Rome and Persia at War and Peace: Competition and Contact in the Near East, 193 to 363 AD 1472418174, 9781472418173

Table of contents :
Blank Page
Blank Page
Blank Page
Blank Page

Citation preview

Rome and Persia at War

This book focuses on conflict, diplomacy and religion as factors in the relationship between Rome and Sasanian Persia in the third and fourth centuries AD. During this period, military conflict between Rome and Sasanian Persia was at a level and depth not seen mostly during the Parthian period. At the same time, contact between the two empires increased markedly and contributed in part to an increased level of conflict. Edwell examines both war and peace – diplomacy, trade and religious contact – as the means through which these two powers competed, and by which they sought to gain, maintain and develop control of territories and peoples who were the source of dispute between the two empires. The volume also analyses internal factors in both empires that influenced conflict and competition between them, while the roles of regional powers such as the Armenians, Palmyrenes and Arabs in conflict and contact between the two “super powers” receive special attention. Using a broad array of sources, this book gives special attention to the numismatic evidence as it has tended to be overshadowed in modern studies by the literary and epigraphic sources. This is the first monograph in English to undertake an in-depth and critical analysis of competition and contact between Rome and the early Sasanians in the Near East in the third and fourth centuries AD using literary, archaeological, numismatic and epigraphic evidence, and one which includes the complete range of mechanisms by which the two powers competed. It is an invaluable study for anyone working on Rome, Persia and the wider Near East in Late Antiquity. Peter Edwell is Senior Lecturer in the Department of Ancient History at Macquarie University in Sydney, Australia. He researches and teaches on the relationship between the Roman and Sasanian Persian Empires and in the area of Late Antiquity more broadly. He is currently part of the project Crises of Leadership in the Eastern Roman Empire funded by the Australian Research Council. Dr Edwell is the author of Between Rome and Persia, published by Routledge and has written numerous book chapters and articles focussing on the relationship between Rome and its powerful eastern imperial neighbour.

Rome and Persia at War Imperial Competition and Contact, 193–363 CE

Peter Edwell

First published 2021 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 52 Vanderbilt Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2021 Peter Edwell The right of Peter Edwell to be identified as author of this work has been asserted by him in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record has been requested for this book ISBN: 978-1-4724-1817-3 (hbk) ISBN: 978-1-315-60702-3 (ebk) Typeset in Sabon by codeMantra

For Penny (and Scarlet & Orlando)

Important sites in the Roman and Persian worlds of the third and fourth century

Contents

Preface Acknowledgements

xi xiii

3 Conflict and diplomacy between Rome and Persia from Ardashir to Philip I

57

4 Persian triumph, Roman defeat

86

5 The last years of the reign of Shapur I to the Persian invasion of Carus

111

6 The relationship between Rome and Persia during the reigns of Diocletian, Bahram II and Narseh

132

7 Rome and Persia during the reigns of Constantine and Shapur II

157

9 The Persian invasion of the emperor Julian and its aftermath

212

Bibliography Index

255 275

Preface

As an undergraduate student at Macquarie University in the 1990s, I was fortunate to participate in a seminar unit focussing on the eastern frontier of the Roman Empire. In one of the sessions, the convenor of the unit, Professor Sam Lieu, gave a presentation on the use of aerial photography to help better understand the Roman military presence in the Middle East. When a copy of David Kennedy and Derrick Riley’s book, Rome’s Desert frontier from the Air, was passed around to students in the seminar, the remarkable photograph of Qasr Bshir (Castra Praetorii Mobeni) that adorns the front cover of the book leapt out at me. My fascination with the Roman and Persian presence in the Middle East was born that day and has continued to grow stronger ever since. The seminar taught by Professor Lieu is now one that I convene myself and have done for a number of years. The source book compiled and edited by Professor Lieu and Michael Dodgeon, The Roman Eastern Frontier and the Persian Wars, has always been indispensable to teaching the unit and, as many have noted over the years, makes a major contribution to our knowledge of events and developments during a period of intense competition in the third and fourth centuries AD between two of the ancient world’s most powerful empires. In the process of teaching the subject, I have added an increasing amount of material relevant to the Sasanian Persian side of the story. Material of a numismatic, archaeological and cartographic nature also plays an important role in developing as detailed an understanding as possible of conflict and contact between the two powers. The impact of conflict on the wide range of people inhabiting a considerably diverse landscape from the Caucasus in the north to the deserts of Arabia in the south is another key focus. This book is an attempt to draw all of this material together into a coherent analysis principally focussed on conflict and diplomacy between the Roman Empire and the Sasanian Persian Empire during periods of considerable transformation for both. The Sasanian overthrow of the Parthians in the 220s and the reigns of the early Sasanian Shahanshahs have been regarded in epoch-making terms in the history of Iran, and the transformation of the Roman Empire from the period of the Severans to the dynasty

xii Preface of Constantine has long been regarded as an extraordinary period of crisis, transition and transformation. Conflict and diplomacy between Rome and Persia during this period were both contributors to and products of significant developments in both empires during this period. Thus, the book attempts to analyse military and diplomatic activity in the broader context of the considerable changes that took place in both empires in political, economic, military, religious and demographic terms. It is important to note that I am principally a Roman historian but have done my best to deal with material pertinent to Sasanian Persia. It is unfortunate that the available material for the analysis of the relationship between the two powers results in a perspective skewed more to the Roman side of the story. The increasing depth of scholarship on all aspects of Sasanian Persian history and culture will undoubtedly continue to rectify the balance, and I hope that the coverage in this book does justice in some way to the Persian perspective on military and diplomatic conflicts between the two powers.

Acknowledgements

The successful completion of a book requires the support and assistance of many. I would especially like to acknowledge Dr Ross Burns for his generous assistance in developing and supplying the maps appearing in this book. I would hasten to add that any errors or omissions are entirely my own responsibility. Ross also supplied photographs for which I am very grateful. Effy Alexakis photographed most of the coins published in this book, and I am profoundly thankful to her. Many thanks also to a number of friends and colleagues I have discussed ideas with over the years. These include Dr Mark Hebblewhite, Associate Professor Lea Beness, Associate Professor Tom Hillard, Jeff Tillitzki, Dr Meaghan McEvoy, Dr Caillan Davenport, Professor Ray Laurence, Professor Sam Lieu, Professor Alanna Nobbs, Professor Bronwen Neil, Dr Eva Anagnostou-Laoutides and Dr the Ross Burns. I would also like to acknowledge my ongoing appreciation for the support and encouragement of my career provided by the late Dr Bill Gale and Mrs Janet Gale.

1

Introduction

At another time he was taken up by someone to a place commanding a wide view, and as he gazed down from there upon all the land and all the sea, he laid his fingers on them as one might on an instrument capable of playing all modes, and they all sang together.1

Thus, Septimius Severus dreamt of a new golden imperial age on attaining ultimate political and military power in the Roman Empire in April 193. One of his early orders of business following the formalities of establishing power in Rome and putting down rival claimants was the punishment of the Parthians for their support of Pescennius Niger during the complex civil wars that wracked the Roman Empire in the wake of the death of Commodus the previous year. For the third time in 80 years, a Roman army would sweep down the Euphrates and Tigris Rivers, sack the Parthian capital and add further to the difficulties of the Arsacid Parthian dynasty in whose affairs Roman emperors had meddled for over 250 years. In the provincial reorganisation that followed, the armies of Rome formally garrisoned the upper Tigris; the eastern territorial expansion of the previous decades seemingly set to continue. Seventy years later, however, three Roman emperors would be depicted on rock reliefs at Bishapur in Iran in various states of subjection to an Iranian ruler whose power was likely inconceivable to Septimius Severus at the time of his elevation to the principate. One emperor begs before the Sasanian Shahanshah, Shapur I, another lies lifeless and trampled under his stunningly caparisoned horse, while the third is Shapur’s personal captive, a unique humiliation in the entirety of Roman imperial history. Shapur’s triumphal inscription at Naqsh-i Rustam heralded him as “the Mazdaworshipping divine Shapur, King of Kings of Aryans and non-Aryans, of the race of the gods….” His empire stretched from Peshawar in modern Pakistan to Armenia and the Caucasus and parts of the Arabian Peninsula. The empire of Iran under the rule of the Sasanian Shahanshahs came of age under Shapur and would represent a consistent challenge to Rome’s

2 Introduction presence in the east in ways that the Parthians had only been able to manage on occasion. A little over a century after Septimius Severus metaphorically gazed on a world that was his before sacking the Parthian capital, and only 40 years after Shapur’s triumphant imperial claims were made, an envoy from the recently defeated Sasanian ruler, Narseh, began negotiations with the victorious Roman Caesar Galerius: It is clear to the race of men that the Roman and Persian Empires are, as it were, two lamps; as with eyes, each one should be adorned by the brightness of the other and not forever be angry seeking the destruction of each other. 2 Narseh had been so soundly defeated by Galerius that he fled for his life from the battlefield and his wife, harem and baggage train were still in Roman possession. While clearly in a position of subservience, the Persian envoy communicated the reality of the situation perhaps more than he knew. An enormous commitment to the physical defence of the eastern provinces and reform of the army already begun under Diocletian, which continued under his Constantinian successors, was all the proof needed of Rome’s acceptance that the Sasanian Persian dynasty and the vast empire it controlled were here to stay. If proof was needed of how much had changed since the days of Rome’s imperial forays into the Parthian world and Shapur I’s invasions of the mid-third century, the late 350s and early 360s emphasised these changes with stark clarity. When Constantius II broke down in tears before the ruined walls of Amida in 360, they might just have been tears of joy given how effective the north Mesopotamian defensive structure was in halting the Sasanian advance any further. Similarly, when Julian’s army began to struggle a few years later in the complex maze of irrigation canals in southern Mesopotamia, the extensive upgrading of irrigation and corresponding population growth that accompanied it signalled that a Roman invasion and attack on Seleucia-Ctesiphon was a much more challenging prospect than it was in the second century. In the peace negotiations that followed in the wake of Julian’s death in 363, Persian requests were entirely territorial and practical. Rhetorical claims made in the 230s, and even in the mid350s, to a right by inheritance to the territory of the Achaemenids stretching all the way to the coast of Asia Minor were nowhere to be seen.

Aims of the book This book is principally focussed on the military and political (mostly diplomatic) relationship between the Roman Empire and the Sasanian Persian Empire from the emergence of Ardashir as Shahanshah to the last years of the reign of Shapur II. It not only analyses these closely related and

Introduction  3 intertwined factors but also aims to frame them in the context of other developments and factors internal to both empires. Relevant background to war, politics and diplomacy between the Parthian and Roman Empires in the centuries beforehand is also analysed. There are a number of modern studies of the topic, a majority of them in dedicated chapters in collected volumes and as parts of historical surveys with specific focusses on the Roman or Sasanian Persian Empires.3 Karin Mosig-Walburg’s Römer und Perser vom 3. Jh. bis zum Jahr 363 n.Chr is an exceptionally in-depth study of the textual sources on the topic. It is unfortunate that it is not a more widely accessible publication. Matthew Canepa’s important and distinctive study of representations of kingship in Rome and Sasanian Iran covers the third and fourth centuries in considerable detail and is an informative study for this book.4 Specific source publications and accompanying commentaries, including David Potter’s critical edition and commentary on the Thirteenth Sibylline Oracle, contain some detailed analysis of aspects of the military and political relationship between the two powers in the concentrated timeframe to which the text belongs.5 Potter also provides a broader historical analysis of events in the third century. In a number of editions and commentaries of the SKZ inscription, the wars between Persia and Rome from 243 to 260 also receive some detailed analysis.6 The sources on the topic are complex, ranging from texts in a variety of different ancient languages through to evidence in divergent mediums such as epigraphy, archaeology and numismatics. The source extract book by M.H. Dodgeon and S.N.C. Lieu, The Roman Eastern Frontier and the Persian Wars, is an important contribution to writing a book such as this one.7 Since its first appearance in the 1990s, Dodgeon and Lieu has been recognised as indispensable to our knowledge of the impact of Sasanian Persia on the Roman provincial and imperial presence in Armenia, Mesopotamia, Syria and Arabia from the 220s to the 360s. The collection of mostly literary source extracts comes from a range of Greek, Latin, Persian, Syriac, Armenian and Arabic texts written in different genres in many different chronological contexts. It is especially valuable in making relevant extracts from texts available that are otherwise difficult to access. The collection of mostly literary source extracts in Rome and Persia in Late Antiquity: Neighbours and Rivals by B. Dignas and E. Winter is also an important contributor to the topic.8 This source book covers the entirety of the Sasanian period and attempts to present a more “eastern” perspective, thus paying more attention to the “non-Roman” sources. It is structured thematically compared with the chronological structure of Dodgeon and Lieu and provides comparatively more analysis of the relevant texts and historical developments to which they relate.

Sources and historiography The textual sources for the military and political relationship between Rome and Sasanian Persia in the third and fourth centuries come from a

4 Introduction range of different genres, languages and timeframes. It is unfortunate but inescapable that the bulk of these surviving texts are in Latin and Greek and skew the narrative towards a Roman/Byzantine perspective on events. For the third century and the first half of the fourth century, no contemporary narrative historical text focussing on Roman history survives and the texts that do survive from this period exist in a variety of genres, which presents some interpretative challenges. There are a number of Armenian textual sources relevant to the topic, but they originated in the fifth century and later, and contain troubling chronological errors. For the third century, there are, thankfully, some lengthy Persian inscriptions that are closely contemporary with the events they refer to, notably the SKZ, KKZ and the Paikuli Inscription. Arabic works produced in the early Abbasid period (750–850), including Tabari, Dinawari, Ya`qubi, Mas`udi and Hamza Isfahani, preserve now lost Sasanian textual traditions relevant to the topic. Versions of the Khwaday-namag, or Book of Lords, a late Sasanian text, appear to have been the principal source of these writers for Sasanian Persian history, although it is important not to overemphasise their complete reliance on such a source.9 The Khwaday-namag appears to have focussed on legends and romantic tales in its renderings of historical events and it was perhaps Tabari who did his best to cut through this approach to produce what might be termed “sober history.”10 On a number of occasions throughout the section of his history relevant to the topic of this book (I.813–845), he noted differing traditions and accounts of events that were available to him. On the topic of the conflicts between Rome and the Sasanians in the third and fourth centuries, Tabari provides the most detail. Dinawari, Ya`qubi, Mas`udi and Hamza Isfahani are much briefer and the level of detail in Tabari’s account is more significant for the third century than it is for the fourth.11 Similarly, the Shahnama of Ferdowsi, an Iranian epic poem composed in the late tenth and early eleventh centuries, drew on a version of the Khwaday-namag for its coverage of much of the Sasanian period. It provides only brief coverage of the conflict between Sasanian Persia and Rome in the third and fourth centuries.12

Roman/Byzantine historiography of the third and fourth centuries Specific discussion of historiographical issues is provided at relevant points throughout the book; however, some discussion of these issues in a broader sense is provided here. It has been much lamented in scholarship that narrative historical texts of the third century and first half of the fourth century from the Roman Empire have not survived. After the conclusion of Cassius Dio’s Roman History ca. 230, the generally maligned Herodian continues for the best part of another decade, but it is not until 353 that a broadly contemporary, chronological history of key events in the Roman world is available in the form of Ammianus Marcellinus’ Res Gestae. Fragments and

Introduction  5 references to works of Roman history written in the third century or soon after show that they were composed at the time and were still extant later in antiquity. These include Nicostratus of Trapezus who covered the period 244–260, Eusebius who wrote a history from Octavian to the death of Carus in 283 and Porphyry who composed a universal history of both Greece and Rome beginning with the Trojan Wars and ending with the reign of Claudius Gothicus (268–270).13 Praxagoras of Athens was a prolific composer of historical works during the reign of Constantine, some focussing on the earlier history of Athens and Macedonia, and one that dealt with the reign of Constantine from 306 to the founding of Constantinople in 324.14 Another work of history has been theorised in the Latin tradition and is known in modern scholarship as the Kaisergeschichte. Possibly written during the reign of Constantine, such a work is only thought to have existed due to perceived similarities in the reporting of events by some later Latin writers who appear to have been relying on an un-named and now lost intermediate historical source for events in the third and early fourth centuries.15 Works by Dexippus of Athens, now mostly lost, are viewed by many as important contemporary textual sources for third-century history up to 270. Born in the early years of the third century and writing mostly from ca. 250 to 270, Dexippus’ reputation owes much to comments by the Patriarch of Constantinople, Photius, in his ninth-century summary of Dexippus’ then extant works in the Bibliotheca: “His style is free from redundancies, massive, and dignified; he might be called a second Thucydides, although he writes more clearly.”16 Among other works, Dexippus wrote a historical chronicle covering a thousand-year period up to the death of Claudius Gothicus and a work known as the Scythica, which covered the history of Rome’s wars with the Goths in the third century. Both works covered events and a timeframe which Dexippus witnessed and were used in later extant texts, one notable example being the Historia Augusta, thought only to be reliable for events in the third century when it relies on Dexippus. The New History of Zosimus, thought to have been composed towards the end of the fifth century and/or beginning of the sixth century, relied heavily on the no longer extant history of Eunapius of Sardis for events of the latter part of the third century up to the early fifth century.17 Writing in the latter part of the fourth century and early years of the fifth century, Eunapius’ history was designed to continue from where Dexippus finished in 270.

Other genres Some surviving Latin and Greek texts written in genres other than history or historical chronicle, and from which details on Rome’s military and political relationship with Persia in the third and fourth centuries can be gleaned, fall within unusual or highly rhetorical genre classifications. The contemporary regional perspective on conflict between Rome and Persia in the 240s and 250s provided in the Thirteenth Sibylline Oracle is one such

6 Introduction example. It is a very useful text but needs careful reading to be of historical use. Some of the most important contemporary Greek and Latin texts from which information about the relationship between Rome and Persia in the third and fourth centuries can be extracted survive in the genre of panegyric. The panegyrics and orations of Libanius are among the most important for contemporary events from the 340s to the 360s. Oration LIX delivered in Antioch ca. 348 is an example worth noting because it provides information on Constantius II’s wars with Persia prior to the major Persian invasions of the late 350s. Julian’s panegyrics on Constantius II written not long after Libanius’ Oration LIX are of similar importance. Libanius’ Funerary Oration and Lament over Julian demonstrate with clarity the strong rhetoric employed by pagans to defend Julian in the wake of his death in Persia. Conversely, St Ephrem’s Hymni Contra Julianum and Gregory Nazianzenus’ Orations (Invectives) are strong, rhetorically driven condemnations of Julian and the Persian campaign. In all of these examples, whether highly praising panegyrics or strongly condemning invectives, they are contemporary with the events and when read carefully provide information on events otherwise not available. The tendency to lament the lack of surviving narrative historical texts from the mid-third to mid-fourth centuries has overshadowed the comparative richness of other categories of sources relevant to the relationship between Rome and Persia, especially for the third century.18 The third- century Persian epigraphic texts have already been noted and it is important also to emphasise the richness of Sasanian rock reliefs as evidence for historical events in the third century, especially in relation to military conflicts between Rome and Persia. Similarly, the reliefs of the Arch of Galerius in Thessaloniki, which depict Galerius’ victory over the Sasanian Shahanshah Narseh in 297/8, are of great importance to our understanding of how this victory was retailed by the tetrarchs to the Roman population. Inscriptions from this period are also useful and are valuable as indicators of victories claimed by Roman emperors over their Persian counterparts. Coins produced at a number of Roman imperial and provincial mints together with those of the Sasanian mints provide another contemporary perspective, one that has received comparatively little prior attention. Archaeology, such as that for the Sasanian siege at Dura Europos, provides graphic and unparalleled insight into the grim realities of military conflict between Rome and Persia. The extensive remains of tetrarchic and later efforts to fortify Syria, Arabia and Mesopotamia provide a richer contemporary testimony on that endeavour than any historical text could.

The impact of Christianity on ancient historiography The imperial patronage of Christianity begun under Constantine and followed by all but one of his successors skews the nature of the surviving Roman textual material referring to events of the fourth century and earlier

Introduction  7 towards a Christian perspective and this has implications for the analysis of the relationship between Rome and Persia. This is noticeable, for example, in the way events in the reigns of Diocletian as a persecutor and Julian as an apostate are reported. Claims made by contemporary and later Christian writers regarding the role of Christians and Christianity in military, political and diplomatic engagements between Rome and Persia also need to be weighed carefully. While we should not doubt that Christianity became a feature of the defence of northern Mesopotamian cities and fortifications such as Nisibis, and that the religion was a factor in diplomacy during the reigns of Constantine and his successors, it is important not to overemphasise its significance given the tendency of Christian writers to place Christianity and the church as close to the centre of history as possible. Similarly, claims in Christian texts of persecution at the behest of Zoroastrian leaders and their imperial counterparts in Persia need to be read with care. This is especially the case in references in Syriac texts to Christians in Persia looking to Constantine and his imperial Christian successors as potential liberators. As a “pagan” ruler and apostate, Julian’s brief but tumultuous reign became a focal point of debates and discussions between Christians and pagans for decades afterwards. This had an impact on historiography generally but also more specifically in the ways ancient texts discussed the military and political relationship between Rome and Persia from Constantine onwards. Pagan historians and panegyrists attempted to lay ultimate blame on Constantine for starting a war with the Persians towards the end of his reign that would end so disastrously with Julian’s death a quarter of a century later. Some also attempted to blame Constantius II for an overly defensive policy approach that did not adequately deal with the Persians and gave them confidence to mount the attacks they did in the late 350s. Other pagan writers sought to lay the blame for territorial losses stemming principally from the difficult position the invading Roman army faced at the time of Julian’s death on his Christian successor Jovian. Christians had an obvious and straightforward means of attacking Julian in relation to the Persian campaign. As an apostate, Julian had raged against God and was driven on by demons in his campaign against the Persians. For someone such as St Ephrem, who it could be argued was forced out of Nisibis as a direct consequence of Julian’s loss in Persia, the contrast with earlier times was stark. Nisibis had withstood three Persian sieges in which Christians were depicted as central to the defence of the city. Now, the fortress had been ceded to the Persians as part of one of Rome’s most humiliating territorial losses, which Jovian had little choice but to agree to. The fact that Jovian attempted to retail events in Persia as an imperial success, however, gave the pagans some ground to work with.

The Res Gestae of Ammianus Marcellinus The surviving books of Ammianus Marcellinus’ Res Gestae provide a select narrative of Roman history from 353 to 378 and include an invaluable

8 Introduction eyewitness account of events directly relevant to the relationship between Rome and Persia in the late 350s and early 360s. The narrative of Ammianus on the Persian invasions of the late 350s is unrivalled in detail for any Sasanian Persian invasion of Roman territory. During this period, Ammianus was privy to high-level information on strategy and events due to his attachment to the staff of a senior military commander in the eastern provinces, Ursicinus. While he was not privy to this level of information on Julian’s campaign in 363, Ammianus provides a detailed narrative of the invasion from its beginnings in Antioch to its disastrous conclusion north of Ctesiphon. Supplemented by Zosimus, who had access to another eyewitness in Julian’s physician, Oribasius, together with references to the invasion by other contemporaries such as Libanius, our knowledge of Julian’s invasion of Persia is by far the most detailed of any Roman invasion of the Persian or Parthian Empires. Ammianus’ text is, of course, a rhetorically motivated piece of work, which at times inflates its author’s own role in events. In coverage of the Persian invasions of the late 350s, one of its central points is the defence of Ursicinus against the blame attributed to him by Constantius II for the fall of the fortress of Amida in 359. Ammianus’ coverage of the entirety of Constantius’ reign from 353 onwards is influenced by the fate of Ursicinus. Ammianus’ account of Constantius’ approach to war with Persia was also designed to contrast with Julian’s invasion of Persia. As a means of defending Julian from the attacks of Christian writers after his death, Ammianus sought to apportion blame for the disaster of Julian’s Persian campaign to Constantine for starting the war and Julian’s Christian successor Jovian for accepting a shameful peace.

Non-Roman/Byzantine sources on the fourth century While the third century is comparatively rich in Sasanian epigraphy and rock reliefs, very little of a similar nature survives from the fourth century. A rock relief from Taq-e Bostan depicting a bearded Roman, almost certainly Julian, lying face down beneath a scene depicting the investiture of Ardashir II is the only extant contemporary Persian reference to Persia’s conflict with Rome in the fourth century. Unfortunately, as noted earlier, Tabari’s History is much more limited on Persia’s conflict with Rome in the fourth century than it is for the third century. The Armenian texts are similarly more limited and contain some problematic chronological inconsistencies. Sasanian coins continued their general uniformity of design from the third century to the fourth century and from the post-reform era of Roman coinage under Diocletian onwards, Roman coin images and legends are less specific than they were in the third century in relation to campaign preparations and claimed outcomes.

Introduction  9

Geography and topography of territorial competition between Rome and Persia The kingdoms of the Caucasus – Iberia, Colchis and Albania The geography and topography of the territory across which military and political competition between Rome and Persia took place was complex and variable. The nature of this geography and topography is important in understanding how, when and why the two powers fought wars and competed with each other for political influence and territorial control during the third and fourth centuries. It is also important to understanding why the two powers reached limits of the geographical power they were able to exercise, often reflected in the agreements they made with each other. The three kingdoms or principalities that lay between the Caucasus Mountains and the kingdom of Armenia were Colchis, Iberia and Albania. Following Pompey’s subjugation of Colchis in 65 BC, which comprised the western section of modern Georgia and parts of north-eastern Turkey, the kingdom was mostly attached to Pontus in the imperial period.19 Colchis possessed ports on the east coast of the Black Sea, which made it of potential strategic importance and also in relation to trade. Strabo observed in the first century AD that Iberia to the east of Colchis was mostly settled with cities and farmsteads and that its capital, Mtskheta, was a sophisticated and developed city. 20 Winter still witnesses heavy snowfall at times and snow melt in spring makes movement through the valleys of Colchis and Iberia more challenging. In Iberia, as well as Colchis, heavily guarded passes in the Caucasus Mountains channelled the movement of people on a north-south axis. This was an important factor in controlling the movement of tribal pastoralists from Ukraine and southern Russia who without effective monitoring had the capacity to sweep further south towards Armenia and Roman and Persian territories. Concerns over the potential devastation that tribal groups such as the Alans and Sarmatians could wreak in Asia Minor, Media and further south led at times to mutual, although grudging, agreements between Rome and Persia over guarding the Caucasus passes. On occasion, these tribal groups were even used as weapons. In the wake of Ardashir’s overthrow of the Parthian Arsacids, for example, the Armenian texts referred to a confederation of Armenians and Iberians who were accompanied by Alans and Huns who had been allowed to pass from southern Russia to join in an invasion of Persia. 21 To the east of Iberia lay the kingdom of Albania (modern Azerbaijan), which opens into flatter plains as its territory stretches to the Caspian Sea. An important geographical feature of Colchis, Iberia and Albania is the Kura river valley, which cuts in a west-east direction from the east of Colchis, through Iberia and Albania to the shores of the Caspian Sea. The Kura initially flows into Colchis from the Armenian highlands in the south

10 Introduction before turning east into Iberia. The direction of the river flow and relatively wide valley through which it flows created an important means of transport and movement in antiquity. This was the route Pompey took in 65 BC after defeating Mithridates in Armenia and marching into Iberia and Albania, where he almost reached the shores of the Caspian Sea. 22 Albania looked predominantly to the Iranian world and is mentioned on occasion in the ancient texts providing support to the Persian army or as part of the Sasanian Persian Empire. Iberia to its west, however, was an area of imperial contention, sometimes within the Roman orbit as a client-kingdom and at others under Persian influence or direct control.

Armenia Armenia’s geographical location and topography was of key importance to the military and political relationship between Rome and Persia and ensured that the kingdom was the focus of intense competition in both the Parthian and Sasanian periods. Armenia was a kingdom which comprised many mountains and fertile valleys and regularly suffered harsh winters. Areas of elevation in ancient Armenia reached more than 3,000 m in places and the mountainous topography of the kingdom was an important factor in Armenian politics. Chieftains and regional nobles in Armenia were able to wield considerable power from their bases in the mountains and valleys that criss-crossed the kingdom and this allowed them autonomy, creating ongoing challenges for a centralised Armenian state authority. 23 Armenia’s location and the loyalty of its king were especially strategic to both Rome and Persia. This led to serious conflicts between the two powers and sometimes mutual agreements over Armenia’s governance. The challenging nature of the topography of Armenia made direct control of the kingdom difficult to sustain for either Roman or Persian rulers when they made attempts to do so. Armenian rulers and nobles would sometimes play Roman and Persian rulers off against each other due to this factor and with factionalism often a strong element in Armenian politics, the involvement of Roman and Persian imperial powers in the kingdom became complex. The terrain of Armenia was not ideal for the movement of large armies across the kingdom, either infantry-based or cavalry-based, due to the mountains and valleys of the kingdom mostly running in a north-south direction. This did not stop the rulers of both Rome and Persia sending armies against each other via Armenia due to the proximity of the kingdom to important “enemy” territory, which for the Romans lay in Media and for the Persians lay in provincial Cappadocia and the northern parts of the province of Syria. Armenia was also strongly connected to the principalities of Parthian and then later Roman provincial territory in northern Mesopotamia. When Roman armies attacked Persian territory from Armenia, the return journey was a difficult one, especially if it was attempted late in the campaign season when low temperatures accompanied

Introduction  11 by snowfall caused serious difficulties. On a number of occasions, Roman and Parthian/Persian armies fought battles against each other in Armenia. The second-century Roman invasions of Parthia, for example, began with military engagements in Armenia and Galerius’ victory over Narseh in 297/8 was scored in Armenia. Among the most important urban centres in Armenia from the first century BC to the fourth century AD were Artaxata, Tigranocerta and Arsamosata. Artaxata was located in the central eastern portion of Armenia and situated in a strategically defensive location with river courses on three sides and a heavily defended palisade on the fourth. 24 The city was the Armenian capital from its foundation ca. 176 BC and was, according to Strabo, a large and impressive city in the Augustan period.25 Artaxata’s strategic significance is indicated in the Peutinger Table, which shows that the city was connected by roads to northern Mesopotamia to the south-west and Media to the south-east (see below for more on the Peutinger Table). It was also connected by roads to the ports and cities of the Black Sea. During the reign of Tigranes II (the Great), a new capital of Tigranocerta was built to replace Artaxata and was settled with captives from 12 cities in Mesopotamia and Syria.26 Tigranocerta was destroyed by Lucullus in 69 BC in spite of its impressive defences and Artaxata once again became the capital despite Pompey’s later rebuilding of Tigranocerta. 27 Artaxata was then largely destroyed by the forces of Nero in 59; however, following an agreement over the government of Armenia with the Parthians, Nero sponsored its reconstruction a few years later. 28 Artaxata would once again suffer extensive damage in a Roman siege during the wars of Lucius Verus in the 160s and was replaced in its role as a capital for a time by the city of Kainepolis (Valarshapat), approximately 70 km to the north-west. Kainepolis was heavily fortified by the Romans in an attempt to take more effective control of Armenia in the wake of Lucius Verus’ war in Armenia aimed principally at expelling the Parthian client-king recently installed there.29 Sinclair refers to evidence for a road from Kainepolis in the direction of Satala on the upper Euphrates where a Roman legionary base had been located since the reign of Vespasian.30 Artaxata would once again become the capital of Armenia in the third century and remained an important urban centre after the settlement with Jovian in the wake of Julian’s death in Persia in 363. Arsamosata was located in the west of the ancient kingdom of Armenia in Sophene, and like Tigranocerta, its modern location has been the subject of debate. Sinclair argues that its remains lie partly submerged under the Keban Dam at Haraba on the upper Euphrates River in Turkey.31 Cohen points out that Pliny referred to Arsamosata as one of the important cities of Armenia ca. 70 and that Tacitus referred to it as a fortress.32 As an important city in Sophene, Arsamosata likely came under Roman control in the late third century and was later the metropolis of one of the four Roman provinces of Armenia after the kingdom’s partition in the 380s.33

12 Introduction

Northern Mesopotamia and Syria South of Armenia, northern Mesopotamia was a more manageable corridor of large-scale military movement throughout the whole period of the Roman and Parthian/Persian relationship.34 This less mountainous but still fertile zone dominated by the Tur ‘Abdin (Mt Masius/Mt Izala) and plains to its west and south was more advantageous to large-scale military and other types of east-west movement in a number of ways. Larger armies could navigate this less mountainous area than Armenia and it lies north of the 200 mm isohyet, often marked on topographical maps to indicate a generally agreed-upon minimum annual rainfall allowing for cereal crop cultivation without the need for irrigation.35 An issue that would regularly confront armies in the northern Mesopotamian plains, however, was searing heat in the summer when the campaigning season was underway. This was problematic for an infantry-based army like that of the Romans but it was also challenging for the heavily clad Cataphracts serving in the Parthian and Persian armies. The heat could be so intense for the Cataphracts that the Roman soldiers referred to them as clibanarii; a clibanarius being an army field oven. While the heat of summer was a serious inconvenience to the armies, the concentration of rainfall in the winter months made operations for a large army in northern Mesopotamia a difficult and perhaps even dangerous prospect.

The Tur ‘Abdin, the Regiones Transtigritanae and Adiabene The area of northern Mesopotamia between southern Armenia and the Euphrates requires some specific analysis as an element in the military relationship between Rome and its eastern neighbour. The flow of the Tigris and the gorge it cuts in a west-east direction from the vicinity of Amida (modern Diyarbakir) to the area just east of Bezabde (near modern Cizre) together with the escarpment immediately to its south known today as the Tur ‘Abdin is an important area to understand with regard to its fortification, especially in the fourth century.36 The flow of the Tigris from its source in the Armenian highlands to turn in an easterly direction in the vicinity of Amida and from there to Bezabde was advantageous to the Romans when directing large-scale military campaigns.37 It was also advantageous in terms of defence. The Tigris cuts a narrower and deeper gorge in this section than the Euphrates does further to the south. With fortifications such as Bezabde and Amida, either founded or significantly upgraded under Constantius II in the fourth century, any invading army faced a significant challenge. Immediately to the south of this length of the Tigris lies the Tur ‘Abdin, an east-west stretch of higher ground, termed as an escarpment by some but given the name Mt Izala in antiquity, which was also a challenge to any large-scale invading army. An invading Persian army of the fourth century

Introduction  13 either needed to confront Bezabde and move up the Tigris taking other fortifications before reaching Amida or travel to the south of the Tur ‘Abdin where it would confront the fortress of Nisibis.38 If it went around Nisibis, it would then need to confront Rhesaina and Constantia.39 (See Figure 8.1, Chapter 8: Google Earth image of key defensive sites in Roman Mesopotamia under Constantius II.) Further to the west lay the cities of Carrhae and Edessa in ancient Osrhoene. Edessa became a particularly important forward base for the Romans and its acquisition, along with the rest of Osrhoene, in the second half of the second century was of major strategic importance to the Romans.40 An attempt to travel down the Khabur River and direct an invasion via the Euphrates would potentially implicate the garrison at Singara and would ultimately require the fortress of Circesium to be assailed. The fortress of Singara’s location at the foot of Jebel Sinjar was highly strategic as it was the first major fortification an invading Persian army would meet on crossing the Tigris at Nineveh. As a high point in otherwise flat terrain, Jebel Sinjar offered early warnings of enemy movement. In political terms, this was also a complex area and became even more so when the Romans pushed beyond the Euphrates to the Tigris in the second and third centuries AD. Between southern Armenia and northern Mesopotamia, along the west-east stretch of the Tigris, lay a number of principalities referred to collectively in the third and fourth centuries as the Regiones Transtigritanae. The separate identity these principalities enjoyed had origins in the independence asserted during the late Seleucid period when the empire began to disintegrate.41 Marciak notes that these principalities then became part of the “Parthian Commonwealth” before some of them (likely all), including Sophene and Corduene, became part of Tigranes II’s Greater Armenia ca. 70 BC.42 Following the demise of Tigranes, however, Corduene came under the effective control of Adiabene to its south, essentially a part of the Parthian Empire. Adiabene was, according to Marciak, at the height of its powers in the first and second centuries AD and was at times attacked and even captured by Roman invading forces.43 The king of Adiabene was present at the conference of Rhandeia in 63 when Rome and Parthia reached an agreement over the rulership of Armenia.44 Adiabene was a significant feature of the wars of Trajan in which it was subjugated by the Romans twice. Septimius Severus and Diocletian were both granted the title Adiabenicus in recognition of their defeat of the kingdom in the context of campaigns directed at Parthia and Sasanian Persia. While Adiabene was often a feature of the conflicts between Rome and Parthia/Persia, Marciak emphasises that it was never under formal Roman control, other than for a brief period of time as part of Trajan’s short-lived province of Mesopotamia or Assyria. In the Sasanian period, Adiabene was more intensively “Iranised” and “its local Semitic cults may have largely succumbed to the spread of state-sponsored Zoroastrianism.”45 The Regiones Transtigritanae became especially relevant in the relationship between Rome and Persia when they were specifically named in a treaty

14 Introduction between the two empires at the end of the third century. In this treaty, the Tigris was named as a boundary between Rome and Persia and the Regiones Transtigritanae were to be Roman client-kingdoms, their collective name indicating that they lay substantially beyond the Tigris. There is some debate on the identification of the individual Regiones Transtigritanae. Five named territories appearing in the treaty of 299 between Rome and Persia were Intelene, Sophene, Arzanene, Corduene and Zabdicene, but when Ammianus referred to the Romans giving up the Regiones Transtigritanae in 363 as a consequence of Jovian’s treaty with Shapur II, he provided a different list of names.46 Arzanene, Corduene and Zabdicene appeared on Ammianus’ list but in place of Intelene and Sophene, Ammianus included Roximene and Rehimene. Zosimus provided a different list again, not naming Moxoene or Arzanene, but including one called Zalene.47 The proliferation of regional names such as these in the area of northern Mesopotamia and the capacity for them to change over time is a likely contributor to the confusion but the important point to note is that they played the part of a buffer zone and were keenly contested by Rome and Persia in the third and fourth centuries.

The Euphrates and Tigris Rivers At a broader level, the Euphrates and Tigris Rivers, whose upper reaches form the loose boundaries of northern Mesopotamia, were two of the key geographical features in military and political competition between Rome and Parthia/Persia. Parts of the rivers became boundaries between the two powers, the Upper and Middle Euphrates serving in this role during much of the Parthian period and, as discussed earlier, the upper Tigris in the third and fourth centuries.48 Both rivers have their headwaters in ancient Armenia and in their upper reaches are narrower and deeper than their flows through the deserts of Syria and Mesopotamia further to the south. The directional flow of the Euphrates out of western Armenia through territory mostly under Roman control in Cappadocia and Syria before flowing into central and southern Mesopotamia meant that the river was an ideal means by which Roman armies attacked the Parthian and Persian capital of SeleuciaCtesiphon. With the directional flow of the Tigris running west-east as it flowed out of ancient Armenia, the river played an important defensive role in the fourth century before it turned to flow south into Adiabene and then on to Assyria and southern Mesopotamia. The Tigris was also a means of the Romans sending armies against Seleucia-Ctesiphon in support of those moving down the Euphrates. There were times when the Euphrates River valley was used by invading Parthian and Persian armies to threaten the rich cities of Roman Syria, including Antioch. The western bend of the  Euphrates in the vicinity of Barbalissos was especially important in this respect. From this point, Antioch and other cities were vulnerable and it would prove a distinctively strategic location for Parthian and Persian armies during attacks on Roman territory on a number of occasions.

Introduction  15 In addition, the broader course cut by the Euphrates from the point at which the river bends in a south-easterly direction near Barbalissos through the Syrian desert and all the way to the vicinity of Seleucia-Ctesiphon allowed a significant and continuous band of irrigation over hundreds of kilometres. The breadth of the irrigation band along this section of the Euphrates supported fortified cities and towns and the movement of armies and traders in either direction. One of the most important settlements and fortifications on the Euphrates below its confluence with the Khabur River (see more below) was Dura Europos, an early Seleucid foundation that was captured by the Parthians ca. 100 BC and later became a fortified town under the Romans in the second and third centuries AD. Below it lay a number of settlements and smaller fortifications dating to the Severan period stretching as far south as ancient Becchufrayn (modern Kifrin) in Iraq.49 On the basis of textual and archaeological evidence, these were the furthest permanent Roman military structures located on the Euphrates. An increased focus in the Sasanian period on irrigation on the lower Euphrates as it bends towards the Tigris in the vicinity of Seleucia-Ctesiphon created a challenging barrier to an invading Roman army attempting to attack Seleucia-Ctesiphon and its surrounds. When Julian’s army invaded in 363, it is clear that a whole complex of settlements and fortifications stretched from Pirisabora on the Euphrates along the canals to SeleuciaCtesiphon on the Tigris. Along with Vologaesias, which appears to have been founded as a type of market town in the middle of the first century AD by Vologaeses I, and the ongoing importance of Babylon as an urban settlement, this whole area of southern Mesopotamia became even more populous in the Sasanian period.50

Bridge crossings of the Euphrates and Tigris Permanent bridge crossings of the Euphrates and Tigris in antiquity were few in number, partly because they were difficult to maintain due to the spring snow melt flowing from the Armenian highlands. By far, the most important crossing of the Euphrates was at Zeugma on the upper reaches of the river in northern Syria (modern Turkey). Zeugma was established as a permanent crossing early in the Seleucid period and it remained as such through the whole of the Roman/Parthian/Sasanian period. The Zeugma bridge crossing gave access to the north Mesopotamian plain from northern Syria and vice versa and was also important to Roman armies marching to Armenia and the upper Tigris. It is possible that a bridge referred to in ancient texts at Capersana lay further north of Zeugma, although the identification of Capersana with modern Ayni on the Euphrates, some 40 km north of Zeugma, is speculative.51 There is some evidence to suggest a permanent bridge crossing in the vicinity of Sura, some 80 km downstream from the bend of the Euphrates at Barbalissos in modern Syria, although this crossing may not have existed

16 Introduction until the third century AD or later.52 Some military crossings of the Euphrates (such as that of Julian’s in March 363) were made via temporary bridges constructed using boats. Military crossings of the Khabur River at its confluence with the Euphrates (first century BC and fourth century AD) were also made via bridges constructed of boats, which was the case when Julian’s army crossed the Khabur in 363.53 Bridges of boats had their limitations and could not be used when the rivers were flooded. They were comparatively narrow, meaning that a large army (such as that of Julian’s) could take days to cross, thus leaving it vulnerable to attack during that time. The most important bridge crossing of the Tigris was at Nineveh (Mosul). The use of this crossing is evident in the mid-fourth century invasion of Roman Mesopotamia by Shapur II; however, it was possibly used in the third century by the invading Roman army of Severus Alexander. The bridge crossing at Nineveh connected Adiabene and Assyria with northern Mesopotamia and the Euphrates River via the Khabur. Its military importance is obvious but the bridge also connected these areas for the purposes of trade and would have been important to the trading kingdom of Hatra some 70 km south-west of Mosul.

Tributaries of the Euphrates The other rivers of any significance in northern Mesopotamia are the two tributaries of the Euphrates, the Khabur and the Balikh (ancient Belias). These tributaries are fed by smaller river water courses and streams flowing from southern Armenia and in and around the Regiones Transtigritanae and the Tur ‘Abdin. The Balikh is a smaller, more intermittent river than the Khabur and has its headwaters in the vicinity of ancient Edessa. While the Balikh’s confluence with the Euphrates is moveable (like that of the Khabur), it has consistently been in the vicinity of ancient Callinicum/ Nicephorium (modern Raqqa) where a fortress was located in the Roman period.54 The Khabur is a more substantial tributary of the Euphrates and its headwaters are located in the Tur ‘Abdin in northern Mesopotamia. One of its feeder rivers in antiquity was the Mygdonius, the main water source for the city and fortress of Nisibis. The Khabur’s confluence with the Euphrates was identified by the Romans as a key defensive point where the fortress of Circesium was significantly upgraded during the reign of Diocletian and maintained at least until Justinian’s reign.55 A smaller fortification existed at the confluence prior to the strengthening activity of Diocletian’s reign and was significant enough to be referred to in the SKZ inscription of Shapur I. During the Parthian period, the Khabur confluence was recognised as a point at which Roman and Parthian spheres of influence met and during Julian’s Persian campaign there is a strong sense in the sources reporting the campaign that the crossing of the Khabur represented a point where the campaign entered enemy territory.56 The Khabur supported villages along its entire length during the Roman period and this is also shown

Introduction  17 archaeologically much earlier in the Bronze Age. One of the Khabur’s roles was to act as a conduit of military and other movement between eastern Syria and northern Mesopotamia. The fortress of Singara, at the foot of Jebel Sinjar in northern Iraq, was one of the important links to be made via the Khabur from the Euphrates and eastern Syria.

The oases of Palmyra and Hatra The deserts of Iraq and Syria are typically thought of as little more than wastelands in antiquity other than the settlements that formed around artesian water sources such as Palmyra and Hatra. The oasis created by the Efqa spring at Palmyra was capable of supporting a citizen population numbering tens of thousands together with animals, agriculture and itinerant traders. Archaeological evidence suggests, however, that the hinterland of Palmyra was more capable of agricultural production and supporting human population than previously thought. A recent project undertaken by Jorgen Meyer focussing on the steppe to the north-west of Palmyra indicates evidence for farmsteads and small villages dating to the Roman/Byzantine period over an extensive area.57 Evidence for agricultural activity includes remains of crops such as barley and olives. This type of agriculture was likely supplementary to pastoralism and it was made possible by water harvesting techniques, which took advantage of the little rain that fell in winter. The steppe lands and jebels to the north of Palmyra were the catchment area for the subterranean water sources that fed the spring at Palmyra. The overall conclusion of Meyer is that the Palmyrene hinterland to the north should be referred to as dry steppe rather than desert. Similar observations might also be made of the territory to the north of Hatra, the other desert trading principality whose significant remains rival those of Palmyra.58 The water supply at Hatra came principally from the Wadi Tharthar, an intermittent system of wadis emanating from the foothills of Jebel Sinjar, some 70 km to the north of the city. Evidence for wells and an internal dam at Hatra indicate the possibility of an artesian source of water.59 It is likely that the territory around Hatra was similarly capable of supporting some agricultural activity that complemented pastoralism. While the deserts and steppe areas of Syria and Iraq likely supported more human and animal life than has sometimes been suggested, they continued to represent significant challenges to population movement of any significance. No large army could march through such territory without facing serious challenges.60 It was, of course, possible for trade to be conducted through these areas and the Palmyrenes became experts in this achievement, especially in the first and second centuries AD. The oasis was located precisely half-way between the Mediterranean coast and the Euphrates River, making it an ideal stopping point in the Syrian desert for long-distance and regional trade between the Roman world and further east to Parthia/Persia, and India via the Persian Gulf.

18 Introduction

Arabia The Arabs of Syria, Mesopotamia and the Arabian Peninsula would become an increasingly important element in the military, political and diplomatic relationship between Rome and Sasanian Persia. The zenith of their significance came in the sixth and seventh centuries in the form of proxy wars fought between the Jafnid and Nasrid allies of Rome and Persia and other alliances formed deep within the Arabian Peninsula by both parties.61 The kingdom of Aksum (Ethiopia), a Roman ally from the fourth century onwards, was also a factor in the affairs of south Arabia, especially from the third century onwards. From a Persian perspective, Arabia was of interest partly with regard to the security of the Persian Gulf. The Parthian regime struggled to maintain control of the gulf with kingdoms such as Characene operating with virtual independence. The gulf coast of southern Iran appears to have slipped from Parthian control by the time of the Sasanian removal of the Arsacids in the 220s and Ardashir made it a priority to regain control of the whole of the gulf, establishing new cities and ports and mounting punitive military actions against Bahrain and the Arabian coast of the gulf soon after he came to power. It appears that Ardashir later went even further than this, campaigning in Oman and further west in Yemen.62 In fact, Mazun, the name by which Oman was known to the Sasanians, appears on the SKZ inscription at the end of the list of provinces over which Shapur I claimed to rule ca. 265.63 Approximately a century later, Shapur II mounted a major campaign against the Arabs in response to problems in the Persian Gulf, which took his forces to Yathrib (Medina) and beyond, a distance of over 1,000 km.64 This was the first military campaign of Shapur II’s majority, which is indicative of the priority the Arabian Peninsula represented to the early Sasanians. Rome’s acquisition of Egypt in the wake of the Battle of Actium in 31 BC saw direct Roman interaction in the Arabian Peninsula soon after with the campaign of Aelius Gallus deep into the peninsula in 24 BC. Rome’s later annexation of the Nabataean kingdom in AD 106 took the empire’s direct territorial interests as far south as Hegra (Medain Saleh) in modern Saudi Arabia. The acquisition of Egypt also resulted in direct contact with Nubia and Ethiopia to the south. While Aelius Gallus was on campaign in Arabia, his ultimate successor as the Prefect of Egypt, Caius Petronius, embarked on campaigns against Ethiopia after the Ethiopians had attacked Rome’s southern Egyptian garrisons.65 Pliny claimed that Petronius captured a number of Ethiopian cities and that this war was in effect a continuation of the wars long fought between Egypt and Ethiopia. Pliny noted that Nero sent an expedition into Ethiopia with a view to invading the country as part of his wars of expansion.66 If the motive reported by Pliny is correct, Nero was weighing up the prospects of military action against Ethiopia at the same time he was dealing with difficulties in Armenia.67

Introduction  19 By the time of the Sasanian overthrow of the Parthians in the 220s, military and political competition between the two powers extended over a considerable geographical range. The means by which this happened is dealt with in more detail in the next chapter. Under the Sasanians, this geographical range extended to include the Arabian Peninsula and Roman imperial expansionism, especially evident in the second half of the second century, was not only challenged by Sasanian rulers, but matched by them.

An ancient conception of the geography and topography of Syria, Mesopotamia, Arabia and Persia The Peutinger Table, on which considerable scholarship has been undertaken for centuries, is an ancient cartographic conception of the world of the Roman Empire, Persia and India that is worth some brief consideration, given the topic of this book. The version on which all modern copies of the map have been produced is generally agreed to date to the thirteenth century and, while there is a debate on the original version, the most recent suggestion of the date of its composition is ca. 300, that is, during the period of the tetrarchy.68 Others believe that it is more likely from the third century (one suggestion is that it was produced for the celebration of Rome’s Millennium in 247), while arguments have been put placing its composition as late as the eighth century in the Carolingian period.69 In the form we have it, the Peutinger Table is an unusual map by modern standards, measuring almost 7 m in length but only a little over 30 cm in height. This feature of the map, presuming that it indicates its original dimensions, is sometimes used to suggest its purpose as a publicly displayed document, perhaps even in an imperial palace as Talbert proposes.70 Many admirable attempts have been made to identify sites on the ground with those that are illustrated and named on the map and how we might use the indication of distances between sites indicated on the map to better understand road and communication networks across the empire. The publication of the map in an online format by a team headed by Richard Talbert is an extraordinary achievement and augments the usefulness of the map markedly.71 Closer examination of the map indicates a number of apparent errors and repetitions of the same sites, while some important sites were either not included or appear to be inexplicably minor. Alexandria in Egypt and Jerusalem in ancient Palestine are examples of the latter. For the purposes of this study, the absence of Armenia is perhaps the most surprising and difficult to explain. Antioch-on-the-Orontes contains by far the largest place-marker of any settlement in the east and for that matter on the entire map. It is larger than Constantinople and larger than Rome. Rocco Palermo has recently undertaken an admirable analysis of the roads and locations in northern Mesopotamia, which are marked on the map.72 My intention here is to comment on how the Peutinger Table as a

20 Introduction document with possible origins in the late third/early fourth century might be used to make more general observations about late Roman imperial perspectives on Armenia, Syria, Mesopotamia Arabia and Persia; that is the area that was the focus of conflict and diplomacy between Rome and Sasanian Persia in the third and fourth centuries. There are ways in which the map is quite clearly nuanced, other than its strikingly unconventional dimensions, and it must be emphasised that there will always be considerable speculation about its origins, purposes and the extent to which the thirteenth-century copy is in any way reflective of a late antique original. The areas of Syria, Arabia, Mesopotamia and Persia are represented on segments 7 and 8 of modern publications of the map, the first undertaken in 1598. Mountains in this area are only marked sparingly and are indicated primarily by a single chain of mountains running in an undulating fashion across the northern section of the map possibly designed to correspond very roughly with the Taurus and Zagros mountain chains. A small sector of mountains marked as Mons Parverdes are possibly the Parihedri mountains in northern Armenia and another smaller one with the city of Edessa at its foothills is possibly an erroneously placed conception of the Tur ‘Abdin.73 Due in part to the extreme compression of the north-south axis of the map, deserts and steppe zones are barely depicted despite their actual dominance of large parts of the landscape to the south. This is possibly an indication that the map was designed to have more of a focus on urban and/or fortified settlements, although it is more likely due to the key focus appearing to lie on roads and the distances between them. In the section of segment 7 that roughly corresponds with eastern Syria, there are some interesting regional markers: one indicating Aree fines Romanorum (“Areas that are the Roman frontier”) and another beneath it marking Fines excercitus Syriatice et commertium Barbaror (“Limit of the army [based in] Syria and [place for] commercial exchange with

ç Figure 1.1 Parts of Segments 7 & 8 of the Peutinger Table detailing Syria, Mesopotamia, Arabia and Persis. Image Peter Edwell.

Introduction  21 the barbarians”).74 The latter is marked at the end of a road beginning in Apamea in Syria that ran via Palmyra. Another area of interest that is marked Campi Deserti et in habitabiles propter aqv(a)e inopiam (“Desert plains, uninhabitable because of lack of water”) lies between the upper Euphrates and Tigris Rivers. While marked a long way north on the map, it likely represents the Syrian desert south of the Euphrates in modern Syria. The very unusual vertical marking of Trogoditti Persi to the east of Zagurae and Singara on the west bank of the Tigris River would appear to mark a conception of where Persian territory began. At the very eastern end of segment 8, and thus the entire map, the words Hic Alexander Responsum accepit usq:quo Alexander (“Here Alexander received the response: ‘Only this far, Alexander’”) mark the eastern limit of Alexander’s exploits. It is potentially revealing that the eastern end of a map of the known world likely produced in late antiquity was marked by this observation. Quite clearly, the Euphrates and Tigris Rivers are strong features of segments 7 and 8 and are among the most significant natural features marked on this section of the map. The joining of the two rivers in their lower reaches and the complex system of canals in the vicinity of Seleucia, Ctesiphon and Babylon is also prominently marked. The Persian Gulf is clearly compressed and serves mostly to indicate the islands located in it. Thylos (modern Bahrain) was perhaps the best known to the late Roman world. As noted briefly earlier, the most practical feature of the map is the marking of roads with distances between settlements along them. The roads all appear to run roughly in a west-east direction, which is partly explained by the strong compression of the north-south axis. Roads running from Cappadocia and Syria to the Euphrates are numerous as are those running from the Euphrates to the Tigris. Those to the north in the more mountainous regions are less frequent as are those to the south in desert areas, although this is not as obvious due to the limited inclusion of desert areas. The focus on roads and distances between settlements on them is strongly suggestive of a linear map or itinerary in terms of the map’s practical use, if, indeed, it was designed with this purpose in mind. How representative the map might be of an imperial, elite or any other view of the world in the late antique period will continue to be debated; however, there are ways in which the Peutinger Table indicates ancient conceptions of geographical and topographical connectivity, which, in turn, potentially informs which areas were of principal imperial importance.

Conclusion The analysis undertaken in this book of the military and political relationship between Sasanian Persia and Rome in the third and fourth centuries contributes to understanding how imperial power changed and transformed both internally and between the two super-powers of south-west Asia in this period. At the time that the Sasanian Persian dynasty emerged under

22 Introduction the leadership of Ardashir in the mid-220s AD, the Roman and Parthian Arsacid Empires had shared a rivalry stretching back over many centuries. It was a rivalry that had worked mostly in Rome’s favour, especially over the preceding century. With some significant internal developments in the third century that were partly brought on by the challenges of the renascent Persian Empire, Rome faced its most serious existential challenges since it left the Italian Peninsula and began building an empire. The changes and reforms that took place in the reigns of Ardashir and Shapur I in the Persian Empire contributed in part to Sasanian successes against Rome and established the dynasty as a viable successor to the Arsacid dynasty for centuries to come. Sasanian successes against Rome and Rome’s ability not only to survive, but eventually adapt were of fundamental importance to both powers through the rest of their histories.

Notes 1 Cassius Dio 75.3.2–3. 2 Peter the Patrician, Banchich 2015, F 201. 3 See for example, Christensen 1939, 109–37; Christensen 1944, 213–52 (Third and Fourth Centuries); Frye 1983, 116–80; Potter 2004, 215–99; Frye 2005, 461–80; Sartre 2005, 343–63; Edwell 2013, 840–55; Greatrex 2013; Winter 2013; Kulikowski 2016, 117–299. 4 Canepa 2009. 5 Potter 1990. 6 Sprengling 1940, 1953; Kettenhofen 1982; Huyse 1999. 7 Dodgeon & Lieu 1991, revised 1994. See also Greatrex & Lieu 2002 in which Chapter 1 contains source extract material on the agreement between Jovian and Shapur II in 363. 8 Dignas & Winter 2007; originally published in German as Winter & Dignas 2001. 9 See Bonner 2015, 81–91 on the complex question of the sources available to the Arabic Chroniclers. 10 Bosworth 1999, xix–xx. 11 See Hoyland 2018 for analysis and translations of the relevant sections of Ya`qubi, Mas`udi and Hamza Isfahani. 12 Levy 1967, 283–4. 13 See Millar 1969, 15. 14 Photius, Bibliotheca, 62. 15 See Barnes 1970. 16 Photius, Bibliotheca, 82. 17 See Croke 2012 and Simelidis 2012. 18 See Edwell 2010 for more detail. 19 See Vitale 2013. 20 Strabo 11.3.1; Braund 1994, 205ff for a discussion of the archaeological evidence to support this claim. 21 Agathangelos 1.19–22. 22 Plutarch, Pompey, 34.1ff. 23 The most successful Armenian king in this respect was Tigranes “the Great,” who not only united Armenia during the first half of the first century BC but with his father-in-law Mithridates of Pontus, expanded Armenia’s power to that of a regional empire. Tigranes took the title “king of kings,” which partly

Introduction  23

24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

48 49 50 51 52 53 54 55 56 57 58

reflected the reality of his position, but was influenced by the title some of the Parthian rulers of Iran took. See Gregoratti 2013 for further discussion of the geographical issues confronting the rulers of Armenia. Hewsen 2011. Strabo 11.14.6. Strabo 12.2.9. Appian, Mithridatic Wars 115. Cassius Dio 58.6.5. Cassius Dio 71.3.1. Sinclair 1987, 434. Sinclair 1989 (vol. 3), 112. Pliny Natural History 6.26; Tacitus, Annals 15.10. On Sophene see especially Marciak 2014. Cohen 2012, 45. Sinclair 1989, 139–40. For a detailed analysis of the geography and climate of northern Mesopotamia, see Palermo 2019, 9–23. See Kennedy & Riley 1990, 25 for a good indication of the 200mm isohyet. This publication is indispensable to any consideration of the Roman presence on the landscape of northern Mesopotamia, Syria and parts of Arabia. On Amida see Sellwood 1989/2011; Bezabde see Comfort 2017, 194–7. See further discussion below on the Tigris. On Nisibis, see Palermo 2014; Cameron 2019, 287–96; Palermo 2019, 71–9. On Singara, see especially Palermo 2019, 79–89. See also Kennedy & Riley 1990, 125–7. On Rhesaina, also see Palermo 2019, 89–95. On Edessa and its foundation, see Ross 2001, chapter 1 and Segal 1970, chapters 1 & 2. On the Roman period at Edessa, see Ross 2001, chapters 2–6 and Segal 1970, chapters 3 & 4. On Carrhae, see Shahbazi 1990. See Marciak 2017, 419–25. Marciak 2017, 426. Marciak 2017, 427. See Chapter 2, p. 37. Marciak 2017, 428–9. Peter the Patrician, Banchich F 202 for the text of the treaty. Ammianus Marcellinus 25.7.9. Zosimus 3.31.1. The simple solution of combining all of the principalities together, which renders the lists of Peter the Patrician and Ammianus defective, is rejected by Blockley 1984, 32 who provides a complex argument on why Peter the Patrician’s list should be accepted as accurate for 299. See Mosig-Walburg 2009, 134–47 and Den Boeft et al. 2005, 234–7 for a detailed discussion of the issue. See Edwell 2013 for analysis of the development of the Euphrates as a boundary between Rome and Parthia in the first century BC and first century AD. See Edwell 2008, 68–73. See especially Fiey 1967 for detailed archaeological evidence on this topic. See Comfort et al. 2000. Konrad 2001, 118; Edwell 2008, 88–9. Isidore of Charax, Parthian Stations, 1; Ammianus Marcellinus 23.5.1–4. Edwell 2017, 132. See Edwell 2017, 133–4. Isidore of Charax 1; Ammianus Marcellinus 23.5.1ff. Meyer 2017. See Palermo 2019, 96–105; Gawlikowski 2013, 73–4 notes the similarities in the landscapes surrounding Hatra and Palmyra, referring to Hatra’s location as “dry steppe.”

24 Introduction















References Catalogue of the Greek coins of Arabia, Mesopotamia and Persia in the British Museum (BMC 28), G.F. Hill , ed., 1922, London. Corpus Inscriptionum Latinarum , (CIL), C. Caesaris , ed., Berlin 1863-Present Open source: https://arachne.uni-koeln.de/drupal/?q=en/node/291 Coin Hoards , 1975. Volume I: Royal Numismatic Society, London. Dura Europos Papyri (P. Dura) , The Excavations at Dura Europos, Final Report Volume V, Part 1: The Parchments and Papyri, C.B. Welles , R.O. Fink & J.F. Gilliam , eds. 1959, New Haven: Yale University Press. Gnecchi, F. , 1912, I medaglioni Romani. 3 Vols. Milan. Inscriptiones Latinae Selectae (ILS), Vols 13, H. Dessau , ed., Berlin 18921916. LAnne pigraphique (AE), R. Cagnat eds, 1888-Present, Paris. Moneta Imperii Romani, (MIR) R. Gbl , 2000, Die Mnzprgung des Kaiser Valerianus I / Gallienus / Saloninus / (253/268), Regalianus (260) und Macrianus / Quietus (260/262), Vienna: Verlag d. sterreich. Akad. d. Wiss. Palmyrene Aramaic Texts (PAT) D.R. Hillers & E. Cussini , eds. 1996, Baltimore: Johns Hopkins University Press. Roman Imperial Coinage, (RIC) Volumes 110, 19231994, C.H.V. Sutherland et. al., eds. London. Roman Provincial Coinage Online (RPC Online) https://rpc.ashmus.ox.ac.uk/project. The versions of texts listed below are the most accessible available in English translation. References to ancient texts are in the Loeb Classical Library editions unless otherwise indicated below. Texts referred to in the versions reprinted in M.H. Dodgeon & S.N.C. Lieu , eds, 1994, The Roman Eastern Frontier and the Persian Wars AD 226363, London & New York: Routledge are referenced separately in relevant endnotes. Agathangelos , History of the Armenians, R.W. Thomson , trans., 1976, Albany: State University of New York Press. Agathias , Agathias on the Sassanians, A.M. Cameron , trans., 1969/70, Dumbarton Oaks Papers, 23/24, pp. 69183. Aphrahat , Demonstrations, A. Lehto , trans., 2019, Piscataway: Gorgias Press. Aurelius Victor , De Caesaribus, H.W. Bird , trans., 1994, Liverpool: Liverpool University Press. 256 Chronicle of Seert, A. Alcock , trans., 2014, http://archive.org/details/AlcockChronicleOfSeertET/page/n23 Codex Theodosianus, C. Pharr , trans., 1952, New York: Greenwood Press. Constantine , Oration to the Saints, M. Edwards , trans., 2003, Constantine and Christendom, Liverpool: Liverpool University Press. Dinawari , Kitab al-Ahbar al-Tiwal, M.R.J. Bonner , trans., 2015, Bures-sur-Yvette: Groupe pour letude de la civilisation du Moyen-Orient. Epitome de Caesaribus, (3 rd edition), T.M. Banchich , trans., 2018, Buffalo, NY: Canisius College: http://www.roman-emperors.org/epitome.htm Eusebius , Life of Constantine, A.M. Cameron & S.G. Hall , trans., 1999, Oxford: Oxford University Press. Eutropius , Breviarium, H.W. Bird , trans., 1993, Liverpool: Liverpool University Press. Ferdowsi , The Epic of the Kings: Shah-nama (Khudanameh), R. Levy , trans., 1967, London: Routledge & Kegan Paul. Festus , Breviarium, J.W. Eadie , trans., 1967, London: Athlone Press. Gelasius of Cyzicus Historia Ecclesiastica, G.C. Hansen , trans., 2008, Turnhout: Brepols. Gregory Nazianzenus , Two Invectives, trans., C. W. King , 1888, London: George Bell. Isidore of Charax , Parthian Stations, W.H. Schoff , trans., 1976 (reprint), Chicago: Ares Publishers. John Malalas , Chronicle, E. Jeffreys , M. Jeffreys & R. Scott , eds, 1986, Melbourne: Australian Association for Byzantine Studies. Justin , Epitome of the Philippic History of Pompeius Trogus, J.S. Watson , trans. 1853, London: Henry G. Bohn. KKZ Inscription, M. Boyce , trans., 1984, Textual sources for the study of Zoroastrianism, Chicago: University of Chicago Press, pp. 112113. Lactantius , De mortibus persecutorum, J.L. Creed , ed., 1984, Oxford: Oxford University Press. Moses Khorenatsi , History of the Armenians, R.W. Thomson , trans., 1978 = Dodgeon & Lieu 1994, pp. 314327. Oracula Sibyllina 13, D.S. Potter trans., 1990, Oxford: Oxford University Press. Orosius , Historiarum adversus paganos libri VII/ The seven books of history against the pagans, R.J. Deferrari , trans., 1964, Washington, DC: Catholic University of America Press.

Paikuli Inscription, H. Humbach & O. Skjrv , eds, 1978, The Sassanian Inscription of Paikuli, Wiesbaden; I: Supplement to Herzfelds Paikuli, by H. Humbach , 1978; II: Synoptic Tables, 1980; III.1 Restored Text and Translation and III.2 Commentary by P. O. Skjrv, Wiesbaden, 1983. Panegyrici Latini, C.E.V. Nixon & B.S. Rodgers , trans., 1994, In Praise of Later Roman Emperors: The Panegyrici Latini, Berkeley: University of California Press. Peter the Patrician , Fragments, T. Banchich , trans., 2015, London & New York: Routledge. Peutinger Table , Histoire des Grands Chemins, 1736, Bruxelles: Jean Leonard. Photius , Bibliotheca: A selection, N.G. Wilson , trans., 1994, London: Duckworth. Pseudo-Joshua the Stylite , Chronicle, J.W. Watt & F.R. Trombley , trans., 2000, Liverpool: Liverpool University Press. 257 Res Gestae Divi Augusti, P.A. Brunt & J.M. Moore , trans., 1967, Oxford: Oxford University Press. SKZ Inscription , P. Huyse , trans., 1999, Die Dreisprachige Inschrift buhrs I. an Der Kaba-i Zardut (KZ), 2 Vols, London: School of Oriental and African Studies. References are to the Greek version of the text unless otherwise stated. Socrates Scholasticus , Historia Ecclesiastica, A.C. Zenos , trans., 1890, Nicene and PostNicene Fathers, Series 2 Volume 2, Grand Rapids: Eerdmans. Sozomen , Historia Ecclesiastica, C.D. Hartranft , trans., 1890, Nicene and Post-Nicene Fathers, Series 2 Volume 2, Grand Rapids: Eerdmans. Tabari , History, C.E. Bosworth , trans., 1999, The History of al-Tabari, Volume V: The Sasanids, the Byzantines, the Lakmids, and Yemen, Albany: State University of New York Press. The Letter of Tansar, M. Boyce , trans. 1968, Roma: Istitutio Italiano per il Medio ed Estremo Oriente. The Martyrdom and History of Blessed Simeon Bar Sabbae, K.R. Smith , trans., 2014, Piscataway: Gorgias Press. Theodoret , Historia Ecclesiastica, B. Jackson , trans., 1892, Nicene and Post-Nicene Fathers, Series 2 Volume 3, Grand Rapids: Eerdmans Theophanes, Chronicle , C. Mango & R. Scott , trans. 1997, The chronicle of Theophanes Confessor: Byzantine and Near Eastern history, AD 284813, Oxford & New York: Oxford University Press. Zonaras , History, T.M. Banchich & E.N. Lane , trans., 2015, London & New York: Routledge. Zosimus , New History, R.T. Ridley , trans., 1982, Melbourne: Australian Association for Byzantine Studies. Al-Aswad, H.B. , 1991, Water sources at Hatra, Mesopotamia, 26, pp. 195211. Albu, E. , 2005, Imperial geography and the medieval Peutinger Map, Imago Mundi, 57, pp. 136148. Alcock, A. (trans), 2014, Chronicle of Seert, http://archive.org/details/AlcockChronicleOfSeertET/page/n23 (accessed on 15 December 2019 ). Alfldi, A. , 1937, Die Hauptereignisse der Jahre 253261 n. Chr. im Spiegel der Mnzprgung, Berytus, 4, pp. 4168. Algaze, G. , 1989, A new frontier: First results of the Tigris-Euphrates archaeological reconnaissance project, Journal of Near Eastern Studies, 48(4), pp. 240281. Alonso-Nez, J.M. , 1979, The Emperor Julians Misopogon and the conflict between Christianity and Paganism, Ancient Society, 10, pp. 311324. Alram, M. , 2007, Ardashirs Eastern campaign and the numismatic evidence, in J. Cribb & G. Herrmann , eds, After Alexander: Central Asia before Islam, Oxford: Oxford University Press, pp. 227242. Alram, M. , 2008, Early Sasanian coinage, in V.S. Curtis & S. Stewart , eds, The Sasanian Era, London: I.B. Tauris, pp. 1730. Altmayer, K. , 2014, Die Herrschaft des Carus, Carinus und Numerianus als Vorlufer der Tetrarchie, Stuttgart: Franz Steiner Verlag. Asmussen, J.P. , 1983, Christians in Iran, in E. Yarshater , ed., The Cambridge History of Iran, Volume 3 (2): The Seleucid, Parthian and Sasanian Periods, Cambridge: Cambridge University Press, pp. 924948. 258 Austin, N.J.E. , 1979, Ammianus on Warfare: An Investigation into Ammianus Military Knowledge, Brussels: Latomus. Aylward, W. , 2013, The rescue excavations at Zeugma in 2000, in W. Aylward , ed., Excavations at Zeugma, conducted by Oxford Archaeology, Vols. 1 and 3, Los Altos, California: Packard Humanities Institute, pp. 154.

Baldus, H.R. , 1997, Uranius antoninus of Emesa: A Roman emperor from Palmyras neighbouring city and his coinage, in M. Gawlikowski , ed., Palmyra and the Silk Road, Damascus: Annales Archologiques arabes syriennes, pp. 371377. Baldus, H.R. , 1971, Uranius Antoninus: Mnzprgung und Geschichte, Bonn: Habelt. Baldwin, B. , 1975, The career of Oribasius, Acta Classica, 18, pp. 8597. Ball, W. , 1989, Soundings at Seh Qubba, a Roman frontier station on the Tigris in Iraq, in D.H. French & C.S. Lightfoot , eds, The Eastern Frontier of the Roman Empire, Oxford; British Archaeological Reports, pp. 718. Ball, W. , 2016, Rome in the East: The Transformation of an Empire, Second Edition, London & New York: Routledge. Balty, J.C. & W. van Rengen , 1993, Apamea in Syria: The Winter Quarters of Legio II Parthica, Bruxelles: VUB Press. Banchich, T. , 2015, The Lost History of Peter the Patrician: An Account of Romes Imperial Past from the Age of Justinian, London & New York: Routledge. Banchich, T.M. , 2009, The History of Zonaras, New York & London: Routledge. Barnes, T.D. , 1970, The lost Kaisergeschichte and the Latin historical tradition, in A. Alfldi , ed, Bonner Historia Augusta Colloquium 196869, Bonn: Beitrage zur Historia-Augusta-Forschung, pp. 1730. Barnes, T.D. , 1976, Imperial campaigns, AD 285311, Phoenix, 30, pp. 174193. Barnes, T.D. , 1976, The Victories of Constantine, ZPE, 20, pp. 149155. Barnes, T.D. , 1978, The Sources of the Historia Augusta, Bruxelles: Latomus. Barnes, T.D. , 1980, Imperial chronology: AD 337350, Phoenix, 34, pp. 160166. Barnes, T.D. , 1982, The New Empire of Diocletian and Constantine, Cambridge, MA: Harvard University Press. Barnes, T.D. , 1983, Two victory titles of Constantius, ZPE, 52, pp. 229235. Barnes, T.D. , 1985, Constantine and the Christians of Persia, Journal of Roman Studies, 75, pp. 126136. Barnes, T.D. , 1998, Ammianus Marcellinus and the Representation of Historical Reality, Ithaca: Cornell University Press. Barnes, T.D. , 2009, The Persian sack of Antioch in 253, Zeitschrift fr Papyrologie und Epigraphik, 169, pp. 294296. Baur, P.V.C. , M.I. Rostovtzeff & A. Bellinger , eds, 1933, The Excavations at Dura Europos: Preliminary Report of the Fourth Season, 193031, New Haven: Yale University Press. Baynes, N.H. , 1955, Rome and Armenia in the fourth century, in Byzantium and Other Essays, London: Athlone Press, pp. 186208. Bennett, J.M. , 1997, Trajan Optimus Princeps: A Life and Times, London & New York: Routledge. Birley, A. , 1987, Marcus Aurelius: A Biography, New Haven: Yale University Press. Bivar, A.D.H. , 1972, Cavalry equipment and tactics on the Euphrates frontier, Dumbarton Oaks Papers, 26, pp. 271291. 259 Bivar, A.D.H. , 1979, The Absolute Chronology of the Kushano-Sasanian Governors in Central Asia, in J. Harmatta , ed., Prologomena to the Sources on the History of Pre-Islamic Central Asia, Budapest: Akadmiai Kiad, pp. 317332. Bivar, A.D.H. , 1983, The political history of Iran under the Arsacids, in E. Yarshater , ed., The Cambridge History of Iran, Volume 3 (1): The Seleucid, Parthian and Sasanian Periods, Cambridge: Cambridge University Press, pp. 2199. Blockley, R.C. , 1975, Ammianus Marcellinus: A Study of His Historiography and Political Thought, Brussels: Collection Latomus. Blockley, R.C. , 1984, The Romano-Persian peace treaties of A.D. 299 and 363, Florilegium, 6, pp. 2849. Blockley, R.C. , 1989, Constantius II and Persia, in C. Deroux , ed., Studies in Latin Literature and Roman History V, Brussels: Latomus, pp. 468489. Blockley, R.C. , 1992, East Roman Foreign Policy: Formation and conduct from Diocletian to Anastasius, Cambridge: Francis Cairns. Bonner, M.R.J. , 2015, Al-Dinawaris Kitab al-Ahbar al-Tiwal: An Historiographical Study of Sasanian Iran, Bures-sur-Yvette: Groupe pour letude de la civilisation du Moyen-Orient. Bosworth, A.B. , 1976, Vespasians reorganization of the north-east frontier, Antichthon, 10, pp. 6378. Bosworth, C.E. (trans.), 1999, The History of al-Tabari, Volume V: The Sasanids, the Byzantines, the Lakmids, and Yemen, Albany: State University of New York Press. Bowersock, G.W. , 1983, Roman Arabia, Cambridge, MA: Harvard University Press. Bowman, A.K. , 2005, Diocletian and the first tetrarchy, AD 284305, in A.K. Bowman , A.M. Cameron & P. Garnsey , eds, The Cambridge Ancient History, Vol. XII: The Imperial Crisis and

Recovery AD 193324, Second edition, Cambridge: Cambridge University Press, pp. 6789. Boyce, M. , 1968, The Letter of Tansar, Roma: Istitutio Italiano per il Medio ed Estremo Oriente. Braund, D. , 1984, Rome and the Friendly King: The Character of the Client Kingship, London & Canberra: Croom Helm. Braund, D. , 1994, Georgia in Antiquity, Oxford: Oxford University Press. Brennan, P. , 2015, Units in the Notitia Dignitatum, in Y. Le Bohec ed., The Encyclopedia of the Roman Army, Vol. 3, Chichester: Wiley-Blackwell, pp. 10491054. Brock, S.P. , 1978, A martyr at the Sasanid court under Vahran II: Candida, Analecta Bollandiana, 96, pp. 167181. Brock, S.P. , 1982, Christians in the Sasanian Empire: A case of divided loyalties, in S. Mews , ed., Religion and National Identity, Studies in Church History, Vol. 18, Oxford: Blackwell, pp. 119. Brok, M.F.A. , 1975, Die Quellen von Ammianus Exkurs ber Persian, Mnemosyne, 28, pp. 4756. Browning, R. , 1975, The Emperor Julian, Berkeley: University of California Press. Bruun, P. , 1965, A law concerning tax payment in gold and the Constantinian solidus, Studi di Antichit Cristiana, 26, pp. 527535 = Studies in Constantinian Numismatics, Acti Instituti Romani Finlandiae Vol. XII, Roma, 1991, pp. 7580. Bruun, P. , 1979, Successive monetary reforms of Diocletian, American Numismatic Society Museum Notes, 24, pp. 129148. 260 Burns, R. , 2016, Diocletians fortification of Syria and Arabia, in S.N.C. Lieu & P. McKechnie , eds, Aspects of the Roman East: Papers in honour of Professor Sir Fergus Millar FBA, Vol. 2, Turnhout: Brepols, pp. 177. Butcher, K. , 2015, Debasement and the decline of Rome, in R. Bland & D. Calomino , eds., Studies in Ancient Coinage in Honor of Andrew Burnett, London: Spink, pp. 181205. Butzer, K.W. , 1995, Environmental change, climatic history, and human modification, in J.M. Sasson , ed., Civilizations of the Ancient Near East, Vol. 1, pt. 2, New York: Scribners, pp. 123151. Cameron, A. , 1969/70, Agathias on the Sassanians, Dumbarton Oaks Papers, 23/24, pp. 69183. Cameron, H. , 2019, Making Mesopotamia: Geography and Empire in a Romano-Iranian Borderland, Leiden: Brill. Campbell, B. , 2005, The army, in A.K. Bowman , A.M. Cameron & P. Garnsey , eds, The Cambridge Ancient History, Vol. XII: The Imperial Crisis and Recovery AD 193324, Second edition, Cambridge: Cambridge University Press, pp. 110130. Campbell, B. , 2005, The Severan Dynasty, in A.K. Bowman , A.M. Cameron & P. Garnsey , eds, The Cambridge Ancient History, Vol. XII: The Imperial Crisis and Recovery AD 193324, Second edition, Cambridge: Cambridge University Press, pp. 127. Canepa, M. , 2009, The Two Eyes of the Earth: Art and Ritual of Kingship between Rome and Sasanian Iran, Berkeley: University of California Press. Canepa, M. , 2010, Technologies of memory in early Sasanian Iran: Achaemenid sites and Sasanian identity, American Journal of Archaeology, 114(4), pp. 563596. Canepa, M. , 2013, Sasanian rock reliefs, in D. Potts , ed., The Oxford Handbook of Ancient Iran, Oxford: Oxford University Press, pp. 856877. Chaumont, M.L. , 1986/2011, Armenia and Iran ii, the pre-Islamic period, Encyclopaedia Iranica, available online at http://www.iranicaonline.org/articles/armenia-ii (accessed on 30 December 2019 ). Chaumont, M.L. , 1969, Recherches sur lhistoire dArmnie de lAvenement des Sassanides a la Conversion du Royaume, Paris: Geuthner. Christensen, A. , 1939, Sassanid Persia, in S.A. Cook , F.E. Adcock , M.P. Charlesworth & N.H. Baynes , eds, The Cambridge Ancient History, Vol. XII: The Imperial Crisis and Recovery AD 193324, First edition, Cambridge: Cambridge University Press, pp. 109137. Christensen, A. , 1944, LIran sous les Sassanides, Second edition, Copenhagen: E. Munksgaard. Christensen, P. , 2016, The Decline of Iranshahr: Irrigation and Environment in the Middle East, 500 BC AD 1500, revised edition, London: I.B. Tauris. Chrysos, E.K. , 1993, Rumung und Aufgabe von Reichsterritorien. Der Vertrag von 363, Bnner Jahrbucher, 193, pp. 165202. Cohen, G.M. , 2012. The Hellenistic Settlements in the East from Armenia and Mesopotamia to Bactria and India, Berkeley: University of California Press. Comfort, A. , 2017, Fortresses of the Tur Abdin and the confrontation between Rome and Persia, Anatolian Studies, 67, pp. 181229. Comfort, A. , C. Abadie-Reynal & R. Erge , 2000, Crossing the Euphrates in antiquity: Zeugma seen from space, Anatolian Studies, 50, pp. 99126.

261 Corbier, M. , 2005, Coinage and taxation: The states point of view, AD 193337, in A.K. Bowman , A.M. Cameron & P. Garnsey , eds, The Cambridge Ancient History, Vol. XII: The Imperial Crisis and Recovery AD 193324, Second edition, Cambridge: Cambridge University Press, pp. 327392. Corcoran, S. , 2000, The Empire of the Tetrarchs: Imperial Pronouncements and Government AD 284324, Revised edition, Oxford: Oxford University Press. Croke, B. , 2013, Zosimus, in R.S. Bagnall , K. Brodersen , C.B. Champion , A. Erskine & S.R. Huebner , eds., The Encyclopedia of Ancient History, Maldon, MA: Wiley Blackwell, pp. 71867187. Crump, G.A. , 1975, Ammianus Marcellinus as a Military Historian, Wiesbaden: Franz Steiner. Dabrowa, E. , 2017, Tacitus on the Parthians, Electrum, 24, pp. 173189. Daryaee, T. , 1995, National history or Keyanid history?: The nature of Sasanid Zoroastrian historiography, Iranian Studies, 28(3/4), pp. 129141. Daryaee, T. , 2002, arhestnh-i rnar: A Catalogue of the Provincial Capitals of rnar, Costa Mesa: Mazda Publishers. Daryaee, T. , 2009, Sasanian Persia: The Rise and Fall of an Empire, London: I.B. Tauris. Davis, D. , 1996, The problems of Ferdowss sources, Journal of the American Oriental Society, 116, pp. 4857. De Blois, L. , 1975, Odenaethus and the Roman-Persian War of 252264 A.D., Talanta, 6, pp. 123. De Blois, L. , 1976, The Policy of the Emperor Gallienus, Leiden: Brill. De Blois, L. , 2019, Image and Reality of Roman Imperial Power in the Third Century AD: The Impact of War, London & New York: Routledge. de Jonge, P. , 1977, Philological and Historical Commentary on Ammianus Marcellinus XVII, Groningen: Bouma. de Jonge, P. , 1980, Philological and Historical Commentary on Ammianus Marcellinus XVIII, Groningen: Bouma. de Jonge, P. , 1982, Philological and Historical Commentary on Ammianus Marcellinus XIX, Groningen: Bouma. De Nardis, M. , 1989, Seneca, Plinio e la spedizione neroniana in Etiopia, Aegyptus, 69, pp. 123152. Debevoise, N.C. , 1938, A Political History of Parthia, Chicago: University of Chicago Press. Den Boeft, J. , D. den Hengst & H.C. Teitler , 1987, Philological and Historical Commentary on Ammianus Marcellinus XX, Groningen: E. Forsten. Den Boeft, J. , D. den Hengst & H.C. Teitler , 2002, Philological and Historical Commentary on Ammianus Marcellinus XXIV, Leiden: Brill. Den Boeft, J. , J.W. Drijvers , D. den Hengst & H.C. Teitler , 1998, Philological and Historical Commentary on Ammianus Marcellinus XXIII, Gronigen: E Forsten. Den Boeft, J. , J.W. Drijvers , D. den Hengst & H.C. Teitler , 2005, Philological and Historical Commentary on Ammianus Marcellinus XXV, Leiden: Brill. Dignas, B. & E. Winter , 2007, Rome and Persia in Late Antiquity: Neighbours and Rivals, Cambridge: Cambridge University Press. Dillemann, L. , 1962, Haute Msopotamie Orientale et Pays Adjacents, Paris: Geuthner. Dodgeon, M.H. & S.N.C. Lieu , eds, 1994, The Roman Eastern Frontier and the Persian Wars AD 226363, London & New York: Routledge. 262 Dodgeon, M.H. & S.N.C. Lieu , 1994, Libanius and the Persian wars of Constantius II, Acta Orientalia Belgica, 9, pp. 83109. Drijvers, J.W. , 2018, Magnus of Carrhae, in O. Nicholson , ed., The Oxford Dictionary of Late Antiquity, Vol. 2, Oxford: Oxford University Press, p. 945. Drijvers, J.W. , 2009, Rome and the Sasanid Empire: Confrontation and coexistence, in P. Rousseau , ed., A Companion to Late Antiquity, Malden & Oxford: Wiley-Blackwell, pp. 441454. Drinkwater, J. , 2005, Maximinus to Diocletian and the crisis, in A.K. Bowman , A.M. Cameron & P. Garnsey , eds, The Cambridge Ancient History, Vol. XII: The Imperial Crisis and Recovery AD 193324, Second edition, Cambridge: Cambridge University Press, pp. 2866. Duncan-Jones, R.P. , 1996, The impact of the Antonine Plague, Journal of Roman Studies, 86, pp. 108136. Edwell, P. , 2008, Between Rome and Persia: The Middle Euphrates, Mesopotamia and Palmyra and the Coming of Rome, London & New York: Routledge. Edwell, P. , 2010, The sources for Romes wars with Shapur I: Eurocentric and Eastern perspectives, Ancient West & East, 9, pp. 155179. Edwell, P. , 2013, The Euphrates as a boundary between Rome and Parthia in the late Republic and Early Empire, Antichthon, 47, pp. 191206.

Edwell, P. , 2013, Sasanian interactions with Rome and Byzantium, in D.T. Potts , ed. The Oxford Handbook of Ancient Iran, Oxford: Oxford University Press, pp. 840855. Edwell, P. , 2015, Arabs in the conflict between Rome and Persia, AD 491630, in G. Fisher , ed., Arabs and Empires before Islam, Oxford: Oxford University Press, pp. 214275. Edwell, P. , 2017, The Middle Euphrates River in the Seleucid and Roman periods, in E. Minchin & H. Jackson , eds, Text and the Material World: Essays in Honour of Graeme Clarke, Uppsala: Astrom Editions, pp. 129141. Edwell, P.M. , 2013, Osrhoene and Mesopotamia between Rome and Arsacid Parthia, in J.M. Schlude & B.B. Rubin , eds, Arsacids, Romans and Local Elites: Cross-Cultural Interactions of the Parthian Empire, Oxford: Oxbow, pp. 111136. Edwell, P.M. , 2013, Trajan: Imperialism victorious and frustrated, in D. Hoyos , ed., A Companion to Roman Imperialism, Leiden & Boston: Brill, pp. 251262. Ehling, K. , 1996, Der Ausgang des Perserfeldzuges in der Mnzpropaganda des Jovian, Klio, 78, pp. 186191. Ensslin, W. , 1949, Zu den Kriegen des Sassaniden Schapur I, Munich: SBAW. Fiema, Z.T. , 2018, Area 34: The Roman Fort. Preliminary report on the 2017 season, in L. Nehme , ed., Report on the 2017 Season of the Medain Saleh Archaeological Project, pp. 1736, https://hal.archives-ouvertes.fr/hal-01804965. Fiey, J.M. , 1967, Topography of al-Madain, Sumer, 23, pp. 338. Fisher, G. , 2011, Between Empires: Arabs, Romans, and Sasanians in Late Antiquity, Oxford: Oxford University Press. Fornara, C.W. , 1991, Julians Persian expedition in Ammianus and Zosimus, Journal of Hellenic Studies, 111, pp. 115. Fowden, G. , 1993, Empire to Commonwealth, Princeton: Princeton University Press. Fowden, G. , 1994, The last days of Constantine: Oppositional versions and their influence, Journal of Roman Studies, 84, pp. 146170. 263 Frendo, D. , 2001, Constantines letter to Shapur II: Its authenticity, occasion, and attendant circumstances, Bulletin of the Asia Institute, 16, pp. 5769. Frye, R.N. , 1977, The Sasanian system of walls for defense, in M. Rosen-Ayalon , ed., Studies in Memory of Gaston Wiet, Jerusalem: Hebrew University of Jerusalem, pp. 715. Frye, R.N. , 1983, The political history of Iran under the Sasanians, in E. Yarshater , ed., The Cambridge History of Iran: The Seleucid, Parthian and Sasanian Periods, Vol. 3 (1), Cambridge: Cambridge University Press, pp. 116180. Frye, R.N. , 2005, The Sassanians, in A.K. Bowman , A.M. Cameron & P. Garnsey , eds, The Cambridge Ancient History, Vol. XII: The Imperial Crisis and Recovery AD 193324, Second edition, Cambridge: Cambridge University Press, pp. 461480. Gariboldi, A. , 2010, Sasanian Coinage and History, Costa Mesa: Mazda Publishers. Gariboldi, A. , 2016, The birth of the Sasanian Monarchy in Western sources, in V.S. Curtis , E.J. Pendleton , M. Alram & T. Daryaee , eds, The Parthian and Early Sasanian Empires: Adaptation and Expansion, Oxford & Philadelphia: British Institute of Persian Studies, pp. 4752. Gawlikowski, M. , 1994, Fortress Hatra: New evidence on the ramparts and their history, Mesopotamia, 29, pp. 147184. Gawlikowski, M. , 2013, The development of the city of Hatra, in L. Dirven , ed., Hatra: Politics, Culture and Religion between Parthia and Rome, Stuttgart: Franz Steiner Verlag, pp. 7389. Gawlikowski, M. , 1985, Bijan in the Euphrates, Sumer, 42, pp. 1521. Ghirshman, R. , 1956, Bchpour, Vol.2, Les Mosaques Sassanides, Paris: Geuthner. Ghirshman, R. , 1962, Iran: Parthians and Sassanians, London: Thames & Hudson. Ghirshman, R. , 1971, Bchpour, Vol. 1, Paris: Geuthner. Gbl, R. , 1971, Sasanian Numismatics, Braunschweig: Klinhardt & Biermann. Gbl, R. , 1974, Der Triumph des Ssniden ahpuhr ber die Kaiser Gordianus, Phillipus und Valerianus: Die ikonographische Interpretation der Felsreliefs, Wien: Akademie der Wissenschaften. Gbl, R. , 1983, Sasanian coins, in E. Yarshater , ed., The Cambridge History of Iran: The Seleucid, Parthian and Sasanian Periods, Vol. 3 (1), Cambridge: Cambridge University Press, pp. 322339. Goltz, A. & U. Hartmann , 2008, Valerianus und Gallienus, in K.P. Johne , U Hartmann , T Gerhardt , eds, Die Zeit der Soldatenkaiser: Krise und Transformation des Rmischen Reiches im 3. Jahrhundert n. Chr.(235284), Berlin: De Gruyter Akademie Forschung, pp. 223295. Greatrex, G. & S.N.C. Lieu , 2002, The Roman Eastern Frontier and the Persian Wars: Part II AD 363630, New York & London: Routledge. Greatrex, G. , 2013, Persia and Byzantium, in R.S. Bagnall , K. Brodersen , C.B. Champion , A. Erskine & S.R. Huebner , eds., The Encyclopedia of Ancient History, Maldon, MA: Wiley Blackwell, pp. 51835184.

Gregoratti, L. , 2013, Between Rome and Ctesiphon: The problem of ruling Armenia, in Proceedings of the Conference Armenia Iran: History and Culture: The modern perspectives of progress, Moscow: Moscow State University, pp. 134141. Gregoratti, L. , 2017, Corbulo versus Vologases: A game of chess for Armenia, Electrum, 24, pp, 107121. 264 Grull, T. , 2014, Expositio totius mundi et gentium. A peculiar work on the commerce of the Roman Empire from the mid-fourth century compiled by a Syrian textile dealer? in Z. Csabai , ed, Studies in Economic and Social History of the Ancient Near East in Memory of Pter Vargyas, Ancient Near Eastern and Mediterranean Studies 2, LHarmattan: PcsBudapest Department of Ancient History, pp. 629642. Guey, J. , 1961, Autour des Res Gestae Divi Saporis 1: Deniers (dor) et (de compte) anciens, Syria, 38, pp. 261274. Guzmn, J.O. , 2014, Rex Armeniis Datus? Nero, Parthia and the Armenian Question, in B. Antela-Bernrdez & J. Vidal , eds, Central Asia in Antiquity: Interdisciplinary Approaches, BAR International Series 2665, Oxford: Archaeopress, pp. 8598. Harper, K. , 2015, Pandemics and passages to Late Antiquity: Rethinking the Plague of c. 24970 described by Cyprian, Journal of Roman Archaeology, 28, pp. 223260. Harper, K. , 2017, The Fate of Rome: Climate, Disease, and the End of Empire, Princeton & Oxford: Princeton University Press. Hartmann, U. , 2008, Claudius Gothicus und Aurelianus, in K.P. Johne , U. Hartmann , & T. Gerhardt , eds, Die Zeit der Soldatenkaiser: Krise und Transformation des Rmischen Reiches im 3. Jahrhundert n. Chr.(235284), pp. 297323. Hartmann, U. , 2008, Das palmyrenische Teilreich, in K.P. Johne , U. Hartmann , & T. Gerhardt , eds, Die Zeit der Soldatenkaiser: Krise und Transformation des Rmischen Reiches im 3. Jahrhundert n. Chr.(235284), pp. 343378. Hauser, S.R. , 2012, The Arsacid (Parthian) Empire, in D.T. Potts , ed, A Companion to the Archaeology of the Ancient Near East, Vol. 2, Maldon & Oxford: Wiley-Blackwell, pp. 10011020. Hauser, S.R. , 2013, Where is the man of Hadr, who once built it and taxed the land by the Tigris and the Chaboras? On the significance of the final siege of Hatra, in L. Dirven , ed., Hatra: Politics, Culture and Religion between Parthia and Rome, Wiesbaden: Franz Steiner Verlag, pp. 119139. Henning, W.B. & S.H. Taqizadeh , 1957, The dates of Manis life, Asia Major, 6(1), pp. 106121. Henning, W.B. , 1939, The great inscription of pr I, Bulletin of the School of Oriental and African Studies, 9, pp. 823849. Herrmann, G. & R. Howell , 1980, The Sasanian Rock Reliefs at Bishapur 1, Bishapur III, Triumph Attributed to Shapur I, Iranische Denkmler 9, Berlin: Dietrich Reimer Verlag. Herrmann, G. & R. Howell , 1981, The Sasanian Rock Reliefs at Bishapur 2, Bishapur IV, Bahram II receiving a delegation, Bishapur V, The Investiture of Bahram I, Bishapur VI, The Enthroned King, Iranische Denkmler 10, Berlin: Dietrich Reimer Verlag Herrmann, G. & R. Howell , 1983, The Sasanian Rock Reliefs at Bishapur 3, Bishapur I, The Investiture/Triumph of Shapur I and Sarab-i Bahram, Bahram II enthroned; The Rock Relief at Tang-i Qandil, Iranische Denkmler 11, Berlin: Dietrich Reimer Verlag. Herrmann, G. & V. Curtis , 2002, Sasanian Rock Reliefs, Encyclopaedia Iranica, online edition, 2002, available at http://www.iranicaonline.org/articles/sasanian-rock-reliefs (accessed on 15 December 2019 ). 265 Herrmann, G. , 1969, The Darabgird relief: Ardashir or Shapur, Iran, 7, pp. 6388. Herrmann, G. , 1977, The Iranian Revival, Oxford: Phaidon. Herrmann, G. , 1998, Shapur I in the East: Reflections from his victory reliefs, in V.S. Curtis , R. Hillenbrand , & J.M. Rogers , eds., The Art and Archaeology of Ancient Persia, New Light on the Parthian and Sasanian Empires, London: I.B. Tauris, pp. 3851. Herrmann, G. , D.N. Mackenzie & R. Howell , 1989, The Sasanian Reliefs at Naqsh-i Rustam, Naqsh-i Rustam 6, The Triumph of Shapur I, Iranische Denkmaler 13, Berlin: Dietrich Reimer Verlag. Hewsen, R.H. 2011, Artaxata, Encyclopaedia Iranica, online edition, available at http://www.iranicaonline.org/articles/artaxata (accessed on 24 December 2019 ). Hewsen, R.H. , 19789, Introduction to Armenian Historical Geography I, Revue des tudes armniennes, New Series 13, pp. 7797. Hill, G.F. , 1922, Catalogue of the Greek Coins of Arabia, Mesopotamia and Persia, BMC 28, London: British Museum. Hillers, D.R. & E. Cussini , 1996, Palmyrene Aramaic Texts, Baltimore: Johns Hopkins University Press. Hitchins, K. , 2001/12, Georgia ii. History of Iranian-Georgian Relations, Encyclopaedia Iranica, available online at http://www.iranicaonline.org/articles/georgia-ii-history- (accessed on 20 December 2019 ).

Honigmann, E. & A. Maricq , 1953, Recherches sur les Res Gestae Divi Saporis, Bruxelles: Latomus. Hoyland, R. , (trans.), 2018, The History of the Kings of the Persians in Three Arabic Chronicles, Liverpool: Liverpool University Press. Hoyland, R. , 2001, Arabia and the Arabs, London & New York: Routledge. Huttner, U. , 2008, Von Maximinus Thrax bis Aemilianus, in K.P. Johne , U Hartmann , T Gerhardt , eds, Die Zeit der Soldatenkaiser: Krise und Transformation des Rmischen Reiches im 3. Jahrhundert n. Chr.(235284), Berlin: De Gruyter Akademie Forschung, pp. 161221. Huyse, P. , 1999, Die Dreisprachige Inschrift buhrs I. an Der Kaba-i Zardut (KZ), Vol. 2, London: School of Oriental and African Studies. James, S. , 2004, The Excavations at Dura-Europos 19281937, Final Report VII: The Arms and Armour and other Military Equipment, London: British Museum. James, S. , 2011, Stratagems, combat, and Chemical Warfare in the Siege mines of DuraEuropos, American Journal of Archaeology, 115(1), pp. 69101. James, S. , 2013, Roman Partho-Sasanian martial relations: Testimony of a cheekpiece from Hatra and its parallels, in L. Dirven , ed, Hatra: Politics, Culture and Religion between Parthia and Rome, Wiesbaden: Franz Steiner Verlag, pp. 217234. James, S. , 2019, The Roman Military Base at Dura-Europos, Syria: An Archaeological Visualization, Oxford: Oxford University Press. Johnson, M.J. , 2008, Observations on the Burial of the Emperor Julian in Constantinople, Byzantion, 78, pp. 254260. Keaveney, A. , 1981, Roman Treaties with Parthia c.95-c.64 BC, American Journal of Philology, 102, pp. 195212. Kelly, G.A.J. , 2008, Ammianus Marcellinus: The Allusive Historian, Cambridge: Cambridge University Press. Kennedy, D. & D. Riley , 1990, Romes Desert Frontier from the Air, Austin: University of Texas Press. 266 Kettenhofen, E. , 1982, Die rmisch-persischen Kriege des 3. Jahrhunderts n. Chr. Nach der Inschrift Shpuhrs I. an der Ka be-ye Zartot (SKZ), Wiesbaden: TAVO. Kettenhofen, E. , 2004, Trajan, Encyclopaedia Iranica, available online at http://www.iranicaonline.org/articles/trajan (accessed on 30 December 2019 ). Konrad, M. , 2001, Resafa V: Der sptromische Limes in Syrien, Mainz: P. von Zabern. Krner, C. , 2002, Philippus Arabs, Berlin & New York: De Gruyter. Kreucher, G. , 2008, Probus und Carus, in K.P. Johne , U. Hartmann , & T. Gerhardt , eds, Die Zeit der Soldatenkaiser: Krise und Transformation des Rmischen Reiches im 3. Jahrhundert n. Chr.(235284), Berlin: De Gruyter Akademie Forschung, pp. 395423. Kreyenbroek, P.G. , 2013, Zoroastrianism under the Sasanians, in K. Rezania , ed., Teachers and Teachings in the Good Religion: Opera Minora on Zoroastrianism, Wiesbaden: Harrassowitz Verlag, pp. 1950. Kropff, A. , 2016, New English translation of the Price Edict of Diocletianus, available at http://www.academia.edu/23644199/New_English_translation_of_the_Price_Edict_of_Diocletia nus (accessed on 30 December 2019 ). Kropp, M. , 1993, Vassal - Neither of Rome nor of Persia. Mar-al-Qays The Great King of the Arabs, Proceedings of the Seminar for Arabian Studies, 23, pp. 6393. Kuwlikowski, M. , 2016, Imperial Triumph: The Roman World from Hadrian to Constantine, London: Profile Books. Laubscher, H.P. , 1975, Die Reliefschuck des Galeriusbogens in Thessaloniki, Archeologische Forschungen 1, Berlin: Mann. Lauffray, J. , 1983, Halabiyya-Zenobia Place Forte du Limes Oriental et La Haute-Msopotamie au Vie sicle, Vol. 2, Paris: Geuthner. Leadbetter, W.L. , 2009, Galerius and the Will of Diocletian, London & New York: Routledge. Lee, A.D. , 2013, Roman warfare with Sasanian Persia, in B. Campbell & L.A. Tritle , eds, The Oxford Handbook of Warfare in the Classical World, Oxford: Oxford University Press, pp. 708725. Lenssen, J. , 1999, The Persian Invasion of 359: Presentation by suppression in Ammianus Marcellinus Res Gestae 18.4.118.6.7, in J. Willem Drijvers & D. Hunt , eds, The Late Roman World and Its Historian: Interpreting Ammianus Marcellinus, London & New York: Routledge, pp. 3745. Lepper, E.A. , 1948, Trajans Parthian War, Oxford: Oxford University Press. Levy, R. (trans.), 1967, The Epic of the Kings: Shah-nama by Ferdowsi, London: Routledge & Kegan Paul. Lieu, S.N.C. & D. Montserrat , 1996, From Constantine to Julian: Pagan and Byzantine Views, New York & London: Routledge.

Lieu, S.N.C. , 1985, Manichaeism in the Later Roman Empire and Medieval China, Manchester: University of Manchester Press. Lieu, S.N.C. , 1986, The Emperor Julian: Panegyric and Polemic, Liverpool: Liverpool University Press. Lieu, S.N.C. , 1986, Captives, Refugees and Exiles: A study of cross-frontier civilian movements and contacts between Rome and Persia from Valerian to Jovian, in P. Freeman & D. Kennedy , eds, The Defence of the Roman and Byzantine East, BAR, S297, Oxford: British Institute of Archaeology in Ankara, pp. 475505. 267 Lightfoot, C.S. , 1986, Tilli: A Late Roman equites fort on the Tigris, in P. Freeman & D. Kennedy , eds, The Defence of the Roman and Byzantine East, Oxford: British Archaeological Reports, pp. 509529. Lightfoot, C.S. , 1990, Trajans Parthian war and the fourth century perspective, Journal of Roman Studies, 80, pp. 115126. Lightfoot, C.S. , 2005, Armenia and the Eastern Marches, in A.K. Bowman , A.M. Cameron & P. Garnsey , eds, The Cambridge Ancient History, Vol. XII: The Imperial Crisis and Recovery AD 193324, Second edition, Cambridge: Cambridge University Press, pp. 481520. Littman, R.J. & Littman, M.L. , 1973, Galen and the Antonine Plague, American Journal of Philology, 94(3), pp. 243255. Loriot, X. , Les premires annes de la grande crise de IIIe sicle: De lavnement de Maximin le Thrace (235) la mort de Gordien III (244), ANRW II/2, Berlin: De Gruyter, pp. 657787. Luttwak, E. , 1976, The Grand Strategy of the Roman Empire, Baltimore: Johns Hopkins University Press. MacCormack, S. , 1990, Art and Ceremony in Late Antiquity, Berkeley: University of California Press. Marciak, M. , 2014, The Cultural Landscape of Sophene from Hellenistic to Early Byzantine Times, Gttinger Forum fr Altertumswissenschaft, 17, pp. 1356. Marciak, M. , 2017, Sophene, Gordyene, and Adiabene: Three Regna Minora of Northern Mesopotamia between East and West, Boston & Leiden: Brill. Matthews, J.F. , 1986, Ammianus and the Eastern frontier in the fourth century, in P. Freeman & D. Kennedy , eds, The Defence of the Roman and Byzantine East, BAR S297, Oxford: British Institute of Archaeology at Ankara, pp. 549564. Matthews, J.F. , 1989, The Roman Empire of Ammianus, London: Duckworth. Mattingly, H. & E. Sydenham , 1926, Roman Imperial Coinage, Vol. II: Vespasian to Hadrian, London: Spink. Mattingly, H. & E. Sydenham , 1968, The Roman Imperial Coinage Vol. IV(3): Gordian III Uranius Antoninus, London: Spink. Mattingly, H. , 1966, Coins of the Roman Empire in the British Museum, Vol. 3: Nerva to Hadrian, London: British Museum. Mayor, A. , 2010, The Poison King: The Life and Legend of Mithradates, Princeton & Oxford: Princeton University Press. McDonough, S. , 2013, Military and society in Sasanian Iran, in B. Campbell & L.A. Tritle , eds, The Oxford Handbook of Warfare in the Classical World, Oxford: Oxford University Press, pp. 601620. Mennen, I. , 2011, Power and Status in the Roman Empire: 193284, Leiden: Brill. Meyer, J.C. , 2017, Palmyrena: Palmyra and the Surrounding Territory from the Roman to the Early Islamic Period, Oxford: Archaeopress. Millar, F.G.B. , 1969, P. Herennius Dexippus: The Greek World and the third-century invasions, Journal of Roman Studies, 59, pp. 1229. Millar, F.G.B. , 1993, The Roman Near East: 31 BC AD 337, Cambridge, MA: Harvard University Press. Mitford, T.B. , 1974, Some inscriptions from the Cappadocian Limes, JRS, 64, pp. 160175. Mitford, T.B. , 1977, Cappadocia and Armenia Minor, ANRW II.7.2, Berlin: De Gruyter, pp. 11941209. 268 Mori, S. , 1995, The narrative structure of the Paikuli inscription, Orient, 30(1), pp. 182193. Morley, C. , 2016, Beyond the digression: Ammianus Marcellinus on the Persians, Journal of Ancient History and Archaeology, 3(4), pp. 1025. Morley, C. , 2017, The Arabian frontier a keystone of the Sasanian Empire, in E.W. Sauer , ed., Sasanian Persia: Between Rome and the Steppes of Eurasia, Edinburgh: Edinburgh University Press, pp. 268283. Mosig-Walburg, K. , 2009, Rmer und Perser vom 3. Jh. bis zum Jahr 363 n.Chr. Gutenberg: Computus Druck Satz & Verlag. Mousavi, A. & T. Daryaee , 2012, The Sasanian Empire: An archaeological survey, in D.T. Potts , ed., A Companion to the Archaeology of the Ancient Near East, Vol. 2, Malden, MA: Wiley-

Blackwell, pp. 10761094. Mouterde, R. , 19301, La Strata Diocletiana et ses bornes milliares, Mlanges de lUniversit SaintJoseph, 15, pp. 219233. Musil, A. , 1927, The Middle Euphrates: A Topographical Itinerary, New York: American Geographical Society. Napoli, J. , 2000, Les Remparts de la Forteresse de Djazla sur le Moyen-Euphrate, Syria, 77, pp. 117136. Nedergaard, J. , 1988, The four sons of Phraates IV in Rome, in T. Fischer-Hansen , ed, East and West: Cultural Relations in the Ancient World, Acta Hyperborea. Vol. 1, Copenhagen: Museum Tusculanum Press, pp. 102115. Nehme, L. , 2018, Introduction, in L. Nehme , ed., Report on the 2017 Season of the Medain Saleh Archaeological Project, pp. 516, https://hal.archives-ouvertes.fr/hal-01804965. Nixon, C.E.V. & B.S. Rodgers , 1994, In Praise of Later Roman Emperors: The Panegyrici Latini, Berkeley: University of California Press. Olbrycht, J.M. , 2012, The political-military strategy of Atabanus/Ardawn II in AD 3437, Anabasis, 3, Studia Classica et Orientalia, pp. 215237. Olbrycht, J.M. , 2016, Germanicus, Artabanos II of Parthia, and Zeno Artaxias in Armenia, Klio, 98(2), pp. 605633. Olbrycht, M.J. , 2016, Dynastic connections in the Arsacid Empire and the origins of the house of Ssn, in V.S. Curtis , E.J. Pendleton , M. Alram & T. Daryaee , eds, The Parthian and Early Sasanian Empires: Adaptation and Expansion, Oxford & Philadelphia: British Institute of Persian Studies, pp. 2335. Oppenheimer, A. , 1983, Babylonia Judaica in the Talmudic Period, Wiesbaden: TAVO. Overlaet, B. , 2009, A Himyarite diplomatic mission to the Sasanian Court of Bahram II depicted at Bishapur, Arabian Archaeology and Epigraphy, 20, pp. 218221. Overlaet, B. , 2009, A Roman Emperor at Bishapur and Darabgird: Uranius Antoninus and the black stone of Emesa, Iranica Antiqua, 44, pp. 461530. Overlaet, B. , 2017, Shapur I: Rock Reliefs, Encyclopaedia Iranica, online edition, available at http://www.iranicaonline.org/articles/shapur-I-rock-reliefs (accessed on 15 November 2019 ). Overtoom, N.L. , 2016, The Rivalry of Rome and Parthia in the sources from the Augustan age to Late Antiquity, Anabasis, 7, Studia Classica et Orientalia, pp. 137174. Palermo, R. , 2014, Nisibis, capital of the province of Mesopotamia: Some historical and archaeological perspectives, Journal of Roman Archaeology, 27, pp. 456471. 269 Palermo, R. , 2019, On the Edge of Empires: North Mesopotamia During the Roman Period (2nd 4th c. CE), London & New York: Routledge. Palmer, A. , 1990, Monk and Mason on the Tigris Frontier: The Early History of the Tur Abdin, Cambridge: Cambridge University Press. Panaino, A. , 2016, Kirdr and the Re-organisation of Persian Mazdeism, in V.S. Curtis , E.J. Pendleton , M. Alram & T. Daryaee , eds, The Parthian and Early Sasanian Empires: Adaptation and Expansion, Oxford & Philadelphia: British Institute of Persian Studies, pp. 5360. Parker, S.T. , 2006, The Roman Frontier in Central Jordan: Final Report on the Limes Arabicus Project 19801989, 2 Volumes, Washington, DC: Dumbarton Oaks. Parker, S.T. , 2015, Strata Diocletiana, in Y. LeBohec , ed., Encyclopedia of the Roman Army, Malden, MA: Wiley-Blackwell, pp. 929933. Parkin, T.G. , 1992, Demography and Roman Society, Baltimore, MD: Johns Hopkins University Press. Patterson, L. , 2015, Antony and Armenia, Transactions of the American Philological Association, 145, pp. 77105. Payne, R.E. , 2015, A State of Mixture: Christians, Zoroastrians and Iranian Political Culture in Late Antiquity, Berkeley: University of California Press. Peeters, P. , 1924, S. Demetrianus eveque dAntioche?, Analecta Bollandiana, 42, pp. 288315. Pkary, T. , 1961, Le tribut aux Perses et les finances de Philippe lArabe, Syria, 38, pp. 275283. Pfeiler, I. , 1973, Drei westliche Munzstatten unter Sapur I anhand zweier Funde, Schweitzer Munzblatter, 23, pp. 129145. Piacentini, F. , 1985, Ardashr I Ppakn and the wars against the Arabs: Working hypothesis on the Sasanian hold of the Gulf, Proceedings of the Seminar for Arabian Studies, 15, pp. 5777. Pink, K. , 1935, Antioch or Viminacium, Numismatic Chronicle, 15(38), pp. 94113. Poidebard, A. , 1934, La Trace de Rome dans le Dsert de Syrie, 2 Volumes Paris: Geuthner. Pond-Rothman, M. , 1977, The Thematic Organization of the panel reliefs on the Arch of Galerius American Journal of Archaeology, 80, pp. 427454. Potter, D.S. 1990, Prophecy and History in the Crisis of the Roman Empire, Oxford: Oxford University Press.

Potter, D.S. 2004, The Roman Empire at Bay: AD 180395, London & New York: Routledge. Potts, D.T. , 1990, The Arabian Gulf in Antiquity, Vol. 1, Oxford: Oxford University Press. Potts, D.T. , 1997, The Roman relationship with the Persicus sinus from the rise of Spasinous Charax (127 BC) to the reign of Shapur II (AD 309379), in S. Alcock , ed., The Early Roman Empire in the East, Oxford: Oxbow, pp. 89107. Potts, D.T. , 2008, The Sasanian relationship with South Arabia: Literary, epigraphic and oral historical perspectives, Studia Iranica, 37, pp. 197213. Pourshariati, P. , 2008, Decline and Fall of the Sasanian Empire: The Sasanian-Parthian Confederacy and the Arab Conquest of Iran, London: I.B. Tauris. Preudhomme, N.J. , 2018, 261 CE Revolution in Caucasian Iberia, I. Javakhishvili , ed., Proceedings of the Institute of Georgian History, XIV, Tbilisi, GE: Tbilisi State University, pp. 4266. 270 Rapp, S.H. , 2017, The Sasanian World through Georgian Eyes: Caucasia and the Iranian Commonwealth in Late Antique Georgian Literature, London & New York: Routledge. Reiner, E. , 2006, The Reddling of Valerian, Classical Quarterly, 56(1), 325329. Rezakhani, K. , 2018, ReOrienting the Sasanians: East Iran in Late Antiquity, Edinburgh: Edinburgh University Press. Robin, C. , 1984, La Premiere Intervention Abyssine en Arabie Meridionale (de 200 270 de lre chrtienne environ, in T, Beyene , ed., Proceedings of the Eighth International Conference of Ethiopian Studies, Vol. 2, Addis Ababa: University of Addis Ababa, pp. 147162. Robin, C. , 2015, Himyar, Aksum and Arabia Deserta in Late Antiquity, in G. Fisher , ed., Arabs and Empires before Islam, Oxford: Oxford University Press, pp. 127171. Rohrbacher D. , 2002, The Historians of Late Antiquity, London & New York: Routledge. Rose, C.B. , 2005, The Parthians in Augustan Rome, American Journal of Archaeology, 109(1), pp. 2175. Ross, S.K. , 2001, Roman Edessa, London & New York: Routledge. Rostovtzeff, M.I. , A. Bellinger , C. Hopkins & C.B. Welles , eds, 1936, The Excavations at Dura Europos: Preliminary report of the Seventh and Eighth Seasons: 19331934 and 19341935, New Haven: Yale University Press. Rostovtzeff, M.I. , A. Bellinger , F. Brown & C.B. Welles , eds, 1944, The Excavations at Dura Europos: Preliminary Report of the Ninth Season, 19356, Part 1: The Agora and Bazaar, New Haven: Yale University Press. Rubin, Z. , 1998, The Roman Empire in the Res Gestae Divi Saporis , in E. Dabrowa , ed., Ancient Iran and the Mediterranean World, Krakow: Jagiellonian University Press, pp. 177185. Rubin, Z. , 2000, The Sasanid Monarchy, in A. Cameron , B. Ward-Perkins & M. Whitby , eds., The Cambridge Ancient History, Vol. 14: Late Antiquity: Empire and Successors, AD 425600, Cambridge, Cambridge University Press, pp. 638661. Sartre, M. , 2005, The Arabs and desert peoples, in A.K. Bowman , A.M. Cameron & P. Garnsey , eds, The Cambridge Ancient History, Vol. XII: The Imperial Crisis and Recovery AD 193324, Second edition, Cambridge: Cambridge University Press, pp. 498520. Sartre, M. , 2005, The Middle East under Rome, Cambridge, Mass.: Harvard University Press. Schiettecatte, J. , 2016, The political map of Arabia and the Middle East in the third century AD revealed by a Sabaean inscription, Arabian Archaeology and Epigraphy, 27(2), pp. 176196. Schippmann, K. , 1986, Arsacids ii. The Arsacid dynasty, Encyclopaedia Iranica, II/5, pp. 525536, available online at http://www.iranicaonline.org/articles/arsacids-ii (accessed on 30 December 2019 ). Schippmann, K. , 1990, Grundzge der Geschichte des sasanidischen Reiches, Darmstadt: Wissenschaftliche Buchgesellschaft. Schlude, J.M. & B.R. Rubin , 2017, Finding common ground: Roman-Parthian embassies in the Julio-Claudian period, in J.M. Schlude & B.R. Rubin , eds, Arsacids, Romans and Local Elites: Cross-cultural Interactions of the Parthian Empire, Oxford & Philadelphia: Oxbow Books, pp. 6591. 271 Schneider, R.M. , 2007, Friend and Foe: The orient in Rome, in V.S. Curtis & S. Stewart , eds, The Age of the Parthians, The Idea of Iran, Vol. 2, London: I.B. Tauris, pp. 5086. Segal, J.B. , 1970, Edessa: The Blessed City, Oxford: Oxford University Press. Sellwood, D. , 1989/2011, Amida, Encyclopedia Iranica, I/9, p. 938, available online at http://www.iranicaonline.org/articles/amida-pers (accessed on 30 December 2019 ). Sellwood, D. , 1995, The Victory drachms of Phraates IV, American Journal of Numismatics, 7(8), pp. 7581. Shahbazi, A.S. , 2002/17, pur I: History, Encyclopaedia Iranica, online edition, available at http://www.iranicaonline.org/articles/shapur-i (accessed on 24 August 2019 ). Shahbazi, A.S. , 1988/2016, Bahrm II, Encyclopaedia Iranica, III/5, pp. 514522, available online at http://www.iranicaonline.org/articles/bahram-02 (accessed on 30 December 2019 ).

Shahbazi, A.S. , 1990, Carrhae, Encyclopedia Iranica, V/1, pp. 913, available online at http://www.iranicaonline.org/articles/carrhae-harran-town-in-mesopotamia-where-in-may-53-b (accessed on 30 December 2019 ). Shahbazi, A.S. , 2004/12, Hormozd II, Encyclopaedia Iranica, online edition, available at http://www.iranicaonline.org/articles/hormozd-ii (accessed on 27 December 2019 ). Shahbazi, A.S. , 2005, Sasanian Dynasty, Encyclopaedia Iranica, online edition, available at http://www.iranicaonline.org/articles/sasanian-dynasty (accessed on 27 December 2019 ). Shahd, I. , 1984, Byzantium and the Arabs in the Fourth Century, Washington, DC: Dumbarton Oaks. Shahd, I. , 1984, Rome and the Arabs: Prologomenon to the Study of Byzantium and the Arabs, Washington, DC: Dumbarton Oaks. Sharankov, N. & I. Hristov , 2019, A milestone of Emperor Philip the Arab from the road Oescus Philippopolis found at the Eastern wall of the castellum of Sostra, Archaeologia Bulgarica, III(2), pp. 5770. Shavarebi, E. , 2019, Sasanians, Arsacids, Aramaeans: Ibn al-Kalbs account of Ardashrs Western campaign, in P.B. Lurje , ed., Proceedings of the Eighth European Conference of Iranian Studies, St. Petersburg: State Hermitage Publishers, pp. 363378. Shayegan, M.R. , 2011, Arsacids and Sasanians: Political Ideology in post-Hellenistic and Late Antique Persia, Cambridge: Cambridge University Press. Shayegan, M.R. , 2012, Hormozd I, Encyclopedia Iranica, available online at http://www.iranicaonline.org/articles/hormozd-i (accessed on 18 December 2019 ). Sherwin-White, A.N. , 1984, Roman Foreign Policy in the East: 168 BC to AD 1, London: Duckworth. Sidebotham, S.E. , 2011, Berenike and the Ancient Maritime Spice Route, Berkeley: University of California Press. Simelidis, C. , 2013, Eunapius of Sardis, in R.S. Bagnall , K. Brodersen , C.B. Champion , A. Erskine & S.R. Huebner , eds., The Encyclopedia of Ancient History, Malden, MA: Wiley Blackwell, pp. 25592560. Sinclair, T.A. 1987, Eastern Turkey: An Architectural and Archaeological Survey Volume I, London: Pindar Press. 272 Sinclair, T.A. 1989, Eastern Turkey: An Architectural and Archaeological Survey Volume III, London: Pindar Press. Skjaervo, P.O. , 1983, The Sassanian Inscription of Paikuli, Parts 3.1 and 3.2, Wiesbaden: Reichert. Smith, K.R. , 2014, The Martyrdom and History of Blessed Simeon Bar Sabbae, Piscataway: Gorgias Press. Smith, K.R. , 2016, Constantine and the Captive Christians of Persia, Oakland: University of California Press. Sommer, M. , 2013, In the twilight. Hatra between Rome and Iran, in L. Dirven , ed, Hatra: Politics, Culture and Religion between Parthia and Rome, Wiesbaden: Franz Steiner Verlag, pp. 3344. Sommer, M. , 2018, Palmyra: A History, London & New York: Routledge. Sprengling, M. , 1940, Shahpur I, the Great on the Kaabah of Zoroaster, American Journal of Semitic Languages and Literature, 57, pp. 341429. Sprengling, M. , 1953, Third Century Iran: Sapor and Kartir, Chicago: Oriental Institute. Stoneman, R. , 1994, Palmyra and its Empire, Ann Arbor: University of Michigan Press. Sutherland, C.H.V. , 1955, Diocletians reform of the coinage: A chronological note, Journal of Roman Studies, 45, pp. 116118. Syme, R. , 1971, Emperors and Biography, Oxford: Oxford University Press. Syvnne, I & K. Maksymiuk , 2018, The Military History of Third Century Iran, Siedlce: Siedlce University of Natural Sciences and Humanities. Talbert, R.J.A. ed., 2000, The Barrington Atlas of the Greek and Roman World, Princeton: Princeton University Press. Talbert, R.J.A. , 2010, Romes World: The Peutinger Map Reconsidered, Cambridge: Cambridge University Press. Thomsen, P. , 1917, Die rmischen Meilensteine der Provinzen Syria, Arabia, und Palstina, Zeitschrift des Deutschen Palstina-Vereins, 40, pp. 1103. Toumanoff, C. , 1969, The Third-Century Armenian Arsacids: A Chronological and Genealogical Commentary, Paris: Librairie C. Klincksieck. Trombley, F. , 1999, Ammianus Marcellinus and fourth-century warfare: A protectors approach to historical narrative, in J. Willem Drijvers & D. Hunt , eds, The Late Roman World and Its Historian: Interpreting Ammianus Marcellinus, London & New York: Routledge, pp. 1626.

Trmpelmann, L. , 1975, Das Sasanidische Felsrelief von Drb, Iranische Denkmler Lieferung 6, enthaltend Reihe 2, Iranische Felsreliefs B, Berlin: Dietrich Reimer Verlag. Vitale, M. , 2013, Kolchis in der Hohen Kaiserzeit: Rmische Eparchie Oder Nrdlicher Aussenposten des Limes Ponticus? Historia, 62(2), pp. 241258. Von Gall, H. , 1998, The representation of foreign peoples on the rock relief of Bishapur II, in V.S. Curtis , R. Hillenbrand , and J.M. Rogers , eds., The Art and Archaeology of Ancient Persia, New Light on the Parthian and Sasanian Empires, London: Bloomsbury, pp. 5257. Warmington, B.H. , 1977, Objectives and strategy in the Persian war of Constantius II, in J. Fitz , ed., Limes: Akten des XI internationalen Limeskongresses, Budapest: Akadmiai Kiad, pp. 509520. Warmington, B.H. , 1981, Ammianus Marcellinus and the Lies of Metrodorus, Classical Quarterly, 31, pp. 464468. 273 Watson, A. , 1999, Aurelian and the Third Century, London & New York: Routledge. Weber, U. , 2012, Narseh, Knig der Knige von rn und Anrn, Iranica Antiqua, 47, pp. 153302. Weber, U. , 2016, Narseh, Encyclopaedia Iranica, online edition, available online at http://www.iranicaonline.org/articles/narseh-sasanian-king (accessed on 15 November 2019 ). Weisweiler, J. , 2009, Christianity in war: Ammianus on power and religion in Constantius Persian war, A. Cain & N. Lenski , eds, The Power of Religion in Late Antiquity, Burlington: Ashgate, pp. 383396. Welles, C.B. , 1941, The Epitaph of Julius Terentius, Harvard Theological Review, 34, pp. 79102. Welles, C.B. , R.O. Fink & J.F. Gilliam , 1959, The Excavations at Dura Europos, Final Report Volume V, Part 1: The Parchments and Papyri, New Haven: Yale University Press. Whitehouse, D. & A. Williamson , 1973, Sasanian Maritime Trade, Iran, 11, pp. 2949. Widengren, G. , 1965, Mani and Manichaeism, C. Kessler (trans.), London: Weidenfeld and Nicolson. Wiesehfer, J. , 2001, Ancient Persia, London: I.B. Tauris. Winter, E. & B. Dignas , 2001, Rom und das Perserreich. Zwei Weltmchte zwischen Konfrontation und Koexistenz, Berlin: Akademie Verlag. Winter, E. , 1988, Die ssnidisch-rmischen Friedensvertrge des 3. Jahrhunderts n. Chr. Ein Beitrag zum Verstndnis der Aussenpolitischen Beziehungen zwischen den beiden Grossmchten, Frankfurt: P. Lang. Winter, E. , 2013, Persia and Rome, in R.S. Bagnall , K. Brodersen , C.B. Champion , A. Erskine & S.R. Huebner , eds., The Encyclopedia of Ancient History, Malden, MA: Wiley Blackwell, pp. 51875191. Wood, P. , 2013, The Chronicle of Seert: Christian Historical Imagination in Late Antique Iraq, Oxford: Oxford University Press. Yarshater, E. , 1971, Were the Sasanians Heirs to the Achaemenids?, in Atti del convegno internazionale sul tema:La Persia Nel Medioevo, Rome: Accademia Nazionale dei Lincei, pp. 517531. Young, G. , 2001, Romes Eastern Trade, London & New York: Routledge. Zouberi, J. , 2017, The role of religion in the foreign affairs of Sasanian Iran and the Later Roman Empire (330630 A.D.), Historia i wiat, 6, pp. 121132.

2

Rome and Parthia Conflict and diplomacy from Sulla to Caracalla

The long period of interaction between the Roman and Parthian Empires is important to consider in some detail before undertaking an analysis of the military and political relationship between Rome and the Sasanian Persian Empire. A number of key issues of contention between Rome and the Sasanians had their origins in the Parthian period and had continued to be of importance to conflict and diplomacy between the two powers across a period of more than three centuries. The most important of these was ongoing contention over Armenia, which went back to the very beginning of the relationship between Rome and Parthia. Roman acquisition of territory across the Euphrates in northern Mesopotamia in the second century was also a significant point of contention, which was clearly on display immediately after Ardashir came to power in the 220s. Roman invasions of Parthia in the second century not only resulted in territorial expansion but also became part of the imperial psyche and expectation. These invasions were still emulated by Roman emperors in the third and fourth centuries and were specifically invoked as examples to follow when invasions of Persia were in preparation or underway. The successes of Rome against the Parthians in the second century were undoubtedly exaggerated in the Roman texts but the capture of Seleucia-Ctesiphon three times in 80 years and the territorial gains that accompanied these successes engendered a certain arrogance in Roman imperial leadership towards the Parthians. As we will see in more detail in the next chapter, some Roman rhetorical representations of the Persians in the third and fourth centuries saw them rendered as Parthians who were simply rebels against their Macedonian overlords of whom Rome was the legitimate imperial successor. While such ideas could be promoted in panegyrics and other types of imperially sanctioned rhetoric, the reality of Sasanian Persia's military strength was undeniable as a consequence of Shapur I's invasions of the 250s. Rome's capacity to sponsor difficulties in the ruling dynasty, as it had done routinely with the Parthians, also changed markedly. Disaffected members of the family of the Sasanian dynasty were less prepared to appeal to Roman rulers for assistance when disputes turned into more serious divisions. This largely removed a mechanism by which Roman rulers were previously able to contribute to ongoing instability in the Parthian imperial leadership.

26

Rome and Parthia

The first meeting between Romans and Parthians During the latter part of Mithridates II's long reign as the Parthian monarch, the first direct diplomatic engagement between Parthia and the Romans took place on the Euphrates in either 96 or 95 BC (the date of 92 BC is now generally discounted) when a Parthian envoy, Orobazes, met with the Roman general Sulla and Ariobarzanes, the Roman nominee to the throne of Cappadocia. 1 Sulla had gone to Cappadocia to support the elevation of Ariobarzanes as a counterweight to concerns over the territorially expansive activity of Mithridates of Pontus and Tigranes of Armenia, which threatened Rome's position in Asia Minor and Cappadocia. According to a number of Roman texts, Orobazes requested amicitia between Parthia and Rome at the meeting with Sulla but was later executed on the orders of Mithridates II for appearing subordinate and too deferential to the Roman general. 2 Whether a foedus (treaty) was established between Parthia and Rome under Mithridates and Sulla, either as a result of the meeting or at some stage afterwards, hinges on a later claim that the Parthian king Oracles II sent envoys to Crassus in 53 BC in the lead up to the Battle of Carrhae reminding him of the f oedus struck ,.., earlier between Rome and Parthia under Sulla. J

Sinatruces and instability in Arsacid Parthian rulership Following the Euphrates meeting between Orobazes and Sulla, there is no evidence for any direct engagement between Parthia and Rome for approximately 30 years. For the Parthians, internal discord was already a problem in the latter years of Mithridates II's reign and it escalated following his death in 87 BC. The most pressing issue that occupied the Parthians on their Central Asian frontier came in the figure of Sinatruces who had emerged with the support of a tribal group known as the Sacaraucae (Sakas) and was making claims to the kingship of the Parthian Empire. The Sacaraucae had earlier been a problem for the Parthians during the reign of Phraates II who died in battle with them in 126 BC. 4 Both his brother and successor Artabanus I met the same fate a few years later. 5 Sinatruces was another brother of Phraates II and Artabanus I and in the latter years of the long reign of his nephew, Mithridates II, he emerged as a serious imperial contender. He had either been captured by the Sacaraucae during the battles of the late 120s or had fled to them during a purge of rival imperial family members on the succession of Mithridates II. Sinatruces' imperial lineage and the support he enjoyed from the Sacaraucae set off a serious and sustained division in Arsacid Parthian imperial rulership for most of the next four decades. Eventually, the imperial line descended from Sinatruces would win out after a series of civil wars and intra-dynastic murders. His grandson, Oracles II (57-37 BC), would finally bring the long period of instability in Parthian leadership to an end, marked by the fratricidal elimination of his brother in 54 BC.

Rome and Parthia 27 Tigranes of Armenia and Mithridates VI of Pontus The latter years of Mithridates II's reign also saw problems emerge in other parts of his vast empire and they were a further catalyst and cause of Sinatruces' actions. It would appear that Babylonia was under the control of another claimant to the Parthian throne, Gotarzes, for a number of years before Mithridates' death and on the north-western frontier, developments in Armenia started to become especially problematic. 6 Tigranes of Armenia, who had been a hostage at the Parthian court since the defeat of his father by Mithridates II in 105 BC, had traded a substantial portion of the eastern part of the kingdom (the so-called "Seventy Valleys") on his father's death in 95 BC in exchange for his repatriation from the Parthian court in order to succeed to the Armenian throne. 7 With an escalation in the activities of Sinatruces following Mithridates' death in 87 BC, Tigranes of Armenia took advantage of these problems and set about recovering the territories he had ceded to Mithridates in 95 BC. Tigranes went further than this and occupied large sections of northern Mesopotamia, including the key city of Nineveh. 8 In 83 BC, Tigranes effectively replaced the Seleucid king in Syria where he ruled via a governor based in Antioch. With some other military successes in the eastern Mediterranean and the Caucasus, Tigranes carved out a substantial regional empire, which was problematic for the Parthians and also caused concern for the Romans. The concerns for both the Parthians and the Romans in relation to Tigranes became more pronounced when he and Mithridates VI of Pontus formed an alliance. During Mithridates' first war with the Romans (88-84 BC), Tigranes provided a relatively low level of support for the Pontic king but in 69 BC, during what would become known as the Third Mithridatic War, Mithridates fled to Armenia after his defeat at the hands of the Roman general Lucullus. The two kings established a formal alliance, cemented by the marriage of Tigranes to the daughter of Mithridates. This was clearly an ominous development for both Parthia and Rome because Tigranes' strength at this time was considerable. 9

Lucullus, Pompey and Parthia The internal problems that plagued the Parthian Empire during the first half of the first century BC were known to the Romans, but the Roman focus during this period was more on Mithridates and Tigranes. According to Plutarch: the Parthian power was not so great as it proved to be (later) in the time of Crassus, nor was it so well united, rather, owing to intestine and neighbouring wars, it had not even strength enough to repel the wanton attacks of the Armenians. 10

28

Rome and Parthia

Lucullus, initially as consul in 74 BC and then as the governor of Cilicia and Asia, headed the Roman effort to deal with the expansionary activity of Mithridates and once the alliance with Tigranes had been established, Lucullus was also dealing with the Armenians. In 68 BC, Lucullus sent an envoy to the Parthian king, Phraates III (70-57 BC), requesting that the Parthians remain neutral, while Roman military activity was underway to subdue both Mithridates and Tigranes. 11 This clearly suited Phraates, given the ongoing problems with rival imperial claimants. The fact that the Romans were expending considerable resources dealing with Mithridates and Tigranes was also advantageous to the Parthians. Indeed, the alliance between Mithridates and Tigranes may well have been viewed by the Parthians as a positive development because it now made Tigranes a sworn enemy of Rome. 12

Pompey, Crassus and Carrhae The treacherous nature of Roman politics at the time saw Pompey successfully manoeuvre Lucullus out of the command of the Mithridatic war. Following his replacement of Lucullus, Pompey struck an alliance with Phraates III in 66 BC, which was aimed primarily at isolating Tigranes. 13 This arrangement was updated in a revised agreement in 64 BC, which was eventually approved by the Roman senate in 59 BC. 14 In the process of negotiating this agreement, Phraates requested that the Euphrates River be recognised as a boundary between Parthia and Rome. The request was made principally because one of Pompey's legates, Aulus Gabinius, had led a raid into northern Mesopotamia in 65 BC and this had caused concern for the Parthians.15 Phraates was informed by Pompey, with what is often suggested to be a certain degree of hubris, that a just boundary would be adopted. 16 It is important to bear in mind, however, that Pompey was strongly focussed on Mithridates and Tigranes at this time and that it was prudent not to strike an agreement with the Parthians that was too definitive, given the complex military and political situation. Pompey would also have been mindful of the complexities of Roman politics, which in theory left matters related to foreign policy to the senate. What does seem apparent is that the Parthians were beginning to focus on what the geo-political situation might look like after the Romans had dealt with Mithridates and Tigranes. Some of the details of Pompey's spectacular two-day triumph held in Rome in 62 BC are indicative of the extent to which Tigranes and Mithridates were seen from a Roman perspective as a more significant enemy and threat than the Parthians at the time. In his account of Pompey's triumph and summary of Pompey's achievements, Appian emphasised the length of the Mithridatic and Tigranian wars - they lasted for 42 years - and that the Romans expanded direct and indirect territorial control in the region as a consequence: In the same war that part of Cilicia which was not yet subject to them (the Romans), together with the Syrian countries, Phoenicia,

Rome and Parthia 29 Coele-Syria, Palestine and the country inland as far as the Euphrates, although they did not belong to Mithridates, were gained by impetus of the victory over him and were required to pay tribute, some immediately and others later. 17 In broad summation, Appian continued, "Such and so diversified was this one war; but in the end it brought the greatest gains to the Romans, for it pushed the boundaries of their dominion from the setting of the sun to the river Euphrates." 18 Plutarch provided a long list of conquered nations that featured in the triumph and Parthia was not among them because Pompey had not scored any direct victories over the kingdom. 19 A decade later, however, Parthian power had clearly become a concern for the Romans and Rome's success in dealing with Mithridates, and especially Tigranes, likely played a hand in this. Tigranes' removal as a problem for the Parthians on their northwestern frontier was one less problem the Parthian leadership needed to deal with. The dynastic divisions that had plagued the Parthian leadership since the mid-90s were also finally dealt with soon after Orodes II came to power in 57 BC. Rome played a part in this episode and it demonstrated the possibility of fostering instability in the Parthian leadership. Installing his brother, Mithridates, on the throne of Media on coming to power after murdering his own father, Orodes II turned on Mithridates soon after. Mithridates, as a consequence, fled to Aulus Gabinius, by then governor of Syria. 20 Gabinius provided military assistance to Mithridates on his return to Parthia and by 55 BC Mithridates had taken control of both Babylon and Seleucia. 21 The following year, however, Orodes II recaptured Seleucia and executed Mithridates soon afterwards. 22 This was an important episode because it established a precedent for Roman involvement in the complexity of Parthian royal politics that would be repeated many times in future. Crassus' infamous campaign against Parthia, which ended in disaster at Carrhae in 53 BC, needs to be analysed in the context of Mithridates' defeat at Seleucia. The battle at Carrhae took place within a short time of Orodes' defeat of Mithridates and given the length of time for the planning and execution of a large-scale invasion such as that of Crassus', it was likely begun to provide support to the besieged Mithridates further south on the Tigris. Crass us' invasion also aimed to take advantage of the internal division. Seleucia figured prominently in the rhetorical exchanges between Crassus and ambassadors of Orodes prior to Crass us' invasion. 23 It appears that the Parthians, under the generalship of the Surena, were successful against Mithridates just in time to be able to deal effectively with Crassus' invasion. While a Parthian army under the command of the Surena dealt with Crassus and his army on the plains of northern Mesopotamia, a larger Parthian force under Orodes II went to Armenia. Orodes succeeded in luring the Armenian king, Artavazdes, away from an alliance previously struck with Rome and married his son Pacorus into the Armenian royal family. 24

30

Rome and Parthia

The confidence of the Parthians in the wake of Carrhae clearly grew and Rome began to take the Parthians more seriously. 25 Schippmann believes that the Parthian victory at Carrhae "produced a mighty echo amongst the peoples of the East." 26 The third-century Roman author, Cassius Dio, claimed that the defeat of Crassus in 53 BC marked the beginning of Rome's war with the Parthians and a selection of Cicero's letters that are contemporary with events suggests that this was how the Romans conceived of the relationship with Parthia soon after Carrhae. Writing from Cilicia where he was the governor in September 51 BC, Cicero expressed concern about an imminent Parthian attack and when he wrote to Cato four months later, Cicero reported with a degree of alarm that a Parthian attack on Antioch was expected. 27 In another letter, this time to Atticus, written the following month, Cicero mentioned a rumour that Pompey was on his way to take command of the Parthian war and later in the same letter was in virtual panic about an imminent Parthian attack. 28

The civil wars of the late Roman Republic and the involvement of Parthia An important feature of the imperial rivalry between the two powers from this time was involvement in each other's political and dynastic squabbles and difficulties. The Romans had already done this with Orodes and his brother Mithridates before Carrhae and as the centuries unfolded similar events took place on many occasions. This involvement was often invited by one of the factional claimants to the Parthian throne and there were also times when Roman generals and politicians attempted to enlist Parthian aid against internal political rivals. Pompey sought Parthian aid against Julius Caesar in the lead up to the battle of Pharsalus, the final showdown between the two powerful Roman magnates that was fought in 48 BC. 29 Pompey again sought the assistance of the Parthians following his defeat at Pharsalus and Dio claimed that Pompey even contemplated fleeing to the court of Orodes II. 30 The ongoing nature of the rivalry and conflict between Rome and Parthia was further indicated by the plans of Julius Caesar to lead a Roman campaign against the Parthians at the time of his assassination. 31 Caesar was apparently in full preparation for the war and had already dispatched some of his troops for the purpose. Following Caesar's assassination, the Parthians provided support to the main conspirators, Brutus and Cassius, whose armies at the Battle of Philippi in 42 BC included Parthian troops. 32 Further reflective of ongoing Parthian involvement in the civil wars of the Romans during this period, Quintus Labienus, who had been sent as an envoy by Cassius to the Parthians, joined forces with Orodes after the defeat of Brutus and Cassius at Philippi. 33 A Parthian invasion of Roman Syria in 40 BC was led by Labienus and the heir to the Parthian throne, Pacorus, when they crossed the Euphrates with a large army and attacked

Rome and Parthia

31

the city of Apamea. This attack failed but Labienus rallied some of the local garrisons to the Parthian cause and defeated the governor of Syria, L. Decidius Saxa, who had been recently installed by Marcus Antonius. Pacorus then conquered much of the eastern Mediterranean coastline and hinterland through to Palestine. Labienus pursued Saxa, who he defeated and killed in Cilicia. In 39 BC, a Roman counter-attack under P. Ventidius Bassus in Asia Minor defeated Labienus, who was subsequently captured and executed, while in the following year Pacorus was killed in Syria. 34 This Parthian military action would remain one of the most threatening Parthian invasions of Roman territory until the middle of the second century AD. Phraates IV came to power in 37 BC as a consequence of the murder of his own father Orodes. He soon undertook a brutal purge of other family members, including his brothers and even his own son. With sections of the nobility also targeted, the ensuing instability encouraged Antonius to attack the Parthians in 36 BC when he marched into Atropatene with 113,000 men. 35 One of Antonius' motives was the return of the standards lost by Crassus and also those taken from Saxa by Labienus and Pacorus a few years earlier. While the rear of Antonius' army suffered a serious defeat, Antonius himself pressed on and began prosecuting a siege of the fortress of Phraaspa. 36 The siege was eventually abandoned and thousands of Roman and auxiliary soldiers died during the retreat due to the cold winter. 37 Antonius' losses against the Parthians effectively marked almost 20 years of Parthian success in blunting Roman imperial ambitions and involvement in Rome's internecine civil wars. During the final playing out of Rome's civil wars between Octavian and Antonius, regional kingdoms and principalities situated on the upper Euphrates River backed Antonius and received assistance from the Parthians in doing so. 38

The role and importance of Armenia between Rome and Parthia Of crucial importance to the Parthian and Roman Empires in the first century BC and for centuries onwards was their ongoing rivalry over the kingdom of Armenia. Pompey's subjugation of Tigranes II and subsequent treaty arrangement with him marked the beginning of a permanent rivalry between Parthia and Rome over control of the kingdom. With Tigranes' death in 55 BC, Armenia's loyalty to Rome after the old man's agreement with Pompey in 65 BC was less certain. Tigranes' son and heir Artavazdes demonstrated that he was either unable or unwilling to provide assistance to Crassus in 53 BC, mostly because the Parthian king Orodes II had occupied the kingdom. 39 In fact, Artavazdes' sister was married to Orodes' eldest son Pacorus soon after Crassus' defeat at Carrhae and, according to Plutarch, both kings were present at the famous performance of Euripides' Bacchae at the Armenian capital, Artaxata, when Crassus' head was gruesomely employed as a stage prop.

32

Rome and Parthia

In the longer term, however, Artavazdes managed to establish a degree of independence between the two powers and this would become a feature at times of the relationship between Rome and Parthia over Armenia in the following centuries. The geography and political complexity of Armenia meant that either Parthia or Rome needed to invest considerable resources in maintaining a military presence in the kingdom and this was not always possible. Armenia was successful at times in playing both powers off against each other as a result. With the primary focus on civil conflict within the Roman Empire from ea. 50 to 30 BC, and Parthian imperial attempts to take advantage of this, rivalry between Parthia and Rome in Armenia does not appear to have been a significant issue until 36 BC when Artavazdes supplied troops for Antonius' Parthian campaign. Their quick withdrawal, however, after the unsuccessful siege of Phraaspa drew Antonius' enduring anger and resulted in Artavazdes' capture and eventual execution in Egypt in 34 BC. 4 For much of the following decade, a tussle ensued between Rome and Parthia for domination of Armenia.

°

Phraates IV and Rome Following the defeat of Antonius at Actium in 31 BC and his subsequent murder in Egypt, Octavian made no immediate changes to his erstwhile foe's client arrangements in the east but it would not be long before an opportunity presented itself to take advantage of ongoing internal difficulties in Parthia. In 33 BC, an individual named Tiridates rebelled against Phraates IV (37-2 BC). The rebellion continued at a relatively low level over the following years but in 31 BC, both Phraates and Tiridates sent envoys to Octavian in the wake of his success at Actium. 41 Tiridates seized power in the following year, forcing Phraates to flee to the Scythians. 42 Phraates would return and eventually emerge victorious over Tiridates who fled to the Romans in Syria soon after May 27 BC. 43 Importantly, and for reasons not immediately obvious, Tiridates brought one of Phraates IV's sons with him, also known as Phraates, as a hostage. Soon afterwards, Tiridates returned to Parthian territory with Roman support, invading down the Euphrates and taking Phraates IV by surprise. He occupied Seleucia for a few months, minting coins describing himself as "Philoromaios" in May 26 BC. 44 Justin claimed that Tiridates would take the throne on behalf of Augustus if l1e was successful. 45 Phraates IV eventually succeeded in gaining the upper hand and Tiridates again fled to Augustus, which found resonance in the Res Gestae. 46 By August 26 BC, Phraates IV was minting coins again at Seleucia. 47 In 23 BC, Augustus agreed to return Phraates' son, who Tiridates had brought with him to Rome in 27 BC, on the understanding that prisoners and the standards lost by Crassus at Carrhae be returned to Rome. 48 Delayed responses from Phraates saw a Roman military build-up in Armenia

Rome and Parthia

33

and Augustus himself came to Syria in 20 BC as part of putting pressure on Phraates to agree to a settlement. 49 The suggestion was that the standards be returned or risk an invasion and Marcus Agrippa was a key player in the diplomacy. With an agreement finally reached, the return of the standards in 20 BC was represented as a military victory by the Romans. 50 Augustus' Res Gestae claimed that the Parthians were compelled to return the standards and the Augustan poets used the return of the standards to great rhetorical effect. 51 A considerable amount of artistic and sculptural evidence survives for the way in which the Augustan regime sought to retail the arrangement with Parthia as a victor y. 52 Coins advertising the return of the standards show a subservient Parthian kneeling and presenting a standard in supplication and the return of the standards is depicted as a central element on the breastplate of the Prima Porta statue of Augustus. 53 Thea Musa, a female Italian slave, was given as a gift by Augustus in partial thanks for Phraates IV's return of the standards. Musa bore the Parthian king a child, known colloquially as Phrataaces and formally as Phraates V, following his father's death. Phraates IV had a number of other wives despite wiping out his harem in 26 BC in the face of Tiridates' advance down the Euphrates, and sons and grandsons by these wives were later sent as hostages to Augustus as a means of guaranteeing friendship. 54 Musa appears to have been behind sending the royal hostages to Rome as a means of securing the succession for Phrataaces. The Parthian princes were prominent at Rome and treated well according to a number of accounts despite two of them dying in Rome. 55 Two others returned to Parthia, with one named Vonones becoming king in AD 6. The other, named Phraates, died soon after arriving in Parthia to contest the Parthian throne. 56 Whether Augustus planned it or not, the Parthian succession following Phraates IV's death in 2 BC was seriously affected by the presence and influence of Musa at the Parthian court and the release of the Parthian royal hostages from Rome.

Figure 2.1 Denarius, Rome Mint . Augustus. 19/18 BC. Obv- TVRPILIANVS III

VIR / FE-RON: Diademed and draped bust of Feronia to right. Rev CAESAR AVGVSTVS SIGN RECE: Parthian kneeling in submis sion to right, holding out a standard with a vexillum inscribed with X. RIC I Augustus 288. Courte sy Noble Numi smatics.

34

Rome and Parthia

The situation in Armenia also continued to be an important factor in relations between Rome and Parthia. In 20 BC, as part of the settlement between Augustus and Phraates, Tigranes, the brother of the Armenian ruler Artaxes, was requested to return from Rome where he was also being held as a hostage, to rule the kingdom. Tigranes would rule as Tigranes III until his death in 6 BC when Tigranes IV was installed without Roman consent and with considerable support from Parthia. 57 Augustus countered this move by declaring one of Tigranes III's brothers, Artavazdes, king of Armenia instead.

Parthia and the Julio-Claudian dynasty of imperial Rome In 2 BC, Phraates IV was poisoned in a likely conspiracy of Musa and Phrataaces with the latter now acceding to the Parthian throne as Phraates V (2 BC-AD 4). The new ruler sent an embassy to Augustus suggesting a renewal of peace, which included a request for the release of Phraates IV's sons. 58 Problems over Armenia, however, saw Augustus dispatch his grandson Caius in the same year as part of preparations for a Parthian campaign. This was in response to Phrataaces expelling Augustus' nominee, Artavazdes, and placing Tigranes IV on the Armenian throne. 59 Tigranes was clearly concerned about Caius' arrival and formally requested Augustus' recognition of his position as the Armenian king. The situation appears to have been defused when the request was accepted and Tigranes was formally invested by Caius in Syria. 60 With the heat taken out of the situation by way of this compromise, a formal meeting between Caius and Phrataaces took place on an island in the middle of the Euphrates River. Witnessed by Velleius Paterculus, the meeting was depicted as a dazzling spectacle with the two empires essentially represented in Velleius' account as equals. 61 Phrataaces and Musa did not last long and were expelled from Parthia in AD 4, their fates the subject of speculation but ultimately unknown. Orodes III (AD 4-6) succeeded to the throne briefly before Vonones I (AD 6-12), a son of Phraates IV who had been a hostage at Rome, became the Parthian king. The Res Gestae referred to the appointment of Ariobarzanes, who had also been a hostage at Rome, as the king of the Medes. 62 While these events allowed Augustus to boast of the supplication of the Parthians in their requests for the various royal hostages, they were all short-lived rulers and this was likely in part due to their links to Rome. 63 The emergence of Artabanus II (AD 10-38), following a prolonged period of rivalry with Vonones, saw stability return to the Parthian leadership. Armenia would continue to be an issue for Parthia and the empire's relationship with its Roman rival. When Cermanicus was sent by Tiberius on an eastern mission in AD 18 to secure a better arrangement regarding the appointment of the king of Armenia and also to oversee the annexation of the client-kingdoms of Cappadocia, Cilicia and Commagene, Artabanus

Rome and Parthia

35

II sent a delegation to Germanic us with the aim of renewing friendship. 64 An agreement between Germanicus and Artabanus whereby Zeno, the step-son of the Roman-client king of Cappadocia, was appointed to the Armenian throne settled imperial rivalry for the foreseeable future. Over a decade later, however, Artabanus took the Romans by surprise and successfully installed his son Arsaces on the Armenian throne. 65 In this process, Artabanus had developed a broader strategy aimed at consolidating power in Armenia and blunting the previous expansion of Roman power in the kingdom. 66 Both Tacit us and Dio mentioned that Artabanus sent an embassy to Tiberius, which spelled out a desire to restore the former western boundaries of the Achaemenid empire. 67 As discussed in later chapters, there is a scholarly debate on whether Parthian and Sasanian rulers actually made such claims or whether they are more indicative of Roman-era authors placing such claims in their mouths as part of a classicising approach to constructing and writing history. 68 It is also possible that Parthian and Sasanian rulers made such claims for rhetorical purposes only in their communications with Roman emperors and their representatives. Tiberius' overall approach to Parthia was essentially little different to that of Augustus and included continued attempts to take advantage of rivalries and intrigues within the Parthian dynasty. Rivals for the Parthian throne were clearly aware of the Roman approach and appeared in Roman territory, often in Syria, on a reasonably regular basis seeking support. Towards the end of his reign, Tiberius took advantage of instability in Parthia, which resulted in Artabanus II fleeing to the Scythians in the face of a rebellion at Seleucia-Ctesiphon. 69 Vitellius, then governor of Syria, supported the return of Tiridates to Seleucia-Ctesiphon where he was crowned in AD 36. Tiridates' rule in Parthia was short-lived as he was unable to gain a sufficient foothold, mostly due to the perception of foreign backing and in the following year Artabanus II was reinstated. With regard to Armenia, Tiberian policy introduced the Iberian kingdom as an important factor when he formed an alliance between Rome and the Caucasian kingdom and placed the Iberian king, Mithridates, on the Armenian throne. 70 This was designed to squeeze the Parthians in Armenia and a brief war between Parthia and Iberia ensued. Following Artabanus' restoration to the Parthian throne, however, Armenia appears to have come under a form of de-facto partition between Roman and Parthian interests. Rome came to terms with Artabanus in AD 37, reflected in another symbolic meeting on a bridge in the middle of the Euphrates. 71 The outcome of this meeting and the recognition of Artabanus by Rome, which the meeting represented, was in part a tacit recognition of Parthian power in Armenia. The diplomatic outcome allowed Tiberius and Vitellius a somewhat hollow boast of Parthian submission with one of Artabanus' sons, Darius, sent to Rome as a hostage. 72 Caligula (AD 3 7-41) eventually imprisoned Mithridates, the king of Iberia previously backed by Rome to rule in Armenia. 73 The purpose of this

36

Rome and Parthia

move was to allay Parthian concerns over Mithridates' intentions in the wake of the agreement of AD 37. Claudius (AD 41-54), however, reverted to the approach of Tiberius with Rome once again seeking to support a rival to the Parthian throne who would be sympathetic to Roman interests. Artabanus II's death in AD 38 set off another round of instability in the Parthian leadership. Three brothers essentially vied for the throne and it took most of the next decade for the second-eldest of the three, Gotarzes, to establish security in power. 74 Claudius firstly exploited this instability by reinstating the imprisoned Mithridates of Iberia in Armenia in AD 49. 75 The venture did not end well with Mithridates' murder at the hands of his nephew, Rhadamistus, in AD 51. Rhadamistus established himself on the Armenian throne to the chagrin, it seems, of both Rome and Parthia. With regard to the Parthian rulership, Claudius also sought to exploit the arrival in Rome in AD 48 of an unofficial delegation from Parthia requesting support to overthrow Gotarzes due to the excessive cruelty he had shown in consolidating power. The delegation requested that Meherdates, a grandson of Phraates IV, who was still resident in Rome, be supported by Rome to replace Gotarzes. 76 This venture did not end well either despite Meherdates' initial enjoyment of support from Abgar, king of Edessa, and the Parthian governor of Mesopotamia. 77 According to Tacitus, the support Meherdates enjoyed was illusory and quickly turned to treachery once he was back in Parthian territory. 78 In AD 49, Meherdates was defeated in battle by Gotarzes and treated brutally in reprisal. With the accession of Nero in AD 54, Roman attempts to cause instability in Parthian leadership came to a temporary end, partly because the new Parthian king, Vologaeses I (AD 51-78), quickly went about implementing reforms. Vologaeses' reign eventually saw the return of stability to Arsacid Parthian rule, which had an impact on the relationship with Rome, especially during the latter period of Nero's reign and the period of instability in Roman imperial leadership after Nero's death. 79 The main focus for both imperial powers was on the situation in Armenia and this would see direct conflict between the two powers over the kingdom. Soon after coming to power, Vologaeses set out to install his younger brother Tiridates on the Armenian throne. At that stage, Rhadamistus was still the king of Armenia. A struggle ensued that lasted the next three years and in the same year Nero came to power, Tiridates emerged as the king of Armenia. The response to the Armenian situation under the new emperor was swift with an invasion planned under the command of Cn. Domitius Corbulo.

Parthia, Nero and Armenia Any direct conflict between Rome and Parthia over Armenia was put on hold when Vologaeses was forced to deal with the revolt of his son, Vardanes. 80 The revolt lasted from AD 56 to 58, during which time Vologaeses sent familial hostages to Nero. This was designed to placate the Romans

Rome and Parthia

37

in the midst of dealing with the revolt and also to remove other potential rivals. 81 Vologaeses also had to deal with problems resulting from the expansion of the Kushan Empire into eastern Iran and the subsequent revolt of the Hyrcanians. 82 It was the latter development that gave Corbulo the upper hand as he prepared to invade Armenia and confront Tiridates. The Roman general captured and destroyed the Armenian capital, Artaxata, in AD 59 after which ambassadors from the Hyrcanians arrived with the aim of striking an alliance with Corbulo. 83 According to Tacit us, the ambassadors were sent on to Rome to deal directly with Nero before their return via the Red Sea and India. The following year, Tigranocerta capitulated to the forces of Corbulo and a Roman nominee, Tigranes VI, was invested with full military pomp at Tigranocerta. 84 Tigranes soon directed damaging attacks at Adiabene making it more vulnerable to a Roman attack. Corbulo remained in Armenia to support Tigranes VI over the following two years while Vologaeses and Tiridates made a number of attempts at recovering Armenia during this period. They were more successful after Corbulo's withdrawal to Syria to take up the governorship of the province in AD 62. Corbulo's successor, Lucius Caesennius Paetus, had been given orders to take direct control of Armenia but soon found himself in difficulty. Vologaeses and Tiridates placed Paetus' troops, located at the fortress of Rhandeia, under siege and despite Corbulo's response to an appeal for help from Paetus, the Roman general capitulated. 85 Once news of these events reached Rome, partly in the form of a letter from Vologaeses, which laid claim to Armenia by divine right, war was decided on and Corbulo was once again placed in command. 86 The rhetoric surrounding Corbulo's projected war swiftly dissipated, however, and a compromise with the Parthians was reached that was more favourable to them than might have been expected a year earlier. This was reflective of the reality of the gains Vologaeses and Tiridates had made at the hands of the hapless Paetus. The essential outcome was that Tiridates would be reinstated as king, but the investiture would take place in Rome and the ceremony be performed by Nero himself. At the formal conference at Rhandeia, both Vologaeses and the king of Adiabene gave hostages to Corbulo and Tiridates paid due reverence to Corbulo and an image of Nero. 87 Arrangements were made for the investiture of Tiridates by Nero in Rome and Nero celebrated a lavish triumph in celebration of what was advertised as a great victory. 88 The events were carefully stage-managed with both Nero and Tiridates paying due respect to one another. At the formal investiture ceremony, Tiridates, addressing Nero as a god, declared that he was the descendant of Arsaces, brother of kings Vologaeses of Parthia and Pacorus of Media and that he was Nero's slave. 89 In response, Nero declared Tiridates to be king of Armenia and reminded him that as emperor, he had the power to both bestow kingdoms and take them away. It seems that the deference of both sides towards each other was in some way lasting with Vologaeses' successor, Vologaeses II (AD 77-80), sending

38

Rome and Parthia

ambassadors to Rome on the occasion of his succession and requesting Nero's memory be honoured some nine years after Nero's death. 90 The treaty of Rhandeia would serve as a benchmark for relations between Rome and Parthia over Armenia in the decades to come but as a compromise it was not always going to last.

Vespasian, Armenia and Parthia It is perhaps surprising that the treaty of Rhandeia held until the reign of Trajan. If anything, it was a less than satisfactory arrangement between the two powers and it essentially took on the broader role of a peace agreement between Rome and Parthia. The survival of the arrangement can be initially explained by the period of internal instability surrounding the latter half of Nero's reign and the eventual emergence of Vespasian as emperor, but it was also due to the Roman preoccupation with the Jewish war, which was not fully settled until ea. AD 73. It is also possible to discern a distinct change in long-term Roman imperial policy during the Flavian period. The failure from a Roman perspective of previous arrangements in Armenia and the compromise with the Parthians that the Rhandeia agreement represented was a contributor to the Romans losing faith in the client-king arrangements in the Near East. Another, and perhaps the most significant, was the failure of the arrangements in Judaea, which had played out in the Jewish war. Most of the Near Eastern client-king arrangements, which had by-and-large held since the days of Pompey and through much of the Julio-Claudian period, were brought to an end under the Flavians. The client-kingdoms of Commagene, Chalcis and Emesa were subsumed into Roman provincial territory and Judaea was thoroughly reorganised in the wake of the war. 91 One of the most significant concerns for the Romans was the ability of the Parthians to interfere with client relationships. As an example, Josephus indicated that Commagene was annexed because its king, Antiochus, had colluded with the Parthians even after recognising Vespasian as Augustus and also because its principal city of Samosata was of potential strategic significance to the Parthians. 92 This approach extended through to the early reign of Trajan with the annexation of the N abataean kingdom and its incorporation into the province of Arabia in AD 106. Concerns over the reality and potential threat of Parthian power in Armenia together with the eradication of client-kingdoms and ongoing Roman imperialist attitudes saw a major military build-up and reorganisation on the upper and middle Euphrates during the reign of Vespasian. This saw the establishment of four legions on or close to the Euphrates in the province of Cappadocia, Legio XII Fulminata at Melitene and Legio XVI Flavia Firma at Satala, Legio III Gallica in Commagene at Samosata and Legio IV Scythica at Zeugma.93 Considerable work was also done across the Caucasus with numerous auxiliary forts constructed and garrisoned there during Vespasian's reign. 94 There is also evidence for Roman road construction in Cappadocia and to the

Rome and Parthia

39

north of Palmyra. 95 In terms of conflict between Rome and Parthia during the Flavian period, there is some evidence for threats of war and planned campaigns but no evidence that anything significant eventuated. Vologaeses threatened an invasion of Syria in AD 76 and Domitian may have been planning an attack on Parthia towards the end of his reign. 96 What does seem clear is that the period from the last years of the reign of Vologaeses I (who died in AD 78) up to the first decade of the second century saw considerable instability once again in the imperial rule of Parthia. 97 Rome will at least have monitored this but it was not until the last years of the reign of Trajan that significant developments took place.

Trajan's Parthian war While the Parthian wars of the emperor Trajan marked a distinct development in the military and political relationship between Rome and Parthia, they were in some ways an extension of developments over the previous century and a half. 98 At the fullest extent of Trajan's Parthian invasion, a Roman army led by its general and emperor had marched all the way to the Persian Gulf and Roman provincial administration was established as far as southern Mesopotamia. Rome now had an Alexander-type example all of its own. A number of future Roman emperors would claim the emulation of Trajan as a goal, and future Parthian rulers were on notice regarding the imperial intentions and capacity of Roman armies. 99 Trajan's Parthian campaign was in some ways the next step in the long running rivalry and dispute between Rome and Parthia over Armenia and an attempt to directly control the installation of the Parthian king rather than the continued pursuit of supporting claimants to the Parthian throne. Trajan's actions also meant that the ramifications of Rome's ongoing tensions with Parthia were felt more directly in regional communities east of the Euphrates. In northern Mesopotamia, the rulers of regional kingdoms and principalities experienced the full force of the Roman military and the ramifications of a Roman emperor's displeasure. They also became an important focus of territorial competition between Roman and Parthian imperial interests in addition to the rivalry over Armenia. While Dio emphasised Trajan's desire for glory as the main motive for the emperor's eastern campaign, Parthian actions in Armenia were the immediate catalyst. One of the contenders for imperial rule in Parthia, Osroes, deposed the agreed-upon Armenian king, Tiridates, without reference to Rome. 100 Trajan's response was to march east in October 113 following the conclusion of the second Dacian war and despite an embassy from Osroes suggesting that a compromise be reached, Trajan continued to march. Concerned about the imminent Roman threat, Osroes removed his original replacement of Tiridates and elevated Parthamasires as the king of Armenia.101 Trajan arrived in Antioch early in 114 and when the campaigning season began, it did not take long for the Romans to assert control

40

Rome and Parthia

over Armenia. The two legionary bases of Cappadocia established by Vespasian at Melitene and Satala were the key points from which the invasion of Armenia was launched. Parthamasires gained an audience with Trajan near Elegeia in Armenia but the emperor rejected his request for recognition as king and was ruthlessly killed on Trajan's orders soon after. 102 Rulers of other principalities quickly submitted to the emperor, which is represented on Roman coinage by three rulers standing before the emperor having their kingdoms reassigned to them. 103 Trajan's success in Armenia resulted in embassies from various kings along the east coast of the Black Sea and in the Caucasus who confirmed themselves as clients of Rome. 104 These client-king arrangements across the Euphrates represented a similar approach to that taken towards kingdoms and principalities west of the Euphrates under the Julio-Claudians. Having met with limited resistance in Armenia and developing an increasing understanding of problems within the Arsacid royal succession at the time, Trajan turned his mind to the broader possibility of an invasion of northern and southern Mesopotamia. 105 There is numismatic evidence to suggest a Parthian dynastic struggle between Osroes based in Seleucia and Vologaeses III in Iran. 106 Dynastic struggles would be commonplace in Parthia throughout the second century, just as they were for much of the first century but the Roman approach was more directed at exploiting these divisions to achieve territorial expansion rather than simply sponsoring ongoing divisions within the Parthian dynasty in an effort to weaken it. At some stage before arriving in Mesopotamia, Trajan gave the rulers of the various northern Mesopotamian principalities the opportunity of an audience, essentially to establish friendship, but there was particular reluctance to do so by Abgar of Edessa, Mannos of Arabia, and Sporakes of Anthemusia/Batnae. 107

Figure 2.2 Sestertius, Rome Mint. Trajan. AD 115/116. Obv- IMP CAES NER

TRAIANO OPTIMO AVG GER DAC PARTHICO PM TR P COS VI PP: Bust of Trajan, laureate, draped, right. Rev- REGNA ADSIGNATA S C: Trajan, bare-headed, in military dress, seated left on platform on right with prefect and soldier; three kings standing before him. RIC II Trajan 666. Courtesy Classical Numismatic Group.

Rome and Parthia

41

Soon after his departure from Antioch in 115, Trajan crossed the Euphrates and captured Anthemusia/Batnae followed by Nisibis, for which he was granted the title "Parthicus." 108 The capture of Batnae and Nisibis essentially isolated the city of Edessa, which forced its king, Abgar, to present himself to Trajan on learning that the emperor was making his way to the city. Friendship was established between the two rulers before Trajan continued on to Adiabene. Broadly contemporary with these events, Lucius Quietus, one of Trajan's most dependable generals, defeated the king of Adiabene, Mebarsapes. 109 Soon afterwards, Quietus occupied the strategically important city of Singara at the foot of Jebel Sinjar in modern northern Iraq. Following these successes, Trajan retired to Antioch for the winter of 115/6. Trajan set out in the spring of 116 for the Tigris with the aim of again subduing Adiabene whose king had re-asserted his authority during the winter. 110 The army was divided and part of it captured the key Adiabenian cities of Nineveh, Arbela and Gaugamela, after which the province of Assyria was formed. 111 One section of the army made its way south and captured Babylon, while Trajan himself appears to have doubled back to the Euphrates in order to lead a force downriver as part of attacking Seleucia-Ctesiphon. 112 It was likely at this time that Roman troops occupied Dura Europos on the Middle Euphrates and a triumphal arch dedicated to Trajan was constructed outside it. 113 On arriving at the point where the Euphrates and Tigris flow closest to each other in the vicinity of the N aarmalcha canal (at that time blocked up), Trajan ordered his ships hauled overland to the Tigris and then captured Ctesiphon with virtually no resistance.114 The emperor sailed down the Tigris to the Persian Gulf and won over Athambelus V who was the king of Mesene/Characene, the principal city of which was Spasinou Charax. 115 Both Mesene and its capital, Spasinou Charax, were closely connected to Palmyra through trade at this time. On returning to Babylon, Trajan learned that much of the captured Parthian territory was in open rebellion. The emperor appointed and crowned a son of Osroes, Parthamaspates, as a Roman client-king of Parthia in the wake of the rebellions. The event was considered important enough by the Romans to mint coins with the legend REX PARTHIS DATUS and Parthian coins depicting Parthamaspates were also minted. 116 On the way back to Antioch, Trajan laid siege to the city of Hatra in what proved to be an unsuccessful attempt to punish the city's disloyalty. 117 Arriving in Antioch soon afterwards, and suffering increasingly from ill-health, Trajan departed for Rome and died at the port of Selinus in Cilicia on 8 August 117.118 Hadrian soon gave up all of Trajan's gains beyond the Euphrates, reverting to a purported Augustan policy of keeping the empire within its natural boundaries.11 9 Despite the quick reversal of fortunes prior to Trajan's death and Hadrian's abandonment of territorial gains, Trajan's Parthian campaign was remembered for centuries in the Roman world. Hadrian himself saw to this with the establishment of the Parthian games in Rome, which Dio claimed continued to be held for many years afterwards. 120 In later Roman invasions of Parthian and Sasanian Persian territory, Trajan's name was

42

Rome and Parthia

often invoked in speeches at the beginning of the campaigns as someone to emulate. Trajan's military and administrative activity from 114 to 117 placed a new emphasis on competition between Rome and Parthia in the kingdoms and principalities of northern Mesopotamia. Some of this activity was also related to the ongoing struggle between the two imperial powers over Armenia. Since the decline of the Seleucids and the rise of the Parthian Empire in the mid-second century BC, kingdoms and principalities based on the cities of Edessa, Carrhae, Nisibis, and Singara and their surrounding territory had been part of the Parthian Empire. This was true in military, political and cultural terms. With Rome's invasions of Parthia on three occasions in the second century, a marked change in the politics and eventually the culture of these entities in northern Mesopotamia became evident. While Trajan's Parthian war had an impact, Rome's enduring territorial expansion into northern Mesopotamia was more evident after the Parthian war of Lucius Verus in the mid-160s. Trajan's successes against the Parthians and subsequent territorial organisation signalled to the ruling elites of northern Mesopotamia that theirs was potentially disputed territory despite territorial gains being given up by his successor Hadrian. This gave local ruling elites potential leverage and when Roman imperial power arrived east of the Euphrates more permanently from the 160s than it had earlier under Trajan, some were able to take advantage. When opportunities arose, though, some of these rulers would assert greater independence.

Northern Mesopotamia and Armenia from Trajan to Lucius Verus In the midst of the rebellion against Trajan's administrative arrangements, Vologaeses III of Parthia succeeded in gaining recognition as king of Armenia where he was recognised as Vologaeses I. This arrangement was endorsed by Hadrian, effectively marking a return to the status quo of the Treaty of Rhandeia. 121 It is possible that Vologaeses and the Roman governor of Cappadocia, Flavius Arrianus, co-operated to repulse an invasion of Alan tribesmen ea. 136. 122 In an event indicative of ongoing imperial interest in Armenia and its surrounding territories, the king of Iberia, Pharasmanes II, visited Rome ea. 142 during the reign of Antoninus Pius. 123 Pharasmanes, members of his family and high-ranking troops were warmly welcomed in Rome and given high honours. It also appears that Antoninus Pius warded off a Parthian invasion of Armenia at approximately the same time. 124 In the late fourth century, the Historia Augusta indicated that Antoninus Pius (137-61) was able to exert influence in the kingdom of Osrhoene by sponsoring the return of Abgar VII to the throne at Edessa during a period of Parthian internal instability. 125 With regard to the relationship between Rome and Parthia more broadly, there is some evidence that direct Roman involvement with the royal family of

Rome and Parthia

43

Parthia continued when Hadrian returned the daughter of Osroes to the Parthians in 128/9. 126 Osroes' daughter had been captured when Trajan sacked Ctesiphon in 116. The Historia Augusta claimed that the royal throne had been captured by Trajan and that Hadrian promised to return it at the same time. This appears not to have eventuated, however, because Antoninus Pius refused a later request from Osroes to return the throne. 127 In a further indication of internal difficulties faced by the Parthian rulership, Antoninus Pius received an embassy from the Bactrians and Hyrcanians of Central Asia. 128

Lucius Verus' Parthian war and its aftermath Lucius Verus' Parthian war was the result of developments in Armenia, which bore similarities to those that marked the build-up to Trajan's war half a century earlier. In 161, the Parthian king, Vologaeses IV, placed a member of his own family on the Armenian throne without the agreement of Rome, provoking a predictable response. 129 Roman and Parthian forces initially clashed on the upper Euphrates with the Parthians gaining the advantage. The destruction of a Roman legion and the death of the governor of Cappadocia resulted in the province of Syria coming under threat from the Parthians. 130 Despite this initial setback, the Romans made substantial gains against the Parthians throughout 163, and by 165 Armenia was ruled by the Roman nominee, Sohaemus. 131 The result of the war was that the Parthians had been ejected from northern Mesopotamia, SeleuciaCtesiphon sacked once again, and the key city of Nisibis was occupied by Roman troops. The kingdom of Osrhoene, which occupied the western portion of northern Mesopotamia, was a dependent kingdom of Rome, and control of the Euphrates River was considerably extended in the direction of southern Mesopotamia. 132 Much has been made, especially in older scholarship, of the impact and longer-term effects of a plague outbreak that began during this war. Bivar identified the disease as smallpox and suggested that it was responsible for the Parthians being less capable of defending Seleucia-Ctesiphon against the Roman attack. 133 Dio emphasised the impact of the disease on the Roman troops and the fourth-century account of Ammianus Marcellinus retailed a story that attributed the disease to religious sacrilege committed by Roman troops while ransacking a temple in Seleucia-Ctesiphon. 134 According to Ammianus, the pestilence spread throughout Persia and across the Roman Empire as far as Gaul. An influential modern analysis of the impact of the plague on the Roman Empire concurs with Ammianus' claims. 135

The city of Edessa and Osrhoene Developments in the territory of Osrhoene as a consequence of Lucius Verus' war against the Parthians were among the most important in the longer term with regard to ongoing rivalry between Rome and Parthia and

44

Rome and Parthia

would have important consequences for the relationship between Rome and Sasanian Persia. From this time, Osrhoene was firmly within the Roman orbit as a client-kingdom, and in the early Severan period, much of the kingdom was converted into a Roman province. The most revealing evidence of Roman and Parthian competition at Edessa in the context of the Romano-Parthian wars of the 160s comes in the form of coins. In the years 163-65, bronze coins were minted at Edessa indicating the installation by Vologaeses IV of an individual named Wael as the king of Osrhoene. 136 Vologaeses removed Ma'nu VIII in favour of Wael, likely because of the sympa thies of the former towards Rome. This took place at approximately the same time as the Parthian installation of Pacorus on the Armenian throne, which was the key factor in the Romans prosecuting a war against the Parthians. The victories of Lucius Verus over the Parthians in 165 saw the return of Ma 'nu VIII to the throne of Edessa. Ma'nu had fled to the Romans after his deposition and on his reinstatement minted bronze coins depicting himself on the obverse and a reverse legend describing himself as Philoromaios. 137 Some rare silver drachms have been attributed to an imperial mint operating at either Edessa or Carrhae in the wake of the Roman victory over the Parthians and they add further to the evidence for the Roman takeover of Osrhoene after 165.138 The coins contain Greek legends and were first minted in the names of Marcus Aurelius and Lucius Verus and their respective wives Faustina Junior and Lucilla. They continued to be minted into the reign of Commodus. 139 The reverse legends follow a relatively uniform formula of VIIEP NIKHC POMAH1N/ROMAI and VIIEP NIKHC TQN KVPIQN C€BAC(TQN), indicative of the Roman victory over the Parthians. Ma'nu VIII's second reign, which began on his reinstatement in 165, lasted another 14 years until his death in 179. His son, Abgar VIII (later known as the Great), ruled until 214, and during this period provincial bronze coins were minted at Edessa depicting Abgar on the obverse with Commodus and Septimius Severus, respectively, on their reverses. 140 Some

Figur e 2.3 Drachm, Edessa/Carrhae Faustina Junior. AD 165. Obv- (AV )CTINA CEBACTH(N): Drap ed bust right of Faustina Jnr. Rev- VITEP NIKHC P.QMAH.1N.Venus (?) standing left holding round object (globe or apple?) in right hand; sceptre in left hand. Unpublished: cf RPC Online 4.10749; BMC 6- 7. Photograph: Effy Alexakis.

Rome and Parthia

45

of these coins name Abgar as Lucius Aelius Septimius Abgaros and Aelius Aurelius Septimius Abgaros, conflations of the imperial titulature of Commodus and Septimius Severus. Bronze coins bearing the same reverse inscription, VITEP NIKHC POMAH1N, as the silver coins of the Aurelius/ Verus joint reign were minted under Commodus and also are thought to have come from Edessa or Carrhae. 141 At Carrhae, bronze coins were minted bearing the bust of Marcus Aurelius on their obverses with reverse legends proclaiming KAPHNWN PHIL RWM. 142 There is no depiction of a local ruler accompanying the reverse legend, suggesting that the city itself was a "friend of the Romans" during the reign of Marcus Aurelius. Rome's defeat of the Parthians in the mid-160s was decisive in swaying Edessa and the kingdom of Osrhoene away from their traditional political leanings towards Parthia. Given the Roman victory over the Parthians in the 160s and the long-term effects this had on Edessa and Osrhoene and northern Mesopotamia more broadly, there was little choice for local rulers and elites but to accept the status quo. When civil war began to brew in the Roman Empire in the early 190s, however, some of these rulers attempted to take advantage, even attempting to go back to the Parthian fold. Loyalty was conditional and aspirations to independence appear to have remained.

The Roman military presence in Northern Mesopotamia after the Parthian war of Lucius Verus While Roman political power across the Euphrates in Osrhoene and northern Mesopotamia during the 170s and 180s was increasingly strengthened, evidence for a Roman military presence there is more difficult to discern. Dillemann suggested, therefore, that the Romans exercised "occupation sans annexion." 143 There is, however, indirect evidence that points to a build-up in the Roman military presence in the 180s. Firstly, Dio indicates that the Osrhoeni and Adiabeni attacked Nisibis prior to the civil war between Severus and Niger in 193, suggesting that Nisibis had received a Roman garrison before that date. 144 Dio also reported that after Niger's death at Byzantium in 195, the Osrhoeni and Adiabeni attempted to negotiate with Severus and justified their attack on Nisibis by claiming that this was part of undermining Niger. Importantly, in their attempts to negotiate, the Osrhoeni and Adiabeni refused to hand over forts they had captured while also demanding the removal of Roman garrisons, which were still there. On the evidence of Dio, then, Roman forts and garrisons existed across the Euphrates, including at Nisibis, prior to the beginning of the civil war between Severus and Niger. While conjectural, it is reasonable to conclude that these forts and garrisons had their origins at least in the reign of Commodus or perhaps even earlier. The late reign of Commodus and the period of civil war soon after his death in 192 saw upheaval in Roman imperial leadership that had not been experienced since the end of the Julio-Claudian period. The Osrhoeni,

46

Rome and Parthia

along with their allies in Adiabene, took clear advantage demonstrating that despite the reinstatement of Ma'nu VIII with Roman support ea. 165 and a seemingly close relationship between Roman and Osrhoenian rulers in the decades that followed, they were prepared to take advantage of instability in Roman imperial leadership. The Parthians faced similar problems at this time, which perhaps further emboldened both the Osrhoeni and Adiabeni. Political independence from the Roman and Parthian Empires was perhaps always an aspiration, although rarely a realistic one. When the chance came, the Osrhoeni and their emboldened neighbours took it. When Septimius Severus finally emerged from the war with Niger and established his rule securely, northern Mesopotamia and the Osrhoeni would be dealt with.

The Severan Wars with Parthia and territorial outcomes Following Pescennius Niger's self-proclamation as imperator at Antioch in 193, Septimius Severus confronted and defeated him the following year. 145 Osrhoene's treachery before and during the civil war, in alliance with Adiabene and possibly Hatra, resulted in the conversion of much of the kingdom into a Roman province under the governorship of a procurator in 195. 146 The city of Edessa was established as a client-kingdom with Abgar VIII confirmed as king. It is likely that the province of Mesopotamia was also established at this time, although it is possible that it was formed a few years later in 198. 147While the role of the Parthians in the actions of Osrhoene and other north Mesopotamian principalities may only have been nominal due to revolts in Seleucia-Ctesiphon and eastern Iran at this time, Septimius Severus sought to punish the Parthians soon after securing his rule. 148Towards the end of summer 197, following the successful elimination of the usurper Albinus in Rome, Severus turned east again and this time dealt directly with the Parthians. The three important texts that provide details on the Parthian campaigns of Septimius Severus are those of Cassius Dio, Herodian and the Historia Augusta. 149 There are clear inconsistencies between the three accounts with Herodian's seemingly the most troublesome and Dio's considered to be the most reliable. Herodian and the Historia Augusta emphasised glory as the key motivation of Severus, while Dio suggested a Parthian attack on Mesopotamia, especially Nisibis, as Severus' main reason for the invasion.150 Herodian proposed alternatively that Severus used his desire to exact revenge on Hatra as retribution for its support of Niger during the civil war of 193 as a prelude to mounting a full-scale attack on Armenia and then the Parthian Empire. 151 On capturing Nisibis, Severus proceeded down the Euphrates with a fleet of boats with the aim of attacking Seleucia-Ctesiphon. An un-named brother of Vologaeses V accompanied the Romans, indicating how fractured the Parthian royal dynasty was at this stage. 152A further indication of the difficulties faced by the Parthians was that Severus easily captured

Rome and Parthia

47

Seleucia and Babylon, which had both been abandoned, and Ctesiphon appears to have been captured easily as well. Severus was not interested in remaining in southern Mesopotamia for long and quickly withdrew after plundering Ctesiphon. It was at this point, likely in early 199, that the first and brief siege of Hatra took place. 153 Realising the difficulties in besieging such an imposing fortress , Severus withdrew and returned the following year to prosecute a lengthier and better-prepared siege. 154 Despite more extensive preparations for the second siege of Hatra, probably in 200, Severus was unable to capture the city.155 Dio believed that Severus wished to capture Hatra because it alone had withstood his forces during the military actions against Parthia in 198/9. 156 The fortress of Hatra came to develop a formidable reputation as a result of holding out against Trajan in 116 and now repulsing the forces of Septimius Severus twice. It would also hold out against the Sasanians in 230 before finally falling to the new Persian regime in 241. Septimius Severus then journeyed to Antioch before returning to Rome at the end of 202 via Egypt where he visited Pompey's tomb and made a sacrifice. 157 The emperor celebrated his decennalia in lavish style in Rome with gladiatorial games and distributions of grain to the populace. 158 A victory arch celebrating the success over the Parthians was constructed in the Forum and dedicated in 203. Denarii depicting Victory over the Parthians were minted in abundance; some types were in the name of Septimius Severus and others in the name of his young son Caracalla. 159 Dio undoubtedly kept his opinions to himself at the time , but at the relatively safe distance of a couple of decades from the events, he was strongly critical of Severus' provincial acquisition of territory in Mesopotamia. 160 It was an expensive province to maintain and yielded little by way of taxation. Of even greater concern for Dio was that the province brought Rome closer to peoples who were aligned with the Parthians, which resulted in unnecessary

Figure 2.4 Denarius, Rome. Septimius Severns. AD 202-210. Obv- SEVERVS PIVS AVG: Laureate bust right. Rev- VICT PART MAX: Victory , winged, draped, advancing left, holding wreath in extended right hand and palm sloped over left shoulder in left hand. RIC IV Septimiu s Severns 295. Photograph: Effy Alexakis.

48

Rome and Parthia

involvement in foreign wars. The latter criticism can hardly be justified in the context of events in the decades before Severus came to power but it is an interesting and rare contemporary opinion on Roman imperial activity. Roman territory had been extended along the Euphrates at least as far as Dura Europos under Lucius Verus ea. 165, and the city of Nisibis likely held a Roman garrison prior to Severns' elevation. Roman territorial involvement across the Euphrates and towards the Tigris in an increasingly direct sense had been developing before Septimius Severus came to power and can be traced as far back as the reign of Trajan. The numismatic evidence especially points to established relationships between Ma'nu VIII and Abgar VIII of Edessa through the reigns of Marcus Aurelius, Commodus and Septimius Severus.

Caracalla's Parthian war Following Caracalla's murder of his brother Geta on 26 December 211, the now sole emperor quickly began comporting himself in the style of Alexander the Great. Dio referred to Caracalla's enthusiasm for Alexander at length, and the Historia Augusta attributed a desire to imitate Alexander to many of Caracalla's actions. 161 Caracalla purportedly went so far as to raise a phalanx of Macedonians, which was 16,000 strong. An aggressive foreign policy was another element of Caracalla's emulation of Alexander and this was directed primarily at the Parthians, although it also encompassed military activity on the Danube. Although the hostile source tradition towards Caracalla emphasised his treachery in actions towards the Parthians and also in his dealings with Armenia and Osrhoene, Caracalla had good reason to take an active interest in events on the eastern frontier. Soon after the murder of Geta, Osrhoene, Armenia and the Parthian Empire were in a state of turmoil. Late in 213, Abgar IX of Edessa was summoned to Rome on the grounds that he had overstepped his authority by seeking to extend his influence throughout northern Mesopotamia. This was met by the king's removal and subjugation of his territory. 162 The Armenian king was similarly deposed and his heirs became bitterly divided against each other. A similar situation prevailed in Parthia from 213 when Artabanus V and Vologaeses V were in open warfare, the former based in Media and the latter in Ctesiphon. 163 The outcome of turmoil in Armenia was always unpredictable from Rome's perspective, and it could equally be so when leadership in Parthia was in dispute. According to Dio, Caracalla appealed to Alexandrian imagery as he made his way east with the ultimate aim of dealing with the situation in Armenia and prosecuting a Parthian campaign. In spring 214, he departed for the east, stopping to deal with issues in Dacia and Raetia on the way. 164 In the autumn, he crossed the Hellespont and sacrificed to Achilles at his tomb, then made his way to Pergamon before establishing winterquarters in 214/5 at Nicomedia. 165 At Nicomedia, the emperor had two enormous siege engines constructed in preparation for a war against Armenia and Parthia, and he keenly oversaw the preparation of a "Macedonian

Rome and Parthia

49

Phalanx." 166 In April 215, Caracalla arrived at Antioch before heading to Alexandria in Egypt to deal with a range of problematic issues there. 167 He departed Alexandria by February 216 and returned to Antioch to undertake preparations for a full-scale Parthian invasion. One of the claimants to the Parthian throne , Vologaeses V, was at this time harbouring Tiridates, a pretender to the Armenian throne, along with a Roman deserter, Antiochus, who had previously attached himself to Tiridates and fled to Vologaeses with him. Caracalla used this as a pretext to invade Parthia. Vologaeses surrendered the two fugitives, and Caracalla called off the campaign. 168 The general Theocritus was sent against Armenia, but was soundly defeated, and by the end of 215, the situation was much the same as it had been at the beginning of 214. 169 Later in 216 , Caracalla sought to further isolate Vologaeses by requesting the hand in marriage of Artabanus' daughter. 170 Dio interpreted the move cynically , inferring that Caracalla was simply looking for a pretext to attack Artabanus, knowing full well that the Parthian monarch would refuse. Indeed, in Dio's account, Artabanus did refuse the proposal, which was followed swiftly by a Roman attack on Media. This resulted in the capture of numerous fortresses and the destruction of Parthian royal tombs at Arbela, perhaps another attempt to draw link s with Alexander the Great as Arbela wa s the site of Alexander's last great battle with Darius III. Denarii and antoniniani were minted in 217 in Caracalla's name advertising VICT PARTHICA as the reverse legend and a denarius includes Caracalla's titulature on the reverse along with VIC PART in the exergue. 171 Herodian provided a distinctly different account, claiming that Caracalla 's initial marriage request to Artabanus was rejected but later accepted. 172 It is difficult to judge which version is more reliable, although Dio's is more in keeping with a strategy aimed at invasion. In early summer 216, Caracalla crossed into Parthian territory and made his way to Artabanus' palace near

Figure 2.5 Denarius, Rome. Caracalla. AD 217. Obv- ANTONINVS PIVS AVG

GERM: Bust of Caracalla, laureate, right. Rev- VICT PARTHICA (VO/XX on shield): Victory, winged, draped, seated right on cuirass, holding shield inscribed VO XX; behind her, shield; at her feet, helmet. RIC IV Caracalla 314a. Photograph: Effy Alexakis.

50

Rome and Parthia

Arbela where he was greeted by the king, his court and the local population.173 When the Parthians appeared to be at their most vulnerable, the emperor gave the order to his army to attack. 174The Parthians, including Artabanus, fled in pandemonium. The Parthians retreated and Caracalla's army ravaged the countryside before retreating to Roman Mesopotamia. Artabanus eventually gathered a large army north of the upper Tigris in the spring of 217. 175This was aimed at avenging the events of late 216 with the probable aim of attacking Roman possessions in northern Mesopotamia. On 8 April 217, Caracalla was murdered between Edessa and Carrhae due to a conspiracy among senior military commanders and the praetorian guard. 176 Caracalla's successor, Macrinus, immediately sought to come to terms with Artabanus but the Parthian ruler rejected the proposal, advancing to Nisibis and confronting Macrinus in a three-day series of skirmishes. 177 The Parthians were clearly in the ascendant and Macrinus sued for peace at a cost of 200 million sesterces. 178Despite this, Macrinus had denarii struck in 218 that advertised VICTORIA PARTHICA. 179In June 218, Macrinus was overthrown, and his son, Diadumenian, was killed at Zeugma while reputedly attempting to flee to Artabanus requesting sanctuary. 18 Caracalla's war against Parthia essentially had been fought to a stalemate despite the turmoil in the Parthian leadership. In northern Mesopotamia, the most significant territorial outcome was the abolition of the client-kingdom of Edessa.

°

Conclusion From the first direct military confrontation between Rome and Parthia, the internal politics of Parthian rulership was a factor. The Parthians, in turn, became enmeshed in the disputes and civil wars in the Roman Empire that came to define the last decades of the Republic. Through much of the Julio-Claudian period, Roman emperors played active roles in destabilising Parthian political leadership, often supporting members of the Arsacid Royal family who had fled to Rome or who had been sent to Rome as hostages guaranteeing agreements between Rome and Parthia. Armenia was the immediate flashpoint and stage for direct military conflict and competition between the two powers during this period, but the treaty of Rhandeia between Nero and Vologaeses I would eventually prove unsatisfactory. In attempting to convert Armenia into a Roman province, establishing provinces as far as Assyria and placing a preferred nominee directly on the Parthian throne, Trajan's approach represented the next step for the Romans. Trajan's campaigns and the short-lived territorial organisation that accompanied them placed a heightened emphasis on Roman power and influence east of the Euphrates, especially in the cities of northern Mesopotamia. On occasion, ruling elites in this area saw the opportunity to play off long-standing Parthian interests against the more recent interests of the Romans in an effort to maintain a degree of autonomy from the two opposing superpowers. This had been the situation in Armenia since the first

Rome and Parthia

51

century BC when rulers and elites took advantage of the tensions between Rome and Parthia over control of the kingdom. In the wake of Lucius Verus' campaign against the Parthians in the mid-160s, Roman power in northern Mesopotamia was strengthened considerably, and rulers in the city of Edessa received Roman backing and announced on their coinage that they were friends of Rome. It is telling that when civil war broke out between Septimius Severus and Pescennius Niger in the 190s, some of the ruling elites in the cities of northern Mesopotamia attempted to take advantage, some backing Niger and others soliciting the patronage of the Parthians. In the turmoil following Caracalla's murder of his brother Geta in 211, Abgar IX of Edessa wasted no time asserting more power in northern Mesopotamia, and the Armenian king also attempted to take advantage. The long-term wresting of influence and control in northern Mesopotamia by Rome from the Parthians is, therefore, an important indicator of the development of Roman political and military power further east from the reign of Trajan onwards. The political and military relationship between Rome and Parthia is important to understand, therefore, in any analysis of the relationship between Rome and the Sasanians. While the routine portrayal of the Parthian rulership as plagued by internal dissension and strife was undoubtedly overplayed in the hostile Roman source tradition, there are many specific and verifiable examples of Parthian imperial claimants and fugitives fleeing to the Romans who were then used to destabilise the Parthian regime. By contrast, up to the Severan period, Roman imperial leadership enjoyed relative stability. In the few notable examples of Roman imperial instability, the Parthian imperial leadership similarly attempted to take advantage, most notably in the civil wars of the late Republic, the instability surrounding the end of Nero's reign and the civil war between Pescennius Niger and Septimius Severus. Related to this was the considerable territorial expansion of Rome in the eastern provinces, especially across the Euphrates in the direction of the upper Tigris in the second century AD. The wars of Trajan against the Parthians marked intensified military aggression on the part of Rome and became an example to follow for centuries despite the immediate outcome delivering little by way of direct territorial gains. The campaigns proved that northern and southern Mesopotamia and the Parthian capital of Seleucia-Ctesiphon were both vulnerable and militarily achievable targets but the changes wrought by the Sasanians changed that potential markedly.

Notes 1 Plutarch, Sulla, 5.3-4; see Keaveney 1981. 2 Plutarch, Sulla, 5.7-8; Livy, Epitome 70; Festus, Breviarium 15. 3 Florus 1.46.4-5; Keaveney 1981, 197 believed that there was a formal alliance under Sulla, whereas Sherwin-White 1984, 119-20 thought that arrangements between the Parthians and Sulla (and later Lucullus) were informal only.

52 4 5 6 7 8 9 10 11 12

13 14 15 16 17 18 19 20 21 22 23 24 25

26 27 28 29 30 31 32 33

Rome and Parthia Justin 42.2.1. Justin 42.2.2. Bivar 1983, 42. Strabo 11.14.15. Strabo 11.14.16. See Mayor 2010, 262-314. Plutarch, Lucullus, 36.6. Plutarch, Lucullus, 30; Appian, Mithridatic Wars, 87; Cassius Dio 36.3. Shayegan 2011, 311-30 argues that the Arsacids employed a subtle strategy of influence from the mid-90s to the mid-60s BC in the western holdings that Tigranes had wrested from them rather than simply abandoning imperialistic activity due to internal difficulties. On this analysis, the intervention of the Romans in Cappadocia and Armenia was initially seen by the Parthians as a positive development in a long-term desire to retrieve these territories. Cassius Dio 36.45.3. Plutarch, Pompey, 33.6; Appian, Mithridatic Wars, 104; Pliny the younger, Epistles, 10.78.1; 112.1. Strabo 12.3.34; Appian, Syrian Wars 51.2.57. Plutarch, Pompey, 33.6; Cassius Dio 37.5.5. Appian, Mithridatic Wars, 118. Appian, Mithridatic Wars, 119. Plutarch, Pompey, 32. Cassius Dio 39.56.2. Bivar 1983, 48. Justin 42.4.1; cf Plutarch, Crassus, 21.6 who emphasised the role and importance of the Surena in these events. Bivar 1983, 50; Plutarch, Crassus, 16.8. Bivar 1983, 56. Overtoom 2016, 139ff argues that Rome and Parthia were isolated from each other prior to Carrhae but the evidence suggests that this is an overstatement. Overtoom uses the observations of theorists of international relations to label what he calls the first Romano-Parthian war that culminated with Crassus' defeat at Carrhae as a "hegemonic war," which resulted in the "violent merger between two previously separate interstate systems." As a result, "Crassus' disastrous invasion of Mesopotamia (54-53 BCE) caused the Mediterranean and Eastern systems to merge suddenly and violently into what we may call the new and expansive "Med-Eastern system." This suggestion overlooks Rome's sustained involvement in Pontus, Armenia, Mesopotamia, Syria and Arabia as part of attempting to deal with Mithridates and Tigranes for approximately four decades beforehand. It also ignores the diplomatic contact and agreements between Rome and Parthia for more than a decade prior to Carrhae. It is important also to place Crassus' defeat at Carrhae in the context of internal developments in Parthia into which the Romans had been drawn with the flight of Mithridates to Rome four years prior to Carrhae. The application of modern theories and models developed in other disciplines may have benefits but there are many potential pitfalls that accompany it. Schippmann 1986. Cicero, Letters to Atticus, 5.18; Letters to Friends (Cato) 14.4. Cicero, Letters to Atticus, 1.3. Cassius Dio 41.55.3-4, 42.2.5; Caesar, Civil Wars, 3.82-4, 88-96. Appian, Civil Wars, 2.83; Cassius Dio 42.2.5. Appian, Civil Wars, 2.110. Appian, Civil Wars, 4.59, 63, 88; Cassius Dio 47.30.3. Cassius Dio 48.24.3-27; Plutarch, Antonius, 28.1.

Rome and Parthia

53

34 Cassius Dio 48.41, 49.19.2-3; Strabo 16.2.8; Frontinus, Stratagems of the Aqueducts, 1.1.6. 35 See Schippmann 1986. 36 Schippmann 1986 refers to the widespread belief that Phraaspa 1s likely 37 38 39 40 41 42 43 44

45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64

65 66 67 68

Takht-i Soleiman. Plutarch, Antonius, 37. Vergil, Georgics, 1.509; 4.561. Plutarch, Crassus, 22.2. Plutarch, Antonius, 39; Cassius Dio 49.25.4. See Patterson 2015 for an analysis of Antonius' approach to Armenia, which Patterson argues was driven by his desire to establish Armenia as a Roman client-kingdom. Justin 42.5-6; Cassius Dio 51.20. Bivar 1983, 65. Sellwood 1995, 75. Sellwood, 1995; The story is also given some credence by the broadly contemporary Isidore of Charax, Parthian Stations, 1, who referred to Phraates killing his concubines on an island upstream from Anatha as Tiridates advanced down the Euphrates. Justin 42.5.8. Res Gestae Divi Augusti 32. Bivar 1983, 66. Cassius Dio 53.33. Cassius Dio 54.7.4. See especially Cassius Dio 54.8.2. Res Gestae Divi Augusti 29; Horace, Odes, 1.12.53-4; Propertius 4.6.79. See Schlude & Rubin 2017, 80ff. See especially Rose 2005. RIC I.287 for coins. See Schneider 2007 for a consideration of the Prima Porta statue and other depictions of Parthians in Roman art. Strabo 16.1.28; Res Gestae Divi Augusti 32. CIL 6.1.1799; ILS 842. See Nedergaard 1988. Tacitus, Annals, 2.4. Josephus,Jewish Antiquities, 18.42; Bivar 1983, 68; Debevoise 1938, 147. Velleius Paterculus 2.100, Res Gestae Divi Augusti 27, Tacitus, Annals, 2.4; Cassius Dio 55.10.18. Cassius Dio 50.21. Velleius Paterculus 2.101.1-3. Res Gestae Divi Augusti 32. Tacitus, Annals, 2.2.1 ff; Dabrowa 2017, 173-4; see also Bivar 1983, 68-9. Tacitus, Annals, 2.56.2-3. See Olbrycht 2016 for a discussion of the Armenian settlement under Tiberius and Artabanus II. Olbrycht emphasises the ultimate failure of Augustus' approach to Armenia, which is evident especially in Tacitus' coverage. Tacitus, Annals, 6.31. See Dabrowa 2017, 174-7. See Olbrycht 2012. Tacitus, Annals, 6.31; Cassius Dio 59.27.2. Shayegan 2011, 293-5 postulates that Tacitus' account of this claim by Artabanus II has some distinctive differences to later reports by Dio, Herodian and Ammianus Marcellinus of Sasanian attempts at leveraging the Achaemenid past in diplomatic communication with Rome. In Tacitus' account, Artabanus invoked the Achaemenid king Cyrus and then Alexander the Great; "at the same time he (Artabanus) threw vain talk and menaces with respect to the old boundaries of the Persians and the Macedonians, and that he would seize

54

69 70 71 72

Rome and Parthia what was held first by Cyrus and afterwards by Alexander." In the other classical accounts, Alexander is not mentioned and references to the Achaemenids are more general by not naming a specific king. This may be more reflective of the origins of the Parthian dynasty in the post-Alexander Hellenistic world. Shayegan (331) concludes that the Babylonian literary tradition was primarily responsible for preserving and then conferring the Achaemenid title "King of Kings" on the Parthian monarchs. Further to this, he attributes the Pontic kingdom with "safeguarding the memory of its Achaemenid heritage (which) furnished the substance of what was to become the Arsacid political ideology." See Chapter 3 and Chapter 10 for further discussion of the claims made by Dio, Herodian and Ammianus. Tacitus, Annals, 6.36; see also Bivar 1983, 73. Tacitus, Annals, 6.43.2. Josephus, Jewish Antiquities, 18.101-3; Suetonius, Vitellius, 2.4; Bivar 1983, 74 who suggests that this took place early in the reign of Caligula. Cassius Dio 59.17.5; Suetonius, Caligula, 19.2; Josephus, Jewish Antiquities,

18.4.5. 73 74 75 76 77 78 79 80

81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103

Cassius Dio 60.8.1. Tacitus, Annals, 11.9ft; Dabrowa 2017, 177-80. Cassius Dio 60.8.1; Tacitus, Annals, 12.9.1. Tacitus, Annals, 12.10. Tacitus, Annals, 11.10.4; 12.10.1. Tacitus, Annals, 12.14. See Hauser 2012, 1004. See Guzman 2014 for a detailed consideration of the Roman literary sources for the wars in Armenia during Nero's reign. Gregoratti 2017 undertakes a detailed analysis of Tacitus' coverage in the Annals of the war between Rome and Parthia in Armenia. Tacitus, Annals, 13.9; Bivar 1983, 81. See Dabrowa 2017, 180-5 for an analysis of Tacitus' account of Nero's dealings with Parthia over Armenia. Bivar 1983, 82. Tacitus, Annals, 13.37. Tacitus, Annals, 14.26.1. Tacitus, Annals, 15.14; Cassius Dio 62.21.2-3. Tacitus, Annals, 15.24. Cassius Dio 62.23.2. Cassius Dio 63.2ft. Cassius Dio 63.5. Suetonius, Nero, 57. Braund 1984, 129-65; Edwell 2008, 19. Josephus, Jewish Wars, 7.219-23. see Edwell 2008, 18-20. See Bosworth 1976. Mitford 1974; Edwell 2008, 18. Epitome de Caesaribus 9.12 (142); Debevoise 1938, 215. See Bivar 1983, 86-7. The most detailed study of Trajan's Parthian campaign remains that of Lepper 1948. See also Bennett 1997, 183-204; Edwell 2013; Kettenhofen 2004. See especially Lightfoot 1990. Bivar 1983, 87. Cassius Dio 48.7.1; Ross 2001, 30-1. Bivar 1983, 88. This is the so-called "Regna Adsignata" coinage of Trajan: BMCRE 1966, No. 1043; RIC II 666.

Rome and Parthia 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139

140 141 142 143 144 145

55

Eutropius, Breviarium, 8.3. Ross 2001, 32. Bivar 1983, 88 claims that it was Vologaeses II. Cassius Dio 68.21.1-3. See Palermo 2019, 25-32 on Trajan's capture of northern Mesopotamia and efforts to provincialise it. Cassius Dio 68.18.2; 23.2. Cassius Dio 68.22.2. Cassius Dio 68.26.1-2. Eutropius 8.3.2; Festus 14.20. Cassius Dio 68.26.4. Baur et al. 1933, 56-65. Cassius Dio 68.28.2-3. It was likely at about this time that an aureus with the legend PARTHIA CAPTA was minted: RIC II Trajan 324, Cassius Dio 68.28.4. RIC 11.667 (Roman). BMC 46; Sellwood 81/1 (Parthian). Cassius Dio 68.30.2-3. Bivar 1983, 91 claims that Hatra's success in defending itself against the Roman siege marked the beginning of a period of prosperity for the city. Cassius Dio 68.33.3. Bennett 1997, 203. Cassius Dio 69.2.3. Chaumont 1986; Historia Augusta, Hadrian, 23.10. Cassius Dio 69.15.1-2; Debevoise 1938, 242. Cassius Dio 69.15.1. Historia Augusta, Antoninus Pius, 9.6. Historia Augusta, Antoninus Pius, 9.6; see also Debevoise 1938, 245. Historia Augusta, Hadrian, 13.8. Historia Augusta, Antoninus Pius, 9.7. Epitome de Caesaribus 15.4 (151). See Millar 1993, 111-4; Birley 1987, 121-33, 144-7. Cassius Dio 71.2.1. Debevoise 1938, 249. Edwell 2013, 116. See Palermo 2019, 32-5 for a consideration of the archaeological evidence for extension of Roman control in northern Mesopotamia in the wake of this campaign. Bivar 1983, 93. Cassius Dio 71.2.4; Ammianus Marcellinus 23.6.23. See Littman & Littman 1973 for more on the impact of the disease on Roman troops. Duncan-Jones 1996. BMC Arabia 28, 91-2, no. 2 and 3. BMC Arabia 28, 92, no. 4 and 5. Babelon 1893, 234; BMC Arabia 28, xcviii-ix. The obverse legends name the emperor or empress (e.g. Faustina Sebaste (RPC Online IV.3 temp. No. 3498), Autokrator Markos Aurelios Antonin (RPC Online IV.3 temp. No. 6496), Loukilla Sebaste (RPC Online IV.3 temp. No.6502), Autokrator Lukios Aurelios Veros Sebastos (RPC Online IV.3 temp. No. 8034). BMC 28 Arabia, "Edessa," 12 (Commodus), 14-35 (Septimius Severns); see also introductory discussion at c-ci. RPC Online IV.3 temp. Nos 8046, 8047, 10746. BMC 28 Arabia, lxxxix-xc; 82, no. 1. Dillemann 1962, 197. Cassius Dio 75.1.1-3. Cassius Dio 75.6.lff.

56

Rome and Parthia

146 Wagner 1983, 111-6. See Palermo 2019, 37-41 for a detailed analysis of epigraphic evidence for the conversion of Osrhoene into a province over the following two decades. 147 Kennedy 1979, 255; Palermo 2019, 35-6. 148 See Debevoise 1938, 255. 149 Cassius Dio 75.9.1-12.5; Herodian 3.9.1-12; Historia Augusta, Severus, 15.1-16.9. 150 Herodian 3.9.1; Historia Augusta, Severus 15.1; Cassius Dio 75.9.1. 151 Herodian 3.9.1-3. 152 Cassius Dio 75.9.3 - Artabanus according to Herodian. 153 Cassius Dio 75.10.1. 154 Cassius Dio 75.11.lff. 155 Cassius Dio 75.11.1-12.5. 156 Cassius Dio 75.11.1. 157 Cassius Dio 75.13.1. 158 Cassius Dio 77.1.1-5. 159 The reverse legend VICT PART MAX surrounding a depiction of Victory appears on RIC IV-1.295 Severns (for example) and RIC IV-1.144 for Caracalla. 160 Cassius Dio 75.3.2-3. 161 Cassius Dio 78.7.1-8.3; Historia Augusta Caracalla 2. 162 Cassius Dio 77.12.1. 163 Cassius Dio 77.1-2. 164 Historia Augusta, Carus, 5.4. 165 Cassius Dio 77.16.7-8. 166 Cassius Dio 77.18.1. 167 Herodian 4.8.6-9. 168 Cassius Dio 77.19.1-2, 21.1. 169 Cassius Dio 77.21.2. 170 Cassius Dio 78 .1.1. 171 RIC IV.1 314a denarius; RIC IV.1 3146 antoninianus; RIC IV.1 297e. 172 Herodian 4.11.1. 173 Herodian 4.11.3-4. 174 Herodian 4.11.5-7. 175 Historia Augusta, Carus, 6.4. 176 Cassius Dio 78.1.4, 78.5.4; Herodian 4.13.3-8; Historia Augusta, Carus 6.6-7.2. 177 Cassius Dio 78.26.7; Herodian 4.15.4. 178 Cassius Dio 78.27.1. 179 RIC IV Macrinus 96. 180 Debevoise 1938, 267.

3

Conflict and diplomacy between Rome and Persia from Ardashir to Philip I

Introduction The emergence of the Sasanians as the supreme dynasty ruling over the lands of Iran in the third century AD wrought fundamental and longlasting changes in the Iranian Empire and had important consequences for the military and political relationship with Rome. Beginning with an internal rebellion against the rulers of the kingdom of Persis, the Sasanians eventually overthrew the Arsacids as rulers of the lands of Iran and established one of the most powerful dynasties of the ancient and late antique worlds. Ardashir, a descendant of Sasan, the eponymous founder of the dynasty, defeated the more powerful of two rival Arsacid Parthian rulers, Artabanus V, at Hormizdagan on 28 April 224. He then captured Ctesiphon and was crowned there also taking the title Shahanshah or King of Kings later in 224 or as late as 227. 1 The initial consolidation of this victory, including the apparent elimination of the last Arsacid imperial claimant, Vologaeses VI, in the years immediately following, quickly gave way to imperial expansion a hallmark of the early Sasanian dynastic rulers. This expansive approach is evident in both the western and eastern portions of the Iranian Empire and there was also an intensified focus on the Persian Gulf and Arabia. 2 The expansive approach of the Sasanians, coupled with some significant political and economic reforms, stood in contrast to the rulers of the Roman Empire who began to face difficulties in the 230s. This would give way to increasingly obvious and rapidly developing problems from circa 250, which are typically identified as a full-blown crisis by the 260s. Some of these problems were both caused and exacerbated by military conflict with Sasanian Persia. Difficulties in managing threats from different kinds of tribal groups in the provinces of the middle and lower Danube Rivers became an increasingly serious problem from the 230s. The emergence of Maximinus Thrax as the emperor in 235 would come to mark the beginning of a proliferation of the so-called "soldier emperors" hailing from the periphery of the empire over the following century. Indications of increasingly serious economic difficulties also belong to the period of the 230s and 240s and localised usurpations in the late 240s might be seen as indicative of more serious rebellions to come.

58

Ardashir to Philip I

Internal factors in Persia and Rome affecting the relationship from the 220s to the 240s The internal political and economic problems in the Roman Empire of the third century AD have received extensive attention in modern scholarship and have traditionally been part of defining the period from 235 to 285 as the "Third Century Crisis." A long series of short-reigning emperors who hailed more often than not from the periphery of the empire and whose fates were in many cases decided abruptly at the hands of their own military companions remains a marker of Rome's internal problems during this period. Coupled with a number of prolonged usurpations resulting in breakaway regional "empires" and the effects of invasions on the Rhine, Danube and eastern frontiers, there were times in the 250s and 260s that anyone could be forgiven for doubting the capacity of the Roman Empire to survive. These factors had clear implications for how Roman leaders responded to Persian military threats and activity. By the mid-240s, there are signs that Rome's internal challenges might have been about to accelerate, although the extent to which contemporaries saw it this way is a matter of debate. 3 Gordian III's death on campaign against the Persians in 244 and financial requirements of the subsequent treaty caused problems in the eastern provinces and were an early indicator of the difficulties Rome would experience in its dealings with Sasanian Persia. 4 The challenges on the Danube with Gothic tribal groups would draw increasing amounts of imperial attention and while the small-time usurpations of Jotapian in the east and Pacatian on the Danube were not serious problems, they perhaps presaged the more significant usurpations of the 260s and 270s. Roman imperial coinage was potentially symptomatic and causative of Rome's political and economic difficulties as the third century unfolded and is often used as a marker of Rome's cumulative problems. 5 With increasing demand on government revenues due to the requirement to place large armies on the Danube and in the eastern provinces, problems with tax receipts due to breakaway regional "empires," especially in the 260s and 270s, and no system of public borrowing, imperial administrations had little choice but to manipulate the silver currency in terms of quantity and fineness so as to meet the requirements of expenditure. 6 Manipulation of the currency is noticeable half a century earlier with the reduction in silver content of the denarius by approximately one-third during the reign of Septimius Severus (193-211). Caracalla's introduction of the antoninianus ea. 215 is also indicative of manipulation of the silver currency. With a notional value of between 1.5 and 2 denarii but containing the same silver content as the Severan denarius, the new denomination was also a means by which the silver currency was debased.7 While its production was stopped under Caracalla's successor, Elagabalus, production of the antoninianus got underway again in 238 and it soon became

Ardashir to Philip I 59 the principal imperial silver unit. Debasement of the antoninianus took place at a rapid rate in the 250s and 260s. The long-term cumulative effect was the reduction in silver content of coins from approximately 60% early in Septimius Sever us' reign to little more than 1% under Aurelian in the early 270s. 8 It was the 250s and 260s that saw an especially rapid and steep decline. The requirement to produce more coins to pay defensive and offensive armies also meant that there were more coins in circulation. Modern monetary theory suggests that problems with inflation are caused by unregulated growth in the supply of money, physical or otherwise. Together with falling confidence in the antoninianus due to increasing reduction in silver content as the third century progressed, high levels of inflation were an almost inevitable outcome. Problems stemming from manipulating the currency unit in which the soldiers were paid were poorly understood in antiquity (and even until quite recently in modern economies) but the potential impact on prices, which were at least partly affected by poor control of currency manufacture and distribution, is spelled out in an albeit later document, the preamble to Diocletian's prices edict. 9 Extensive evidence for hoarding of silver coins in the third century is another indicator of a growing inflationary problem and was, in turn, a contributing factor to it. With the silver content of denarii and antoniniani dropping, savers of silver currency (often soldiers) selected coins of higher silver content for hoarding in an attempt to hedge against future inflation. This had the undesirable effect of withdrawing higher quality silver coins from circulation, which, in turn, made the problem worse. This is one of a number of problems with what are referred to as "commodity" currencies, which rely to a significant extent on their precious metal content as a means of maintaining faith in them as a means of exchange. The Roman denarius had been a commodity currency since it was first minted in the days of the Republic. During the Julio-Claudian period, silver content, or fineness, of the denarius was upwards of 90% until the reign of Nero. While debasement under Nero is observable, the rate of the reduction in fineness of silver currency is particularly notable from the reigns of Septimius Severus and Caracalla onwards. In effect, this resulted in the silver currency of the latter half of the third century, namely the antoninianus, shifting from its role principally as a commodity currency to operate more as a fiat currency; a currency whose value attains based on factors other than its precious metal content. With little intrinsic value in coins and paper money, modern fiat currencies rely exclusively on sentiment, which in terms of exchange rates is driven substantially by measures of the performance and stability of the economies of which they form a part. Sentiment is also driven by other factors, including political, military or even cultural ones. With the silver currency unit, initially the denarius and then the antoninianus, operating as the main source of remuneration for the soldiers, attempts to maintain or even build faith in

60

Ardashir to Philip I

it needed to be based more on messaging from the imperial administration than on intrinsic values based on silver content. This is one explanation for a proliferation of different reverse types on the antoniniani of the imperial mints from the 250s to the 270s, especially the mints producing coins for armies defending imperial frontiers on the Danube and in the east. The Antioch mint, for example, produced large volumes of antoniniani with widely variant military themes and other messages promoting concepts of imperial stability and even claims of military victories when the opposite was quite clearly the case.

The contrast between Rome and Persia By contrast, there are signs that the Sasanian political and economic system was beginning to strengthen in the 230s and 240s. From the outset, the Sasanians re-organised the imperial coinage, which sent an early signal internally and possibly externally that a new regime was in charge. While the models for Sasanian coinage came primarily from Alexander the Great and the Seleucid and Parthian periods that followed, Gobl remarked, "But the numismatic history of the Sasanians viewed as a whole is that of a thoroughly distinctive type within the monetary world of Antiquity and of the Orient." 10 Part of the new approach to coinage was the minting of gold coins, which did not take place under the Parthians. The first gold dinars of Ardashir weighed as much as 8.5 grams, approximately 2 grams more than contemporary Roman aurei of the 220s. By the end of his reign in the early 240s, Sasanian gold dinars weighed between 7.2 and 7.4 grams and rarely deviated from this range through the rest of the dynasty's existence. Roman aurei continued a sharp decline in weight from a range of 6-6.5 grams under Severus Alexander to as little as 3.0-3.5 grams under the emperors of the 250s and 260s. In terms of messaging, Ardashir was described as "King of the Iranians" on early gold dinars and this soon gave way to an expanded title, "King of Kings of the Iranians." 11 By the end of his reign, he was referred to as "of the seed of the gods," which is also shown on silver drachms. While Alram believes that Sasanian gold coinage was "purely prestigious" and did not have the economic significance that it did in the Roman Empire, Gariboldi believes that there was a connection between Sasanian gold and silver denominations; that is, an established ratio existed between the two. 12 The standard silver denomination minted by Ardashir and all of his successors was the drachm, of which there were also fractions. The Parthian kings initially struck drachms on the Athenian weight standard of 4.25 grams and with approximately 95% fineness but in the decades prior to the Sasanians coming to power, weights of Parthian drachms had dropped to 3. 70 grams with fineness of 70 % .13 By the end of Ardashir's reign, the Sasanian silver drachm contained approximately 90% silver and weighed 4.2 grams, thus resembling the old weight standard of the Athenian silver drachm. 14 Silver content of Sasanian drachms increased even further

Ardashir to Philip I

61

to approximately 94% under Shapur I while retaining a similar weight. 15 The only exception is a discreet series of drachms minted under Shapur I that contained substantially lower silver contents (as little as 12 % ) and are thought to have been minted from recycled Roman coinage captured in the wars of the 250s/60s. 16 With these coins as one of very few exceptions, and a clear demonstration of the contrast between Roman and Sasanian currencies in the middle of the third century, the Sasanian silver drachm would remain stable with silver content of 90% or higher for much of the rest of the dynast y's rule. The iconography of the Sasanian silver coinage was equally stable with comparatively little variat ion during the third century and later. While a unique double dinar of Shapur I likely depicts the Shahanshah with a subservient Philip or Valerian on its reverse, this type of iconography is nowhere to be seen on the silver coinage despite the significance of Shapur's victories over the Romans proudly displayed in rock reliefs and the SKZ inscription of the 260s. 17 Gobl postulated that the lack of the "propagandist element" on Sasanian coins might even have been defiant of the Roman numismatic traditions of the time, which saw many and varie d reverse types aimed at promoting imperial reliability. 18 Gobi suggested further that some aspects of Sasanian numismatic design, especially the engraving of the Shahanshah's name and titles close to the edge of the coin's obverse, were not only adopted from the Roman imperial numismatic tradition, but were "a deliberate challenge to the old enemy." The ruler depicted on the obverses of Sasanian silver drachms was denoted primarily by crowns, which were distinctive for each ruler and the reverses were structured around a fire altar as the central element of the design. 19 While this relatively uniform design of Sasanian silver coinage is often dismissed in terms of its capacity to convey imperial messaging, it had the capacity to send a continued message of impe rial stability and centralised authority, while promoting confidence in the currency.

Figure 3.1 Drachm, Ctesiphon. Ardashir I. AD 224-242. Obv- "The Mazda worshipping, divine Ardashir, King of Kings of Iran who is descended from the Gods": Bearded crowned bust of king to right. Rev-"Fire of Ardashir": Fire altar with ribbons no attendants. Gobl type 111/2/2. Cour tesy Noble Numismatics.

62

Ardashir to Philip I

Sasanian economic development and military reforms Indicators of an increased capacity for agricultural output, building activity and manufacturing are in evidence during the reigns of the early Sasanian rulers of the third century and this was important to underpinning a longer-term capacity to expand the army. Ardashir and Shapur I were famed for their foundations of cities and expanded irrigation programmes and there is material evidence to support these claims. 20 According to the accounts of Tabari and Hamza Isfahani, Ardashir and Shapur I were attributed with founding as many as 18 cities, most of them in Persis, Mesopotamia and Khuzistan but some also in locations such as Characene in the Persian Gulf. 21 Some of these urban foundations were enlargements of previously existing settlements of the Parthian period or earlier and a number would become important population centres in the Sasanian period. Ardashir began an urban expansion project in the broader environs of Seleucia-Ctesiphon that would continue under his successors and with Shapur I's strengthening and enlargement of the strategic site of Meshike (renamed Peroz-Shabur - Pirisabora in Latin), the Sasanian capital would be more difficult for an invading Roman army to subdue and capture. Some of the population growth that took place in these cities came from deported populations from cities of the Roman eastern provinces as a consequence of the invasions of Shapur I in the 250s. These deportations appear to have targeted engineers, builders and those with craft-based skills such as glass-making and textile-manufacturing. This enhanced the capacity for manufacturing in some urban centres, which, in turn, created an enhanced capacity to trade higher value manufactured items. 22 Cities such as Gundeshapur, which was newly built according to Tabari, were noted especially for their concentrations of Roman captives. 23 Shapur I's palace complex at Bishapur shows material evidence for the presence of Roman captives, especially in mosaics that demonstrate Antiochene and/or Edessan designs and styles. Along with urban foundations and enlargements, expanded irrigation projects, including dams, were also significant under the early Sasanian rulers and this, in turn, had the capacity to stimulate population growth over time. In the process of analysing the development of irrigation in southern Mesopotamia from neo-Babylonian times onwards, Christensen refers to the Sasanian period as the zenith of irrigation in this region. 24 Other irrigation expansion activities are in evidence in Khuzistan, while from the Iranian plateau to the eastern portions of the Sasanian realm qanat irrigation techniques and technologies were employed at greater levels than before to support population growth and larger urban centres. In relation to Fars (Persis), Christensen notes, "As in Mesopotamia and Khuzistan, this expansion involved the typical Sasanian combination of city-founding, construction of irrigation works, deportations and forcible resettlements." 25

Ardashir to Philip I

63

While outlays on larger armies and the administrative apparatus under Ardashir and Shapur I increased, the potential to collect taxes from newly acquired territories also increased. With evidence for construction and enlargement of ports in the Persian Gulf, which were likely military in their purposes, the capacity for increased trade and accompanying taxation was also possible. Ardashir was credited with the foundation or enlargement of no less than eight ports in the Persian Gulf and the rivers of lower Mesopotamia. 26 The site of Siraf on the Iranian coast of the Persian Gulf was among the most important of these port foundations and there is evidence indicating Sasanian expansion throughout the gulf under Ardashir and through the rest of the dynasty's history. On the east Arabian coast of the gulf, a colony and port at Bitn Ardashir was also founded following a military expedition. 27 With this expansion, access to luxury trade originating in the gulf but also in south Arabia and via the sea routes with India had the capacity to increase. With larger urban populations in Iran and Mesopotamia, demand for such items was a likely by-product. This not only had the potential to increase the Sasanian capacity to tax trade, but the consequent diversion of trade bound for the Roman Empire had possible consequences for the Roman administration in its collection of taxes or tariffs on trade. 28

Sasanian military reforms and structures The extent and nature of early Sasanian reforms to the military are a challenging topic due to the nature of the surviving source material. Much of what is known of the structure of the Sasanian army (spah) and the tactics it employed on campaigns comes from western sources or Islamic sources and mostly pertains to later conflicts with Rome/Byzantium and other enemies. There are some earlier references but they are mostly from hostile source traditions. Herodian, for example, described a Persian army encountered during Severus Alexander's campaign of 232/3 as "a horde of men rather than an army" and that it was difficult for the Sasanians to assemble an army because "it was not an organized standing force." 29 This hostile representation has had a marked impact on how the Sasanian army has been viewed in modern scholarship. In the most detailed written account of the structure of an invading Sasanian army, which survives in the Res Gestae of Ammianus Marcellinus, it is clear that the army was highly capable and the successes of Shapur I's invasions in the 250s are another indication. In the third and fourth centuries, generally speaking, the emphasis was on siege warfare but there were occasions in which field battles took place. As had been the case under the Parthians, cavalry was the key element of the Sasanian military. There were some developments in the reigns of the early Sasanian rulers, which saw more heavy-armoured cavalry units (kataphractoi) form part of the Sasanian cavalry forces. McDonough believes that the mail armour worn by the kataphractoi was borrowed from

64

Ardashir to Philip I

Roman-style chain mail armour as a consequence of military engagements with the Romans. 30 Farrokh relies on Herodian's report of a battle outside Ctesiphon in 233 between the invading army of Severus Alexander and Ardashir's defending force to provide some indication of early Sasanian tactics in the field against a Roman army. 31 Severus Alexander's army was taken by surprise by the Persians, surrounded on all sides and trapped "like fish in a net." The Persians then destroyed the whole army according to Herodian. Farrokh emphasises here the likely role of the horse archers, although Herodian does not specifically mention them, and also suggests that the Persian cavalry may have been using a "scatter-coalesce" technique adopted from the Central Asian frontier. 32 This technique was a development of the Parthian technique of quick withdrawal from an engagement, which encouraged the Romans to pursue before the cavalry employed the Parthian shot and then quickly rounded and enveloped the pursuing army. This was the technique most famously employed against Crassus at Carrhae. The scatter-coalesce approach relied on the Persians coalescing at pre-determined points, which required considerable forward planning and co-ordination. The other field engagements, which are known to have taken place, are the battle of Meshike between Gordian III and Shapur in 244, the battles at Barbalissos and Edessa/Carrhae in 252 and 260 respectively, the battle of Satala in 297 and the Battle of Singara in 344. The relatively little that is known of tactics employed by the Sasanians in these battles is considered in later chapters and it is only the Battle of Singara from which sufficient information survives. Interpreting information about the Battle of Singara is challenging because it comes from two panegyrics that are heavily pro-Roman. Perhaps the most detailed knowledge available of Sasanian military capabilities relates to siege warfare, especially in the 250s and late 350s. The Sasanian siege of Dura Europos of the mid-250s provides unparalleled archaeological evidence for an ancient siege and the siege of Amida in 359, reported by Ammianus Marcellinus as an eyewitness, remains one of the most detailed accounts of a pre-modern siege. The relevant details of these sieges are considered in the specific chapters to which they relate but in overall terms it is important to emphasise that the Sasanians possessed advanced and sophisticated abilities in siege warfare. The ability to operate in close-quarters during sieges is a possible indicator of an improved infantry capacity of the Sasanian army, certainly when compared with that of the Parthians, despite Ammianus' derogatory claim that the Sasanian infantry was comprised of little more than slaves conscripted from the peasantry. 33

"State religion" and the early Sasanians Significant religious changes in relation to the state are often highlighted as a hallmark of the early Sasanian regime and combined with imperial patronage of Christianity in the Roman Empire in the fourth century are

Ardashir to Philip I

65

used to indicate an important development in the nature of the military and political relationship between Rome and Persia. It is reasonably clear that Ardashir, Shapur I and all of their successors placed an emphasis on Zoroastrianism as an important element in promoting their legitimacy as rulers. Longstanding links between Sasan and Papak, as centrally important figures in the establishment of the Sasanian dynastic line, and the fire temple of Anahid at Istakhr in Persis formed an important element in Sasanian imperial emphasis on Zoroastrianism as part of a programme designed to advertise imperial legitimacy. In the Sasanian historical tradition, the Zoroastrian priestly hierarchy emerged under the Sasanians in a powerful position in relation to the Shahanshahs. This became evident with the emergence of individuals as leaders of the Zoroastrian religion in the third century from the reigns of Ardashir to Narseh. The Zoroastrian priest (Herbed) Tansar is reputed to have taken the lead early and was an important figure in Ardashir's early patronage of the Zoroastrian religion and its elevation to the status of what some refer to as a "state religion." In a surviving work known as the Letter of Tansar, which Mary Boyce argued was authentic to the period of Ardashir I's reign, Tansar claimed "For Church and State were born of the one womb, joined together and never to be sundered." 34 Tansar purportedly worked to promote this idea and as part of this attributed Ardashir with the restoration of the Avesta, the Zoroastrian religious texts destroyed by Alexander the Great who was still described as the "cursed" for doing so. Ardashir made a clear choice to retail the close links between the dynasty and Zoroastrianism via a reverse image on silver drachms of his reign depicting a lit fire altar, likely the regnal fire lit at the beginning of his reign as Shahanshah, a practice that would continue under his successors. While it is clear that Zoroastrianism was of importance to Sasanian rulers, there is a debate regarding its status as a "state" religion and the ramifications this had for other religions, especially Christianity and Manichaeism. Some believe that Zoroastrianism quite clearly enjoyed the status of "state religion." For example, Kreyenbroek observes, "The Sasanian era was clearly a vitally important phase in the history of Zoroastrianism. As a state religion, and one of the pillars of Sasanian ideology, Zoroastrianism was able to develop practically without restraints, and it was in this period that the Zoroastrian tradition assumed or developed some of the features that later came to seem characteristic of Zoroastrianism generally. It has been said with some justice that this was the period in which a Zoroastrian church came into being - an ecclesiastical organisation based upon a coherent hierarchical system of priests and fire temples which covered much of the Sasanian Empire." 35 Farrokh points to potential military significance in the adoption of Zoroastrianism, noting the claims of Islamic-era writers that Ardashir appointed spiritual supervisors to each unit of 1,000 savaran (cavalry archers) as part of encouraging bravery and "martial ardour" on the battlefield. 36

66

Ardashir to Philip I

Daryaee similarly emphasises Zoroastrianism as a state religion from Ardashir onwards. 37 The significance of Zoroastrianism and imperial legitimacy is linked by Daryaee to a tradition that extended back to the Achaemenids, which Kreyenbroek similarly holds to: "The link between the king 's authority and his righteousness in the eyes of God is strongly implied in the Achaemenian inscriptions, and also in Avestan mythology is said to depart from the king who becomes a sinner." 38 Some of this imagery is indicated in Ardashir's investiture relief at Naqsh-i Rustam where he receives a diadem from Ahura Mazda, while the horse he rides tramples what is believed to be the last Arsacid Parthian monarch, Artabanus V. Ahura Mazda is also astride a horse and trample s a representation of Ahreman, the source of all evil in Zo roastrianism. In th e SKZ inscript ion of Shapur I, also from Naqsh-i Rustam, Shapur I refers to himself as "the Mazda-worshipping divine Shapur. .. of the race of the gods, son of the Mazda-worshipping divine Ardashir. .. of the race of the gods" providing an indication of the development of this strong message of dynastic legitimacy linked to the Zoroastrian religion. 39 There are some who believe, however, that the term "state religion" is not appropriate to describe Zoroastrianism under the Sasanians. Drijvers

Figure 3.2 The Ka'ba-ye Zartosht (Kaaba of Zoroa ster) at Naqsh-i Rustam, Fars Province , Iran. Photograph: Ross Burns.

Ardashir to Philip I

67

notes opposition to the idea among scholars such as Schippmann, Rubin and Wiesehofer and points out that the close relationship that existed between Zoroastrianism and the state under Bahram II, for example, was not permanent across the entire period of Sasanian Persia's existence and was more exceptional than representative. 40 He also holds that "Zoroastrianism never became the state-sponsored religious orthodoxy. The kings seem only to have sought support of the Zoroastrian priesthood and given Zoroastrianism a certain degree of domination over other religions in times of internal and political problems." 41 On this analysis, Kerdir's domination during the reigns of Bahram I and II was unusual rather than typical and Shapur II's persecutions of Christians in the fourth century were only politically motivated. There were "long periods of religious openmindedness in which Zoroastrians had to accept the existence of other religious groups and competitors within the Sasanid Empire." 42 Two recent studies make some important points on religion in the Sasanian empire of the third and fourth centuries and further inform the debate on the role of religion in the relationship between Rome and Persia. Richard Payne refers in broad terms to Zoroastrian intolerance of minority religions, including Christians, as a "myth," even during the tenure of Kerdir as mowbed, emphasising a role for Christians (and Jews) "as long as they refrained from challenging the superiority of the Good Religion." 43 Kyle Smith undertakes a detailed contextual reading of relevant fourth-century and later texts to propose that the Christians of Persia were not a significant motivator for Constantine's projected Persian war and that the identification of Constantine as a possible liberator has been misunderstood. 44 These debates are considered in further detail in the relevant chapters of this book but it is important to note them here as part of a broader consideration of the key issues.

Early Sasanian imperial expansionism and Roman responses to it While Herodian described Ardashir as a bold military protagonist in the east the poor state of Roman preparedness for any military ramifications of Ardashir's emergence as Shahanshah received most of the attention in surviving Roman texts. 45 Cassius Dio believed that the new ruler "was of no particular account himself," blaming any Roman fear of the Sasanians and their military capacity on a distinct lack of military preparation and discipline. 46 Evidence from other traditions indicates that the Romans offered support to those who sought to challenge Ardashir soon after he came to power, especially the rulers of Armenia and Hatra who were still connected to the Arsacid regime and accepted some royal and noble fugitives after the overthrow of Artabanus V. The Arsacid family and the nobility surrounding it had long been deeply divided and some members, it would seem, pledged allegiance to the new Sasanian regime. While over time the

68

Ardashir to Philip I

Sasanians were largely successful in obscuring the Arsacid Parthian contribution to Iranian national history, the Parthian noble families survived with enough influence to emerge later as part of what one scholar has recently termed "the Sasanian-Parthian Confederacy." 47 The ongoing role and importance of the Surena (head of one of the most important Parthian noble families), for example, as a military leader in the Sasanian period is clear in a number of texts of the period and later.

Armenia as an early source of contention Resistance offered against the emergence of Ardashir as the supreme ruler of Iran by some surviving members of the Arsacid Parthian family was nowhere more obvious than in Armenia, whose king, Tiridates, was a close relative of the overthrown Arsacids. 48 This resistance would have immediate implications for the relationship between Rome and Sasanian Persia. Arrangements that had previously existed between Rome and Parthia over Armenia typically involved compromises over the nomination and formal investiture of the Armenian king. The Parthians had regularly nominated family members to the position, which created potential difficulties for themselves and Roman rulers. With the Sasanian overthrow of the Parthians, a distinct change took place in relation to Armenia, where Roman emperors now actively supported Arsacid claimants and their descendants to the Armenian throne against the wishes of the Sasanian Shahanshahs who sought to install close family members as rulers of Armenia. Arsacid claimants and their families sometimes appealed to Roman rulers for assistance or fled to them when the Sasanians threatened or succeeded in removing them. If the Armenian historians of the fourth and fifth centuries are to be believed, Armenia not only played a centrally important role in resisting the Sasanian overthrow of the Parthians, but its rulers sought on a number of occasions to reverse it. An emphasis on the close blood ties between the Armenian king, Tiridates, and the last Parthian ruler, Artabanus V, was emphasised by the Armenian historians, while this was also a factor in resistance from the city of Hatra. 49 According to Agathangelos, in the year following Ardashir's defeat of Artabanus V (i.e. 225/6), Tiridates appeared at the head of a confederacy of warriors from Armenia, Iberia and Albania together with tribal groups from north of the Caucasus and invaded Persian territory all the way to Ctesiphon. 50 In this instance, the Alans and the Huns were deliberately allowed to pass through the heavily guarded Caspian gates to be used against the Sasanians. Only a few years later, in 227/8, Tiridates purportedly led a similar attack on Sasanian Persia with claims that Seleucia-Ctesiphon was captured a second time. 51 Ardashir was preoccupied during this period with problems in his eastern domains but sought to deal with the ongoing difficulties in Armenia by mounting a punitive attack on the kingdom. This failed due to the efforts of an alliance

Ardashir to Philip I

69

of Armenians and exiled Parthians, including the sons of Artabanus V. 52 Ardashir's attempts to deal with the Armenians by issuing military threats against them, and perhaps also scoring a victory over them, saw Tiridates explicitly request the assistance of the Roman emperor Severus Alexander ea. 229. 53

Hatra and the Persian Gulf While Ardashir sought to deal with the troublesome Armenians, he also turned attention to the principality of Hatra. The fortified city of Hatra had held out against Roman sieges prosecuted by Trajan and Septimius Severus in the second century, its formidable defensive circuit frustrating the efforts of the Romans on three separate occasions. Dio claimed that an attack by Ardashir on Hatra in 229/30 was designed to provide a forward base for the Persians to attack Roman territory further west. 54 In reality, however, the city was too remote to be effective in this respect, sitting approximately 200 km east of the Euphrates and 100 km south-east of the nearest Roman fortification at Singara, which was itself difficult for the Romans to supply. In an Arabic tradition reported by Tabari, the Hatrene king at the time, Daizan, controlled a large amount of territory extending all the way to the Euphrates and Ardashir sought revenge on the king for his treachery while the newly established Sasanian ruler was away fighting in Chorasan on the Central Asian frontier. 55 Like the Roman assaults on Hatra in the second century, Ardashir's attack failed and in 235 there is clear and datable evidence placing a Roman auxiliary unit, Cohors IX Maurorum, at the city. 56 It is, of course, possible that the Romans were supporting Hatra earlier and that this was a factor in Ardashir's reasons for attacking the city. 57 Sartre believes that Hatra should be considered a Roman ally from ea. 227 onwards, while Sommer holds that it was a client-kingdom from approximately the same time. 58 Further to Sasanian attempts to deal with Armenia and Hatra, while also managing issues in the eastern dominions, Ardashir turned attention to the Persian Gulf. Piacentini discusses the early interest Ardashir took in the coastal areas of Persis that lay on the gulf and there is evidence for sustained interest by the early Sasanian rulers throughout the gulf more broadly. 59 During much of the Parthian period, the kingdom of Characene had wielded considerable power throughout the gulf and had enjoyed varying levels of independence from the Parthians, even looking favourably towards Rome at times. A number of later Arabic texts referred to Ardashir's military actions in Oman (Mazun), with one claiming that he captured the port of Ash Shihr in eastern Yemen and also installed the dominant tribal group in Mazun. 60 While Potts cautions that this is a "remarkably early projection of Sasanian power well beyond the Persian Gulf," he indicates that there is reasonably solid support for the claim via the SKZ inscription of Ardashir's son, Shapur I, who claimed overlordship of Mazun (in the

70

Ardashir to Philip I

Parthian version). Tabari referred to the Tanukhid Arab ruler Jadhima and his nephew 'Amr ibn Adi who were based at Hira and ruled over southern Iraq and the Badia in the Hijaz from ea. 250, as allies whom the Persians relied on "to keep the adjacent Arabs under control." 61 The origins of these arrangements possibly lay in Ardashir's reign. It is noted earlier that Ardashir was attributed with the foundation of eight ports in the Persian Gulf, and it is probable that they had a primarily military purpose. 62 Ardashir's actions in consolidating Sasanian power in the gulf and military activity in Oman would have increasingly important ramifications for the Arabian Peninsula as the third century unfolded and eventually contributed to a new front in competition between Rome and Persia.

Ardashir and the rhetoric of war with Rome Following attempts to deal with ongoing elements of Parthian resistance in Armenia and Hatra, Ardashir made it clear that the new regime would not tolerate the Roman presence in northern Mesopotamia. In much-discussed passages in the works of Cassius Dio and Herodian, Ardashir wrote to Severus Alexander ea. 229 claiming that territory as far west as the coastline of Asia Minor, which had been under Achaemenid rule prior to the invasions of Alexander the Great, belonged to the Iranian Empire by inheritance. 63 This was echoed over a century later in a similar claim recorded by Ammianus Marcellinus in which the then Shahanshah, Shapur II, made demands of Constantius II as part of the diplomacy preceding the war of 359/60. 64 Tabari attributed a more personal motive to Ardashir but linked it with the Achaemenids, claiming that the Shahanshah wished to avenge the death of the last Achaemenid ruler, Darius III: "He (Ardashir) arose, as he maintained, to avenge the blood of his cousin Darius ... who made war against Alexander and who was murdered by his two servants." 65 In the last confrontation between Rome and the Parthians in 217, Caracalla had styled himself as Alexander the Great and the very name taken by one of his successors, Severus Alexander, was designed to link him to memory of the still-revered Macedonian King and General. 66 Referring to events two centuries earlier in the reign of Tiberius (ea. 18), Tacitus and Dio provided an account of an embassy sent by the Parthian ruler, Artabanus II, to Tiberius, which spelled out a desire to restore the former western boundaries of the Achaemenid empire (see Chapter 1).67 Some reject the authenticity of the claims purportedly made by Sasanian rulers suggesting that Dio, Herodian and later Ammianus were essentially putting such claims in their mouths. This was driven by a classicising model of writing history that sought to cast conflict between Rome and Persia in terms of the great conflicts between Achaemenid Persia and Classical Greece in the fifth century BC. 68 Proponents of this approach suggest that the Sasanians had little knowledge of the extent of Achaemenid territorial possessions prior to Alexander the Great's invasions with some suggesting

Ardashir to Philip I

71

that they looked more to the Kayanids of eastern Iran as their ancestors or that their origins may even have lain with the Indo-Parthians. 69 Shayegan takes a different position by proposing that the identification by Roman authors of Caracalla with Alexander the Great during his war with the Parthians, and Severus Alexander's adoption of his name, invited the idea of similarity or even "parity" between the Sasanians and the Achaemenids who Alexander had triumphed over as a worthy foe. 70 When the Sasanians under Ardashir proved to be more successful against Roman armies than had been initially thought possible, the Roman authors switched their rhetoric to one in which the Sasanians were the equals of the earlier Achaemenids who had enjoyed success over the Greeks. Shayegan points out that an examination of Shapur I's SKZ inscription carved sometime in the 260s made territorial claims that were only limited to Mesopotamia, Syria, Cilicia, Armenia and Cappadocia, territories Shapur conquered in the 250s. 71 There are no references in the SKZ to the Achaemenids or claims to the territory they once controlled all the way to the borders with Classical Greece. It is difficult to accept that the Sasanians were ignorant of the Achaemenid Persian past and the territorial supremacy they once enjoyed and especially how this might be used to rhetorical effect in diplomacy with their Roman enemies. Given the importance of Achaemenid-era monuments located in Persis, especially Naqsh-i Rustam, to the Sasanian dynasty and the ongoing use of the Greek language in the Parthian and Sasanian worlds, knowledge of the powerful Achaemenid past was surely available to them. 72 The coinage of the pre-Sasanian rulers of Persis also provides hints that there were ongoing references in the kingdom to the Achaemenid past, which the Sasanians themselves adopted. Aramaic legends were used on the pre-Sasanian coinage rather than Greek ones and the names of some rulers harked back to the powerful Achaemenid kings. 73 Reverse iconography depicting a fire altar and the king standing next to it was a clear link to Mazdaean worship practised by the Achaemenids, and likely influenced the distinctive reverse iconography of Sasanian imperial coinage. In terms of imperial rhetoric and claims, there was much to gain by appealing to the powerful past of the Achaemenid Persian Empire, especially when Rome's presence on the landscape of the region was relatively recent. The most sensible suggestion is that of Canepa who proposes that the Sasanians did include these territorial claims in their early diplomacy with Rome but that they were primarily rhetorical. 74 Dio and Herodian embellished the claims of Ardashir perhaps with the purposes attributed to them by Shayegan. Certainly, by emphasising or embellishing these claims Dio and Herodian indicated just how much they believed the stakes had been raised in terms of the threat Rome now faced. This was in spite of the earlier claim of Dio that Ardashir was of no account himself and that Rome was vulnerable due to its own poor preparation and ill-discipline in the Roman army. 75 A related issue to this is the Roman use of the term

72

Ardashir to Philip I

"Parthians" to describe the Sasanians despite clear evidence for knowledge of the Sasanian overthrow of the Parthians. This is discussed further in Chapter 8 and especially in the conclusion.

Initial Roman responses to Sasanian diplomatic and military threats Ardashir's threats sparked panic across the Roman eastern provinces, according to Dio, and even resulted in mutinies among some of the Roman garrisons. 76 In response, Severus Alexander reminded the Sasanian ruler of Rome's victories over the Parthians, especially those of the second century by Trajan, Lucius Verus and Septimius Severus, suggesting that a similar outcome would transpire if Ardashir invaded. 77 Ardashir ignored the diplomatic warnings and invaded anyway, laying siege to the fortress of Nisibis, metropolis of the Province of Mesopotamia while also raiding Armenia, Syria and Cappadocia. 78 In 230/1 Severus Alexander began planning a retaliatory campaign, which did not get underway until the following year. During this period of preparation large issues of denarii were minted in Rome with reverse legends and images depicting strong military themes. These included IOVI PROPVGNATORI, MARS PROPVG and MARS VLTOR. 79 Necessary and expected optimism was also advertised with bronze issues depicting VICTORIA. 80 The army, which set out from Rome late in 231, comprised raw recruits and apparently veterans of both Septimius Severus' Parthian campaigns and those of Caracalla. 81 If Herodian is to be believed on this claim, some of Severus' veterans were aged in their 50s or 60s. The eastern provinces, including Egypt, contained 11 legions and numerous auxiliary units at the time, which could be drawn upon in the event that the campaign became a full invasion of Persian territory. 82 Some troops from the Danubian provinces were added during the march from Rome to Antioch. The Armenian historian, Moses Khorenats'i, summed up the military preparations of Severus Alexander, claiming that he raised troops from the Black Sea to Egypt.83 Following the emperor's arrival in Antioch in the winter of 231/2, the mustered army was, indeed, significant, signalling that the campaign was not simply designed to repel the attacks of Ardashir but potentially sought to prosecute a full-scale invasion. The rudiments of diplomacy between the emperor and Shahanshah were exercised but neither side appears to have been genuine in their hopes for its outcome. Severus Alexander offered peace, friendship and intimidation all at once, while Ardashir responded with the same rhetoric he had employed in earlier diplomacy; the Romans should withdraw from Syria and Asia and hand back what rightly belonged to the Persians by virtue of inheritance from the Achaemenids. A 400-strong embassy of tall and immaculately dressed Persians, possibly

Ardashir to Philip I

73

designed to impress upon the emperor the claim to Achaemenid heritage, was summarily dismissed by Severus Alexander. The unfortunate members of the embassy, we are told, were reduced to servitude and resettled in the wilds of Phrygia. 84 The campaign was directed across a broad geographical area and has been variously analysed. 85 According to Herodian, who provided the most contemporary extant account of the campaign, one part of the army attacked Media via Armenia, a second column went south, possibly via the Euphrates to the Tigris, Seleucia-Ctesiphon and the Persian Gulf, while a third part of the army under the command of the emperor may have attacked via Singara. 86 It is likely that the first column had the support of Armenian and Iberian forces and was thus the most successful. 87 While the northern army achieved successes, the Persian army was by-and-large able to fend off the attacks directed further south, particularly on SeleuciaCtesiphon. With little success to be gained from continuing the campaign and problems requiring immediate attention on the Rhine and Danube frontiers, a hasty Roman retreat took place. The mixed successes and failures of both armies saw the emergence of different traditions about the outcome of the war afterwards. Herodian's lack of exactness in reporting some important aspects of the campaign raises doubts about his own sources for it; nonetheless, his version became the basis of a tradition, mostly in Greek, that the campaign was an ill-omened failure. 88 A tradition preserved in Latin texts of the latter half of the fourth century, which are thought to have relied on a now lost intermediate Latin account referred to in modern scholarship as the Kaisergeschichte, reported the campaign as a success. 89 It was hardly a memorable one if the Historia Augusta's reports of a somewhat underwhelming triumph in Rome are believed.90 Severus Alexander purportedly took the title Persicus Maximus, although he would not have been the first emperor to take a grandiloquent title in the face of underwhelming achievement. Neither the Greek nor Latin traditions reported a treaty following the war, which is itself suggestive of a stalemate. The Persians did not engage in any military actions directed at Roman territory for the following four years, while the Roman retreat could be seen as a victory for the Sasanians. This first major confrontation between Roman and Sasanian imperial rulers and their armed forces had some similarities to those between Rome and the Parthians in the second century but there were considerable differences regarding the outcome. The immediate cause of conflict followed by invasions of Parthia by Trajan, Lucius Verus and Septimius Severus was unrest in Armenia and in the latter two cases, threats to Roman territorial claims in northern Mesopotamia. In the case of Severus Alexander's campaign, a Persian attack on northern Mesopotamia together with instability in Armenia preceded Persian attacks in Syria and even Cappadocia. The war prosecuted by Severus Alexander was essentially fought to a stalemate, which might partly be explained by poor planning and execution on the

74

Ardashir to Philip I

part of the Romans but is also an indication of the renewed strength of the Sasanians. Severus Alexander's underwhelming campaign against the Persians in 232/3 was also associated with growing difficulties in the west. Early in 234, he received urgent dispatches from the Illyrian governors requiring his presence to deal with tribal raids across the Rhine and Danube but this cannot have been the first news he had received of what was clearly a growing problem. 91 The following year, a serious revolt of the soldiery in the Danubian provinces of Pannonia and Moesia, unhappy at Alexander's response to the emergency, saw the end of the Severan dynasty and the elevation of the first of the soldier-emperors, Maximinus. Maximinus would spend almost his entire reign fighting against the Germanic tribes on the Rhine and the Sarmatians and Dacians on the Danube. 92 While the apparent quietness in Armenia and Mesopotamia following Severus Alexander's Persian campaign was a likely outcome of the war and Roman preoccupations in the Illyrian provinces, it was also due to Ardashir's necessity to deal with ongoing issues on the Sasanian eastern frontier. 93 Ardashir's ongoing support for the emerging Kushano-Sasanian dynasty, an associated branch of the Sasanian royal family, in Central Asia and northern India at times received greater focus than the situation in Armenia and Mesopotamia in the west. The situation on much of the eastern frontier of the Sasanian empire was fluid during the 220s and 230s with the Kushan dynasty forced out over time in favour of the Kushano-Sasanians as part of Ardashir's consolidation of Sasanian power. 94 There were a number of occasions in which the old Kushan dynasty attempted to reinstate itself with varying levels of success and this required the direct attention of both Ardashir and his successor Shapur I. It is also possible that the Sasanian conquest of the southern shores of the Persian Gulf belongs to the period of the mid- to late 230s and with Ardashir's activity in Oman difficult to date precisely, it may also belong to this period. 95

Ardashir attacks Roman Mesopotamia In the late 230s, Ardashir was in a position to launch a new campaign in northern Mesopotamia and this was to have lasting consequences. Sometime during the reign of Maximinus Thrax (235-8), the metropolis of Provincia Mesopotamia, Nisibis, together with Carrhae were captured by Ardashir. 96 Maximinus was in no position to respond, given the near-constant military exigencies of the Rhine and Danube frontiers, as the many VICTORIA GERMANICA coin issues of Rome during his reign amply demonstrate. 97 The short-lived emperor, Pupienus, had been about to prosecute a campaign in 238 against the Persians when he met an untimely death. 98 Hauser prefers to place the Sasanian capture of the Mesopotamian cities of Nisibis and Carrhae in the context of the siege and capture of Hatra in 240/1. 99 Dura Europos on the middle Euphrates provably came under

Ardashir to Philip I

75

Persian attack in April 239 according to a dated grafitto inscription from the site and Julius Terentius, the Tribune of Cohors XX Palmyrenorum based at Dura lost his life in battle against the Sasanians likely, in the context of this attack. 100 The capture of Hatra was apparently undertaken initially at Ardashir's direction and in the same year that his son, Shapur I, assumed the diadem, apparently as a co-ruler of the empire, which would place the siege and capture of the formidable fortress in 240/1. 101It appears that Sasanian attacks on Roman Mesopotamia were sustained from 238 to 241 and that the combined evidence for the attack on Dura and the capture of Nisibis, Carrhae and Hatra suggests a programmatic Sasanian approach to taking control of Roman Mesopotamia. 102 The siege of Hatra was a momentous event, according to Tabari, who noted one later Islamic report that claimed it took four years, although he personally preferred a different account suggesting it took two years. 103 As both Hauser and Gawlikowski point out, the surviving wall circuit and archaeology suggests that significant resources were required to capture the city. 104 Tabari attributed the Sasanian attack on Hatra to the treachery of its king, while Shapur was earlier occupied on the eastern frontier in Chorasan. This is the same accusation that was levelled at the Hatrene king by Ardashir in 229 if we are to believe Tabari. The removal of Hatra as a powerful regional presence in the north Mesopotamian steppe was to have significant longer term consequences, especially in relation to the control of Arab tribes. 105 This was similar to the situation Rome experienced in the steppe and desert regions around Palmyra after its demise at the hands of Aurelian in 272/3.

The Roman response - Gordian III's Persian campaign It would not be until 242 that a concerted Roman response to the Sasanian capture of most of Mesopotamia eventually got underway. The Persian invasion of Gordian III would increase the stakes in military competition between Rome and Persia and mark the first clear defeat experienced by a Roman emperor on campaign against the Parthians or Persians since the dying days of the Republic. According to a near-contemporary text, the Thirteenth Sibylline Oracle, it could be argued that the war took on a "global" significance when Gordian responded to "an uprising of enterprising Persians, together with Indians, Armenians and Arabs." 106 The reference to Indians possibly reflects Ardashir's ability to draw troops from an expanded area of Sasanian control in northern India via the Kushano-Sasanians and the reference to Arabs is an early indication of alliances with Arabs, possibly as a result of earlier Sasanian military activity in the Persian Gulf and Oman (Mazun). 107 Gordian III's campaign was the first conflict between Rome and Persia that received attention in a Persian text, namely Shapur I's SKZ inscription.108 The SKZ demonstrates a decidedly different position to that of the surviving Roman texts and links the outcome of the war with Gordian to

76 Ardashir to Philip I

the wars Shapur fought with Rome in the 250s. Much has been made of the SKZ as a source since its discovery in the 1930s, partly because it is seen as providing an "eastern" perspective in contrast to the dominance that the "western," that is predominantly Latin and Greek, texts usually enjoy in reporting and interpreting events. There is a tendency in scholarship, therefore, to privilege the SKZ because it is unusual to have a non-Roman perspective on events that the Romans were profoundly involved in. The SKZ is also broadly contemporary with events having been produced sometime in the 260s. As a trilingual inscription carved in Greek, Middle Persian and Parthian, it was intended for a broad audience, possibly including Roman diplomats. The inscription provides a considerable amount of valuable information, which is not available in other sources, and it allows a reconstruction of some of the key events in the Persian invasion of Gordian III and the invasions of the Roman eastern provinces by Shapur I in the 250s. The inscription also provides a detailed account of how Shapur I conceived of his empire late in his reign. It is important, however, to avoid "privileging" the SKZ in some way because it runs the obvious risk of obscuring problems with the source and its claims. 109 The SKZ makes no reference, for example, to setbacks suffered by Shapur or victories scored by the Romans and as a piece of imperial rhetoric, we would not expect it to. Its perspective is a decidedly imperial one and is good evidence for the similarities in imperial rhetorical programmes of both Roman and Sasanian rulers. The surviving Roman texts focus almost entirely on the fate of Gordian III and the apparent treachery of his Praetorian Prefect Philip. Ammianus Marcellinus referred to a successful battle by the Romans over the Persians at Rhesaina, the base of Legio III Parthica in Mesopotamia, which probably took place in spring 243. 110 Ammianus also noted that Gordian III's tomb was still visible at a location referred to as Zaitha on the left bank of the Euphrates, upstream from Dura Europos, when the army of Julian was invading Persia in 363. 111 There is little else available in the Roman texts that sheds light on the campaign. Some analysis of Roman coinage is potentially useful in attempting to better understand and date the progress of Gordian's army as it made its way to Antioch prior to the campaign. He had been detained for a not insignificant period on the way to the east by serious problems in Dacia, Moesia and Thrace following attacks across the Danube by the Goths, Carpi and Sarmatians. 112 According to numismatic research by Kettenhofen and earlier by Karl Pink, Gordian departed Rome in 242, marching via various cities of Asia Minor before arriving at Antioch in Syria. 113 The mint at Antioch produced a series of antoniniani at this time that have been linked to Gordian's war against the Persians and this is potentially useful as evidence for campaign preparations. Pink identified 13 types of antoniniani with reverse legends and iconography suggestive of military activity and identified them as issues of the Antioch mint. These were later incorporated into Roman Imperial Coinage Volume IV.3 as RIC Gordian III nos.

Ardashir to Philip I

77

206-19. 114 The coins include reverse legends such as FIDES MILITVM, MARTI PACIFERO, ORIENS AVG and three VICTORIA types, while the reverse iconography is unmistakably military. Pink was convinced that these coins were indicative of a reopening of the mint and celebration of the recapture of Antioch by Gordian reported by the Historia Augusta. 115 With the Historia Augusta's claim of the capture of Antioch by the Persians in the late 230s and its subsequent recapture by Gordian long dismissed, the "Persian war series" of the Antioch mint dating to the period 242/3 can now be seen as evidence for the Roman emperor's organisation and promotion of the impending campaign against Shapur I. Paying the soldiers and producing time-honoured images of military success fulfilled an important practical concern while expressing necessary confidence in the outcome. In spring 243, the army departed Antioch and marched towards Carrhae via the bridge crossing at Zeugma. 116 The Roman victory at Rhesaina over the Persians reported by Ammianus likely followed. The battle at Rhesaina is a critically important piece of evidence for several reasons. Firstly, the presence of a Persian army there confirms that the Sasanians were in control of much of northern Mesopotamia and it is possible that they held Rhesaina as well as Carrhae and Nisibis (and probably Singara too). The resumption of coin minting at Nisibis in Gordian III's name and the commencement of minting in his name at Singara is possible confirmation of Roman success in forcing the Persians out of Mesopotamia in the wake of the success at Rhesaina .117 The SKZ made no reference to the defeat at Rhesaina and focussed instead on a great victory by Shapur over Gordian and his army at Meshike on the lower Euphrates, which can be dated to sometime between mid-January and mid-March 244. 118 In the wake of this victory, Meshike was refounded,

Figure 3.3 Medallion, Rome. Gordian III. AD 243-244. Obv- IMP GORDIANVS PIVS FELIX AVG: laureate, draped, and cuirassed bust left. RevADLOCVTIO AVGVSTI: Adlocutio scene: Gordian, assisted by soldier behind him, standing right on low dais, addressing four soldiers, two of whom hold spears and shields, standing left; vexillum, signum and aquila behind. Gnecchi II, 4 (pl. 103, 3). Courtesy Classical Numismatic Group.

78

Ardashir to Philip I

according to the SKZ, as Peroz-Shabur (Pirisabora (Latin), modern Anbar), meaning "Victorious is Shapur." The Roman texts focussed almost entirely on the treachery of Philip, then Praetorian Prefect, as the key reason for the young Gordian's death and made no reference to it occurring in the context of a battle with the Persians. Only the Thirteenth Sibylline Oracle indicated that Philip's treachery took place in the context of a battle on the Euphrates with the Persians: A young Ares, he (Gordian) will lead spearmen against the Assyrians. Unto the Euphrates, deep flowing, silver shall warlike Ares stretch forth his spear for the sake of vengeance. Betrayed by his colleague he will fall down in the ranks, smitten by gleaming iron. 119 Ammianus Marcellinus visited the tomb of Gordian at Zaitha in 363 during the emperor Julian's ill-fated march into Persia and recreated a speech made by Julian on the occasion. 120 It reflected a belief that Gordian had been killed at the location where he was buried, "in hoe ubi sepultus est loco" (that is Zaitha) sometime after his victory over Shapur at Rhesaina. Zaitha was located approximately 200 km north of Meshike/Pirasabora and considerably south of Rhesaina, which poses some questions about Ammianus' account. If Gordian was killed at Zaitha due to a conspiracy led by Philip on the way down the Euphrates from Rhesaina, did Philip as the emperor continue the invasion into Persia before meeting Shapur at Meshike where he and the army were defeated? How would this account reconcile with the SKZ's claim that Gordian was the emperor defeated at Meshike and that a number of rock reliefs put up by Shapur in the 260s depict a dead Roman emperor being trampled under Shapur's horse, who can only be Gordian? The most likely explanation is that Gordian and the army, with Philip as Praetorian Prefect following the death at some stage of the previous Praetorian Prefect, Timesitheus, had marched down the Euphrates after the victory at Rhesaina before being defeated in battle at Meshike where Gordian was killed. Following negotiations led by Philip, the army returned via the Euphrates, bringing back Gordian's body where he was buried at Zaitha which was, at the time, considered to be within Roman territory. Ammianus mistakenly presumed that Gordian had been killed at Zaitha when Julian's army visited the tomb 120 years later. It is also worth noting that Zosimus believed Gordian died in the enemy's land and that the Epitome de Caesaribus claimed that while Gordian died in Ctesiphon his Cenotaph was on the borders between Rome and Persia. 121

The agreement between Philip and Shapur With Gordian III dead at Meshike, his replacement, Philip, was left with little option but to strike an agreement with Shapur. 122 It appears that a treaty followed, which contained provisions primarily of a financial nature but may have included a non-interference clause with regard to Armenia.

Ardashir to Philip I

79

A one-off ransom payment of 500,000 denarii, possibly for the return of captured soldiers but probably part of secur ing safe passage for the army out of Persia, was an explicit part of the agreement. 123 It is generally accepted that the reference to denarii in the inscription was to gold dinars or aurei. 124 The agreement also included a requirement to pay annual tribute. Some later Byzantine writers claimed that Mesopotamia and Armenia were also given up, although there is no direct indication of this in the SKZ. 125 The suggestion of the cession of Armenia may find indirect support in the SKZ because Shapur later used what he called Roman interference in Armenia as the pretext for his momentous invasions of Roman territory in the 250s. The cession of Mesopotamia is more problematic and scholarly arguments in support of it rely on complex discussions using numismatic, archaeological and papyrological evidence. 126 Philip sought to retail a different version of events on his return from Persia with the army. At Antioch, antoniniani were minted with obverse legends in the name of Philip and reverses related to the events just concluded in Persia. 127 One of these types included the reverse legend PAX FVNDATA CVM PERSIS, which is unique in Roman coinage. The iconography accompanying this legend depicts the goddess Pax holding a branch in one hand and a spear in the other. The peace that had been established with Persia was thus won through military strength. The obverse legends of these coins contain a possible allusion to Philip taking a related victory title. Philip's titulature on these coins is either IMP. C.M. IVL. PHILIPPVS P.F.AVG. P.M or IMP. IVL. PHILIPPVS PIVS FEL AVG. P.M. P.M. is unlikely to mean Pontifex Maximus because the standard inclusion of this title on coins of emperors preceding Philip and, indeed, of other coins of Philip himself is at the beginning of the imperial titulature followed by the years of Tribunician Power and Consulships. 128 There is also epigraphic evidence indicating that Philip took the titles Persicus Maximus and Parthicus Maximus on his return. 129 A milestone in Philip's name recently discovered near the remains of the fortress of Sostra

Figure 3.4 Antoninianus, Antioch. Philip I. AD 244. Obv- IMP CM IUL PHILIPPVS P F AUG PM: Radiate, draped, curaissed bust right. Rev- PAX FVNDATA CVM PERSIS: Pax standing left holding branch & tran verse sceptre. RIC IV Philip I 69. Photograph: Effy Alexakis.

80

Ardashir to Philip I

in Central Bulgaria depicts the titles Persicus Maximus and Parthicus Maximus as part of his titulature. 130 This not only demonstrates a programme retailing victories that were actually scored by Gordian III in the early part of the campaign but the taking of Parthian and Persian victory titles separately. It is possible that this was an attempt by Philip, as it was of other third-century emperors, to retail (untruthfully) a victory that emulated the victorious Roman emperors of the second century and also those of Alexander the Great.

Conclusion In the early confrontations between the new Sasanian Persian regime and Rome, some elements of the old confrontation between Rome and the Parthians remained. Armenia was a significant point of contention and would remain so, as was the Roman province of Mesopotamia. Roman responses to Sasanian military threats to both were similar to the responses of the second century. Initial punitive military actions became attempts at major invasions. In the cases of Severus Alexander and Gordian III, however, there were no resounding successes, mostly failures despite the imperial rhetoric. With Armenia as a central point of elements of Parthian resistance to Ardashir and the new regime, a distinct change is noticeable in competition between Rome and Persia for control and influence in the kingdom. Rome supported Arsacid claimants to the throne, while the Sasanians consistently sought to install one of their own. Sasanian attacks on Hatra and eventual success in this endeavour in the early 240s were at least partly motivated by Roman military support for the rulers of the city and its surrounding territory. Rulers of both Rome and Persia were dealing with ongoing issues on opposing frontiers, although the momentum in these endeavours was decidedly different. Sasanian support for the emerging Kushano-Sasanians in the east was expansionary, as was their activity in the Persian Gulf. For Roman rulers, military challenges and responses on the Danube and Rhine were serious and mostly defensive in nature. Roman rulers, however, were still able to muster two large armies in the space of a decade to confront Sasanian activity in Armenia and Mesopotamia and there are some indications that the Sasanians were stretched militarily at times due to their commitments on the other frontiers. It is important to remove a sense of inevitability in the success and strength of the Sasanian rulers in comparison with those of Rome that is so obvious later in the 250s and 260s. Ardashir's urban foundations and expanding of irrigation projects would be unlikely to bear fruit in terms of additions to the army until the reign of his son, Shapur I. The resources used on them, however, were required immediately. Any potential economic and taxation benefits from port construction in the Persian Gulf would also have taken

Ardashir to Philip I

81

time to develop. While there are some indications of impending difficulties for the Romans, what would unfold in the 250s and 260s was, of course, impossible to foresee. When the Jews of Dura Europos sponsored the painting of dozens of scenes from the stories of the bible on the walls of their synagogue circa 245, they likely looked forward to future generations using them for spiritual instruction and edification despite the Persian attack on the town six years earlier. Their confidence would not suggest an imminent belief in the collapse of the empire or the thought that in the next decade, their town and place of worship would cease to exist.

Notes 1 Agathias 4.24.1; Zonaras 12.15 [572]; Tabari 1.813-23. Shavarebi 2019, 365 notes the claim of Agathias that Ardashir was crowned in 226/7. Frye 1983, 118-9 outlines the case for two acceptable dating schemes for Ardashir's accession either in 224 or in 227. 2 See Morley 2017, 268-78. 3 De Blois 2019, 65 refers to the years 231-249 as a period of the "onset of crisis" but not yet of full-blown crisis. 4 See Potter 2004, 238-9. Potter discusses some of the difficulties that the financial aspects of the treaty together with the costs associated with the Philippopolis project caused for Philip's brother, Julius Priscus, who Philip had left behind as the governor of Syria. 5 See De Blois 2019, 159-62 for a sensible and accessible account of the currency debasement problem in the Roman Empire of the third century. 6 Corbier 2005, 391-2. 7 Corbier 2005, 333. 8 Butcher 2015, 184-5. 9 Butcher 2015, 185 warns that there is no direct evidence of complaints about the problem during the key period of the debasement of the Roman silver denominations (i.e. 215-260), but this is a period in which contemporary Roman texts are especially limited, particularly after 235. The Prices Edict of Diocletian was issued in 301 and set maximum prices on approximately 1,200 items. Fragmentary pieces of the inscription have been found in nearly 40 locations, all but one of them in the eastern provinces; Corcoran 2000, 205. It is possible that the inflationary problems it sought to address were brought on by Diocletian's currency reforms of a few years earlier but there is a debate on this topic. One of Diocletian's specific concerns was the impact of rampant price inflation on the soldiers. Part of the preamble to the edict spells this out: Whatever way everyone's shared security demands our armies be directed, through villages or towns and on every route, effrontery goes to meet them with a spirit of thievery. It ratchets up the prices of things for sale, not fourfold or eightfold but so much that the human tongue's reckoning cannot untangle what to call the accounting and the deed! In sum, meanwhile, by the purchase of one thing a soldier is deprived of his bonus and his salary: he yields to the detestable profits of robbers all the tax the whole world pays to support the armies. By their own hand our soldiers seem to give up the expectation of their own service and the labors they have completed to those who steal from everyone. In this way, day after day, the plunderers of the state itself carry off so much they don't know they have it! Kropff (2016)

82 Ardashir to Philip I

10 11

12 13 14 15 16 17 18 19 20

21 22 23 24 25 26 27 28 29 30 31 32 33

34 35 36

The principal problem, therefore, in the eyes of imperial authorities was greed. Supply of goods was manipulated or even withheld by "greedy" traders eager to turn an opportunistic profit, especially when a marching army temporarily swelled the population of cities. In a market economy, however, this only tells one half of the story. The demand element of the equation places similar emphasis on buyers competing for available goods and services as a factor in pushing up prices. With more coins in circulation, especially when donatives were paid to the soldiers, prices inevitably increased. The problem that the Prices Edict sought to address can be demonstrated more broadly in the Roman economy of the third century. Gobl 1983, 323. Alram 2008, 17. Alram 2008, 22; Garibaldi 2010, 49-50. Schindel 2013, 824. Alram 2008, 20. Alram 2008, 20. Alram 2008, 23. See also Gobl 1971, 25. Alram 2008, 23. Gobl 1983, 323-4. The notable exception during the reign of Bahram II (276-293) was the inclusion of the queen and crown prince on some obverses. In the Khudanameh (Book of Kings), the third-century Sasanian rulers are presented as city builders and sponsors of enlarged irrigation works. Ardashir founded cities such as Veh-Ardashir near Seleucia and Firuzabad in Fars (Persis) (see Christensen 2016, 67). Ardashir also refounded the cities of Charax and Forat as Astarabad-Ardashir and Bahman-Ardashir, respectively, in Mesene in the Persian Gulf. Shapur I refounded Meshike as Peroz-Shabur (Pirisabora, Latin) on the lower Euphrates and also founded or expanded urban settlements in Mesene. In the SKZ inscription, Shapur referred to the transfer of Roman prisoners to Persia, Parthia, Susiana and Assuristan (lines 34-6) and there is evidence for their presence at locations such as the Bandar-i Kaisar at Shushtar. Shapur's palace at Bishapur also demonstrates the presence and skills of these captives. Mousavi & Daryaee 2012, 1077 emphasise the archaeological evidence for this activity. Christensen, 2016, 30; Mousavi & Daryaee 2012, 1078 also emphasise the movement of populations to the cities in these areas from elsewhere in the empire and thus a greater focus on urbanism in the Sasanian period. Mousavi & Daryaee 2012, 1078. Tabari I. 826-7. Christensen 2016, 57-67. Christensen 2016, 176. Whitehouse & Williamson 1973, 31-2. Whitehouse & Williamson 1973, 45. See Young 2001. Herodian 6.7.1. McDonough 2013, 601-2. Farrokh 2017, 151; Herodian 6.5.5-10. Farrokh 2017, 151-2. Ammianus Marcellinus 23.6.83. See McDonough, 2013, 604. Lee 2013, 713-4 believes that Ammianus' account and a later absence of Sasanian infantry in accounts of conflict between Rome and Persia is evidence enough for the Sasanians never developing any prowess in the capacity of infantry. Boyce 1968, 22; Letter of Tansar, para 8 = Boyce 1968, 33-4. Kreyenbroek 2013, 19. Farrokh 2017, 17.

Ardashir to Philip I 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56

57 58 59 60 61 62 63 64 65 66 67 68

69 70 71 72

73 74 75 76 77 78

83

Daryaee 2013, 69-70. Kreyenbroek 2013, 26. SKZ lines 1-2. Schippmann 1990, 92-8; Rubin 2000, 647-51; Wiesehofer 2001, 199ff; Drijvers 2009, 444-5. Drijvers 2009, 444. Drijvers 2009, 444-5. Payne 2015, 57. Smith 2016. Herodian 6.2.7. Cassius Dio 80.4.1. Poursharia ti 2 0 0 8. See Mosig-Walburg 2009, 67-74 for a detailed textual analysis of Armenia during the reign of Ardashir I. See, for example, Moses Khorenats'i 2.71. Syvanne & Maksymiuk 2018, 67. Agathangelos 1.19-22. See Syvanne & Maksymiuk 2018, 67-8. Cassius Dio 80.3.3-4; Moses Khorenats'i 2.71-3. Moses Khorenats'i 2.71-2. Cassius Dio 80.3.2. Tabari 1.827-8. L'Annee epigraphique 1958; 238, 239, 240 = Dodgeon & Lieu 1994, 1.4.5 (p. 33). See also James 2013 on a cheekpiece discovered at Hatra that he believes is unequivocally from a Roman helmet of the third century and may be further evidence for Roman military support for the city. See Palermo 2019, 99-103 for further analysis of the relationship between Rome and Hatra. Sartre 2005, 510. Sommer 2013, 38. Piacentini 1985. See Potts 2008, 198. Tabari 1.768. Whitehouse & Williamson 1973, 31-2. Cassius Dio 80.4.1-2; Herodian 6.2.1-2. Ammianus Marcellinus 17.5.5. Tabari I. 813-4. Herodian 5.7.3. Tacitus, Annals 6.31; Cassius Dio 59.27.2. Yarshater 1971; Rubin 1998, 179-80; Potter 1990, 376. See also Shayegan 2011, 30-8. Potter states his position clearly, "He (Shapur) does not, in fact, seem to have been aware that these lands belonged by right to anyone but the Romans." See also Mosig-Walburg 2009, 26-7. Davis 1996; Edwell 2013, 842; Rezakhani 2018, 41-5; Olbrycht 2016, 31-2. Shayegan 2011, 369-70. Shayegan 2011, 37. See Fowden 1993, 28-36 for a detailed discussion on why the claims of Dio and Herodian should be accepted as genuine. See also Daryaee 1995 for a discussion of why the Sasanians were more aware of the Achaemenid past than has sometimes been accepted in modern scholarship. Garibaldi 2010, 7-8. Canepa 2009, 49. See Canepa 2010 for a detailed analysis of the Sasanian employment of the Achaemenid past in the rhetorical development of a Sasanian national vision. Cassius Dio 80.4.1-2. Herodian 6.2.4. Herodian 6.2.1-6; Zonaras 12.15.

84 Ardashir to Philip I 79 80 81 82 83 84 85 86

87 88 89 90 91 92 93 94 95 96

97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115

RIC IV.234-241; RIC IV.244a-b; RIC IV.245c-248c. RIC IV.521-524. Herodian 6.3.3. Cassius Dio 55.23. Moses Khorenats'i 2.72. Herodian 6.4.4-6; Zonaras 12.15. See Edwell 2008, 160-7. Herodian 6.5.1-2; Edwell 2008, 163-4. An inscription from Palmyra (PAT 0278) dated AD 242 refers to the earlier visit of Severns Alexander and accompanying legions. It is presumed that the imperial visit was associated with the campaign against the Persians but marching an army across open desert from Palmyra to the Euphrates would be fraught with logistical difficulties. Syvanne & Maksymiuk 2018, 71-2. Edwell 2008, 163-7. Barnes 1970; Barnes 1978, 90-4; Syme 1971, 221-36 on the Kaisergeschichte. Historia Augusta, Severus Alexander, 57.1.4-5. Herodian 6.7.2; see also Campbell 2005, 26. Drinkwater 2005, 30. For further details on the military challenges faced by Maximinus, see Huttner 2008, 161-71. See Alram 2007 on the eastern campaigns of Ardashir, which are placed by a number of scholars in the mid-230s. See especially Rezakhani 2018, 68-75. Syvanne & Maksymiuk 2018, 74. Syncellus, p. 443, 3-9 = Dodgeon & Lieu 1994 2.2.1 (pp. 43-4); Both Zonaras 12.18 [581] and Historia Augusta, Gordian III 26.6 referred to the later recapture of Nisibis and Carrhae by Gordian III in 243. Kettenhofen 1982, 28-31 linked an observable break in minting coins at the provincial mints of Nisibis and Carrhae under Maximinus with the claim of Syncellus to place the date of the Sasanian capture of Roman Mesopotamia in 238. RIC IV.2.70-4; 90-4; 115, 116, 121. Historia Augusta, Maximus (Pupienus) and Balbinus 13.5. Hauser 2013, 137; See also Loriot 1975. Baur et al. 1933, 112-4. See Rostovtzeff et al. 1944, 176-85 and Welles 1941 for a discussion of the undated epitaph of Julius Terentius. Codex Manichaicus Coloniensis 18. 1-16 = Dodgeon & Lieu 1994, 1.5.1 (p. 33). Baur et al. 1933, 112-4. Tabari 1.828. Hauser 2013, 119-39; Gawlikowski 1994, 147-84. Sommer 2013, 44. Oracula Sibyllina 13, lines 11-12. Potter 1990, 189-211 covers the Persian war of Gordian III, his death on campaign and the subsequent elevation of Philip in detail but believes that the author of the Oracle was not very well informed on these events. For a detailed consideration of the SKZ as a historical source, including the history of analysing it since its discovery in the 1930s, see Edwell 2010, 166-71. Edwell 2010. Ammianus Marcellinus 23.5.17. Ammianus Marcellinus 23.5.7. Drinkwater 2005, 35. Kettenhofen 1982, 23-5; Pink 1935. Mattingly & Sydenham 1968. Pink 1935, 97-8; 104-5. Historia Augusta, Gordian III 26.6.

Ardashir to Philip I

85

116 Mosig-Walburg 2009, 31-42 provides a comprehensive account of the conduct of Gordian 11l's Persian campaign based principally on the textual sources. 117 Kettenhofen 1982, 29-31. 118 See Dodgeon & Lieu 1994, 355; Huttner 2008, 188. 119 Oracula Sibyllina 13, lines 15-20. 120 Ammianus Marcellinus 23.5.7-8, 17. 121 Zosimus 3.14.2; Epitome de Caesaribus 27.3 122 SKZ line 9. 123 SKZ line 8. 124 Guey 1961; Pekary 1961. 125 Evagrius Historia Ecclesiastica 5.7 = Dodgeon & Lieu 1994, 2.2.3 (p. 45) referred to the eastern portion of Armenia being ceded. Zonaras 12.19 [583] claimed that both Mesopotamia and Armenia were given up by the Romans. 126 See Edwell 2008, 176-8; Palermo 2019, 45-7 provides some detailed analysis of the situation in northern Mespotamia in the wake of Gordian 11l's campaign. 127 PAX FVNDATA CVM PERSIS, SPES FELICITATIS ORBIS and VIRTVS EXERCITVS; RIC IV.3 Philip 1 Antioch No. 69-74. 128 See, for example, RIC IV.3.3-7 P.M. TR.P. III COS. P.P. Pink 1935, 101, 107-8 concluded that P.M. meant Parthicus Maximus and given the Pax Fundata Cum Persis legends of some of these coins, one cannot discount it meaning Persicus Maximus. 129 See Huttner 2008, 190; Korner 2002, 131ft. 130 Sharankov & Hristov 2019.

4

Persian triumph, Roman defeat

Introduction There is little doubt that events of the 250s and 260s combined to present the Roman Empire with some of its greatest challenges, the invasions of Shapur I among the most serious. Confronted with multiple difficulties of an economic, political, military and even epidemic nature, the office of emperor and the various mechanisms of imperial government would face a potentially existential crisis. The very survival of the empire through this period in a form still resembling its past is characteristic of its resilience and capacity to adapt. By contrast, for the Sasanians under the leadership of Shapur I, the dynasty that emerged supreme in the 220s could be said to have come of age. With victories over the Romans stretching all the way to coastal Cilicia, the defeat of two large Roman armies in battles and the capture alive of the Roman emperor himself, the rhetorical and territorial opportunities were enormous. Well may Shapur have emphasised the claim to be "of the seed of the gods." Despite the successes of Shapur and the grandiloquent claims of the SKZ and rock reliefs in various locations in the Iranian world, the long-term outcomes of the invasions of the 250s might be seen at one level as limited. There is no clear indication that the Persians sought to retain the territories and cities they sacked and captured beyond short-term arrangements to organise the deportations of captives. 1 This is sometimes spoken of in negative terms but it was in all likelihood a wise decision on Shapur's part. Captives and plunder were of greater immediate use than a long-term investment in garrisoning and organising territory at a considerable distance from the heartlands of the Sasanian Empire. Trajan and his successors had learned the difficulties of such an undertaking in reverse. Similarly, the establishment of Mesopotamia as a province by Septimius Severus was a source of criticism from the likes of Dio because it was too costly to maintain and embroiled the Romans in ongoing wars with the Parthians. 2 If Shapur I's invasions demonstrated anything with clarity in the longer term, it was that the limits of Sasanian imperialism in the west essentially lay at the upper Tigris, roughly the same practical limit of Roman imperial attempts to expand east. While the Roman Empire was suffering its greatest

Persian triumph, Roman defeat

87

existential challenge since the days of its emergence from the Italian peninsula, it was not as simple as kicking in the door and watching the whole rotten edifice crumble. We saw in the previous chapter that the so-called "third century crisis" of the Roman Empire is conventionally held to have begun circa 235 with the end of the Severan dynasty and the elevation of Maximinus Thrax, the first of the "soldier emperors." The problems confronting Roman rulers and their administrations became more pronounced and complex as the following half-century unfolded, especially in the third quarter of the third century. 3 The rate of decline in silver content in the antoninianus was precipitous during the 250s and 260s and by 270 its silver content was confined to the thin wash that everyone knew was the barest disguise for its lack of precious metal content. We have already seen the potential ramifications of a commodity currency rapidly declining in precious metal content coupled with the increased minting of the same coins. The fact that the soldiers were the primary recipients of these coins saw attempts to improve the standing of the silver currency via reverse images depicting military themes and appealing to political stability at a time when many were witnessing disintegration. The contrast with the Sasanian silver coinage of the same period could not have been starker. We also saw in the previous chapter that Ardashir was renowned in Sasanian history as a founder of cities and that increased agricultural output was possible via various irrigation improvements, especially in the western portion of the empire. His establishment of direct control over the Persian Gulf and construction of ports connected the gulf and eastern Arabia more effectively with Iran and southern Mesopotamia. Under Shapur I these activities continued, the texts and archaeological sources suggesting an expanded programme of city foundation. During the 250s, this was augmented by the deportation of captives from the cities of Rome's eastern provinces and its defeated armies; some cities, such as Gundeshapur, apparently constructed anew for the purpose. This would essentially strike a double blow against the Romans. Captives enhanced the already increased economic capacity of the Sasanians while depleting Roman efforts to rebuild in the wake of the invasions of the 250s. The archaeological evidence from Zeugma, one of the dozens of cities captured by the Sasanians in the 250s, attests to long-term decline in the wake of the invasions and at Dura Europos we have clear evidence for the total abandonment of the Euphrates settlements below the Khabur confluence after the withdrawal of Persian forces in 260.

Disease as a factor afflicting the Roman Empire at the time of the invasions of Shapur I Climatic challenges and disease were another set of difficulties experienced in the Roman Empire with apparent intensity in the third quarter of the third century. Kyle Harper's recent work emphasises what he sees as the

88

Persian triumph, Roman defeat

most neglected disease epidemic in the study of Roman history as a fundamentally important consideration among the stressors afflicting Rome in the mid-third century, the so-called plague of Cyprian. 4 Harper points to a considerable body of textual evidence for the plague as the same disease that spread across most of the empire with the earliest datable references to it in Alexandria in 249. In the following year, there are attestations of a serious disease outbreak in Carthage and in Rome by 251 leading to the suggestion of an emerging pandemic disease that spread west and north from Egypt. Other textual references suggest that disease outbreaks were also attested at Athens and Antioch in the 250s, and there were some ancient claims of an empire-wide disease outbreak. 5 There was a belief in antiquity that the disease had its origins outside the empire, specifically in Ethiopia. 6 The disease is claimed not only to have been severe in its initial impact, but to have then ravaged many parts of the empire over the next 15-20 years.7 Harper suggests further that the origins of the disease outside the empire's "endemic pool of native diseases" combined with global climatic upheavals in the 240s brought about by dramatic changes to monsoon and other ecological systems to play an important part in the disease's seemingly sudden onset and pandemic qualities. 8 He emphasises the disease's potential demographic impact in the medium to long term and how this formed an important factor in administering shocks to the Roman system at a time when it was already struggling with a number of other difficulties. 9 The disease was thus an "exogenous shock" that came on top of other shocks that combined to have a transformational effect on the political, military and economic organisation of the Roman Empire through the third quarter of the third century. 10 If the disease played its part in a devastating and transformational impact such as that described by Harper, the plague of Cyprian would have had important consequences for political and military conflict between Rome and Sasanian Persia. The impact of this conflict, especially in the 250s, in turn, placed its own stresses on the Roman system, which Harper emphasises as a geo-political factor. It is obvious but still important to note that such a devastating pandemic disease would not respect imperial borders, just as the Antonine plague of the second century didn't, and that the Persians (and the tribal groups of the Rhine and Danube) would be equally susceptible to it. The cities, towns and provinces of the Roman Middle East had more in common climatically and perhaps even demographically with southern Mesopotamia than they did with their Roman imperial counterparts in western Europe or north Africa. Rome occupied provincial territory as far east as northern Iraq and as far south as northern Arabia and there is ample evidence for population movement in both directions across the imperial frontiers. The connectedness of Syria, northern Mesopotamia and Armenia with southern Mesopotamia via the Euphrates and Tigris Rivers in the form of trade and relatively regular military movement demonstrates the point. Shapur I's soldiers marched in their thousands across Syria, Mesopotamia

Persian triumph, Roman defeat

89

and Cappadocia and as far west as Cilicia in the 250s and thousands of Roman captives were transported from the cities captured by Shapur back to cities in Mesopotamia and other parts of the Iranian empire. A pandemic disease of the type described by Cyprian must surely have been a significant issue across the Roman/Persian frontiers, especially if it had the capacity to reduce the population of a city such as Alexandria as dramatically as suggested in modern scholarship. 11 A specific example from the Roman/Persian Wars of the 250s potentially indicates that not all references to disease at this time were necessarily identical or even similar to the devastating plague described by Cyprian in Carthage. Shapur I's armies ravaged the Roman provinces of Mesopotamia, Syria, parts of Cappadocia and Cilicia throughout the 250s but there is no indication of disease afflicting them or even hindering them during this time. Conversely, some ancient writers attributed the impact of plague on the army of Valerian as a factor in Shapur's decision to invade the Roman eastern provinces in 260 rather than stay away. 12 In this case, it appears that the effects of disease on Valerian's army actually encouraged the Persians to attack and they then undertook prolonged raiding as far west as Selinus in Cilicia. If the plague that had affected Valerian's army was the same as the plague described by Cyprian, it is almost inconceivable that the Persian army did not contract it as well and that it did not, in turn, have a significant effect on invasions and deportations. Was the plague reported in 260 and its impact overstated in the ancient sources? Was it a different disease in the eastern provinces to the plague of Cyprian, or was it a disease that the Persian army was less susceptible to? Disease and its demographic impacts should not be dismissed as serious issues for Rome to deal with in the third quarter of the third century but some caveats need to be observed on how it is analysed as a factor in contributing to "crisis" and especially as a factor in conflict between Rome and its enemies, including Persia.

Religion as a factor in Shapur I's invasions The role of Zoroastrianism in legitimising the Sasanian regime continued under Shapur I, and its role as a "national" element in Sasanian Persian rhetoric and practice was potentially on display during the invasions of the Roman eastern provinces in the 250s. The Zoroastrian priest Kerdir eventually occupied a powerful position at the head of an elaborate Zoroastrian hierarchy with strong influences on the mechanisms of government under Bahram I (271-74) and his successor, Bahram II (276-92). By his own testimony, Kerdir originally found favour under Shapur as a herbad (priest) and while it is possible that he later overstated his importance under Shapur's rule, he was important enough ea. 265 to be mentioned in Shapur's SKZ inscription. 13 In his own inscription put up at Naqsh-i Rustam ea. 280 (known as the KKZ), Kerdir claimed that Shapur had given him "authority and independence of action regarding the affairs of the Divine

90

Persian triumph, Roman defeat

Beings among the priesthood at court, throughout the provinces and in all places." 14 Kerdir was also given responsibility by Shapur for ensuring the security and welfare of Zoroastrian communities in the Roman provinces and other non-Iranian lands conquered by the Sasanians in the 250s. 15 This responsibility took on what, in Kerdir's mind at least, was a religious element in Sasanian imperialist activity: And I caused many fires and priestly colleges to flourish in Iran, and also in non-Iranian lands. There were fires and priests in the non-Iranian lands which were reached by the armies of the King of Kings (Shapur I). The provincial capital of Antioch and the province of Syria, and Cilicia, and the districts dependent on Cilicia; the provincial capital of Caesarea and the province of Cappadocia, and the districts dependent on Cappadocia, up to Pontus, and the province of Armenia, and Georgia and Albania and Balasgan, up to the 'Gate of the Alans' - these were plundered and burnt and laid waste by Shapur, King of Kings, with his armies. There too, at the command of the King of Kings, I reduced to order the priests and fires which were in those lands. And I did not allow harm to be done them, or captives made. 16 This politicising of minority religions in the Iranian world of the third century would later feed further into the notion of an interdependence between Zoroastrianism and the state, and Kerdir was the primary figure in this. Under Ardashir, Shapur I and Hormizd I, it would appear that religions other than Zoroastrianism were at least tolerated. Indeed, Manichaeism was more than simply tolerated by Shapur I, appearing to have received some form of imperial favour if Mani's claim that he was given licence to preach his message across the empire is true. 17 Like Kerdir, Mani also accompanied Shapur on the invasions of the Roman Empire in the 250s, suggesting some form of enduring imperial support for Manichaeism and perhaps indicating a possible role for religion in the prosecution and outcome of the invasions. 18 While potentially attractive as evidence for the development of religion as a factor in political and military conflict between Rome and Persia, it is important to weigh these claims carefully because they rely exclusively on the claims of Kerdir and Mani. We saw in the previous chapter that some legitimate questions have recently been put in relation to the sources for religious persecution in third-century Persia and Christian sources of the fourth century identifying Christianity as a factor in Constantine's Persian War.

The lead-up to Shapur I's "second great contest" with the Romans While Shapur, like his father before him, was dealing with difficulties in the eastern part of the empire in the mid-240s, Philip and his successors,

Persian triumph, Roman defeat

91

Decius (249-51) and Trebonianus Gallus (251-3 ), spent most of their comparatively short reigns dealing with problems in the west on the Rhine and Danubian frontiers. There was little choice for these emperors but to deal with the immediate threats of the Carpi and Quadi in Dacia and Pannonia and increasing problems with the emergence of the Goths. Philip spent much of his reign dealing with these issues and may have beaten a hasty retreat from Persia in 244 due to them. The Carpi, together with their Gothic allies, had commenced fresh raids directed at Dacia and the Balkans late in 243 and within only a few months of returning from Persia and securing rule of the empire for himself and his son, Philip departed for Thrace where he established headquarters at Philippopolis. Most of the next two years would be spent dealing with the Carpi before the necessities of celebrating Rome's 1,000th anniversary saw his return to Rome in the latter half of 247. In 248, however, the revolt of Pacatian in upper Moesia appears to have encouraged further tribal attacks on Pannonia, this time by the Quadi and Iazyges. With the whole region of the middle and lower Danube in a state of sustained instability, the Goths began more serious and sustained attacks, which saw them and their allies surge into Moesia Inferior early in 248. 19 Sending Decius, likely equipped with a special command over the Pannonian and Moesian provinces aimed at dealing with the situation, ultimately led to Philip's demise because Decius was, in fact, highly successful in doing so. Decius' whole reign of approximately 20 months was spent dealing with the same problems and culminated in his death in battle against the Goths at Abritus in Moesia Inferior in August 251. Decius' successor, Trebonianus Gallus, had little choice but to strike a deal with the Goths, which included an agreement to pay tribute, a humiliating necessity in the wake of Decius' death. The unfulfilled expectations of the Danubian soldiers that Gallus would avenge the death of Decius by punishing the Goths appear to have been one factor that brought about his own demise in the middle of 253. While Shapur was in an undeniable position of strength after the defeat of Gordian III and the negotiations with Philip, events on the Sasanian eastern frontier required his attention and would do for some time. In fact, it is possible that these problems had begun to emerge during the invasion of Gordian III in 243. The Armenian historian, Moses Khorenats'i, referred to a revolt by a military commander named Perozmat Karin in Abarshar in western Chorasan ea. 243. 20 It is possible that the revolt took place in the wake of Shapur's defeat at Rhesaina in the same year. 21 Perozmat was purportedly successful against Shapur initially and the rulers of Gilan on the Caspian Sea and Sakastan also revolted. Perozmat was killed in a conspiracy ea. 24 7 and in the wake of this, Shapur spent some time suppressing the other rebels in the east associated with him. The appointment of his son, Bahram, as the king of Gilan and Hormizd, Shapur's eventual successor in 2 70, as the governor of Abarshar was an important part of dealing with these problems.

92

Persian triumph, Roman defeat

Armenia and Iberia under Persian control When Shapur I's actions against Nisibis and Armenia began ea. 250, followed by invasions of the Roman eastern provinces in subsequent years, the Romans were largely preoccupied and poorly prepared for what would become one of the most serious and sustained military challenges for the Roman Empire in its entire history. Shapur pointed specifically to Roman interference in Armenia some time before the invasions began as justification for them. 22 If we are to believe the Byzantine chronicler, Zonaras, who claimed that there were territorial provisions in a treaty between Shapur and Philip in 244, Philip recovered Armenia (and also Mesopotamia), in contravention of the agreement, soon after his return from Persia. Shapur's pretext may have referred to ongoing Roman involvement and interference in Armenia rather than one specific action. The Armenian king requested assistance from Philip some time before Philip's death in 249 and the Sasanians were responsible for his removal slightly later. 23 Another Armenian historian, Agathangelos, claimed that the Persians killed most of the Arsacid royal family of Armenia and that one of the Armenian princes, Tiridates, fled to the Roman court. 24 Chaumont proposed that the asylum given to Tiridates by the Romans was the source of Shapur's grievance regarding Rome's breaking of the agreement of 244. 25 As a result of this activity, Hormizd, Shapur's son and eventual successor as Shahanshah in 270, was made king of Armenia. 26 These events are difficult to date precisely due to chronological inconsistencies in the Armenian historical accounts but they likely belong to 250/1. 27 Hormizd later led a targeted campaign directed at Roman cities in eastern Cappadocia during his father's Syrian campaign in 252/3 and may have played a part in the Persian attack on Antioch. 28 Following his succession as Shahanshah on his father's death in 270, Hormizd's brother and eventual successor to the Sasanian throne in 293, Narseh, was appointed king of Armenia indicating the importance of imperial control of Armenia to the Sasanians. The kingdom of Iberia, which had enjoyed a close relationship with Rome at times prior to the third century came under direct Persian rule at approximately the same time. Like Armenia, Iberia was included in the SKZ's list of territories directly under Persian control, as was its eastern neighbour, Albania. 29 There is some evidence indicating that the Iberian king accompanied Shapur on his invasions of the Roman eastern provinces in the 250s. The Iberian ruler at the time the SKZ was carved had an Iranian name, Amazasp, and may have been related to the Sasanian royal family. 30 Over the following decades, the Persians would achieve what Braund refers to as a coup in Transcaucasia, especially in Iberia, and this extended to activity of a religious nature at the hands of Kerdir in the 280s. 31 Attempts by Kerdir to enforce the practice of Zoroastrianism in Iberia in the 270s and 280s caused local resentment and were a possible factor in Iberia looking more in the direction of Rome by the 290s.

Persian triumph, Roman defeat

93

Sources and historiography of Shapur I's campaigns against the Romans The most detailed texts for the invasion are the SKZ inscription and the somewhat enigmatic Thirteenth Sibylline Oracle, a roughly contemporary text with events, likely written in Syria. 32 The genre of the oracle makes its use challenging and one of its most illuminating contributions is an indication that at least some elements of the Syrian urban elites placed more faith and trust in local power brokers and usurpers than the often short-lived and somewhat distant emperors. Potter believes that the bulk of the text was compiled by summer 253 before the final lines were added sometime after 261. 33 It is, therefore, most useful for Shapur's "second contest" with Rome and appears to have been written, at least in part, to glorify the usurper Uranius Antoninus who is thought to have been successful in defeating a Persian army outside of the city of Emesa in 253. 34 These texts can be supplemented by later Armenian, Arabic, Latin and Greek texts together with archaeological and numismatic evidence to provide a reasonably detailed account of Shapur's momentous invasions. Using the SKZ, the Sasanian invasions of the 250s can be grouped into two overall campaigns. The inscription refers to the campaign of 259/60 as the "third contest" with the first represented by the war with Gordian III in 243/4. The second contest principally involved the Persian invasion of Syria, Cilicia and Cappadocia and commenced in 252. It would appear, however, that a number of targeted campaigns took place during the 250s and that these were included on the inscription in either the second contest or the third. While the inscription makes no references to setbacks and defeats, there are indications in other sources of victories won over the Sasanians by Roman forces together with Uranius Antoninus of Emesa and Odenathus of Palmyra.

The second contest of Shapur I with the Roman Empire It is reasonably clear that the invasion of the Roman eastern provinces by Shapur I began during the short reign of Trebonianus Gallus (June 251August 253). The most likely year is 252 and it appears that this first main thrust of the campaign continued into 253. 35 Gallus' reign was marked by ongoing difficulties with the Goths in Thrace following the death of Decius in battle against them in mid-251 and an outbreak of plague in Rome, which even claimed the life of Decius' son, Hostilian, whom Gallus had adopted. While it appears that Gallus never left Rome after he arrived there a few months after his elevation following Decius' death, issues of Adventus coins at Antioch for him and Volusian potentially suggest otherwise. 36 The numismatics of the 250s is suggestive of numerous Roman imperial visits to Antioch in contrast with the suggestion that the Thirteenth Sibylline Oracle viewed the emperors as distant figures who essentially left the imperilled eastern provinces with their own devices during this period. 37 In 251, Decius' son, and Caesar at the time, Hostilian, is attested at Antioch via

94

Persian triumph, Roman defeat

Adventus coinage, while his other son, Herennius Etruscus, is similarly attested in 250/1 while still Caesar. 38 Even Decius' wife, Herennia Etruscilla, had Adventus coinage minted at Antioch in her name. That Decius himself is not attested with Adventus coinage is telling and clearly indicative of the emperor's preoccupation with the Gothic invasions that would eventually claim his life. The presence of numerous representatives of the imperial families of Decius and Gallus at Antioch from 250 to 252/3, therefore, signalled the importance of developing problems with the Persians and the Adventus coinage is an indicator of this. The invasion routes and proposed details of the progress of the invasion have been analysed many times using the order in which captured cities are listed on the SKZ. 39 The SKZ also includes the claim that a major battle was fought at Barbalissos on the easterly bend of the Euphrates in which

Antioch. Trebonianus Gallus. AD 251-2. Obv- IMP C C VIB TREB GALLVS P F AVG: Radiate draped, curaissed bust right. Rev- ADVENTVS AVG: Emperor on horseback left, raising right in salute, transverse sceptre in left. RIC IV Trebonianus Gallus 79. Photograph Effy Alexakis.

Figure 4.1 Antoninianus,

Antioch. Herennius Etruscus. AD 250. Obv- HEREN ETRV MES QV DECIVS CAESAR: Radiate bust, right, two dots beneath. Rev- ADVENTVS AVG: Herennius Etruscus (Trajan Decius? ) on horseback left, hand raised, holding spear. RIC 4 Trajan Decius 156. Photograph: Effy Alexakis.

Figure 4.2 Antoninianus,

Persian triumph, Roman defeat

95

the Sasanians were victorious over a Roman army of 60,000 men. Latin and Greek texts record details relevant to some aspects of the invasion but no coherent narrative of these events has survived from the Roman side. The extensive and detailed archaeological evidence discovered at Dura Europos, one of dozens of cities captured as part of the second contest, reveals the grim nature of siege warfare between Rome and Persia and underscores the significant resources both sides contributed to waging war and defence. 40 Archaeological evidence from Zeugma, another of the cities captured in Shapur's invasion of 252/3, shows extensive damage attributed to the Persian sack of the city. Referring to the evidence as a "snapshot of an urban catastrophe," the excavators revealed the extent of the damage still visible at the time of the rescue excavations in 2000 and noted the longer term difficulties the city likely experienced in recovering from the attack due to its severity. 41 The bulk of the cities listed in the second contest were captured in a campaign directed at Syria that began with the capture of Anatha on the lower Euphrates and progressed with reasonable speed to Barbalissos on the bend of the Euphrates where a large battle was fought as noted earlier. 42 Cities were then attacked to the north and south of Barbalissos before a major assault on Antioch. The port city of Seleucia also fell, likely taken before Antioch as part of cutting the city off from its port, while some cities in Cilicia were attacked and captured. An Arabic tradition referred to Shapur beginning his invasion with the capture of Nisibis but Nisibis is not named on the SKZ. 43 It would make strategic sense for Nisibis to have been attacked, given its military and administrative importance in Roman Mesopotamia and it was clearly dangerous to leave the fortification behind while the invasions took place further west. It is possible that the city was captured some time before the main invasion began, perhaps even in 250 and in concert with the overthrow of the Armenian king. Both Tabari and Eutychius claimed that Shapur put Nisibis under siege but was forced to put the siege on hold due to problems in Chorasan. In these accounts, he returned later and successfully captured the city.

The Persian capture of Antioch The capture of Antioch-on-the-Orantes, the metropolis of Syria, dominated in the surviving Latin and Greek texts with some stories associated with the event surviving centuries afterwards. There is a debate as to whether Antioch was captured in both the second and third contests because a city named Antioch was included in both lists of captured cities on the SKZ. There are questions about whether this meant Antioch-on-the-Orantes in both cases or whether an Antioch in Cilicia was the one captured in 260. The SKZ lists Antioch after Seleucia and before Cyrrhus, making it likely that Antioch-on-the-Orantes is meant here. 44 In the later list of cities ascribed to the third contest of 259/60, the SKZ lists Antioch again along with the capture of five other cities that were all located in Cilicia, suggesting that the second Antioch was a different city of the same name. 45

96

Persian triumph, Roman defeat

The most likely candidate in this case would be Antioch ad Cragum, which was located in Cilicia. Barnes notes that two captures of Antioch-onthe-Orontes are generally preferred in scholarship, although Millar only accepted the second capture of the city "on balance." 46 There has also been debate about the dating of the first capture with some preferring to place it in 252, 253 and even 256. 47 In a number of accounts, Antioch was betrayed by a deserter called Mariades who at an earlier time had been a member of the boule of the city and had fled to the court of Shapur after serious factional disputes broke out at the city. 48 Mariades was portrayed as a centrally important figure in Shapur's ability to successfully capture Antioch, although his reward afterwards was gruesome execution on Shapur's orders. One of the better known surviving stories regarding the capture of Antioch was told almost 150 years later by Ammianus Marcellinus who claimed that some of the residents of Antioch attending a performance at the theatre were so unprepared for the attack that the Persians surprised them by raining arrows down on them as they watched a stage performance. 49 While Trebonianus Gallus was in obvious difficulty in mounting a defence against Shapur's invasion of 252, there was clearly some effort to do so despite the problems he faced on the Danubian frontier and in Rome. There is an indication of the preparations of a Roman army to meet the Sasanian invasion in the numismatics of the period. At Antioch, prior to its capture by Shapur in 252/3, a series of antoniniani were minted in the name of Trebonianus Gallus and another in the name of his son and co-Augustus, Volusian, and they have all the hallmarks of coin series produced at Rome and Antioch prior to earlier Roman imperial campaigns against the Parthians and Persians. 50 Reverse types of the Gallus series depict familiar military legends and iconography such as MARTEM PROPVGNATOREM and MARTI PACIFERO, the first illustrating Mars with spear and shield and the latter with Mars holding an olive branch in one hand and a spear in the other. 51 Other reverse types with familiar military links include ROMAE AETERNAE, SAECVLVM NOVVM and a number of VICTORIA types. 52 The Antoniniani of Volusian from Antioch depict similar reverse types to those of Gallus. Among them are MARTEM PROPVGNATOREM and MARTI PACIFERO together with SAECVLVM NOVVM, ROMA AETERNAE and multiple VICTORIA types. 53 As noted earlier, antoniniani were also minted in the names of Gallus and Volusian with the reverse legend ADVENTVS AVG. 54 Potter believes that Gallus was in no position to respond to Shapur's invasions due to an inference that the author of the Thirteenth Sibylline Oracle despised Gallus because he did not offer protection from the enemy. 55 The author was instead praising of the usurper Uranius Antoninus of Emesa who offered resistance to Shapur's invasions. While it could be argued that the ADVENTVS types are more indicative of hopefulness at Antioch than the reality of imperial visits by either Gallus or Volusian, someone senior had to be responsible for marshalling the army that met the Persians at Barbalissos, claimed to have

Persian triumph, Roman defeat

97

Figure 4.3 Antoninianus, Antioch. Trebonianus Gallus. AD 252-3. Obv- IMP C

C VIB TREB GALLVS PF AVG: Radiate, draped, curaissed bust right. Rev- MARTI PACIFERO: Mars, helmeted, in military attire, advancing left, holding olive-branch in right hand and spear in left hand. RIC 4 Trebonianus Gallus 85. Photograph: Effy Alexakis.

numbered 60,000 in the SKZ, and the Adventus coins may be indicative of an imperial visit for this purpose. A possible production break at the Antioch mint may also indicate its capture by the Persians, although this is a fraught topic in scholarship and will likely remain so. Alfoldi proposed in the 1930s that there was an identifiable break in the minting activities of Antioch associated with the capture of the city and that it was observable from the last days of Gallus' reign in 253 through to the spring of 254 by which time Valerian's joint rule with Gallienus had taken effect. 56 This was based on distinctive differences in styles and minting quality in series of antoniniani that suggested the operation of a different mint at the time. Antoniniani, for example, from the short reign of Aemilius Aemilianus (June-September 253) are indicative of this based on poorer quality iconography and unusual spelling variations in their reverse legends. 57 It has been proposed that this mint operated at Samosata or Emesa during the period in which Antioch was occupied by the Persians as neither of these important urban centres were listed on the SKZ in relation to this invasion. 58 It is also possible that the mint at Samosata continued in production after the re-opening of the Antioch mint in 254. Debate will continue because these conclusions are based primarily on observed stylistic differences on the coins. While most of the cities were captured in a traceable order in 252/3, smaller, more specific operations in certain geographical areas appear to have taken place simultaneously or in the years immediately afterwards based on where they were listed on the SKZ. Five cities are listed in Cappadocia and sit at the very end of the list of cities captured in the second contest suggesting a separate and specific campaign. This was likely directed in 252/3 from Armenia as part of Shapur's aim of securing the Persian position in Armenia after the overthrow of the Arsacid king and the installation of his son, Hormizd , as king. 59 Indeed, Hormizd is thought to have led this campaign himself.

98

Persian triumph, Roman defeat

Valerian, Gallienus and the war with Persia With the demise of Trebonianus Gallus in August 253, his successor Aemilius Aemilianus proposed a campaign against the Persians as a priority but his rule of only four months ensured that it was no more than a proposal. 60 The emergence of Valerian and Gallienus as joint rulers after Aemilianus' demise provided the capacity for more effective imperial management of the ongoing problems in the Danubian provinces while also responding to the Sasanian invasions. 61 Early in 254, Gallienus moved to establish himself on the Danube, perhaps at Viminacium, to monitor events in Illyricum and on the lower Danube. 62 With these areas sufficiently calm by 256, he moved further north to the Rhine in 257, leaving his young son Valerian II as Caesar in nominal control on the middle Danube. This was likely due to the removal of troops from this area for his father Valerian's envisaged actions against the Persians and subsequent activity in the area of groups such as the Franks and Alamanni. Problems emerged in 259, however, on the middle Danube in Illyricum with the revolt of Ingenuus, forcing Gallienus back to the region to deal with it. In 260, the Iuthungi invaded Italy, even reaching Rome, while Alamannic raids reached Milan leaving Gallienus with little choice but to return to Italy to deal with them. While Gallienus' commission was a daunting one, Valerian's assignment was perhaps even more challenging. He departed Rome for the east at about the same time his son headed to the Danube and is firmly attested at Antioch in January 255. 63 The circumstances of the Roman recovery of Antioch, likely due to its abandonment by the Persians, are not known. What is often left out of the accounts of Valerian's struggle with Shapur in the mid-250s is the emergence of a serious issue in the Black Sea region, which had implications not only for that area but also for the provinces of Asia. This problem appears to have occupied Valerian for at least part of 254 and continued to be a problem in following years. A group known as the Borani, who were of Scythian origin and hailed from an area north of Crimea, began attacks on the east coast of the Black Sea ea. 253/4. 64 Another raid the following year affected the north coast of Asia Minor and in 256, a neighbouring tribe to the west of the Borani sailed across the Bosphorus and attacked Bithynia. This ongoing threat caused more than just a distraction for Valerian and will likely have exercised Gallienus' mind as well. The myriad problems confronting Valerian and Gallienus on various fronts presented a significant challenge in attempts to deal with the Persian attacks during this period.

The capture of Dura Europos and Circesium Another two cities, which are listed together in this section of the SKZ, are Dura Europos and Circesium on the Middle Euphrates. 65 The two fortified cities appear to have been taken in a specific and targeted campaign aimed

Persian triumph, Roman defeat

99

at the Middle Euphrates below the Khabur confluence. Of exceptional archaeological significance, and likely representative of dozens of similar events at the time, is the evidence for the Sasanian siege of Dura, which dates to a period later than the invasion of 252/3. Among the evidence for the final siege of the city were antoniniani found on the bodies of victims of the siege, which were minted at Antioch in the name of Gallienus between 255 and 257, suggesting that the siege cannot have taken place before these dates. 66 The evidence for the siege has been analysed and interpreted many times ever since its dramatic discovery in the 1930s and it sheds important light on the realities of Shapur's invasions of the 250s, mostly lost in the ancient litera ry texts. There was extensive analysis undertaken at the time of the discoveries in the 1930s and more complete analysis since has fleshed out the detail and corrected some earlier misconceptions. The descriptions of the siege and more recent analysis undertaken by Simon James provide the most comprehensive and up-to-date accounts, including an excellent discussion of earlier scholarly interpretations of the evidence. 67 The techniques employed by the Sasanians in attacking the city together with the attempts of the inhabitants to withstand them demonstrate sophis ticated knowledge of siege warfare on both sides. Jame s compares the archaeological evidence for the siege with advice available in antiquity in military manuals whose origins lay principally in the Greek and Hellenistic periods. James also refers to a late Sasanian military manual, known as the book of Ayin, and its contents suggest that it drew on earlier Sasanian traditions of siege warfare. The inclusion of Dura and Circesium on the SKZ separately and towards the end of the cities captured in the "second contest" combined with the antoniniani of 255-7 gives rise to the somewhat perplexing possibility that the two fortifications were by-passed as Shapur's army initially made its way up the Euphrates in 252/3 from Anatha, onl y to be captured at some

Figure 4.4 The Western Wall of Dura Europos from the desert.

100

Persian triumph, Roman defeat

stage afterwards. The obvious problem with this schema is that Shapur took the risk of leaving the two fortifications behind him as the army made its way further west in the direction of Barbalissos and Antioch. There is some evidence from the site to suggest that Dura was indeed captured and held briefly by the Persians in this first part of the invasion in 252/3. 68 For unknown reasons, Dura and apparently Circesium were perhaps abandoned by the Persians, or even recaptured in a military offensive by the Romans, before a Roman reoccupation, which is supported by the discovery at Dura of a divorce document dated 30 April 254. 69 Dura and likely Circesium were then recaptured by the Sasanians in a later campaign as indicated by the extensive evidence for a siege at Dura, in either 255, 256 or 257. This would explain why the cities were recorded on the SKZ later than the first part of the campaign. Listing the cities twice on the inscription would have given rise to the suggestion that the Sasanians had at some stage suffered a setback during the campaigns. While arcane discussions surrounding the precise dating of the city's fall, and even that there were likely two occupations of the city, are of perhaps great significance to Duraphiles, the proposed scenario provides an insight into the complexities of the invasions and that the textual and archaeological evidence do not always neatly align.

Possible setbacks for the Persians Further indicative of possible Persian setbacks at the hands of Roman forces are coins minted in the 250s of Valerian, Gallienus and Valerian II that include VICT PART or VICTORIA PART in their reverse legends. Analysis of the coinage of the joint reigns of Valerian and Gallienus and the sole reign of Gallienus is complicated by distinctive differences in mint attributions and dating in Roman Imperial Coinage Volume V (RIC V) compared with that of the comprehensive study by Robert Gobl in Moneta Romani Imperii (MIR). Gobl's attributions and dating are preferred in scholarship and are thus referred to here first. Antoniniani of Valerian and his son and Caesar, Valerian II, minted at Antioch and dated 256-8 contain a reverse depiction of Victory presenting a wreath to Valerian II with the accompanying legend VICTORIA PART. 70 Another antoninianus attributed to the Antioch mint in 256/7, this time in the name of Gallienus during the joint reign with Valerian, similarly depicts Gallienus receiving a wreath from Victory with the same accompanying legend, VICTORIA PART. 71 Antoniniani depicting RESTITVT ORIENTIS were minted at Samosata in the names of Valerian and Gallienus from as early as 255 and as late as 260. 72 Antoniniani were minted in the name of Valerian senior at western mints that also advertised a victory over the Parthians. At Viminacium, antoniniani in the name of Valerian were minted in 256 with the reverse legend VICT PART accompanied by Victory holding a wreath and palm branch with a seated captive. 73 At Cologne, an antoninianus of Valerian with the reverse legend VICT PARTICA accompanies an image of Victory

Persian triumph, Roman defeat

101

Figure 4.5 Antoninianus, Antioch. Valerian II. AD 256-8. Obv-P LIC COR VALERI-

ANVS CAES: Bust of Valerian II, radiate, draped, right. Rev- VICTORIA PART: Victory, winged, draped, standing right, presenting wreath to Valerian II and holding palm; Valerian II holding globe and spear. MIR 1604d; RJC V Valerian II 54. Photograph: Effy Alexakis.

holding a wreath/palm branch and is thought to have been minted in 257.74 While it is best to reserve a level of scepticism about the magnitude of such claims, these coins at least demonstrate that imperial claims to victories over the Persians (Parthians) were made during this period and it is possible that there was some, likely modest, substance behind them. There is also some indication of resistance to the Persian invasions in this first campaign in the surviving texts. According to the Thirteenth Sibylline Oracle, Hierapolis and Beroea were triumphant over Shapur's forces, although it should be noted that both cities are listed on the SKZ as being captured immediately after Barbalissos. 75 The Thirteenth Sibylline Oracle and Malalas indicate an Emesene victory over Shapur at this time and there is epigraphic evidence suppor tive of such an event.76 Emesa is not listed on the SKZ and a victory over the Persians outside "the city of the Sun" referred to in the Thirteenth Sibylline Oracle is thought to have been led by Uranius Antoninus. 77

The third contest between Persia and Rome The second campaign of Shapur's into Rome's eastern provinces (or "the third contest" as it was called in the SKZ) was prosecuted over 259 and 260 and the SKZ's account of it focussed on four main points: The size and composition of a Roman army defeated in battle beyond Carrhae and Edessa; the capture of Valerian, army commanders and the Praetorian Prefect; the destruction of territory and capture of dozens of Roman cities; and the taking of thousands of captives. 78 While the second contest had focussed more on Syria and the Euphrates, the third contest was directed more at northern Mesopotamia, Cilicia and Cappadocia, although some territory in Syria also came under attack. This section of the SKZ began by emphasising that both Edessa and Carrhae had been put under siege before

102

Persian triumph, Roman defeat

Valerian arrived with an army numbering 70,000 soldiers. A significant battle between the two forces took place and the Persians were victorious. Reflecting intimate knowledge of the composition of the Roman army, thus demonstrating that Shapur's victory was over an army drawn from all over the empire, the SKZ listed different provincial areas from which the soldiers in the Roman army originated. The Danubian provinces of Raetia, Noricum, Dacia, Pannonia and Moesia were listed first, followed by Spain and the north African province of Mauritania. Provinces located in Asia Minor including Thrace, Bithynia, Pamphylia, Galatia, Phrygia and Lycia then followed. Syria, Judaea, Arabia and Mesopotamia were also included, while the only areas not represented were Egypt, Gaul, Britain and Italy. The primary purpose of listing a diverse range of territories from which Roman soldiers were drawn was to demonstrate the magnitude of the Persian victory.

The capture and fate of Valerian The capture of Valerian was one of the most triumphant moments in Sasanian history and undoubtedly one of the grimmest for the Romans. The SKZ referred to Valerian being taken prisoner "with our own hands" and in a number of rock reliefs from the ancient Iranian world Valerian was depicted in personal subjection to Shapur. This is likely depicted in the famous Paris cameo, which illustrates Shapur I capturing Valerian in battle. 79 In a wellknown and likely spurious account of Valerian's captivity in Persia, Lactantius told the story in the early fourth century of the emperor acting as the personal footstool of Shapur whenever the Sasanian ruler mounted his horse. 80 Valerian's captivity and his fate continued to be of great interest in the Roman world, especially in the fourth century, when it was used by Christian writers as one of a number of historical lessons on how persecutors of the religion had been punished for their wrongs. 81 Zonaras told the story preserved in some accounts still available to him in the twelfth century that Valerian had willingly surrendered to Shapur because his own army was mutinous and about to overthrow him. 82 Ghirshman boldly identified the remains of a building at Shapur I's palace complex at Bishapur that he believed was a smaller second palace designed to house the captured Valerian, thus indicating that the emperor was treated relatively well in captivity. 83 Most of the Roman accounts referred to Valerian's death in captivity at an advanced age and reserved little sympathy for him, although some accused Shapur of treachery by tricking the emperor at the time of his capture. 84 Only the Historia Augusta, universally derided but often referred to when other sources prove inadequate, gave an indication that the Roman leadership under Gallienus had any interest in retrieving Valerian from captivity. 85 The Historia Augusta also claimed that the Iberian king and rulers of principalities in Persia promised to do what they could to free Valerian. Shapur, of course, was given little credit by the Roman sources for the enormous victory he scored over the Roman army at this time; the focus instead was on Roman incompetence.

Persian triumph, Roman defeat

103

The SKZ also made the claim that Roman military commanders, the Praetorian Prefect and senators were taken into captivity. Mennen proposes that the Praetorian Prefect in this case was Sucessianus who had led reconstruction efforts at Antioch on behalf of Valerian in 256/7. 86 In a later Arabic tradition, Tabari referred to the transportation of Valerian to the newly built city of Gundeshapur along with a large number of Roman soldiers. 87 Tabari claimed that Valerian oversaw the construction of a dam at Shushtar in Khuzistan with the help of Roman soldiers from Anatolia. Tabari was aware of different Persian traditions regarding Valerian's fate, one in which Valerian was freed after paying a large financial indemnity but suffered facial mutilation and another in which Shapur ordered his execution. This may be reflected in the sixth-century AD account of Agathias who claimed that Shapur ordered Valerian flayed alive. 88 This, in turn, may have drawn on a tradition first recorded by Lactantius in which Valerian was flayed but only after his death. 89 According to Lactantius, the body was dyed with vermilion and kept as a grim type of trophy in a temple that was frequented by Roman ambassadors to the Sasanian court. The story is in part recounted by Peter the Patrician who referred to Galerius scolding the then Persian king, Narseh, after defeating him in 298 with the words "Even after death, with loathsome art you kept his (Valerian's) skin and inflicted undying insult on his dead body." 90 In the Oration to the Saints of ea. 323, the emperor Constantine addressed Valerian as follows: But you Valerian, who showed the same murder-lust toward those who heeded God, you made the holy judgment manifest when you were caught and led as a prisoner in bonds with your very purple and all your royal pomp, and finally, flayed and pickled at the behest of Sapor the King of the Persians, you were set up as an eternal trophy of your own misfortune. 91 Reiner supported the account of Lactantius, from which the Oration to the Saints reference was likely drawn, by pointing out precedent for such treatment identified in a cylinder seal reference to the Assyrian king Sargon II (721-705 BC) treating a defeated king in a similar way when he "dyed the skin of the rebel Illubi'di as red as wool." 92

Captives The significance of a key objective of Shapur's campaigns, the taking of captives, has been under-appreciated in scholarship. 93 Following a list of dozens of captured cities beginning with Samosata in Syria, followed by cities mostly located in Cilicia and Cappadocia, the SKZ adds: We led away into captivity men from the Empire of the Romans, non-Iranians, and settled them into our Empire of Iranians, in Persia, in Parthia, in Susiana and in Asorestan and in every other nation where our own and fathers' and our forefathers' foundations were. 94

104

Persian triumph, Roman defeat

Figure 4.6 Remains of a bridge (Band-e Kaisar) at Shushtar, Iran on which Roman captives and the emperor Valerian were claimed to have worked by

Tabari and other later Arabic writers. Photograph: Ross Burns.

While this claim is made after listing the details of the third contest, it implies that this was a feature of all three of Shapur's contests against Rome, especially the second and third. We have already seen that a story emanating from the Sasanian tradition of Valerian's captivity told of the emperor and captive Roman soldiers working to build a dam at Shushtar, while the SKZ referred to military leaders and other officials being taken captive along with Valerian and his Praetorian Prefect. Captive Roman soldiers potentially served many purposes and so did non-military captives from the dozens of cities captured in the 250s. The early Sasanian rulers, especially Ardashir and Shapur, were renowned for their establishment and enlargement of cities, palaces and irrigation works across the empire, and the need for skilled workers in them was likely considerable. Ghirshman believed that Captive Roman technicians were set to developing the fertile lands of Susiana (the modern province of Khuzistan), bordering on Babylonia ... On the three great rivers that water the plain Shapur I had large dams built, which also did duty as bridges; vestiges of them still testify to the size and scope of the public works he carried out. The progress of urban development under the same king also owed much to the manpower furnished by prisoners of war. 95 Of Shapur's urban foundations, Bishapur, Shushtar and Gundeshapur were the main centres in which Roman captives were settled. 96 Ghirshman noted what he believed to be the Roman "military camp" style of the

Persian triumph, Roman defeat

105

layout of Gundeshapur and at Bishapur highlighted the Hippodamian grid plan of the city, which stood in stark contrast to earlier Iranian city foundations.97 Further to this probable Roman influence on urban planning due to the participation of skilled Roman captives, Roman influence on the styles and construction of mosaics in the Sasanian palace at Bishapur likely came from the work of resettled Roman captives. 98 The appearance of the mosaics from Bishapur bears a strong resemblance to the styles of mosaics from Syria and northern Mesopotamia of the second and third centuries AD but Ghirshman also noted what he believed were elements suited more to local Iranian tastes. There were other Roman influences at the Bishapur palace complex, including apparent identical masonry techniques in Philip I's Philippopolis imperial complex in his hometown of Shahba in Arabia to an enormous temple to the goddess Anahid built near the palaces at Bishapur. Even a commemorative monument set up to honour Shapur contained Roman-style Corinthian capitals and masons' marks in Greek letters. Further to the textual and archaeological evidence for Roman captives in Sasanian Persia, evidence from Dura Europos may provide an indication of deportations of both soldiers and civilians. With the generally agreed proposition that Dura was captured the first time by the Persians some time in 252/3 before coming back under Roman control sometime in 254 before a second Persian capture of the city as early as 255 or as late as 257, there are possible hints at the fate of the garrison and civilian population of the city during this period. Papyri, namely P. Dura 153 and 154, together with dipinti from the Synagogue indicate some degree of Sasanian organisation at the city in the wake of its first capture. The fate of the garrison and civilian population at this time is not specifically indicated; however, the taking of captives would be a likely scenario. Following the second capture of the city, the only later evidence of human occupation at the site is transient and opportunistic suggesting that it was abandoned by the Persians and not reoccupied by the Romans. The fate of its garrison and whatever civilian population the city had following its reoccupation in 254 was likely deportation to Persia. At Zeugma, which was attacked and captured in the wake of the battle of Barbalissos in 252/3 but before the attack on Antioch, archaeological evidence points not only to a serious level of destruction, but also to preparations by the population beforehand: Indeed, the tell tale signs of a surprise attack are absent from Zeugma ... But there is abundant evidence for a protracted interruption of domestic life concluded by a hurried escape. Several hoards of coins found in the houses are among the many indications that the refugees hoped to return. To these one can add rooms with abandoned furnishings and household accoutrements. Two categories of objects are particularly forthcoming about life in the city on the eve of the attack: vessel

106

Persian triumph, Roman defeat glass, if only for the sheer volume in which it was found, speaks to the amount of tableware left behind ... ; locks and padlocks for doors and cabinets, also found in abundance, suggest a preoccupation with security ... Perhaps most revealing is the bronze statue of a nude and helmeted Mars, laid away in a storeroom along with a large bronze candelabrum in the so-called Villa of Poseidon ... The deposit suggests that the inhabitants were preparing to endure a threat from which they expected to recover, as if sandbagging before a flood, or boarding up windows before a hurricane. The behaviour of the residents might also suggest a rather acute understanding of the enemy's disinterest in occupation, thereby inspiring hope that they might return and rebuild after the attack ... The pattern of abandonment witnessed by the rescue excavations of 2000 suggests that the residents of Zeugma were familiar with the Sasanian penchant for enslaving prisoners, and that they may have feared capture in equal measure to death. 99

Later Arabic traditions also made reference to Shapur's deportations and emphasised the extent of them. 100 According to Eutychius, Shapur and the Persians captured Nisibis, curiously not referred to on the SKZ, as part of the 252/3 campaign, slew the garrison and took away prisoners and a great deal of wealth. 101 Eutychius continued on to discuss Shapur's devastation of Syria and that he took many captives. The tenth-century Chronicle of Se-ert referred to Shapur taking Roman captives back to Iraq (Mesopotamia), Susiane and Persis, while building new cities such as Gundeshapur where many Christian deportees were settled. 102 Among these deportees, according to the Chronicle, was Bishop Demetri(an)us of Antioch who had been taken captive in 252/3 and was deported to Gundeshapur where he served as the bishop of the captive Christians before dying of grief. 103 In the city of Rev-Ardashir, according to the Chronicle, two separate Christian churches became identifiable as a result, one a Greek-speaking church stemming from the captives and the other a Syriac-speaking church tracing its origins to the missionaries of the second century AD. Candida, a daughter of captives taken during Shapur's wars, purportedly became the favourite wife of one of Shapur's successors, Bahram II, but was martyred due to her refusal to renounce Christianity. 104 There is some indication in the texts, also, of the immense organisational efforts required to transport prisoners back to Persia. Zonaras tells the sobering story of the poor treatment of captives on their way back to Persia, at least in part due to their overwhelming numbers. 105 In overall terms, there is plentiful evidence for Shapur's captives and they should be seen as an important purpose of the invasions. The extensive city building programmes of Ardashir and Shapur, together with irrigation and other agricultural projects, required skilled labour and this would have been in abundance in the cities of the Roman East in the mid-third century.

Persian triumph, Roman defeat

107

Conclusion While Shapur's successes over the Romans in the 250s provided ample opportunity to boast of unequalled victories over the Romans at the time, territorial gains were not one of the outcomes. The taking of captives, both military and civilian in origin, was of clear benefit to the Sasanians in the enlarged urban centres of both Ardashir and Shapur. This contrasted with devastation across the Roman eastern provinces, the depletion of human resources and other serious challenges faced by the Romans indicated in a lack of recovery in some cities and complete abandonment in others. Losses suffered in two significant battles together with the capture of Antioch, military and administrative leaders and the emperor himself were major blows at a time when there were many other challenges. The Gothic and other tribal threats on the Danube together with direct threats to Italy and the attacks of the Borani in Asia Minor combined with the setbacks against Shapur to produce a particularly harrowing period for Roman imperial rulers. Apparent ongoing problems with disease outbreaks across the empire added to the difficulties. A precipitous decline in silver content in the main denomination in which the soldiers were remunerated was both symptomatic and causative of ongoing problems. Some profound changes would come as a result. The Sasanian dynasty of Persia came of age against a western enemy that had many times worsted its Parthian predecessors. Shapur had set an example for those who would come afterwards, similar to the examples set by Roman invaders of Parthia in the second century, especially Trajan. In the decades that followed, many challenges lay ahead for the Romans, while Shapur and his successors would continue to toast the victories of the 250s. While the Sasanians under the leadership of Ardashir and Shapur had achieved so much, the Roman system would find a way to survive, adapt and in time recover. An equilibrium would eventually be reached between the two powers and ensure a rivalry that would last for centuries.

Notes 1 There is some evidence to indicate that elements of the Sasanian invading force were planning to stay longer. Grenet 1988 believes, for example, that the Sasanians planned to retain Dura Europos after it was first captured in 253, even installing a satrap there. Hoards of uncirculated Sasanian silver drachms found at Gaziantep and in Cilicia may indicate that some believed they were, indeed, staying for longer. Pfeiler 1973 discussed two hoards of Sasanian coins found in what appear to be contexts associated with the invasions of the 250s. The first, found at Gaziantep, contained 29 coins in total; three from the reign of Ardashir and 26 from the reign of Shapur. The second had a provenance described only as Cilicia and contained 23 drachms of Shapur I. Pfeiler suggested a date of 253 for the deposition of the first hoard and 260 for the second. Another hoard likely associated with Shapur I's invasions is noted in the collection of the British Museum with an origin simply stated as "Middle

108 Persian triumph, Roman defeat

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

East." This hoard contained 48 drachms, all of Shapur I. See Coin Hoards 1975, 70-1. In pragmatic terms, however, any long-term effort to retain the conquered cities and their surrounding territories would have stretched available resources and become too costly - exactly the same problem the Romans had faced under Trajan during a short-lived and futile attempt to organise conquered territory in central and southern Mesopotamia. Cassius Dio 75.3.2-3. See De Blois 2019, 65-86 for a detailed narrative account of the years 249-68 as the years of Rome's greatest "crisis." Harper 2015; Harper 2017, 136-49. Harper 2015, 231. Harper 2017, 137. Harper 2017, 138 suggests 249-62 as the initial timeframe with the possibility of waves of the disease then occurring up to 270 when the emperor Claudius Gothicus himself is claimed to have succumbed to it. Harper 2017, 144. Harper 2017, 140-1. Harper 2015, 250. Parkin 1992. Peter the Patrician, Banchich F 173; Zosimus 1.36.1-2, likely relying on Dexippus. SKZ, line 34 (Persian), line 28 (Parthian). KKZ, lines 1-2. KKZ, line 11. KKZ, lines 11-13. See Widengren 1965, 30-5. Alexander Lycopolitanus, Contra Manichaei Opinions Disputatio, 2 = Dodgeon & Lieu 1994, 3.3.2 (p. 65) refers to Mani accompanying Shapur on his military campaigns during the reign of Valerian. See also Lieu 1985, 58-9 for a discussion of the claim that Manichaeism developed such a profile during Shapur's reign that it potentially challenged Zoroastrianism as the statesanctioned religion. Drinkwater 2005, 37. Moses Khorenats'i 11.87. Syvanne & Maksymiuk 2018, 83. SKZ, line 9. See Mosig-Walburg 2009, 75-84 on Armenia during the reign of Shapur I. Moses Khorenats'i 11.71; 78. Agathangelos 1.36. Chaumont 1969, 83-6. Frye 1983, 126; SKZ, lines 41 & 48; Chaumont 1986/2011; Shayegan 2012; Syvanne & Maksymiuk 2018, 85. Syvanne & Maksymiuk 2018, 85 argue for a date of 252. SKZ, lines 18-19; Shayegan 2012. SKZ, line 3; see Braund 1994, 240ff. SKZ, line 25; Hitchins 2001/2012 refers to him as a "high dignitary of the Sasanian realm." Braund 1994, 244; see also Preud'homme 2018 and further discussion m Chapter 5. See Edwell 2010 on the sources for Shapur's invasions of the 250s. Potter 1990, 142. Edwell 2010, 164. Archaeological evidence from rescue excavations at Zeugma indicates a date range of early 252 to mid-253 for the extensive destruction layer attributed to the Sasanian sack of the city, which is evident at the site. See Aylward 2013.

Persian triumph, Roman defeat

36 37 38 39 40 41 42 43 44 45 46 47

48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64

109

See also Balty & van Rengen 1993, 14-5 who refer to numerous gravestones of cavalrymen at Apamea in Syria dated to 252. Their deaths are believed to have been associated with Shapur's invasion. See Edwell 2008, 184-6 for further discussion of the dating. RIC IV.3 Gallus 79; 214; 224. Potter 1990, 48; Potter 2004, 250-2. RIC IV.3 Decius 193; RIC IV.3 Decius 156 - it is possible that they were in Antioch at the same time. Sprengling 1940; 1953; especially Kettenhofen 1983; Huyse 1999; Dodgeon & Lieu 1994, 360-3, notes 5-19; Edwell 2008, 186-98; Mosig-Walburg 2009, 44-52; Palermo 2019, 45-7. See especially James 2011 for the archaeological evidence for the siege. Aylward 2013, 29-31. SKZ, line 12. Tabari I, 826; Eutychius, Annals, 5 = Dodgeon & Lieu 1994 (pp. 295-6). SKZ, line 15. SKZ, line 31. Barnes 2009; Millar 1993, 159-63. Potter 1990, 292 prefers 252. The suggestion of 253 owes much to Alfoldi's dating using evidence from the coins of the Antioch mint while 256 was proposed by Ensslin 1949, 33 and Honigmann & Maricq 1953, 132-42 based primarily on linking the fall of Antioch to the siege evidence from Dura Europos, which is typically dated 256/7 also on numismatic evidence. Potter is correct to question the date of 256, partly because the numismatic evidence used to date the fall of Dura now falls into a broader timeframe of 255-7 than the original analysis of the coins suggested in the 1930s and also because the listing of Antioch the first time in the SKZ can be placed in the context of the movements of the Sasanian army in the vicinity of northern Syria more than it can be linked to the capture of Dura Europos. Malalas 12.26; Historia Augusta, Triginta Tyranni 2; Anonymous Continuator of Dio, Frag. 1 = Dodgeon & Lieu 1994, 3.1.5 (p. 53). See also Oracula Sibyllina 13, lines 89-90; Edwell 2008, 182-4. Ammianus Marcellinus 23.5.3. RIC IV.3 Gallus 79-96; RIC IV.3 Volusian 214-2396. RIC IV.3 Gallus 84 & 85. RIC IV.3 Gallus 89, 90, 91, 93-6. RIC IV.3 Volusian 219, 220, 222, 235, 236; 221; 238, 239a & b. RIC IV.3 Gallus 79; RIC IV.3 Volusian 214, 224. Potter 2004, 252. Alfoldi 1937. See Dodgeon & Lieu 1994, 364, note 23 regarding the minting break at Antioch. RIC IV.3 Aemilianus 23-9. Samosata is not listed on the SKZ until the third contest in 259/60 and Emesa is not listed at all. SKZ, lines 18-9; see Dodgeon & Lieu 1994, note 19, Chapter 3 which suggests that Shapur's son Hormizd is thought to have led this campaign. Zonaras 12.22 [591]. See especially Goltz & Hartmann 2008, 223-95 on the reigns of Valerian and Gallienus with special attention to the chronology of military and other key events in their reigns. Drinkwater 2005, 43ff. Kienast 2017, 203. Drinkwater 2005, 42; Oracula Sibyllina 13, line 36: "When the Roman Ares destroys the German, defeating the life-destroying Ares of the Ocean, then there will be a long war for the Persians .... "

110 Persian triumph, Roman defeat 65 SKZ, line 17. 66 James 2004, 22-3; RIC V.1 Gallienus 436-60. 67 James 2004, 30-9; James 2011, 69-101. See also James 2019 for a comprehensive account of the Roman military base at Dura and how it functioned in the third century AD. 68 James 2004, 23-4. 69 Welles et al., 1959, 166-71; P. Dura 32. 70 MIR 1604d (RIC V.1 Valerian 54 AD 255); MIR 1604a (RIC V.1 Valerian 291 AD 259). 71 MIR 16046 (RIC V.1 Gallienus (joint reign)) 453. 72 MIR 1677m (January 255-mid-256); MIR 1685m(l) (4) (mid-256 to end 256); MIR 17006(2) 257-260. 73 MIR Viminacium, 847a, b, c, d, (VICT PART). See MIR 846a (VICT PARTICA) 256. (RIC V.1 Valerian 262). 74 MIR 881h Cologne 257 (RIC V.1 Valerian, joint reign, 22 258/9 Lugdunum). 75 Oracula Sibyllina 13, line 129; SKZ, line 13. 76 Oracula Sibyllina 13, lines 150-4; Malalas 12.26; Dodgeon & Lieu 1994, 56. 77 Potter 1990, 323-28; Zosimus 1.38.1. 78 SKZ, lines 19-37. 79 Some scholars believe that the cameo represents Shapur II and Jovian; see, for example, Gobl 1974, 33-8. Gariboldi 2010, 21-2 appears to believe that it does represent Shapur I and Valerian, although does not say so explicitly. There is a possibility that it was carved by a Roman artist; see Wiesehofer 2001, Plate XXVIIb who claims that the cameo is "perhaps" of Valerian and Shapur I. 80 Lactantius, De Mortibus Persecutorum, 5. 81 For example, Lactantius, De Mortibus Persecutorum 5; Jerome, Chronicon, s. aa. 258-60, p. 220, 12-9 = Dodgeon & Lieu 1994, 3.3.1 (p. 59); Orosius, Adversos Paganos, 7.22.3-4. 82 Zonaras 12.23 [594]. 83 Ghirshman 1962, 149. 84 Aurelius Victor 32.5; Zosimus 1.36.2. 85 Historia Augusta, Valerian 1.1-4.1. 86 Mennen 2011, 161. See also Dodgeon & Lieu 1994, note 37, 365. 87 Tabari I, 826-7. 88 Agathias 4.23.7. See Cameron 1969/70, 120-1, 138. 89 Lactantius, De Mortibus Persecutorum, 5.6. 90 Peter the Patrician, Banchich F.201, 133-5. 91 Oration to the Saints, 24. 92 Reiner 2006, 327. 93 Lieu 1986, 476. 94 SKZ, lines 34-6. 95 Ghirshman 1962, 137. 96 See Ghirshman 1956, 25ff. 97 Ghirshman 1962, 135, 139; Lieu 1986, 478. 98 Ghirshman 1956, 147-8. 99 Aylward 2013, 30-1. 100 Asmussen 1983, 929; Lieu 1986, 477-8. 101 Eutychius, Annals, 5 = Dodgeon & Lieu 1994, 295-6. 102 Wood 2013, 35; Lieu 1986, 478. 103 See Peeters 1924. 104 See Brock 1978. 105 Zonaras 12.23 [594].

5

The last years of the reign of Shapur I to the Persian invasion of Carus

Introduction With the senior Augustus in captivity, the eastern Roman army surviving in patchy remnants and a section of the Persian army ranging across the Cilician coastline seemingly at will, Rome's problems in the eastern provinces may well have seemed insurmountable in 260. The surviving and now sole emperor, Gallienus, was fully occupied with difficulties in Raetia, which threatened to spill over into Italy, and some of Valerian's senior commanders and administrators were beginning to make moves that spelled likely demise for the son of the first Roman emperor to be taken captive. At about the same time, a revolt on the Rhine under the leadership of the general Postumus took place, marking the beginning of more than a decade of the so-called Gallic Empire in Gaul, Britain and Spain. In the east, however, events unfolded in the years immediately following the demise of Valerian that contributed to a stabilisation of Gallienus' imperial fortunes. Shapur's forces became vulnerable as they splintered into smaller raiding parties as far west as Selinus on the southern coast of Asia Minor. The difficulties of operating this far west, combined with the actions of remnant Roman forces under various would-be usurpers and the Palmyrene leader, Odenathus, contributed to a Persian withdrawal. Odenathus' loyalty to Gallienus and competent command of the remnants of the eastern Roman army in the proceeding years largely kept the Persians at bay. In the west, Postumus would prove to be content with controlling the territories he initially gained, preferring to pass himself off as a legitimately recognised partner in imperial rule than attempting to overthrow Gallienus. Difficulties of an economic and demographic nature, however, hampered a sustainable recovery and the short-term benefit of Odenathus' command in the east sowed the seeds of an eventual Palmyrene rebellion taking in almost the entirety of the eastern provinces, including Egypt. During the 260s, the silver content of the antoninianus plummeted for all intents and purposes to zero, marking the beginning of a permanent loss of faith in silver as a mechanism of remunerating the army. Difficulties associated with disease outbreaks appear to have continued, demonstrated poignantly by the story told in one tradition of the death of Claudius Gothicus of plague in 270. The Sasanian deportations of the 250s and 260s hampered recovery

112

Shapur I to Carus

among the devastated cities of the eastern Roman provinces. The death of Shapur I in 270 and a subsequent period of introspection in Sasanian leadership, however, appear to have sparked a period of intra-dynastic rivalry in Persia. By comparison, the reign of Aurelian (270-5) saw a return of stability in some form, especially in relation to the Palmyrene and Gallic usurpations. All but one emperor from Aurelian to Carus would plan, and in one case actually carry out, a Persian invasion. Religious developments under Bahram I and Bahram II in Persia have traditionally been seen as significant indicators of the emergence of religion as a factor in imperial rivalry and conflict in the fourth century. The Zoroastrian herbad and later mowbed, Kerdir, who rose by his own account to a place of considerable influence under Shapur became instrumental, it would seem, in the rise of Bahram I and Bahram II to their positions of Shahanshah. If there was any doubt as to the primacy of Zoroastrianism in relation to the state before Shapur's death, there would not be under the Bahrams. Kerdir would reach the highest ranks of the nobility under Bahram II and become the most powerful judge in the land. He also attacked religious rivals, especially the Manichaeans, and promoted the idea of an orthodox approach to Zoroastrianism. The extent to which Kerdir and the role he attained was exceptional is an important subject of debate. Aurelian's strong emphasis on the cult of Sol Invictus might also be seen as presaging developments in Roman imperial approaches to religion, perhaps even contributing in some way to an eventual adoption of a monotheistic cult by Constantine and his successors in the fourth century.

Shapur's withdrawal For Shapur and the Persian forces accompanying him, the defeat of Valerian and the Roman army near Edessa in 260 lay open the rich Roman provinces and cities to the west. With the cities of Syria captured in the first campaign but only a handful of Cilician and Cappadocian cities attacked at that time, Shapur perhaps believed that he had unfinished business in those provinces. 1 The SKZ lists the capture of a number of cities in Cilicia suggesting a coordinated attack on that province aimed at capturing as many resources as possible. 2 The first signs of potentially serious difficulties for Shapur, however, were also encountered during this period. Some later Byzantine writers claimed that the Persian forces became ill-disciplined and that this marked the beginning of setbacks for them. 3 While besieging Pompeiopolis near Zephyrion, for example, they were caught unawares by a Roman force that attacked them from the sea under the command of a general named Ballista (Callistus). With Gallienus preoccupied at the time of his father's capture by a serious attack of the Alammani in the Italian Alps, Ballista and the comes sacrarum largitionum, Macrianus, emerged in control of some of the surviving troops after Valerian's defeat. 4 Ballista not only inflicted a defeat on the Persians at Pompeiopolis but was also credited with capturing Shapur's

Shapur I to Carus

113

harem. Dodgeon and Lieu believe that Shapur now withdrew east, pillaging other Cilician cities as he did so. 5 Another section of the Persian army, which had likely already been detailed by Shapur to operate further west, continued to attack cities further along the Cilician coast. The furthest west this section of the army would reach was Selinus, the port where the emperor Trajan died while returning from his Parthian campaign in 117.6 It is possible that the Persian capture of Selinus, where a temple and cult of the Divine Trajan was still active in the mid-third century, was a deliberate aim of Shapur's. The symbolism of capturing this location given Trajan's extensive invasions of Parthia almost 150 years earlier was likely not lost on him. Shapur's withdrawal appears to have become more urgent with the emergence of Odenathus of Palmyra in command of another surviving remnant of Valerian's troops in 261. Zonaras claimed that an alliance was struck between Gallienus and Odenathus who pursued the retreating Persian troops along the Euphrates and inflicted losses on them.7 A sixth-century Byzantine text tells an account of a somewhat harried Shapur paying off the remaining Roman garrison at Edessa in order to let his troops pass unmolested on their way back to Persian territory. 8 The lack of retention of territory and fortifications, even in areas such as northern Mesopotamia where Roman provincial and military organisation since the second century had been a bugbear for Parthian and Sasanian administrations alike, has led to subtle or even overt criticisms of Shapur's campaigns as little more than "razzias" and raids. 9 This position is clearly influenced by the Roman source tradition that sought to cast the momentous losses suffered by the Romans as driven chiefly by internal problems. It is important to emphasise, however, that the Sasanians also had ongoing issues to deal with on their own eastern frontier, just as Rome was dealing with ongoing problems in the west, especially on the Danube. As noted in

Herennia Etruscilla. AD 249-251. Obv€P€NNIAN AIC0POYCKIAAAN CE:'.B:Draped bust of Herennia Etruscilla to right. Rev - (TPAIAN CE:'.AINO)/ THC IE:'.PA:Tetrastyle temple enclosing statue of Trajan, shown as Zeus , seated left, holding thunderbolt in his right hand and sceptre in his left; on architrave, 0€/0Y TPAI. RPC IX 1308.1. Image: Peter Edwell.

Figure 5.1 Follis, Selinus-Traianopolis.

114

Shapur I to Carus

earlier chapters, Shapur laid claim in the SKZ to the realm of the Kushans up to Purushapura (modern Peshawar) on the south-eastern side of the Hindu Kush. The Sasanians had done this by supporting the Kushano-Sasanian dynasty and expansion of their control into northern India would continue after Shapur's death ea. 270. 10 The first two Kushano-Sasanian rulers, Ardashir I (ea. 230-45) and Peroz I (ea. 245-75), required ongoing support in maintaining their power and dealing with setbacks against the Kushans. 11 Some Arabic writers claimed that Shapur put a siege of Nisibis on hold ea. 250 due to the need to deal with problems in Chorasan. 12 There had been problems in the preceding years with a revolt in Abarshar in western Chorasan, which spread to other areas and Shapur had spent a number of years dealing with it early in his sole reign (see Chapter 4). While little detail is known of developments in this respect through the 250s, Shapur was likely required at times to deal with difficulties on the Central Asian and Indian frontiers and this may be reflected in his relatively quick withdrawal from the Roman eastern provinces in the early 260s. 13

Shapur's commemoration of the victories over the Romans Between his return to Persia in 260 and his death in 270, Shapur commissioned a number of monuments celebrating what he believed were the great achievements of his reign, including the momentous victories over the Romans. Carved rock reliefs at Naqsh-i Rustam and Bishapur are the most informative visual prompters of Shapur's victories over the Romans, while, as we have already seen, the trilingual inscription carved on the Ka'aber of Zoroaster at Naqsh-i Rustam, known as the SKZ, provides a unique insight into Persian claims of victories over the Romans. 14 A carved relief at Darabgird is more controversial with some believing that it depicts Shapur's victories over Rome but others suggesting that it is actually a relief of Ardashir. The choice of Naqsh-i Rustam to depict the captivity of Valerian and submission of Philip on a rock relief (Naqsh-i Rustam 6) together with carving the SKZ on a prominently placed fire altar suggests the identification of the site as an important one to Sasanian royal and imperial rhetoric. 15 As noted in Chapter 3, the importance of Naqsh-i Rustam in the Achaemenid period as a royal burial site (at least four Achaemenid royal tombs are located there) together with its proximity to Persepolis is often used to suggest that the Sasanians consciously sought to link themselves to the Achaemenids as part of an ongoing programme of legitimating the dynasty. 16 At Bishapur, a city and palace foundation of Shapur I established in the mid-260s, three rock reliefs (Bishapur I, II, III) include depictions of Shapur's victories over the Romans. 17 Bishapur I, the most seriously degraded of the three reliefs, retains enough detail to depict a Roman emperor keeling on the ground in subservience between two horses, one likely mounted by Shapur, while another figure lies dead directly beneath the horse. The figure begging is probably Philip but may be Valerian and the prostrate figure is Gordian III.

Shapur I to Carus

115

Bishapur II is the best known of the reliefs depicting Shapur's victories over the Romans and likely shows Philip kneeling before Shapur's horse, Valerian standing behind the horse and held by Shapur by the wrist, while Gordian III lies prostrate under the horse. To the left of the central arrangement are two panels of mounted Persian nobles and to the right what are thought to be defeated representatives of the Kushans and other peoples. 18 Bishapur III is a more complex rock relief than Bishapur II but essentially depicts the same story with more detail, especially in relation to what some believe may be defeated Kushans paying tribute. 19 An alternative theory put more recently by Bruno Overlaet is that the four registers to the right of Bishapur III depict a progressive account of the submission of the usurper Uranius Antoninus of Emesa to Shapur. On this analysis, the figure bending in supplication before Shapur and his horse is Uranius Antoninus as is the figure behind the horse being held by the wrist as an indication of vassal status and gossibly also the dead figure under the horse after the arrangement failed. 2 Alternatively, for Overlaet, the dead figure is a personification of the defeated Roman Empire. The relief would thus depict events of late 253/early 254 during Shapur's Syrian campaign but with a distinctly different outcome to that inferred in the Thirteenth Sibylline Oracle and John Malalas. 21 These texts indicate a defeat for Shapur at the hands of a priest, named Sampsigeramus by Malalas, outside "a city of the sun" claimed in the Oracle. This is thought by most to be Uranius Antoninus, who is not directly named in these texts, but who was producing coins at Emesa depicting himself as Augustus, which are dated

Figure 5.2 Rock relief (Naqsh-i Rustam 6) depicting Shapur I and Roman emperors

Philip I and Valerian. Naqsh-i Rustam, Persis. Photograph: Ross Burns.

116 Shapur I to Carus

Figure 5.3 Rock relief (Bishapur II) depicting Shapur I and emperors Philip I and

Valerian. Gordian III prone under Shapur's horse. Bishapur, Persis: Photograph: Ross Burns.

Figure 5.4 Rock relief (Bishapur III) depicting Shapur I and emperors Philip I and

Valerian at centre. Gordian III prone under Shapur's horse. Kushans and others bearing tribute to right. Romans bearing tribute to left . Bishapur, Persis: Photograph : Ross Burns.

Shapur I to Carus

117

25 3/4. 22 Malalas' clearly confused account of the events even claims that Shapur was killed due to the treachery of Uranius Antoninus. Overlaet bases his interpretation of Bishapur III principally on the belief that the delegation arriving to the right carries a number of offerings, including the Black Stone of Emesa (registers 2 and 4 ), the central cult object of the Temple of Elagabal at the city, for which the priest of Sampsigeramus, that is Uranius Antoninus, was responsible. Overlaet's theory would at least explain the fate of Uranius Antoninus, which is otherwise unknown but there are some potential problems with such an interpretation. The fact that Emesa is not named on the SKZ, which we would expect if the scene depicts events as Overlaet believes, is one such issue. In visual depictions of the Black Stone of Emesa on coins, it was clearly a large stone that required a quadriga to move it in the case of Elagabalus' temporary relocation of it to Rome, and in bronze coins of Uranius Antoninus it is a prominent feature of the temple of Elgabal at Emesa. The offering identified by Overlaet as the stone is carried by a single attendant. At Darabgird (see Figure 5. 7), an especially interesting relief depicts a Sasanian king on horseback with a dead figure (traditionally thought to be Gordian Ill) lying beneath, an elderly figure with head bowed (Valerian?) towards the king who places an outstretched hand on his head and another likely Roman military figure standing with an outstretched hand in supplication towards the king. 23 Behind the king and his horse is an array of Persian nobles and behind the elderly figure is a host of togate, bareheaded Romans. The main source of debate in this relief is the identification of the Persian king whose crown in this case is that typically depicted for Ardashir I. The obvious problem here is that the victories depicted appear to

Figure 5.5 Rock relief (Bishapur III) close up of figure bearing "Black Stone" of

Emesa. Photograph: Ross Burns.

118 Shapur I to Carus

Figure 5.6 Follis , Emesa. Uranius Antoninus. AD 253-4. Obv- AVTOK COVLP

ANTWNINOC CE: Laureate, draped, curaissed bust of Uranius Antoninus right. Rev- EMICWN KOLWN; EXin exergue: Hexastyle temple containing the conical stone of Elagabal, ornamented with facing eagle, shaded by two umbrellas (parasols), pediment ornamented with crescent. Baldus 38-42. Photograph: Effy Alexakis.

be those of Shapur I over Gordian and Valerian. Either Shapur chose for some reason to be depicted in the relief wearing his long-decea sed father's crown or the scene actually depicts an unidentified event or series of events between Romans and Persians during the reign of Ardashir I. 24 Georgina Herrmann preferred to see the Darabgird relief as the depiction of a victory of Ardashir during his invasions of Mesopotamia in the 230s. 25 It may also be possible to link the relief to Severus Alexander 's invasions and the fate of parts of his army in the early 230s. While the rock reliefs and SKZ provide valuable information on the Sasanian Persian wars with Rome and the representation of them to a largely internal Persian audience, the inscription also provides a contemporary snapshot of Sasanian familial and dynastic strength across Persia by the 260s. This was an important factor in allowing Shapur to undertake and sustain the campaigns against the Roman Empire. Hormizd , Shapur's son and ultimate successor as Shahanshah, was the King of Armenia, another son and later successor, Bahram, was the King of Gilan, while Narseh was the King of the Sakas in the east. A fourth son, Shapur, was the King of Mesene in the Persian Gulf, while two brothers of Shapur named Ardashir ruled in Adiabene and Kerman. 26 This tight management of rulership in Persia not only allowed Shapur to better raise troops and reduce the possibility of rebellions while he was on campaign, but also provided a mechanism for succession. The latter may have been a double-edged sword because there were potentially many claimants when Shapur died.

Odenathus of Palmyra and Roman attempts at recovery With Macrianus and Macrianus junior killed in 261 in the process of a failed attempt at invading the western provinces to remove Gallienus,

Shapur I to Carus

119

Figure 5. 7 The Sasanian relief at Darabgird. Ghirshman 1971, plate XXII.

Odenathus set out to deal with Ballista and Macrianus' other son, Quietus, who had established themselves in Emesa. Quietus was defeated in battle and Ballista was killed by his Emesene hosts, which saw Odenathus established as the most powerful figure in the Roman eastern provinces. Numerous texts indicate that Odenathus was now the principal figure in rallying what remained of Roman troops, together with his own Palmyrene militia, who were the main means by which the Persians were forced to withdraw. Odenathus remained loyal to Gallienus and was most likely recognised as Rector Totius Orientis due to his loyalty and military achievements. The granting of this title came on top of his appointment as Dux Romanorum the previous year, the combined titles giving Odenathus the authority to issue orders to the eastern provincial governors while also holding supreme command over Roman military forces. 27 At Palmyra, he was known by numerous titles, including the Lord of Tadmor and King of Kings, the latter taken due to his victory over the Persians. 28 Odenathus' activities prior to the early 260s are speculative, given the state of the evidence but it is probable that he had already played a part in Roman attempts to defend the eastern provinces during Shapur's invasions. 29 De Blois proposes that Odenathus was supporting Roman defensive efforts against the Persians as far back as 252/3 and that he acted briefly in concert with Uranius Antoninus in 254. 30 Sommer suggests that he may have served as the governor of Syria Phoenice prior to 259/60 and continued

120

Shapur I to Carus

in the office after 260. 31 It appears, however, that Odenathus' loyalty to Rome was not always guaranteed because at some stage, which cannot be dated precisely based on the evidence, Odenathus sought to engage in diplomacy with Shapur. According to a later, fragmentary text, Odenathus sent an embassy to Shapur, which included letters and gifts. 32 The embassy was angrily rejected by the Persian king who retorted that he was Odenathus' master anyway before ordering the destruction of the letters and dumping the gifts in the Euphrates. Banchich proposes a date of 260/1 for the embassy based on a sensible analysis of events. 33 De Blois proposes an earlier date coinciding with Shapur's success in capturing Dura Europos and Circesium ea. 256. 34 On De Blois' analysis, with the overture rejected, Odenathus threw his lot in fully with the Romans and in 259, when Valerian arrived to challenge the Persian siege of Edessa, Odenathus marched down the Euphrates and inflicted losses on Babylonia. The last observation is dependent on the identification of Odenathus in a Jewish textual reference that referred to attacks on the towns of Nehardea on the Middle Euphrates, Sekansiv (location unknown) and Mahoza, which has three possible locations on the Euphrates or Tigris Rivers. 35 For De Blois, Odenathus returned from this campaign just in time to attack the Persians as they made their retreat. Some later texts referred to Odenathus' recapture of Nisibis and Carrhae from the Persians ea. 262, while a number of fourth-century Latin texts referred to Odenathus and his army then invading Persian territory, making it all the way to Seleucia-Ctesiphon. 36 Zosimus confirmed this and indicated that the Palmyrene attack on the Persians was a sustained one, likely lasting from 262 to 264. 37 The author of the Historia Augusta claimed that Odenathus' chief purpose in the war against Persia was to retrieve Valerian, although the principal purpose of this claim in the text was to attack Gallienus for making no attempt to repatriate his father. 38 Moreover, the author of the Historia Augusta was keen to point out that Gallienus celebrated a triumph in Rome on the back of Odenathus' victories. Bronze asses minted at Rome in 264 depicting Gallienus in a quadriga are possible indications of the celebration of such a triumph. 39 Epigraphic evidence discovered to the south of Rome and dating to 265 indicates that Gallienus took the title Persicus Maximus and it is possible that Gallienus attempted to emulate Alexander the Great in celebrations of the victories won over the Persians at the hands of Odenathus. 40 Malalas made the claim that Gallienus himself marched against the Persians before striking a treaty and it is possible that this finds numismatic support with the minting of antoniniani in Gallienus' name at Antioch or another eastern mint containing the reverse legend PAX FVNDATA (see Figure 5.8). 41 The accompanying reverse iconography of this coin details a trophy with two bound captives seated beneath it. As with other sections of Malalas' Chronicle that cover the middle part of the third century, however, this passage concludes with a spurious historical claim that diminishes the overall reliability of the text. According to Malalas, Gallienus marched against Persia and then made war on Odenathus who he killed in battle

Shapur I to Carus

121

before returning to Rome. This runs counter to all other surviving accounts, which referred to Gallienus' recognition of Odenathus and reliance on him in the wake of the Sasanian invasions and that Odenathus' death was the result of familial treachery ea. 267. The PAX FVNDATA coins were possibly minted at the behest of Odenathus and may be linked to more ambitious imperial intentions. Sommer believes that Odenathus aimed to "completely get rid of the Sasanians and make Palmyra the powerhouse of a new eastern empire," and that this laid the groundwork for the rebellion under Zenobia a few years later. 42 In 267, just before his murder either at the hands of a cousin, Maeonius, in Emesa, or due to a conspiracy involving the emperor, Odenathus undertook another raid on Ctesiphon. 43 Well over a century later, the Antiochene rhetorician, Libanius would recount the sustained successes of Odenathus over the Persians in the 260s. 44 While it is possible to observe a reprieve for Gallienus in the east following the capture of Valerian, mostly due to the efforts and loyalty of Odenathus, the reality for Gallienus during the years of his sole reign was stark. He had multiple serious issues to deal with, any of which had a distinct chance of ending his own reign prematurely. The revolt by Macrianus senior in favour of his sons Macrianus junior and Quietus in the latter half of 260 together with the usurpation of Postumus in the west had the potential to boil over. When the two Macriani marched on Ital y in 261 and were defeated by Aureolus, one of Gallienus ' generals, the emperor's prospects improved and improved further still when Odenathus decided to remain loyal and defeated the remaining usurper, Quietus and his praetorian prefect Ballista at Emesa in the same year. While Postumus ' rebellion would take in Gaul, Britain and most of Spain, he appears not to have been interested in Italy or the Danubian provinces. Odenathus' loyalty in the east and his management of the situation in relation to the Persians added further to Gallienus' capacity to strengthen his position in the territory immediately under hi s command in the central provinces of Italy, the Danube, north Africa and Egypt. Towards the end of Gallienus' reign in 267 , the Goths

Figure 5.8 Antoninianu s, Antioch. Gallienus. AD 265. Ob v- GALLIENVS AVG: Radiate , draped and cuirassed bust right. Rev- PAX FVNDATA: Two captives bound and seated at foot of trophy; branch in exergue. RIC V Gallienus 652. Photograph: Effy Alexakis.

122

Shapur I to Carus

on the Danube operated at some level of coordination with the so-called "Black Sea Goths," which resulted in serious difficulties in Thrace and Greece. Gallienus' unsuccessful attempts to deal with Postumus the year before had also seriously weakened him. Gallienus' death in September 268 due to a plot of his senior generals was perhaps inevitable by that time.

The rebellion of Palmyra and implications for the relationship between Rome and Persia In time, the strength and authority Odenathus established and maintained on Gallienus' behalf in the east would serve as a factor in the rebellion of Zenobia and Vaballathus. Numerous texts suggested that the death of Odenathus at the hands of a Roman imperial conspiracy was the immediate reason for the rebellion but the strength of the military forces under Odenathus allowed the rebellion to build into a serious problem for Gallienus' successors, Claudius Gothicus and Aurelian. 45 The rebellion of the Palmyrenes, which lasted from 268 to 273, eventually took in most of the eastern Roman provinces, including Syria, Egypt, Arabia, Mesopotamia and a large part of Asia Minor. 46 Coupled with the ongoing success of the "Gallic Empire," the 260s were a challenging time for Roman emperors who also faced near-constant invasions and threats from the Goths, Alammani, Heruli, Iuthungi and Vandals in the Balkan provinces. Claudius II would spend most of his short reign from 268 to 270 dealing with problems with the Goths on the Danube and in Greece and famously scored a great victory over them from which the epithet "Gothicus" would attach permanently to his name. The Persians appear to have paid close attention to the activities of Odenathus, who demonstrated to the very end that the Palmyrenes could pose a real and genuine threat. With the rebellion of Zenobia and Vaballathus, and the death of Gallienus soon after, Shapur likely sensed an opportunity, although with succession moves perhaps already under way for his son, Hormizd, it was not an ideal time to send princes, generals and troops to participate in an inevitable clash with the Romans. Inscriptions from Arabia and Syria indicate a claim by Vaballathus of a victory over the Persians but the extent of any such victory and when it might have happened are difficult to establish with any chronological precision. Milestones dated 270-2 from Bostra and Philadelphia (Amman) include Persicus Maximus in the titulature of Vaballathus. 47

Shapur I's death and the immediate aftermath Shapur I's death in 270 sparked a period of relative instability that would continue until Bahram II's rule became established in the second half of the 270s. 48 The sixth-century Byzantine writer, Agathias, highlighted the brevity of the reigns of Hormizd and Bahram I following the death of

Shapur I to Carus

123

Shapur and the belief among the Romans that neither achieved much. 49 Hormizd was the King of Armenia at the time of his elevation and in military terms would only be known for a war of unknown proportions against the Sogdians in Central Asia. 50 While Sasanian coins display comparatively minor variations in the third century compared with the Roman coinage, Hormizd's silver drachms demonstrate some changes, the most significant being the addition of Aneran to his official title, extending it thus to King of Kings of Iran and non-Iran, which had already been part of the title that his father Shapur I took in the SKZ inscription. Under Hormizd's successor, Bahram I (271-4 ), the seeds of dynastic dispute continued and contributed further to instability in Sasanian leadership at times over the following two decades. Frye discussed evidence to suggest friction between Bahram and his brother Narseh, who was the King of the Sakas at the time of Bahram I's death and believed that he should succeed as Shahanshah. 51 The accession of Bahram in 2 71 had likely caused a rift but with the accession of Bahram's son in 274, the rift widened and led to problems in the late 270s and 280s. Frye proposed that part of the problem was that Bahram I was the son by Shapur I of a lesser queen or even one of his concubines and that this was a factor in stirring dynastic rivalry in the wake of Shapur's death. This was also the period in which the power of Kerdir, the chief priest of the Zoroastrian religion, grew considerably and possibly played a role in the elevations of both Bah ram I and Bahram II. 52 Daryaee discusses Kerdir's role in Bahram I's elevation and notes that in the last year of his reign (274), he successfully orchestrated imperial condemnation of Mani that played out in more dramatic circumstances early in the reign of his son and successor. On this analysis, Kerdir's involvement in Bahram II's elevation to power had assisted in the successful bypassing of Narseh. Kerdir's dividend for supporting both Bahrams, especially Bahram II, was his own increased power over the Zoroastrian religion and the enforcing of the religion's supremacy over other religions, including Manichaeism, Christianity, Judaism and Buddhism. Daryaee also notes that "it is during this period that the Sasanian kings lost much of their religious power as caretakers of the Anahid fire-temple to Kerdir." 53

The emperor Aurelian and Persia Part of this period of Sasanian dynastic instability in Persia overlapped with the period in which the Roman emperor Aurelian was dealing with serious difficulties in the Balkans and northern Italy, together with the Palmyrene rebellion. With Aurelian's armies marching to take on the Palmyrenes in 272, Zenobia sought Persian assistance together with that of Armenia and the Arabs. Aurelian foiled the arrival of Persian reinforcements to assist Zenobia while successfully winning over the Armenian and Arab troops, which had been sent to support Zenobia. 54 With Aurelian's victory over the

124

Shapur I to Carus

Palmyrene forces imminent, Zenobia set out to appeal personally for further Persian assistance but was captured by Aurelian's horsemen. Zosimus referred to Persian mercenaries active in Aurelian's army when Palmyra was put under siege and they were employed by Aurelian for their advanced skills in archery. 55 In the wake of Aurelian's initial victory over the Palmyrenes in 272, the Persians sent gifts to him, including a dazzling purple cloak, as an apparent offering of peace. 56 For the purposes of his own aggrandisement, perhaps by virtue of this embassy, Aurelian took the titles Persicus Maximus and Parthicus Maximus, which is demonstrated in a number of inscriptions from his reign. 57 There is also one extant epigraphic reference to Aurelian taking the title Palmyrenicus Maximus. 58 Aurelian faced the immediacy of dealing with attacks by the Vandals and Carpi on the Danube together with threats by the Alamanni and Iuthungi on northern Italy. After putting down the second rebellion of Palmyra in 273, he was forced to return quickly to Italy to deal with yet another raid of the Alamanni and in 274 successfully confronted the new Gallic ruler, Tetricus, in Gaul. For the first time since the ill-fated expedition of Valerian in 259/60, a Roman emperor was in a position to plan and even begin executing a war against the Persians. 59 Aurelian was murdered in the early stages of preparation for the campaign due to a household conspiracy in 275. 60 The same passage claims, however, that Aurelian had already defeated the Persians by virtue of his victory over Zenobia. This elision of the victory over Zenobia and the Palmyrenes with a victory over Persia may reflect Persian support for Zenobia or a Roman conception at the time that grouped the Palmyrene enemy with the Persian enemy and that the Persians were providing support to Zenobia's Palmyra. There are no explicit signs in Aurelian's coinage of a claim to victory over the Persians but there is numismatic evidence for preparations for the war against the Palmyrenes and the following settlement of the eastern provinces. The minting of extensive series of antoniniani at Rome, Milan, Siscia, Serdica and Cyzicus with reverses depicting Oriens with bound captives at his feet is suggestive of preparation for the war against the Palmyrenes and settling the east but may also have extended to preparations for a war against the Persians. 61 The same mints produced large numbers of antoniniani containing legends typically associated previously with impending military activity against the Parthians or Persians. 62 Large numbers of reverses, especially at Serdica, were also minted with the legend RESTITVT ORBIS and iconography depicting Victory passing a wreath to Aurelian who holds a spear. 63 Aurelian's reign was also well known for coins depicting Sol Invictus, either on the reverse iconography of the PACATOR ORBIS series or accompanying the reverse legend SOLi INVICTO. Not all of these and other military types will have been associated with the wars against the Palmyrenes or an impending Persian campaign, but many of them contained messages previously associated with eastern campaigns and some had distinctively eastern iconography and/or legends.

Shapur I to Carus

125

Aurelian's successors and Persia The prosecution of a war with Persia continued as an expectation of Roman emperors and in the year following Aurelian's planned Persian invasion and death, the short-lived emperor Florian was about to attack the Persians when he was assassinated near the city of Tarsus in Asia Minor. 64 Florian's successor, Probus, received envoys from the Persians who sought peace following the emperor's successes in dealing with an attack on Egypt by the Blemmyae. 65 Probus sent the embassy on its way with a reply that roundly rejected the gifts the Sasanian king had sent and threatened an invasion. Apart from the usual suspicions associated with the Historia Augusta, problems exist with this passage because it refers to Narseh as the Persian Shahanshah with whom Probus corresponded. The possibility of a Persian embassy to Probus finds some resonance in the Armenian History of Moses Khorenats'i, but here the Persian ruler was identified as Ardashir, which is an even more egregious chronological error. 66 Probus spent most of his reign preoccupied with events on the Rhine and Danube. On the Rhine, the Franks and Alamanni devastated Gaul, once again an imperial responsibility due to the reduction of the Gallic Empire under Aurelian. Gaul, the middle Danube and Italy entirely occupied Probus until 279. In 280, Probus arrived at Antioch, likely as part of preparations for a Persian campaign but was forced to depart early in 281 to deal with a serious military revolt based at Cologne. Having successfully dealt with this, Probus was murdered in 282 at Sirmium in Pannonia while in the early planning stages of a Persian campaign. 67 If we are to believe the Historia Augusta regarding the fates of Aurelian, Florian and Probus, marching on Persia had become a virtual guarantee of an emperor's demise.

The Persian invasion of Carus The most significant military engagement between Roman and Persian forces following Shapur I's withdrawal in 260 and prior to the wars between Galerius and Narseh in the 290s was the Persian expedition of Probus' successor, Carus, in 283. 68 Carus ruled as emperor for a little under a year but in keeping with the trends reported since the last year of the reign of Aurelian set his sights on an invasion of Persia. In Carus' case, the invasion actually took place and initially met with success before the emperor's untimely death at the hands of a bolt of lightning while in camp not far from Seleucia-Ctesiphon. He was able to plan and prosecute the campaign because his son Carinus by-and-large managed the Rhine and Danubian problems, although Carus undertook some raids against the Quadi and Sarmatians before the Persian campaign. There are numerous ancient accounts of the invasion but all of them are from the latter half of the fourth century and later. Some were written in the immediate wake of Julian's Persian campaign and formed part of the

126

Shapur I to Carus

rhetorical debates associated with Julian's reign and his death in Persia. In the accounts of a number of the Latin writers of the second half of the fourth century, Carus was strongly criticised for going beyond what the gods were pleased with. 69 Both Aurelius Victor and Festus described Carus as "immodest" and accused him of defying the gods and Oracles. His death by lightning strike was his reward. Ignoring omens and advice from soothsayers was also one of the criticisms levelled at Julian. In accounting for Carus' demise for a similar reason, the gods could still be portrayed as powerful, an especially important point to make in the face of vehement Christian criticism of the disaster that befell Julian. A brief but valuable reference to Carus' campaign is made by Ammianus Marcellinus who was with Julian's army in 363 when the emperor and his guard inspected a deserted city near Seleucia, which had been destroyed by Carus. 70 In later Byzantine texts, there are claims that Carus was successful against the Persians before dying in battle against the Huns. 71 The aftermath of Carus' death in Persia was not accompanied by a military disaster such as that of Gordian III or Julian suggesting that the army was in a reasonably strong position at the time of his demise. There is some confusion about which of Carus' two sons, Numerian and Carinus, fought another brief war with the Persians soon after but both took the title Persicus Maxim us as Augusti. Some ancient texts claimed that N umerian campaigned against the Persians in northern Mesopotamia before he was defeated, captured and killed near Carrhae in November 284. 72 Others claimed that Carinus campaigned against the Persians, one retailing the story in a similar vein to that of N umerian. 73 In the wake of Carus' death, coins were minted advertising his consecration as a god. Among them were coins of the mint at Siscia in Pannonia demonstrating that Carus had been granted the title Parthicus posthumously (see Figure 5.9). 74 Another type with the same legend was also minted at Lugdunum. 75 Of interest in the case of the deified Carus are two types minted at Rome that contain the title Persicus and these are the only Roman coins known to contain the title (see Figure 5.10). 76 The existence of coins commemorating the deification of the same emperor but using two different titles celebrating the same victory is further indicative of the approach of celebrating victories over the Parthians in emulation of the emperors of the second century and over the Persians in emulation of Alexander the Great. As Augusti, Carinus and Numerian also took the title Persicus Maximus, which was a clear attempt to bolster their own legitimacy by linking to their father's victory over the Persians.7 7 When Carus invaded Persia in 283, Bahram II was dealing with a significant problem in the east, "domestica seditione" according to the Historia Augusta. 78 There is an indication that this was a sustained problem in a reference in a panegyric of 291 in honour of Maximian that referred to a rebellion led by Ormies (Hormizd), believed to have been a brother of Bahram II. Hormizd led this rebellion with the support of the Saci (Sakas),

Shapur I to Carus

127

AD 283-5 Obv-DIVO CARO PARTHICO: Radiate bust of Carns right. Rev- CONSECRATIO AVG (SMSXXI in exergue): Lighted altar A in right field. RIC V Carns 111. Photograph: Effy Alexakis.

Figure 5.9 Antoninianus, Siscia. Numerian/Carinus.

Rome. Numerian/Carinus. AD 283-4. Obv- DIVO CARO PERS(ICO): Radiate bust of Carns right. Rev- CONSECRATIO (KA crescent A in exergue): Eagle, standing right, head left or upwards. RIC V Carns 48. Photograph: Effy Alexakis.

Figure 5.10 Antoninianus,

Rufii and Geli (Gilan). 79 On the coins of Hormizd, which were minted at Kabul, Balk, Herat and Merv, he took the title Kushanshah and in at least some of them styled himself iconographically like the Kushan rulers. 80 Bivar believed that this Hormizd was actually the Kushano-Sasanian ruler of the time with the same name but Rezakhani dismisses this suggestion. 81 Whether the specifics of the panegyric of 291 also reflect the situation at the time of Carus' invasion in 283 is difficult to judge but there is sufficient evidence to suggest a sustained period of difficulty for Bahram II in relation to Hormizd through the 280s, which was not resolved until shortly before Bahram's death. If rock reliefs carved at Naqsh-i Rustam do indeed depict Bahram II scoring victories over both Carus and Hormizd, there was a clear attempt at retailing a different version of events to the Latin texts. 82

128

Shapur I to Carus

Conclusion While problems existed for Gallienus at many different levels in the early 260s, some of them perhaps existential, the successes of Shapur would not deal a "knock-out" blow to his reign. The actions and loyalty of Odenathus were the most important elements in Gallienus' immediate survival but the successes of the would-be usurpers, Macrianus and Ballista, against the Persian raiding parties in the immediate wake of the defeat and capture of Valerian should not be dismissed either. By 270, however, the situation in Rome's eastern provinces had deteriorated almost unimaginably with a Palmyrene rebellion that took in almost everywhere east of the Bosporus. The loss of imperial control of Egypt had the potential to cause serious difficulties for the whole empire, given its significance for grain supply. With the Gallic Empire still strong and Claudius Gothicus dying of the plague, Aurelian was likely daunted at the challenges he faced on coming to power in 270. Shapur I's death at about this time would come to mark a period of comparative instability and challenge for the Sasanian imperial leadership. In what appears to have been a struggle between Shapur's descendants to succeed him as Shahanshah, Bahram I emerged in power following the brief reign of his brother, Hormizd, and this was due in part to the support he received from Kerdir. Kerdir would then play an important role in the reign of his son, Bahram II. Bahram II was confronted by a serious insurrection in the eastern portions of the Sasanian realm early in the 280s and this was a likely factor when Carus invaded Persia in 283. Following the successes that Aurelian achieved in putting down the Palmyrene rebellion and bringing the Gallic Empire to heel, a Persian invasion once again became a Roman imperial prerogative. Aurelian and a number of his successors were in the final stages of preparation for a Persian campaign but were assassinated before it could be prosecuted. Only Carus would succeed in undertaking such a campaign in 283 but it would be cut short by misfortune. Bahram II's apparent difficulties with an eastern rebellion made responding to the invasion more challenging. When Diocletian came to power in 284 in the wake of the downfall of Carus' short-reigning sons, there was nothing inevitable about the reforms that his reign would become known for. There was likely little more expectation that his reign would see a significant victory over the Persians. The previous two decades had seen some of the most significant challenges that Roman imperial rulers would face. Some fundamentally important changes solidified during this period, especially with the origins of emperors lying in the Danubian provinces where they had often proven themselves as career military leaders. Despite the problems the emperors still faced, even after the Palmyrene rebellion and Gallic Empire had been dealt with, a Persian invasion became a rhetorical and even practical aim. Revenge for the successes of Shapur I was one driver of this as was the need to establish legitimacy. In Persia, the role of Kerdir in rulers coming to power and the

Shapur I to Carus

129

development of his own political power and position within the nobility would mark a high point in the relationship between church and state in Sasanian Persia.

Notes 1 See Henning 1939, 843; Sprengling 1953, 108-9. 2 SKZ, lines 28-31. 3 Syncellus, 466, 15-23 = Dodgeon & Lieu 1994, 3.3.5 (66); Zonaras 12.25 [602]. 4 Macrianus was based at Samosata at the time of the battle in which Valerian was captured and had refused a desperate request for troops and material support from Valerian; Anonymous continuator of Dio, frag 3 = Dodgeon & Lieu 1994, 3.3.4 (65-6). 5 Dodgeon & Lieu 1994, 366, note 42. 6 SKZ, line 31. 7 Zonaras XII.25 [603]. 8 Peter the Patrician, Banchich F.176, 116-7. 9 See Ball, 2016, 3. 10 See Rezakhani 2018, 68-80. 11 See Rezakhani 2018, 68. 12 Tabari I. 826; Eutychius, Annals 5 = Dodgeon & Lieu 1994, 295-6. 13 Herrmann 1998, 45-6 believes that the Bishapur III rock relief of Shapur's, which depicts his victories over the Romans in its central register, also shows representatives from Central Asia and India bringing gifts. 14 Huyse 1999, vol. I, 10-14 argues for a date of 260-2 for the composition of the SKZ. 15 Herrmann, Mackenzie & Howell 1989. 16 See Chapter 3. 17 Herrman 1977; Hermann 1980; Herrmann & Howell 1981; Herrmann & Howell 1983. Canepa 2013 provides an excellent table of all known Sasanian rock reliefs and provides a good discussion of the reliefs relevant to Persia's conflict with Rome. 18 Canepa 2013, 866. Von Gall 1998 believes that Arabs bringing gifts are depicted in this relief. 19 See Herrmann 1998, 45-6 regarding the Kushans. 20 Overlaet 2009, 477. Also Overlaet 2017. 21 Malalas 12.20 [589J; Oracula Sibyllina 13, lines 147-54. 22 Baldus 1971; Baldus 1997. 23 Tri.impelmann 1975. 24 Herrmann & Curtis 2002. 25 Herrmann 1969, 63-88. 26 SKZ, lines 18-20; Frye 1983, 126. 27 Stoneman 1994, 107; Sommer 2018, 151. There is some debate regarding the exact title, partly because Zonaras XII.24 referred to Gallienus granting Odenathus the title of Strategos of the East. Potter 1990, 390-4 believes that this is actually reflective of the title Rector Totius Orientis. He provides a detailed analysis of the titles taken by Odenathus and why he believes Rector Totius Orientis is the most likely title bestowed on him by Gallienus. 28 See Sommer 2018, 152. 29 See especially Sommer 2018, 145-50 for a detailed discussion of Odenathus' rise to prominence prior to 260.

130 Shapur I to Carus 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

49 50 51 52 53 54 55 56 57

58 59 60

De Blois 1975. Sommer 2018, 150. Peter the Patrician, Banchich F175. Banchich 2015, 115-6. De Blois 1975. Iggereth Rav Sherira Gaon, 82 = Dodgeon & Lieu 1994, 4.3.1 (70-1). Dodgeon & Lieu 1994, 370, note 7 discuss the proposal of Neusner 1966, 48-9 that the date for these events should be 262 or 263. Historia Augusta, Gallienus 12; Zosimus 1.39.1 (Nisibis only). On the following invasion, see Festus, Breviarium 23, 64, 13-18; Eutropius 9.10; Jerome, Chronicon, s.a. 266, 221, 10-2 = Dodgeon & Lieu 1994, 4.3.2 (72). Zosimus 1.39.1-2; See also Sommer 2018, 153. Historia Augusta, Gallienus 10.4. RIC V.1 412,413. AE 2002, 00290. De Blois 1976, 136-7 discusses with caution the claims by some scholars regarding Gallienus' emulation of Alexander. Malalas 12.27; RIC V.1 Gallienus 652. Sommer 2018, 154. Sommer 2018, 155-6; Stoneman 1994, 108. Libanius, epistle, 1006. See Watson 1999, 57-9. Watson 1999, 60-6. Thomsen 1917, 1973, 1996. See Hartmann 20086, 366-7 for further discussion of Vaballathus' titulature and the place of Persicus within it. There is an ongoing debate about the year of Shapur's death. Tabari 1.831 indicated that there was a debate about the length of Shapur's reign even in antiquity and this has ramifications for establishing the exact year of his death. According to Tabari, some believed the length of Shapur's reign was 30 years and 15 days, while others claimed he ruled 31 years, 6 months and 15 days. With the added complication of Shapur's joint rule with his father commencing in 240 and sole rule in 242, and Tabari giving no indication as to whether the estimates of the length of Shapur's reign he quoted were from the beginning of joint rule or sole rule, it is possible that Shapur died as late as 273. Shabazi 2002/17 prefers 270 (in the month of May to be precise), based on the analysis by Henning & Taqizadeh 1957 of the dates of Mani's life and death. Agathias 4.24.5. Daryaee 2013, 10 believes he was the king of Chorasan based on Tabari 1.833. Frye 1983, 127. Panaino 2016 discusses the importance of Kerdir implicit in his ability to put up at least four inscriptions on his own authority. Like Payne 2015, Panaino advises caution regarding the image of Kerdir as a violent persecutor. Daryaee 2013, 11. Historia Augusta, Aurelian, 28.1-3. Zosimus 1.54.3; Watson 1999, 77. Historia Augusta, Aurelian, 29.1-3. CIL 17-02, 00172 (AD 271-5 - Gallia Narbonensis); CIL 17-02, 00184a (AD 274 - Gallia Narbonensis); AE 1936, 00129 (AD 271-5 - Syria). See Hartmann 2008, 317 on Aurelian's titles, which also included Parthicus Maximus and Arabicus Maximus. CIL 5. 4319; Hartmann 2008, 317. Historia Augusta, Life of Aurelian 35.4. Watson 1999, 103-4 doubts the plausibility of a Persian campaign by Aurelian at this time.

Shapur I to Carus

131

61 RIC V.1.17-18, 61-5, 67 (Rome); 134-7, 150 (Milan); 151 (Ticinum); 187-8, 230, 246-55 (Siscia); 276-83 (Serdica); 360-5 (Cyzicus). On the coins of Aurelian related to the war with Palmyra, see Hartmann 2008, 316-21. 62 MARTI PACIFERO, RIC V.1.33-34 (Rome), 133 (Milan), 270-1 (Serdica), 359 (Cyzicus); FIDES MILITVM, RIC V.1.28, 46 (Rome), 92-4 (Milan/Ticinum), 124-7 (Milan), 328-9, 344 (Cyzicus); CONCORDIA MILITVM, RIC V.1.59-60, 82 (Rome), 120 (Milan), 215-9, 244-5 (Siscia), 273 (Serdica), 356 (Cyzicus). 63 RIC V.1. 53 (Rome), 139 (Milan), 287-306 (Serdica), 347-8 (Cyzicus), 386 (Antioch). 64 Malalas 12.32. 65 Historia Augusta, Probus 17. 1-6. 66 Moses Khorenats'i 2.77. 67 Historia Augusta, Probus 20.1. 68 See especially Altmayer 2014, 87-120 for a detailed consideration of the textual, epigraphic and numismatic evidence for the campaign and its aftermath. 69 Aurelius Victor 38.2-4; Festus, Breviarium 24, 65, 6-11; Eutropius 9.18.1; Jerome, chronicon, sa 284,224, 23-225.1 = Dodgeon & Lieu 1994, 5.1.6 (113); Epitome de Caesaribus 38.1; Historia Augusta, Carus 7.1, 8.1-9; Orosius, Adversos Paganos, VII.24.4. See Kreucher 2008, 420-2 for more specifics on Cams' progress on the Persian campaign. 70 Ammianus Marcellinus 24.5.3. 71 Malalas 12.34 claimed that Cams was killed fighting the Huns after invading Persia; Syncellus, 472, 11-2 = Dodgeon & Lieu 1994, 5.1.6 (115); Cedrenus, I, 464, 6-9 = Dodgeon & Lieu 1994, 5.2.1 (119); Zonaras 12.30 [611] who indicates that there was a dispute about Cams' death in antiquity either at the hands of the Huns or a bolt of lightning in Persia. 72 Malalas 12.35; Zonaras 12.30 [611]. 73 Nemesianus, Cynegetica 63-75 = Dodgeon & Lieu 1994, 5.2.1 (116-7) a contemporary source which claimed a great victory over the Persians by Carinus but made no mention of a defeat; Chronicon Paschale, 510.2-15 = Dodgeon & Lieu 1994, 5.2.1, 118-9. 74 DIVO CARO PARTHICO, RIC V.2 108-13. 75 RIC V.2 30. 76 DIVO CARO PERS, RIC V.2 48, 50. Coins minted at Antioch in the name of Philip in 244 contain the legend Pax Fvndata cvm Persis and these are the only numismatic references to Persis on Roman coins. A unique medallion of Galerius from Siscia, dated ea. 302, contains the legend VICTORIA PERSICA (see Chapter 6). See Altmayer 2014, 106-7. 77 CIL 14.126 dating to 283/4 near Ostia; CIL 8.2717 dated 284 at Lambaesis, Numidia. CIL 6.40708 dating to 283-5 in the name of Carinus only. 78 Historia Augusta, Carus, 8.1. See also Eutropius 9.18 who referred to an insurrection among the Persians at the time of Cams' invasion. 79 Panegyrici Latini 11 17.2; Frye 1983, 128 believed that the Rufii should be viewed as the Kushans. 80 Shabazi2002/201Z 81 Bivar 1979, 325-7; Rezakhani 2018, 81. Shahbazi 1988/2016 also dismissed this suggestion. 82 See Bivar 1972, 279ff.

6

The relationship between Rome and Persia during the reigns of Diocletian, Bahram II and Narseh

Introduction Diocletian's elevation to the principate in November 284 would eventually come to mark an important turning point in Roman imperial history but at the time many could have been forgiven for thinking that it might end in short time and in much the same way as the reigns of emperors since Gallienus' demise 16 years earlier. This is not the place for a full assessment or critique of the reforms of Diocletian over the following two decades; however, some consideration of the importance of Diocletian's reforms is appropriate because they had a profound impact on the military and political relationship between the Roman and Persian Empires at the time and also on how the two empires would interact in the future. Diocletian's attempts to deal with the issues that had plagued the many short-reigning emperors who preceded him have typically been hailed as innovations, which brought the instability in the Roman world of the previous half century to heel. Fergus Millar summed this up with characteristic clarity in relation to the Roman Near East in 1993: Beyond that, in the Tetrarchic period, from 284 onwards, there came rearrangements of provincial boundaries; a new taxation-system; what seems to have been a more active and positive attitude on the part of government (reflected in this period in a long series of boundary-decisions in the name of the Emperor); the appearance of a new tier of regional government, the twelve 'dioceses', each embracing a number of provinces; and above all the formalization of the geographical division of the Empire between (eventually) four Emperors: two senior Augusti and two Caesares. It was within that framework that the repeated role of Antioch as an Imperial residence was in effect transformed into that of a regular capital. It was also in this period, in 298/9, that a treaty with Persia gave the Empire in the Near East its greatest-ever extension, to and apparently across the Tigris. Moreover, virtually all along the zone between the settled area and the steppe literary evidence, inscriptions and the physical remains of Roman forts and roads combine to provide the most detailed picture we have for any period of a Roman 'desert frontier'. 1

Diocletian and Narseh

133

Some of the innovations Diocletian's reign has become known for are better interpreted as reactions to current and ongoing issues rather than abstractly designed strategies designed to ward off future problems. They also represented developments in processes already underway, some that went back almost a century. Diocletian's lack of a son saw him choose a colleague in imperial rule on merit rather than filial connection and as it transpired Maximian was an excellent choice and worked well with Diocletian in sharing the imperial office. Within months of his elevation as Caesar, Maximian was dealing with a serious disruption in Gaul caused by the Bagaudae, while Diocletian departed for the east at about the same time. This was not unlike the division of responsibilities between Valerian and his adult son, Gallienus, in 253. Altmayer has recently analysed the rule of Carus, Carinus and Numerian in the years immediately preceding Diocletian's coming to power as a forerunner of the tetrarchy that would eventually emerge in the 290s. 2 In 286, Maximian faced the beginning of what would become a long usurpation in Britain and Gaul when Carausius, one of his officers from the lower Rhine, revolted. Maximian's elevation from Caesar to Augustus was a likely response to this. The years 287-8 saw tribal groups, including the Alamanni and Heruli, undertake significant invasions across the Rhine. 3 With Maximian occupied by the revolt of Carausius, Diocletian came west in 288 to deal with problems in Raetia and in 289/90 the Danubian frontier also required his attention. Thus, the situation in the west was not at all unlike that confronted by many a third-century emperor between the years 235 and 284, and it required the attention of both emperors at times. The military reforms that Diocletian is often credited with may not have been as revolutionary as is sometimes thought. Campbell prefers to analyse the changes that took place under Diocletian largely as developments already underway in the third century, some harking back to the Severan period. 4 The increased number of provinces under Diocletian and the defensive undertakings in Syria, Arabia and Mesopotamia in some ways forced an approach whereby legions were required to be smaller and more versatile and they were already becoming smaller beforehand. Auxiliary cohorts and cavalry vexillations with experience operating in frontier regions were the more appropriate garrisons for the forts that studded the countryside from the Red Sea to the Euphrates and this approach was already in evidence before Diocletian came to power. If the Notitia Dignitatum is an accurate reflection of the distribution of the various units of the Diocletianic army, the eastern provinces contained 28 legions, 70 cavalry vexillations, 54 alae and 54 cohorts. 5 The Danube contained the next highest concentration with 17 legions and an unknown number of vexillations, alae and cohorts, although they were likely to have been in similar proportion to the legions as they were in the east. This overall distribution was not considerably different to the situation prior to Diocletian's reign but the divisions perhaps were, especially in relation to the garrisoning of forts such as those constructed along the Strata Diocletiana.

134

Diocletian and Narseh

Significant currency reforms took place under Diocletian in an attempt to arrest the ongoing problems with the silver currency unit, especially obvious from the joint reigns of Valerian and Gallienus onwards. Currency reform had been attempted during the reign of Aurelian but appears to have largely failed. While he restored the weights of aurei, antoniniani (now Aureliani) and denarii to the ratios under Caracalla, the problem of low levels of silver content persisted, thus continuing to undermine confidence in the silver units. 6 Diocletian attempted currency reform based on the system of Nero but its success is difficult to judge. It is possible that the inflationary problems spelled out so dramatically in the Edict on Maximum Prices were in part caused by the manipulation of the silver units under Diocletian. 7 An important change that came about in the Diocletianic currency reform was a more uniform approach to iconography that by-and-large attained across the empire of the tetrarchs. Gold coins depicted the deities associated with Diocletian and Maximian (Jupiter and Hercules), argentei contained political and military themes, while the nummus contained what Corbier refers to as "the ecumenical image of the Genius of the Roman people." 8 Nummi depicted other deities often associated with prosperity. This more generalised approach to numismatic iconography means that the coinage is less useful as a source for specific historical events from the 280s onwards, especially in relation to imperial claims of victories over enemies. Perhaps the most important shift relating to the currency as a result of currency reforms and developments in the tetrarchy was the beginning of a move away from silver currency in preference for gold as a means of remunerating soldiers, the final element of this shift put into place under Constantine with the minting of a new gold unit, the solidus.

The return of Tiridates to Armenia In the east, Diocletian spent the years 285-7 overseeing the beginning of improved defences against the Persians and Arabs, and importantly during this period, struck an agreement with the Persians. In a panegyric of 289 in honour of Maximian, the author referred to gifts that were sent by the Persians to Diocletian and the author claimed with rhetorical flourish that the Persians submitted their empire to him. 9 Bahram II was even said to have made supplication to Diocletian. 10 We saw in the previous chapter that Bahram was dealing with a serious revolt in the east led by Hormizd at the time of Carus' invasion in 283 and with the continuation of these difficulties, he likely had little choice but to strike the agreement with Diocletian. The treaty's most important provision was the return of Tiridates to Armenia to rule over at least a portion of the kingdom, a major concession given the strength the Persians had enjoyed in Armenia since the 250s. Tiridates had originally fled to the Romans as a child ea. 250 and for the next four decades, Armenia was under direct Persian control, usually by a close relative of the Shahanshah himself. Tiridates apparently

Diocletian and Narseh

135

found a large Persian army occupying Armenia when he returned to the kingdom with Roman support. 11 He then spent most of his reign consolidating his position in Armenia and attacked Persian territory as far south as Assuristan. 12 Bowman proposes that the province of Mesopotamia may also have been restored to Rome at this time; however, there is no reason to believe that it had not remained in Roman and/or Palmyrene hands after Odenathus recovered Nisibis and Carrhae in 262 (see Chapter 5). 13 The return of Tiridates to Armenia was a possible reason for Diocletian taking the title Persicus Maximus. 14 The return of Tiridates to Armenia in placation of Diocletian not only bought time for Bahram in his necessity to deal with the revolt of Hormizd in the east, but may also have been an attempt to weaken the position of his uncle, and long-time rival for the position of Shahanshah, Narseh, who was at the time King of Armenia. Daryaee believes that Tiridates was only able to take control of the western portion of Armenia, while Narseh retained control of the larger eastern portion. On any analysis, the return of Tiridates weakened Narseh's position. 15 There is also evidence at this time for other developments in the Caucasus, which appear to have strengthened Bahram II's hand. Following a complex period of political manoeuvring at the hands of the Arsacids, who had again become resurgent in Iberia following Shapur I's death in 270, Bahram sponsored the elevation of a child king, Mirian III, to the throne of Iberia and purportedly travelled to the capital, Mtskheta, to oversee the coronation himself .16 In the Georgian Life of the Kings, Mirian was an illegitimate son of Bahram but Rapp prefers the suggestion that Mihran/ Mirian was from the powerful Mihranid house, one of the seven great Iranian families. 17 The years 290-92 are poorly attested in both the Roman and Persian sources. Most activity undertaken by Diocletian and Maximian was concentrated on the Rhine and Danube. 18 Diocletian undertook a war against the Saracens in 290, in which the Persians also appear to have played a role and it was perhaps during this period that Bahram was finally successful in putting down the revolt of Hormizd. 19

Diocletian and N arseh The elevation of Narseh as Shahanshah in Persia in 293 and the implementation of the tetrarchic imperial college by Diocletian in the same year would be important to both empires internally and would also prove important to the military and diplomatic relationship between the two empires. Narseh's elevation followed an internal power struggle, which had preoccupied the Persian leadership since the death of Bahram II earlier in 293, and was part of a rivalry that had been in play since the years immediately following the death of Shapur I. As we have already seen, Bahram II's acquiescence to Diocletian's request for the return of Tiridates to Armenia in 287 may have

136 Diocletian and Narseh formed part of strategic manoeuvring on Bahram's part in relation to his uncle Narseh. The impact of Kerdir's religious and political manoeuvring, especially in the elevation and entire reign of Bahram II, also appears to have been a factor in the rivalry between Bahram and Narseh. On Bahram II's death in 293, he was initially succeeded by his son Bahram III, then king of the Sakas in the east, but serious divisions within the Sasanian royal family and nobility came to the fore. The Paikuli inscription from northern Iraq provides the details and means by which Narseh claimed he came to power and it is potentially revealing. 20 The influence of a leading noble called Wahnam and the King of Meshan in the elevation of Bahram III had threatened to lead to civil war before a number of princes, nobles and apparently Kerdir, turned to Narseh. According to the inscription, Narseh made his way from Armenia to the borders of Assuristan, where he was met by the princes, nobles, and Kerdir, before overthrowing his grand-nephew Bahram III. 21 At the end of the Paikuli inscription, Narseh included a list of rulers and nobles with whom he claimed to be in peace and friendship. 22 Among them was Tiridates of Armenia and the list also included kings of central Asian principalities such as Khwarazm (Chorasmia) and Baluchistan. 23 This part of the inscription also mentioned a king of the Kushans. Rulers of Arab tribes were listed as well but as Frye pointed out, none of the Sasanian princes were referred to and neither were the rulers of the centrally important principalities of Kerman, Merv, Gilan and Meshan. 24 The King of Meshan, of course, had been supportive of Bahram III, so there is little surprise at his lack of inclusion. The Paikuli inscription stands in stark contrast to the SKZ by indicating that Narseh's principal support came from the periphery of the Sasanian Empire, possibly indicating his "outsider" status in relation to the Sasanian royal family during the reigns of the Bahrams, despite being a son of Shapur I. In the first line of the final paragraph, the inscription includes the statement, "And We claimed the whole realm anew," suggesting that Narseh's rule would not mark a continuation of the rule of the Bahrams. 25 Daryaee proposes that a rock relief of Narseh's from Naqsh-i Rustam indicates that Narseh regained control of the fire temple of Anahid at Istakhr, which had been given to Kerdir by the Bahrams and was likely indicative of a "religio-political shake-up in the Sasanian Empire" as a result of Narseh's coming to power. 26 Herrman and Howell point to good evidence that the investiture scene of Bahram I at Bishapur (Bishapur V) suffered a Persian form of "Damnatio Memoriae" and was in essence "usurped" by Narseh with his name written over that of Bahram's. 27

The eastern defensive reforms of Diocletian While it is possible that Diocletian began planning a significant strengthening of fortifications and roads in Syria and Arabia soon after his arrival in the east in 285, much of the activity to achieve this belonged to the period

Diocletian and Narseh

137

from 293 onwards. 28 The principal purpose of this activity was to contain future Sasanian threats and potential threats from Arab tribal groups with some evidence to indicate concerns related to a joint threat from both of them. In northern Syria, there is direct evidence for concerns about future Persian threats in the strengthening of the fortress of Circesium at the confluence of the Khabur and Euphrates Rivers. Circesium was important enough to the Persians to be mentioned in the SKZ inscription of Shapur I but Ammianus Marcellinus claimed that it was only small and exposed to danger before Diocletian strengthened it. 29 Ammianus related the improved defences at Circesium to preventing a future Sasanian capture of Antioch as Shapur I had done in 252/3: This place (Circesium), which was formerly small and exposed to danger, Diocletian, alarmed by a recent experience, encircled with walls and lofty towers, at the time he was arranging the inner lines of defence on the very frontiers of the barbarians, in order to prevent the Persians from overrunning Syria, as had happened a few years before with great damage to the provinces. He went on to tell the story of the Persian assault on Antioch in 252/3 during which the theatre at the city came under attack. Diocletian's commitment to Antioch is demonstrated in a lengthy description by John Malalas of the emperor's building programme at the city: He built there (Antioch) a large palace, having discovered foundations which had been laid formerly by Gallienus Licinianus. Diocletian also built a public bath, which he called Diocletianum, in the level part of the city near the old hippodrome. He also built granaries to store corn, and he gave everyone measures for corn and all other commodities on sale, so that none of the market traders should be intimidated by the soldiers. He also built what is known as the stadium at Daphne for the Olympic athletes ... He also built a temple of Olympian Zeus in the stadium at Daphne, and at the curved end of the stadium he built a temple to Nemesis. He reconstructed the temple of Apollo, decorating it with different kinds of marble. He also built a subterranean temple to Hekate with 365 steps. He built a palace also at Daphne for visiting emperors to stay in, since formerly they had put up tents in the grove and stayed there. 30 In the sixth century, Procopius referred to Diocletian's construction of forts in the province of Euphratesia (Euphratensis in Latin) and these were probably associated with the defence of Antioch. 31 Euphratesia included the bridge-crossing at Zeugma, the former royal capital of Commagene, Samosata, and its metropolis was at Cyrrhus. Further to these textual references, Malalas referred to the construction of an arms factory at Edessa

138 Diocletian and Narseh

CILll'IA

- Rcsalna

l.l'ugnu

r ( .•irrhat·

\nrir,ch

0·1·uncin1r

Sdeucia Piena

.\kppo

c f l:mLI

SYRIA

I· .'rr".i r ,. ·r J