How did Trajan succeed in subduing Parthia where Mark Antony failed?

di Graham Wylie

The Ancient History Bulletin, 4/2 (1990), pp. 37-43



The facts of Antony's expedition of 36 B.C. against Parthia are briefly these. Taking advantage of internal dissension in Parthia, he launched an offensive through Armenia and Media in alliance with the Armenian king Artavasdes. But the siege train which was following him was caught and destroyed by the Parthians, with the result that he was forced eventually to raise the siege of Phraaspa, where he had planned to winter, and retreat through Media to Armenia, harassed by the Parthian horse-archers and losing a large part of his force. Contrary to the fairy tales perpetuated from ancient sources by earlier modern historians,1 the expedition was not bungled from the outset by an Antony besotted with love for the Egyptian queen, but seems to have been well planned, well organized, and well led by the foremost general of his day (App., BC. 5.14) who indeed deserves special commendation for his inspiring leadership during the retreat, which probably saved his whole army from annihilation. Yet he failed to subdue Parthia, whereas Trajan succeeded about 150 years later. Why was this?

A short answer might be that Trajan did not succeed in subduing Parthia, which erupted into revolt as soon as he turned his back. We might also note that his last personally directed operation, against the desert fortress of Hatra, was as total a failure as Antony's siege of Phraaspa. But he, or his lieutenants, were certainly more successful than Antony against the Parthians in the field. Why was this?

Our information on Trajan's Eastern campaigns in A.D. 114-117 is scanty and sometimes conflicting. The main extant sources are the epitomized book 68 of Dio's History, the relevant parts being mostly due to Xiphilinus; the Chronographica of Malalas of Antioch, a sixth century historian of dubious accuracy;2 and some fragments of Arrian's lost Parthica (on which Malalas' account appears to be based), which are, however, so fragmentary that it is not always possible to associate them with any particular incident or campaign.3 There is also numismatic and epigraphical evidence.

Even the chronology of the campaigns appears to be still in dispute. Longden,4 who is followed by Debevoise,5 considers that the conquest of Upper Mesopotamia and Adiabene (Dio 68.21-23.1) took place in 114 immediately after that of Armenia, and that Ctesiphon was captured in the following year. Garzetti,6 however, regards Longden's placing of the Babylonian campaign in 115, rather than in 116, as 'fairly questionable'. He considers that 115 was a relatively 'empty year' in which Upper Mesopotamia and Adiabene were subdued and a definite frontier line drawn between the Euphrates and the Tigris. The matter turns largely on the dating of an earthquake which occurred in Antioch while Trajan was wintering there (Dio 68.24-5) and which the Dio/Xiphilinus narrative places before the conquest of Mesopotamia. Malalas gives the date as 23 December 115, but Longden thinks it should be a year earlier.7

These campaigns, even if actuated partly by Roman imperialism or 'love of glory' (Dio 68.17.1), had the more practical objectives of ensuring the security of Syria and Roman control of the caravan trade routes to the East, and establishing an eastern frontier to the Empire. The Parthian king Chosroes had provided an excuse by deposing Axidares, the Armenian king nominated by Nero, in favour of his brother. Trajan left Rome in October 113, ignoring Parthian attempts at conciliation, reached Antioch in the New Year and advanced into Armenia in the following spring. After being joined later by forces from Cappadocia and Galatia, his total strength would have been about 11 legions.8

Dio's epitome gives no details of the Armenian campaign, but it appears that Trajan marched up the Murad valley to Arsamosata without meeting any resistance at all. About the end of spring he reached Satala, where he was joined by the rest of his forces and received pledges of peace from most of the tribal chiefs from the Caucasus and Caspian regions. Lusius Quietus, a brilliant Moorish officer, was sent with a column to subdue the hostile Mardi east of Lake Van. Advancing east to Elegeia, near Erzerum, Trajan interviewed and deposed the Armenian king Parthamasiris, informing him bluntly that Armenia was now to be a Roman province (Dio 68.19.3). Soon afterwards the monarch was killed, probably by Trajan's orders. The conquest and pacification of Armenia was completed by the end of the campaigning season,9 apparently with few hostilities (Dio 68.18.3b), many of the chiefs surrendering voluntarily. L. Catilius Severus was appointed governor and Trajan then moved southward (c. Sept. 114) through the Bitlis pass to his next objective, Upper Mesopotamia, then consisting of a number of petty states more or less under Parthian domination.

Having captured Nisibis (Dio 68.23.2) Trajan probably used it as a base, and set up garrisons elsewhere. Local princes Mebarsapes of Adiabene, his vassal, Mannus, ruler of the Arabs centred on Singara, and Manisarus of Gordiene were understandably nervous of him (or of the Parthians), and communicated with him only through envoys. The emperor visited Abgarus the king of Osroene at his capital Edessa, and the two became friends (21.2-3). It was probably at Abgarus' suggestion that Sporaces, ruler of Anthemusia, was dispossessed and his capital Batnae occupied (23.2). Singara and other places (Libbana and Thebeta) were occupied by Lusius 'without a battle' (22.2). The kings of Gordiene and Adiabene remained safely behind the Tigris. No help came from Parthia, which was in the throes of a dynastic struggle.

Whether Trajan wintered in Edessa or in Antioch is uncertain, but at the beginning of spring he 'hastened into the enemy's country' eastward towards Gordiene and Adiabene. Meanwhile boats had been constructed in the forests around Nisibis for the purpose of bridging the Tigris. These were taken apart, loaded into waggons and transported to the river. The point selected for crossing was well up in the highlands, 'opposite the Gordiene mountains', probably because the river was narrower there.

This portion of Dio's narrative (68.26.1-3) gives a little military detail. Building a stable bridge of boats across a fast-flowing mountain stream must have been no easy matter, especially under fire from an enemy army assembled on the far side. 'But Trajan had a great abundance of ships and soldiers' enough apparently to distract the barbarians while the bridge was flung across. Once a bridgehead was established, enemy resistance collapsed, and a large tract of territory fell into the hands of the Romans, though it is not clear whether one or two campaigns were required. The new province of Assyria included Gordiene, Adiabene and the Kirkuk region.

Information on the subsequent invasion of Babylonia is extremely slight. It is not even certain whether there was one Roman army or two. One would expect that after occupying Adiabene and establishing garrisons as far south as Libbana (near Nineveh), Trajan would have struck direct at Ctesiphon along the 'Alexander road' through Gaugamela and Arbela. But Dio (68.26.42) says: 'After this [conquest of Adiabene] they advanced as far as Babylon itself. . . quite free from molestation', which would indicate a switch from the Tigris to the Euphrates.10 And later (68.28) Trajan's final advance on Ctesiphon seems to have been across the Tigris from the direction of Babylon. It appears fairly certain that there was a Roman army on the Euphrates (Dio 68.28; Amm. Marc. 23.5.17); a triumphal arch was erected by legion III Cyrenaica at Duro-Europus (Thapsacus), though of doubtful date,11 and the emperor reviewed his troops at Ozogardana further south, near modern Hit (Amm. Marc. 24.2.3).12 If a second army was used for the previous campagin on the Tigris, Trajan may have left it there to occupy the conquered territory, reserving his main force for a thrust down the Euphrates valley, which again he commanded in person. Another winter may have intervened before this final operation.

Trajan would have advanced down the east bank of the Euphrates, the route followed by Severus in A.D. 195 and by Julian in 363, who both like Trajan approached the river from the north-east. Severus started from Nicephorium (Callinicum) and Julian further south on the junction with the Chaboras river, but Trajan's starting point is uncertain. Trajan, whose Danube fleet had played an essential role in the rapid transport of materials and men in the Dacian wars, was here again accompanied by a large fleet, which carried stores, siege equipment, spare cavalry mounts and pack animals, etc., and could be used for blockading towns and fortresses, bridging streams and ultimately in the siege of Ctesiphon, if a way could be found to the Tigris through the intricate network of irrigation canals in lower Babylonia. The army marched along the narrow cultivated strip beside the river heavier going, perhaps, for foot-soldiers, but more difficult for horse-archers. Greek peltasts had got the better of the Persian cavalry on this same soft ground at the battle of Cunaxa (Xen., Anab. 1.10.7). It might also provide some shade on a march probably made in late spring, for the Euphrates overflowed its banks and inundated the countryside in May after the spring thaws, presumably making the ground impassable for several weeks.13

As we noted, Trajan seems to have reached Babylon without meeting any opposition at all. Julian later was forced to take elaborate precautions against ambuscades, stragglers from his army were cut off by enemy horsemen, and he had to fight a major engagement before reducing the fortified city of Perisabora near the Royal Canal (Amm. Marc. 24.1.13,16; 24.2.5-21). Certainly Julian had incurred the hatred of the local population by letting his troops, who were by no means as well disciplined as Trajan's,14 loot, burn and murder to their hearts' content. But the main reason for Parthian inactivity in Trajan's time was the absence of a stable central organization. Two rival monarchs, Chosroes and Vologeses, had contested the throne for several years past,15 and the only resistance to the Romans came from loyal vassals such as Mebarsapes of Adiabene. In some provinces, e.g. Characene, Trajan was actually welcomed.

The Royal Canal near Babylon, which had once connected the Euphrates with the Tigris (Herod. 1.193), had apparently fallen into disrepair. Dio states (68.28.1-2) that Trajan planned to reopen it in order to use his fleet to cross the river at Ctesiphon; but finding that the then level of the Tigris (which does not rise till July) was much higher than that of the Euphrates, and fearing flooding, he hauled the ships overland. Ammianus Marcellinus (24.6.1) confirms this, and indicates that Trajan did actually carry out some excavation, continued later by Severus. At all events, Trajan crossed the river, and the city fell without a struggle.

The war was concluded in the summer; in autumn Trajan sailed to the Persian Gulf and received the submission of Characene, returning to winter in Babylon. But meanwhile rebellion had flared up in the occupied areas, even as close as Seleucia, and was aided by a large-scale invasion of Armenia and Mesopotamia from Media. This was to be expected. Chosroes was still at large, and the nerve centre of the Parthian empire was not Babylonia, despite its fertility, but Media. Indeed, Ctesiphon was near the western frontier of the empire, as far from its eastern end as from Rome. Peace was restored by a combination of force and diplomacy. Trajan split the notoriously disunited Arsacids by installing Chosroes' son Parthamasapates as client-king of Babylonia in return for his help in defeating the invaders. He bought off another Arsacid prince by granting him part of Armenia. Lusius put down the revolt in Mesopotamia with a heavy hand. But the new Roman frontier, probably extending from near Nineveh to Singara and thence down the Chaboras river to the Euphrates,16 followed no natural line of defence and required a large permanent garrison. Trajan died in 117, and his successor Hadrian's first act was to withdraw from all the conquered territories except Armenia, which was transferred to Parthamasapates. 'Thus,' as Dio says (68.33.1), 'the Romans had undergone their hardships and dangers all for naught.' Later emperors penetrated even further; but not for long.

The respective contemporaries of Trajan and Antony would have agreed that Trajan's expedition had been an outstanding success, while Antony was lucky to get out alive. But the two expeditions are hardly comparable. They were initiated under different circumstances, with different military resources, against a different enemy, by a different route, with a different strategy. Whether different tactics were used we do not know, since little is known of those used by Trajan against the Parthians. Paradoxically, the results achieved by the two expeditions were not so different. After the dust settled, neither seems to have greatly altered the status quo.

Trajan started with considerable advantages over Antony. His position was unquestioned, he was immensely popular and he could command what resources he chose money, men, equipment. We do not know what forces he took to Parthia other than infantry Dio mentions cavalry only twice (68.19.4; 68.31.3) and other arms not at all but given his expertise in the conduct and logistics of large-scale warfare, we may be sure that he had no reason to fear the Parthians in the field under any conditions. The Roman army was a much more formidable instrument than 150 years earlier: better armed, better equipped, with cataphracts and archers to match those of the Parthians, and a wider range of siege equipment and auxiliary services, especially engineers. Trajan's Column is described as a 'unique monument to the excellence and versatility of the Roman army at its peak'. If Trajan and his successors never succeeded in taking over the Parthian empire, the main reason was simply an insufficiency of troops to hold its vast tracts of desert and semi-desert.

Trajan too was experienced in amphibious warfare from the Dacian wars, and had a fleet. Not that Antony needed one, or would necessarily have changed his plan of campaign if he had had one, but it did enlarge Trajan's range of strategies as well as providing rapid transport for heavy equipment and stores, and consequently a faster and less arduous march for the army.

Did Antony, then, choose the wrong strategy? Would he have fared better if he had elected to advance down the Euphrates valley, with or without a fleet? But this option was not really open to Antony. He would have had to fight his way across the river at Zeugma, to begin with. That would have been difficult without a fleet. And the Parthians were waiting on the far side, ready to harass him every foot of the way, whichever route he took. Even if he had had a fleet, he knew little of waterborne operations, as was seen at Actium, and would have inevitably have run into trouble. None of these objections applied to Trajan. The situation, as Antony would have seen it, was that Crassus had met disaster by rejecting Armenian help and making a direct strike south. The only alternative in sight was to make a surprise attack on Media in the hope of capturing Phraaspa as a base, which would bring him within striking distance of either Ctesiphon or Ecbatana. It would no doubt have been wiser to annex Armenia and Mesopotamia first, as Trajan did, but Antony could hardly do that if he wanted to use Armenian cavalry. Moreover, he had not the time to spare that Trajan had. I think it likely that Caesar, faced with the same situation, would have made the same decision.

Antony himself blamed Artavasdes17 for the failure of his enterprise (Plut., Ant. 50). This was perhaps unjust. His real ill fortune lay in finding the Parthian monarch firmly in the saddle and able to hold his forces together, despite Antony's efforts to create dissension. If, like Trajan, he had invaded a kingdom rent by internal strife, with nobody at all to organize a defence, he almost certainly would have succeeded. As it was, he failed more by bad luck than by bad judgment.

To appreciate Trajan's achievement, it is more instructive to compare his campaign with another equally distant from it in time but more closely resembling it in detail that of Julian against Sassanid Persia in A.D. 363, for which we have the vivid eye-witness account of Ammianus Marcellinus (books 24 and 25). Their armies were of similar size, both took the same route, and there was possibly not much to choose between them as military commanders; but while Julian was hopelessly romantic and superstitious,18 lending a ready ear to soothsayers and argumentative philosophers (Amm. Marc. 23.5.10-11), Trajan was a shrewd and able negotiator who was always willing to compromise. He managed to retain a large part of the territory he had won, while Julian threw away his life, and his successor Jovian weakly returned all the Roman gains. Even Gibbon (Decline, ch. 24), who admired Julian, could not understand why he gave up the siege of Ctesiphon and withdrew northwards (Amm. Mar. 24.7.1) refusing even to discuss peace terms with the Persian king (Socrates, Hist. Eccl. 3.19-21)19 after a brilliant night attack across the Tigris and the defeat of a large Persian army had brought the Romans within the very gates of the city (Amm. Marc. 24.6.13). Unfavourable omens from the following morning's sacrifice (Amm. Marc. 24.6.17) may have deterred the superstitious emperor!20 One cannot imagine Antony, Trajan or the hard-nosed Severus throwing away the fruits of victory to go seeking the bubble reputation, Alexander-like, somewhere in the desert. His burning of the ships is more comprehensible. They could not be rowed or sailed upstream, and he would not leave them to the enemy.



1     Plut., Ant. 37.6; Livy, Perioch. 130; but not Dio or other ancient writers. Cf. earlier moderns: G. Rawlinson, The Five Great Monarchies of the Ancient Eastern World (London 1871) vol. 2, 208. Th. Mommsen, The Provinces of the Roman Empire, from Caesar to Diocletian, trans. W.P. Dickson (London 1909) vol. 2, 27-28. P. Sykes, History of Persia, 3rd ed. (New York 1969) vol. 1, 360. The more recent criticisms of Antony's strategy by A.N. Sherwin-White, Roman Foreign Policy in the East, 168 B.C. to A.D. 1 (London 1984) 315ff., likewise appear to have little substance.

2     See R.P. Longden, "Notes on the Parthian Campaign of Trajan", JRS 21 (1931) 29ff., for discussion in relation to von Stauffenberg's commentary on Malalas.

3     Longden, ibid., 12, n.3.

4     Longden, ibid., 4.

5     N.C. Debevoise, History of Parthia (New York 1968) 229, n.80.

6     A. Garzetti, From Tiberius to the Antonines: a History of the Roman Empire AD 14-192, trans. J.R. Foster (London 1974) 680.

7     For discussion, and details of the rival schedules, see Longden, ibid., 2ff.; Garzetti, ibid., 366-370, 680.

8     For details of legions see Debevoise, ibid., 220-221; Garzetti, ibid., 366.

9     Longden, ibid., 10, n.4.

10     He could not have taken a fleet down the Tigris to Ctesiphon, as the river, which does not flood till July, would still be too low.

11     Garzetti, ibid., 368.

12     Debevoise, ibid., 232, n.95.

13     According to Dio (40.20.3), Crassus had intended in 53 B.C. to advance from Zeugma to Seleucia 'along the banks of the Euphrates and on its stream, so as to reach there safely with his army and provisions.' Dio says he was dissuaded by a treacherous ally who said it 'would take a long time', but his reason may have been the spring floods. The bridge at Zeugma collapsed while Crassus' army was crossing it (Dio 40.18.5), which would suggest flood waters, and he might well have decided not to wait until they receded. Hence the disaster at Carrhae, which occurred on 9 June (Ovid, Fasti 6.465-9).

14     W.E. Kaegi, Jr., "Domestic Military Problems of Julian the Apostate", Polychordia, Festschr. F. Dolger (Amsterdam 1967) 247-264.

15     For more details see Longden, ibid., 12-13; Debevoise, ibid., 228-229. From numismatic evidence it appears that the two alternately controlled the Ctesiphon-Seleucia mint from 105/6 onwards.

16     M.A. Stein, Geog. J. (1938) 62ff., quoted by M. Cary, History of Rome (London 1954) ch. 38, n.5.

17     Artavasdes with his heavy cavalry deserted the siege train when it was attacked by the Parthians.

18     Montaigne (Essays, bk. ii, ch. 19) says 'he was besotted with the art of divination'. Cf. Amm. Marc. 22.1.1, 23.3.3, 24.8.4, 25.23-4.

19     Socrates attributes his refusal to 'having put his trust in certain divinations which his associate the philosopher Maximus produced, and dreaming of assuming and even surpassing the glory of Alexander'. There is apparently 'an extensive lacuna' in Ammianus' MS. at this point (Rolfe, 467, n.1 in Loeb edition).

20     He may have feared being trapped between Ctesiphon and the Persian main army if he besieged the city. But if so, why go there at all? See R. Browning, The Emperor Julian (London 1974) 207-208.