Ancient History & Civilisation

Chapter 6

Disaster at Carrhae (53 BC)

In order to understand the course of the battle and the tactics used by both sides, we need to first analyse the armies and assess their strengths and weaknesses.

The Roman Army at the Battle of Carrhae

The first issue we need to consider is the size of the Roman force, and here the accounts vary. Once again we are faced with the fact that we have no contemporary source for this information. Appian has by far and away the greatest figure when he quotes Crassus’ army as 100,000 strong.187Such an army had not been seen since the days of Hannibal and would never have been raised for such a campaign. Again we must turn to Plutarch (and his unknown source) for a more realistic figure. Plutarch informs us that Crassus crossed into Mesopotamia in 53 BC with an army of seven legions of infantry, four thousand horsemen (of which 1,000 were Gallic and the rest native auxiliaries) and an equivalent number of auxiliary troops.188 If we follow the standard estimates that each of Crassus’ legions was roughly 4,800 men strong, then we have a figure of just under 34,000 legionaries.189 Add the 4,000 cavalry and 4,000 auxiliary infantry and we have a total of some 42,000 men.190

There are several problems with taking this figure as an exact one. Prior to the Imperial era, the size of the legion was not an absolute and we know that Crassus had problems recruiting legionaries, so he may not have been able to fill seven whole legions. Added to this is the rough nature of Plutarch’s calculation of the number of auxiliary infantry. Thus we are working with a rough estimate of 38,000 infantry (split between legionaries and auxiliaries; a difference which will be explored below) and 3,000–4,000 cavalry (of which only 1,000 were Gallic).

These numbers do not represent a homogeneous body of men. Of this figure, 34,000 were full Roman legionaries. These legionaries were the elite infantry of Crassus’ army, armed with javelins (pila) and short sword (gladius), with shields, helmets and chest armour for protection. In close order combat, the Roman legionary had proved to be superior to any other infantry in the ancient world. As detailed earlier, they had defeated the Macedonian phalanx and the Armenian foot-soldier. However, this did not mean that they were without weaknesses. For the legionaries to be at their most effective, the battle would have to be fought at closequarters, where the short Roman sword would be most effective. Aside from the javelin, the standard Roman legionary had little in the way of distance weaponry. In terms of defence, the helmet, shield and chest armour were again effective defence at close quarters, but this still left much of the body undefended and vulnerable to weapons of range.

Aside from weaponry and armour, we must also examine the nature of their training and ability. On the whole it appears that the bulk of Crassus’ legionaries were raw recruits in 55 BC, along with a smattering of experienced legionaries (most probably distributed in the junior NCO ranks of the legion, such as the centurions). The bulk of the men would not have seen a major battle before. Nevertheless, too much can be made of the supposed inexperience of these men. They had the autumn, winter and spring of 54–53 BC in which to be trained and they had been blooded in battle in 54 BC, when they defeated the Parthian satrap, Silaces. Given Crassus’ previous focus on his men’s training and an unwillingness to give battle unless he had total confidence in their abilities (as seen in the Spartacus campaign), we can safely assume that they were up to the expected Roman standard.

The other section of Crassus’ infantry, however, was composed of native auxiliaries. In the case of auxiliary forces there were no strict rules as to their composition, numbers, or weaponry, as it depended entirely upon where they were raised; which in this case we don’t know. It is probable that they were raised from the Roman territories in the east and the Roman allies of the region. This would give them experience of the region and local warfare, but as to their weaponry and armour, we can only speculate. It is likely that they were lightly armoured and possessed a mixture of spears, swords and light bows. We are told at one point that there were at least 500 native archers in the army.191 Certainly they would not have been able to match the Roman legionaries in either offensive or defensive capabilities. Nevertheless, such a mixture and balance was typical for Roman armies of the period and would have mirrored the armies of Lucullus and Pompey, and thus been more than a match for the armies that they were expecting to encounter in the region.

If there was a weakness in Crassus’ army, then it lay in his cavalry. Roman armies of the period rarely had large numbers of cavalry and Crassus’ army was no exception. It appears that he took no cavalry with him from Italy. Of his 4,000 cavalry, just 1,000 were non-native and these were the Gallic cavalry loaned by Julius Caesar. The Gallic cavalry are described by Plutarch as being lightly equipped with short spears and having little armour.192 This compared badly to the Parthian heavily-armoured cataphract. Of the remaining 3,000 native cavalry we are not given any detail, but the assumption is that these too were light cavalry rather than heavily-armoured ones, given the criticism of the sources. Of either group’s training or experience we know nothing, though we must again assume that they would have been brought up to scratch by Crassus and his son during the winter months.

This brings us onto another topic that needs examining before we progress, namely the quality of the Roman commanders. We have already looked at Crassus himself, but one aspect that is rarely commented on is the nature and quality of his junior officers. First and foremost were his two deputies, Publius Licinius Crassus and Gaius Cassius Longinus. Publius Crassus (Crassus’ youngest son) appears to us in the sources as being everything that his father was not. Cicero, eight years later, describes him to Julius Caesar thus:

Out of all our nobility, the young man for whom I had the highest regard was Publius Crassus; and while I had entertained great hopes of him from his earliest years, I began to have quite a brilliant impression of him when the highly favourable opinions you [Caesar] had formed of him became known to me193

and elsewhere says:

Publius Crassus, son of Marcus, who at an early age sought the circle of my friendship, and I exhorted him with all my power to follow that straight path to renown which his ancestors had trodden and made smooth for him. For he had enjoyed excellent upbringing and had received a thorough and complete training. His mind was good, if not brilliant, his language choice abundant, and in addition he had dignity without arrogance and modesty without sloth.194

These refrences of Cicero’s regarding Publius Crassus are two out of just five he makes to the Battle of Carrhae in total, throughout all his extant works (the other three being comments on the supposed ill omens that occurred). As well as impressing Cicero, Publius served under Julius Caesar in Gaul, where in 57–56 BC he distinguished himself as a legionary commander in Aquitania.195 Thus he appears to us from the sources (most of which are hostile to his father) as being a model Roman aristocrat; brave in battle, yet modest about it. In our surviving sources, and amongst the Roman aristocracy, especially Caesar and Cicero, it is his loss at Carrhae that is felt more keenly than that of his father.196

Yet, Publius Crassus appears to be typical of the type of officer that Marcus Crassus took on this campaign. As he had done all through his political life, and as he clearly showed during his Spartacus campaign, Crassus cultivated the best of the young Roman aristocrats; this time by giving them positions on the general staff of this supposedly glorious and profitable campaign. As well as Publius, we are given a host of names of aspiring young Roman aristocrats, such as repre-sentatives of the distinguished families of the Marcii Censorini, Octavii, Petronii, Roscii and the Vargunetii.

Added to these names is that of Gaius Cassius Longinus, who served as Crassus’ quaestor (official deputy) during this campaign. Cassius was later to achieve immortality as one of the two leaders of the conspirators that assassinated Julius Caesar in the Roman Senate house in 44 BC (the other being Brutus). This campaign is the first time that we hear of young Cassius, but his role is a significant one. Plutarch’s account of the whole campaign places Cassius at the centre of events, always urging Crassus not to follow what turns out to be the wrong, and often disastrous, course of action. Given the later blackening of Cassius’ name (due to his role in Caesar’ assassination) this is highly curious (see appendix two on the possible sources for this anomaly). Of the three main commanders, Crassus, his son, and Cassius, only the latter survived to tell the tale, which makes any account he gave, including his heroic role, questionable to say the least. Nevertheless he does appear to have been yet another young and talented Roman commander.

Therefore, we can see that Crassus, regardless of later sources’ views on his own abilities as a commander, undeniably had a talented and energetic command staff surrounding him. Regarding his army, though, a closer examination of their composition does reveal a number of potential flaws and weaknesses. Nevertheless, this was still a powerful Roman army and one which, on past form, was widely expected to replicate the results of the armies of Lucullus and Pompey in fighting the armies of the east. In order to understand the reason that they failed so spectacularly we must now turn our attention to the Parthian army of Surenas.

The Parthian Army at the Battle of Carrhae

Not only do we have fewer descriptions of the Parthian army at Carrhae than of the Romans, but the issue is further clouded by some noticeable differences between Parthian armies in general and the one which Surenas fielded at Carrhae, differences that hold a key significance.

Dio (writing in the third century AD) provides us with our best general description of the Parthian military and it is with him that we should start:

But I will describe their equipment of arms and their method of warfare; for the examination of these details properly concerns the present narrative, since it has come to a point where this knowledge is needed. The Parthians make no use of a shield, but their forces consist of mounted archers and lancers, mostly in full armour. Their infantry is small, made up of the weaker men; but even these are all archers. They practise from boyhood and the climate and the land combine to aid both the horsemanship and archery.197

Justin, an even later Roman source, gives us the following description of the composition of the Parthian army:

They have an army, not like other nations, of free men, but chiefly consisting of slaves, the numbers of whom daily increase, the power of manumission [the freeing of slaves] being allowed to none, and all their offspring, in consequence, being born slaves. These bondsmen they bring up as carefully as their own children, and teach them, with great pains, the art of riding and shooting with the bow.198

He then elaborates upon their tactics:

Of engaging with the enemy in close fight, and of taking cities by siege they know nothing. They fight on horseback, either galloping forward or turning their backs. Often too they counterfeit flight that they may throw their pursuers off their guard against being wounded by their arrows. The signal for battle among them is given, not by trumpet, but by drum.199

And gives this detail of their armour:

Their armour, and that of their horses, is formed of plates, lapping over one another like the feathers of a bird, and covers both man and horse entirely.200

Lucian, a second century source tells us that the Parthians fought in units of 1,000 known as ‘dragons’, due to the symbol they fought under.201

From these later descriptions it is possible to create an image of a generic Parthian army from this period, which would be composed of three types of fighting man. The elite of the army, most probably the noble or free men, would be the heavily-armed cavalrymen, known as cataphracts. Then there would be the lightly-armed horse archers and the light infantrymen, armed with bows. Both of the latter two categories would be serfs, taken from the estates of the nobility.

Surenas awaited the Roman army at Carrhae with a force composed of just 10,000 men, which would be ten dragons (if we accept Lucian’s’ definition of a basic Parthian unit). Of these there were apparently 1,000 cataphracts, 9,000 horse archers and no infantry. All of these men came from Surenas’ own estates. In addition, Plutarch furnishes us with one crucial detail, namely that there were 1,000 baggage camels laden with spare arrows.202 It is these last two facts that mark Surenas’ army out from a standard Parthian army of the era, and we need to understand both their cause and their effect.

The lack of infantry has rarely been commented upon and, when it is, it is usually dismissed as being a side effect of Orodes taking the bulk of the army into Armenia.203 Yet the Parthians had no single standing army as such. Each landowner was responsible for raising troops and supplying them to the king. In Surenas’ case, he raised and fought with his own army, manned from his own family estates in eastern Parthia. It is unlikely that he would have split this army and even if he had, then why would the king take all of his infantry? To my mind the lack of infantry is not a passing detail or a side effect of the army’s division. It is far more logical to see that the army that Surenas put into the field to fight Crassus in 53 BCwas deliberately created without any role for infantry.

Surenas had a year to study the Roman method of warfare and could consult Silaces, the defeated satrap of Mesopotamia, for first hand experience of how they fought. As the Romans had demonstrated time and again, in close order fighting they were virtually invincible. The Armenians, who fought in a similar style to the standard Parthian manner, had met with heavy defeat in 69 BC. Given everything we know about Surenas, it is clear that he would have been well aware that Orodes was intending to sacrifice him to slow down the Romans by letting him face them first, and it is equally clear that he would not meekly wait for his supposedly ‘inevitable’ destruction. It is obvious that Surenas did not meet the Romans in battle blindly, but had worked out a strategy that he hoped would bring him victory. To accomplish this he needed to avoid playing to the Roman strengths, whilst utilising those of his own army. In this case, the Roman strength was close-quarter infantry fighting, whilst his army’s were speed and long-range weapons.

Therefore, it appears that Surenas spent the winter months modifying the standard Parthian army and way of fighting into a force capable of defeating a Roman army. One key element of this plan would be the complete lack of infantry, with his whole army being composed of nothing but cavalry. Thus his army would be able to engage the Romans at speed and avoid getting entangled with the legionaries on the ground.

However, whilst the lack of footsoldiers would allow him to avoid getting entangled in a close-quarter battle, this alone would not bring him victory. Disposing of the infantry element of his army was nothing more than removing a negative aspect from his force. Of his remaining force of 10,000 the majority were lightly-armoured horse archers, who on the face of it would never be able to defeat an infantry army on their own, as they traditionally had one key flaw; once they had emptied their quiver of arrows then they would be useless at a distance and would have to attack the Romans at close quarters, for which they were not armed or armoured. It is here that Surenas introduced the key element of his battleplan and one which (as far as we can tell) was unique to him. This is of course the addition of the baggage train of 1,000 camels laden with tens of thousands of additional arrows. In addition, this baggage train would be at the front line, or just behind it, allowing the horse archers to re-arm at the battlefront, rather than having to ride to the back of the army, dismount, re-arm and then return. The whole process could be done whilst still mounted, near the battle-line and would therefore take far less time.

There is one further element that was crucial to the success of this plan, namely the quality of the arrows themselves and the bows used to fire them. Here we are operating in the near-complete absence of any evidence for the type of arrow used at Carrhae. All we know is that they were barbed and completely penetrated the Roman shields and armour. Now this cannot be a coincidence, and raises two interesting aspects. The Parthians and Romans had never fought before, yet Surenas had total faith that his arrows would penetrate Roman armour. Furthermore the Romans had fought eastern armies before (the Seleucids, Pontines and Armenians), and never encountered the same problems with arrows that they did at Carrhae. The first issue can be answered with reference to Surenas’ attacks on the Roman garrisons during the winter of 54–53 BC, which would have had more to do with the Parthians testing of their arrows’ abilities on Roman armour, than a serious attempt to retake the towns. We might recall that Plutarch relayed the Roman soldiers’ claims that ‘strange missiles are the precursors of their appearance, which pierce through every obstacle’.204The strangeness of these arrows may be more than Plutarch’s dramatic turn of phrase and may well illustrate that the Romans had never encountered that particular type of arrow before. Certainly Surenas went into the battle well aware of the devastating capabilities of his arrows against Roman armour. However, we must not discount the contribution made by the Parthian compound bows either. As seen in the illustration of the horse archer (figure 15), the Parthians used a short compound bow, which must have given the arrows a tremendous velocity. We have little exact evidence for the bows, other than descriptions, and shorter bows were common throughout eastern armies. Nevertheless, it is clear that the combination of this short compound bow and the barbed arrows produced devastating results on this occasion and may well have been aunique combination.

Surenas’ army was fronted by one thousand cataphracts fully clad in heavy armour and armed with long lances, superficially resembling medieval knights and far superior to the Roman cavalry. These shock troops formed an advance guard for the 9,000 horse archers armed with the armour-penetrating arrows and supported by a thousand baggage camels, allowing for near instantaneous rearming on the move. Therefore, we can see that it was an army designed for fighting a battle at speed and at distance, which was just the type of fighting that did not suit the Romans.

Furthermore, Surenas’ tactics played to the strengths of his men in terms of upbringing. The horse archers were all serfs from his estate and would have all been trained in horseback archery from childhood. They would have been used to following and obeying their feudal lord from birth and would have had the winter to practise the new tactics that they had been given. In short, they were the perfect body of men to learn these new tactics and carry out their master’s modified version of Parthian warfare.

Thus the army that the Romans faced at Carrhae was not there as a consequence of chance, but had been designed with fighting them specifically in mind. It was not designed to fight a long campaign, but to defeat this particular Roman army in a battle. This army reflected the genius of its commander and showed the Parthian system of private armies and devolved commanders at its best. It is clear that Orodes would not have thought out or executed these tactics. The uniqueness of this force and its difference to the standard Parthian method of fighting gave Surenas another major edge in that Crassus was not expecting it. Surenas had taken the opportunity to study the Roman army and how it fought and had been given the time to modify his own force accordingly. As far as Crassus was concerned, the army that he would soon be facing would fight in exactly the same way as had the one the year before, and as the Armenians had a decade before (who after all had comprehensively defeated the Parthians themselves, a generation earlier). What he did not know is that Surenas had created a new and unique method of warfare, designed specifically to win the upcoming battle.

It is highly unlikely that Crassus would have been able to discover Surenas’ new tactics before it was too late. Even his scouts would not have been able to see much difference in Surenas’ army at a glance. They could report seeing little in the way of infantry, but not know that there were in fact none at all. They could report a baggage train, but then such things were common in armies; they would not have been able to tell that it contained nothing but arrows. To all intents and purposes it would have looked like the army that Crassus was expecting to face. The only warning sign he had were the soldiers’ stories of strange arrows raining down on them during the winter clashes, but whether he would have given them any greater significance is doubtful. When battle was joined, he would have been unaware of how truly unique a Parthian force he faced. Thus Surenas went into the battle knowing his enemies tactics, but not vice versa.

The Dio Variation of the Battle

Of the battle itself, we have two detailed descriptions from Plutarch and Dio; neither is contemporaneous and they differ in some important ways. Of the two, the more detailed and knowledgeable is Plutarch’s (see appendix two for the possible reasons why). In order to gain the full picture of events though, we must look at both accounts and the best place to start is with the shorter variant of Dio.

Dio’s version has Crassus’ army being led directly into the path of Surenas’ by the Arab traitor Abgarus (though Plutarch states that he had left Crassus’ army by this point205). In effect it is a classic ambush, with the Parthian army being concealed, awaiting the arrival of the Romans (though this account ignores any presence of Roman scouts). Dio states that this was accomplished by the Parthians hiding in dips and woods, despite the fact that there was no woodland in this area.

Nonetheless, when the Romans were led into this trap, the Parthian army revealed themselves, at which point Publius Crassus suddenly broke ranks and led his cavalry at the Parthian ranks, which then appeared to break, with Publius giving chase. This however was a feint (which was an old tactic even in this century) and when they had led Publius away from the main army, the Parthians turned, surrounded and annihilated him.

This concluded Dio’s first phase of the battle. The second phase commenced with what is described as an almost suicidal charge by the Roman infantry who did so, according to Dio, ‘to avenge his [Publius Crassus’] death’.206 The Roman infantry were then devastated by the Parthian cataphracts, whose heavy lances broke the Roman ranks. Again Dio takes a scathing line on the Roman troops when he states that ‘many died from fright at the very charge of the lancers’.207 With their lines broken, the Roman soldiers were then slaughtered by the Parthian archers.

The final defeat came in the third phase, which began with the final treachery of Abgarus, who not only led the Romans into this ambush, but at the appropriate point apparently turned his allied forces (which are presumed, but not mentioned prior to this point) against the Roman lines, attacking them from the rear. The Romans, apparently unable to face two enemies at once, then turned their line around and exposed themselves to a Parthian attack from the rear.

for Abgarus did not immediately make his attempt upon them. But when he too attacked, thereupon the Osroeni themselves assailed the Romans on their exposed rear, since they were facing the other way, and also rendered them easier for the others to slaughter.208

Dio then concludes this brisk battle description with a wonderfully dramatic picture of the Roman plight:

And the Romans would have perished utterly, but for the fact that some of the lances of the barbarians were bent and others were broken, while the bowstrings snapped under the constant shooting, the missiles were exhausted, the swords all blunted and most of all, that the men themselves grew weary of the slaughter.209

Dio would therefore ask us to believe that the Parthians ran out of weapons and ammunition (in his account there is no mention of Surenas’ ammunition train) and then decided to take it easy and have mercy on the Romans, who they had grown tired of killing. It is not this aspect of his account that we find hard to believe. Dio’s account is a catalogue of staggering incompetence and failures on the Roman part.

Firstly, Marcus Crassus walks the Roman army into an ambush, led along by Abgarus. Then Publius Crassus breaks with all known Roman discipline, not to mention common sense, and races off to attack the Parthians on his own and is slaughtered. Third, we have the Roman infantry rushing headlong into attacking the Parthian army, seemingly for no better reason than revenge. Fourth, we have the Romans being taken completely unawares by the treacherous attack of Abgarus’ allied soldiers. Fifth, the Romans were seemingly unable to fight on two fronts and managed to get themselves twisted and turned around until they did not know which way they were facing. Marcus Crassus’ role in this sequence of errors is unclear, for we hear nothing more of him once he has led his men into the trap.

Aside from the catalogue of Roman failings, Dio’ account is short, devoid of any clear detail, and introduces a number of new elements which we do not find in any earlier source. They range from the significant (the treachery of the Arab allied contingent), to the bizarre (Surenas hiding his army in the woods – on a dusty north Mesopotamian plain).210 From start to finish, this battle narrative was designed to show the incompetence of the Roman army and especially it’s leadership, in the form of the Crassi. Actually, the Parthians do not come out of this narrative particularly well either. It seems that they won through a mixture of underhand tactics, treachery, ambushes and feints, combined with Roman ineptitude. Given the poor state of the Parthian Empire in his own day (third century AD), this is not perhaps surprising, but as an historical record it leaves much to be desired.

If we are to find out how the Roman Republic met such a catastrophic defeat in the east, then we need to turn to Plutarch, who presents us with a more detailed and logical sequence of events, which appear to have been based on a source with first-hand experience of the battle itself.

The Initial Clash

Throughout his account, Plutarch presents us with a far more realistic depiction of the Battle of Carrhae, and it is this one that we must accept as being the closest to the true sequence of events, as far as can be determined.

Rather than walking into a trap, Plutarch tells us that Crassus had sent his scouts out looking for Surenas’ army. By mid-afternoon, just beyond the river Belikh, they found what they were looking for. Given that Surenas’ battleplan was based on a significant element of misinformation, not in terms of location, but in terms of his army’s unusual formation and potential method of attack, it is no surprise that his own advance guard inflicted heavy casualties on the Roman scouts.211The fact that some survived to report their presence is also not a surprise as Surenas’ plan involved the Romans advancing onto his chosen ground.

Here we can see both the brilliance of Surenas as a tactician, and where Dio gets at least one of his oddest pieces of information from. Plutarch reports that Surenas had concealed the bulk of his army behind an advance guard. Therefore, an approaching force would only see the front of the army, in its width, rather than its depth. Thus Surenas concealed the bulk of his army from Crassus until battle was engaged, but not in the bizarre method that Dio states. Plutarch tells us that;

the enemy came in sight, who, to the surprise of the Romans, appeared to be neither numerous nor formidable.212

Furthermore, Surenas had ordered his heavily-armoured cataphracts to wear concealing robes and skins over their armour, in order to disguise their true nature. To an observer they would appear to be ordinary cavalrymen, rather than cataphracts. Surenas’ plan was obviously to lure Crassus into battle before he knew the number and type of force he was truly facing. It is at this point that Crassus made a decision that with hindsight may have proved to be a mistake. Plutarch reports that when the Parthians were located nearby, the Roman officers wanted to camp and give battle at day break. It is possible that this break would have allowed the Romans time to scout out the Parthians more thoroughly and therefore discover that the army which they were about to face was not a typical Parthian one. Crassus, however, wanted to push on immediately and Plutarch states that he was urged on by his son Publius, who was eager for battle.213 It is obviously this statement that led Dio into making his claim that Publius Crassus broke away from the army at the beginning of the battle and launched himself at the Parthians.

Even if Crassus had camped for the night and attempted to scout the Parthian army, there is nothing to indicate that they would have been any more successful than their predecessors, who had been dispatched with heavy casualties (a process made easier by the massed Parthian archers). All that a further scouting mission would have been able to tell Crassus is a rough estimate of the numbers, which would give the Romans a clear four to one advantage, and that the majority of them were mounted. They would not have been able to tell him how many were cataphracts (he would have been expecting a number of them anyway), nor that the baggage train of camels actually contained a large number of spare arrows, nor that there were no infantry. When Crassus advanced upon the waiting Parthians, he did so in full confidence that his army would easily outmatch the supposedly-inferior Parthian army (both in numbers and type). He had no reason to believe that he was in fact playing right into the hands of Surenas, who had chosen his ground – mostly flat with little cover, ideal for a fully mobile attack – and had concealed his true tactics.

Plutarch also gives us the Roman formation as they advanced upon the Parthians. At first Crassus adopted a linear formation with his army strung out across the plain in a long line and his cavalry divided between the two wings. Crassus commanded this formation from the centre, with the two wings commanded by Cassius and Publius Crassus. Plutarch tells us that he did this in order to avoid being surrounded by the enemy and that it was Cassius’ idea; the implication here being that if Crassus had stuck to this formation then the Parthians would not have been able to ride around the army and attack them from many sides. 214 Quite why he was expecting them to do this at such an early stage we are not told.

However, Plutarch then tells us that Crassus altered this formation and advanced upon the Parthians in a square formation:

Then he changed his mind and concentrated his men, forming them in a hollow square of four fronts, with twelve cohorts on each side.215 With each cohort he placed a squadron of horse, that no part of the line might lack cavalry support, but that the whole body might advance to the attack with equal protection everywhere.216

Plutarch does not give us the reasons why Crassus changed his tactics. In fact the whole passage is an odd one. Plutarch (or his sources) is attempting to alert us to the fact that he believed that Cassius’ formation was the best one and that by changing it Crassus made a mistake. We are told that Cassius’ formation would have prevented the Parthians from surrounding the army, but given that the Romans only had 4,000 cavalry, compared to the Parthians’ 10,000, this is an ambitious statement to say the least. Furthermore, Plutarch or his source are using hindsight here as prior to the battle no-one knew that the Parthians were going to surround the Roman army, as the Romans did not know the size of Surenas’ cavalry force or his tactics.

In fact there is nothing at all wrong with Crassus’ chosen formation, which as Plutarch states gave the Romans strength on all sides and would prevent an enemy from exploiting a weak area.217 As for why Crassus chose to ignore the advice of his vastly less-experienced junior officer (Cassius), we will probably never know, but it does perhaps show a greater degree of caution, for which he was known. The battle commenced with a thunderous wall of noise from the Parthians. Plutarch describes the scene well:

the signal was raised by their commander, first of all they filled the plain with the sound of a deep and terrifying roar. For the Parthians do not incite themselves to battle with horns or trumpets, but they have hollow drums of distended hide, covered with bronze bells, and on these they beat all at once in many quarters, and the instruments give forth a low and dismal tone, a blend of wild beast’s roar and harsh thunder peal. They had rightly judged that, of all the senses, hearing is the one most as to confound the soul, soonest rouses its emotions, and most effectively unseats the judgement.218

Utilising this battle cry to full effect, Surenas opted to begin the battle with a full-scale cavalry charge at the Roman army, with the cataphracts at the front, followed by his archers. Leading the charge himself, he then had his cataphracts remove the coverings which had been hiding their armour as they were galloping. This would have added to the dramatic effect of the charge, as their highly-polished bronze and steel armour would have caught the sun. The Romans would suddenly have realised that they were facing a full charge by heavilyarmoured cavalry. Surenas was clearly using every psychological trick he could to unnerve the enemy.



However, if he was hoping for the Roman line to break, either in panic or under the force of his heavy cavalry, then he was to be disappointed. For unlike in Dio’s account of the battle, the Roman line held strong. As they had been trained to do, the Romans soldiers locked their shields together and maintained their discipline and composure. We can see that in this respect Crassus had trained his army well. To maintain your discipline in the face of a cavalry charge was one thing, but given the added drama that Surenas had brought to this charge, it is a testament to the Roman discipline that they stood their ground.

This was incidental to Surenas’ plan; if the Roman line had broken then all the better, but it is doubtful that he ever believed it would do so. Rather than charge into the Roman line, Surenas actually diverted his cavalry around the Roman square, on both sides, until they had the Romans surrounded, taking the Romans by surprise. Crassus, however, soon recovered from this unusual tactic and, aware that he was being surrounded, ordered his auxiliary troops to charge at the Parthians and break their flanking manoeuvre. But they were met with a hail of arrows that forced them back into the square, taking heavy casualties in the process.

We can see that Surenas’ battleplan had worked beautifully thus far. Rather than attack the Romans head on and get involved in a static mêlée, which would have favoured his enemy, he encircled them at speed and deployed the bulk of his force, his 9,000 horse archers, to devastating effect. Now the Parthian archers began to unleash a barrage of arrows at the Romans from all sides. Given the penetrative capabilities of the arrows the Parthians were using, the Roman army was soon being slaughtered. Plutarch again captures the scene well,

But the Parthians now stood at long intervals from one another and began to shoot their arrows from all sides at once, not with any accurate aim, for the dense formation of the Romans would not suffer an archer to miss his man even if he wished it, but making vigorous and powerful shots from bows which were large and mighty and curved so as to discharge their missiles with great force. At once the plight of the Romans was a grievous one; for if they kept their ranks, they were wounded in great numbers, and if they tried to come to close quarters with the enemy they suffered just as much. For the Parthians shot as they fled and it is a very clever thing to seek safety while still fighting and to take away the shame of flight.219

Thus the Roman army, despite its numerical superiority, was trapped, huddled in a square and coming under a constant barrage of arrows. If the Romans moved to engage the archers, they would turn and retreat whilst still firing. The Roman soldiers could not get near enough to the archers to engage them in close combat. This tactic became known as the ‘Parthian shot’, the ability to still attack your opponents whilst retreating. Once Crassus had recovered from the initial shock of the Parthian tactics, however, he still had several reasons to be hopeful. Although his army was taking casualties, he must have sensed that if this was the best the Parthians could do, then he could still carry the day. The Parthian army seemed to be composed of nothing but horse archers, supported by a relatively low number of cataphracts. The Romans had already shown that they could withstand a full cavalry charge, the Parthians had no infantry, and once the archers ran out of arrows then the Romans could advance and force their retreat.

In this regard Crassus would normally have been quite correct. Under the usual terms of battle, the horse archers would soon have emptied their quivers and the Parthian cavalry would then have had to attack the Romans legions at close quarters (or withdraw). However, it is at this point that the true masterstroke of Surenas’ plan was brought into play – namely mobile re-arming. Having surrounded the Romans, Surenas deployed his camel train to replenish the archers. Thus the Parthian archers would only need a short break to ride up to one of the camels, take a fresh quiver of arrows, return to their positions and continue shooting. So long as the archers did this at slightly different times, and as long as the camels were well spaced amongst the surrounding archers, then the barrage would continue indefinitely.

It appears that Crassus soon became aware of this development. Perhaps he observed it actually happening, or he simply deduced that the rain of arrows was not weakening. Once he was aware of it though, he realised that his only hope now lay in breaking the encirclement. To that end, he sent a message to his son, out on one of the wings (we do not know which), ordering him to lead a breakout and engage the enemy at close quarters with his cavalry. If the Roman cavalry could drive off the Parthians, even in one area, then it would give the main army time to regroup. This breakout and the engagement that followed would deter-mine the outcome of the whole battle.

The Breakout and the ‘Battle within a Battle’

Publius Crassus gathered together as many troops as he could muster on his wing. Plutarch tells us that he had 1,300 cavalry (including his own 1,000 Gauls), 500 auxiliary archers and eight cohorts of legionaries (just under 4,000 men).220 Publius then led this force and charged the Parthian cavalry ahead of him. Plutarch also records that with him leading the charge were two young aristocratic friends of his, Censorinus and Megabacchus.221 At first it appeared that the plan had worked successfully as the Parthians appeared to break, turn and retreat. Not wanting to lose the initiative and sensing victory, Publius chased after the enemy, with both cavalry and infantry, hoping to finish the Parthians off.

Whether the Parthians on Publius’ wing did genuinely break or not, we will never know. Plutarch certainly raises it as a possibility.222 Publius’ charge would certainly have taken them by surprise and it was conducted with a large number of Roman and allied cavalry, backed up by archers and legionaries. Such a force was a formidable combination of speed, firepower and close-order infantry. However, the retreating Parthians wheeled their horses away from the main Roman army and towards their cataphracts. At that point the retreating Parthians turned, were joined by the cataphracts and attacked the oncoming Romans.

Whilst it appeared that the Romans still had the numerical advantage, and had a good mix of cavalry and foot, once again the Parthians adhered to the battle-plan of their master and placed the cataphracts between the Romans and their archers. This would have allowed the archers to continue to fire at the Romans as the two cavalry forces engaged each other, in the first, and only, close-order clash of the battle.

Although the Romans had the numerical advantage in this encounter, the Parthians had by far and away the advantage in terms of weaponry. The Roman cavalry were lightly armoured and only had short spears, whilst the Parthian cataphracts were heavily armoured and carried long lances. They were supported by mounted archers, whilst the Roman archers were on foot and would not have been able to keep up with the mounted clash. The same goes for the 4,000 Roman legionaries present. Nevertheless it is said that Publius Crassus led the charge into the Parthian cataphracts with great bravery and determination, backed up by his Gallic cavalry.

Plutarch gives a testimony to the bravery of the Gallic cavalry:

with these [the Gauls] he did indeed work wonders. For they laid hold of the long spears of the Parthians, and grappling with the men, pushed them from their horses, hard as it was to move them owing to the weight of their armour; and many of the Gauls forsook their own horses, and crawling under those of the enemy, stabbed them up in the belly. These would rear up in their anguish, and die trampling on riders and enemy indiscriminately mingled.223

Thus Plutarch paints a harrowing picture of the chaos that was a battle within a battle. Strategy went out of the window, replaced by a mêlée where it came down to hand-to-hand fighting between Gauls and Parthians. When the dust had literally settled, despite their bravery and savagery, it was clear that the Gallic cavalry had been well beaten. Those that remained were all wounded, including Publius Crassus himself, and they retreated to the relative protection of the Roman legionaries that had accompanied them. This force then moved to a nearby hillock to make a determined last stand, with the horses in the centre and a ring of legionaries, with locked shields, on the outside to protect the wounded. This, of course, did not save them from a fresh barrage of arrows from the Parthian horse archers.

Plutarch reports that despite being advised to either flee or surrender, Publius Crassus was determined not to desert his command.224 Seeing that they were surrounded on that hillock and that defeat was inevitable, and unwilling to be taken alive, he resolved to choose a more dignified exit. Being unable to pick up a sword due to an arrow wound to the hand, he ordered a soldier to strike a sword into his side, killing him instantaneously. Plutarch also tells us that Censorinus did likewise, whilst Megabacchus still had the strength to take his own life, as did the other surviving officers.225 The rest of the men fought on until the Parthian cataphracts charged the hillock, butchering them with their long lances. Of a force of around 5,500, less than 500 were taken alive226. The Romans had lost over a quarter of their cavalry (including all of their best Gallic cavalry), and a good number of their archers, along with a number of the key junior officers. It was a defeat that sounded the end for Roman hopes at Carrhae. With this force defeated, the Parthians chopped off Publius’ head, stuck it on top of a lance, and returned to the main battle. Before we return to the battle though, we need to dwell on this most important encounter within the Battle of Carrhae, as ultimately it decided the fate of the battle.


This episode has often been explained as being nothing more than Publius Crassus falling for one of the oldest traps in existence: a faked retreat to draw him away from the main body of the army, leading him into heavier Parthian forces, which then turned on him and cut him down. Yet this view overlooks a number of key elements. Firstly, the Romans had to attempt a breakout or they would have faced total annihilation. Secondly, the Parthian cavalry surrounding the Roman army was mostly horse archers; they had only 1,000 cataphracts to protect 9,000 horse archers from 40,000 Romans. Publius took with him all of Rome’s best cavalry (the Gauls) as well as a number of archers and legionaries in support.

The question of whether it was an intended trap depends on what orders Surenas had given. He must have expected the Romans to attempt to break out of his encirclement and we must ask ourselves what strategy he had prepared for this eventuality. Given the appearance of a large force of cataphracts, it is more than likely that Surenas had held them in reserve, following the initial charge and encirclement, so that they could be deployed against any breakout. With careful observation the cataphracts could be sent to wherever the Romans broke out of. All the horse archers had to then do was retreat, whilst still firing, and lead the Roman force towards where they knew the reserve force of cataphracts would be. The trap would then close in on them.

Again, this shows the brilliance of Surenas. Not only did he have an initial strategy, but he had a counter strategy to deal with any Roman breakout. It also demonstrates the severe threat that the Parthians still faced from the Romans, despite the successful encirclement and the barrage of arrows. Had the Roman cavalry successfully broken out of Surenas’ trap, then they could have put the horse archers to flight and allowed the army to extricate themselves. It is unlikely that it would have brought them victory, but it would have given them time to retreat and regroup.

The aim of Surenas’ plan must have been a clear and total victory on the day. Anything less than the destruction of the Roman army would have allowed them to withdraw and fight another day, and Surenas was only ever going to fool them with his modified way of fighting once. For Surenas it was all or nothing; winning the day would not be enough, he had to win the war in one battle. Without total victory at Carrhae, the Romans would return, stronger than before.

Even though the breakout had been planned for, the fighting itself was still going to be close. The Romans broke out with 1,300 cavalry and over 4,000 foot. Given that Surenas only had 1,000 cataphracts in total (and we do not know how many were deployed against Publius) the result was never going to be a foregone conclusion. As it was, the superior Parthian cataphracts carried the day, which meant that the key encounter of the battle was lost due to the poorer quality of the Roman cavalry. For all the tactical planning and innovations, in the end it came down to that one factor. The Romans were not lacking in courage, on the part of Publius or his Gauls; they simply were outmatched in terms of weaponry.

The Final Stage

Initially at least, the breakout that Crassus ordered appeared to have worked. A large part of the Parthian army encircling the main Roman force was drawn away, either fleeing from Publius or riding hard to catch up with him. Crassus used this let-up wisely and staged a withdrawal, whilst still under intermittent arrow fire. The Roman army, laden with casualties, regrouped on nearby sloping ground, which would at least give them some protection from the Parthian cavalry. Here Crassus was faced with a difficult decision, exacerbated by a lack of information, as he needed to know how his son was doing. If Publius had routed the Parthians opposed to him, then he could have possibly advanced and cleared the rest of the Parthian cavalry away, or at least retreated back to the safety of one of the garrisoned towns and regrouped. However, he was not able to come to any decision until he had this information, to which ends he sent messengers out, to try to reach Publius’ position.

Plutarch records that the first one was intercepted and killed, but that the second messenger not only reached Publius’ position, but was able to assess thesituation and mange to return to the main army. When he did so, he informed Crassus that his son was surrounded and being cut to pieces.227 To say that this left Crassus with a dilemma would be an understatement. On a military basis, he knew that the breakout would fail unless he took the main army to link up with Publius. However, this meant gambling with his army and putting them back into the mess that they had only just managed to extricate themselves from. Even if they got there in time, there was no reason to assume that they would be victorious, as the rest of the Parthian army would also converge there.

On the other hand, if he turned and retreated he was not only condemning his son to death – a death that would have been his responsibility – but as the majority of the Roman army was on foot and the Parthians were mounted, there was no reason to believe that they would reach safety in time. Given the number of casualties that they had already sustained, their progress would not have been swift. Furthermore, if the main body of the Parthians did catch them up, they would be strung out in columns and with their backs to them. For whatever reason, military or personal (or both), Crassus resolved that the only move open to them was to advance and meet up with Publius’ beleaguered force.

But, before they had advanced far, they were met with the sight and sound that told them that the encounter between Publius and the Parthians was over. Coming towards them was a cloud of dust accompanied by the beating of war drums. When the Parthians did come into view, they were preceded by the severed head of Publius Crassus. Plutarch tells us that Roman morale sank.228 Not only had a large number of their colleagues been slaughtered, depriving them of most of their cavalry support, but they knew that the battle was about to be rejoined. Despite his grief, it was at this point that Crassus showed his qualities as a general and tried to rouse his men with an impassioned speech:

Mine, O Romans, is the sorrow, and mine alone; but the great fortune and glory of Rome abide unbroken and unconquered in you who are alive and safe. And now if you have any pity for me, thus bereft of the noblest of sons, show it by your wrath against the enemy. Rob them of their joy; avenge their cruelty; be not cast down at what has happened, for it must needs be that those whose aim at great deeds should also suffer greatly. It was not without bloody losses that even Lucullus overthrew Tigranes, or Scipio overthrew Antiochus; and our fathers of old lost a thousand ships off Sicily; and in Italy many imperators and generals, not one of whom, by his defeat, prevented them from afterwards mastering his conquerors. For it was not by good fortune merely that the Roman state reached its present position of power, but by patient endurance and the valour of those who faced dangers on its behalf.229

Now, whilst we have to admit that it is highly unlikely that anyone had the time or the materials to note the speech down word for word, there were enough survivors to have noted the general contents of the speech. Furthermore, as it is reported by Plutarch, who takes a fairly hostile line on Crassus over Carrhae, we can have some confidence that the speech is a fairly accurate representation of what Crassus said.

Nevertheless it was going to take something greater than a stirring speech to save the Romans from the impending slaughter. True to his plan, Surenas (and we are not told whether he was directly involved in the defeat of Publius) employed his tried and tested tactics. The cataphracts again charged the Roman army, forcing them to form closely together, and then the horse archers were brought back into the fray. The Roman army was subject to a constant barrage of arrows and lances, slowly whittling down their numbers.

Only one thing saved the Roman army from total annihilation that day at Carrhae, and that was the arrival of dusk, whereupon the Parthians withdrew for the night. Even though they had the Romans surrounded, the Parthians were unwilling to risk fighting at night. Aside from the traditional reluctance they had of fighting after dark, the conditions made continuing highly risky. They were in the middle of a plain with little natural light and the danger of getting too close to the Romans, or even of friendly fire, was too great.


Thus despite the slaughter and the total defeat they had suffered, the Romans still had a glimmer of hope. The Parthians withdrew and camped nearby, and made no attempt to block their escape. This may seem odd to us today, especially given that the Romans still numbered some 20,000 men (including their wounded) and Crassus himself was still alive and unwounded (in the physical sense anyway). Surenas knew that he had won a spectacular victory, the likes of which no one but he had thought possible, yet he still faced problems. Although the Romans had been comprehensively defeated, a large number of them yet remained, who, if they made for the safety of Roman-held territory, would have been able to recover and regroup. Furthermore, Crassus, the architect and driving force of the Roman invasion, was likely to be more determined than ever to avenge the death of his son. As long as Crassus remained free, the danger to Parthia was not over. Plutarch hints that the Parthians sent an embassy to the Roman army when night fell, to discuss terms of surrender. All he actually says is that:

they would grant Crassus one night in which to bewail his son, unless, with a better regard for his own interests, he should consent to go to Arsaces (Orodes II) instead of being carried there.230

Taking Crassus alive would have been a major prize for Surenas. Yet, due to the Parthian inability or unwillingness to fight at night, the prize could still have eluded Surenas and if Crassus escaped then it would tarnish the remarkable achievements of that day. Ironically, Crassus’ decision to fight immediately in the afternoon, rather than next morning, actually saved the Roman army from utter annihilation, though the Romans had clearly suffered a devastating defeat. Half of their army was dead, and they had been comprehensively outfought. Yet all was not lost. As Crassus himself had pointed out in his rousing speech, Rome had been defeated many times in battle and yet had always emerged victorious in the end. Half the army lay dead on the field of Carrhae, but half yet remained. If they could get safely back to the series of Roman-controlled Mesopotamian towns and then ultimately back into Syria itself, they could re-group for the winter.

It was still possible for Crassus to turn the clock back a year. Rome still held the bridgehead of garrisoned towns in northwestern Mesopotamia. If Crassus wintered in Syria, he could allow his injured soldiers time to heal, raise fresh troops (he was still one of the three men who dominated the Roman Republic after all) and rebuild his army. Certainly his reputation would have taken a battering, but his powerbase was secure. His command extended until 50 BC so there was plenty of time for a fresh campaign in 52 BC. Furthermore, Surenas could only play his masterstroke once. Crassus was not going to fall for that trick twice and could send to Rome for fresh forces, especially additional cavalry. He could plan a new route of invasion, perhaps taking the cities of Babylon, Seleucia and Ctesiphon, which would rebuild shattered Roman morale and then tackle Surenas in his own time and fashion. Thus, as night fell on the battlefield of Carrhae, the Romans had lost the battle, but not the war; the whole campaign was still in the balance, dependant upon the Romans making it to safety.

Before we commence an analysis of the Roman retreat we must pause and comment on the one major discrepancy between the accounts of Plutarch and Dio, that is the treacherous attack of the Osroene leader, Abgarus. Plutarch, writing a century earlier than Dio and seemingly using a first hand account of the campaign, had no such attack take place. Crassus was accompanied for a time in Mesopotamia by an Arab chieftain, whom he names as Ariamnes.231 Even allowing for confusion over names, there is the fundamental point that Plutarch records the Arab chieftain left Crassus’ army before the Battle of Carrhae.232 Furthermore, in what is a very detailed account of the battle itself, at no point does Plutarch mention that a native allied contingent betrayed the Romans and attacked them, which we must expect to find if it actually happened. Given its absence from this, our best source for the battle, we must assume that this treacherous attack did not occur. Where Dio got this from we will never know, but, as far as is possible to do so when dealing with ancient sources, we must clearly note that this treacherous attack by Abgarus in the Roman rear did not take place and was a later fiction copied by Dio into his account.

The Retreat to Carrhae

Again, Plutarch and Dio disagree on the finer details of the retreat. Nevertheless, the first stage of the Roman retreat was to get safely back to the town of Carrhae itself and the security of its walls and Roman garrison. Plutarch tells us that the Romans looked to Crassus for leadership, but that he was lying on the ground in despair, which meant that the escape had to be organised by the two most senior surviving Roman officers: Cassius and Octavius.233 Dio omits this and states that Crassus led the survivors on the retreat.234

It is clear that the journey itself was a perilous one. In the dead of a cold Mesopotamian night, 15,000–20,000 men, a good many of them injured, had to walk the route back to Carrhae. In fact it was no mean feat that they were still able to navigate their way back to the town in darkness and following the hardship of the day’s battle. A hard decision had to be taken that night, in regard to what was to be done with those men who were too seriously wounded to walk. Given that time was of the essence and that they had to be at the walls of Carrhae before dawn, the brutal decision was made to leave the seriously wounded behind. Plutarch provides us with a dramatic description of their journey;

Then the sick and wounded perceived that their comrades were abandoning them, and dreadful disorder and confusion, accompanied by groans and shouts filled the camp. And after this, as they tried to advance, disorder and panic seized upon them, for they felt sure that the enemy was coming against them. Frequently they would change their course, frequently they would form in order of battle, some of the wounded who followed them had to be taken up, and others laid down, and so all were delayed235

Not only were a number of men left behind, numbering some 4,000 it is estimated, but a number would have died on route to Carrhae, from untreated wounds and fatigue.236 For many it was a march of death. The first Romans to reach the town of Carrhae were the remnants of the Roman auxiliary cavalry, about 300 in number. They were led by a Roman nobleman by the name of Egnatius. However, when they reached the town an event occurred that was to set the tone for the whole Roman retreat. Upon reaching the walls of Carrhae, Egnatius gained the attention of the Roman guards on the walls, shouting to them to tell their commander (a Roman officer by the name of Coponius) that a great battle had taken place between Crassus and the Parthians. At that point he and his men promptly rode off and headed towards Zeugma and the crossing back into Roman Syria, without even identifying who he was.

This was an ominous sign: a Roman officer deserting his commander and the whole campaign and riding as fast as possible for the safety of a Roman province. Plutarch tells us that Egnatius was forever tainted by this act of cowardice and we can find no further trace of him in subsequent Roman political or military life.237 Nevertheless, despite its brevity, the message actually had the desired effect and Coponius, realising that something catastrophic had occurred, immediately led an expedition out from Carrhae, located the column of Roman survivors and escorted them back into the town.

For Crassus at least, the first stage of the retreat had been accomplished and the bulk of the Roman survivors had reached safety. Exactly how many men reached the relative safety of Carrhae is difficult to estimate, as we are not given a clear figure by Plutarch. However, it does seem, judging from some of the later figures that Plutarch gives us, that between 15,000–20,000 men reached the town. Actually, this raises one of the most surprising and neglected aspects of the whole Carrhae campaign, namely how many Romans were killed during the battle and how many were killed during the aftermath. As we shall see the balance between the two is actually quite surprising.

When dawn broke, the Parthians advanced upon the site of the Roman army’s last stand, and as they expected found that the bulk of the army had fled. What they also found were the 4,000 seriously wounded Roman soldiers, who had been left behind. Surenas, unwilling to show any more mercy to them than their comrades had, promptly had these men slaughtered. He then set upon the task of locating the bulk of the Roman army. During this day his cavalry came across a number of Roman stragglers, who had either been separated from, or fallen behind, the main group (an easy thing to do given the state of the retreat at night). In all but one case they too were easily dispatched.

There was, however, one notable exception, which Plutarch chooses to highlight and so should we. One of Crassus’ legates was an officer by the name of Vargunteius, who hailed from a minor senatorial family. During the retreat he was in command of four cohorts, less than 2,000 men (especially given the losses of the previous day), but became separated from the main group. When day broke and the Parthian cavalry located them, they decided to make a last stand on a small hillock. Given the overwhelming odds there was only ever going to be one outcome, yet they fought and died hard to such an extent that the Parthians noted them for their bravery, not something that had been in great supply from the Romans during the retreat. As they were down to the last twenty men (not including Vargunteius, who had already fallen) they charged the Parthians in a last defiant gesture. So impressed were the Parthians with their defiant stand that they parted and allowed them to continue to Carrhae unmolested.238 Such tales of heroism in this retreat were few and far between.

As stated earlier, we therefore have recorded incidents of over 6,000 Roman soldiers surviving the battle, but dying the next day. Given that these are only two such incidents (many more not being recorded due to the absence of any surviving witnesses) we can begin to appreciate the scale of the Roman losses that occurred in the days after the battle.

The Retreat to Syria

At this point, both Crassus and Surenas were locked in an odd game of cat and mouse. Surenas was not exactly sure of where Crassus was, whilst Crassus and his army had to evade the Parthians and seek the refuge of either Armenia or Syria. Although Carrhae was the most logical place for Crassus to make for, Surenas could not be certain. Added to this, Plutarch states that Surenas received a report (from whom we are never told, nor are we told how Plutarch’s source got to know of this) that Crassus was not in Carrhae and was in fact heading for the border.239This would have left Surenas in something of a dilemma. However, he soon came up with a plan to resolve it by sending a man up to the walls of Carrhae and requesting a peace conference between himself and Crassus, to organise a truce and a safe withdrawal of the Roman forces from the towns and cities of Mesopotamia. Whilst the evacuation of the occupying Roman garrisons was a necessary move for the Parthians, Surenas needed to locate Crassus, dead or alive, even more. Plutarch reports that Cassius took the bait and reported back to Surenas’ emissary that Crassus would be willing to meet with him, which only served to confirm Crassus’ presence within the town.240 By this simple ruse and by Cassius’ short-sightedness, the Parthians now knew where to end this war and Surenas moved his entire army towards the town of Carrhae.

For Crassus, Cassius’ stupidity had left him with an even bigger headache. Given the strength of the Roman forces in Carrhae (a garrison, plus 15,000–20,000 survivors) he would have been able to resist a Parthian siege, not that Surenas’ army was equipped for storming a city. The problem was that although the Parthians could not get in, soon the Romans would not have been able to get out and they did not know how long the food and water would last, given the size of the Roman forces within. Crassus could have adopted a policy of waiting it out if he knew help was going to arrive to alleviate a siege, but where would this help come from? Assistance would not soon be forthcoming from Roman Syria, given the few forces that remained there, which only left Armenia. However, as Crassus was not able to rely on the Armenians to help him when he was in a position of power, it was highly unlikely that he could do so now in such a weakened one. Although he was never to know it, this assessment proved to be a highly perceptive one, as only a few days later King Artavasdes would meet with King Orodes to discuss a peace treaty between Armenia and Parthia.

This left Crassus with only one viable option; he would have to break out of Carrhae, evade the waiting Parthians and make for Syria or the Armenian foothills. It appears that the Roman army was divided up into groups, each led by one of the senior surviving commanders. We know of groups led by Crassus, Octavius and Cassius, but there must have been more. It is probable that each group had a different destination and different route, to divide and distract the Parthian pursuers. The move had to be made at night, so as to slip past the Parthians and had to be done when there was no full moon, in order to keep as much cover as possible.

Although we know what happened next, why it happened is the subject of much conjecture. The facts, ultimately are that whilst Cassius’ group made it to Syria, Octavius’ and Crassus’ did not. Plutarch ascribes this to Crassus once again relying on, and being betrayed by, a native guide, this time a man known as Andromachus. According to Plutarch, Andromachus offered to guide Crassus and Cassius from Carrhae, but planned to lead them on a circuitous route and delay them, so that the Parthians would be able to find them by daybreak.241

Plutarch’s version of the event also has Cassius realising that they were being led into a trap, then breaking away and returning to Carrhae without telling Crassus242. If this was true then it was desertion of the highest order. It would seem to be either a daring double bluff or foolish in the extreme to return to the town of Carrhae, past the Parthians once more and hope that they rode off after the other groups. Dio, naturally, has none of this detail. He has Crassus making for the Armenian foothills and Cassius safely reaching Syria independently.243 When day broke and the Parthians realised that the Romans had evacuated Carrhae, they set off after them once more. Again Dio reports that many groups did not escape the Parthian cavalry, though it seems that on this day a number of them were taken prisoner (perhaps this was due to Surenas wanting Crassus alive or at least to confirm that they had killed the right man).244

Of the three main groups, we know that Crassus’ got bogged down in a marsh, whether at the hand of a treacherous guide or by simple misfortune, and thus when day broke he was still out in the open and some way off safety. Octavius and the 5,000 men he commanded had reached the relative safety of the mountains at Sinnaca before daybreak. Cassius it seems disappears from the picture and only turns up again safe and sound in Roman Syria, the only one of the key Roman commanders to do so.

By now the Parthians, led by Surenas, had spotted Crassus’ group and were moving in on them. However, he was saved by the intervention of Octavius, who could see the relative position of both groups from his high position. Unlike many of the Roman officers in that retreat, he appears not to have thought of his own safety, but his duty to his commander and led his force of 5,000 men (some of them unwillingly) to rescue Crassus from the advancing Parthians, who were far less in number than the Romans. Thus Crassus finally reached the safety of the foothills, where the Parthian cavalry were far less potent and where Roman numbers would count.

For Surenas, the situation was serious. Certainly he had defeated the Roman army at Carrhae and he had inflicted further heavy casualties on them during the retreat, but if Crassus should escape, even with a force of 10,000 men back into Syria, then the war would continue. In desperation, he tried one last stratagem. He either sent an embassy to the Romans in the hills, or went himself, stating that he wanted a peace conference to offer the Romans the opportunity to evacuate all territories east of the Euphrates. The details of this treaty were to be worked out at this meeting between the two men, along with a few officers from either side, on neutral ground between the two forces. Plutarch reports that he went and delivered this offer himself and reports his words:

I have put your valour and power to the test against the wishes of the king, who now of his own accord shows you the mildness and friendliness of his feelings by offering to make a truce with you if you will withdraw and by affording you the means of safety.245

Now, Dio and Plutarch report very different reactions by Crassus to this offer. Dio reports that:

Crassus, without hesitation, trusted him. For he was in the very extremity of fear, and was distraught by the terror of the calamity that had befallen both himself and the state.246

According to Dio, therefore, Crassus was eager to meet Surenas and accept whatever deal he offered, and so walked right into his trap. Dio’s account would have us believe that the experienced general and the cynical political manipulator that Crassus was, fell for this ruse due to the pressures he had been under during the last few days. Plutarch however reports a very different Crassus and one more in keeping with the man we know. He reports that:

Crassus, who’s every discomfiture at the hands of the barbarians had been due to fraud, and who thought the suddenness of their change a strange thing, would not reply, but took the matter into consideration.247

This description fits the cunning and cynical Crassus that is more familiar to us. Even after all that had happened to him, he was still very much in control of his faculties. He would have been well aware that he had lost the battle, but not the war. However, he was not prepared for what happened next. Although he and his officers saw through Surenas’ ruse, the surviving legionaries, trapped on a desolate Mesopotamian hilltop, and with the Parthian force below, apparently did not. In yet another example of the lack of discipline that had plagued the retreat from the start, the troops mutinied and demanded that Crassus attend the peace negotiations. They had survived the calamitous day at Carrhae and the two near-disastrous retreats and now it appeared that their officers wanted more hardship for them, rather than a negotiated settlement. Plutarch reports that Crassus once again attempted to reason with them, arguing that they could make good their escape into the hills, but to no avail.248 In all fairness he had led them on what turned out to be a disastrous campaign and we could hardly blame the legionaries for having little faith left in his abilities or judgement. Thus Crassus was forced to meet Surenas, for what he believed would be his death, rather than his soldiers’ salvation.

Plutarch reports that before he descended to meet Surenas, he made one final and prophetic speech to his two senior surviving commanders:

Octavius and Petronius and you other commanders of Rome here present, you see that I go because I must and you are witnesses of the shameful violence I suffer; but tell the world, if you get safely home, that Crassus perished because he was deceived by his enemies, and not because he was delivered up to them by his countrymen.249

With that he descended to meet Surenas. Once again though, Octavius did not let him down and he and Petronius and some other of the officers went with Crassus, in order to protect him. When Crassus sent two legates ahead of him to meet with Surenas and see what protocol was to be observed, neither returned. Plutarch names them as the two Roscius brothers.250 Nevertheless, Crassus and his retinue continued onwards. When Surenas and his officers met with Crassus they noted that they were on horses whilst he was on foot and offered him the use of a spare horse, which they had brought along. When Crassus mounted the horse, the Parthian grooms attempted to gallop the horse away towards the Parthian lines, with Crassus still on top of it. At once Octavius stepped in and killed one of the grooms, but was in turn struck down by the other one. Petronius too entered the fight and was killed by his commander’s side. It is reported that Crassus was the last to fall in this unedifying struggle, killed by a Parthian soldier named by the sources as either Promaxathres or Exathres.251

Upon the death of Crassus and most of his senior officers, Surenas sent word to the Romans up in the hills, who had witnessed this assassination (which they had greatly been responsible for), and called for their surrender, pledging that they would not be ill treated. Amazingly, a number of them actually believed Surenas’ offer, despite what happened to Crassus, and did surrender. They were added to the growing tally of Roman prisoners. Understandably a number of the remaining soldiers did not accept Surenas’ offer and made away under the cover of night. Plutarch reports that the majority of them were hunted down and killed, whilst Dio states that the majority escaped through the mountains and reached safety in Roman territory.252

Thus died Marcus Licinius Crassus, one of the three leading men of Rome; assassinated in an ignominious scramble over a horse. Within a decade he was joined by the other two members of the triumvirate: Pompey, assassinated on an Egyptian beach in 48 BC and Caesar, four years later, assassinated in the Roman Senate House by a group of his so-called supporters (who incidentally were jointly led by Cassius, the man who had let Crassus down on so many occasions).

It was here, in the hills of Sinnaca, that Surenas finally completed his victory. With Crassus dead the Roman campaign was over and the war had been won. Surenas seized the chance to celebrate and did so in a vindictive style. He had Crassus’ head cut off (as he had done with Publius’) as well as his hand, and sent Silaces (the satrap of Mesopotamia, whom Crassus had defeated in 54 BC and who was at the Battle of Carrhae) to convey both trophies to King Orodes. Before doing so it is alleged that he poured molten gold into the mouth of Crassus’ head, mocking his great wealth.253 Crassus’ body was then apparently left to rot on a heap of Roman corpses.254

Before the head reached the king he arranged a victory parade in the city of Seleucia (which he had retaken the previous year from the rebel Mithradates III and which was known to harbour pro-Roman sympathies). He paraded the Roman captives through the streets of Seleucia in a mockery of a Roman triumph. At the head of the procession he placed a Roman prisoner who was said to resemble Crassus and had him dressed in a woman’s robe and forced him to pretend to be Crassus.255 Behind him he had men carrying Crassus’ fasces (the ceremonial bundle of rods and axes which symbolised a consul’s authority), but now they were crowned with freshly-severed Roman heads. Next came the captured Roman legionary eagles, the symbol of Roman military might, which were then distributed amongst unnamed Parthian temples and hung there as trophies for the next thirty years.256 Following the prisoners were a number of Seleucid musicians who sang songs ridiculing Crassus for his cowardice and effeminacy. Surenas even brandished a number of parchments of the Milesiaca, a noted erotic work, found amongst the possessions of one of the Roscius brothers, to ridicule the Romans’ weaknesses.

In Armenia, Silaces arrived with his special delivery just as King Orodes and King Artavasdes of Armenia were conducting a treaty of alliance. There are no reports of whether any fighting actually took place between the Armenians and the Parthians. Given this silence and Artavasdes’ vacillating mood earlier in 53 BC, it is most probable that the Armenians gave in without a fight. It is possible that Artavasdes was hoping that this would only be a temporary treaty and that he could break it when Crassus defeated Orodes and then try to explain away his actions.

As it turned out, both kings at the meeting were in for a shock. Under the terms of the treaty with Parthia, Armenia would return to the vassal status that it occupied in the time of Mithradates II, with Parthia acknowledged as the stronger, but Armenia retaining its territorial integrity. Once again the treaty was sealed with a marriage alliance, with Artavasdes’ sister being married to Orodes’ eldest son, Pacorus. Ultimately Crassus’ invasion had allowed Orodes to turn the clock back on Parthian-Armenian relations and restore the old balance of power. It was at the feast to celebrate this alliance that Silaces arrived with Crassus’ head; to be more precise, it was during a theatrical performance of the Bacchae, by the famous Greek playwright Euripides (both the Parthian and Armenian kings had developed a taste for the mainstream Hellenistic culture). During a pause in the singing, it is reported, Silaces entered and, after making his bow to the king, cast Crassus’ head into the space where the singer stood. At which point the singer, named as Jason of Tralles, picked the head up and recited the verse from the play:

We bring from the mountain, a freshly cut twist of ivy to the palace, a prosperous spoil.257

To the Parthians it seemed fitting; for Crassus it was the final humiliation, his head being use as a theatrical prop in a Greek drama.258 However, when the rejoicing was over, both kings would have realised that they now had growing problems. For Artavasdes, rather than playing the Romans off against the Parthians and thereby maintaining an independent Armenia, now found himself with Rome defeated and Parthia in the ascendant. What he must have hoped would be a temporary treaty to avoid the Parthian army had now turned into a permanent position of vassalage to a resurgent Parthia. The Parthian heir now had a clear claim to his throne and he had clearly miscalculated when he did not provide Crassus with the cavalry he needed.

For Orodes, the utter surprise and joy at the news must have soon soured when he realised just how the invasion had been defeated. On the one hand, not only had Armenia been brought back under the Parthian wing (as it was prior to 87 BC), but the looming threat of Rome had been met and comprehensively defeated, with the ultimate Parthian prize of Syria (which they had quested after for nearly one hundred years) now lying open and defenceless. On the other hand, however, he will have soon realised just how this had been accomplished and that, although he had eliminated one threat to his throne, he had just greatly increased another.

It is probable that Orodes sent Surenas to meet the Roman invasion purely in order to slow it down, and it is highly unlikely that he was expecting Surenas to win such a decisive victory. Prior to Carrhae, Surenas was already the second most powerful man in Parthia; his family was the strongest of the noble houses outside of the Arsacids themselves. Furthermore, Surenas had been responsible for putting Orodes on the throne in preference to his brother, and then responsible for ending the ensuing civil war by defeating said brother. Now, if that were not enough, Surenas had actually managed to comprehensively defeat the Romans in battle (in their worst defeat for 150 years), kill one of Rome’s leading men and single-handedly not only end the Roman invasion, but stop the juggernaut that was the Roman Republic. The acclaim that Surenas would receive from all non-Roman quarters, never mind the Parthian people, army and nobility was going to be immense. No king could stand such acclaim for another and certainly not one as weak as Orodes.

For Orodes, if he was to keep his throne and stop the House of Suren replacing the House of Arsaces on the Parthian throne, there was only one possible answer. Within a year, Surenas, the man who had done what no other had done for generations (defeat a Roman invasion), was put to death on the orders of the king. We do not know the details of how he managed to do this, but the charge used was treason. Possibly he lured Surenas away from his forces with the promise of more honours and then had him swiftly executed. In any event, the man who had accomplished so much was murdered by an undeserving monarch who would soon regret the disposal of his best general.

In the end, therefore, there was only one winner to emerge from the Carrhae campaign. It was neither Crassus, nor Surenas; both had met ignoble ends, rather than death on the battlefield. The only clear winner was Orodes II, who began this war as a weak monarch in charge of a weak empire and ended it as the unquestioned ruler of the region’s leading superpower. All that lay ahead was the resumption of Parthian westward expansion and the accomplishment of the long term Parthian goal of reaching the Mediterranean.

Summary – The Battle and the Retreat

We can now see the full scale of the disaster that befell Rome during the Carrhae campaign. The Romans had lost battles before, but never one in such a comprehensive manner and followed by such a comprehensive rout. At the end they were literally chased out of Parthian territory in abject disarray, with their vaunted Roman discipline abandoned and with an ‘every man for himself’ attitude being the order of the day. The retreat from Carrhae was as disastrous as the battle itself and must count as one of the great disastrous retreats in history. The only clear estimates we have for Roman casualties are from Plutarch, who puts the Roman dead at 20,000, with 10,000 captured (see appendix one) and Appian, who merely reports that less than 10,000 escaped to Syria.259

One aspect that is rarely noticed is just how many of these dead and captured resulted from the retreat, rather than the battle itself (at least 6,000 were killed on the day following the battle). This is not as surprising as it sounds, as there was little hand-to-hand fighting during the battle; it was mostly a barrage of arrows, most of which disabled rather than killed outright. The only close-quarter fighting occurred during Publius Crassus’ breakout, during which less than 6,000 Romans died. For the rest of the battle, the Roman casualties were from arrow strikes. Given the prolonged nature of the Roman resistance and the random barrage of the Parthian arrows, it appears that a great many of the Roman casualties were not immediate fatalities, but men who suffered multiple wounds of varying degrees. Many of these would have succumbed to their injures after the battle, due to the fatigue and blood loss, rather than during the battle itself.

Of the Parthian casualties we have no word, though again the only close-quarter combat which the Parthians took part in was during Publius’ breakout. Given that the bulk of this fighting was done by the Parthian cataphracts and the ferocious nature of the battle, even with their heavy armour we can expect them to have taken a considerable number of casualties. The difference here is that Surenas would have taken the bulk of his casualties from amongst his 1,000 cataphracts, rather than across the army evenly. This still gave him more than enough horse archers available to hunt down fleeing Romans, but may explain his apparent inability to tackle the force that assembled around Crassus at the end.

What can be learnt from the battle itself? It certainly would appear that whilst the Romans had the overall numbers they lacked depth in certain areas, most notably the cavalry. This, however, was not an intrinsic flaw of Crassus’ preparations. As the wait until 53 BC showed, Crassus knew that his army was weak in cavalry. This shortage only became the crucial issue because Surenas choose to exploit a known Roman weakness. For the battle he was expecting, Crassus had enough cavalry to keep the Parthian cataphracts occupied. Yet for the battle that Surenas engineered; a highly mobile and missile-based one, he was hopelessly outclassed.

Nevertheless, it must be pointed out that the Roman loss at Carrhae was down to one man. Unlike traditional views of the battle, it was not lost because of Crassus’ incompetence, but because of Surenas’ brilliance. Surenas realised that he could not defeat Rome over the length of a campaign; past history had taught him that. He did realise, however, that Rome could be defeated in a single battle, if he prepared for it properly. If that defeat was a heavy one, both in terms of the psychological damage and the number of casualties, then the war would be over. Added to this was his realisation that the Roman Republican system had mutated to such an extent that it began to resemble Parthia, in so much as the whole campaign was reliant on a single commander. If he captured or killed Crassus then the invasion would be over. Certainly there would be likely to be another dynast along at some point in the future (most likely to be either Pompey or Caesar), but that would be a different war.

Crassus and the Romans were undone at Carrhae by Surenas’ tactics of turning the battle into a fast-paced cavalry engagement, with no infantry and a total reliance on missile fire. Had the Romans got close enough to the Parthians in sufficient numbers, then their numerical and military superiority at close quarters would have shown. Surenas’ genius lay in stopping the Romans from doing this. Nevertheless, for the Romans the battle itself was not as catastrophic as many would believe. This was not a typical Parthian army that they faced, but one that very much reflected the genius of its commander. As Publius’ breakout had shown, at close quarters the Romans were still a force to be reckoned with, and there must have been points when the outcome of the ‘battle within a battle’ was still in the balance. Furthermore, Surenas’ tactics could only be used once, after which the Romans would be ready for them. It is interesting to note that when Caesar was preparing for his Parthian campaign (which was abandoned following his assassination) the sources note that his proposed force was heavy in cavalry.260

What really did the damage for the Romans, and what turned a terrible defeat into a catastrophic one, was the retreat, or as we should say the retreats. These shambolic manoeuvres doubled the numbers of men lost, either killed or captured. The Roman general was killed, along with the majority of his young aristocratic officers. Both retreats were plagued by a complete breakdown of discipline. During the first retreat, to Carrhae, Crassus’ advance guard did not stay to provide cover, which could have allowed the stragglers to catch up, or to find the groups that had become detached from the main force (such as the force led by Vargunteius). Instead, they deserted their post and fled back to Roman Syria. Of the two officers who are known to have survived, both could be, and indeed were, accused of desertion. Furthermore, there are excellent comparisons to their contemporaries who died. Whilst Vargunteius died fighting a brave last stand, Egnatius fled Parthia and survived in ignominy. Whilst Cassius betrayed Crassus and reached Syria safely, Octavius died fighting to defend him, when he too could have put his own life first. On too many occasions the Roman army was beset by indiscipline from both officers and men. This was an ominous sign for the Roman Republic.

The combination of the defeat and the retreat made the Parthian campaign a total disaster for Rome, the likes of which had not been seen since Hannibal crossed the Alps into Italy during the Second Punic War. Of an army of 40,000 plus, barely a quarter of them returned back to Syria. The seemingly unstoppable Roman juggernaut had come off the road altogether. Thus in the first battle and the first war between the two great superpowers of the east, Rome was the clear loser. Given that their rapidly-expanding empire had been built on an almost legendary invincibility, this defeat had serious implications. Not only had the Roman Empire been prevented from advancing, but it was now in clear danger of retreating.

If you find an error or have any questions, please email us at Thank you!