Procopius and Dara

Procopius and Dara Author(s): Brian Croke and James Crow Source: The Journal of Roman Studies, Vol.
73 (1983), pp. 143-159 Published by: Society for the Promotion of Roman Studies Stable URL: http://www.jstor.org/stable/300078 . Accessed: 31/01/2014 19:27
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Society for the Promotion of Roman Studies is collaborating with JSTOR to digitize, preserve and extend access to The Journal of Roman Studies.
http://www.jstor.org
This content downloaded from 93.55.139.145 on Fri, 31 Jan 2014 19:27:56 PM All use subject to JSTOR Terms and Conditions
PROCOPIUS AND DARA *

By BRIAN CROKE AND JAMES CROW Plates XI-XII
in urban life fromantiquityinto later times The question of the degree of continuity has recentlyemerged as a preoccupation among scholars of the late antique and early medieval periods.1 Considerable attentionis currently being devoted to the fate of the to features classical cityand of the traditional patternsof urban life,with special reference such as reductionof physicalarea in individualcities,decline of population,changes in use or physical decay of classical buildings and the emergenceof new social organizationsas reflectedin urban topographyand development.2 Attemptsto refineand analyse these stemmedfromthe recentupsurgeofinterest issues withgreater precisionhave largely in late in such major excavationsas those at Sardis, Aphrowell demonstrated Roman archaeology, disias, Caesarea, Carthage, Ephesus and Thessalonika.3 Yet the archaeological evidence can only tell part of the story. We must still relya greatdeal on literary evidence, and the and the archaeologicalmaterialis the subject of this article. of the literary interconnection For all studies of late antique cities the single most importantcontemporary literary of Procopius of Caesarea. Where it provides(as often) or De Aedificiis, textis the Buildings, evidence for a particularsite we need to be able to evaluate its contributhe only literary few detailed discussions of the value tion critically. So far there have been remarkably or the archaeologicalstandpoint.4 As a result,the of its evidence eitherfromthe literary witnesseda major effort to secure and refortify traditionalview that the reignof Justinian based on the information the defencesof the easternfrontier, provided by Procopius, still largelyholds the field,despite the availabilityof othersources of information which could be used to clarify Procopius' picture.5 The Buildingswas designed as a panegyricon the Emperor Justinian. Its date is conin A.D. 554.6 Of the six books which comprisethe troversial, but it may have been written work (it is perhaps unfinished)at least the first, coveringConstantinopleand its environs, appears to have been designed for the imperial court; 7 the work as a whole draws on a mixtureof personal observationand what seems to be archivalmaterial.8 Withinthe constraintsof panegyric,Procopius sets out to list and describe the building activities of
* Unless otherwisestated all references are to the Buildings, and translations are generally those of Loeb Classical H. B. Dewing, Procopiusvii, Buildings, I971). Library (1940, reprinted 1 See, for example, A. Kazhdan and A. Cutler, in ByzantineHistory', 'Continuity and Discontinuity Byzantion 52 (I982), 437 ff.; more generally, R. Hodges, Dark Age Economics. The Originsof Towns and Trade (I982). 2 For an independent view and a guide to recent see AverilCameron, ' Images of Authority: literature Byzantium', Elites and Icons in Late Sixth-century 3-35 (reprinted in Past and Present 84 (I979), Byzantium, and Change in Sixth-century Continuity I98I) ; and P. Brown,' A Dark Age Crisis: Aspects ', EHR 88 (I973), of the Iconoclastic Controversy (reprinted in Society and the Holy in Late 1-34 Antiquity(I982), 25I-301). 3 Bibliographyon individual cities in R. Stillwell (ed.), The PrincetonEncyclopaediaof Classical Sites (1976) and subsequently in the annual volumes of Arch. Bibl. 4 This gives rise to such tvpical statements as that of J. A. S. Evans to the effectthat the Buildings ' provides a full and remarkably accurate account of Justinian'sbuilding programme' (Procopius (1972), 77). Soon to appear is a majornew studyofProcopius: Averil Cameron, Procopiusand theSixth Century. In the meantime the best guide to Procopius and his Buildings is B. Rubin, ' Prokopios von Kaisareia ', I RE 23. I (I957), 572-87 and Das Zeitalterjfustinians by W. see too the commentary (I960), I75-7;
5e.g. R. Krautheimer,Early Christianand Byzantine Architecture (I981), 27I : ' The securityof the Empire under Justinianand his successors entailed a vast building programmelasting until the turn of the century ... Procopius presents an impressive picture of the building programme'; or P. Brown, The World ofLate Antiquity (197 I), 154: ' Fromthe Black Sea to Damascus the emperor's foresight was crystallizedin stone '. An archaeologist'sview may be representedby R. E. M. Wheeler, 'The Roman Frontier in Mesopotamia', The Congressof Roman FrontierStudies,ed. E. Birley(1949), 124: 'Here in the Buildingsis a documented basis for the study of sixth-century withwhichthe fieldworker fortification, must familiarizehimselfat the outset '. 6G. Downey, ' The Composition of Procopius' De Aedificiis', Trans. Am. Phil. Soc. 78 (I947), 171-3. For the question of the date: E. Stein, Histoire du Bas-Empire II (I949), 837; and forthe emperor's reputationas a builder: J. Irmscher,'Justinian als Bauherr in der Sicht der Literatur seiner Epoche ',
Buildings (1977).
PiAlhornin 0. Veh's German translation of the
Klio 59 (I977),
as a Builder', Art Bulletin 22 (1950), 262-6. 7 G. Downey, 'Notes on Procopius, De Aedificiis, Book I ', Studies Presentedto David M. Robinson2 ' La Place des listestopony8cf. M. Perrin-Henry, miques dans l'organisationdu livre IV des Edifices de Procope ', Geographica Byzantina (Byzantina Sorboniensia 3, I980), 93-I06.
(I953), 719-25.
225-9
and G. Downey,'Justinian
144
BRIAN CROKE AND JAMES CROW
disbelievetheywere the workof a single man ' (i. I. I7). Justinian ' lest futuregenerations So vast was his work,says Procopius,that' you mightsay thatGod's provisionforthe safety of the Roman people allowed Justinian to succeed to the empirejust to be a builder ' (2. 9. thisimpression, he listshundredsofcitiesand forts i i). In the courseofreinforcing throughout the empire which Justinian supposedly eitherbuilt or restored. on imperial Although the Buildingshas no extant parallel as such, its concentration in imperialpanegyric. Yet this literary building formspart of a long tradition influenceon the structureand contentof the Buildingshas as yet hardly been properlyrecognized or analysed.9 The work as a whole still needs to be subjected to an adequate literaryand historicalcritique. That is not, however,our purpose here. Rather, we have chosen to evidence which can shed lighton Procopius' on the archaeologicaland literary concentrate account of Justinian'sbuildings in the eastern frontier region, and especially his lengthy of Dara in Mesopotamia. Such a studyprovides insightinto the descriptionof the fortress the achievements of Justinian techniquesused by Procopius in magnifying and will supplementthe detailedanalysisofindividualpassages initiated by Downey's studyofthe evidence of the Buildingson Antioch.10 to justify. From its foundationin The choice of Dara forsuch a studyis not difficult on the Persian border, and the literary A.D. 507, it was the major Roman fortress sources forit are relatively abundant and detailed. The cityhad a brieflife-just over a centuryto complicatethe archaeologicalrecord. Finally,while and thereis littlelater construction the physical remains at Dara are not as well preserved as those of other contemporary easternfortresses, notably Sergiopolis (Resafa) and Zenobia (Halibiye), the combination of extant literary sources and clearlydefinedphysical remains is unique, and makes it an obvious candidate for a comparativestudy.1"
I. THE EASTERN FRONTIER REGION IN THE BUILDINGS OF PROCOPIUS
of Justinian'sbuilding activity Systematicresearchinto Procopius' presentation in the east has so far been concentratedon his evidence for Antioch.12 His methodshere are instructive. Rather than discuss the rebuildingof Antioch afterthe damaging earthquakes of A.D. 5z6, 528, 553 and 557, he concentratesinstead on the aftermath of the Persian sack of the city in A.D. 540, which provides bettermaterialfor his panegyricalpurpose. He exaggerates inflicted the destruction by thePersiansand conveniently neglectsto mention that theyhad actuallyleftthe walls intact.13 We findselectiveomission and misrepresentation of this kind throughoutBook 2, which covers the cities and fortsof the eastern that is manifestly provinces(Fig. i). Sometimes construction earlierin date is attributed to Justinian. For instance,Procopius describesthe cityof Batnai (Suruc) in Mesopotamia as being unwalled beforeit was fortified by Justinian (2. 7. i8); yet it was certainly walled in 504, when the Persians seized it (Jo. Styl. 63), and it was rebuiltsome time afterby the Emperor Anastasius (ibid. 89). Likewise, in Book i, Procopius ascribes to Justinian the church of SS. Peter and Paul in the capital (I. 4. i), but since the church was certainly
9 Except now for Averil Cameron, op. cit. (n. 4 above), chap. 4. 10 For Sergiopolis: WV. Karnapp, Die Stadtmnauer von Resafa in Syrien (Deutsches Archiiologisches Institut,DenkmiilerAntikerArchitekturII) (1976); for Zenobia: J. Lauffray, ' El-Khanouqa, preliminairesgeographiques a la publication des fouilles faites a Zenobia par le Service des Antiquites de Syrie ', Annales Archeologiques de Syrie I (1951), 4158. See also the plan of the towersin Karnapp, figs. 11The literature on Dara is limited. See especially W. Ensslin, 'Zur Griundungsgeschichte von DaraAnastasiopolis ', Byz.-neugriech. 3b. 5 (1927),
I 00-9.
La fondation de Daras (Anastasiopolis) en Mesopotamie ', Melanges offerts a G. Schlumberger (1924) I, 55-60; C. Capizzi, L'imperatoreAnastasio (Or.
342-7 ; P. Collinet, ' Une " ville neuve " byzantineen 507:
13 Downey, op. cit. (n. 12), 36 1-78. The much vaunted Justinianic restoration appears to have fared no betterthan its predecessors,for in 573 the walls had largelycollapsed (Evagr., HE 5. 9).
ofMethodin theDe Aedificiis', Byz. I4
bibliography). For the remains at Dara (all with plates): C. Preusser, Nordmesopotamische Baudenkmtiler. Altchristlicher tand islamischerZeit (Wiss. Veroffent. d. deutschen Orient. Ges. I7) (I9II), 44C. Mango, Byzantine Archi5, fig. I2, pls. 54-7; tecture (1976), 24, 39; M. Mundell, 'A Sixth Century Funerary Relief at Dara in Mesopotamia', d. Ost. Byz. 24 (I975), 209-27 and J. G. Crow, _Jahrb. 'Dara. A Late Roman Fortress in Mesopotamia', Yayla 4 (I98I), 12-20. 12 G. Downey, ' Procopius on Antioch: A Study
(I939), 362.
Christ.Anal. 184,
I969),
2I6-2I
(includesfurther
PROCOPIUS AND DARA
I45
BLACK
SEA
Trapezus
Eucliaita
Coloneia Satala"
_ Sebasteia
, ~~~~~~~~~~Theodosiopolis
_Z
~ ~ ~ ~ ~ ~ ~~~~~~~~~Citharizon ~~~ 83
Viransehir
Mvelitene., _
<
Caesareia
3 Martyropol
Sozopetra
iS ts
Samosata
/~~~~~~~~~Zerz.evanl. Rabat
Constantina *
I Iisarkav Rhabdion ?
Anazarbus
Bezabde
-Birta*
Europus
Cyrrhus Antioch *Beroea Chalcis Apamea
;>
DARA
Resaina Batnae *ane Rsia . Carrhae
~~~~Nisibis
Thannouris
Singara D
Androna
Sra
Zenobia
Resapha 't * Palmyra
FIG. I. MAP OF THE EASTERN FRONTIER SHOWING THE PRINCIPAL DRAWN BY B. V. WILLIAMS.
150km
a~~~~~~~~~~~~~~Circesium
bvw
LATE ROMAN FORTRESSES.
146
standing in June 5I9, it seems clear that it was built before Justinian,probably by Anastasius.14 So too Justinianis given creditfor the harbour of Eutropius at Chalcedon (I. II. 22) whereas in factit was the work of a late fifth-century Eutropius.15 Another frequentProcopian device in the Buildings(as in the Secret History) is to count Justinian's reignfromSi8, the beginningof that of his uncle JustinI (I. 3. i). This allows the panegyristto credit all Justin's building activitiesto Justinian. So it is that Justin'spopular renovationof the walls of Edessa (Urfa) afterthe ruinous flood of 525 (Jo. Mal. 4I8) is listed as an achievementof Justinian, along with a considerablenumberof Edessan churches(I. 4. 29). Further,Procopius exaggerateseven more by suggestingthat the walls of Edessa were alreadydilapidated beforethe flood(2. 7. i i-i6), which is difficult thatthe walls were repaired, to believe since we knowfroma local contemporary description and a new outer wall built, just a few years beforehand(Jo. Styl. 89). Not only does Justinianacquire Justin'sedificesin Procopius' account, but the author also adds to the undertakenby Anastasius but only finallyfinished emperor's tally several constructions afterhis death. The walls of Melitene (Malatya) are a good example (3. 4. 19). in the Buildings are less easily detectable. Constantina Other misrepresentations withmud (2. 5. 2), together (Viran?ehir)is describedas a shoddilybuiltcity,itswalls thrown which is hard to accept in the case of a fortressthat withstood a Persian siege in S0I/2 and was the headquarters of the dux of Mesopotamia before Dara was built in S07/8.16 The same considerationapplies to Circesium (Buseire), the headquarters of the dux of Osrhoene; 17 this had the additional advantage of being surroundedby riverson all sides except one, which was completelyfenced offwith a long wall. These natural defences forChosroes in 540 (Wars 2. 5. 2-3). Nonetheless,Procopius claims proved too formidable that its walls had been ruined with the passing of the years until Justinianrestoredthem that this situationshould have been allowed (2. 6. 2). Quite apart fromthe improbability base like Circesium,the notion of walls crumblingthrough to develop at a major military age and neglect is a regular literaryand panegyricaltopos in the Buildings(e.g. I. 4. 9; 2. 8. 8-25; 3 4. 7; 3 4. 11 ; 3. 6. 14) We 2. 2. I; 2. 5. I; 2. 6. 2; 2. 7. 5; 2. 7. 17; have no means of assessing the credibilityof most of these examples. In the case of Callinicum, however,Procopius would have us believe that Justinianwas forced to dismantle the city's antique walls (2. 7. 17, cf. 2. 9. 2). CertainlyCallinicum, previously Nicephorium,had been a Hellenistic foundation,but it had been completelyrebuilt late in the fifth centuryand renamed Leontopolis afterits new imperial founder(Mich. Syr. 9. i (Chabot ii, 126)). Again, Procopius' blanket statementthat the city'swalls were in ruins needs to be treatedwith caution. it has been observed that Justinian'sfortifications, More generally, as described by after Procopius,were probablynot so important all, since much of the workascribedto JusIn supportof thisthe cityof Zenobia would appear to have been original.18 tinianis unlikely was responsibleforsome construction to providesome evidence. Certainly there.'9 Justinian Procopius provides an extended descriptionof his work (2. 8. 8-25) but there remain in his account, for it is inconceivablethat the isolated hill west of the city inconsistencies should not have formedpart of the earliercircuit. In practicethe citadel hill (2. 8. ZI-2) a fortified settlement at this point in the Euphrates was the primefactorwhich determined valley. Procopius is simplyemployinghere the same toposof high ground threatening the defenceswhich he uses at Antioch(2. I0. I2), ArmenianTheodosiopolis (3. 5. 9) and Dara (2. I. 26-7, with n. 6o below). It must be observed,however,that in the case of Zenobia
14 At least it is mentioned as complete in a letter of June 519 (Coll. Avell. 2i8, pp. 679-80). 15 Patria in. i66 (Preger,267). There is no reason to preferthe evidence of Procopius to that of the Patria as does PLRE 2 (I980), xxxvii, s.v. ' EutroEutropius 'protospatharios and quaestor' pius '. may be more accuratelyassigned to the early fourth (I964), byzantine2 century(R. Janin,Constantinople 16 Jo. Styl. 58. It was earlier the headquarters of the legion I Parthica (Not. Dig. Or. XXXVI. 29 (Seeck, were constructedin the early 78)). The fortifications fourthcentury by Constantius Caesar at the same
238-9).
time as the work at Amida (Amm. Marc. I8. 9. I). For a descriptionof the defences see J. G. Taylor, ' Journalof a Tour in Armenia,Kurdistanand Upper Mesopotamia with notes of Researches in the Deyrsim Dagh in i866', Journ. Royal Geog. Soc. 38 (i868), 28I-36I, esp. 354; much less survives to be seen today. 17 Also the headquartersof the legion IV Parthica (Not. Dig. Or. xxxv. 24 (Seeck, 76)). 18 E. Honigmann, RE Iv A, I 7 I 6 s.v. ' Syria'. 19Lauffray,op. cit. (n. io). See comments by
Karnapp, op. cit.(n. IO),
28
n. 99.
PROCOPIUS
AND DARA
147
building,because in 540 during thereis independentsupportforthe notionof Justinianic and run the Persian invasion Chosroes quickly bypassed Zenobia as being unimportant down(Wars 2. 5. 7). clear incongruities in the Buildings. For example, There are, on the otherhand, further the walls of Sergiopolisare describedas barelyadequate to deterSaracens,thoughmud walls fortresses, however, for that (2. 9. 3-4). Like so many of the frontier normallysufficed Sergiopolis had been rebuilt by Anastasius and renamed Anastasiopolis. It has even the work been argued that the impressiveremainsat Sergiopolis are substantially recently of Anastasius, ratherthan Justinian.20In any event,Procopius can be convictedof missince in 542 the city'sdefenceswere said to be sufficiently stoutto ward off representation, Chosroes (Wars 2. 20. IO ff.). In the BuildingsProcopius says that the walls of Sura (Suriya) were totallyinadequate and that it could be captured easily (Z. 9. i), yet in 540 (Wars 2. 5. 8-27). its walls were in fineshape and the citycould onlybe takenby treachery Procopius claims too that the walls of Satala (Sadak) were carelesslyconstructed(3. 4. 2) but in 530 the walls were secure enough to keep the and could be felled withouteffort, in a former at bay(Wars i. I5. 9-I7), just whatwe wouldexpect legionary headPersians work of quartersand the base of the dux of Armenia.2' At Chalkis (Kinnesrin) Justinian's can be dated preciselyto According to Procopius in the Buildings 550/Irt22 restoration the walls of the city were completelydecrepit(2. 2. i), yet only a decade earlier Chalkis off Chosroes (Wars 2. I 2. i). Once again it looks as ifProcopius and fought was well fortified the extentof Justinian'sworkat a particularsite. has over-stated These examples sufficeto show the extent to which Procopius distortedand exagthe emperor'sreputation. It suggeststhat one geratedin the Buildingsin orderto amplify containedin it, particularly in order to date must be cautious in handlingthe information thatbecause Procopius physicalremains. It also exposes the dangerofthe circularargument date for a particularconstruction, then remains of similar style in indicates a Justinianic other locations must also be Justinianic. It has been held, for example, that the walls of Amida (Diyarbekir)mustbe Justinianic simplybecause theyare similarto thoseof Dara, and building Procopius cites both as the work of Justinian(2. 3. 27). But the fragmentary name and providesno independentevidence recordsan official's at Amida merely inscription of Amida's defences.23 date forthe restoration a Justinianic to confirm This situationthenleads us to inquireabout the lightthatarchaeologicaland epigraphic evidence are able to shed on the Buildings. Relevant excavation reportsand studies are for the eastern provinces, but what we do possess only erodes our patchy, particularly was responsible stillfurther.Althoughit is apparent,forinstance,thatJustinian confidence and in the Balkans, the evidence for a good deal of building along the Danubian frontier to date suggests that it was Anastasius who was most responsible for the extensive reof places like Tomi and fortification programmethere.24 In particular,reconstruction ascribes to Justinian Histria which Procopius specifically were, it is now thought,actually the work of Anastasius.25 As faras the east is concerned,recentresearchhas concluded, as noted above, that the walls of Sergiopolis were built by Anastasius, not Justinian;26 can now be seen while Procopius' descriptionof North Africansites he knew at first-hand to be deceptivelyselective.27 of Justinianic Turning now to the epigraphicevidence,thereare eighteeninscriptions
Karnapp, op. cit., 5I-3. The legion involved was XV Apollinaris (Not. 59 Dig. Or. XXXVIII 13 (Seeck 84)) with RE2 (I92I), s.v. ' Satala '. For the remains see T. B. Mitford, 'Roman Frontierin Cappadocia' in D. Haupt and Roms ii H. G. Horn, Studien zu den Militdrgrenzen 0I977), 50I-I 6. 22 IGL Syr. 2, 348, 349. For the defences see R. Mouterde and A. Poidebard, Le Limes de Chalcis, Organisationde la Steppe en Haute Syrie Romaine (Bibl. arch. et hist., 38) (I945), 8-9, pl. i. 23 D. van Berchem,' Recherchessur la chronologie des enceintes de Syrie et de Mesopotamie ', Syria See, however A. esp. 262-7. 254-70, 3I (I954), Gabriel, Voyages archeologiquesdans la Turquie
20 21
orientale (1940), and D. Oates, Studies I34-5, I75-82 in theAncientHistoryofNorthern Iraq (i 968), I03-6. 24 J. Barnea, ' Nouvelle Contribution 'a l'histoire de la Dobrudja sous Anastase Ier ', Dacia n.s. ii (i 967), 3 55-6; V. Velkov, Citiesin Thraceand Dacia in Late Antiquity (I977), 47, io8, 2I3 and J.G. Crow, ' The Late Roman Frontier of Lower Moesia ', in The Frontiersof the Roman Empire (forthcoming). 25 I. Barnea, ' Contributionsto Dobrudja History under Anastasius I ', Dacia n.s. 4 (I960), 363-74. 26 Karnapp, op. cit. (n. I0). 27 AverilCameron, 'Byzantine Africa-the Literary Evidence ', in Universityof Michigan Excavations
at Carthage vii (I982),
3I,
33-6.
I48
of fortifications in Syria.28 Only threeof these-at Ma'an,29 date recording the construction Chalkis 30 and Cyrrhus31-are imperial inscriptionsrecording either the name of the emperor or of an imperial official. The remaininginscriptionsattest works of private armies in the sixthcentury.32The small defence,a symptomof the decline of the frontier is in sharp contrastwithepigraphicevidence from numberof imperialmilitary inscriptions the reconquered provinces of North Africa,where twenty-eight imperial inscriptionsare knownfromfortsand defencesof Justinianic date.33 A possible explanationforthe relative fromSyria may be found in the identification of the agency lack of imperial inscriptions which put up the fortifications. In Africawe may assume thatthe workof construction was largelyorganized by the occupyingfieldarmy,but for Syria and Mesopotamia it is clear that the Church frequently organized the constructionof defences,which were paid for by imperialgrants.34But even if this was the normalprocedure,it is unlikelythat the imperial authoritieswould have allowed such expenditureto have gone unrecorded. It can only be assumed that the small number of imperial inscriptionsin Syria actually reflects of fortification a limitedeffort and by Justinian comparable to the widespread contraction forcesthere,togetherwith a building up of alliances with local weakeningof the military tribes.35 fromCyrrhusprovidesa further The inscription illustration of the state of Justinianic defences. It is of special importancebecause, unlike those fromChalkis and Ma'an, it is of the acropolisgate. The defences in its originalposition,cut in the voussoirsand keystone to the reignof Justinian.36 of the acropolis may be dated withcertainty Anotherinscription of Justinianis reportedfromCyrrhus,37 but remainsunpublished. It is unclear whether was carriedout on the lower citywalls; possiblythe acropolis workof fortification further fortifications representthe main Justinianic work. The cityhad been in decline since the fourthcenturyat least 38 and it is quite probable thatthe vast earliercircuitwas abandoned by the sixthcentury. Contractionof the defendedcircuitis a featureof Justinianic fortificationsattestedin the law codes (e.g. Cod. Just.27. 2. I4) as well as by Procopius,even in the Buildings.39 For Cyrrhus,however,Procopius does not mentionany such featureof the defences; rather,they are described in extravagantterms (2. I. 4). One is, therefore, entitledto suspect that here too he is using an emptyrhetorical formulato describe what was now littlemore than a defendedhilltop. All thesefactsthrowdoubt on our traditional pictureof the relativeroles of Anastasius the variousfortresses and Justinian in reconstructing in the network of defencesin the East. Above all they compel us to look more closely at what Procopius at least appears to have most marvellousdeeds-the rebuildingof Dara. It is doubly regarded as one of Justinian's valuable to scrutinizeProcopius' picture of Dara because it is a place he himselfknew at first hand.
II. SITE AND CONSTRUCTION OF DARA
When the Persians finally capturedAmida in January503 aftera protracted siege, the Roman empire lost a crucial base forits expeditionsin Mesopotamia. Even aftertheyhad
28 Many of the Syrian defences are discussed in W. Liebeschuetz, 'The Defences of Syria in the Roms sixth century', Studien zu den Militdrgrenzen ii
archaeological evidence. 29 IGL Syr. 4, I809 (dated to 547/8). " IGL Syr. 2, 348, 349 (dated to 550). 31 IGL Syr. i, I45, I46, I47 (datedto c. 542). 32 Liebeschuetz, op, cit. (n. 28), 495-9. 33 D. Pringle, The DefenceofByzantineAfrica from 3ustinian to the Arab Conquest (B.A.R. 99), 1980. See also J.Durliat, Les dedicacesd'ouvragesde defense dans l'Afriquebyzantine (Coll. de l'Ecole fran9aisede Rome 49) (I98I), where one findsin the inscriptions from the reign of Justinian (Nos. I-23) the same tendency to exaggerate by representing a reconstructionor partial constructionas a complete one (e.g. No. 9, with commentary,pp. iio-i (cf. ILS
83I)).
34
(n.
2I
above), 490-3,
with references to the
3ahrhundert(I969), I37-8; A. J. Festugiere, 'La vie de Sabas et les tours de Syrie-Palestine', Rev. Bibl. 70 (I963), 92-3. 35 Liebeschuetz, op. cit. (n. 28). Although recognizing a contraction in the militaryforces in Syria, Liebeschuetz (491) maintains confidence in Procopius as a source, and regretsthat the Buildings ' becomes increasinglyselective so that the fortifications of Euphratensis do not include mere forts' 36 IGL Syr. 4, 1809. .3 E. Frezouls in J. Balty, Apamee de ,Syrie. Bilan de recherches archeologiques I965-I968 (I969), 90 n. 2. The originallocation of this inscriptionis unknown; it was found reused in the cardo. 38 E. Frezouls, ' Recherches sur la ville de Cyrrhus', Ann. Arab. Arch. Syrienne 3/4 (954),
39 Reduction at Leptis Magna (6. 4. 2-3) and Caesarea (5. 4. 7-I4); movementuphill at Mocissus and Bizana (5. 4. 15-I8) (3. 5. 15).
io6-i i.
D.
Claude, Die byzantinische Stadt im 6.
PROCOPIUS
AND DARA
I49
east, particularly purchasedAmida back the Roman armycould make littleheadwayfurther militum complainedto theemperorAnastasiusthat againstNisibis (Nusaybin). The magistri theycould notbe moresuccessfulagainstNisibis withouta nearerbase forsupplies and reinforcements, and refuge too when necessary; they needed a new fortressfurthereast of Amida and Constantina. Actingon thisadvice Anastasiustookthe decisionin 505 to turnthe small villageof Dara into a large,fortified and well-providedcitywhich would assume the role of a forwardbase for future Roman campaigns against the Sassanians.40 Dara was preferred to Ammudis,the othernominatedvillage,probablybecause it best combinedthe military requirements of good watersupplies with a naturallydefensiveposition.4'
<
~~~~~~~North
' .
0. '
.water
ditch quarries
Necropolis
(~bridge ;water gate
500 m
FIG. 2.
SKETCH PLAN OF DARA AFTER PREUSSER, I9II, REDRAWN BY B. V. WILLIAMS.
FIG. I2.
Dara was located eighteen km north-west of Nisibis and fivekm from the Persian frontier on the edge of the Mesopotamian plain just northof the modernNusaybin-Mardin highway. It stood on three hills and was isolated by a narrowstrip of low ground from the main scarp ofthe Tur Abdin plateau,whichrisessharplyto the north(Fig. 2). Through the centre flows the Dara9ay (a tributaryof the Habur), originatingfive km away at
40 Capizzi, op. cit. (n. 2I7-8 summarizes the II), sources. 41 In the dry season of 1903 British militaryobservers reported that Dara was the only watering point suitable for military purposes between Nisibis and Mardin (Admiralty War Staff Intelligence
(I917), and CentralKurdistan Mesopotamia
Division, A Handbook of Mesopotamia iv. Northern 268). (1895), in a Syrian monastery 0. H. Parry,Six months 159, notes thatDara is one of the best wateredvillages in Mesopotamia.
I50
Kordis; 42 like the late Roman fortressat Viran?ehir-Kalek6yin eastern Cappadocia,43 the site was chosen because the physical featurescombined to provide a good, naturally defended position for a copious wateringpoint. The circuitof the walls is irregularas it follows the crest of threehills. The internalarea is fairlysmall, with a maximuminternal diameterof about iooo m, and the area suitableforhabitationis limitedby the steep, rocky ground, especially on the east side of the valley. As a small village Dara came under the ecclesiasticaljurisdiction of the bishop of Amida and the incumbentof the see of Amida at the time of Anastasius,Thomas, played an instrumental role in the construction ofthe new city.44 At the same timethe distinguished Roman official Calliopius was placed in charge of the construction of the fortress and he arrivedthere afterthe conclusion of war in 505, first tracingout the perimeterof the new cityin the age-old way, with a hoe.45 The construction of a citysuch as Dara so close to the frontier was a blatantviolation of the treatyof 441 which restrainedboth sides from constructing fortresses within the border region (Proc., Wars i. 2. 15). The Persians protestedvigorouslybut were in no position to obstructthe building of the new city,since theywere more urgently distracted by the EphthaliteHuns (Wars io. IO. I5). Nonetheless,thePersiansdid despatchskirmishing partiesfromNisibis to interfere withprogress. As a resultthe Romans moved a contingent under Pharesmanes fromEdessa to Amida, in order to provide additional protectionfor those engaged in the construction of Dara (Jo. Styl. go). Furthermore, we are informed that when the Persian king Kavad was no longer preoccupied with the Huns and resolved to put a stop to the construction, the walls were already sufficiently high to defend those seekingrefugebehind them(Zach. Mit., HE 7. 6). The citywas presumably well advanced by November 506 when the magister militum, Celer, togetherwith Calliopius, who was apparentlynow stationed there, negotiated the final details of peace with the Persians (Jo. Styl. ioo). Zachariah informs us that the citytook two to threeyears to build (HE 7. 6). Since construction appears to have begun late in 505, it was presumablynot completed until late in 507 at the earliest. It was formidableand elegant,comprisingpublic baths, porticoes,storehouses,cisterns,a palace, churches,columns of Anastasius and many other buildings.46The emperor,following a traditional pattern, named the cityafterits imperial founder-Anastasiopolis-and bestowed on it the rankof metropolis, and it became the new base of the dux of Mesopotamia.47 The subsequent historyof Dara confirmedthe excellence of the original site and fortification.The city was besieged duringthe first Persian war of Justinian'sreign and was the scene of Belisarius' first victoryover the Persians in 530 (Wars I. 13. 9-14. 55). During the second Persian war (539-44), in the campaign of 540 which saw the capture of Antioch by Chosroes, Dara alone was able to resist a Persian siege.48 Apart fromthe restorations carriedout by Justinian, further work was undertakenon the waterworks and elsewhere early in the reign of JustinII (565-78).49 It was only in 573 that the cityfell for the first time to Persian assault. The conduct of the Roman commandersshows how the cityhad become a symbol of Roman resistancein the east. No attemptwas made to buy offthe Persians or to conclude a truce once theyhad gained controlof the defences. Rather,to the horrorof the citizens,the gates remainedlocked; therewas to be no escape. For seven days thefighting withinthe citycontinuedand the resultant carnagewas horrific.50 This blow to Roman prestigewas feltnot only on the frontier but also in the capital; it pushed the emperorJustinover the thresholdof insanity.51 From its capture in 573 Dara remained under Persian control. It could not be won
42
43 M. Restle, 'Viran~ehir-Kalek6y, ein befestiger Platz in Kappadokien', J7ahrb.d. 6st. Byz. 24
Reichs (I935'),
(I975),
44
E. Honigmann,Die Ostgrenzedes byzantinischen

ii
n. 6.
Zach. Mit., HE 7. 6 with Capizzi, op. cit. (n. 217-8. 45 Marcell. com. Chron. (MGH. AA. xi, ioo) with B. Croke, 'Marcellinus and Dara: A Fragment of his lost de temporumqualitatibus et positionibus locorum', Phoenix 37 (I983).
I I),
I96-207.
49 Menander, fr. 15 (FHG IV 220). 50 Jo. Eph., HE (Payne-Smith, 382-3) with P. Goubert, Byzance avant l'Islam I (I95i), 69-71. 5' Theoph. Sim. 3. II. 2-3; Mich. Syr. (Chabot II, 312).
46 Jo. Mal. 399. 15-17; Evagr., HE 3. 37; Zach. Mit., HE 7. 6. 47 Jo. Mal. 399. 20; Procop.,WarsI. 22. 3. 48 Liebeschuetz, op. cit. (n. 28), 487-99, esp. 498.
PROCOPIUS AND DARA
151
back by negotiation,nor by force,until it was finallyconceded to the Romans in 59I as part of the price paid forRoman supportby the Persian king Chosroes II in his attemptto recoverhis thronefromthe usurper Vahram.52 Thirteen years later a Persian armywith all its elaborate machinerywas once again at the gates of Dara; and when they finally took the city in 6o6, afteran eighteen-month siege, the Persians proceeded to dismantle the walls.53 Dara changedhands once again duringthe victoriouscampaignsof the emperor Heraclius in the 62os, but finally fellto the Arab armyunder lyad ibn-Ghanm,along with the restof Roman Mesopotamia,in 639.54 No longera frontier to its former city,it returned status,althougha Syrianbishop is attestedthereuntilthe thirteenth century.55As withso manyof the new citiesof the sixthcentury throughout the empire,urban lifeat Dara ceased once its preciserole was removed. By contrast, the ancientcitiessuch as Nisibis and Edessa continued as urban centresthroughoutthe middle ages until the present day. Nowadays the name of Dara is preservedonlyby a small villagescatteredwithinthe shell of the sixthcenturyfortress.56 The most informative source for the foundationof Dara is a lengthyaccount in the historyof Zachariah of Mitylene (preserved in a Syriac redaction) emphasizing the role of Thomas and the clergyof Amida as the agentsof Anastasius. There is anotherdetailed and contemporary account by a former aide of Justinian, Marcellinus comes, which is preserved in the St. Omer manuscriptof his Chroniclebut which appears to derivefromhis lost volumes ' On the Locations of Places'. It is quite possible that Marcellinus himself had actuallyvisitedDara.57 Otherwise, thereis a briefnoticein the sixth-century chronicle of JohnMalalas (repeated in the Chronicon Paschale, Theophanes and Cedrenus), to which may be added Evagrius Scholasticus(probablybased on Malalas) and the detailed regional history ofJoshuathe Stylitewritten in 507 duringthe city'sconstruction.The mostdetailed eye-witness description of Dara comes, however, from Procopius who accompanied Justinian's general Belisarius there in 529/30, although Procopius tells us less about Anastasius' original constructionof Dara than about Justinian'sextensive rebuildingof the city.
III. OUTER AND INNER WALLS
Procopius' account of Dara opens the second book of the Buildings(2. I. 4-3. 26), which describesJustinian's workof construction in the frontier provincesof Mesopotamia and Osrhoene. It is the longestdescription ofany individualsite,withthe exceptionof Constantinople; more than double the lengthof the account of Antioch (2. IO. 2-25) and four times thatof Zenobia (2. 8. 8-25). In its detail,it is comparable only withthe description of the Great Church of Constantinople,St. Sophia. Both by its position in the Buildings and by the degree of detail given, Procopius presents Dara as a model for Justinian's achievementin restoringthe securityof the empire. The account follows the standard order which Procopius employs in the Buildings-fortifications, waterworks,public buildings and churches-and opens with a briefsummaryof the foundationof the cityby Anastasius,referring the reader back to his own account in the Wars (I. 9. 20). Even the most cursoryreadingwill revealobvious distortions.For example,Procopius would have us believe that Justinian was responsibleforbuilding barracksfor the soldiers at Dara (2. 3. 26), when we would expect them to have formeda natural component of the originalconstruction of what was essentially a military base. More seriously,he credits the emperorwith the two main churchesat Dara: that of St. Bartholomewand the main church of the city,the so-called ' Great Church ' (2. 3. 26). The fact is, however,that
52 Theoph. Sim. 5. 3. Io with Goubert, op. cit. (n. 50), I67-8. 53 A. Stratos, Byzantiumin the SeventhCentury I.
54
Theophanes A. M. 6130 (de Boor, 340. 25) with A. Stratos, Byzantium in the Seventh Century II. 634-4I (1972), 85. 55 R. Janin, ' Dara', Dict. d'hist. et de geog. eccl. 14, 83-4; Mundell, op. cit. (n. I I), 225-7. 56 A list of some of the travellers' accounts of Dara
6IO-34 (I968),
61-2.
is to be found in Mundell, op. cit. (n. iI), 2I2 ff. Photographs taken by Gertrude Bell are a valuable record of the site seventyyears ago. These are now kept in the Bell Collection, Universityof Newcastleupon-Tyne. Photographs from a more recent explorationof Dara by Cyril Mango, Ihor 9ev6enko and Marlia Mundell Mango are housed in the Byzantine Photographic Collection at Dumbarton Oaks, WashingtonD.C. 57 Croke, op. cit. (n. 45).
I52
St. Bartholomew'swas certainly the work of Anastasius (Theod. Anag., HE 2. 57 (Hansen I57. 9-II)), while Zachariah of Mitylene (HE 7. 6) leaves us in no doubt that the 'Great Church ' was a chiefpreoccupationof the bishop and clergyof Amida involved in Dara's construction. Justinianis praised firstof all for his rebuilding of the walls of Dara. Procopius describes how the haste of the originalbuilding led to a number of structuraldeficiencies in the defences: the walls were too low and the masonrywas loosely laid withpoor quality mortar. This faultyconstruction was especiallyapparent in the towers,which had begun to collapse because of the processes of weathering (2. i. 6-io).58 Recognizingthe strategic importanceof Dara and the special danger fromthe sophisticatedPersian siege engines, Justinianordered the defences to be restored(2. I. II-14). The emperor provided the followingremedies: ofall he rendered first thewall(which, as I havesaid,was very lowand therefore veryeasy for anyenemy to assault)bothinaccessible and wholly impregnable foran attacking force. For he contracted the original apertures of the battlements by inserting stonesand reducedthemto very narrow slits, leaving onlytraces ofthem in theform oftiny windows, and allowing them to openjust enoughfora hand to pass through, so thatoutlets wereleftthrough whicharrows couldbe shotagainst assailants.Then abovethesehe addedto thewall a height ofaboutthirty feet, notbuilding theaddition uponthewholethickness ofthewall,lestthefoundations should be overloaded by the excessive weight whichboreupon them, so thatthe wholeworkwould suffer someirreparable damage, but he enclosedthespace at thatlevelwithcoursesof stones on the outsideand constructed a colonnaded stoa running all aroundthewall, and he placed the battlements abovethisportico, so thatthewall really had a doubleroofthroughout: and at thetowers there wereactually three levelsforthemenwho defended thewall and repelled attacks uponit. For at aboutthemiddle ofeachtower he addeda rounded structure uponwhich he placed additional battlements, thusmaking the wall three-storeyed (2. I. 14-17, cf. IT7ars
2. 13. 17).
Although many of the towers were in ruins, they could not be demolished lest the enemy suddenly attack. To overcome this problem, Justinianordered that the defective towersshould be reinforced with externalstonework cladding, rectangular in shape. Each tower was heightenedby 'an extra turretplaced on its summit (2. I. I5-22). The outer wall (-rrpoTEiXlXapa) was also increasedin height,and on the south side, where the approach was level and the ground suitable forsiege mines,59 he built a crescent-shaped, water-filled ditch with an outer wall to preservethe circuitwalls fromassault (2. I. 23-5). A mound which lay between the main walls and the new outworks had covered enemy mining As a prefaceto his account of Justinian'srenovationof the walls of Dara, Procopius claims that the walls of Anastasius' city were hastilythrowntogetherand inadequate, so that within a short time the elementsalone reduced them to total ruin (2. I. 7-I9). We have alreadyobserved how Procopius utilizesthis panegyricaltoposeven when it is clearly untrue (e.g. 2. 5. 2, 2. 6. 2, 3. 4. I9, 3. 5. 7). It is no less false and exaggeratedin the case of Dara. Zachariah's account of Anastasius' building of Dara (HE 7. 6) and Joshua the Stylite'sclaim thatthe walls were certainly strongenough to deterthe Persians even before theywere completed(Jo. Styl. go) can be contrastedwith Procopius' versionin which the original workmenwere so fearfulof imminentPersian attack that:
58 Procopius notes that the earlier walls were too low at a number of sites: 2. 5. 2 (Constantina); 3. 5. 6 (Theodosiopolis (Erzerum), with the restoration described as similar to that at Data (3. 5. IOI2)); 3. 2. io (Martyropolis). That the walls had been poorly built is a common assertion, e.g. at Amida (2. 3. 27-8), Constantina (2. 5. 2), Chalkis
and thiswas levelled(2. operations
I.
26-7).60
tionsat Dibsi Faraj, northern Syria,1972-1974:
were rebuilt in hard stone. Excavations at Dibsi Faraj-Neocaesarea showed that the earlier towers built of a soft limestone were encased in new walls of conglomerateand brick (R. P. Harper, ' Excava-
(2. 9. i) and Hemerium (2. 9. io), where the walls
preliminarynote on the site and its monuments', DOP 29 (I975), 3I9-37, esp. 326-8). The rebuilding is dated after453 and may be Justinianicin date. 59In his account of the siege of 540, Procopius states (Wars 2. Q3. i6) that the Persian siege tunnel lay to the east, the only approachable side. He tacitly correctsthis in the Buildings(2. I. 24). 6' This mound was probably a tell, the remainsof earlier settlementon the site. The plain south of Dara is still filled by the remainsof prehistoricand later settlementsof this form.
PROCOPIUS
AND DARA
I53
extreme with care,sincethehasteinspired bytheir they did notcarry outthebuilding eagerness to keep company with work. For stability is neverlikely detracted from the stability of their outtheconstruction of swiftness. Theytherefore carried speed,noris accuracy wontto follow but raising it only madeit fit to withstand theenemy, thecircuit-wall in great haste, nothaving carenecessary;indeedtheydid notlay thestonesthemselves to such a height as was barely should,or bindthemproperly at thejointswithmortar.So or fit themtogether as they fully, thesnowsand theheat could not in anywaywithstand within a short time,since the towers it came aboutthatthe mostof themfellintoruin(z. i. becauseof theirfaulty construction,
7-10).
There are two reasons for accusing Procopius of error and misrepresentation here: firstly, it is scarcely conceivable that such neglect would have gone unchecked at the base againstthe Persians,especiallynotin thecase of Dara, Romans' mostimportant forward now theheadquartersofthedzixof Mesopotamia; secondly,it plainlycontradicts Procopius' own reportselsewhere. Far from being dilapidated so soon afterbeing built, the walls to have feetforJustinian must have been standingsolidly to theiroriginalheightof thirty the fact simply built on top of them, as Procopius says he did (2. I. I 6). -Furthermore, that in his eye-witnessaccount of the battle outside Dara in mid-530 Procopius gives no hint that the defenceswere inadequate in any way (Wars I. I3. 9-14. 55) should suggest that they were in perfectand well-maintainedcondition afterall. Although stationedin Dara himselfin 530 Procopius never returnedto the city after so for subsequent incidentsin and around Dara he the expiryof Belisarius' appointment, would have had to rely on the reports of other participantsand perhaps some wvritten sources.61 If his information is true that when Chosroes attemptedto take Dara again a decade later (540) the walls were sixtyfeet high and the towers one hundred feet (Wars 2. 13. I7), just as described in the Buildings(2. I. I6), then Justinian'sprogrammeof refortification must have been undertakenduringthe 530s. In otherwords the Dara that Procopius knew at first-hand was the originalAnastasian city completed after507. So for the account of Justinian'swork at Dara contained in the Buildingshe must have relied sources. What All thiswe can deduce fromour literary solely on second-handinformation. remainsat Dara? It is to this further light is thrownon the matterby the architectural question thatwe now turn. The fortifications side of the city at Dara are best preservedalong the north-eastern systemcomprisesa tall curtainwall, flankedat inter(Fig. 2, P1. XI, I). Here the defensive vals ofabout 50 m by largeU-shaped projecting towers towers. Betweenthesemain interval are small rectangular turrets, normallytwo in number. The U-shaped towershave a high, domical vaulted,circularchamberat first floorlevel (P1. XI, I), whichhas fivewide, arched windows probablynarrowedto the fieldby a thin curtainof facingstone.62 In the northeasternsectormuch of the innerface of the curtainis obscured by modernvillage houses, but at the east tower flanking the southwatergate are preservedthe remainsof open stairs builtwithmonolithic supportedby corbelsand partly bonded into the rearface treadspartly of the towerwall.63 Even if no traces of these stairswere preserved,the massive domical stairs. vault withinthe towerwould suggestthem,as it is unlikelyto be pierced by internal The outer wall is well preservedin the north-eastern parallel to sector; it runs straight, the main curtain withoutany expansions in frontof the towers as are found at Amida, Resina (Theodosiopolis) and Singara (Sinjar) (Fig. 2, P1. XI, i). A rock-cutditchis seen at the north-eastern angle. At the northwatergate (P1. XI, i), as distinct fromthe northgate, the curtainsurvives have remained. The up to the coping stones of the parapetwalk, althoughno crenellations
616. 7. I8 (writtenand oral sources); 2. 4. I-5 (based on directlocal experience). Archivalmaterial is probably the source of his detailed lists (4. 9 4. i i), cf. Perrin-Henry, op. cit. (n. 8.) 62 See the photographsin Mango, op. cit. (n. iI), pl. 37 and in S. J. Hill, GertrudeBell (i868-i926) (I976), fig. Io. The blocking of these windows is probablypost-Roman,when the towerwas converted for domestic use. 63 Similar open stairsbehind the towersare known fromAmida and Sergiopolis. For Amida see Gabriel, op. cit. (n. 23), 96 ff.and for Sergiopolis, Karnapp, op. cit. (n. io). Staircases at the rearof towers,on the inside, are found at Zenobia (Karnapp, op. cit.), Dibsi Faraj (Harper, op. cit. (n. 58), fig. c) and de l'archiAntioch (G. Rey, Etude sur les monuments en Syrie et dans 1'Ile de tecturemilitairedes croise's Chkpre(187 I), I88-9, figs.48, 49). It is unlikelythat of date between there is any significantdifference the two types of staircase.
154
facing for the curtain is of well-dressed,smooth ashlar blocks interruptedonly by the conduits for the river and on the inside by the arched entrance for an intervalturret. The wall core consists of hard, mortaredrubble with vertical bonding courses of ashlar headers. This use of verticalbondingcourses is reminiscent of ' opus punicum ' commonly found in Africa, but probably unique in Mesopotamia and Syria. Horizontal bonding coursesofashlarheadersand stretchers oflate Romanconstruction werean establishedfeature methods, identical in functionto brick bonding courses; 6 it is possible that vertical bonding courses were designed as a precautionagainst mining,since theywould limitthe section of wall which could collapse because of subsidence.65 Similar facingmay be seen between the northwatergate and the north-eastern angle, but the verticalbonding is only certainin the east tower flanking the northwater gate. The south water gate is identical in construction to the north(P1. XII, 2). West of the northwatergate the curtainclimbsthe steep north-western hill; in general it is poorly preservedand it is not possible to determinewhetherthe systemof flanking towers and turrets was continued,as only one U-shaped tower is well preserved(Fig. 2). On the west side of the enceinte,the foundationsof U-shaped towerswith circularinternal chambers were seen in the saddle between the north-western and south-westernhills. But confirmation that the systemof turretsand U-shaped towers continued may be seen at the south-western corner. Althoughthe curtainwall survivesonly in fragments of wall core, it was set on a foundation of bedrock,withcuttings on the surfaceto receivethe ashlar facingstones. These rockfoundations werecut so thatthe quarryface was flushand vertical with the curtainwall and towersabove, and acted as a socle of bedrockat least 4 m high.66 The outline of the quarryface showed that the curtainwas defendedby U-shaped towers and rectangular turrets, identicalin plan to those fromthe north-eastern curtain. The main evidence for the structuralhistoryof Dara is found at the survivingUshaped tower on the north-eastern curtain,west of the river(P1. XI, 2). It is located at a point where the curtaindeviates slightly towardsthe west. Examination of this structure shows it to be the westtowerofthe northgate. The gate is of a simple design,a single portal gatewayflankedby two U-shaped towers. The east tower survivesonly in the jambs of the doorway and in the scar of the robbed east wall of the curtain. In frontof the west towerare tracesof the outerwall, which is positionedon the edge of the northscarp of the hill. As the northcurtainis set back fromthis edge, this positionwould suggestthat the inner and outerwalls are contemporary. Two distincttypesof masonrywere seen at the gate. Type A is of well dressed,plain ashlar blocks, facingto the frontand rear a compact, mortaredrubble core. The mortar is uniformly hard and has survivedwell in the joints of the facingstones and of the core where it was covered by the later work. The gate curtain and the west tower, which is bonded into the curtain,are built in this style. Type B abuts the rear face of the gate curtain. It is faced only to the rear, as the core abuts the inner face of the Type A gate curtain. The ashlar facing stones are larger and less regularlylaid than Type A and the stonesof the core were larger,bonded witha softer, less durable mortar, whichsurvived poorlyin thejoints of the facingstones. Type B masonryran across the rearface of the gate curtainand west tower. At the gateway, the jambs of Type B construction showed that it had been narrowedand the springersindicatedthatthe archhad been lowered (P1. XII, i). Furthertraces of the additionalwall of Type B masonrywere found east of the gate, low down the slope. Whetherthis additional wall was continuousis unclear. Elsewhere along the circuitmuch of the curtainsurvivesonly as a tumble of stone.
64 For stone and brick horizontalbonding courses see J. B. Ward-Perkins,' Notes on the structure and building methods of early Byzantine Architecture' in D. Talbot Rice, The Great Palace of theByzantine Emperors, Second Report (I958), 52-I04. Stone bonding courses are to be seen on the walls of Antioch. 65 In Greek fortifications, towerswere not bonded into the curtainfora similarreason,see F. E. Winter,
Greek Fortifications (I97I),

66
Deep rock-cut ditches were a feature of late
I58
n.
3I.
Roman fortificationsin Mesopotamia; see the description of Kale Hetmi Tay, possiblythe Rhabdios described by Procopius (2. I. I-I3), in J. G. Taylor, ' Travels in Kurdistan with notices of the sources of the eastern and westernTigris, and ancient ruins in their neighbourhood', Yourn. RQval Geog. Soc. 35 (I865), 2I-58, esp. 5z; and Rabbat Kalesi in Taylor, op. cit. (n. i6 above), 360-I, and G. Wiessner, Nord-mesopotamische Ruinenstdtten (Studien zur spiitant.u. fruhchrist. Kunst, bd. z) (I980), pls. I02I.
PROCOPIUS
AND DARA
155
But at two places, in the saddle, betweenthe south-west and north-west hills, and similar immediately east of thesouthwatergate,largeashlarfacing stoneswerenoted, to thetypeB construction from thenorth gate. On theupperpartofthecurtain wall ofthe north of ashlar water gate,adjacent to theflanking U-shapedtower, maybe seen a section masonry, which is larger and lessregularly laidthantherest ofthecurtain.This secondary work is similar toTypeB construction andprobably represents a repair totheuppercurtain.67 The structural sequenceat the north gateis clear. In Phase I, the gatecomprises a singleportalgateway flanked by U-shapedtowersof Type A construction.Phase II represents modifications to the gate by thickening the curtain to the rear and slightly narrowing thegateway.It is builtofType B masonry, inferior in quality to thatof Type A. In theabsenceof anyadditional archaeological or epigraphical evidence we mustturn to theliterary sources in order to datethesephases. Two mainperiods ofconstruction on thewallsaretheworkofAnastasius and Justinian. PhaseI is clearly primary and we may reasonably assumethisto be Anastasian.Phase II is likely to be Justinianic. How far do these structural observations comparewithProcopius'accountof the Justinianic work at Dara ? First, thetowers: Procopius doesnotspecify theoriginal shape of thetowers, but he is clearwhendescribing the rectangular cladding added to manyof thetowers becauseoftheir disrepair (2. I. i9). None ofthetowers examined showedany trace ofsucha repair but,sincenotall thetowers arenowextant, itis possible thatcladding was carried out on a few. Furthermore, he unfavourably compares theinferior quality of theearlier work with that ofJustinian suggest that theopposite (2. I. 7-IO), buttheremains wastrue, for thePhaseI masonry usesa harder mortar than thePhaseII additions. Second, theaddition to thecurtain wall: Procopius records thatthecurtain wallwas raisedwitha vaulted gallery (2. I. 15-17). At onlyone pointdoes thecurtain survive to itsfullheight, at the north watergate. This is clearly of one build,without anyextragallery, although from its position it maybe atypical. However,it standsin excessof 30 feet, the figure whichProcopius considered to be theheight of the original wall.68 The evidence of the additional wallofType B construction at thenorth gatecouldbe interpreted as theinternal base forthe addition of thevaulting whichProcopius describes, but alternatively, sinceit does notsurvive as a continuous wall,it maybe considered to be a widening of the gate curtain and theprovision of a staircase behindit. Two photographs takenby Gertrude Bell during hervisitto Dara in igiI show an uppercurtain wall rising abovethefront ofthemaincurtain at thesouthwatergate. No traceof thisupperworknowsurvives, but it is possibleto estimate thatthiswall roseto a height of about i 8 m or 6o feet, a figure similar to thatrecorded by Kinneir and which agreeswiththeheight givenby Procopius in the Wars.69 Fromthetwo Bell photographs none of the specific features whichProcopius detailsin the Buildings can be recognized. The outerfaceofthewall appearsto be of one buildwithout thetell-tale blocking of the embrasures, and theinner facedoes notdisplay anytraceofcorbels or vaulting to support an upperrampart walksuchas survives at Resafeand other frontier.70 siteson theEastern At no pointdoes the structural record correspond to the description givenby Procopius exceptforthe totalheight of the wall,a factattested in the Warsbut notthe Buildings. thatthe outer Third, the repairsresulting froma greatflood: Procopiusrecords wallwas destroyed, further theinner wallrequired repairs and thenorth gatewas removed up thehill(2. 3. 22-3). No comment maybe madeon theouter wallas it does notsurvive north of the north watergate. The maincurtain appearsto survive in its primary state;
67 Rough ashlar is also seen on the inner face of the north water gate, low down, in the spandrels of,the arched conduits; this is distinct from both Types A and B. A detailed surveywould no doubt reveal further variationin masonrystyles. 68 As noted above, Procopius states that the walls were sixty feet high in 540 (Wars 2. 13. i6), from which we are obliged to assume that their original height was thirtyfeet. In practice, however, the curtain is unlikely to have been a constant height overall, given the varied nature of the terrain. At
Amida, Gabriel (op. cit. (n. cussed by Crow,op. cit.(n.

II, 12;
that the curtain varied from 8-I2 m. 69 The Bell photographs are published and disii
23), 96-I I3)
records
Antioch (Rey,op. cit. (n. 63), I92, figs.50,
an earlier description of the south water Asia gate is found in W. Kinneir, tourneythrough Minor, Armenia and Koordistan (i8I8), 440-l. 70 Karnapp, op. cit. (n. Io above) ; also at Hisarkaya, (Wiessner, op. cit. (n. 66), pl. 5) and at
51).
above), 17-18, figs.
I56
the formof the turretsand the type of construction are both identical with the curtain wall to the east. Also the main curtainis identicalin construction to the south water gate and thereis no suggestionthat this required any renewal. Secondary workis restricted to the masonrysimilarto the Type B construction described above. From its position,this later work is too high up the curtainto have been occasioned by flooding; ratherit was probably required afterbombardmentby stone-throwing artillery. As to Procopius' new northgate,we have seen thatthe northgate which survivesis primaryand is located above the level of flooding. No trace of an earlier gate closer to the riverwas observed. The descriptionof the northgate would appear to be mistaken,but it may provide an insight into how Procopius composed his account. Learning fromsome source, probably official, thatthe gate had been rebuilt,he elaboratedthisfactby combiningit withthe storyof the flood and the emperor's devices to preventfurther damage. Certain aspects of the descriptioncannot now be verified fromthe survivingremains. None of the towersremainsto its originalheight,so it is not possible to look for evidence of the additional turrets which Procopius records. No trace of the additional outworkon the south side was seen, althoughit mightshow up when the fieldsare under crop. However, a line of triangulardressed stones was observed close to the modern road leading south from Dara. These are probably the coping stones for the merlons of the parapet. They may represent tracesof thisadditionaloutwork, the more regularashlarblocks having been removedforre-use in modernbuilding. Finally, it is worth noting one omission in Procopius' descriptionof the walls that would appear to be significant.In his contemporary account of Anastasius' originalconstruction, Marcellinusdescribeswhatwas knownas the' Herculean tower' (turris Herculea) which appears to have formedpart of the northwall and was probably located at the top of the slope in the north-west sectorof the city. This enormoustowercommanded a panoramic view and acted as a lookout in the directionsof Amida and Nisibis.71 There is no mention of this tower in Procopius' descriptionof the decayed state of the Anastasian walls, yet the towerwas still standingtall and formidablewhen the Persians tried to build a siege towerhigherthan it in 573 (Jo. Eph., HE 6. 5). The factthatthis Anastasiantower was still the highestpoint in the city'sfortifications afterJustinian'srebuilding,combined with Procopius' failureto mention it, suggests that the original walls had by no means suffered fromtimeand the elementsto the extentsuggestedby the panegyrist (2. I. 7-IO). In most cases, therefore, where we can directlycompare Procopius' account of the are either fortifications with the structuralremainsat Dara it appears that his statements literally false or else at least to be treatedwith reservation.In fact,hardlyany detail of his descriptioncan be positivelyconfirmed, althoughcertainpoints can no longer be verified because the physicalremainseitherdo not surviveor are unclear. So it is difficult to avoid the conclusion that for Dara Procopius is not a reliable witness. Furtherconsiderationof Procopius' account only serves to reinforce this conclusion.
IV. DARA S WATER SYSTEM
Having consideredthe fortifications, Procopius' account continueswith a description of Dara's water systemwhich exemplifiesanother Procopian topos: that is that in the BuildingsJustinianis constantly creditedwith redirecting or blockingthe course of rivers (2. 7. 7-I0; 8. i6-i8; io. 6-8, i6-i8), like Queen Semiramis at Babylon (Herodotus I. or else with constructing I85); new channels, cisternsand aqueducts in cities and forts where watersupplies were always precarious(2. 4. I3, 22-4; 5. 9-II ; 9. 6; II. 5-7). In the time of Justiniancisternswere constructed, Procopius informsus, between the inner and outerwall and near the churchof St. Bartholomew in orderto storewater (2. 2. I-2),72 cominginto the cityfromthe river,which flowedfromthe northand could not be diverted fromsupplyingthe citybecause no level groundintervened.The riverwas canalized before
71 Chron.(MGH.AA. xi, Ioo) with Croke, op. cit. (n. 45). 72 Since the citywas alreadyprovided wvith cisterns (Zach. Mit., HE 7. 6), those built between the walls were probably to provide a water supply for cattle
penned therein time of siege (cf. Proc., Wars 2. I3. I8). At Beroea (Aleppo) in 540, cattle exhausted the water supply in the citadel and the citizens were forced to surrender(Proc., Wars 2. 7).
PROCOPIUS AND DARA
I57
it enteredthe city and at the curtainwall the conduitswere secured by stout iron bars.73 and were led out of the cityto the The watersflowedthroughthe city,filledthe reservoirs plain beyond in a manner similarto that described for the northwater gate (2. 2. 3-6).74 of controlling the watersupply and how Procopius' accountthenturnsto the difficulty this was resolved,the chiefproblemsbeing the need to deny an enemyaccess to the water supply duringa siege and to preventflooddamage to the cityand its defences(2. 2. 7-9).75 one ofthe membersof the garrison He thengoes on to explainhow, in the timeofJustinian, at Dara had a large pit dug down at a certainspot inside the city. He therebydiscoveredan undergroundriver ' eitherin consequence of a dream or led to do it of his own accord' It just so happened thatat the timeheavy rains broughtinto floodthe riverthat (2. 2. IO). normallyflowedthroughDara. When the cityitselfwas awash, it was observedthat rather on its normalcourse out of the southernside of the walls, the riverflowed than keep flowing into the pit instead and then went underground,resurfacing days later about fortymiles away near Resaina-Theodosiopolis ' and it was recognized by the objects which it had carried offfromthe houses of that city' (2. 2. I6).76 This was a useful discovery,for it meant that in futurewhen the city was under siege fromthe southernside, the river's flow could be halted by closing sluice gates in the citywalls and forcing the riverinto its undergroundcourse (2. 2. I7-I8). Procopius then cites two occasions when the besieging armywas hard-pressedby lack of waterand forcedto abandon the siege ' outwittedby the While thereis no reason to disbelievethe factsof this account,since such underground riverswere widelyknownin Mesopotamia (Philostorgius, HE 3. 9), thereis reason to doubt Procopius' integrity in listingthe exploitationof the discoveryamong the achievementsof Justinian.If thatwere the case, it would probablyhave formed partofJustinian's rebuilding programmeat Dara after530. Yet Procopius goes on to say thatthis engineering featwas employed against the Persians who came to besiege Dara ' during the reign of Cabades' when Procopius himselfwas there. One's suspicions are (2. 2. I9) that is in 530, precisely a later campaign at Dara immediatelyaroused,especiallybecause in the course of narrating versionof this phenomenon: he offers quite a different But as soon as thisriver wall (irpoTEiXtoya),itflows abouttheentire getsinsidethecircuit city, and thenflows its cisterns, it fallsintoa chasm, filling out, and veryclose to the circuit-wall whereit is lostto sight. And where it emerges from there has becomeknown to no manup to thistime. Now thischasmwas notthere in ancient theemperor times, but,a longtimeafter builtthiscity,natureunaidedfashioned Anastasius and placed it there, and forthisreason it comesaboutthatthosedesiring to drawa siegeaboutthecity ofDara are veryhardpressed ofwater by scarcity (Wars8. 7. 8-9). What this passage makes clear is that when Procopius was writing the eighthbook of his history of Justinian's wars,thatis in 553/4,77he was not aware of the versionhe gives in the Buildingsabout how and when the pit was built and the flood occurred. It therefore followsthat the eventsreportedby Procopius cannothave takenplace by 530 when he was at Dara, yetthe device fordiverting the riverto an undergroundcourse was certainly being used in 530 (2. 2. I9). Procopius, therefore, only learntof the floodstorybetween writing
73 Joshua the Stylite(52) reportsthat the conduits at Edessa were similarly guarded with iron grilles and this was normal practice. At Dara holes to receive the iron bars are seen at both the northand south water gates (see pl. XII, 2). 74 At this point in his descriptionProcopius does not explain how the reservoirswere filled (cf. n. 82 below). The largest survivingcistern is placed on theslopes of the north-west hill (see fig.2 and Mango, op. cit. (n. II), pl. 37). J. B. Tavernier (Les six voyages I(I7I2), 233) noted seven or eight ruined churches. He saw two great cisterns to the north of the northchurch and a crypt,perhaps a cistern, beneath. 75 Procopius also wrongly attributesto Justinian
ofthe Romanemperor'(2. foresight
2. I9-2I).
Egeria's Travels (I 971), 284-7) and Antioch (2. I0. 15-I8, but see Downey, op. cit. (n. I2), 371 ff.). In an early photographof the Iron Gates at Antioch at least three structural phases are apparent: R. Dussaud, P. Deschamps and H. Seyrig, La Syrie
76 Resaina has great springs, one of the principal sources of the Habur. The chasm which Procopius describes is most probably a natural swallow-hole, so some undergroundwater systemconnectingDara and Resaina is possible, although it is most unlikely that large objects were washed down it only to resurfacelater elsewhere. Unfortunately, Crow was unable to investigatethe water system in any detail. 77 Stein, op. cit. (n. 6), 7I7.
et medievale, antique illustree (I93I).
thewaterworks ofEdessa(2.
7. 2-IO,
cf.J.Wilkinson,
I58
the eighthbook of the Wars and the compositionof the Buildings. To distortthe factsby the device to a fortuituos discoveryin the reign of Justiniansuited his paneattributing gyricalpurpose. we combine the two Procopian passages just discussed, the underground If, therefore, of the cityin 507/8 riverpit musthave been discoveredsome timebetweenthe construction and the firstknown use of the diversionscheme against the Persians in 530. Although Procopius does say that the undergroundriverwas discoveredand the pit fashionedsome wroNAN$ time after(xpovcw 6JC'TrpOv) Anastasius built the city (Wars 8. 7. 9), this statement does not rule out a date in the reign of Anastasius. Indeed, such a date would appear to be confirmed by anotherbriefdescription of Dara fromlate in the reignof Anastasius. In the work of a certainTheodosius, entitledde situ terraesanctae, which formspart of the popular traditionof pilgrim accounts and handbooks of the Holy Land in the sixth century,we find: Anastasius builta citycalledDara. It has a length theemperor In Mesopotamia of3 mileson whentheycame to plunder accountof the factthatit was wherethe Persians, the emperor's are notfound at all exceptthere. The river madea ditchbecausewaters province, emerges in backunderground so thatthisentire and at theend ofthecity thecity river plunges is enclosed wall(de situterrae bythe sanctae, 29 (CCL i75, I24)). one thingis certain: here Despite the looseness of this possibly lacunose description, notice of Dara that explains how the riverthat flowsinto the citygoes is a pre-Justinianic underground just inside the southernwall.78 The versionin Procopius' Buildingsis therefore a calculated misrepresentation.Moreover, the very fact that there survive from the originalAnastasian wall conduits in the south water gate with sockets for a double metal grill and sluice gate to shut the conduits off79 suggests that the functionthey served of stemmingthe river's flowformedpart of the Anastasian design. Procopius followshis account of how the riverwas divertedwith a long description of how the emperor and his advisers arrivedat a solution to controlthe floodwatersand The Alexandrian master-builder preventfurther damage (2. 3. I-I5). Chryses,who was apparentlyresponsible for much of Justinian'swork in Mesopotamia, heard about the damage caused by the forceof the riverat Dara and in a dream was advised of a plan to dam up the riverin the gorge beforeit reached the city (2. 3. 2-6), while Justinianwas inspired in a dream with the very same plan but, Procopius assures us, Justinian'splan prevailed (2. 3. I4). The scheme involved the building of a barrage with sluices at the feetin frontof the outer wall.80 point where the riveremergedfromits gorge some forty In addition,the northgate which had been flooded was blocked up with verylarge stones inundation (2. 3. 22-3). and moved to a higher position up the hill, safe from further There is no evidence available to evaluate Procopius' account here, but his highlyliterary lead one to doubt and the co-incidenceof the schemes of Chrysesand Justinian description to the emperor. The dam was probably part of the Anastasian design the role attributed as well. Procopius goes on nextto explainhow the emperorprovidedthe citywithboth cisterns in the original it had neither and an aqueduct, since previously (2. 3. 24). The lack ofcisterns that the Anastasian citywas well provided with plan is contradicted by specifictestimony cisterns(Jo. Mal. 399. I7; Zach. Mit., HE 7. 6), presumablythose whose impressive as the builderof the first remainsstillstand at Dara. Likewise, to claim creditforJustinian aqueduct in the city is patentlymisleadingsince we know the aqueduct too was built by thatJustinian was responsible Anastasius(Zach. Mit., HE 7. 6).81 It is possible,nonetheless, extraones. the cisternsand aqueduct or forconstructing forrepairing
78 The de situ terraesanctae is dated to the reign of Anastasius because the latest constructions mentioned in the work are Anastasian (cf. A. Heisenberg Grabeskirche und Apostelkirche i (I908), I o6-I o). 7 See n. 73 above. 80 From fig. 2, the plan of C. Preusser, fig. I2, the distance fromthe inner wall to the mouth of the gorge is at least 50 m. Even though the outer wall
appears to swingout as it reachesthenorthwatergate, Procopius has underestimatedthis distance. 81 In his first attemptto describe the water supply Procopius says that it came direct from the river (2. 2. 3-6). Since the survivingcisternsare located on high ground, this would have been impossible withoutsophisticatedhydraulicequipment. He later correctsthis statement(2. 3. 24), but still attributes the work to Justinian.
PROCOPIUS
AND DARA
I59
If Procopius' descriptionof Dara is not to be trustedthen we are also entitledto be suspicious about his account of Justinian'swork at other sites where he claims a general at Dara. Armenian Theodosiopolis (Erzerum) is the with the reconstruction similarity most conspicuous example in this category. According to Procopius the curtain wall of and not thickenough to feethigh by the time of Justinian, Theodosiopolis was only thirty be built on; nor did the citypossess any outworksor moat (3. 6-8). What Justiniandid was to have a deep ditch dug around the city(3. 6. 9) and, as at Dara, he constructedon the walls an extra galleryand outworks; he also narrowedthe embrasuresin the towers and modifiedthe towers so that each became a separate fortress (3. 6. i2). This account of Procopius is extremelysuspect when considered alongside a local descriptionby the ArmenianhistorianMoses of Chorene of the foundationof Theodosiopolis in 42I by the Roman generalAnatolius.82 First of all, Anatolius surroundedthe site with a large ditch, forthewalls werelaid out. The ramparts includedhigh-pointed and deep broad foundations towers ' like the prows of ships ' on the northernand southern sides, while the eastern and westernsides had large rounded towers. In the elevated centreof the city a square, was laid out. The city was built on thermalsprings and water known as an Augusteon, was conducted around it by undergroundchannels. Finally, it was filledwith both arms and troops. As with Dara in the age of Justinian, Theodosiopolis which was reinforced and renamed by Anastasius can scarelyhave been in the conditionclaimed by Procopius. the studyof an individual The aim of thisstudyhas been to shed light,mainlythrough on the eastern site, on the value of Procopius' Buildingsas a record of the fortifications frontier in the reignofJustinian.Both hereand in othercases whereit is possible to evaluate its evidence by comparisonwith other sources, it has frequently been found to be exaggerated,misleadingand sometimescontradictory.Recent researchpoints ratherto overall of Roman military forcesin Syria duringJustinian's runningdown and withdrawal reign; 83 a close analysisof the Buildingsitselfcan be seen to implythat Justinianic similarly, work at Dara and elsewhere in the east contributedlittle more than a slight modification and repair of the Anastasian defences. Throughoutthe Buildings,Procopius was influencedboth by his panegyricalpurpose and by literaryconsiderations. In this artfulpresentation, Justinianacquires the credit for construction which belonged more properlyto Anastasius and JustinI, and what was is made to seem like major new building. The detailed analysis in factminorrefurbishing of Dara is enough to suggestwhat is coming to be more widely of Procopius' presentation recognizedon generalgroundstoo, namelythatthe Buildingsis a sophisticatedtextworthy in its own right,and only to be used with the utmostcaution as of far closer consideration 84 on Roman forts of factualinformation and citiesin the sixthcentury. a potentialrepository Sydney The National Trust,Hadrian's Wall Estate
82 Hist. Arm. 3. 59 (Langlois ii, i66-7), cf. F. H. Weissbach, RE 5A (I934), I924-6 S.V. 'Theodosioat towers pentagonal For the surviving pOliS, 2 '. Theodosiopolis, see R. H. Unal, Les monuments anciensde la ville d'Erzerumet de sa rigion islamiques (Bibl. arch. et hist. de l'Inst. fran,ais d'Archeol. d'Istanbul 22) (I968), I6, fig.3.
Liebeschuetz, op. cit. (n. 28). We should like especially to thank Averil Cameron forbringingus together forthisprojectand forher subsequent advice, as well as the Leverhulme Trust for making possible Crow's visit to Dara.
83 84
JRSvol.
LXXIII (I983)
PLATE XI
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~ =_~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
__ ! . ....... 9.?* .....
w ^ .n e < @ = .,. : w>. WW_
~~~~~~~~~~~~~~~~~~~~~~~~~~~~........
: ii ; ............... s !:z iiS l 11 _D will N lulllfiiliii
2I WAwC:{,c_..
__
* | _--*-i
_ | 3W-U-i:
fiiZ
--]-s
fi
gs g| IfiS . -lr| S-
5A_~~~~~~~~1 _ais ^4
x"w
iS
'w'P'f~~~~t
:m's'S ^;
(2).4r_i
_l|
JRS vol. LXXIII
( I 983)
PLATE
XI I
_^
.
!l .
X
_ -. < v.
,=
r
............
_S
_
o
__
"zE
n;
w-XiS
_ E____
_E
a_|> ffi9f
swRRe_
.4'.
--Xd
--
,.
E_ w
n,
..o.'
.^s
sheag v:-4 ,6 I ,E ___ =, 9 AV
t_g__
(I)
A-
LL.*v2
-ts
C:
.so?$'
rqP l!
b.
t.aYtk
__
i_D __ffi_
_,;=_
bN
53
_=_
_
X
l_
_ ............................. _e
R ,,r xW
b
_
b
iR
__
s
_
....
... _
t_
............
'- 3-q*
(2)
S|
(I)
NORTH GATE FROM THE SOUTH-EAST SHOWING TYPE B MASONRY. W'EST TURRET FROM THE NORTH.
(2) G. Crow. Photo7. G. Crow. 3t. Photo
SOUTH WATER GATE AND

Procopius and Dara

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Procopius and Dara

Uploaded by

Copyright:

Available Formats

Procopius and Dara Author(s): Brian Croke and James Crow Source: The Journal of Roman Studies, Vol.

PROCOPIUS AND DARA *

PiAlhornin 0. Veh's German translation of the

BRIAN CROKE AND JAMES CROW

ofMethodin theDe Aedificiis', Byz. I4

PROCOPIUS AND DARA

Resaina Batnae *ane Rsia . Carrhae

BRIAN CROKE AND JAMES CROW

Karnapp, op. cit.(n. IO),

BRIAN CROKE AND JAMES CROW

with references to the

Claude, Die byzantinische Stadt im 6.

(~bridge ;water gate

SKETCH PLAN OF DARA AFTER PREUSSER, I9II, REDRAWN BY B. V. WILLIAMS.

(I917), and CentralKurdistan Mesopotamia

BRIAN CROKE AND JAMES CROW

43 M. Restle, 'Viran~ehir-Kalek6y, ein befestiger Platz in Kappadokien', J7ahrb.d. 6st. Byz. 24

E. Honigmann,Die Ostgrenzedes byzantinischen

PROCOPIUS AND DARA

BRIAN CROKE AND JAMES CROW

and thiswas levelled(2. operations

tionsat Dibsi Faraj, northern Syria,1972-1974:

(2. 9. i) and Hemerium (2. 9. io), where the walls

BRIAN CROKE AND JAMES CROW

Greek Fortifications (I97I),

Deep rock-cut ditches were a feature of late

Amida, Gabriel (op. cit. (n. cussed by Crow,op. cit.(n.

23), 96-I I3)

Antioch (Rey,op. cit. (n. 63), I92, figs.50,

above), 17-18, figs.

BRIAN CROKE AND JAMES CROW

PROCOPIUS AND DARA

ofthe Romanemperor'(2. foresight

et medievale, antique illustree (I93I).

BRIAN CROKE AND JAMES CROW

JRS vol. LXXIII

sheag v:-4 ,6 I ,E ___ =, 9 AV

(2) G. Crow. Photo7. G. Crow. 3t. Photo

SOUTH WATER GATE AND

You might also like