Skip to main content Accessibility help
Hostname: page-component-6c8bd87754-9k7mv Total loading time: 0.196 Render date: 2022-01-17T08:28:19.237Z Has data issue: true Feature Flags: { "shouldUseShareProductTool": true, "shouldUseHypothesis": true, "isUnsiloEnabled": true, "metricsAbstractViews": false, "figures": true, "newCiteModal": false, "newCitedByModal": true, "newEcommerce": true, "newUsageEvents": true }

Article contents

The Garrisoning of Mesopotamia in the Late Antonine and Early Severan Period

Published online by Cambridge University Press:  08 May 2015

D.L. Kennedy*
University of Sheffield


The garrison of Mesopotamia in the later second century is unknown. What follows, therefore, is inevitably somewhat speculative. Nevertheless, with little likelihood of any foreseeable accretions of much new epigraphic or archaeological evidence, it may not be unproductive to resume such evidence as is available and consider the probabilities by reference to other comparable regions.

Research Article
Copyright © Australasian Society for Classical Studies 1987

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)


* The genesis of much of this article is to be traced to a period of Study Leave spent in the Department of Classics and Ancient History at the University of Western Australia in 1984. Subsequently, some of this material was presented at Seminars in that Department and the Departments of Classics and Ancient History in the University of Queensland and of Classics at the Australian National University. I am grateful to all those who participated and offered discussion, as well as to Glen Bowersock, Philip Freeman, Ben Issac and Tom Parker for helpful advice and comments at earlier stages.

1 The frontiers of Syria, Iraq, Iran and Turkey all come together in this area in which is to be found too the large Kurdish population spanning the borders of all four countries. Field-work in so sensitive an area is, therefore, at best difficult.

2 Dillemann, L., Haute Mésopotamie orientale et pays adjacents (Paris1962), 197.Google Scholar

3 Dio 75.1.1–3

4 Wagner, J., ‘Provincia Osrhoenae’, in Mitchell, S. (ed.),Armies andFrontiers in Roman and Byzantine Anatolia [BAR S156] (Oxford 1983), 106114.Google Scholar

5 ILS 9117 for 115/6; ILS 394 for 184/5; CIL 3.13627A for amilestone of Commodus found some 20 miles south-west of Kainepolis and presumably on a road linking that garrison to Cappadocia. We might recall too that under Claudius a cohort had been stationed at Gorneae 12 miles south-east of Artaxata (Tacitus, Ann. 12.45.3;c/ Dio 61.6.6); after Corbulo’s installation of the Roman nominee in Armenia in 60 he left a force of 1000 legionaries, 3 cohorts and 2 alae of auxiliaries. This was at a time when the nearest major Roman garrisons were rather more distant than a century later. From the reign of Domitian we have now an inscription cut on a rock face near Baku on the Caspian and recording a vexillation of XII Fulminata (AE[1951] 263; cf. Grosso, F., Epigraphica 16 [1954], 117 f.).Google Scholar

6 Dio 71.3 — dated to 172 by vonPremerstein: see Magie, D., Roman Rule in Asia Minor (Princeton 1950), 1536 n. 14.Google Scholar

7 Dio 76.9.1–2.

8 Kennedy, D.L., ‘Ti. Claudius Subatianus Aquila, “First Prefect of Mesopotamia”ZPE 36 (1979), 261 f.Google Scholar

9 French, D., ‘R.E.C.A.M. Notes and Studies No. 2. Ti. Claudius Subatianus Aquila’, Anatolian Studies 27 (1977),191 f.CrossRefGoogle Scholar

10 Wagner, op. cit. (n.4), 107–13.

11 Dio 55.24.4.

12 Mann, J.C., ‘The raising of new legions during the Principate’,Hermes 91 (1963), 483;Google Scholar cf. id., Legionary Recruitment and Veteran Settlement during the Principate (London 1983), 63.

13 Most conveniently studied — in conjunction with other texts naming Pacatianus — in Pflaum, H.-G, Les carrièresprocuratoriennes équestres (Paris 1960-1), II no. 229Google Scholar. In a fragmentary career inscription, almost certainly of this man, he is described as procurator et praeses Alpium Cottiarum. I am grateful to the anonymous reader of this article for the cautionary observation that the cursus of Pacatianus, cut after Severus’ death, does not necessarily prove that the title Parthica was current in 195, only that that was its title a generation later.

14 Herodian 3.3.10.

15 Mann, op. cit. (1963).

16 Smith, RE., ‘The army reforms of Septimius Severus’, Historia 21 (1972), 486 n.28.Google Scholar

17 Mann, op cit (1983), 63 and 145; Speidel, M.P., ‘Legionaries from Asia Minor’, ANRW 2.7.2 (1980), 730 ff.Google Scholar (esp. 737). Cf. Speidel, M.P. and Reynolds, J.M., ‘A veteran of legio I Parthica from Carian Aphrodisias’, Epigraphica Anatolica 5 (1985), 33 f.Google Scholar These latter cautiously decline to see the deceased of their inscription as an original recruit. I do not believe the interpretation of Birley, E., ‘Promotions and transfers in the Roman army. II The centurionate’, Carnuntum Jahrbuch 1963/64 (1965), 2133Google Scholar that the two centurions at Lambaesis of CIL 8.2877 and 2891 were engaged in the original formation and training of the legion, or some part of it (accepted by Speidel, M.P., ‘Severiana as the title for army units and the three Legiones Parthicae of Septimius Severus’, PACA 17 (1983), 120 f.).Google Scholar Rather they seem to be examples of what Birley himself, in relation to equestrian commanders, has called’impending movement’. I hope to develop this question of ‘recent or impending movement’ for soldiers elsewhere.

18 Herodian 3.1.3: ‘A large number of the lower class in Antioch, particularly the young men who were unthinking, enthusiastic supporters of Niger, joined the army only really on an impulse rather than because of their experience’; 3.4.1 : ‘a vast number of people, including almost the entire youth of Antioch, presented itself for service …’; cf. 3.2.10 and 3.4.6. With so many killed at Issus alone (Dio 75.8.1 says 20,000 of Niger’s followers), Severus would have been reluctant to release from service those who may have seen themselves enlisting only ‘for the duration’ even if not drawn from the traditional sources of recruits.

19 OCD2 s.v. ‘Legion’; Mann, op.cit. (1983), 160; Ritterling,RE 12.1435; Magie, op. cit. 1544.

20 Amm. Marc, 20.6.8; ILS 9477 on which the name of a legion, called ‘Severiana Antoniniana’, is unfortunately missing. More of the text has now been found and supplies the name: M.P. Speidel and J. Reynolds, op. cit (n.17), 31–5.

21 F. Schachermeyr, RE 15.1161.

22 Castelin, K.O, The Coinage ofRhesaena in Mesopotamia (New York 1946), 6 et passim.Google Scholar

23 Segal, J.B., Edessa, ‘The Blessed City’ (1910), 10, 21 n.4, 79.Google Scholar

24 Before being ceded to the Persians in 363, Nisibis figures far more prominently in Ammianus’ account than does Rhesaina: Nisibis is referred to 10 times, Singara 6, Rhesaina once — and that a reference back to Gordian Ill’s victory there a century before. For the importance of Nisibis see now B. Isaac apud A. Oppenheimer, Babylonia Judaica (Wiesbaden 1983), 319–34. D. Oates, ‘A Note on Three Latin Inscriptions from Hatra’, Sumer 11 (1955), 40 places the legions at Singara and Nisibis. For a later date, the Notitia Dignitatum records (Or. 26.29) under the Dux Mesopotamiae: Praefectus legionis primae Parthicae Nisibenae, Constantino (= Constantia), indicating presumably that the I (sic) Parthica had some association with Nisibis before it was ceded to the Persians in 363.

25 See for example the vexilla of the four Syrian legions on the coinage of the colony of Ptolemais Ace, never the base of even one of these legions:Head, B.V., Historia Numorum, 2nd ed. (Oxford 1911), 793.Google Scholar Later, the colony of Neapolis struck coins for.Y Fretensis (Barag, D., ‘Brick stamp impressions of the legio X Fretensis’,BJ 167 [1967] 244 n.l).Google Scholar See now the references and discussion for colonies and vexilla issues by Issac, B., ‘Roman Colonies in Judaea: the Foundation of AeliaCapitolina’, Talanta 12/13 (1980/81), 40 f. andn.37.Google Scholar

26 None are in fact known in Severus’ army in Gaul; cf. Whittaker, ad Herodian 3.7.3, n.3 (Loeb). However, CIL 6.210 gives the career of a soldier of legio VI Ferrata transferred into the Guard, probably after Issus in 194.

27 Harrer, G.A., Studies in the Roman Province of Syria (Princeton 1915), 8790:Google Scholar AE (1930) 141 for 194 and naming Ti. Manilius Fuscus as leg. pr. pr. praesespro[vinci]ae Syriae Phoeni[ces]; cf. Murphy, G.J., The Reign of the Emperor L. Septimius Severus from the Evidence of the Inscriptions (Philadelphia 1945), 43.Google Scholar

28 Graham, A.J., ‘The Division of Britain’, JRS 56 (1966), 104.Google Scholar

29 Wagner, J., Seleukeia am Euphrat/Zeugma (Wiesbaden 1976), 132–46; IGLS 1.3942.Google Scholar

30 ND, Or. 33.23, 28.

33 Gracey, M.H., The Roman Army in Syria, Judaea and Arabia, unpubl. diss. (Oxford 1982), 92–6;Google ScholarPoidebard, A., La trace de Rome (Paris 1934), 3584.Google Scholar

32 References collected inWelles, al., The Excavations at Dura Europus. Final Report, V.l The Parchments and Papyri (New Haven 1959), 25.Google Scholar

33 Discussed byKennedy, D.L., ‘The frontier policy of Septimius Severus. New Evidence from Roman Arabia’, in Hanson, W.S. and Keppie, L.J.F. (eds.), Roman Frontier Studies, XII, 1979 [BAR S71] (Oxford 1980), 879–88.Google Scholar

34 SHA, Sev. 11.7 for the date; cf. Fink, al., ‘The Feriale Duranum’, YCS 7 (1940), 306 f.Google Scholar

35 In winter the news is unlikely to have been conveyed by sea. The land distance from Antioch to Lugdunum would, even at the very fastest speed, require some 25 days; probably the news travelled slower and took as much as 8 weeks.

36 SHA, Sev. 10.7; Clod. Alb. 9.1-2; Dio 75.6.2. See Graham, A.J., ‘The numbers at Lugdunum’, Historia 27 (1978), 628 n.28 and 629 for the extent of support for Albinus on the continent.Google Scholar

37 Dio 76.6.1–7.2; Herodian 3.7.2–6.

38 Platnauer, M., ‘On the date of the defeat of Pescennius Niger’, JRS 8 (1918), 148;Google Scholar followed by G.J. Murphy, op. cit. (n.27), 80: Birley, A.R., Septimius Severus (London 1971), 177;Google Scholar Herodian 2.7.7-8 nn. 1-2 (Whittaker): Speidel, M.P., ‘The Roman Army in Arabia’, ANRW 2.8 (1978), 722.Google Scholar Cf. Bowersock, G.W., Roman Arabia (Cambridge, Mass. 1983), 112 f.Google Scholar

39 Herodian 3.4.7 reports the flight of many of Niger’s soldiers ‘beyond the Tigris’. Pflaum, op. cit. II no. 263 and add for P. Vibius Marianus, transferred from his centurionate in the frumentarii into legio III Gallica as primus pilus, possibly, Pflaum suggests, in the wake of Niger’s defeat.

40 Graham, op. cit. (n.36), argues that the 150,000 on the field at Lugdunum reported by Dio (75.6.1 ) must — and could — have been fairly evenly divided. It is not necessary however to assume parity of numbers to explain the close result —the army of Britain was notoriously tough and experienced (cf. Herodian 2.15.2). The ferocity and closeness of the battle reflect this and the high stakes both men fought for. It is surely inconceivable that Severus, with so much gained already but again thrown into the balance, and with superior manpower resources, would have allowed himself to be matched at all closely in strength on the battle field.

41 Ritterling, RE 12.1513.

42 Speidel, op. cit. (n.38), 722; Kennedy, D.L., ‘The date of the Arabian governorship of Q. Scribonius Tenax—, ZPE 37 (1980), 26.Google Scholar

43 Alfoldy, G., ‘Septimius Severus und der Senat’, BJ 168 (1968),119 n.52;Google Scholar cf. Speidel, op. cit. (n.38), 722. ‘Seem’ because we may now be able to redate Q. Scribonius Tenax to this very period: Kennedy, op. cit. (n.42), 35-7; cf. Bowersock, op. cit. (n.38), 161.

44 Speidel, op. cit. (n.38), 722 — ‘astonishing risk’; Kennedy, op. cit. (n.42) — ‘remarkable’; Bowersock, op. cit. (n.38), 116— ‘bizarre’.

45 Bowersock, op.cit. (n.38), 116 f.

46 Birley, op. cit. (n.38), 11 ff.

47 I Italica was at Novae in Moesia; III Italica at Lauriacum in Noricum and///Itàlica at Castra Regina in Raetia. Another possibility, III Gallica, at, probably, Raphanaea in Syria, was later to stage a lone mutiny against Elagabalus c. 219 (Dio 80.7.1 and 3).

48 See for example Dio68.21.1,31.1;75.1.1,12.2; Pliny,.N7f5.85-6;Plut. Ant. 37.4. See in general Dillemann, op. cit. (n.2), 88 f; Shahid, I.,Rome and the Arabs (Washington, D.C. 1984), 7, 99. Cf. P. Dura 22, 25.Google Scholar

49 Eutropius 8.18.4; Festus 21; SHA, Sev. 9.9, 18.1; Herodian 3.9.3; Zosimus 1.8.

50 Drijvers, H.J.W., The Book of the Laws of Countries (Assen 1964), 56 f.Google Scholar

51 Bowersock, op. cit. (n. 3 8), 79n. It might even be a reference to the Caracallan annexation of the rump of the kingdom of Osrhoene (I owe this observation to Philip Freeman).

52 SHA, Sev. 9.11; and see for example ILS 418 and 445, both for 196.

53 Lessing, C., Scriptorum Historiae Augustae Lexicon (Leipzig 1901–6), 313.Google Scholar I am grateful for this point to the anonymous reader of my original submission to Antichthon.

54 SHA, Comm. 6.2; cf. Dio 79.9.2–3.

55 Dio 80.4.1–2.

56 A less likely explanation in my opinion but one at least deserving of consideration is that legio Arabica is to be understood as the actual name of a legion as with the various legions called Italica, Parthica, Gallica, Cyrenaica, Scythica,Macedonica. No such legion is of course otherwise known and the best one might propose is that the new legions of Severus, or some of them, originally bore names other than I, II and III Parthica. The two new legions of Marcus Aurelius, for example, were originally called Pia and Concors (ILS 2287), and, as we have seen, the earliest reference to one of the new Severan legions called it simply legio Parthica. Is it possible that Severus originally called his new legions Arabica,Parthica and, perhaps, Adiabenica?, only later, after the Second Parthian War, changing the names? However, as the referee of this article observes to me, the usage of the Historia Augusta requires that the legion’s name be normally complemented either by its numeral (e.g. Hadr. 3.6) or its domicile (Sev. 5.1, 6.7; Av. Cass. 5.5).

Cited by

Send article to Kindle

To send this article to your Kindle, first ensure is added to your Approved Personal Document E-mail List under your Personal Document Settings on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part of your Kindle email address below. Find out more about sending to your Kindle. Find out more about sending to your Kindle.

Note you can select to send to either the or variations. ‘’ emails are free but can only be sent to your device when it is connected to wi-fi. ‘’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.

Find out more about the Kindle Personal Document Service.

The Garrisoning of Mesopotamia in the Late Antonine and Early Severan Period
Available formats

Send article to Dropbox

To send this article to your Dropbox account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your <service> account. Find out more about sending content to Dropbox.

The Garrisoning of Mesopotamia in the Late Antonine and Early Severan Period
Available formats

Send article to Google Drive

To send this article to your Google Drive account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your <service> account. Find out more about sending content to Google Drive.

The Garrisoning of Mesopotamia in the Late Antonine and Early Severan Period
Available formats

Reply to: Submit a response

Please enter your response.

Your details

Please enter a valid email address.

Conflicting interests

Do you have any conflicting interests? *