-
We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
Sorry to hear, I hope it's all worked out. Enjoy the holiday and thank you for your writings (in this and other threads, particularly about the Soviets)
.png){kind=avatar}
I miss the best thread in this forum... 
Anyone knows some thread-revival invocations in Ancient Persian?
Anyone knows some thread-revival invocations in Ancient Persian?
8.6 THE ROMAN COUNTERATTACK OF 360 CE.
8.6 THE ROMAN COUNTERATTACK OF 360 CE.
The Sasanian summer campaigns of 359 and 360 CE had been a huge success; for the first time after decades of failed attempts Šābuhr II had managed to design and implement an offensive strategy that could realistically achieve his ultimate goal of recovering the Roman province of Mesopotamia. Instead of butting his head unfruitfully against the great fortified city of Nisibis, this time the Šahān Šāh chose a slower, indirect approach consisting of cracking open first the outer, more vulnerable ring of Roman border fortresses that shielded Nisibis from the south, east and north, and in two years he had destroyed two of these fortresses (Amida and Singara) and taken and garrisoned another one (Bezabde). Now, the Sasanian king could finally organize a proper attack against Nisibis without the constraints of time imposed by having major Roman garrisons on his back.
Obviously, Constantius II and his ministers and generals were perfectly conscious that if they did not react in any way, the next year the Sasanian king would launch a direct attack against Nisibis and then the only way to stop him would be by risking an open battle on the flat plains of northern Mesopotamia, a terrain that historically had usually favored the cavalry-heavy Iranian armies. Of special significance in this context was the decision by Šābuhr II of installing a Sasanian garrison at Bezabde and repairing its walls. This fortress controlled one of the major crossing points across the Tigris, and so one of the major routes of invasion. Given that the Sasanian king had simply razed to the ground all the other captured Roman fortresses, this decision hinted at his intention of using Bezabde as the main launching point for next year’s campaign.
The catastrophic outcome of the Sasanian invasion of 360 CE also presented another (and possibly, in Constantius II’s mind, more serious) danger: it damaged seriously the reputation of Constantius II as sole augustus of the Roman empire while his cousin Julian was leading a major rebellion in the West. And here lies the heart of the question, because the movements of Constantius II and his praesentalis army during the campaign of 360 CE depend on the chronology that we choose to follow, which is quite confused in Ammianus’ account. In turn, this is tied closely to Constantius II’s reactions to Julian’s moves and his crucial decision to prioritize the armed response against the Sasanians while trying to deal with his rebel cousin by political and diplomatic measures, a decision almost unparalleled in the history of the Roman empire. Where were Constantius II and his army while Šābuhr II took these two Roman fortresses?
Constantinople was the main residence of Constantius II. View of the area occupied by the Hippodrome and the Great Palace, behind it, with terraced gardens that descended towards the Bosporus. Restitution by the architect and archaeologist Jean-Claude Golvin.
In a long and convoluted narrative that I will not be posting here, Ammianus tells us that Constantius II spent the winter in Constantinople reinforcing his army, training it and gathering supplies and animals for a major campaign, while Julian consummated his usurpation in Gaul and sent envoys to his cousin to inform him officially of the fait accompli and that a solar eclipse happened in the East (together with a long and confused excursus about eclipses). Notice that in his account Ammianus seems to imply that all this happened while Constantius II was still in Constantinople.
Ammianus doesn’t return to the details of Constantius II’s war against the Sasanians until this passage (Rerum Gestarum Libri XXI, XX, 11.1-5):
Securing his northern flank by ensuring the Armenian cooperation was a key precaution for Constantius II that shows once more that he was a sound strategist. And he did so with bribes or subsidies (that Ammianus diplomatically describes as “presents”. From Ammianus’ account, it can be surmised that Constantius II met the Armenian king in Cappadocia, where the Roman augustus had advanced with his army from Constantinople. The meeting could have taken place at Caesarea Mazaca, the provincial capital, or maybe in one of the legionary fortresses in the border. The former possibility seems to make sense considering that Julian’s letters reached Constantius II at Caesarea according to another passage of Ammianus, but the latter possibility also makes sense considering that the route that Constantius II took to reach Mesopotamia was not the usual one, via Tyana in Cappadocia, Tarsus in Cilicia and then Antioch in Syria. Instead, according to Ammianus he followed the upper course of the Euphrates via Melitene (modern Malatya in Turkey), Lacotena, and Samosata, and then crossed the Euphrates (there was a crossing point at Samosata) and reached Edessa which like in previous campaigns was to act as the Roman main base in this campaign.
Route of Constantius II’s army from Constantinople to Edessa (in red). I’ve underlined in yellow the main military road from the Bosporus to Antioch. Map from The Digital Atlas of the Roman Empire, by the University of Lund.
Again, following Ammianus’ account, Constantius II seems to have remained for a long time in Edessa, and he only moved “after the autumn equinox”, that is on 21 September at the very earliest, which is quite late in the campaign season. And then he moved to the destroyed city of Amida in the upper course of the Tigris, which only makes sense if his intention from the start was to recover Bezabde. In other words: at this time, Šābuhr II had already concluded his destructive invasion and had been able to retreat back into Ērānšahr with impunity, while Constantius II reached Mesopotamia and remained in Edessa.
The Roman army then marched directly from Amida to Edessa, following the course of the upper Tigris valley, and reached the city (Rerum Gestarum Libri XXI, XX, 11.6-7):
Constantius II inspected personally the state of the fortress and realized that not only had the Sasanian defenders repaired the breach caused by their siege ram, but they had also repaired any other weathered sections of the wall. In short: now the fortifications of Bezabde were even stronger than before the Sasanian conquest. As Constantius II was a seasoned warrior, upon seeing this he opened talks with the defenders, trying to arrange a negotiated surrender with them: either abandon the fortress and return to Ērānšahr unmolested or enter Roman service and be richly rewarded for it (in other words; he offered them bribes). Unsurprisingly, the defenders refused; as we will see, according to Ammianus’ account Šābuhr II dealt in an extremely harsh ways with traitors and their families. And considering the importance of Bezabde and the spirited defense that the garrison put up against the Roman siege, it’s evident that the Šahān Šāh left in there a garrison of elite troops, highly skilled and motivated.
British reenactor Nadeem Ahmad dressed in the full garb of a Sasanian Šahān Šāh or a high-ranking spāhbed. The battle axe was a personal combat weapon typical of eastern Iran and Central Asia, although it seems to have been less used in western Iran.
The Romans then made their preparations and launched their assault, but as we will see in Ammianus’ account, the Romans fared no better than the Sasanians when assaulting a strong fortress with a determined and skilled garrison within it (Rerum Gestarum Libri XXI, XX, 11.8-10):
It's almost as if we were reading the Sasanian sieges of Nisibis, Amida or Singara all over again. It’s interesting to see how, according to Ammianus’ description, the first tactic that the Roman soldiers employed was the classical testudo; clearly this formation had not gone out of fashion even if the Roman had long since abandoned the rectangular scutum in favor of lighter oval shield. In reliefs in Trajan’s Column in Rome more than two centuries earlier, legionaries can be seen doing exactly this while attacking a Dacian fort. It failed in face of what seems to have been a veritable deluge of projectiles from the Sasanian defenders, and when they tried to use wicket mantlets, the result was the same, as the defenders simply dropped heavier projectiles upon the attackers, tearing large gaps in the mantlets. What seems to me is that here the Romans pressed forward an improvised attack (probably due to the lateness of the season) to take to resolve things quickly, and they discovered the hard way that the Sasanians could defend a fortress as well as they did. So, the first three days of the assault ended in a Roman defeat, possibly with serious losses. Let’s see how the Romans renewed their assault (Rerum Gestarum Libri XXI, XX, 11.11-15):
The tides have turned here. Now it’s the Romans who have to resort to all sorts of elaborate siege techniques and absorb huge losses in their ranks while the defenders manage to counteract each and every one of their efforts. This is a long passage, that Ammianus managed to garble it considerably (either him or the scholars who have restored the text of the only mutilated manuscript that was rescued in the XV century and published in Florence), so careful reading is necessary. Also, some interesting bits of information appear in it.
Remains of the walls of the (supposed) site of ancient Bezabde, near Cizre in southeastern Turkey. The river in the background is the Tigris.
Following the standard practice of the time, the Romans protected this huge ram and other “smaller rams” so that the Sasanian incendiary devices could not set them on fire (like the Sasanians had done with their siege towers at Nisibis and Amida), but at the hour of truth, the defenders were able to use a “subtle device” with ropes that immobilized the head of the ram, while its crew lay exposed to a deluge of projectiles from the walls, Ammianus’ explanation is frustratingly vague, but in ancient and medieval times a common technique to counter rams was to grip its head with ropes or chains and then lift them upwards from the walls, rendering them useless. The siege now entered its bloodiest and most hardly fought period (Rerum Gestarum Libri XXI, XX, 11.16-19):
As usual with Ammianus, the text is full of dramatism; he was a soldier and had fought in similar engagements and knew well the dangers and fear to which the fighters were exposed in this sort of encounters. As we can see, the defenders did not hesitate to sally twice from the walls and engage the Roman legionaries in hand-to-hand combat. Let’s remember that this was Constantius II’s praesentalis army, the crème de la crème of the Roman mobile armies, so the fact that the Sasanian defenders were able, or even willing to come to grips with them is an indication that the Sasanians, or at least their elite forces, were perfectly able to fight in close fighting against the dreaded Roman infantry if necessary. Notice also that the Sasanians seem to have a considerable degree of familiarity with incendiary devices and employed such techniques with a considerable effectiveness against the Roman besiegers. As we will soon see, they kept using fire against the Romans during this siege with success. The second sally seems to have left the rams badly damaged, even the giant one, but the Roman army just redoubled its efforts against Bezabde. Clearly, Constantius II considered the control of Bezabde a vital matter (or he was desperate for a victory, given the political situation within the empire), and his army’s morale was high enough to endure it with its will to fight intact, a tribute to Constantius II’s thorough preparations and the respect his soldiers had for their augustus. Let’s continue to the next point in Ammianus’ tale (Rerum Gestarum Libri XXI, XX, 11.20-23):
Siege ramps could be built with earth (the surest but slower method) or with bundles of wood, as Ammianus says was the case here. And you can see what the risks were in this case: they were vulnerable to the incendiary attacks of the enemy. Once more, the Sasanians had made a huge gamble with a risky sally, and it had ended in success. After this, the siege petered out, as Constantius II decided to try to starve the garrison by hunger, which is quite puzzling given the lateness of the season (Rerum Gestarum Libri XXI, XX, 11.24-32):
In short: the last assaults were followed by the start of Fall rains and both fear of the discontent and mutiny among his troops if he carried on with the siege, and of the events happening in the West, forced him to lift the siege and retire back to Syria, where he set up his winter quarters in Antioch. The Roman siege of Bezabde had failed, and the road to Nisibis lay open for Šābuhr II.
Gold solidus of Constantius II. On the obverse, CONSTANTIVS AVG(ustus). On the reverse, GLORIA ROMANORVM. Mint of Antioch.
The Sasanian summer campaigns of 359 and 360 CE had been a huge success; for the first time after decades of failed attempts Šābuhr II had managed to design and implement an offensive strategy that could realistically achieve his ultimate goal of recovering the Roman province of Mesopotamia. Instead of butting his head unfruitfully against the great fortified city of Nisibis, this time the Šahān Šāh chose a slower, indirect approach consisting of cracking open first the outer, more vulnerable ring of Roman border fortresses that shielded Nisibis from the south, east and north, and in two years he had destroyed two of these fortresses (Amida and Singara) and taken and garrisoned another one (Bezabde). Now, the Sasanian king could finally organize a proper attack against Nisibis without the constraints of time imposed by having major Roman garrisons on his back.
Obviously, Constantius II and his ministers and generals were perfectly conscious that if they did not react in any way, the next year the Sasanian king would launch a direct attack against Nisibis and then the only way to stop him would be by risking an open battle on the flat plains of northern Mesopotamia, a terrain that historically had usually favored the cavalry-heavy Iranian armies. Of special significance in this context was the decision by Šābuhr II of installing a Sasanian garrison at Bezabde and repairing its walls. This fortress controlled one of the major crossing points across the Tigris, and so one of the major routes of invasion. Given that the Sasanian king had simply razed to the ground all the other captured Roman fortresses, this decision hinted at his intention of using Bezabde as the main launching point for next year’s campaign.
The catastrophic outcome of the Sasanian invasion of 360 CE also presented another (and possibly, in Constantius II’s mind, more serious) danger: it damaged seriously the reputation of Constantius II as sole augustus of the Roman empire while his cousin Julian was leading a major rebellion in the West. And here lies the heart of the question, because the movements of Constantius II and his praesentalis army during the campaign of 360 CE depend on the chronology that we choose to follow, which is quite confused in Ammianus’ account. In turn, this is tied closely to Constantius II’s reactions to Julian’s moves and his crucial decision to prioritize the armed response against the Sasanians while trying to deal with his rebel cousin by political and diplomatic measures, a decision almost unparalleled in the history of the Roman empire. Where were Constantius II and his army while Šābuhr II took these two Roman fortresses?
Constantinople was the main residence of Constantius II. View of the area occupied by the Hippodrome and the Great Palace, behind it, with terraced gardens that descended towards the Bosporus. Restitution by the architect and archaeologist Jean-Claude Golvin.
In a long and convoluted narrative that I will not be posting here, Ammianus tells us that Constantius II spent the winter in Constantinople reinforcing his army, training it and gathering supplies and animals for a major campaign, while Julian consummated his usurpation in Gaul and sent envoys to his cousin to inform him officially of the fait accompli and that a solar eclipse happened in the East (together with a long and confused excursus about eclipses). Notice that in his account Ammianus seems to imply that all this happened while Constantius II was still in Constantinople.
Ammianus doesn’t return to the details of Constantius II’s war against the Sasanians until this passage (Rerum Gestarum Libri XXI, XX, 11.1-5):
Securing his northern flank by ensuring the Armenian cooperation was a key precaution for Constantius II that shows once more that he was a sound strategist. And he did so with bribes or subsidies (that Ammianus diplomatically describes as “presents”. From Ammianus’ account, it can be surmised that Constantius II met the Armenian king in Cappadocia, where the Roman augustus had advanced with his army from Constantinople. The meeting could have taken place at Caesarea Mazaca, the provincial capital, or maybe in one of the legionary fortresses in the border. The former possibility seems to make sense considering that Julian’s letters reached Constantius II at Caesarea according to another passage of Ammianus, but the latter possibility also makes sense considering that the route that Constantius II took to reach Mesopotamia was not the usual one, via Tyana in Cappadocia, Tarsus in Cilicia and then Antioch in Syria. Instead, according to Ammianus he followed the upper course of the Euphrates via Melitene (modern Malatya in Turkey), Lacotena, and Samosata, and then crossed the Euphrates (there was a crossing point at Samosata) and reached Edessa which like in previous campaigns was to act as the Roman main base in this campaign.
Route of Constantius II’s army from Constantinople to Edessa (in red). I’ve underlined in yellow the main military road from the Bosporus to Antioch. Map from The Digital Atlas of the Roman Empire, by the University of Lund.
Again, following Ammianus’ account, Constantius II seems to have remained for a long time in Edessa, and he only moved “after the autumn equinox”, that is on 21 September at the very earliest, which is quite late in the campaign season. And then he moved to the destroyed city of Amida in the upper course of the Tigris, which only makes sense if his intention from the start was to recover Bezabde. In other words: at this time, Šābuhr II had already concluded his destructive invasion and had been able to retreat back into Ērānšahr with impunity, while Constantius II reached Mesopotamia and remained in Edessa.
The Roman army then marched directly from Amida to Edessa, following the course of the upper Tigris valley, and reached the city (Rerum Gestarum Libri XXI, XX, 11.6-7):
Constantius II inspected personally the state of the fortress and realized that not only had the Sasanian defenders repaired the breach caused by their siege ram, but they had also repaired any other weathered sections of the wall. In short: now the fortifications of Bezabde were even stronger than before the Sasanian conquest. As Constantius II was a seasoned warrior, upon seeing this he opened talks with the defenders, trying to arrange a negotiated surrender with them: either abandon the fortress and return to Ērānšahr unmolested or enter Roman service and be richly rewarded for it (in other words; he offered them bribes). Unsurprisingly, the defenders refused; as we will see, according to Ammianus’ account Šābuhr II dealt in an extremely harsh ways with traitors and their families. And considering the importance of Bezabde and the spirited defense that the garrison put up against the Roman siege, it’s evident that the Šahān Šāh left in there a garrison of elite troops, highly skilled and motivated.
British reenactor Nadeem Ahmad dressed in the full garb of a Sasanian Šahān Šāh or a high-ranking spāhbed. The battle axe was a personal combat weapon typical of eastern Iran and Central Asia, although it seems to have been less used in western Iran.
The Romans then made their preparations and launched their assault, but as we will see in Ammianus’ account, the Romans fared no better than the Sasanians when assaulting a strong fortress with a determined and skilled garrison within it (Rerum Gestarum Libri XXI, XX, 11.8-10):
It's almost as if we were reading the Sasanian sieges of Nisibis, Amida or Singara all over again. It’s interesting to see how, according to Ammianus’ description, the first tactic that the Roman soldiers employed was the classical testudo; clearly this formation had not gone out of fashion even if the Roman had long since abandoned the rectangular scutum in favor of lighter oval shield. In reliefs in Trajan’s Column in Rome more than two centuries earlier, legionaries can be seen doing exactly this while attacking a Dacian fort. It failed in face of what seems to have been a veritable deluge of projectiles from the Sasanian defenders, and when they tried to use wicket mantlets, the result was the same, as the defenders simply dropped heavier projectiles upon the attackers, tearing large gaps in the mantlets. What seems to me is that here the Romans pressed forward an improvised attack (probably due to the lateness of the season) to take to resolve things quickly, and they discovered the hard way that the Sasanians could defend a fortress as well as they did. So, the first three days of the assault ended in a Roman defeat, possibly with serious losses. Let’s see how the Romans renewed their assault (Rerum Gestarum Libri XXI, XX, 11.11-15):
The tides have turned here. Now it’s the Romans who have to resort to all sorts of elaborate siege techniques and absorb huge losses in their ranks while the defenders manage to counteract each and every one of their efforts. This is a long passage, that Ammianus managed to garble it considerably (either him or the scholars who have restored the text of the only mutilated manuscript that was rescued in the XV century and published in Florence), so careful reading is necessary. Also, some interesting bits of information appear in it.
- First, that the Romans had to resort to an abandoned Sasanian siege ram in order to attack Bezabde. This seems to imply that, for all their preparations, the Romans had not brought with them equipment heavy enough for an all-out siege. Given the amount of time that Constantius II took to reach Mesopotamia, and then at Edessa, that’s quite surprising.
- Second, that they resorted to an abandoned Sasanian ram, which suggest that at least in rams, the Sasanians had heavier equipment than the Romans. Obviously, this is the famous ram that Šābuhr I had used in the 250s against the walls of Antioch when he took the city a second time; 100 years later, it was again reassembled and brought to Bezabde from Carrhae (a heavy and time-consuming task in itself) to bring down the walls of this fortress.
Remains of the walls of the (supposed) site of ancient Bezabde, near Cizre in southeastern Turkey. The river in the background is the Tigris.
Following the standard practice of the time, the Romans protected this huge ram and other “smaller rams” so that the Sasanian incendiary devices could not set them on fire (like the Sasanians had done with their siege towers at Nisibis and Amida), but at the hour of truth, the defenders were able to use a “subtle device” with ropes that immobilized the head of the ram, while its crew lay exposed to a deluge of projectiles from the walls, Ammianus’ explanation is frustratingly vague, but in ancient and medieval times a common technique to counter rams was to grip its head with ropes or chains and then lift them upwards from the walls, rendering them useless. The siege now entered its bloodiest and most hardly fought period (Rerum Gestarum Libri XXI, XX, 11.16-19):
As usual with Ammianus, the text is full of dramatism; he was a soldier and had fought in similar engagements and knew well the dangers and fear to which the fighters were exposed in this sort of encounters. As we can see, the defenders did not hesitate to sally twice from the walls and engage the Roman legionaries in hand-to-hand combat. Let’s remember that this was Constantius II’s praesentalis army, the crème de la crème of the Roman mobile armies, so the fact that the Sasanian defenders were able, or even willing to come to grips with them is an indication that the Sasanians, or at least their elite forces, were perfectly able to fight in close fighting against the dreaded Roman infantry if necessary. Notice also that the Sasanians seem to have a considerable degree of familiarity with incendiary devices and employed such techniques with a considerable effectiveness against the Roman besiegers. As we will soon see, they kept using fire against the Romans during this siege with success. The second sally seems to have left the rams badly damaged, even the giant one, but the Roman army just redoubled its efforts against Bezabde. Clearly, Constantius II considered the control of Bezabde a vital matter (or he was desperate for a victory, given the political situation within the empire), and his army’s morale was high enough to endure it with its will to fight intact, a tribute to Constantius II’s thorough preparations and the respect his soldiers had for their augustus. Let’s continue to the next point in Ammianus’ tale (Rerum Gestarum Libri XXI, XX, 11.20-23):
Siege ramps could be built with earth (the surest but slower method) or with bundles of wood, as Ammianus says was the case here. And you can see what the risks were in this case: they were vulnerable to the incendiary attacks of the enemy. Once more, the Sasanians had made a huge gamble with a risky sally, and it had ended in success. After this, the siege petered out, as Constantius II decided to try to starve the garrison by hunger, which is quite puzzling given the lateness of the season (Rerum Gestarum Libri XXI, XX, 11.24-32):
In short: the last assaults were followed by the start of Fall rains and both fear of the discontent and mutiny among his troops if he carried on with the siege, and of the events happening in the West, forced him to lift the siege and retire back to Syria, where he set up his winter quarters in Antioch. The Roman siege of Bezabde had failed, and the road to Nisibis lay open for Šābuhr II.
Gold solidus of Constantius II. On the obverse, CONSTANTIVS AVG(ustus). On the reverse, GLORIA ROMANORVM. Mint of Antioch.
The yazatas have heard my prayers and my inspiration has returned
Praise Ahura Mazda!
8.7 THE DEATH OF CONSTANTIUS II AND THE PREPARATIONS AND PLANNING FOR JULIAN’S CAMPAIGN AGAINST ĒRĀNŠAHR.
8.7 THE DEATH OF CONSTANTIUS II AND THE PREPARATIONS AND PLANNING FOR JULIAN’S CAMPAIGN AGAINST ĒRĀNŠAHR.
After his failure against Bezabde, Constantius II found himself in an increasingly desperate situation. As soon as the weather conditions allowed for it, Šābuhr II would invade again, and this time with good chances of achieving his dreamed prize of Nisibis. But even more worryingly, in the West Julian launched that winter a swift campaign against Constantius II, ending the uneasy truce that had prevailed until then, with the intention of finishing Constantius II off now that he was weakened after his failure against Bezabde and two years of Sasanian successes.
As usual, Ammianus narrative is convoluted and the chronology is unclear, but it seems that Constantius II had foreseen this, for he dispatched his trusted official Florentius (who had fled from Gaul) as Praetorian Prefect for Illyricum, and together with Taurus (the Praetorian Prefect for Italy) was appointed consul for the year 361CE, in a clear attempt to secure the loyalty of both men and reward them for past services. He also dispatched his secretary Gaudentius to Africa in order to secure it against any attempts of invasion from Gaul or Hispania.
The beleaguered augustus though still had to deal with one last threat from the East. While he was in Antioch preparing to march against Julian, he received news that Šābuhr II had massed his army and was approaching the passes of the Tigris. Constantius II reacted in the same way, recalling his scattered forces that had wintered across Syria and gathering them at Antioch. From here, he crossed the Euphrates at Capersana (slightly north of Zeugma) and marched to his usual operations base at Edessa, where he encamped his army and waited for Šābuhr II’s actions.
But then, instead of crossing the Tigris and invade, Šābuhr II remained in the eastern bank of the river, playing a waiting game with his rival. He was probably aware that due to Julian’s rebellion Constantius II could not remain inactive for long, so time played in the Šahān Šāh’s favor (Rerum Gestarum Libri XXXI, XXI, 13.1-5):
But while Constantius II was encamped in Edessa waiting for Šābuhr II’s next move, events in the West took a frankly alarming turn for him: Julian invaded Illyricum and the local troops either were defeated or joined his side. The consuls and praetorian prefects of Illyricum and Italy Florentius and Taurus fled to the West and Italy also fell in Julian’s hands except for the strategic city of Aquileia, which remained loyal to Constantius II. Julian entered Sirmium and then proceeded directly against Constantinople. Upon knowing this, Constantius II abandoned Mesopotamia and recalled his advanced troops, gathering the whole army at Nicopolis (location unknown, probably in Syria or western Mesopotamia) where he addressed his troops telling them he would be marching now against the usurper.
From there the army marched to Antioch, where it took the main military road across Anatolia to the Bosporus. While on the march, and after several omens (according to Ammianus), Constantius II died at Mobsucrenae in Cilicia at the foot of the Taurus mountains on October 5, 361 CE (Julian calendar) after a short illness. At this point, Constantius II’s death was a godsend for the Romans, as it avoided yet another bloody and destructive civil war and left Julian as the sole uncontested augustus of the Roman empire, as he was the sole male heir of Constantine I alive and in the five previous years had managed to gather enough personal military prestige with his campaigns against the Franks and Alamanni to be accepted by the army.
After receiving the news of his cousin’s death, Julian advanced directly on Constantinople, which opened its gates to him. Julian entered the city of the Bosporus in triumph on December 11 361 CE (Julian calendar). A thorough purge of Constantius II’s loyalists in the civilian and military structures followed swiftly, with plenty of executions and exiles.
Julian was a very different character from his cousin. He had been brought up in isolation and constantly fearing for his life, at the whim of Constantius II. Because of this, he’d turned to books and had become a very well-read young man, with an extensive knowledge of philosophy and history. And it’s from history books that he drew most of his inspiration for his campaigns, in the texts of Thucydides, Herodotus, Plutarch, Livius, Caesar, Arrian, etc. His service in Gaul gave him what he lacked: real experience with commanding an army in campaign. And here he showed that even when confronted with the harsh realities of war, he was still willing to emulate what he considered was the “proper way” of waging war, the one that had brought success and glory to Alexander the Great and the great Roman conquerors of old, through unrelenting offensive warfare.
Gold solidus of Julian. On the obverse: FL(avius) CL(audius) IVLIANVS P(ius) F(elix) AUG(ustus). On the reverse: VIRTVS EXERCITVS ROMANORVM (“The ‘virtus’ of the army of the Romans”). Mint of Antioch.
In support of his plans of launching a large-scale invasion of the Sasanian empire following the steps of Trajan or Septimius Severus, he could argue that the defensive strategy conducted for the last twenty-five years by Constantius II had not stopped Šābuhr II’s attacks; on the contrary, the defensive system carefully designed and built by Diocletian in Mesopotamia had been finally breached, and the province was in serious risk of being lost, while it’s very possible that after the loss of Amida, Singara and Bezabde the satrapies beyond the Tigris (Regiones Transtigritanae) that the Treaty of Nisibis had ceded to Rome had already been lost by that time.
And Julian wasted no time in organizing and launching his Expeditio Persica: he invaded Ērānšahr in March 363 CE, just eighteen months after becoming sole augustus of the Roman empire. It was a propitious moment for such an enterprise, as the European borders of the empire were secure after Julian and Constantius II’s strenuous efforts, while the Goths kept themselves loyal to their agreement with the House of Constantine and now as sole Roman augustus Julian could bring the full might of Rome against the Sasanians. This invasion has a special place in history because it was the last full-scale invasion of Ērānšahr by a Roman army until Heraclius’ one in the VII century CE, and the last one launched by a united Roman empire.
John Harrel points out that Julian had enjoyed success in Gaul with a policy of confrontation, pursuit, and devastation against the Germanic confederations, while the twenty-five years of war had possibly convinced Julian that a decisive victory could not be achieved in the Northern Mesopotamian or Armenian theatres of war. Campaigning in these fortified zones resulted in costly, fruitless sieges and indecisive battles. Continuing with his argument, Harrel also points out that as evidenced by Constantius II’ protracted defensive stance, Julian was possibly also well aware that winning defensive battles did not lead to victory but, rather, a long war of attrition, and that as a scholar versed in classical history, Julian would have studied earlier attacks on the city of Ctesiphon and would have been well aware of the manpower and logistics requirements for such an endeavor. He would also have known that prior augusti had achieved victory by achieving two objectives: taking Ctesiphon and defeating any Arsacid or Sasanian army that tried to intervene. But according to Harrel, Julian seems to have had a third objective, one that had eluded all previous successful Roman leaders fighting the Iranian empires. To obtain a decisive victory, Julian had to break into the Iranian Plateau and/or replace Šābuhr II with a client king.
Harrel continues by saying that while Julian’s intent to lead the attack against the Sasanian empire was understood by his contemporaries, his strategic goals were unclear. The sources disagree as to whether Julian’s intended operation was a larger version of his Gallic campaigns, which were actually raids across the Rhine against the Frankish and Alamannic tribes, or a serious attempt to overthrow Šābuhr II and subjugate Ērānšahr to Rome. In the V century CE, the Christian chronicler Socrates of Constantinople suggested that Julian actually believed he was Alexander reincarnated, destined to extend the empire to India. Harrel thinks that the fact that the Sasanian prince Hormazd was one of Julian’s commanders means that the option of replacing Šābuhr II, as Constantine the Great had planned in 337 CE, was seen as a viable course of action. There is a reference to Hormazd in a letter from Libanius implying that Julian planned to put the exiled prince on the Sasanian throne. So, according to Harrel, despite the wild claims of his Christian enemies, Julian had an achievable goal to end the war and that was to turn the western part of the Sasanian empire into a client kingdom under prince Hormazd, which for Harrel would have caused the wuzurgān to revolt and torn the eastern kingdom apart, thus destroying the Sasanian empire.
Continuing his analysis, Harrel points out that in June 363 CE Julian’s army marched north to the Diyala (Douros in Greek) River, which would have allowed him an access route to advance toward the Persian heartland. The main Silk Road followed the Diyala River onto the Iranian Plateau and, by June, the passes in the Zagros Mountains should have been passable. This was the axis of advance that Muslim armies followed to break onto the Iranian Plateau and destroy the Sasanian empire in the VII century CE. And to Harrel, it’s no surprise that after Julian’s arrival at this strategic avenue of approach into the Iranian heartland, on or about 17 June 363 CE, the Sasanian and Roman main armies clashed on the banks of the Diyala River.
The Tigris River Basin. The Diyala River is the longest of its tributaries (445 km) and along history it has been the main land route between the Mesopotamian plain and the Iranian Plateau. It joins the Tigris very near the conurbation of Ctesiphon, and also very near the modern city of Baghdad.
I’ll keep on following Harrel’s detailed analysis of the campaign and Julian’s preparations and alleged objectives here. According to Harrel, whatever his strategic goals may have been, Julian’s operational plan was very similar to Constantine I’s plan in 337 CE. Two armies would attack along separate axes, each initially threatening a key Sasanian region and eventually converging on the enemy capital: Ctesiphon. His forces would then mass and capture the Sasanian capital. This opening phase would have led to the defeat in battle of the main Sasanian army commanded by the Šahān Šāh, and then the second phase would start with the Roman army breaking onto the Iranian Plateau, and ultimately replacing Šābuhr II. It was assumed that Julian’s Praesentalis Army alone or massed together with the secondary invasion army (that Harrel identifies with the field army of the East) would have a reasonable chance of defeating the Šābuhr II’s main army in open battle. The main effort, under Julian, would attack directly down the Euphrates River towards Ctesiphon. The supporting attack would cross the Tigris River, march through the Regiones Transtigritanae, re-establish Roman rule and alliances in them and then move into Ērānšahr proper, marching along the east bank of the Tigris, and south toward Ctesiphon.
Harrel then moves on to make an educated guess about the size and composition of Julian’s invasion forces. The composition and size of Julian’s Praesentalis Army is unclear. Zosimus was the only historian to reference the size of the army noting that when Julian reviewed the army in March 363 CE it mustered 65,000. Gibbon accepted this figure and added the field army of the East’s 20,000 to 30,000 men to the force for a total of between 85,000 and 95,000 soldiers committed to the invasion. A number of historians subtract the forces of the field army of the East from Zosimus’ 65,000, reducing the Praesentalis Army to between 30,000 and 45,000, depending on which source they follow for the size of the field Army of the East. If these numbers were accurate, Julian’s army would have numbered between 35,000 and 47,000 in strength and would have been insufficient to accomplish the campaign objectives. As Harrel points out, the problem with this small figure for the praesentalis army is that most historians overlook the fact that 20,000 men from Julian’s army were engaged in manning and providing security for the supporting fleet, which would have left only 15,000 to 27,000 men for land operations.
Harrel carries on with his analysis by stating that, assuming Vegetius’ ratio of one horseman to four infantrymen, the 65,000-man Praesentalis Army would have contained 13,000 cavalrymen and 52,000 infantrymen. The infantry consisted of Roman legions from Gaul, the Balkans, and the East, as well as auxilia units of “Gauls” and Germans from the Rhine River valley. The core of Julian’s infantry was his loyal auxilia from the Army of Gaul who had declared him augustus. Of the fourteen infantry units named by the sources, eight units were auxilia regiments from Gaul, four were elite legions (two from Gaul), one legio comitatense from an unknown home region (possibly Thrace), and one limitanei legion from the East.
Harrel follows by pointing out that Julian’s cavalry consisted of regular Roman units supported by Goth and Saracen foederati, Armenian allies, and Iranians who had changed sides. Ammianus follows the Graeco-Roman literary convention of referring to the Goths as “Scythians”, which has been the cause of so much confusion. The only regular cavalry units named by the sources are the Promoti and Tertaci with the guard units Scutarii and Protectores Domestici. The exiled Sasanian prince Hormazd would have had a contingent of followers at least as a bodyguard if not a full cavalry unit. The Armenian general Zawray commanded 7,000 Armenian horsemen and Harrel guesses that he was probably present at the initial March 363 CE mustering of the invasion force.
The next point touched by Harrel is the logistics of the campaign, pointing that Ammianus provides some details about the logistics build-up for the invasion but little analysis about the overall logistics plan for the invasion. According to this author, the cities of Roman Mesopotamia were transformed into massive supply depots. He also writes that a large fleet of over 1,000 supply boats and barges was built to support the main effort down the Euphrates River. These boats and barges carried the army’s siege train, replacement weapons, and armor, but most of them were loaded with the expedition’s rations. While vast quantities of sour wine and other foodstuffs were carried, the majority of the weight would have been grain, of which a large percentage would’ve been baked into biscuits for the army’s rations thus reducing the amount of wood needed for the campaign. Fifty war ships protected the fleet and fifty pontoon boats were also included to build bridges, which as we will see would be very necessary in this campaign.
Map of Northern Mesopotamia in Late Antiquity. The border between Rome and Ērānšahr on the Euphrates stood at Circesium. From there it followed the course of the Khabur river in a northeastern direction and the continued directly east along the Jebel Sinjar and reached the Tigris River to the north of Nineveh (very near modern Mosul).
Following Ammianus’ account, Harrel points out that on land the army was supported by a massive baggage train that, while not mentioned by the sources, must have included a large herd of horses for cavalry remounts and cattle for rations. Mules were required to carry a contubernium’s daily camping equipment (tents, cooking utensils, entrenching tools, rations, etc.). It took 1,200 mules to support 5.000 men which could have been replaced by 350 ox wagons; 50,000 infantrymen would have required 12,000 mules or 3,500 wagons (pulled by 7,000 oxen). Light artillery and other equipment and supplies needed to establish a camp would have required wagons. According to Harrel, the exact number is impossible to calculate, but the supply train for an army of 60,000 would have been “massive”.
And most importantly, Harrel also points out that the water requirements for this force dictated the Romans’ axis of advance and confined it to rivers or major canals. In the temperate climate of Europe, 2 liters of water per person were required per day. Cavalry mounts and pack animals required approximately 15 to 20 liters of water per day. Spring and summer operations in Mesopotamia doubled that requirement. So, according to Harrel, the daily water requirement for 60,000 men would have been 120,000 liters, and water for the army’s animals exceeded 600,000 liters. There was no way that the army could transport even one day’s water rations, and the location of a campsite with an adequate water source was a primary task for the Roman scouts and outriders. Harrel states that moving this massive force was difficult and that while well organized, the Romans “set no speed records”. When compared to march rates of early industrial armies the Romans made comparable daily marches. Both forces, centuries apart, averaged 20 km/day.
As for Julian’s strategy, for Harrel it seems clear from Julian’s march down river that he did not plan to use the Euphrates as a supply line or even as a line of communication. He started the advance south in March 363 CE, three months before June harvest (in the area of Ctesiphon), and he bypassed key Sasanian fortresses guarding the Euphrates. These fortresses were no direct threat to his army but were a significant threat to any river traffic between the last major Roman fortress on the Euphrates (Circesium) and Julian’s army. For Harrel, these facts indicate that Julian intended to feed his army during its march southwards from the supplies on his riverboats supplemented from stores captured on the way to Ctesiphon, and that once it reached the enemy capital, the Roman army’s primary source of supply would be the region’s June harvest.
Harrel states that this deployment put Šābuhr II in a strategic dilemma. Until the Praesentalis Army marched south of Circesium, it posed a double menace for the Sasanian monarch: should he protect the Iranian Plateau (where the military heart of the empire was located) or Ctesiphon and with it the provinces of Āsūrestān and Xūzestān (which provided the Sasanian crown with the lion’s share of its income). Until Julian marched south of Circesium, there was the possibility that he could have turned east along the Singara road directly to the Zagros passes into the Iranian homeland. But it would also have been novel (and very un-Roman, which for a devoted scholar of Roman history like Julian would’ve been completely out of character) if the Romans had marched onto the Iranian Plateau without first sacking Ctesiphon. Yet the Romans striking directly onto the Iranian Plateau would’ve been the most dangerous course of action. The only ancient source indicating Šābuhr II’s location in March 363 is Magnus of Carrhae, who placed the main Sasanian army in “Persian Armenia” and not near Ctesiphon (more about this later). Harrel thinks that Šābuhr II, who by now was a mature general, probably chose to place himself and his main army in a central position from which he could defend Bezabde and block the northern route onto the Iranian Plateau, or march southwards to support the garrison of Ctesiphon.
Harrel gives a number for Šābuhr II’s army at the start of the campaign of 40,000 men. He doesn’t elaborate on the reason for this estimate, although I’d guess that he bases it on two points:
Satellite view of Mesopotamia in winter, with the Zagros Mountains and the highlands of the Iranian Plateau covered in snow.
Thus, in the absence of the armies of the Iranian magnates and his vassals and allies the Sasanian field army was smaller Julian’s and would have been formed by the professional forces that Šābuhr II paid directly from the Treasury with the income of the royal estates, tribute from vassals and the taxes he was able to levy across the empire (essentially on trade). it would have included among other contingents the elite forces of the royal guard and the royal corps of elephants. But apart from the field army of the Šahān Šāh, there were other Sasanian forces present in Mesopotamia in March 363 CE. Ancient sources (like Ammianus himself) describe many fortresses and fortified towns and cities dotting the land, all of them well garrisoned with royal troops, which would have been professional forces paid for by the king, with the main concentration happening in the great conurbation of Ctesiphon, where several large fortified cities stood very near to each other (of which Ctesiphon and Vēh-Ardaxšir were the most prominent).
In addition to that, other sources like the Babylonian Talmud and Christian Syriac texts state that the citizens of large cities were also obliged by law to organize and arm themselves as militias to defend their cities in case of attack (the people responsible for this where the richest and most prominent members of the community: the richest merchants, the heads of religious communities, etc.). Although these militias were only useful when defending fortifications and by law they could not be forced to fight outside their cities, they nevertheless alleviated a lot of pressure from the defensive role of the royal army, whose effectives would have been stretched thin in order to cover the very dense network of fortresses and fortified cities in Mesopotamia.
Upper valley of the Diyala/Sirwan River in Iranian Kurdistan.
Despite his youth, Julian was a staunch traditionalist in every way, and so he followed the old and beloved Roman custom of appointing two commanders for the Army of the East. He put it under the joint command of two people who couldn’t be more different in character and experience. The first one was his cousin on the maternal side Procopius, who (according to Ammianus) must have been then 37 years old and completely inexperienced in military matters (he’d been a civil servant and diplomat, and he’d been a member of the embassy that Constantius II sent to the Sasanian court in 358 CE). His only qualification for the post must have been then (according to Harrel) that he was loyal to Julian. The other commander was a very different character: Sebastianus, who had been Dux Aegypti under Constantius II, and who had been now raised to the rank of comes. Sebastianus is described by Ammianus as a “young man”, which for the Romans meant he was under 30; he also seems to have been a man without powerful patrons, and if at that age he had managed to reach such an important rank that means he must’ve been a very competent military leader to escalate the command ladder so quickly, without family connection and under such an experienced military commander as Constantius II had been (and indeed, later Valentinian I and Valens continued to appoint him for important commands). It seems quite clear that these two characters didn’t mix well, and indeed the performance of the Army of the East during this campaign was quite lackluster, as it failed to attain all of the goals that Julian had assigned to it. Quite obviously, either Julian didn’t trust Sebastianus and appointed his cousin Procopius as joint commander to ensure his loyalty or wanted to ensure that in case of success the imperial house could claim the credit for it, or both. To Harrel’s opinion, I would add that another reason for this strange appointment could have been Procopius’ diplomatic credentials, as the Army of the East was expected to collaborate and join forces with the army of king Arshak II of Armenia, and this kingdom, previously a staunch Roman ally, had been acting in an increasingly detached way under Šābuhr II’s relentless pressure.
The kingdom of Armenia in the IV century CE found itself in an increasingly precarious position sandwiched between two powerful empires that fought to assert their control over it. The consequence was a progressive weakening of the kingdom, as its unruly nobility became divided in pro-Roman and pro-Sassanian parties, which meant that Armenia spent most of the time either in civil war or fighting foreign invasions. To the southwest, marked in close green/red stripes, the satrapies ceded to Rome in the First Treaty of Nisibis in 299 CE. But this was only the beginning of the territorial disintegration of the kingdom: during the second half of the IV century it was forced to cede chunks of territory to Rome, Ērānšahr and even to its northern neighbors the kingdoms of Iberia and Albania.
As noted above, the location of Šābuhr II’s main army is stated only in one of the surviving passages of Magnus of Carrhae (who’d been an officer in the Roman army and had taken part in the campaign):
Harrel speculates that the assembly area for Šābuhr II’s army at the start of the 363 CE campaign remains unidentified but, based upon the timing of its appearance in the vicinity of Ctesiphon in June, it must have been in the vicinity of the junction of the Tigris and the Greater Zab rivers. This assembly area would have covered the Diyala Pass, the main pass across the Zagros, and it would have been near “Persian Armenia” and near Nineveh, thus covering also Bezabde and the southernmost of the Regiones Transtigritanae. Harrel quotes the above passage by Magnus of Carrhae in support of his hypothesis; according to it, Šābuhr II would’ve been surprised by the appearance of two Roman campaign armies, which is debatable (it would’ve been impossible to hide such preparations from the considerable (and well documented) network of spies that the Sasanians kept in the Roman East.
What’s more credible though is that he was surprised by the early date at which the Romans began the invasion. It’s possible that the Šahān Šāh first adopted a passive stance behind the Tigris, in order to prevent its crossing by the Army of the East in order to reach the Zagros passes, or any attack against Bezabde (both options would have become a dangerous possibility if king Arshak II of Armenia joined his army with the Roman force led by Procopius and Sebastianus). Šābuhr II would have later abandoned it in favor of moving south to cover Ctesiphon from Julian’s advance. To Harrel it seems clear however, that the Sasanian king declined to cross the Tigris to engage the Army of the East and while he waited further developments and later began moving south towards Ctesiphon along the eastern bank of the Tigris he and left the opposition of Julian’s advance to his general, Surena, and his small border cavalry army.
Ammianus names this general as “the Surena”, and there’s some controversy around this name. For some scholars (a minority, and no Iranologists amongst them), this was an official title, the second after the king. But the majority view sees in this person a nobleman, the head of the Sūrēn Pahlav family, the oldest and most prestigious family of the wuzurgān. In times of the Arsacids, the head of the House of Sūrēn held the hereditary title of Aspbed (“Commander of the Horse”), which was the highest commanding rank in the army, second only to the king. A certain Pērōz the Aspbed still appears at the ŠKZ under Šābuhr I, but this title disappears from the records afterwards. Still, many scholars are tempted to see in this “Surena” who acted as an independent commander of a large cavalry force (Harrel makes a guess for its strength between 3,000 and 5,000 cavalrymen) in such a key mission as covering Ctesiphon from the advance of the main Roman army and delaying it as much as possible, no other that the Aspbed of Ērānšahr, who must have been the head of the House of Sūrēn at the time.
This force though was not formed exclusively by Iranian warriors. Ammianus wrote that it was also formed by a force of Arabs under the command of (Rerum Gestarum Libri XXXI, XXIV, 2.13):
Many scholars see in the ethnonym “Assanitic” a Latin corruption or maybe an early attestation of the Ghassanid Arab tribal confederation that would later become a long-standing ally of the Romans. “Phylarch” was a Greek title that the Roman authorities bestowed on Arab chieftains that they recognized as powerful enough to promote and keep diplomatic relations with. As for “Malechus”, there’s two disputed possibilities. One is that “Malechus” is the Latin corruption of the Arabic “Malik”, meaning “King”. Thus, he would have been “King Podosaces”. But some scholars don’t agree with this, noting that Ammianus wrote “Podosaces” in the genitive case, and so the correct translation should be “Malik, son of Podosaces”. This translation is also possible, as “Malik” has also been used as a male name in Arabic across centuries. Podosaces is the most difficult part of the name. Irfan Shahîd proposed that it could be a Latinization (via Greek) of the Arabic Fadawkas (meaning “little fox”), or perhaps a corruption of the Middle Persian title Bidaxš.
After his failure against Bezabde, Constantius II found himself in an increasingly desperate situation. As soon as the weather conditions allowed for it, Šābuhr II would invade again, and this time with good chances of achieving his dreamed prize of Nisibis. But even more worryingly, in the West Julian launched that winter a swift campaign against Constantius II, ending the uneasy truce that had prevailed until then, with the intention of finishing Constantius II off now that he was weakened after his failure against Bezabde and two years of Sasanian successes.
As usual, Ammianus narrative is convoluted and the chronology is unclear, but it seems that Constantius II had foreseen this, for he dispatched his trusted official Florentius (who had fled from Gaul) as Praetorian Prefect for Illyricum, and together with Taurus (the Praetorian Prefect for Italy) was appointed consul for the year 361CE, in a clear attempt to secure the loyalty of both men and reward them for past services. He also dispatched his secretary Gaudentius to Africa in order to secure it against any attempts of invasion from Gaul or Hispania.
The beleaguered augustus though still had to deal with one last threat from the East. While he was in Antioch preparing to march against Julian, he received news that Šābuhr II had massed his army and was approaching the passes of the Tigris. Constantius II reacted in the same way, recalling his scattered forces that had wintered across Syria and gathering them at Antioch. From here, he crossed the Euphrates at Capersana (slightly north of Zeugma) and marched to his usual operations base at Edessa, where he encamped his army and waited for Šābuhr II’s actions.
But then, instead of crossing the Tigris and invade, Šābuhr II remained in the eastern bank of the river, playing a waiting game with his rival. He was probably aware that due to Julian’s rebellion Constantius II could not remain inactive for long, so time played in the Šahān Šāh’s favor (Rerum Gestarum Libri XXXI, XXI, 13.1-5):
But while Constantius II was encamped in Edessa waiting for Šābuhr II’s next move, events in the West took a frankly alarming turn for him: Julian invaded Illyricum and the local troops either were defeated or joined his side. The consuls and praetorian prefects of Illyricum and Italy Florentius and Taurus fled to the West and Italy also fell in Julian’s hands except for the strategic city of Aquileia, which remained loyal to Constantius II. Julian entered Sirmium and then proceeded directly against Constantinople. Upon knowing this, Constantius II abandoned Mesopotamia and recalled his advanced troops, gathering the whole army at Nicopolis (location unknown, probably in Syria or western Mesopotamia) where he addressed his troops telling them he would be marching now against the usurper.
From there the army marched to Antioch, where it took the main military road across Anatolia to the Bosporus. While on the march, and after several omens (according to Ammianus), Constantius II died at Mobsucrenae in Cilicia at the foot of the Taurus mountains on October 5, 361 CE (Julian calendar) after a short illness. At this point, Constantius II’s death was a godsend for the Romans, as it avoided yet another bloody and destructive civil war and left Julian as the sole uncontested augustus of the Roman empire, as he was the sole male heir of Constantine I alive and in the five previous years had managed to gather enough personal military prestige with his campaigns against the Franks and Alamanni to be accepted by the army.
After receiving the news of his cousin’s death, Julian advanced directly on Constantinople, which opened its gates to him. Julian entered the city of the Bosporus in triumph on December 11 361 CE (Julian calendar). A thorough purge of Constantius II’s loyalists in the civilian and military structures followed swiftly, with plenty of executions and exiles.
Julian was a very different character from his cousin. He had been brought up in isolation and constantly fearing for his life, at the whim of Constantius II. Because of this, he’d turned to books and had become a very well-read young man, with an extensive knowledge of philosophy and history. And it’s from history books that he drew most of his inspiration for his campaigns, in the texts of Thucydides, Herodotus, Plutarch, Livius, Caesar, Arrian, etc. His service in Gaul gave him what he lacked: real experience with commanding an army in campaign. And here he showed that even when confronted with the harsh realities of war, he was still willing to emulate what he considered was the “proper way” of waging war, the one that had brought success and glory to Alexander the Great and the great Roman conquerors of old, through unrelenting offensive warfare.
Gold solidus of Julian. On the obverse: FL(avius) CL(audius) IVLIANVS P(ius) F(elix) AUG(ustus). On the reverse: VIRTVS EXERCITVS ROMANORVM (“The ‘virtus’ of the army of the Romans”). Mint of Antioch.
In support of his plans of launching a large-scale invasion of the Sasanian empire following the steps of Trajan or Septimius Severus, he could argue that the defensive strategy conducted for the last twenty-five years by Constantius II had not stopped Šābuhr II’s attacks; on the contrary, the defensive system carefully designed and built by Diocletian in Mesopotamia had been finally breached, and the province was in serious risk of being lost, while it’s very possible that after the loss of Amida, Singara and Bezabde the satrapies beyond the Tigris (Regiones Transtigritanae) that the Treaty of Nisibis had ceded to Rome had already been lost by that time.
And Julian wasted no time in organizing and launching his Expeditio Persica: he invaded Ērānšahr in March 363 CE, just eighteen months after becoming sole augustus of the Roman empire. It was a propitious moment for such an enterprise, as the European borders of the empire were secure after Julian and Constantius II’s strenuous efforts, while the Goths kept themselves loyal to their agreement with the House of Constantine and now as sole Roman augustus Julian could bring the full might of Rome against the Sasanians. This invasion has a special place in history because it was the last full-scale invasion of Ērānšahr by a Roman army until Heraclius’ one in the VII century CE, and the last one launched by a united Roman empire.
John Harrel points out that Julian had enjoyed success in Gaul with a policy of confrontation, pursuit, and devastation against the Germanic confederations, while the twenty-five years of war had possibly convinced Julian that a decisive victory could not be achieved in the Northern Mesopotamian or Armenian theatres of war. Campaigning in these fortified zones resulted in costly, fruitless sieges and indecisive battles. Continuing with his argument, Harrel also points out that as evidenced by Constantius II’ protracted defensive stance, Julian was possibly also well aware that winning defensive battles did not lead to victory but, rather, a long war of attrition, and that as a scholar versed in classical history, Julian would have studied earlier attacks on the city of Ctesiphon and would have been well aware of the manpower and logistics requirements for such an endeavor. He would also have known that prior augusti had achieved victory by achieving two objectives: taking Ctesiphon and defeating any Arsacid or Sasanian army that tried to intervene. But according to Harrel, Julian seems to have had a third objective, one that had eluded all previous successful Roman leaders fighting the Iranian empires. To obtain a decisive victory, Julian had to break into the Iranian Plateau and/or replace Šābuhr II with a client king.
Harrel continues by saying that while Julian’s intent to lead the attack against the Sasanian empire was understood by his contemporaries, his strategic goals were unclear. The sources disagree as to whether Julian’s intended operation was a larger version of his Gallic campaigns, which were actually raids across the Rhine against the Frankish and Alamannic tribes, or a serious attempt to overthrow Šābuhr II and subjugate Ērānšahr to Rome. In the V century CE, the Christian chronicler Socrates of Constantinople suggested that Julian actually believed he was Alexander reincarnated, destined to extend the empire to India. Harrel thinks that the fact that the Sasanian prince Hormazd was one of Julian’s commanders means that the option of replacing Šābuhr II, as Constantine the Great had planned in 337 CE, was seen as a viable course of action. There is a reference to Hormazd in a letter from Libanius implying that Julian planned to put the exiled prince on the Sasanian throne. So, according to Harrel, despite the wild claims of his Christian enemies, Julian had an achievable goal to end the war and that was to turn the western part of the Sasanian empire into a client kingdom under prince Hormazd, which for Harrel would have caused the wuzurgān to revolt and torn the eastern kingdom apart, thus destroying the Sasanian empire.
Continuing his analysis, Harrel points out that in June 363 CE Julian’s army marched north to the Diyala (Douros in Greek) River, which would have allowed him an access route to advance toward the Persian heartland. The main Silk Road followed the Diyala River onto the Iranian Plateau and, by June, the passes in the Zagros Mountains should have been passable. This was the axis of advance that Muslim armies followed to break onto the Iranian Plateau and destroy the Sasanian empire in the VII century CE. And to Harrel, it’s no surprise that after Julian’s arrival at this strategic avenue of approach into the Iranian heartland, on or about 17 June 363 CE, the Sasanian and Roman main armies clashed on the banks of the Diyala River.
The Tigris River Basin. The Diyala River is the longest of its tributaries (445 km) and along history it has been the main land route between the Mesopotamian plain and the Iranian Plateau. It joins the Tigris very near the conurbation of Ctesiphon, and also very near the modern city of Baghdad.
I’ll keep on following Harrel’s detailed analysis of the campaign and Julian’s preparations and alleged objectives here. According to Harrel, whatever his strategic goals may have been, Julian’s operational plan was very similar to Constantine I’s plan in 337 CE. Two armies would attack along separate axes, each initially threatening a key Sasanian region and eventually converging on the enemy capital: Ctesiphon. His forces would then mass and capture the Sasanian capital. This opening phase would have led to the defeat in battle of the main Sasanian army commanded by the Šahān Šāh, and then the second phase would start with the Roman army breaking onto the Iranian Plateau, and ultimately replacing Šābuhr II. It was assumed that Julian’s Praesentalis Army alone or massed together with the secondary invasion army (that Harrel identifies with the field army of the East) would have a reasonable chance of defeating the Šābuhr II’s main army in open battle. The main effort, under Julian, would attack directly down the Euphrates River towards Ctesiphon. The supporting attack would cross the Tigris River, march through the Regiones Transtigritanae, re-establish Roman rule and alliances in them and then move into Ērānšahr proper, marching along the east bank of the Tigris, and south toward Ctesiphon.
Harrel then moves on to make an educated guess about the size and composition of Julian’s invasion forces. The composition and size of Julian’s Praesentalis Army is unclear. Zosimus was the only historian to reference the size of the army noting that when Julian reviewed the army in March 363 CE it mustered 65,000. Gibbon accepted this figure and added the field army of the East’s 20,000 to 30,000 men to the force for a total of between 85,000 and 95,000 soldiers committed to the invasion. A number of historians subtract the forces of the field army of the East from Zosimus’ 65,000, reducing the Praesentalis Army to between 30,000 and 45,000, depending on which source they follow for the size of the field Army of the East. If these numbers were accurate, Julian’s army would have numbered between 35,000 and 47,000 in strength and would have been insufficient to accomplish the campaign objectives. As Harrel points out, the problem with this small figure for the praesentalis army is that most historians overlook the fact that 20,000 men from Julian’s army were engaged in manning and providing security for the supporting fleet, which would have left only 15,000 to 27,000 men for land operations.
Harrel carries on with his analysis by stating that, assuming Vegetius’ ratio of one horseman to four infantrymen, the 65,000-man Praesentalis Army would have contained 13,000 cavalrymen and 52,000 infantrymen. The infantry consisted of Roman legions from Gaul, the Balkans, and the East, as well as auxilia units of “Gauls” and Germans from the Rhine River valley. The core of Julian’s infantry was his loyal auxilia from the Army of Gaul who had declared him augustus. Of the fourteen infantry units named by the sources, eight units were auxilia regiments from Gaul, four were elite legions (two from Gaul), one legio comitatense from an unknown home region (possibly Thrace), and one limitanei legion from the East.
Harrel follows by pointing out that Julian’s cavalry consisted of regular Roman units supported by Goth and Saracen foederati, Armenian allies, and Iranians who had changed sides. Ammianus follows the Graeco-Roman literary convention of referring to the Goths as “Scythians”, which has been the cause of so much confusion. The only regular cavalry units named by the sources are the Promoti and Tertaci with the guard units Scutarii and Protectores Domestici. The exiled Sasanian prince Hormazd would have had a contingent of followers at least as a bodyguard if not a full cavalry unit. The Armenian general Zawray commanded 7,000 Armenian horsemen and Harrel guesses that he was probably present at the initial March 363 CE mustering of the invasion force.
The next point touched by Harrel is the logistics of the campaign, pointing that Ammianus provides some details about the logistics build-up for the invasion but little analysis about the overall logistics plan for the invasion. According to this author, the cities of Roman Mesopotamia were transformed into massive supply depots. He also writes that a large fleet of over 1,000 supply boats and barges was built to support the main effort down the Euphrates River. These boats and barges carried the army’s siege train, replacement weapons, and armor, but most of them were loaded with the expedition’s rations. While vast quantities of sour wine and other foodstuffs were carried, the majority of the weight would have been grain, of which a large percentage would’ve been baked into biscuits for the army’s rations thus reducing the amount of wood needed for the campaign. Fifty war ships protected the fleet and fifty pontoon boats were also included to build bridges, which as we will see would be very necessary in this campaign.
Map of Northern Mesopotamia in Late Antiquity. The border between Rome and Ērānšahr on the Euphrates stood at Circesium. From there it followed the course of the Khabur river in a northeastern direction and the continued directly east along the Jebel Sinjar and reached the Tigris River to the north of Nineveh (very near modern Mosul).
Following Ammianus’ account, Harrel points out that on land the army was supported by a massive baggage train that, while not mentioned by the sources, must have included a large herd of horses for cavalry remounts and cattle for rations. Mules were required to carry a contubernium’s daily camping equipment (tents, cooking utensils, entrenching tools, rations, etc.). It took 1,200 mules to support 5.000 men which could have been replaced by 350 ox wagons; 50,000 infantrymen would have required 12,000 mules or 3,500 wagons (pulled by 7,000 oxen). Light artillery and other equipment and supplies needed to establish a camp would have required wagons. According to Harrel, the exact number is impossible to calculate, but the supply train for an army of 60,000 would have been “massive”.
And most importantly, Harrel also points out that the water requirements for this force dictated the Romans’ axis of advance and confined it to rivers or major canals. In the temperate climate of Europe, 2 liters of water per person were required per day. Cavalry mounts and pack animals required approximately 15 to 20 liters of water per day. Spring and summer operations in Mesopotamia doubled that requirement. So, according to Harrel, the daily water requirement for 60,000 men would have been 120,000 liters, and water for the army’s animals exceeded 600,000 liters. There was no way that the army could transport even one day’s water rations, and the location of a campsite with an adequate water source was a primary task for the Roman scouts and outriders. Harrel states that moving this massive force was difficult and that while well organized, the Romans “set no speed records”. When compared to march rates of early industrial armies the Romans made comparable daily marches. Both forces, centuries apart, averaged 20 km/day.
As for Julian’s strategy, for Harrel it seems clear from Julian’s march down river that he did not plan to use the Euphrates as a supply line or even as a line of communication. He started the advance south in March 363 CE, three months before June harvest (in the area of Ctesiphon), and he bypassed key Sasanian fortresses guarding the Euphrates. These fortresses were no direct threat to his army but were a significant threat to any river traffic between the last major Roman fortress on the Euphrates (Circesium) and Julian’s army. For Harrel, these facts indicate that Julian intended to feed his army during its march southwards from the supplies on his riverboats supplemented from stores captured on the way to Ctesiphon, and that once it reached the enemy capital, the Roman army’s primary source of supply would be the region’s June harvest.
Harrel states that this deployment put Šābuhr II in a strategic dilemma. Until the Praesentalis Army marched south of Circesium, it posed a double menace for the Sasanian monarch: should he protect the Iranian Plateau (where the military heart of the empire was located) or Ctesiphon and with it the provinces of Āsūrestān and Xūzestān (which provided the Sasanian crown with the lion’s share of its income). Until Julian marched south of Circesium, there was the possibility that he could have turned east along the Singara road directly to the Zagros passes into the Iranian homeland. But it would also have been novel (and very un-Roman, which for a devoted scholar of Roman history like Julian would’ve been completely out of character) if the Romans had marched onto the Iranian Plateau without first sacking Ctesiphon. Yet the Romans striking directly onto the Iranian Plateau would’ve been the most dangerous course of action. The only ancient source indicating Šābuhr II’s location in March 363 is Magnus of Carrhae, who placed the main Sasanian army in “Persian Armenia” and not near Ctesiphon (more about this later). Harrel thinks that Šābuhr II, who by now was a mature general, probably chose to place himself and his main army in a central position from which he could defend Bezabde and block the northern route onto the Iranian Plateau, or march southwards to support the garrison of Ctesiphon.
Harrel gives a number for Šābuhr II’s army at the start of the campaign of 40,000 men. He doesn’t elaborate on the reason for this estimate, although I’d guess that he bases it on two points:
- Šābuhr II’s army did not engage Julian’s army until the latter had reached Ctesiphon and it was more or less forced to give battle. And even then, the Romans prevailed despite the fact they had to cross a major water course to attack the Sasanian army; nevertheless, the latter was able to put up a spirited fight for almost a whole day and retreat in order after the battle. This could indicate that the army of the Šahān Šāh was smaller than the Roman one, but not so small as to turn any fight into a sure defeat for the Sasanians. Also, if the force of the Roman secondary invasion army was of 40,000 men, it must have been tailored to be a match to Šābuhr II’s army.
- Julian attacked very early in the campaign season, in March. And this was probably deliberate on his part, because a substantial part of Šābuhr II’s army was formed by “feudal” levies from the Iranian Plateau (the private armies of the wuzurgān) and from contingents sent by his vassals and allies in Central Asia, Eastern Iran and India (like the armies of his brother Šābuhr Sakān Šāh and the ones of Bahrām Kušān Šāh, the possible “puppet king” of the Chionites in Ṭoḵārestān). These armies would take time to mobilize and travel the huge distances across the empire, which also involved crossing one or various mountain ranges. As for the range that all of them would need to cross, all the passes down to the Diyala Pass (that led directly from Hamadan to Ctesiphon) were blocked by snow until June, although the southern passes between Pārs and Xūzestān were passable, this increased considerably the length of the travel. In practice, this meant that by attacking early in March, Julian had numerical superiority guaranteed for three months. Here, the continued advantage of Rome’s professional army was fully exploited by the young augustus.
Satellite view of Mesopotamia in winter, with the Zagros Mountains and the highlands of the Iranian Plateau covered in snow.
Thus, in the absence of the armies of the Iranian magnates and his vassals and allies the Sasanian field army was smaller Julian’s and would have been formed by the professional forces that Šābuhr II paid directly from the Treasury with the income of the royal estates, tribute from vassals and the taxes he was able to levy across the empire (essentially on trade). it would have included among other contingents the elite forces of the royal guard and the royal corps of elephants. But apart from the field army of the Šahān Šāh, there were other Sasanian forces present in Mesopotamia in March 363 CE. Ancient sources (like Ammianus himself) describe many fortresses and fortified towns and cities dotting the land, all of them well garrisoned with royal troops, which would have been professional forces paid for by the king, with the main concentration happening in the great conurbation of Ctesiphon, where several large fortified cities stood very near to each other (of which Ctesiphon and Vēh-Ardaxšir were the most prominent).
In addition to that, other sources like the Babylonian Talmud and Christian Syriac texts state that the citizens of large cities were also obliged by law to organize and arm themselves as militias to defend their cities in case of attack (the people responsible for this where the richest and most prominent members of the community: the richest merchants, the heads of religious communities, etc.). Although these militias were only useful when defending fortifications and by law they could not be forced to fight outside their cities, they nevertheless alleviated a lot of pressure from the defensive role of the royal army, whose effectives would have been stretched thin in order to cover the very dense network of fortresses and fortified cities in Mesopotamia.
Upper valley of the Diyala/Sirwan River in Iranian Kurdistan.
Despite his youth, Julian was a staunch traditionalist in every way, and so he followed the old and beloved Roman custom of appointing two commanders for the Army of the East. He put it under the joint command of two people who couldn’t be more different in character and experience. The first one was his cousin on the maternal side Procopius, who (according to Ammianus) must have been then 37 years old and completely inexperienced in military matters (he’d been a civil servant and diplomat, and he’d been a member of the embassy that Constantius II sent to the Sasanian court in 358 CE). His only qualification for the post must have been then (according to Harrel) that he was loyal to Julian. The other commander was a very different character: Sebastianus, who had been Dux Aegypti under Constantius II, and who had been now raised to the rank of comes. Sebastianus is described by Ammianus as a “young man”, which for the Romans meant he was under 30; he also seems to have been a man without powerful patrons, and if at that age he had managed to reach such an important rank that means he must’ve been a very competent military leader to escalate the command ladder so quickly, without family connection and under such an experienced military commander as Constantius II had been (and indeed, later Valentinian I and Valens continued to appoint him for important commands). It seems quite clear that these two characters didn’t mix well, and indeed the performance of the Army of the East during this campaign was quite lackluster, as it failed to attain all of the goals that Julian had assigned to it. Quite obviously, either Julian didn’t trust Sebastianus and appointed his cousin Procopius as joint commander to ensure his loyalty or wanted to ensure that in case of success the imperial house could claim the credit for it, or both. To Harrel’s opinion, I would add that another reason for this strange appointment could have been Procopius’ diplomatic credentials, as the Army of the East was expected to collaborate and join forces with the army of king Arshak II of Armenia, and this kingdom, previously a staunch Roman ally, had been acting in an increasingly detached way under Šābuhr II’s relentless pressure.
The kingdom of Armenia in the IV century CE found itself in an increasingly precarious position sandwiched between two powerful empires that fought to assert their control over it. The consequence was a progressive weakening of the kingdom, as its unruly nobility became divided in pro-Roman and pro-Sassanian parties, which meant that Armenia spent most of the time either in civil war or fighting foreign invasions. To the southwest, marked in close green/red stripes, the satrapies ceded to Rome in the First Treaty of Nisibis in 299 CE. But this was only the beginning of the territorial disintegration of the kingdom: during the second half of the IV century it was forced to cede chunks of territory to Rome, Ērānšahr and even to its northern neighbors the kingdoms of Iberia and Albania.
As noted above, the location of Šābuhr II’s main army is stated only in one of the surviving passages of Magnus of Carrhae (who’d been an officer in the Roman army and had taken part in the campaign):
Harrel speculates that the assembly area for Šābuhr II’s army at the start of the 363 CE campaign remains unidentified but, based upon the timing of its appearance in the vicinity of Ctesiphon in June, it must have been in the vicinity of the junction of the Tigris and the Greater Zab rivers. This assembly area would have covered the Diyala Pass, the main pass across the Zagros, and it would have been near “Persian Armenia” and near Nineveh, thus covering also Bezabde and the southernmost of the Regiones Transtigritanae. Harrel quotes the above passage by Magnus of Carrhae in support of his hypothesis; according to it, Šābuhr II would’ve been surprised by the appearance of two Roman campaign armies, which is debatable (it would’ve been impossible to hide such preparations from the considerable (and well documented) network of spies that the Sasanians kept in the Roman East.
What’s more credible though is that he was surprised by the early date at which the Romans began the invasion. It’s possible that the Šahān Šāh first adopted a passive stance behind the Tigris, in order to prevent its crossing by the Army of the East in order to reach the Zagros passes, or any attack against Bezabde (both options would have become a dangerous possibility if king Arshak II of Armenia joined his army with the Roman force led by Procopius and Sebastianus). Šābuhr II would have later abandoned it in favor of moving south to cover Ctesiphon from Julian’s advance. To Harrel it seems clear however, that the Sasanian king declined to cross the Tigris to engage the Army of the East and while he waited further developments and later began moving south towards Ctesiphon along the eastern bank of the Tigris he and left the opposition of Julian’s advance to his general, Surena, and his small border cavalry army.
Ammianus names this general as “the Surena”, and there’s some controversy around this name. For some scholars (a minority, and no Iranologists amongst them), this was an official title, the second after the king. But the majority view sees in this person a nobleman, the head of the Sūrēn Pahlav family, the oldest and most prestigious family of the wuzurgān. In times of the Arsacids, the head of the House of Sūrēn held the hereditary title of Aspbed (“Commander of the Horse”), which was the highest commanding rank in the army, second only to the king. A certain Pērōz the Aspbed still appears at the ŠKZ under Šābuhr I, but this title disappears from the records afterwards. Still, many scholars are tempted to see in this “Surena” who acted as an independent commander of a large cavalry force (Harrel makes a guess for its strength between 3,000 and 5,000 cavalrymen) in such a key mission as covering Ctesiphon from the advance of the main Roman army and delaying it as much as possible, no other that the Aspbed of Ērānšahr, who must have been the head of the House of Sūrēn at the time.
This force though was not formed exclusively by Iranian warriors. Ammianus wrote that it was also formed by a force of Arabs under the command of (Rerum Gestarum Libri XXXI, XXIV, 2.13):
Many scholars see in the ethnonym “Assanitic” a Latin corruption or maybe an early attestation of the Ghassanid Arab tribal confederation that would later become a long-standing ally of the Romans. “Phylarch” was a Greek title that the Roman authorities bestowed on Arab chieftains that they recognized as powerful enough to promote and keep diplomatic relations with. As for “Malechus”, there’s two disputed possibilities. One is that “Malechus” is the Latin corruption of the Arabic “Malik”, meaning “King”. Thus, he would have been “King Podosaces”. But some scholars don’t agree with this, noting that Ammianus wrote “Podosaces” in the genitive case, and so the correct translation should be “Malik, son of Podosaces”. This translation is also possible, as “Malik” has also been used as a male name in Arabic across centuries. Podosaces is the most difficult part of the name. Irfan Shahîd proposed that it could be a Latinization (via Greek) of the Arabic Fadawkas (meaning “little fox”), or perhaps a corruption of the Middle Persian title Bidaxš.
Last edited:
Fantastic update! Now we've come to the part of the story I've been waiting for all this time. Even though I know how Julian's expedition ends, I can't wait to read how it gets to that conclusion...
8.8 THE MARCH OF JULIAN’S ARMY TO CTESIPHON (363 CE).
8.8 THE MARCH OF JULIAN’S ARMY TO CTESIPHON (363 CE).
Julian’s eastern campaign is probably the best documented war amongst all the armed conflicts between Romans and Sasanians. Apart from Ammianus’ account (which, as we will see, is not entirely accurate) there’s also plenty of information to be found in other sources: Libanius (who was a personal friend of Julian), Zosimus, fragments of the lost chronicle of Magnus of Carrhae (who, like Ammianus, took part personally in the expedition), together with minor apports from other authors, mainly Christian ones (generally hostile to Julian).
Despite all this wealth of testimonies, there’s some important gaps. First of all, obviously, the lack of Iranian sources. Clearly, Šābuhr II’s victory over Julian made a lasting impression, because it’s echoed in two very late works: those by Ferdowsī and Tabarī. But, as they are removed more than six hundred years from the events, and through unclear chains of transmission (perhaps even oral history), theirs reports are completely inaccurate and full of fantastic details; in the case of Ferdowsī, he even conflates the characters of Šābuhr I and Šābuhr II and narrates Valerian’s capture and defeat and Julian’s defeat and death as a single, continuous conflict.
Although the general assessment of Julian’s campaign is clear to modern scholars (a Roman disaster, or at least a very heavy defeat), there are serious differences amongst them about what were Julian’s strategic objectives and also about the design of the Sasanian strategy of defense. As in previous chapters, in the analysis I follow mostly the works of John Harrel and Ilkka Syvänne, but in order to reconstruct the chronologic and geographical frame I’m going to follow the seminal work about the Roman-Sasanian wars: The Roman Eastern Frontier and the Persian Wars (AD226–363): A Documentary History by Michael H. Dodgeon & Samuel N.C. Lieu, in which they reconstruct the chronology of the invasion and provide the modern equivalents for the places mentioned by Ammianus. In this work, the reconstruction is mainly based on Ammianus (the main source, as he was a soldier who took part personally in the expedition) but also on the sources I mentioned above:
Syvänne as usual, is quite unorthodox in his views and focuses on the whole political aspects of Julian’s reign, while Harrel restricts himself mostly to military events. Syvänne underlines that Julian’s political position was anything but secure, and that the nine months delay in launching the campaign was probably due to the necessity to consolidate minimally his hold on power and to conduct a purge of Constantius II’s loyalists in the army and the administration. He also points out that Julian’s anti-Christian policies must’ve caused considerable turmoil, as according to both Christian (Gregory of Nazianzus) and pagan (Libanius) authors most soldiers in the eastern armies were Christian by then. And on top of that, his winter stay in Antioch was a difficult one; as the behavior of the troops and of Julian himself (something that’s condemned even by Ammianus) did not endear him to the populace and probably did nothing to endear him to the army (as Antiochenes began to openly mock him in the streets and public events).
They also disagree from the start about the role to be performed by the northern army led by Procopius and Sebastianus. To Syvänne, it’s clear that their army had not enough men to be able to lead independent action against Ērānšahr, and that only if it reunited with the Armenian army led by king Aršak could it pose a real menace. But according to Syvänne it was Julian who actually prevented this from happening; as Aršak’s army would’ve been larger than the Roman one, he would’ve been expected to command the invasion, something that (to Syvänne) would’ve been unacceptable to Julian. According Syvänne’s reading of the ancient sources, the purpose of these northern forces was to lead a secondary invasion of the Sasanian empire, with the Nisaean Plain around Ecbatana (modern Hamadān) in Media as their final objective; one of the richest parts of the western part of the Iranian Plateau, but the division of command imposed by Julian, his unclear orders and his refusal to allow the Armenian king to command the Roman force meant that the Roman army of Procopius and Sebastianus did absolutely nothing during the campaign, while king Aršak’s incursion achieved success and reached deep into Sasanian territory once Šābuhr II’s main army headed towards the south.
The Nisaean Plain, near ancient Ecbatana (modern Hamadān, in Iran).
Julian left Antioch and headed to Hierapolis (modern Manbij, in Syria) via Litarbae (-el Athareb, Syria- 5 March) and Beroea (-Aleppo, Syria- 6 March). According to Ammianus (XXIII, 2, 6) a colonnade collapsed at Hierapolis (where arrived on March 9) and killed fifty soldiers of the army. Ammianus’ account (who was a direct eyewitness) of the campaign is peppered with such “ominous” events. Most scholars think that this is merely a literary device common in ancient pagan writers: with the benefit of hindsight, they interpreted such events (and in some cases even invented them outright) as “omens” from the pagan gods, which was a firmly held belief in classical pagan religion (Ammianus was pagan himself, and he wrote his work when he was retired in Rome, and his public would have the still mostly pagan senatorial upper class of the Urbs). But Syvänne, who as usual is rather unorthodox here (and who has a rather pointed dislike for “classicists”) choses to see in this events and in others that followed it something more “sinister”: that possibly there was, from the start of the campaign, an ongoing plot against Julian’s life, either from disaffected Christians or from loyalists of Constantius II who couldn’t accept the way Julian had seized power. This is something impossible to ascertain. Ammianus himself says nothing of the sort and doesn’t even hint at it. He was a pagan, and a fervent admirer of Julian since the time he served under his command in Gaul. After Julian’s death, rumors spread in the Roman army that he’d been killed by a Roman javelin and not by a Persian one and they were repeated by other sources, but Ammianus clearly and unequivocally dismisses them.
Still, I sort of agree with Syvänne that it was strange that there were so many “accidents” threatening the life of Julian during the campaign. And that he’d certainly done much in the months between his rise to the purple and the start of the campaign to alienate vast swathes of the civilian and military elites of the empire, in his fiscal and religious policies, as well as in his purging of the administration and the army of the East of loyalists of Constantius II.
He remained in Hierapolis for three days (according to Syvänne, Hierapolis would’ve been the concentration point for the army) and, after crossing the Euphrates, he reached Batnae (modern Suruç in Turkey) on 12 March. There another fifty soldiers were killed while foraging (Res Gestae Libri XXXI, XXIII, 2, 7–8):
From there, Julian’s army arrived to Carrhae on 18 March after a forced march. According to Theodoret of Cyrrhus, he avoided Edessa because of its strong Christian connections. At Carrhae he divided his forces (Ammianus XXIII, 3, 4, corroborated by Zosimus and Sozomen) between the main army led by him and the secondary army led by Procopius and Sebastianus. He feigned a march towards the Tigris and at some point between Carrhae and Nisibis he turned sharply south towards Callinicum on the Euphrates, which he reached on 27 March (Ammianus XXIII, 3, 6–7, corroborated by Zosimus). The next day he received a delegation of “Saracen” chieftains who presented him with a gold crown (Ammianus XXIII, 3, 8.). Presumably, his army was joined here by the forces of the Roman Arab foederati led by these chieftains, who fought for the Romans as long as the Romans paid for their services. As we will see, later in the campaign Julian stopped paying “subsidies” to them, and they promptly deserted the Roman side. At Callinicum, he was also joined by the large transport fleet under the command of the comes Lucillianus (Ammianus XXIII, 3, 9, corroborated by Zosimus).
Possible reconstruction of the perimeter of the Roman walls of Circesium, on the modern site of Buseyr, Syria. Located at the confluence of the Euphrates and Khabur rivers, this was the last Roman city on the Euphrates.
Julian left Callinicum for Circesium (the last large Roman fortress on the Euphrates) where he crossed the river Khabur on a bridge of boats (Ammianus XXIII, 5, 1 and Zosimus III, 14, 2). At Circesium he received a letter from Sallustius (the Praetorian Prefect for Gaul) asking him to call off the campaign (Res Gestae Libri XXXI, XXIII, 5, 4):
Once more, Ammianus treated this letter from Sallustius as yet another bad omen, but Syvänne thinks that what could have been prompted Sallustius’ letter was a renewed unrest among the Germanic peoples across the Rhine, as Julian had brought with him to the East most of the Army of Gaul, leaving the West largely undefended. To Syvänne, this could suggest that there was also some sort of opposition against the campaign among some of Julian’s pagan loyalists. After the crossing (which signaled the symbolic start of the invasion and the entry into Sasanian territory)
Upon leaving Circesium, the army marched southwards along the western bank of the Euphrates and passed Zaitha (modern Qalat es-Salihiyah in Syria) on 4 April (Ammianus XXIII, 5, 7 and Zosimus III, 14, 2). At this point (just when the army was crossing the border), Julian was urged on by his theurgist/philosopher friends in his retinue to continue the campaign despite yet more adverse omens (Ammianus XXIII, 5, 10; Socrates III, 21, 6) and decided to address the troops in a speech (Ammianus XXIII, 5, 16–23; also Magnus of Carrhae F 2–5) and gave them a donativum of 130 silver coins per soldier, which suggests that there could’ve been a certain lack of enthusiasm among the troops, even among the Gallic forces (after all, they’d acclaimed him as augustus because he promised he would not send them to fight in the East). The army then entered the Sasanian province of Āsūrestān, and the army adopted a marching order appropriate for marching across hostile territory (Res Gestae Libri XXXI, XXIV, 1, 1-4):
Harrel makes some comments about this passage by Ammianus. It might seem surprising that Ammianus wrote that Julian considered his army “small” and that he chose to arrange his army in an artificially extended column to make it seem bigger than it was. By all accounts, an army of 65,000 men was not small at all and would’ve been adequate for the task. But Harrel points out that the reason for this was probably the fact that approximately 20,000 men would’ve been needed to man the fleet, so the marching part of the army would’ve had about 45,000 men actually, which would have made advisable for Julian to try to cover this fact by unorthodox marching arrangements. A marching Roman army of 60-65,000 men would have occupied a march column ten Roman miles long. With only 40-45,000 men marching in the main body, with the infantry in two columns, the army would have occupied considerably less than ten miles. In this extended march formation, it would have taken three to four hours for the rear guard to close with the head of the column when the army stopped. Syvänne though makes no comments about this point, other than stating that in his opinion Julian’s army would have been formed by 65,000 men but that the 20,000 men manning the fleet wouldn’t have been included in this amount (or in other words, that Julian’s total strength would’ve reached 85,000 men).
Marching arrangements of Julian’s army for the 363 CE invasion, according to John Harrel (The Nisibis War: The Defence of the Roman East AD 337-363)
Some of the officers mentioned by Ammianus in this passage had served under Julian in the Army of Gaul and in some cases had risen meteorically thanks to him, so they would’ve been diehard loyalists of the young augustus. Foremost among them was Nevitta, who had been appointed by Julian magister equitum in 361 AD and Roman consul the next year; Ammianus did not think highly of him: he considered him to be “of low extraction”, “boorish” and “cruel”. Judging by his name, he was probably of Germanic origin (maybe Frankish) and most scholars think he would’ve been a pagan (although I don’t know on which grounds, other than the fact he was trusted by Julian). Dagalaifus was a similar case. He was an officer from the Army of Gaul, pagan and of Germanic origins, who had been raised to the post of comes domesticorum (i.e. head of Julian’s personal guard), a post that showed that Julian trusted him completely. Unlike Nevitta, who disappears from the records after Julian’s death, Dagalaifus had a successful career as a military commander under Valentinian and Valens.
Flavius Victor on the contrary seems to have been an officer who made his career in the East, and thus under Constantius II. Due to this and to his Christian faith, some historians consider him to have been opposed to Julian, or at least to some of his policies. But Julian’s acts seem to indicate otherwise, as he raised him to the rank of comes and to the post of magister equitum in Constantinople before the start of the campaign. He was also of “barbarian” origins (Sarmatian, in this case) and he also had a long and successful military career under Julian’s successors. Flavius Arinthaeus was a similar case to Victor. His rank was also that of comes. He also would enjoy a successful career under Julian’s successors. He was made magister peditum by Valens, and in 372 CE he was appointed as consul.
Notice also that the army was marching along the eastern bank of the Euphrates, clearly in order to avoid having to stage a crossing later in the face of possible enemy resistance, although this also forced the army to guard its open eastern flank, which would have exposed to attacks by the Sasanian cavalry.
The army passed in front of the ruins of Dura Europos (which stood on the opposite bank of the Euphrates) on 6 April. The army then reached the first Sasanian fortress on the river, Anatha (modern Anah in Iraq), which was captured on 11 April after a show of strength (Ammianus XXIV, 1, 6–10; Zosimus III, 14, 2–3; Libanius, Oration XVIII). Let’s review Ammianus’ narration of this event (Res Gestae Libri XXXI, XXIV, 1, 6-10):
Anatha was a riverine fortress located on an island in the middle of the Euphrates, and thus it could have opposed resistance. That the fortress surrendered so quickly after Hormazd’s mission seems to indicate that there was willingness from the defenders to collaborate. Most scholars think that the garrison could’ve been supporters of Hormazd against his brother Šābuhr II, but Ilkka Syvänne offers (and quite unsupported) theory. He suggests that among the “numerous offspring” of the old Roman soldier who lived in Anatha could’ve been Pusaeus himself, the commander of the garrison. Syvänne only has very slight circumstantial “evidence” to support his theory: “Pusaeus” is not an Iranian name, so he must’ve been a local Aramaean speaker and that he was later allowed to reach command posts in the Roman army (although this means little or nothing; at the time the Roman army was stuffed with foreign soldiers and officers). Against it there’s the obvious fact that Ammianus says nothing of the sort despite going at length to tell the old soldier’s story.
Map of Julian’s invasion of Ērānšahr, according to Ilkka Syvänne (Military History of Late Rome, 361-395). Compare the several possible locations of the Naarmalcha canal; I will post other maps in this chapter.
Nevertheless, the garrison ran a huge risk with his desertion, because Ammianus tells us how did Šābuhr II deal with traitors: they were executed, along with their families. So, in this sense it’s logical that Julian resettled all of them with their families to Syria. The fact that they lived there with their families could imply also that they were local levies of limited combat value, and perhaps with a limited sense of loyalty towards the Sasanian monarchy; also as their families were with them they had little to fear in the way of reprisals from the Šahān Šāh.
Presumably while it was still at Anatha (Ammianus doesn’t specify) Julian’s army was hit by a “hurricane” (presumably a sandstorm) that caused serious disorder and tore apart many tents. And the very same day, another disaster struck (Res Gestae Libri XXXI, XXIV, 1, 11):
Both Harrel and Syvänne agree that this sudden flood was probably manmade, and that it could have been the Sasanian forces that manipulated the irrigation systems of the area to flood the way of the Roman army, something that Ammianus himself realized at the time. After these incidents, the army continued its advance and captured and burnt an unnamed city (Ammianus and the other sources offer no details).
On March 13 (presumably, according to Dodgeon and Lieu), the Roman army reached the Sasanian fortress of Thilutha (which Dodgeon and Lieu identify with modern Telbes, a rocky islet 14 km. south of Ana). According to Ammianus, this was an almost impregnable fortress (Res Gestae Libri XXXI, XXIV, 2, 1-2):
In other words, the Romans bypassed the fortress, and the defenders made no attempt at attacking the Roman fleet while it sailed downriver under their walls. The attitude of the Sasanian defenders allowing the Roman fleet to sail under their walls unmolested is noteworthy, but their promise to remain neutral until the end of the campaign sounds quite dubious. As Syvänne and Harrel noticed, in bypassing Thilutha, Julian was acting like Šābuhr II in his campaigns against Roman Mesopotamia in the 340s: he was advancing directly against the main enemy base (then Nisibis, now Ctesiphon) leaving enemy garrisons behind him directly on his supply route. And now (like in the 340s) this meant that if the invading army failed to conquer its objective before it ran out of supplies, it would have to live off the land or retreat in haste with its troops going hungry and being harassed by the enemy (which is exactly what happened in the end). I haven’t seen any scholar make a hypothesis about the possible motivations behind the behavior of Thilutha’s garrison. Considering as it wasn’t a city but a purely military fortress, it’s possible that it was garrisoned by professional troops and not by a local militia, and so they should have acted more aggressively. Syvänne puts forward the opinion that it was part of a deliberate tactic by the Sasanians to drag the Romans deep into Ērānšahr while denying them to maintain open supply routes. It’s a possibility worth considering but given the size of the Roman army and the effectiveness they showed at siege warfare (as we will see soon enough when the army reached the key Sasanian fortified city of Pērōz-Šābuhr).
This procedure was repeated at the next fortress the army encountered, Achaiacala. This is an unidentified place; Alois Musil referred to it in his 1927 work Arabia Deserta: A Topographical Itinerary as probably situated somewhere in the vicinity of modern Haditha, about 75 km downriver from Ana (or 45 km. in a straight line), which was also bypassed after the garrison “promised” Julian they wouldn’t intervene further in the war. Two days after this episode (it would’ve been mid-April according to Dodgeon and Lieu) the army crossed to the western bank of the Euphrates at a place that Ammianus called Baraxmalcha. Musil could not locate this place but he hypothesized it could be a corruption of the Arabic word for “ford” farad, which combined with the Arabic for “king” malek would mean “the royal crossing”. The army then entered the town of Diacira which had been abandoned by its inhabitants. Diacira was identified by Musil as modern Hit in Iraq, as he translated the name as “bitumen-giving”. After this, according to Dodgeon and Lieu the army passed by Sitha (neither Dodgeon and Lieu nor Musil offer any possible locations for this place) and must’ve then crossed back to the eastern bank of the Euphrates at Megia (these places only appear in Zosimus’ account, not in Ammianus’, and neither Dodgeon and Lieu nor Musil offer any hypothesis about their possible location).
Around April 22, the army reached Ozogardana (named as Zaragardia in Zosimus’ account; identified by Musil with modern Sari-al-Hadd in Iraq); where the army rested for two days. Here, the Romans first encountered the cavalry force lead by Sūrēn who would shadow them constantly from this moment on; the renegade Sasanian prince Hormazd led a foraging/reconnaissance party that was ambushed by the cavalry force of Sūrēn and the Arab phylarch Podosaces, as narrated by Ammianus and Zosimus (Res Gestae Libri XXXI, XXIV, 2, 4-5):
Ammianus’ claim that this was an ambush seems improbable, because from his description of the encounter it seems clear that the Sasanian force was located across the Euphrates at a place where the river could not be forded, and so this would have been a very inept attempt at an ambush indeed. As Syvänne noted, it seems more probable that this was a fortuitous encounter. But it was to foreshadow things to come; from now on, the Roman column would need to be constantly on guard against surprise attacks, and foraging parties would need to be reinforced if they didn’t want to be annihilated by Sūrēn’s force, which as we will see, would show itself consistently superior to the Roman cavalry. In addition to the difficulties for foraging, this would also mean that from this moment on, Julian’s army would advance without being able to conduct a proper reconnaissance of what lay in its way.
Thus, shadowed by enemy forces, the Roman army reached the town of Macepracta (thanks to the mention by Ammianus of the remains of a rampart at this location helped Musil to locate it near the beginning of an ancient rampart at modern Ummu-r-Rus in Iraq, which stretches from the north bank of the Euphrates northward as far as the Tigris), near the start of the Naarmalcha (also known as the “Royal Canal”) and crossed it to reach the heavily fortified city of Pirisabora. The crossing of the canal was not without difficulties (Res Gestae Libri XXXI, XXIV, 2, 6-8):
Here we see how the Roman cavalry failed in the attempt to force the crossing of the Naarmalcha, but the day was saved by the Roman auxilia, probably Julian’s veterans from the Army of Gaul, who were very adept in these sort of missions, as can be seen in Ammianus’ account of Julian’s campaign in Gaul. Pirisabora has been identified easily by modern scholars with Pērōz-Šābuhr, the ancient Misiche, which Šābuhr I renamed after is victory over Gordian I’s army a century earlier. After the Arab conquest of Mesopotamia, it was renamed as Anbar and under the Umayyads and Abbasids it still retained its importance as a main military base and one of the main arsenals of the Caliphs’ armies, under the name of al-Anbar. The ancient city has not been excavated because it lies today under the waters of the Euphrates. It was of essential importance to Julian’s invasion because it controlled the entry to the Naarmalcha from the Euphrates, and controlling it was necessary if the fleet was to follow the army to Ctesiphon. The city was taken by assault after a short and violent siege by the Roman army, as stated by Ammianus, Zosimus, Libanius and the Liber Suda. Ammianus’ account of the events (Res Gestae Libri XXXI, XXIV, 2, 9-22):
The fall of Pērōz-Šābuhr was an important blow to the Sasanian defensive system, as according to Ammianus it controlled access to the Naarmalcha (meaning “the Royal Canal”) canal. The exact location and course of the Naarmalcha has been the subject of quite a lot of confusion and misgivings even in modern accounts of the campaign, in part because the ancient sources themselves offer contradictory information about this watercourse. While writing these chapters about the events in the Late Antique Near East I check whenever possible the places and locations where events took place in the Barrington Atlas of the Greek and Roman World, which is perhaps the most complete compilation published of the Classical world geography-wise. And when I reached this point in the campaign, I noticed that this scholarly work situated the start of the Naarmalcha a long wat downriver from Pērōz-Šābuhr, which goes against all the accounts of Julian’s campaign; this surprised me. And then I checked other modern resources and noticed a similar state of confusion (the Digital Atlas of the Roman Empire, an online resource supported by the University of Lund in Sweden, does not even draw it clearly on the ground).
The Naarmalcha canal, according to François Paschoud.
Finally, I found a paper that seemed to settle the subject, in a clear and precise way. And once more to my surprise though, this was not exactly a new research at all. In 1978, the French scholar François Paschoud wrote a detailed paper about this ancient canal which pretty much solved my doubts about its location and history (Le Naarmalcha: à propos du tracé d’un canal en Mésopotamie Moyenne). I’ll be following Paschoud’s paper here. basically, the Naarmalcha was a canal that took water from the Euphrates and the flowed on a south-east direction, following the natural incline of the middle Mesopotamian plain. Its main function was to act as a irrigation channel, but it also allowed the navigation by riverine vessels. According to Paschoud, its main branch did not reach the Tigris; after reaching its environs some kilometers south of Seleucia-Ctesiphon, it followed roughly a parallel course to the Tigris until all its water was distributed in lateral irrigation channels. But according to Paschoud, when comparing the ancient sources (Ammianus, Zosimus, Libanius and Cassius Dio -when describing Trajan’s 116-117 CE Parthian campaign-) in this last sector of the Naarmalcha there was a side channel that connected the Naarmalcha with the Tigris upstream from Ctesiphon and Vēh-Ardaxšīr (this is important, because for example it disagrees with Harrel’s reconstruction of the campaign, but not with Syvänne’s) which was known to Graeco-Roman authors as “Trajan’s Canal”, despite the fact that Cassius Dio makes clear that Trajan found this canal already in existence, and so it wasn’t built by him). We’ll return to this point later.
It’s noteworthy how the Roman army managed to storm so quickly the walls of the city; the mention of a helepolis (perhaps the most elaborate and costly siege engine in Antiquity) makes clear that Julian army was well equipped for sieges, and that Julian and/or his generals knew how to use them. Nevertheless, it’s clear that by now Julian (used to his “commando-like” operations in Gaul and in Germanic lands) was getting impatient, as he chose to arrange a negotiated surrender with the defenders of the citadel. About the issue of the surrender of Pērōz-Šābuhr though there’s a very important difference between the accounts of Ammianus (see above) and Zosimus (New History, 3.18.4):
First, Zosimus says there were five thousand defenders, not two thousand and five hundred. And he adds the very important details that Julian actually bought them off, which was always a very unpopular measure with Roman soldiers and officers. in order to resort to it, Julian must have thought that time was really of the essence. Possibly because of this, Julian thought it to be advisable to distribute money among his troops (Zosimus, New History, 3.18.5-7):
According to Syvänne, Julian’s profligacy would have serious consequences later on. The distribution of money among the Roman soldiers is also noted by Ammianus, although in less favorable terms to the emperor (Res Gestae Libri XXXI, XXIV, 3, 3):
On the day after the fall of Pērōz-Šābuhr, a Roman reconnaissance party was ambushed and routed by the cavalry led by Sūrēn. Its commanders were relieved of their command by Julian (Ammianus XXIV, 3, 1–2, Zosimus III, 19,1–2 and Libanius, oration XVIII, 229). The army marched along the canal past Phissenia (identified by Musil with modern ‘Akar an-Na Jeli, lying about two kilometers from the Euphrates on the left bank of the canal) but the advance was slowed by the Sasanians who breached the dams and turned the low ground into marshland (Ammianus XXIV, 3, 10–11, Zosimus III, 19, 3–421 and Libanius, Oration XVIII, 222–6 and 232–4). Ammianus is quite confusing in this part of his account, because he doesn’t state clearly where did the army start to follow the Naarmalcha towards Ctesiphon. As we’ve seen above, he explicitly states it bifurcated from the Euphrates at Pērōz-Šābuhr, but later on he writes as if the army kept on following the Euphrates downriver from this city, hence the confusion among scholars.
At this point the Euphrates was divided into many small streams (Ammianus XXIV, 3, 14). In this district they found an abandoned Jewish settlement (Ammianus XXIV, 4, 1). The Jewish settlement was probably Nehardea, the great center of Jewish rabbinical learning in “Babylonia”.
Julian headed the bridging operation and led his army to Bithra (a problematic location; Musil identified it tentatively with Ibrahim al-Halil, about 25 km. south-east of Phissenia) where they found the ruins of a palace (Zosimus III, 19, 49). Maiozamalcha (The name in Aramaic means “royal fort” or “royal port”, so called because of its position on the Naarmalcha. Musil noticed a ruin mound -modern Han az-Zad- about 18 km. west of Ctesiphon; The “ma’oz” part of the name could be perhaps derived from “Mahoza”, the name given by Jews and Christians to Vēh-Ardaxšīr), a well-fortified city, was taken after a laborious siege (probably on 10–13 May 363 CE) as related in Ammianus XXIV, 4, 2–30, Zosimus III, 20, 2–22, 7 and Libanius, Orations XVIII, 235–42. The storming of Maiozamalcha was another outstanding achievement by the Roman army in siege warfare (Res Gestae Libri XXXI, XXIV, 4, 2-30):
I’d like to comment on two points of Ammianus’ account. First, that obviously the Roman army had launched its invasion very well provided with siege machinery and that like at Pērōz-Šābuhr it was employed here to devastating effect (even better here, perhaps because of the lack of a citadel, which is strange, as it’s a common feature in practically all excavated Sasanian cities, whether in the eastern or the western part of the empire). The second point is an ominous one for the Romans: Ammianus mentions that Sūrēn’s cavalry attacked the pack animals of the army. As we will see, this would later have consequences, and I’m inclined to agree with Syvänne who thinks this attack was part of a planned strategy by the Sasanian defenders.
Map of Sasanian Iraq (provinces of Āsūrestān and Nodšēragān in 363 CE). The irrigated plains began south of the Naarmalcha canal.
Julian’s eastern campaign is probably the best documented war amongst all the armed conflicts between Romans and Sasanians. Apart from Ammianus’ account (which, as we will see, is not entirely accurate) there’s also plenty of information to be found in other sources: Libanius (who was a personal friend of Julian), Zosimus, fragments of the lost chronicle of Magnus of Carrhae (who, like Ammianus, took part personally in the expedition), together with minor apports from other authors, mainly Christian ones (generally hostile to Julian).
Despite all this wealth of testimonies, there’s some important gaps. First of all, obviously, the lack of Iranian sources. Clearly, Šābuhr II’s victory over Julian made a lasting impression, because it’s echoed in two very late works: those by Ferdowsī and Tabarī. But, as they are removed more than six hundred years from the events, and through unclear chains of transmission (perhaps even oral history), theirs reports are completely inaccurate and full of fantastic details; in the case of Ferdowsī, he even conflates the characters of Šābuhr I and Šābuhr II and narrates Valerian’s capture and defeat and Julian’s defeat and death as a single, continuous conflict.
Although the general assessment of Julian’s campaign is clear to modern scholars (a Roman disaster, or at least a very heavy defeat), there are serious differences amongst them about what were Julian’s strategic objectives and also about the design of the Sasanian strategy of defense. As in previous chapters, in the analysis I follow mostly the works of John Harrel and Ilkka Syvänne, but in order to reconstruct the chronologic and geographical frame I’m going to follow the seminal work about the Roman-Sasanian wars: The Roman Eastern Frontier and the Persian Wars (AD226–363): A Documentary History by Michael H. Dodgeon & Samuel N.C. Lieu, in which they reconstruct the chronology of the invasion and provide the modern equivalents for the places mentioned by Ammianus. In this work, the reconstruction is mainly based on Ammianus (the main source, as he was a soldier who took part personally in the expedition) but also on the sources I mentioned above:
- Libanius, who was also a contemporary of the events and a personal friend of Julian. He was also a pagan like Ammianus and a staunch supporter of Julian. Ammianus as a whole, though a pagan and an admirer of Julian, was not overly anti-Christian or overtly supportive of Julian’s internal policies as Libanius was, but this increased impartiality is overshadowed by two facts: Libanius writings are closer in time to the events than Ammianus’ ones (his work was probably published in Rome only in 391 CE at the earliest) and Libanius was a much better writer, had a superior education and a higher social status than Ammianus, meaning that he could have had access to better sources.
- Zosimus, who wrote in Constantinople during the late V century CE (or very early VI century CE). He was a pagan and ardently anti-Christian and the part of his work that covers Julian’s campaign is based on the work of another pagan (and anti-Christian) historian, Eunapius of Sardes, who wrote during the late IV century CE or early V century CE, and was thus reasonably close to the events.
- Fragments of the lost history of Magnus of Carrhae, who is supposed by scholars to have been also an officer in Julian’s expedition (a similar case to Ammianus).
- And finally, fragments in a scattering of (generally extremely hostile to Julian) Christian authors. Despite the obvious bias of their authors, they can be useful when contrasted to the sources listed above.
Syvänne as usual, is quite unorthodox in his views and focuses on the whole political aspects of Julian’s reign, while Harrel restricts himself mostly to military events. Syvänne underlines that Julian’s political position was anything but secure, and that the nine months delay in launching the campaign was probably due to the necessity to consolidate minimally his hold on power and to conduct a purge of Constantius II’s loyalists in the army and the administration. He also points out that Julian’s anti-Christian policies must’ve caused considerable turmoil, as according to both Christian (Gregory of Nazianzus) and pagan (Libanius) authors most soldiers in the eastern armies were Christian by then. And on top of that, his winter stay in Antioch was a difficult one; as the behavior of the troops and of Julian himself (something that’s condemned even by Ammianus) did not endear him to the populace and probably did nothing to endear him to the army (as Antiochenes began to openly mock him in the streets and public events).
They also disagree from the start about the role to be performed by the northern army led by Procopius and Sebastianus. To Syvänne, it’s clear that their army had not enough men to be able to lead independent action against Ērānšahr, and that only if it reunited with the Armenian army led by king Aršak could it pose a real menace. But according to Syvänne it was Julian who actually prevented this from happening; as Aršak’s army would’ve been larger than the Roman one, he would’ve been expected to command the invasion, something that (to Syvänne) would’ve been unacceptable to Julian. According Syvänne’s reading of the ancient sources, the purpose of these northern forces was to lead a secondary invasion of the Sasanian empire, with the Nisaean Plain around Ecbatana (modern Hamadān) in Media as their final objective; one of the richest parts of the western part of the Iranian Plateau, but the division of command imposed by Julian, his unclear orders and his refusal to allow the Armenian king to command the Roman force meant that the Roman army of Procopius and Sebastianus did absolutely nothing during the campaign, while king Aršak’s incursion achieved success and reached deep into Sasanian territory once Šābuhr II’s main army headed towards the south.
The Nisaean Plain, near ancient Ecbatana (modern Hamadān, in Iran).
Julian left Antioch and headed to Hierapolis (modern Manbij, in Syria) via Litarbae (-el Athareb, Syria- 5 March) and Beroea (-Aleppo, Syria- 6 March). According to Ammianus (XXIII, 2, 6) a colonnade collapsed at Hierapolis (where arrived on March 9) and killed fifty soldiers of the army. Ammianus’ account (who was a direct eyewitness) of the campaign is peppered with such “ominous” events. Most scholars think that this is merely a literary device common in ancient pagan writers: with the benefit of hindsight, they interpreted such events (and in some cases even invented them outright) as “omens” from the pagan gods, which was a firmly held belief in classical pagan religion (Ammianus was pagan himself, and he wrote his work when he was retired in Rome, and his public would have the still mostly pagan senatorial upper class of the Urbs). But Syvänne, who as usual is rather unorthodox here (and who has a rather pointed dislike for “classicists”) choses to see in this events and in others that followed it something more “sinister”: that possibly there was, from the start of the campaign, an ongoing plot against Julian’s life, either from disaffected Christians or from loyalists of Constantius II who couldn’t accept the way Julian had seized power. This is something impossible to ascertain. Ammianus himself says nothing of the sort and doesn’t even hint at it. He was a pagan, and a fervent admirer of Julian since the time he served under his command in Gaul. After Julian’s death, rumors spread in the Roman army that he’d been killed by a Roman javelin and not by a Persian one and they were repeated by other sources, but Ammianus clearly and unequivocally dismisses them.
Still, I sort of agree with Syvänne that it was strange that there were so many “accidents” threatening the life of Julian during the campaign. And that he’d certainly done much in the months between his rise to the purple and the start of the campaign to alienate vast swathes of the civilian and military elites of the empire, in his fiscal and religious policies, as well as in his purging of the administration and the army of the East of loyalists of Constantius II.
He remained in Hierapolis for three days (according to Syvänne, Hierapolis would’ve been the concentration point for the army) and, after crossing the Euphrates, he reached Batnae (modern Suruç in Turkey) on 12 March. There another fifty soldiers were killed while foraging (Res Gestae Libri XXXI, XXIII, 2, 7–8):
From there, Julian’s army arrived to Carrhae on 18 March after a forced march. According to Theodoret of Cyrrhus, he avoided Edessa because of its strong Christian connections. At Carrhae he divided his forces (Ammianus XXIII, 3, 4, corroborated by Zosimus and Sozomen) between the main army led by him and the secondary army led by Procopius and Sebastianus. He feigned a march towards the Tigris and at some point between Carrhae and Nisibis he turned sharply south towards Callinicum on the Euphrates, which he reached on 27 March (Ammianus XXIII, 3, 6–7, corroborated by Zosimus). The next day he received a delegation of “Saracen” chieftains who presented him with a gold crown (Ammianus XXIII, 3, 8.). Presumably, his army was joined here by the forces of the Roman Arab foederati led by these chieftains, who fought for the Romans as long as the Romans paid for their services. As we will see, later in the campaign Julian stopped paying “subsidies” to them, and they promptly deserted the Roman side. At Callinicum, he was also joined by the large transport fleet under the command of the comes Lucillianus (Ammianus XXIII, 3, 9, corroborated by Zosimus).
Possible reconstruction of the perimeter of the Roman walls of Circesium, on the modern site of Buseyr, Syria. Located at the confluence of the Euphrates and Khabur rivers, this was the last Roman city on the Euphrates.
Julian left Callinicum for Circesium (the last large Roman fortress on the Euphrates) where he crossed the river Khabur on a bridge of boats (Ammianus XXIII, 5, 1 and Zosimus III, 14, 2). At Circesium he received a letter from Sallustius (the Praetorian Prefect for Gaul) asking him to call off the campaign (Res Gestae Libri XXXI, XXIII, 5, 4):
Once more, Ammianus treated this letter from Sallustius as yet another bad omen, but Syvänne thinks that what could have been prompted Sallustius’ letter was a renewed unrest among the Germanic peoples across the Rhine, as Julian had brought with him to the East most of the Army of Gaul, leaving the West largely undefended. To Syvänne, this could suggest that there was also some sort of opposition against the campaign among some of Julian’s pagan loyalists. After the crossing (which signaled the symbolic start of the invasion and the entry into Sasanian territory)
Upon leaving Circesium, the army marched southwards along the western bank of the Euphrates and passed Zaitha (modern Qalat es-Salihiyah in Syria) on 4 April (Ammianus XXIII, 5, 7 and Zosimus III, 14, 2). At this point (just when the army was crossing the border), Julian was urged on by his theurgist/philosopher friends in his retinue to continue the campaign despite yet more adverse omens (Ammianus XXIII, 5, 10; Socrates III, 21, 6) and decided to address the troops in a speech (Ammianus XXIII, 5, 16–23; also Magnus of Carrhae F 2–5) and gave them a donativum of 130 silver coins per soldier, which suggests that there could’ve been a certain lack of enthusiasm among the troops, even among the Gallic forces (after all, they’d acclaimed him as augustus because he promised he would not send them to fight in the East). The army then entered the Sasanian province of Āsūrestān, and the army adopted a marching order appropriate for marching across hostile territory (Res Gestae Libri XXXI, XXIV, 1, 1-4):
Harrel makes some comments about this passage by Ammianus. It might seem surprising that Ammianus wrote that Julian considered his army “small” and that he chose to arrange his army in an artificially extended column to make it seem bigger than it was. By all accounts, an army of 65,000 men was not small at all and would’ve been adequate for the task. But Harrel points out that the reason for this was probably the fact that approximately 20,000 men would’ve been needed to man the fleet, so the marching part of the army would’ve had about 45,000 men actually, which would have made advisable for Julian to try to cover this fact by unorthodox marching arrangements. A marching Roman army of 60-65,000 men would have occupied a march column ten Roman miles long. With only 40-45,000 men marching in the main body, with the infantry in two columns, the army would have occupied considerably less than ten miles. In this extended march formation, it would have taken three to four hours for the rear guard to close with the head of the column when the army stopped. Syvänne though makes no comments about this point, other than stating that in his opinion Julian’s army would have been formed by 65,000 men but that the 20,000 men manning the fleet wouldn’t have been included in this amount (or in other words, that Julian’s total strength would’ve reached 85,000 men).
Marching arrangements of Julian’s army for the 363 CE invasion, according to John Harrel (The Nisibis War: The Defence of the Roman East AD 337-363)
Some of the officers mentioned by Ammianus in this passage had served under Julian in the Army of Gaul and in some cases had risen meteorically thanks to him, so they would’ve been diehard loyalists of the young augustus. Foremost among them was Nevitta, who had been appointed by Julian magister equitum in 361 AD and Roman consul the next year; Ammianus did not think highly of him: he considered him to be “of low extraction”, “boorish” and “cruel”. Judging by his name, he was probably of Germanic origin (maybe Frankish) and most scholars think he would’ve been a pagan (although I don’t know on which grounds, other than the fact he was trusted by Julian). Dagalaifus was a similar case. He was an officer from the Army of Gaul, pagan and of Germanic origins, who had been raised to the post of comes domesticorum (i.e. head of Julian’s personal guard), a post that showed that Julian trusted him completely. Unlike Nevitta, who disappears from the records after Julian’s death, Dagalaifus had a successful career as a military commander under Valentinian and Valens.
Flavius Victor on the contrary seems to have been an officer who made his career in the East, and thus under Constantius II. Due to this and to his Christian faith, some historians consider him to have been opposed to Julian, or at least to some of his policies. But Julian’s acts seem to indicate otherwise, as he raised him to the rank of comes and to the post of magister equitum in Constantinople before the start of the campaign. He was also of “barbarian” origins (Sarmatian, in this case) and he also had a long and successful military career under Julian’s successors. Flavius Arinthaeus was a similar case to Victor. His rank was also that of comes. He also would enjoy a successful career under Julian’s successors. He was made magister peditum by Valens, and in 372 CE he was appointed as consul.
Notice also that the army was marching along the eastern bank of the Euphrates, clearly in order to avoid having to stage a crossing later in the face of possible enemy resistance, although this also forced the army to guard its open eastern flank, which would have exposed to attacks by the Sasanian cavalry.
The army passed in front of the ruins of Dura Europos (which stood on the opposite bank of the Euphrates) on 6 April. The army then reached the first Sasanian fortress on the river, Anatha (modern Anah in Iraq), which was captured on 11 April after a show of strength (Ammianus XXIV, 1, 6–10; Zosimus III, 14, 2–3; Libanius, Oration XVIII). Let’s review Ammianus’ narration of this event (Res Gestae Libri XXXI, XXIV, 1, 6-10):
Anatha was a riverine fortress located on an island in the middle of the Euphrates, and thus it could have opposed resistance. That the fortress surrendered so quickly after Hormazd’s mission seems to indicate that there was willingness from the defenders to collaborate. Most scholars think that the garrison could’ve been supporters of Hormazd against his brother Šābuhr II, but Ilkka Syvänne offers (and quite unsupported) theory. He suggests that among the “numerous offspring” of the old Roman soldier who lived in Anatha could’ve been Pusaeus himself, the commander of the garrison. Syvänne only has very slight circumstantial “evidence” to support his theory: “Pusaeus” is not an Iranian name, so he must’ve been a local Aramaean speaker and that he was later allowed to reach command posts in the Roman army (although this means little or nothing; at the time the Roman army was stuffed with foreign soldiers and officers). Against it there’s the obvious fact that Ammianus says nothing of the sort despite going at length to tell the old soldier’s story.
Map of Julian’s invasion of Ērānšahr, according to Ilkka Syvänne (Military History of Late Rome, 361-395). Compare the several possible locations of the Naarmalcha canal; I will post other maps in this chapter.
Nevertheless, the garrison ran a huge risk with his desertion, because Ammianus tells us how did Šābuhr II deal with traitors: they were executed, along with their families. So, in this sense it’s logical that Julian resettled all of them with their families to Syria. The fact that they lived there with their families could imply also that they were local levies of limited combat value, and perhaps with a limited sense of loyalty towards the Sasanian monarchy; also as their families were with them they had little to fear in the way of reprisals from the Šahān Šāh.
Presumably while it was still at Anatha (Ammianus doesn’t specify) Julian’s army was hit by a “hurricane” (presumably a sandstorm) that caused serious disorder and tore apart many tents. And the very same day, another disaster struck (Res Gestae Libri XXXI, XXIV, 1, 11):
Both Harrel and Syvänne agree that this sudden flood was probably manmade, and that it could have been the Sasanian forces that manipulated the irrigation systems of the area to flood the way of the Roman army, something that Ammianus himself realized at the time. After these incidents, the army continued its advance and captured and burnt an unnamed city (Ammianus and the other sources offer no details).
On March 13 (presumably, according to Dodgeon and Lieu), the Roman army reached the Sasanian fortress of Thilutha (which Dodgeon and Lieu identify with modern Telbes, a rocky islet 14 km. south of Ana). According to Ammianus, this was an almost impregnable fortress (Res Gestae Libri XXXI, XXIV, 2, 1-2):
In other words, the Romans bypassed the fortress, and the defenders made no attempt at attacking the Roman fleet while it sailed downriver under their walls. The attitude of the Sasanian defenders allowing the Roman fleet to sail under their walls unmolested is noteworthy, but their promise to remain neutral until the end of the campaign sounds quite dubious. As Syvänne and Harrel noticed, in bypassing Thilutha, Julian was acting like Šābuhr II in his campaigns against Roman Mesopotamia in the 340s: he was advancing directly against the main enemy base (then Nisibis, now Ctesiphon) leaving enemy garrisons behind him directly on his supply route. And now (like in the 340s) this meant that if the invading army failed to conquer its objective before it ran out of supplies, it would have to live off the land or retreat in haste with its troops going hungry and being harassed by the enemy (which is exactly what happened in the end). I haven’t seen any scholar make a hypothesis about the possible motivations behind the behavior of Thilutha’s garrison. Considering as it wasn’t a city but a purely military fortress, it’s possible that it was garrisoned by professional troops and not by a local militia, and so they should have acted more aggressively. Syvänne puts forward the opinion that it was part of a deliberate tactic by the Sasanians to drag the Romans deep into Ērānšahr while denying them to maintain open supply routes. It’s a possibility worth considering but given the size of the Roman army and the effectiveness they showed at siege warfare (as we will see soon enough when the army reached the key Sasanian fortified city of Pērōz-Šābuhr).
This procedure was repeated at the next fortress the army encountered, Achaiacala. This is an unidentified place; Alois Musil referred to it in his 1927 work Arabia Deserta: A Topographical Itinerary as probably situated somewhere in the vicinity of modern Haditha, about 75 km downriver from Ana (or 45 km. in a straight line), which was also bypassed after the garrison “promised” Julian they wouldn’t intervene further in the war. Two days after this episode (it would’ve been mid-April according to Dodgeon and Lieu) the army crossed to the western bank of the Euphrates at a place that Ammianus called Baraxmalcha. Musil could not locate this place but he hypothesized it could be a corruption of the Arabic word for “ford” farad, which combined with the Arabic for “king” malek would mean “the royal crossing”. The army then entered the town of Diacira which had been abandoned by its inhabitants. Diacira was identified by Musil as modern Hit in Iraq, as he translated the name as “bitumen-giving”. After this, according to Dodgeon and Lieu the army passed by Sitha (neither Dodgeon and Lieu nor Musil offer any possible locations for this place) and must’ve then crossed back to the eastern bank of the Euphrates at Megia (these places only appear in Zosimus’ account, not in Ammianus’, and neither Dodgeon and Lieu nor Musil offer any hypothesis about their possible location).
Around April 22, the army reached Ozogardana (named as Zaragardia in Zosimus’ account; identified by Musil with modern Sari-al-Hadd in Iraq); where the army rested for two days. Here, the Romans first encountered the cavalry force lead by Sūrēn who would shadow them constantly from this moment on; the renegade Sasanian prince Hormazd led a foraging/reconnaissance party that was ambushed by the cavalry force of Sūrēn and the Arab phylarch Podosaces, as narrated by Ammianus and Zosimus (Res Gestae Libri XXXI, XXIV, 2, 4-5):
Ammianus’ claim that this was an ambush seems improbable, because from his description of the encounter it seems clear that the Sasanian force was located across the Euphrates at a place where the river could not be forded, and so this would have been a very inept attempt at an ambush indeed. As Syvänne noted, it seems more probable that this was a fortuitous encounter. But it was to foreshadow things to come; from now on, the Roman column would need to be constantly on guard against surprise attacks, and foraging parties would need to be reinforced if they didn’t want to be annihilated by Sūrēn’s force, which as we will see, would show itself consistently superior to the Roman cavalry. In addition to the difficulties for foraging, this would also mean that from this moment on, Julian’s army would advance without being able to conduct a proper reconnaissance of what lay in its way.
Thus, shadowed by enemy forces, the Roman army reached the town of Macepracta (thanks to the mention by Ammianus of the remains of a rampart at this location helped Musil to locate it near the beginning of an ancient rampart at modern Ummu-r-Rus in Iraq, which stretches from the north bank of the Euphrates northward as far as the Tigris), near the start of the Naarmalcha (also known as the “Royal Canal”) and crossed it to reach the heavily fortified city of Pirisabora. The crossing of the canal was not without difficulties (Res Gestae Libri XXXI, XXIV, 2, 6-8):
Here we see how the Roman cavalry failed in the attempt to force the crossing of the Naarmalcha, but the day was saved by the Roman auxilia, probably Julian’s veterans from the Army of Gaul, who were very adept in these sort of missions, as can be seen in Ammianus’ account of Julian’s campaign in Gaul. Pirisabora has been identified easily by modern scholars with Pērōz-Šābuhr, the ancient Misiche, which Šābuhr I renamed after is victory over Gordian I’s army a century earlier. After the Arab conquest of Mesopotamia, it was renamed as Anbar and under the Umayyads and Abbasids it still retained its importance as a main military base and one of the main arsenals of the Caliphs’ armies, under the name of al-Anbar. The ancient city has not been excavated because it lies today under the waters of the Euphrates. It was of essential importance to Julian’s invasion because it controlled the entry to the Naarmalcha from the Euphrates, and controlling it was necessary if the fleet was to follow the army to Ctesiphon. The city was taken by assault after a short and violent siege by the Roman army, as stated by Ammianus, Zosimus, Libanius and the Liber Suda. Ammianus’ account of the events (Res Gestae Libri XXXI, XXIV, 2, 9-22):
The fall of Pērōz-Šābuhr was an important blow to the Sasanian defensive system, as according to Ammianus it controlled access to the Naarmalcha (meaning “the Royal Canal”) canal. The exact location and course of the Naarmalcha has been the subject of quite a lot of confusion and misgivings even in modern accounts of the campaign, in part because the ancient sources themselves offer contradictory information about this watercourse. While writing these chapters about the events in the Late Antique Near East I check whenever possible the places and locations where events took place in the Barrington Atlas of the Greek and Roman World, which is perhaps the most complete compilation published of the Classical world geography-wise. And when I reached this point in the campaign, I noticed that this scholarly work situated the start of the Naarmalcha a long wat downriver from Pērōz-Šābuhr, which goes against all the accounts of Julian’s campaign; this surprised me. And then I checked other modern resources and noticed a similar state of confusion (the Digital Atlas of the Roman Empire, an online resource supported by the University of Lund in Sweden, does not even draw it clearly on the ground).
The Naarmalcha canal, according to François Paschoud.
Finally, I found a paper that seemed to settle the subject, in a clear and precise way. And once more to my surprise though, this was not exactly a new research at all. In 1978, the French scholar François Paschoud wrote a detailed paper about this ancient canal which pretty much solved my doubts about its location and history (Le Naarmalcha: à propos du tracé d’un canal en Mésopotamie Moyenne). I’ll be following Paschoud’s paper here. basically, the Naarmalcha was a canal that took water from the Euphrates and the flowed on a south-east direction, following the natural incline of the middle Mesopotamian plain. Its main function was to act as a irrigation channel, but it also allowed the navigation by riverine vessels. According to Paschoud, its main branch did not reach the Tigris; after reaching its environs some kilometers south of Seleucia-Ctesiphon, it followed roughly a parallel course to the Tigris until all its water was distributed in lateral irrigation channels. But according to Paschoud, when comparing the ancient sources (Ammianus, Zosimus, Libanius and Cassius Dio -when describing Trajan’s 116-117 CE Parthian campaign-) in this last sector of the Naarmalcha there was a side channel that connected the Naarmalcha with the Tigris upstream from Ctesiphon and Vēh-Ardaxšīr (this is important, because for example it disagrees with Harrel’s reconstruction of the campaign, but not with Syvänne’s) which was known to Graeco-Roman authors as “Trajan’s Canal”, despite the fact that Cassius Dio makes clear that Trajan found this canal already in existence, and so it wasn’t built by him). We’ll return to this point later.
It’s noteworthy how the Roman army managed to storm so quickly the walls of the city; the mention of a helepolis (perhaps the most elaborate and costly siege engine in Antiquity) makes clear that Julian army was well equipped for sieges, and that Julian and/or his generals knew how to use them. Nevertheless, it’s clear that by now Julian (used to his “commando-like” operations in Gaul and in Germanic lands) was getting impatient, as he chose to arrange a negotiated surrender with the defenders of the citadel. About the issue of the surrender of Pērōz-Šābuhr though there’s a very important difference between the accounts of Ammianus (see above) and Zosimus (New History, 3.18.4):
First, Zosimus says there were five thousand defenders, not two thousand and five hundred. And he adds the very important details that Julian actually bought them off, which was always a very unpopular measure with Roman soldiers and officers. in order to resort to it, Julian must have thought that time was really of the essence. Possibly because of this, Julian thought it to be advisable to distribute money among his troops (Zosimus, New History, 3.18.5-7):
According to Syvänne, Julian’s profligacy would have serious consequences later on. The distribution of money among the Roman soldiers is also noted by Ammianus, although in less favorable terms to the emperor (Res Gestae Libri XXXI, XXIV, 3, 3):
On the day after the fall of Pērōz-Šābuhr, a Roman reconnaissance party was ambushed and routed by the cavalry led by Sūrēn. Its commanders were relieved of their command by Julian (Ammianus XXIV, 3, 1–2, Zosimus III, 19,1–2 and Libanius, oration XVIII, 229). The army marched along the canal past Phissenia (identified by Musil with modern ‘Akar an-Na Jeli, lying about two kilometers from the Euphrates on the left bank of the canal) but the advance was slowed by the Sasanians who breached the dams and turned the low ground into marshland (Ammianus XXIV, 3, 10–11, Zosimus III, 19, 3–421 and Libanius, Oration XVIII, 222–6 and 232–4). Ammianus is quite confusing in this part of his account, because he doesn’t state clearly where did the army start to follow the Naarmalcha towards Ctesiphon. As we’ve seen above, he explicitly states it bifurcated from the Euphrates at Pērōz-Šābuhr, but later on he writes as if the army kept on following the Euphrates downriver from this city, hence the confusion among scholars.
At this point the Euphrates was divided into many small streams (Ammianus XXIV, 3, 14). In this district they found an abandoned Jewish settlement (Ammianus XXIV, 4, 1). The Jewish settlement was probably Nehardea, the great center of Jewish rabbinical learning in “Babylonia”.
Julian headed the bridging operation and led his army to Bithra (a problematic location; Musil identified it tentatively with Ibrahim al-Halil, about 25 km. south-east of Phissenia) where they found the ruins of a palace (Zosimus III, 19, 49). Maiozamalcha (The name in Aramaic means “royal fort” or “royal port”, so called because of its position on the Naarmalcha. Musil noticed a ruin mound -modern Han az-Zad- about 18 km. west of Ctesiphon; The “ma’oz” part of the name could be perhaps derived from “Mahoza”, the name given by Jews and Christians to Vēh-Ardaxšīr), a well-fortified city, was taken after a laborious siege (probably on 10–13 May 363 CE) as related in Ammianus XXIV, 4, 2–30, Zosimus III, 20, 2–22, 7 and Libanius, Orations XVIII, 235–42. The storming of Maiozamalcha was another outstanding achievement by the Roman army in siege warfare (Res Gestae Libri XXXI, XXIV, 4, 2-30):
I’d like to comment on two points of Ammianus’ account. First, that obviously the Roman army had launched its invasion very well provided with siege machinery and that like at Pērōz-Šābuhr it was employed here to devastating effect (even better here, perhaps because of the lack of a citadel, which is strange, as it’s a common feature in practically all excavated Sasanian cities, whether in the eastern or the western part of the empire). The second point is an ominous one for the Romans: Ammianus mentions that Sūrēn’s cavalry attacked the pack animals of the army. As we will see, this would later have consequences, and I’m inclined to agree with Syvänne who thinks this attack was part of a planned strategy by the Sasanian defenders.
Map of Sasanian Iraq (provinces of Āsūrestān and Nodšēragān in 363 CE). The irrigated plains began south of the Naarmalcha canal.
Yeah, kinda breaks in the middle of the story
Granted that you spoiled the ending so it's not a real cliffhanger. What struck me was how much better the Romans seemed to be at siege warfare (defensive and offensive)... Who would they practice it on? Each other?
Granted that you spoiled the ending so it's not a real cliffhanger. What struck me was how much better the Romans seemed to be at siege warfare (defensive and offensive)... Who would they practice it on? Each other?
Yeah, kinda breaks in the middle of the story
Granted that you spoiled the ending so it's not a real cliffhanger. What struck me was how much better the Romans seemed to be at siege warfare (defensive and offensive)... Who would they practice it on? Each other?
I had to break it somewhere, and it was a real headache where to do it. Finally I decided that the arrival of Julian's army would have been a good place, from a nartive point of view. Following Ammianus' account is a real pain in the ass, because he writes atrociously, and his account has important contradictions. Thus the need to compare it with Libanius' and Zosimus' accounts. Zosimus' writing is thankfully more concise and clear, as he's much less prone to rhetorical florileges, but he wrote removed almost 150 years from the events. Libanius is perhaps the most reliable source, fact-wise, but he was a rhetor, and his texts are unbearably rhetorical for a modern reader. Plus, all three were admirers of Julian, so caution is advisable when reading them (curiously, Ammianus is perhaps the most impartial of the trio). Critical reading and resorting to notices from Christian authors (all of them markedly anti-Julian) is of the essence in order to get a balanced image.
Yeah, kinda breaks in the middle of the story
Granted that you spoiled the ending so it's not a real cliffhanger. What struck me was how much better the Romans seemed to be at siege warfare (defensive and offensive)... Who would they practice it on? Each other?
As for the Romans and siege warfare; let's not forget they bypassed two Sasanian fortresses (Anatha and Thilutha) as they were fortified islands and taking them would've been very costly. This implies that, just as the Sasanians had done in the 340s, Julian was racing towards his main objective (Ctesiphon) leaving enemy garrisons behind them, which meant that he could only rely on the supplies he carried on his fleet; there were no open supply routes to his rearguard. And at Pirisabora, according to Zosimus, he basically bribed the defenders, a risky move considering that already before crossing into Sasanian territory there were symptoms of discontent in the army (he already had to distribute a donativum after crossing the Khabur).
Great thread, thanks.
It seems to me that you've been somewhat sniffy about the quality of Ammianus Marcellinius. Which is interesting, because I know Gibbon rates him quite highly as a source. Care to comment a bit on that?
It seems to me that you've been somewhat sniffy about the quality of Ammianus Marcellinius. Which is interesting, because I know Gibbon rates him quite highly as a source. Care to comment a bit on that?
Great thread, thanks.
It seems to me that you've been somewhat sniffy about the quality of Ammianus Marcellinius. Which is interesting, because I know Gibbon rates him quite highly as a source. Care to comment a bit on that?
Purely as a source, he's invaluable, as he was an officer and took part personally in the events. But that doesn't mean he doesn't have his faults:
- Scholars consider he wrote his Res Gestae in Rome after he retired from the army, and that it wasn't finished/published before 391 CE. That means that these are the memoirs of an old man writing from memory, and at certain points, it shows. He also wrote it for a very specific public: the pagan clique of the Roman Senate led by the likes of Nichomachus Flavianus and Q. Aurelius Symmachus during the reigns of Gratian, Valentinian II and Theodosius II. These traditionalist senators were frontally opposed to the increasingly anti-pagan measures of these emperors, and obviously Ammianus had to be diplomatic towards their political bias.
- Ammianus was not a member of the elite of the empire, unlike other ancient writers (both Pliniii, Cassius Dio, Festus, etc.) which means that when he talks about political decisions taken by emperors and ministers, or strategic movements decided by the emperor and the top brass, he's writing as an outsider, and we don't know his sources for that. So we must assume he wrote from hearsay.
- Although he's usually considered to have been a member of a rich family from Antioch, we really know nothing about his background, other than that he was an officer in the Roman army, and that he served mostly in the East, under the patronage of Ursicinus. But scholars nowadays consider he can't have been a member of a very wealthy family, for two reasons: he doesn't say so (and in the Roman society, this was something that one needed to brag about) and his writing style is quite plain and poor compared for example to Libanius; which means he didn't have access to the sort of exclusive education that was offered to the scions of the very wealthy. This could have resulted in a more readable text for modern readers, without so many literary references, but the problem is that Ammianus is simply a very poor writer. He repeats himself, he contradicts himself at times and he tries repeatedly to insert classizicing literary topoi or strange tales about the most bizarre subjects in the most unexpected places, cutting the pace of the narrative. It makes for a very heavy reading at times. Perhaps this is influenced by the fact Ammianus was not writing in his native language (he was a native Greek speaker, but he wrote in Latin).
- When considering his value as a source, there's something important to consider: he was not an impartial witness (no ancient historian was). He was an admirer of Julian and although he tried to keep a modicum of impartiality, he rarely succeeded (same as with Ursicinus). Gibbon appreciated him so much because Gibbon also admired Julian (he considered Christianity to have been a major factor in the decline of the empire), but ironically, although Ammianus was a pagan, he was not overly anti-Christian in his work, and avoided showing any sectarian bias.
Last edited: