Collections: Phalanx’s Twilight, Legion’s Triumph, Part IIIa: Peak Pike-Phalanx

This is the first part of the third part of our four(ish) part (Ia, Ib, IIa, IIb, IIIa, IIIb, IVa, IVb, IVc, V) look at the triumph of the Roman legions in the third and second century over the Hellenistic armies of the heirs of Alexander. Last time, we looked at some of the operational and strategic advantages that the Roman legions possessed. Now we turn from theory to the actual combat record of these armies. We’ll begin not with the failure of Hellenistic armies, but with some of their notable third century successes, looking especially at situations where these armies succeeded at the things they are, in theory, supposed to be bad at.

The reader may be pardoned for assuming, at the end of the previous part, that we would pick up with Roman legions screaming towards a Hellenistic sarisa-phalanx,⁶¹⁸ but before we get to those victories in the second century BC, I think it is worth pausing for a moment to look at the performance of the Macedonian phalanx in the third century. In particular, because I think it is possible to come away from a lot of discussion of Hellenistic armies with the somewhat simplistic assumption that these armies were such fragile, exquisite creatures that they could only fight on perfectly flat, featureless plains against entirely cooperating enemies.

And as we’ll see, that simply isn’t the case. Hellenistic armies can, and do, win battles on rough terrain. They can, and do, compel reluctant enemies into pitched battles on favorable terms. They can, and do, defeat tactically challenging adversaries like horse archers and withstand the crushing rush of Gallic warriors. In short, before we can see how Roman armies defeat this military system, it is worth noting that it worked and indeed, worked quite well.

So we are going to look at, in as much as the sources permit, some of the successes of Hellenistic armies in the third century. In particular, I want to cover the notion that Hellenistic armies could be easily foiled by even modestly rough terrain with the Cleomenean War (228-222), that they were easily beaten by horse archers, focusing on Antiochus III’s invasion of Parthia (210-208), before looking at the combination of tactical innovation and strategic weakness that defined Pyrrhus of Epirus’ campaign in Italy against the Romans (280-275). In order to keep that scope manageable (as much for my writing time as anything else) we’re going to split this into two sub-parts, dealing with hills and horse archers this week, and Pyrrhus of Epirus next week.

Now I should say at the outset that our sources for this period are quite bad. We often have very incomplete descriptions of these battles or conflicting accounts (often in much later sources). So our ability to reconstruct details may vary here, though we can be quite confident about outcomes. In particular, I want to note that my reading of the always confusing and often confused sources for Pyrrhus’ three major battles follows P.A. Kent, A History of the Pyrrhic War (2020) which is also absolutely the thing you should read if you want to understand the course of Rome’s famous war with Pyrrhus of Epirus. I’ve also leaned here in N.L. Overtoon’s Reign of Arrows (2020) and M.J. Taylor, Antiochus the Great (2013), the former with rather more reservations (some of which will come up) than the latter.

I want to note at the outset that while this seems like a tremendous amount to cover, in a lot of cases what we can say is sharply limited with how few sources we have. Often all we know about these battles is that the Hellenistic armies won them, or we have conflicting and confusing accounts from ancient sources. In many cases, the battle itself is beyond confident reconstruction. Nevertheless, Hellenistic armies are able to win in a wide variety of difficult circumstances and that is the core point here: this military system worked, generally speaking.

But, as always, if you want to help this project reach its goal of recruiting 20,000 phalangites reaching more readers, please share this essay and others, as I rely on word-of-mouth for gaining new readers. And, if you want to support the growing ACOUP military-tributary complex, you can contribute your tribute by supporting the project on Patreon. If you want updates whenever a new post appears, you can click below for email updates or follow me on twitter (@BretDevereaux) for updates as to new posts as well as my occasional ancient history, foreign policy or military history musings, assuming there is still a Twitter by the time this post goes live. I am also on Bluesky (@bretdevereaux.bsky.social) and (less frequently) Mastodon (@bretdevereaux@historians.social).

Email Address

Subscribe!

The Hills of Sellasia

One of the common things said about the Macedonian phalanx and the Hellenistic armies it depended on is that it simply isn’t functional in rough terrain. And there’s not nothing to this. The claim is based on Polybius (18.31.5-12) who says as much, that ditches, ridges, trees and water will all disrupt the phalanx and that it is hard to find a battlefield without such features. That passage of Polybius, in turn, seemingly gets endlessly exaggerated and essentialized in the pipeline from specialist literature to textbook to introductory course to history teacher to high school course, leading to a popular perception that these formations just stop working on any terrain more broken up than a football pitch.

But this is a case where we ought to be fairly cautious of Polybius’ simplification here. Hellenistic armies (and their Macedonian phalanxes) do operate in rough terrain, sometimes using it to their advantage, sometimes being disadvantaged by it. Now, we ought not swing too hard the other way: the Roman legions do, as we’ll see, seem to be less hindered by broken ground. But this is a fighting system developed in ancient Macedonia and Greece (the former famously forested and both broken up by numerous streams, mountains and hills) rather than the vast open Eurasian Steppe. Dealing with rough terrain some of the time was always part of the bargain.

Via Wikipedia, a map of the Antigonid kingdom (labeled Macedonia, in orange) overlayed over a topographical elevation map, which ought to give a good sense of just how broken up the homeland of these armies was.

Terrain disruption and roughness seems to have been a factor in both Pyrrhus’ two victories (Heraclea in 280 and Asculum in 279). At Heraclea, Pyrrhus seems to have used the river Siris to his advantage, forcing the Romans to ford it to attack him (Plut. Pyrrh. 16.6; Zon. 8.3).⁶¹⁹ At Asculum, the rough terrain prevents Pyrrhus’ elephants and cavalry from operating, but his infantry functioned just fine, leading to a bloody but inconclusive first engagement (Plut. Pyrrh. 21.5),⁶²⁰ whereas on the following day, Pyrrhus seizes the heights before the engagement to force the battle into more open space where his elephants could be decisive.

Via Wikipedia, a topographic map of Iran today with the Alborz mountains (at the southern edge of the Caspian Sea) marked.

Perhaps an even more dramatic demonstration of the ability of large Hellenistic armies to manage difficult terrain when well-generaled is Antiochus III’s expedition against Arsaces II (210-208). Antiochus first surprises Arsaces by marching through the northern Iranian deserts (Polyb. 10.28), causing the later to retreat into Hyrcania, which required Antiochus to cross the difficult and defended mountain passes through the Alborz mountains. This is truly rough country, as Polybius notes (Polyb. 10.29.2-3), not just rolling hills but mountains with narrow passes. And here the composite nature of the Hellenistic army comes into play, because Antiochus is clearly well-furnished with light and medium troops to screen his advance. Arscaes’ forces blocked the main pass with wooden barricades, but Antiochus’ picked detachments ascended the heights around the pass to clear those defenses by flanking them, clearing the way for the main body (Polyb. 10.30) which remained strong enough to prevent the Parthians from resorting to a pitched battle to disperse the screening forces (as we’ll see in a moment).

Via Wikipedia, the Alborz Mountains – this is the terrain that Antiochus III’s Hellenistic army is fighting on.

But perhaps the most direct example of a Hellenistic army performing effectively on rough ground is the Battle of Sellasia, the final and decisive battle of the Cleomenean War (228-222), which I am of course also excited to talk about because it is one of the most thorough and complete drubbings the Spartans ever take.

The brief background is this: Cleomenes III (r. 235-222), one of the hereditary kings of Sparta, seems to have realized – as Agis IV (r. 245-241) had before him – that the Spartan system was in dire need of reform. At the same time, the Achaean League (a federal league of Greek states, a common sort of thing in this period), led by Aratus of Sicyon, was aggressively pushing to consolidate the Peloponnese. Cleomenes thus embarks both on a set of military campaigns against Aratus and the Achaeans as well as sidelining the ephors to institute an ambitious set of reforms, raising the number of Spartiates – which had been just a few hundred – to 4,000; these he armed in the Macedonian manner, at least according to Plutarch, with the sarisa (Plut. Cleom. 11.2), presumably rather than the lighter thureophoroi troops common in many poleis.⁶²¹ He also seems to have been bankrolled by the Ptolemies (looking to stir up trouble for their Antigonid rivals, perhaps) and so was able to supplement his army with mercenaries.

Via Wikipedia, a map of the key Greek powers in the Peloponnese for the Cleomenean War, with the Archaean League in red.

Once Cleomenes starts winning, however, the potential of having one state – Sparta – consolidate the Peloponnese and possibly reach beyond was an unacceptable security threat for the Antigonus III Doson, who rules Macedon (as part of the Antigonid dynasty, the smallest and weakest of the ‘big three’ Hellenistic states), leading to Macedonian intervention in the war to try to bail out the Achaeans and contain Cleomenes III. Cleomenes tries to block Antigonus at the base of the Isthmus of Corinth, but this fails when Antigonus is able to sail some troops around that land-based blocking force and support a revolt in Argos, behind the lines. A series of raids and smaller sieges follows which effectively confine Cleomenes III to Laconia, leading to the final, decisive engagement at Sellasia in Laconia (that is, Sparta’s home territory on the road to Sparta itself) itself in 222.

We have the battle described by Plutarch twice (Cleom. 28 and Philop. 6) and even more usefully in detail by Polybius (2.65-70); of these, Polybius is generally to be preferred, writing about 50 years after the battle (and about 250 before Plutarch) and being quite a bit more familiar with military affairs as well. Cleomenes, as the defender, chose the ground and chose it well: the battlefield consisted of two hills, bifurcated by a river (the Oenous) which he then fortified with field-works. Antigonus initially avoids attacking directly, instead trying to maneuver Cleomenes out or bait him into a foolish attack, but logistics will have been decisive in the event. Cleomenes, on his home territory with Sparta behind him, could stay on his hills forever, but Antigonus could not remain camped in hostile territory for very long, so he opts to force the issue by storming Cleomenes’ fortified hilltop position.

Via Wikipedia, a diagram of the dispositions of the armies as described by Polybius.

Cleomenes occupies the right-hand-side hill, named Olympus (not to be confused with Mt. Olympus) with his Spartans and mercenary troops, while on the other hill, Euas, he has the Perioikoi and Greek allied troops under the command of Eucleidas. Finally, stretched over the riverbanks in the center he deploys cavalry and some of his remaining (probably lighter) mercenary troops. The battlefield here is thus nothing but broken up ground: hills and riverbanks. Antigonus responds by forming two main striking forces: on his right (against Euas) the chalkaspides (the more senior part of the phalanx) supported by Illyrian troops (medium infantry) in ‘alternating lines’ – suggestive of an articulated phalanx, supported in turn by Arcarnaeans, Cretans and the Achaeans. Opposite Olympus, he deployed the rest of his phalanx (probably the less senior leukaspides) along with his mercenary troops, presumably including his 1,000 Galatians; this is clearly the ‘junior’ wing (on the left-hand side, as one would expect) but Antigonus opts to lead it himself, seemingly because he was gunning for Cleomenes in particular (Polyb. 2.66.8). In the middle he strings his cavalry, supported by more allied troops.

The result was a battle effectively in three ‘lanes’ which interact with each other. On Euas, the Illyrians and Achaeans open the attack by pushing up the hill, but the Achaeans get into immediate trouble as Cleomenes’ mercenary troops in the center turn their flank; Philopoemon, a young man and junior officer rescues the issue by driving forward with a portion of the cavalry (for which Philopoemon is commended by Antigonus after the battle, while he criticizes his own cavalry commander for failing to do so). That danger averted, the chalkaspides then push up the hill and quite roughly shove Cleomenes’ forces off of the hill, inflicting heavy losses. The fighting in the center is heavy, but indecisive, while on Olympus, a see-saw fight between the Spartans and Antigonus’ leukaspides develops. According to Polybius, the terrain was more confined and so the Macedonian formation here was more compact, with no intervals (2.66.9), perhaps preventing the creation of a similar articulated phalanx as on the other hill. The two lines meet, Cleomenes initially is able to make ground and push forward, but then Antigonus doubles up his phalanx and decisively crushes Cleomenes’ line. The result is a crushing victory in which the Antigonids won on both hills separately. In both cases the Antigonid phalanxes were able to advance effectively over rough, hilly terrain to force a successful engagement with enemy infantry in prepared positions on high ground.

To sum up then, it is certainly the case that, as Polybius notes, the Macedonian phalanx could be disadvantaged by rough ground, but it does not follow that this was a formation that become utterly useless in hilly terrain. On the one hand, the Hellenistic army contained all sorts of other kinds of troops which could better handle extremely rough terrain like mountains (where, I will note, Roman legions would also struggle to operate), and on the other hand for modestly difficult terrain, the phalanx could still perform when well-handled. It was not enough merely to fight a battle near a hill or a river to gain an advantage against the phalanx, as Pyrrhus’ victories demonstrate.

The Archers of Parthia

Well, if the confines of mountains won’t fatally cripple the phalanx, what about archers – or better yet horse archers – and wide open spaces? Here the assumption is generally that the Parthian military system, in particular, represented a fatal weakness to the Hellenistic military system. And at 20,000 feet, you can see the case: the Parthia ends up breaking away from the Seleucid Empire and then absorbing most of its eastern provinces, so clearly they could win against Seleucid armies. And, of course, in Total War games, horse archers are ‘hard counters’ for the slow-moving sarisa-phalanx units.

As always, history is more complicated.

Via Livius.org, a small model of a Parthian mounted archer dating to the first or second century CE. I’m going to note that these guys sometimes lose, but it must be noted that the Parthian Army was a real threat that could defeat poorly led Hellenistic or Roman armies, even if the Parthians themselves generally come off the losers in their major wars with Rome.

Alexander’s campaigns, for one, had brought him to the very edge of the Central Steppe (a part of the broader Eurasian Steppe) in Bactria and Sogdiana and he fought both armies of Scythian steppe nomads and armies with Scythian auxiliaries in them successfully, albeit with substantial difficulty (Arr. Anab. 4.4-4.22). The largest engagement was a battle on what Arrian refers to as the River Tanais (Arr. Anab. 4.3.6-4.5.1) against a force of Scythians. In the battle that results, we can see Alexander leaning quite heavily on the supporting arms of his army: he clears the riverbank to make a crossing safe using his siege artillery (read: torsion catapults) and while his phalanx does cross the river, effectively all of the fighting is done by his light infantry and his cavalry. A useful reminder of why those later Hellenistic armies (especially the Seleucid ones) have all of those supporting arms.

The Seleucid record against the Parthians is more mixed, but ‘mixed’ doesn’t mean ‘nothing but defeat.’ Instead, the pattern we see – admittedly with a very limited sample set – is that when the Seleucid Empire is united and able to focus East, the Parthians aren’t generally able to resist Seleucid power but that – primarily because of events happening in the West of the Seleucid kingdom, the Seleucids are, after c. 250, rarely so united and able to focus. Now that is a real weakness of the Seleucid system – it is a problem with Hellenistic monarchy that it is so easy to destabilize it – but that is not the same as Hellenistic armies being categorically incapable of coping with horse archers.

Parthia forms out of such a moment of division: the Seleucid satrap Andragoras rebels in 245 in the context of a larger civil war involving Antiochus Heirax (running from 246 to 236) creates the opportunity for the nomadic Parni to push into the satrapy of Parthia and ‘set up shop.’ This is the context first for Arsaces I’s unsuccessful effort first to set up shop in the satrapy (read: province) of Margiana) in 245 (being repelled by Diodotus I, satrap and soon to be king of Bactria, with a Hellenistic army), before successfully ‘picking off’ Andragoras in Parthia in 239 while the main Seleucid army was distracted, off dealing with (or more correctly, failing to deal with) a civil threat from Antiochus Heirax in Anatolia.⁶²² Seleucus II (r. 246-225) is only going to be able to get around to trying to do anything about the Parthians in 235.

Seleucus II’s effort to subdue Parthia, when it comes in 235, is clearly a failure. Evidently there was a battle, but our sources for it are almost non-existent. Quite clearly the Parthians – potentially aligned with Diodotus I’s Bactrians – won, and the Seleucid effort to reassert power initially fails. Seleucus’ Antiochus Heirax problems weren’t over either, which may have divided or weakened the Seleucid effort. It is frustrating we don’t have better details about the failure of this effort to get a sense of how it differed from Antiochus III’s much more successful campaign.

The next effort will come with Antiochus III (r. 223-187), who though he comes to power in 223, waits until 210 to deal with the Parthians, being occupied in the intermediate space by a host of other problems, including losing the Battle of Raphia in 217. That said, by 210, Antiochus has managed to more-or-less stabilize his kingdom and consolidate power and so now the Parthians, under Arsaces II (r. 217-191), face effectively the full power of the Seleucid war machine.⁶²³ If simply having horse archers was enough to solve that problem, Arsaces II should have been alright, because the Parthians were excellence horse archers, the Parni themselves having come from the Steppe. But it is not enough and Arsaces II gets, quite frankly, utterly wrecked.

Our sources for this are not great (mostly Polybius 10.27-31), but we through him we can chart the basic campaign. Antiochus surprises Arsaces by making an operational desert crossing, getting his army into the core of Parthian territory intact. Arsaces could at that point, have offered battle in the relatively open ground. That he doesn’t suggests that the Seleucid field army – clearly quite large, even if we don’t trust Justin’s numbers – was sufficiently strong that Arsaces didn’t think he could win even on his home territory. Instead, he withdraws into the mountains, hoping – it seems – as much to exhaust Antiochus’ supplies (and perhaps money, as he has to be paying all of these soldiers and has been looting temples to do it) as to use the rough ground to his advantage. As we noted above, Antiochus’ light infantry components were perfectly capable of forcing the relevant passes over the Alborz mountains into Hyrcania, allowing Antiochus III to pursue Arsaces II successfully.

At this point, the Parthians try a pitched battle (Polyb. 10.31.1-6) at the foot of Mount Labus, evidently a strong defensive position. The description of the battle is minimal, but apparently the Parthians tried to force the phalanx (Polyb 10.31.3), failed to succeed at that and were then flanked and routed by Antiochus’ lighter troops who occupied the heights around the battlefield. Overtoom (op. cit., 115) tries to read this as an episode of feigned flight – that the Parthians skirmished with the phalanx and then withdrew – but this reading does not hold. For one, we have just the one source on this battle and Polybius says πτοηθέντες ὥρμησαν πρὸς φυγή, “terrified, they made to flee” (Polyb 10.31.3) which is just very clear. One can argue that perhaps Polybius has misunderstood, but Arsaces’ army does flee and much of his army (but not, perhaps, Arsaces?) ends up holed up in Sirynx, a fortified town close by, and neither Arsaces nor his army risks a pitched battle. That is not a thing you do if your harassment tactics are working, but it is a thing you do if you just got badly mauled in a pitched battle.

In any case, having now pinned Arsaces’ army down, Antiochus sets down to besiege Sirynx. The Parthians evidently attempt to flee encirclement but fail to break out when it is clear Antiochus will breach the town (Polyb. 10.31.11), their breakout fails and then Antiochus takes the town. The fighting was fierce but while Overtoom (op. cit., 127-8) insists the siege must have been long, Polybius explicitly notes it wasn’t – the moats which protected the town were filled ταχέως (‘swiftly’) and then the wall was undermined and breached. Polybius then cuts off, frustratingly but from Justin we know the outcome of the campaign, which is that Arsaces was forced into vassalage under Antiochus (who would then go on to also subordinate Bactria in the same way by 206). It seems fairly likely that, after a lost pitched battle and a lost contested siege, Arsaces was effectively out of army and thus forced to negotiate from a position of relative weakness. Vassalizing, rather than destroying, the rulers of his wayward provinces seems to have been Antiochus III’s default policy, so I’m not sure we can even assume that Arsaces had a particularly strong bargaining position after the fall of Sirynx, but he may have.

Subsequent affairs were typified more by Seleucid weakness as a result of internal weakness or external western threats than Parthian strength. Antiochus III gets into a disastrous war with Rome (192-188) with crushing results in both lost soldiers and a heavy financial indemnity. Antiochus III then dies in 187 and the internal collapse of Seleucid power begins. Antiochus III’s son, Seleucus IV is assassinated in 175 and Antiochus IV, a usurper, seizes the throne. Antiochus IV’s efforts to expand into Egypt meet with initial success, but are checked by Rome, which threatens to join the war if Antiochus IV continues, so he retreats; his death of natural causes in 164 kicks off a series of Seleucid dynastic struggles from that point until the disestablishment of what was left of the kingdom by the Romans in 63. In the chaos, the Parthians were well-positioned to gobble up whatever Seleucid vassals or satraps were nearby and did so, with relatively little direct Seleucid intervention.

Demetrius II (r. 145-138, 129-126) makes an effort in 139 to bring the Parthians to heel, but does so in the context of an on-going civil war with Diodotus Tryphon. Demetrius seems to have won several initial engagements (so says Justin (36.1.4)) but then fumbles the campaign and is captured – a good example of that weakness we’ve noted where Hellenistic armies often have to have the king himself present. Antiochus VII Sidetes (r. 138-129) then tries himself 130. He reoccupies Mesopotamia, defeats several Parthian armies (once again, from Justin, 38.10.5-6) and pinned the Parthians back to their core territory in Parthia, but then like Demetrius, it seems, overextended, gets cut off by a revolt of the Mesopotamians and is defeated and killed in 129. The Seleucids fall back into civil war, their last real effort to bring Parthia to heel having failed.

There are a few things to note here. First, Seleucid armies can beat Parthian armies and do so, repeatedly. Antiochus III’s campaign demonstrates that well-commanded and backed by the united might of the Seleucid Empire, a Seleucid army could defeat Parthian armies, force its way into Parthia proper and compel a favorable peace. If we believe our sources – and I think we should – evidently Demetrius II and Antiochus VII were also able to win victories in pitched battles against Parthian forces, albeit not decisive ones.

On the other hand, Parthian armies can clearly beat Seleucid armies too. I often find myself reacting against the assumption of Parthian super-armies based on a handful of their most notable victories – the Parthians, more often than not, lose to Roman armies (the Romans sack Ctesiphon three times: 116, 164, 197) – but we shouldn’t go overboard in the other direction either. Parthian armies were properly dangerous on their home territory and could and did destroy poorly commanded armies which were unprepared. It is unfortunate but perhaps to be expected that the surprising Parthian victories are the ones that get described in detail in our sources, so we have a good sense of what a Parthian army winning looks like, but not a good sense of what a Parthian army losing does, despite the latter happening about as much as the former.

As for the Seleucids, it seems to me that the Parthians alone were not enough to cause the collapse of Seleucid rule and that the spark which lights the fuse of Seleucid self-destruction is not the arrival of the Parthians in 245, but rather the arrival of the Romans in 192, as the Romans will continually meddle in Seleucid politics, blocking expansion, supporting Seleucid rivals and generally making a mess of their succession disputes. Of course the Romans might have found this more difficult had the Parthians not been causing their own problems, but it seems worth noting that while the Romans proved capable of defeating a maximum-strength-deployment Seleucid army (under Antiochus III at Magnesia, no less) the Parthians mostly thrive picking off weak Seleucid leaders bedeviled by civil wars and disunion.

Can the Parthians beat a Seleucid phalanx? Yes. Do they always do so, in the manner of a Total War ‘hard counter?’ No. Of course, the same might well be said of Parthian performance against Roman legions as well: they can beat them, but most certainly do not always do so. And speaking of Roman legions: next week we will, at long last, turn to Italy and take a closer look at why Pyrrhus of Epirus’ effort to face Rome both met with initial success and ultimately failed.