How many legions and equivalents at Zama?

Nov 2011
1,051
The Bluff
Nick, you have three ongoing threads in three different forums going, here, RAT and ACG.

Did you not get the answers you were looking for in any of the three?
Not the answer he wanted we must assume and no answer on RAT. I've been a member of RAT for some twelve years and although most of their discussions now take place on Facebook, I can't see any more favourable responses coming from there I'm afraid. Should he get one.
 
Oct 2018
1,498
Sydney
DIocletion'sBetterThanYou has Walbank's HCP. From recollection, he states the figures are exaggerated.
Indeed, Walbank does. Walbank 2.427: 'Clearly the figures given now for the two camps are too large: they are larger than any army Carthage ever assembled, and Hasdrubal can hardly have doubled his force since Scipio landed. It is likely that at some stage the figures have been inflated to the glory of Scipio, and De Sanctis (iii. 2. 584; cf. Scullard, Scip. 320) plausibly suggests 30,000 infantry plus 3,000-5,000 cavalry for the combined force.'
 
Nov 2011
1,051
The Bluff
If I take Appian’s 23,000 infantry and divide this by four legions, I get 5,750 infantry. If I presume Livy’s figure of 6,200 infantry includes 500 cavalry, this would allocate each legion 5,700 men. Livy’s figure of 2,000 cavalry when divided by four legions, would allocate each legion 500 cavalry. Now let’s remember that Appian states the Roman infantry was “about” 23,000 men, which means not exactly 23,000 men...

Numbers in the ancient sources have not been properly studied...
Numbers in the ancient sources have had many studies focused on them - generally for specific campaigns, though Brunt's Italian Manpower comes to mind as a more extensive treatise. There is a fetish to reconcile numbers across sources but such is generally a pursuit of accord in vain. The numbers various sources give do not have to marry up in every instance, though it would be nice if they did. For instance, they do not with Alexander's invasion army. Livy is very clear that there existed many figures for this invasion army leaving Sicily from 16,000 foot to so many that the birds fell out of the air. Some Roman annalistic sources were prone to such exaggeration and it wouldn't surprise that Polybios' figures (Utica, Hasdrubul above) come from Fabius Pictor.
 
Last edited:
  • Like
Reactions: Edratman
Nov 2011
1,051
The Bluff
While on the subject of numbers, it should be noted - as DiocletianIsBetterThanYou has pointed out - that Polybios' figures for the Roman casualties at Cannae are exaggerated. But that is not all. Polybios provides us with figures of eight overstrength legions leading to 80,000 men in the field. This was not necessarily the case and Polybios may well be incorrect either from confusing the material available to him or from the source he chose to utilise. Livy (22.36.1-4) again tells us there are conflicting numbers provided by the authorities available to him:

The armies also were augmented. But how large were the additions of infantry and cavalry I should hardly venture to declare with any certainty —so greatly do historians differ in regard to the numbers and kinds of troops. [2] Some say that ten thousand new soldiers were enlisted as replacements; others that four new legions were enrolled, so that they took the field with eight. [3] Some assert that the legions were also increased in the numbers of their infantry and cavalry, and that each received an additional thousand foot and a hundred horse, bringing up the total of every one to five thousand [p. 321]foot and three hundred horse; [4] and that double the1 number of horse and an equal number of foot were furnished by the allies.
On balance, four overstrength legions with 10,000 supplumenta plus their alae (approximately 50,000) makes more sense in terms of how the battle played out.
 
Oct 2015
894
Virginia
Brunt, Hallward, DeSanctis et al (and me, for what it's worth) agree with you and Livy's unnamed sources. Walbank, Toynbee and most modern writers (who are either not classicists or to whom military history is uninteresting) prefer Polybius' 89,000. Though how even Hannibal could annihilate an army twice the size of his own is hard to understand; let alone the command, logistical and literary problems involved with the bigger number.
 
Last edited:
Nov 2011
1,051
The Bluff
Brunt, Hallward, DeSanctis et al (and me, for what it's worth) agree with you and Livy's unnamed sources.
Nice to know (and good to have you as well!). I assume Brunt's assessment comes from his Italian Manpower? From whence come Hallward's and DeSanctis's rationales (save me searching)?

What's clear enough to me is that there were various reported figures for the army of Cannae - just as there were for the Army of Scipio invading Africa. Unfortunately, 216 is one year where Livy omits his full listing of appointments (consuls, et al) and legion levies and their deployments/dispositions. We are left to assemble the Rubik's Cube from the notices of legion actions and theatres. While it is possible that the Senate did indeed order four extra overstrength legions so as to put eight such into the field at Cannae, the real question is, how probable is it? We have two conflicting notices: a reinforcement of 10,000 infantry for the four legions presently in the field against Hannibal or those legions brought up to overstrength and added to by four similarly overstrength newly raised legions. In the aftermath of Cannae, Rome is reduced to raising legions of boys "over seventeen and some who still wore the purple-bordered dress of boyhood" as well as "eight thousand young and stalwart slaves" (Livy, 22.57.9-11).

Now, not all decisions of a state can be enacted and some may be changed or cancelled. Sparta, for example, had a significant fleet in 404 having taken many of Athens' triremes. Despite ordering 120 east less than ten years later to confront the Persian fleet, only 85 are found in battle (due to lack of seaworthiness and equipment). We may well have something similar here. The Senate may well have ordered four more legions. In the event this proved impracticable, the following decision may have been to send those available - the 10,000 odd (a little over two legions) - bringing the four in the field to the overstrength size noted. Separate decisions, one essentially amending or cancelling the first, has perhaps led to the confusion over numbers in the sources.

More to the thread's point, there were clearly more than the two "legions of Cannae" serving in Sicily. After Scipio departs with those two legions, we still have two legions in the province (Livy, 30.27.8) which have to be supplemented after his African predations (see above). Scipio has not only taken the two legions of Cannae but some of the other forces in Sicily as well as his 7,000 volunteers. This, to me, makes Livy's statement of him bringing the legions up to 6,200 foot (and allies) credibly though several historians find this too large a number for the time. To be fair, they point out that such numbers are not seen until the legions of Pydnaawhere they are brought up to 6,000 (so, 24,000 - legions plus alae). But what if this is the first such attestation of such? Given the 3,000 reinforcements sent otSicily after Scipio's departure, I find that to be the more likely.
 
Last edited:
Oct 2018
1,498
Sydney
The topic of numbers at Cannae is not one that I have thought too deeply about, certainly not enough to have a strong opinion, but to show the other side of the debate, I'll quote Lazenby 1978, Hannibal's War, 75-76:

'But the most important question concerns the number of legions Varro and Aemilius Paullus eventually had at Cannae. Polybius states quite unequivocally (3.107.9) that it was decided "to maintain the struggle with eight legions, a thing which had never happened among the Romans before, each of the legions having up to 5000 men, apart from the allies," and this, if true, means that it was decided to raise four new legions, each of 5000 men, and to bring the existing four under Servilius and Atilius up to the same strength. He also believed that an equal number of allied infantry was levied, and consistently with this says that the Romans had 80,000 infantry at Cannae (3.113.5). These figures have been doubted, because Livy (22.36.1-4), although he reports this tradition, also records a variant one - that only 10,000 fresh troops were raised - and it is argued that the smaller is to be preferred a priori, and that Hannibal's tactics at Cannae are unintelligible if he was really outnumbered by nearly 2:1 in infantry. But the a priori argument is not necessarily to be accepted - it is very difficult, for example, to accept the lowest of the three estimates Livy gives (29.25.1-4) for the army Scipio took to Africa in 204, and some scholars would accept the highest. But a more serious objection to accepting the tradition that only 10,000 fresh troops were raised for the Cannae campaign, is that this would have only given the Romans a slight numerical advantage, even if the four existing legions under Servilius and Atilius were strengthened to 5000 men, and that as many allied soldiers were levied as Roman. For even if we accept that all the 10,000 were citizen soldiers, that would still only have meant 30,000 citizen troops in all, with 30,000 allied infantry in addition. But Hannibal had 40,000 infantry (3.114.5), and some of the Roman infantry were bound to be left to guard the camp on any day of battle - Polybius indeed believed that 10,000 were left to guard the larger of the two camps, and Appian (Hannibalic War 4.26) mentions 5000 in the smaller, for what it is worth. Thus Varro and Paullus, we are being asked to believe, were left to face Hannibal's 40,000 infantry with possibly as few as 45,000 of their own, while their 6000 cavalry (3.113.5) were actually outnumbered by Hannibal's 10,000 (3.114.5). If this was the case, the Romans had certainly learnt nothing from the Trebbia, for there they had tried to fight Hannibal with a slight numerical superiority in infantry, and with fewer cavalry, and they had lost. Moreover, it might be argued that Hannibal's tactics, far from being unintelligible if he were really outnumbered by something like 2:1 in infantry, would be unintelligible if his infantry on the field was nearly as numerous as the enemy's.

Finally, it is perhaps significant that Livy's circumstantial account of the Roman losses at Cannae, and of those taken prisoner or who escaped, presupposes the higher number: in particular, if twenty-nine military tribunes were really killed (22.49.16), there must have been at least five legions present, since there were only six such officers to a legion. In other words, Livy's figures for the Romans killed, and for those who survived whether as prisoners or fugitives, support Polybius' figures for the size of the Roman army, and it cannot be too strongly stressed that Polybius is quite emphatic about his figures, though he obviously realised that they were unusually high.'
 
Nov 2011
1,051
The Bluff
The topic of numbers at Cannae is not one that I have thought too deeply about, certainly not enough to have a strong opinion, but to show the other side of the debate, I'll quote Lazenby 1978, Hannibal's War, 75-76 '
Polybios is, indeed, insistent on his eight overstrength legions and, as I say, both scenarios are possible. Polybios has either chosen to follow that tradition or, as I say, he's confused the lot. Is it possible the annalistic tradition Livy reports is post Polybios? Reducing Roman numbers reduces Roman embarrassment I suppose. Eight unprecedented legions decimated certainly plays well with boys and slaves being enlisted. But there obtrudes the thought that Polybios is happy with some 90,000 Carthaginian troops and allies in Africa.

That aside, I find it interesting that Livy's notice of Scipio bringing the two legions he took from Sicily up to 6,200 foot each is rejected by several historians. The basis is that we "don't hear of such until the Pydna campaign" or similar. Apparently, we're not hearing of it here although Livy plainly attests this and I would argue strongly this goes back to Polybios. Livy is plain that Scipio wished to raise a levy in Rome on taking his consulship for Sicily/Africa and was refused; having to settle for the "armies in Sicily" and permission to take only volunteers. Scipio takes his volunteers and selects from the "armies in Sicily" the legions of Cannae which he proceeds to strengthen with both his 7,000 volunteers and others from provincial forces which results in Rome sending 3,000 supplementa to bring the remaining two legions back to paper strength. I find it only too credible that Scipio, stymied by the Senate, got his way in any case: something in the order of 24,000 foot for Africa. It wouldn't be the only time Africanus worked a swifty on the Senate.
 
Last edited:
Oct 2018
1,498
Sydney
That aside, I find it interesting that Livy's notice of Scipio bringing the two legions he took from Sicily up to 6,200 foot each is rejected by several historians. GThe basis is that we "don't hear of such until the Pydna campaign" or similar. Apparently, we're not hearing of it here although Livy plainly attests this and I would argue strongly this goes back to Polybios. Livy is plain that Scipio wished to raise a levy in Rome on taking his consulship for Sicily/Africa and was refused; having to settle for the "armies in Sicily" and permission to take only volunteers. Scipio takes his volunteers and selects from the "armies in Sicily" the legions of Cannae which he proceeds to strengthen with both his 7,000 volunteers and others from provincial forces which results in Rome sending 3,000 supplementa to bring the remaining two legions back to paper strength. I find it only too credible that Scipio, stymied by the Senate, got his way in any case: something in the order of 24,000 foot for Africa. It wouldn't be the only time Africanus worked a swifty on the Senate.
Yes, I agree. I see no clear reason to reject the 6200 foot.
 

Duke Valentino

Ad Honorem
Jul 2017
2,319
Australia
The strength of the Roman army at Cannae is an interesting topic. On Livy's 10,000, as much as Salaminia will chafe at me mentioning, I would like to contribute Delbruck's thoughts on it, who also rejects the smaller figure as incorrect:

But the 86,000 figure for the Romans stems, as Appian proves, drawing on Roman sources, from their own side, and as we shall establish further below in a closer examination of the composition of the Carthaginian army, no objective reason exists to cause us to doubt the strength of that army.

Warfare in Antiquity, 326.

Delbruck proposes the composition of the Roman army thus:

Hoplites in the battle: 55,000
Hoplites in the camp: 2,600
Rorarii in the battle: 8,000
Rorani serving as orderlies behind the front: 7,000
Rorarii in the camp: 7,400
Cavalry: 6,000

Total: 86,000 men.

Do these figures seem roughly correct to everyone else?