Onward to Ctesiphon!


Emperor Julian is killed during his failed invasion of Sassanian Persia in June 26, 363 AD. Above is a recreation of Sassanian Persia’s elite cavalry, the Savaran, as they would have appeared during Julian’s failed invasion. Note the heavily armored Sassanian elite guardsman (Pushtighban) whose lance has pierced a Roman infantryman. Further right is a Savaran officer whose sword is drawn in what is now known as the “Italian grip” but Sassanian in origin. To the far right can be seen a Zoroastrian or Mithraist Magus brandishing a Sassanian era symbol. Also of interest are the armored elephants in the background. Armored elephants were especially prized as their cabs afforded very high elevation over the battlefield, which was ideal for Sassanian archery.


The army marched on past a palace built in Roman style and the Persian king’s game preserve, whose wild inhabitants the Roman cavalry slaughtered. Julian rested his army for two days, while the emperor himself detoured to see the site of a city destroyed by the emperor Carus (ad 282–83) during his march upon Ctesiphon. Such tourism was exactly appropriate for an emperor with so close a relationship to the past; in Julian’s own account of his trip from Antioch toward the Euphrates to begin the campaign (which survives in a letter) the emperor showed an acute interest in the antiquities he passed. But this visit in Babylonia turned grisly: the ruined city was a place of execution, and Julian found crucified there, as a warning to others, the kinsmen of the Persian officer who had surrendered Pirisabora to the Romans.

Persian marauders began to harass the Romans, killing baggage animals and foragers: in driving them off, Julian was nearly shot down by an engine on the wall of a Persian fortress, which he then spent two days besieging. Julian directed the attack from the front, placing himself once again in peril, and the fortress was taken. But raiders kept ambushing the Roman troops. The Roman camp was more carefully fortified, and Julian gave his army another day of rest.

The fleet of Roman supply boats had been accompanying the army up a canal that had emerged from the Euphrates and now ran parallel to the Tigris. This canal was connected to the Tigris above Ctesiphon by a dry cross-canal, which Julian had read about in his books: Trajan had tried to use it, and Severus had used it. The Roman army dug out the blockage, and the Roman boats sailed into the Tigris. But Ctesiphon was on the other side of the river, and the Persians planned to contest the crossing. It may have been now that Julian held his Alexander-like games, in order to distract the Persians from noticing his preparations for crossing. First Julian sent over five ships, which the Persians set on fire. Julian deftly lied to his army that the flames were the agreed-upon signal that a lodgment had been gained on the other side and sent the rest of the fleet over: after hard fighting, the bank was captured.

Now that Julian had defeated all the natural obstacles that separated him from Ctesiphon, the Persians were finally obliged to meet him in battle. The relative passivity of the Persian army during Julian’s nearly two-month invasion of their territory is one of the insoluble mysteries of the campaign: even now King Sapor himself had not arrived to command the force defending his winter capital. Perhaps Julian’s feint toward the Tigris at the beginning of the war had been successful; perhaps the force Julian had left on the Tigris had distracted the king; or perhaps the Persian royal army took a long time to assemble: Julian had marched out early in the year.

Here before Ctesiphon, finally, were the formidable Persian cataphracts, “blinding the eyes of onlookers with their gleam.” Here were the Persian infantry, with their long shields. Here were the Persian elephants, “looking like hills on the march, the motion of their gigantic bodies threatening doom to any who came near.” Julian arrayed his Roman army in three lines, with the weakest troops in the second: this, as Ammianus points out, was the well-known “Homeric array,” so called after Nestor’s disposition of his troops in the Iliad just before he gave the order that his contingent should fight as a mass rather than as individuals. This disposition had been alluded to in Classical Greece, and Pyrrhus had picked it up from Homer and used it, although he interpreted it as a matter of left, right, and center rather than front, middle, and back. (So well known was this ruse that the term was used in rhetoric to denote a speech with a weak middle section.) Julian, the better philologist, corrected Pyrrhus: by Julian’s day it was important to get such things right.

The armies approached each other, javelins were thrown, and a cloud of dust swirled up. Then, with battle cries and a blast of trumpets, the lines met in hand-to-hand combat. The Persians began to waver, and wavering became flight: the Romans pursued the enemy up to the very walls of Ctesiphon, but a Roman officer managed to restrain them from pressing into the gates behind the Persians. Twenty-five hundred Persians fell, Ammianus says, and only seventy Romans. As after the taking of Maozamalcha, Julian rewarded the brave with crowns, again reviving a practice a long time dead.

At last, Julian stood beneath the walls of Ctesiphon, his objective, the walls that Trajan, Avidius Cassius, Severus, and Carus had pierced before him. All that remained was to capture the great city. But now Julian went into council with his officers, and that council produced an astonishing decision: Ctesiphon would not be attacked. The council was urged that Ctesiphon was “impregnable by its site alone” and that the army of King Sapor was approaching. The force of this logic, or any logic to the same end, is difficult to comprehend: four Roman marshals had found Ctesiphon perfectly pregnable before. And why, other than to besiege Ctesiphon, had Julian carried such formidable siege equipment in his fleet—such an equipage that none of his previous assaults on strong places had taken him longer than four days? And can Julian really have expected to capture Ctesiphon without having to fight the King’s army? This baffling decision was, moreover, related to another, equally perplexing. Julian did not accept the offers of a treaty of peace which may have been made by Sapor at this time. But the Roman army was neither to stay before Ctesiphon and await Sapor, nor retreat into Roman territory up the convenient Tigris, nor withdraw by way of the Euphrates. Instead, it was to strike north away from the river into the interior. This, in turn, occasioned another fateful decision: the fleet of boats that carried the supplies and the siege machinery was to be burned.

The orator Libanius, Julian’s panegyrist, was the first to try to justify this decision. Had the boats not been burned, he said, they would have fallen into the hands of the enemy, or it would have taken too many men to pull them back into Roman territory against the current. These and other justifications have been offered by modern students as well. But Libanius admits that the decision was denounced at the time; Ammianus thought it dire; the army in Babylonia was horrified; and indeed the emperor came to regret it, and the order was soon countermanded —but too late to save the burning boats.

The twin decisions not to attack Ctesiphon and to march north can be explained if they are combined with a detail provided by Libanius: Julian was moving to seek out the army of Sapor, which was in fact marching from the north. There was still a considerable Persian force in Ctesiphon, and Julian no doubt preferred to fight the two Persian armies separately, rather than united. Perhaps he feared being attacked from the rear by Sapor while besieging the Persian capital. It may well be that Julian expected to return to attack Ctesiphon once Sapor had been driven away. Or perhaps he expected Ctesiphon to yield without a fight when the Great King was vanquished—as some of the forts Julian had bypassed had promised to do. But these considerations merely make the burning of the boats more perplexing: if Julian had intended to return to the Tigris, why did he not simply protect the boats with a garrison or tow them out of harm’s way, perhaps to one of the strong places he had captured on his march?

In burning the boats, Julian claimed a place for himself in a high tradition. Alexander had dismissed his fleet when he turned inland in Asia Minor, an act so famous that it had received the compliment of explicit imitation by Agathocles of Sicily, when he landed in North Africa to fight the Carthaginians (310 bc). It was Agathocles who added the refinement that the fleet should be burned to deny his soldiers any hope of retreat, and Agathocles’ act entered the tradition of stratagems, and perhaps the literary tradition as well, where it may have encouraged Vergil to give space in the Aeneid to the old legend that Aeneas’s wanderings were delayed (or halted) by the wayfarers’ women burning the Trojan ships. In burning his boats, Julian was behaving as he had at Pirisabora and all through his campaign: emulating the past. And this instance of emulation was not a fantasy of Ammianus’s: dumbfounded at the folly of the decision, the historian failed to make the connection to Alexander and Agathocles.

Upon leaving the Tigris, Julian’s expedition almost immediately came to grief. King Sapor’s army was in the area—Persian cavalry harassed the Romans, and Persian cataphracts were seen—but the Persian host would not fight the Romans, preferring to starve them. The Persians fired the crops along the Roman line of march, and so great was the conflagration that the Romans had to camp while it died down. It was now, in order to reassure his troops, that Julian employed the venerable stratagem of presenting starving enemy captives to them naked, to show how feeble the enemy were. In vain: the army was becoming unmanageable, and the soldiers took advantage of the assembly to cry out that they should return the way they had come. There was nothing to eat back there, the emperor and his officers replied: all that territory had been wasted. The route of march was left up to the gods, whose opinion was solicited by sacrifice—but the gods were asked only about routes of retreat (and refused to endorse any): Sapor’s unwillingness to fight and the army’s perilous supply situation had now doomed Julian’s expedition. The army turned west, crossed the Dyala (a tributary of the Tigris), and trudged back toward the great river. The Persians had burnt all the crops. A sand-storm came up, and Julian sent his army into camp to prevent a panic (was the King’s host approaching?). The Roman army was now thoroughly unnerved.

Dawn, and with it so large a force of cataphracts that the King himself must be nearby. But although gathering in larger bodies, the Persians would still not commit to a pitched battle, merely to demonstrations and harassment and delaying actions to run down the Roman supplies. In a skirmish, Macameus, a Roman unit commander, killed four foes; then the Persians swarmed him and he was mortally wounded. His brother Maurus avenged him by slaying the man who had struck him down, and although shot through the shoulder with an arrow, terrified those who approached, and got Macameus’s body away with the last traces of life still in it. This episode of frontline fighting by senior officers was one of many: Julian placed himself repeatedly in peril, and the unit commander was the second man out of the tunnel at Maozamalcha. More broadly, it appears that in the fourth century heroic leadership had become usual among higher officers to a degree never before seen in the Roman tradition: what had been a (possibly eccentric) choice in the time of Titus was now common behavior. While in Julius Caesar losses were listed in men and centurions, in Ammianus they are listed in men and tribunes—who commanded regiments at least six times as large as the centuries of Caesar’s centurions. Where an estimate is possible, the tribunes of Julian’s army, like the centurions of Caesar’s, had a much higher chance of being killed than the soldiers they led. The emperor Constantius himself practiced with all sorts of weapons and was especially good at infantry hand-to-hand combat drill. This habit of heroic command was to last into the sixth century, where it flowered again into a tradition of combat by challenge between the lines before battle. One of Justinian’s generals, although over seventy and fat, was the first up the ladder during an attack on a town: he fell from the top of the ladder and crashed to the ground. The enemy all turned their missiles upon him, and his guards had to cover him over with their shields and drag him away by the foot. But when he was towed out of range and had been set upon his feet again, the tough old warrior went right back up the ladder.

On the day Macameus and Maurus proved themselves heroes, the Romans drove off the Persian attacks. Soon the Romans arrived at the Tigris but found the bridge burnt by the Persians. Since nearly all their bridging equipment had been destroyed with the fleet, the Romans would have to make their way home up the east bank of the river. But the Romans did find an unplundered estate, and there rested for two days. On the first day of their renewed march up the river the Persians attacked the rear of the marching column. In the fighting a Persian satrap was killed, and the Roman who killed him stripped his body and presented the arms to Julian for a reward: another glimpse of the revival (or survival) of the rituals of single combat. The Romans marched on through the ravaged landscape, their supplies steadily dwindling: now a Persian force tried to block their passage with cataphracts, horse archers, and elephants. Julian drew up his army in a crescent, with wings curving forward, and ordered a fast advance to deprive the Persians of a chance to use their arrows. Ammianus does not make the connection, but the combination of these two tactics suggests that Julian was emulating Miltiades’ famous plan for the Athenians at the battle of Marathon in 490 bc. The Persians were driven back, but the Romans did not hold the field and returned to their tents after the battle. Such was the hurt of the Romans that they had to agree to a three-day truce, which they could ill afford because of their scarcity of food: the Roman beasts were now so worn out that they could not carry even what food the Romans had left.

Julian was now seeing ominous visions by night, and a falling star added to his foreboding. When the Romans set out after the truce expired, the Persians pressed them hard from ambuscade. The rear guard was attacked: Julian flew to its aid without his armor. Then report came that the van was threatened as well. Now the center was assailed, and from somewhere came flying a spear and took Julian through the body. The emperor was dying. But he would die as he had lived, forever racing with the admired ancients. As he lay in his tent he held a discussion about the immortality of the soul, as dying men with philosophical pretensions had done for centuries, on the model of the immortal Socrates. By midnight, Julian was dead.

The army of the Romans now lacked both food and a leader. The second deficiency, at least, could be made good: Jovian, an officer of the bodyguard, was hurriedly made emperor by the soldiers. Julian’s death and the succession were soon reported to Sapor by a deserter, and the King bade the harrying of the Romans continue. The army reeled north again, along the bank of the river, attacked by elephants, cavalry, and missiles and followed by taunts that Julian had been murdered by his own men. Three tribunes were killed fighting bravely. Now the Persians were becoming bold enough to attack the Roman marching camps. Finally, half a month after the beginning of the retreat, the Romans could go no further: for four days they tried to march on but turned back in the face of Persian resistance. The soldiers demanded to be allowed to cross the Tigris to the safety of the west bank, and some Gauls and Germans, powerful swimmers, managed to get over by forging through the torrents, but the rest waited vainly, hungrily, for rafts to be contrived from the hides of the last pack animals.

It was at this point that Sapor offered Jovian an unequal peace, requiring him to surrender Roman land and border fortresses, including the great Roman bastion of Nisibis, which the King had attacked three times in vain. To extract the army, which starved even as the negotiations dragged on, and to escape the East before a rival to his power might arise, Jovian was compelled to accept Sapor’s terms. Now the Romans were no longer attacked by the Persian army, but many soldiers were still lost in the march back to Roman territory—starved, drowned, or taken as slaves by marauding Persians and Saracens. And so it was the remnants of Julian’s expedition came home.