Tigranocerta: 69 BC

The battle came about as as side-issue in the on-going struggle between Mithradates V, king of Pontus, and Rome; here represented by Lucullus. Mithradates, whose name was usually mis-spelt Mithridates by classical authors, had married one of his many daughters to Tigranes, king of Armenia, and had taken refuge with him; Lucullus pursued Mithradates into Armenia, threatened Tigranes' new capital, Tigranocerta, and engaged the Armenian King's forces there, inflicting a serious defeat. Despite another successful encounter with Tigranes the following year, Lucullus was however nonetheless obliged to withdraw from Armenia, and his political enemies back home managed to have him deprived of his commission. He had to wait many years before he could eventually celebrate a triumph for his battlfield successes in the east.

Of the sources detailing the battle. Plutarch is the standard reference (hereafter 'Luc.'), being the most detailed, as one of his 'Lives' is that of Lucullus. Appian gives a shorter account as part of his 'Mithridatic Wars' (hereafter 'MW'), and the almost-forgotten Phlegon of Tralles, a portion of whose 'Collection of Chronicles and List of Olympian Victors' was epitomised by the medieval Byzantine bibliophile Photius, includes some brief but important details. Plutarch's account of the campaign seems to be mostly derived from lost accounts by Livy and Sallust, as he quotes both in several places; he also quotes Strabo and one Antiochos the Philosopher. He several times refers to Lucullus' letters to the Roman Senate, although it is not clear if he had access to them, or, as seems much more likely, he is quoting references to them in Livy or Sallust. Appian is similar to Plutarch and is evidently drawing upon much the same source tradition, if not always quoting exactly the same sources at the same time. Other sources give almost no useful information: Dio Cassius would have been useful, but there is a gap in Dio's extent works that picks up just after the battle, however the late epitome ('Breviarum') of Eutropius contains a few snippets.

Plutarch and Appian both make what seem, at least to modern eyes, outlandish claims as to the size of the Armenian army (300000 men, MW 85; 260000 men, Luc. 26.2) and almost equally absurd claims for small size of the Roman force (2 legions and 500 horse, MW 84; 11000 men, Luc. 27.2). Internal evidence from both disproves these figures and also allows reasonable estimates as to the true quantities to be derived. Phlegon's numbers are markedly different, not to say more realistic in appearance, and are obviously drawn from an entirely different source tradition; they are an important check.

The Roman side

According to Appian, Lucullus had started his initial campaign against Mithradates with a single legion from Rome, took charge of two legions already in Asia that had previously been commanded by Fimbria, and added two others by which time he had 30000 foot and 1600 horse (MW 72). Livy records that Lucullus' initial legion was raised personally (Luc. 7.1), and that on arrival in Asia he added not just Fimbria's forces, but the other Roman forces in the area (Luc. 7.2). It is tempting to conclude that the 'two other' legions in Appian are those that were then under the command of Cotta, who was also campaigning in the general area against Mithradates at the time. However, Plutarch says (Luc. 8.5) that Lucullus had the above-mentioned forces, some 30000 foot and 2500 horse, before he relieved Cotta. Therefore after relieving Cotta, Lucullus could call upon 5 legions in addition to those previously under Cotta's commanded, numbers unknown but likely two, this being the typical size of such armies in the area, eg. Fimbria's forces after 85 BC. Accounting for the 4000 casualties Cotta had lost in the meantime (Luc. 8.2), Lucullus therefore had something like 7 legions (two very depleted however) of perhaps 25000 men given some legions fresh, and presumably full-strength, but others more combat-weary. When Appian says 'added two more' this presumably then refers to two legions being recruited locally; this being the era in which recruitment of locals, even non-citizens, was just beginning in earnest. One could presume legions of local 'Italians' are meant, since raising legions of non-citizens could still cause political complaints 20 years later, and Mithradates had earlier reputedly massacred 80000 Italians in Asia, so there was likely a large pool of potential recruits in the area despite his depradations in the meantime. However, Galatians (or Gallo-Greeks as Livy meaningfully calls them) and Greeks are at least as likely; and probably a mixture of all three. Certainly Lucullus emphasised to the Senate his success in wage the war by relying on local military resources rather than having to draw upon Rome (Luc. 13.4).

Despite having access to 7 legions, not all would have been at the battle however. Two substantial forces were absent. The first, 6000 strong (Luc. 24.1) under the command of Sornatius and Triarius, was in Pontus trying to hold down that region in Mithradates' absence. As Sornatius at one point (Luc. 17.1) commanded a force of 10 cohorts on a grain-procuring expedition; these 6000 would appear to comprise a legion plus enough auxiliaries to make the numbers up to 6000. The second force was under Murena, the future consul, besieging Tigranocerta itself. It too consisted of 6000 men, and, as befits a besieging force, is specifically said to be infantry (Luc 27.2). It too was likely also a legion and associated auxiliaries. Thus Lucullus likely had 'only' 5 legions at the battle.

This conclusion appears to be at odds with various statements in Livy and Appian. Appian for one says that Lucullus marched against Tigranes with two picked legions and 500 horse (MW 84) and Plutarch (Luc. 24.1) says 12000 foot and less than 3000 horse; likewise Frontinus gives Lucullus 'not more than 15000' (2.1.14). Even if accurate, these figures are contradicted elsewhere, eg. Eutropius (6.9) gives 18000 men, and even within Plutarch and Appian there are inconsistencies: after detaching Murena's 6000 men, Plutarch for instance says Lucullus led out to battle (as opposed to marching out at the start of the campaign) 24 cohorts of 10000 heavy infantry plus 'all' 1000 of his slingers, archers and horse (Luc. 27.2), perhaps therefore 500 horse and 500 light infantry, and clearly you need more than 12000 men to detach 6000 from and still have at least 10000 left. The conclusion is that the men Lucullus 'led' were those he personally led, as opposed to those of his subordinates whom nethertheless were part of his army. This is partially confirmed by Plutarch (Luc. 28.3) who reports that the infantry that Lucullus led onto the hill from which he charged down, whom in Appian are intimated to be all of 'his' infantry, were two speiras strong, usually translated as cohorts. As it seems extremely unlikely that Lucullus would have entrusted himself to the protection of a mere two cohorts in open battle, probably less than 1000 men, let alone used such a miniscule force to attack the main enemy strength, it seems apparent that two legions, which Appian reports as Lucullus' entire infantry force, are meant rather than two cohorts. Speiras is essentially the same word in Greek as maniple is in Latin, a 'handful' of troops. Maniples had ceased to be tactical units in the 1st century BC, so it is not clear what unit is being referred to here: a century seems far too small, as indeed does the larger cohort. To my mind legion is the only unit that makes sense, even if a legion was more usually called a telos ('host') in Greek.

In addition to these 'Roman' forces of conventional legionary heavy infantry, Lucullus had access to various kinds of auxiliary forces. Plutarch (Luc. 28.2) refers to Galatian and Thracian cavalry; the latter possibly being local Bithynians (Luc. 14.1) rather than European Thracians. The Galatian ruler Deiotaros independently fought against Mithradates' forces (MW 75) and at one point Lucullus made direct use of 30000 Galatian porters (Luc. 14.1) so it is possible that in addition to the two locally raised legions who may have been or included Galatians, there may have been additional Galatian foot. However as Deiotaros is recording as raising no more than three legions at a time from his own subjects 20 years later, when his territories were more extensive, two legions' worth of infantry at this time may have been the limit of his capabilities. Plutarch (Luc. 27.2) also mentions archers and slingers, and these too would most likely have been recruited locally. Appian notes (MW 84) that when Lucullus moved down the Euphrates he only required supplies from the locals rather than military assistance, with the clear implication that other areas he subdued had to supply men; and before the battle Lucullus had subdued very extensive portions of Asia Minor. Pontic archers are particularly likely, as Machares, the son of Mithradates and the governer of Pontus, had already asked to be admitted as a Roman ally before the battle (Luc. 24.1). Kappadokians are mentioned by Strabo as assisting against Mithradates (12.2.1) and therefore most likely against Tigranes too.

The number of auxiliary forces is harder to estimate than that of the legions. Auxiliaries seem to have only rivalled legionary numbers in Spain, but Spanish auxiliaries were particularly well-regarded. Earlier in the campaign, Lucullus' cavalry was reported to be inferior to a Mithradatic force including 4000 horse (MW 80, Luc. 15.1,3) and the greatest number of horse Lucullus is said to have is 'less than 3000' (Luc. 27.2); there seems little reason to dispute this particular number. If Murena and Sornatius both commanded 6000 men including a legion, it would seem the total number of all auxiliaries might be reckoned at something like 50% of that of the legionaries. However, these forces quite likely had a higher proportion of auxiliaries than would have had the army as a whole, given their missions. I would guess conservatively the army at Tigranocerta had perhaps only 1 auxiliary per 3 legionaries all told. This would tie in with Plutarch's statement (Luc. 8.5) that Lucullus had 30000 foot at a time when he had only five legions; the legionary component of these might well have been over 20000, if not by very much.

With regard to officers, we are told that sometime before the battle (MW 79) Lucullus had a cavalry commander (Pomponius?, Luc. 15.2), but this position does might not seem important enough to warrant classification as a DBM general. We know the names of several of Lucullus' legates: Sornatius, absent defending Pontus; Murena, the future consul, absent besieging Tigranocerta; Triarius, also absent in Pontus (possibly and/or Delos); Fannius; Hadrianus and Sextilius. The latter three are not mentioned in connection with events far from the battle at the time, and so were probably present (although Appian has Sextilius besiging Tigranocerta, with no mention of Murena).

Lucius Fannius later operated idependently (Dio 36.8) and got into trouble with Tigranes before Lucullus sent him assistence and thus seems suitable qualified for sub-general status. M. Fabius Hadrianus later operated in Cabira and run into trouble against Mithradates before being relieved by Triarius, he had previously inflicted a severe defeat on Mithradates' cavalry forces (Luc. 17.1). He is probably the strongest candidate for a sub-general; because of his defeat of the enemy mounted, he might qualify as a mounted commander. Sextilius (possibly the same as the Secilius of Dio 36.3.2-3) had before the battle commanded a mixed force of horse and foot serving as a screening detachment (Luc. 25.3), and might similarly qualify as a sub-general. Lucullus himself seems to have led his legionaries on foot, "enduring all the fatigue of a foot-soldier" (Luc. 28.3).

The battle was quite a large one, so it might be impractical to use the standard DBM scale of 1 element equals 250 men. At one element equals approximately 500 men, or, for legionaries, one cohort, the Roman force can be postulated as follows:

Lucullus' command

1 x Lucullus - Reg Bd (S) C-in-C
19 x Ex-Fabian Legionaries - Reg Bd (S)
1 x Archers and slingers - Irr Ps (O)
1 x Macedonians - Irr Cv (O)
22 elements, 21 1/2 ee, breakpoint: 7 1/2 ee

1 x Fannius - Reg Bd (O) Sub-general
3 x Italian Legionaries - Reg Bd (O)
1 x Archers and slingers - Irr Ps (O)
4 x Pontic Archers - Irr Bw (I)
4 x Kappadokian auxiliaries - Irr Ax (O)
13 elements, 12 1/2 ee, breakpoint: 4 1/2 ee

1 x Sextilius - Reg Bd (O) Sub-general
9 x Ex-Cottan Legionaries - Reg Bd (O)
6 x Italian Legionaries - Reg Bd (O)
16 elements, 16 ee, breakpoint: 6 ee

1 x Hadrianus - Reg Cv (O) Sub-general
3 x Bithynian and Thracian cavalry - Irr Cv (O)
10 x Ex-Cottan Legionaries - Reg Bd (O)
4 x Bithynian auxiliaries - Irr Ax (O)
1 x Archers and slingers - Irr Ps (O)
19 elements, 18 1/2 ee, breakpoint: 6 1/2 ee

Although some of the legions of the seven available to Lucullus had been relatively recently raised, they had by this stage seen extensive campaigning (several years and many encounters with the enemy), and as such deserve (O) rather than raw (I) status. In any case, one or both may have been absent from the battle. The two 'Fimbrian' legions (called in Dio 36.15 'Valerians'; Fimbria having illegally taken them over from Valerius Flaccus in 85 BC) were extremely seasoned veterans of the Sullan-era; albeit ill-disciplined in Lucullus' hands, and were mostly likely present. I accordingly grade them as (S), but not irregular - their ill-discipline seems to have been of the off-field variety. For the purposes of this reconstruction, I shall assume the two most-recently recruited legions are the two absent from the field, but the classifications would not change if they were not. The ex-Cottan legionaries were veterans too, but not of as long standing; and their numbers were very much depleted; hence I have kept them as (O). Plutarch has Lucullus with 10000 legionaries in 24 cohorts; Appian two units. I reconcile these by assuming two legions under direct command of Lucullus, and a smaller group under a lesser officer: Fannius. The cavalry are rated as Cv rather than LH as they were ordered to attack the enemy catafracts with their swords (Luc. 28.2); Appian however has the Roman cavalry harrass the enemy and thus draw them on into an unmanageable pursuit; the two need not be contradictory. Macedonian horse are mentioned by Frontinus (2.7.8), albeit with no indication of time and place, as being unreliable and attempting to desert Lucullus; and are thus graded irregular; Thracians actually did desert Hadrianus in 68 BC.

In addition to these troops, the Romans should need a fortified camp; however it would seem that this was sufficiently far enough away to not appear on a DBM-sized battlefield.

The Armenian side

Appian (MW 85) records 250000 foot and 50000 horse. Plutarch (Luc. 26.6) records 55000 horse, of which 17000 were catafracts; 20000 slingers and archers; 150000 heavy infantry, and 35000 pioneers and other labourers brought along to swell the apparent size of the army. The 150000 heavy foot were said to be partly organized into phalanxes, and partly into cohorts; Appian however says (MW 87) that it was only after this battle that Mithradates organized Tigranes' forces into squadrons (ilas) and cohorts (speiras) 'as nearly as possible in the Italian manner'.

The overall numbers are clearly bogus. Feeding such a horde would have been almost impossible, especially since the major food-storage site in the area - Tigranocerta - was under siege at the time; and given that Mithradates advised Tigranes to avoid battle (Luc. 26.3) and concentrate on cutting Lucullus' supplies, one must conclude that Tigranes' force was only a fraction of its purported size (it would have been Lucullus aiming to cut Tigranes supplies, as he had done before to Mithradates, if Tigranes' army was of such an enormous size). This is bolstered by the inclusion of the labourers to bulk up the army's apparent size. An army that so vastly outnumbered its foe would have little need of making itself look even bigger; and indeed, it is hard to see how such men could have been seen, stationed as they were at the rear, through such an awesome throng. In any case, 35000 extra men, large enough number as it is, represents less than a 15% increase in size from the purported totals above!

It seems to me that the Armenian numbers (save perhaps for the labourers) have been greatly inflated to make the victory sound all the more impressive. It is not hard to see how and why; the source for the numbers is Lucullus himself, in his letter to the Senate describing the victory (Luc. 26.6). Naturally, Lucullus had a vested interest in promoting his victory in a favourable light; and we know from another letter of his that he was all too capable of exaggerating his achievements with regard to Tigranes (Luc. 35.5). The Senate was far away, and how was it to know any better? If Caesar could play a similar numbers game in much nearer Gaul a few years later, Lucullus certainly could have done so in distant Armenia. That the Armenians outnumbered the Romans by over twenty to one, as Livy claimed (Luc. 28.7) is blatant propoganda, albeit following in the best of classical traditions. Later sources, such as Eutropius' 'Breviarum' are even worse: 600000 catafracts and 100000 archers and other troops. The casualty discepancies are if anything even more marked: Plutarch (Luc. 28.6) claims well over 100000 dead Armenians compared to a mere 5 Roman dead and 100 wounded!

Fortunately, we have one important literary check on these numbers - Phlegon of Tralles. The passage in Photius is short, but informative: "At that time [ie. the 177th Olympiad, 72-69 BC] Lucullus was besieging Amisus, but having left Murena with two legions to carry on the siege, he himself set out with three others against the territory of the Cabiri, where he went into winter quarters. He also ordered Hadrian to make war against Mithradates, who was defeated... In the fourth year of this Olympiad Tigranes and Mithradates, having collected an army of 40,000 foot and 30,000 horse, who were drawn up in the Roman order of battle, engaged Lucullus and were defeated; Tigranes lost 5000 killed, a large number of prisoners, besides a promiscuous rabble".

This confirms that Lucullus' available manpower for the campaign as a whole was considerable: in addition to the three legions he had led against the Cabiri (Cabira in Plutarch, where it is a place rather than a people), and the two under Murena, there were sufficient military resources for Hadrian to mount his successful campaign against Mithradates. According to Plutarch (Luc. 17.1,2) it was two of Mithradates' subordinates who were defeated by Hadrian, and it was Hadrian's subsequent ostentatious display in front of Mithradates' camp that caused the King's forces to disintegrate. Hadrian was on a grain-procuring mission as Sornatius had been before, and likely commanded a force of at least similar size, and possibly greater if he had been truly ordered to provoke Mithradates.

The mention of a 'promiscuous rabble' 'besides' the other other casualties backs up Plutarch's account of civilian labourers being somehow involved in the fighting. According to Plutarch (Luc. 29.1), Mithradates was not present in person at the battle; Appian (MW 85) is not clear on the point. On balance it seems preferable to follow Plutarch's more detailed account; Appian and Plutarch stress that Mithradates was certainly represented and had a hand in things, but this need not mean he was actually there in person; this may be echoed in Phlegon when it is said that Tigranes (as opposed to Tigranes and Mithradates) lost men.

The notice about being drawn up in the Roman order (and I have not been able to locate a Greek original to check the exact wording here) would seem to disagree with Appian regarding the post-battle institution of Roman military practice in Tigranes' forces, and confirm Plutarch. Perhaps the truth lies in a mixture of all three: the battle formation was Roman, but only a portion of the men were (as yet) organised in Roman-style units. Incidentally, Frontinus (2.1.14) has Lucullus attacking before the Armenians had actually drawn up their order of battle properly.

The casualty count, 5000 men, is of an entirely different order to those of Appian and Plutarch; and an entirely believable one at that, and makes me believe that the estimate of the army's size might well be credible. 40000 foot (not counting the 'promiscuous rabble', it seems) is certainly credible for an army drawn from not just Armenia but also Gordyene, Arabia, the Caucasus and Media amongst other places (Luc. 26.4). 30000 horse might appear to be a bit excessive, given how many horsemen Achaemenid Persia or the Successors of Alexander could draw from some of these places, but perhaps not wildly so; Eumenes was able to recruit at least 5000 horsemen from seemingly just Kappadokia in quite a short period.

Having said that however, a ratio of somewhat over 40% horsemen is undoubtably high. In contrast, the ratios given in Appian and Plutarch are a fraction under 20%. Even 20% exceeds the ratios found in typical classical armies such as that fielded by Lucullus, but this is only to be expected; Armenia, despite its elevated geography, produced many horsemen, and was itself close to the steppe with its abundant horse archers. Indeed, the ratios in Plutarch and Appian are paerhaps troubling for their smallness if anything, just as their overall numbers are for their greatness. For instance, after the battle, the new soldiers enrolled in the Armenian army were said to be 1/3 cavalry (MW 87; reputedly 70000 foot and half that many horse), and we have no reason to believe that cavalry were expressly emphasized at this time; quite the contary, given the emphasis on them being in Italian-style formations.

Furthermore, Plutarch relates (Luc 28.1) that the Romans needed to "come to close quarters with men who fought with long range missiles, and eliminate, by the rapidity of their onset, the space in which archery would be effective", and yet allocates a mere 20000 slingers and archers on foot out of over 200000 footmen; and at most, 38000 out of 55000 cavalry as missile-armed. Similarly, Dio, describing an army led by Mithradates in Armenia and probably composed mostly, if not entirely, of Armenians, says "they were horsemen and bowmen for the most part" (36.49.6). Thus the amount of heavy infantry in Plutarch's army composition has been massively inflated compared to the other arms (Appian gives no breakdown in terms of heavy and light infantry).

Assuming for the moment that Phlegon's figures are accurate, and that the 20000 archers and slingers on foot in Plutarch are also true, that would leave 20000 heavier infantry, plus the 'promiscuous rabble'. Since the missile-armed troops would typically be formed in front of the other troops, they would not add to the length of the battle line (a few, such as Arabs, could have formed up in denser formations and have been put into the line of battle, as they were at Raphia in 217 BC, but there were probably not many Arabs present, as 'most' had been killed by Sextilius in a previous skirmish (Luc. 25.2-3). Accordingly 35000 labourers, as reported by Plutarch, would certainly be adequate to appear as a second line bulking up the numbers of the army.

What then of the mounted troops? 6000 horsemen were said (MW 95) to have been dispatched to Tigranocerta before the battle; taking these off the 30000 of Phlegon leaves a still large 24000 effectives present at the battle. Mithradates had fled to Tigranes with 2000 horse (MW 82) who might well have been present even if their commander wasn't - Plutarch (Luc. 26.3) has Mithradates' general Taxiles joining with his 'forces' which implies a certain number of men in addition to himself; Taxiles was positioned with Tigranes during the battle and thus these men should be under his command. Of the other 22000 horse; I assume 1/3 would be catfracts, since that is the ratio (if not numbers) Plutarch reports, and the rest horse archers. Most of these catafracts were apparently on the Armenian right wing (Luc. 27.6) under the Median Darius (Appian) or another Mithradates (Dio 36.15.2).

The left wing

1 x King of Adiabeni - Irr Kn (I) ally-general
3 x Adiabene and other catafracts - Irr Kn (I)
15 x Assorted horse archers - Irr LH (F)
2 x Arabs - Irr Bw (I)
2 x Greek mercenaries - Reg Sp (I)
8 x Archers and slingers - Irr Ps (O) @ 2
31 elements, 27 ee, breakpoint: 9 ee

1 x Tigranes - Irr Kn (I) C-in-C
3 x Armenian catafracts - Irr Kn (I)
4 x Taxilies' Pontic cavalry - Irr Cv (O)
20 x Archers and slingers - Irr Ps (O) @ 2
16 x Phalangites - Reg Pk (I)
4 x Imitation Legionaries - Reg Bd (I)
8 x Other trained foot - Reg Ax (O)
17 x Labourers - Irr Hd (O)
6 x Irr Bg (I)
79 elements, 50 1/2 ee, breakpoint: 17 ee

1 x King of Media - Irr Kn (I) ally-general
6 x Median and Armenian catafracts - Irr Kn (I)
14 x Assorted horse archers - Irr LH (F)
8 x Untrained foot - Irr Ax (O)
12 x Archers and slingers - Irr Ps (O) @ 2
41 elements, 35 ee, breakpoint: 12 ee

The catafracts are graded as (I) and not (X) for three reasons. Firstly, Armenian catafracts were not 'fully' armoured; their unarmoured thighs were a weak point when attacked from the flank (Luc. 28.4) - presumably they were protected frontally by the standard built-up saddle of the area, and thus to not fit into the Kn (X) category. Secondly, when attacked by Lucullus' cavalry, they were drawn out of formation: in DBM Kn (X) do not follow-up recoiling enemy, but Kn (I) do. Thirdly; their performance at this battle can only be described as 'inferior' - the only thing exceptional about it was how exceptionally bad it was! It is worth noting that prior to the next battle Tigranes enrolled some Iberian lancers because he deemed them more warlike than his own. Most of the Greek mercenaries were being besieged at Tigranocerta (MW 86), but I assume a few might have been present at the battle.

Fighting the battle:

Turning to the battlefield itself, two major terrain features defined the battlefield. Central to the course of the battle in both Plutarch's and Appian's accounts was a large hill somewhat to the rear of bulk of the Armenian army, from which Lucullus, having taken an apparently unobserved route over it with his infantry (MW 85), charged down from into the Armenians as they were distracted by his cavalry from another direction. The approach to the hill as the Romans saw it was said to be neither rough nor steep and the top of the hill was flat and broad (Luc. 28.2; Frontinus 2.2.4). The second feature was the river which the Romans crossed in its shallowest portion to reach the battlefield. The Armenian army was on the eastern side; the shallower portion was where the river 'turns westward' (Luc. 27.4). Evidently the width of the fordable part was not too restrictive, as the Roman cohorts formed up in maniples to cross (as opposed, say, to crossing in single files). The westward bend was such that the Romans appeared to be moving further away from the Armenian host as they prepared to cross it; Tigranes is said to have thought them retreating until Taxiles pointed out they were in full battle-dress, not in their marching gear.

Accordingly the main part of the battlefield can be reconstructed as below, scaled at 15mm scale to a 2400 by 1200 mm table (8' by 4'):

In DBM terms, the hill should count as bare good-going, with gentle slopes. The top part should be counted as flat, with a crest separating it from the slopes reaching down to the plain represented by the rest of the table-top. The river should count as tricky, and just slightly less than 100p wide, except in the north-west portion depicted in above map as speckled, which should count as easy.

Plutarch tells us that Lucullus led out his force at daybreak on October 6th (Luc. 27.4,7); so the campaign season is autumn. As the above battlefield is somewhat too small to accommodate all of the pre-battle force positioning, we need to allow some time for this. Accordingly, the battle should start at 0700 hours. The weather is clear with no unusual effects; for a more "what-if" scenario, you may wish to dice for both time and weather.

Special rules:

Lucullus' command is allowed to make a flank-march, in contradiction to the usual rule that CinCs may not flank march. In this case, each Roman general must dice independently until the CinC's element is on the table; however, the CinC may swap his PiP die with any one sub-generals' dice even when off-table. This means he has a very good chance of arriving when he wants to. Furthermore, if he flank marches on the left, his approach is covered by the hill, so arrives immediately, rather than the following bound.

No special victory conditions are necessary, and apart from the above flank-marching rule, no special deployment rules are required either. As the Romans are invading, the Armenians start to deploy first, and the Romans take the first bound of play.

References:

Appian's Mithridatic Wars, Loeb reprint of 1972 (published 1912), translated by Horace White.
Plutarch's Lucullus, Bill Thayer's on-line version of Bernadotte Perrin's 1914 Loeb edition.
Photius' Bibliotheca, including the Collection of Chronicles and List of Olympian Victors, translated by Roger Pearse.
Cassius Dio, book 36, also on Bill Thayer's site.
Eutropius, book 6, from an 1886 translation on the Forum Romanum site.
Frontinus' Stategemata, Bill Thayer's on-line version of Charles Bennett's 1925 Loeb edition.
An updated PDF version of the 2nd volume of the Magistrates of the Roman Republic.