,chambers,and passagesdating from the
1st century had been incorporatedon both sides of the PanathenaicWay.108Being largely under-
groundthey survivedthe Herulian attackand continuedin use at least throughthe 5th centuryand
perhaps into the 6th, but no later. Another source of water in the 6th century was a huge cistern
constructedagainst the inner face of the Post-Herulian Wall, apparentlya successorto one of the
classical period immediately below the Klepsydra, called by A. W. Parsons the Paved Court.109
Parsons attributedthe rehabilitationof the Klepsydrasystem in this much reducedform to Jus-
tinian's interest in the problemof the water supply (Procopius,de aed. LV.2.24). He could find no
evidence for the duration of the re-use beyond the fact that after an apparent period of disuse the
ancient sourcewas sought and found in the 10th or 11th century.
Mention should be made too of an independentwater line whose sole purpose appears to have
been to furnish water for three mills along the east side of the Agora.10 The other majorsourceof
water was the great numberof wells sunk in the Agora itself but more especiallyin the environs.

Drainage
The Great Drain, with its main north-south line along the west side of the Agorathroughoutthe
classical period, continued to function, but with decreasing efficiency, through the 5th century
(PI. 17:b). In the beginning it had been frequently cleaned out, but as time went on the layers of
accumulateddebris and silt became thicker.A stretchof the south branch, after it crosses Piraeus

106 Below, p.
81.
107
Below, pp. 117-120, for these buildings.
108
Hesperia 7, 1938, p. 329 and 8, 1939, p. 218 (Shear, Sr.).
109
"Klepsydraand the Paved Forecourtof the Pythion,"Hesperia 12, 1943, pp. 191-193.
110Below, pp. 80-81.
 30 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

Street, may be taken as representative.111 Here at the bottom lay a late Hellenistic deposit only
0.20 m. thick. Above it was a layer of pre-Herulian times 0.50 m. thick and on top of that a post-
Herulian accumulation0.60 m. thick with coins dating down to the middle of the 3rd century.
Debris gradually settled on top of this, rising very nearly to the top of the drain. That water con-
tinued to run at this high level is demonstratedby the heavy water wear and lime deposit on the
undersurfacesof the coverslabs. Coins fromthis level went down as late as the 5th century.Wheth-
er the drain continuedin use into the 6th centuryis uncertain;if so, it could have allowed no more
than a trickle, and it was certainlynot functioningafter the 6th century.This patterncan be paral-
leled in the stretchof the drain in front of the Tholos.112
© American School of Classical Studies at Athens

Baths
For personal use only. License: CC-BY-NC-ND.

The Agora area was well supplied with unpretentiousbaths fromthe end of the 4th centuryuntil
the end of the 6th. With
the in connectionwith the Palace of the
exception of the small bath builtthe
Giants, all were located in the southwestern section.113
The baths had much in common.At least
two were built on the sites of earlier baths. All except one were small, compact establishments
intended for use by only a few people at a time. All conformedto standardpractice in their in-
stallationswhile differingin minormattersof arrangementand decoration.The hypocaustcolumns
were invariablymade of circulartiles. In most cases evidencewas found, in the presenceof tegulae
mammatae, for the circulationof warm air through the wall cavities as well as under the floors.
Their use in all the baths may be presumed.
Methods of constructionwere similar, at least for the roomsrequiringthe use of water, the walls
being made of rubblemasonrywith hard, gray lime mortar.The pools themselveswere lined with a
hard, pinkish cementwith much ground-uptile mixed with it.
In the center of the industrialdistrict,in the low groundat the northwestend of the Areopagus,
were two small baths, both in use during the 4th andth hcenturies.114The more elaborateof the
two, the East Bath, was originally built in the 2nd or early 3rd century,destroyedin the Herulian
raid, and rebuilt in the form in which it was found sometime in the 4th century (PI. 19:b). No
evidenceremainedof an earlier phase for the West Bath on accountof the deep disturbancesmade
all aroundit in Byzantinetimes, but it is not impossiblethat the two buildingswere contemporary.
In view of this probabilityand their proximityit has been suggestedthat the East Bath was for men
and the West for women.115
The East Bath overlay the Great Drain, its east foundations resting on the east wall of the
drain.116The building measuredca. 10 x 20 m. overall and consistedof the normal elements.The
main entranceled into the apodyterionfromthe west by way of a long rectangularcourtyard.From
the apodyterion one passed successively through twofrigidaria, the tepidarium, and the caldarium.
The first and largerfrigidarium containedtwo pools, one in the west, apsidal, end and another in
the east. The smallerfrigidariumhad an apsidal pool at the east end;at the west a doorwayled into
a service court in which was a well, undoubtedlyfor the use of the bath. The tepidariumwas
111Young, "IndustrialDistrict,"p. 285.
112
p. 121.
Hesperia, Suppl. IV,
113For this
bath, below, pp. 107-108; for the bath in House C see below, p. 87.
see
114
Young, "IndustrialDistrict,"pp. 279-283, with detaileddescriptionsof both baths. Cf. Hesperia 17, 1948, plan,
p. 164, fig. 6 (Thompson).
115Thompson, op. cit., p. 169.
116This section of the drain had gone out of use after Sulla's attack in 86 B.C. (Young, "Industrial District,"
pp.279-283).
 THE BUILDINGS 31

comparativelysmall, almost square. The much larger caldariumhad two adjacentpools along the
north wall.
The bath, although small, seems to have been fairly luxuriouslyappointed,even in its final phase.
The apodyterionwas paved with marble slabs of various colors, laid in cement. The largerfrigi-
darium was paved in marble between the two sunken pools, and a thin revetmentof white, green,
and pink marble coveredthe walls. The concretebedding with impressionsof the paving tile and
slots for the revetmentwas preservedonly in the easternend but presumablycoveredthe entire area
(PI. 20:a).
The West Bath, which lay to the west of the Street of the Marble Workers (PI. 3: B 18), was
© American School of Classical Studies at Athens

apparently smaller and simpler, although it must have been larger than the present remains
For personal use only. License: CC-BY-NC-ND.

indicate. When discovered,it consistedof only three rooms, two along the east side and anotherat
the northwest. That the southernmostof the east rooms was the caldarium was shown by the
passage from the furnace at the south end and part of the east wall of the furnaceitself, with layers
of ash covering the groundjust south of it. Two rows of hypocausttiles remained in place at the
south end of the caldarium, and their tile bedding was preserved throughout the whole room
(PI. 20:b). The adjoining room to the north would have been the tepidarium,and the remaining
room,on the west, therefore,thefrigidarium.The apodyterionmay have occupiedthe empty corner
of the rectangle. The presence of tegulae mammataeand pieces of marble revetmentin the de-
structionfill indicateda similar wall treatmentto that in the East Bath.
The survivalof only three rooms may be explained by the fact that for the rooms in which water
was used the constructionwas of heavy rubble and concrete,whereas, as demonstratedby the East
Bath, the rooms which had no need of waterproofingwere built of dry stone and many re-used
blocks. These rooms were used as quarries by Byzantine builders who stripped the West Bath
entirely and the East Bath in large part.
Another small bath lay on the Areopagusslope at I 17 about midway between the large bath at
D-F 17-18 and the Philosophical School A (Pls. 3, 19:a).17 In size it is comparableto the East
Bath, but its plan is more compactbecause,whereas the roomsof the East Bath are stretchedout in
a row, here the apodyterionis tucked in on the west side between the apses of thefrigidarium and
the caldarium.Some earlier structureof unknownpurpose underlaypart of the bath, and re-use of
at least one of its walls may perhaps account for the awkward junction of the south wall of the
apodyterionwith the west apse of the caldarium.
The apodyterion,which was paved with marbleslabs, was enteredthrougha doorwayin its west
wall; another doorway in the northeastcornerled into thefrigidarium.Thereafterthe progression
was regular, from north to south, from the frigidarium, through the two tepidaria, and into the
caldarium, the last two doorways being on axis, the first only slightly off. The frigidarium had
apsidal pools at both east and west ends, both floors and walls being paved and revettedwith Hy-
mettian marble. The room was not quite symmetrical:the apses were not on axis either with the
central part of the room or with each other, and the nearly square central area tapered slightly
towardthe north.At the southeastcornerthe wall was angledback slightly to allow sufficientwidth
for the doorway into the first tepidarium. The tepidaria were small, the first measuring ca.
2.60 x 1.45, the second ca. 2.10 x 1.60 m. Furnaces were found outside the east wall of each. An

The bath had alreadybeen discoveredand partially excavatedduring the investigationof the Areopagusslopes in
117

1897 by W. Dorpfeld, who called it a Roman villa. Most of the overlyingand surroundingfill was removedat that time,
and the foundations were left exposed until they were finally covered with an accumulationof earth and rubbish.
Consequentlythe building, re-excavatedin 1933, can now be dated only within very broadlimits.
 32 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

apse on the west side of the caldariumand rectangularprojectionson the east and south provided
space for three small plunge baths, each of which was equippedwith its own furnace.
The original east and south walls were roughly stuccoedon the outside and intendedto be seen.
At some later time they were reinforcedby a skin of rubble masonry, probablyto protect them
against erosionfromwater tricklingdown the slopes. In both inner and outer masonry,gray mortar
of good quality was used, and there need be no great differencein time between the two.
The problemof water seepage from the springs abovewas serious, and ingeniousmeasureswere
taken to cope with it. A rectangularterracottachannel, measuring0.14 m. wide and 0.12 m. deep,
was laid under the building on its north-south axis, beginning at the south end. Two transverse
© American School of Classical Studies at Athens

drains, entering from east and west under the caldariumon its axis, met at the main channel and
pouredtheir contentsinto it. The main channelcontinuednorthwardas far as the middleof the first
For personal use only. License: CC-BY-NC-ND.

tepidarium,where it met another east-west channel. This westward drain then carriedthe whole
flow beyondthe limits of the building towardthe northwest.
As noted above,most of the chronologicalevidencefor the constructionand life of the bath disap-
pearedwith the 19th-centuryexcavations;what little remainedpointedto constructionin the 3rd or
4th centuryand use throughthe 5th and probablyinto the 6th.
The largestand most complexof all, the SouthwestBath (Pls. 6, 19:c,d, 20:c), was a public bath,
a sprawling structureoccupying an irregular, more or less triangular space boundedby Piraeus
Streeton the north,AreopagusStreeton the east, and on the west a high retainingwall which, since
the 4th century B.C., had supportedthe terraceon which the bath was built. From the beginning,
any building erectedon the terracewas limitedto an irregularform,on the one hand by the scarpof
rock along the Areopaguson top of which the road always led down from the southwest,and on the
other hand by the sharp drop in the groundlevel towardthe west which had led, from early times,
to the constructionand maintenanceof the retainingwall. This accountsfor the somewhathetero-
geneous characterof the plan of the building, especiallytoward the middle, since the roomson the
east side had been aligned with AreopagusStreet while those on the west had taken their line from
the retainingwall.
The bath in the Late Roman period followed a long traditionof bathing establishmentson the
same site.18 It was in its third, 2nd-centuryphase that it stood in 267, when some parts suffered
total, others partial destruction(PI. 20:c). The whole building lay in a semiruinousstate for some
75 years, to judge by a consistentsprinklingof sherdsas late as the middleof the 4th centuryfound
in the top of the thick layer of fill otherwisetypical of Herulian destruction.
Although the generaloutlines of the bath remainedthe same, some radicalchangeswere made in
the internal arrangements(PI. 19:d). The main entrance was now moved to the east, using the
ruined south tepidariumas a broad passageway to which access was gained through a doorway
opened in the east wall. The northwestcornerof the complex was completelyreorganized,and all
the rooms involvingthe use of water were concentratedalong the north side of the building. With
this change,the remainingpart of the building becamequite distinctin character.The most impos-
ing feature was the vaulted west hall, createdby throwing together two rooms and coveringthem
with a marble-chipmosaicfloor.At the south a large exedra,similarlyfloored,was added.Benches
were built all aroundthe exedra, against all wall surfacesin the hall itself, and against the walls of
all the "dryrooms".
The presence of benches around the "dryrooms"suggests a common purpose. Although com-
monly associatedwith apodyteria,their suddenproliferationat a time when actualbathingfacilities
118
For the earlier as well as the later periodssee Hesperia 38, 1969, pp. 394-415 (Shear,Jr.). In view of the detailed
presentationof the bath in this article, it will be necessaryto describeit here in only summaryform.
 THE BUILDINGS 33

were so curtailedseems illogical. Anotherexplanationis more plausible, i.e., that they were used as
places of instructionby teachers who had not yet achievedthe status of having buildings of their
own, such as those on the slopes of the Areopagusfarthereast.119
The bath as such met its final destructionin the late 6th century,undoubtedlyduring the Slavic
invasion of the 580's. Not long after, probablyduringthe first half of the 7th century,it was rebuilt
in a sketchy manner, but not as a bath. Some parts went out of use entirely; the west hall was
divided into two rooms of unequal size, separatedby a flimsy crosswall. The marble-chipmosaic,
where it survived,once more servedas a floor, but where it was badly damagedit was not repaired
substantially.The purpose of the building in its final form is obscure,but one may conjecturethat
© American School of Classical Studies at Athens

again it might have been a school,built to replaceone or morethat the Slavs had destroyed,and was
by now completelyChristianized.120
For personal use only. License: CC-BY-NC-ND.

Probablythe most elegant of all the baths in the Agora area was one closely connectedwith the
Palace ofthe Giants (Pls. 6, 20:d), at the cornerof the south complex and overlyingthe foundations
of the Middle Stoa. Thefrigidarium occupiedthe northernhalf: a long rectangularroomwith three
water basins, or plunge baths, along its north side. One of these was semicircular,the otherssmaller
and rectangular.To the south were two rooms heated by hypocaust;the larger, apsidal, was the
caldarium,the other, circular, probablythe sweating room (laconicum).The rooms to the east of
these rooms were probablyapodyteria.The main rooms were luxuriously fitted with marble re-
vetment for walls, floors, and even the lining of the plunge baths. The door jambs also were of
marble.121A large cistern to the southeastfurnishedwater for the bath.

Latrines
Athens, as far as presentevidenceindicates,was poor in public latrinescomparedto othercities of
the Roman world and earlier. Only two of any consequence are known.122 The first was an im-
posing square marble structurewith four Corinthiancolumnssurroundingan open court. It stood
just outside the Roman Market near the Tower of the Winds (PI. 5) and could accommodateabout
60 persons.123The other, near the southwest entranceto the Agora, was built in the 1st century
after Christ. The first was destroyedin the Herulian attack and was never rebuilt. The second
suffered severe damage at the same time but was restored to use in successively curtailed
forms according to the vicissitudes of the Southwest Bath, with which it was closely connected
(Pl. 21 :c).124It was poor by comparisonwith the establishmentnear the Tower of the Winds. Its
walls were of brick,the floor was of stucco,and the proportionswere modest.In its latest phase the
stucco floor was replaced with one of pebble mosaic. By this time its capacity had shrunk from
about 20 to about 14. The trencheswere flooredin the traditionalway with pan tiles except where
the two main drains of the bath could be used directly.

119 Below, pp.

37-48.
In this connectionit is well to recall that Theodoreof Tarsus, beforebecomingArchbishopof Canterburyin 669,
120

had studied philosophy in Athens and was well versed in Greek and Latin literature,both sacred and secular. (For
references,see Frantz, DOP, p. 199, note 78).
121 See Agora XIV,
p. 212. For a fuller description,see below, pp. 107-108. For a small bath (PI. 21:a) in the final
phase of House C, see below, p. 87.
122 Remains of a third structure,
previouslyidentifiedas a latrine,just behind the Stoa of Attalos (Hesperia 29, 1960,
p. 340, fig. 4 [Thompson]),are now attributedto the colonnadeon the north side of the East ColonnadedStreet (Broad
Street), below, p. 67.
123 Travlos, PDA, pp.
342-344.
124
Hesperia 38, 1969, pp. 405-406, 412-415 (Shear, Jr.). Another,of which only traces remain, was also conven-
iently placed near the southwest entranceto the Agora but did not survivethe attackof 267.
 34 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

Inside the Post-Herulian Wall, at the junction of the East ColonnadedStreet (BroadStreet) and
the narrower street leading off to the south, and just before the monumentalstairway leading up
toward the Roman Agora,125was a latrine evidently,from its location,public, but small (PI. 21:e).
Originally built in the late 4th or 5th century,it was repairedand slightly modifiedin the 7th.
Some of the larger buildingswere equippedwith minimal facilities.The only provisionfor these
needsto be discoveredin the whole of the Palace of the Giants is a small installationin the northeast
cornerof a room opening off the north side of the southeast (smallest) court (Pl. 21:f). It is of the
standardtype used in the houses of an earlier period, flooredwith pan tiles sloping down gently to
empty into an already existing drain which led from the courtyardand on out through the north
© American School of Classical Studies at Athens

wall. The size and locationmake it probablethat this was for the use only of the officialwho had his
For personal use only. License: CC-BY-NC-ND.

quartersin the southeastwing. No trace was found of any larger establishmentfor general use.
A latrine of similar type, but with two trenchesat right angles to each other, was installed in a
room to the right of the entrance to House A on the Areopagus (Pls. 21:b, 26:a).126Water was
providedby the pipeline that descendedfrom the springhouseat the southwest cornerof House B
and ultimately zigzagged across the courtyard of House A. There were some indications of a
plumbinginstallationto the left of the entranceto House B, but the erosionof the whole northedge
of the building preventedcertainty. The northern edge of both houses to the south was also too
chopped up to preserveany traces of the latrines that must almost certainlyhave existed. The late
(7th-century) level in House C, however, preserveda primitive one in the room at the southeast
cornerof the great peristyle (Room 4).
A 6th-centuryhouse, southwest of the exedra of the Great Court of the Palace of the Giants, is
tenuous in plan; neverthelessit had a well-built latrine connectedwith it, with a drain leading
directlyfrom it to the Great Drain (PI. 21:d).127

DOMESTICARCHITECTURE
Private Houses
House plans in Athens changedlittle from classicaltimes throughLate Antiquity. In its simplest
form the typical house consisted of a few rooms grouped around a small courtyard,and this ar-
rangement continued to be the norm in the poorer or more crowded districts in a city. Further
developmentwas by elaborationrather than fundamentalchange.128In areas where more space
was available,houses tendedto be larger, and a peristylewas addedto the court.This type became
established in the Hellenistic period and continued as the standardin the Mediterraneanworld
throughoutantiquity. The only striking innovationwas the introductionin the 1st centuryB.C. of
125
Hesperia 42, 1973, p. 390 (Shear,Jr.).
126
Below, p. 39.
127A numberof
examples of the same type have been found in privatehouses to the west and south of the Areopagus.
It may be assumed that this was a unit commonto the averageAthenian house in Late Antiquity. For latrines in the
houses of Athens in earlier periods, see Hesperia 29, 1959, pp. 101-102 (Thompson).
128 For the
developmentof the Greek house, see D. M. Robinsonand J. W. Graham,Excavationsat Olynthus,VIII,
iii, The Hellenic House, Baltimore 1938; Travlos, PDA, pp. 392-401; C. Krause,"Grundformendes griechischen
Pastashauses,"AA (JdI 92), 1977, pp. 164-179. See also J. C. Balty, "Notes sur l'habitation romaine, byzantine et
arabe d'Apamee,"ColloqueApameede Syrie, 29-31 mai 1980, Brussels 1981, pp. 471-503.
For a survey of the houses of the Roman period in Greece, both with and without peristyle, see J.-P. Sodini,
"L'habitaturbain en Grece a la veille des invasions,"in Villeset peuplementdans l'Illyricumprotobyzantine(Actesdu
Colloque organise par l'Ecole Fran,aise de Rome, Rome, 12-14 mai 1982), Rome 1984, pp. 341-397.
 THE BUILDINGS 35

an apse at the end of the largest room in the house,129and this became increasinglyfrequent in
succeedingcenturies.
In general the inhabitants of post-Herulian Athens followed their forebearsnot only in their
choice of residentialsites but, so far as possible, in the actual dwellings.130Of the existing remains
of private houses in the Agora area itself only a few, chiefly dating from the 6th century, were of
totally new construction.The bulk of the houses made serviceableafter 267 were in the densely
populateddistrictat the northwestfoot of the Areopagus.For centuriesthis area had been occupied
by unpretentioushouses and industrialestablishmentscrowdedclosely together. Much rebuilding
was undertakenhere after Sulla's destructionin 86 B.C.and again afterthe Herulian attack(but not
© American School of Classical Studies at Athens

until the 4th century). Many houses were put in shape again, their plans and general appearance,
where possible, changing little from those of their prototypes,but more often in a much curtailed
For personal use only. License: CC-BY-NC-ND.

shape (P1. 22:a).131

One house, on the lower slope of the west end of the Areopagus,may be taken as an example.
Built in the Hellenistic period, it originally coveredan area of 300 sq. m. and consistedof six rooms
and a peristyle with eight columns surroundingan apsidal fountain. It was destroyedin 267 but
was reconstructedin part in the 4th century.132Painted stuccodecorationwas foundon the walls of
several of the rooms, the best being in the large room at the north cornerof the house. Here alter-
nating panels of red and yellow surmounteda low blue dado and a black string course.The earth
floor overlay a pebble mosaic which probablybelongedto the original house. In the districtwere
three public baths and at least one latrine.133
In Himerius' farewell to Ampelius, Proconsulof Achaea in 359/60,134the sophist speaksof Am-
pelius' generosityin restoringa part of the city from squalor.He is precisein its location:"anarrow
street called Kollytos in the very heart of the city, taking its name from the deme and very well
known because it served as a market place."'35In spite of its precision,however, this statementis
tantalizing because the exact location of the deme Kollytos, even in classicaltimes, is uncertain.Is
Kollytosthe districtjust describedor is it farthersouth or north?It was certainlynot far from here,
and even in earlier times the boundariesof the demes were not always clear.136By the 4th century
after Christ the distinction might have become still looser, and a road "takingits name from the
deme"could have originatedin it but led fartherafield. Himerius' descriptionwould fit well one of
the streetsin the districtunder discussion,137 and it is not altogetherimpossiblethat it was the scene
129
Balty (op. cit., p. 490, note 53) rightly questionsthe blanketdesignationof these roomsas triclinia and prefersthe
more neutral term "oecus".
130 Because of the difficultyof differentiationin the date of private construction,houses of both the 4th and the 5th

centuriesareincludedin this section.Evidenceforspecificdateswill be givenwherepossible.

131 For this whole area, see Young, "IndustrialDistrict,"esp. pp. 272-288. The area was never excavatedentirely,
andit was impossibleto recovermorethana few completeplans.
132
Hesperia 9, 1940, pp. 272-273 (Shear, Sr.).
133
Above, pp. 30-33.
134
Above, pp. 20-21.
135 YTEVW'TO T v Ko,Xvros ovrw
rts ev To) AeoalTarTq TS 7roAEWC,8 ov
KaAovpEvos9 eIvyJAvo
ICwv , ayopas
8b Xpe?arTct Evosg (Ecl. xxxI.9 [Diibner, p. 35]; also apud Photius, Bibl., Cod. 243 [Henry, VI, p. 119]).
136
Cf. Strabo, 1.4.7, using Kollytos itself as an illustration of the difficulty. For various interpretationsof the am-
biguous literary referencescf. Young, "IndustrialDistrict,"pp. 141-142; W. K. Pritchett, "The Attic Stelai, Part I,"
Hesperia 22,1953, pp. 271,275-276; D. M. Lewis, "Noteson Attic Inscriptions,"BSA 50,1955, p. 16; W. E. Thomp-
son, "Notes on Attic Demes," Hesperia 39, 1970, pp. 64-67; J. Traill, Hesperia, Suppl. XIV, The Political Organi-
zation of Attica, Princeton 1975, p. 40.
137
Young, "IndustrialDistrict,"p. 136, fig. 1.
 36 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

of Ampelius' benefaction.The recordfor longevityis perhapsheld by a large, amorphoushouse on

the slope of the Areopagus,just west of House A (below, pp. 38-39; P1.22:b). Built in the 5th cen-
tury B.C., it was probablydamagedby the Herulians but not abandoneduntil the time of the con-
structionof the Post-Herulian Wall.'38The southernpart, at least, was rebuilt and occupiedin the
4th and 5th centuries.The plan of the house in its last phase is not at all clear, nor is its purposeat
that time. The one large roomof the original house survivesin its earliestform. The presenceon its
floor of one Doric and three Ionic capitals and one Ionic base might indicatea peristyle court, but
the sizes are so disparatethat it is hard to believethat they would satisfy even the relaxedstandards
of this period. Possiblythey came from differentroomsof the predecessor.The tenement-likechar-
© American School of Classical Studies at Athens

acter of the house is at variance with the comparativeelegance of its immediate neighbors.If, as
For personal use only. License: CC-BY-NC-ND.

seems possible,the latterwere residencesof teachers(below, pp. 44-47), the house underdiscussion
might have served as a rooming house for some of their pupils. A few small crude marble figures
were found on the site, e.g., a statuette of Asklepios, a head of Sarapis, and a statuette of Tyche
(P1. 23:a-c).139
More splendid lodgings were not lacking in the Agora area. A 4th-centuryB.C. residencebeside
the PanathenaicWay near the southeastcornerof the Agoraemergedafter many vicissitudeswith a
few rooms rebuilt, one paved with marble slabs and another with an elaboratemosaic floor with
arabesquesand geometricdesigns in red, blue, and pale green tesserae.140More impressivewas a
house in the southwest corner of the Agora area, salvaged from a pre-Herulian house. Its most
noteworthyfeaturewas a room, probablya triclinium,with a handsomemosaicfloor (P1.23:d).141
Perhaps the most imposingof all the privatehouses of which any trace survivesstoodabout half-
way up the north slope of the Areopagus, on an artificial terrace cut back into the hillside. The
southern part of the building lies under the Areopagus park, the excavationof which is not con-
templatedunder the Agora program,so that the plan as recoveredis far from complete.The princi-
pal featureof the house as exposedwas a squarecourtpavedwith marbleslabs and surroundedby a
flower bed ca. 0.50 m. wide. Around the garden ran a shallow basin of approximatelythe same
width, also pavedwith marbleand containedbetweentwo stuccocurbs.A fountainat the middleof
the southeast side of the court poured water into the basin and thence, through a lead pipe in the
middle of each side, into the middle of the garden (Pls. 24:a, 25:a). Both the garden and the sur-
rounding basin stopped short of the north cornerto provideaccess to a doorway on the northeast
and to form a niche at the northwest.Of the house itself only parts of the southwestand southeast
walls and the east cornerremainedvisible. Similar gardencourtscan be found in North Africaand
elsewhere (cf. one in a well-preservedhouse in Dougga, Tunisia, P1.24:b;also in Ostia).
It is uncertain whether the house is pre- or post-Herulian. The style of the court would be at
home in the 2nd or 3rd century,but there is nothing in the surroundingfill to indicatea date earlier
than the middle of the 4th century for its construction.In any case, it was certainly in use at that
time; a lamp of the 4th century (L 4034) and a coin of ConstantiusII were found on the floor.

138A coin of Aurelian (A.D.270-275) was found in the destructiondebris.

Asklepios (S 875), height 0.195 m.; Sarapis (S 885), height 0.06 m.; Tyche (S 871), height 0.265 m. For this last
139

see E. B. Harrison, The AthenianAgora,XI, Archaicand ArchaisticSculpture,Princeton 1967, no. 119.

Hesperia8, 1939,pp. 217-218 (Shear,Sr.).
140
141
Hesperia17, 1948,pp. 169-170 (Thompson).Althoughthe star-and-diamond patternis commonfromthe 3rd
centuryon,theblacklinebetweentwowhitelinesbordering date,as in
thestarsevidentlydenotesan early5th-century
the Churchof S. Sofiain Sofia(B. Filov,SofiskataTsirkva,Sofia1973,pp. 49 ff. andpls. XV-XVII. Forthe date,see
alsoR. F. Hoddinott,Bulgariain Antiquity,New York1975,pp.271-277.1 amgratefulto ErnstKitzingerforpoint-
ing out this feature and for the reference).
 THE BUILDINGS 37

A curious feature of the house is the use of three large blocksof Egyptiangranite in the masonry,
roughly squared but not finally worked. The presenceof so uncommona material high up on the
hillside can hardly be attributedto chance. It seems most likely that there was a sanctuaryof some
Egyptian divinity near by, perhaps never finished,for which the blockswere broughtup at consid-
erable effort.142This probabilityis reinforcedby the discoveryonly a few metersto the east of part
of a bronze disk with engraveddesigns, Egyptian in character(PI. 25:b).143
Constanterosion of the steep slope of the Areopagushas carriedaway some of the buildings that
must have stoodthere, leaving only faint clues to their identity.The discoveryof two piecesof sculp-
ture associatedwith Mithraism suggeststhat a Mithraeum once stoodin this area.144One is a head
of Helios of the 2nd century,evidentlysalvagedfrom the Herulian destructionby the occupantsof
© American School of Classical Studies at Athens

one of the later houses on the slopes.145The other is a head of Selene in relief, found in a well lower
For personal use only. License: CC-BY-NC-ND.

down the hillside (PI. 25:c).146These slight evidencesof foreigndivinitiespermitthe cautiousconjec-

ture that unorthodoxreligions might have gravitatedto this area as being more hospitablethan the
Agora, which in any case was overcrowdedby the time they made their appearance.
Houses on the Areopagus: Architecture
Apart from the strictly privatehouses is a group of buildingson the north slope of the Areopagus
which seem, by their size and relationshipto each other, to have serveda commonpurposebeyond
the ordinarydemandsof Athenian life and yet to be unsuitedto officialpublic use. There is little in
general plan to distinguishthem fromthe importanthouses of the majorcities of the Empire,e.g. in
North Africa, of the same period. In scale, detail, and workmanship,however, there can be no
comparison.Only when they are viewed in the context of the general run of houses in Athens can
their real importancebe appreciated.While the groundplan of Athenianhouses throughthe whole
of antiquity averages 130-150 sq. m., with a few of those of the Roman period in the outskirts
reaching335-420 sq. m., the Areopagushouses coveredfrom 1000 to 1350 sq. m. each.147From the
exceptional size and commandingposition of these buildings it seems obvious that their owners
were men of wealth and eminence.
In some at least of the lesser cities of the Empirethe wealth tendedto be concentratedin the hands
of a specificclass of society, dependingon the factorsresponsiblefor whateverimportancethey had
achieved.Ostia, for example, owed its prosperitychiefly to the grain trade with North Africa, and
its finest houses reflect strong North African influence. The goddess Fortuna Annonaria had a
conspicuousplace in its cults, and all the majorrenovationsand other public works were creditedto
variouspraefectiannonae.148 In Syria, the city of Apamea,becauseof its positionat the intersection
of importanttrade and military routes, enjoyeda long period of prosperityfollowing the reign of
Trajan. The grand houses that adornedthe city are attributedto a "localelite"who becamemore
and more active in the life and administrationof the provinceand thereforeincreasinglyrich.149
Athens had neither commercialnor political incentivesto prosperity.With the virtuallycomplete
cessationof the export of sarcophagi,the lamp industrywas almostthe only sourceof foreigntrade.
142
See Hesperia 17, 1948, pp. 162-163 (Thompson).
143B 904.
144
H. A. Thompson in an unpublishedmanuscript.
145
S 2356 (published as Alexander the Great): Hesperia 40, 1971, pp. 273-274, pl. 58:b (Shear, Jr.) and footnote
161 below, p. 41. For the building, see below, pp. 40-41.
146
S 857. T. L. Shear, Sr., "Excavationsin the Athenian Agora,"AJA 41, 1937, p. 184 and fig. 10.
147
As noted by Balty (footnote 128 above, p. 34), p. 486.
148 G.
Becatti, Case Ostiensedel tardoimpero, Rome 1948, pp. 44 ff.
149
Balty (footnote 128 above, p. 34), p. 490.
 38 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

Moreover, Athens had been receiving a yearly distributionof grain from Constantine until the
transferof the capital to Constantinople.This caused an acute grain shortageas that city monopo-
lized the sourcesformerlydrawn on by Athens.150
The contrastbetween Corinthand Athensofferedby the authorof the Expositiototiusmundi (ca.
A.D. 359) is revealing: Corinth very active in commerce,Athens noted for its centers of higher
learning and its history.151All the evidence points to an upper class composedalmost entirely of
intellectuals. The Neoplatonic School was still enjoying the fruits of the Platonic Academy, the
biaboX&Ka, whatevertheir actual sourceand amount,and the resourcesof the originalschooland its
successorswere apparentlycontinuallyreplenishedby other bequests.152
© American School of Classical Studies at Athens

The four buildings constitutingthe Areopagus group stood on the lower slopes of the hill, on
For personal use only. License: CC-BY-NC-ND.

terracesleveled for their predecessors(PI. 6). Their siting and plans were conditionedby the two
east-west streets that ran through the area and by the terrain itself. The northernmost,House A
(PI. 26:a), was contiguousto the South Road, which formsthe southernboundaryof the Agora,but
with a very slight differencein orientationso that its northwestcornerencroacheson the road by
about a meter. House B (PI. 26:b) is about 15 metersto the southeast,a little fartherup the hill; the
easternhalf was built against the retainingwall of the Upper South Road. House C (PI. 27:a) lies
still fartherup the slope, directlyacrossthe road from House B. The south edge of the road there-
fore determinedthe line of its northernwall while a scarpin the hard rockof the Areopaguslimited
furtherexpansionto the south. Of House D only the apse remains,ca. 35 meterswest of House C,
and 15 metersto the north,
ther start of what are apparentlythree roomswhich might have opened
on a peristyle,and as mosaicfloorwith a simple pattern(PI. 37:a). The interveningempty space
between the apse and the three rooms suggests that the courseof the Upper South Road may have
been divertedafter the destructionof the building and causedto pass throughthe ruins.
kind in the area. The presenthouses
It seems unlikely that these were the first buildingsoft their
all seem to be contemporary,probablybuilt in the last quarterof the 4th century.A numberof bits
of concretefoundationswere found scatteredaround in the vicinity, sometimesincorporatedin the
new buildings, necessitatingslight irregularitiesin plan. In other cases they may representprede-
cessorseither swept away by the rains washing down the hillside or else deliberatelyclearedaway
to make a clean sweep for the new construction.The earlier remains, although sparse, extended
over an area wide enough to lead to the belief that the predecessors,whatevertheir nature,occupied
almost as much groundas their successors.
Aside from the general similaritiesof plan, the one elementcommonto all is a large room,or hall,
enteredfrom a peristyle court by way of an antechamberand in most cases terminatingin an apse.
This was the focal point of the house. The whole unit (exceptfor the apse itself) was surroundedby
smaller rooms. In some cases these rooms led to an adjacentcourt, with or without peristyle, and
often there was an additionalcourton the opposite side of the building. Exigenciesof terrain led to
variationsin the arrangementsof the rooms,but the basic elementsremainedconstant.
House A in its present state is the smallest and simplest of the group, measuringca. 27 x 35 m.
(includingthe apse; PI. 26:a). A roughly square building, it consistsof the basic nucleus of apsidal
unit (includingantechamber),peristyle, and surroundingrooms.The west side has sufferedheavy
damage,with the loss of the entire outer wall. It is not improbablethat a secondarycourt occupied
the now devastatedarea on that side.
150
Day, p. 263.
151
Above,p. 22; also Day, pp. 262-263.
152
On this whole subject,with extensivereferences,see Glucker, Antiochus,chap. 5.
 THE BUILDINGS 39

Only in the apsidal room are the walls preservedabove groundlevel (PI. 28:a). In the apse itself
the wall was massive enough (0.90-1.00 m. thick) to support a half dome; the wall was further
strengthenedby the abruptly rising bedrockagainst which it was set. The wall was built of large
ancientblocksof brecciaset on end and spacedat intervalsof ca. 0.75 m. The rest of the wall was of
brick and stone masonryset in firm concretecharacteristicof the Late Roman period.Three niches
were set in the wall of the apse about a meter above the floor. Only the central one is preserved
almostto its full height;it is five-sidedand carefullyconstructedof brickslaid horizontally.The two
lateral niches, of which only the floor survives,were semicircularand approximatelythe same size
as the central niche.
© American School of Classical Studies at Athens

At the junction of the apse with the straightside walls the thicknessof the masonryis reducedto
0.65 m. Traces of stucco still adhereto the inner faces of both apse and side walls, and much disin-
For personal use only. License: CC-BY-NC-ND.

tegrated stucco was found in the destructionfill inside the building. The floor was paved with
square terracottatiles of which a few remain in place.
The apsidal roomwas enteredfromthe peristylevia an antechamberthrougha broaddoorwayin
the north end. Only part of the stylobate of the peristyle remains, but it can be restoredas a rec-
tangle of which the east side, however,is not quite symmetrical.It measuresroughly 8-9 x 9.50 m.,
with four columns on the north and south sides and five on the east and west. The depth of the
colonnadewas 2.50 m. on all sides except the south, where, as befittedthe entranceto the chief part
of the complex, it was widened to 3.00 m. Around the peristyle, includingthe northernpart of the
apsidal unit, ran a series of small rooms (PI. 28:b), ca. 4.50 m. deep on the east and west sides,
ca. 3.00 m. on the north. The roomsflankingthe apsidal unit were small, more like alcoves.At least
one of these is probablya shrine, as in the house probablyto be identifiedas Proclus'house.153The
southwest cornerof the complex was closed off by some earlier rooms,clumsily incorporatedin the
new plan. The largest of these had a sturdyfloor of pebble mosaic.
The foundationsfor all the partitionsin the east wing are preserved,showing a divisioninto four
rooms of unequal size. The west wing has lost not only the entire outer wall but also all the inner
partitions except one. In this is a carefully finished slit, 0.20 m. wide, which might have been to
affordpassage for a water pipe. A gap, presumablyarched,of 2.20 m. in the wall of the peristyleon
the same side preservedaccessto a very deep well, which was in use continuouslyfromthe middleof
the 1st centuryuntil the end of the 6th.154 Of the northouter wall only a small bit survives,but with
it, along the inner face, the remainsof a doublelatrine, convenientlylocatedin a roomto the right of
the large entrancevestibule. No evidence,whether in the form of stairs or columns, was found for
an upper storey, but its existence is probableon the analogy of House C.
House B is much larger (ca. 30 x 60 m.) and more complex than House A (Pls. 26:b, 29). The
line of its south side was determinedby the retaining wall of the Upper South Road, which served
also as the south wall of the eastern half of the complex. The steep slope of the hill at this point
necessitatedstringing out all the elements in a row from east to west, so that the peristyle court
adjoinedthe apsidal unit on its east side instead of the north, thus leaving space for an additional
antechamberor forecourtat the north. Access from the peristyle to the main unit was through a
narrow doorwayin the side of the smaller antechamber.On the opposite (east) side of the peristyle
was a smaller court (ca. 7 x 10.50 m.) surroundedby a colonnaderesting on a marble stylobate.At
about midpoint in the court was a well. Each of the courts was surroundedby a series of small
rooms. The main entranceto the house was through a central doorwayflankedby two large vesti-
bules, one of which, as in House A, was probablya latrine. The constructionis of mixed stone and
153
Below, pp. 42-44.
154 M 17:1; see below, pp. 42, 47.
 40 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

brick, and as in House A, three niches of brickwere sunk in the wall of the apse about a meterfrom
the floor.
Communicationbetween the two courtswas through one room only. This had doorwayson all
four sides and might have been regardedmerely as a passageway;otherwisethe smallercourt with
its surroundingrooms formedan independentunit. An oven (PI. 30:a), probablyfrom a later use,
was built into the cornerof a room in the main unit, and a circularhearthoccupiedthe centerof the
room at the southeasterncorner (PI. 30:b). The only independentcommunicationbetween the east
wing and the outside was througha wide passage near the southeastcornerof the building, down a
few steps and through a narrow doorwaywhich led to the street.
© American School of Classical Studies at Athens

The western part of the house has suffered the most from erosion, so that the plan cannot be
For personal use only. License: CC-BY-NC-ND.

determinedaccurately.Only the rooms on the south side can be made out with certainty.Two of
these are adjacentto the apsidal hall and are separatedfrom each other by an east-west crosswall
with a connectingdoorway. Diagonal channels were cut in the south and west walls of the south-
ernmost room, apparently to bring water from two wells or cisterns immediately outside. The
channel from the west wall continuedacrossthe cornerof the room and turned northwardto pass
through the crosswall. A third room, structurallyquite independentbut evidentlybelongingto the
complex, lies some five metersto the west, separatedby a steeply sloping rise in the bedrock,which
may have been used as a ramp for reachingthe Upper South Road.
The rest of the west wing evidentlyconsistedof a long rectangularcourt (ca. 16 x 4 m.) running
from east to west. It had no peristyle, but a foundation along its south side may have carried a
colonnadefor a porticofour metersdeep, frontingthe south rooms,or it may have been for a wall of
a narrow room or corridor.Three rooms probably opened off the court on the north side. The
emphasis on the water supply and the more informalnature in generalof the west wing lead to its
identificationas a servicearea, perhaps kitchenand pantries.
House C is the most complex of the Areopagusbuildings. Its overall dimensionsare almost the
same as those of House B (ca. 60 x 30 m.), and the plans of the two are basicallysimilar. The west
wing of C is more formal, however,with a third peristylecourt insteadof the plain servicecourtof
House B (Pls. 27:a, 31, 43:a, b).155
The most importantfeatureof House C is a nymphaeumleadingdown fromthe southeastcorner
of the central peristyle by two marble steps flanked by marble columns to a small triclinium (ca.
3 x 3.50 m.; PI. 32). On its east side this lookedinto an apsidalroomhousinga semicircularpool, the
transitionmarkedonly by a marblestrip in the floor.The two steps leadingdown into the pool were
coatedwith fine cement,the edge of each markedby a red stripe (PI. 33:a). The motivationfor this
constructionwas obviouslythe readyavailabilityof water froma fine fountainhouse into which the
water from a spring higher up the hill had been channeledsince the 2nd or 3rd century (PI. 36:a).
Accessto the fountainhousewas gainedby breakingthroughits west wall and spanningthe gap with
a wide brick arch of quality inferior to the excellent brickworkof the fountain house (PI. 33:b). A
well just south of the outsidewall of the nymphaeum,which had been in use sincethe fountainhouse
was built, was brought into direct relationshipwith the pool by an archedopening with a parapet
(PI. 34). The well ceasedto be used for domesticpurposesat the time the house was built (end of the
4th century)but was continuedin use as a supplementarysourceof water for the pool.

For a detaileddescriptionof the house as excavated(except the west wing), see the excavationreportsof 1970 and
155

1971 in Hesperia 40, 1971, pp. 266-270 and 42, 1973, pp. 156-164 (Shear,Jr.). See also below, pp. 88-89.
 THE BUILDINGS 41

Adjoiningthe tricliniumon the opposite side, for most of the remainingwidth of the peristyleand
extending back as far as the rock,was a room with a floor level 1.30 m. abovethat of the triclinium
but with no dividing wall (Pls. 32:a, 35), only a marble fence with sileni as fence posts (PI. 39:b;
S 2363). The angle between the triclinium and the larger room was filled by a nondescriptroom,
roughly square. The original purpose of the larger room is not clear, except that it was obviously
related to the nymphaeum.156Two water channels of uncertain purpose ran across the room
from south to north. The proportionsof the room are far too meagerfor an apsidalhall of the usual
type. The nymphaeum was evidently regardedas of prime importance,which may explain some
unusual features in the west wing: the introductionof a third peristyle court and the long room
© American School of Classical Studies at Athens

along the south side, against the Areopagusrock.The wall is massive (ca. 1.50 m. thick), with five
rectangularniches of varying sizes groupedsymmetricallyalong its length. As noted above in con-
For personal use only. License: CC-BY-NC-ND.

nection with House B, the limitationsof space imposedby the two east-west roads, the Areopagus
rock, and the contours of the hillside necessitatedmodificationsin what might be consideredthe
standardplans of houses of this kind. In House B the apsidal unit had to be placed alongside the
peristyle instead of in front of it. The formalityof the west wing of C, with its only slightly smaller
peristyle court, and the massive back wall of the large room, with niches suitable for statuaryas in
the apses of Houses A and B, make it seem quite likely that this room servedthe same purposesas
the apsidal rooms of A and B.157A suggestion that the room was a library, while attractive,is
untenable for, had the room been intendedas a library,one would have expectedprovisionagainst
groundwater such as an outside wall, leaving an interveningspace to protectthe booksfrom damp.
Also, the discrepanciesin the sizes of the niches, both in width and depth, are more suitable for a
group of statues or portraits,or both, than for books.
All three houses were rich in sculpture. Some, the better preservedpieces, were found in wells,
others in the destructiondebris over and around the houses. The most significantcollectioncame
from two wells in House C, especiallyone in the east wing.158This well was empty except for four
pieces of sculpture, all in an excellent state of preservation.Three were portraitbusts: Antoninus
Pius (PI. 40:a), a young woman (PI. 38:a), and an elderlywoman (PI. 38:b);there was also a small
statue of Herakles (PI. 38:c).159A statue of Hermes at the same scale was found intact many years
ago in a well in the adjacenthouse (B; PI. 38:d) which was abandonedat just about the same time,
along with a head of Nemesis and a statuetteof a seatedphilosopher(PI. 39:a, c).160In anotherwell
in House C (PI. 37:b), in the main courtyard,were thre e heads:one of Helios, one of Nike,
and a portraitof a beardedman (PI. 40:b-d).161
In the destructiondebris or in late Roman fill in immediatethevicinit following: a
draped headless bust of reddish marble PI.
(S 1026, 41:a) and a Doric capital, diam. 0.345 m. (both
in the doorway between Rooms 1 and 3); a small female bust (S 1044, PI. 41:b); a small statue of
Asklepios (S 1068, PI. 39:d);also many architecturalfragments(Pls. 42, 43:d).
156 In alater period the relationshipwas close;below, p. 89.
157
An apse cut back into the rock at the west end of House C is evidentlyfrom an earlier building.
158
Well Q 21:4. For the circumstancesand significanceof these depositssee below, p. 48.
159
S 2436, S 2437, S 2435, and S 2438, respectively.Hesperia 42, 1973, pp. 170-173, pls. 37, 38 (Shear,Jr.).
160 S
1054,S 1055,andS 1053,respectively; Well P 18:2.Hesperia8, 1939,pp. 214-215,236-241 (Shear,Sr.).
Well P 21:2.S 2356 (Helios),S 2354 (Nike),S 2355 (Beardedman);Hesperia40, 1971,pls. 58-59 (Shear,Jr.).
161

See p. 37 above.The identityof S 2356 as Helios(notAlexander,as published)is establishedby a rowof 15 drilled

holesforthe insertionof metalspikesto forma crown,theregularattributeof Helios(Guide3,pp. 292-293,326). For
the possibilitythatthe Nikeheadis copiedfromthe Nikeof theAthenaParthenos,see E. B. Harrison,"TwoPheidian
Heads: Nike and Amazon,"in The Eye of Greece,D. Kurtz and B. Sparkes,edd., Cambridge1982, pp. 53-88.
 42 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

Houseson the Areopagus:WaterSupply

For their water supply the houses of the Areopagus group dependedchiefly on the numerous
wells on the slopes, all, or almost all, of which had been sunk in much earlier times.162
House A. The main sourceof water for this house was a very deep well (M 17:1) in the line of the
west colonnadeof the peristyle.Since the contentsshowedthat it had been in almost continuoususe
fromthe 1st throughthe 6th century,it must be supposedthat the buildersrespectedit by spanning
its position with an arch. This probably provided drinking water, while utilitarian needs were
servedby a well a little fartherup the slopes, near the southwestcornerof House B (N 19:2). For its
re-use it was coveredwith a square, brick-vaultedchamberfrom which a vaultedpassageled off to
© American School of Classical Studies at Athens

the north, protectinga small rectangularwater channel set in a deep cut in the slope of the hill.163
For personal use only. License: CC-BY-NC-ND.

As the terrain leveled off, the passage stopped abruptly, and the cutting for the water channel
becameprogressivelyshalloweruntil it enteredthe building,barelybelow the level of the floor.The
point of entrancewas about the middleof the east wing, where the channel changeddirectionat an
obtuse angle for the purpose. It now crossedinto the east peristyleand zigzaggedinto and along the
north peristylewhere its coursewas lost in the upheavalssufferedby the north side of the building,
but it very likely providedwater for the latrine by the entranceto the building.
House B. House B too had an ample water supply. A clusterof wells in and aroundit are proba-
bly to be associatedwith it (0 18:1, 0 19:1, P 18:1, P 18:2, P 19:1). In addition,a water channel
was built to skirt the southeastcornerof the building. Waste water was carriedoff by a tile drain
leading north from the middleof the east side of the peristyle.
House C. An elaborateand still usable installationof the 3rd centuryaffordedHouse C an ample
water supply.164Springs fartherup the slope were tapped by a large rock-cutshaft from which the
water was led off through a channel protectedby a vaulted tunnel cut through bedrockand paved
with rectangulartiles. The channel continuedunder the floor of the springhouseand emptied into
the pool of the nymphaeum.A subsidiarysourceof water was the well just outsidethe south wall of
the nymphaeum(Q 21:5).
With the spring producingwater steadily, particularlyin the winter months, there must have
been dangerof flooding.To avertthis an ingenioussystemwas devised.A large overflowchannelat
the northwest corner of the pool carried the excess water to a drain under the floor of the east
peristyleof the main courtyard.At a point oppositeWell P 21:2, in the colonnadeitself, the channel
bendsslightly to the west. A little fartheron it picks up again to resumeits northerlycourse.Appar-
ently the overflowfrom the pool was collectedin Well P 21:2, presumablyfor purposesother than
drinking (washing, gardening?)to ensure maximum use of the water, but with a built-in arrange-
ment to preventfloodingwhen the water was overlyabundant.This perhapsexplains the existence
of two wells in the courtyard,both in use at the same time:Well P 21:3 in the colonnadeof the west
peristyle, with its fine marble wellhead, would provide clean drinking water, while Well P 21:2
would furnishwater for more general uses.165Two other channelscarriedgroundwater diagonally
acrossthe courtyardsouthwestto northeast;their furthercoursehas been destroyedby erosion.
"Houseof Proclus"
In an effort to determinethe exact role of the Areopagusgroup, an importantfactor is a large
building on the south side of the Acropolis,betweenthe Odeionof HerodesAtticusand the Theater
For the chronologyof the wells as evidencefor the historyof the houses, see below, pp. 47-48.
162

163
Hesperia 27, 1958, p. 147 (Thompson).
164
Above, p. 40, and Hesperia 40, 1971, p. 268 and 42, 1973, pp. 159-160 (Shear, Jr.).
165
I am indebtedto John Camp II, the excavator,for his (unpublished)descriptionand explanationof these arrange-
ments. See also below, p. 87.
 THE BUILDINGS 43

of Dionysos, which will serve as a point of referencefor the houses. It was discoveredin 1955 in the
course of shifting the line of the road running along the south side of the Acropolis.Only a part of
the house was excavated, and most of that was immediatelycoveredover; the rest still lies under
modernbuildings in the vicinity, especiallyto the south, but the key elementsof the plan are visible,
making possible a general reconstruction(Pls. 27:b, 44:a).166
This was no ordinaryhouse by Athenianstandards.A large roomopens into a wide apse (6.60 m.
wide, 4.40 m. deep); the lower part of the wall of the apse was surfacedwith marble revetment
slabs. Abovethe revetmentthe thicknessof the wall diminishes,and in it were seven niches suitable
for sculpture (as in the Areopagus houses).167The floors of both parts of the room were covered
© American School of Classical Studies at Athens

with mosaics in elaborategeometricpatterns (PI. 36:b), the apse being emphasizedby having the
For personal use only. License: CC-BY-NC-ND.

floor laid at a slightly higher level. Against the outer face of the east wall of the apse was a small
shrine of Cybele, identifiedby a statuetteof the goddessin a niche in the wall (PI. 44:b). A statue
base with a funeraryrelief carvedon the front servedas an offering table. Both pieces of sculpture
were re-used in these positions.
An important passage in the biography of Proclus, the third Head, or Diadoch, of the Neo-
platonic School at Athens, by his successorMarinus makes it seem probablethat this was the house
of Plutarch, the founder of the school, and of its successiveheads,168and that it served as head-
quartersfor the school for more than a century,until the school was closed in 529. In a rare bit of
topographicalpinpointing Marinus says that Proclus enjoyedhis house because his predecessors
Plutarch and Syrianus had lived in it and for its location "nearthe Asklepieion made famous by
Sophocles and the Temple and Theater of Dionysos" and that it "could be seen, or at least
discerned,by someonestandingon the Acropolisof Athena"(KaLyap 7rpobrols aAAosoevrvTx41ao-L
apiLoLWTraTrrl avrT KaLrL OlK)TLSOV7r7jpgEV, ?XV KaL 0 7T7arpavrov 1vpLavosKaCL0 rpo7arTp, W
avTos' EKaet, FIAovrapXos WKro-av, yElrova LoevovoavPTOV 7TOToOKAo aovS a 'AOKA-
TLELOV, KaL TOv TrpOS Tr() OEaTpO) ALovvOlov, OpWAVt l c a \w aaio-r7Tri
tX y'yVOp?EV7rV T'l
'AKpo7rT'oAeL7'AO7va^s).169
The house in question fits all the topographicalspecificationsin the VitaProcli,and furthermore,
its site, as far as it could be estimated from its scatteredknown parts, precludes the existence of
anything comparablein the area. Various objectsfound in or near the building support the view
that it was related to philosophyin some way:170a fragmentof an inscription,apparentlyof philo-
sophical content,with the words co4obt' and /oTov,'71 and part of a head of a philosophertype.'72
166I. Miliades, <<AvaorKaa'LvorlwSrjsv 'AKpo7roA?,co>, 1955, pp. 46-52,
HIpaKTLKa with pls. 3-7.
For sculpture in niches, cf. Ostia, Domus della Fortuna Annonaria, Becatti (footnote 148 above, p. 37), pp. 23,
167

44-46, fig. 50.

168
First suggestedby Miliades (footnote 166 above).
169
Marinus, Vita Procli 29 [Boissonade].See also L. J. Rosan, The Philosophyof Proclus, New York 1949 (with an
English translationof the Vita) and Proclus le philosophe,trans. A. E. Chaignet, I, Paris 1900. The phrase 7 a\xaws
alo-Or7TrTv yLyvoeLvr7vrj7 'A po7Tr6Xt rfjs 'AOqrvas demands some re-examination.ProfessorHarold Cherniss, who
kindly looked at the passage with me, suggested that the dative, unless it is simply bad grammar,is used to emphasize
the fact that the viewer is standingon the Acropolis. "Orat least discerned"limits the preceding"visible",rather than
offering a senseless alternative"otherwiseperceived"(Rosan's translation),and implies that someonestandingon the
Acropoliscould see it with some difficulty.Homer Thompson, who happenedto be in Athens at the time the problem
arose, respondedto a query whether the factsjustified this interpretationwith the following:"Lookingoverthe present
top of the south wall of the Acropolisone has no difficultyin seeing the supposedsite of the house;but in Late Antiquity
one would presumablyhave had to climb up to a sentry walk."
170Miliades
(footnote 166 above), p. 49.
171
Republishedby W. Peek, Attische Versinschriften,Berlin 1980, p. 36, no. 38.
172 This is
probably the head found in the Frankish wall (Rizokastro), in the section just south of the building
 44 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

The fine head of a philosopherof the early 5th centuryin the AcropolisMuseum (P1.44:c)173is said
to have come from the vicinity, but the report is unsubstantiated.It has been thought to represent
Plutarchhimself. Part of a statue of Isis, life size or over, of dark gray marble,was also found near
by, perhaps the cult statue from the Isis sanctuaryof the Hadrianic period on the terrace of the
Asklepieion,not far from the house. The statue had been damaged,probablyin 267, and recut to
form a bust, when it might have been used to adorn Proclus'house.174
Finally, in the westernmost room of the house, there was a carefully constructedand well-
preservedgrave of a year-oldpiglet, 0.70 m. long, with a numberof graveofferings:severalvases, a
lamp of the 5th centurywith a running Eros on the disk, and most importantof all, the sacrificial
© American School of Classical Studies at Athens

iron knife in the neck of the victim,175perhapsfrom some esotericritual, althoughthere is no record
that pigs and Neoplatonismhad anythingin common.176
For personal use only. License: CC-BY-NC-ND.

All the aboveare indicativeof a close connectionof the house with philosophyor pagan ritual, or
both. The shrine of Cybele, the Mother of the Gods, focusesthe identificationon Proclus himself.
Proclus had a special affinity with the Mother of the Gods, "towhom he prayed especially and in
whom he rejoicedexceedingly."'177 He also purified himself every month by the practicesof the
worship of the Mother of the Gods (Vita Procli, i 9). The shrine is in a key positionin the house and
is perhaps the very place where Proclusofferedhis prayers.
On presently available evidence it is impossible to date the constructionof the house precisely.
The presumptionis that it was after Alaric'sinvasionin 396, but that is not certain.178It is improb-
able, however,that Plutarch would have built such a grandiosebuilding immediatelyafter found-
ing the Neoplatonic School. Also to be consideredis the probabilitythat the house was built by
Plutarch on a much smaller scale and then greatly enlarged by Proclus and, perhaps, by his suc-
cessors,as the importanceand prestigeof the Academygrew. In fact, at least two of the roomsin the
excavated portion are obviously additions, and most of the preservedmosaic floors could hardly
have been laid beforethe middle of the 5th century.179A large bath, of which parts were discovered
during building operationssouth of "Proclus'house",was believedby its excavatorto be connected
with it.180

Houses on the Areopagus:Identification

Although far less monumentaland less elaboratein their appointments,the houses on the Areo-
pagus bear a close affinityto the "Houseof Proclus"in the emphasis on the apsidal room with the
(Serpentze), and published by G. Rodenwaldt, GriechischePortrdtsaus dem Ausgang der Antike (Programmzum
Winckelmannsfeste76), 1919, p. 4, no. 2, pl. III.
173No. 1313; cf. G. Dontas, "Kopfeines Neuplatonikers,"AM 69/70, 1954-1955, pp. 147-152.
174See S.
Walker, "A Sanctuaryof Isis on the South Slope of the Acropolis,"BSA 74, 1979, pp. 243-257.
175 Dontas
(footnote 173 above).
176 owe the pig lore to David
Pingree.
177 Vita
Procli, 33, where Marinus expatiateson the subject.
178 Miliades
(footnote 166 above,p. 43), p. 48.
179The study and chronologyof floor mosaicshave been greatly advancedby new evidencefrom Stobi. I am grateful
to Ruth Kolarik for sharing with me her conclusionsbased on that materialwhich will be the subjectof a forthcoming
study. Also to Ernst Kitzinger, who concurs in the dating and in the hypothesis that Proclus' house was originally
Plutarch's but subsequently greatly enlarged. Accordingto present views, the mosaics of "Proclus'house" show a
slightly later developmentthan those of the tetraconchin the Library of Hadrian (below, pp. 72-73) and therefore
should be dated sometimein the third quarterof the 5th century.Cf. now P. Asimakopoulou-Atzaka,tZvrayMa r&v
T4lboTwrv
FIaXaLoxpporraavLKw^& rfis EXXaabosII. HIeAoro^vv71os-repea 'EAAaia, Thessalonica 1987, pp. 121-
123. This book appearedwhile the presentvolume was in press.
180G. Dontas, <<'vacrKa47 oLKO7rebov'AyyEAo7rovovo, AeAr 17, 1961-1962, A', pp. 88-89. See also Travlos,
PDA, s.v. Balaneia (Bath A).
 THE BUILDINGS 45

niches for sculpture and also in their size in relation to the average Athenian house. Detailed
comparisonof the plans as a whole is impossible,with so much of Proclus' house still lying under
modern buildings, but the scatteredremains indicate a general similarity in the dispositionof the
rooms. A counterpartto the shrine of Cybele may perhaps be seen in a small room adjacentto the
formal room in the west wing of House C, which has a shallow niche on the west side. The niche
was carefully revettedwith marble slabs and, though small, was clearly a feature of some impor-
tance. An alternativeexplanation,that this was a small fountain,is excludedby the absenceof pipes
or any provisionfor them.
Unlike "Proclus'house",in which the effect of opulence is conveyedby the mosaicfloorsin every
© American School of Classical Studies at Athens

room of the house yet explored and by the marble facing of the apse, little remainsin the houses on
For personal use only. License: CC-BY-NC-ND.

the Areopagus to give any comparableimpression. House C alone, with its well-matched set of
peristyle columns, its carefully constructedpool, and adjacentchamberwith mosaic floor and re-
vetted walls, approaches"Proclus'house"in magnificenceor size.181No trace of mosaic floorswas
found in any other of the Areopagus group except for a floor with a simple pattern of interlacing
circles associatedwith the exiguous remains of House D (PI. 37:a). Nor, except for the marble-
revettedwalls of the triclinium in House C, were there any other signs of such elaboratetreatment.
The apse of House A has a plaster bed treated for a stuccoed surface, and this may have been
standard for the whole group. Time and the sloping terrain have dealt harshly with all the floor
surfacesexcept in the pool complex, where the greaterdepth affordedmore protection,and it is not
inconceivablethat some mosaics have disappeared,although one would expect some tracesto have
remained.
Given the general similarity in plan and appointmentsbetween the Areopagushouses and "Pro-
clus' house",and the fact, noted above,that the upper class in Athens seems to have been composed
chiefly of intellectuals,it is not unreasonableto carrythe argumentone step furtherby associating
the Areopagus houses with the teachers who amassed considerablewealth from the fees they
chargedtheir studentsand who are known to have taught in their own homes.
The basic literary testimonyis Eunapius' statementthat the sophists (of the 4th century)made a
practiceof teachingin their own homesbecausethe dissensionbetween studentsand townsmenwas
such that "not one of the sophists dared to go down into the city and discoursein public, so they
lecturedto their studentsin their own privatetheaters"(EvTOLt8LOTLKOlS Oeaarpols).182 In the same
passage Eunapius describesthe house of the sophist Julian as "smalland shabby"but "breathing
the spirit of Hermes and the Muses and differingin no way from a holy temple."In it were statues
of associateswhom he had admiredand a "theaterof polishedmarble,imitatingthe public theaters
but smaller, and of a size suitable to a house."'183
The apsidal room, which had come into common
use as the major element in houses in many parts of the Empire in the 4th century, might have
commendeditself to the sophists as a lecture room and occasionedEunapius' referenceto Julian's
"privatetheater".
Studentswere sometimestaken in as lodgers.184This practicewas no innovation;Epicurus may
have followed it, although there is some dispute whether his house and "The Garden" were

181
For a more detailed descriptionof House C, see Hesperia 42, 1973, pp. 156-164 (Shear,Jr.).
182 Eunapius, VS, 483.
183
Cf. also the house of the father of Herodes Atticus with busts of famous orators in the corridors(Philostratus,
VS I.21 [521]).
Cf. the crisis precipitatedby the sea captain who deliveredan unexpectedlylarge contingentof potentialpupils to
184

Prohaeresiusin the middle of the night. Prohaeresiushad to call on his relativesto take care of the overflow(Eunapius,
VS, 485-486).
 46 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

adjacent or at some distance from each other.185At the end of the 2nd century the flamboyant
Proclus of Naucratis had "a library in his own house which was open to his students and sup-
plementedthe teachingin his lectures."'186
In addition to the similarity of plan to that of "Proclus'house",the niches for sculpture in the
apses and in the main room of House C and the sculptureitself, found in all the houses, reinforce
the hypothesis of their identity as houses of private teachers.Two statues, in particularthe com-
panion pieces representingHerakles and Hermes (above, p. 41; PI. 38:c, d), are vivid reminders
that Herakles and Hermes were the patron divinities of educationand also recall the sophist Ju-
lian's affinitywith Hermes.'87The size (abouttwo-thirdslife) would fit them well for a positionin
the peristyle of a grandiosehouse. Although all the sculpturefound in the houses would have been
© American School of Classical Studies at Athens

at home in the villa of any affluentcitizen, there is a markedacademicor religious flavorhere that
For personal use only. License: CC-BY-NC-ND.

would be suitable for this environment.188

From the size of the houses and yet the compactnessof the area into which they are squeezedone
would hesitate to attribute this group to wealthy Athenians, perhaps of senatorial rank or with
priestlyconnections,or to high officialshaving a secondresidenceaway fromthe capital. For these a
more likely district would be the outskirtsof the city, especially south of the Acropolis (probably
also the northwest,althoughlittle is known of it). There a numberof large villas with mosaicfloors,
baths, and, above all, space around them have been found.189 (The sprawling remains of a large
villa in the National Gardens,with mosaicfloors,gives an impressionof what might be expectedfor
this class of citizens.) The Areopagushouses, except for House C (above, pp. 40-41), which per-
haps had a prior phase as a private residencebefore being incorporatedin the group, are distin-
guished more for size than for opulence.
The close interrelationshipof the Areopagushouses seems to indicate some sort of coordinated
plan and thereforea commonpurpose. Other buildings of the same kind may, and probablydid,
exist between the known buildings and the PanathenaicWay, and also to the west. But the terrain
has been subject to constant erosion with increasing force as the descendingtorrents gouged out
gullies in the hillside, and for the most part even the foundationsof whateverstructuresstoodthere
have been washed away.
These indications of interdependencemay have some relation to the fact that the schools of
Athens, even those that were nominally private, were subject to governmentcontrol at both the
municipaland the imperiallevel.190The law requiredthat at Athens therehad to be "manyto teach
and many to listen"('Set yap 7roXAoVefvaL, Kara TOv vo ov TOv LPoauovK, 'AO ?-Ln',TOVS pv
AEyovras, rovs be aKoVovrTa).191 And a decreeof the EmperorJulian requiredthat "if any man
shall wish to teach ... he shall be approvedby the judgment of the municipal senate."192 On the
highest level, when competitionfor the chair of the sophistJulian was intense and acrimonious,the
proconsulhimself steppedin. 193With the authoritiesmaintainingsuch supervision,it would not be
185
For differentviews on its location,see below, pp. 65-66.
186 Philostratus, VS 11.2I (604).
Many examples of dedicationsto both divinitiesin associationwith gymnasia are given by M. Luni ("Storiadell'
187

istituzione gymnasialee dell' attivitaatletica in Cirenaica,"Quadernidi archeologiadella Libia 8, 1976, pp. 223-284).
Cf. also J. Delorme, Gymnasion,Paris 1960, pp. 337-340.
188 For the fate of the
sculpture,see below, pp. 87-88.
189
Below, p. 49 and Asimakopoulou-Atzaka(footnote 179 above,p. 44), chap. II, 'A^ijva.
190On the complicatedquestion of the organizationof the schools,see Marrou (footnote15 above,p. 17), chap. VIII.
191Eunapius, VS, 487.
192
CTh xIII.3.5, A.D. 362 (trans. Pharr).
193
Eunapius, VS, 488-489.
 THE BUILDINGS 47

surprisingif they also took a practicalinterestin a coordinatedplan for the buildingsthemselves.It

has been suggestedthat the scatteredbits of concretefoundationsfound throughoutthe area of the
Areopagus houses were the remains of pre-Herulian predecessors,and House A shows definite
signs of a pre-Herulian phase. This might explain a window tile from a pre-Herulian context
(PI. 41:d) whose sophisticatedform was probablythe model for the many ineptly executed copies
which have been found in the vicinity (PI. 41:e, g, h).194
The later history of the houses will be discussedin Chapter IV (below, pp. 87-90).

Houses on the Areopagus:Chronology

© American School of Classical Studies at Athens

All five of the buildings were probablyerectedabout the same time, i.e., in the secondhalf of the
For personal use only. License: CC-BY-NC-ND.

4th century. With the exception of House C all probablywent out of use permanentlyjust before
the middle of the 6th century.
House A. The evidencefor actual constructionconsistedof a handfulof sherdsfrom undisturbedfill
into which the foundationsof the apse were set down, all dating from the second half of the 4th
century.In addition,a lime-slakingpit three meterswest of the apse, probablyused in the construc-
tion of the building, had at the bottom a firm layer of earth containing some pottery, also of the
secondhalf of the 4th century,including a lamp (PI. 41 :f) and part of a terracottafigurine.195
For the life span of the building as well as its construction,evidencewas foundin a stratifiedwell
in the west colonnadeof the peristyle.196The well had been in use from the 1st century and con-
tinued into the 10th, but with some interruptions.One of these breaksoccurredin the secondhalf of
the 4th century, quite probablymarkingthe time of the constructionof the house; the next can be
dated ca. 540-550, followed by a thick layer of silt and, finally, pottery of the 9th or 10th century
thrown in to close the well.
House B. A parallel situationwas observedin a well in the northwestcornerof House B,197where a
break representedby a layer of stones separateda pre-Herulian layer from one beginning in the
late 4th century and continuinginto the 6th, and after anotherbreak, into the 7th. The other wells
of House B reflect the same general picture as follows (all these dates were provided by John
Hayes):
0 18:1. In the west (service)courtyard.Continuousfill from the end of the 4th or beginnfiing of
the 5th centuryuntil the early 6th.
P 18:1. Near the northeast corner of the peristyle. Cleaned out in the second half of the 5th
century and used until sometimein the 6th.
P 19:1. Near the apse. Continuous use from the 5th to the early 6th century, then sealed by
abandonmentfill.

194A similar structure found southeast of the Acropolis has been proposed as related to the Areopagus group
(. Travlos, "The LawcourtElII HAAAAAIMI," Hesperia 43,1974, p. 509). The identificationis on the groundsthat
it is near the ruins believed to be those of the Palladium, which is included among the sites of schools mentionedin a
passage of Plutarch (de exilio, 14). Roughly contemporarywith the Areopagusgroup, it might possiblybe thoughtof as
a continuation of the earlier tradition. An intermediate link might be found in the unfinished apsidal room near
"Proclus'house",whose constructionseems to have been halted by the Herulian attack (Miliades [footnote166 above,
p. 43], pp. 46-47 and pl. 4:a; idem, <<'Avao-Kaa'L1. 'AOijvaL.a) NOTL'co T719 To "Epyov 1955, p. 7).
'AKpo7roA(0eo>,
5323 and T 1361. The figurinewas later found to join with anotherpiece, not fromthe pit but froma similar fill
195 L
around the walls, resulting in reidentificationas a lamp, L 5324 (PI. 41:c).
196 M 17:1;
Agora V, p. 125: Group M, but with reviseddating by John Hayes, who inspectedthe pottery.
197 P 18:2. All the wells of House B are includedin the
deposit list in AgoraV, p. 126.
 48 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

Q 19:1. East (service)courtyard.A breakca. 540 was followed by an accumulationof nearly a

meter of snail shells, and on top a small amount of potteryrunning down to the end of the 6th
century.
Q 17:4.At the north edge of the Areopagusgroup. In use from the 1st centuryto the 6th, and
again in the 8th century.
House C. This building providedthe most abundanttestimony,both for constructionand for later
history. A well at the south side of the apsidal room with the pool,198originally used in connection
with the adjacentspringhouseof pre-Herulian date, containedpotteryfrom the first half of the 2nd
century down to the second half of the 4th, when it ceased to serve domesticuses and was made
© American School of Classical Studies at Athens

accessible to the newly constructedpool room as an auxiliary water supply. A date, also in the
For personal use only. License: CC-BY-NC-ND.

secondhalf of the 4th century,for the constructionof the whole house was given by pottery under
the floors of two roomsoff the east side of the peristyle,as well as from the fill dumpedbehind the
exterior south wall at the time of construction.199
The break noted in the continuityof the life of Houses A and B is shown more dramaticallyin
House C becauseit was of a short,clearlymarkedduration,illustratedespeciallyby the re-layingof
the floors in three rooms, also in the peristyleand in the servicecourt, overlyinga layer of pottery,
the latest dating from the early 6th century, and also by the pottery from a drain channel leading
from the pool whose course was divertedat the same time.200The wells associatedwith House C
were apparently all, with the exception of that in the pool room, cleaned out at the time that the
house was built. They are, however,instructivefor the latest phases of the building.The following
wells were in use:
P 21:3. In the west colonnadeof the peristyle. Shows abandonmentat the time of the destruc-
tion of the building, with architecturalmembers(PI. 73:c), etc., topped with loose dirt, stones,
and tiles. Finally a 7th-centuryre-use, with a limestonecollar built up over the marble well-
head to bring the mouth of the well up to the post-destructionlevel.
P 21:2. In the east colonnadeof the peristyle. The same sort of fill as in the precedingwell
(P 21:3).
Q 21.:4.In Room 7, in the east wing of the building. Empty except for four pieces of sculpture,
almost intact, in an excellent state of preservation(Pls. 38, 40). The well was sealed with a
marble slab flush with the floor and partially overlaidby a bath in the next period. This well,
the most significant for the history of the house, will be discussedin detail in a later chapter
(below, p. 87).
ATHENS IN A.D. 396
A brief summaryof the appearanceof Athensjust beforethe Visigothic invasionof A.D. 396, as
far as it can be determinedat present,may be in order.
Enteringthe city at the Dipylon, the visitorwould pass betweena continuousseries of colonnades
flankingthe PanathenaicWay, presumablyall containingshops, until the streetenteredthe Agora.
The PanathenaicProcessionwas still following this route on its way acrossthe Agora and on up to
the Acropolis. The citadel showed some of the ravages of time, but its main buildings appeared,
superficiallyat least, much as they had two centuriesbefore.
Leaving the PanathenaicWay at the entranceto the Agora the visitormight take the road along
the west side of the square, dominatedby the Hephaisteion on the hill, KolonosAgoraios,above.
198Q21:5.
199See
Hesperia 42, 1973, pp. 159-160 (Shear,Jr.).
200Ibid., p. 161.
 ALARIC IN GREECE (A.D. 396) 49

Along the streetat the left, the Stoa Poikile, on the right the Stoa Basileios,the Stoa of Zeus, and the
Temple of Apollo Patroos were still standing;the Tholos, badly damaged by the Heruli, had its
walls reinforcedby a concretering, and the Metroon, almosttotally ruined,had now been rebuiltin
a reducedform, probablyas a synagogue.
The oppositeside of the Agora was completelytransformedby the incorporationof what was left
of the Stoa of Attalos and the Libraryof Pantainosinto the Post-Herulian Wall. The south side and
the entire centerof the Agora were empty, with the ruins of the classicalbuildingsprobablycovered
by vegetation.The evidencefor the north side is still incomplete(1988), but it seems probablethat
little except the Stoa Poikile survived.
© American School of Classical Studies at Athens

East of the Agora the rest of the main part of the city was entirely enclosed within the Post-
For personal use only. License: CC-BY-NC-ND.

Herulian Wall. Inside the enclosure the major structuresstill standing were the south half of the
Roman Market with its two gateways,the Doric on the west, the Ionic on the east, and immediately
to the east, the large marble latrine and the Tower of the Winds.201The north flank of the wall,
where it met the southwest corner of the Library of Hadrian, evidently took a series of turns to
incorporatethe Library itself as part of the circuit, the south wall probablyhaving been destroyed
in 267. Of the interior, only the rooms at the east end and the southeast corner of the peristyle
remained.
On the south slopes of the Acropolisonly the Theater of Dionysos, which, of the buildingson the
slope, had suffered the least, had been restored and remodeledand resumed its function as the
official assembly place.
Of the outskirtslittle can be said except that a numberof villas and baths were scatteredaround,
particularly on the south side of the Acropolis. Some of these had survivedthe Herulian attack,
others had been damagedbut were susceptibleof repair,and some were entirelynew. How many of
these were in use in the 4th century and how many only in the 5th is uncertain.Since they are all
chance finds, no solid conclusionscan be drawn about their numbersor distribution.202

ALARIC IN GREECE (A.D.396)

THE THREAT

That some preoccupationwith dangerwas felt in Greece as a whole even as early as the reign of
Constantineand remainedall throughthe 4th centuryis attestedby a numberof inscriptionsdeal-
ing with the repair of the walls, usually in the form of dedicationshonoring the official (or indi-
vidual) responsible. Two inscriptions are of especial interest, each on the base of a statue of a
proconsul (of Achaea) "Phosphorius".In the first, from Megara, the Megarians are recordedas
having made and erected a marble statue of Phosphoriusin recognitionof his justice and because,
by fortifying the cities, he delivered the inhabitants from the cruel enemy.203The second in-
scription, found near the agora at Argos, is a simple epigram to the effect that the statue of the
201 Travlos, PDA,
p. 31, fig. 39.
202
Remains of many of these were discoveredin the second half of the 19th century, briefly noted at the time in
'E#'ApX (passim), and either covered up or destroyedand never published. Some are representedonly by mosaic
floors, a few of which were salvaged and are now published, with illustrations and references,in Asimakopoulou-
Atzaka (footnote 179 above, p. 44), chap. II (e.g., a fine one in the courtyardof the Byzantine Museum, found in
Euripides Street, no. 63).
203 IG VII, 96.

[0]]wor4foplov Meyapies appw-Tovoolo KaJ.ovrES

eLKova Xacver7v 0rO70-av e7T' Ev&Ks[at]s
ovveKa 7rvpycooas 7roAias KparepaAyea OoSpo[v]
TreVv araptS7rov b'ov Evva&r[aL],.
 50 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

renownedproconsulPhosphoriuswas set up by Archeleosas a gift to the Argives,by decreeof the

Council.204
The identityof Phosphoriusand the date of his proconsulshiphave been mattersof dispute.That
he was to be identified with Aurelius Valerius Tullianus Symmachus (PLRE, Symmachus6),
consul in 330 and probably Proconsul of Achaea in 319, was first proposedby T. Reinach and
supported by A. von Premerstein.205The identificationwas based on the extreme rarity of the
signum Phosphorius, which was borne by Symmachus' son L. Aurelius Avianius Symmachus
(PLRE, Symmachus2). It is not known for the father, but since there is no room for a procon-
sulship of Achaeain the long and well-documentedcareerof the son, it is conjecturedthat he inher-
© American School of Classical Studies at Athens

ited the signum from his father according to common practice. The only serious challenge to
Reinach's proposal has been from E. Groag, who rejectedboth the identificationand the date and
For personal use only. License: CC-BY-NC-ND.

proposeda date of ca. 379-382, with the otherwiseundesignated"Phosphorius"remaininganony-

mous.206Groag'smain objectionwas that there was no barbarianthreat in Greeceduringthe reign
of Constantineand that the epigrammust refer to the disturbancesafter the Battle of Adrianoplein
378. He thereforeequated Archeleos,the donor of the statue in Argos, with Archeleos,the priest
who first celebratedthe ceremonyof the tauroboliumin Athens.207
More recently G. Polara has made a convincingcase for revertingto Reinach's theory.208He
observedthat the Megarian epigram could equally well refer to the general fears of barbarian
invasionthat had been prevalentin Greecesince 267, e.g. a threat fromthe Sarmatiansin 322. Part
of Groag's case was based on a stylisticcomparisonof the two altars of the tauroboliumin Athens,
the one of Archeleos,the other dated386, which he found "verysimilar"and thereforevery close in
date, hence the date of 379-383, to bring Adrianopleinto the argument.But althoughthe two altars
resemble each other closely in content and arrangement,it is quite clear that the altar of 386 is a
somewhat crude copy of that of Archeleos209 and could have been made at any time that the earlier
altar was visible. If a suggestion made above is valid, that the tauroboliumwas introducedinto
Athens in the reign of Julian,210the differencein date would be more than twenty years.
More useful is a considerationof Archeleoshimself. The name, with the spelling 'ApXe'XeAo,
instead of the more common 'ApxA'Xaos, occurs on both the Argive statue base and the Athenian
altar, probably also in a third inscription,found at Lerna and now lost.211All three have Argive
connections.The priest of the taurobolium"livedin Argos"(vaLEraleL ev "Apyet), where he held
priestlyofficesconnectedwith the cults of Demeter and Hera, also at Lerna. Perhapshe cameup to
Athens specificallyfor the inaugurationof the taurobolium.The Archeleosof Lerna also had close
connectionswith the cults. Both men show pride in their lineage:the celebratorof the taurobolium
204IG IV, 1608.
EiKova coofoptovAeyaKv8fos avOv7raroLo
'APxf'Ae?WAavaot oTrorE xapt4o'^pvoS.
*(7rL(ip,aTr),B(ovAfis).
Reinach,"Unnouveauproconsuld'Achaie,"BCH 24, 1900,pp. 324-328; Premerstein(footnote67 above,
205 T.

p. 10), pp. 71-80.

206 Groag,Achaia,pp. 54-55.
207
Above,p. 19.
208 G.
Polara,"IInonnodi Simmaco," La paroladelpassato29, 1974,pp. 261-266, with a detailedanalysisof the
and
epigraphical documentary evidence,includingthedateof Symmachus'proconsulship.I owethisreferenceto T. D.
Barnes.
209 Cf. Svoronos (footnote32 above,p. 19),pl. LXXX forthetwo.
210Above,p. 19.
211 IG IV, 666;Anth.Pal. III (Cougny),p. 45, no. 279. The missingthirdsyllablehasbeenrestoredwithan alpha
('Apxe[Xa]ov)becauseof the final syllable -ov, but there seems little doubt that this is a memberof the same family.
 ALARIC IN GREECE (A.D. 396) 51

"rivaledhis ancestors with his own deeds";the Archeleos of Lerna had "the same name as his
father"(oywvMvov 7arpi). The two might, as Svoronosbelieved,212be one and the same, but the
Lerna priest might rather be consideredas the father of the tauroboliuminnovatorand the epony-
mous son of the donor of the statue of Phosphorius.The third Archelaos (and chronologicallythe
first of the three) might have been one of the causes of the conspicuouspride, and, if he can be
consideredthe grandfatherof the celebrantof the taurobolium,his careerwould fall conveniently
around 320, the only time at which any evidencehas been producedfor a Proconsulof Achaeawith
the signum Phosphorius.
The dangerwas intensifiedafter the Battle of Adrianoplein 378, when bandsof victoriousGoths
© American School of Classical Studies at Athens

roved freely through the Balkan peninsula and threatened the cities of Greece, many of whose
fortifications were already weakened by the disastrous earthquake of 375.213 The threat reached its
For personal use only. License: CC-BY-NC-ND.

culmination in the 390's when an army of Visigoths under the leadershipof Alaric emergedfrom
beyond the Danube and invadedThrace. Immediatelyrelated to that danger, it seems, is a pair of
epigrams on a re-used herm shaft found in the Agora, honor m ings"adorned
Athens with Wisdom"and for building "thecity walls and its towers"(PI. 45:a).214
lamblichus, a friend and correspondentof Libanius, made two shorttrips to Athens, one in 357,
the otheher
in 358. He returnedin 362/3 and very probablyremained hisd until
there th, at least as
late as 391, and it is to this periodthat the epigrammust refer.215Since the large materialgift rather
than the contemplationof philosophy is likely to have been the motivatingforce behind the dedi-
cation and the monument was set up only after his death (ucera7roTrov), it is probable that the
erectionof the wall and its towers took place not long after this, that is, when the Visigothic danger
was becomingserious.
The fears were well founded,for by 395 Alaric was on his way towardthe invasionof mainland
Greece. Pas singeasily by way of Thermopylae, where his passage was apparently facilitatedby
bribery of the proconsul Antiochus,216he arrived in Attica in 396 and promptly captured the
Piraeus in order to cut off supplies for Athens. Accordingto Zosimus, Alaric thought it would be
easy to take Athens because its great size and small populationwould make it impossiblefor those
inside the walls to defendit.217This statementcould hardly be applied to the crampedspace within
the Post-Herulian Wall, but it would be intelligibleif it referredto the old city wall as reconstituted
under Valerian and repairedby lamblichus (and later to be restoredby Justinian).218
Up to the point of Alaric's arrivalbeforethe gates of Athens and after his departurefrom Attica
there is substantial agreement among the chroniclers.What happened in between is less certain.
The only detailed account is Zosimus' picturesque description. Accordingto this familiar tale,
when Alaric approachedwith his army he saw Athena in full armor"asshe appearsin her statues"
(cf. Frontispiece),walking aroundthe rampartswhile Achilles stoodbeforethe walls. The sight so
terrifiedhim that he abandonedall efforts to besiege the city. His offer of peace was accepted,and
he entered Athens where he bathed, dined in state, and received gifts, after which he departed,
Svoronos(footnote32 above, p. 19), p. 482.
212

Groag, Achaia, p. 55; P. Lemerle, "Invasionset migrationsdans les Balkans (depuis la fin de l'epoque romaine
213

jusqu'au VIIIe siecle),"Revue historique211, 1954, pp. 265-308.

214 I
3542: A. E. Raubitschek,"lamblichosat Athens,"Hesperia 33, 1964, pp. 63-68. For the correctidentificationof
this lamblichus as the namesake rather than the grandsonof the great Syrian philosopher,see Alan Cameron (foot-
note 42 above, p. 20). For a revised reading, ibid. and J. and L. Robert, "Bulletin epigraphique,"REG 78, 1965,
pp. 104-106,no. 155.
215 Raubitschek,op. cit., pp. 65-68. lamblichus is mentionedin letters of Libanius written in 390 and 391.
216
Zosimus, v.5-6. Eunapius (VS, 476) attributeshis successto treasonon the part of Christianmonks.
217 Loc. cit.
218 Below, p. 58.
 52 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

leaving both Athens and all Attica unharmed.219But three contemporarywriters on the event220
took a more pessimisticview: Claudian,referringto Stilicho'srecall while he was preparingto take
the field against Alaric, speaks of the lost opportunityand the consequentdevastationin the Pelo-
ponnese and enslavementof Athens:
Si tunc his animis acies collatafuisset,
prodita non tantas vidisset Graecia caedes,
oppida semoto Pelopeia Marte vigerent,
starent Arcadiae, starent Lacedaemonis arces,
non mare fumasset geminum flagrante Corintho
© American School of Classical Studies at Athens

necfera Cecropiae traxissent vincula matres.221

For personal use only. License: CC-BY-NC-ND.

And St. Jerome: Quid putas nunc animi habere Corinthios, Athenienses, Lacedaemonios, Arcadas,
cunctamque Graeciam, quibus imperant barbari? et certe paucas urbes nominavi, in quibus olim
fuere regna non modica.222And finally, Philostorgius: 'AAAaptXos... . crAOEv
T rT E4abt, KaL
ras 'Afvas KaLMaKbozvas KaLrov
etAEXV Aa?Aladras cEi
7rTpoeXe^Ls aro.223
Archaeologicalevidencehas been graduallybuilding up to reconcilethe two extreme views, i.e.,
that Alaric included Athens among all the rest of the cities of Greece in his conquest,on the one
hand, and that he inflictedno damagewhateveron Athens or even in Attica, on the other. For some
time it has been clear that the area of the Dipylon sustained heavy damage during the reign of
Arcadius,presumablyfrom Alaric's army.224It now seems certain that he did succeedin entering
the city with his soldiers,at least as far as the Agora,where he destroyeda numberof buildingsthat
had escapedthe Heruli, and struckterrorinto the minds of whateverinhabitantshad not been able
to take refuge within the Post-Herulian Wall, beforegoing into the city to strike a bargainfor the
ransomof the city.225
Zosimus,v.6.I-3. To the pagan Zosimus,Athena and Achilles were an invincibleteam who, 20 years earlier,had
219

deliveredAthens and Attica from an earthquakethat had causedheavydamageto all the rest of Greece (IV.i8.2-4). For
pagan and Christian attitudes on divine influence (especially in relation to the Battle of the Frigidus), see J. Straub,
Reallex.f. Antikeund ChristentumVI, cols. 869-875, s.v. Eugenius.
220 Zosimus is generallyconcededto have written his historybetween498 and 520 (Paschoud,Zosime I, p. xvi; R. T.
Ridley, BZ 65, 1972, p. 280, with earlier references;A. Chastagnol, "Zosime II, 38 et l'Histoire Auguste,"Bonner
Historia AugustaColloquium,1964/5, p. 75, specifies"secondhalf of the reign of Anastasius".
221 In Rufinum II.I86-191.
222
Epist. LX.I6. It is worth noting that this letter was written in 396, the year of Alaric'sattack(for the date, see J. N.
D. Kelly, Jerome, London 1975, p. 215). Destruction at this time over a wide area is attested by archaeological
evidence:at Argos (M. Pierart and J.-P. Thalmann, "Travauxde l'Ecole Fran~aiseen 1974,"BCH 99, 1975, p. 705);
Corinth (Broneer [footnote 13 above, p. 2], p. 159; T. Gregory, "The Late Roman Wall at Corinth,"Hesperia 48,
1979, p. 269; C. K. Williams, II and O. H. Zervos,"Corinth,1981: East of the Theater,"Hesperia 51, 1982, p. 118 and
idem, "Corinth,1982: East of the Theater," Hesperia 52, 1983, pp. 23-24); and Isthmia (0. Broneer,Isthmia, III,
TerracottaLamps, Princeton 1977, p. 3, note 9).
223 12.2.
J. B. Bury (History of the Later Roman Empire, Dover ed., 1958, I, p. 119, note 2), relying on Zosimus'
statementthat Athens escapedcapture,takes Philostorgius'accountto refer to Alaric'scaptureof the Piraeus, but the
destructionof the walls of the sanctuaryat Eleusis at that time (G. Mylonas, Eleusis and the Eleusinian Mysteries,
Princeton 1961, p. 186) seems to confute Zosimus'claim that "all of Attica"was excludedfrom Alaric'sdepredations.
For an analysis of Zosimus' accountof Alaric's invasion of Greece and his dependenceon Eunapius' fragmentary
UniversalHistory, see Paschoud'scommentaryon these chapters(Zosime III, notes 8-13, pp. 91-102).
224AgoraVII, pp. 53, 63-64.
225 For the
archaeologicalevidence,see below, pp. 53-54. Athena and Achilles (or their human agents) need not be
utterly banished. Not to put too fine a point on it, the walls of the Acropoliswould have been a much more suitable
place for them to appearthan the outer walls, and for this the Agorawould providean admirablevantagepoint. A clas-
sically minded defender of the city might have recalleda similar ruse, when the friendsof Peisistratos,in orderto bring
the tyrant back from exile, dresseda young woman as Athena in full armorand put her in a chariotto drivearoundthe
 ALARIC IN GREECE (A.D. 396) 53

Scholarshipmust have suffereda severe,if temporary,setbackat the time of the invasion. Hila-
rius, who was staying near Corinth, was captured and beheaded, along with his slaves. Priscus
died, presumably of natural causes, at the age of ninety or more. lamblichus, who contributed
toward the fortifications,was honored "afterdeath",probablybetween 391 and 396, "and there
were many who in their grief threw away their lives, while others were slaughteredby the bar-
barians." Among these was Proterius, from the island of Cephalonia, "as to whose worth and
probity there is good evidence."226
It was against this backgroundthat Synesius of Cyrene paid his first and only visit to Athens,
probablybetween 395 and 399,227 and recordedhis famous impressionof disillusionmentin letters
© American School of Classical Studies at Athens

to his brother,228cursing the captain of the ship that broughthim to Athens, a city which had now
only its famous name. Formerly the center of wisdom, now desertedby philosophy, no more than
For personal use only. License: CC-BY-NC-ND.

the empty hide of an animal, reveredonly by bee-keepers,and with only a pair of Plutarcheanswho
drew young men to the theaters, not by the fame of their eloquencebut by the jars of honey from
Hymettus. Although his disenchantmentwas doubtless heightened by comparisonwith Alexan-
dria, where he had just finished his studies with the glamorousHypatia, he must have seen Athens
at its lowest ebb since the Herulian invasion.

DESTRUCTIONIN THE AGORA

The comprehensiveview of the Agora gained by 40 years of excavationhas necessitatedsome
revisionin assessing the the
damagecaused by he
Heruli.d Heruli. In the beginning,when
the the magnitudeof the
disasterof 267 became evident,the destructionof any building of Late Antiquity for which no pos-
itive evidence could be found was attributedto the Heruli by association.Now it has become in-
creasinglyclear that a few buildings escapedwith only minor damageand remainedstandinguntil
the end of the 4th century, and cumulativeevidencenow shows that a lesser, but still serious, de-
structionmay be associatedwith the arrivalof Alaric and the Visigoths in 396. Since the conclusion
runs counterto previouslypublishedopinions the evidencemust be presentedin some detail.
Of the buildings
the along
a the west side of the Agora only the Metroon and the Bouleuterioncan be
demonstratedto have been victims of the Heruli. The Tholos was damaged,but it was rebuilt and
surviveduntil the end of the 4th century.229The evidencebecomesmuch clearer in the case of the
Stoa of Zeus and the Temple of Apollo Patroos,where analysisof the coinsand the lamps leadsto the
inescapableconclusionthat neitherof these two buildingscouldhave been destroyedin 267, but that
neithercouldhave outlastedthe 4th century.In contrastto the situationprevailingaroundstructures
certainlydestroyedin the Herulian attack,the fill over and aroundthe Stoa of Zeus and the Temple
of Apollo Patrooswas almost totally lackingin coins and other objectsof the 3rd century.Instead,a
fill ca. 0.50 m. thick containingmany coins and lamps of the secondhalf of the 4th centurycollected
over the "classical"floor of the Agora. Embeddedin the surface of this layer were a few pieces of
brokenarchitecturalmembersof the Stoa of Zeus, their surfacesfresh, their colorsbright. More of
these appearedin profusionin the layer above,along with many chips fromtheir breakingup. The

city, biddingthe Athenians to welcome Peisistratos,whom Athena "wasbringingbackto her own citadel"(Herodotus,
i.60). Examples of such divine interventionwere not uncommon.Cf. also the miracleof Zeus Panamaros(P. Roussel,
"Le miracle de Zeus Panamaros,"BCH 55, 1931, pp. 70-116, with related incidents).
226
Eunapius, VS, 482 (trans. Wright).
227
For the intricaciesof this dating, see C. Lacombrade,Synesiosde Cyrene,hellene et chretien,Paris 1951, chap. VI,
esp. pp. 75 ff.; also PLRE II, Synesius 1, and J. Bregman, Synesiusof Cyrene:Philosopher-Bishop,Berkeley 1982,
p. 41.
228
Epist., 54 and 135.
229 For all three
buildings, see above, pp. 4, 25.
 54 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

top of the lower layer thereforerepresentsthe groundlevel at the time of the destruction.Although
the upper fill still containedmuch materialof the 4th century,the lamps found in it were predom-
inantly of the first half of the 5th century.The softnessof the earth in both layers, the absenceof a
hard impenetrablesurfaceseparatingthem, and the many disturbancesfrom Byzantinetimes per-
mittedmuch filteringdown of coins and small objects,so that over most of the area the stratification
providesonly a general indicationof its chronology.The coins from a numberof uncontaminated
spots, however, included 19 that were legible and relevant (i.e., later than the 2nd century after
Christ). Of these, all exceptthree were datednot laterthan 395. Of the otherthree,the latestwas one
fromthe period402-408, at the very top of the layer, in freshcondition.
© American School of Classical Studies at Athens

There is probablya direct relation between the breakingup of marble aroundthe Stoa of Zeus
and the Temple of Apollo Patroos and the inception of the major building programof the early
For personal use only. License: CC-BY-NC-ND.

years of the 5th century discussedin the next chapter. Useful building material derivedfrom the
structuresruined by the Heruli would have been exhausted long since. Now a new supply had
becomeavailable,and that at least some of it went into the new SquareBuilding in the middleof the
Agora (below, p. 66) is confirmedby the discoveryin its foundationsof a number of pieces of the
marble geison and a fragment of a triglyph from the Stoa of Zeus. Larger blocks from the same
buildings may have been used in the superstructureof the Square Building, which has long since
disappeared. Several lime-slaking pits were found midway between the Stoa of Zeus and the
Square Building, evidentlyfor use in the constructionof the latter. After the building was finished
the lime pits were left open as refuse pits, first, perhaps,for the kitchensof the new buildings,since
their bottomswere coveredwith a thick layer of bones, later as general rubbishbins.
The speed with which destructionwas followedby rebuildingis furtheremphasizedby the posi-
tion of the statue of Nike, one of the akroteriafromthe Stoa of Zeus,230apparentlylying where she
fell in front of the south wing and incorporatedin the top of the wall formingthe northwestcorner
of one of the lime-slaking pits used for the constructionof the Square Building. The Nike, lying
exposed among other debrisof the late 4th century,was evidentlytaken as a convenientlimit by the
diggers of the pit. Many fragmentsof her and her companionswere found in the same late 4th-
centuryfill.231
The above evidenceis indicativeof a date at the end of the 4th centuryfor the destructionof the
temple and the stoa. That the actual cause was the invasionis made almostcertainby the discovery,
only a short distanceaway, of a small pit dug through a well-definedfloor in the north side of the
stoa on the Panathenaic Way.232In the pit was the head of a herm,233in good condition and
obviously carefully laid there, and with it eight coins, the latest being one of Arcadius,struck be-
tween 383 and 395.234The pit showed signs of heavy burning (Pl. 45:b). The layer throughwhich
it was dug also containedcoins of the same period, with the same cut-off point of 395.
The later history of the Stoa Basileios was in all probabilitysimilar to that of the Temple of
Apollo and the Stoa of Zeus. As with the Stoa of Zeus, the lowest depositoverits floorwas of the 5th
century, with no sign of Herulian destruction,235and fragments of its terracottaacroteriawere
found in plunderedfoundationtrenchesin the Annex of the Stoa of Zeus and in the excavationof
230 S 312.
231
Hespenria4, 1935, p. 376 (Shear, Sr.); AgoraXIV, p. 99, note 91.
232
Hesperia 42, 1973,
pp. 370-382 (Shear,
Jr.), and above,p. 28.
233 S
2499. Ibid., p. 406, pl. 76:b.
234 The others were Valentinian
II, period 383-393 (2); Valentinian II, year 383 (1); Gratian or colleague, period
378-383 (1); Valens, period 366-375 (1); 4th century,unidentified(1); 2nd century (1).
235AgoraXIV, p. 210.
 ALARIC IN GREECE (A.D. 396) 55

the Stoa itself. Also the close proximity of the two buildings makes it likely that they shared the
same fate.236
The fact that Alaric may now be safely held responsiblefor a considerableamountof damageand
destructionhitherto attributedto the Heruli does much to reconcilethe apparentlyconflictingevi-
dence for the history of the Stoa Poikile, the foundationsof which have now been discoverednear
the northwest cornerof the Agora. Both archaeologicaland literary evidenceargue for its survival
through the 4th century. It contributedno recognizablematerial to the parts of the Post-Herulian
Wall so far explored. On the other hand, architecturalmembers, some of which very probably
belonged to the Poikile, including moldings with painted decorationas well as pieces from other
© American School of Classical Studies at Athens

buildings, were found in profusionin a long wall of the secondhalf of the 5th century after Christ
which borderedthe east side of the PanathenaicWay.237The existenceof these fragmentsin the one
For personal use only. License: CC-BY-NC-ND.

short stretch of wall, as well as the freshness of the color, is a strong indicationthat the building
from which they were drawn stoodnear by and was still standingnot long beforethe constructionof
the wall.
The survivalof the Stoa Poikile into the 4th centuryis corroboratedby the testimonyof Himerius
and Synesius of Cyrene. Although Athenian rhetoricof the period demandedfrequent allusions to
the institutions and sometimesthe buildings of the illustriouspast, with some occasionalconfusion
between ancient symbols and present conditions,the accountsof both men contain specificdetails
implying that in each case symbolism is reinforcedby actual observation.Himerius, addressinga
group of Ionian visitors to the city (Or. x [LIX]), promises to show them the Stoa Poikile, with its
painting of the Battle of Marathon, before taking them up to the Acropolis.Whether he was ac-
tually proposing a conductedtour or merely a verbal descriptionis open to doubt, but the circum-
stantial detail with which he describesthe painting of the battle, and a similarlyvivid accountof the
battle itself in an earlier oration (II [VI].20-2I), at least suggest that the paintings were still visible.
But even if the descriptionby Himerius be rejectedas rhetoricallicense,it is harderto dismissin the
same way the more matter-of-factstatementof Synesiusthat studentsreturningto Alexandriafrom
Athens put on airs becausethey had seen the Academyand the Lyceumand the Stoa Poikile, adding
that the stoa was no longer decoratedbecausethe proconsulhad taken away the painted panels.238
Synesius visited Athens between 395 and 399, or shortly after the death of Himerius.239The
paintings may then be assumedto have been visible during most of Himerius' residencein Athens
(from ca. 330) and to have been removedbefore,but probablynot long beforeSynesius'visit. There
is insufficientevidenceto say what promptedthe removalof the paintingsby the proconsul,whether
it was damageinflictedon the building (by Alaric?),offeringa pretextfor the removalof the panels,
or simple avarice on the proconsul'spart. From the tone of Synesius' remarksit seems not to have
been the general decayof the building. The natureof the paintings (the Battle of Marathonwas the
most important subject treated) probably rules out Christian zeal in eradicating paganism, al-
though the possibility should not be eliminated entirely. A prototypefor the proconsul might be
seen in Verres, the notoriousgovernorimmortalizedby Cicerofor his crimesof bribery,corruption,
and plundering. By coincidence,a man of similar reputation,Antiochusthe son of Musonius, held
the office of Proconsulof Achaea in 395/6, although there is no referenceto his carryingoff works
236
Thompson,"WestSide,"p. 37.
237
For the Long Late Roman Wall, see below, p. 81. For the identificationof the building on this evidence,Hesperia
19, 1950, pp. 327-329 (Thompson) and L. S. Meritt, "The Stoa Poikile,"Hesperia 39,1970, pp. 233-264. For the dis-
coveryandanalysisof the building,Hesperia53, 1984,pp. 5-12 (Shear,Jr.).
238 Epist., 54 and 135.
239 For a recent study of Synesius' career,with earlier bibliography,see Bregman(footnote227 above, p. 53).
 56 III. FROM DIOCLETIAN TO THE VISIGOTHIC INVASION

of art.240It is hard to believethat Synesiuswould have referredto the removalof the paintingsif the
building had not been standingwhen he wrote, and it is equally hard to escape the impressionthat
he was speaking of a recent episode. It seems a reasonablehypothesis thereforethat the Poikile
remained standing at least until 396, when it was damagedbut not destroyedby Alaric, and that
shortlyafter that (still in 396?) the paintingswere removedby the proconsul,whetherAntiochosor
another. It might have remainedin a dilapidatedstate, but protectedas a venerablerelic, until the
middle of the following century,241when it was completelydestroyedor dismantledand much of it
built into the wall of the secondhalf of the 5th centuryon the east side of the PanathenaicWay.242It
is not impossiblethat its end was due to a Vandal attackin 467 or 476.243
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

This is the same Antiochusmentionedby Zosimusas havingconnivedat Alaric'sinvasionof Greece (v.5.2-5). For
240

the identity of Musonius, possibly a professorof rhetoricat Athens, cf. PLRE I, Musonius 1, and Paschoud,Zosime
III, p. 88.
241 It was still standing when a Late Roman stoa (5th century) was built against its west wall (Hesperia 53, 1984,

pp. 15-17, fig. 10 [Shear,Jr.]).

242 Below, p. 81.
243 Below, p. 78.
 IV
PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)
PROSPERITY (A.D.396-450)
THE TIME OF THE NEOPLATONISTS
We have seen that one of the most importantelementsin the life of 4th-centuryAthens had been
the schools of philosophy and rhetoric. The Neoplatonic School of Athens, establishedprobably
© American School of Classical Studies at Athens

toward the end of the 4th century,came into its own only in the early 5th as Plutarch, its founder,
For personal use only. License: CC-BY-NC-ND.

gained in statureand influence.From then on it dominatedthe scene,with an unbrokensequenceof

Diadochs, or Heads, from Plutarchto Damascius,who was in officein 529 when Justinian forbade
the teaching of philosophy in Athens.1
Recent studies have shown that the term "Academy"is misleading, implying as it does a direct
connectionwith Plato's Academy,but the term will be retainedhere for convenienceas denotingthe
school foundedby Plutarch. There is no evidencefor a "GoldenChain"of Diadochs from Plato to
Plutarch or for direct inheritanceof the material assets of the Academy.2Some revisionshave been
made in the accepted sequence of Diadochs: Plutarch, Syrianus, Domninus, Proclus, Marinus,
Isidore, Hegias, and Damascius. Domninus must now be excluded, although he may have had a
"secondchair",and probablyHegias as well.3
Although the philosophers gave little attention to material concernsin their philosophicaltrea-
tises, some color and substanceis added by the biographiesof two of the Diadochs. Marinus' Vita
Procli is completethough brief and concernsitself more with Proclus'characterand outlookon life
than historical fact or chronologicalsequence. The biographyof Isidore by Damascius is so frag-
mentary that the lacunae usually prevent full comprehensionof the stated facts;4but for all their
deficienciesthe two shed welcome light on the Academyand its place in Athens.
Between Proclus' birth in 4125 and his death in 485 there is little chronologicalevidencebeyond
the fact that he came to Athens late in 430 or early in 431. He was immediatelytaken to see Plu-
tarch, who, on accountof his advancedage, had alreadygiven overthe regularteachingof the school
to Syrianus,his designatedsuccessor.Proclusshowed such exceptionalpromise,however,that Plu-
tarch took him as a special studentuntil his death in 432. Syrianus,who died ca. 437, was succeeded
immediatelyby Proclus,who would then have been abouttwenty-fiveyears old. For the first part at
least of his tenure as Diadoch Proclus took an active interest in the affairs of Athens, appearingat
public meetings, "discussingquestions of justice with the officials"and "giving his opinion on a
variety of matters."6Suddenly, "a storm broke"and he went to Lydia for a year.7Presumablythe
cause was friction with the Christians, although this is not specified. Proclus studiously avoided
1 For the sequel to this action, see below, pp. 84-92.
2
Below, with references,pp. 86-87.
3Saffrey and Westerink, pp. xxxiii-xxxiv; Glucker,Antiochus,
p. 307, note 23.
4R. Asmus, Das Leben des Philosophen Isidoros, Leipzig 1911; C. Zintzen, Damascii Vita Isidori Reliquiae,
Hildesheim 1967, with Greek text and related fragments.
5The accepteddate. For this and other probabledates in his career,see Saffreyand Westerink,introduction.
6
Marinus, Vita Procli, I5.
7 This was
probablysome time after 450 (Saffreyand Westerink,p. xxii), but there is no way of knowinghow much
later.
 58 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

giving provocationto the Christians,but a number of veiled allusions have been collectedby Saf-
frey.8 Relations between the two gradually deteriorated,reaching their lowest point with the re-
moval of the statue of Athena from the Parthenon,when "thatstatue of hers that had stood for so
long in the Parthenon was taken away by those people who move things which should not be
moved."The story is well known but bears repetition. "For it seemed to the philosopherthat a
beautiful woman appeared to him in a dream and commandedhim to prepare his house quickly
'becausethe Lady of Athens wishes to come to live with you'."9

AFTERALARIC
REVIVAL
© American School of Classical Studies at Athens

Stateof the Fortifications

For personal use only. License: CC-BY-NC-ND.

What was the actual state of the outer fortificationsof Athens at the time of Alaric's attack and
how much damage was inflicted by him can only be conjectured.As noted above,10the old outer
circuit (the Wall of Valerian) had undergonesubstantialrepairstowardthe end of the 4th century,
and the fact that after 400 some importantbuildingswere being erectedoutsidethe Post-Herulian
Wall permits the assumption that the outer circuit was once more regardedas secure. The only
certaintyis that the Wall of Valerian was largelyrebuiltin the 6th centuryby Justinian,T implying
that by that time it was in a weakenedstate. The modestscale of destructionfromAlaric'sforcesby
comparisonwith that causedby the Heruli might argue a smaller invadingforcewhich could enter
the city merely by storminga few gates or, at worst, breachingthe wall in a few places. Superficial
repairs after that might have been enough to satisfy the need for securityin the absenceof an im-
minent threat, but they could not provideenough stability to withstand a Slavic attack when that
dangerbecamereal a centuryand a quarterlater.

Rehabilitationof Public Buildings

Alaric'sdeparturehad removedthe immediatethreatto Athens. It had also left depredationsboth
inside and outside the city, which it was necessaryto repair along with the still untidy mess caused
by the Heruli. But this time the damage was less and recoverymore rapid. The first efforts at
rehabilitationwere only stop-gap measures undertakenby a city that was impoverishedand still
endangered;but they preparedthe way for a resurgenceon a relatively monumentalscale which
reflects,if not a return to real prosperity,at least a sense of securitygreaterthan had been known
for two hundredyears. 2
Of the buildings with their roots in the past the most importantto be rebuilt was the Metroon.
We have seen above (p. 25) that after the Herulian invasion it was put back into use only in the
chez Proclus:le diadoqueplatonicien,"
D. Saffrey,"Allusionsantichretiennes
8 H. Revuedes sciencesphiloso-
59, 1975,pp. 553-563.
phiqueset theologiques
9Marinus,VitaProcli,30 (Rosan'stranslation).Whatrelationthe statueboreto Pheidias'workis a mootpoint
and irrelevanthere.The originalmayhavebeenreplacedseveraltimes.At any rate,somestatuestoodin the temple
throughoutLateAntiquity,andas an objectof venerationit stillretainedits oldpower.Its ultimatefateis notknown,
For a summaryof the knownfacts,see Frantz,"Julian,"p. 401,
but it was probablynot takento Constantinople.
note54.
10P. 51.
11Procopius,de aed. IV.2.24. For evidenceon the ground, see, inter alia, Thompson and Scranton(footnote4 above,
p. 1), p. 376.
12 In the it mustbe remembered
followingdescription thatwhiletheresultsfromtheexcavationof theAgoramaybe
regardedas reasonablydefinitivefor that area,for the restof Athenswe are dependenton chancediscoveriesmade
underlesscarefullycontrolledconditions.Nevertheless, thesesporadicremnantssupplementandcorroborate the con-
clusionsdrawnfromthe Agoraandwill be interwovenwiththemwithoutfurtherexplanation.
 PROSPERITY (A.D.396-450) 59

shabbiestpossible way. Now it was transformedbeyondrecognition.13Only the two northernmost

rooms are sufficientlywell preservedto offer any clue to their appearanceand purposeat that time,
but there are some indications that the two other rooms were in use also, and perhaps the front
porch, with or without its columns. By prolongingwestward the north and south colonnadesof the
north room, it was transformedinto a basilica-like structure.The three small rooms which orig-
inally opened off the west side of the peristylewere not rebuilt, but the area of the middleroomwas
now occupiedby an apse lined with marblebenchesof the Hellenistic periodtaken from elsewhere
(PI. 6). The adjacentroom to the south, the "dining room",was made more formal by laying a
simple mosaic over the previous floor and the broiling pits. The mosaic consistedof a centralpanel
of intersectingcircles with lozenges at their centers,borderedfirst by a simple running ivy tendril
© American School of Classical Studies at Athens

and finally by a series of plain circlesconnectedby straightlines (PI. 45:c). An unflooredstrip along
For personal use only. License: CC-BY-NC-ND.

the south wall may have been occupied by a bench. The lowered level of the dining room was
retained by the builders, and accessto it was gained by a broad stairway leading down into its east
end. Numismatic evidence gave a clear indication of date for the transition from the first, rude,
rehabilitationof the Metroon to the final form. In a thin layer of earth above the burning in the
larger pit were 14 coins, the latest being four of Theodosius I, one struckbetween 383 and 392, all
in fresh condition.14The mosaicfloor lay directlyover the thin layer of earth. Consideringthe fresh
conditionof the latest coins underneath,very little time could have elapsed before the mosaic was
laid, giving a date ca. 400 for the reconstruction.
The resemblanceof the basilica-like plan of the reconstructedMetroon to that of a Christian
church is obvious,15but the westward orientationis an obstacle.A more attractivetheory was put
forward by H. A. Thompson. On the basis of a fragment of marble revetmentwith a menorah
carvedon its surface,16he suggestedthat here was a synagoguetucked into an ancient monument,
much like a more imposing one built into the Gymnasiumat Sardis.17
Jewish communitiesexisted in Greece from late Hellenistic times, especially in trading centers
such as Corinth, Sparta, Thessalonica, and a number of the islands. The evidence for Athens is
slight, but there was presumably a Jewish communitythere in the time of the Apostle Paul. His
stay in the city was brief, however, and the communitywas obviouslyless activethan elsewhere in
Greece (Thessalonica, Beroea, Corinth). A few lamps with Jewish symbolshave been found in the
Agora, the earliest being from the 4th century (Agora VII, nos. 353, 1065, 2385, 2440). Some
epitaphs with apparentlyJewish names have appearedin various parts of the city.18
The Stoa of Zeus was never rebuilt, although it may possibly have served some temporaryuse,
like the Tholos but in even less substantial form. The large numbers of lamps of the 4th century
found lying over the floor are indicativeof some sort of habitation,but the presenceof recognizable
parts of the building in the foundationsof the large Late Roman Square Building directlyeast of
the Stoa precludeits existence after the 4th century.
13
Nothing can be addedto the observationsmade by H. A. Thompson when he excavatedthe Metroon in 1936 and
published his results in 1937 ("WestSide,"pp. 195-202). His descriptionis summarizedhere for convenience.
14
The otherswere ConstantiusII or colleagues,period351-361 (3); Valens and Valentinian II, period367-378 (5);
Arcadius,uncertaintype (1); illegible (1).
15
Cf. D. Pallas (Oco,oyL'a31, 1960, p. 348), who offers a lawcourtas an alternative.
16 A
4546, Godsand Heroes (Agora Picture Book No. 19), Princeton 1980, p. 30, ill. 59.
17
The suggestion was published by A. T. Kraabel, "The Diaspora Synagogue,"Aufstiegund Niedergang der ro-
mischen Welt 11.19.1,Berlin/New York 1979, pp. 477-510 (pp. 483-488 for Sardis,pp. 505-507 for Athens). Further
explorationis neededto check the hypothesis.
18 L. Urdahl, "Jews in Attica,"SymbOsi43, 1968, pp. 39-56. The number has been greatly reducedby
J. and L.
Robert, "Bulletinepigraphique,"REG 82, 1969, pp. 453-454. For more on the Jewish communityin Late Antiquity,
see Frantz, Reallex.f. Antike und Christentum,s.v. Athen II (in press).
 60 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

New Construction
It was perhaps at this time (early 5th century) that a large round building was erected on the
PanathenaicWay just beforeit enters the Agora. Only about one quarterof the circumferencewas
preserved,most of the rest lying in the cut for the Athens-Piraeus Railway (Pls. 6, 46:a).19But
drawings made at the time the railway was laid recorded another segment of the circle. The
northern part of the foundation was heavily buttressed at close intervals to take account of the
rapidly falling ground level and perhaps to support a domed roof. A single coin of the late 4th or
early 5th centuryin the foundationsprovidesthe only clue to the date of construction.The life of the
building was in any case short because it was destroyed,certainlybeforethe end of the 6th century
and more likely toward the end of the 5th, perhapsby the Vandals.20
© American School of Classical Studies at Athens

Identificationof the building is complicatedby the presenceof two interiorsupportswhose spac-

For personal use only. License: CC-BY-NC-ND.

ing has no relationto any normalinteriorarrangements.The piers are representedonly by massive

foundationscomposedof miscellaneousmaterial,includingseveralfragmentsof a statue of the 2nd
or 3rd century:a horse protome,a hand with an orb, and a few piecesof drapery(PI. 46:b, c).21The
asymmetricalspacing might be attributedto a repair necessitatedby the collapse, real or threat-
ened, of a portion of the domedroof, but this would have been a clumsy deviceat best. The scanti-
ness of the remainsleaves the purposeof the "RoundBuilding"enigmatic.That its functionwas an
importantone is made clear by its siting in a conspicuouspositionat the very entranceto the Agora,
at the intersectionof the PanathenaicWay and the road along the west side.
The public characterof the Agora in the 4th and 5th centuriesis emphasizedby a group of four
monumentbases, all dating from this time, aroundthe northeastcornerof the Metro6n.22Three of
these are on a north-south line between the Metroon and the southeast corner of the Temple of
Apollo Patroos;the fourth,with the same orientation,lies immediatelyto the northof the northeast
cornerof the Metroon. Around the bases and stretchingaway to the east was a hard-packedroad
surface containing many coins of the late 4th and early 5th centuries. There is no evidence as to
what statues were carriedon these bases, but it may be notedthat the statuesof Apollo Patroosand
Hadrian and also a head of Antoninus Pius were all found in the immediatevicinity.23
Another base, for a statue of very considerablesize, lies just west of the PanathenaicWay at the
point where it shifts its courseto a more southerlydirectionafter enteringthe Agora (PI. 46:d). The
overall dimensionsof the base at the lowest finished course (of which only one block is preserved)
are about 4.10-4.20 m. east-west and 4.15-4.25 m. north-south. The base is set at an angle to the
street, perhapsfor greatervisibility from both north and south as the streetbends.The foundations
are sunk in a fill of the 3rd and 4th centuries.It seems likely, therefore,that the statue was erected
sometimein the 4th century and continuedto stand throughmost of the 5th but, to judge from the
absenceof a 6th-centurylayer, not longer. It, the Square Building east of the Stoa of Zeus, and the
Round Building might all have been casualtiesof a Vandal invasion in 467.24The size of the base
and the prominence of its position make it probable that it was an emperor who stood on top,
perhapson a column.25
19
Hesperia 42, 1973, p. 126 (Shear, Jr.).
20
Below, pp. 78-79.
21 S
2529 (horse);S 2531 (hand).
22
Thompson, "WestSide,"pp. 202-203.
23
Thompson has recently suggested (personal communication)that an arch may have been built here in Late An-
tiquity (cf. the one between the Stoa of Attalos and the Library of Pantainos). This would help accountfor the well-
definedeast-west road to the east.
24 Below, p. 78.
25 After the destructionof the statue the base was overrunby a narrow wall set in firm concrete,entering from the
 PROSPERITY (A.D.396-450) 61

The vicissitudes of the buildings on the west side of the Agora, especially the Tholos, the Me-
tro6n, the Round Building, and the large statue base, are now seen to be closely interwoven.Their
history may perhaps be reconstructedas follows:
1) The Tholos and Metroon suffer heavily at the hands of the Heruli in 267.
2) The Tholos is made usable, probablyduring the first half of the 4th century,but the Metroon
remains in ruins.
3) The Tholos is damaged beyond repair by Alaric in 396. Almost immediatelythe Metroon is
rebuilt, probably once more as a synagogue. At the same time the Round Building is erectedin a
conspicuousposition by the PanathenaicWay.
4) With the whole area reclaimed for public use the streets are given a more formal aspect by
© American School of Classical Studies at Athens

placing large statues in key locations:at the northwestcornerof the reconstructedMetroon and the
For personal use only. License: CC-BY-NC-ND.

entranceof the PanathenaicWay into the Agora.

5) At some unspecifiedtime, but before the end of the 6th century,the Round Building is ruined
forever, and a curved line of an aqueduct is laid from the North Mill, passing over the ruined
monumentbase and the Round Building.
Probablyrelatedto these activitiesis the constructionin the 5th centuryof a stoa, at the end of the
Panathenaic Way before it turns into the Agora, its east end built against the still standing Stoa
Poikile.26
While these sometimes makeshift operationswere going on in the Agora, where mere clearing
away of the ruins was a formidabletask, at least one building of a more monumentalcharacterwas
being erected on the foundationsof an earlier structure,probablyjust east of the post-Herulian
enceinte. All that remains is the epistyle of a porch (PI. 46:e) bearing the dedicationof a building
"with its porches" to the Emperors Arcadius and Honorius by Severus Aetius, Proconsul (of
Achaea).27There is some reason to believe that this may be the Hellenistic gymnasium,the Dio-
geneion, the headquartersof the ephebes from at least as early as the 2nd centuryB.C., repairedor
rebuilt severaltimes, most recentlyin the Antonineperiod,until it was finally rebuilt, still underits
own name, even though the ephebes had ceasedto exist as an institution.28
Whether or not the identificationis correct,the building itself is evidencethat substantial con-
struction was being encouragedat a high level almost immediatelyafter Alaric's departure.This
impression is supported by another dedicatoryinscription from the same general area, of which
only a fragmentremains, with letters very similar in size and shape to those of the Diogeneion.29
Anotherbuilding is shadowybut important.It is representedby three inscribedblocksrelatingto
philosophy, known for many years but never satisfactorilyexplained. Two bear, respectively,the
words TflIKiIN and KYNIKI2N (PI. 47:a, c). The third is a fragmentfrom the right side of the

easton a curvedline whichcontinuedon to meetthe northeastcornerof the SquareBuilding(below,p. 109).Hereit

followed around or across the cornerand resumedthe same curvedcourse through the middle of the Round Building.
The line of the curve when projectedeastwardwould fetch up near the point where the Late Roman aqueductcoming
through the Agora changes its course to run westward (below, p. 81). The wall probably carried a branch of that
aqueduct and would have been built to provide power for the mill which must have been destroyedby the railway
(below, p. 81). The firmnessof the concreteprecludesa date later than the 6th century. It may be contemporarywith
the mainlineof the watersystem(ca.475) or elsea lateraddition.In Byzantinetimesmuchof theinteriorof thestatue
basewas dugout anda vaultedtombconstructed inside.
26 1984 10.
Hesperia53, (Shear,Jr.), p. 16, fig.
27
IG II2, 5205.
28
For a fuller account, with reasons for the identificationand date, see A. Frantz, "A Public Building of Late An-
tiquityin Athens(IG II2,5205),"Hesperia48, 1979,pp. 194-203andpls. 62-64.
29E.M. 1861.Frantz,op. cit.,pl. 64:f.
 62 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

block,with only the letters . . . IfN preserved,but the close correspondenceof the three piecesin all
other respectsmakes the restoration[EIIKOYPE]IfIN almost inevitable.30
W. Judeich was the first to considerthe blocksas a group and to attributethem to a single build-
ing.31He was followed by A. von Premerstein,who repeatedJudeich's brief descriptionof them as
epistyles.32But a slight taper in the blockinscribedITfIKfN shows that it must have been a door
jamb.33Its projectedlength (originallyheight) of ca. 3.50 m. is close to the height of the doorsof the
Stoa of Attalos.The fragmentaryEpicureanblock,too shortfor the taper to be detected,is similar in
its other dimensionsand was probablyalso a jamb. The Cynicsblock,with a length of only 2.57 m.,
was probably a lintel or a crowning member of some sort. The surfacesthat would have abutted
© American School of Classical Studies at Athens

againsta wall are roughpicked,as for rubble,thus makinga Romandateprobable.The inscriptions

must have been addedafterthe destructionof the buildingfor which the blockswere made.A sugges-
For personal use only. License: CC-BY-NC-ND.

tion that they were carvedwhile they were still servingas a doorframeis untenable,not only because
of the unparalleledverticalarrangementof the letters (as would have been the case) but also for the
uncomfortablyclose associationof Stoics,Cynics, and Epicureansin one portal.
The date of the blocks in their re-used form is difficultto determine.As is so often the case, the
letter forms can be matched from the 2nd century all through Late Antiquity, with formal in-
scriptionsof a single word continuingto be carvedwith greatercare than are the longertexts, which
tend to becomecrampedand irregularin the later centuries.Close resemblancesmay be seen in the
Panatheniusinscriptionof the secondhalf of the 3rd century,34but that does not necessarilyexclude
a later date for those discussedhere.
Also puzzling is the way in which the blockswere re-used. Certainlythey were laid horizontally.
Two of them are too long and thin for lintels, unless they were embeddedin the masonry of the
walls abovethem;probablythis would also be true for epistyles. Otherwise it would be temptingto
suggest that they were placed over the entrancesto differentdepartmentsin one building housing
several branchesof philosophy.An attractivesuggestion35is that they were built into the masonry
over alcovesfor booksin a library,to designatethe categories.One such librarymight be that in the
Ptolemaeion,to which the ephebes were requiredto presentbookson the completionof their stud-
ies.36What the booksswere is unknown,but the fact that philosophywas an importantpart of their
education,as attestedby inscriptions,37suggeststhat philosophicalworks were well represented.Or
the inscriptionsmight have been used simply as appropriatedecorationon either a facade or an
inner wall to emphasize the intellectualnature of the contents,as sometimeson modernbuildings,
e.g. the National ArchivesBuilding in Washington. But the genitive in all these inscriptionsmakes
that less likely.
A much earlier parallel for this type of use may perhapsbe seen in a series of inscriptionspainted
on a plastered wall in the Gymnasium at Taormina, with brief descriptionsor biographies of
30 70. First noted in the Metochi Penteli on Metropolis Street; later in the Stoa of Attalos; now in the
TWLKwUv,E.M.
courtyardof the EpigraphicalMuseum. D. Kambouroglou,'AOi7va2, 1980, p. 154, also ?roa 1, 1981, pp. 47 ff.
KVVLKwV,IG II2, 5184. Found in the excavation of the Church of the Serapheim on the north slope of the
Acropolis.P. Kavvadias, 'E4'ApX 1898, cols. 17-18, no. 15. Presentwhereaboutsunknown.
'
E7rKovpeLwv.Now among the marblesgatheredin front of the Asklepieion,on the south slope of the Acropolis.
31Topographievon Athen, 1st ed., Munich 1905, p. 100, note 4; Judeich2,p. 105, note 5.
32 "Der Parthenonfriesund die Werkstattdes panathenaischenPeplos,"OJh 15, 1912, pp. 30-35.
33Noted and measuredby J. Travlos. The diminutionis 10 cm. from bottomto top.
34IG II2, 5201, above, p. 10; PI. 15:c.
3SMade by Robert Sinclair.
36M. N. Tod, "Sidelightson Greek Philosophers,"JHS 77, 1957, p. 139.
37Ibid., p. 137.
 PROSPERITY (A.D.396-450) 63

historians.38But all this must remain speculative.What seemsto be certainis that the three inscrip-
tions, on blocks taken from an earlier structure,were used in some way in one building devotedto
intellectual pursuits.

Building Activityof Herculius

Some ten or twelve years after the rebuilding of the "Diogeneion",what was apparentlya con-
certedand intensive building operationwas launchedin Athens, probablyby Herculius, Prefectof
Illyricum from 408 to 410.39
The key to the action, both in the Agora and elsewhere in the city, and to its initiator is probably
to be found in the Library of Hadrian. The Libraryhad evidentlybeen out of use since the Herul-
© American School of Classical Studies at Athens

ian attack, when much of the inner colonnade was destroyed. It was reconstructedin Late An-
For personal use only. License: CC-BY-NC-ND.

tiquity, as shown by several bases in the east colonnade, carved in crude imitation of their fine
Hadrianic predecessors.40An inscriptioncarvedhigh on the wall of the porch of the main (west)
entrance, to the left of the doorway, leaves little doubt of the identity of the sponsor of the recon-
struction. This is a dedicationof a statue of Herculius by Plutarch, who signs himself a sophist
(PI. 47:c)
rTO OEO,v ravTa'7 v vfapXov
EpKOVA&oVayvov
ca-rr7(reoourrfr.41
nIAovrapXos0p.Ucv0 TrajuA7v
The height at which the inscriptionis set indicates a statue of over life size. In its prominentpo-
sition beside the main entranceit would have been conspicuousto everyvisitor to the library.
The appearanceof Plutarch, a sophist, in the epigram raises some questions. Three Plutarchs
achieved distinction of various sorts in Athens at this time; or else one man (or perhaps two) ac-
quired fame for three different reasons. The three (if there were three) had most of their known
characteristicsin common.They were at least near contemporaries,devotedto intellectualpursuits,
and pagan. Probably all were wealthy. The most famous of the three was the founderof the Neo-
platonic School in Athens, who died ca. 432.42 Another was the donor of the statue of Herculius,
and a third, very probably the same man, was honored by the city of Athens for three times de-
fraying the cost of the sacredship at the Panathenaia(PI. 47:e).
Az os?[' E]pex o8atP
os A^i[a]Xoywv avEOflKEV
nIXovrapxov a-rTaOEp7^s
Yp,uao-aoppocvvs!
o KatLrp\s Irorvqrov 'AOqva?T;seeAaa-crev
vavv ecAarasLep/v, rXAovrov bAov7rpoXeas.43
The question is, can the last two be equated with the founder of the Neoplatonic School?44The
donor of the statue describes himself as a sophist. Although even in classical times the terms

38 G.
Manganaro, "Una bibliotecastoricanel ginnasio di Tauromeniume il P. Oxy. 1241,"La parola del passato29,
1974,pp. 389-409;also idemin A. Alfoldi,RomischeFrimhgeschichte, Heidelberg1976,pp. 83-96, pls. I-IV. I am
gratefulto J. F. Gilliamforthesereferences.
39 Thesedatesare secure.Referencesin PLRE II to the CodexTheodosianus mayextendthe spanto 407-412.
40 Someof the have
replacements longbeenvisible(cf. Frantz[footnote33 above,p. 5], pl. 91). Morerecentlynew
evidencein the formof morebasesandotherarchitectural fragmentshas beenuncoveredduringexcavationson the
southsideof the Library(H. W. Catling,JHS Archaeological Reports1983-1984,pp. 8-9).
41IG II2, 4224. I am gratefulto Glen Bowersockfor illuminatingcommentson this questionand for the inter-
pretationofferedhere.
42 PLREI, Plutarchus5 (andII, Plutarchus1).
43IG II2,3818,carvedon the backof IG II2,3692 (PI.47:d).(Theepsilonandalphain line 1, bracketedin IG, are
clearlyvisiblein the photograph.)
Bothmenareregardedas the sameindividualin PLREII, Plutarchus2.
44 As
by Wilamowitzap. G. Kaibel,Epigrammata Graeca,Berlin1878,p. 376.
 64 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

philosopher, sophist, and rhetor were not always mutually exclusive,45most commentatorsnow
rejecta commonidentity on the groundthat a philosopherwould not style himself a sophist.46The
essence of the epigram, however, is a subtle comparisonof the dual and contrastingroles of both
honorand,"stewardof laws and gentle administrator",and donor, "stewardof myths"(with over-
tones of Platonicphilosophy)and "sophist".The benefactorof the Panathenaiais honoredas "king
of logoi"(,3ao<L&evsAoywv),presumablysophist, but not incompatiblewith "philosopher".
The intractability of LXAoo6o40osin elegiac or hexameter verse as an argument for allowing more
flexibility to o-oLo-r's '47 would have much weight in earlier times but less in the less careful 5th
century.At any rate, it might be a factoragainst excludingPlutarchthe Neoplatonistas the sophist
© American School of Classical Studies at Athens

who honoredHerculius.
The field is limited to some extent by the one fixed chronologicalpoint, 407-412 (the outer limits
For personal use only. License: CC-BY-NC-ND.

of Herculius' term of office)or thereabout,for the erectionof the statue. The donormust have been
a man of considerablewealth, distinction,and influence for such a conspicuousgesture, especially
if, as is generally agreed, he was also responsiblefor three times mounting the Panathenaicpro-
cession. Plutarchthe Neoplatonist at age about sixty fits all these specifications.No sophistsof the
same name are known. Among the few Plutarchslisted in PLRE for the appropriatetime only the
otherwise unknown father of the philosopherHierius has been suggested,but he is mentionedby
Damasciusonly in a contextof philosophers(Vita Isidori88 [EpitomaPhotiana]+ Frag. 289), and
so his claim is no more valid than that of the famous Plutarch. Lacking further evidence the
problemmay never be solveddefinitely,and the questionremainsopen.48
Incidentally,the impressionof Herculius as a man of letters (and also a pagan) is strengthened
by a statue base bearing the dedicationof another statue of Herculius by another sophist, Apro-
nianus (PI. 47:f), this one on the Acropolis,beside the statue of Athena Promachos.49
At about the same time as the rehabilitationof the Libraryof Hadrian, the centerof the Agora,
which had escaped the modest beginnings of reconstructionon the west side, was finally cleared
of the most obstructiveof the Herulian debris. Here an imposing complex was erected over the
45For a numberof examples see Tod (footnote36 above,p. 62), p. 138, note 93. An outstandingexample in the 4th
century is the philosopher-sophistThemistius (PLRE I, Themistius 1). For the possibility that even Syrianus, in his
commentaryon Hermogenes, referredto himself as a sophist, see RE IV A, col. 1732. On the intricaciesof the dis-
tinctions between sophist, rhetor, and philosopher,see G. W. Bowersock,GreekSophistsin the Roman Empire, Ox-
ford 1969, pp. 10-16. See also G. Fowden, "The Pagan Holy Man in Late Antique Society,"JHS 102, 1982, pp. 43-45.
46 E.g., Robert (footnote38 above, p. 6), pp. 73, 95; Saffrey and Westerink,p. xxx; Millar, "Dexippus,"p. 17 and

note 64. PLRE follows Robertand Saffrey-Westerinkand gives him a separateidentity.

47 Observedby Tod ([footnote36 above, p. 62] as in his note 127, p. 141).
48 A suggestionmade by W. Peek ("Zwei Gedichteauf den NeuplatonikerPlutarch,"ZPE 13, 1974, pp. 202-204)
that the Neoplatonist, or at least anotherphilosopherof the same name, appearson an inscriptionin the Epigraphical
Museum, no. 4713, must be rejected.The drawing accompanyingthat article shows that the propername is restored.
(Cf. the photographof the fragment in M. Mitsos, <'ArobTOVSKaraAoyovs 'AOqvaicov'E ,83owvo,'ApX'Ej 1971,
pl. 7.) This was pointed out to me by G. W. Bowersock.For anotherPlutarch "PraesesInsularum",see A. Chaniotis,
"Plutarchos,Praeses Insularum,"ZPE 68, 1987, pp. 227-231. But he presumablywould not qualify.
49IG II2, 4225.
Tiro]v&7rafrLv
[r]ov 7rpoplaxov0e0/rzhv 'E[px]ovA&o[v
f'C0ovov OCKwV[V]1O6OEv
a[L7r]vrrTawv
a'o f[tor]rT7s
[8]e&vbos'AOTqvawv'A7rpomavo' -E?
[1o]T?Jc?7rapa IHpo,AuaX)
HaxAabt KfKpoT7l['7r].
Cf. A. Frantz, "Herculiusin Athens: Pagan or Christian?"Akten d. VII Int. Kongr.f. Christl.Arch. Trier, 5.-1i1.
Sept. 1965, Vatican 1969, pp. 527-530. Groag (Achaia, p. 73) suggests that Apronianus, unknown except for this
epigram, may have been the predecessorof Leontius in the chair of sophisticin Athens.
 PROSPERITY (A.D.396-450) 65

foundationsof the Odeion of Agrippa and adjacentbuildings to the south. Formerlyknown as the
"Gymnasiumof the Giants",it will henceforthbe referredto as the "Palaceof the Giants".50
Conclusiveevidencefor the constructionof the Palace between about 410 and 425 was provided
by pottery and coins.51Considering its proximity to the Library of Hadrian, it is perhaps not
stretchinga point to suggest that Herculius might be creditedwith this establishmentas well.52
Probablyto be connectedwith the Palace is a statue found built into a modernwall not far from
the entrance (PI. 66:b). The statue is slightly over life size, weighing about half a ton, and so it had
probablynot strayedfar from its place of origin. The costume,the toga trabeata,betokensan official
of high rank.53
© American School of Classical Studies at Athens

In considering the setting of the Palace complex, notice should be taken of a series of statues,
probably made for the Odeion of Agrippa in its Antonine phase and salvaged from the ruins to
For personal use only. License: CC-BY-NC-ND.

adorn the environmentof its successor.Three were seated figures, one recognizableas Epicurus,
which may have originallyoccupiedthree bases in frontof the entranceto the Odeion.54Sevenmore
are of a single type, a standing figure wearing a himation.55Five of the seven were found in a
stragglingrow just opposite the east wall of the great courtof the Palace and moreor less coinciding
with its length. The other two were found in fragmentsnear by. Their position makes it probable
that all seven had stood against the aqueduct of the late 5th century on the other side of the Pan-
athenaic Way.56Since the aqueductis some sixty years later than the Palace it seems probablethat
during those sixty years the statues were more closely relatedto the Palace and were later shiftedto
their final position, perhaps to mask the utilitarian aqueductwhere it rose above groundlevel.
A similar situation might be observedin a series of five seated figures discoveredoutside the city
wall, northwest of the Dipylon Gate. Four were found in 1968 built into a Late Roman wall; the
fifth, discoveredin 1963, was in a Late Roman building close by.57Although headless, two are
identifiedas Epicurus himself, one as possiblythe EpicureanphilosopherColotes;the othertwo are
unidentifiedphilosophers.In the re the foundationsof a Late Roman building. This leads
to the speculationthat here might have been the "Gardenof Epicurus"in a 2nd-centuryrebuilding
prompted by the encouragementof Hadrian and his mother Plotina, which gave rise to a strong
revival of Epicureanism.58Opinion is divided on the location of the Garden, whether within or
outside the walls. Dontas, relying on Cicero's description (de fin., 5.1-3) favors a location out-
side the walls. R. E. Wycherley, on the basis of a passage in Pliny, the other majorliterary source
50 For the
reasons, and a detailed descriptionof the building, see ChapterV.
51
Including 41 coins of which 14 (the latest) were struck in the decade 383-392 (reigns of Theodosius I and Ar-
cadius), found in a lime-slaking pit obviouslyused in the constructionof the building.John H. Kroll was kind enough
to examine the coins in detail and concludedthat all had been in circulationfor more than a decadeand that they had
been thrown into the pit "somewhatafter 400, certainly in the first quarter of the 5th century."An adjacentpit pro-
duced24 coins of which the latest were 9 of Theodosius II and 2 of Arcadius,also struckin the decade383-392. For the
pottery, see Thompson, "Odeion,"p. 134.
52 It should be noted that Herculius' benefactionswere not limited to Athens. He was honored
by the people of
Megara for the gift of an aqueduct,presumablyto replaceone destroyedby Alaric (IG VII, 93).
53S 657. For a fuller
descriptionof the statue and its significance,with references,see below, p. 113.
54 S 826, S 930, S 1304. Thompson, "Odeion,"pp. 124-125. S 1304 is representedonly by a fragmentof the chair.
55 Agora I, nos. 57-63.
56 Agora I, pp. 74-75 (note by H. A. Thompson).
57 G. Dontas, <<?EKOVLrTLKa B'?, AEAr 26, 1971, A' (1973), pp. 16-33, where the statues themselves and the sig-
nificanceof their finding place are studied in detail.
58 Cf. ILS 7784 ( = SEG III, 834) and J. H. Oliver, "An InscriptionConcerningthe EpicureanSchool at Athens,"
Transactions of the American Philological Association 69, 1938, pp. 494-499.
 66 IV. PROSPERITY, DECLINE, AND DISASTER (A.D. 396-582)

(NHxIx.5o-5sI), referring to the Garden as in ipsa urbe, would place it within the walls.59 Dontas
counterswith the suggestionthat in view of the crowdedconditionswithin the walls Epicurus'own
house, with a very small garden,might have been inside, while the actual Gardenwas outside.The
archaeologicalevidenceseems to favor Cicero'saccount.60Dontas notes that when a series of por-
traits occurs it denotes either a public building, e.g., library, gymnasium, odeion, etc., or else a
privatehouse of exceptionalimportance.The statuesfromthe Odeionmight well fall into this same
category.
A footnote to this monumental undertakingmay perhaps be read in an event mentioned in a
single sentencein Olympiodorus'History,when the Athenians,confrontedwith the need to under-
© American School of Classical Studies at Athens

stand the method of transcribingbooks accordingto the K&AoV, awardeda statue, or at any rate a
herm or even a painted portrait (uKdova),to Philtatius, a friend of the author and a grammarian,
For personal use only. License: CC-BY-NC-ND.

who was able to explain it to them.61A crisisthat couldprovokesuch a need could seem only to have
been some large-scalereorganizationof libraries. The year was 416/7. Once the reconstructionof
the Libraryof Hadrian was finished,replenishmentof the books,which must have sufferedjust as
much as the building, would have been an immediate necessity. This would have provided an
occasionto modernizethem by transcribingthe new copies accordingto the K&Aov,a system that
had fallen into disuse but had recentlybeen reintroduced.It is quite possiblethat this was the direct
outcomeof the wave of renovation,of whatever sponsorship,that took place in the first quarter of
the 5th century, and that the sponsor, Herculius or another,was personallyresponsiblefor seeing
that the bookswere replenishedas well.62
Another building, closely related to the Palace in location and date, was possibly part of Her-
culius' general scheme:a large structure,about twenty-fivemeters square, not far from the north-
east cornerof the Palace (below, p. 109). Its plan is uncertainin some details, as is also its purpose,
but the regularityof the plan and the scale of the "SquareBuilding"are more in keepingwith some
public edifice than a private house, and its location makes its public characteralmost certain. It is
not improbablethat it was connectedin some way with the Palace.
In contrastto the often nebulousand almostalways imprecisedating attendingbuilding activities
in Athens in Late Antiquity, the sharp focus of the elements of this building programof the first
quarterof the 5th centuryis refreshingand significant.63
Of the area between the Greek and Roman Agoras,just inside the Post-Herulian Wall, only a
small part has been excavatedup to the time of writing (1988). A large building of the 2nd century
behindthe Stoa of Attalos, its north side aligned with the ancientstreet,was reclaimed,probablyin
the early 5th century (Pls. 6, 48:a). But it was in a curtailedform, since the Post-Herulian Wall
overran its northeast corner. Some bits of mosaic similar to that in the revived Metroon have
appeared, but most of the building lies under the scarp forming the present limit of the Agora
Excavationsin that particulararea. Its continuation,however, has been picked up farthereast, on
the site of the Church of St. Thomas.64
59 "The Garden of
Epicurus,"Phoenix 13, 1959, pp. 73-77.
recently, M. L. Clarke ("The Garden of Epicurus,"Phoenix 27, 1973, pp. 386-387) chooses a location
60 Most
outsidethe walls becauseof an episode in Heliodorus,AethiopicaI.i 6.5, which may be more entertainingthan accurate.
61Olympiodorus,frag. 32 Dindorf.
62
For a fuller discussion,see Frantz (footnote33 above,p. 5), pp. 377-380. A precedentfor includingthe bookswith
the building will be recalledwhen, some centuriesearlier, Pantainospresentedthe city with "a librarywith the books"
(T2rv P,1tXAoL07JK77vlAera rTv fi8ltkiv); Agora XIV, p. 115.
63 For the possibility that the rebuilding of the Parthenon was the crowning feature of Herculius' program, see

Frantz, "Julian,"pp. 395-401.

64
P. Lazaridis, AeAr 26, 1971, B' 1 (1974), p. 63 and fig. 1.
 PROSPERITY (A.D.396-450) 67

On the south side of the east-west street leading from the PyrgiotissaGate to the gateway to the
Roman Market the excavationcarriedout so far has yielded results of considerableimportancefor
the history of the city in Late Antiquity. Here a grandiose marble-pavedand colonnadedstreet
(PI. 70) was laid out about the time of Trajan as an approachto the Roman Market (the Broad
Street).65But Herulian damage was severe, and the colonnadeon the souththe side of the street (the
only side so far uncovered)was totally destroyed.The Libraryof Pantainostoo was left in ruins. A
section of the Post-Herulian Wall was built over the western part of the colonnade,using as con-
struction material many blocks fromincluding the including the dedicatoryinscription.66The rest
was left to the mercy of looters.
No attempt was made to clean up the debris until the first quarter of the 5th century,when the
© American School of Classical Studies at Athens

stoa of the Broad Street was rebuilt for its entire length, using columns of miscellaneouscharacter
For personal use only. License: CC-BY-NC-ND.

and sizes to replacethose that had been destroyedor carriedaway. One sectionof the colonnadewas
used to form the principal facadeof a large, apparentlypublic building (PI. 50:a), the main part of
which had as its
its ground level the high terracebehind, to the south. Te chief feature of the piano
nobile was a large room with an apse at the end, set off by a pair of columns. At the opposite end
was a square peristyle and on the sides a group of smaller rooms (Pls. 48:b, 49). The scale of the
building and its interior arrangementsare suitable to an audiencehall, perhaps for the praetorian
prefect or the proconsul.The rebuilt colonnadeat the lower level, with its three lavishly appointed
rooms with marble-revettedwalls and niches for sculpture (P1. 50:b), would make an appropriate
formal entranceand waiting room where a visitor might be receivedbeforegoing up to an audience
with some high official. At present no trace of a staircasehas been found, but there seems a like-
lihood that a wooden one led up from one of the smaller rooms on
the north side (Room G). The
building extended up to the end of the Broad Street at its intersectionwith a narrowerstreet enter-
ing from the north. A very modest latrine was built against the outer wall of the stoa at this point
(P1. 21:e).67
A date of constructionca. 420 was determinedon the evidenceof a considerableamountof pottery
and, in a footing trench for one of the walls, a coin of Theodosius I, minted between 383 and 395
but worn to such an extent that it must have been in circulationfor about a generation.Five iden-
tical coins were later found in the same trench, along with one each of Theodosius I and Arcadius,
both struckin 383.
A drain of the 2nd century along the south side of the apsidal building showed the course of the
street on the upper terrace, which remainedin use in the 5th century, even though by that time it
came to an abrupt end at the Post-Herulian Wall.68Across the street are the remains of a bath of
the Late Roman period, most of which lies under the modern houses just beyond the present ex-
cavation zone. This raises the possibility that other structuresmight have been erectedon the ter-
race in connectionwith the apsidal building that was obviouslythe centralelement in the program.
Soon after the departure of Herculius from office the constructionof large secular buildings
seems to have tapered off or ceased utterly, so far as present evidencepermits a judgment. Lack of
motive and withdrawal of funds may have been partly responsible;69the mounting threat from
65
Hesperia 42, 1973, pp. 389-391 (Shear, Jr.). The west end of the street on the north side and some of the marble
paving were exposed by Greek excavatorsin 1898-1899 (TipaKrtKa 1899, pl. II), when it was misidentifiedfirst as a
Roman villa and later as a latrine.
66
AgoraXIV, pp. 114-115; Hesperia 42, 1973, pp. 385-395 and 44, 1975, pp. 332-345 (Shear,Jr.).
67
Hesperia 44, 1975, pp. 332-343, figs. 1-3, pl. 78 (Shear,Jr.). For the later historyof the building and the stoa, see
below and Hesperia 42, 1973, pp. 395-398 (Shear, Jr.).
68
Hesperia 44, 1975, p. 337 (Shear, Jr.).
69
Even the Palace of the Giants was not completelyfinished;the floor of the great court was evidentlypreparedfor
mosaic, which was never laid (see below, pp. 105-106).
 68 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

various barbariantribes, who were making increasinglyfrequentassaults on Greece both by land

and by sea, may have played an important part as well.70 Later works tended more toward the
utilitarian, such as water mills and aqueducts, also the ecclesiastical.71Private building, on the
other hand, seems to have continuedthrough much of the 5th century, at least the third quarter,
sometimeson a fairly grandscale. The remainsare fragmentary,often consistingof only as much as
was discoveredin a single building lot during modern construction,and usually covered up or
destroyedpromptly. The most importantof these are in the southern and eastern sections of the
city, a favoredresidential district for the affluent in Imperial times. Many, perhaps the majority,
are reconstructionsof earlier buildings, usually of the 2nd century,an exceptionbeing a large new
© American School of Classical Studies at Athens

villa in the National Gardens with mosaic floors of about the middle of the 5th century.72Of the
For personal use only. License: CC-BY-NC-ND.

others,most are baths, some, at least, connectedwith gymnasiaor similar establishments.The most
importantexamples are the following:
1) Part of a nymphaeumat the cornerof Nike and Apollo Streets(Spiro 22). Describedin detail by
I. Threpsiadesin <?PoAai'K?1 &'7ravXAts
Ev'AOvvat?>>, loX4eulwv5, 1952/3, pp. 126-141. Probablybe-
longs with a bath (Travlos, PDA, s.v. Balaneia, Bath M, p. 181), discoveredin 1921 (B. Schweitzer,
AA [Jd137] 1922, p. 258).
2) A large bath destroyedwhen the Zappeion was built, leaving no traces. Reportedto have had a
mosaic floor resembling one in the villa in the National Gardens. Perhaps part of a gymnasium
(Spiro 21; Travlos, ibid., Bath K).
Two baths were found by chancein the northwesternpart of the ancient city:
3) Euripides Street (Spiro 24; Travlos, ibid., Bath R; M. Chatzidakis,AeXA 22, 1967, B' 1 [1968],
p. 18). The mosaic is now displayedin the courtyardof the ByzantineMuseum.
4) Sarre Street, north of the Hephaisteionand northeastof the Kerameikos.No mosaic.Two coins
of the second half of the 4th century were found in it (AZAr 20, 1965, B' 1 [1967], pp. 49-51;
Travlos, ibid., Bath S).
THE RISE OF THE CHRISTIAN CITY

The pagan momentumcarriedalmost uncheckedthrough the first half of the 5th century. The
willingness of the city officialsto seek the counsel of Proclusargues for a pagan-orientedsociety,at
least at the upper level. The Christian populationwas undoubtedlygrowing, but it is represented
only by semiliterategravestones73and Christiansymbolson lamps and pottery.
It is not to be supposedthat there was a sharp line of demarcationbetweenpagans and Christians
in general. Opportunism and expediency undoubtedlyplayed a part in many conversions,as in
the case of Theagenes, archon and close associate of Neoplatonists in the 480's.74 Potters and
Lemerle (footnote213 above, p. 51), pp. 297 ff. For a convenientsummaryof the activitiesof some of these
70 Cf.

tribes,see A. A. Bon, Le Peloponnesebyzantinjusqu'en 1204, Paris 1951, pp. 14-17, and below, pp. 78-79. An edictof
424 (CTh xvi.5.4), addressedto the Prefect of Illyricum,exemptedthe people of Delphi, on accountof "newkinds of
losses" from exhibiting "the usual spectacles to the people of the Eternal City" (trans. Pharr). Might this reflect a
similar situation in Athens?
71
Below, pp. 79-82.
72
Discoveredand partiallyexcavatedin 1846. Descriptionand illustrationsof the mosaicsin Spiro, CorpusI, Attica,
nos. 14-20.
73For a selectionof these, see C. Bayet, De titulis AtticaeChristianaeantiquissimis,Paris 1878 and J. S. Creaghan
and A. E. Raubitschek,"EarlyChristian Epitaphs from Athens,"Hesperia 16, 1947, pp. 1-54. It is unfortunatelyim-
possible to fix the limits of their dates more closely than from the middle of the 4th centuryto sometimein the 7th.
74
Damascius, Vita Isidori,frag. 4, 258 (Zintzen, p. 209; Asmus, p. 93); Frantz, DOP, p. 197, note 60.
 PROSPERITY (A.D. 396-450) 69

lampmakerscatered to both factions, with the same shops offering a choice of Christian symbols
and pagan scenes,75and pagans and Christianswere buried in the same cemeteries.76
The Church itself has left little recordof this period. For the 5th centurythe names of only three
bishops are known, and these only for having attended church councils: Modestus Ephesus in
431; Athanasius I, Corinth in 458; and Anatolius, Constantinoplein 459. A fourth, Clematius, is
known only from his tombstoneof uncertain date, found in the area of a Christian basilica near
Lykabettos.77
Legislationfor the Establishmentof Christianity
The legislation for the suppression of paganism when taken as a whole shows that the main
© American School of Classical Studies at Athens

preoccupationof the Imperial governmentwas neither with doctrine, nor with the temples and
For personal use only. License: CC-BY-NC-ND.

their contents,but with sacrifices.Between 345 and 435 sixteen decreeswere issued on this subject,
far more detailed,explicit, and intemperatein languagethan those dealingwith the buildings,with
appropriatesentencesprescribedfor differenttypes of violation.78Frequent reiterationraises seri-
ous doubts about the efficacy of the laws. Their impotence is in fact often demonstratedby the
phraseology:". . this regulationhas been very often decreedby repeatedsanctions"(on the issue of
cult images);79also on the extinction of pagans: "the regulationsof constitutionsformerlyprom-
ulgated shall suppress any pagans who survive, although We now believe that there are none"
(April, 423), but "proscription... and exile shall restrainthe pagans who survive"(June, 423);80
and on the same day "Christians... shall not... dareto lay violent hands on Jews and paganswho
are living quietly and attempting nothing disorderlyor contraryto law."81Finally, in 438, three
years after the order for the destructionof the temples, came a wordy, sometimesincoherentdia-
tribe by Theodosius II demanding"watchfulnessover the pagans"and their "heathenenormities",
"stubborninsolence",and open contemptof the laws in spite of the heavy penalties prescribed."If
they cannotbe reformed,at least they might learn to abstainfromtheir mass of crimesand from the
corruptionof their sacrifices."82
The actual status of the temples of Athens up to the end of the 5th century is obscure.One fact
alone is clear, namely that the blanketcall for destructionof all the templesby Theodosius II in 435
had little effect, as demonstratedby the continuedexistence of all the major temples of Athens.83
Edicts both before and after Theodosius show varying attitudes toward the buildings themselves.
One, issued in 399, orders that the "templesin the countrydistricts ... shall be torn down,"84but
75E.g.,the lampmakerXLOvr/, whoseshopflourishedin the lastquarterof the 5thcentury(AgoraVII, pp. 55-56).
For a similartendencyin the West,but on a higherlevel,cf. E. Kitzinger,ByzantineArtin theMaking,Cambridge,
Mass. 1977,pp. 34-35. The humblerAthenianwaresweredesignedsimplyto pleaseall customers.
76
Representingthe longestspan of continuoususe is a cemeterydiscoveredin 1972 southof the Olympieion.It
contained42 graves,veryclosetogether,with offeringsrangingin datefromthe Geometricperiodto the 7th century
afterChrist(AeXAr 28, 1973,B' 1 [1977],p. 32, no. 13 andpls.21-25). Another,betweentheTheaterof Dionysosand
the Odeionof Pericles(andextendingoverits ruins),was in use fromthe timeof the HerulianinvasionthroughLate
1876, p. 21 and 1931, pp. 28-30).
Antiquity (HIpaKrLKa
77EMME I, p. 21 forthe list;p. 56, fig.43 forClematius'tombstone.
78
CTh xvi.io.2, 4-II, 13,1 5, i8, 20, 23, 25; and esp. o0.12.
79 CThxvI. o.i9. TranslationsarefromPharr.
80 CTh xvI. I
.22-23.
81CTh xvI. 1 0.24.
82Novels Theodosius in.8.
of
83Frantz,DOP.
84 CThxvi.io.i6.
 70 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

four, issued by the same emperors(Arcadiusand Honorius), decreethat the temples should not be
destroyed,adding (in two decrees)that they shouldbe turnedto public use.85Finally, 23 yearsafter
the supposedly final order for destruction,a lengthy edict of Leo and Majorian providedfor the
preservationand protectionagainst cannibalizationof "all the buildings that have been foundedby
the ancients as temples and as other monuments."86Only one decree calls for the closing of the
temples,87although this may be implicit in others specifyingthat they be turnedto public use; and
three more prohibit the approachto or entranceinto any temple.88An interestingexceptionto the
one existing edict on the removalof cult images may be noted in a decreeissued in 382 by Gratian,
Valentirian II, and Theodosius I,89 providing that "the temple shall continually be open . .. in
which images are reportedto have been placed which must be measuredby the value of their art
© American School of Classical Studies at Athens

rather than by their divinity."All celebrationsand festivitiesare to be preserved,and the temple is

For personal use only. License: CC-BY-NC-ND.

to be open "in such a way that the performanceof sacrificesforbiddentherein may not be supposed
to be permittedunder the pretet smple.uch temple."Althoughthis edict was addressedto
accesstoAlthough
the
Palladius, Duke of Osrhoene and so probably referred specificallyto the temple at Edessa, a city
of Osrhoene,90officials in other cities, not least in Athens, would have been quick to exploit the
loophole.

Closingof Temples
It is likely that only after Christianitywas firmly establishedand acceptedby the majorityof the
population were the major temples of Athens convertedinto churches.91A period of transition
seems to have intervened, during which the temples were perhaps officially closed but actually
accessiblein an inconspicuousway. Before the end of the 5th century probablyall the cult statues
had been removed.This phase by its very nature leavesno archaeologicalevidence,but a chapterin
Marinus' VitaProclisheds some light.92This is the descriptionof Proclus'healing of Asclepigeneia
by his prayersin the temple of Asklepios.While he was praying,the girl's conditionsuddenlytooka
turn for the better, and when he finished he walked over to her and saw that she was completely
restoredto health. Such an act, the biographercontinues, was performed"escapingthe notice of
most people, and providingno pretext for those wishing to plot against him."93The incident has
been used to support the theory that all temples were closed, that such prayersas were offered in
them had to be in the greatestsecrecy,and even that Procluaswasdeniedentranceto the Parthenon.
The first two suppositionsare mutually exclusive. If the temple of Asklepios was closed, then the
event could not have taken place at all, no matter how secretly.All the passage shows is that the
deed was performedunostentatiously.Its spectacularnature (Oav^iayap o`VrsWKaLaKova-aL)is
intensified by the attribution of godlike powers to the doer (pft^a yap 6 o-wHrjp,WoTre?Oo, laro), all
of which, in the minds of ill-intentionedpersons,would amountto magic, one of the greatestcrimes
85 CTh 19.
XVI.10.3, 15, i8,
86 Novels
of Majorian Iv.i.
C87 Th xvI. 0.4.
88CTh xvi.I. IO, II, 13.
89 CTh xvi.io.8.
90
Pharr, p. 473, note 17 to CTh xvi.io.8.
91 Frantz, DOP.
92 Vita
Procli, 29.
93
Evxop,evov oeavrov
roy apXaLorepoV rpo7roV,
paorTIva6poa K.?ETao^A7 7rep&rTiv KOopJVEfalvero Kat )7aj vf]
o
yLyveTO' pela yap 'rvcwp, o oTre
OeoS, laro. v.TA^p(oevrwv oe ro8TWepo)V, 7rpoS r7Tv 0Ao^KA^tyyvetav e?a8l&e Ka&

KareXaf4pavev avrT7v
aV r apTL rf. TW)oV7TrpLEo'oToTWV ro or'Soa XAV,\V.EV7v7raoov, ev vyLELvtj Oe KararaoE botayovorav.
'
KCLTOLOVTOvepyov boE'pa6aro OVKaAAoCs 7X Kavravua rovs 7TOAOVs Aav6avwv Kal ove z.av eTaov-poqao'&vr ToSs
XeveLVeOEAovo-L rapaor-Xv.
 PROSPERITY (A.D.396-450) 71

in the eyes of the Christians.That might have been the reasonfor avoidingpublicity.And the aorist
(LE7rpa4arTo) implies that this was an occasional,not habitual occurrence.The edicts quotedabove
give the impressionthat the pagans, providedthat they abstainedfrom sacrificing,went about their
business in comparativefreedom.
At the time of Proclus'"miracle"the temple of Asklepioswas intact,with its contents,as Marinus
says at the beginning of his account.94But by the time he wrote, the year after Proclus' death, the
statue must have been removed.
The Athena Parthenos is the only other cult statue known specificallyto have been removed,as
describedin the next chapter of the Vita Procli; this within Proclus' lifetime.95Within the time
© American School of Classical Studies at Athens

limits of Proclus' return from Lydia after 450 and his death in 485 it is hard to be more precise
about the date. It seems likely that the events describedwere part of a general, though unofficial
For personal use only. License: CC-BY-NC-ND.

program and that both took place later rather than earlier in that period, for it would be hard to
reconcileMarinus' statementthat Proclusenjoyedhis house becauseof its proximityto the Temple
and Theater of Dionysos and the Temple of Asklepios96with the presenceof buildings, especially
revered by him, that had been stripped of all their significance.The result, coming after all the
fruitlessedicts for the suppressionof paganism,was probablybroughtaboutby the inexorableforce
of Christianitywhich finally made its way up into the highest and most powerful levels of society.

Conversionof SecularBuildings
The building commonlyknown as the Agoranomeionin the Roman Market may possibly have
been one of the the first ancient monumentsin Athens to be convertedinto a church, but without
architecturalchange.97The evidenceis meager-only a few Christiansymbolscarvedon the walls.
But taken in conjunctionwith the adjacentbuilding, the Tower of the Winds, or Horologium of
Cyrrhestes,it becomesmore plausible. The earliestknown referenceto the Horologiumas a church
is found in the 15th-centurydescriptionby the Anonymousof the AmbrosianLibrary:"al presente
e una chiesa dei greci et e opera molto degna, tutta di marmoro."98 Two Maltese crossescarvedon
the jambs of the doorway,however,show that it had servedsome Christianpurposemuch earlier.99
Its octagonalplan and convenientlocationhave given rise to the belief that here was a baptisteryfor
the church(?)in the Agoranomeion.100 The fact that the Tower of the Winds was alreadyequipped
with an elaboratewater systemfor the clockworkswould have been an addedinducement,and since
both were secular buildings, the Christiansmight have been less hesitant to claim them beforethey
achievedcompletesupremacyin the city.
It might be that a church in the Agoranomeionand the 7th-centurybasilica (below, p. 73) under
the Fetiye Djami are related chronologically,the Agoranomeionservingbriefly as a churchbefore
the constructionof the basilica. If such were the case, the conversionof the Horologium perhaps
took place between these two events, as the area gained a more formalplace in the Christiancity. In
94K(a yap 7)VTVXI Xf 7oAS Kr TOre LI ElXeV ET aITop7T0rov rTOTov 0 wTjpos lepov. For the interpretation of
as "undespoiled",i.e., with the cult statue still in place, ratherthan that the temple was undestroyed,see
"a1ropOrjTov"
Frantz, DOP.
95Above, 58.
p.
96 Vita Procli, 29.
97AETr 27, 1972, B' 1 (1976),
pp. 17-21.
98 E. Ziebarth, "Ein griechischerReiseberichtdes XV. Jahrhunderts,"AM 24, 1899, 77.
p.
99See H. S. Robinson, "The Tower of the Winds and the Roman Market-Place,"AJA 47, 1943, 291, note 1.
p.
100First suggestedby A. Orlandos,
'ApX'E4 1964, HapapT71f.a, pp. 58-59, figs. 110-112. Cf. also D. B. Small, "A
Proposal for the Reuse of the Tower of the Winds," AJA 84, 1980, pp. 96-99. For both buildings see Travlos,
Ho\eobopiLK , p. 139; idem, <<Xpl-TravrKai
OpoXKpEVTrKo'AOivaL>, Kat HLDI 'EyKvKAo7radkia, Athens 1962,
pp. 726-727.
 72 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

keeping with this theoryis the form of the Maltese crossesin circles,a late but still Early Christian
form.101The presence of a number of Christian gravestonesin the vicinity also suggests a con-
version of the two buildings in the Early Christianperiod, but there is little to distinguishin these
between the 5th, 6th, and 7th centuries,as is also the case in the graffitiof a bird,a crossmonogram,
and a fish on one of the columns.102

First ChristianChurches
The earliest Christians in Athens left no trace of their places of worship. Very likely, as else-
where, they confinedthemselvesto privatehouses, perhapseven throughthe 4th century.Repeated
decrees aimed at enforcing the laws against paganism were widely ignored in Athens, even such
© American School of Classical Studies at Athens

seemingly iron-clad statutes as the Edict of Theodosius II in 435, calling for the destructionof all
For personal use only. License: CC-BY-NC-ND.

temples.103The continuedexistenceof the majortemplesof Athens as well as in many other partsof

Greece is testimonyto the impotenceof the laws. Even officialrecognitionof the helplessnessof the
Imperial government,at least for some parts of the Empire, is implicit in the clause entrustingthe
implementationto "thelocal magistrates",who were surely, in the case of Athens, pagans.104
But the mounting pressure from Christianitybecame acutely perceptibletoward the middle of
the 5th century. The truce which seems to have prevailed between the two factions became pro-
gressivelymore tenuous as the centuryneared its end. The Christianpopulationwas undoubtedly
growing, and it is obviousthat it was no longer confinedto the indigent.We have seen that Proclus
was seriously, if only temporarily,inconveniencedby being obligedto leave Athens for a year.
A substantialnumber of churchesmust have been built in the 5th and 6th centuries,to judge by
the abundantexamples of carvedarchitecturalornamentfound in the city, only a few, however, in
their original context.'05The places of worship in the 5th, 6th, and 7th centuries fall into two
categories:1) totally new structures,usually basilicas,and 2) templesand otherancientmonuments
convertedinto churches,with or without structuralremodeling.The churchesof category2 tend to
be later than those of category 1 and will be discussedin ChapterVI.106
More recent investigationshave necessitateda revision of previouslypublished lists,107with the
eliminationof some churchesfor which no satisfactoryevidenceof their existenceremainsor which
are now seen to be too late to qualify for inclusionin the presentvolume.
Probablythe earliestof the new churchesin Athens was the tetraconchin the courtyardof the Li-
braryof Hadrian (PI. 51). A long controversyas to whether it was originallya secularor ecclesiasti-
cal building was finally settled in 1980 when John Travlos discoveredthe beginning of a narthex
well bondedinto the masonryof the main building.108The previouslyaccepteddate in the early 5th
101Orlandos,op. cit., p. 59, fig. 112. Similar crossescan be seen on the doorwayof the late, Christian, Hephaisteion.
102 Ibid., figs. 110 and 111.

103CTh
XVI.IO.25.
104This
whole topic is discussedmore fully in Frantz, DOP.
Probablymost of the basilicas, on the outskirtsat any rate, were destroyedin the Slavic invasionsof the late 6th
105

centuryand later, leaving the ruins as a quarryfor later builders.Many pieces, after successiveuses all overtown, may
have ended up far from their places of origin. Most of these are now exhibited in the courtyardof the Byzantine Mu-
seum;others are in the Agora collection.
106Category2 will also include churchesbuilt over the ruins of temples previouslydestroyed,e.g. the Asklepieion.
107
E.g., Travlos, HoXeo8bo1lKLI, pp. 137-144 and <<XptrTavLKa''AO6jvaL>, cols. 721-733 (footnote100 above,p. 71).
108 For the new identification,see
J. Travlos, ?To rerpaKcovXoolKobo,t tAdarTs Bt3AtoOBQKrS rov 'Abptavov>,
^f4XLa
"E7r7els F. E. MvXAwvav, Athens 1986, I, pp. 343-347. The previous literature includes A. M. Sisson, "The Stoa of
Hadrian at Athens,"Papers of the British Schoolat Rome 11, 1929, pp. 66-71; G. Sotiriou, <?AwaAa&oxp&L<rTav&Kal
3aoLALKal Tr?S 'EXAAa/ov,'ApX'E4 1929, pp. 173-175; Travlos (footnote 33 above, p. 5), pp. 41-63 and PDA,
pp. 244-252.
 PROSPERITY (A.D.396-450) 73

centuryshould now be revisedin the light of recentprogressin the datingof mosaics.109It now seems
likely that the building was erectedin the secondquarterof the century.It might thus be the earliest
of the churchesbuilt in Athens as the firsttangiblesymbolof the triumphof Christianityin that city.
In its complexityand sophisticationsurpassingall the basilicasin the city, it seems to reflectinterest
at an imperial level, whether or not the workmen themselveswere imported. (The quality of the
mosaicshardlymeasuresup to Constantinopolitanstandards.)110 The churchsufferedheavily from
fire, probably from the Slavs in 582, and was rebuilt as a three-aisledbasilica,perhapsfor the visit of
Constans II in 662/3. Tradition makesthis the first cathedralin Athens.
Next in chronologicalorder of the known churches in Athens is probably the Basilica on the
Ilissos, occupyinga little island formedby two branchesof the river,southeastof the Olympieion.11
© American School of Classical Studies at Athens

Kitzinger and Kolarik see here a slightly later developmentof some of the motifs occurringin both
For personal use only. License: CC-BY-NC-ND.

the tetraconchand the Ilissos basilica.

The southeasternquarterof the city, well away from the centerof town, was especiallyfavoredas
a location for the early churches.12 In the vicinity of the Basilica on the Ilissos was anotherlarge,
three-aisled church between the periboloswall of the Temple of Olympian Zeus and an elaborate
bath of the 2nd century;this last continuedin use into the 7th century,probablyin conjunctionwith
the church,as a baptistery.113Anothermay be representedby the late constructionover the floor of
the large caldarium(with apsidal end) of a bath under the Church of the Soteira Lykodemou(the
present Russian churchon Philhellene Street)."14
In 1963 the remains of a three-aisle basilica with three apses were uncoveredin the Roman
Agora. Only a small part of the east end is visible; the rest lies under the mosque, the Fetiye
Djami.115From the shoddy character of the masonry and the quantity of re-used material, the
church appears to date from the 7th century.
Of another basilica only the plan remains, published by A. Lenoir as "in Athens"but without
specifying the locality.116 Some churches for whose existence there is no evidence rest on such
widely acceptedtraditionthat they should be mentioned.The most persistentof these traditionsis
one which places a very early churchof St. Philip on the site of the presentchurchof the saint at the
north edge of the Agora."17The sole source of the belief appears to be the Anonymousof Vienna,
referring to the episode in the life of St. Philip in which he convertedand baptized an Ethiopian
eunuch (Acts 8, 26-40). But this incident which, accordingto the Acts, took place on the road
from Jerusalem to Gaza, has here been transferredto Athens, to a vague area northeast of the
109See R. Kolarik, footnote 179 above, p. 44.
110See Krautheimer,Architecture,
pp. 126 and 241 for its resemblanceto S. Lorenzo in Milan and the influence of
Asia Minor on that church.Also, Deichmann, Studien (footnote97 above, p. 27), p. 36.
111First published in full by Sotiriou (footnote69 above, p. 23), pp. 1-31; idem (footnote108 above,p. 72), pp. 208-
210. Some revisionswere made by M. Chatzidakisafter furtherinvestigation,"Remarquessur la basiliquede 'Ilissos,"
Cahiers archeologiques 5, 1951, pp. 61-74 and 6, 1952, p. 192. Brief publicationby J.-P. Sodini, "Mosaiquespaleo-
chretiennesde Grece,"BCH 94, 1970, p. 702.
112 A previoussuggestion (Travlos in E. P.
Blegen, "News Items From Athens,"AJA 50, 1946, pp. 373-374; Frantz,
DOP, pp. 193-194) of clearly definedspheres of influence,with the pagans holding the center of town and the Chris-
tians on the outskirts,must be treatedas a tendency,not a hard and fast demarcation.
113J. Travlos, <<'AvaoKa4LKaLe'pevvaL
irapaTro 'OAv,JLTErLov>,IlpaKrLKa 1949, pp. 26-43; Travlos, PDA, p. 181 and
figs. 238-244.
114 K. S. Pittakis,
'Eb'Apx 1856, pp. 1449-1456, with plan; Travlos, PDA, p. 142.
115AErT 19, 1964, B' 1 (1966), p. 96; 20, 1965, B' 1 (1967), p. 22.
116 Architecture
monastique, Paris 1853-1856, I, p. 248, fig. 160, and republished by Sotiriou (footnote 69 above,
p. 23), pp. 5-6.
117Travlos, IlOAECobo,ULtK,
p. 142; A. Mommsen, Athenae Christianae, Leipzig 1868, p. 105, no. 125.
 74 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

Hephaisteion.1I8It will be noted that there is no mention of a church in this account.A churchof
the 16th centuryon the site of the presentchurchwas seen in a dilapidatedconditionand drawn by
Couchaudin 1842.119 By 1866, when Mommsen saw it, it had been renovated.The possibilitythat
an earlier churchoccupiedthe site was eliminatedin 1963, when the presentchurchwas extensive-
ly remodeledand deep digging under the foundationsrevealedno earlier remains.
The tradition of a church on the north slope of the Areopagus in honor of St. Dionysios the
Areopagite is less nebulous. The church whose remains are now visible, however, dates from the
16th century,and intensiveexcavationin and aroundit discoveredno trace of a predecessor.120
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

DECLINE (A.D.450-529)
REMOVAL OF WORKS OF ART TO CONSTANTINOPLE

Since the time of Sulla the art treasuresof Athens had been a target for despoilersand collectors.
The wholesale removalof statuesto embellishConstantine'snew capitalwas widely commentedon
from Late Antiquity through the Byzantine period. Lists occur of famous statues removed and
importantcities denuded,but rarely are the two in conjunction.12
There can be no doubt that vast numbers of statues were carriedoff to Constantinople,where
they adorned palaces, fora, baths, the Hippodrome, etc., but attributionsto Athens as a place of
origin are rare, as indeedthey are for most cities, and hardevidenceis almostnon-existent.In a few
cases the only certain fact is that the informationis wrong.

The Obelisk of Theodosius

To begin with the certaintiesof misinformation:"themonolithwhich standsthere was brokenoff
from the one standing in the Hippodrome. It was brought from Athens by Proclus in the time of
Theodosius the Younger."'22This is wrong on three counts:the date, the conditionof the obelisk,
which is intact, and the source. The only element of truth is that Proclus erected the obelisk, as
stated in the inscription.But the contextshows that it was Theodosius I, not II, who was responsi-
ble,'23and the actual date is providedby the Chronicleof Count Marcellinus: "ValentinianoAug.
IV. et Neoterio coss .. . obeliscus in circo positus est" (A.D. 390).124 Athens as a place of origin is
excluded by the EmperorJulian's letter to the Alexandriansbegging them to expedite the delivery
of the obelisk which Constantius II had greatly desired for Constantinople,even to the point of

11 The relevantpassage (3 a) reads:KaTa apKTOV e TOVTOV (sc. "Theseum")v'7rijpXev- 7rpwrn ayopa rTs 7rXoAEw,
el d v o aWoo-roAog DlAiL7r7
ro ypaTJo.yaTea
V eftOove. The complete text of the Vienna Anonymous with annotation
may be found in Wachsmuth, I, pp. 731-741.
119 A. Couchaud, Choix d'eglises byzantines en Grece, Paris 1842, pl. 2. For the date, see Travlos and Frantz, "St.
Dionysios,"pp. 174-179.
120
Travlos and Frantz, "St. Dionysios,"p. 164.
121
The fullest and most systematic,though not necessarilythe most accurate,accountis that of the pseudo-Kodinos,a
corpus of tracts relating to the antiquities of Constantinopleput togetherca. 995. This forms Part II of the three-part
HarpLa Kwvo-TravTLvov7rToAXEws (= Patria), in Scriptores originum Constantinopolitanarum, T. Preger, ed., Leipzig
1901. The complicatedquestion of the interrelationshipof the Patria and other texts is the subjectof a forthcoming
work by Averil Cameronet al. I am indebtedto ProfessorCameronfor a preliminaryview of the material.
122 HiEp rTOV?rparL'yiov . . . Mov6AOov o AVraeoV EKELcaoKAata BVy TOV aT.EVOVELSTO ITKOV
8e aro 'A07vasg 7TapaIlpOKXov7raTpLKLov
K?E TOV&LKpoV
ev TOLsXPOVOLs Oeo8oai'ov (Patria II, p. 183).
123 R.
Janin, Constantinople byzantine, 2nd ed., Paris 1964, pp. 189-191.
124PL LI, col. 919.
 DECLINE (A.D.450-529) 75

having a boat built for its transport.125Julian's plea, too, proved unavailing, and the projectwas
accomplishedonly in 390. It is hardly likely that in the intervening28 years the obelisk's course
would have been interruptedby shipwreckand that the obelisk itself would have lain in Athens (or
Piraeus) until it was rescuedby Proclus. More probablythe informationis simply erroneous.126
"Aboutthe Elephants"
This is the heading in the Patria for a group of statues of elephants observedat the Porta Aurea
by a number of travelers, but the Patria alone give "the Temple of Ares at Athens"as their pro-
venience.127Cedrenusleaves open the questionwhether the elephantshe saw (or reportedon) were
the same as those broughtby Theodosius or only "like".128 In the 9th century(880-890?) the Arab
© American School of Classical Studies at Athens

travelerHarun b. Jahja saw five elephantson the gate and a man standingbesidethem, holdingthe
For personal use only. License: CC-BY-NC-ND.

reins.129By the beginning of the 13th centurythe numberof elephantshad dwindledto two, as seen
by Robert de Clari. He alone mentionsthat the statues were made of bronze.130
The difficultyof acceptingAthens as the provenienceof the elephants lies in the mention of the
Temple of Ares. This had been in ruins since the Herulian invasion of 267, and in the time of
Theodosius II it would have been completelyburied under destructiondebrisand accumulationof
silt. Whether its identity would still have been known after a century and a half of desolation is
questionable.131If such a remarkablegroup had actually stoodthere in Pausanias'time one would
have expected him to mention it. One possible referenceto it might be found in the "statuesof
Egyptian kings"which "stoodbeforethe entranceof the theaterwhich they call the Music Hall," all
bearing the name of Ptolemy but each with his own surname:Philometor,Philadelphos,Soter.132
Elephants would be appropriatein such a context,but the circumstancesare againstthe probability
of any connection.
An element of credibilityenters with the statue of Constantinethat stood on top of the porphyry
column in the Forum of Constantine (the "BurntColumn").The three existing descriptionspro-
vide a consistentimage, using much of the same phraseology,for the basic information:the column
stood in the open space of the forum (7repL,Ascrrosk),and the statue was erectedby Constantine'in
his own name" (rw T'O 'vo.Aaavrov), suggesting that it was re-used. In fact, the Patria state that it
125 'O \eXovcEtLva7rap vilv aKOVW ELSv'Uos LKavovE7rrTp.LEvov,e7rTrT?I 1Ol0ovoZo7rwcpaAAo TL TCoVaTL-
AXLiOVOV
'E' ro'vrov evavinry1aTr
.OTCTwv Eppli4LEVOV. Kos
oao .aKaplir] KavoiTaVTo, cs E7aT4wv avrov eL rTv
'A74v 7rarpLbaKwo-Tavrivov TroXLV (Ep. 59 [58] 443 B).
126 Cf. G. Bruns: "Der
Ursprung der Nachricht, dass der Obelisk aus Athen gekommensei, ist nicht mehr festzu-
stellen, die Uberlieferung halt sich aber" (Der Obelisk und seine Basis auf dem Hippodrome zu Konstantinopel,

for recent bibliographyon the obelisk.

127
llepl TCV?X?favrTcv. A' o r)AaL TWV
Tria
wopra^ e4eavTwv Xpvo-tdar ?KariLv (KTOVvaov TOV"ApecoS a7o
'A6rOvas,7rapa 3
Oo0800V TOV
roI.Kpov (de signis, p. 27; Patria II, p. 182).
128
"OTL OLEAEavTres ol EV -J'VO7lropTfv
TnL O'p.OIOiLOLL
ELOLV raAaL
Oeo7o0toa e7Toas
e21 rvT7r 7roLv e'ijXaoa-ev
(Cedrenus,323 D [CSHB, I, p. 567]).
129
J. Marquart, Osteuropaischeund ostasiatischeStreifziige, Leipzig 1903, pp. 206, 215.
130
C. Hopf, Chroniquesgreco-romainesinedites ou peu connues, Berlin 1873, p. 69. Cf. also Janin (footnote 123
above, p. 74), p. 271. If there is any kernel of truth in the tradition, the elephants were probably taken from some
triumphal arch. See RE, s.v. Triumphbogen,col. 475, line 37 (Kahler).
131 H. A. Thompson points out that the elephantsmight have cometo light in the building operationsfor the Palace of

the Giants over the Odeion (Thompson, "Odeion,"p. 136, note 1). He also findssome evidencethat the temple had been
repairedafter the Herulian raid and that it was still standing 150 years later (footnote9 below, p. 97).
132 i.8.6. J. P. Binder suggests a possible connectionwith the "Battleof the
Elephants"in 275 B.C.when AntiochusI
Soter, with the help of a small herd of elephants, was victoriousover the Galatians (Lucian, Zeuxis, 8-12).
 76 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

was a statue of Apollo. Cedrenusalone says that it was "thework of Pheidiasand was broughtfrom
Athens."Accordingto the Patria and Cedrenusthe column itself was broughtby Constantinefrom
Rome. 133
By combining the statement in the Patria that the statue originally representedApollo and
Cedrenus'that it was a work of Pheidias and was broughtfrom Athens, we might cautiouslyasso-
ciate it with the only known bronze statue of Apollo by Pheidias, the Apollo Parnopius, which,
accordingto Pausanias,134was "onthe Acropolis,opposite(7repav)the Parthenon,and said to be by
Pheidias."This has been taken as possibly the same statue as that to which John Tzetzes referred
(ca. A.D. 1170): avrov re rov avO)jXLov'AIroAAcovaeKELVOV.This line is followed by an enumer-
© American School of Classical Studies at Athens

ation of several other works of Pheidias and the remark that "somehave been lost, some melted
down, some victims of time,"and then the lines:
For personal use only. License: CC-BY-NC-ND.

\ A , ,, ei
rwa o ELO'Kat rTravrTarTWL7rTrLKKaL<opW
KcEfaA) '8 'ArXoAAtvoso
avrw rTW IaAarTw.135
In short, identificationof Constantine'sstatue with the Apollo Parnopiusof Pheidiasis possiblebut
by no means certain.

The Athena Promachos

In 1203, on the eve of the Fourth Crusade, a mob of drunken rioters in Constantinoplepulled
down a bronzestatue of Athena in the Forum of Constantineand brokeit up into many pieces. The
scene was vividly describedby Niketas Choniates,togetherwith a meticulouslydetaileddescription
of the statue as he must have seen it almost daily for many years.136The agreementof Niketas'
observationswith Pausanias'all too brief accountof Pheidias'bronzestatue of Athena Promachos,
supplementedby coins and lamps of the 2nd and 3rd centuries, forms the basis of the generally
accepted belief that this indeed was Pheidias' Promachos,removed from the Acropolis at some
undeterminedtime in Late Antiquity and set up on a column in the Forum of Constantine at
Constantinople. The attribution to Pheidias was reinforcedby a commentaryon an oration of
Aelius Aristeides (Orat., 50) by Bishop Arethas of Caesarea (9th-10th centuries). On Aristeides'
statementthat Pheidias' masterpieceswere the Zeus at Olympia and three statues of Athena, "one
of ivory,one of bronze,and the Lemnian,"Arethascomments"I believethat this [i.e., the bronze]is
the one in the Forum of Constantine at the entrance to the senate, with Thetis opposite, on the
right" (boKel MOLav'r7 EO-TVL pp
(veVr4o Kc0varravrLvov avaKEqLMEVt/
KaL TOLS 7Tpo'rvAa1sL
... ).137 Niketas' account corresponds also with the very brief description by Cedrenus a cen-
tury earlier, to the effect that the Lindian (sic) Athena stood at the west end of the Forum of
133 KaL o 7ropfvpovs KaL 7TepL/,3AErTosKL'WV, <' OVtrcpLbpvrOatLKwvoravrTiov
Hesychius (Patria I, p. 17): ...

OpSWe?V8IK7v7 7AXLov7TpoXAajltrovraT roAtrat.

Patria II, p. 174: KaL 7Trept3AETrros
btLp&V OVtSo KLCOVKat 7) OT?2A7)rot 'A7oAAXvos,
avT'lv 'HAtov E'oT7faev o
KwvotravTrtvoS ElSrTO voAa avrov,
ME7yas o-~rasev Tfr Kf)aXA Xovs (K
TlW'V Xpo-rro
rovTO biLK7vaxrltvC, wg "HAL&o
TroT 7roALTratL
eKAap7Trv.
Cedrenus,295 D-296 A (CSHB, I, p. 518): ... KaLev avrW T) 4f6pq KlOVa
K'L .OVO\iLOO oAXo7op4vpov orT7)0e?V, (K
Prsaoayvrov vw avbopavTa er' oVOTIaTL avTro, ev w y.ypa.rrat
<,KwvoTavTiv0os.
E'Aa/uev 'r7A\lovbLK7V,os Iv "'pyov ,uEv (ccE&ov, r7''X) be cf 'AOr7v&wv.
134 1.24.8.
135
Tzetzes, Chil. VIII, lines 333 and 338-339. Cf. J. Overbeck,Antike Schrnftquellen,
Leipzig 1868, p. 139, no. 17.
136 Niketas
Choniates, Historia (CSHB, pp. 738-740).
137 For details of the
argument,with additionalevidence,see R. J. H. Jenkins, "The BronzeAthena at Byzantium,"
JHS 67, 1947, pp. 31-33. Arethas is quoted by S. Kougeas,AaoypaL'{a4, 1913, p. 240.
 DECLINE (A.D.450-529) 77

Constantine, wearing a helmet, with the Gorgoneion and snakes around her neck.138The only
discrepancyis Cedrenus'identificationof the statue with the Athena of Lindos. But this seemsto be
either sheer absentmindednessor else confusionamong a number of notes. On the precedingpage
he has alreadydescribedthe Lindian Athena as having stood,not in the Forum of Constantine,but
in the Palace of Lausos (near the Hippodrome):a much smallerstatue, only four cubitshigh and of
green marble, the work of Skyllis and Dipoinos, and the gift of Sesostris,Tyrant of Egypt, to Kleo-
boulos, Tyrant of Lindos.139The imperfect'l'o-rarois used advisedly,becausethe Lindian Athena
appears later (35i C-D [CSHB, I, p. 616]) as amongthe many statueslost in the fire of 475/6 which
destroyedthe whole quarterof the city in which the Lauseion stood.The descriptionhere is briefer,
and the donor is Amasis instead of Sesostris. Butrecipient the is the same, and both passages un-
© American School of Classical Studies at Athens

recipient
questionably refer to the same statue. Not so the third, which correspondsin place and dress to
For personal use only. License: CC-BY-NC-ND.

Niketas' statue and, along with its pendant, Amphitrite, at the other end of the Forum, was still
standing ('lo-ravraL).The suggestionthat "Lindian"is an author'sor copyist'sslip for "Lemnian"
must be dismissedbecause of the last sentence,referringto Amphitrite:"thistoo was broughtfrom
Rhodes,"which would make sense only in relation to another statue from Rhodes, i.e., the "Lin-
dian". But her appearancehere was simply a transferencefrom another note against the author's
better knowledge.
However and wheneverthe Promachosgot to Constantinople,it was certainlynot taken by Con-
stantine, because it was seen by St. Jerome when he visited Athens in 372.140 On the Acropolis,
"iuxta simulacrum Minervae,"he was shown a great bronze sphere used by athletes to test their
weight-lifting powers. And still later, ca. 410, the sophistApronianuserecteda statue of Herculius,
Prefect of Illyricum, "besidePallas the Promachos." 141
Bishop Arethas'statementthat he believedPheidias'bronze statue of Athena to be the one stand-
ing in the Forum of Constantineis strengthenedby a miniatureof the 11th centurybut which goes
back to a 9th-centuryprototype,in which a statue of Athena correspondsclosely with the written
descriptions.142 The outer limits for the removalof the Promachosmay thereforebe taken as ca. 410
and the middle of the 9th century.After Constantine,Theodosius II is mentionedmost frequently
as a remover of works of art. Justinian also figures in this role. Soon after his reign the Slavic
invasion put an end to Athens as a fertile groundfor depredation,and so the most likely time for the
removal of the Promachos is during the last three quarters of the 5th century or the first three
quartersof the 6th. But it may be possibleto be moreprecise.It has been suggestedthat the occasion
for setting up the statue "at the entranceto the senate-house"was the rebuildingof the senate not
long after it was destroyedby fire in 462. This might place the removal of the Promachosfrom
Athens at about 465-470.143

L38 TavTral e TpOs T7V TOV of po v 7rXaTELav yaaX.ara bvo, 7rpos .ev bvL To is ALwbas 'A01 vas, Kpavos E'Xov
To
KaLTOr opyoveLov rTpas KaL oep1?s TrpL TOTOVrpaX ov E.'7TE7rAy.uevovs (ovr'Tco yap To eL'8wov avrT7s ol 7raXAaLo
'
loroTpovv), 7rpos b8eT77VavaToXjv 'A TirpL'T'.... '1xjU be avrinT
a?ro 'Pobov (Cearenus, 322 D-323 A [CSHB, I,
p. 565]).
139
322 B (CSHB, I, p. 564): loTaro bc Kal rToayaA,a T7rs Avblias 'A7rvasTrerpa7rxv
Cedrenus, (K Xl0ov
opapaybov, E'pyov KaL
?KvAXXLbos AL7TOL'VOVTQW ayaX.aroTVpyWV, o7Tep 7TOTEbipov 7TEfjAf VE-E(T-fTplS Alyvt'rTov
Tvpavvoo KAeoAovAX TWATvbL rTpavvvw.
140
Commentaryon Zachariah,II111.12.3.For the date of Jerome's visit, cf. Kelly (footnote222 above, p. 52), p. 37.
See above, p. 64.
141

142
Jenkins, loc. cit. (footnote 137 above, p. 76).
143
Saffrey and Westerink,p. xxiii, note 1. The statue is confusedhere with the Parthenos,which is of courseimpos-
sible becauseof the material. For the source of this commonconfusion,see Frantz, "Julian,"p. 401, note 54.
 78 IV. PROSPERITY, DECLINE, AND DISASTER (A.D. 396-582)

THE VANDALSIN ATHENS

LiteraryEvidence
Apart from the activityof the schoolsfew historicalfacts are certainfor the last two thirds of the
5th century. The last known proconsulof Achaea is Hesychius (A.D. 435), and the rest of the cen-
tury "disappearsinto murkytwilight."144But a few elements,archaeologicaland epigraphical,shed
some rays of light. Althougha numberof raidswere madeon the Greekislandsand mainlandin the
courseof the 5th century,145no literaryevidencehas been adducedthat Athens itself suffered.The
chroniclersof the importantand extensive Vandal campaignsagainst Spain and Africa were spar-
ing in their references to lesser activities which could only be regarded as incidental. A single
© American School of Classical Studies at Athens

sentence of Procopius, however, offers the possibility that Athens might not have escapedutterly:
rs rTeaXAs ' EXaos Ta TrAfWra KaLLo'a-z
For personal use only. License: CC-BY-NC-ND.

IIAXvpLovs ovv EX71fCTOKaLrTrs T HeAoToorVVfoTov

aVrj vfjooETlOKLvTa.146 This admittedlyinconclusivepassage, taken in conjunctionwith some
hitherto unexplained evidenceof destructionin the third quarter of the 5th century in the Agora,
raises the possibilitythat a Vandal raid should not be discounted.

ArchaeologicalEvidence
In brief, the evidenceis as follows:over and in frontof someof the buildingsof the west side of the
Agora was found an accumulationof ash and debrisdated by coins and potteryto the period about
460-485. Much of this overlaythe north-south road,which went out of use fromthat time until the
Middle Byzantine period. A similar deposit was concentratedover the southeast corner of the
Palace of the Giants-a deposit indicatinglocal damage predatingthe general destructionof that
building by a century.This destruction,whateverits cause, was followed almost immediatelyby a
suddenresurgenceof building activity,contrastingwith the apparentlethargyof the precedinghalf
century. Part of these operations was utilitarian: an aqueduct and at least three (more probably
four) water mills running from south to north along the PanathenaicWay, togetherwith the wall
along the east side of the street; also perhaps the rebuilding of the second storey of the southeast
cornerof the Palace.147Other buildingswere moreformal:a pair of stoas borderingthe Streetof the
Libraryof Hadrian (below). The southernof the two stoas overlaya considerableaccumulationof
potteryof the same date as in the depositsjust mentioned.148
The Square Building mentionedabove in connectionwith the Palace of the Giants (p. 66) may
have been one of the casualtiesof the attack.Another may have been the Stoa Poikile. As we have
seen (p. 55), it probablysuffereddamagefromAlaric'sforcesin 396, after which time the paintings
were removedby the proconsul.But the building continuedto be cherished,even in a ruinous state,
until the middle of the 5th century, when it was completely destroyedand some of its members,
along with pieces from other buildings, were built into a wall alongside the above-mentioned
aqueduct.149

144
Groag, Achaia,p. 76 (quoting Seeck).
E.g., the Huns advancingas far as Thermopylaein 447 (Bury [footnote223 above,p. 52], p. 275, quoting Priscus,
145

frag. 3 Dindorf) and Vandals invadingthe Ionian islandsand the west coastof the Peloponnesein 467 and again in 477
(Procopius,bell. [Vand.]111.5.23;II11.22.i6; cf. also Groag, Achaia,p. 78 and Bon [footnote70 above,p. 68], p. 14, with
otherreferences).
146 Bell.
(Vand.) III.5.23. Groag gives the date as 467.
147 Below, p. 106.
148
Many fragmentsof the ceiling coffersand the original inner colonnadeof the Parthenonwere used in the founda-
tions of this stoa. All were badly broken, but their surfaceswere remarkablyfresh (see W. B. Dinsmoor, Jr., "New
Fragmentsof the Parthenonin the Athenian Agora,"Hesperia 43, 1974, pp. 132-155 and Frantz, "Julian,"p. 395).
149Below, p. 81.
 DECLINE (A.D.450-529) 79

To remove this hypothesis of a Vandal invasion in 467 from the realm of pure speculation,we
may look to an inscription from Megara recounting the benefactionsof a certain Diogenes, de-
scribed as a KoA7)s/ and benefactor "of the Greek cities."150Groag argues that Diogenes was not
Proconsulof Achaea and even maintainsthat he held no officialposition in Greecebut was a high-
ranking military commanderto be identifiedwith a Count Diogenes who was in the serviceof the
Emperor Zeno in 476 and had bestowed similar favors elsewhere.151Groag further places his
service in Greece either in the reign of Leo I (A.D. 457-474) or in the first decadeof Zeno's reign
(A.D. 474-491) and associateshis rebuildingin Megara with destructioncaused by the Vandals in
476. If the Vandals reachedMegara, on the main road from the west leading to Athens, they might
© American School of Classical Studies at Athens

reasonablybe expectedto have gone the rest of the way.152A parallel situationto Diogenes' contri-
butions may be seen in the case of Herculius, who gave the Megarians an aqueductafter, and most
For personal use only. License: CC-BY-NC-ND.

probablybecause of, Alaric's invasion.153

THE STREETOFTHELIBRARYOFHADRIAN
As shown above,the streetalong the easternhalf of the northernborderof the Agora had been put
out of use by the constructionof the large basilica in the time of Hadrian. Destroyed in 267 and
apparentlyabandonedto predatorsseekingbuilding material,the basilica was now finally stripped
to the foundations,and the road was once more opened up. A charactersimilar to that of the Pan-
athenaicWay was given to it by the constructionof anotherstoa, or more likely a pair of stoas. The
foundationsof the back wall and stylobateof a colonnadedbuilding along the south side of the street
are preservedat the northernedge of the presentexcavations;the counterpartpresumablylies in the
yet unexcavatedarea.154The southern stoa was built over the south foundationsof the basilica. A
plunderedtrench directlyunder the stoa yielded many pieces of marblerevetmentfrom the basilica
along with a considerableamount of potteryof the middle or third quarterof the 5th century.The
line of the street leads to the propylonof the Libraryof Hadrian, which had been restoredearlierin
the century.155It may be consideredthereforethat both the street and its stoas were conceivedas a
monumentalapproachto the latter building.

THE BEGINNINGOF INDUSTRIALIZATION

By this time the characterof the Agora was changing. Competitionfrom small industries was
threateningthe formal buildings. Perhaps the first responsewas the protectionof the area around
the Palace of the Giants. From the south corners of the great court, walls were carried south-
eastward and southwestwardin such a way as to enclose the whole south section of the complex,
together with a considerableamount of the surroundingground, especially to the west. Both east
and west wings can be traced as far as the two ends of the Middle Stoa, where the wings were
carriedon its foundations.They are then lost sight of, but it is probablethat they were continued
southward and that the back wall of South Stoa II, which was still standing, formed the south
boundaryof the enclosure.The walls were not an integral part of the Palace court, but how much
later they were added is impossibleto say. The extra space free of buildings in the western part of
150IG VII, 26 (= SEG XIV, 379). Specifically mentioned are the repair of the fortificationsof
Megara and the
renewal of a bath for which he contributed2200 square feet of marble.
151 Groag, Achaia, pp.
77-78.
152
Procopius (bell. [Vand.] 111.I.17), in a totally irrelevantcontext, defines a day'sjourney as "two hundredand ten
stades, or as far as from Athens to Megara."
153IG VII, 93. See footnote 52 above, p. 65.
154Hesperia 40, 1971, pp. 264-265 (Shear, Jr.).
155Above, p. 63.
 80 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

the enclosurewas perhapsto allow for a gardenor grove. It might also have been to protectagainst
the industrieswhich were springingup near by, e.g. metalworkingestablishments.

Bronze Foundries
The south side of the Agora had been an industrialarea fromthe time of the Sullan destructionof
86 B.C. to the early 2nd centuryafter Christ, when the partial reconstructionof severalof the major
buildingsrestoredits more formalcharacter.156 But immediatelyafter the Herulian invasiona new
incentivewas providedby the bronze statues thrown down during the raid:an invitationto metal-
workersto turn them into commercialassets by melting them into pots and pans, trinketsand small
ornaments,or utilitarianobjects.In the area immediatelysouth of the Palaceor even underits walls
© American School of Classical Studies at Athens

were found one foundryand two dumps of furnacewaste, all dating from the late 3rd into the late
For personal use only. License: CC-BY-NC-ND.

4th centuryand put out of use by the new buildingoperations.157 The foundrywas situatedin what
was later to becomethe angle betweenthe west courtof the Palace and the adjacentbath,just north
of the Great Drain (P1. 68:a). Several molds were found in situ, buried to their rims in earth or
sand.158The arrangementsare closely paralleled in a modern bronze-castingestablishment in
Rome (PI. 69:a).159The crude and unsymmetricalconstructionof the firing chamberis perhaps
explained by its temporarycharacter,if one may press the modernparallel. The similar structures
in Rome, except for their underpinnings,were used for only one firing, after which they were
dismantledand re-erectedon the same base. Judging from the many fragmentsof molds found in
the area of the Agora foundry, this one producedonly small objects,even pins. Waste from the
foundriesof the 3rd to 4th centurieswas founddumpedat the south edge of the Law Court160at the
southwestcornerof the Palace of the Giants, also in a well underone of the west roomsof House A
on the Areopagus.161
The only survivingfoundryof post-Herulian times designedfor the casting of large statuesover-
lay the north wall of the Mint, outsidethe Palace precinct(PI. 68:b).162The remainsconsistof two
brick piers and some oval holes presumablyfor buryingthe molds for a statue of life size or larger.
A consistently6th-centuryfill over the remainsof the establishmentgives a terminusante quem for
its life. Being outsidethe limits of the Palace precinctit was very likely allowed to remainas long as
there was a demandfor bronze statues, i.e., until the demiseof the Palace itself.

Watermills
The powerful influence of the Palace of the Giants, reinforcedby its sturdy concreteprecinct
walls, providedan effective hedge against encroachmentof small and untidy industries in its vi-
cinity during most of its life. The earliest of the post-Herulian establishmentswere related to the
general cleaningup of the whole Agora area and were removedto make way for the constructionof
the Palace. From then until the 6th centurynothing except a series of watermillsfor grindingflour
was allowed to encroachon its precinct,and these, with the aqueduct(PI. 6) which poweredthem,
clung discreetlyto the opposite side of the PanathenaicWay. Three mills were strung along the
entire east side of the Agora proper to its northernedge. Of the South Mill only the wheel race is
156Footnote 20
above,p. 3.
157
These and all the other foundriesin the Agora have been publishedby Carol Mattusch (Mattusch);also C. Mat-
tusch, Bronze Workersin the AthenianAgora (Agora Picture Book No. 20), Princeton 1982.
158 Mattusch M, p. 370.
159 That of the Fratelli
Nicci, 22 Via San Michele. I owe these photographsto the hospitalityof the Nicci brothers,
who allowed me to spend three days in 1957 photographingthe sequenceof stages of bronze-casting.
160
Mattusch K, p. 368. There is reasonto believe that some minting may have taken place here in Late Antiquity.
161
Mattusch L, pp. 369-370 and Y, p. 376.
162Mattusch N, p. 373.
 DECLINE (A.D.450-529) 81

preserved,set into the ruined pronaos of the SoutheastTemple (PI. 70). Nothing of the mill room
was identifiable,but some fragmentarywalls spread over the temple seem to be contemporaryand
are perhaps to be associatedwith the mill.
The Central Mill is a rare survivingexample of a type whose principleswere laid down by Vitru-
vius and which must have been commonin antiquity.163The preservedremains consistof the mill
race,wheel race (PI. 71:c),and tail raceas well as the mill room,which was built into an obtuseangle
formedby the Post-Herulian Wall and the south face of Tower W 4 (Pls. 14:d,71:a). The overshot
wheel, with a diameterof 3.20 m. as restoredfromthe scratchesmadein the wall due to faulty align-
ment, drovetwo gear wheels, one vertical,the other horizontal,which in turn drovethe mill itself.
The water was carriedaway by the tail race set below groundlevel along the base of Tower W 4 and
© American School of Classical Studies at Athens

thence northwardby the mill aqueduct.A survivalof the type, still in operation,can be seen on the
For personal use only. License: CC-BY-NC-ND.

island of Thasos (PI. 71:b). On the basis of coins, pottery,and lamps found in the gravelly fill into
which the mill roomand wheel racewere set, it was possibleto datethe constructionof the mill to the
third quarterof the 5th century,probablyin the reign of Leo I (A.D. 457-474).164
The wheel race and the mill room of the North Mill must have been in the very middle of the
railway cutting, for the northernpart of the aqueduct,up to the point where it was cut off by the
railway line, was carriedon arches (PI. 69:b). It then reappearsas a tail race at a much lower level
immediatelynorth of the railroad,having made a sharp turn westward.A fourthmill was probably
turned by a branchleading westward from the main aqueductclose to the North Mill. This branch
survivesonly as a curvedwall running acrossthe foundationsof the Round Building (p. 60, above).
If this was the case the wall would have carriedarches to providethe necessaryheight. This may
have been a later addition,or it was perhaps intendedto run alternatelywith the North Mill.
It is in this connectionthat accountshould be taken of the Late Roman wall in which fragments
of the Stoa Poikile were found.165This is a long wall which abuttedagainstthe northwestcornerof
Tower W 5 of the Post-Herulian Wall (Pls. 5, 6). From that point it ran northwestward,almost
parallel with the PanathenaicWay, and ended at, or just before, the west end of the south wall of
the stoa on the Street of the Libraryof Hadrian. A fragmentof a stampedplate (PI. 72:d)166found
embeddedin the concretemasonryof the wall could be dated to the middle or third quarterof the
5th century. For part of the course of the wall the aqueduct was built against the west face;
elsewhere a narrow intervalwas left between them (PI. 69:b).
Although the smoothly plasteredwest face of the wall showed that it predatedthe aqueduct,the
difference in date may have been only a matter of weeks rather than of years, since the ceramic
evidencefrom the wall pointedto a date of about the third quarterof the 5th century.
The purpose of the wall is not clear. Its location on the opposite side of the PanathenaicWay
removes it from direct relationshipwith the Palace of the Giants and relates it rather to the long
slim triangle between the Stoa of Attalos and the PanathenaicWay. The north end of the triangle
was occupied by a large sprawling building, much of which has disappeared.It consistedof two

163
Vitruvius, X.4.I, 5.1 and 2. The Central Mill was excavated and published by the late Arthur W. Parsons, "A
Roman Water Mill in the Athenian Agora,"Hesperia 5, 1936, pp. 70-90. Readers are referredto this article for a
detailed descriptionof the mill and also for a reconstructionof its workings and assessmentof its fidelity to Vitruvian
principles. For a recent analysis see R. J. Spain, "The Roman Watermill in the Athenian Agora:a New View of the
Evidence,"Hesperia 56, 1987, pp. 335-353. See, in general,T. S. Reynolds,Strongerthan a HundredMen, Baltimore
1983, esp. pp. 36-38.
164
Parsons, op. cit., p. 88. Many coins were found under the mill, chiefly of the late 4th and early 5th centuries.The
latest were four of Marcian (A.D.455-457) and three of Leo I.
165Above, p. 55.
166 p 20037.
 82 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

parts or more:an eastern,a western,and possiblya northern,loosely connectedarounda courtyard.

Its dimensionsare ca. 39 metersfrom east to west and 16 meters,as preserved,fromnorth to south.
The northernside was cut off by the railroad,makingit impossibleto determineits full extent with
certainty,but it is probablethat little more than the north wall is missing and that the house would
have ended at the Streetof the Libraryof Hadrian, past the north end of the Stoa of Attalos, which
at that periodwas Tower W 7 of the Post-Herulian Wall.
The plan of the house is irregular (PI. 6). There are severaljogs in the outside walls, and the
roomsare of differentshapes and sizes. The westernwing consistedof three pairs of roomsto which
a small, more or less triangular room was later added to the south. A rectangularwater channel
© American School of Classical Studies at Athens

outside connectedwith round pipes inside the room through a hole in the wall, bridgedby round
stones. From these waterworksand the presenceof animal bonesthe annex couldbe presumedto be
For personal use only. License: CC-BY-NC-ND.

a kitchen.
Of the east wing, which was much less well preservedthan the west, we can be sure of only three
rooms along the north side; their walls, too, were removedby the railway cutting. At the south the
two wings were connectedby a wall. A gap of about 12 metersbetween the wings on the north side
would have providedeasy accessto the building.
Materials used in the buildingvariedfrom field stonesto re-usedblocksto freshlyquarriedstone.
Except in the "kitchen"they were laid in lime mortar,not very firm. Some at least of the walls were
plasteredinside and out with a coarsegray plaster, striated.In one room were parts of three tiles of
standarddimensions(ca. 0.54 m. square by 0.03 m. thick).
The plan of the building, large for a private house, would have providedliving quarters for a
considerablenumberof people. It might perhapsbe relatedin some way to the whole systemof the
water mills.

DISINTEGRATION AND DISASTER (A.D.529-582)

THE REIGN OF JUSTINIAN

Apart from the closing of the schools,'67Justinian's influence on Athens is obscure. The testi-
mony of Procopiusis contradictory.On the one hand, he statesthat Justinian rebuiltthe walls of all
the cities of Greece below Thermopylae, including those of Athens.168On the other, he charges
Justinian with confirmingthe unpopular measuresof the LogotheteAlexanderwhich resulted in
the fact that "in all Greece, and not least in Athens itself, no public building was restorednor could
any needful thing be done."'69The first of these statements,that the walls were rebuilt, is corrobo-
rated by archaeologicalevidence.For the second,there is no way of knowingto how long a periodit
refers.170Nothing has yet been discoveredin the Agorato suggestthat there was any constructionof
consequenceduring his reign; but it must be rememberedthat of that period little of the center of
the city has been uncovered;also that, failing numismaticor ceramicevidence,there is little to dis-
tinguish ordinarybuilding techniquesof the 4th, 5th, and 6th centuries.
Procopius' statementthat Justinian rebuilt the walls of Athens, which had "sufferedfrom the
long passage of time, while no man in the whole world took thought for them,"'17has generally
been taken to mean that the outer circuithad passedcompletelyout of use by the 6th century. 72But
167 On this controversial
subject,see below pp. 90-92.
aed. IV.2.23-24.
168 De
169Anecdota
xxvI.33 (trans. Dewing).
170Cf. Frantz, DOP, p. 197.
171 De aed.
IV.2.24 (trans. Downey).
172Procopius'phraseologyis reminiscentof Zosimus' accountof the activitiesof Valerian (1.29.3, above,p. 1), when
the city walls had fallen into neglect since the time of Sulla. Could Procopius, who could not have observedall of
 DISINTEGRATION AND DISASTER (A.D. 529-582) 83

evidencefrom chance discoveriesat various points in the line of the wall shows that for most of its
circuitthe "Wallof Valerian"is actually a rebuildingof the ThemistocleanWall, and that although
it was badly damaged,certainlyby the Heruli and probablyalso by Alaric's forces, it remainedin
use until the time of Justinian, who rebuilt and reinforcedit drastically.173

NEW CONSTRUCTION
Private Houses
As has been noted above,most of the privatehouses in use in the 4th and 5th centuriesdate from
pre-Herulian times, even though they underwent considerablechanges. In the Agora at least,
© American School of Classical Studies at Athens

houses of totally new constructiondate only from the 6th centuryand later. The most importantof
these, if indeedit is a privatehouse, was a large structurebuilt againstthe southeastquadrantof the
For personal use only. License: CC-BY-NC-ND.

Tholos.174By this time te Tholos itself was in ruins, but a survivingstretchof wall providedan
anchor for three walls leading off from the main body of the house and enclosing two courtyards
(Pls. 6, 73:d). The house measured 16 x 21 m. overall and consistedof eight rooms in additionto
two courtyards,ranging in size from ca. 2 x 3 m. (B) to ca. 4 x 10 m. (A). The interior arrange-
ments were irregular,but the foundationsfor the frontform a continuousline along the west forkof
the ancient road. The northeast corner overlay the Great Drain, which was by now almost com-
pletely clogged; but it still had enough space to carry away the water from two small terracotta
drains leading out of the cornerroom. In seven of the roomspart of the floor was found in place.
The rubblewalls containedmany ancientfragmentsand were bondedwith a stronggray mortar.
Their surfaces were finished with two coats of plaster:a backing layer with a striated surface, of
which a great deal was found in place, and a thin finishingcoat, most of which had crumbledaway.
The reveals of at least one doorwaywere constructedof bricks.
The date of constructionseems well established,by coins and pottery found under the original
floor, at ca. 500 or very soon after.175It is thereforeruled out as part of the building programof the
precedingcentury, and its purpose, whether as a governmentbuilding or a private house of more
than averageproportions,remains obscure.
Toward the end of the 6th century the house suffered heavy damage from fire. Eleven coins of
Justin II (A.D. 565-578) and one of Tiberius II (A.D.578-582) found among the burned debris
probablyfix the Slavic attackof the 580's as the cause of the destruction;there was also a fragment
of a plate (PI. 76:f) datedby J. H. Hayes 570-600. The house was later rehabilitated,and the floor
levels were substantiallyraised.
With a similar history but far less imposing in characterwere a few houses clusteredaroundthe
Palace of the Giants, one of which was built against the apse on the west side.176From the scanty
remains it was impossibleto determinethe plan beyondthe fact that it had two and possibly three
rooms,or even more. Under the floorwas a fairly elaborateplumbingsystemwhich showedsigns of
hard use and repairs;its relationto the building and to the Palace remainsobscure.A small latrine
was constructedwith some care in the southwest cornerof the house; it emptied into a rectangular

Justinian's work at first hand, have unconsciouslyapplied Zosimus' evaluation of the 3rd-centurywalls to his own
time?
173
pp. 144-148.
Travlos, IoAeo8ojL,uK7j,
174
Hesperia, Suppl. IV, pp. 121-126. Once more, the following accountis based entirelyon Thompson'smeticulous
excavationand interpretation.
175 The latest coins are from the
reign of Anastasius I (A.D. 491-518); four of these were found in significantplaces.
For an analysis of all the coins found in the house, from both its constructionand destruction,see Metcalf, "Slavonic
Threat," pp. 141-143.
176
Hesperia 22, 1953, p. 42 (Thompson).
 84 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

terracottadrain which ran underthe outsidewall and into the Great Drain of the Agora (P1.21:d).
The entrance to the house was through a porch on the north side, with a threshold laboriously
workedfrom a Roman monumentbase.177
Apart from the houses of more or less recognizableplan discussedabove, evidencefor habitation
from the 4th to the 6th centuriesis to be found in a few scatteredremains of concretefoundations,
probably socles to carry mud-brickwalls, on the slopes of Kolonos Agoraios and the Areopagus.
That the Arroped
Areopagus slopes were
the residential district is further shown by the heavy
concentrationof wells, at least a dozen having been in use in the 4th century,and by the consider-
able amount of potteryand coins found in the same area.
© American School of Classical Studies at Athens

Shops
For personal use only. License: CC-BY-NC-ND.

Few shops have been identifiable as such in the post-Herulian Agora. A row of rooms added
along the front of some earlier houses in the southwesterndistrict (PI. 3: G-H 18-19) had access
directlyfrom the street and could hardly have been used for any other purpose. It is also likely that
the partitioningof the south colonnadeof the stoa northwestof the Agora on the PanathenaicWay
servedthe same need. 78

New CommercialConstruction
The only building recognizablydesignedfor commercialor some other utilitarianuse was a fairly
large structure,ca. 10 x 20 m., overlying the east end of South Stoa II (PI. 3: N-O 15), its south
wall beddedon the remainsof the backwall of the stoa and the north wall lying immediatelynorth
of the stylobate.The buildingwas dividedlongitudinallyby a wall along the medianline, and three
crosswalls (one missing) gave four pairs of roomsof equal size, i.e., 4.60 m. fromnorthto south and
3.80 m. from east to west. The rigidity of the plan, a series of equal-sized rooms at the north ad-
jacent to and presumablyopening into a similar series to the south, suggeststhat these were used as
shops. The builders of the east survivingcross wall came on a well encroachingon their line at the
south end and spanned its western edge with a flat arch of bricks set in very soft, yellow mortar
similar to that in a bit of the adjacentwall. The well was only 1.50 m. deep, and the waterjugs from
its period of use dated from the 5th and 6th centuries;it is not entirely clear whether the well con-
tinued in use after the constructionof the building. But the south side of the shop building abutted
againstthe northwall of a 4th- or 5th-centurystructurewhich was alreadyin ruins, and so the shop
building can hardly be earlier than the 6th centuryand may possiblybe as late as the 7th.179
DECAY OF EDUCATIONALINSTITUTIONS
The long continuity which the Neoplatonic School had enjoyedwas abruptly ended when Jus-
tinian, accordingto John Malalas, sent an edict to Athens prohibitingthe teaching there of phi-
losophy and law. 180 In spite of Malalas' reputationfor inaccuracythis statementhas generallybeen
accepted as true; it is also universally interpretedas referringto the general laws in CodexJus-
tinianus .II. I 0.2 and 1.5.1 8.4, in which Athens is not mentionedspecifically.How far and how per-
manentlythis act disruptedthe materialand intellectualwell-being of Athens has recentlyreceived
the attentionof several scholars,with conflictingconclusions.'8'
177 I 6477.
178
Hesperia 42, 1973, pp. 373-382 (Shear,Jr.).
179 Although the formal plan of the building suggests shops, Eugene Vanderpooloffers an alternativesolution:that it
was erectedtowardthe end of the lifetime of the Palace of the Giants for the use of the superintendentof the complexas
a toolshedand general storehousefor its maintenance.
180
ChronographiaxvIII.i87 (CSHB, p. 451).
181 For the
complex arguments,cf. the following:Alan Cameron,PCPS;J. P. Lynch, Aristotle'sSchool:A Studyof a
 DISINTEGRATION AND DISASTER (A.D.529-582) 85

The immediatesequel is told by Agathias, our only sourcefor the episode,writing in the 570's.182
Accordingto the bare text, seven philosophers, Damascius (Syrian), Simplicius (Cilician), Eula-
mius (Phrygian), Priscianus (Lydian), Hermeias and Diogenes (from Phoenicia), and Isidore of
Gaza, "the very flower of the philosophersof our time," went to Persia where they visited King
Chosroes because 'they did not share the view of God prevailing under the Romans and thought
that the Persian state was far better . . . [and they were] persuadedby the widespreadtale that the
Persian governmentwas supremelyjust, the union of philosophy and kingship."Disillusionment
soon set in, and they returnedhome undera safe conductarrangedbetweenJustinian and Chosroes
as part of a peace treaty signed in 532. Although Athens is not mentionedas the point of departure
© American School of Classical Studies at Athens

for Persia, Damascius was Diadoch of the NeoplatonicAcademyat the time, and it is a generaland
probablysafe assumptionthat the others were membersof the school.183
For personal use only. License: CC-BY-NC-ND.

This short passage poses some questions, superficiallyminor but having a bearing on the assess-
ment of the facts. First the date of departure, and whether Chosroes was actually on the throne
when the philosophers went to Persia. The sequence of Agathias' narrative seems to imply his
accession as a reason. After condemning Chosroes' gullibility toward the charlatan Uranius he
continues "and yet [Chosroes]had seen really eminent philosophersbefore who had gone to see
him" (KaLTo 7rTporepov apLfrov' w' aAL0S6 ireTEaro
vLAoXrooVs E'vbeV avrov a#LKOIE'-
vovs). That would fit the notion of the "unionof philosophy and kingship"which they found so
attractive. But this interpretationimposes difficult chronologicallimits on the whole adventure.
Chosroes came to the throne in Septemberof 531, which would allow only one year for him to
establish a reputation as a lover of philosophy strong enough to reach as far as Athens and induce
the philosophersto leave for Persia; to give them time to learn that their hopes had been deceived
and persuade Chosroesto let them depart;and finally for Justinian to be inducedto write into the
peace treaty (September532) a safe conducthome for all seven.
Agathias says merely that the philosophers"cameto see"Chosroes.It is perhapsconceivablethat
they were alreadyin Persia at the time of his accession.In spite of the stormyeventsthat markedthe
reign of his father Cavades,many reformswere set in motion during that reign for which Chosroes
receivedmuch of the credit.184The reportof these might have reachedthe west, perhapsduringthe
negotiations(which fell through) for the adoptionof Chosroesby Justin I, and becomesufficiently
"widespread"by 529 to instill in the philosophers'minds the idea that Persia would be a congenial
place.185
Whether the philosophersdepartedimmediatelyon learningof Justinian's edict or waited a year
or two (without teaching) to see if the new law had any more force than those of the past aimed at
suppressing paganism, Agathias makes it clear that the cause was the new legislation combined
with the news from Persia. After describing their idealistic expectations he continues, "believ-
ing this to be true, and being forbiddenby the laws to practice freely the rights of citizenship

GreekEducationalInstitution,Berkeley/Los Angeles 1972, chap. VI; H. J. Blumenthal,Byzantion48, 1978, pp. 369-

385; Glucker, Antiochus, pp. 247-255, 322-329, 377-379; I. Hadot, Le probleme du neoplatonismealexandrin:
Hierocleset Simplicius, Paris 1978, chap. II. For a more optimisticopinion of Malalas' accuracysee now M. Peachin,
"JohannesMalalas and the Moneyers' Revolt,"Latomus180, 1983, pp. 325-335.
182Agathias, Historiae II.30-31. For analysis of Agathias' excursus on Persia, with translation,see Averil Cameron,
"Agathiason the Sassanians,"DOP 23/24, 1969/70, pp. 67-183, esp. pp. 164-176. For the date of Agathias'History,
cf. Averil Cameron, Agathias,Oxford 1970, p. 9.
183 See Averil
Cameron, Agathias(footnote 182 above), p. 101 and DOP, pp. 172-175.
184
See A. Christensen,L'Iran sous les Sassanides,2nd ed., Copenhagen 1944, pp. 353-361.
185
Also, the possibility might be consideredthat the lure of the "philosopher-king"and the union of philosophyand
kingship enteredAgathias' mind only in retrospect.
 86 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

(e,utroAtrevo-Oat)because they were not conformingto the existing order, they immediatelywent
abroad (av'rKa aTLovreS poXovToes a&Xoba7ra... M6vi)."186
What teachers were affected by the law? Malalas specifies philosophy and law as the subjects
that could not henceforthbe taught at Athens. The actual edicts, however, are more general. The
prohibitionis against the teachers,not the subjects."We forbidthe teachingof any subject(7rav8
&aOripa)by those sufferingfromthe insanityof the impious pagans, so that they shouldnot pretend
to educatethose who wretchedlyresortto them, while actuallythey are corruptingthe souls of their
so-called pupils . . . nor shall they receiveany public financial support" (Cod. Just. . I I .10.2). The
penalty for violationof the orderis confiscationof propertyand exile. In anotheredictof 529, aimed
© American School of Classical Studies at Athens

primarilyat heretics,pagans are also excludedfrom militaryserviceor holdingany office (1.5.18.4).

For personal use only. License: CC-BY-NC-ND.

Although literary evidenceexists only for the closing of the NeoplatonicAcademy,from the word-
ing of 1.1I1.10.2 it seems that all pagan teachers,sophists, and rhetoricians,as well as philosophers
and lawyers, fell under the ban, while in addition,those who held public chairs would have their
funds cut off.187
More important than the date of departure from Athens is the question of the philosophers'
return. Under the terms of the treaty they all went "home" (iaravrTs oKabf a7revoaorrT7aav).But
where was home? Although all presumablyhad taught in Athens, none, apparently,was a native
Athenian.188From Agathias'text we may infer that safe conductwas extractedfromJustinian only
on Chosroes' insistence (ov yap aviJKevo Xopo7'pi p7l ovxt KaLerL Tr8f 'vo'rTjvaLKaLKpareLvrT?V
eKeXELpiav). The treaty required that the philosophersbe allowed to return home and live safely
"bythemselves"(eW'eavrots) without being requiredto adopt opinions which they did not hold or
to changetheir own faith (Averil Cameron'stranslation).This suggeststhat it was only to ensure a
safe return to private life and not to teaching.
Agathias ends his narrativewith the departureof the philosophersfrom Persia, with no mention
of their ultimate destination. It has been suggestedthat Simplicius returnedto Athens, but not to
teach.189If so, he would have been following his own preceptthat if a philosopheris at odds with
the authoritiesand cannotkeep his faith and his self-respecthe must withdraw from public affairs.
If possible he will retire to another state, otherwise he should hide, continuingto teach himself or
anyone he can.'90Where the others went is unknown.
The strongestcase for a resumptionof the Academy'sactivitiesafter only a shortinterruptionis a
statement of Olympiodorusthat the Academy'sendowment funds (8bLaoxLKa)have lasted to the
present day in spite of the many confiscationswhich are taking place, a passage which Cameron
sees reason to believe was not written before 560.191

(DOP [footnote182 above, p. 85], pp. 173-175) believes that reportsof Chosroes'promise as a
186 Averil Cameron

philosopher-kingmight have reachedthe west beforehis accession.

187Procopius (Anecdotaxxvi.i) cites the cutting off of funds as part of Justinian's destructionof "themarks of dis-
tinction and all things which confer honor and beauty both in Byzantiumand in every other city" (trans. Dewing). A
curiousassumptionby Alan Cameron(PCPS, p. 8) that the schoolsof rhetoricwere exempt from the orderbecauseby
then "mostof the rhetoriciansare likely to have been Christians anyway" is tenuous, given the uncooperativespirit
which the Athenians of Late Antiquity displayed toward efforts to impose Christianity (Frantz, DOP, pp.187-188,
200).
188Thus continuingthe ancienttradition.In the 4th centuryB.C.,at least, few of the famousphilosophersand sophists
were native Athenians.
189 Alan
Cameron,PCPS, pp. 21-25.
190Quoted ibid.,p. 16.
191
Ibid., pp. 11-12. The passage has been analyzed in detail by Blumenthal ([footnote181 above, p. 85], pp. 369-
372, 384), Lynch ([footnote181 above, p. 84], pp. 177-189), and Glucker (Antiochus,pp. 248-251).
 DISINTEGRATION AND DISASTER (A.D.529-582) 87

Of what the diadochikaconsistedis uncertain.If some of the many bequeststo the Academyhad
been in the form of landholdings,might not the confiscationshave been applied piecemeal,regard-
less of the status of the Academy?The excavatorsof "Proclus'house"reportedthat the stratigraphy
indicatedabandonment"in the 6th century,"but it was impossibleto be more precise.The prestige
and influence of the Neoplatonic School might have been sufficientto protectthe propertyduring
the absenceof the philosophersin Persia (?) and to maintainit with the dwindlingdiadochikafor a
few more years, even without pupils.
It is clear that the literarytestimonyprovidesno final prooffor either side of the argument.Not to
be dismissedlightly, however, is the archaeologicalevidence,which shows a sharp breakin activity
© American School of Classical Studies at Athens

in the complex of houses on the north slope of the Areopagus about 530.192That there is good
reason to associatethese with educationis superfluousto the present argument.193
For personal use only. License: CC-BY-NC-ND.

All the buildings under discussionwere abandonedsuddenly,not as a result of general decline,

neither asa result of violence. The most telling piece of evidnce
evidenceis providedby House C of the
Areopagus group, discoveredin 1970, long after the first tentative identificationof others in the
group was first put forward.'94The new evidencehas been published briefly but will be reviewed
here in the context of all the houses of the group.195
House C, like the othersin the group,was built in the secondhalf of the 4th century,probablyas a
privatedwelling, since a well at the edge of what was later to becomea nymphaeum(and still later a
baptistery)containedwaterjars from the middleofthe he 2nd centuryto thee end of the 4th, when that
part of the building was evidentlyturnedto more formaluse.196It stoodunchangedfor more than a
century,when a radicaltransformationtook place, showing signs of hasty abandonmentby the first
owners. For the date at which the abandonmentand remodelingtook place the most importantevi-
denceis fromthe least thatconspicuouselement: then floorsin threerooms,also in the main
peristyle and in the small southwest service court. The latest potteryfrom between the floorsin all
these contextswas uniformlydated to the early 6th century.197A well (Q 21:4) in a room adjoining
the easternservicecourtyardwas coveredwith a marbleslab laid flush with the surroundingoriginal
floor level in such a way as to suggest that the well was deliberatelysealed when the house was
abandoned. It was this well and another (P 21:2) in thehe main courtyardof the same house that
containedr fre thedescribedin the previouschapter.198The lack of potterysuch as accu-
mulates in disused wells indicates that well Q 21:4 was in use up to the time the sculpture was
depositedand sealed immediatelyafter. It was never again accessible,for a wall of the bath which
was includedin the remodelingwas laid directlyover the slab and the adjacentfloor.199
192Alan Cameron
(PCPS), writing beforethe most importantdiscoverieswere made, attributesthis state of affairsto
general decline in Athenian prosperity rather than to Justinian's act (p. 8). But this judgment is underminedby an
opinion expressed later (pp. 28-29) that "so far from Athens being already eclipsed by Gaza and Alexandria when
Justinian stepped in, both Gaza and Alexandria were in full decline while Athens seemed set for a fresh period of
prosperity at this period. Students were flocking back to the Academy,now under the directionof the colourful Da-
mascius. Could it not be this that causedJustinian to act? Not becauseAthens was decliningbut anxiety becauseit was
reviving?"
193For the reasons, see above, pp. 44-47.
194
By H. A. Thompson, "Twilight,"p. 68.
195 Reports of the discoveryof House C appear in Hesperia 40, 1971,
pp. 266-270, 273-275 and 42, 1973, pp. 156-
164, 168-173 (Shear, Jr.); Frantz (footnote30 above, p. 19), pp. 29-38. Cf. also above, pp. 40-41.
196
For the chronologicalevidence,see above, p. 48.
197
Hesperia 42, 1973, p. 161 (Shear, Jr.). For the remodelingitself, see below, pp. 88-89.
198
P. 41.
199 Well P 21:2 also was in use
up to the time when the sculpturewas deposited.This well, however,continuedin use
until the destructionof the house because the sculpturewas two meters below the destructiondebris of the 580's. The
 88 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

As notedin the previouschapter,the courseof the historyof all the Areopagushouses,A-D, up to

and including the sharp break ca. 530, was similar. Houses A, B, and D, however, went into a
steady decline from which they never emerged,while House C was quickly taken over but under a
differentdispensation.200
As noted above, all the sculpture was evidently a collectionmade by a lover of classical art.201
None of this is ordinarydebris such as is commonlyfound in wells. All was in an excellent state of
preservationand was obviouslyplaced in the wells for purposes of concealment.Considerationof
the time and the conditionof the sculptureleads to the conclusionthat the two events,the actionof
Justinian and the concealmentof the sculpture,are closelyrelated.The occupantseither left Athens
© American School of Classical Studies at Athens

or disappearedfrom sight, while still keeping the hope that their fortunes would be reversedand
For personal use only. License: CC-BY-NC-ND.

that they would returnto their old surroundings.

Parallels for this sort of activitycan be found in other periodsand places. First of all, the Varva-
keion copy of the Athena Parthenos(Athens, N.M. 129) was found protectedby a brickvault in a
house of the Roman period in the northernpart of Athens.202With it were a statuetteof Asklepios
and a woman's head. A collection of sculpture dating from the late 2nd to the 4th century was
buried in a room in a villa in Antioch under apparently similar circumstances.203 And a statue
found under the house of the Brunelleschi family in Florence is describedby Ghiberti: "la quale
statua quando sormentola fede cristianafu nascosain quel luogo da qualche spirito gentile ... fece
murare una sepoltura di mattoni e dentrovi seppelli detta statua. . . .204 Most recently, a large
collectionof heads and other fragmentsof sculpture from Notre Dame was discoveredin a secret
cave where it had been hidden at the end of the 18th centuryto escape the "revolutionaryenthu-
siasm [which]was readyto destroythem as symbolsof tyrannicalgovernment."205 And in the center
of London (Walbrook), in an apparently privately owned Mithraeum, was found a cache of
sculpture,carefully concealed,evidentlyfor protectionagainst Christian fanaticismas early as the
4th century.206
The natureof the changesmade at the time the floorsof House C were relaid leavesno doubtthat
they representeda transitionto Christianuse of an officialcharacter.The majorinnovationwas the
transformationof the southeast cornerof the house into a baptistery,to which the plan lent itself
well. At first glance it might be thought that the complex had been a baptisteryfrom the start, so
well may have been sealed when the sculpturewas put in, but the need for water may have obliged the new owners to
reopen it, thus accountingfor the two meters,representingthe periodof use, betweenthe sculpture,which was foundat
the very bottomof the well, and the final destructiondebrisof the 580's (see below, p. 90).
200 For details of this
chronology,see above, pp. 47-48.
201 P. 41. It was not confinedto House C:
cf., for example, the Hermes from House B, also from a well, but none of
the rest had been disposedof with such care.
202 W. H. Schuchhardt,Antike Plastik II, Berlin 1963, pp. 31-32, and K. Lange, "Die Athena Parthenos,"AM 5,

1880, pp. 370-372. Judith P. Binder has kindly called my attentionto this.
203 D. M. Brinkerhoff, A Collection of Sculpture in Classical and Early Christian Antioch, New York 1970, esp.
pp. 35, 63-72. It is suggestedthat the concealmentwas occasionedby the Persian sack of Antioch in 540 or the Arab
conquestin 637.
204
L. Ghiberti, I Commentarii,0. Morisani, ed., Naples 1947, p. 55.
205 X. Muratova, "Currentand
ForthcomingExhibitions,"Burlington Magazine 119, 1977, pp. 725-726. I owe
these last three references to the kindness of Joseph Alsop. See also A. Erlande-Brandenburg,"Les sculptures de
Notre-Dame de Paris decouvertesen 1839," Revue du Louvreet des musees de France 29, 1979, pp. 83-89, for an
earlierdiscovery.
206 R. Merrifield in Roman
Life and Art in Britain, J. Munby and M. Henig, edd., London 1977, pp. 375-401. See
also M. J. Vermaseren, Corpus inscriptionum et monumentorum religionis Mithriacae, The Hague 1956, I, p. 285,
no. 814.
 DISINTEGRATION AND DISASTER (A.D.529-582) 89

convenientlydoes it comply with the requirements.But the elementsdenotingChristianuse can be

seen to be later additions, and it seems certain that the room with the pool was originally a nym-
phaeum. The most importantchange was the removalof a mosaicpanel, probablycontainingsome
offending scene, from the middle of the floor of the rectangular section of the room and its re-
placementwith a panel of the same size with a simple cross of red stone, its angles filled with mis-
cellaneous bits of marble (Pls. 34, 43:c).2?7Although the use of the cross on a floor was generally
frowned on at that period, E. Kitzinger has shown that it was often used as an apotropaicsymbol,
especially in key places such as thresholds,to deny accessto the powers of evil which were thought
to lurk particularlyat doorwaysand other openings.208This would have been particularlyappro-
© American School of Classical Studies at Athens

priate where the new plaque was replacingan emblemaof pagan content.The substitutionwas not
For personal use only. License: CC-BY-NC-ND.

carriedout entirely successfully,as the edges show plainly. The surroundingmosaic,on three sides
only, makes it evident that this part of the room was a triclinium. The combinationof triclinium
and nymphaeum,both on a palatial scale (e.g., Hadrian's villa at Tivoli and Domitian's palace on
the Palatine) and in private houses, is well attestedin classicalantiquity.209
A wide variety of forms was allowed in nymphaea:rectangularor apsidal, with or without foun-
tains, with or without steps. The apsidal shape, reminiscentof an original cavern, was the most
widely favored.In some cases the pools were very shallow, in others deeper, sometimeswith steps
leading down, as in the Areopagushouse.210The lavish use of marbleon walls and pavements,seen
also in the Areopagushouse, is typical of nymphaeain the Late Empire.211
Great flexibility was admitted in the disposition and number of the ancillary rooms in baptis-
teries, the most importantbeing the catechumeneum,where candidatesfor baptism would partici-
pate in a preliminary service. This room was commonlyused also for funeral or commemorative
banquets. The large room adjacentto the triclinium on the west side (in the Areopagus house)
would serve this use well and might explain the late addition of benches along the east and west
walls for the participantsto recline or sit on (Pls. 27, 31). Color is lentthis
to this view by the discovery
in the Slavic destructionfill of two lamps with Christian symbols and a sigma table, about half
preservedbut broken (PI. 72:e, f, a).212The room opens into a small room south of the baptistery
proper which probablyservedas an apodyterion,or dressingroom.213Anothersignificantfind was
part of a circulartray of dark red stone such as were used in religious services(ST 781, PI. 72:c). A
very small bath was installed over parts of three rooms to the northeast of the nymphaeum
(PI. 21:a), obviouslya later additionbecause it partly coveredthe well which containedmost of the

207
For a comparablecross without finials, cf. the baptisteryin the Catacombof Pontianusin Rome (F. Cabrol, Dic-
tionnaired'archeologiechre'tienneet de liturgie, Paris 1922-1953, i, col. 407, fig. 1322). Also on numerouslamps of
the period (AgoraVII, passim).
208 E.
Kitzinger, "The Threshold of the Holy Shrine,"Kyriakon.FestschriftJ. QuastenII, Munster 1970, pp. 638-
647, noting a border with three simple crosses on the thresholdof the entranceto the central square of the cruciform
Martyrion of St. Babylas at Antioch.
209 See N.
Neuerburg, L'architetturadellefontane e ninfei nell' Italia antica, Naples 1962, p. 240 and figs. 81, 82
(Hadrian's villa); pp. 221-222 and figs. 77, 78 (Domitian's palace);passim (privatehouses).
210
E.g. Pompeii, Neuerburg, op. cit., p. 120 and fig. 48.
211 R.
Meiggs, Roman Ostia, 2nd ed., Oxford 1973, caption to pl. XIII.
212
L 5628, L 5630, A 3869. For a study of all known examples of sigma tables see G. Roux, Salaminede ChypreIV,
Paris 1973, pp. 133-196, esp. 174-180. The evidence for the purpose of these, beyond the fact that they have some
liturgical significance,is inconclusive.They are found in both pagan and Christian contexts.
213 For
comparablearrangements,among many others,cf. the baptisteryin the cruciformchurchat Antioch (R. Still-
well, Antiochon the OrontesII, Princeton 1938, pp. 29-34). The apsidal pool is especiallynoteworthy.
 90 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

sculptures (Q 21:4, above, p. 41). The inclusion of a bath in a baptistery,although not universal,
was common.214
The fate of the two pieces of sculpture left above ground in the house, a fine relief of the 4th
century B.C. representingHermes presentingthe infant Dionysos to the Nymphs, with the heads
deliberately hacked off, and a torso of Athena turned upside down and used as a step block
(PI. 39:e), emphasizesthe differentspirit which promptedthe treatmentof the two sets of sculpture
and reinforces the belief that the statues in the well were placed there from reverence,not for
desecration.215
The case of the related buildings on the Areopagausslopes is similar if less dramatic. In their
© American School of Classical Studies at Athens

earlier phases (i.e. constructionand main periods of use) Houses A and B, probablyalso D, run a
parallel course to that of C.216For the half centurybetween 529 and 582 the differenceis striking
For personal use only. License: CC-BY-NC-ND.

and important.Thanks to recentrefinementsin the datingof Late Roman potteryit is now possible
to discriminatebetweenthe progressivestages with some precision.Houses A and B show the same
breakin continuity,which in effect representedthe end of the useful life of the houses. Unlike their
neighbor C, however, which was immediatelyremodeledon a fairly grand scale, A and B show a
more ragged line of demarcation,with activity tapering off for 15 or 20 years and then ceasing
utterly before550. These were essentiallyprivatehouses, and their owners might have continuedto
live in them for a while even though they had fallen on evil times, deprivedpresumablyof the ample
fees that successfulteacherscommanded.
There is no reason to suppose that all philosophers and sophists reacted in the same way to
Justinian's edict beyond obeying the compulsoryinjunctionnot to teach. In fact, unanimity not
being a Greek characteristic,it would be surprisingif they had respondeduniformly.The owner of
House C, perhapsbecauseof having moreto lose, immediatelyburiedhis treasuresand left Athens.
The occupantsof the other houses may have elected to wait out the storm in Athens, with notable
lack of success.To carry speculationa little further,is it possiblethat all the residentsof the Areo-
pagus houses held chairs of philosophy, rhetoric, or sophistic and that their funds were cut off
immediatelyin accordancewith the provisionsof Codex Justinianus 1. I I. 1.2?
The walls of Houses A and B apparentlyremainedstanding,althoughprobablycrumblingfrom
lack of maintenance.The accumulationof pottery over the floors and the material salvaged from
most of the wells run down to the decade 540-550; other wells went out of use slightly earlier, ca.
530.217The amount of potteryand coins from the third quarterof the 6th centurywas very scanty,
indicatinghabitationonly by occasionalsquatters(who apparentlydinedexclusivelyon snails). And
so the houses remained,in a decayingcondition,until 582, when all of them, A, B, C, D, were en-
gulfed in a fiery destructionthat left its markfor posterityin the tricliniumof House C, with its floor
buriedunderan undisturbedlayer, 30 centimetersthick, of ash and charcoal(PI. 73:a, 74:a).218
E.g. at Djemila (J. Lassus, "Les edifices du culte autour de la basilique,"Atti del VI congressointernazionaledi
214

archeologiacristiana,Ravenna, 23-30 settembre1962, Ravenna 1966, pp. 597-599.

215 The relief: I 7154, Hesperia 42, 1973, p. 168 and pl. 35:c (Shear,Jr.); idem, OpusRom9, 1973, pp. 183-191. For

the identificationof the scene, H. A. Thompson, "DionysosAmong the Nymphs in Athens and Rome,"Journalof the
WaltersArt Gallery 36, 1977, pp. 73-84. The statue:S 2337, Hesperia 40, 1971, p. 275, pl. 59:b and 42, 1973, p. 163
(Shear,Jr.).
216 See above, pp. 47-48. The erosion of the hillside has carriedaway most of the evidencefor D; what little remains

points to the same conclusion.

217 For a
descriptionof the wells and their contents,see above, pp. 47-48.
218
A thick layer of destructiondebris,uniformin character,over the entire area of the houses atteststo a date ca. 580
for the disaster. Especially to be noted are a coin of Justin II, 574/5, found on the latest floor of House C and, among
many lamps, six which were identicalwith others found in the destructionfill of the watermill (AgoraVII, nos. 2402,
 DISINTEGRATION AND DISASTER (A.D.529-582) 91

In some respectsthe later historyof the Palace of the Giants was similar to that of the Areopagus
houses but without the dramaticbreak shown by House C. A brief period of abandonmentbegin-
ning about 530 is established by the presence of a tiled grave below the latest floor (P1. 75:a). A
short period of final occupationfollowed during which the Palace was put to use on a very modest
scale, and evidentlyonly in the basementrooms.Most of the evidenceis confinedto the southeastern
unit: several fragmentsof sigma tables (A 2266 and A 2293) and some lamps of the secondhalf of
the 6th centurywith Christian symbols,as well as severaltypically Christianampullae (PI. 72:b).
What the precise functionof the Palace was at this time is not entirely clear except that it was for
Christian use, as indicated by a cross carved on a marble column shaft found in the southeast
basement.The sigma table, in this contextwith exclusivelyChristianobjects,219
© American School of Classical Studies at Athens

suggestscommunal
worship, at least on a small scale. (Fragmentsof two otherswere found in the vicinity.) There is no
For personal use only. License: CC-BY-NC-ND.

evidenceon the ground for any monasticestablishment,but one hitherto unexplainedfeature may
cast some light. After the abandonmentof the complex in its original form, a series of small build-
ings or roomswas addedagainst the east wall of the southernsection. In the smallestof these there
were a large number of stamped unguent flasks or ampullae, concentratedin a little over a meter
square of soft earth and confinedto a depth of only about fifteen centimeters.220 The potteryin the
fill was of the mid-6th centuryand two coins of Theodoric (A.D.493-526) were found close by, at a
slightly lower level. All the ampullae belongto a type recentlyestablishedby John Hayes as coming
from a single source of production,probablyPalestine, althoughthey are found all over the eastern
Mediterranean.221They are cigar shaped, averagingca. 0.20 m. in height. Near the bottom is a
rectangularor circular stamp with a monogramor other device (P1. 73:b). Because of their heavy
fabric, suitable for export, and convenientshape for holding in the hand, Hayes concludesthat they
were made as containers,probablyfor sanctifiedoil from one of the great shrinesof the Holy Land,
to be exported all over the Christian world.222Because of the large numbersthat have been found
Hayes believes that they were intended not only for liturgical purposes but also for private use.
They flourishedfrom ca. 550 to ca. 650, their disappearancebeing "presumablyone of the conse-
quences of the Arab conquest of Palestine and the changes in the pattern of trade which resulted
from it."
In addition to the one concentrationof these ampullae in the Palace, a dozen or more examples
have been found at scatteredpoints in the Agora area. One might conjecturethereforethat all or
most of the Palace was turned into a monasticcomplexof some sort and that the monkspursuedan
independent (but appropriate)sideline in distributingthe ampullae with their sanctifiedcontents
among the neighborhood.223The ultimate fate of both the Palae and "Proclus'house"224was
differentfrom that of the houses on the Areopagus.Both buildingsevidentlyescapedthe Slavic fire;
instead, they were gradually overlaid with silt which accumulated
withey all through the rest of the 6th
century. The Palace was a more formal building than the others; its main floor (in the southeast
wing) had alreadybeen put out of use during the Vandal(?) invasion, and the basementroomshad

2578,2811, 2812,2813,2830). For the destructionof the mill see below, p.93. For the final, ignominious,revivalof the
houses, see below, p. 123.
219
See footnote212 above, p. 89.
220
Twenty-two were catalogued;fragmentsof many others are storedwith the contextpottery.
221
J. W. Hayes, "A New Type of Early Christian Ampulla,"BSA 66, 1971, pp. 243-248.
222
For their distribution,see ibid., p. 246, fig. 3.
223
Hayes notes: "To what extent the Church participatedin their distributionand sale is a matter for conjecture,
though one suspectsthat a profitablesideline such as this would not be left entirely to outsiders."
224 For
this, see above, pp. 42-44.
 92 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

been modifiedin such a way that they seem to have been intendedonly for storagespaceand so were
unsuited for human habitation.
That the effect of Justinian's edict reached beyond the sphere of educationis illustratedby the
fate of the Palace of the Giants. One can only conjecturethat with the eclipse of the schools, the
main industryof Athens in Late Antiquity,the whole city suffereda declineand that funds were no
longer availableto maintain the Palace.
In the light of the evidencepresentedabove,the situationin Athens after 529 may be summarized
as follows:Almost immediatelyafter the publicationof the decreesome philosopherswent abroad.
That they ever returnedto Athens is open to question, but if they did it was not to teach. Others,
© American School of Classical Studies at Athens

including also sophists, stayed on but without pupils and so without any means of livelihood,
For personal use only. License: CC-BY-NC-ND.

whether from imperial funds or private fees. Death or diminishing resources finally caused the
abandonmentof such buildings as had not already been taken over by the Christians, but the
buildings that had been turned to Christian use surviveduntil they were destroyedby the Slavs in
582.225The date ca. 530 for the end of effectiveteachingin Athens, attestedby such a considerable
amount of physical evidence,can hardly be coincidence.There is no other event in recordedhistory
to accountfor the suddenbreakat this time. A reasonableconclusionis that now, finally, a law was
passed with sufficientforcebehind it to exact compliancein such a way as to have a profoundeffect
on the material prosperityof the city.

CONVERSIONOF TEMPLESINTOCHURCHES
As previouslystated, the conversionof pagan temples into Christianchurchestook place far later
in Athens than elsewhere in the Empire except Rome.226The Christiancommunitywas still in no
position to challengethe powerful pagan upper classes;it was perhaps also motivatedby a distaste
for moving into what might have been regardedas contaminatedquarters.There is no evidenceof
any pagan temple being turned to Christian use beforethe late 6th century.The conversionof the
Parthenonand the Asklepieionmay probablybe assignedto late in the reign of Justinian.227The
Hephaisteion and the Erechtheion,on the other hand, very likely retainedtheir pagan, or at least a
neutral, status until well on in the 7th century.228The Ionic temple on the Ilissos, the only other
temple in Athens known to have been transformedinto a church,229was totally destroyedin the
18th century;the only recordof its final phase is Stuart and Revett'sbrief accountand drawing of
the exterior.230

Lynch ([footnote181 above,p. 84], p. 167) is of the opinion that it was the Slavic invasionthat was responsiblefor
225

the fact that "these Neoplatonists were the last generationof pagan philosophersin Athens."But the archaeological
evidenceis so strongthat this view is no longer tenable.
226 P. 70 above; also Frantz, DOP, pp. 185-202. For the similar situation in Rome, where, as in Athens, trans-

formationof secular buildings long precededthe conversionof temples, see R. Krautheimer,Rome: Profile of a City,
Princeton 1980, pp. 71 ff.
227 F. W. Deichmann, "Die Basilika im Parthenon,"AM 63/4, 1938/9, pp. 127-139. For the Asklepieion, see

Frantz, DOP.
228 Hephaisteion, Frantz, DOP, pp. 202-205. Erechtheion,The Erechtheum,J. Paton, ed., Cambridge,Mass. 1927,

pp. 492-523.
229
The Temple of Olympian Zeus incorporatedsmaller places of worship but was never convertedin toto.
230J. Stuart and N. Revett, Antiquitiesof Athens I, 2nd ed., London 1825-1830, chap. II, pl. VII, showing a low
dome, which suggests a remodelingin the 15th or 16th century, although, like the Hephaisteion, it may have had an
earlier Christian phase as well.
 DISINTEGRATION AND DISASTER (A.D.529-582) 93

THE SLAVIC INVASION

The disintegrationof Athens which began from neglectwas completedby a temporarybut devas-
tating invasion by a band of Slavs and Avars, probablyin 582.231This attack was only part of a
general incursion into Greece which has left its mark in coin hoards in Isthmia, Corinth, Ken-
chreai, the western Peloponnese (near Olympia), and elsewhere.232It is unlikely that Athens was
actually occupied by Slavs at this or any other time, but evidence from several parts of the city
indicatesdestructionon so wide a scale as to keep the Agora, at least, and probablymost of the rest
of the city from revivingin any appreciableway for the rest of the century.233
Referenceto the Slavic destructionhas been made repeatedlyin the precedingchaptersin connec-
© American School of Classical Studies at Athens

tion with individual buildings. A summaryof the most importantpieces of evidencemay be useful.
For personal use only. License: CC-BY-NC-ND.

A frameworkis providedby six hoardsor concentrationsof coins in the Agora and two elsewherein
the city. The Agora deposits encircle the east, west, and north sides; and a large area of intensive
destruction by fire, documented by coins and pottery, completes the circuit on the south side
(Pls. 73:a, 74:a). These deposits are as follows:
1) Over the floor of the secondperiod of the large house southeastof the Tholos (18 coins). Latest,
Tiberius II, A.D. 578-582.
2) A savings hoard at the northwest cornerof the Agora, close to the PanathenaicWay (18 coins).
Latest, Tiberius II.
3) A hoardon the north side of the Agora along the PanathenaicWay, apparentlyburiedin a series
of holes dug through a hard road level (341 coins). Latest, Tiberius II, A.D.579.
4) The flour mill (Central Mill) near the south end of the Stoa of Attalos (400 coins). Latest,
Tiberius II?
5) The late Roman building on the Broad Street, east of the Post-Herulian Wall (Room G).
Contentsof a purse (6 coins). Latest, Justin II, A.D.565-578.
6) The same building, Room 8 (244 coins).234Latest, Anastasius I, A.D.491-518.
The most extensive area of destructionby fire in the last quarter of the 6th century is that of
the "School"buildings on the north slope of the Areopagus, which has already been described
231
Metcalf ("SlavonicThreat")established the dateon thebasisof five hoardsof coins from the Agora and elsewhere
in the city. The date has since been corroboratedby the discoveryof three more hoards (see below).
232
For referencesto these, see J. H. Kroll, G. C. Miles, and S. G. Miller, "An Early Byzantine and a Late Turkish
Hoard from the Athenian Agora,"Hesperia 42, 1973, pp. 302-303. The complex problem of the dates, extent, and
durationof Slavic infiltrationinto Greece in general has been debatedfor over a centuryand is outsidethe scope of this
book. For a convenient summary, with extensive bibliography, see G. L. Huxley, "The Second Dark Age of the
Peloponnese,"AaKovxKalcYIrovbal 3, 1977, pp. 84-110. For the Slavs in Greecewith special referenceto Thessalonica,
see S. Vryonis, "The Evolution of Slavic Society and the Slavic Invasionsin Greece,"Hesperia 50, 1981, pp. 378-390.
233 Metcalf's caution in
minimizing the destructionand the duration of the "interruptionof ordinarytown life" as
"onlytemporary"is excessive. This judgment was based on the fact that "coinsstruckin the reigns of Maurice and his
successorshave been discoveredin the Agora in considerablenumbers."But a distinctionmust be made between the
reign of Maurice himself, which coversthe 20 years after the invasion (A.D. 582-602) and accountsfor only 25 coins
(1.2 per year) and those of his successors:Phocas (A.D. 602-610), 48 coins (6 per year); Heraclius (A.D. 610-641), 232
coins (7.4 per year); and Constans II (A.D. 641-688), 817 coins (30 per year). These figures are sustained by other
archaeologicalevidence in the Agora, accordingto which recoverywould have set in only in the beginning of the 7th
century.
234 Deposits 1, 2, and 4 were published by D. M. Metcalf ("SlavonicThreat"),deposit 3 by Kroll, Miles, and Miller

([footnote232 above], pp. 301-305). Deposits 5 and 6 are describedbriefly in Hesperia 42, 1973, p. 396 (Shear,Jr.).
Shear notes that in deposit 5 "thelarge proportionof clipped,heavily worn and illegible pieces gives the impressionthat
the group was depositeda good deal later."
 94 IV. PROSPERITY, DECLINE, AND DISASTER (A.D.396-582)

(pp. 44-48, above).The depositsoutsidethe Agora are a large hoardof 598 coins from the Dipylon
and 6 gold coins from an osteothekesoutheastof the Acropolis.235
The significanceof the Slavic attack on Athens is well illustratedby the destructionof the most
imposing of the buildings on the Areopagusand the only one of the group to survivethe transition
from paganismto Christianity.As the year 267 signaledthe end of Athensas one of the most impor-
tant cities of classicalantiquity, so 582 marks its end as a provincialcultural center.
235
These are included in Metcalf, "SlavonicThreat." The latest in the Dipylon hoard, A.D. 577/8; from the osteo-
theke, Maurice, probablyA.D. 582-584.
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.
 V
THE PALACE OF THE GIANTS
EXPLORATION AND IDENTIFICATION
The most prominent and probably also the latest product of the burst of building activity that
occurredin the area of the ancientAgora in the 4th and 5th centurieswas the great centralcomplex
which we have now labeled the Palace of the Giants (Pls. 52-54).) After a useful life of little more
than a century (ca. A.D. 410-530), the building was abandonedand subsequentlyvery thoroughly
© American School of Classical Studies at Athens

strippedof its masonry.The part that survivedlongest was a series of colossalfigures,Tritons and
For personal use only. License: CC-BY-NC-ND.

Giants, that had adornedits facade (Pls. 56-58). They were savedby their sheerbulk and awkward
shape which rendered them unattractivefor re-use as building material. One of these figures, a
Giant, still stands where it was placed soon after A.D. 400 (PI. 58:c, d). For centuries it kept a
solitary watch over the desolate area of the ancient Agora while the foundationsof the enormous
building to which it belonged were gradually coveredby silt and, from the 11llthcentury on, by
private houses. As a result of damage done to these houses in the War of Independencethis Giant
and his fallen companionscame to light and soon attractedthe attentionof the Greek archaeological
authorities.2
In repeated campaigns (1858, 1870, 1895-1896, 1912) the ArchaeologicalSociety in Athens
cleared the facade of the building to its full width togetherwith the four piers that had supported
the colossal figures.3It was now evident that six of these had flankeda triple entrancewayleading
into the building proper.The Societyre-erectedthe torsosof two Tritons that had fallen from their
pedestalsand replacedon one of these statues a head that had been carriedto Eleusis.
Thus matters stood for over two decadesduring which time scholarswere free to speculateas to
the form and function of the structure. But under the circumstances to speculate was still to grope in
the dark. Little wonder then that the most detailed reconstruction proposed in those years has
proved also to be the farthest from the mark: a large basilica lying east-west across the middle of the
Agora.4Walter Judeich, the reveredmaster of Athenian topography,writing in 1931, was more
reserved but more prescient. In his opinion the building (Hallenbau) was probably to be thought of
as a kind of central point in the area as it was in Late Antiquity without relevance to the boundaries
of the marketplace of the classicalperiod.5
The American excavators reached the area of the Giants in 1934, in the fourth season of the
Agora Excavations. In this and the following two seasons the whole extent of the Late Roman

I A draft of this chapterwas written by John Travlos in 1975. At that time the initial identificationof the complex
as a gymnasium was still accepted,and this affectedthe interpretationat many points. The present version takes into
accountthe new identificationas an officialresidence.But the physical descriptionremainslargely unaltered,and only
a few minor adjustmentshave been made in the splendidseries of plans left by Travlos.
2For the history of the Giants up to the middle of the 19th centurysee C. van Essen, BCH 50, 1926, pp. 210-212;
Judeich2, pp. 230-231; Thompson, "Odeion,"pp. 137-139. Especially interestingis the impact made by the standing
Giant on Cyriac of Ancona in the 15th century:B. Ashmole, Proceedingsof the British Academy45, 1959, pp. 40-41,
pls. XIV-XVI. For the Giant emerging from the ruinous houses see the water color by the Swiss artistJ. J. Wolfens-
berger, dated 1834: Wycherley (footnote6 above, p. 1), frontispiece.
3 EVvorTTLKr "EEKuO-Er rv Ilpa6eWv Tis9 'ApxaLoAoyLKs9 EraTaLpedag 1858/1859, pp. 14-19; IIpaKTLKa1870/
1871, pp. 12, 33 and 1895, p. 19; AM 21, 1896, p. 109; HIpaKTLKa1912, pp. 91-99.
4 Van Essen
(footnote2 above), pp. 183-212.
5Judeich2,p. 330.
 96 V. THE PALACE OF THE GIANTS

complex was stripped of its heavy overburden,and the ground plan became fairly clear.6As yet
attention was concentratedon the earlier buildings the ruins of which underlaythose of the Late
Roman complex, viz. the great Hellenistic stoas and the Odeion of Agrippa. A detailedaccountof
the Odeion appeared in 1950. This publicationcontaineda brief descriptionof the late complex
with emphasison the historyof the colossal figureswhich had servedboth structures.7
The first parts of the great complexto becomeintelligiblein the courseof the excavationwere the
colonnadedcourtyard,of which the colossalfiguresformedthe fagade,and a bathingestablishment
farthersouth. Since the combinationof courtyardand bath is one commonlyfound in gymnasium-
bath establishmentsof Roman Imperialtimes,8the complex was tentativelyidentifiedas a gymna-
© American School of Classical Studies at Athens

sium, and it was suggested that the rooms opening out of the colonnadesmight have served as
classroomsin a predominantlyeducationalinstitution.
For personal use only. License: CC-BY-NC-ND.

More completeexcavationand study, however,have made it apparentthat the whole southeast-

ern part of the complex with its two additionalcolonnadedcourtyardseach borderedby rooms is
unquestionablyof a residentialcharacter,quite comparablewith the large houses of the same peri-
od that have now come to light on the slope of the Areopagusjust to the south (PI. 6). It has now
becomeobvious,too, that the bath was scaledto the needsof a residenceratherthan of a gymnasium
of the size suggestedby the great north courtyard.Since all these elementswere contemporaryand
parts of one design, the inference seems inescapablethat the north courtyardis to be regardedas
ancillary to the residence,serving ceremonialneeds. Since in its overall scale the establishmentas
thus envisagedfar exceeds any conceivableprivate residence,we are driven to the conclusionthat
we have to do with an official institution, the functionof which will be explored below (pp. 110-
116). These considerations,we believe,justify the changeof name from "Gymnasium"to "Palace".
LOCATION
The Palace was planted boldly in the very middleof the area once occupiedby the Agora (Pl. 6).
At the time of its constructionthe east side of this area was firmly defined by the Post-Herulian
Wall and the west side by the Hellenistic Metroon, now partially rebuilt probablyas a synagogue
(above, p. 59). Southward the new structurewas carried up to the still lofty back wall of South
Stoa II which at this time supportedan aqueduct.Northwardthe great complex looked down the
PanathenaicWay.
In the placing of the Palace marked attention was paid to the Panathenaic Way, the age-old
thoroughfare which still ran diagonally through the area from northwest to southeast (above,
pp. 26-29). For the visitor, normally approachingfrom the northwest, the upward grade of the
road made the facadeof the new structureseem truly monumental(PI. 55:a). The terrainhad been
similarlyexploited alreadyby the buildersof the Odeionof Agrippawhich had previouslyoccupied
the site (PI. 55:c). It was also the pull of the PanathenaicWay that induceda striking departure
from axial symmetryin the overall design of the Palace:the southernpart was shifted eastwardso
6
Hesperia 4, 1935, pp. 360-363, pi. III; 5, 1936, pp. 6-15, fig. 13; and 6, 1937, pp. 357-358, pl. IX (Shear, Sr.).
Details of the Palace complex were elucidatedin the course of subsequentexplorationof the earlier buildings on the
site, e.g., Hesperia 22, 1953, pp. 40-41 and 35, 1966, p. 45 (Thompson).
7Thompson, "Odeion,"esp. pp. 134-139. For brief general accountsof the Palace see Agora XIV, pp. 211-213;
AgoraGuide3,pp. 110-115; J. Camp, The AthenianAgora,London 1986, p. 200.
8 On the blend of gymnasium and thermae, see Ward-Perkins,Architecture,pp. 292-296 with bibliography.The
phenomenonis well illustratedby the great gymnasiaof the cities of Asia Minor. See for example F. Miltner, Ephesos,
Vienna 1958, pp. 58-68 (Gymnasiumof Vedius) and F. K. Yegul, ArchaeologicalExplorationof Sardis,Report3: The
Bath-Gymnasium Complex, Cambridge, Mass. 1986, esp. pp. 147-151. See also W. Heinz, R6mische Thermen,
Munich 1983, passim, and bibliography,pp. 209-215.
 EXPLORATIONAND IDENTIFICATION 97

that the suite of rooms (33-41; see PI. 54) in the southeast block might overlookthe famous old
street. This arrangementalso facilitatedaccessto the "kitchenentrance"below that suite.
The continuingimportanceof the PanathenaicWay is evidentalso in the placingof the structures
to the north of the Palace (PI. 6). The fact that a freestandingwall was carriednorthwardfrom the
northwest cornerof the Palace to the site of the Temple of Ares indicatesthat some structurestood
at this time on the foundationsof the temple, probablythe temple itself.9North of the temple site an
enigmatic square building with peristyle court faced eastward on the Panathenaic Way (above,
p. 66), while the round or apsidal structurestill farthernorth was placed in the angle between the
Way and the ancient west road of the Agora (above,p. 60).
© American School of Classical Studies at Athens

This row of new buildings that borderedthe west side of the PanathenaicWay definitelyturned
For personal use only. License: CC-BY-NC-ND.

their backs on what was probably a rather bleak area to the west. On the east they were to be
screenedbeforelong from an equally unattractivearea by a freestandingwall againstwhich was set
the aqueductthat carriedwater to a series of flour mills (above,pp. 80-82).
An interesting and significant element in the overall design of the Palace was the inclusion of
extensive "garden"areas to east and west of the southernhalf of the building complex (Pls. 6, 55:a).
The bounding walls, of poor rubble concrete,extendedout from the southeastand southwest cor-
ners of the North Court. Thence the walls could be tracedto the northeastand northwestcornersof
the Middle Stoa. Farther south no actual remains have survived,but both walls undoubtedlycon-
tinued south to the line of the ancient roadway that borderedthe south side of the ancient Agora.
Although no tracesof planting have been detectedin these much ravagedareaswe may supposethat
they were indeed planted with trees and shrubsfor the delectationof the residentsof the Palace.
With its overall limits thus establishedthe Palace is seen to occupy most of the triangularspace
bounded to east, west, and south by three of the oldest roads known in Athens, the lines of which
were perpetuatedin importantthoroughfaresuntil our own times.
The Palace sprawled over the thoroughlystrippedfoundationsof a number of ancientbuildings
(PI. 53): the Odeion of Agrippa, the Middle Stoa, South Stoa II, and the two shadowytemples that
once rose between those two stoas.10For the most part the buildersof the 5th centuryafter Christ
paid little attentionto the ancientfoundations.In the case of the Odeionof Agrippa,however,there
appears to have been some consciousrelationshipbetween the old and the new. Not only did the
North Court of the Palace assume the same axial emplacementas the Odeion, but in its shape and
dimensionsit closely approximatedits predecessor.One is temptedin fact to supposethat the origi-
nal intention was to have the North Court coincidecompletelywith the Odeion, but that for some
reason the court was shovedthe northwardin the final design by 13.60 m. Inasmuchas the emplace-
ment of the North Court in relation to the Odeion resulted in little or no saving in the cost of con-
struction,the close physical relationshipmay be takento suggesta deliberateeffortto revivein some
measure a famous ancient monument.

OVERALL PLAN

The plan of the Palace comprisedfour majorparts:the North Court, the South Court complex,
the SoutheastCourt complex, and the Bath. In additionmay be notedthe elaborateconnectinglink
9The Temple of Ares certainly suffered in the Herulian raid of A.D. 267 as shown
by the re-use of some of its
marble in the Post-Herulian Wall. See W. B. Dinsmoor, Hesperia 9, 1940, p. 52; M. H. McAllister, Hesperia 28,
1959, pp. 1-64, esp. p. 41. But the fragmentsfound in the Wall derivefrom the ceiling of the temple, probablyall from
its east end. Fragmentsof other parts of the building (steps, walls, columns,entablature,roofing,and akroterion)were
found scatteredaround the foundations.One must thereforeconsiderthe possibilitythat the temple had been repaired
after the Herulian raid and that it was still standing 150 years later when the Palace was built.
10Agora Guide3,p. 160.
 98 V. THE PALACE OF THE GIANTS

between the North and South Courts (Rooms 9-12) and the walled, presumablygarden areas to
east and west of the main block. All these parts belong to the original plan; only minimal adjust-
ments and repairs have been detected.
The North Court (Rooms 1-8), monumentalin scale and in its sculptural adornment,has the
appearanceof an area for receptionsand formal exercises. The smaller South Court (Rooms 13-
20) also has a suggestionof formality,especiallyin the presenceof the spaciousroomopeningoff its
south colonnade. The still smaller Southeast Court, borderedby rooms on three sides (Rooms
22-41), has an intimate, genuinely residential character.We shall find reason to believe that the
principal living roomsof the master and his family occupiedthe piano nobile of the eastern side of
the Southeast Court complex;these rooms looked out across the PanathenaicWay. The bathing
© American School of Classical Studies at Athens

facilities to the west of the South Court (Rooms 42-49), balancingthe SoutheastCourt complex,
For personal use only. License: CC-BY-NC-ND.

were presumablyintended primarily for the use of the residents,but the plan suggests that they
were also meant to be convenientlyaccessiblefrom the North Court.
The overall dimensionsof the Palace are 84 meters from east to west and 129 m. from north to
south. The building proper covered an area of ca. 5000 sq. m. while the whole establishment,
includingthe enclosedgardenspaces,measuredca. 13,500 sq. m., i.e., aboutone quarterof the total
area of the old Agora.

STATE OF PRESERVATION
The present state of the building comes out in general from the photographs(Pls. 52, 57). The
foundationsof rubble concreteare preservedthroughoutwith few gaps, thus permittingan almost
complete restorationof the plan. In only a very few places, however, do the walls still rise above
floor level:in the east wall of the North Court, in the Bath, and notablyin the semibasementrooms
in the east block of the SoutheastCourt complex where they are preservedto a height of 2.60 m.
(PI. 60:b). In none of the courtyards,however,is the stylobatesufficientlypreservedto indicatethe
spacing or the form of the columns,and in only a few cases is there evidencefor the preciseposition
or width of the doorways.
Slight remnantsof marble wall revetmentremain in the Bath (Pls. 64, 65), and a few detached
fragmentsfound in the east blockof the SoutheastCourtcomplexmust comefroman upperroomin
that area (PI. 63:a). A few tesserae (PI. 63:b) from floor mosaic found along with these bits of
revetmentare the only evidencefor the use of mosaicin the whole complex, a surprisingfact. Some
terracottafloor tiles are still to be seen in Rooms 14, 17, and 18, openingoff the south colonnadeof
the South Court (PI. 59:a) and in the open area of the Southeast Court (22). Of the six colossal
figures re-used in the entrancefacadeonly one, a Giant, as noted above, has survivedin place;the
torsos of two Tritons found on the spot have been re-erected.The head of one of these Tritons has
been found at Eleusis and reattachedto its torso; the head of another Triton, likewise found at
Eleusis, is now exhibitedin the Stoa of Attalos. Severalnon-joiningfragmentsfoundon the site are
likewise preservedin the Stoa.

CONSTRUCTION
The Palace with its multiple parts was laid out on the ground with a fair degree of care. The
North Court, to be sure, is ca. 0.40 m. longer on the east than on the west side, but this deviation
would not have been noticeable.The masonryalso, insofar as it is preserved,shows good, honest
workmanship.
The foundationswere everywherecarrieddown to bedrock.Beneath the contemporaryground
level they consistedof small, unworkedstones set in an abundanceof strong,white mortarmade of
 EXPLORATION AND IDENTIFICATION 99

sand and lime. This part was pouredin a trenchwithout forms and so is readily distinguishableby
reason of its rough and irregular faces from the carefully finished part above ground level. The
clear line of demarcationis a useful indicationof the ground level at the time of construction.The
foundationsare normally only a few centimetersthickerthan the upper wall.
The walls above floor level consist for the most part of unworked stone set in the same white
mortaras that used in the foundations.The stone masonryis interruptedat intervalsby coursesof
red burnt brick measuring 0.42 x 0.21 x 0.04 m. (Pls. 60:b-62:a). The bricks are well made and
were apparently manufacturedexpressly for this building. No stamps have been noted. Bricks
occurnormallyin single courses,occasionallyin two, rarely in three. The individualbrickswere set
© American School of Classical Studies at Athens

in both faces of the wall, sometimesas stretchersand sometimesas headers.The normalthicknessof

For personal use only. License: CC-BY-NC-ND.

the wall is 0.67-0.68 m. with slight variations.Bricks were also used extensivelyat special points:
especially to reinforcecornersand dooropenings,in the constructionof vaults, and probablyalso in
the arcading of the peristyle courts. In addition to the rectangularbricks, use was also made of
square bricks of the same consistencymeasuring0.30 m. to the side;these occur, apparentlyindis-
criminately,in certain parts of the building.
In the constructionof the foundationsand upper walls, in additionto the unworkedstone, many
ancient blocks were employed. Some of these were salvagedfrom the ruins of the earlier buildings
on the site. They were normallyfragmentary:small piecesof limestoneor porospreservingonly one
worked face. A number of step blocks from the Middle Stoa were also incorporated,notably in the
foundation for the east wall of the Southeast Court (PI. 60:b). A few fragmentarycolumn drums
and capitals from the Odeion have been noted in the foundationsof the North Court. Finally, some
large river boulderswere includedin the concreteof the walls.
One would have expectedthe wall surfacesabovefloorlevel to be plastered,yet no traceof plaster
has been observedon either the inside or outside faces of the, admittedlyvery slight, remainsof the
walls. This may be due to chance in survival,to economyin construction,or to incompletion,most
probably the last. A dozen fragmentsof marble revetmentof various kinds found in the basement
rooms of the Southeast Court complex undoubtedlyderive from the piano nobile above. The same
may be said of a few fragmentsof floor mosaicfound in the same place. Marble revetmentwas used
freely in both the floors and the walls of the Bath.
Throughout most of the complex the floors were apparentlyof clay. This applies to both of the
larger courts and to most of the adjacentrooms. In the small SoutheastCourt, however, the exca-
vators found remnants of terracotta-tileflooring, as also in the five rooms (14-18) adjoiningthe
middle part of the south side of the South Court (PI. 59:a). For the curioustreatmentof the floor in
the northernmostbasementroom of the SoutheastCourt complex see below (p. 106).

NORTH COURT

The North Court (1) served as a monumental vestibule to the Palace (PI. 55:a, b). Its north
facade was pierced by three broad openings which gave access to a large open area, measuringca.
29.40 x 37.80 m., borderedby colonnadeson its east, west, and south sides. The entranceswere
flankedby colossal figures of Tritons and Giants salvagedfrom the ruins of the earlier building on
the site, viz. the Odeion of Agrippa. The massive piers supportingthese figures measured2.40 x
3.30 m. in plan. They were carefully constructedof ancientblocksin horizontalcourseswith some
mortar between. Among the re-used material in the piers is a fragmentof one of the double Ionic
columns from the upper storey of the facadeof the Stoa of Attalos (A 1401). The lateral piers were
crownedwith re-used marble geisa placed upside down (A 1303-A 1307).
 100 V. THE PALACE OF THE GIANTS

Only three of the great torsos have been found by the excavators(Pls. 57, 58).1 It is assumed,
however, that all six of the figures made in the Antonine period for the new facade of the Odeion
(PI. 55:c), having lain for a centuryand a half among the ruins of that building, were salvagedand
re-used by the builders of the Palace. The marblepedestalsof the easternmostand westernmostof
the figures now standing are ancient;the pedestal second from the east is a modernrestorationin
poros. The fact that the figure which has remaineduprightstoodclose to a cornerof its pier may be
taken to indicatethat this and the other middlepier each carriedtwo pedestalswith their respective
statues, leaving only single figures for the lateral positions.An anta may confidentlybe restoredon
either extremityof the series as in the Odeion of Agrippa.
© American School of Classical Studies at Athens

Since nothing apart fromthe colossalfiguresand their pedestalshas survivedfromthe fagade,the

restorationof the architecturalsuperstructuremust be largely conjectural.It may be taken as cer-
For personal use only. License: CC-BY-NC-ND.

tain, however,that the openingswere arcuatedratherthan trabeated.In the Odeionfor which they
were originally designed the pilasters behind the figures are assumed to have carrieda horizontal
entablature. But the interaxial span in the facade of the Odeion was only 3.83 m. whereas in the
facade of the Palace the centralopening measuredca. 5.60 m. and the lateral openings ca. 4.20 m.
Such spans called for arches.
The three arched openings must have been set in a wall resting on top of the piers behind the
sculpturedpedestals (PI. 55). The availablespace, as indicatedby the positioningof the one Giant
that has remainedin place, would have accommodateda wall with a thicknessequal to that of the
outer walls of the Palace, i.e., ca. 0.68 m. The sculpturedpedestalsbearingthe Corinthianpilasters
each with its Atlas-like Giant would thus have stoodin frontof the wall but would have been tied to
the wall by protrusionsof the entablature,the result being a decorativescreen analogous to that
familiaron fully developedmonumentalarchessuch as those of SeptimiusSeverusand of Constan-
tine in Rome. So much may be regardedas reasonablycertain,but for the restorationof the upper-
most part of the facade no evidence exists. It is possible, however, that the projectingsections of
entablature supported freestandingstatues as on the Arch of Contantine. If so, the wall would
presumably have been carried up high enough to serve as a backgroundfor the statues and to
constitutean attic. But this has seemedtoo speculativeto be includedin the restoreddrawing.
The complete disappearanceof the wall behind the Giants may seem puzzling. It may be as-
sumed, however,that in this most prominentpart of the building, and one in which great strength
was required,the architectwould have employedthe best availablematerial,eithermarbleor brick;
either of these would have invitedthe attentionof vandalsafter the abandonmentof the Palace.
The ground level in the line of the facade is indicatedby the constructionof the piers below the
"Giants"(PI. 57). These piers were topped with two low coursesof re-used marble blocks,below
which the masonrywas unsightly and surely not intendedto be visible. We may thereforerestorea
ground level between the piers 0.30-0.40 m. below the bottomsof the pedestalsfor the Giants. No
trace of thresholds or other arrangementsfor gates has been observed.Such may have vanished
along with the wall behind the Giants, but it seems more probablethat the passagewayswere not
intendedto be closed and that consequentlythe North Court was meant to be readily accessibleto
the public.
At the time of constructionthe general ground level to the north of the North Court lay ca.
1.80 m. below the floor level within the Court. One would have expectedthe differenceto be made
goodby a broadflight of steps leading up to the facade,but no remainsof such have cometo light in
the excavations.A less formal arrangementtherefore seems probable, viz. a simple earth ramp.
This is probablyto be construedas one of severalsigns of incompletion(below, p. 108).
11For a detailed accountof the colossal figures, see Thompson, "Odeion,"pp. 103-124.
 EXPLORATION AND IDENTIFICATION 101

Little can be said about the towers that flankedthe facadeon either side (3, 4). They undoubtedly
rose higher than the centralpart, howeverthat may have been finishedabove,and in all probability
they had plain facadesprovidingeffectivecontrastwith the "baroque"middlepart. Each tower had
the east-west width of the lateral colonnades in the North Court. The interior measured ca.
1.60 x 4.40 m. The south foundationhad the same depth and thicknessas that for the outer walls
and stylobate of the North Court. This makes improbablethe restorationof an exedra with an
archedfront like those in the Stoa of Attalos. More probablythe small roomscontainedstairs. Since
access must have been from the south, the available space could scarcely have accommodateda
stairway for the public, but it would have sufficedfor a narrow stair for servicepurposes.
© American School of Classical Studies at Athens

The North Court proper was severely simple in plan. The open space was borderedby colon-
nades on three sides, 5.10 m. wide on east and west, 6.00 m. on the south. A little of the visible part
For personal use only. License: CC-BY-NC-ND.

of the east outer wall survivesnear the north end; its thicknessis 0.68 m. The foundationsbeneath
the colonnadeshad a normalthicknessof 0.80 m., but in their northernpartsthe lateral foundations
were thickenedto 1.20 m. because of the greater depth due to the sloping ground level. Since the
stylobate could not have exceeded ca. 0.80 m. in width there is little possibility of a second step
below the stylobate.
None of the stylobatehas survived,nor has anything been recognizedfrom the superstructureof
the colonnades,neither columns,capitals,nor entablature.Restorationmust thereforebe little more
than guesswork.A clue, however, is providedby a peculiarityof constructionin the foundationsat
the southeast and southwest corners of the courtyardwhere protrudingspurs exist for arches set
diagonally across the portico. The most plausible explanation is that the diagonal arches were
intended to take up the thrust of arcuated constructionin both the north-south and east-west
colonnades. Similar spurs occur in the South Court at its northwest and northeast corners, al-
though, surprisingly,not in the south corners.In the SoutheastCourt no such constructionhas been
observed, and here presumably the entablaturewas trabeated. In the two larger courtyardsthe
arches in the colonnadeswere presumably of brickwork.In order to account for the excavators'
failure to find any remains of columns,John Travlos was preparedto considerthe possibilitythat
the arches had been carriedon rectangularpiers of masonry,and this is the constructionindicated
in the restoreddrawing of the North and South Courts (PI. 54). It must be recognized,however,
that such economicalconstructionwould have been very unusual in an importantbuilding of this
period, especially in one of its most prominent parts; the possibility should be admittedthat the
arches were supportedby round marble columns which provedirresistibleto those who looted the
Palace after its abandonment.
Masses of masonry projectingfrom the outer face of the south outer foundationof the North
Court mark the positions of niches presumably for statuary. Four may be restored with a size
suitable for single figures of life size. A larger niche facing the length of each of the lateral colon-
nades could have sheltereda pair of statuesor a sculpturalgroup. No sculptureof suitable date has
been found within the limits of the Palace.
Rooms protrudedfrom the sides of the North Court:one apsidal and one rectangularon the west,
two rectangularon the east.12In all four cases the walls have been stripped down to the founda-
tions, nor has anything of the superstructuresbeen found. We may assume, however, that all the
rooms are to be restored as exedrae opening on the colonnades through arched or colonnaded
fronts.13Such exedrae were a commonfeature of the palaestraein the great baths of Rome and the
12
Two short remnants of foundation projectingfrom the west wall of the Court near its southwest corner may
representanother small exedra (P1. 53).
13 Less probableis the restorationof doorwaysas envisaged,for instance,in the perspectiverestorationin Plate 55:a.
 102 V. THE PALACE OF THE GIANTS

provinces,but a model was closer at hand in the Libraryof Hadrian in Athens where two apsidal
and one rectangularexedra extendedout from either side of the court (PI. 51:b).
The two western exedrae were thrust out into a completelyopen space. The smaller of the two
was centeredon the west side of the North Court while its larger neighborwas squeezed into the
space that remainedto the north. This apsidal exedra would undoubtedlyhave lookedout into the
court through a colonnade.The thinness of the outer wall calls for wooden roofing rather than a
semidome.14
On the east side of the North Court the architecthad a less free hand. The proximityof the Pan-
athenaic Way ruled out the possibility of a semicircularexedra symmetricalwith that on the west
side. Farthersouth he was inhibitedby the presenceof a large monumentof early Roman date with
© American School of Classical Studies at Athens

a tomblike basement chamber,which was still respected.The solution was to constructnear the
For personal use only. License: CC-BY-NC-ND.

middle of the east side two small rooms sharing a party wall; the plan of the northernroom was
distortedout of respectfor the earlier monument.
A large square room, the foundationsof which came to light at the southeastcornerof the North
Court, is earlierthan the Palace, as shown by the fact that it extendedsouthwardbeyondthe south-
ern side of the North Court (PI. 53). The west wall of the room presumablyrestedon the surviving
foundationof the east wall of the Odeion.The south foundationof the roomwas overlaidby the gar-
den wall of the Palace. The room measuredinternallyca. 7.40 m. to the side. The concretefounda-
tions, although not identical to those of the North Court, are indubitablyalso Late Roman. The
structureconceivablyservedsome temporarypurposein the early part of the building program.

PASSAGE FROM THE NORTH COURT TO THE SOUTH COURT

Accessto the South Court of the complexwas by way of an openingin the middleof the south side
of the North Court (PI. 54) and thence through a rectangularvestibule (9) measuring internally
4.80 x 5.60 m. Traces in the mortar bedding for the threshold of the vestibule call for a pair of
columns. From the vestibule one proceededinto a larger rectangularchamber measuring 8.20 x
10.10 m. internally (10) and thence on southwardinto a semicircularcorridor22.60 m. in diameter
and ca. 4 m. wide (11). The inner side of the corridoris presumablyto be restoredwith a colonnade
borderinga semicirculararea open to the sky. The corridorwas paved with marble slabs as indi-
cated by imprints in the slight remains of the mortarbedding.The extremely ruinous state of this
interesting suite of rooms is regrettable,not least because it leaves us with no indicationof what
provisionwas made for the differencein level between the North Court and the South Court. This
differencewas considerableinasmuchas the stylobateof the North Court was 59.03 m. above sea
level while the floor of the semicircularcorridorlay at + 60.50 m. In the absenceof evidencewe
may restoretwo or three steps leading up from the vestibuleinto the large chamberand again from
the large chamberinto the semicircularcorridor.
The semicircularcorridorgave accessto a long gallery measuring38.80 m. in length and ca. 5 m.
in width (12). The south side of this gallery formedthe north wall of the South Court. The ends of
the gallery can be fixed from slight remains of their foundations,but its north wall was utterly
destroyedby those who robbedthe north foundationsof the Middle Stoa on which that wall had
stood. The survivingremains preserveno precise indicationof how the long gallery communicated
14
Beneaththe foundationfor the semicircularexedra is a concretefoundationfor the northwestcornerof an unfin-
ished rectangularroom projectingfrom the wall of the North Court; it measures ca. 8.20 m. east-west, ca. 3.70 m.
north-south (P1. 53).
 EXPLORATION AND IDENTIFICATION 103

with the south part of the complex. From the relationshipof the parts we may assume, however,
that through openings in the south wall of the gallery one might have enteredboth the peristyleof
the South Court and the Bath.

SOUTH COURT

The South Court was appreciablysmallerthan its northernneighbor.It was borderedon all four
sides by porticoessurroundingan open space measuring 19.60 x 22.40 m. Since the porticoeswere
4.70-4.80 m. deep, the overall dimensionsof the Court were ca. 30.60 x 33.20 m. The differenceof
2.80 m. in the length of the stylobate on the short and the long sides probably represents the
interaxial spacing of the columns; if so this indicates a plan with 8 x 9 columns. Although some
© American School of Classical Studies at Athens

11.90 m. of the south stylobateis preservedin place, it is of little help in refiningthe plan, since it
For personal use only. License: CC-BY-NC-ND.

consistsof a variety of re-used blocks,the cuttings in their tops dating from previoususe.
Apart from this run of stylobatelate pillagers have strippedthis part of the building down to its
bare foundations,leaving no remnantof the superstructure.Here, as in the North Court, arcuated
constructionseems more probable than trabeated. In the northwest and northeast corners of the
peristyle court are concrete foundations suitable for the support of the oblique arches intended,
perhaps here as in the North Court, to sustain the thrust of terminalarchesin arcuatedcolonnades,
although it is puzzling here, as in the North Court, why all cornersof the peristylewere not treated
alike. Slight remains to the east of the southeast corner of the peristyle (32) suggest an entrance
from the garden area.
Escape for rainwater was apparentlyprovidedby a terracottadrain that ran southwardfrom a
point near the middle of the south porticoto empty into the Great Drain (PI. 53).
The focal point of the South Court was the entrance to the large room (14), measuring
8.00 x 9.78 m., that openedon the south colonnade.This imposingchamberwas set on axis. It was
flankedon either side by a pair of smallerroomswith which there was probablydirectcommunica-
tion, although the present state of the foundationsdoes not permit the certainlocationof doorways.
That these five roomsconstituteda group may be inferredfrom the commonuse of flooringmadeof
square terracottatiles measuring 0.545 m. to the side (PI. 59:a). To east and west of this central
group the cornersof the South Court were occupiedby small roomsof equal size which presumably
opened directlyon the south colonnade.We have no evidenceregardingthe treatmentof the open-
ings into these various rooms,but we may assume that the great centralchamberhad a colonnaded
front.
The west side of the South Court was borderedin its northernhalf by the Bath (42-49), of which
more below (pp. 107-108). South of the Bath the theruins of Palae periodare extremelyexiguous,
so much so that it has been difficultto establish even the outlines of the structures.There appear,
however, to have been two buildings, one small (50, 51) and one large (52-54). In neither case is
there evidenceof directcommunicationwith the court.Thrust out into
thest"back yard",these build-
ings probably served some modestpurpose relatedto the domesticneeds of the Palace.
The west wall of the South Court was carried south to what had been the south wall of South
Stoa II but which now supportedan aqueduct(above,p. 14). In the angle were insertedtwo rooms,
one L-shaped and one rectangular(21). This pair of rooms,which communicatedwith each other,
was enteredfrom the gardenarea through a broaddoorway(PI. 54). The locationof the roomsand
the unusual width of the doorway (2.20 m.) have suggestedthe possibilityof a stable and carriage
house. Nothing in the way of stalls or mangershas cometo light, but the setting correspondsnicely
with the recommendationsgiven by Xenophon and Vitruvius: the stable should have a sheltered
 104 V. THE PALACE OF THE GIANTS

location;there should be roomfor exercisein the open;the place shouldbe within view of the owner
to enable him to keep an eye on the health of the animals and to reducethe dangerof theft.15

SOUTHEASTCOURT
Adjoiningthe east side of the South Court is the third majorelement in the plan of the Palace:the
SoutheastCourt (Pls. 60-63). The two courtsapparentlycommunicatedwith one anotherthrough
a doorway in the east wall of the South Court via Room 31. The Southeast Court was certainly
accessiblefrom outside through a doorway in its east wall near its southeast corner. The overall
dimensionsof this wing were 26.20 x 32.30 m. The open area of the court was square, 6.85 m. to
© American School of Classical Studies at Athens

the side, and was borderedon all four sides by colonnades3.70 m. wide. Although the base foun-
dations afford no certain evidence for the column spacing, a reasonable solution would be four
For personal use only. License: CC-BY-NC-ND.

columns to a side. Nothing remains of columns or entablature.The open court was paved with
terracottatiles 0.565-0.58 m. square. A well opened in the middle of the stylobate of the north
colonnadeof the court.
The north side of the court was borderedby an assemblageof seven roomsin two east-west rows
(24-30). The conditionof the foundationsdoes not permitthe sure locationof doorwaysso that the
arrangementon the plan is admittedlyarbitrary.Nor have we evidencefor the use of the individual
rooms except in the case of the northeasternmemberof the group (29). In the northeastcornerof
this room is preservedthe lower part of a latrine of the simple type familiar in the houses of the
period (above,pp. 33-34; P. 21 :f). The arrangementconsistedof an east-west trench,flooredwith
two large, concavepan tiles set against the north wall of the room.The inner side of the trenchwas
supportedby a wall of rubble masonrythe top of which was level with the floor of the room. The
trenchopenedat its west end into a rectangularterracottadrainthat camedown fromthe south and
continuednorthwardinto the eastern"gardenarea"of the Palace. The drain, in additionto serving
the latrine, presumablyalso carriedoff rainwater from the SoutheastCourt, a sensible plumbing
arrangementcommonlyemployedin antiquity.
Set axially on the south side of the Court was the principalroomin this sectionof the Palace (23).
Measuring 4.30 x 5.02 m., it was a good deal smallerthan the correspondingchamberin the South
Court (8.00 x 9.78 m.). We have no indicationof how the roomopenedon the south colonnade,but,
in view of the limited width of the room,probabilityfavorsan archedopening ratherthan columns.
The remaining foundationson the south side of the court permit the recoveryof an outline plan,
but they yield no indicationfor the placementof doorways.Hence the restoredplan on Plate 54 is to
some extent arbitrary.Most puzzling is Room 31, the angular "corridor"ca. 1.40 m. wide wrapped
aroundthe south side of Room 23. At its west end we assumethat it permittedpassagebetweenthe
South and the Southeast Courts. At its east end one might have expected it to join the east-west
corridor (33-35) that led in from the doorway in the east wall of the complex. But there is no
indicationof a doorway at the critical point, and a considerabledifferencein floor levels virtually
rules out the possibilityof continuitybetween the two corridors.
The mostprobableinterpretationof Room31 is that it housedin its southand east partsa stairway
leading up to the upper storeythat is to be restoredabovethe east blockof the SoutheastCourt com-
plex (see below, p. 106). This would have permitteda desirablemeans of communicationbetween
that upper floor, which appearsto have been the "mastersuite"of the whole establishment,and the
Xenophon, de re equestri Iv.i; Vitruvius, vi.6.4. For a good discussion of Greek and Roman stables, see
I5
Rostovtzeffet al., Dura Europos,pp. 73-75. J. K. Anderson(AncientGreekHorsemanship,Berkeley 1961) deals with
stable management in chapter VII. For detailed documentation,see G. Lafaye in C. Daremberg and E. Saglio,
Dictionnairedes antiquitesgrecqueset romaines,Paris 1875, II, i, s.v. equile, pp. 743-745; s.v. equitium,pp. 791-794.
 EXPLORATION AND IDENTIFICATION 105

South Court, not to mention the Bath adjoiningthat court. No trace of a stair has been observedin
Room 31. Some external evidence,however, supportsthe restoration.From the southeastcornerof
the room rectangular foundations of rubble masonry project, one eastward and one southward.
These conceivablycarriedbuttressesto counterthe thrustof vaults supportinga stairwayof masonry
rising in two flights. Sincethe structuresrestingon the rectangularfoundationshad been demolished
at a late date but while the Palace was still in use, the stair also may have been removedat the same
time, a time when the upper storeyof the east blockmay no longerhave existed.
The east block of the SoutheastCourt complex compriseda series of rooms (33-41), the floorsof
which lay at a level 1.10 m. below that of the SoutheastCourt. This suite was accessiblefrom the
© American School of Classical Studies at Athens

courtyardthrough a doorway near its southeast corner.The suite could also be reachedfrom out-
side the building through a doorway in its east wall. On entering through this door one descended
For personal use only. License: CC-BY-NC-ND.

by a flight of steps, some six or seven in number,of which the bottomtwo remain in place. The stair
was roughly put together of miscellaneous re-used blocks;the steps show heavy wear from foot
traffic. From the bottomof this stair a broadcorridorled westward up a gentle slope into the court-
yard (33-35).
Through the north wall of this east-west corridora doorway led into Room 37, a somewhat
broader north-south corridor3.30 m. wide, which served as a common lobby for the four rooms
that opened on it (38-41). Rooms 38-40 had earthen floors; their walls were unplastered.The
individualchambersdifferedmarkedlyfrom one anotherin both shape and dimensions.
The southernmost(38) was square in plan, 4.00 m. to the side; it opened on the lobby through a
narrow doorway.Next to the north was a largerroom (39) measuring4.00 x 5.20 m., with an open
front supportedby two columns,the square bases of which have left stains on the marblethreshold.
Within the room were found two unfluted column shafts of Pentelic marblethat probablystoodon
these bases. One measures2.39 m. in height, 0.325 m. in lower diameter;its workmanshipsuggestsa
date in the 4th centuryB.C. The other,with a height of ca. 2.475 m. and lower diameterof 0.295 m., is
probablyof early Roman date. Whereas both these columnshafts had obviouslybeen re-used in the
Palace, an anta capital found in the same area may well have been carvedfor the Palace (PI. 63:c).16
Its sides are adornedin low relief with acanthus leaves, both smooth and rough, alternatingwith
lotus leaves.The marbleis Pentelic,the carvingremarkablysure and fine. It is possible,indeedprob-
able, that both the column shafts and the anta capital derivenot from the groundfloorbut from the
upper floor,the plans of the two floorspresumablybeing substantiallyidentical.
Next we come to 40, a room of unusual plan: square front part enteredthrougha broaddoorway;
back of this an eastward projectionmeasuring 1.20 x 5.50 m. The north and south ends of the
extension to a length of 1.50 m. were roofed with barrel vaults built of brick, whereas the middle
part rose higher, conceivablyto serve as a kind of flue (PI. 61:a).
Finally, at its north end, the lobby broughtone to the largest and obviouslythe principalroom of
the suite (41). One entered through a doorway ca. 1.70 m. wide. The main area of the room had
inside dimensionsof 6.10 (width) x 6.70 m.; but the rear part, 3.50 m. in depth,was contractedto a
width of 4.55 m. The floor of the room lay on a filling of earth ca. 0.50 m. above roughly dressed

A 2268. A. Frantz, The Middle Ages in the Athenian Agora (Agora Picture Book No. 7), Princeton 1961, ill. 8.
16

Height 0.284 m.; resting surface 0.296 x 0.35 m.; top surface0.385 x 0.464 m. I am not aware of any close parallel to
this capital either in the scheme of decorationor in quality of workmanship. For approximationssee R. Kautzsch,
Kapitallstudien, Berlin/Leipzig 1936, p. 83, pl. 17, no. 244 (Corinth); A. Orlandos, 'H _vAoXorryo IaXalaLoxpr-
rLaVLKJl Bao-LXLK?
II, Athens 1954, pp. 318, 320, fig. 276 (Nea Anchialos, Basilica A). For a recent extended study of
this type of capital see V. Deroche, "L'acanthede l'Arcd'Hadrien et ses deriveesen Grece propre,"BCH 111, 1987,
pp. 425-453, again without close parallels.
 106 V. THE PALACE OF THE GIANTS

bedrock.This packing was surfaced with a thin layer of small field stones and broken roof tiles
carefully laid in north-south bands with east-west borderson the north and south sides (P1. 59:b).
The pebbles and sherds were beddedin crumblylime mortar;as found they were overlaidwith a
thin layer of clay. This can scarcelyhave been the intendedsurfaceof the floor. Most probablythe
layer of field stones and brokentiles was meant as underpinningfor a marble floor laid in bands.
The marblewould surely have been beddedon a thin layer of mortar.17The operationwould seem
to have been interruptedafter the laying of the layer of field stones and brokentiles. The layer of
clay may be regardedas an economicalway of making the floor usable. This then may be taken as
one more indicationof incompletionin the original project(below, p. 108).
Southwardof the east-west corridor,at the extremesoutheastcornerof the SoutheastCourt com-
© American School of Classical Studies at Athens

plex, is an octagonalroommeasuring7.04 m. fromside to side externally(36). The preservedlower

For personal use only. License: CC-BY-NC-ND.

parts of the wall show the most careful constructionto be found anywhere in the Palace (P1. 60).
Large, re-used poros blocks were set at intervals,and the interveningspaces were then filled with
brickwork.The walls are unique in the Palace also becauseof their great thickness:0.86 m. Such
constructionimplies unusual height for the octagon:it was undoubtedlya tower. The doorwayhad
neitherjambs nor threshold,and so, presumably,there was no door. The floorwas roughlycobbled
with small stones.Againstthe middleof the east wall openeda well, undoubtedlycontemporarywith
the Palace and convenientlysituatedto supply the needsof the suite to the north.
The high-rising tower was a familiar architecturalform in Late Antiquity, especiallyin villas.18
In our case, however,the idea may have been suggestedby the well-preservedTower of the Winds,
of Hellenistic date and just a little larger than the tower in the Palace: greatest width 7.80 vs.
7.04 m.
Various indicationstaken togetherattest the existence of an upper storey above the rooms in the
easternblock of the SoutheastCourt complex. In the first place it is hard to see why, if there was to
be only a single storey, its floor level should have been set so much lower than the floor level of the
adjoiningarea to the west or than the groundlevel outdoorsto the east. Moreover,the walls in this
area are unusually thick and built with special care; they were certainly capable of carrying a
second storey. Finally, a number of fragmentsof wall revetmentof coloredmarble and numerous
chunksof floor mosaicwere found among the debrisof abandonmenton the floorsof the roomsjust
described(P1. 63:a, b). Since there is no indicationthat either floor mosaic or wall revetmentwas
employedin these ground-floorrooms,that materialmay be assumedto have fallen from an upper
storey. The floor plan of this upper storey must have followed closely that of the lower, but the
preciousmaterialsnoted abovesuggest that the upper floor was the piano nobileof the Palace. The
rooms at this higher level would have been more open to summer breezes;they would also have
commandedsplendidviews eastwardtowardthe Acropolisand Mount Hymettos, northwarddown
the PanathenaicWay.
A comparablelayer of potsherdswas found beneaththe mortarimmediatelyunderlyingthe marblefloor slabs in
17

the orchestraof the Odeion of Agrippa: Thompson, "Odeion,"p. 59, pi. 42:a. Cf. also the underpinningfor the or-
chestrafloor of the Theater of Dionysos: E. Fiechter,Das Dionysos-Theaterin Athen I, Stuttgart 1935, pp. 55-60.
18 P.
Grimal, "Les Maisons a tour hellenistiqueset romaines,"Melanges d'archeologieet d'histoiredel'Ecolefran-
faise de Rome 56, 1939, pp. 28-59; R. Paribeni, "Ledimoredei potentioresnel basso impero,"RM 55, 1940, pp. 131-
148; Dunbabin, Mosaics, pp. 119-123, figs. 109, 111-113 (villa mosaicsat Carthageand Tabarka); McKay, Houses,
Villas and Palaces, pp. 236-237, pl. 76, fig. 75. Among the Late Roman additionsto the Forum Baths at Ostia is an
octagonal room with wide windows intended undoubtedlyas a pleasant place to sit and watch the passing scene, al-
though the room was apparentlyonly one storey in height: Meiggs (footnote211 above,p. 89), p. 414, fig. 29. Octago-
nal foundations,apparentlyfor a tall structure,are emergingin the Villa of Theseus at Nea Paphos on Cyprus:W. A.
Daszewski, Dionysos der Erloser, Mainz 1985, p. 17, fig. 1; idem, Report of the Departmentof Antiquities, Cyprus
1984, p. 296.
 EXPLORATION AND IDENTIFICATION 107

No trace of a stairway leading to an upper storey has come to light within the east block of the
SoutheastCourt complex. It appears thereforethat the upper floor was accessibleonly by the stair
that we have found reason to restorein Room 31.

BATH
The small but elegant bathing facility at the northwestcornerof the South Court (Pls. 64-66:a) is
undoubtedlya part of the original design of the Palace. This is shown by the interlockingplans as
well as by the identity of materials and masonry. Although the foundationsare preservedalmost
throughout, thus assuring the overall plan, they stand high enough in only one or two cases to
© American School of Classical Studies at Athens

indicate the position of doorways;hence the relationshipamong rooms as proposedin the restored
plan (Pls. 53, 54) is all too often arbitrary.
For personal use only. License: CC-BY-NC-ND.

The heated area of the bath (caldarium)is readily recognizablein the southwesternpart of the
block, a location recommendedby Vitruvius (v.io.i) centuries earlier (43-45). This position as-
sured full exposure to the sun in the afternoons,the usual time for bathing. The adjoiningroom to
the north (42), a long rectangle in plan, has no provision for heating, nor have the three water
basins that opened off its north side. This will have been the cool or cold room (frigidarium).The
dressing room (apodyterium)is probably to be recognized in the large room (46) between the
caldarium and the South Court, while the series of three small rooms to the south (47-49) pre-
sumably servedsome of the practicalneeds of the bath and its personnel.
How the Bath proper was entered is problematic.We may assume, however, that the principal
approachwas from the South Court, probablythroughRoom 46. But the fact that the long corridor
north of the South Court (12) was carriedwestward to overlapthe Bath stronglysuggeststhat one
could enter the Bath also from the corridor.
All three rooms of the caldariumwere heated by hypocausts.Each compartmentreceivedhot air
from a furnace set at a low level against its south side, while the west room,much the largest, had a
second furnace on its west side. The hypocaustsbeneath all three rooms communicatedwith each
other (PI. 53).
The pillars supportingthe floorsof the heatedroomswere of brick,some round,some square.All
the tiles disappearedduring World War II, but a small remnantof the flooringhas survivedin the
northeastcornerof Room 45 (PI. 65:a). The constructionwas normal:the brickpiers carriedlarge
tiles above which was a layer of field stones 0.15 m. thick beddedin gray mortar;on this rested a
layer of mortar0.10 m. thick containingmany fragmentsof tile and much poundedtile; this served
as the beddingfor floor slabs of Pentelic marble 0.02 m. thick.
The walls of this part of the bath were also revettedwith thin slabs of Pentelicmarblebackedby a
layer, 0.04 m. thick, of mortar containing pounded brick, continuouswith that below the marble
flooring.At the sides of the doorway leading into the central room of the caldariumthe lower ends
of two marblejambs have survivedin situ (PI. 66:a).
It may be assumed that the three niches opening off the western room in the caldarium(44) and
the one at the south end of the eastern room (45) containedwarm-water basins for bathing. The
water would have been warmed by the hypocaustswhich ran continuousbelow the niches.
The massive nature of its foundationsindicates that the central, round room (43) was covered
with a dome provided,no doubt,with an opening (oculus)in its top and a damperto controlventila-
tion. The ground plan leaves little doubt that the caldariumcomprisedtwo distinctparts:the cen-
tral and west rooms (43, 44) that intercommunicatedand were enteredfrom the frigidarium (42),
through a doorway certainly attested in the semicircularniche, and the smaller eastern suite (45)
that seems to have had a separate entrancefrom the frigidarium.It is tempting to suppose that the
two suites were intendedrespectivelyfor men and women as in the Baths of Pompeii.
 108 V. THE PALACE OF THE GIANTS

In the large east-west roomthat we take to be thefrigidarium(42) a nearly squarearea is marked

off at each end by a transversefoundationon which we have restoredpairs of columnsto produce
exedra-like spaces. The three water basins that borderedthe north side of the large room were
revetted,both floors and walls, in Pentelic marble applied in the same technique as that described
for the caldarium.The massivesemicircularniche clearly implies a half dome over the niche.
Water was availablefor the bath in close proximityand at a suitable level (55). The supply came
throughthe aqueductof the Antonineperiodthat had been restoredin the 5th centuryon top of the
wall originallyerectedin the 2nd centuryB.C.by the buildersof South Stoa II (above,p. 14). Water
was drawn from the aqueduct into a basin erectedabove the ruins of South Stoa II at its west end
© American School of Classical Studies at Athens

(PI. 62:b). The basin is square, measuringca. 4.60 m. to the side internally.The walls are massive:
ca. 1 m. thick, in masonryidenticalwith that of the Palace and supportedby buttresses.The inlet is
For personal use only. License: CC-BY-NC-ND.

not preserved,but the outlet may still be seen: a terracottapipe 0.10 m. in diameter internally,
which leads off at the level of the floorof the basin in the directionof the Palace bath ca. 40 m. to the
northeast.The interveningarea has been much disturbedin later times, nor have we any clue as to
the precise point at which the water enteredthe bath.19
Drainage from the bath was providedby a terracottachannel the pillaged courseof which could
be traced (PI. 53). Passing from east to west throughthe servicearea south of the bath, the channel
swept around the southwest cornerof the bath and then swung toward the northwestto pour into
the ancient Great Drain beneaththe west edge of the Agora.20

MARKS OF INCOMPLETION

Severalindicationsleave little doubtthat the Palacewas not quite finishedin accordancewith the
original design, although it was certainlyused for at least a century.The most strikinglack was the
absence of an adequate approach to the facade of the North Court. As noted above (p. 100), the
differencein level between the Court and the groundto the north called for a broad stair perhaps
rising to a terraceimmediatelyin frontof the facade.Since no traceof such has cometo light, we can
only supposethat one ascendedto the courtover a simple earth ramp, an approachunworthyof the
monumentalconceptionof the entrance.
Although marble revetmentwas used freely on the walls of the Bath and apparentlyalso in the
piano nobile of the east block of the SoutheastCourt complex, the absenceof wall plaster has been
noted in places where it might have been expected in the ground-floorrooms of that same block
(above,p. 105).
Mosaic flooringseems certainlyto have been used in the piano nobile(above,p. 106), but we have
found reasonto believe that an intentionto lay a marblefloor in one of the ground-floorrooms (41)
of the same blockwas not carriedout (above,p. 106). The very limiteduse of mosaicflooringis also
perhaps attributable to incompletion. In a building of such scale and obvious importance it is
scarcelyconceivablethat at least the principal rooms, such as the long corridor(12) and those fac-
ing on the south sides of the South and Southeast Courts (14, 23), were not intended to receive
mosaicflooring.The terracottatiling that was actuallyused in Rooms 14, 16, and 17 seems a cheap
substitute.

A similar reservoirassociatedwith the public bath at the west foot of the Areopagus(above,pp. 31-32) must have
19

drawn from the continuationof the Antonine aqueduct although nothing remains of the aqueduct between the two
reservoirs.
Plate 53 shows three terracottadrains that flowed southwardfrom the area of the Palace into the Great Drain.
20

The two largerantedatethe Palace;the smallest,startingat the south edge of the South Court (13), presumablycarried
off the rainwaterfrom that court.
 EXPLORATION AND IDENTIFICATION 109

A striking deficiency in an otherwise well-organized establishmentis the lack of an adequate

latrine. The primitive facility found in Room 29 of the SoutheastCourt complex seems completely
out of scale with its context. It is possible, however, that the remains of a more adequate facility
have eluded detection in the very ruinous area to the west of the South Court (47-49). The long
terracottadrain that was carriedaroundthe Bath to empty into the Great Drain had its beginning
in this area (above, p. 108). There is no apparent connectionbetween the Bath and the terracotta
drain, but the drain could well have serviceda latrine in one or other of Rooms 47, 48, or 49. So
placed, the latrine could have drawn on the same sourceof water as the Bath.
© American School of Classical Studies at Athens

STRUCTURES TO THE NORTHWEST OF THE PALACE

Brief reference has already been made to the contemporarybuildings to the northwest of the
For personal use only. License: CC-BY-NC-ND.

Palace (above,p. 97; PI. 6). Distressingly little more can be added.The northwardextensionof the
west wall of the North Court to the very edge of the foundationsof the Temple of Ares clearly
implies the continuedexistence of some structureson the site of the temple. Although, as we have
seen, much of the marble ceiling was stripped away after the Herulian sack, the body of the
building may well have continued to stand on the site with or without columns. Before the 5th
century it would, of course, have ceased to function as a pagan sanctuary,but we have nothing to
indicatewhat use it servedsubsequently.The existenceof the connectingwall does, however,imply
associationwith the Palace.
The building (Square Building) that was crowdedclose against the north side of the site of the
Temple of Ares is almostequally enigmatic.Ceramicand numismaticevidenceindicatesa dateearly
in the 5th century,closelycontemporary,that is, with the Palace.A few architecturalfragmentsfrom
the Stoa of Zeus Eleutherios imbeddedin the foundationsof the building probablyresulted from
damage done to the Stoa by the Visigoths in their raid of A.D. 396. From the fabricof the building
only the foundationsremain;they are of massiveconcretesimilarto that used in the Palace. None of
the floor has survived,nor have any elements of the superstructurebeen recognized:the area was
very thoroughly stripped by the house builders of the llth-12th centuries. Measuring ca. 25 x
27.50 m., the building was curiously symmetricalin plan about an east-west axis. A central court
colonnadedon the north, south, and west sides was borderedby 11 roomsof varioussizes, the north
and south suites being identicalwith one another.The entrancewas presumablyat the middleof the
east side, flanked to north and south by the slightly projectingends of the north and south suites.
Excavationfailed to reveal any traceof a foundationbetweenthose two projections.
The curiously symmetricalplan would be extremelyunusual in a privatehouse, nor is there any
distinctivelydomestic feature about the building. The layout is equally unsuited to a commercial
establishment,in which the shops tend to be closely uniform in size and plan. We are thus almost
driven to conjecturesome public function, conceivablyas an office building, but whether for civic
use or as an adjunctto the Palace we can scarcelyhope to know.21
The massivesquare base that probablysupporteda columnarpedestalfor a statue ca. 25 m. to the
east of the building of 11 rooms is too remote and is wrongly oriented to be associatedwith that
building. As for the circularor semicircularstructure("RoundBuilding")north of the building of
11 rooms, we can only hope that furtherstudy will yield clues to plan and purpose.
An intriguing parallel for the building of 11 roomsis to be found in the "Maisondes Jardinieres"at a conspicuous
21

place just outside the northeast corner of the forum of Timgad: S. Gsell, Les monumentsantiques de l'Algerie, Paris
1901, I, p. 213, fig. 36:F; Princeton Encyclopedia,s.v. Thamugadi, p. 900 (J. Lassus); P. MacKendrick,The North
African Stones Speak, Chapel Hill 1980, pp. 233-240. It has been repeatedlysuggestedthat this may have been an
officialresidence,perhapsof the legate of Legio III Augustawho also actedas governorof the provinceof Numidia. But
the plan of the building, very similar to ours, looks no more suitable for a residencein Algeria than for one in Athens.
 110 V. THE PALACE OF THE GIANTS

In general the way in which this group of three contemporarystructureswas thrust, cheek by
jowl, into the space betweenthe Palace and the line of the PanathenaicWay on the north,combined
with the linkage betweenthe Palace and the site of the Temple of Ares, suggeststhat all four may be
parts of one building program.

PLAN AND PURPOSE OF THE BUILDING

The southernhalf of the great complex differs little from the large houses on the north slope of the
Areopagus (P1. 6: A, B, C). Like them it representsthe final developmentof the typical Athenian
house, the history of which can be traced from the classical period into Late Antiquity (above,
© American School of Classical Studies at Athens

p. 34). The core element was a colonnadedcourtyardaroundwhich were groupedroomsof various

For personal use only. License: CC-BY-NC-ND.

sizes and shapes as requiredby the needs of the household.Normally the largest room in the house
was placed on the long axis of the courtyardfrom where it commandedthe best view of the court.
The entranceled from the street directlyinto the court and tendedto be modest.Sanitaryfacilities
were of the simplest, and the inclusionof a bath within the house came late, as in House C where it
was added only in the final period (above, p. 87). A house with a long history might eventually
comprisenotjust one but two or even three peristyleunits (Houses A, B, C). There is little evidence
for more than one storey in the normal house, nor has any trace been observedof enclosedgarden
areas outsidethe building proper.
The residentialpart of the Palace obviouslyfollows this same old tradition.In size it differslittle
from Houses B and C. The inclusionof a bath in the original designof the Palace is in keepingwith
its relatively late date. The existence of a partial upper storey is only slightly unusual. More re-
markableis the spacioussetting with enclosedand presumablyplanted areas on all sides.
But the feature that most clearly distinguishes the Palace from the more or less contemporary
houses on the slopes of the Areopagus,and indeedfrom those found anywhereelse in Greece, is the
North Court, an integralpart of the original design. The conceptof a peristyleon this scale with its
exedrae both rectangularand apsidal is utterly foreign to the traditionof domesticarchitecturein
this part of Greece. Equally strangeto that traditionis the studiedmonumentalityof the facadeand
main entrance. One is reminded of the much discussed mosaic representationin the Church of
Sant'ApollinareNuovo22in Ravenna dating from A.D. 490 (P1. 67:c). The extremely prominent
tripartite facade clearly labeled "Palatium"is generally acceptedas referringto the palace of the
builder of the church, the Emperor Theodoric. Although the precise architecturalinterpretation
may be debated,there can be no doubtthat the strikingfacadeis to be taken as symbolicof the seat
of imperial power.
As observedalready (above, p. 100), the facade of the Giants had much in common with tri-
umphal arches of the most developedform. Thus it may be supposedto have evokedsome associa-
tion with the emperor,who was normallythe objectof the honorrepresentedby the arch. Inasmuch
as a triumphal arch, or an adaptationof an arch, commonlymarkedthe entranceto an important
public area such as a city, sanctuary,or forum,the choiceof this architecturalform for the entrance
to the great court of the Palace investedthe establishmentwith a certainaura.23

E. Dyggve, Ravennatum Palatium Sacrum. La Basilica ipetrale per cerimonie. Studii sull'architettura dei palazzi
22

della tarda antichita, Copenhagen 1941; W. F. Volbach, Friihchristliche Kunst, Munich 1958, pp. 28-29, pl. 152;
Krautheimer, Architecture, pp. 196-197, figs. 148-149; G. de Francovich, II Palatium di Teodorico a Ravenna, Rome
1970, passim; N. Duval, Corsi di cultura sull'arte ravennate e bizantina XXV, 1978, pp. 93-122.
23 On these aspects of the triumphal arch, see H. Kahler, RE VII a, 1939, s.v. "Triumphbogen,"cols. 472-474;
M. Pallottino,EAA I, 1958, s.v. "arcoonorarioe triomfale",p. 590.
 PLAN AND PURPOSE OF THE BUILDING 1ll

Equally telling is the placement of the Palace in what must have been still recognizableas the
center of civic life in the days of Athens' greatness.Significantalso is the relationshipbetween the
Palace and the PanathenaicWay at a time when their national festival, and especiallythe Proces-
sion, still meant a great deal to Athenians. We have noted how the fagade of the North Court
dominatedthe ancient course of the Procession,while the large room that is to be restoredin the
upper storey at the northeastcornerof the SoutheastCourt complexmust have seemedlike a "royal
box" designed expressly for watching the processionas it approachedfrom the north, passed, and
made its way up toward the Acropolis. One is remindedof the fact that many of the Roman Im-
perial palaces adjoined circuses: the Palatine in Rome, the palace of Diocletian in Antioch as
© American School of Classical Studies at Athens

described by Libanius, the palaces in Constantinople,Milan, Sirmium, and Trier.24The Pan-

For personal use only. License: CC-BY-NC-ND.

athenaic Way had provided the track for some of the most spectacularequestrian events in the
Panathenaicfestival,25and still the Processionitself was regardedas a great spectacle.
Even on normal days the occupantsof the piano nobile in the Palace enjoyedan impressiveview
of the Acropolis with Mt. Hymettos in the distance. From the upper windows of the gazebo-like
octagonal tower one could also overlookthe rest of the Palace to savor the view of the Temple of
Hephaistos then, as now, rising against the backgroundof Mt. Parnes. Here again one is reminded
of the attention paid to the location of Imperial and official palaces: in Antioch the palace over-
lookedthe Orontes, in Spalato the Adriatic,at Dura-Europosthe Euphrates.26
Anotherfactorthat argues against our complexbeing a privateestablishmentis its sheer size. We
are woefully ignorant, to be sure, of villas in Greece, but there appears to be nothing of this scale
known at present.Even HerodesAtticus'favoritevillas at Marathonand Kephissiaare represented
only by a few scatteredsculpturesand inscriptions.27
In view of the flourishing state of the schools of philosophy in Athens of the 5th century, one
might ask whether the Palace might have served as the seat of one of the schools. The only school
that could be consideredat this period,however,was the Neoplatonic,and the residenceof the head
of that school is known to have stoodto the south of the Acropolis(above,pp. 42-44). In view of the
still ambiguousstanding of the Christian church in contemporaryAthens, it is far from likely that
the complex could have been intended as a bishop's palace. Equally remote is the possibility that
any civic officialwould have been providedwith such a palatial seat in the Athens of Late Antiqui-
ty. Nor is there anything to suggest a military presencein Athens at this periodthat would havejus-
tified residentialaccommodationof this nature.
Can any more plausible function be proposedfor our building? Perhaps the most probableuse
was as an officialresidence,maintainedby the Imperialgovernmentand intendedfor the occasional
accommodationeither of the emperorhimself or of high-rankingImperialofficialswhen on tours of
duty. Such establishmentsare the subject of several edicts in the Theodosian Code which were
promulgatedin a period (A.D. 313-348) close in time to the constructionof our building.28There is
24
Ward-Perkins, Architecture, p. 450.
25
E. Vanderpool,"Victoriesin the Anthippasia,"Hesperia 43, 1974, pp. 311-313; AgoraGuide3,pp. 104-106.
26
W. L. MacDonald in his discussionof the significanceof Domitian's Palace in Rome emphasizes the splendid
location of the site in relation to the city: The Architecture of the Roman Empire, New Haven 1965, pp. 69-74.
27 AulusGellius,Noctesatticae1.2.2;Philostratus,
VSn.i (562);Graindor(footnote4 above,p. 16), pp. 115, 147,
227-228; S. Karusu,"DieAntikenvomKlosterLukuin derThyreatis,"RM 76, 1969,pp. 253-265;W. Ameling,
HerodesAtticus,II, Inschriftenkatalog,
Hildesheim/Zurich/NewYork 1939, In a recentgeneralaccountof villas,
includingthoseof Late Greece
Antiquity, doesnot appearamongthe many"villa-bearing" districts:J. Percival,The
RomanVilla:An HistoricalIntroduction, Berkeley/LosAngeles,1976.
28 TheodosianilibriXVI cumconstitutionibus T. Mommsen,ed., Berlin1954.The translationsof
sirmondianis2,
passages below are taken from Pharr. Cf. T. Mommsen, "Praetorium,"Hermes 35, 1900, pp. 437-442. For the
 112 V. THE PALACE OF THE GIANTS

some fluidity in nomenclature.The buildings are designatedsometimesas palatia, sometimesas

praetoria,this at a time when a palatium might house someoneotherthan an emperorand a praeto-
rium was no longerexclusivelythe residenceof a militarycommander.In Edict VII. I 0.2 (A.D. 407) it
appears that the term praetorium is construedas the permanent residence of the governorof a
province while palatium denotes an official residence for occasional use. In an edict of A.D. 405
(vii. i 0. i) the buildings are referred to simply as "sacred dwellings" (sacrae domus).
The primarypurposeof the institutionwas undoubtedlyto provideconvenientand suitabletem-
porary accommodationfor high-ranking governmentofficials. But the emperors were also con-
cernedto protectlocal authoritiesfrom the extravagantdemandsof lesser functionaries.Thus in an
edict of A.D. 406 (vii. I . i) municipal councilsand municipalitieswere relievedof the necessity "to
© American School of Classical Studies at Athens

heat any private baths for the use of tribunesor of minor counts. . . . this serviceis grantedonly to
For personal use only. License: CC-BY-NC-ND.

counts with the rank of illustriousand to mastersof soldiers,if they should so wish."29
The governorsof provincesand the vicarsof the prefecturewere chargedwith the maintenanceof
the palaces (vii.io.i of A.D. 405). These officials were responsible not only for maintaining the
buildings in good condition but also for seeing that admission was refused to any unqualified
persons who boldly relied "on high rank or fortune."The edict continues,"Of course we judge it
unnecessaryto prohibit horses that are not Ours from the stablesof Our palaces."The question of
responsibilityfor maintenancehad been a matterof long standing.Alreadyin A.D. 396 (xv.i.35) it
had been decreedthat "if any palace (palatium),officialresidenceof a governor(praetorium),state
storehouse,or stable and shelteringplace for public animals should fall into ruins, such structures
shall be repairedout of the resourcesof the governors."
Athens of the 5th centurywas, to be sure, neither an administrativecapital nor a military base.
The governorof the provincehad his seat in Corinth,while the Prefect of Illyricum, under whom
the provincewas included,would presumablyhave spent most of his time in faraway Sirmiumor,
later, in Thessaloniki. Consequentlyone would not expect to find a permanentresidence(praeto-
rium) in Athens. But prefects, governors,and many other high-ranking Imperial officials must
have been obligedto make officialvisits to Athens fromtime to time, while some of them, in keeping
with their tastes, undoubtedlyfound it possible to spend longer periodsbecauseof the cultural and
social climateof the ancient city.
On such occasionsour building would have servedadmirablythe convenienceof both the visitor
and the host city. The east blockof the SoutheastCourt complexofferedcomfortablelodgingfor the
guest and his immediatecircle with space also for a permanentstaff in the SoutheastCourt proper.
The South Court complexcomprisedmodestbut adequatefacilitiesfor formalmeetingsand meals.
Relaxation was readily available in the well-furnishedbath, and, if we are right in our identifica-
tion of the rooms outside the southwest cornerof the South Court (21), the guest's horses could be
stabled in a safe place with plenty of room for exercise. The North Court, we may assume, would
have servedfor large public receptionsor ceremoniesin connectionwith an officialvisit. In between
such special occasionsthe same court with its spacious colonnadesand exedrae would apparently
have been open to the citizens, a token revival of the great days of old when the whole Agora, sur-
roundedby its splendidstoas, had been the proud preserveof the Athenians.

distinctionbetween principia and praetoriawhen the terms are applied to military establishments,see R. Fellmann,
Die principia des Legionslagers Vindonissa und das Zentralgebaude der romischen Lager und Kastelle, Brugg 1958,
esp. pp. 89-92.
29 For troubles brought on the provincials in an earlier period by high-ranking officials, their staffs, and especially
their wives, see Tacitus, Annales 111.33-34. The technical term for the staff of a provincial governor was cohors; see for
example Juvenal, VIII.I27. For travel arrangementsof the wealthy and important,cf. L. Casson, Travelin the Ancient
World, London 1974, pp. 180-181.
 PLAN AND PURPOSE OF THE BUILDING 113

From the Codex Theodosianuswe know who was responsiblefor the maintenanceof such estab-
lishments, but who paid for their construction?It was conceivablysome governmentdepartment,
more probably, however, some wealthy individual. We have already seen examples of generous
benefactionsmade to Athens, as to other ancient Greek cities, by high-ranking Imperial officials
who had some personal interest in Greek culture, and one possible name has been mentioned
(above, p. 65). As noted above (p. 65), the discovery of an imposing portrait statue of a high
Imperial official in close proximity to the Palace is suggestive of its Imperial connections
(PI. 66:b).30In Greece such honorswere normallyaccordedeither in gratitudefor or in expectation
of substantialbenefactions.
© American School of Classical Studies at Athens

Whatever the identity of the donor, the choice of site and many aspectsof the architectureof the
Palace bespeaka close and sympatheticknowledgeof the Greek, more particularlyof the Athenian
For personal use only. License: CC-BY-NC-ND.

scene, and a nostalgic view of the old city's glorious past. The emplacementof the complex in the
very middle of the ancient Agora and in very close relationto the PanathenaicWay is too preciseto
have been accidental.Symptomaticalso was the decisionto resurrectand re-use the "Giants"in the
most prominentposition in the new structure.In the early 5th centurythere must have been some
persons in Athens still aware that these great statues had been carved in the Antonine period as
characteristicallyclassicizing adaptationsof pedimental figures of the Periclean Parthenon. We
have noted also the possibilitythat the combinationof rectangularand semicircularexedrae in the
North Court was inspired by similar features in the Library of Hadrian which was itself under-
going restorationat about this same time (above, p. 63). Likewise the octagonal tower that rose
above outheast cornerof the Palace looks like a reflectionof the Tower of the Winds.
Nor did the architectwho drew up the overall plan of the Palace need to go outside Greece for
models. As we have already noted (p. 110), the strictlyresidentialpart of the complex is little more
than a well-organized and slightly larger than normal example of a traditional Athenian house
type. Peristyle courtyardsof large scale had been familiar in Athens at least from the Hellenistic
period. The practiceof combininga large peristylewith anotherarchitecturalunit had also been in
vogue for long in Greece, notably in the planning of the palaestra-gymnasiumtype of complex.
Currently, i.e., in the 5th century, a similar combinationof elements was being employed in
Greece in the design of early Christian churches.31In a number of these buildings the sanctuary
properin the form of a basilicawas approachedthroughan atrium,a large courtyardset againstthe
30S 657. Agora I, no. 64. Preservedheight 1.33 m., i.e.,
approximatelylife size. The statue was found in 1936 built
into a modernwall ca. seven metersnorthof the northeastcornerof the Palace. The associationwith the Palace is not of
course certain, but such an ungainly mass of marble is not likely to have been carriedfar for re-use. In the 5th century
the only conceivablecontext in this area for such
the a statue is Palace. For its more preciselocationthe only clue is the
treatmentof the back, which was only roughly blockedout and certainlynot meant to be visible. The loss of attributes
precludesa sure identificationof the subject'srank beyondthat proposedin the original publicationon the basis of the
costume:"the portrait of an unidentifiedsenator".The statue appears to date from the late 5th century,too late to be
connectedwith the foundationof the Palace.
In the National Museum in Athens is a portraitbust of a beardedman, N.M. 423, reducedin antiquity from a full-
length statue, that has been datedby comparisonwith a consulardiptychof A.D. 428 to about that time, close, that is, to
the date of foundation of the Palace. See P. Graindor, Byzantion 1, 1924, pp. 245-250; J. Kollwitz, Ostromische
Plastik der theodosianischenZeit, Berlin 1941, pp. 91,104,106,125-128, pl. 41; AgoraI, p. 80. The place of findingof
the bust is not recorded.If by chance it could be associatedwith the Palace the damage that presumablyled to its re-
duction from full length to bust might have occurredin the Slavic raids of the 580's that are well attestedin the area of
the Agora (above, p. 93).
31On the plans of basilicas of the 5th century in Greece, see G. A. Soteriou, Xplo-rLavLKil Ka'LBvCavrtv' 'ApXaLo-
Xoyia, Athens 1942, pp. 290-319; Orlandos (footnote 16 above, p. 105), pp. 89-151; Krautheimer,Architecture,
pp. 90-101, fig. 34 (Epidauros, ca. A.D. 400), fig. 35 (Nea Anchialos, Basilica A, ca. A.D. 470), fig. 37 (Nikopolis,
Basilica A, 6th century), fig. 38 (Lechaion,St. Leonidas,A.D. 450-460[?]), and pp. 518-527.
 114 V. THE PALACE OF THE GIANTS

west frontof the basilicaand colonnadedon three sides.Occasionallythe entranceto the atriumwas
given a modestlymonumentaltreatment.In the atrium were performedthe matins,the collectionof
offeringsand otherpreliminarypartsof the Christianservice,afterwhich those qualifiedproceeded
into the churchproper while catechumens,repentantsinners, and the mentally ill remainedin the
atrium. The passage from the atrium into the body of the churchwas normallyeffectedthrough a
long, narrow corridor,the narthex, which played the same role as the long corridor(12) in our
Palace. Such similarities in architecturallayout are not to be construedas having any religious
significance.At a time when very few substantialbuildingsother than churcheswere being erected
in Greece it would be strangeif the gifted architectwho was responsiblefor our Palace had not also
designedsome churches.
© American School of Classical Studies at Athens

The search for significantparallels for our establishmentis difficult.32Those buildingsthat can
For personal use only. License: CC-BY-NC-ND.

with certaintyor even with probabilitybe identifiedas officialresidencesare few in number,and of

those few a large proportionare so ruinous or so little exploredthat their plans are far from com-
plete. The difficultiesof the searchare exacerbatedby what we know fromthe Codex Theodosianus
of the checkeredfortunes of such establishments:neglect and deterioration,repair and re-assign-
ment, sale at public auction.33Furthermore,as we have learned from the study of the Athenian
building, one must reckon on the possibility of strong influence from local tradition so that com-
parison among buildings in variousparts of the Empire calls for caution.
It would be pointlessto attemptcomparisonwith establishmentsthat are known to have been the
permanentresidenceof the emperorwhether in Rome, Ravenna,Milan, Constantinople,Cologne,
Trier, Sirmium, Split, Salonica,or Antioch. Quite apart from the fact that we have no recordof an
emperorresidingin Athens in the time of the Palace, the Athenianestablishmentcould not possibly
have accommodatedan imperial household.
A more appropriatecategoryfor comparisonis that of official residencesfor governorsof prov-
inces or for commandingofficers (duces) of regions. The most promisingexample is an establish-
ment at Gortyn in Crete, which is attestedepigraphicallyas the seat of the governorof the province
of Crete and Cyrene at a time when Gortyn was the principal city of Crete.34Despite repeated
explorationand study many uncertaintiespersist regardingthe plan and the very complexchronol-
ogy extending from the Augustan period into Late Antiquity. As in our establishment,however, a
clear distinctioncan be discernedbetweena residentialpart, includinga bath, and an officialpart in
which the role of our North Court seems to be filled by a large basilica-likehall.
The Polish excavators of the "House of Theseus" at Nea Paphos in Cyprus have tentatively
32On palaces in general, see Swoboda, Palste3; McKay, Houses, Villas and Palaces, with extensive bibliography
pp. 259-268. Many valuable observationsare to be found in Ward-Perkins,Architecture.N. Duval has expressed
gravedoubtsas to the existenceof anythinglike a "standardplan"for palacesin Late Antiquity:"Existe-t-ilune 'Struc-
ture Palatiale' propre a l'AntiquiteTardive?"(Le SystemePalatial en Orient,en Greceet a Rome [Actesdu Colloque
de Strasbourg19-22 Juin 1985], E. Levy, ed., Leiden 1987, pp. 463-490).
On villas of Late Antiquity, especially in relationto their historicalsetting, see Percival (footnote27 above,p. 111).
Wilson (Piazza Armerina,pp. 69-85) gives an excellent brief account of Late Roman villas with many plans. J. J.
Wilkes (Diocletian'sPalace, Split: Residence of a Retired Roman Emperor, Sheffield 1986), in his discussionof the
function of the "Palace",concurswith N. Duval in regardingthe building at Split as more nearly the equivalentof a
modern"chateau"than of the palace of a reigning emperor.
R. Fellmann ([footnote28 above, p. 112], pp. 90-92), in discussing the organizationof a Roman military camp,
emphasizesthe distinctionbetween the principia as the administrativequarterand the praetoriumas the residenceof
the commandingofficer.
33 CTh X.2.I and 2.
34 A. M. Colini, AeAr 29, 1973-1974, B' (1979), pp. 908-912, fig. 2 (plan); PrincetonEncyclopedia,pp. 362-363

(K. Branigan); I. Sandars, Roman Crete, Warminster 1982, pp. 77-80, 158; Sodini (footnote 128 above, p. 34),
pp. 343-344, with bibliographyand plan.
 PLAN AND PURPOSE OF THE BUILDING 115

proposed to recognize in that building the official residence of the proconsul of the province of
Cyprus.35The Cypriotebuilding (2nd-5th centuries)containsmany more roomsbut is more com-
pact than the Athenian; it consists of deep ranges of rooms enclosing all four sides of a large rec-
tangular courtyard.Although no specificevidenceof the officialnature of the establishmenthas yet
come to light, the proposedidentificationis supportedby the great size of the establishment,unique
in Cyprus, by the distinctionbetween modestquarterssuitable for a permanentstaff and a suite of
larger, more formal rooms magnificently floored with mosaics that might well have served as a
4"stateapartment".Perhaps also significantis the insertionat a late date in the history of the build-
ing of a formal, three-doorwayentrance.
At Apollonia, the one-timeport of Cyrene, a building dating from ca. A.D. 500 and much the
© American School of Classical Studies at Athens

largest in the city has been tentatively identifiedas the official residenceof the governorof Libya
For personal use only. License: CC-BY-NC-ND.

Pentapolis.36Here an obviouslydomesticpart is separatedby a roofedpassagewayfrom a block in

which rooms clearly intended for formal use are groupedarounda colonnadedcourt. The interior
decoris as austere as that of the Cypriotebuilding was lush. Despite the lack of specificevidencethe
identificationseems
onvincing.
Much the same may be said of the "House
of the
"Governor" in Volubilis, a Roman municipium
in Mauretania.37The house stood in the fashionable northeast quarter of the town and was re-
markable for its outstanding scale and decor. Its claim to an official status rests on an inscription
recordingits rebuidin g by M. Ulpius Victor, procuratorof the city under Gordian III (A.D. 238-
244). Suites of rooms grouped around three "atria"could well have accommodateddomesticand
official life.
In central Europe, among the establishmentsthat have strongif not certainclaim to identification
as governors palaces are those at Carnuntumnear Vienna (2nd-4th centuries)38and andat Aquincum
near Budapest.39In neither case, however, is the plan sufficiently complete to permit a useful
comparison.
Another great establishmentthat invites comparisonwith our Athenian complex is the Villa at
Piazza Armerinain central Sicily (PI. 67:b).40The Sicilianvilla is approximatelya centuryearlier,
35See the excavationreportsby W. A. Daszewski in Reportof the Departmentof Antiquities,Cyprus,1970, pp. 112-
132; 1972, pp. 204-236; 1976, pp. 185-226; 1984, pp. 294-304; idem, DionysosderErloser (footnote18 above,p. 106)
for the mosaics with a plan of the building, p. 17, fig. 1.
36 Princeton Encyclopedia,pp. 71-72 (D. White); R. G. Goodchild,
Antiquity34, 1960, pp. 246-258; idem, "The
Palace of the Dux,E in Apollonia, the Port of Cyrene, Excavationsby the University of Michigan 1965-1967, J. H.
Humphrey, ed., Tripoli n.d., pp. 245-260 (Goodchild's manuscript was completed in 1966-1967); J. W. Hayes,
"Notes on the Palace Pottery,"in ibid., pp. 260-265.
37 R. Thouvenot, Volubilis,Paris 1949, pp. 47-48; McKay, Houses, Villas and Palaces, p. 225.
38 Swoboda,Paldste3,p. 296, note 74: "Die Bestimmungdes Bauwerks-ob Palast oder Teil einer
Baderanlage-ist
zunachst nicht entschieden."PrincetonEncyclopedia,pp. 198-199 (R. Noll): "The purposeof the complex, as well as
of the individualrooms, is still undetermined."
9 E. B. Thomas, Romische Villen in Pannonien,
Budapest 1964, pp. 107-108; Swoboda, Palaste3,pp. 284-289,
fig. 111; McKay, Houses, Villas and Palaces, p. 202.
Much is being learned through the continuing exploration of another importantvilla of the 3rd-4th centuries at
Gamzigrad,ancient Romuliana, in easternYugoslavia.Within a fortifiedenclosureare two loosely connected"palatial
residences"together with two temples and several "servicebuildings".Since the site has now been identified as the
burial place of the EmperorGalerius within his native Dacia Ripensis the complex in its fully developedform is per-
haps to be regarded,like Split, as a residencefor a retiredruler. See M. Mirkovicand V. Popovic,in Palastund Huiitte,
D. Papenfuss and V. N. Strocka,edd., Mainz 1982, pp. 486-492; Wilson, Piazza Armerina,pp. 84-85,91, fig. 50:D;
Wilkes (footnote32 above, p. 114), pp. 66-70, fig. 11 with bibliographyin note 183. I am gratefulto ProfessorSlobo-
dan Curcic for helpful discussionof the Yugoslavvillas.
40 G. V.
Gentili, La Villa Imperiale di Piazza Armerina (Itinerari, no. 87), Rome 1954; Dunbabin, Mosaics,
pp. 196-212; Ward-Perkins, Architecture,pp. 461-463, figs. 312-314; A. Carandini et al., Filosofiana,la Villa di
 116 V. THE PALACE OF THE GIANTS

but the two have much in common.Both exhibit an easy disregardof preciseaxial symmetryin the
overall layout. Each was entered through a triple gateway that formedthe facade of a colonnaded
court.The main body of the Sicilian villa, like the Athenian,comprisedtwo courtcomplexes,to one
of which was appended a bath. Three of the rooms are distinguishedby stately scale and plan
(PI. 67:b: nos. 30, 39, 46). It would be difficult, however, to distinguish a residential from a cer-
emonial, a privatefrom a public part in this great complex.This accordswith a growing conviction
among scholarsthat the villa at Piazza Armerinawas not an Imperialor officialresidencebut, more
probably,the seat of some wealthy Sicilian magnate.
A more profitablecomparisonmay be drawn from a more remote source,viz. Dura-Europos on
the Euphrates River.41The largest building on the site in the Roman period (PI. 67:a) has been
© American School of Classical Studies at Athens

identifiedwith virtual certaintyas the official residenceof the Dux Ripae, i.e., the commanderof
For personal use only. License: CC-BY-NC-ND.

the Euphrates frontierzone. Built sometimeafter A.D. 211, the building shared the destructionof
the city in A.D. 256. Like our Athenian building the Dura palace was given a striking location:set
high on the north edge of the city, it overlookedthe river and the plain. But the most significant
feature shared by the two buildings was the dual nature of their plans, which comprisedin each
case readily distinguishableresidentialand ceremonialparts.
In the Palace of the Dux the residential elements are grouped around the large north court.
Outstandingamong the parts was the suite of rooms centeredabout the large apsidal chamberin
the north side. The suite faced outwardthrougha colonnadeacrossa broadterracefrom which the
residentsmight enjoy fresh breezesand a splendidview. This element in the Dura building may be
comparedwith the east block of the SoutheastCourt in the Athenian. In each case we are clearly
dealing with the "mastersuite".In the Palace of the Dux this suite was adjoinedon the east side by
a bath, on the west by a latrine. The roomsset aroundthe other three sides of the north court were
clearly intendedfor the domesticstaff. The large room at the southwestcornerof this courtand the
adjoiningannex have been identifiedas stables.
The large south courtconstitutesa well-definedsecondpart of the ducal palace. From outsideone
entered through a substantial propylon centered in the south side and wide enough to admit a
procession.Passage from the south into the north court led through an axially placed, wide door-
way. A secondarydoorwaywith propylonopenedin the middleof the west side of the south court;it
would have providedconvenientaccess to the stables. From the east side of the court projecteda
large room reminiscentof the semicircularexedraopening fromthe west side of the North Court in
the Athenian building. The excavatorsof Dura regardedtheir south court as a suitable place for
troop reviews, formal audiences,the trial of lawsuits, and other such exercises.
It thus appearsthat in two widely separatedplaces in the Late Romanworld the requirementsof
an officialresidencewere met in much the same way, one in the 3rd centuryand one in the 5th. The
solution was the skilful combinationin one carefully organized complex of a domesticestablish-
ment, a guest suite, and a ceremonialfacility. Little wonder that, as we learn from the Imperial
edicts, accommodationin such palatia was eagerly sought.
Piazza Armerina,Palermo 1982; Wilson, Piazza Armerina;N. Duval, "Commentreconnaltreun palais imperial ou
royal? Ravenne et Piazza Armerina,"Felix Ravenna 115, 1978, pp. 27-62.
41 Rostovtzeffet al.,
Dura-Europos;C. Hopkins, The Discoveryof Dura-Europos,New Haven 1979, pp. 222-225;
S. B. Matheson, Dura-Europos:The Ancient City and the Yale Collection,New Haven 1982, p. 22; Ward-Perkins,
Architecture,pp. 352-353. The most specificevidencefor the identificationis the occurrenceof the name and title of the
Dux Ripae in dipinti on a first coat of plaster in Room 7: Rostovtzeffet al., op. cit., pp. 30-40, 93-94. The exploration
of Dura-Europoshas been renewedrecentlyby a Frenchmission:Syria 63, 1986, passim. In the spring of 1988 sound-
ings were carriedout in the palace underthe directionof ProfessorSusan B. Downey. The results call forsome correc-
tions and adjustmentsin the plan and history of the complex, but they do not affect its relevanceas a parallel to the
Athenian building. I am grateful to ProfessorDowney for the information.
 VI
THE DARK AGES (A.D. 582-700)
RECOVERY
Recoverybegan on a modest scale at the turn of the century and continuedthrough the first two-
thirds of the 7th century, as indicatedby the numbersof coins noted above (p. 93). It was probably
dictatedby external circumstances.Athens, with Attica, was one of the few partsof Greecesouth of
Thessalonica not occupied by Slavs, and inhabitants of some of the neighboringareas may have
© American School of Classical Studies at Athens

taken refuge there as well as on the islands.1

Even this modestrecoverywas interruptedat least once during the reign of Heraclius (A.D. 610-
For personal use only. License: CC-BY-NC-ND.

641). The evidenceis providedby a group of 86 coins in the debrisfrom destructionby fire in Room
8 of the Broad Street Building (which had alreadysufferedin the invasionof 582). The latest piece
in the group dates to year 24 of the reign, i.e. 633/4, and so the troublemay probablybe assignedto
the late 30's of the century. This interruption is therefore too late to be reckonedwith the dis-
turbances in the whole Aegean area, which occurred,probably, in 614.2 Whether it was part of
another such, unrecorded,operationor some purely localized disaster,it is impossibleto say. The
rest of the building very likely sufferedas well, but it was rebuilt later in the century,and no signs
of destructionremained.The adjacentarea is still unexcavated.
Recoveryreachedits peak in 662/3 when Constans II, in the courseof his campaignagainst the
Slavs in the interior of Greece, made Athens his headquartersfor the winter. The dramaticrise in
the number of coins from his reign is made even more impressive by the fact that, as noted by
Charanis, of 817 coins published in Agora II only five could be said definitelyto have been minted
after 663.3
It may have been at this time that the Hephaisteion and Erechtheion were remodeled as
churches,4and it is perhaps not mere coincidencethat a lead seal of Theophylaktos, Bishop of
Athens in the 7th century,was found on the Areopagus.5Also, fromthe cemeteryon the Areopagus
(below, p. 118) came a bronze reliquarycross of the late 6th or early 7th century.6
THE BUILDINGS
The shoddy nature of the 7th-century buildings in Athens makes it difficultto recaptureany co-
herent topographicalplan of the Agora. The buildingswere undoubtedlysparse,and many of those
1For the islands, see S. Hood, "An
Aspect of the Slav Invasionsof Greece in the Early Byzantine Period,"Sbornik
NarodnikoMuzea v Praze A 20, 1966, pp. 165-171, and idem, "Islesof Refuge in the Early Byzantine Period,"BSA
65, 1970, pp. 37-45. T. Gregory, however, will show, in a forthcoming article in AeXTrlov r Xplo'TtaVtLK7
'
'ApXaLoAo0yLKs EraLpipas, that some at least of these island settlementswere actually part of a widespreadcommer-
cial network that included both island and coastal sites and had been in operation from the 4th through the 7th cen-
turies. On this general subject,Charanis, "Observations,"and "Significance."
2 On this
major upheaval, see Metcalf, "SlavonicThreat," pp. 14-23. M. Thompson (AgoraII, p. 4) suggests that
the Bulgarians might have laid siege to Athens on their way to Corinth, ca. 640. For the debateconcerninga possible
invasion which included the capture of Corinth at that time, cf. K. M. Setton, "The Bulgars in the Balkans and the
Occupation of Corinth in the Seventh Century," Speculum 25, 1950, pp. 502-543 and 27, 1952, pp. 351-362;
P. Charanis, "On the Capture of Corinth by Onogurs and its Recapture by the Byzantines,"Speculum 27, 1952,
pp. 343-350.
3Charanis, "Significance,"pp. 163-172. For more on ConstansII in Greece, see Frantz, DOP, pp. 197-199.
4 Above, p. 92, and Frantz, op. cit..
5 IL 529. V.
Laurent, Le corpusdes sceaux de 1'empirebyzantin V, Paris 1963, pp. 437-438, no. 585.
6 Travlos and
Frantz, "St. Dionysios,"p. 168 and pl. 43.
 118 THE DARK AGES

that did exist must have vanishedwithout a trace. But a surveyof the fragmentaryremainsgives the
impression that although the west side of the Agora was occupied to some extent, the favored
Romanthe
district was in the south, from the RomanRomanMarket to theRoman bath (Southwest Bath) at the
southwest entrance to the Agora (PI. 3). The chief concentrationwithin this area was at the east
end, from a strip of unimpressivehouses along the innerthe cPost-Herulian
face of Wall above
Tower W 2 to the western edge of the "Schools".The great majorityof the coins of Constans II
found in the Agora come from this region and over the whole course of the Panathenaic Way.7
These include many from a gully torn out by the torrentialrains pouring down the slopes of the
Acropolisand Areopagus,between the PanathenaicWay and the preservedlimits of the schools.A
similar proportionof the coins of the earlier 7th centurysuggests some importantinstallationfar-
© American School of Classical Studies at Athens

ther up the slopes, and we may conjecturethat as the Slavic threat once more became acute in the
For personal use only. License: CC-BY-NC-ND.

beginning of the 7th century an effort was made to achieve a strong position by establishing a
garrisononon the west slope of the Acropolis.8
A fortressin that locationmay have existed even beforethe constructionof the Beule Gate, if the
opinion of Beule himself is correct.9Somesupportis given to this view by a statuebase (IG 12, 3193)
found "onthe Acropolis,west of the Propylaea"(PI. 74:b). After an erasureof seven or eight lines,
presumablybearingthe name and titles of the honorand,the inscriptionrefersto a fortress"restored
to order"(KOO'O . . . KaraeKevaaev). The editorsfollowed Dittenbergerin dating the structurein
question in the 2ndd century with a rebuildingin the 3rd. If this dating is correct,the raid of the
Kostobokoica. A.D. 170, when they caused great damagein Eleusis, might have been the incentive
for the first phase and the Herulian attackfor the second.One might hazarda guess that the fortress
was allowed to fall into disrepairin peacetimeand "restoredto order"'in time of danger. 1
Housing for the garrisonmay have been providedby the remodeledschoolsand the flimsy build-
ings along the east side of the Post-Herulian Wall. A cemeteryon the Areopagusslope, near the
later Church of St. Dionysios the Areopagite, is probably to be associatedwith the military in-
stallation.'2 A number of graves contained belt buckles of the type known as "Avar".Much has
7 P. Charanisnotes that this volume of Heracliusbearingthe
may be increasedby the inclusione
figures of Heraclius and Constantine, grandfather and father of Constans. Charanis believes that Constansmay have
kept these in circulationto emphasize the legitimacyof his own succession,which his violent conductmight have belied
(Charanis, "Significance,"p. 167).
8 Mabel
Lang came to the same conclusionindependentlywhile studying the inscriptionspainted on many of the
water jars found in the vicinity.
9 E. Beule, L'Acropole d'Athenes I, Paris 1853, pp. 106-108. The puzzling and interestinghistory of the Gate is

beyondthe scope of this volume.

0 The text of the
inscriptionfollows:
[xapLv av>T ?V?py?OLSv ME?]-
yaxLv avrifbfo0LV JAC]-
vazoLnJ HoALtbtavE6r7Ka[v].
KOa.Jov TQ)fpovpLw o avr[0o]
OLKELOLsavaXoM1a0rLV
Karea'Kevac-ev.
1 For the wordaKoO-MLOV in a similar context, with a discussion of its meaning, see above, p. 10 and n. 64. A
suggestion by P. Graindor ("Un episode de la vie d'Herode Atticus,"Le musee beige 16, 1912, pp. 73-74) that the
original donor might have been Herodes Atticus has met with little favor. For the evidence of the activities of the
Kostobokoi,see A. von Premerstein,"Untersuchungenzur Geschichtedes Kaisers Marcus,"Klio 12, 1912, pp. 147 ff.
The fortressis mentionedby R. K. Sherk, "RomanImperial Troops in Macedoniaand Achaea,"AJP 78, 1957, p. 61.
12 See Travlos and Frantz, "St. Dionysios,"pp. 166-168 and Frantz, DOP, p. 198.
 THE BUILDINGS 119

been written on the source and significanceof these buckles. The more elaborate examples un-
doubtedlyfind their prototypesin Eastern Europe and the area of the Black Sea.13Of those found
in Greece some, at least, must have been brought by invaders at whose death or capture they fell
into the hands of the defenders.The quality of the workmanshipvaries considerably,and it is not
impossiblethat many are copies or adaptationsmade in the foundriesof Athens or Corinth.In some
small buckles from the Areopagus cemeterythe plate is cruciform;they thereforeare probablyof
local manufacture.
The most important building activity that can be associatedwith Constans' visit concernsthe
Broad Street Building, constructedin the early 5th centurybehind the colonnadeof the street. As
© American School of Classical Studies at Athens

noted above, it had incurredheavy damage in 582 and again, to a lesser degree,in the 630's. It was
now reconstructedon a fairly drasticscale, includingsubdivisionof much of the original colonnade
For personal use only. License: CC-BY-NC-ND.

into small rooms.14The date is well attestedby six coins of Constans II under the newly laid tile
floor, all dating between 646 and 658. Constansembarkedon his Slaviccampaignin 658. Sometime
in the next four years he must have made his plan to use Athens as his headquartersfor the next
phase, and we may assume that it was then that the massive debriswhich filled much of the build-
ing was clearedaway and the building made the suitable for Imperialneeds. Even so, it was appar-
ently not entirely ready on the Emperor's arrival, for a large number of roof tiles were found
stackedagainst the wall of one of the rooms, and there they remainedafter his departure.
Probablycontemporarywith this phase was a very small and primitivelatrine at the intersection
of the Broad Street and the road coming in from the north (PI. 21:e).15The building, which had
been hastily and flimsily built for the emergency,fell into decaysoon after Constansmovedon. The
marble colonnadeof the secondstorey collapsed,and much of it fell into the street,which was soon
silted over. This fact, togetherwith the spectaculardrop in the numberof coins in the whole Agora,
reflectsthe transitorynature of the rebuildingprogram.
Among the flimsy structureseast of the Panathenaic Way possibly to be associatedwith Con-
stans' visit was a small laundry establishmentjust inside the Post-Herulian Wall at the Hypapanti
(South) Gate (PI. 74:c). Built into the southwest cornerof a building of the Hellenistic period, the
operative part of the establishment consisted of a long rectangular basin ca. 1.00 m. wide and
3.35 m. long from east to west, revettedwith slabs of Hymettian marble.In a recessat the south was
set a tub hollowed out of an ancient of block Pentelic marble 1.60 m. long and 0.75 m. wide,
standing ca. 0.40 m. above the floor. Water flowed in from the south, obtained apparently by
dammingthe ancient channelbetween the establishmentitself and the streetto the south and letting
it overflowinto the tub. It is to be notedthat no attemptwas made to put back into use the elaborate
waterworks of the early years of the Christian era even though the laundry was very close to the
vaulted brick chamberof that system.16Sixty-one coins in all were found in the vaulted chamber.
The laundry was destroyedby fire along with a numberof buildings in the vicinity. It was over-
laid with a fill of broken tiles with much burning and a considerableamount of pottery of the 7th
century. The settlementand the conflagrationthat destroyedit could be dated with some precision
to the reign of Constans II or very soon after by the many coins found in the destructiondebris,
13
G. D. Weinberg, "AWandering Soldier'sGrave in Corinth,"Hesperia 43,1974, p. 513, note 3, for a summaryof
the problem,with references.
14 Hesperia 42, 1973, pp. 395-397 (Shear, Jr.).
15
Above, p. 34.
16
For the water system, see Hesperia 8,1939, pp. 218-220 (Shear, Sr.). Six were of the 1st to 2nd centuryor earlier;
all the rest were of the 4th to 5th century.
 120 THE DARK AGES

which included 42 from the reigns of Phocas and Heraclius (A.D. 602-641), 67 of Constans II
(A.D. 641-668), and only three from the remainingyears of the century (A.D. 668-705).
Another building rehabilitated (partially) in the 7th century, this one perhaps in the reign of
Heraclius, was the Southwest Bath. It was almosttotally destroyedin 58217and was now rebuilt in
only the sketchiestpossible manner. It was noted above (p. 32) that in its penultimate,post-Heru-
lian phase the bathing facilities were sharply curtailed,being then limited to the north side of the
building.18In the final, post-Slavicperiod even these were eliminated.Some floors in the southern
section were repavedat a much higher level. The west hall was in use again, minus its exedra;its
marble-chipfloor, cleared of debris, served again though damagedin part. A flimsy crosswall di-
© American School of Classical Studies at Athens

vided the room into two unequal parts.

It has been suggestedthat the additionof benches in all the rooms in the southernsection of the
For personal use only. License: CC-BY-NC-ND.

post-Herulian building indicatedthat in the final phase it was used as a school. One might conjec-
ture that it continued to function as such through the 7th century, but completelyChristianized,
although with a classical flavor. It was in Athens that Theodore of Tarsus laid the foundations
for the distinguished ecclesiasticalcareer that took him to Britain as Archbishopof Canterbury
(A.D. 669-690), "well versed in Latin and Greek."'19
Some impressionof the povertyof Athens in the 7th centurymay be gained by an examinationof
the state of the water supply at that time. Of some 15 wells in use during the 6th century in the
general area of the schools, only four continuedinto the 7th century.20In addition, the running-
water systemsappear to have gone permanentlyout of use in the late 6th century.The earth which
clogged the vaulted channel (as well as the vaulted passage) leading from the springhouseat the
southwest cornerof House B containedpottery of the same characteras that found in the general
destructionfill; and the fresh-water channel which skirted the southeast corner of House B was
evidentlycut off and never restored.
INDUSTRIAL ESTABLISHMENTS
Several tile kilns in the area of the 7th-centuryoccupationshould be consideredas adjunctsto the
constructionactivitiesdescribedabove rather than as commercialenterprises.All those yet discov-
ered are in the vicinity of buildings which were extensively remodeledat the time and so were
probablyonly temporaryaffairs for the sole purposeof providingfor the needs of remodeling.The
best preserved,measuringca. 2.75 x 5.50 m., was at the west side of the SouthwestBath, using the
west wall of the hall with the mosaicfloor as its own east wall (PI. 75:b). Only the very edges of the
floor of the stacking chamberremained,with a little of the stuccofacing of the west wall. Most of
the fire chamberwas preserved,consistingof a passage 0.80 m. wide running down the middle of
the kiln, which distributedthe heat to the stackingchamberabove by means of six lateral flues on
each side. The plan, though simpler, is not unlike that of a tile factoryin Corinthof a much earlier
period.21This and the absenceof potterywasters indicatethat the Athenian furnacealso was used
exclusivelyfor makingtiles. The identificationis strengthenedby the discoveryin the ash fill of over
fifty leaf-shaped kiln supports such as are used for separating tiles in small modern tile-making
17 For the evidence,see Hesperia 38, 1969, p. 415 (Shear,Jr.).
18 For the plan, see PI. 19:d.
19 The Venerable Bede, Historia ecclesiastica iv, chap. 1; Pope Zacharias, PL LXXXIX, col. 943C.
N 20:3; 0 18:1;Q 17:4;S 22:2. This last lies somewhatoutside the area, to the southeast,and more likely served
20

some of the late houses near the PanathenaicWay.

21 Ancient Corinth: A Guide to the Excavations, 6th ed., Athens 1960, p. 87; A. Orlandos, Les materiaux de con-
structionet la techniquearchitecturaledes anciensgrecs, Paris 1966, pp. 71-72. On kilns in general, D. P. S. Peacock,
Potteryin the Roman World,London 1982, p. 69.
 INDUSTRIALESTABLISHMENTS 121

establishments.The source of the clay for the tiles was probablysome large amorphouspits in the
vicinity. Two similar tile kilns, tracesof which were foundjust to the north, were probablyused for
the same purpose.22
The kiln can be datedwithin fairly close limits. Fill antedatingthe kiln underthe northend of the
bath producedpotteryup to the late 6th century,but the kiln had alreadygone out of use by Byzan-
tine times when a floor formedacross the area. In view of the proximity of the kilns to the South-
west Baths it is probablethat they were built to make roof tiles for the 7th-centuryreconstructionof
the bath complex, their location being well suited to the sourceof the clay on the one hand and the
destinationof the finishedproducton the other.
© American School of Classical Studies at Athens

The final capitulationto industryis illustratedby the olive pressesinstalledin or near two build-
ings with a distinguishedpast:the Palace of the Giants and the Metroon. It has been seen that after
For personal use only. License: CC-BY-NC-ND.

the abandonmentof the Palace complexa series of small roomswas built againstits east wall, in one
of which was found the deposit of Early Christian ampullae mentioned above (p. 91). Sometime
later the southeasternmostof the roomswas enlargedand a new room addedto the north to accom-
modate an oil-pressing installation. In the northeastcornerof the new room was a carefullymade
marblebase for a press (PI. 76:a), and overthe foundationsof what was apparentlythe original east
wall of the room was laid the foundationfor a round crusheror mill. A large millstone of volcanic
stone of the same diameter,found to the northwestof the Palace, has now been set up on the round
foundation,as shown in Plate 76:b. A date after 582 seems likely for the installationof the press in
view of the fact that the east wall of the original series of roomshad alreadybeen destroyed.
In the Metroon the third room from the south was last seen with a mosaic floor, related in some
fashion to the colonnadedcourt to the north and togetherwith it formingan element in the monu-
mental building period of the 5th century; it was provisionallyidentified as a synagogue.23The
room with the mosaic was now divided by interior walls built of ancient blocks into three parts
(PI. 76:c). The actual pressing took place in the largest (southeast)room, where the press bed still
stands, raised on an ancient block so that the oil could run through the spout into a circularplas-
tered basin. Adjacentto the basin was a rectangulartrough, also plastered,perhaps for the pulp of
the olives waiting to be pressed. The two other rooms bear no traces of their purpose but were
presumablyconnectedwith other parts of the operation.
That such installations were in common use in Late Antiquity is well illustrated by a similar
press and crusherin the village of Olympos in Attica (PI. 76:d, e).24This is a small room,ca. five or
six meters square, perhapsa baptistery,attachedto the west end of an Early Christianbasilica.The
change from baptistery to olive press was made after serious damage to the basilica necessitated
reducingits width from three aisles to one. This occurrenceimmediatelysuggestsa post-Slavicdate
for the installation of the olive press, and the date is supportedby the 7th-centurypottery found in
graves within the basilica. Both press and crusher are still in situ, the crusher supported on a
circular foundation much like that in the Palace press room. Also in situ is a large basin of black
stone below the spout of the press.
The sketchyprovisionfor collectingthe oil fromthe press in the Palace is at variancewith the fine
workmanshipof the marblepress bed itself. There is no traceof a stuccoedbasin underthe spout, as
in the Metroon, nor one of stone as at Olympos. Instead,a rough and irregulardepressionperhaps
22 The
excavation suggested that the pits had been opened by persons intent on getting clean clay such as could be
found in the lower levels in the masses of disintegratedmud bricks from the collapsedwalls of earlier buildings. The
craters were refilled with small stones, probablyremovedfrom the clay by screening,and with other coarse material.
Similar pits have been found under the colonnadeon the Broad Street:Hesperia 42, 1973, p. 397 (Shear,Jr.).
23 Above, 59.
p.
24 For the
descriptionI am indebtedto the late Margaret Crosby.
 122 THE DARK AGES

held sometemporarycontainer.The differencebetweenthe pressand its surroundingsindicatesthat

the press had been made for an earlierperiodof use. It is not impossiblethat the traditionof oilpress-
ing in the area is as old as the Palace itself. There was ample spacefor a small olive grovein the west-
ern part of the enclosuresouth of the great court,and a modestpressingestablishment,adequatefor
this privateproduction,might easily have been maintainedsomewhereon the premises.25
The Palace press in its present form must date from the late 6th or more likely the 7th century,
and it is probablethat the press in the Metroon is of similar date. From the great numberof mill-
stones found in the central part of the Agora it is likely that the presses of the Metroon and the
Palace were only two of a numberthat gatheredto exploit the olive groveof the defunctPalace, but
no tracesof others have been found.
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

25
Cf. an ancillary industry (winemaking) in the Archbishop'sPalace at the 16th-centuryChurch of St. Dionysios
the Areopagite,for a similar combinationof activities (Travlos and Frantz, "St. Dionysios,"p. 186).
 EPILOGUE
The departure of Constans II marked the end of the brief period of prosperitythat Athens had
enjoyed during his stay there.1A few wells just inside the west flank of the Post-Herulian Wall
yielded a small amount of shoddy potteryof the 8th century. It is possible that some of the modest
structuresa little fartherup the hill which were destroyedby fire in the 7th centurywere followed
by equally flimsy buildings that left no trace.2
© American School of Classical Studies at Athens

A hitherto unexplained phenomenonof the incidenceof 88 coins of the period 711-741 may be
For personal use only. License: CC-BY-NC-ND.

explored briefly here.3 The compositionof the group was equally surprising:61 coins of Philip-
picus (A.D. 711-713), 4 of Anastasius II (A.D. 713-716), 23 of Leo III (A.D. 717-741). The most
striking factor was the high number from the brief reign of Philippicus, especially in view of the
rarity of his coins.4 But there is nothing in the reigns of either Philippicus or Anastasius II to
accountfor this burst of activity in Athens. It is better thereforeto considerthe three reigns as one
period, in which case the emphasis shifts to Leo III. This 30-year span might be called an oasis in a
numismaticdesert,because for the 26 years prior to the reign of Philippicusthe Agora Excavations
produced only 8 coins, and for the 126 years between the death of Leo III and the accessionof
Basil I (A.D. 867) the total was only 14.
The great majorityof the coins of the 8th centurywere found along the PanathenaicWay, from
its entrance into the Agora and on up the Acropolis slopes. Most of the rest came from irrelevant
contexts in the area south of the Agora proper, probablyhaving been washed down the slopes over
the years. There is an interesting parallel in the 7th century, when Constans II made Athens his
headquartersduring the winter of 662/3 while he was campaigningagainstthe Slavs. Of well over
800 coins of his reign found in the Agora the majoritywere discoveredalong the PanathenaicWay
or in its immediatevicinity. This led to the conjecturethat Constanshad establisheda garrisonon
the Acropolis.5
A similar situation might be recognizedin the 8th century. In 726 Leo III issued the first edict
banning sacred images, thereby touching off the IconoclasticControversy.The following year the
inhabitantsof Greece and the Cycladesreactedviolentlywith a revoltso seriousthat it requiredthe
Byzantine fleet, with the aid of Greek fire, to quell it.6 No city or island was specifiedby name, but
since the avowedpurposeof the edict was to eradicateall tracesof paganismthere is little doubtthat
the insurrectionwould have had at least the moral supportof the Athenianseven if they were not in
a position to offer much material aid. One may conjecture,then, that after the suppressionof the
revolt the authorities, fearing its renewal, re-established the garrison on the Acropolis (above,
p. 118). But no new outbreak occurred,and as the danger receded the garrison was disbanded,
probablystill during Leo's reign, and Athens lapsed once more into decline. Its only other claim to
1 Out of well over 800 coins from his
reign found in the Agora more than 700 are known to have been struckbefore
his departure(AgoraII, pp. 70-71, 106-107).
2
Above, p. 90.
3First publishedby M.
Thompson, "SomeUnpublished Bronze Money of the Early Eighth Century,"Hesperia 9,
1940, pp. 358-380; the numbersquoted here are those of her revisionin Agora II, pp. 71, 108.
4 Cf.
Thompson, Hesperia 9, 1940 (footnote3 above), p. 362.
5 Frantz, DOP, p. 198, and above, p. 118.
6
Nicephorus, Breviarius,37 C-D (CSHB, pp. 64-65); Theophanes, p. 405 (de Boor).
 124 EPILOGUE

distinctionin the 8th century was that it was the birthplaceof the Empress Eirene7and that she
found it a fitting place of exile for her husband'sover-ambitiousstepbrothersin 797.8
The situation in Greece as a whole began to improvetoward the end of the 8th centurywhen, in
783, Eirene's general, Stauracius,conducteda victoriouscampaign against the Slavs in the Pelo-
ponnese. This was followed by other successesuntil, by the middle of the 9th century, Byzantine
authority was establishedthroughout Greece.9By that time a new spirit had infused the land; it
was brought in from the east and exemplifiedby three churches,all built within a space of three
years and convenientlydated by founders'inscriptions:Athens, St. John Mangouti, 871; Thebes,
St. GregoryTheologus, 872; Skripou, Dormition of the Virgin, 873.10The churchat Skripou (the
© American School of Classical Studies at Athens

ancient Orchomenos)typifies the break with the past and the new vigor which presaged the de-
veloped Byzantine style of architectureand sculpturethat was to becomethe norm in Greece.The
For personal use only. License: CC-BY-NC-ND.

classicallyorientedbasilica was replacedby the cross-in-square,and architecturalornament,once

firmly rooted in classical motifs, gave way to textile-inspireddesigns enlivenedby animals which
owed nothing to the classicalpast. In the somewhatclumsy proportionsand crude carvingof Skri-
pou can be found the seeds of the sophisticatedcross-in-squarechurchesthat were to becomethe
glory of 11th- and 12th-centuryAthens when it once more enteredon a periodof prosperity.11

7Theophanes, p. 444 (de Boor).

8
Theophanes, p. 473 (de Boor).
9 D. Obolensky,"The Empire and its Northern Neighbors,"CambridgeMedieval History IV, i, Cambridge1966,
p. 489.
10M. G. Sotiriou, "'O vaos rijs xKplTroiV "
BoLtoria,"'Apx'Ef 1931, pp. 119-157 (Skripou);G. A. Sotiriou, O
ev Ot,3asLvaos Fp7RyopLov rov QeoXoyov,"'ApX'Ef 1924, pp. 1-26 (Thebes); idem, EMMEI, pp. 85-87 (St. John
Mangouti).
1 See Frantz, AgoraXX. This and the ensuing periodswill be the subjectof anothervolume in this series.
 APPENDIX
THE POST-HERULIAN WALL
The Post-Herulian Wall enclosedan area of 120,300 square meters,that is, barely one-nineteenth
of the extent of the city at the time of Valerian (Pl. 5). The enclosed area formed an irregular
rectanglenorth of the steep rock of the Acropolis,the Acropolisitself formingthe south side of the
enclosure.The dimensionsof the new wall were as follows:west flank, 425 m.; north, 470 m.; east,
© American School of Classical Studies at Athens

255 m.; south, 300 m. Also included within the wall was the area of the Library of Hadrian, as
For personal use only. License: CC-BY-NC-ND.

shown not only by the way in which the Librarywas connectedwith the northflank of the wall, but
also by a passage in a descriptionby the Anonymousof the AmbrosianLibrary, referringto the
rectangleof the Library as a part of the city wall.1
The Post-Herulian Wall protectedthe central part of the city and the Acropolis.Its only short-
coming was that it was not directly connectedwith the entranceto the Acropolis, a disadvantage
that was felt in Byzantine times. This was remediedin the 11th century by the constructionof a
new wall, the so-called Rizokastro,which surroundedthe Acropolisandjoined the Post-Herulian
Wall on its east and west sides.
It may be noted that up to the beginning of the 20th centurythe area enclosedby the wall con-
stitutedthe centerof the city of Athens. Even today this area remainsas the most interestingdistrict
of the city because in it are preservedmonumentsof all periods:the Classical, Hellenistic, Roman,
and Byzantine, as well as the Turkish and after the liberationfrom the Turks.
So far as we know, the only referencesto the Post-Herulian Wall are by Cyriac of Ancona,who
visited Athens in 1444 and 1465,2 and the Anonymousof the AmbrosianLibrary, who was prob-
ably in the city in 1466 or 1470.3 It also seems probablethat Symeon Kavasilas,who, in a letter to
Martinus Crusius written in 1578, referredto Athens as being "three-fold"and "whollyinhabited",
had in mind the Wall of Valerian, the Post-Herulian Wall, and the Rizokastro.4
After the capture of Athens by the Turks in 1456 the city spread out once more outside and
around the Post-Herulian Wall, which was abandonedsoon after. At any rate, it is certain that it
was no longer in use in the 16th and 17th centuries,since on top of it were built churcheswhich can
be dated to that time.5 With the passage of time large sections of the wall were destroyed.Many
parts were systematicallydemolishedby the inhabitantsin orderto recoverthe stones for building
material for their houses. Of the parts preservedup to the middle of the 19th century many were
deliberatelydemolishedby archaeologistsin searchof importantinscriptions,sculpture,and archi-
tectural members,or else to recoverthe earlier buildings lying underneath.
1"Itema u(n) canto de li muri della terra, che e tra ponentee tramontana,e un grandissimoaedificioet e in f(orm)a
quadrata;alla fazza per ponente e una porta bella con un portico con 4 colonne, per la qual al presente s'entra nella
terra;in la parte senestradella detta porta sono 5 belle colonnee grandial ornamentodel muro, alla destrae solamente
el muro di marmoriquadrati, et e tutta la detta fazza cerca un tirar d'arco;l'altra fazza che guarda verso tramontana,
non ha alcuno ornamentoessceptoil cornisodi sopra, ma e un bel muro e sono 3 torre, una per cantoe una in mezzo, la
qual fazza e piu di un tratto (a) balestralonga. Questi due fazze sono tutte intiere, ma le altre doi sono ruinate, mostra
esser stato un grande et bello aedificio"(Ziebarth [footnote98, p. 71], p. 77).
2 Bodnar (footnote 52 above,
p. 9), p. 40.
3Judeich2, p. 16, note 2 and Ziebarth
(footnote98 above, p. 71), pp. 80-83.
4
M. Crusius, TurkograeciaeLibri Octo,Basiliae 1584, VII, 18; also Cte. De Laborde,Athenesaux XVe, XVIe, et
XVIIe siecles I, Paris 1854, p. 58.
5 Travlos and Frantz, "St. Dionysios,"p. 176.
 126 APPENDIX: THE POST-HERULIAN WALL

These archaeologicaloperationsmade it possible to identifysuch buildingsas were preserved,as

well as to determinethat they had found their way into a later fortificationwall which stretched
northwardfrom the Acropolis.In 1851/52, when the ArchaeologicalSocietyexcavatedaroundthe
building then believed to be the Bouleuterion,it was recognizedfor the first time that the remains
did not belong to the Bouleuterionbut to a later fortificationwall, without, however, any certain
determinationof the date of its construction.6Also, when in 1861 S. Koumanidesexcavatedaround
the Stoa of Attalos and the Panagia Pyrgiotissa,as well as on the site of the Churchof St. Demetrios
Katiphori,he recognizedthe existenceof a circuitwall and gave a full descriptionof its course.7
It is worth noting that in 1852, when publishing an inscriptionfound in the neighborhoodof the
© American School of Classical Studies at Athens

Church of the Hypapanti, KyriakosPittakis wrote that beforethe Revolutiona traditionwas cur-
For personal use only. License: CC-BY-NC-ND.

rent among the older Athenians that when the barbariansattackedAthens the inhabitantsbuilt a
wall around the district north of the Acropolis because the Acropolis was not sufficientfor their
safety. He also noted that in that area there were many ancientstonesand that the inhabitantsused
them to build their houses.
Thus, immediatelyafter the middleof the 19th century,the Post-Herulian Wall began to appear
on various plans of Athens, sometimes accuratelydrawn8and on others with many errors.9The
beginningof the excavationsof the AmericanSchool,however,providedthe opportunityto discover
in all possible detail the whole of the western flank, while at the same time chance finds elsewhere
in the city recoveredin the courseof modernbuildingoperationsmade it possibleto determinemore
1
accuratelythe courseof the wall on the other two flanks.
CONSTRUCTION OF THE WALL
The method of construction of the Post-Herulian Wall resembles closely that of the Wall of
Valerian, the only possible differenceperhaps being the total absence of iron clamps such as are
used in the Wall of Valerian.
Like the Wall of Valerian, the Post-Herulian Wall, whose thicknessfluctuatesbetween 2.50 and
3.50 m., is not built entirely of stone; only the two outer faces are constructedof ancient squared
blocksand architecturalmembers.The wall thereforeconsistsof two quite separate,parallel walls,
varying in thickness from 0.40 to 0.70 m., dependingon the size of the ancient blocks. The inter-
vening space was filled with large and small unworkedstones and architecturalmembers,mostly
columns, capitals, Ionic bases, geisa, and generally stones with irregular surfaces (PI. 10:a). The
spaces between the stones were filled with small chips of architecturalmembers,roof tiles, sculp-
ture, and inscriptions,all bondedtogetherby lime mortarof a poor consistencyand not very strong.
The wall was built, so far as possible, on the foundationsof earlier buildings. Where it was
necessaryto build directlyon the groundthe earth was dug away down to bedrock,in which a deep

K. Pittakis, D. Charamis,and P. Eustratiades,'E7rLypa4al a&?K?oroL, fasc. 2, Athens 1852, pp. 7-8.

6

1861, pp. 8-18.

7 rFEVLK,VEAEvos
8 E. Curtius, AttischeStudien I, Gottingen 1862, pi. II; idem, Sieben Karten zur Topographievon Athen, Gotha
1868,pl. 3; Wachsmuth, II, i,pl. I.
9 E. Curtius, Stadtsgeschichtevon Athen, Berlin 1891, pi. VI. G. Fougeres,Athenes, Paris 1914, p. 15. G. A. Soti-
riou, EMMEI, plate opp. p. 26. Judeich2,pl. IV.
10The plan of the Post-Herulian Wall madeon the basis of the newly discoveredsectionswas firstpublishedin 1946
(Travlos [footnote112 above,p. 73], pl. 30). Cf. also Bodnar(footnote52 above, p. 9), pi.III; Travlos,
IloEo58oSKuc7,
fig. 221.
 DESCRIPTION OF THE WALL BY SECTIONS 127

trench was opened to a depth of 0.50-0.70 m., in orderto providea firm and level surfaceon which
to lay the foundationsof the wall.
The faces of the wall were built of large ancient blocks laid in regular courses,and whereas the
outer face was built with great care, with perfectjoints, the inner face was carelessly laid and
consistedof all sorts of architecturalmemberswith their carvedsurfacesprojectingfrom the face of
the wall. On this face column drums were used in such a way that their curved surfaces were
contiguous to each other and visible in the face of the wall. The spaces between the drums were
filled with small stones and strong lime mortar (see Pls. 7:a, 8).
Characteristicof the constructionis the total absenceof brick. Instead, pieces of marbleor terra-
© American School of Classical Studies at Athens

cotta roof tiles were used to fill the gaps or to make horizontalbeddingsfor the wall courses.
Some sections of the wall were built entirely of marble;others of poros. This difference,which
For personal use only. License: CC-BY-NC-ND.

makes a striking impressionon first view, has no relationeither to the solidityor to the appearance
of the wall. It arises from the availability of material from marble or poros buildings, whichever
were closest to that particularsectionof the wall.
The mortar used in the constructionwas not always of the same composition.In general, the
mortar used in the inner filling of the wall is deficient in lime, whereas that used for setting the
blocksin the two faces is very strongand consistsof lime, sand, and small piecesor grainsof red clay
from ground-up roof tiles. This is the mortarwhich came into use in Athens at the end of the 1st
centuryB.C.
The rectangulartowers projectingfromthe walls were built in the same way, but their sides were
thicker, 1.40-1.90 m. Their interiorswere filled with architecturalmembersof all sorts,taken from
earlier buildings.
The wall is preservedto its greatestheight behindthe Stoa of Attalos, where it rises to 7.50 m. Its
actual height, however, must have reached 11.50 m. This could be determinedbefore the recon-
structionof the Stoa of Attalos from a sectionof the shop-frontwall (betweenshops 2 and 3), where
it was preservedto its original height of 11.50 m. The Post-Herulian Wall must certainly have
risen to that point, because it was built over the shop-frontwall for the entire length of the Stoa.
The battlementswould have begun above that point.
The gates of the Post-Herulian Wall, estimatedas eight in number,were protectedat intervalsby
towers to guard the main streets of the city.

DESCRIPTION OF THE POST-HERULIAN WALL BY SECTIONS

For a better understandingof the constructionof the wall we shall follow its course in detail on all
three sides, i.e., the West, North, and East.

WEST FLANK
Sections of the west side of the wall were first discoveredin the middle of the 19th century;the
entire flank was uncoveredand later systematicallyexcavatedby the AmericanSchoolof Classical
Studies.
This flank of the wall begins at the northwestcornerof the Acropolisat the retainingwall in front
of the Propylaia and ends at the northwestcornerof the Stoa of Attalos (425 meters).On accountof
the steep slope of the terrain, however, there is a great difference(66.80 m.) in the elevationat the
two ends.
The west flank is the most interesting,being the only side which has been uncoveredover its full
length. This will be describedin four sectionsdeterminedby the three gates on that side.
 128 APPENDIX: THE POST-HERULIAN WALL

First Section
The first sectionbegins at the northwestcornerof the retainingwall of the Propylaiaand ends at
the Hypapanti Gate. It has a length of 135 meters and is punctuatedat about midpointby a pro-
jecting tower (W 1), rectangular,6 x 7 meters. The thicknessof the wall in this section is stable,
measuring2.95 m. Only at the beginning,near the Klepsydra,does it have a thicknessof 4.50 m. on
accountof the steep slope of the rock down to the line of the road (the Peripatos).
Up to this point the lower courses of the wall were preserveduntil 1937, when it was found
necessaryto demolish them in order to expose the front of the Klepsydraand the west edge of the
paved forecourt.1 All the rest of this section up to the Hypapanti Gate is visible today except for a
© American School of Classical Studies at Athens

small stretchof 22 meterswhich, after its excavationand study, was coveredover again in orderto
For personal use only. License: CC-BY-NC-ND.

restore the Peripatos.'2This portion was preservedto a height of 2.25 m. and consistedof three
courses.Both faces were built of isodomicmasonry,with large porosblockslaid as orthostates.The
poros blocks, some the soft stone from Piraeus, some hard, were taken from various earlier build-
ings dating from the 6th century B.C. (At least two of these blocks came from the Kara quarries,
which are known to have been in use at the time of Peisistratosand his successors.)Some of the
blocks used as orthostateswere curved, taken from some unknown round building, 3.54 m. in
diameter;they were 0.86 m. high and 0.81 m. thick. In this section of the wall in general, archi-
tectural memberswere few; also in the filling, which consistedexclusivelyof small fieldstonesand
lime mortar,the absenceof fragmentsof marblesculptureand inscriptionswas noticeable.
The section south of the Peripatos up to the Klepsydrawas first noted by P. Kavvadiasin the
course of excavationsin 1896-1897.'3 It was excavatedsystematically,however, by the American
Schoolin 1937, when it was discoveredthat it followedthe easternmostof the two retainingwalls of
the PanathenaicWay,14which servedto support the mass of earth as the street ascendedfrom the
Eleusinion to the Peripatosand the Klepsydra.
In this area only the lower courses of the Post-Herulian Wall were preserved,seated in a deep
foundationtrench. The destructionof this sectionof the wall must be dated in the Byzantineperi-
od, about the middle of the 11th century, when it appears that systematicdemolitionwas under-
taken to provide material for the constructionof the Rizokastro,15the new wall which encircled
the Acropolis.
North of the Peripatosthe Post-Herulian Wall was preservedto a height up to 2.90 m. At that
point the northwest corner of the Rizokastrowas preserved,showing that at the time of the con-
structionof the Rizokastroall the remainingnorthwardsectionof the Post-HerulianWall was still
in use (PI. 5). This sectionof the wall down to the Hypapanti Gate, which was excavatedin 1937, is
built of all sortsof miscellaneousstones and has undergonerepairsover the long periodof its use.16

SecondSection
The secondsectionof the wall, with a length of 78 meters,stretchesfrom the Hypapanti Gate to
the Christ Church Gate. Only the outer face is preserved.It rests everywhereon the natural rock
and at many points rises to a height of 3.20 m. The inner face was based on the buildings of the
Eleusinion, that is, on the Roman stoa which fixes the southern boundaryof the temenos, on the
1Hesperia 7, 1938, pp. 332-333 (Shear, Sr.).
12Ibid., pp. 330-332.
13 P. Kavvadias,
ILpaKruta 1896, p. 18 and 1897, p. 10.
14 Hesperia 7, 1938, pp. 333-334 (Shear, Sr.).
15 A. W. Parsons recordedthe same opinion ([footnote109 above, p. 29], pp.
259, 263, fig. 40). He, however,dated
the new wall to the end of the 12th or the beginningof the 13th century.
Hesperia 7, 1938, pp. 327-328 (Shear, Sr.).
16
 DESCRIPTION OF THE WALL BY SECTIONS 129

propylonof the Eleusinion, and on a large part of its circuit up to the northwestcornerof the large
retaining wall of the terraceof the Eleusinion. It appearsthat the temple of the Eleusinion was not
destroyedat that time to provideconstructionmaterialfor the wall. The one existing blockfromthe
temple built into the wall was certainlyplacedthere in a later, Mediaeval,repair. Beyondthis point
as far as the Christ Church Gate the wall was based on various monumentsof shell limestone.
The whole of the survivingouter face of the wall was built of porosstonesof all sorts and of small
dimensions, most of them coming from the retaining wall of the Eleusinion. Generally speaking,
this section is preservedonly in poor condition, and the upper portions have undergonein later
times many repairs with small stones and bricksbondedwith mortar.
© American School of Classical Studies at Athens

To the right of the Christ Church Gate (PI. 14:b) are preservedthe foundationsof a rectangular
For personal use only. License: CC-BY-NC-ND.

tower (W 2) projectingfrom the wall 5.86 m. and with a width of 7.65 m. (PI. 14:a).The walls are
1.90 m. thick and are built of large poros blocks and ancient marbles, including a doorjamb of
Pentelic marble. Filling for the interiorconsistedof architecturalmembersincludingthree epistyles
(one of these also carries the lintel of a door or window); also in the filling was a single column
which would have carrieda statue. Four drumsof the column are preserved,of Hymettian marble,
on one of which was carvedthe inscription I 3299 (= IG II2, 3504). Also preservedare the Ionic
base and Ionic capital, both of white marble.17These membersmake it possibleto restorethe total
height of the column as 8.15 m. and the diameteras 0.82 m. The column apparentlystood in the
vicinity,as well as anothercolumn, fluted,of white marble,foundbuilt into the wall near by. Many
similar, isolated columnscarryingstatues seem to have stood in the area of the Agora, especiallyon
the east side. At any rate, a large numberof column drums from such monumentshave been found
in the Post-Herulian Wall in the section from the Christ Church Gate to the north end of the Stoa
of Attalos. Exactly in front of the northwestcornerof the Stoa were four fallen drumsof a column
monumentwith an inscriptionof the 1st centuryB.C. (Agora I 3948).18
The wall and especially Tower W 2 were first excavatedin 1851 by the Greek Archaeological
Society. In their report were described,apart from the above-mentionedcolumn and epistyles, the
following:a roof tile with a gorgoneionin the middleand beneathit the inscription[A]IFIrIYPOY
as well as water spouts with lion's heads, some of Pentelic marble,othersof terracotta.And finally,
80 inscriptions.19Most of the inscriptionscame from the Agora, some stating "to be set up in the
Metro6onor "in the Bouleuterion".These led the first scholarsof Athenian topographyto look in
this vicinity, between the Churchesof the Hypapanti and Christ, for the group of the Metro6n, the
Bouleuterion,and the Tholos,20which are mentionedtogetherby Pausanias.
The whole section of the wall between the Hypapanti Gate and the Christ Church Gate, as well
w as systematicallyexcavatedin the years 1936, 1937, and 1958, bringing a better
understandingof the discoveriesof the mid-19th century.21During this new investigationmany
inscriptionsand architecturalmembers,as well as pieces of sculpture,came to light. Still preserved
in its original position is the inscription I 4713 which stood in front of the propylon of the Eleu-
sinion and was incorporatedin the face of the wall.22 In addition to simple building material,

17
Found in 1851. See Pittakis et al. (footnote6 above, p. 126), fasc. 1, p. 4; fasc. 2, pp. 5-6.
18
Hesperia 19, 1950, p. 318 (Thompson); B. D. Meritt, "Greek Inscriptions,"Hesperia 23, 1954, pp. 254-255,
no. 36.
19Pittakis et al. (footnote6 above, p. 126), fasc. 1, pp. 3-5.
20K. S. Pittakis, L'ancienneAthenes,Athens 1935, pp. 31-31. Cf. also idem,
'E4'Apx 1853, p. 880, note 1.
21 Hesperia 6,
1937, p. 359 and 7, 1938, pp. 326-328, 329 (Shear, Sr.). Thompson, "Odeion,"pp. 95-96.
22J. H. Oliver, "GreekInscriptions,"Hesperia 10, 1941, pp. 78-82, no. 34.
 130 APPENDIX: THE POST-HERULIAN WALL

inscriptionsI 5250,23 I14712, which is still built into the wall, and a large statue base with a dedi-
cation by Herodes Atticus to his friend M. Ceionius Civica Barbarus(I 4924)24 were found.
Also from this sectionof the wall came characteristicfinds fromthe Eleusinion, such as sculpture
and inscriptions,includingmany fragmentsof the Alcibiadesstelai, which are known to have stood
in the Eleusinion. In the filling of the wall near the Eleusinionwas foundthe fine lintel with paint-
ed decoration,of the 5th centuryB.C., now displayedin the Agora Museum (Lioness Lintel).25

ThirdSection
This section of the Post-Herulian Wall extends from the Christ Church Gate to the Pyrgiotissa
© American School of Classical Studies at Athens

Gate (Pls. 7, 8:a). It has a length of 100 metersand a consistentthicknessof 3.50 m. The outer face
was laid over the stylobateof the SoutheastStoa and the Libraryof Pantainos,while the inner face
For personal use only. License: CC-BY-NC-ND.

was beddeddirectlyon bedrock.

The southernhalf of the wall, which was uncoveredin 1958, is preservedto a height of 1.90 m.
and is carelessly built of large poros blocks. Almost all the architecturalmembers used in this
section come from the Southeast Stoa, on which it was based. On the inner face use was made
chiefly of epistyles from the Stoa, while in the filling were chiefly large column bases, Ionic bases,
column drums, and geisa of this unusual colonnade.26
The northernhalf of this section, on the other hand, built over the Libraryof Pantainos,is con-
structedwith exceptionalcare, almostentirelyof architecturalmembers,most fromthe buildingsof
the Agora. Its preservedheight of 3.40 m. was uncoveredin 1933 and 1939.27 Most of the material
used in the constructionwas taken from the Libraryof Pantainos.Especiallyinstructivefor deter-
mining the original position in the Library of the various architecturalmembersis the method of
placing them in the wall. For example, the Ionic base of the south pilaster in the colonnadeof the
Libraryremainsin its original position,and the shaft of the pilasterwas laid horizontallynear it in
the outer face of the wall, near Tower W 3. The crowningmoldingof the pilaster was found near
by; thus the three pieces together gave accuratelythe height of the pilaster. The most important
find, however,which gave the certainidentificationof the building underthe wall as the Libraryof
Pantainos, was the inscriptionI 848,28built into the inner face of the wall about midway between
Towers W 3 and W 4 and which was the lintel of the main entranceto the Library.Epistyles of the
Librarywere used as building materialin both faces of the northernhalf of that sectionof the wall.
Also, in the filling, were column bases, pieces of columnsof Hymettian marble, Ionic capitals, and
many geisa, some of which belongedto the colonnadeof the Library.29
In additionto materialfromthe Library,architecturalmembersof otherbuildingswere found in
this section of the wall, built into both faces and used as filling for the core. Other importantfinds
were sculpture and inscriptions.The architecturalmaterial, in conjunctionwith the ruins of the
buildings in the Agora, permitted their accurate reconstruction.Among the buildings providing

Hesperia 8, 1939, p. 218 (Shear, Sr.).

23

24 Hesperia 7, 1938,
p. 328 (Shear, Sr.).
25 G. P. Stevens, "Lintelwith the Painted Lioness,"Hesperia 23,
1954, pp. 169-184.
26 Thompson, "Odeion,"p.
95. Hesperia 29, 1960, pp. 344-345 (Thompson). For the architectureof the Southeast
Stoa, R. Ross Holloway, "Explorationof the Southeast Stoa in the Athenian Agora,"Hesperia 35, 1966, pp. 79-85.
Travlos, PDA, pp. 436-438.
27Hesperia 4, 1935, pp. 329-334 and 9, 1940, p. 294 (Shear, Sr.).
28 A. W. Parsons, "A Family of Philosophersat Athens and Alexandria,"Hesperia,
Suppl. VIII, Commemorative
Studiesin Honor of TheodoreLeslie Shear, Baltimore 1949, pp. 268-272.
29For the architectureof the Library,see Travlos, PDA, pp. 432-435; AgoraXIV, pp. 114-116.
 DESCRIPTION OF THE WALL BY SECTIONS 131

material for the wall were the Metro6n,30the Temple of Ares,31the Odeion of Agrippa,32the
Southeast Temple,33and the Southwest Temple.34In a great many cases, however, the source of
many of the architecturalmembershas not yet been determined.
About midpointbetween the Christ Churchand PyrgiotissaGates, Tower W 3 projected6.94 m.
from the wall (PI. 9:b). Its width was 8.77 m.; its walls, 1.55 m. thick, were built of ancient blocks
and architecturalmembers.Of special interestwere some Ionic columnsof the 5th centuryB.C. from
some still unidentifiedbuilding.35One of these columns, 6.67 m. in height, has been re-erectedand
now standsin the south stairwell of the Stoa of Attalos. Insidethe sametower were foundsignificant
numbersof pieces of sculpture.36The architecturalmembersfound in the tower, along with others
© American School of Classical Studies at Athens

fromthe wall itself, are now displayedin frontof the wall, between Towers W 3 and W 4.
Still anothertower (W 4) stoodin the northernpart of the sectionof the wall underdiscussion.Its
For personal use only. License: CC-BY-NC-ND.

dimensions were 9.70 x 7.50 m. and it was preservedto a height of 3.45 m.37 It was built over a
large monument base measuring 5.15 x 5.35 m. The still preservedfilling consists of drums of a
monumentalcolumn similar to that found in Tower W 2 and poroscolumndrumsfromthe Middle
Stoa. Into the north side of the tower is built a large curvedblock of white marble which, from its
dimensions, may certainly be attributedto the triumphal arch, of which the bases are preserved,
leading from the Greek to the Roman Agora. In the northeastcornerof the tower are incorporated
many inscribedbases (I 754, I 928, I 5850, I 6737, I 6738).
Fourth Section
This sectionof the wall, the last on the west flank, stretchesthe entire length of the Stoa of Attalos
and is no longer preserved.Some elements,however, survivewhich enable us to follow in detail its
course and method of construction. The
first investigationin this vicinity took place in the years
1859-1862,1868-1869, 1874,1890, and 1898-1902 underthe auspicesof the Greek Archaeologi-
cal Society.38The sections of the wall and its towers discoveredat that time were demolishedin
orderto uncoverthe remains of the Stoa of Attalos and also to removefrom the masonryof the wall
the inscriptionse,sculpture,and architecturalmembers built into it. These were not taken to the
National Museum and its storeroomsbut were laid out near the Stoa and on its floor, coveringit
completely. Smaller fragments of special interest were stacked in piles to a height of two meters
inside the three southernmostshops of the Stoa.
The excavation of this section of the wall as a whole was accomplishedlater by the American
School, especially in the work precedingthe restorationof the Stoa of Attalos.39To begin with, all
the ancient material uncoveredduring during the earlier work of the ArchaeologicalSociety was
removedfrom the site, and systematicexcavationof the ruins of the wall was started.These were
30
Thompson, "WestSide,"p. 185.
31
Dinsmoor (footnote9 above, p. 97), p. 3, note 9; McAllister (footnote9 above,p. 97).
32
Thompson, "Odeion,"pp. 31-141.
33 Agora XIV, pp. 167-168; Dinsmoor (footnote 19 above, p. 3), pp. 410-452 (for a revision of the previous iden-
tificationof the temple as the Temple of Demeter).
34 AgoraXIV, pp. 165-166; Dinsmoor, op. cit.
35Hesperia 29, 1960, pp. 351-356 (Thompson);Agora XIV, p. 166.
36 Harrison, (footnote46 above, p. 7), pp. 373-378. Hesperia 29, 1960, pp. 350-359
(Thompson).
37 Stephen G. Miller, "A Roman Monument in the Athenian Agora,"Hesperia 41, 1972, p. 94.
38 rFEVLK7 vveAXEvo-s1860, pp. 8-14; 1861, pp. 14-17; 1862, pp. 7-10; 1869, pp. 7-9; 1870, pp. 7-8. 'XLATrWp 1,
1861, pp. 265-272. flpaKTrKd 1874, p. 20; 1890, pp. 9-10; 1898, pp. 65-68; 1899, pp. 73-75; 1900, pp. 32-34; 1902,
p. 46.
39Hesperia 19, 1950, pp. 317-322 (Thompson).
 132 APPENDIX:THE POST-HERULIANWALL

then demolishedin orderto uncoverthe remainsof the Stoa of Attalos. Later, duringthe restoration
of the Stoa of Attalos, all the ancient materialsrecoveredby the Greek excavationsand those of the
American School were arrangedin classifiedpiles in front of the Stoa, coveringthe whole area on
either side of the PanathenaicWay for the length of the Stoa and the Odeion of Agrippa as far as
the Church of the Holy Apostles. The smaller pieces of sculptureand architecturewere placed on
shelves behind the Stoa and in the empty space at its north end.
This section of the wall, the northernmostof the western flank, coveredthe whole of the shop-
frontwall of the Stoa and also its narrowsouth end (PI. 11). Two towers at the ends (W 5 and W 7)
and one in the middle (W 6) completedthe fortificationof this most importantpoint in the north-
© American School of Classical Studies at Athens

west cornerof the Post-Herulian Wall.

The Stoa of Attalos, whose ruins were incorporatedin the wall, had been destroyedin the Heru-
For personal use only. License: CC-BY-NC-ND.

lian attackof 267. The building had been set on fire, with the result that the roof collapsedand the
stone was turned to lime, destroyinglarge sectionsof the wall with many of its architecturalmem-
bers. In spite of the fire, however,at the time of the constructionof the Post-Herulian Wall the Stoa
stood in comparativelygood condition, in which it remained even in some places to its original
height until the reconstructionof the Stoa.
At the time of the constructionof the Post-Herulian Wall, it seems that all the interior walls of
the Stoa, with the dividing shop walls, were still standing.These were probablyre-used to house
the guards and for other necessarypurposes. Only the exterior and interior colonnadeswere torn
down at that time, to be used in the constructionof the wall. The same fate befell the retaining
wall of the terrace in front of the Stoa and the monumentsalong its length, all stripped to their
foundations.
The Post-Herulian Wall was laid beside the shop-front wall of the Stoa of Attalos and rested
everywhereon the soft bedrockin spite of the great differencein its level, which at the north end
was 5.50 m. below the floor of the Stoa. On accountof the stability of the rock the trenchin which
the wall was beddedwas dug 0.20-0.40 m. deeperthan the foundationsof the Stoa. Also, because
the foundationof the shop-frontwall projected0.75 m., the thicknessof the Post-Herulian Wall
near the bedrockwas 1.20 m., whereas higher up, over the floor of the Stoa, it was 1.94 m. To this
last dimensionshould be addedthe thicknessof the shop-frontwall, i.e., 0.66 m. to reachthe actual
thicknessof the Post-Herulian Wall, that is, 2.60 m.
To the south of Tower W 6 the material used for the constructionof the wall was for the most
part poros column drums from the Middle Stoa, laid at right angles to the wall, one next the other
with their surfacestouching (Pl. 8:b). To the north of Tower W 6, however,the building material
was exclusuvely architecturalmembers from the Stoa of Attalos: Doric column drums, capitals,
epistyle blocks, triglyphs, and geisa from the lower storey of the Stoa, as well as Ionic columns,
parapets, epistyles, and geisa from the upper storey. Among them were many fragmentsof bases
and capitals from the inner ordersof both colonnades.
It is to be noted that it was in this section,in front of the tenth shop from the north, to be precise,
that in 1862 the discovery was made of the Doric epistyle with the inscription BAEIAEYE
ATTAAO? . . . which establishedwith certaintythe identificationof the building as the Stoa of
Attalos. Near by were found other inscribedfragmentsof the epistyle.40If the place of discoveryof
the epistyle correspondsto its original position, it should be possible to determinemore accurately
the relationof the inscriptionto the axis of the colonnade.41
?rvvAte?vws1860, pp. 7-10. Cf. also hAtlrrwp 3, 1862, pp. 365-367.
40 rFVLiKu

On the inscription and a probable restoration, see B. D. Meritt, "Greek Inscriptions,"Hesperia 26, 1957,
41

pp. 83-88.
 DESCRIPTIONOF THE WALLBY SECTIONS 133

Many ancient blocks were found also along the narrow side of the Stoa of Attalos. And here, for
the beddingof the wall, a trench 0.89 m. deep and varying in width between 2.32 and 2.47 m. was
opened in the bedrockwhich at that point formedthe floorof the Stoa. In the trenchwere foundthe
large blocks of white marble from the crowningmemberof a base, which are now displayedalong
with other similar blocks east of the East Building, near the Church of the Holy Apostles. In the
same trench were found pieces of Ionic columns from the upper storey of the Stoa of Attalos and,
most important,the crowningmoldingof a pilaster belongingto the neighboringarchedopening of
the south exedra of the Stoa. This molding, in spite of the damage by fire, has been used in the
restorationof the Stoa, in its original position.
© American School of Classical Studies at Athens

Tower W 5. Of the three towers which existed in this sectionof the Post-Herulian Wall over the
length of the Stoa of Attalos, the southernmost(W 5), together with Tower W 4, constitutesthe
For personal use only. License: CC-BY-NC-ND.

PyrgiotissaGate. Tower W 5, whose walls apparentlywere still standingto a considerableheight,

was demolished during the excavations of the ArchaeologicalSociety in 186042to uncover the
southwestcornerof the Stoa of Attalos,the ruins of which had been incorporatedin the tower. Thus
the entire south pilaster of the Stoa with its crowning molding was revealed along with the sole
survivingsection of the crepidomaof the colonnade,includingthe stylobate,the two steps, and the
poros drain in front of them. Also preservedto a height of 1.40 m. is a sectionof the terracewall of
the Stoa.
The tower measures 7.50 x 9.90 m., with the thickness of the walls estimated at 1.50 m. The
masonry of the lower part consistedentirely of blocks coming from the base of a large monument,
4.841 x 3.899 m., while in the upper part, arrangedas orthostates,were placedslabs inscribedwith
ephebic lists, 1.617 m. high, taken from the same monument.43These orthostateswere crowned
with a tainia of the same stone on which, on the west side of the towers, was carvedthe inscription
IG II2, 5200, referring to the constructionof a wall believed to be the Post-Herulian.44The in-
scription and the architecturalmemberstaken from the monumentand used in the constructionof
the tower are now displayedto the west of the position of Tower W 5.
During the Turkish period and after the abandonmentof the wall the interior of the tower
was used as the Church of the Virgin, who quite naturally took on the epithet in this case of
Pyrgiotissa.45
Tower W 6. Tower W 6 stood in about the middleof the Stoa of Attalos and was built aroundthe
base of column 11 of the inner colonnade,countingthe columnsfromthe south (PI. 11:a). The ruins
of the tower, which projected7.30 m. from the wall, had a width of 6.60 m. and are estimatedto
have been preservedto a height of 7.85 m. fromthe floorof the Stoa. The tower, except for its foun-
dations, was demolishedin 1900 (PI. 12).46It was built of regularblocksof marble from the quar-
ries of Hymettus all taken from one and the same monument.These were gatheredtogether in a
heap a little to the west of their position.
As far as can be judged from surviving drawings and an old photograph,the lowest part of the
tower, to a height of 3 meters, consistedof two coursesof orthostatesdividedby a low string course.
The upper masonry,of which ten coursessurvived,was built in a pseudo-isodomicfashionof blocks
alternating in height between 0.65 and 0.325 m. On one of these blocks on the west side of the
42 FEVLK) o'vvefXEVO't
1860, p. 9.
43Miller (footnote37 above, p. 131).
44For the inscription,see above, p. 9.
45 For the church, see Mommsen (footnote 117 above, p. 73), pp. 92-96, no. 110; A. Xyngopoulos, EMME I,
pp. 110-111.
46
1900, pp. 33-34.
K. Mylonas, IHpaKTrKa
 134 APPENDIX: THE POST-HERULIAN WALL

tower, as shown on the old drawings,was carvedthe inscriptionIG II2, 4209, mentionedby several
travelersof the 18th and 19th centuries.
In the course of studying the blocks at the beginning of the reconstructionof the Stoa of Attalos
two more inscriptionswere found on cornerblocks (Agora I 6120 a, b), giving the beginning and
end of one inscription. From these it was possible to determinethe width of the monumentfrom
which the blockswere taken, probablyabout 4.10 m.
Finally, it was establishedthat the blocks in the tower, 115 in number, came from a large, high
monument which originally stood exactly in front of the position of the tower, on a large base
5.80 m. square, near the retainingwall of the Stoa of Attalos. The monumentwas dedicatedto the
© American School of Classical Studies at Athens

donor of the Stoa, King Attalos II, who was representedin a four-horse chariot. Almost all the
blockson which tracesare preservedof the attachmentsfor the feet of the horseswere placedon the
For personal use only. License: CC-BY-NC-ND.

base of the monument. In the Roman period, as shown by the inscription, the monument was
dedicatedto the EmperorTiberius.47A similar case may be seen in the monumentof Agrippa on
the Acropolis,which was originally dedicatedto King Eumenes II of Pergamon.
The foundationsof Tower W 6, which were found at a depth of three metersbelow the floor of
the Stoa of Attalos, were uncoveredin 1949 and subsequently demolished.48They were 1.40-
1.60 m. thick and were built of poros architecturalmembersfrom the Middle Stoa. The identi-
ficationof the sourceof the materialwas certainbecausesimilar memberswere found in the whole
area of the Middle Stoa.
Inside Tower W 6 and the adjacentportion of the wall were found 97 column drums of the
Middle Stoa, which were built into the side of the wall. In additionto the column drumsthe tower
producedDoric capitals,crowningblocksfromthe high parapetbetweenthe columnsof the Middle
Stoa, epistyle blocks, triglyphs, metopes, and Doric geisa; the capitals and metopes are of white
marble,the rest are of poros. Finally, there were many fragmentsof terracottaantefixeswith lions'
heads. Today, except for the epistyles and capitals, which have been placed on the terraceof the
Middle Stoa toward its east end, all the architecturalmembersfound inside the tower are displayed
on their own shelves behind the Stoa of Attalos.
In additionto the materialfrom the Middle Stoa the demolitionof the foundationsof Tower W 6
producedarchitecturalmembersfrom the Temple of Ares and the Odeionof Agrippa;49also, there
were about 50 poros blocks, 1.30 x 0.57 x 0.43 m. On the short side of these were preservedthe
remains of a good and strong plaster, similar to that used on the inner surface
the of walls of the
southeastcornerof the cryptoporticusof the Odeion of Agrippa. All these blocks,except two which
have been kept as samples,have been used in the constructionof the large retainingwall behindthe
Stoa of Attalos. It seems likely that they were originallyused in the inner walls of the cryptoporticus
formingthe rectangleof the Odeion.
Tower W 7. The wall built in front of the shop-frontwall of the Stoa of Attalos did not reach to
the northern end of the Stoa but stopped over the vertical foundationwhich marked the end as
planned before a change was made in the course of construction,adding three shops to the north.
Directly over this foundation(3.80 m. in thickness)there remaineduntil the end of the 19th century
a section of the Post-Herulian Wall, 1.70 m. thick, which appears on an old photograph5?and

47 On the
inscriptionand the monument,see E. Vanderpool,"AthensHonors the EmperorTiberius,"Hesperia 28,
1959, pp. 86-90.
48Hesperia 19, 1950, p. 317 (Thompson).
49
Ibid., p. 319.
50 F.
Adler, Die Stoa des Konigs AttalosII. zu Athen, Berlin 1875, pl. I.
 DESCRIPTIONOF THE WALLBY SECTIONS 135

which is included on a plan drawn by A. Lykaki in 1899 in the course of excavations by the
ArchaeologicalSociety.51
This section of the wall comprisesthe south side of a large tower, 17.20 x 14.70 m. Its north side
followed the northernstairway of the Stoa of Attalos, with a thicknessof 2.80 m.; the boundaryof
the west side was the crepidomaof the stylobate of the Stoa and was 1.80 m. thick. It seems that
originally, in the constructionof the tower, it was planned to reinforcethe east and west walls by
building new walls. Deep foundationtrenches were found during the excavationsof 1954, one in
front of the shop-front wall and the other behind the stylobate. The trenches were 1.40-1.50 m.
wide; the first reacheda depth of 2.10 m. and the second3.60 m. from the floor of the Stoa. Only a
few column drums from the Stoa of Attalos, however, were laid in one trench, the western, when
© American School of Classical Studies at Athens

work stopped and the trenches were covered with Herulian destructionfill. This fill contained
For personal use only. License: CC-BY-NC-ND.

ashes, carbonized wood, and many bits of bronze and iron as well as many small fragments of
architecturalmembersof the Stoa of Attalos. It is clear fromthe nature of the fill and from the coins
found in it that the trencheswere filled in the courseof the constructionof the Post-Herulian Wall.
The height of this tower must have been impressivebecauseat the north end of the Stoa the floor
was 6.55 m. higher than the groundin front and the ancient road approachingit. Given that at the
time of constructionof the wall the terraceoutside the Stoa had been destroyedas well as the cre-
pidoma as far as Tower W 5,52Tower W 7 must have risen to a height of about 18 meters.
The masonry of the tower, to judge from the old photographsof this section (PI. 11), is very
characteristic.As in the other towers, the lower part was laid in orthostates,the upper with regular
blockslaid horizontally.In the interiorof the tower, spacedat regularintervals,were discoveredthe
foundationsof four rectangularbases, the length of whose sides variedfrom 1.10 to 1.70 m. (PI. 13).
These bases, which were built of large and small stones with an abundanceof strong lime mortar
and ash-colored clay, were not seated on the bedrockbut on the firm hard-packedfilling of the
remains of the classical law courts 3.60-4.45 m. below. These bases supportedsquare piers, esti-
mated at 0.80 m. to the side, to carryan upper storey.
Each storey would have consistedof one large room, 12.60 m. square, and on its east side three
rooms correspondingto the ancient shops on the first and secondfloorsof the stoa. Communication
between the two floorswould have been providedby the still existing northstairwayof the Stoa. The
whole complex of the tower would have been accessiblefromthe east, that is, fromthe city side, by a
doorwaywhich must have been openedfor that purposein the east wall of the third shop of the Stoa.
The tower must have had a secondsmall doorwayin the south side. The existenceof a small doorway
at this point may be hypothesizedfrom the constructionof the south wall of the tower, which ter-
minateswith doorjambstowardthe east.53The width of this doorwaydid not exceed 1.40 m., that is,
a little narrowerthan the doorwayfound at the east end of the north side of the wall.54
The floor of the ground-floorroom of the tower was pavedwith a mosaicof shoddyconstruction,
consisting of small irregular pieces of white marble set in an underpinningof small stones and
abundant mortar, gray in color. It is not certain whether this floor was contemporarywith the
constructionof the tower or slightly later. Nevertheless, the fact that it was confined within the
51 IlIpaKTLKa1899, pl. II.
H. A. Thompson (Hesperia 19, 1950, p. 320) writes that the blocks of the Stoa stylobatewere taken away in the
52

10th and 11th centuries. But fill excavatedat this point more recentlygives no indicationof the time of the removalof
the blocks.
53 The doorjambwas found at a distance of 3.17 m. from the shop-front wall of the Stoa. See R. Bohn, Die Stoa
Konigs Attalosdes Zweiten zu Athen, Berlin 1882, pl. I. Four blocks of the jamb were preservedin the 19th century.
See Adler (footnote50 above, p. 134).
54Below, p. 137.
 136 APPENDIX: THE POST-HERULIAN WALL

limits of the tower and especially that it was laid 0.10-0.14 m. lower than the stylobateof the Stoa
permits the assumption that it was more likely to have been contemporarywith the tower.55A
similar mosaic has been found in the East Building of the South Square (Hesperia 22, 1953, p. 37
[Thompson]) and in Temple II of Zeus and Athena on the west side of the Agora (Thompson,
"WestSide,"pp. 86, 90).
Two bases of white marble slabs, placed near the northernpiers at the time the mosaic was laid
and reinforcedwith large rectangularblocks,were obviouslyused to receivecertainobjects.We do
not know what purpose was servedby the first-storeyroom of the tower. It is, however,very prob-
able that some civic officesor religiouscults were installedhere to replacesometemplesor buildings
destroyedby the Heruli.
© American School of Classical Studies at Athens

In any case, the tower complex must have had some special significance since, according to
For personal use only. License: CC-BY-NC-ND.

Athenian tradition, in this northernmostpart of the Stoa of Attalos, formerlyknown as the Gym-
nasium of Ptolemy,there was said to have been the residenceof the Metropolitanof Athens and the
seat of his bishopricin the Byzantineperiod.56

NORTH FLANK
The north flank of the wall has not yet been discoveredover its entire length. From the preserved
sections,however, and from other bits discoveredduring new constructionit is possible to trace its
course accurately.
The north flank measured470 meters in a straight line and consistedof three separatesections:
the first, the western, with a length of 146 meters,stretchedfrom the northeastcornerof the Stoa of
Attalos to the Libraryof Hadrian. The second,or centralsection,includedthe Libraryof Hadrian,
and the third and last began at the northeastcornerof the Library and continuedfor a distanceof
206 metersto the east, where it turned southward.

WesternSection
The wall takes off at an acute angle from the east side of the north end of the Stoa. From that
point it is preservedfor a length of 50 meters,with a thicknessof 3.50 m. and a height of 7.50 m. It
is built of ancient blocks of all sorts, most of them architecturalmembers of poros, including a
triglyph from the Middle Stoa. For the first 18 metersthe outer face is beddedon the north wall of
the classical law courts,whereas the rest is based on the natural rock. This first sectionwent with-
out a break to the east until it met the southwest cornerof the Library of Hadrian, whose entire
area of 114 x 76 m. is enclosedwithin the new fortificationwall.

CentralSection
At the time of the constructionof the Post-Herulian Wall the Libraryof Hadrian must have been
in very good condition. Still in the 15th century, as can be deduced from the descriptionby the
Anonymousof the AmbrosianLibrary,who, as we have seen, was in Athens in 1466, the whole of
the porch of the Library was preserved,with its four columns,and the whole of the marblewall of
the west facade;only the columnsof the southernpart had disappeared.
55H. A.
Thompson (The Stoa of AttalosII in Athens [Agora Picture Book No. 2], Princeton 1959 and Agora XIV,
p. 104) considersthe mosaic to have been the original floor of the Stoa of Attalos. I have many reservationsand would
like his opinion. In any case, it must be remarkedthat in the first plans the mosaic was indicatedin the east end of
Tower W 7 and above the trench dug, as we have seen, for the constructionof the Post-Herulian Wall. (Cf. Bohn
[footnote53 above, p. 135], pl. I; HpaKcrtca1899, pl. I.)
56 Pittakis
(footnote 114 above, p. 73), p. 1469, no. 3035. Mommsen (footnote 117 above, p. 73), p. 94. See also
D. Kambouroglou, Al fIaAaLa' 'AO^ivat, Athens 1922, p. 154.
 DESCRIPTION OF THE WALL BY SECTIONS 137

The most importantcontributionof the descriptionof the Anonymous,however, is that it shows

clearly that the whole of the Libraryof Hadrian was part of the defense circuit.The author writes
that one entered the fortified
area the through propylon and characterizesthe three projecting
exedraeof the north wall as towers of the circuit.57In his descriptionof Athens the Anonymousalso
remarksthat many ancient remains could be seen both inside and outsidethe recentlyfortifiedarea
(terra ultimamente murata). Earlier students of the topographyof Athens, relying on this obser-
vation of the Anonymous, date the wall in the Frankish period. There is no longer any doubt,
however, that the new fortificationwas the Post-Herulian, which excavationshave shown under-
went drasticrepairs in the Frankish period.
© American School of Classical Studies at Athens

The outer walls of the Library constitutedthe centralsection of the north flank of the wall. It is,
however, very probablethat one side of the Library,the south, which is still hidden under modern
For personal use only. License: CC-BY-NC-ND.

houses, was strengthenedfor further defense. Communicationbetween the Library and the area
enclosedby the Post-Herulian wall seems to have been achievedby a door in the south side of the
quadrangle,probablyat the position of the middle rectangularexedra.
Eastern Section
From the southeast corner of the Library the wall continued fart he only sections
visible in this area today are the part preservedin the courtyardof the Medresse and some bits
along Hadrian Street (below). The section in the Medresse has a thickness of 3.50 m., and its
construction of large poros architecturalmembers follows in general the characteristicmethod
describedabove.58
Although few bits of the wall are preservedtoday, nevertheless its course eastward from the
Library of Hadrian can be determinedsafely thanks to discoveriesmade during constructionof
houses along Hadrian Street. Thus as shown by the most recent excavations,at a distance of 50
metersfrom the southeastcornerof the Libraryof Hadrian, for the continuationof the wall use was
made of the north side of another older building, 66 meters long, the Pantheon.59 Some remains of
this were uncovered at 78 Hadrian Street. Still another fragment of the wall was discovered farther
east in the cellars of the houses to either side of the cornerof Mnesikles and Hadrian Streets.This
was preservedto a length of 10 meters, bedded on the hard-packedcrepidomaof the porch of the
Pantheon.60Also a small gateway, 1.65 m. wide, along with a piece of the wall, 3.40 m. thick, was
found in 1973, during the constructionof a house at 94 Hadrian Street.

EAST FLANK

The total length of the east flank has been estimated at 255 meters. Only a small section of 70
meters, near the no longer existing Church of St. Demetrios Katiphori, was standing until 1861
when it was demolished by the Greek ArchaeologicalSociety. Only a small bit at each end was
spared.The wall at this point was standingto a height of five meters,approximately.Beforethe ex-
cavationit was consideredto be part of the Prytaneion,but after the excavationit was recognizedas
belonging to a fortificationwall. In it were found many architecturalmembers,pieces of sculpture,

57Ziebarth (footnote98 above, p. 71), p. 77.

58
The section inside the Medresse was discoveredin 1914 by A. Keramopoullosduring the demolition of the
Medresse. He, however, describedonly the architecturalmembersand sculpture found inside the Medresse without
mentioning specifically anything about the wall: AeAT 1, 1915, Hapaprnjua,p. 55 and idem, <<'AvarKaa'L
7rapaTO
QflpoXoytov 'Av8povLKOv TOV Kvpp2To-Tov?, 1paKTLKa 1914, pp. 125-126.
59Travlos, PDA, pp. 439-443.
60 G.
Dontas, "More about the New Hadrianic Building in Athens,"AAA 2, 1969, pp. 1-3.
 138 APPENDIX: THE POST-HERULIAN WALL

and inscriptions.6 Many large marbleswere left on the spot and are still keptthere in a fencedenclo-
sure,62while the more importantsculptures,inscriptions,and fifty or so inscribedhermswere trans-
ferredto the archaeologicalcollectionsof that periodand later to the National Museum.63
In view of the discoverywithin the wall of inscriptionsrelative to the Gymnasiumof Diogenes,
that district has been identified as the Diogeneion, in spite of the fact that no trace of an ancient
building has so far been discoveredthere.64In some otherinscriptionsthe Gymnasiumof Ptolemyis
mentioned, in another, the Stoa of Romaios (IG II2, 958), so that these buildings, too, must be
sought in this district.65The most important piece of sculpture was a group of Theseus and the
Minotaur (E.M. 1664 + 1664a) which, along with other indications,permits the assumptionthat
© American School of Classical Studies at Athens

the famed Theseion was in the same area.66

For personal use only. License: CC-BY-NC-ND.

Except for the section of the wall just described,its entire course is unknown becausethe whole
districtis coveredwith modernhouses. Only a few blocksin situ south of the Panagios Taphos can
be thoughtto have belongedto a continuationof the wall. The courseof the wall on the east side has
been indicatedhypotheticallyon the plan. It is a fact, however,that it must have terminatedat the
northeastcornerof the Acropolis.
In concludingthis descriptionof the wall it should be notedthat some scholarsof the topography
of Athens have tried to connectthe Post-Herulian Wall with the Mediaevalwall discoveredin 1914
during the excavationof the Odeion of Perikles.67The opinion was expressedthat the Post-Heru-
lian Wall turned eastwardand includedthe south side of the Acropolis.The wall discoveredon the
south side of the Acropolis,however, is securely dated by its constructionnot earlier than the 11th
century and is part of the so-called Rizokastro,which surroundedthe Acropolison all sides in the
Byzantine period.68Consequently, it has no relation to the Post-Herulian Wall, since both west
and east flanks of the earlier wall could be seen to terminateat the steep rockof the Acropolis.

THE GATES
Of the eight gates estimatedto have existed in the Post-Herulian Wall, the positionsof only five
are certain. Since these five all coincidewith the point of entranceof importantarteries,however,
the positions of the other three have been restoredon the plan following the same principle.There
are no literaryreferencesto the gates, nor even local traditionsabout them or their names. As noted
above,they are designatedhere by the names of the churcheserectedover or besidethem as follows:
Hypapanti, Christ Church, Pyrgiotissa(west flank);Asomatasta Skalia, Megali Panagia, Panagia
Krystalliotissa(north flank); St. Demetrios Katiphori,Panagia Sarres (east flank).

o-vve'Xevors 1861, pp. 18-20, 37. A picture of the wall before demolition and a brief description were
61rEvlK7

published by E. Breton, Athenes, Paris 1862, pp. 261-262. Many inscriptionsfrom St. Demetrios Katiphori were
publishedby Koumanoudesin tAtXl'orwp 2-4, 1862.
62 The piles of small marbles in the enclosurecome from the demolitionof houses in the neighborhoodof the Plaka.
63 For the herms, see P. Graindor, "Les cosmetesdu musee d'Athenes,"BCH 39, 1915, pp. 241-401; E. Lattanzi,
I ritrattidei cosmetinel Museo Nazionale di Atene, Rome 1968.
64
FeVLK,tJo-VVEXevoa-s1863, p. 7.
65
Travlos, PDA, pp. 578-579.
66 For the discovery the sculpture,P. Pervanoglu,Archiologische
of Zeitung 24, 1866, pp. 160-161, no. 208. For the
Theseion, Travlos, PDA, pp. 578-579.
67 P.
Kastriotis,<<To 'f2ISEoV rov I epLKXeovs>,'ApX'E4 1914, pp. 148-151; G. Guidi, "II muro Valeriano a
S. Demetrio Katiphorie la questionedel Diogeneion,"Annuario4-5, 1921-1922, p. 35.
68Above, pp. 125, 128.
 DESCRIPTION OF THE WALL BY SECTIONS 139

WestFlank
Hypapanti Gate. Discovered in 1938 under the floor of the church during its demolition.The
gate was located at the entrance of an ancient road, six meters wide, on the north side of the
Eleusinion; underneath it runs a water channel of the Roman period. The gate had a width of
2.94 m. Both sides were preservedand also part of the marblethreshold,on which were tracesof its
stonejambs, as well as the rectangularsocketfor the pivot of the north leaf of the door.69
During the demolition of the church, built into the northernanta of the interior colonnade,in
front of the sanctuary,were found the elements of a marble door frame, richly decorated.Similar
architecturalmemberswith the same decorationwere used to frame the sides of the Krystalliotissa
© American School of Classical Studies at Athens

Gate in the north flank of the wall. The use of these in the repair of the latter gate, which is known
For personal use only. License: CC-BY-NC-ND.

to have occured in the time of Justinian,70permits the assumptionthat the Hypapanti Gate was
repairedat the same time.
Christ Church Gate. Discovered in 1959.71The gate cuts diagonally through the wall and is
2.87 m. wide (PI. 14:b).The angle was imposedby the directionof an incomingroadof great antiq-
uity, under which runs the aqueductof Peisistrateantimes. The road comes in on the north side of
the Eleusinion, where it appears to join the other ancient road which enteredthrough the Hypa-
panti Gate. The two roads then continuedeastwardas one road, six meterswide. This is evidently
the Street of the Tripods, which is known to have terminatedat the Odeion of Perikles and the
Theater of Dionysos.
The Christ Church Gate was protectedfrom the south by Tower W 2.72 Part of the marble
threshold was preserved,along with traces of the attachmentsfor the doors and door jambs. In
front, in the angle formed by the gate and the tower (PI. 14:c), was the small shrine of Hekate
describedabove (p. 8).
The gate must have been out of use during the Mediaeval period because it was discoveredto
have been closedby a wall of 13th-centurydate composedof small stones and clay.73
Pyrgiotissa Gate. This was the most importantof the entrancesin the west flank of the wall. It
was built on the great paved street connectingthe Greek and Roman Agoras (PI. 14:d). With a
width of 2.90 m., it was protectedby two towers (W 4 and W 5), only one of which survives,the
other having been demolishedin 1860.74
Under the south edge of Tower W 4 and contemporarywith the gate was a rectangulardrain to
carryoff rain water. Behind it were preservedthe first steps of a stairwaywhich apparentlyled up
to the top of the wall and to the two towers of the gate. Similar stairwaysmust have existed at other
points in the wall and the towers.
North Flank
Asomatasta Skalia Gate. The gate in question is the propylon and entranceto the Library of
Hadrian which, as is clear from the descriptionby the Anonymous of the Ambrosian Library,
continued even into the 15th century to be one of the main gates leading into the area enclosed
within the Post-Herulian Wall. At that time the entire porch was intact, with its four Corinthian
columns frontingthe large entranceway, 3.10 m. wide.
69
Hesperia 8, 1939, p. 221 (Shear, Sr.).
70
Below, p. 140.
71
Thompson, "Odeion," p. 95.
72 Above, p. 129.
73Thompson, "Odeion,"pp. 95-96.
74Above, p. 131.
 140 APPENDIX: THE POST-HERULIAN WALL

In the 1l1thor 12th centurythe Byzantinechurchof the Asomatawas erectedover the north side
of the propylon,includingthe steps, from which the churchtook the epithet "staSkalia".
Gate of the Megali Panagia. In the southeastcornerof the Libraryof Hadrian we should expect
to find anothergate, whose positionshould probablyhave been determinedby an importantancient
east-west road linking the east and west parts of the city. The road apparentlypassedbetween the
Library of Hadrian and Roman Agora. Sectionsof the road, whose width must have exceededsix
meters,have been found along the south side of Hadrian Street.75
KrystalliotissaGate. This gate is found in a courtyardon the south side of Hadrian Street,ca. 100
meters east of the Library, and is preserved in excellent condition. Both door jambs and the
© American School of Classical Studies at Athens

thresholdsurvive,giving a width of 1.98 m. and a height of 2.50 m.

Both the gate, which is richly decorated,and the adjacentsection of the wall can be dated, as
For personal use only. License: CC-BY-NC-ND.

G. Sotiriourightly observed,in the time of Justinian.76The constructionof this sectionof the wall
and of the gate differs sharply from that of the Post-Herulian Wall. The wall at this point has a
thickness of about one meter and is built of large regular blocks interspersedwith smaller stones
and bricks.This Justinianian section was placed 8.50 m. south of the outer north wall of the Pan-
theon, which, as we have seen, originallyconstitutedthe continuationof the Post-Herulian Wall.
It is very probable that the Justinianian gate was built in this location as a substitute for an
earlier one. The question can be resolvedonly by the acquisitionand excavationof the property.
In the 16th or 17th century,when the wall and gate had gone out of use, the Churchof the Pana-
gia Krystalliotissawas built inside. Its ruins are still preserved.77
East Flank
Saint Demetrios Gate. The placing of a gate in the modern KyrrhestesStreet is justified on the
one hand by the need of an entrance in the east flank and on the other hand by the inscription
IG II2, 5199, which was foundbuilt into the wall near the now demolishedChurchof St. Demetrios
Katiphori.78
The inscriptiondates from the secondhalf of the 3rd centuryand refersto the buildingof a wall.
That is, it correspondsto the inscriptionIG II2, 5200, which was carvedon a block from Tower
W 5, in front of the PyrgiotissaGate.79One mayjudge then that IG II2, 5199, too, was placed in a
conspicuousposition, surely near anothergate of the Post-Herulian Wall.
Panagia SarresGate. With even greatercertaintyperhapswe may place anothergate in the east
flank of the wall. This should be near the now demolishedchurchof the Panagia tou Sarre,80at the
point where the ancientStreetof the Tripods passed (above,p. 139). The streetmust have still been
an importantarterywhen the Post-Herulian Wall was built.

Smaller Gates
In additionto the gatesjust discussed,smallerand less importantgates must have been openedin
the wall, whose numbercannotbe determined.One of these, with a width of 1.65 m., was found in
75G. Dontas,AEAT24, 1969,B' 1(1970),p. 23.
76 ic^ IXoAi', Athens 1920, pp. 3-13; idem,
G. Sotiriou, 'E/38op0jKovrTa7revraaerf7ps rtjs 'PtCapedov'EKCKA7raorLKt
EMMEI, pp. 28-30.
77Xyngopoulos(footnote45 above,p. 133),p. 106.
1861,p. 37. Cf. Bodnar(footnote52 above,p. 9), pp. 175-176.
78 rFEVK) oVVwAEwLvS
79Above,p. 9.
80
Travlos, 0HoAeoboLK?I, pl. XII:23.
 DESCRIPTION OF THE WALL BY SECTIONS 141

1973, during the constructionof a house at 94 Hadrian Street.Anothersmall gate probablyexisted

in the south side of Tower W 7, at the north end of the Stoa of Attalos.81

COMMUNICATION BETWEEN THE CITY AND THE ACROPOLIS

With the building of the Post-Herulian Wall the city was limited to the north side of the Acro-
polis, and it is clear that there was no direct communicationwith the Acropolisitself. It is a fact,
however, that in case of need accessto the Acropoliscould be providedby the small ancient gate in
the north side of the Acropoliswall, near the Caves of Pan and Apollo V67r MaKpas'.82Still another
entrance was providedin the north side of the Acropolis rock by the secret stairway through the
© American School of Classical Studies at Athens

Aglaurion. This passage was originally used for descentto the Mycenaean spring in the depths of
the cave; later, and up to the Turkish period, it was used as a way up to the Acropolis.Part of the
For personal use only. License: CC-BY-NC-ND.

steps of the Mediaeval stairway is still preserved.83

81Above, p. 135.
82P. Kavvadias, 'A06vIw>v?, 1897,pp. 27-28.
<To7roypaquKa 'Ecp'ApX
83
Ibid., pp. 28-32. 0. Broneer,"A MycenaeanFountainon the AthenianAcropolis,"Hesperia8, 1939,
pp.320-322, 326.
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

INDEXES
 INDEX I: GENERAL INDEX
ACADEMY 55. See also Neoplatonic School, Platonic Arethas (Bishop of Caesarea)76
Academy Argos 28, 49-50, 52222
Achaea 28, 11 Aristeides,Aelius 76
Achaea, Provinceof 316,16 Asklepieion44, 92
Achilles 51, 52225 Asclepigeneia70-71
© American School of Classical Studies at Athens

Acropolis20, 64, 76, 77, 125, 141 Athena 51, 52225;statue of, re-used 90
Adrianople,Battle of 50, 51 Athena, Lindian 76-77
For personal use only. License: CC-BY-NC-ND.

Aegina 21 Athena Parthenos 58, 77143, 88

Agathias 85 Athena Promachos20, 64, 76-77
Aglaurion 141 Athenaeus 10, 11
Agoranomeion71 Athanasius I (bishop) 69
Agrippa, monumentof 134. See also Odeion of Agrippa Athenodoros(pirate) 5
Alaric, invasionsof 14, 18, 25, 26, 27, 49-56, 6552,78, Atrium 113-114
79, 83 Attalos II, base of monumentdedicatedto 134
Alexander (Logothete) 82 Augustus 316
Alexandria 23, 53, 55, 74, 87192 Aurelian 6
Amasis 77 Avar(s) 93; buckles,so-called 118-119
Ampelius 20, 21, 35-36
Amphion 9 BAPTISTERY
88-90
Amphitrite (statue) 77 Barbarus,M. Ceionius Civicus 130
Ampullae, Christian 91 Basil 1 123
Amyklaion (Sparta) 21 Basil the Great 20
Anastasius I 83175, 93 BALIAEYEATTAAOI 132
Anastasius II 123 Basilica(s):near Lykabettos69; near Olympieion73; on
Anatolius (bishop) 69 Ilissos 2369,73; plan of, publishedby Lenoir 73;
Anatolius (Prefect of Illyricum) 16, 25 quarriedafter Slavic destruction72105;Roman Ba-
Anconia Fabia Paulina 19 silica (Northeast Stoa) 4, 15, 79; 7th-century,
Anemurium 638 under Fetiye Djami 71, 73; under Churchof the
Anonymousof the AmbrosianLibrary 71, 125, 136- Soteira Lykodemou73; outside Athens 11331
137, 139 Baths 30-33, 49, 68, 90, 107-108, 110; in Athens:on
Anta capital (in Palace of the Giants) 105 Areopagusslope 31-32, 10819,on Broad Street
Anthemius (priest of Athena) 20 (unexcavated)67, East 30, 31, in House C 87, 110,
Antioch 20, 23, 111, 114 in Palace of the Giants 30, 107-108, Southwest
Antiochus I 75132 32-33, West 30-31, south of Proclus'house 44; on
Antiochus (Proconsulof Achaea) 51, 55-56 outskirtsof Athens 49
Apamea 37 Belt buckles ("Avar")118-119
Apollo (statue in Forum of Constantine)76 Beroea 3, 59
Apollo Parnopius (statue) 76 Beule Gate 118
Apollo Patroos (statue) 60 Boule 12
Apollonia 115 Bouleuterion4, 53, 129
Apronianus 64, 77 Broad Street Building 73, 117, 119
Aqueducts:Antonine 320, 14, 22, 29, 103, 108; Mill 28, Bronze disk 37
29, 5925, 61, 65, 78, 80-81, 97 Bronze foundries80
Aquincum 115 Brunelleschi,statue found beneath house of the 88
Arcadius2691,54, 5914,61, 6551,66, 70 Building behind Stoa of Attalos 66
Arch of Constantine(Rome) 100 Bulgars 1172
Archeleos 19, 50, 51 "BurntColumn"75
Areopagus, Council of the 12, 17 Byzantium28
 146 INDEX I: GENERAL INDEX

CALENUS1 Diadochika38, 86-87

Carnuntum 115 Diadoch(s) 18, 43, 57, 85
Carthage 10618 Diocletian 16, 111
Cavades85 Diogeneion 61, 63
Cave of Apollo V7roMaKpais141 Diogenes (Count) 79
Cave of Pan, Acropolis 141 Diogenes (philosopher)85
Celsus 18 Dionysius (Bishop of Corinth) 19
Cemeteryon the Areopagus 118 Dipylon 3, 4, 15, 23, 52, 94
Cervonius20, 21, 22, 23, 24 Djemila 90214
Chalkis 21 Domninus 57
Chione (lampmaker)6975 Dougga, Tunisia 36
© American School of Classical Studies at Athens

Chosroes85, 86 Dromos. See Streets

For personal use only. License: CC-BY-NC-ND.

Christianity 18-20, 24, 68-74, 88-89, 91, 92 Dura-Europos 111, 116

Churches (Athens):Asomata 138, 139-140; Hypapanti Dux Ripae 116
8, 126, 138-139; Megali Panagia 138; Panagia
Krystalliotissa138, 140; Panagia Pyrgiotissa7, EAST BUILDING320, 136
126, 133, 138; Panagia Sarres 138, 140; St. Deme- Edessa (Osrhoene)70
trios Katiphori 126, 138, 13861,140; St. Dionysius Edict of Galerius 1824
the Areopagite74, 118, 12225;St. John Mangouti Egypt:Syringesin 16; influenceof, in privatehouse 37
124; St. Philip, traditionof early church 73-74; St. "EgyptianKings"75
Thomas 66; in Agoranomeion71; Tower of the Eirene (Empress) 124
Winds as church71 Elephants, Battle of the 75132
Cicero 55, 65-66 Elephants,statues of 75
Clematius (bishop) 69 Eleusinion 5, 7, 8, 128-130
Cleodamus 1067 Eleusis 3, 2155, 52223, 118
Cologne 114 Ephebes 12, 17
Colotes 65 Epicureans62
Constans I 16 Epicurus45, 65-66
Constans II 73, 93233,117, 118, 119, 120,123 Epidauros,basilica 11331
Constantine(the Great) 9, 14, 16, 38, 50, 75, 76, 77, Epirus 28
1187 Erechtheion92, 117
ConstantineII 28 Eulamius 85
Constantinople16, 20, 38, 589, 74, 111, 114 Eumenes 134
ConstantiusII 16, 22, 23, 28, 36, 5914, 74 Eunapius 1715,18, 24, 45
Corinth 2, 28, 16, 22, 38, 52222, 59, 93, 1172, 120 Euodius (consul) 19
Couchaud74 Ezekiel 19
Croesus 210
Cult images 44, 58, 69, 70, 123 DUKESOFATHENS:Acciaiuoli 6
FLORENTINE
Cybele 43, 44, 45 Florian 6
Cybele and Attis, cult of 19 FortunaAnnonaria37
Cynics 61-62 Fountain house 40
Cyprus 114-115 Fourth Crusade76
Cyrene 114 Franks 6
Cyriac(us) of Ancona 952, 952, 125
Cyrrhestes.See Tower of the Winds GALERIUS11539
Cyrus 210 Gallenius 315, 4, 6, 10-11
Cyzicus 28, 11 Gamzigrad,Yugoslavia 11539
Gardencourts 36, 97
DAMASCIUS
57, 64, 85, 87192 "Gardenof Epicurus"65-66
Decius 19 Gates, city: Athena Archegetis 15, Piraeus 4, Sacred 15;
Delos 5 Post-Herulian Wall: 8, Asomatata Skalia 139,
Delphi 6870 Christ Church (Middle) 7, 8, 14,15, 128, 129, 139,
Demeter and Hera, cults of, at Lerna 50 Hypapanti (South) 14, 128, 139, Krystalliotissa8,
Dexippus 2, 12, 167, 17 139,140, Megali Panagia 140; Panagia Pyrgiotissa
 INDEX I: GENERAL INDEX 147

(North) 7, 8, 15, 133, 139, Panagia Sarres 140, St. Isidoreof Gaza 57, 85
Demetrios Katiphori (East) 4, 7, 8, 9, 140; other Isthmus (of Corinth), Isthmia 1, 52222,93
gates 140-141. See also Dipylon
Gaza 87192 JEROME, ST. 20, 77
George (Bishop of Alexandria) 20 Jewish communities59
Gortyn 114 John Chrysostom,St. 20
Gordian III 115 Julian (Emperor) 16, 18, 19, 20, 21, 23, 24, 46, 50, 74,
Goths 17, 315,11, 51 75
Gratian 54234, 70 Julian (of Cappadocia,sophist) 17, 18, 45, 46
Great Drain 7, 29, 30, 34, 83, 84, 103, 108 Justin I 85
Gregoryof Nazianus 20 Justin II 83, 90218, 93
© American School of Classical Studies at Athens

Gymnasia 68; of Diogenes 138; of the Giants. See Palace Justinian 6, 8, 29, 51, 58, 77, 82, 83, 85, 86, 87192,92,
For personal use only. License: CC-BY-NC-ND.

of the Giants;of Ptolemy 136, 138; at Taormina 62 139, 140; closes schoolsof philosophy 13, 57, 84,
86, 90-92
HADRIAN 1,3, 13, 65
Hecate, priestessof, on Aegina 19 SYMEON125
KAVASILAS,
Hecate, shrine of 8, 19 Kenchreai93
Hegias 57 Kerameikos3, 2691
Hephaisteion. See Temple of Hephaistos Kleoboulos(tyrantof Lindos) 77
Heraclius 93233, 117, 1187, 120 Klepsydra29, 128
Herakles 46. See also SculpturefromAreopagusHouses Kollytos21, 35-36
Herculius 20, 63-66, 67, 77, 79 KolonosAgoraios 13, 48
Hermeias 85 Kosmon 118
Hermes 46. See also Sculpturefrom AreopagusHouses Kostobokoi118
Hermogenes, Flavius 22
Herodes Atticus 13, 111, 130; father of 45183 LAKE MEOTIS 28
Heruli: invasion 1-3; destruction3-5 Lampmakers12, 26, 27, 37, 68-69
Hesychius (Proconsulof Achaea) 78 Lapethos,Cyprus 9
Hierius 64 Late Roman Wall (Long) 28, 55, 81, 97
Hilarius 18, 53 Latrine(s):at cornerof Broad Street 34, 67, 119; in
Himerius 18, 20, 22, 23, 24, 26, 27, 28, 35, 55 House A 34, 39; in House C 34; in House west of
Honorius (consul) 19 Palace of the Giants 34, 83; in Palace of the Giants
Honorius (Emperor)2795,61, 69 34, 104, 109; in private houses 34127; near south-
Horologium of Cyrrhestes.See Tower of the Winds west corner of Agora 33, 33124;near Tower of the
House(s): Areopagusgroup (Houses A, B, and C) 38- Winds 33, 49
42, 44-48, 80, 87-90, 123; of Epicurus 65-66; of Laundry 119
Proclus 42-44, 87, 91; other 34-37, 83-84; typol- Law Court 320, 80
ogy of 34-35, 37 Law courts,classical 135, 136
Huns 78145 Lechaion,St. Leonidas 11331
Hypatia 53 Lemnos 28
Leo I 70, 79, 81165
IAMBLICHUS(II) 18, 51 Leo III 123
Iconoclasticcontroversy123 Leonides (martyr) 19
Ilissos, basilica on the 2369, 73 Leonticus,Cn. Claudius 9, 10
Illyricum 21 Leontius (sophist) 6449
Illyrius, Claudius (Leonticus) 9 Lerna 50, 51
Impost blocks 27 Libanius 17, 51, 111
Incompletion(of Palace of the Giants) 108-109 Libraryof Hadrian 3, 5, 6, 15, 49, 63, 64, 66, 79, 102,
Inscriptions:in Argos 49-51; building 9-11, 61-63, 113, 125,136, 137, 139, tetraconch44179, 72
140; of foundersof churches124; in Megara 49, 79; Libraryof Pantainos4, 7, 49, 6662,67, 130
in Theater of Dionysos 25; honoringDexippus 129; Libya Pentapolis 115
of Stoics, Cynics, and (?)Epicureans61-62 Lioness Lintel 130
Isis, sanctuaryof 44 London (Walbrook),Mithraeum 88
 148 INDEX I: GENERAL INDEX

Longinus 1067, 17 Olympiodorus86

Lyceum 55 Olympos,Attica 121
Origen 19
28, 21
MACEDONIA Osteothekesoutheastof Acropolis94
Majorian 70 Ostia 36, 37, 10618
Malalas 84, 86
Maltese cross 71-72 PALACEOFTHEGIANTS30, 65, 75131, 78, 79, 83, 84179,
Mamertinus 21, 23 91, 92, 95-116, 121
Marathon, Battle of (paintingsof) 55 Palladium 47194
Marcian 81164 Palladius 2043
Marinus 43, 57, 70, 71 Palladius (Duke of Osrhoene)70
© American School of Classical Studies at Athens

Market of Caesar and Augustus. See Roman Market PanathenaicFestival (and Procession),Panathenaia12,
For personal use only. License: CC-BY-NC-ND.

Martinus Crusius 125 15, 17,20,23,24,26,63,64, 111

Maurice 93233, 94235 PanathenaicWay 5, 7,13, 15,19, 26-29, 48, 60, 61, 80,
Maximian, coin of 640 96, 97, 111, 118, 123, 128
Megara 49, 79 Panathenius 10, 11, 62
Metroon 4, 25, 49, 53, 58-59, 61, 121-122, 129, 131 PanhellenicUnion 6
Metropolitanof Athens 136 Panhellenion6
Milan 111, 114; S. Lorenzo 73111 Pantheon6, 137
Mills 81-82. See also Aqueducts Parthenon23, 24, 58, 6663,70, 78148, 92, 113
Minucianus, M. lunius 9-10 Paul (Apostle) 59
Mithraeum, Mithraism 37, 88 Peisistratos 52225
Modestus (bishop) 69 Peloponnesus,Peloponnesians1, 21, 52
Mosaic 32-33, 35, 36, 38, 39, 44, 45, 46, 59, 66, 6769, Pergamon18
68, 73, 89, 98, 106, 121, 135-136 Persia 85-86
Mother of the Gods 44 Persian sack 5, 12
Musonius 19, 55, 56240 Phaidros20, 25
Mycenaean spring 141 Pheidias 20, 589, 76, 77
Philip the Arab 9
NARCISSUS (Bishop of Athens;martyr) 18-19 Philip, St. 73-74
Narthex 114 Philippicus 123
Nea Anchialos, Basilica A 11331 Philtatius 66
Nea Paphos, Cyprus 114 Phocas 93233, 120
Neoplatonic School (Academy)18, 38, 43, 44, 57, 63, Phosphorius49, 50, 51
84,85,86,87, 111 Piazza Armerina 115-116
Nicagoras (priest of Eleusis) 9-10, 16 Piglet, grave of 44
Nicagoras (sophist) 9-10, 167 Piraeus 52223
Nicagoras the Younger ( archon) 12 Plato 57
Nicci foundry (Rome) 80 PlatonicAcademy38, 55, 57, 86
Nicomedia 23 Plotina 65
Nike (acroterionfrom Stoa of Zeus) 54 Plutarch (biographer)18
Niketas Choniates 76 Plutarch (Neoplatonist) 18, 43, 44, 57, 63, 64, 70-71
Nikopolis, Basilica A 11331 Pnyx 1,4, 11,24
Northeast Stoa. See Roman Basilica Pompeii, Baths 107
Notre Dame 88 Pompeion318,23, 24, 26-28
Nymphaea 40, 41, 68, 87, 89 Pontus 28
Porphyry 1067, 17
OBELISK 23, 74-75
OFTHEODOSIUS Post-Herulian Wall 5-11, 125-141 and passim
Odeion of Agrippa 3, 4, 65, 75131,96, 97, 99, 131, 134 Praetextatus,Vettius Agorius 19, 22
Odeion of Herodes Atticus 3 Priscianus85
Odeion of Perikles 139 Priscus 18, 53
Olive presses 121-122 Probus 6, 10, 11
Olympia: 3, 5, 93: Bouleuterion 538; Leonidaion 538; Proclus:12, 18, 43, 44, 46, 57, 68, 70-71, 72; House of
Pelopion 538; Statue of Zeus 76; Temple of Zeus 5 42-44, 87, 91
 INDEX I: GENERAL INDEX 149

Proclus Patricius 74, 74122 Skyllis and Dipoinos 77

Procopius78, 82 Skyros28
Prohaeresius166, 17, 18, 20, 45184 Slavic attacks:14, 72105, 73, 92, 93-94, 117, 118, 119,
Proterius 53 123, 124; coin hoardsrelating to 93
Prytany decrees 12 Snails 48, 90
Ptolemaeion62 Sophocles43
Publius (Bishop of Athens;martyr) 18-19 SoutheastBuilding. See Stoa, Southeast
Spalato 111
QUADRATUS (Bishop of Athens; martyr) 18, 19 Sparta 2, 28, 21, 59
Square Building 54, 59, 66, 78, 97, 109
RAVENNA 110, 114 Stable 103
© American School of Classical Studies at Athens

Re-opening of the temples 24 Statues northeastof Metroon 26, 60

For personal use only. License: CC-BY-NC-ND.

Revetment31, 33, 43, 45, 67, 89, 98, 99, 106, 107, 108, Stauracius124
119 Stilicho 52
Rizokastro 125, 128, 138 Stoa(s): Late Roman 56241; Middle 4, 13, 79, 99, 134,
Roads. See Streets 136; Northeast. See Roman Basilica;on Panathe-
Roman Basilica. See Basilicas naic Way 26, 28, 48, 56241,84; Stoa of Attalos 4, 7,
Roman Market 3, 8, 15, 49 8, 49, 126, 129, 131-135; Stoa Basileios 4, 5, 49,
Rome: 92227,100, 111, 114; Palatine, Palace of Domi- 54-55; Stoa Poikile 5, 49, 55, 56, 61, 78, 81; Stoa of
tian 89, 11126 Romaios 138; South Stoa II 32?,14, 79, 103, 108;
Round Building 60, 61, 81, 109 Southeast(Building) 5, 7, 130; Stoa of Zeus Eleu-
therios 4, 49, 53, 54, 59, 109; on Streetof the Li-
114
SALONICA braryof Hadrian 78, 79
Sant'ApollinareNuovo 110 Stobi 44179
Sarcophagi,cessationof export of 12, 37 Stoics 61-62
Sardis, Gymnasium, synagoguebuilt in 59 Streets:Broad Street ( = East ColonnadedStreet) 15,
Sarmatians50 34, 67, 117, 119, 12122;Dromos 14, 23; Peripatos
Schoolsof philosophy 13, 17-18, 20, 45-47, 57-58, 82, 128; Streetof the Libraryof Hadrian 15, 79; Street
84-87, 90-92. See also Diadochika,Diadochs, Neo- of the Marble Workers31; South Road 8, 13, 14,
platonic School 15; Street of the Tripods 139, 140; Upper South
Sculpture:caches of 41, 54, 88, 90; from Areopagus Road 15, 38, 39, 40; West Street 15; roads in the
Houses: Antoninus Pius 41, Asklepios 36, 41, Agora 15. See also PanathenaicWay
Athena 90, beardedman 41, female bust 41, Helios Stucco,painted 35
37, 41, Herakles 41, 46, Hermes 41, 46, Hermes Sulla 1, 12, 30116, 35, 74
with Dionysos and Nymphs (relief) 90, Isis 44, Symmachus,Aurelius Valerius Tullianus 50
Nemesis 41, Nike 41, older woman 41, philoso- Symmachus,L. Aurelius Avianius 50
pher, head of 44, philosopher,seated 41, Sarapis Synagogue25, 59. See also Metroon
36, Selene 37, Silenus 41, Tyche 36, young woman Synesius of Cyrene 53, 55, 56
41; from northwestof Dipylon 65-66; head of herm Syrianus43, 57
from pit 54; Palace of the Giants, re-used in 65, Syringes (Egypt) 16
95-116 passim, portraitof official found near 113.
See also Amphitrite,Athena, Elephants, Olympia, TABARKA10618
Statues, Varvakeion Taormina, Gymnasium62
Scylacius22 Taurobolium 19, 50
Scythians 17, 11 Tebens. See Teres
Sesostris ("Tyrantof Egypt")76 Temples: Apollo Patroos4, 49, 53, 54 ; Ares 3, 75, 97,
Severina6 109, 131, 134; Asklepios70- 71, 72106;Dionysos
SeverusAetius 61 43, 71; Hephaistos 48, 72101,92, 111, 117; Ilissos
Shops 84 92; Southeast3, 81, 131; Southwest3, 131; Zeus
Sigma tables 89, 91 and Athena (Temple II) 136
Simplicius 85, 86 Teres (Claudius) 9-10
Sirmium 111, 114 Tetraconch.See Libraryof Hadrian
Skripou (Orchomenos),Church of the Dormition of the Thasos 81
Virgin 124 Theagenes 68
 150 INDEX I: GENERAL INDEX

Theater of Dionysos 19, 24-25, 43, 49, 71, 139 VALENS 54234, 5914
Thebes, walls 9 Valentinian II 2795, 54234, 5914, 70
Thebes, Church of St. GregoryTheologus 124 Valerian (Emperor) 13, 6, 9
ThemistocleanWall 1, 11, 14-15, 83. See also "Valerian "ValerianWall" 1, 11, 15, 51, 58, 83, 126. See also The-
Wall" mistocleanWall
Theodore of Tarsus 33120, 120 Vandals, attackon Athens 56, 60, 78-79, 91, on Ionian
Theodoric91, 110 islands 78145
Theodosian Code 111, 114 Varvakeioncopy of Athena Parthenos88
Theodosius I 1824,23, 59, 6551, 67, 70, 74 Verres (governor)55
Theodosius II 6551,69, 72, 74, 75, 77 Via Egnatia 16
Theophylaktos(Bishop of Athens) 117 Victor, M. Ulpius 115
© American School of Classical Studies at Athens

Thermopylae 78145 Vienna Anonymous73-74

For personal use only. License: CC-BY-NC-ND.

Thessalonica 1, 16, 59 Villas 111; on the outskirtsof Athens 46, 49, 68; towers
Thessaly 28 in 106
Tholos 4, 25, 49, 53, 59, 61, 83, 129 Visigoths 51, 109
Thrace 28 Vitruvius 81, 103-104, 107
Tiberius 134 Volubilis 115
Tiberius II 83, 93
Tile making 120-121 29, 61, 80-81
WATERMILLS
Timgad 10921 Water supply 22, 29, of Areopagushouses 42
Tivoli, Villa of Hadrian 89 Wolfensberger,J. J. (artist) 952
Tower of the Winds 3, 49, 71, 106, 113
Triarius 5 XENOPHON
103-104
Trier 111, 114
Triumphal arches 110 ZENO(Emperor)79
 INDEX II: ANCIENT AUTHORS
AGATHIAS, Historiae CICERO, Definibus bonorumet malorum[defin.]
R. Keydell, Berlin 1967 J. Martha, Paris 1955
II.30-31 85 5.1-3 65-66
AMMIANUSMARCELLINUS[Amm. Marc.] CLAUDIAN, In Rufinum
C. U. Clark, L. Traube, and W. Heraeus, Berlin M. Platnauer, 1922 (Loeb)
© American School of Classical Studies at Athens

1910-1915 11.186-191 52
23
For personal use only. License: CC-BY-NC-ND.

xvII.7.2-8
xxi.8.i 21 CLAUDIUS MAMERTINUS, Gratiarum actio de consulatu
XXI. 2.25 21 Iuliano [Grat.act. Iul.]
23 E. Galletier, Panegyriqueslatins III, Paris 1955
xxII.9.3-5
21 9 20-21
xxvnII.4.3
XXXI.5.15-I7 28 CODEXJUSTINIANUS [Cod. Just.]

[Anth. Pal.]
PALATINA
ANTHOLOGIA P. Kriiger, Corpus Iuris Civilis, II, Codex Iustinia-
E. Cougny, Paris 1890 nus, Berlin 1954
No. 279 50 1.5. 8.4 84, 86
I.I I.I.2 90
ARETHAS, Commentaryon Aelius Aristeides I.II.I 0.2 84, 86
S. Kougeas, Aaoypaflia 4, 1913, p. 240
ad Orat. 50 76 CODEXTHEODOSIANUS
[CTh]
T. Mommsen, Theodosianilibri XVI cum constitu-
Orationes [Orat.]
AELIUS ARISTEIDES,
tionibussirmondianis,2nd ed., Berlin 1954
W. Dindorf, Leipzig 1829, repr. 1964
C. Pharr, trans., The Theodosian Code and Novels,
50
and the SirmondianConstitutions,Princeton 1952
AULUSGELLIUS,Noctes atticae VII.IO.I-2 112
R. Marache, Paris 1967-1978 VII.I I.I 112
1.2.2 11127 X.2.I-2 114
XIII.3.5 46
AURELIUS VICTOR
xv.I.35 112
P. Dufraigne, Paris 1975
28 XVI.5.4 6870
33
XVI.IO.2-25 1824, 24, 69-70, 72
BEDA VENERABILIS,Historia ecclesiastica gentis Novels of Majorian
Anglorum IV.I 70
J. E. King, 1930 (Loeb) Novels of Theodosius
IV.I 12019 mI.8 6982
CASSIUS Dio Continuatorof Cassius Dio
U. P. Boissevain, Berlin 1895-1901 See Petrus Patricius
XLII. 14.I 1
DAMASCIUS, Vita Isidori
CEDRENUS, GEORGIUS, Historiarum compendium C. Zintzen, Hildesheim 1967
I. Bekker, CSHB, I, Bonn 1838 R. Asmus, Leipzig 1911
259 A 28, 10
266 (Epitoma Photiana) 1819
295 D-296 A 76
77 Frag. 4, 258 6874
322 B
322 D-323 A 77 P. HERENNIUSDEXIPPUS
323 D 75 Jacoby, FGrH 100, 28 (= Muller, FHG III,
351 C-D 77 p. 680, frag. 21) 2
 152 INDEX II: ANCIENT AUTHORS

EUNAPIUS, Vitaesophistarum[ VS] HISTORIAEAUGUSTAESCRIPTORES[Historia Augusta]

J. F. Boissonade,Paris 1849 H. Peter, Leipzig 1884
W. C. Wright, 1961 (Loeb) Vita Gallieni, 13 28, 10-11
476 51216
482 1820, 53 JEROME
483 45 Epistulae [Epist.]
45 F. A. Wright, 1933 (Loeb)
485-486
46 LX. I6 52
487
488-489 46193
Commentariorumin ZachariamProphetamad exsu-
492 166
perium tolosanum episcopum Corpus christiano-
© American School of Classical Studies at Athens

EUSEBIUS rum:series latina LXXVI A, Turnholt 1970

Historia ecclesiastica[HE] III.I2.3 77140
For personal use only. License: CC-BY-NC-ND.

G. Bardy, Paris 1952

1927 JULIAN
IV.23.3
VI.32 1928 J. Bidez et al., Paris 1932-1964
Vita Constantini[v. Const.] Epistulae [Ep.]
PG XX, ii, Paris 1857 59 (58) 2366, ,74-75
I. A. Heikel, Leipzig 1902 Letter to the Athenians 1712
II.48-60 17 Orationes[Or.]
1.6 16
EXPOSITIOTOTIUSMUNDIET GENTIUM
VIII (V).I59 B 25
J. Rouge, Paris 1966
52 22, 38 JUVENAL

GEORGE, BISHOP OF ALEXANDRIA

P. de Labriolleand F. Villeneuve, Paris 1957
See Photius, Bibl., Cod. 96 VIII.I 27 11229

HELIODORUS,Aethiopica LIBANIUS, Epistulae [Ep.]

R. M. Rattenbury, T. W. Lumb, and J. Maillon, R. Foerster, Leipzig 1921-1922, repr. Hildesheim
Paris 1960 1963
1.16.5 6660 627 17
HERODOTUS LUCIAN, Zeuxis
C. Hude, Oxford 1927 K. Kilburn, 1959, repr. 1968 (Loeb)
i.6o 52225 8-12 75132
1.I55 210
Chronographia
MALALAS, JOHANNES,
HESYCHIUS L. Dindorf, CSHB, Bonn 1831
See Patria I
XVIII.I87 84
HIMERIUS
MARCELLINUS,Chronicon
Eclogae [Ecl.]
PL LI, Paris 1861
F. Dubner, Paris 1849
Col. 919 74
xxxi 20-21, 35
Orationes[Or.] MARINUS, Vita Procli
F. Dubner, Paris 1849 J. F. Boissonade, Leipzig 1814, repr. Amsterdam
Numbers in parenthesesrefer to the edition of A. 1966
Colonna, Rome 1951 L. J. Rosan, The Philosophyof Proclus, New York
II (VI).20-21 55
1949
III (XLVII).I 2 23-24
5 57
IV (XXXVIII) 20-21
19 44
vnII(VII) 126
55
23 1710
X (LIX)
XIV (XLVIII) 2258 29 43,70-71
xxvII (xxv) 22 3? 58, 71
LI (Colonna) 2259 33 44
 INDEX II: ANCIENT AUTHORS 153

MICHAEL CHONIATES II.I (562) 11127

S. Lambros,Athens 1879, repr. 1968 11.21 (604) 46
Vol. I, pp. 150-156 1925
PHLEGONOF TRALLES
Breviarius
NICEPHORUS, See Photius, Bibl., Cod. 97
I. Bekker, CSHB, Bonn 1837
PHOTIUS,Bibliotheca [Bibl.]
37 C-D 1236 R. Henry, Paris 1959-1977
Historia
NIKETASCHONIATES, Cod. 96 (George, Bishop of Alexandria) 20
I. Bekker, CSHB, Bonn 1835 Cod. 97 (Phlegon of Tralles) 537

Pp. 738-740 76 Cod. 243 (Himerius) 35135

© American School of Classical Studies at Athens

NOVELS OF MAJORIAN PLINY, Historia naturalis[NH]

J. Beaujeuet al., Paris 1950-1972
For personal use only. License: CC-BY-NC-ND.

See Codex Theodosianus

xIx.50-51 65-66
NOVELS OF THEODOSIUS
PLUTARCH
See Codex Theodosianus
Moralia
NOVUMTESTAMENTUM P. H. De Lacy and B. Einarson, 1959 (Loeb)
Acts 8, 26-40 73-74 De exilio, I14 47194
Vitae
OLYMPIODORUS
C. Lindskog and K. Ziegler, Leipzig 1914-1939,
L. Dindorf, Historicigraeci minoresI, Leipzig 1870
1957-1964
Frag. 32, p. 463 66 12
Sulla, 14
ORIGEN,ContraCelsum
PORPHYRY, Vita Plotini [Plot.]
M. Borret, Paris 1967-1969 P. Henry and H.-R. Schwyzer, Leiden 1951
III.30 19 1067
xvII.83
KnNETANTINOYnOAEnE
IlATPIA [Patria] PRISCUS
T. Preger, Scriptoresoriginum Constantinopolitana- L. Dindorf, Historicigraeci minoresI, Leipzig 1870
rum, Leipzig 1901 Frag. 3 78145
I, p. 17 (Hesychius) 76
PROCOPIUS
II 744121
De aedificiis[de aed.]
II, p. 174 (Pseudo-Kodinos) 76
G. Downey, 1961 (Loeb)
II, p. 182 (Pseudo-Kodinos) 75
IV.2.23-24 5811,82
II, p. 183 (Pseudo-Kodinos) 74 29
LV.2.24
de signis, p. 27 75 De bellis [bell.] (De bello Vandalico[Vand.])
PAUSANIAS H. B. Dewing, 1953 (Loeb)
F. Spiro, Leipzig 1903, repr. 1967 III.I .17 79152
1.8.6 75 III.5.23 78
76 III.22.I6 78
1.24.8
Historia arcana [Anecdota]
1.25.2 2!261
H. B. Dewing, 1969 (Loeb)
1.42.3 167
2167
XXVI.I 86187
PETRUS PATRICIUS,Historia XXVI.33 82
I. Bekkerand B. G. Niebuhr, CSHB, Bonn 1829
PSEUDO-KODINOS
Muiiller,FHG IV, p. 196, 9, frag. (1) 210 See Patria II
Historia ecclesiastica
PHILOSTORGIUS, STRABO
F. Winckelmann,Berlin 1972 A. Meineke, Leipzig 1877
I2.2 52 1.4.7 35136
PHILOSTRATUS,Vitaesophistarum[ VS] SUDA
W. C. Wright, 1961 (Loeb) A. Adler, Leipzig 1928-1938, repr. 1967-1971
1.21 (521) 45183 M 1087 9,10
1.23 (526) 164 N 373 9
 154 INDEX II: ANCIENT AUTHORS

SYMMACHUS,Relationes VITRUVIUS,De architectura

R. Klein, Darmstadt 1972 F. Granger, 1931-1934 (Loeb)
5.I 1818 V.IO.I 107
vi.6.4 103-104
GEORGIUS,
SYNCELLUS, Chronographia x.4.I 81
W. Dindorf, CSHB, Bonn 1829 X.5.I-2 81
38i 11,28
De re equestri
XENOPHON,
Epistulae [Epist.]
SYNESIUS, E. C. Marchant,Oxford 1900, repr. 196i8
R. Herscher, Paris 1873, repr. Amsterdam1965 IV. I 103-104
54 53,55
ZACHARIAS,
POPE,Epistola
© American School of Classical Studies at Athens

'35 53, 55
PL LXXXIX, Paris 1850
For personal use only. License: CC-BY-NC-ND.

TACITUS,Annales Col. 943C 12019

C. Fisher, Oxford 1928
11221
ZONARAS, JOHANNES, Annales
III.33-34 M. Pinder, CSHB, Bonn 1841
11
THEOPHANES,
Chronographia XII.23
C. de Boor, Leipzig 1883-1885 XII.26 28 10, 10

P. 405 1236 Historia nova

ZOSIMUS,
P. 444 1247 L. Mendelssohn,Leipzig 1887, repr. 1963
P. 473 1248 F. Paschoud,Paris 1971-
THUCYDIDES 1.29.2-3 1, 82172
1.39 28
H. StuartJones, rev. J. E. Powell, Oxford 1942
1.42-43 28
1.90.3 5
1.7I.2 640
1.93.I-2 5
III.I I.3 23
TZETZES, JOANNES, Historiarum variarum chiliades Iv.I8.2-4 52219
[Chil.] V.5.2-5 56240
T. Kiessling, Leipzig 1826 v.5.5-6 51216
VIII, lines 333, 338-339 76 v.6.i-3 2(691, 51-52
 INDEX III: INSCRIPTIONS
Inscriptionesgraecae [IG]

IG II2, 958 138 5021 2578

3193 1064, 118 5184 61-63
3198 129 5199 952,56, 140
© American School of Classical Studies at Athens

3504 ( = Agora I 3299) 129 5200 9, 956, 133, 140

For personal use only. License: CC-BY-NC-ND.

3669 128 5201 10-11, 62

3670 129 5205 6127
3689 954 5208 2580
3690 954 IG IV, 209 2258
3692 6343 666 50211
3818 6343 1608 49-50
4209 134 IGV1, 455 2151
4222 128, 1712 IG VII, 26 (= SEG XIV, 379') 79150
4224 6341 93 6552,79153
4225 6449 96 49203

4841 1932 1855 (= SEG XV, 322) 2260

4842 1933 IG XII 9, 907 21 3

Supplementumepigraphicumgraecum [SEG]

SEG III, 834 ( = ILS 7784) 6558 SEG XV, 322 (= IG VII, 1855) 2260
SEG XIV, 379 (= IG VII, 26) 79150 323 2149

Corpusinscriptionumlatinarum[CIL]

CIL VI, 1779 1931, 35

ATHENS,AGORAMUSEUM

I 754 131 I 5250 130

I 928 131 I 5850 131
I 3299 ( = IG II2, 3504) 129 I 6120 a, b 134
I 3542 51214 I 6477 84177
I 3948 129 I 6737 131
I 4712 130 I 6738 131
14713 129 I 7154 90215

I 4924 130

ATHENS, ASKLEPIEION

inscription: [EnIKOYPE]IfnN 61-63

 156 INDEX III: INSCRIPTIONS

ATHENS, EPIGRAPHICALMUSEUM [E.M.]

70 61-63 1861 6128

1664 + 1664a 138 4713 6448

MISCELLANEOUSPUBLICATIONS

'Apx'E E 1919, pp. 29-30 2369 L. Moretti, Iscrizioniagonistichegreche, Rome 1953,

BCH 16, 1892, pp. 102-106 2153 no. 90 2028
50, 1926, pp.1-1 4116 2260 K. Pittakis, et al., 'Ewrypaapalavc[ oroL, fasc. 1,
© American School of Classical Studies at Athens

BSA 26, 1923-25, pp. 225-236, no. 20 2151, 53 Athens 1852, pp. 3-5 12919
For personal use only. License: CC-BY-NC-ND.

Byzantion 20, 1950, pp. 136-139 954 W. Peek, Attische Versinschriften,Berlin 1980,
H. Dessau, Inscriptioneslatinae selectae[ILS], no. 38 43191
no. 7784 ( = SEG III, 834) 6558 REG 78, 1965, pp. 104-106, no. 155 51214
J. des Gagniers et al., Laodiceedu Lycos, Le Nymphee, M. J. Vermaseren, Corpus Cultus Cybelae Attidisque
Quebec/Paris 1969, pp. 339-351 2257 II, Leiden 1982, nos. 389, 390 1932
Hesperia 26, 1957, pp. 83-88 13241
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

PLATES
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

The Athenian Agora and environs. Air view, 1975

 /66nx.-26
EON Pim i :-::-::::':::::::-:
~~~'::'......'"':: \:t: : : : : :
.
API- 7R .....
IEPON .K4.TTOZ

7, , ^y'^ ^^ ).A AAIZ- T AIA,YA:...... HTT'A" \ \

%':'...

^ / 1 ^ ^ ^ "*-***. 86B.C.
Athens, to ..^ 27... .. "
/ /\ j-^ HMOMJION k. ^ ......

/
^^ 11
/ /^y ^^ J^^J^^
l''^<"..:X:.;//~/ .::::"..
© American School of Classical Studies at Athens

"::: ~ ';V....
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. . .
For personal use only. License: CC-BY-NC-ND.

:,::,,%.::.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
......

H0(A IETQOY-V
kYPPHZT0Y:..-~~~~~~~~~~~~

...'~~~~~~~~~~~~~~~~~~nF ............

* AO"O .>U.E I N- / oI...

NYMOF NAPEIOnAPTY/

[...:.:. :
,,,,-~"l'//, flNY O ' ~./.4"~lm'
'~'"1%"-~ i ......: . .
rYMNAZION BAAANE

.4..' ~
~~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
I~~~~~~~~~~~~~~~~~~~IN
E" 'NbAAANE..N

K~~~~~~~~~~~~~~~~~~~~~~~

~":':~..
: :,,? ":::::::..:,

(.. MNHM0ION
KAI QH

86BcT
Athens~ tOAD26
?::.
922~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~15
7?', ons A.D.:2676 B. t
 1

PLATE 3

A|-B' | C| D E | jG | H : ! | KI
I.I,J L M | | P Q R| | T | U V |W
-*-
X-
-** .:Y<l.
'". j4 YK
I
*s8sa
** { o\K-
11-1
.J..... /A.9
\|>s l^ i- .* 'I// . 41* /f-d-r
^../........ . *:
1, 11,
ACCA EXc^atr
; 1, I , I"

N 1,1, 1,
2 I, 1,I t 14
/ c A 2
3 -r
3
--
© American School of Classical Studies at Athens

4 .?????;ip----
--- if
cl ft 4
-
For personal use only. License: CC-BY-NC-ND.

4
5 ii -5
I -
-
r1A .- 1- - -

I
I ' / -1 o I *m> ' EW . ' S -' - A's:--
1""

6 I,,-
10-
7 WV I 7
8 :'""'i
6
I ,
-.= im
I mm
0 1
:.i::'E_.,
9
:;,,,
to m I

11
1NK_'1 -----
-

||,,._.
1-:-- !~ Il
.

,>
I
si
\ ai,'115
I-SI-l"-
4. 1.. EEl
IT
J
L
I - ,,- ?
*
Io
+,

12
l e:vat'-<t\
-X FL /1

12
13 u4/~~~~~~~~~~~~~~~A agES--^ 13
14 14
I _

15
-15
l'^^fc -1.: ' -.C Jl?-
16

.17
',,~.~~,. -16

17
16 tf& .
16
19
19
20 % 2o
21
jor < . 21
22 :22
23 :23
24 24
25 !25
-- \ _-2 I \ I \ |MW\
A\ \ .'-\ \ 11 \ I\-A \ \1-, \ . r- 7~~~~ -'V
- ,

26 26
27
27

26
29
29
o0
J.TDAVLOS
.- I If -r . I
F k{ lj )
o0 L I[M L
1 (
1,i^
,0^^ A , -II-T I -.
K Iri o P IQ R 5 T u Iv lwl X Y.
I

The Athenian Agora. Actual-state plan

 THEMISTOCLAN
WALL
LI?~~~~~~~~~~~~~~~~~~~~~..
ACARHIAN

GATE

GATE~ ~~~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~.GT

7ioJ3;i;zyLKi\:~~~b~n
© American School of Classical Studies at Athens

~ .' POL MUSESS~

OFTHE
M~~~~~~~~~~~~~~ETRp.
''
For personal use only. License: CC-BY-NC-ND.

P/ O:P ' ' ":"' : : " GARDEN

....:

MELI
~~ ~~ ~ ~ ~ ~ ACROPOIA IDE
~~t

AI .,.o ~;
:~~~~~~~~~~~~~~~~~~~~~~~~A OP
. ('~~~~PY
'PH~Lor=^Pr'os~AKLelel
HIPPA~
DIPYo poo-~

Athens, 5th century after Christ

 PLATE 5

W7

o 1.~~~~~~~~~~~~~~~~~~~~~~~~~~
1~~~~~~~~~~~~~~~~~
o ~~~METROPQD i
'16 1
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

Wo 0rA7j,

,A0~~~~~
::....
~~~~~~~~~
Holy "-mi E
Apostles /

e~~~~~

,~;.~~~~C......

Co,
stern
/ 0~~ste'~' ERECHTHEION

Beul~
Gate

NIKE 'Qkt
4~~~ ~ ~ ~~ ~ r

o50 100 20 T~~prAVLtOS J.

4-i~~~~~
ter is n I . [-18
Li Li ~~~~~Cister

isteCrten jji/
~~
I ~-q ~~ I"' 1984

The Acropolis and the Post-Herulian Wall

 © American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.
PLATE 6

The Athenian Agora, 5th century after Christ

A
AGO IA
5THCENTURYA.D.
 © American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. East face

b. Westface
PLATE 7
 © American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.
PLATE 8

a. In the Library of Pantainos

r-..

b. Column drums from the Middle Stoa

1<

-.
tt&
 © American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Interior

b. Exterior
r
7eiJ
u!>'^^

>
--^"tlt^S

;i.
'

--
u-,,;p-
PLATE 9
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

si***

a. Core of wall, south stretch b. Foundationsof Tower W 5 at south end o

The Post-Herulian Wall

 PLATE 11
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Tower W 6 before demolition, 1894 (courtesyof the German ArchaeologicalInstitute, Athens)

b. Tower W 7 at north end of Stoa of Attalos (from old photograph)

 © American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.
PLATE 12

a. From southwest

b. From northwest
? ', A
A
*A
R.

"-' ;.':
-.~e:

'.?.:.- .
.
A
Lr!fi
 PLATE 13
'
X(

'~ -A

,, ..

. ~? . , . ,, :. .. . ~ ".;-.'.-
r~~~~~~~~~~~~~~~~~ -
© American School of Classical Studies at Athens

..,,: =...;.,- ._' ..~ ...:.~-':. ,~:. :..&,.r_ ; ;,...._ : . . . . '

For personal use only. License: CC-BY-NC-ND.

~~~~~~~~~~~~~~Z...
, ,... .. ....... . . . ....~,,,, .~~~~o

:"
:' ~, ~.'..

' "-" '"~"''""+"'?"'~' ......

~=';':'-'" '"'~ "' "'~ '' ~'~ '"' '"
,, ' " " ' " 1~ ' :'*''~
a.Foundations of pie~~~~~~~-rs.o,upotn;pprsoe

p . .

a. Foundations of piers for supportingupper storey

-. _ I I T
%- V.

-'I
L]'

b. Foundationsof piers as excavated

 e __

© American School of Classical Studies at Athens

For personal use only. License: CC-BY-NC-ND.

hrist Church Gate from west

.a...from
Tower no
W

a. Tower W 2 from northeast

Y~~~~~~~~~~~
g~~~_ ...~

u wfim ll | BIB
STOA \ * nAN
f//

- ,,_

:-'" " "tl"""^:^ Y AP

-4---lo 50145

d. Pyrgiotissa Gate

The Post-Herulian Wall

 PLATE 15

a. Building inscription,IG II2, 5200

© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

b. Coins found in Post-Herulian Wall

I- ...
.- - -- -- M - --- .

. , ^. ''
''t.
.pi
' .',,,
it.'._~
..''- .
.^'-or
i~,~..~
. .T

c. Epigram honoring Panathenius, IG II2, 5201 (photographby B. Stamatopoulos)

 © American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.
PLATE 16
-
I
i
.

a. From Acropolis

b. Phaidros bema and orchestra

1*
 PLATE 17
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

,. . '- _Sr' - - '.. ....t .'

"'

a. Theater of Dionysos, Phaidros inscription,IG II2, 5021

4) 7

b. Athenian Agora, Great Drain

 .

PLATE 18
' . X
?Er,

,.. S^iBia

.1%
i. ;4
© American School of Classical Studies at Athens

Ut:e
For personal use only. License: CC-BY-NC-ND.

,
!-
-' *.
L.. . .? %.

--. .- gg.-'22L-

* -. ..?
, I -,'-
lib. 1. '.-:,-
i%bpl*

r
ik

*
B. ...-

WrKw
L*F / : A. '*- -.,

..?IL?'

?
? :*...:,
. i .- . ."
' ""'
': " . '* v , e'" __ ?

~ *:. ....:." , ,,
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.
;.^

; ^^' '' Set '%

.'r , ... . ..
'~, "' ';;'~:'i~;. '; . ,7'
'":tSl '',.. ;.
:,. ,'.' ":',e: ~

The Pompeion from the southeast. Late foundationsat lower right

 PLATE 19
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Bathon Areopagusslope b. EastBath

4NCE

c. SouthwestBath,actualstate d. Southwest LateRoman

 © American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.
PLATE 20

a. East Bath

c. Southwest Bath
b. West Bath

d. Bath in Palace of the Giants

 '

PLATE 21
i*1 : . ....
~'.
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. House C, 6th-century bath

b. Latrine in House A
b. Latrine in House A

. Latrine west of Palace of the Giants

c. Latrine near southwest corner of Agora d. Latrine west of Palace of the Giants

..
:.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.,I,j^ .: \I
e.Ltrn 1a A ... .;

e. Latrine at corner of Broad Street f. Latrine in Palace of the Giants

 PLATE 22
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Industrial section of Agora

wf',,-7,
WSL : ,- ",
,,4.

b. House on lower slope of Areopagus

 PLATE 23
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

b. Sarapis, S 885

a. Asklepios, S 875 c. Tyche, S 871

a-c. Sculpture from a house on the lower slope of the Areopagus

d. Mosaic in house in southeast corner of Agora

 PLATE 24
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Garden court on Areopagus slope

t"1 r
't .
4, tdall

* - * *.' ""*" a-

ft .4v-t * r%.%4 '

k.t
if y -' * *
* %
J7?t.%t:
* .'.-

::w
. -*

b. Garden court in Dougga, Tunisia

 PLATE 25
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

___ _____ __ ___ _-1

0 5 10
a.court
arden on Areopagus slopeI - 1 M.

a. Garden court on Areopagus slope

b. Bronze disk, B 904

c. Relief of Selene, S 857

© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND. PLATE 26

a. House A
Houses on the Areopagus

b. House B
 © American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Areopagus House C

b. "House of Proclus"
PLATE 27
 --i,

PLATE 28
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

b~~~~~~~~~~~~~~~2

* .1k ~ -

a. Fromnorth

-.7j~
- I..-

~,,.r' .i~.
~,. .,; ' ,, -. r
' -

' ^^
..'',d, 'y.-

^ ; 3 ^'L' ..",

_--*
-.5.'f

rv~_
.,

k. -, -,- -

b. Fromnortheast
 © American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Apsidalroomfromnorth

b. Generalview lookingsouthwest
-
EE-.
-w---
PLATE 29
 © American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.
PLATE 30

b. Room with hearth

a. Oven
 PLATE 31

I
, _
, I
I-

- -- - . ... - -. .-
I - I"I - . -IF, I
I

I
© American School of Classical Studies at Athens

14 1
F, -e
For personal use only. License: CC-BY-NC-ND.

Orr, J*- ..."

I
- - .
I . I
A&

I.
'-u .I
._..

4_'
.'A
.
.1
.~
i .f
f' ~:,-J
' g
"_ 4:"- 1

lb"', .J

*
. .. .
i,11 v/. ,.I,up
1X
c?'
.-
..-I x
.ee-.1-7.0
I1 *- e: .4.1.1.
1..4.
.,
ik" F, , .4'.,
.
O' . .4-,

' ,;I *
ji "-'
".L N <*
*; !

t..
^

LI'
fA ~ ' It. .
 - - ~ 'P 2n1
~~~~~~~~~~~'

.:4 .
© American School of Classical Studies at Athens

'~~~~~~~~~~~~~~V
.~ S.~B!~li~.
'~;~~~~~~-. A
For personal use only. License: CC-BY-NC-ND.

-F~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~7.' -

? ', ~ ~ ~ .~.~~~~~~~
~ ~ ~ ~
:
~,.:ij~..,
~ _.r'
k~~~~~~~
~ ~~~~~~~~ ~ ~ ~

0~
C__
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
b ~~~~~~~~~~~~~~~~~~~~~~~~~~~~"'M

vr~~~~ ,
?~" . - - '~
...W ~' '"~. .i'' _~ '~
_..i'.
w.e.-t
looking Trilinum
a~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.- b.'-'tranceto.trcinu

a. Triclinium, looking west b. Entrance to triclinium

Areopagus House C
 © American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

1*
.7
a. Nymphaeum, looking west

b. Fountain house from nymphaeum

-7
N
PLATE 33
 .

PLATE 34
.I . . ;.
.;i W% - '

A -
..-*. i
*IsA
w~~~~~~
''

. t {/ * *s t,. : ...?^ <!

%'^\
.
* .r *.'**-
e..
~>
/' ^ 2 ''~ ~'...

'S;.
© American School of Classical Studies at Athens

r:5
For personal use only. License: CC-BY-NC-ND.

l"-v^^%^~~~~~~

|l;1
f

, - ...- .- i

1~-~
X,

y o
- W~~~~~~~~. t
ON. .-

: ;'v_-* .

-%,
orA,
' ' s l
oL __ _ _
44p !..S .

I I - t.-'. ~ 1
~~~~~~~~~,
II

* tI 4a t ;.
a-- J
Ifo b- il-1.
-

I'.. Aa~e 1
il,
f .-
1% .x
- t ...;. .
1-

'

,t.

.o
?;Is

ik
K I.+,
i}S: Itx

House and arch to well

 PLATE 35

*.
__jI
I -a-
."A
., .

* At
© American School of Classical Studies at Athens

'?
For personal use only. License: CC-BY-NC-ND.

-4
-at 1?
*. -; v
*44
4" .... rtj
I *-q I I.,
* .aCe. Y

I.
'1 t '?- .S*5
r J S
A .'
1
!

:4v

I.
I . .
It !'
r .

Pt'-

;.
... 4, _
;,. I

'. .. a
-S 111S
la.@-,
A C-
.4
'P

-:
*1 14 ,
to J

t fI I

I,. .

' ' "

'; '
'" i: " "**.
II~cr,. '~"
::- 'V,'''

14 - '/.
1
' ' ' It'^l-l% ?*'
. ''.I?L~
.49....
i.
ly
a
1:
f4
4
i
^ , ..
,/ 4tr
I
P .
t e
M
"*^ -
/< , ^'/.4 * .1:
..* X
I . i;*4t\
jA.tt '*
,,

I, , . ,
.
7.
ri f:w
t .
w4.
-'

I',A
! .. 4 t :6i N 11 l.
I
....k
k' I
i. 'k ..
z ,&

Areopagus House C, southeasternquarter. Air view

 PLATE 36

'uf^^SSs
?
30,."'..'
'
VA-- I *X C. . -

5-!tki r -A
,,.L~~~~~~~~~'-.
4
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

>0 o

..

.bv

.4

a. Fountain house next to Areopagus House C

1 '4
A>
j
H- rp*% 4

b. Mosaic in "House of Proclus"

 PLATE 37
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

~ :'.~
," '.
b.',~. -'. *;*- A. -?El

a. Areopagus
a. AreopagusHouse
House D, mosaicfloor
D, mosaic floor

..m I-'N
~~~
-- i^'"~ ** SbA - w r.- !w

Fk'4''w'
 PLATE 38
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Female bust, S 2437, Well Q 21:4 in House C b. Female bust, S 2435, Well Q 21:4 i]n
House C

;I "

c. Herakles, S 2438,Well Q 21:4 in House C d. Hermes, S 1054, Well P 18:2 in House B

 PLATE 39

4?~i
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Nemesis, S 1055, Well P 18:2 in b. Silenus, S 2363, c. Philosopher, S 1053, Well P 18:2
House B House C in House B

\ vl

.4,

d. Asklepios, S 1068, House B

 PLATE 40
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Antoninus Pius, S 2436, Well Q 21:4 b. Helios, S 2356, Well P 21:2

c. Nike, S 2354, Well P 21:2 d. Bearded man, S 2355, Well P 21:2

 PLATE 41

-A
I.
.! f
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Marble bust, S 1026

b. Marble bust, S 1044

c. Lamp, L 5324 d e

d, e, g, h. Fragments of window frames

f. Lamp, L 5323 g h
 _

PLATE 42
© American School of Classical Studies at Athens

3iA-
.3867.
For personal use only. License: CC-BY-NC-ND.

a. A 3867

'"-',, .~...~.;....

b. A 3866
_

_
_l
__

c. A 3862

d. A 3856

e. A 3861

f. A 3863 g. A 3858
 PLATE 43
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

%
a, b. Perspectiverestorations

?
14. -1 ?'
AL ~
- -;, . .

? i . ,

c. Mosaic in triclinium with later insertion d. Ionic column from peristyle, A 3866
 PLATE 44
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

r r e r s " , . ,
a. "House of Proclus",general view from southeast ("Epyov 1955, p. 7, fig. 3)

b Shn inHu oPrlu ( 19 . i. ) c.

Portrait of philosopher, .... .
Acropolis 1313.
Museum
b. Shrinein "Houseof Proclus"("Epyov1955,p. 8, fig. 4) of philosopher,
c. Portrait Museum1313
Acropolis
 PLATE 45
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

b. Herm head, S 2499, in pit showing signs of burning

b. Herm head, S 2499, in pit showing signs of burning

a. Epigram honoring Iamblichus, I 3542

c. Mosaic in reconstructedMetroon
 © American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

S 2531
^lt^ b, c. Fragments of statue from Round Building
-f
i~/:l-. 1||
a. Round Building in northwest corner of Agora

e. Dedication of building by Atius, IG II2, 5205

d. Monument base on PanathenaicWay
 a. Epigraphical Museum No. 70 b. IG II2, 4224
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

KAI^A; 4:100 6

c. IG IF, 5184

,
M.-~
_YC_IIMj MMgriiiirirlliim
I

d. IG II2, 3692 e. IG II2, 3818 f. IG II2, 4225

© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.
PLATE 48

a. Building east of Stoa of Attalos

 PLATE 49

A.>
I
y
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

'

4 /.-J ,
?? /

I,

&hldi

%tf .
j*, I,

/1
/
C?
-I
I ''
.1 %. :;
/ > ./
I / /
'2/I
&f
'fi*/J
ef'
,^Z
i--
I-

;(v ,.,.P,
l ir
'~ '~llr
.
i
.
I.1

qPl-

^*-"A

: ;.
.tv '..:
. .

.
~ -1:...
I
-. t
..it ^r 7F
I '. ..?, ,. .
: . . .: '../
? X.i. -.?
. .4.
'.l~,~ :..

Colonnade of the Broad Street and the over the of Pantainos. Air view
 PLATE 50
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Colonnade of Broad Street

.:
' ~ b.Building
~
over '" ':;' of
~~~~~~~~~~~
Library ."~,,
bse Pantainos , r

b. Building over Library of Pantainos, basement room

 PLATE 51
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

-w.t1U d

- :;&. :

'. I.-

a. Tetraconch in Library of Hadrian from east

-1 III I
_-""'^^ 0 10 20I 30 40 M 1. TPAYAOE
I I iIb t1984 .

b. Library of Hadrian with tetraconch

 PLATE 52

1 .~--- , . .1 :t

tL-2~~

.^' -1l?
© American School of Classical Studies at Athens

4
I
Al.
For personal use only. License: CC-BY-NC-ND.

*
1. \F. ^ _i
tL \: . -

1
,.1
.
4- -. -v I

.,-T

-
-
-

..
,._-

.,--.
'. M
ofI ,.a..-
% L
,4 ---,-
'?I*-
1,
I. .tS
';
.t . _6 .- - .. _ - t.
I im

a .-
4.. * *.
. ,
. 4

| ' IrSp
aJ 4
4
|W _ 9~~~~~~~~~~~~~~~~~'A
I
ir ll
...i
.':,
IFb
't

wi
V' 9;~)I

, 1. i . .'

,
,*- ^- - "

s. "I

.1 . .I. -
.

.k-

N
". - -, -X, -%w L- , - -

The Palace of the Giants. Air view, May 1975

© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

,,,-.---,
... ., . ,. . . . . . ..

-JLi^^^
\\ Ikdl-^
| | |~~~~~~

(0 10 20 30 40 (\ J- l
^.r11^ r"
ll\lt111111 ~ J 1M. 1^1 | | I |

The Palace of the Giants. Actual-state plan showing earlier structures

 PLATE 54
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

* *0

1 *41 * * J 2 -----< F2 810g

04410 2 4 T0RA J V
|
---- k 49
9 39
47 |4i| I 22* 37

* ? I ^1
13 I ?f * * *m
^^^?
51 51 |2|38

52 53
e r~o --- n I 'I , - -- - -. ,I
?~~~~~~~1
* 1 "I1""
| |1-
I
' 7

O 10 20 30 40 J. TRAVLOS
Iltii i iitr I I t M. 1984

The Palace of the Giants. Restored plan

 PLATE 55
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

-_ J.T
31 .3 .7

a. Palace of the Giants. Perspectiverestoration,Panathenaic Way at bottom

b. Palace of the Giants. Tentative restorationof north facade

. f: ,
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND. PLATE 56

4..
,.
.., '~.
t~
,..

,
~~
CI
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

-:

,; , \
The
area ofth
-S;Y
Plceofte
iat.-Ve
rm
h
ot,
1969
I

S
i
.~I--
~_~~
 PLATE 58
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Tritn,f. ont vie

a. Triton, front view

.- - .,r
7

c. Giant, front view d. Giant, side view

 PLATE 59
.- -. . . .

'. 'i,'.i.. .. -

! i',,t,, "'_'

'"....
!;., ,/..^'^-' ,..... ....
.
. . . _~.
,. ..
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

i' R-
' -14 1," . .
n.. f'o" :,

a. Tile flooring in Rooms 14, 17, and 18, from northwest

b. Sherd bedding for floor in Room 41, from north

 © American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.
PLATE 60

b. From
s.outest
a. Air view, 1975

b. From southeast
I.-
._
.v.,.4.
~~~~~~~~~~~~
 © American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

b. Room 41, masonry at northeast corner

a. Room 40, vault above northeastcorner
PLATE 61
 PLATE 62
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

* V
3 A

a. Room 39 from west. Traces of column bases on threshold

:(- Sc :,.
*r' '-i

VF ' .:.6^
~ ~~~~. ~ n.'.. "'- , .,.,, *.,..I *t',3 .'.. .
*...t
-~ . ,

-.taICistern?1'.;a,.;
Ro abov .5 wetedo ot
rmsuth. Oulet or., hole at nr ed of

b. Cistern above west end of South Stoa II (Room 55) from south. Outlet hole at north edge of floor

The Palace of the Giants

 PLATE 63

I .!.!
.I
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

b. Fragments of mosaic found in Room 39

a. Fragments of marble revetmentfound in Room 39

c. Anta capital, A 2268, found in Room 39

The Palace of the Giants

 %r
I
f
!': | - "
--.'-'-:,"-
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

..%%Uakmbl - -
!.a-.,
'P" -
.-; . Ampt-
- - rfri^tc- .^ Ia
0;21f".
. A0
-%% . int-
.S~ I.g ~ . -
AL
*c I

%i
. ~~~~~~~~~~-
-NI"b .,,- -
-
'
. 7 ~.,..k',.~ ..
- zA -
.
A;l J..

, .
t.*!:. r-
%
-l ,-

... .. -I . _

0-:t
-Z^
.
T\4.
<A

o..
-:... '
.
mS ***
,; ~:.~-...-.,_,

The Palace of the Giants. The Bath soon after excavation, from the south
 © American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

b. Room 42, semicircularbasin

a. Room 43, remains of hypocaust
PLATE 65
 PLATE 66
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Bath, south side of Room 42. Bedding for marble floor and remains of marble wall revetmentincluding
door frame (upper left)

b. Marble torso of Roman Imperial official, S 657

 PLATE 67
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Dura-Europos. Palace of the Dux Ripae, before A.D. 256 (Excavationsat Dura-Europos:Ninth Season
1935-1936, New Haven 1952, fig. 7)

c. Ravenna, Sant'ApollinareNuovo. Mosaic showing

facade of palace, ca. A.D. 490 (W. F. Volbach,
FriihchristlicheKunst, Munich 1958, pl. 152)

. ,
/
,.
.-^i
0 10 20 30 40 50 METRES --

b. Piazza Armerina. Plan of villa, first half of 4th century

A. Wilson, Piazza Armerina,Austin 1983, fig. 1.
 PLATE 68

' r. .. . . .
W- ?^

t 3b ~~~~~~~~~~~~~~~~~'
~ i . ---' , ~ ~ ~ ~ ~ ~
.~~~~~~~~~~~~~~~~~~~
- *.

_g~~~~~...,;....
'j'~~~~~~~~~. ., .. ... *BP
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

1 .
. .: 4 -1

. . k

.e~~.,'*~ **Frk
*' *~~ * ''. ' t
*;^^Z,**,:.~'~~~;::~'.... M;
't*^*[*^
?a.~.. ^
..p...f..i.: n. P , , '. .
.11
' ' i- * < - *- 4-

.B
....a,. -., p.t...? P.alac .of the ....G.i.'^frr (.Mat
'* s jS *'
M).
*
tV .
a. Bronz
castingpit nea Palace f the Gants ,'JK (Matusch M
C* t" A ;

, -.0- -
!.IF.-, .,fq
-
.- ,. 40". t;,r't
..
--,i:- -.,
.j . ., - 'i, -,- .:..- ,9.1

li'- -i-..

I~~~~~~~~~~~~~~~~~~~I

.. t
B.rzecstn :..it o .

I
"

b. Bronze-castingpit over Mint (Mattusch N)

 PLATE 69
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Modern bronze foundry in Rome

*
*. .e_ . -. .'
v e _ * .o --,, .
-;: .**-^*';: _.
-:,-;<^.-^,^^^:.?W(^
.-_.*4

b. Aqueduct and Long Late Roman Wall

 < TEPRACE
SOUTHEAST
OF T
0
AGO
k4
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

_ =_ __ = --==== :Z
-ZZ

0 10 2o A-*
3o,,r
I

Southeast corner of the Athenian Agora. Actual-state plan

 PLATE 71
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

I.~

a . Ceta-M m i i

-eA ,it!---4

a.CentralMilU, mill room

a. Central Mill, mill room

b. Water wheel on Thasos

 PLATE 72
.
© American School of Classical Studies at Athens

L 54695479
L 5479
For personal use only. License: CC-BY-NC-ND.

L 5453

G 576 P 27463

b. Objects from final phase, Palace of the Giants

a. Sigma table, A 3869, Areopagus House C, final

phase

-~~~~~~~~~~~~~~~-w.

c. Red stone tray, ST 781, Areopagus House C, d. Fragment of stamped plate, P 20037, Long Late
final phase Roman Wall

e. Lamp, L 5628, AreopagusHouse C, f. Lamp, L 5630, AreopagusHouse C,

 PLATE 73

WoP -s . ' b .. - --

Lo a
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

.
S~~~~~~~~~~~~w;
_. r.:'

a. Areopagus House C. Clearing ashes from floor of triclinium

SS 112 SS 4259 SS 8075 SS 123

b. Stamps on ampullae from Palace of the Giants

* ; - -

V n s>^ G 5 i A
! ~--.
I I i ; "
-
-._.
-_ ~-_-e
>--~----~ ------ - - - .....-'-
..... -....;~-".
~.i-- ..-_ -----"<--:.
F5-f" '
tJ FkAl^
~,,,7 Ilii ^5;a-* C;? "

7
^-'-^^ "''^^ ^^^r-.^-^^~~~~~~1
tl -
1tHt44i - 4 - i- n

c. Well head in courtyardof Areopagus House C, d. Late house next to Tholos

 PLATE 74
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Destruction debris over Areopagus House C

b. IG II2, 3193, referringto fortress (from squeeze)

-
* i '

-
.S 's^s^^^^ss^
 PLATE 75

-
: . - -' '.:., ?'

**'
E *,^'^ *' ^ ^- io.;;,*0
*v "- . * e:' . _
. . .-
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Seventh-centurygrave in Palace of the Giants

vs

-
. -
. 4.

?......?. ** " . ; - r^.A .';

b* ;'

b. Tile kiln near Southwest Baths

 PLATE 76
© American School of Classical Studies at Athens
For personal use only. License: CC-BY-NC-ND.

a. Marble base for olive press in Palace

of the Giants
b. Millstone and foundationfor crusher in Agora

d. Olive crusher and press in Olympos, Attica (photographby V. R. Grace)

c. Oil press in Metroon

f. Fragment of plate from destruc-