Chapter 9
In 72 BC Mithridates was in his sixtieth year, and the seventeenth of his struggle with Rome. Among the levies now facing him might be found the children of the Roman legionaries he had fought in 88 BC. Certainly, one of his opponents - the Murena now cruising the Propontis with the Roman fleet - was the vengeful son of the Murena who had so spectacularly failed to defeat Mithridates in the war of a decade previously.
Mithridates grimly continued to keep faith in the eventual political collapse of the Roman polity. As well as the continued exertions of Sertorius and the depredations of the pirates, Italy itself was now racked by the rampaging Spartacus. It is probable that Mithridates, like almost everyone else, underestimated Spartacus’ military genius. Otherwise he would undoubtedly have attempted to give a fellow enemy of Rome substantial support. However, even if the phenomenon of Spartacus was considered a slave revolt of the type that the Romans quashed with monotonous regularity in Sicily, the sheer scale of this insurrection meant that no reinforcements would be coming to Lucullus in 72 BC. In fact, with Spartacus in Italy, Mithridates in the east, Sertorius in the west, and the pirates right across the ocean in between, the Roman colossus was beleaguered on all sides. Hope was by no means lost.
Out of consideration of the stringency facing Rome, Lucullus had declined an offer from the senate to help his war effort with 3,000 talents of silver. The Roman had decided to use the resources of Pontus to finance the war against that country’s king, but first he had the challenge of getting there. He had by now established that Mithridates had evacuated Bithynia, and had no doubt that the king was preparing to resist his advance through Paphlagonia or Phrygia. This left the choice of routes as either through Galatia or along the Black Sea coastal plain.1
The coastal route was straightforward as far as Heraclea, but going further involved taking that city out of the war, and this probably meant another protracted siege – something that Lucullus could ill afford. His Fimbrians had already endured the gruelling siege of Cyzicus, and were unimpressed by the thanks of the grateful citizenry. After almost a decade in Asia Minor, they wanted plunder and to go home. In the event, the siege of Heraclea was left to Cotta. Through a mixture of Pontic determination and Cotta’s woeful military ineptitude, the siege dragged on for two years. It was only brought to a halt when Menchares, on the other side of the Black Sea, was persuaded to become an official ‘friend’ of the Romans and to stop smuggling supplies to the besieged citizens.
This left Galatia, and across Galatia went the Fimbrians, complaining bitterly about the lack of supplies as they went. Deiotarus had foreseen this problem. As many as 30,000 Galatians followed the Romans, not as reinforcements but as a supply chain. Every one of these men carried a bushel (a measure equivalent to about eight gallons) of wheat on his back, and in this way the Roman army crossed the barren highlands of Galatia, and descended into the unprotected heartlands of Pontus.2 This fertile and well-populated area had been Pontic for generations, and had thrived under the aegis of the Mithridatic dynasty. With Mithridates still attempting to gather an army at Cabira to the east across the river Lycus, the Fimbrians were finally free to plunder the abundance of the countryside to their heart’s content. In the Roman camp the price of an ox dropped to under a day’s wages, and a slave could be had for four drachmas. When the Romans could plunder no more they simply pillaged and despoiled what was left. The degree of human and economic damage that this entailed partly explains why Mithridates had opted to fight Sulla in Greece rather than in his Anatolian conquests, which he had been more confident of holding. Warfare was more damaging than any typhoon for the lands a hostile army passed over.
With plunder aplenty, the Fimbrians became more particular about their choice of loot. Agricultural produce was all very well, but retirement pensions needed gold, and gold was to be had from the cities and towns. These, annoyingly, Lucullus refused to take by storm. Instead he camped before them and negotiated their surrender, giving easy terms to encourage the next fortification down the line to give itself up as well. This far-sighted strategy was both intended to expedite the surrender of Pontus and to allow the province, once yielded, to become as much a cash cow for Rome as it was for Mithridates. The problem lay in explaining all this to an army fixated on pillage and plunder. It did not help that by and large the Roman leadership shared the Fimbrian outlook. Lucullus later attempted to prevent further uprisings in Asia Minor by fixing the payment of the Sullan indemnity with a twenty-five per cent tax (on virtually everything) and freezing interest rates at twelve per cent. As Lucullus also ruled that interest should not accrue beyond the amount of the original debt, the ravages of galloping compound interest were cut at a stroke. It also meant that if Mithridates were to arise, vampirelike once more from defeat, Asia would be less likely to succumb to his promises of financial relief. By these measures the foundations were laid for the economic recovery of Asia Minor, and by way of thanks Lucullus incurred the bitter enmity of those at Rome who had been making their fortunes from the suffering of the region – a group which included many senators and their equestrian partners in extortion.
The Cabira campaign
Mithridates could not prevent the despoiling of his lands, because he lacked an army with which to do so. However, he had the cavalry which had survived Cyzicus, and the largely-intact forces which had been defending the southern approaches to Pontus during the Bithynian campaign. This probably totalled some 40,000 men – not enough to defeat Lucullus, but certainly enough to trouble him.3 Certainly, there was no way that Pontus could be considered conquered whilst this force remained in the field. By way of reminding the Romans that he was far from a spent force, Mithridates began raiding with a goodly proportion of his cavalry over the Lycus. He engaged a Roman force on the other side and soundly defeated it. Plutarch relates that one of the leading Romans captured was a man called Pomponius (possibly a relative of Pomponius Atticus, the friend of Cicero). Mithridates offered to spare the man’s life in return for his friendship. Pomponius defiantly replied that he would remain Mithridates’ enemy as long as Mithridates was an enemy of the Romans. ‘The king was amazed by the man, and did him no harm’, reports Plutarch.4
In the hope of bringing Mithridates to battle, Lucullus settled down to a leisurely siege of Amisus. There was not much else he could do. The ever-strengthening cavalry of Mithridates meant that the Romans could no longer plunder the plains with impunity, and Pontic irregulars would make life difficult in the hills. If pushed, Mithridates would retreat to the fastnesses of Armenia Minor, and Lucullus certainly did not relish persuading the Fimbrians to undertake the unrewarding task of chasing a highly-mobile Mithridates around the mountains. Mithridates knew that Lucullus was already facing criticism for dragging out the war, and believed that Amisus –fortified as it had been – was capable of outlasting Lucullus.*
Therefore Mithridates remained at Cabira and waited for the Romans to come to him. He placed a small garrison, commanded by a relative called Phoenix, on the mountain passes and gave orders for a signal fire to be lit if the Romans moved, as he felt Lucullus would be forced to do. Eventually, in the spring of 72 BC, Lucullus left the siege of Amisus to Murena and two legions, and reluctantly made his way to Cabira with the rest of the army.
Details of the campaign which followed are hopelessly confused, with different sources giving different and incompatible chronologies. However, a number of events are commonly agreed on, though not the order in which they occurred. Militarily, it would appear that Lucullus had first to storm the fortress of Eupatoria, which Mithridates had constructed specifically to guard the approaches to Cabira. It may be that this was where Phoenix was stationed, for we are told that after informing Mithridates that the Romans were coming, he promptly defected to them. However, as the historian Memnon is clear that this fortress was captured at the end of the Cabira campaign, this suggests that Lucullus left the fortress in native hands and at some point it returned its allegiance to Mithridates.5
It is also clear that Mithridates gave Lucullus a warm welcome to Cabira and inflicted a sharp setback on some of his troops in a highly-confused skirmish which seems to have developed from an attack on Roman foragers. Confronting Mithridates in Cabira had been a bold move by Lucullus and, despite this setback, the Roman proceeded to push his luck even harder. Some local Greeks offered to lead the Roman force to an advantageous position and Lucullus accepted. Since this involved passing through a dangerously narrow valley, he took advantage of the fact that he had veteran troops under his command and performed the manoeuvre at night. To slow the enemy reaction, the Romans first lit camp fires and generally gave the impression that they were settled until dawn.
In fact, dawn saw the Romans occupying an old fort on the heights overlooking Cabira. This was a secure position defensively and one from which the Romans could easily move to the plain.
But that narrow defile meant that Lucullus was cut off from his recent conquests in Pontus and was now dependent on supplies from Cappadocia in the south to keep his army fed. The vulnerability of that supply line was soon demonstrated by a violent skirmish with the Pontic cavalry. At first the Pontics got the worse of this confrontation but Mithridates himself led a countercharge, and gave the Romans such a fright that they continued to retreat long after the Pontic forces had pulled back.
Now fighting on his home turf, it was Mithridates who offered battle on the favourable ground of the plain, and Lucullus who declined. While this stalemate dragged on, Mithridates appears to have attempted to decapitate the Roman force with the assassination of Lucullus. He fell out very publicly with a barbarian chieftain called Olthacus, and this man promptly followed the prevailing trend and defected to the Romans. Olthacus gave Lucullus intelligence which was checked and proven trustworthy, with the result that Olthacus himself became a trusted member of Lucullus’ entourage. One midday, the barbarian presented himself at the commander’s tent with urgent news, and as was his custom, a dagger on his belt. An officious attendant told Olthacus that Lucullus was asleep and had ordered that he should not be disturbed no matter how urgent the tidings. Olthacus must wait until the general had awakened. This was far from convenient, because the would-be assassin had parked his horse just outside the camp for a quick getaway, and it was only a matter of time before this suspicious positioning was noted. After being physically prevented from forcing his way into the Lucullan bedchamber, Olthacus gave up the attempt. Since he was hopelessly compromised anyway, he mounted his horse and rode back to Mithridates. As Plutarch commented, many a general has come to grief from being asleep at the wrong moment, but only Lucullus was saved by being asleep at the right time.6
The war of attrition continued. At Cyzicus, Mithridates had been taught that a strong position with poor supply lines was at least as bad as a poor position with good supply lines, and he was determined to give Lucullus the benefit of this lesson. The Pontic cavalry made a violent attack on a supply convoy which was bringing corn from Cappadocia. The convoy had an escort of ten cohorts of infantry, commanded by one Sornatius, and these succeeded in holding off the attack, so Mithridates decided that the next attempt by the Pontics would constitute a mixed force of infantry and cavalry.
Some 4,000 infantry and the Pontic cavalry fell upon the next Roman supply convoy but, possibly due to poor coordination between foot and horse, the battle took place in the narrow valley where the horsemen were almost unable to deploy. The infantry and half the Pontic cavalry were wiped out. The infantry were not a serious loss but, with his cavalry crippled, Mithridates was unable to dominate the plain. Almost certainly, as soon as Lucullus comprehended this he would descend from his hill and either force the Pontic infantry to battle on the plain or, if battle was declined, lay siege to the Pontic camp. Mithridates, harbouring no illusions as to what would happen to his foot soldiers if the Fimbrians got to grips with them, made ready to withdraw.
In order to minimize panic in his camp, the king gave out that in the recent fight his men had merely got the worst of a minor skirmish. He instructed his guards to stop anyone from leaving the camp and discreetly ordered the royal treasure to be evacuated on muleback. Rumours of a major Pontic defeat were exaggerated by the survivors of the battle in the valley as they trickled back into camp. It also became clear from the loaded waggons approaching the Roman camp that the Pontic attack had failed miserably. Those who knew of Mithridates’ plan to evacuate his treasure took the chance to send away their own baggage as well. It did not take long for the soldiers to connect the crush around the camp gates with the reports of a major defeat, and the dissimulation of their commanders led the normally-obedient infantry to conclude that their betters intended to leave them in the lurch.
Mithridates in his tent became aware of general tumult outside. It was too late to rally his troops, who had decided that it was now every man for himself. No one listened to the king; he was knocked from his horse and one of his priests, Hermaeus, was trampled to death by the crowd trying to get out of the gates. With commendable promptness, Lucullus divined what was happening and sent his army to harry the retreating force, giving explicit instructions as to which members of Mithridates’ entourage he wanted captured alive. The Romans reached the camp and slew the few who had remained as they were attempting to pack tidily.
Lucullus knew his men and had specifically told them to ignore the camp and concentrate on killing as much of the Pontic army as possible whilst it was disorganized and vulnerable. However, the Fimbrians finally had the chance to get what they had signed up for – loot in portable, easily-converted form. They stopped to plunder the camp. Even then they might have caught Mithridates, for the king was reduced to just another body in the flow pushing its heedless way out of the gates. Fortunately, one of the king’s entourage observed the plight of his sovereign. He forced his horse through the throng and, at considerable personal risk, gave his mount to the king. Being on horseback once more allowed Mithridates to put some distance between himself and the Romans and to become reunited with his treasure train.
The Roman cavalry were still closing and Mithridates again became in danger of capture. The enemy had him in sight when the king tried a last desperate stratagem. He – with his own hand, by some reports – slashed open the bottom of some treasure sacks and drove off the mules. A fortune in gold and jewellery was scattered across Mithridates’ line of retreat, and no pack of hounds ever swerved so readily after a red herring as the Roman pursuit which took off on the treasure trail.
A frustrated Lucullus learned that Mithridates had slipped through his fingers and that Callistratus, the king’s chief attendant and one of the men whom Lucullus was most eager to question, had been killed by soldiers who suspected him of concealing gold. Of course, that is exactly what Callistratus had been doing – not in his girdle as the soldiers suspected, but by the waggon load in fortresses and strong points out of Roman reach. This fortune had now to be ferreted out item by item. However, with the Pontic army all but destroyed and Mithridates in headlong flight, Lucullus was prepared to dedicate plenty of time to the job.
The exile
Not all fortresses surrendered readily to Lucullus but, as well as treasure, many Roman sympathizers captured by Mithridates over the years fell into Roman hands and were freed from their imprisonment. Also, Nysia, the sister of Mithridates, became a Roman prisoner, which caused Mithridates to determine that the same would not happen to his wives and concubines. These were kept at a place called Phernacia along with Roxane and Statira, two unmarried sisters who, though in their forties, the king was keeping in secluded reserve for a diplomatic marriage if required, and to prevent them from producing any rival heirs for the Pontic throne in any other case. Since it was impossible to evacuate this harem to safety, Mithridates decided that their death should come before the dishonour of falling into Roman hands.
A eunuch was sent with the king’s order that the harem commit suicide. It is reported that reactions varied, with some praising the king for taking the time to consider their honour even in his hour of need, whilst others, as they died, wished their own fate upon their royal husband. Each chose death as she felt most appropriate. Some took poison but one Monime of Miletus* chose to hang herself by her royal diadem. This was because she had held out against the courtship of Mithridates until she received the status of wife and queen. She got her wish, only to discover that the honour meant that she was secluded in a harem and ignored for long periods by ‘more of a keeper than a husband’. The stifling boredom of the royal seraglio had replaced the usual social interactions of a Greek noblewoman and now Monime decided that the the diadem which had brought her so much misery could release her from it as well. Even in this, her crown failed to meet her expectations. It refused to bear Monime’s weight once she put herself into the noose she had attached to it. After spitting on the broken crown, the unhappy girl finally presented her throat to the eunuch’s dagger.
Mithridates, with 2,000 horsemen, headed for the last sanctuary available to him, the kingdom of his son-in-law, Tigranes. Given that family solidarity was not a strong point among the royal families of the region, it is unlikely that Mithridates was counting on Tigranes welcoming a relative fallen on hard times. Rather, once he had publicly made himself a supplicant to the Armenian power, Tigranes had either to accept his inconvenient guest or openly show his subservience to Rome by meekly handing him over. This was something that the man who titled himself the King of Kings would not do readily, and it was this pride rather than family sentiment on which Mithridates gambled for his salvation. Yet family ties were indeed to save him, albeit in an unexpected way.
Unsurprisingly, Tigranes was less than thrilled with the arrival of Mithridates on his metaphorical doorstep, but, as the Pontic king had calculated, Tigranes was not prepared for the damage which his public image would suffer were he to simply abandon his son-in-law and erstwhile ally to his fate. Mithridates was kept in in the manner to which his royal person was accustomed, but in a castle well away from the Armenian court and guarded so closely as to almost be a prisoner. Tigranes hurried himself off to Syria to further distance himself from whatever happened in Anatolia, and to gain time to prepare for the inevitable envoy from the Romans.
This is where family ties came to the help of Mithridates. Lucullus was hosting in his camp the noble, pig-headed and dissolute Clodius, scion of Rome’s ancient and patrician Claudian house, and his brother-in-law. The antipathy which Lucullus rapidly developed for his brash and arrogant relative probably inspired the idea of sending him as envoy to Tigranes the demand the person of Mithridates. Clodius duly set off for Armenia, leaving Lucullus to settle down to conquering Armenia Minor and returning to the sieges of Amisus and Sinope without the distraction of his demanding relative.7
Arrogant and ignorant he may have been, but Clodius was no fool. He soon determined that Tigranes was trying to run him in circles, from the guides who literally attempted to do so with a highly-circuitous route to the capital, to the messengers who regularly informed him that the king would welcome him personally in the very near future. Clodius ditched his guides and made his own way to the capital, and, by way of encouraging the king to speed up his audience, he began intriguing with various members of the Armenian court, promising some aid from Lucullus if they moved against Tigranes.* Clodius also showed admirable dedication to his duty. Regularly offered an abundance of treasure by Tigranes as ‘presents’, Clodius reluctantly accepted but a single cup – not enough even to avoid insulting the king, let alone enough to count as accepting a bribe.
This stubbornness on Clodius’ part boded ill for the meeting with Tigranes. What Tigranes hoped to gain from the meeting will never be known. Perhaps he hoped that Rome would leave him alone in exchange for a promise that Mithridates would never leave his gilded captivity. It is fair to guess that a suitable incentive, such as offering Tigranes suzerainty of Cappadocia, would have seen Mithridates promptly handed over in chains. However, as his refusal of bribes had shown, the Roman had come with a different purpose in mind.
Clodius, who like all Roman aristocrats was essentially a politician, deliberately chose to imitate Gaius Popillius Laenas, who in 168 BC had drawn a line in the sand around King Antiochus IV and bluntly told him to decide on peace or war with Rome before he stepped out of the circle. This Republican plain speaking and scorn of kings played well at home and Clodius had a political career to build. So, he bluntly told Tigranes to hand over Mithridates or face the consequences. No bribes or concessions were offered apart from this naked threat.
Almost all that could be said for Clodius’ style of is that it produced prompt results. The Roman demand was flatly rejected and Clodius was sent packing. After almost twenty months in a marshy backwater, Mithridates was brought to the Armenian court, honoured by Tigranes as his father-in-law and given a place on the royal council. Whether intending to or not, Clodius had immensely helped the cause of Mithridates by making Tigranes an offer he had to refuse and almost certainly bringing Armenia into the war on the side of Pontus against the Romans.
Mithridates celebrated his return to a position of power in typical fashion, by throwing himself into palace politics and ensuring that those who had shown insufficient passion for promoting his cause came to a sticky end. Mithridates would have been dismayed by the news that the last of his strongholds in Pontus had finally fallen, but would have certainly been encouraged by the dogged loyalty shown to his cause by the defenders of cities such as Sinope and Amisus. Nevertheless, by 69 BC, Pontus was effectively in Roman hands, although the Bosporan kingdom remained under the rule of Menchares, Mithridates’ traitorous son, who was now officially a ‘friend of the Roman people’.
The Armenian campaign
It is possible that Mithridates did his best to convince Tigranes that the Romans would never dare to come against him in Armenia – after all, he wanted his son-in-law to assemble his army for a campaign in Asia Minor. Tigranes himself might or might not have been considering a foreign adventure against the Romans, but he literally refused to consider the idea that the Romans might have the temerity to march on him. The first messenger who came with news that Lucullus was indeed on his way was beheaded for being alarmist.
Nevertheless, the news was correct. Whether Lucullus jumped at the opportunity which the undiplomatic Clodius had presented, or whether Clodius had been requested to produce exactly the outcome which transpired will never be known. In either case, it was Lucullus who went onto the offensive. He did this against a major foreign power which had so far stayed scrupulously neutral, and absolutely on his own authority. The senate in Rome was more than somewhat astonished to hear that they were now at war with Armenia, and that Lucullus had taken the army on what many regarded as virtually a massive private plundering raid. Lucullus, already unpopular in Rome for his sensible settlement of Asia, could expect very little sympathy if anything went wrong with his unauthorized invasion, and could confidently expect to have to defend himself in court even if everything went right.8
Clodius did not accompany his brother-in-law. He had fallen out with Lucullus and took his revenge by stirring up the Fimbrians to the point of mutiny. Being unceremoniously booted from Lucullus’ presence, Clodius joined the army of Marcius Rex in western Anatolia. Marcius gave Clodius a minor command against the pirates. After being captured and ransomed by the pirates, Clodius rejoined the Roman land army and this time did successfully instigate a mutiny, in the course of which he almost lost his life. After this it was decided that Clodius and Asia Minor were not suited to each other and Clodius returned to Rome.*
Meanwhile, in the spring of 69 BC, Lucullus handed command of Pontus to Sornatius (who had acquitted himself well in the Cabira campaign) and headed for Tigranocerta, the Armenian capital. He was accompanied by Murena junior, who probably commanded auxiliary troops, and two picked legions under Lucullus’ personal control. In the context of the Fimbrians, ‘picked legions’ meant the two legions least likely to mutiny, though they were far from delighted with the honour which their commander had done them. Overall, Lucullus probably had some 12,000-18,000 foot and 3,000 horse. Given the length of their service, the two legions were probably severely under-strength and the infantry numbers were made up with auxiliaries – Lucullus prided himself on using native resources whenever possible. The Roman general was marching his surly, unwilling little army into the heartland of an enemy who could be confidently expected to raise ten times as many infantry, and at least twenty times the cavalry. No wonder Tigranes was incredulous.
Tigranocerta lay slightly north of due east from Comana in Cappadocia (which was probably the jumping-off point of Lucullus’ invasion). Therefore, there was only one sensible route for the Roman army to follow – which is along the east-west valley of the anti-Taurus which has Edessa on the other side of the mountain range, and which finishes at the head of the River Tigris, where Mosul stands today. This is the route which Crassus later scorned in favour of taking his army along the more direct route to Parthia, 150 miles to the south, which in his case finished at Carrhae in 53 BC when the Parthian cavalry wiped him out.
The initial part of the route wound between snow-covered mountains and torrents still swollen with snowmelt. There was some apprehension at the prospect of crossing the headwaters of the Euphrates into the small, Armenian-controlled principality of Sophene, but it turned out that the Euphrates was at its lowest in living memory.9 The army could simply wade across, something Lucullus did not hesitate to point out as a signal of divine favour. With Tigranes still wilfully ignorant of their coming, the Romans made their way eastward, the Fimbrians muttering darkly about not being allowed to plunder the countryside as they went along. However, Lucullus would not have wanted his army either slowed down by loot, or to have lost its appetite for plunder before it reached its destination. Besides, nothing encouraged a native population to help an army on its way more than if it was friendly but showing signs of changing its mind.
The Romans rounded the headwaters of the Tigris and were set to fall on Tigranocerta itself before one of the favourites of the King of Kings decided to literally risk his neck by bringing his royal master up to date with developments. For his reward, this man, Mithrobarzanes, was given a force of cavalry several thousand strong along with instructions to crush the impertinent Roman army and bring its commander to Tigranes alive.
Mithrobarzanes seems to have decided that his only chance of surviving his mission was to hit the Romans suddenly with all his strength and hope that they had been lulled into complacency by the total lack of opposition along their march so far. However, Lucullus was an old hand at fighting in Asia Minor and, unlike the average Roman general, he appreciated the importance of scouts. Warned of Mithrobarzanes’ approach, Lucullus sent out a force of some 3,500, infantry and cavalry in equal proportion, and told them to stand the Armenians off until the Romans had fortified a camp. This Roman force met the Armenian cavalry, which attempted to disperse it with a headlong charge. Against experienced infantry who knew how to stand their ground such a cavalry charge was suicidal, and the Fimbrians were very experienced. Mithrobarzanes and most of his force duly perished on the spot.
The Battle of Tigranocerta
The brusque treatment of his advance force finally convinced Tigranes that the Romans were both a serious and imminent threat. As his levies had not yet completed their muster, he decided to withdraw from his capital to Taurus and gather his army about him there. En route, he discovered that he would have to do without his Arab levies, as one of Lucullus’ commanders, Sextilius, came across the Arabian camp, attacked it and dispersed the entire force. Murena junior, meanwhile, was hot upon the trail of Tigranes. At one point he was so close on the king’s heels that Tigranes had to abandon his baggage and make a dash for safety. While this was going on, Lucullus settled down to besiege Tigranocerta. This was the standard Roman campaign tactic – find something the enemy valued and march straight at it. The enemy would eventually muster an army to defend the target of the Roman advance and would be defeated. Indeed, the massive walls of Tigranocerta would not have been an easy nut to crack, had Lucullus intended a serious investment rather than a provocation to Tigranes.
At this early stage of the campaign Tigranes was still confident that he could handle the Roman threat, and paid little heed to Mithridates. In fact Mithridates was not even included in the retinue of the King of Kings and had to make do with frantic letters and pleas transmitted through an ambassador called Taxiles. After his experiences at Cyzicus, Mithridates had become a convert to the idea of victory by malnutrition. Now he pointed in vain to the long and exposed Roman supply lines and argued that Tigranes and his hordes of cavalry could starve the Romans into submission without a battle. His proposal fell on deaf ears. With his precious new capital under siege and a massive multi-national army gathering about him, Tigranes was determined to fight. The hard-won experience of Mithridates was construed as envy that someone else was going to crush the Romans, or alternatively as proof that the Romans had broken his spirit.
It is hard to determine the size of the Armenian force, as Roman chroniclers are prone to exaggeration, but Appian gives Tigranes credit for raising the suspiciously-round numbers of 250,000 foot and 50,000 horse.10 Plutarch claims as his source the actual report of Lucullus to the senate, and this allowed him to give a more detailed breakdown. By this account, 35,000 of the Armenian army were a pioneer and engineering corps, supplying their comrades with wood, bridging small rivers and making roads. The backbone of the army was 150,000 heavy infantry arranged into cohorts or phalanxes according to the preferred disposition of the subject people who had supplied the troops. Of the 55,000 cavalry, Plutarch says 17,000 were completely armoured, probably in the style of Parthian cataphracts. At the fore of this massive force were 20,000 light missile troops, a mix of slingers, archers and javelineers, who by themselves easily outnumbered the entire Roman force.11 By way of contrast, an independent account, admittedly late and fragmentary, comes from one Phlegon of Tralles. He puts the number of effectives in the Armenian infantry at 40,000 – perhaps a more realistically-sized, but still formidable, force. Phlegon’s assessment of cavalry numbers is still high at 30,000.12 Even by this lowest estimate, however, the Romans were still massively outnumbered.
The high morale of the Armenians could be seen in the heroics of a force of 6,000 cavalry who broke through the Roman siege lines into Tigranocerta, then broke out again with Tigranes’ royal harem and a good portion of the royal treasury in tow.
The substantial Armenian garrison of Tirganocerta, strengthened by a large number of Greek mercenaries, was yet another problem for Lucullus. If he turned to face Tigranes, he risked having the garrison fall on his back, yet an army of up to a quarter of a million men was impossible to ignore. The only solution was for Lucullus to divide his already-tiny army, leave Tigranocerta to Murena and 6,000 foot, and take the rest against Tigranes. This gave Lucullus twenty-four cohorts, or about 10,000 men, with about another 1,000 mixed cavalry and skirmishers and an unknown number of auxiliary troops. These formed up on the plain beside ‘the river’. Because we do not know the exact location of now-vanished Tigranocerta, it is not certain what this river was, though it was either the upstream Tigris or a tributary thereof. The river today called the Zgran is a very plausible candidate. We know the battle took place within sight of Tigranocerta (wherever it was), but given the paucity of detail about the battle and the incompatibility of the details we do have, only a rough outline of events can be reconstructed.13
It is reasonable to suppose that Lucullus first intended to take on the Armenians as they attempted to cross the river. Since Lucullus had to use a ford when going the other way, it seems certain the river was not something which could be crossed by an army in battle formation. Therefore it would appear that Lucullus intended to hit the relieving force as it came across the river with an eye to defeating its units in detail as they crossed. However, as the huge Armenian force hove into view, another possibility, based on the enemy’s line of approach, presented itself to him. His scouts had shown him a ford perhaps slightly to the north, where the river curved westward, its course affected by a gently rising plateau on the other side.
At this point the Armenians were not fully deployed and, in any case, were working on the not-unreasonable belief that they were going to be the aggressors in the coming battle. Indeed, Tigranes was far from certain that there was going to be a battle at all. Eyeing the diminutive Roman force, he remarked jokingly ‘if that’s a diplomatic mission, it’s too big. If it’s an army, it’s too small’. Consequently, he was unsurprised when the Romans wheeled left, and trotted away along the river bank. Taxiles pointed out that the Romans were not backing off. The men of a retreating Roman army would have put their shields back into their leather covers – not presented them freshly polished. The Romans were dressed for battle, with plumes affixed to their helmets, a sure sign that they were expecting an engagement.
There was considerable confusion as the large and unwieldy Armenian force was reconfigured to fight its battle on the east bank of the river, a confusion which was more acute because Lucullus had screened his infantry with his cavalry. Consequently, Tigranes was unaware that the Roman infantry had not stopped going once it had crossed the river but instead the legionaries were looping round to gain the higher ground of the plateau at the rear of the Armenian right flank. ‘Are they on us?’ asked the bewildered Tigranes. Still uncertain of where his enemy had got to, he closed the main body of his infantry about himself at the centre, and sent a large force of cataphracts to push the Roman cavalry aside.
Cataphracts were a reasonable choice, since these were very heavily-armoured cavalrymen, and thus more capable of standing against the infantry which Tigranes expected to find behind the Roman horsemen. However, rather than stand and fight, the Roman cavalry gave way. By this hypothesis, they gave way rapidly along their river flank, falling back more slowly toward the plateau, whilst the Armenian cavalry probed forward along the line of least resistance looking for the legionaries who should be there. The cavalry nearest Lucullus on the plateau (Lucullus had chosen to fight on foot with his legionaries) were now almost on the right flank of the cataphracts furthest from the river, and Lucullus saw his chance. His cavalrymen were sword-armed Galatian and Thracian irregulars, and hitting the lance-armed cataphracts on the flank meant that the lighter-armed horsemen could use their swords to beat down the Armenians’ lances before they could be swung around.
This was not going to do more than delay the cataphracts, but it would keep them busy long enough for the Roman infantry charge to hit them standing. The Romans did not bother with their usual preliminary shower of javelins, mainly because the Galatians and Thracians were already engaged, and would be far more affected by friendly fire than the well-protected Armenians. The infantry’s intention was to use their swords to slash at the unprotected shins and thighs of the Armenians, but the cataphracts did not give them the chance. Still struggling with the Roman cavalry, they chose the only possible line of disengagement – a line that sent them crashing into the advancing Armenian infantry.
Either at this point or slightly earlier, Lucullus decided that it was time for the rest of his infantry to make their presence known. Raising his sword, he bellowed ‘Soldiers, we are victorious’, and led the downhill charge. The Armenian baggage train was as confused as the rest of the army, and was keeping close behind the right flank on the assumption that the Romans were still some way ahead. The unexpected sight of Lucullus and his legionaries thundering downhill toward them prompted those manning the baggage train to retreat precipitately, directly away from the Roman advance – and again into the right of the Armenian infantry, which was by now in a state of total chaos.
Lucullus had achieved his initial objective – given his enemy’s overwhelming superiority in cavalry and missile troops, the last thing he had wanted was to be pinned by cavalry whilst his forces were picked off by bows and javelins. This was not an unreasonable fear, and exactly what the Parthians later did to Crassus. Instead, Lucullus had brought his men to close quarters with hardly a casualty, and his legions were now chewing through the right of the Armenian army whilst the rest of Tigranes’ force was still working out what had hit them.
Unsurprisingly, the Armenian right broke, and their panic communicated itself to the troops next in line. The baggage handlers were already in full flight along with much of the cavalry and to the unsteady Armenian levies it seemed that the battle was already lost. The same thought was preoccupying Tigranes, who had been off-balance since before the battle commenced, and who was keenly aware that the tumult of the Roman advance was getting uncomfortably close to his own position. This was the moment to make a headlong charge at the Romans and hope that the remainder of his army would take inspiration from his valour. Of course, there was the embarrassing and all-too-likely prospect that this charge and Tigranes’ subsequent death would be in vain, which left the second option – to follow the example of those even nearer to the Romans, and make a timely exit from the battlefield.
Tigranes opted to depart, and with him went the last chance of Armenian arms salvaging anything from the day. The battle became a rout.
The following series of diagrams offer a highly speculative look at how the battle of Tigranocerta might have happened. Though this sequence is consistent with the available evidence, other interpretations are certainly possible. Tigranes’ army is in black, Lucullus in grey. Thin lines represent cavalry in skirmish order.
Battle of Tigranocerta
Phase I
Phase II
Phase III
Phase IV
Phase V
Phase VI
After Tigranocerta
Just as with the numbers of the army before the battle, the number of casualties afterwards are highly unreliable. For a start, it is improbable that anyone was counting, and secondly the Romans pursued the Armenian rout for a considerable distance. Therefore we get a body count as high as 100,000 (Plutarch) and as low as 5,000 (Phlegon). Nor are the two incompatible. If Plutarch drew his figures from claims that Tigranes lost 100,000 men and Phlegon says the Romans killed 5,000, it may be that 95% of the army lost their taste for soldiering and headed directly home without stopping to collect their back pay.
Lucullus was well aware that scattering an enemy army was not the same as destroying it, and he had given very strict orders that killing and not plunder was to be given priority. The Fimbrians, whatever their other faults, were a dependable force on the battlefield. But they would have been aware that tens of thousands of cavalry were still unaccounted for, and were not going to break into very small groups to hunt down their scattering foes.
Indeed, had they done so they would may well have come to grief, because there were probably 2,000 highly-disciplined cavalry close at hand. This was the personal guard of Mithridates who had been late arriving on the battlefield, quite possibly because he had a pretty good idea of what was going to happen there. If pessimism was indeed the cause of his tardiness, he would have been unsurprised to meet first a trickle and then a flood of panicked Armenian troops heading in the opposite direction. In any case, once he had ascertained that there would be no further fighting that day, he directed his energies to finding Tigranes. Given that Tigranes’ negotiating position had just sharply deteriorated, Mithridates now needed the goodwill of his son-in-law more than ever. Therefore he carefully refrained from pointing out that he had told Tigranes so. Instead he concentrated on raising the chastised king’s morale, handing over his personal guard with instructions that they should look after his crestfallen relative.
The Romans now received an unexpected bonus. The commander of the garrison of Tigranocerta, one Mancaeus, had witnessed the dispersal of the Armenian army from the walls. Given that the Romans were now in command of the land outside, he began to doubt the loyalty of his Greek mercenaries. Therefore he ordered them to hand in their weapons. This the Greeks took as a preliminary to their arrest and possible massacre, so they gathered together and started making themselves clubs from whatever heavy objects were at hand. When Mancaeus sent troops to disarm them once more, the mercenaries wound their cloaks about their shield arms and attacked with their makeshift weapons. Success gave them access to real weaponry from the defeated Armenian detachment. The Greeks fought their way to a section of the walls between the towers, and invited the delighted Romans to enter the city as their allies.
Finally the Fimbrians had all the plunder they could carry, though Lucullus reserved the state treasury for himself. For the next week or so he amused himself by having the Greeks in the city put on a series of plays and pageants before allowing any who so wished to return to the original homes from which Tigranes had forcibly transplanted them.
* One of the precautions the defenders of the city had taken was to lay in a stock of wasp nests and wild bears. When the Romans began the traditional attempt to undermine the walls of the city, the Pontic defenders countermined the tunnels and unleashed bears and wasps, singly and in combination, upon the Roman miners.
* This may or may not be the same person as Monima of Stratonice (p.41)
* One who listened to the Roman promises too attentively was King Zarbienus of the Gordyenians, who paid with his life for being a suspected conspirator.
* One of the first things Lucullus did on returning to Rome was to break the family connection with Clodius by filing for divorce.