Number of legions at Zama

[Spartan JKM]

Hello all. I hope it's OK to bring up an old thread, but I was googling for information about Zama, and came across this old thread here on the greatest site for Roman history and culture (IMHO, but I hope many will agree ) .

For the most part, Livy and Appian take a back seat to Polybius, when all three are available, but Appian may be the one to follow for the numbers, and he alone gives Roman figures for the onset of the battle: to reiterate what was stated earlier in this thread, and with my slightly differed view, I think Scipio's army at the onset of Zama, following Appian's deduction, was 23,000 foot and 1,500 horse, accompanied soon by 6,000 foot (almost all Roman and allied) and 4,000 horse from Massinissa, and also mentions 600 Numidian horsemen under a chieftain called Dacamas - for a total of 29,000 infantry and 6,100 cavalry. Appian states in his Roman History (Libyca).41,

"...Proconsul Publius Scipio had about 23,000 foot and 1,500 Italian and Roman horse. He had as allies Massinissa with a large number of Numidian horse, and another prince, named Dacamas, with 600 horse. He drew up his infantry, like those of Hannibal, in three lines..."

What Appian probably 'missed' was that part of Massinissa's force was the 6,000 infantry, mostly Romans and Italians; Appian never mentions the arrival of the 10,000 under him while Scipio and Hannibal were talking (unless some fragments of his work have not been recovered; but substantial 'gaps' are usually identified by later historians).

Basically, Scipio probably had two core legions, of unprecedented strength, with allied contingents of even a little more numerical strength (conjecture), and 4,600 Numidian horsemen. It seems Scipio lost perhaps 700 or more Roman/allied cavalry between his arrival in Africa to the onset of Zama, if we follow this analysis. But Appian's account of the battle itself is generally unconvincing, particularly being that it includes a personal duel between Hannibal and Scipio, and another between Hannibal and Massinissa.

We don't know, without doubt, the exact relative strengths of the two armies, but many feel that Hannibal did not outnumber Scipio as much as Appian claimed. Remember, this was a frontal engagement devoid of any flank or oblique attacks, and though Scipio's army was more experienced (except Hannibal's third line) and of higher quality, the disparity of quality between the 2 armies wasn't quite like that of Alexander's brilliantly constructed instrument against the Persian heavy cavalry, who couldn't even closley match the Companions and Thessalians in close-order fighting, or the motley levies of Asiatic infantry against the disciplined Alexandrian infantrymen. Nor was the difference at Zama akin to Caesar's X or XIII Legion against the unweildy Gauls, or even Belisarius' famed bucellarii against city mobs and Vandals. Thus there is no way Scipio would have been hanging on in the final stages of the battle, as was the case, if his infantry had been outnumbered significantly, such as 50,000 vs. 23,000 as Appian seemingly claims. Remember - the cavalry squadrons of all the mounted participants left the field in the initial stages of the battle, and Gaius Laelius and Massinissa didn't return until a 'providential' moment. It is also improbable that 4,000 Macedonians under one Sopater were part of Hannibal's 2nd line, as Livy would have us believe. Polybius, a Hellene himself, would have mentioned that (I will not go along with anyone's proclamation that he 'missed' that detail).

Again, we do not know the exact quantity of Scipio's invasion force to Africa from Lilybaeum (SW Sicily), and we can only extrapolate from Livy's accounts, along with Polybius' and Appian's later figures. Indeed, the Cannae and Herdonea survivors, the fit ones, now exiled in Sicily, would form the core of Scipio's army. But there were troops he brought from Italy, as was mentioned earlier in the thread, and as Livy tells us in Book 29.24,

"...He (Scipio) then wrote to M. Pomponius requesting him, if he thought it advisable, to come to Lilybaeum that they might consult together as to what legions should be selected and what ought to be the total strength of the invading force. Orders were also sent all round the coast for every transport vessel to be requisitioned and brought to Lilybaeum. When the whole of the military and naval forces in Sicily were assembled there, the town could not afford accommodation for all the men, nor could the harbour hold all the ships, and such enthusiasm prevailed in all ranks that it seemed as though instead of marching to war they were to reap the fruits of a victory already won. This was particularly the case with the survivors of Cannae, who felt quite certain that under no other leader would they be able to do such service for the commonwealth as would put an end to their ignominious condition. Scipio was far from despising these men, he was quite aware that the defeat at Cannae was not brought about by any cowardice on their part, and he knew, too, that there were no soldiers in the Roman army who had had such a long experience in every kind of fighting, and in the conduct of sieges. They formed the 5th and 6th legions. After announcing to them that he would take them with him to Africa, he inspected them man by man, and those whom he did not consider suitable he left behind, replacing them with the men whom he had brought from Italy. In this way he brought up the strength of each legion to 6,200 men and 300 cavalry. He selected the Latin contingent also, both horse and foot, out of the army of Cannae."

These troops brought from Italy were certainly the 7,000 volunteers, presumably including a 600-strong 'cohort' provided by Camerinum, located in not-so-close Umbria (Livy, Book 28.45-46). Livy reports three totals of the number of Scipio's invasion force, Book 29.25,

"As to the number of troops put on board there is considerable divergence among the authorities. I find that some state it to have amounted to 10,000 infantry and 2,200 cavalry; others give 16,000 infantry and 1,600 cavalry; others again double this estimate and put the total of infantry and cavalry at 35,000 men. Some writers give no definite number, and in a matter so uncertain I prefer to include myself amongst them..."

Well, the first figure of 12,200 total is belied by the facts of the strength of the 5th and 6th legions. Theodore Dodge surmises Scipio's invasion force as about 35,000 men, as does Serge Lancel; to quote the French archaelogist/historian, Pg. 164 of his Hannibal,

"...On the overall strength, between the various figures provided by his sources, Livy hesitates to opt for the largest, 35,000 men, infantry and cavalry combined, but it is the most probable for an expedition of this size..."

This is the more likely figure, but it may have been a little less. What we know is that the two core legions of Scipio's army, the 5th and 6th, which were increased to 6,200 foot and 300 horse, made up 13,000 in number, thus that would make the allied contingents 22,000, to bring us to the figure of 35,000 total as the invasion force. If things were conforming to tradition and without any alterations (alterations would not be out of the norm), then the invading force would number 27,200 - two strenghtened legions composed of 13,000 Romans and 14,200 socii; the allies were composed of 1,800 horse to the Romans' 600. Indeed, the Roman military system was always adaptable and flexible, but this 9,000 differential is probably a little too high a proportion (I'm just floating with this). Remember, an allied contingent of a legion was traditionally, at this time, equal in infantry but three times the cavalry, thus if Scipio's 5th and 6th contained 600 horsemen, their allied contingents, on paper, should have numbered 1,800, giving us 2,400 total horse for his invading army in 204 B.C. So, maybe the figure offered by Livy of 16,000 infantry and 1,600 cavalry was indeed the total for the allied contingents. I think 30,000+ is a plausible number for Scipio's invasion force (which was actually higher than what Hannibal had upon his arrival in Italy, but less cavalry) in 204 B.C. - 29,000 total infantry and 2,000+ cavalry. Massinissa joined him in Africa with, according to the majority, 2,000 cavalry. Livy tells us his own sources put it at a mere 200, and this is where I think B.H. Liddell Hart is correct, as he tells us on Pg. 125-126, Scipio Africanus: Greater Than Napoleon,

"...Livy accepts the smaller estimate, for the very sound reason that Masinissa after his return from Spain had been driven out of his father's kingdom by the joint efforts of Syphax and the Carthaginians, and for the past year and more had been eluding pursuit by repeated changes of quarter. An exile, who had escaped from the last battle with only sixty horsemen (Livy Book 29.33) , it is unlikely that he could have raised his band of followers to any large proportions..."

Maybe Massinissa joined Scipio with a number between 2,000 and 200 at this early juncture, but we'll never know for certain. Remember, when Masinissa joined Scipio with 10,000 (6,000 foot and 4,000 horse) men just before Zama, 10 'cohorts' - about 5-6,000 men - were already lent him from Scipio to win back his kingdom. So, Scipio had around roughly 31,000+ (if more, it wasn't much) men at the beginning of the African campaign, which would become at times divided. Allowing for the extreme minimal losses suffered by the Romans and their allies in the initial clashes with two forces, the first just 500 and the second about 4,000 (both commanded by a different commander named Hanno), then at the 'Burning of the Camps', in which the enemy was destroyed while they slept, and at the Great Plains (probably Scipio's most brilliant win), and adding Masinissa's native forces of roughly 5,000 men (he may have had 1,000 or so native infantrymen blended in with his Roman/Italic allies), Scipio had about 34-35,000 men at Zama. Again, we can add Massinissa' 10,000 to Scipio's army of 24,500 (following Appian), but we cannot add 10,000 when accounting for Scipio's numbers since the initial landing, as roughly half of Massinissa's force included soldiers who had been part of the initial landing. Thus, when applying all the math, and accounting for Scipio's losses (there were no reinforcements outside of Africa of troops), we can add only 5,000 from Massinissa, and possibly 600 from one Dacames.

There were plenty of allies in Sicily by 205 B.C., as it was now completely a Roman province, and a perfect stepping-stone to Africa. Scipio was basically given a blank check by the Senate regarding Sicily, and he was supported by the allies in Italy, as Livy tells us, Book 28.45,

"...Scipio did not succeed in obtaining permission to levy troops and indeed he did not press the point, but he was allowed to enlist volunteers. As he had stated that his fleet would not be a charge on the State he was given liberty to accept any materials contributed by the allies for the construction of his ships. The cantons of Etruria were the first to promise assistance, each according to its means. Caere contributed corn and provisions of all kinds for the crews; Populonia, iron; Tarquinii, cloth for the sails; Volaterrae, timber for the hulls and corn; Arretium, 3,000 shields and as many helmets, whilst they were ready to supply as many as 50,000 darts, javelins and long spears. They also offered to furnish all the axes, spades, sickles, gabions and hand-mills required for forty warships as well as 120,000 pecks of wheat and provision for the sailing-masters and the rowers on the voyage. Perusia, Clusium and Russellae sent pine-wood for the timbers of the ships and a large quantity of corn. The Umbrian communities as well as the inhabitants of Nursia, Reate and Amiternum and the whole of the Sabine country promised to furnish men. Numerous contingents from the Marsi, the Paeligni and the Marrucini volunteered to serve on board the fleet. Camerinum, a city leagued on a basis of equal rights with Rome, sent a cohort of 600 men-at-arms. The keels of 30 ships-20 quinqueremes and 10 quadriremes-were laid down, and Scipio pressed on the work so rapidly that 45 days after the timber had been brought from the forests, the ships were launched with their tackle and armament complete..."

Again, Scipio already brought some troops from Italy for his expedition; it seems he was just prohibited from levying heavily. If a 'cohort' was indeed 600, as Livy says (unless an isolated one not conforning to tradition), maybe the deduction of the 10 cohorts Masinissa was given later by Scipio should total 6,000, making Massinissa's native contribution a little smaller, in turn making Scipio's force at Zama about 33-34,000.

I feel Hannibal's plan was sound and workable at Zama; he didn't need a victory of destruction as he wasn't the invader with a smaller force, and he didn't have the cavalry quantity or quality (at least on his right wing). But had his first two lines not fought amongst each other (how often does that come up in the history of recorded battles?), Scipio would have been hard-pressed to use his whole force before Hannibal used his crack third line - perfectly disposed as a 'true' reserve (ie, it was placed far enough back to prevent being outflanked, and close enough to deliver a coup de grace if the oppurtinity arose). But he must have calculated the Roman/Numidian cavalry was coming back soon. But the Roman victory could be assured only if they attacked the rear of Hannibal while he was still engaged. If Scipio's infantry had been defeated, how well could 5,500-6,000 (accounting for the possible few losses they suffered in their pursuit, as there was no substantial clash between the four mounted units) horsemen do against a recently readied body of Hannibal's approx. now probably less than 20,000 (accounting for small losses in the final clash if they had emerged victorious) infantry (certainly, some of the remnants from the first two lines had blended around and into his last line)? What if Scipio's men had turned on each other in the first two lines? Romans, particularly these disciplined ones, would have never done that, but I am simply posing that his chances for victory would have been jeapordized. Clearly, Scipio utilized his advantage very wisely.

Marcus Regulus' defeat at Tunis in 255 B.C. at the hands of the mercenary Xanthippus came mainly because the Carthaginians had much superior cavalry that day. But Regulus, against supposedly nearly 100 elephants in his front, altered the customary checkerwise formation of the maniples by placing them one behind the other, thus creating intervals. We don't know for sure about this, nor do we know if Regulus deployed his ranks with more than the traditional three lines, but he seems to have added to the ranks of each individual maniple. What we do know is that Regulus shortened his front and made his ranks deeper. In this battle, some of the elephants were probably driven through with some success ('lanes'?) to the rear without doing severe harm. Polybius doesn't mention intervals, but says, Book 1.34,

"...while the formation of the main body, owing to the depths of the ranks behind, remained for a time unbroken..."

This probably means many of the elephants were funneled through (100 elephants were certainly not all stopped in their tracks!). But unfortunately for Regulus, the Roman cavalry couldn't hold very long against their better and more numerous adversaries, and many of the Carthaginian cavalry found their way into the lanes, which opened new flanks. The Romans were cut to pieces as a result, as other Carthaginian horsemen swung around the Roman flanks to the rear.

B.H. Liddell Hart, in his terrific (albeit very tendentious) Scipio Africanus: Greater Than Napoleon, describes the elephant charge at Zama, Pgs. 178-179,

"...The remainder of the elephants wrought much havoc among Scipio's velites, caught by their charge in the front of the Roman line. But the foresight that had provided the "lanes" and laid down the method of withdrawal was justified by its results. For the elephants took the line of least resistence, penetrating into the lanes rather than face the firm-knit ranks of the heavy infantry maniples..."

Well, there are no copyrights in warfare, but Scipio probably profited from Regulus' failure, who had inferior cavalry than his enemy at Tunis. It was the opposite for Scipio at Zama. Regulus seems to have shown some original foresight to absorb elephant charges, though he lost the battle. Or perhaps, on the other hand, the tactic of handling elephants was wrought inadvertantly. This doesn't discredit Scipio's foresight, but Hart doesn't mention this peculiar information - that of an elephant charge being absorbed by Roman legionaries with some success in the First Punic War. But maybe that's just trivial (I love detail), and Scipio enhanced this maneuver with the use of trumpets and bugles to frighten the elephants, which worked quite well.

Hart says of Polybius on Pg. 4 in the Introduction of Greater Than Napoleon,

"...The verdict of Polybius is clear, and his facts still more so..."

Maybe I'm pinpricking too much, but Polybius tells us in Book 15.14, of Hannibal's army at the end,

"...while of those who took flight only quite a few escaped, as the cavalry were close on them and the country was level. More than 1,500 Romans fell, the Carthaginian loss amounting to 20,000 killed and nearly the same number of prisoners..."

We have to decide how many are 'quite a few escaped', and 'nearly the same number as 20,000 taken prisoner'. It seems Polybius believed Hannibal had in all a little less than 40,000 men, and if this didn't include Hannibal's horsemen, who were long gone, that would add another 3-4,000 at most.

But Hart states on Pg. 177 of Hannibal's army,

"...His total force was probably in excess of 50,000, perhaps 55,000..."

A little later on Pg. 185 Hart assesses,

"...Polybius and Livy agree in putting the loss of the Carthaginians and their allies at 20,000 slain and almost as many captured...".

That means slightly less than 40,000 slain and captured. But, again, he doesn't mention the "..only quite a few escaped.." part from Polybius, nor does he mention anything about approx. 5-10,000 escapees (or approx. 10-15,000), which would be the case if Hannibal's army had a total of 50-55,000 men at the onset of the battle :?:

Moreover, Hart thinks Scipio embarked for Africa with 17,600 men, of which 1,600 were cavalry. That's clear from Hart (Pg. 120). He tells us the first figure is indeed disproven by the 'previous facts' (ie, the numerical strength of the 5th and 6th legions), but doesn't tells us why he feels the third figure is incorrect; he merely proposes, quite tersely, that the 'second figure as the correct estimate'. But later, when accounting for his troop strength at Zama (Pgs. 177-178) he writes,

"...The Roman strength is less certain (he is saying 'less certain' than Hannibal's 50-55,000), but if we assume that each of Scipio's two legions were duplicated by an equal body of Italian allies, and add Masinissa's 10,000, the complete strength would be about 36,000 if the legions were at full strength. It was probably less, because some wastage must have occured during the earlier operations since quitting his base..."

Clearly, Hart thinks the 'wastage' was minimal. But except for the arriving 10,000 under Massinissa while Scipio and Hannibal were discussing terms, Hart doesn't explain how Scipio's army enlarged from 17,600 to the 'probably less' than 36,000 (maybe 6-7,000 unaccounted for), nor did he seem to allocate the 10 cohorts, part of Massinissa's 10,000, as part of the original invading force. Scipio was not reinforced by any ally or Rome itself, other than Masinissa, during his African campaign. But forgetting all that deep conjecture, I think Scipio had about 35,000 men at Zama, of which 6,000+ were cavalry, of which about 75% were Numidians under Massinissa and Dacames (or a little less, if we discard the existence of Dacamas, which comes solely from Appian). Hannibal opposed him with 40,000+ troops, of which about 3-4,000 were horsemen; Hannibal's left flank was composed of excellent Numidians (the most 'warlike in Africa', according to Polybius) under one Tychaeus, and Appian tells us that one Mesolutus, presumably the same chieftain Livy names as the usurper Mazaetullus, also joined Hannibal. According to Polybius, Tychaeus provided 2,000 horsemen, and I think it is Diodorus who gives a figure of 1,000 under Mazaetullus (Serge Lancel, Pg. 173). Appian also states one Vermina, the son of Syphax as being part of Hannibal's allies at Zama, but Livy doesn't think Vermina was on the verge of joining the Carthaginian cause until after Zama (Book 30.36), in which Scipio routed him in a clash, technically the final clash of the Second Punic War; I feel this defeat of Vermina can be cautiously accepted, but certainly not a result of 15,000 of Vermina's men slain, and 2,700 captured (1,500 Numidian horsemen).

I hope this wasn't too cerebral!!

Thanks, James K MacKinnon

[Matthew]

Very interesting and a good deal more informative than the preceding discussion (no offence intended to those contributors).

I have a question or two and an observation, though:

Where does the figure of 10,000 Foot and 2,200 Horse come from? (i.e. do we know Livy's source?)

Do you think that an Allied or Roman Cohort is likely to have been a stable number of men in the Roman Republic? I rather suspect the number varied quite a bit from unit to unit, much like the strength of a Legion itself.

The difference between 1,600 and 2,200 Horse is 600, which is, I suspect, the basis of your idea that the second figure Livy gives relates to allies only?

Matthew James Stanham

[Gaius Julius Caesar]

"I hope this wasn't too cerebral!!

Thanks, James K MacKinnon [/quote]"


Not at all! while I like to read, I am afraid I don't quite have the patience to copy and type such a response! All respect to you! 8)

Most of what I think and have to say will, thank god, remain in my head! :wink:

[Aryaman2]

I will opt for the minimum figure, not only for Scipio´s army, but also for the Carthaginian army, 35.000 just ignores the difficulties of transporting troops by sea at the time, and that applies also to Hannibal, how was he supposed to ship 20.000 from Italy?
Sources report a wide range of numbers, for instance at Great Plains the total force of Roman infantry is opposed by just 4.000 Celtiberian mercenaries. IMO some discrepancies would arise from the fact that sources usually count units at full strength, while in campaign that is never the case. My guess is that Scipio landed with 12.000, he received some replacements along the campaign that allowed him to keep Roman strength and added the support of Numidian cavalry. I would guess a number of around 15.000 for each army at Zama.
Another thing I would like to note is how sources report casualties, anyone familiar with detailed casualty reports starting in the 18th century knows that WIA/MIA are always much greater than those KIA. In fact, you can get only accurate data mixing reports from both sides. The winner side reports are very accurate with all data regarding his own side (and here the low casualties always reported in Ancient battles for the winner) while they are only estimations, normally highly exagerated, about the enemy casualties, including those KIA (nobody cared much about counting enemy bodies) you have to look at the enemy report for casualties, but they usually only report those present after battle, so that all kind of casualties are put together in a single group, that is probably what the best informed reports of our sources on enemy casualties usually mean when they talk about those enemies KIA, although I would say that is rarely the case.

[Spartan JKM]

Thanks for joining me in the rekindling of this thread you guys!.

Thanks for your note, Byron, as with your help on my question regarding the armor.

Matthew:

...Where does the figure of 10,000 Foot and 2,200 Horse come from? (i.e. do we know Livy's source?)

Do you think that an Allied or Roman Cohort is likely to have been a stable number of men in the Roman Republic? I rather suspect the number varied quite a bit from unit to unit, much like the strength of a Legion itself.

The difference between 1,600 and 2,200 Horse is 600, which is, I suspect, the basis of your idea that the second figure Livy gives relates to allies only?

Hello Matthew. We do not know from whom that first figure came from, but we know it did not come from one of Livy's primary sources - Lucius Coelius Antipitar, because Livy tells us that Antipitar said of Scipio's invading force upon embarking in Sicily, Book 29.25,

...Some authors have not given a figure, and I would prefer to be in their company since the question is moot. While he avoids a precise figure, Coelius still exaggerates and gives the impression of huge numbers. He says that birds fell to the ground when the soldiers' shout went up, and that the numbers boarding the ships were so great that it appeared that no mortal was being left behind in either Italy or Sicily!..."

It's too bad we don't have Antipitar's work (there are fragments), because although he is described as an historian of 'rhetoric', his account of the Second Punic War apparently made use of both Roman and Carthaginian sources. His vivid description cited by Livy of Scipio's established force is obviously one of hyperbole, but probably illustrates that it was indeed a significantly large army, particularly for one ferried across water for two days (they disembarked on the third day). After all, this was perhaps hitherto the most difficult and largest task ever undertaken by a Roman force, led by perhaps the greatest commander ever produced in her history (I would claim 'definitely' if not for the existence of Julius Caesar a century and a half later). True, legions of 6,200 infantry and 300 cavalry were never this large, and not matched again until the Third Macedonian War of 171-167 B.C. But this expedition of Scipio's was perhaps paramount in the military history of Rome, both Republic and Empire (maybe I'm now being a little hyperbolic), against its most dangerous enemy ever; no Roman enemy ever marched or landed an army of their own onto Italian soil and threatened Rome of her very existence, until the fall of Rome in the early 5th century, and by then the political and economic infastructure to support the western army had collapsed, not as a result of any significant military defeat.

Thus, in my opinion, for one to discard the strength of of the 5th and 6th legions as being 13,000 infantry and cavalry is to deny the cardinal ability of the Roman military system to adapt and implement a degree of flexibility, something applied when events necessitated such measures. In terms of augmenting numbers, it was the socii who generally appended the extra numbers, so I feel it is probable the two allied contingents to the 5th and 6th legions were of higher quantity in both horse and foot. Scipio's invading force of roughly 30,000 makes sense, particularly that Scipio could count of Numidian help upon his arrival at the Promonturium Appolinis, which was less than forty miles NW of Carthage itself, and merely 125 miles across from Lilybaeum.

The amount of cavalry Scipio could bring to Africa was indeed limited, quite a bit less than 1/10 of his infantry, a ratio twice as much when he faced Hannibal two years later, as transporting horse by sea was extremely difficult - the reason Hannibal could not bring his horses from Italy in 203 B.C., given his dire situation of limited transports (he had to build some of his own) and the need to bring as many men as possible to Africa in the face of Carthage's greatest peril, following the debacle at the Great Plains.

No Matthew - I do not think the Roman or allied 'cohort' was a staple number in the Republican period. But whatever the contingents were called at the time (I guess it doesn't matter), they were not as varying, in terms of troop quantity, as in later periods. The size of the legion, Roman and/or allied in the Republican period, was increased by simply increasing the number of soldiers in the maniples of the hastati and principes, and increasing the number of velites. But its basic organizational structure didn't change. The term 'cohort' used by Livy is anachronistic, as the term cohort, applying to the tactical sub-unit of the legion we know so well of, was not extant until later, arguably beginning with Gaius Marius. But Polybius uses the term 'cohort' a couple of times; perhaps the translators were also describing the reformed units of the Republican army in hindsight, primarily beginning with Scipio Africanus 'Major', as 'cohorts', though at the time the term wasn't actually applied. But maybe this assessment is wrong, and the term 'cohort' was used as early as Scipio, as he added flexibility to the army with the innovation of a tactical unit of three maniples (about 360-480 men, put into action at Baecula in 208 B.C.), and it was through an evolutionary process throughout the second century B.C., and into the first, that the 'cohort' became the basic tactical unit of the Roman army; the legion of ten cohorts was more flexible than the manipular legion, as its structure allowed it to function more effectively against disjoined tribal brigands etc., as this required smaller scaled fighting. But it could fight in pitched battle easily as well too. The velites ceased to exist, with practically every legionary becoming a heavy infantryman. It reached a point where a citizen militia was unsuitable for long periods of service, which entialed garrisoning in the provinces. But they sure came through in the Second Punic War! They suffered some very painful lessons at the hands of one of history's greatest military leaders, but they began to quickly adapt and adopt - a signature of Roman military greatness. Polybius' account of Scipio's force upon embarkation does not survive, but if Livy had it, he doesn't say so; actually, Livy faintly praises Polybius by name only once in his work of the Second Punic War, amid discussing Scipio's triumph and Syphax's death, at the very end of Book 30,

"...Polybius, who is by no means a source to be disregarded..."

We can guess that one or more of these sources, all used by Livy, was the source for that first number of 12,200 as the invasion force (remember, it wasn't from Lucius Coelius Antipitar):

Gaius Acilius

Marcus Porcius Cato

Lucius Cincius Alimentus

Quintus Quadrigarius Claudius

Quintus Fabius Pictor

Lucius Calpernius Piso 'Frugi'(meaning 'honest')

Aulus Postumius Albinus

Silenos (accompanied Hannibal, and Polybius' source)

Sosylos (accompanied Hannibal, and Polybius' source)

Timagenes of Alexandria (used by the famous ancient geographer Strabo and, possibly, by Livy)

Valerius Antias

Interesting observation on your part of the 600 differential of the 1,600 and 2,200 cavalry, and thinking that may have been a basis for my theory of the second figure from Livy as pertaining to the allies; no, that's a coincidence. The main reson behind my thinking is that Tiberius Sempronius Longus, who would lose to Hannibal on the Trebbia, was commissioned in Sicily at the onset of the war to invade Africa with two legions aboard 160 quinqueremes and 12 cutters: Livy tells us that he had two legions, with the allied contingent bolstered to 16,000 infantry and 1,800 cavalry, but Publius Cornelius Scipio (the great one's father), was given a lesser amount of 14,000 allied infantry and 1,600 allied cavalry to compliment his two legions. Publius was given just 60 quinqueremes, as the Senate believed the enemy could not employ a mode of warfare involving the sea up there. He was to sail for Spain. Now, maybe that meant nothing, but it was an observation that caught my attention. Clearly, unlike with in 204 B.C., Livy tells us with conviction that the Romans would strike at Africa with 26,400 men aboard 172 ships (an average of about 153 men per ship, but 12 were the smaller cutters). The Roman strategy at the onset was practical - a two-pronged attack against Carthaginian power. Hannibal would certainly be serious trouble at first in Spain, they figured, but they would easily (relatively) triumph in Africa; they had already been there and learned what not to do etc. Of course, as we know, Hannibal swiftly unleashed one of military history's most daring and audacious attacks in history, thwarting all their initial plans.

I should have mentioned this, and meant to, but another issue, with all this conjecture, has just hit me: I figured the proportion of 22,000 allied to 13,000 Roman numbers was too much to arrive at the surmised figure of 35,000 for Scipio's invasion force. But I overlooked that in 218 B.C., with these commissions of Scipio the Elder and Longus, the legion was fixed at just 4,000 infantry and 300 cavalry, giving two Roman legions a total of 8,600. Therefore, the proportion of allied contingents for both consuls in 218 B.C. to their two Roman legions was a differential of 7,000 with Scipio the Elder, and 9,200 with Longus. Mmmmmmmm :roll:

Aryaman2:

... 35.000 just ignores the difficulties of transporting troops by sea at the time, and that applies also to Hannibal, how was he supposed to ship 20.000 from Italy?

Hi Inaki. I respectfully disagree. The proposed number from me for the landing was about 30,000, but maybe it was 35,000, and maybe Scipio's losses from his arrival in Africa to Zama were more than I figured; 35,000 was what I think Scipio had at the onset of Zama, and still do. 'Difficulty', particularly when faced by the likes of Scipio and Hannibal, was hardly impossible. In times of war, according to Polybius, a quinquereme could carry more than 125 marines (normally 40), excluding the assigned crew. Two groups of 20 warships (certainly quinqueremes) flanked 400 transports (this has never been doubted to my knowledge) for Scipio's crossing to Africa, which was about 125 miles and just two days away (they disembarked on the third day). These transport ships could carry more men than a warship; approx. 80 legionnaires (a little less on average), maybe five horses, and the varied supplies per transport seems reasonable per ship. Perhaps some were integrated with each quinquereme, as one of them could comfortably carry forty men (not including the rowing crew etc.) To cite two famous examples of difficult sea-crossings of antiquity, the Persians crossed the Aegean to invade Attica (the battle of Marathon) in 490 B.C., aboard about 25,000 men aboard at least 200 ships (sources are extremely sweeping with those numbers, with some saying 60,000 men on 600 ships, which is seemingly incredulous). Julius Caesar's first expedition to Britain was composed of, at most, 10,000 men (a legion with Caesar was usually around 4,000, maybe less, and he took two with him) and about 300 cavalry aboard 98 ships (80 for the troops, 18 for the horses), which would be 125 men per transport and 17 horses for each ship. Moreover, Gnaeus Scipio sailed to Emporiae (modern Ampurias), in NE Spain, from Massilia (modern Marseille) with perhaps 20,000 troops ('most of' his brother Publius' double-consular army of 25,000 men, according to Livy, Book 21.32); he certainly sailed along the shoreline, and we are given no amount of ships. But being Publius had been given 60 quinqueremes for his task upon arriving at the mouth of the Rhone, we can assume that Gnaeus sailed to Catalonia with more than 40 of them.

If Scipio used seperate transports for the horses as Caesar did, and using the mathematical gauge of 29,000 infantry and 2,200 cavalry aboard 400 transports, this means 107 men were on 271 transports, and 17 horses were carried over on 129 transports. My deduction is affected by Caesar's using 18 transports for his approx. 300 cavalry. 'Specialized' transports, presumably larger, did not come until Caesar's second and bigger incursion to Britain in 54 B.C. This may not have been the case at all, as we have no detailed information on the compositional nature of Scipio's boarding measures, but it is not at all implausible that he either put 5 horses amongst 72 soldiers in 400 transports, or had 107 men per transport on 271, and the horses aboard 129 set aside just for them, with 17 each aboard each ship. Remember, they didn't need to travel far in waters not so hostile anymore; the Carthaginian navy was now even beyond a broken reed. I may not be as stubborn with my opinion of approx. 29,000 infantry and 2,200 cavalry, if not for the substantial data of 400 transports. That's a huge armada for tis time.

In my opinion, to opine that the figure most believed by the consensus is off by as much as 60%, is to overlook Livy's detailed information about the distribution and resposibilites amongst varied Italian communities for the construction of extra ships needed for Scipio's invasion in 204 B.C. He specifically tells us that the socii navales (probably based out of Naples and/or Peasentum), not the state, beared the expense of providing the extra ships needed for Scipio's expedition. All this for an invading force of 12,000? 12,000 is slightly less a figure than what Gaius Claudius Nero arrived with in 211 B.C.(13,100), sailing from Puteoli, to relieve the beleaguered 9,000 men tenuously holding the Ebro line under Lucius Marcius and Tiberius Fonteius, and 1,000 less than what Scipio sailed to Emporiae with a year later from the mouth of the Tiber ('skirting' the coasts, as Livy says, Book 26.19, aboard 30 quinqueremes).

Moreover, if one believes just 12,000 arrived in Africa under Scipio in 204 B.C., the figure of 400 transports to cross must be discarded, as well as the belief that this was a huge undertaking, rather than a riad, which Gaius Laelius administered earlier. It would also need to be believed that he received reinforcements from Sicily and/or Italy, of which we have no record. He was indeed supplied with grain and clothing from Sardinia, and Livy tells us that 3,000 troops were raised for service in Sicily, as all the crack troops from Sicily had gone to Africa with Scipio (Book 30.2), but not once did a substantial reinforcement reach Scipio, except the invaluable cavalry of Massinissa and Dacamas (always a question mark with the latter), an element on which Scipio's plans in Africa, particularly with the anticipation of facing Hannibal himself, were centered around. But never do we read of any troop reinforcements except from the Numidians. Sure, perhaps a 'few' filled his ranks when receiving supplies etc., but the numbers would have been negligible, without any mention of them.

As for Hannibal, the quantity of his embarking force from Croton is not known with any given figures (surviving ones). But he had a much longer voyage than Scipio had a year earlier, and certainly with a numerically lesser armada. His force he brought to Africa would constitiute his thrid line at Zama, and judging by his dispositions leading up to and in the battle, he came back with about 15,000 men at most, and his 'Old Guard', was by now a handful. Polybius tells us within his account of Zama, the foremost ranks of Hannibal's third line lowered their spears to not allow remmants of the defeated troops of the first two lines to enter their ranks in the front, meaning they were indeed his African pikemen. Peter Connolly suggests it was only them that came back to Africa, and they numbered just 4,000. But Polybius tells us that the final clash between the infantry of Hannibal and Scipio were about the same in numbers, both armies over 30,000, so if Connolly is right, Hannibal's third line must have been substantially augmented by the fugitives of the first two lines, which isn't entirely implausible: if 12,000 men, approx. half of the first two lines combined, joined Hannibal's third line around the flanks, and his third line was merely 4,000 initially, that would give us the final duel of some 16,000 men each - something believable to some. But not by me - that would entail far too many Roman/allied losses, from a theory of 29,000 infantry to begin with.

The fact that the Roman navy was all but absent in Hannibal's route between Croton and Leptis Minor (80 miles SE of Carthage) suggests that the Senate had no intention of obstructing his path or discouraging him from leaving Italian soil. They were thrilled to see him go!

True, the amount of soldiers enumerated for an army did not all fight in a battle, as some were left to guard the camps, and our original sources identify this issue, albeit not every battle: at Cannae, we are told by Polybius and Appian, in isolated comments between them, that 10,000 soldiers were left in the larger camp, and 5,000 in the smaller, but Livy mentions only that 7,000 escaped to the smaller camp, and 10,000 fled to the larger camp (Book 22.39), perhaps confirming Polybius' view of the 10,000, but raising the question whether the smaller camp was substantially guarded at all from the onset. But all this conjecture could mean absolutely nothing.

...Sources report a wide range of numbers, for instance at Great Plains the total force of Roman infantry is opposed by just 4.000 Celtiberian mercenaries. IMO some discrepancies would arise from the fact that sources usually count units at full strength, while in campaign that is never the case...

Actually, our main sources concur that the Carthaginian/Numidian force was about 30,000 at the Great Plains, so we have to apply a little deductive logic. True, there was a much scantier ability among analysts to procure information back then; revisionists offer what they feel probably happened, and there is at least an account of the Great Plains battle, rather than a succinct mention, to provide for conjecture without pulling too much out of the air. Though we'll never know for sure, it reaches a point where we have to accept an inconclusive explanation (IMHO, of course). It seems, too, that not everybody will be satisfied by another's 'truth'. As Marcus Aurelius proclaimed,

"Everything we hear is an opinion, not a fact. Everything we see is a perspective, not the truth."

But, again, our three primary sources, Polybius, Livy, and Appian (Appian in much more nebulous fashion, as he doesn't mention 'Great Plains'), all tell us that the combined forces of Syphax and Hasdrubal Gisgo, concentrated with the 4,000 Celtiberians, amounted to some 30,000 men. This may be a magnification, as we are told that this force, in the wake of the 'Burning of the Camps' disaster, was mustered up within thirty days. Also, the Carthaginian forces gathered (a troop levy was undertaken in the city of Carthage itself) must have been very low, as Scipio was able to leave a force small enough to merely 'keep up the facade of the blockade' of Utica, as Livy states in Book 30.8, and he himself 'proceeded against the enemy with the main body of his troops' inland towards perhaps what is now the plain of Souk el Kremis (about 80 miles SW of Carthage, 65 miles SW of Utica), near the towns of Bou Salem and Jendouba. But Scipio could also draw soldiers from his crew of 40 ships off Utica. Also, this new area of operations was now closer to Syphax's domain, but Scipio's celerity was primarily to stop the recruiting process of his enemy. Professor H. H. Scullard, following the reconstructive theories of J. Kromayer and Georg Veith, believes the forces of Hasdrubal Gisgo and Syphax amounted to some 20,000, with Scipio's swiftly arriving force into the region numbering 12-15,000 (Scipio Africanus: Soldier and Politician, Pg. 129). But others, such as J. F. Lazenby, opine Polybius can be more readily taken at face value, with the Carthaginian force at 30,000 against Scipio's 20,000 (Hannibal's War, Pg. 209). Whatever the truth, Scipio was almost certainly outnumbered and taking a great risk, as a defeat so far inland would be as disastrous as what befell Regulus' 52 years earlier, as the Roman survivors and their allies would have found themselves isolated far from their camp. Polybius and Livy (almost certainly following Polybius) tell us of the normal battle formation of a Roman army, but, yes, don't seem to know where the enemy infantry units stood.

If indeed Scipio's infantry faced only the Celtiberians, this would mean his cavalry alone was confronting 26,000 men or less; if we allocate for units not being at full strength while on campaign, which is certainly tenable, the lesser numbers could not be significantly less, even if we accept more conservatively surmised modern figures. But again, Scipio wouldn't have dared divide his forces if there were significant enemy forces concentrated near Utica and Carthage; clearly, most of the available Carthaginian forces were sent to join the combined forces of Hasdrubal Gisgo and Syphax, evidenced by the distress caused within Carthage by the subsequent defeat of this newly formed force. Remember, probably 2/3 of Hannibal's army at Zama came from abroad. It is also quite conceivable that Polybius, as he seemingly had done before and after, leaves out any deep reference to the allied contingents in Scipio's force, so perhaps the parts of his account that suggest that all of Scipio's infantry were faced by only the Celtiberians should be considered as only the Romans opposed them. Regardless of any omissions with regards to garrisoned or reserve troops, which would have been, presumably, minimal, given the dire straits Carthage was now increasingly facing, it is absolutely incredulous that Gaius Laelius' and Massinissa's cavalry, perhaps 4,000 between them (Massinissa had not yet won his kingdom 'back'), routed 26,000 (or a little less) enemy troops by themselves. Nor could the 4,000 Celtiberians (or whatever the true number) have faced up to the entire Roman infantry, which would have overlapped them considerably. This is all based on the hypothesis that the original sources are close with their enumerations, mainly with the ratio of Scipio's numbers against the enemy, in which he was probably outnumbered by 30% or so.

Being that Polybius provides the detail that Scipio enveloped the Celtiberians with his rear ranks, in Book 14.8, it is quite possible that Scipio actually brought out at least some of his rear ranks of principes and, if the need arose, triarii, to support his cavalry wings from behind. Or perhaps there were indeed infantry units from Syphax and Hasdrubal Gisgo that were placed on the flanks of the Celtiberians, with the Numidian and Carthaginian cavalry covering the flanks of their respective contingents, meaning they hoped to overwhelm Scipio's cavalry flanks with superior numbers of both their cavalry and infantry units (we are not told by our original sources if they significantly outnumbered Scipio), then surround Scipio's legionairies while they were occupied by the battle-hardened Celtiberians. If so, Scipio's sublime generalship beat them to it, as his horsemen routed the vastly inferior cavalry of the enemy (presumably) on both sides quickly, exposing the flanks of the rest of the army of Hasdrubal Gisgo and Syphax; what is likely is that Scipio now had his hastati hold the Celtiberians, and had some of his principes and triarii, under cover of the hastati, move out to the left and right and fall upon their flanks, which had been stripped by the quick flight of Syphax's and Hasdrubal Gisgo's cavalry. But this would mean that the hastati alone matched the frontage of the Celtiberians, and the rear ranks of Scipio's infantry were needed to be brought up on either flank to envelop the Celtiberians, which means that this action was probably carried out by a single legion of roughly the same numerical strength as the Celtiberians. But almost all of the rest of the army was involved in the battle, so the answer is possibly that Syphax's Numidians, horse and foot, who were on the left (Roman right), were about as numerous as the Celtiberians and Carthaginians combined, horse and foot, to their right (Roman left).

Thus, when Scipio arranged his forces in the routine manner, his right to left, he formed one full legion on the right (opposite the enemy left), with its attached ala to its right, then his second full legion, with its ala to the left (the alae were on the wings of the Roman legions). Gaius Laelius and his cavalry was placed on the right wing of the army, stretched to match Syphax and his numerous Numidians. On the other side of Scipio's army, the left-hand Roman legion was disposed opposite the Celtiberians, and the left ala and Massinissa's Numidians were opposite the Carthaginian foot and horse. So, after the rout of all of Syphax's horse and foot on his left (Roman right) by Gaius Laelius and the entire right legion, and the vanquishing of the Carthaginian horse and foot on the right (Roman left) under Hasdrubal Gisgo by Massinissa and just the ala of the left-sided legion, the Celtiberians, certainly dispirited by the quick flight of their allies, were opposed by the single left-hand legion of approx. the same number of Romans. But Syphax and Hasdrubal Gisgo, and plenty of their men, got away.

The subsequent destruction of the Celtiberian force was effected by a model of tactical brilliance by Scipio, and is possibly a good answer to the question on another thread I was reading - which Roman victory almost matches Cannae as a great victory? Scipio still used the triple-line format, but broke from tradition by not using the principes and triarii to support the hastati in a way that only included strengthening it and filling its gaps if need be; what he probably did was have each 'cohort' (the sub-unit of three maniples innovated by him) of principes and triarii function as self-dependent units, ready to move up and extend the front line of the hastati, thus the chances of the hastati giving way, before an outflanking maneuver by the remainder of his principes and triarii could be effected, were diminished, as their frontage was increased. It is quite possible that Scipio, in witnessing Hannibal's tactics some 13 and 15 years earlier, saw the efficacy of the elasticity of Hannibal's army, which he re-organized (no details) after Trasimene. A Roman force, Scipio figured, needed to achieve similar things. He basically created offensive reserves from the rear ranks.

Forgive all the conjecture; what basically happened, in my opinion, at the Great Plains was that Scipio, certainly with a much superior army, less so over the Celtiberians, swiftly vanquished the flanks of the enemy, which was composed of all its cavalry and the native infantrymen of Syphax and Hasdrubal Gisgo with his two cavalry squadrons, one legion and all the alae; then, with his other Roman legion without its ala, he held the Celtiberian with his hastati, perhaps lengthened on each end by 'cohorts' (probably tactical units of three maniples) from his principes (and possibly triarii), then had his remaining (or entire) 'cohorts' of the principes and triarii, under cover of the front line engaged with the Celtiberian front, turn into column, probably half to the right, half to the left, and then march out to encircle the Celtiberians. The Celtiberians fought courageously, but could not break Scipio's lengthened )possibly) hastati before being attacked in flank and rear. The only puzzling issue is how Scipio so quickly vanquished the enemy horsemen, probably more numerous. Hasdrubal Gisgo, apparently a capable adminstrator, was a terrible commander in the field.

Scipio did not merely develop a detached reserve which would be brought in hurriedly for a necessary contingency, but formed an offensive reserve to prolong his lines if need be, and outflank the enemy. He was unable to this when matched against Hannibal, because at Zama Hannibal thwarted this tactic with his own deployment of a 'true' reserve - one large enough and placed perfectly; far back enough to prevent envelopment, but also close enough to deliver a coup de grace, should the opportunity present itself. But Scipio's foresight in procuring superb cavalry, in improving the Roman and acquiring allies who were inherently great horsemen, enabled him to come through. The problem with all this is that Polybius' accounts of Scipio's battles present problems of slight uncertainty. For example, though the victory at Ilipa was seemingly a tremendously impressive win, where were the Carthaginian cavalry units, some 4,000 of them? There is no explanation. Perhaps it is the result of fragmentation, not embroidered bias, which is still a possibility, as he did live and work under the patronage of the Scipionic Circle.

...In fact, you can get only accurate data mixing reports from both sides. The winner side reports are very accurate with all data regarding his own side (and here the low casualties always reported in Ancient battles for the winner) while they are only estimations, normally highly exagerated, about the enemy casualties...

That doesn't strain credibility, but some of the veritable, 'original' sources of the Second Punic War, those I mentioned earlier, and whose works didn't survive, made use of Carthaginian eyewitnesses too; Polybius' research included interviews with important participants, such as Gaius Laelius and Massinissa. Of course, we certainly do not have a balanced view of the Second Punic War, and the patriotic rhetoric is not slightly embroidered. But because it was such an event on a massive scale, far from everything can be colored to fool future generations. I doubt Polybius pulled his estimations out of the air; he spoke to ex-soldiers who might have given him an insight into enemy (Carthaginian) casualties (though possibly very crude information), and we cannot be sure what proportion from Polybius is pro-Roman, though almost certainly above the halfway point. I feel it is a mistake to suppose that our understanding of the Second Punic War is wholly of Roman origin, or to attempt to make a mechanical seperation, so to speak, of the Roman and pro-Carthaginian passages in the texts of Polybius; the truth can still be applicable even if something 'sounds' bias, and each event has to be scrutinized in itself.

Of course record-taking, correspondence, and tallies etc (all the same, thing really), were much cruder back then, but the armies had sub-units, particularly those of the Roman forces, thus a 120-160 strong maniple could be accurately accounted for, and then ten maniples could be discernibly accounted for after that. Also, a detail such as 400 transport ships can be accurately enumerated. Almost definitely, the exact truth we do not have of the numbers of battle losses, but revisionism has closed the gap of the errors from way off to more closely accurate, hopefully.

I believe in assessing the 'assessments'. For example, at a surfaced glance, it seems incoceivable that Hannibal could surround a Roman army more than twice his numerical strength (approx. 32,000 infantry vs. 65,000) on three sides. It seems he would have had to weaken his lines proportionately to the extent that it would be impossible to withstand the Roman sledgehammer. Thus, the theory that the Romans did not possess the 80,000 total infantry at Cannae, but more like 48,000. But Hannibal's maneuver of an extraordinary 'reverse-refusal' with his Celts and Spaniards would be unintelligible if his infantry on the field had been outnumbered by only a little; with his cavalry, he wouldn't have needed to what he actually did, which carried more risk. His incredible display of battlefield control, in which he enitced the aggressive Romans to attack him, was centered around the two strong blocks of African infantry on the wings, and a little back. This meant that not only would the Romans tend to be sucked into the middle, but if things went amiss, fugitives from the Celtic and Iberian units would also be funnelled into the center where they would be compelled to at least bunch and slow the Roman advance, even if they didn't want to. Thus, the Romans were sucked far enough in, becoming more crowded, so their own relentless forward pressure on the stretched infantry of Hannibal's became retarded, and soon totally contained,a due to the attck on their flanks, which was not effected by Hannibal having his Africans march too far out. There is certainly much luck involved, but it was a masterpiece that has resonated beyond its own age.

Thanks, James K MacKinnon

[Robert Vermaat]

Wow, that was a great article. Laudes!

[Matthew]

Indeed, truly a pleasure to read.

A question about the fixed size of the Legion in 218 BC, though. Is it Polybius you have in mind when citing the fixed size as 4,000 Foot and 300 Horse, as I have seen that passage read a couple of ways and Evelyn S. Shuckburgh's current translation of Polybius Book 6 puts the fixed number at 4,200 Foot and 300 Horse, which I think makes good sense when trying to divide up the troop types [i.e. 600 Triarii, 1,200 Principes, 1,200 Hastati and 1,200 Velites] into Maniples, though it raises certain questions about the reinforced Legion [i.e. what kind of troops would logically make up the 800 Man differential, just Hastati and Principes or Velites as well?]. If that is not the passage you have in mind, I would be grateful if you could point me in the right direction.

There are, of course, some very interesting things that arise out of thinking about the numbers presented by Polybius. In the first case (that of the Legion of 4,200 Foot), if one is inclined to take Polybius at his word and divide the Velites equally into the Maniples, you end up with:

10 Maniples of Hasatati, each containing 120 Hastati and 40 Velites
10 Maniples of Principes, each containing 120 Principes and 40 Velites
10 Maniples of Triarii, each containing 60 Triarii and 40 Velites

When divided into Centuries this means 80 Men per Century of Hastati and Principes and 50 per Century of Triarii.

Once reinforced (excluding the possibility that the Velites were reinforced) a Legion of 5,000 Foot works out as:

10 Maniples of Hasatati, each containing 160 Hastati and 40 Velites
10 Maniples of Principes, each containing 160 Principes and 40 Velites
10 Maniples of Triarii, each containing 60 Triarii and 40 Velites

When divided into Centuries this means 100 Men per Century of Hastati and Principes and 50 per Century of Triarii.

Very entertaining when thinking about the nomenclature of the Century and the 'Marian' reforms that make every Century a uniform 80 Men. Playing with the idea of Cohorts composed of one of each of the three Manipular divisions results in a Non Reinforced Cohort of 420 Men and a Reinforced Cohort of 500.
All supposition, of course, but I find it intriguing when thinking about how Polybius describes how the camp is divided (the Triarii being associated with the Equites) and the practical method by which troops might be embarked on transports (Centuries and Maniples being the natural divisions to use, which suggests several possiblities; A 500 strong Cohort, for instance, can theoretically be nicely broken up into four units of 125 for transport, assuming Officers are not counted or are included in the 500, depending on your view).

I often wonder how much this sort of numerical thinking influences the figures presented by Ancient Authors. Cannae is a good example, as the figure Polybius gives of 80,000 is clearly the result of him calculating 8 Reinforced Legions (5,000 x 8) and an equivalent number of Allies (5,000 x 8). Taking the reduced Cavalry number of 200 per Legion, we can also account for the 6,000 Horse in relation to this figure, as the Legionary Horse would be 1,600 (200 x 8) and the Allied Horse around 4,800 (three times as many), the total being 6,400, which is very close to the figures Polybius gives (6.113. - 'eighty thousand foot and rather more than six thousand horse').
Livy, though, seems more inclined to talk about 50,000 Foot, though without saying that is the total number in the army. Interestingly, if there were 50,000 Foot, the 6,000 Horse would also be explicable in terms of 1,500 Roman Horse [5 x 300] and 4,500 Allied Horse [5 x 900]. Of course, five Legions wouldn't make much sense, but it is possible. Could Livy have been retroactively reducing the number of Infantry based on the number of Cavalry? Was Polybius inflating the number of Infantry in comparison to Cavalry to prove his thesis: 'a lesson to posterity that in actual war it is better to have half the number of infantry, and the superiority in cavalry, than to engage your enemy with an equality in both' (3.117.), which he then goes on to demonstrate the significance of at Zama.

This sort of thing always worries me when thinking about numbers, but all the same, I think Scipio's forces must have been closer to 30,000 than 10,000. I do wonder how Legions of 6,200 Foot and 300 Horse were divided up into Maniples (5,000 + 1,200, maybe? A fourth line?).

Anyway, another very informative and interesting post. Laudes to you.

Matthew James Stanham

[Aryaman2]

JKM
That was an excellent exposition. I will try to defend my position with further arguments
1) From a logical military perspective, before embarking Scipio discharged unfit men from the Cannae legions and brought them up to full strength with new recruits, it makes little sense to bring all those green recruits to overload your transports and logistics, instead it is more logic to transport only your veterans, while the new recruits are left in garrison duty in Sicily, where they could be trained and get seasoned. Then they could be brought in small convoys little by little as replacements, without compromising logistics.
2) As for naval expeditions, there are other examples, Thucydides, a most reliable source, says that the Athenian expedition to Sicily was composed of 134 ships to transport just 5.100 infantrymen. The relation between ships and men is dependant on multiple factors, I will look into further examples later.
3) General numbers, many of your suppositions are based on what another sources say, for instance take Great Planes. You have one source indicating that 4.000 Celtiberians matched the whole Roman Infantry, and other source that the total Carthaginian force was 30.000. You say that 26.000 could not be routed by Roman/Numidian cavalry alone, but other alternative is to think that in fact the Carthaginian army was much smaller, the core being the 4.000 Celtiberians.
The same for Hannibal troops transported from Italy. If you believe Scipio trasnported 35.000, it is not that difficult to believe that Hanibal managed to transport 12.000 to 20.000, but if the lower numer is given credit, then it is difficult to believe that Hannibal brought that many men, 4.000 would be a more credible number, and then the whole battle of Zama could have been fought by much lower numbers.
4) As a general rule, I always turn to what we know for sure about numbers and logistics from other better known periods. I know you are familiar with the 30YW and the difficulties armies just about 20.000 found supplying themselves in Germany, how the 43.000 army Gustavus Adolphus assembled at Alte veste badly starved, imagine then how difficult was to supply armies of 35.000+ in North Africa.

[Aryaman2]

JKM
I would like to discuss your numbers of Caesar expedition to Britain. I think that is a very valuable source, as he is an eyewitness and he is talking about his own forces, you said
"Julius Caesar's first expedition to Britain was composed of, at most, 10,000 men (a legion with Caesar was usually around 4,000, maybe less, and he took two with him) and about 300 cavalry aboard 98 ships (80 for the troops, 18 for the horses)"
There you make the same mistake that Ancient authors, you multiply the number of legions by his paper strength, however units in campaign are almost never at full strength. Caesar himself says that at Pharsalos his cohortes averaged 275 men instead of the authorised strength of 480, and that some legions were at half strength. It is not unreasonable to think that the 2 legions could well have mustered about 2/3 of their strength, that is 3.200 each, that would give us 7.400 infantry+300 cavalry with 98 ships.
However in 54 BC Caesar organized a larger expedition, mustering 5 legions,and 2.000 cavalry, by the same counting about 16.000 infantry and 2.000 cavalry, but he required no less than 800 ships! So the 400 ships of Scipio don´t look too much off the mark, besides remember that he transported also rations for 60 days.

[L C Cinna]

One more thing.

Someone already mentioned Scipio's "losses" before the battle. Even if you get the numbers for troops transported to Africa right it still doesn't tell us anything about how many men Scipio had at Zama.

With "losses before the battle" I mean all troops which couldn't participate. Don't go and underestimate this.

The numbers for peole injured and ill due to fighting or other things like diseases and so on are always pretty high, especially in such an operation overseas in a slightly different climate. When you look at such army strenght reports through all of history you notice that a lot of people will be missing sooner or later. I just remember my own time in the army when we stayed outside for combat practice in winter after 1 week nearly a 5th of the regiment were whether ill, injured or not fully "usable" anymore and that without any enemy around.

The next thing is that you need a lot of troops to do other duties. Strategically important places need to be garrisoned, the supply lines secured, the camps secured to have a place to retreat to and so on.

I'm not an expert on such matters, I just speak of my own experiences, maybe someone knows how to calculate an average of troops which would not have been able to participate. If we remember that Scipio had several major and minor engagements before and needed to secure several strategically important places I'd say that at least 1/4th of his troops would not be present. So if he had over 30.000 paperstrenght before the campaign there would only be maybe 25.000 present at Zama as a maximum I'd say.

Just a guess.

[Aryaman2]

Another example I recall about ships and troops, from an eyewitness account, the Byzantine expedition against the Vandals according to Procopius mustered 10.000 foot and 5.000 horse, and required 500 transport ships plus 92 warships, so again the 400 ships of Scipio expedition don´t seem to be too many.

[gauis alexandrois]

i dunno jow many legions were in it but i read off a site that there was around 35'000 to 40'000 men. . the site is http://www.roman-empire.net/army/zama.html . it tells the tactics used by each side of the battle

[Spartan JKM]

Hello everyone.

My apologies, I just cannot post as much as I would like to.
Matthew:

...Is it Polybius you have in mind when citing the fixed size as 4,000 Foot and 300 Horse, as I have seen that passage read a couple of ways and Evelyn S. Shuckburgh's current translation of Polybius Book 6 puts the fixed number at 4,200 Foot and 300 Horse, which I think makes good sense when trying to divide up the troop types [i.e. 600 Triarii, 1,200 Principes, 1,200 Hastati and 1,200 Velites] into Maniples, though it raises certain questions about the reinforced Legion [i.e. what kind of troops would logically make up the 800 Man differential, just Hastati and Principes or Velites as well?]. If that is not the passage you have in mind, I would be grateful if you could point me in the right direction.

My pleasure Matthew; it is from Livy we get those figures for 218 B.C. Polybius mentions a legion, when discussing the events leading up to Cannae (Book 3.107), as being 4,000 infantry and 200 horse for a Roman legion. Upon a second glance, the figure for a Roman legion of 4,000 foot and 300 horse is stated specifically only for Sempronius Longus' expeditionary force. Polybius does not provide a tally for the Roman dispositions of their plans for Spain and Africa in 218 B.C.

If Polybius and Livy are both available, Polybius' account of something should generally be accepted over Livy. But Livy, a literary, not analytical, genius of writing, would have provided us with the most detailed account of the war even if we had Polybius' work in full; Livy tells us more about troop dispositions and the political scenes throughout the great conflict with Hannibal. No question there was much scaremongering and glamorizing from the patriotic Livy, but he takes time to mention the problem of the discrepancies among the varying sources, and he perhaps tells us more than any other historian of antiquity of his sources; he barely mentions Polybius by name, but when he does, it is with praise, evidenced by his comment in Book 36.19, when discussing the battle losses between Romans and Macedonians, in the fighting at Thermoplylae in 191 B.C.,

"...is the number I have given, on the authority of Polybius. It would be another a matter if we were to credit Valerias Antias!..."

But most of all, his comment in Book 33.10, when discussing the fighting at Cynocephalae,

"...we have in fact followed Polybius, a reliable authority on all Roman history, but especially to be trusted on affairs in Greece..."

I guess my point is, from my view, we need to nitpick at each isolated situation from the two - when there is some question; there is no 'set-rule'. Polybius is almost certainly correct most of the time, if their works do not concur, and he had the experience in military and political affairs. But Livy fills in many gaps not provided by Polybius (though we have to wonder where Livy arrived at conclusions not provided by Polybius), and unless his details are obviously wrong (eg, stating that there were 57,000 Carthaginian deaths at the Metaurus), I think we should accept his works as, on whole, reliable.

Anyway, let's take a look at the Senate's pragmatic plan for the conduct of the opening phase of the war, at which they knew Hannibal was on the march (though they never could have imagined the war could be carried into Italy itself, by sea or land). Polybius tells us, Book 3.40-41,

"...on news reaching them sooner than they had expected that Hannibal had crossed the Ebro with his army, determined to send, with their legions, the Consuls Publius Cornelius Scipio to Spain and Tiberius Sempronius Longus to Africa.

While occupied in enrolling the legions and making other preparations they were pushing on the project of establishing in Cisalpine Gaul the colonies on which they had decided. They took active steps to fortify the towns, and ordered the colonists, who were about 6,000 in number for either city, to be on the spot within 30 days. The one city they founded on one side of the Po, they called it Placentia, the other, which they named Cremona, was on the far side (northern side). Scarce had both these colonies been established when the Boii Gauls, who had been for long as it were lying in wait to throw off their allegiance to Rome, but had hitherto found no opportunity, became elated now by the messages they received assuring them of the near arrival of the Carthaginians, revolted from Rome, abandoning the hostages they gave at the end of the former war which I described in my last Book. Calling on the Insubres to join them, whose support they easily gained owing to their long-standing rancour against Rome, they overran the lands which the Romans had allotted to their colonies and on the settlers taking to flight, pursued them to Mutina, a Roman colony, and there besieged them. Among those shut up there were three men of high rank who had been sent to carry out the partitionment of the country, Gaius Lutatius, a former consul, and two former praetors. On these three requesting a parley with the Boii, the latter consented, but when they came out for the purpose they treacherously made them prisoners, hoping by means of them to get back their own hostages. When the praetor Lucius Manlius, who with his troops was occupying an advanced position in the neighborhood, heard of this, he hastened up to give help. The Boii had heard of his approach, and posting ambuscades in a certain forest attacked him from all sides at once as soon as he reached the wooded country, and killed many of the Romans. The remainder at first took to flight, but on getting to higher ground rallied just enough to give their retreat an appearance of order. The Boii following at their heels shut this force too up in the place called Vicus Tannetis. When the news reached Rome that the 4th Legion was surrounded by the Boii and besieged, they instantly sent off the legions destined for Publius under the command of a praetor to its assistance, ordering Publius to enroll other legions from the allies.

The condition and course of Celtic affairs from the outset up to the arrival of Hannibal were such as I have narrated here and in the previous Book. The two Roman consuls, having made all preparations for their respective enterprises, set sail early in summer to take in hand the operations determined on, Publius bound for Iberia with 60 ships and Tiberius Sempronius for Africa with 160 quinqueremes..."

Livy provides the same events later with a little more detail (Book 21.25-26); in Book 21.17, he tells us,

"...The seat of the campaigns had already been decided; the consuls were now ordered to proceed to the sortition. Spain fell to Cornelius, Africa to Sempronius. It was resolved that six legions should be raised for that year, the allies were to furnish such contingents as the consuls should deem necessary, and as large a fleet as possible was to be fitted out; 24,000 Roman infantry were called up and 1,800 cavalry; the allies contributed 40,000 infantry and 4,400 cavalry, and a fleet of 220 ships of war and 20 light galleys was launched. The question was then formally submitted to the People whether it was their wish and command that war should be declared against the people of Carthage. When this was decided, a special service of intercession was conducted; the procession marched through the streets of the city offering prayers at the various temples that the gods would grant a happy and prosperous issue to the war which the people of Rome had now ordered.

The forces were divided between the consuls in the following way: to Sempronius two legions were assigned, each consisting of 4,000 infantry and 300 cavalry, and 16,000 infantry and 1,800 cavalry from the allied contingents. He was also provided with 160 warships and 12 light galleys. With this combined land and sea force he was sent to Sicily, with instructions to cross over to Africa if the other consul succeeded in preventing the Carthaginian from invading Italy. Cornelius, on the other hand, was provided with a smaller force, as Lucius Manlius, the praetor, was himself being dispatched to Cisalpine Gaul with a considerable force. Cornelius was weakest in his ships; he had only 60 warships, for it was never supposed that the enemy would come by sea or use his navy for offensive purposes. His land force was made up of two Roman legions, with their complement of cavalry, and 14,000 infantry from the allies with 1,600 cavalry. The province of Cisalpine Gaul also received two Roman legions and 10,000 allied infantry with 600 Roman and 1,000 allied cavalry..."

OK. Between our two primary sources, it is clear that Roman preperations in 218 B.C. were based on the supposition that the war would be waged in Spain and Africa. But a discrepancy arises with Livy stating that six legions had been raised, but Polybius telling us that the 'considerable force', according to Livy, dispatched under the praetor Manlius was the 4th Legion; in this time, a consul would normally be given command of two legions, while a praetor usually led one, often for the purpose of judicial governing, as was the case here. This 4th Legion would normally have been under the command of one of the consuls, but was obviously allocated for the troubling event in Cisalpine Gaul, and they must have deemed it serious enough, with war against Carthage just brewing, to send both Scipio's originally assigned legions to Cisalpine Gaul (under another praetor, Gaius Atilius). Scipio, from Polybius, was ordered to levy two more legions, but Livy seems to think that only one of his legions was sent, as he tells a little later in Book 21.26,

"...When the intelligence of this sudden outbreak reached Rome and the senate became aware that they had a Gallic war to face in addition to the war with Carthage, they ordered Gaius Atilius, the praetor, to go to the relief of Manlius with a Roman legion and 5,000 men who had been recently enlisted by the consul from among the allies. As the enemy, afraid to meet these reinforcements, had retired, Atilius reached Tannetum without any fighting. After raising a fresh legion in place of the one which had been sent away with the praetor, Publiu. Cornelius Scipio set sail with his 60 warships and coasted along by the shores of Etruria and Liguria, and from there past the mountains of the Salyes until he reached Marseilles..."

Both historians are a little addled, but it seems that six legions were indeed the amount of newly formed forces raised for the uncoming ventures in Spain and Africa, and to deal with the uprising on the Po. Once Sempronius was diverted quickly to nothern Italy from Sicily; in the wake of the Senate's astonishment that not only had Hannibal arrived in northern Italy, but had also scattered Scipio at the Ticinus River. Both Polybius and Livy believe there were four Roman legions in the north (two were now under Ganeus Scipio in NE Spain), thus what probably occured was that four legions were initially raised for the two major incursions under Scipio and Longus, but the trouble on the Po compelled the Senate to give Manlius one of Scipio's legions and head to Mutina, and subsequently Scipio's second legion was sent to the area under Atilius, compelling Scipio to raise two new (er) legions.

Indeed, Matthew, the standard strength of the Republican legion during this period consisted of 4,200 infantry and 300 cavalry. But it's a mistake for anyone to say it was 'fixed', or rigidly imposed. If military situations may have demanded, it could be strengthened or depleted, and Aryaman2 and L C Cinna are indeed correct that sickness, casualties, and other tasks for legionaries would erode a legion's numerical strength on campaign. A demanding situation even had legions at this time raised to over five and 6,000. But if legions reached a point of being substantially understrength due to such elements (say, 25% or more), I would be inclined to think we would have been told about it, such as Caesar's mentioning of his seven legions numbering merely 15,000 men when he headed to Greece to face Pompey - substantially understrength, but his total troop strength was equivalent to three 'regular' legions. Basically, the longer a legion remained in service, the smaller it became, as the killed or disabled were not replaced by recruits enrolled in the same legion; when an accession of strength was received, it was usually towards new or different corps. Thus, a legion itslef would be reduced in numbers, but more corps could be formed for the whole army on campaign. This was mainly the case in the later Republican period..

Moreover, we hear of a plague-like condition that seriously sapped the Carthaginian army on the outskirts of Syracuse, one originally composed of, according to Livy, 25,000 foot, 3,000 cavalry and 12 elephants, which had embarked at Heraclea Minoa from Africa in 213 B.C.

I don't know the name Evelyn S. Shuckburgh, Matthew. The velites (originally known possibly as rorarii) were presumably not organized into maniples of their own; we are not told how they were organized or commanded in any detail, but they indeed were the the light troops. What is probable is that they, adminstratively, were attached to the maniples of the heavier infantrymen, but in battle they fought as skirmishers in the open, both in preliminary fighting and in support of the heavier infantry and/or cavalry.

I believe the 800 differential (did you mean the standard 4,200 legion as opposed to an augmented 5,000?) you were asking about would be filled amongst all the hastati, princeps, and velites; the maniples would be enlarged with the first two and the number of velites would be increased. The horsemen (equites) and triarii were the wealthiest and more experienced men, thus they were recruited from a more limited pool.

But you seemed to cover all that very sapiently.

We are not sure, but the tactical units Polybius describes as 'cohorts', were the subdivisions of the alae, the allied wings of primarily Latin troops. They numbered between 400-600, and their size was presumably influenced by the size of the legions they were supporting. The crack troops of the allies were known as the extraordinarii, a seperate force of infantry and cavalry placed at the immediate disposal of the consul. This could have been one or more cohorts. The basic function of the extraordinarii was to head up the column during an advance and bring up the rear upon retreat (or 'retirement').

...Cannae is a good example, as the figure Polybius gives of 80,000 is clearly the result of him calculating 8 Reinforced Legions (5,000 x 8) and an equivalent number of Allies (5,000 x 8) . Taking the reduced Cavalry number of 200 per Legion, we can also account for the 6,000 Horse in relation to this figure, as the Legionary Horse would be 1,600 (200 x 8) and the Allied Horse around 4,800 (three times as many), the total being 6,400, which is very close to the figures Polybius gives (6.113. - 'eighty thousand foot and rather more than six thousand horse').
Livy, though, seems more inclined to talk about 50,000 Foot, though without saying that is the total number in the army. Interestingly, if there were 50,000 Foot, the 6,000 Horse would also be explicable in terms of 1,500 Roman Horse [5 x 300] and 4,500 Allied Horse [5 x 900]. Of course, five Legions wouldn't make much sense, but it is possible. Could Livy have been retroactively reducing the number of Infantry based on the number of Cavalry? Was Polybius inflating the number of Infantry in comparison to Cavalry to prove his thesis: 'a lesson to posterity that in actual war it is better to have half the number of infantry, and the superiority in cavalry, than to engage your enemy with an equality in both' (3.117.), which he then goes on to demonstrate the significance of at Zama.

This sort of thing always worries me when thinking about numbers, but all the same, I think Scipio's forces must have been closer to 30,000 than 10,000. I do wonder how Legions of 6,200 Foot and 300 Horse were divided up into Maniples (5,000 + 1,200, maybe? A fourth line?)...

Actually, Livy tells us of two variants - four new legions, to make a total of eight for the campaign, and the other possibility that 'some authors' record a supplementary mobilization of just 10,000, which, in addition to the existing four legions, we would have some 50,000 foot. He doesn't tell us what he thinks, only that the effort and vigor devoted to this campaign against Hannibal was greater than ever in Rome's history. But his list of Roman dead, prisoners, and escapees, which he tells us in a more decisive style, amounts to over 83,000 men. Despite the clamity that befell the Romans, their tactics of reverting their infantry to bring about superior weight, in a fashion akin to a massive phalanx, upon the enemy are incredulous if they outnumbered Hannibal's infantry by a small proportion (perhaps 40,000 vs. 32,000, allocating for troops left in the respective camps etc.). I think the 6,400 total cavalry is plausible, but a different proportion: Polybius tells us that the cavalry was increased, Book 3.107,

"...They decided to bring eight legions into the field, a thing which had never been done before by the Romans, each legion consisting of about five thousand men apart from the allies. For, as I previously explained, they invariably employ four legions, each numbering about 4,000 foot and 200 horse, but on occasions of exceptional gravity they raise the number of foot in each legion to 5,000 and that of the cavalry to 300. They make the number of the allied infantry equal to that of the Roman legions, but, as a rule, the allied cavalry are three times as numerous as the Roman..."

Now, on paper, this would mean the Roman cavalry at Cannae would have been 2,400, which was probably accurate, as 1,600 would have been sheer suicide against what they knew Hannibal could counter them with (2,400 was a sure loss anyway, but they wanted to hold up for as long as possible while the infantry won the battle). The allied cavalry would be 7,200, as Polybius tells us the proportion of 3-to-1 was 'as a rule'; according to him, as you said, the whole cavalry force of Rome at Cannae numbered 'rather more' than 6,000, thus the rule could not be applied this time, and the allied cavalry was at least 4,000 horsemen. Earlier in the campaign Gnaeus Servilius Geminus, along with his lieutenant Gaius Centenius, had lost almost all the cavalry attached to the 1st, 2nd, 12th, and 13th legions, losses which would have been very difficult to replace in full. Obviously, if it had been possible, 7,200 allied horsemen would have been raised to compliment the Roman horse. But the Romans had been depleted in cavalry too, thus if they were smaller than 2,400, perhaps the allied horse numbered 'rather more' than 4,000. They knew Hannibal's cavalry was their paramount concern.

However, Livy tells us in Book 22.36,

"...The armies were increased, but as to what additions were made to the infantry and cavalry, the authorities vary so much, both as to the numbers and nature of the forces, that I should hardly venture to assert anything as positively certain. Some say that 10,000 recruits were called out to make up the losses; others, that four new legions were enrolled so that they might carry on the war with eight legions. Some authorities record that both horse and foot in the legions were made stronger by the addition of 1,000 infantry and 100 cavalry to each, so that they contained 5,000 infantry and 300 cavalry, while the allies furnished double the number of cavalry and an equal number of infantry. Thus, according to these writers, there were 87,200 men in the Roman camp when the battle of Cannae was fought..."

Well, at the risk of too much picking and choosing, segments of Livy can arguably be preferred, particularly that in this case the allies furnished double, not triple (the usual 'rule), the cavalry, and his connotative manner here is one of 'what others tell us' - an honest and professional approach, when one doesn't know for sure (but he clearly follows Polybius more than any other). Following Livy, it is not implausible that the cavalry of the alae numbered 4,800 horsemen at Cannae, presuming this is how many were 'in the camp' when the battle was about to be fought; certainly, not all 87,200 men fought in the battle. However, we have to keep in mind that Hannibal's Numidians, who were used to fight a containment battle with the allied horse of Rome, could not have numbered too much less than the allies opposite them, as they succeeded thoroughly in that very containment. Thus we must ask how many Numidians Hannibal deployed on his right. Hannibal's original number of 6,000 cavalry, upon his arrival in Italy, could not have increased in Numidians yet. But the gauls beared the brunt of the fighting They probably numbered a little less than 4,000, with their counterparts numbering well over 4,000. Against not so vast odds, the more maneuverable Numidians would certainly be able to hold up for a while. But that's just a guess.

I would say Polybius' theory was based on the ingenious battlefield control exercised by Hannibal, particularly with his superior cavalry, and the egregious failure of the Romans at Cannae; even though they pragmatically nullified Hannibal's ability to outflank them, as Varro had disposed his forces so the hilly terrain was on his left, the river on his right, and the sea behind him. Many constantly talk of Hannibal's superiority in cavalry, and the ground that helped him. But given the nature of numbers of the forces at Cannae, infantry also required reasonably level ground, and Varro deserves credit for refusing battle on July 31; we always here of Paullus refusing battle on August 1, as he didn't like the terrain. But Varro made the same decision based on the same situation the day before, on his day to command. The west side of the Aufidus contained terrain which was flatter over a more widespread area, and both consuls saw this. Hannibal may have possessed advantages at Cannae, but as with Alexander, a superior instrument is merely potential when faced against a substantially more numerous opponent. It is the commander's utilization of that instrument is what wins the battle. The east bank (Roman left) was still suitable for cavalry, but provided more hope for infantry against superior cavalry. Hannibal simply came up with the answer. Perhaps Varro could have steadied his cavalry squadrons with some foot, as he had done in an earlier skirmishing success against Hannibal's foragers, but he clearly needed to use his infantry to its maximum potential with what he was trying to achieve, and a wider frontage is less stable. It was indeed a lesson to the posterity of warfare, and Scipio the Younger, who had been here and escaped, surely studied all this to the letter. The very name of Cannae would become a generic noun to soldiers of all ages.

Judging by what happened at the battle, it is not impossible for what Hannibal achieved against so huge a force, being he sucked them in, and his infantry dispositions caused the Roman infantry to funnel in towards the center. Livy's figures of casualites, though piecemeal, add up to Polybius' assessment of what the Romans had at the beginning. Livy tells us that some 10,000 escaped to Canusium, which is entirely consistent with his information that they later were formed into two legions. Polybius' figures do not make sense, as he gives us a higher figure of losses than what he tells us the Romans began with. But Peter Connolly tells us the succeeding part of Polybius' account (that to his casualty list, Book 3.117) is missing, and that his high figures cannot be corroborated by subsequent events (Greece and Rome at War, Pg. 188). We'll never know for sure, and too much extrapolating can be trite. Like I'm doing :? !!

I'll get back to this thread again.

Thanks, James

[Spartan JKM]

OK. Let's continue here.
Aryaman2:

...From a logical military perspective, before embarking Scipio discharged unfit men from the Cannae legions and brought them up to full strength with new recruits, it makes little sense to bring all those green recruits to overload your transports and logistics, instead it is more logic to transport only your veterans, while the new recruits are left in garrison duty in Sicily, where they could be trained and get seasoned. Then they could be brought in small convoys little by little as replacements, without compromising logistics...

That does make sense, Inaki, but I don't think that is what occured: Livy tells us that 3,000 men were raised - a year after Scipio's departure - for service in Sicily for the very reason there were not any crack troops left in Sicily (Book 30.2). This was a year later, and if any training of the newer recruits had been required, and for them to begin sailing to Africa in smaller convoys to join Scipio's army, I cannot believe that we wouldn't have been told this, in whatever form. There is never any mention from our sources that small convoys of troop detachments were sent to Scipio during his campaign in Africa; I am all for revisionism, reading between the lines, and applying deductive logic. Although what you propose makes logistical sense, what happened, according to Livy, does not seem impracticable; he tells us in detail of Scipio being amply supplied from Sicily and Sardinia in 204 B.C. ( Book 29.36). The next year, some 100 'freighters', supported by 20 warships under one Publius Lentulus, reached Africa to supply Scipio - from Sardinia in 203 B.C. The simultaneous supply effort from Sicily under Gnaeus Octavius, with 200 transports and 30 warships, met with bad weather, were swept around the waters too close to Carthage, and rounded up by the Carthaginians (Book 30.24).

Following Livy, in 204 B.C. Scipio and the praetor Marcus Pomponius Matho discussed the logistics for the expedition to Africa. When all the ships were mustered, the harbor of Lilybaeum could not hold all the ships, and the town all the men, 'such was the enthusiasm for crossing to Africa' (Book 29.24). But Scipio had already been in Sicily for a year, and his 7,000 volunteers had come with him. He had immediately formed the 7,000 into 'ranks and centuries' (Livy Book 29.1), which is the process of of forming them into legions, first assigning them, according to age and experience, into the three ranks of hastati, princeps, and triarii, then, presumably, disseminating each rank into 20 centuriae (two centuriae make a maniple).

It was in 204 B.C. that Scipio inspected the Cannae legions, who among the 'fit' ones, were as experienced and seasoned in various forms of combat, including the blockading of towns (remember, many had continued fighting the enemy here in Sicily, of which much was garrison duty). The unfit men from the Cannae legions were replaced with the volunteers, who by now, a year later, were trained and ready (though not battle-hardened). Livy says that other 'supplementary troops' were added to the 5th and 6th to bring the number of each legion up to 6,200 infantry and 300 cavalry. The figure of 17,600, an accepted one from some writers, doesn't make complete sense, being that Scipio's landing in Africa was the only one to occur throughout the three years of this campaign; if indeed that were the case, that figure would mean that the four legions (two Roman, two alae), would be both of basic strength - 4,200 foot and 200 horse. We would expect lower proportions of cavalry usually furnished by the allies, due to the difficulty of transporting horses by sea. But for the allied contingents not to be a little numerically stronger would be peculiar, because there were plenty of allied troops in Sicily upon Scipio's arrival there, given the grumbling they were making in 209 B.C. (Livy Book 27.9). Moreover, Scipio had brought the 7,000 volunteers with him, many of which were allied troops.

...Thucydides, a most reliable source, says that the Athenian expedition to Sicily was composed of 134 ships to transport just 5.100 infantrymen. The relation between ships and men is dependant on multiple factors...

True, we have to be careful with the mathematical deductions - distance, ship size etc. The expedidition to Sicily, which left Corcyra probably in June of 415 B.C., was composed in all of 267 ships (of all kinds, including one horse transport). The total figure of hoplites, Argives, Mantineans and 'other mercenaries', bowmen, slingers, light-armed Megarian 'exiles', and 30 cavalry, amounted to an invasion force of 7,180. They travelled twice as far as Scipio, hugging the instep of Italy along the way, and averaged 26 men per ship aboard vessels much smaller than what the Romans had over two centuries later. Of course, the ship composition was varied. See Thucydides, Book 6.43-44.

Another example I recall about ships and troops, from an eyewitness account, the Byzantine expedition against the Vandals according to Procopius mustered 10.000 foot and 5.000 horse, and required 500 transport ships plus 92 warships, so again the 400 ships of Scipio expedition don´t seem to be too many.

Indeed, Procopius is perhaps the best source of the ancient world, in terms of a real eyewitness and whose works we have in full. This huge expedition also served as a commercial voyage as well as an attack upon the Vandal kingdom, and one grandiose in nature - transporting more than 5,000 horses (there were also some Huns and Heruls with the 5,000 horsemen) was unprecedented, before and after; they travelled from Byzantium to Sicily, receiving diplomatic and logistcal support from the Ostrogothic queen (I don't recall her name), and continued on to Africa. But you are missing an important detail Inaki: there were an additional 30,000 Ionian, Egyptian, and Cilician sailors who went with the army - a total of 45,000+ men. See Procopius Book 3 (I think Ch. 10, 11 or 12; definitely around there).

...I know you are familiar with the 30YW and the difficulties armies just about 20.000 found supplying themselves in Germany, how the 43.000 army Gustavus Adolphus assembled at Alte veste badly starved, imagine then how difficult was to supply armies of 35.000+ in North Africa.

Ah-ha. This is Aryaman from paradoxplaza! I did notice a thread over there discussing the 'incorrect' numbers given by the ancients; but they are only discussing Alesia, specifically the numerical strength of the Gallic relief army - one of the couple of known instances of ludicrous magnification, another being that there a million soldiers under Darius III at Gaugamela.

Gustavus Adolphus was as proficient as any leader in military history with the proper planning of transport and supply. But like Hannibal and Scipio etc., he took it for granted that his men had to the right to be fed by whatever lands they subjugated. Gustavus knew that the horrors of war were inevitable. But he didn't rely soley on plunder: he innovated (or perhaps augmented) a structured system of requisitions, feeding his army from magazines established at suitable centers; his regiments were provisioned in bulk from a regular staff who distributed the supplies. This eliminated his army's need to (substantially) disperse, in order to forage and pillage. Moreover, his system didn't require the constant need to move into fresh regions. This reform, so invaluable for his time, also enabled Gustavus to sustain great mobility, as he was able to reduce his baggage train, because the preponderance of camp-followers, a feature of 17th century armies, was now not so necessary.

But when an attritional campaign endured, the supply situation would greatly be reduced. Both the armies of Gustavus and Wallenstein were starving by the end of August, 1632, at the Alte Veste, but it was Gustavus who attempted to force battle to relieve the problem. Indeed, it had been a contest of will power, famine being the principal weapon, and Wallenstein won; he had smartly fortified his position across Gustavus' lines of communications. Three days after Gustavus retired, Wallenstein devastated Saxony to feed his men. For nearly two months (all of July, most of August, 1632) the two armies faced each other immovably, thus exhausting their supply; Gustavus could not draw from magazines, or even the territory around Nuremberg, used up, and he was not in a position to establish more depots. This is why he sought battle, not only from his prior experience in Ducal Prussia, but to prevent the melting away of his army. But Wallenstein was surprised by Gustavus in early November on the River Rippach, and perhaps only nightfall saved him. The subsequent Battle of Lutzen is rife with conjecture, but what is certain is that this time Wallenstein was driven from his trenches. Whoops, I'm 1,835 years off topic!

Scipio never found himself in such a position in Africa; he had established a good base at Utica, with the navy supplying him from Sicily and Sardinia, the ideal provinces for such a task. Livy tells us that, after the Great Plains battle, Scipio's army was heavily laden with booty even before he captured an abandoned Tynes (modern Tunis), a valuable position in view of Carthage, and possessed of good defence-works (Book 30.9).

I would like to discuss your numbers of Caesar expedition to Britain. I think that is a very valuable source, as he is an eyewitness and he is talking about his own forces, you said
"Julius Caesar's first expedition to Britain was composed of, at most, 10,000 men (a legion with Caesar was usually around 4,000, maybe less, and he took two with him) and about 300 cavalry aboard 98 ships (80 for the troops, 18 for the horses)"
There you make the same mistake that Ancient authors, you multiply the number of legions by his paper strength, however units in campaign are almost never at full strength. Caesar himself says that at Pharsalos his cohortes averaged 275 men instead of the authorised strength of 480, and that some legions were at half strength. It is not unreasonable to think that the 2 legions could well have mustered about 2/3 of their strength, that is 3.200 each, that would give us 7.400 infantry+300 cavalry with 98 ships.
However in 54 BC Caesar organized a larger expedition, mustering 5 legions,and 2.000 cavalry, by the same counting about 16.000 infantry and 2.000 cavalry, but he required no less than 800 ships! So the 400 ships of Scipio don´t look too much off the mark, besides remember that he transported also rations for 60 days.

Sure. Caesar's numbers at Pharsalus were significantly understrength - and he records it. The quantity of his legions for that campaign is not a valid gauge to ascertain the numerical strength of his legions in Gaul 6-10 years earlier. We have no means of knowing how many men he had at certain points, but we can be sure they were larger in Gaul. He harangued his men upon leaving for Greece, The Civil War 3.6,

"...That since they were now almost arrived at the termination of their toils and dangers, they should patiently submit to leave their slaves and baggage in Italy, and to embark without luggage, that a greater number of men might be put on board: that they might expect every thing from victory and his liberality..."

To reiterate, the longer a legion remained in service the smaller it became: later in the Civil War, in Book 3.89 and 3.106, we read that the effective strength of two legions was a total of 3,200 men. Absolutely, units on campaign were depleted from their papaer strength, but if we are not told, for the most part, I think it's a good guess that the depletion wasn't significantly understrength, as Caesar's was at Pharsalus. We can apply a little guesswork from plucking out information amid Caesar's account; The Gallic War, Book 1.42, when Ariovistus wanted a conference with Caesar,

"... as ambassadors were being often sent to and fro between them, Ariovistus demanded that Caesar should not bring any foot-soldier with him to the conference, [saying] that "he was afraid of being ensnared by him through treachery; that both should come accompanied by cavalry; that he would not come on any other condition." Caesar, as he neither wished that the conference should, by an excuse thrown in the way, be set aside, nor durst trust his life to the cavalry of the Gauls, decided that it would be most expedient to take away from the Gallic cavalry all their horses, and thereon to mount the legionary soldiers of the 10th legion, in which he placed the greatest confidence, in order that he might have a body-guard as trustworthy as possible, should there be any need for action..."

OK. He's telling us he's taking away all the Gallic cavalry's horses, and placing his legionaries on them. Earlier he tells us. Book 1.15, when fighting the Helvetii in a cavalry clash,

"...sends forward all his cavalry, to the number of 4,000 (which he had drawn together from all parts of the Province and from the Aedui and their allies), to observe toward what parts the enemy are directing their march. These, having too eagerly pursued the enemy's rear, come to a battle with the cavalry of the Helvetii in a disadvantageous place, and a few of our men fall...""

Given the context ('the', not 'some of', the 10th legion) of the meeting with Ariovistus, the whole 10th legion was taken, mounted upon these 4,000 horses. Thus the famed 10th legion in 58 B.C. was, in campaign, about 4,000 men. In the 5th year of the Gallic War (53 B.C.), while fighting the Nervii, we are told, Book 5.48-49,

"...Caesar, approving of his motives, although he was disappointed in his expectation of three legions, and reduced to two, yet placed his only hopes of the common safety in dispatch...fortifies a camp in the most favorable position he can. And this, though it was small in itself, [there being] scarcely 7,000 men, and these too without baggage, still by the narrowness of the passages, he contracts as much as he can,..."

So, now we have two legions numbering 3,500 men each. The 10th legion was one of the two taken to Britain in 55 B.C. Now, I stated 10,000 at most was the possible number of Caesar' first expedition to Britain, and I should have specified that 4,000 men was the average in Caesar's legion at the time. It's a good guess they were between 8-10,000 (total) aboard the 98 ships. Closer to 4,000 is probably accurate, and perhaps a little less by 55 B.C. Some of his soldiers may have undertook the task of rowers in the warships; Caesar states they did on the second expedition, in The Gallic War, Book 5.8. Between Theodore Dodge, J.F.C. Fuller, and Adrian Goldsworthy, all of whom wrote biographies of Julius Caesar, they all opine Caesar's 7th and 10th legions numbered between 8-10,000 men, and they indeed address the issue of depleted ranks on campaign (not Fuller as much, but he must know of the issueevery bit as much), in relation to their paper strength, and that later in his career was where his legions considerably less in quantity. The only figure Caesar gives us is 2,000 horse left with Titus Labienus and his three legions, to defend the harbors and 'discover' what was going on in Gaul, as he mentions in The Gallic War, Book 5.8,

About Caesar' second British expedition, well, let's break it down. With regards to the new ships for the second expedition, he tells us, The Gallic War, Book 5.1,

"... during the winter as many ships as possible should be built, and the old repaired. He plans the size and shape of them. For dispatch of lading, and for drawing them on shore, he makes them a little lower than those which we have been accustomed to use in our sea; and that so much the more, because he knew that, on account of the frequent changes of the tide, less swells occurred there; for the purpose of transporting burdens and a great number of horses, [he makes them] a little broader than those which we use in other seas. All these he orders to be constructed for lightness and expedition, to which object their lowness contributes greatly..."

OK. The new ships to transport horses and burdens (being weighed down with supplies) were constructed to be a little larger; the troop transports were constructed to be faster and lighter, thus they weren't larger than before. Maybe not necessarily smaller, but they were now lower to the water-line. Obviously, loading and beaching would be facilitated. In the next chapter of The Gallic War (Book 5.2), Caesar tells us,

"...he finds that, by the extraordinary ardor of the soldiers, amid the utmost scarcity of all materials, about 600 ships of that kind which we have described above and 28 ships of war, had been built, and were not far from that state, that they might be launched in a few days. Having commended the soldiers and those who had presided over the work, he informs them what he wishes to be done, and orders all the ships to assemble at Port Itius, from which port he had learned that the passage into Britain was shortest..."

Port Itius was probably located near where modern Wissant is located (the fine beach of Cap Blanc Nez). OK. 600 new ships and 28 warships. Julius then tells us he hastily prevented a potential rebellion among the Treveri, who were 'tampering' with the Germans beyond the Rhine; the Treveri were also troubled by internal discord, and Caesar marched through the Ardennes with four legions and 800 horse, and settled things; he then proceeds, Book 5.5 of the The Gallic War,

"...Caesar went to Port Itius with the legions. There he discovers that 60 ships, which had been built in the country of the Meldi, having been driven back by a storm, had been unable to maintain their course, and had returned to the same port from which they had set out; he finds the rest ready for sailing, and furnished with every thing..."

That's 540 new ships ready to go. Caesar plans to take suspect Gallic chieftains to Britain with him, as hostages, thus they couldn't stir up trouble in his absence. Dumnorix of the Aedui pleads to be left behind, Caesar refuses, and the Aeduan chief is cut down soon after he flees. Caesar tells us alittle later of the second landing, The Gallic War, Book 5.8,

"...All the ships reached Britain nearly at mid-day; nor was there seen a [single] enemy in that place, but, as Caesar afterward found from some prisoners, though large bodies of troops had assembled there, yet being alarmed by the great number of our ships, more than 800 of which, including the ships of the preceding year, and those private vessels which each had built for his own convenience, had appeared at one time, they had quitted the coast and concealed themselves among the higher points..."

Now, obviously he planned to bring Britain, among a few things he possibly wanted to achieve (military adventuresome being one), into the political sphere of Rome; taking private vessels could only mean an act of diplomacy - and he took many of them. Yes, he's telling us that there arrived off the shore of SE Britain, probably near present-day Sandwich, with more than 800 ships. He tells us his five legions contained 2,000 horse ('equal to that he left on the continent'), but he doesn't tell us how many transports were relegated for the horses and their mounts themselves; for the first expedition, he tells us 18 transports were used for the cavalry, but doesn't give us a figure for that cavalry force (the consensus is about 300).

Thus, we have 540 transports and 28 warships - specific figures given to us by Caesar; thus, the remaining ships, to bring us to 'more than' 800, would be 'more than' 232 private vessels. His figures are very puzzling; let's assume, if I may, that the first voyage indeed comprised 8,000 foot and 300 horse, aboard what we know as being 80 transports for the footmen and 18 for the cavalry. That's 100 men per ship for the infantry, and 17 horse and mount per transport. We were apprised that the new ships were built for horses and burdens were a little larger. So, let's say 20 horses and extra supplies could fit in one transport. 2,000 cavalry would entail 100 transports. Because of the new design, and by this deductive logic (not etched in tone), Caesar needed 17 less ships for the cavalry (proportionately) than his first expedition. Now we have 'more than' 700 total ships. Assuming the the 'more than' appended to the 800 total is attributed to the private vessels, that gives us 478 transports (including the 28 warships) for the infantry. As for the warships, we can assume that they were not used for the transport of Caesar's men, due to so many transports being utilized. Most of the supplies were with the horse transports, as those ships were built to accomodate them, and this was hardly a long voyage, as they arrived off Britain the following morning (the 'sun was setting' upon departure the previous day).

Boy, this should be easier, huh? Again, the newer transports for the infantry were designed to beach and embark/disembark easier; they were not built to support a higher capacity. Thus, I am going to allot the same figure of 100 men per transport among the infantry: 20,000 footmen (maybe a little less, and not a little more) crossed to Britain with Caesar aboard 200 transports. The horses and bulk of the supplies were aboard 100 transports. The warships numbered 28, giving us 328 ships. If we add 'more than' 232 private vessels, we arrive at the figure of 'more than' 560 ships in all. That gives us a surplus of 'more than' 240 ships. I am not an expert in military logistics, of any period, but I cannot see the efficacy in Caesar building and additional 240 ships so that his infantry could be more comfortably spread in 440 ships - 45 men per ship. It is possible, being there were almost certainly not 368,000 people (92,000 warriors) of the Helvetii and their allies who migrated SW from their lands in 58 B.C. (The Gallic War, Book 1.29), to avoid the possible attack upon them by Ariovistus, and there were certainly not 258,000 relieving Gauls at Alesia in 52 B.C. (The Gallic War, Book 7.76) - illustrations of Caesar's propensity to exaggerate absurdly - , that there were not 'more than' 800 ships that terrified the Britons; 560 or so would still be more than they ever had seen at once, thus they would still have fled inland to prepare for a resistence, which Casivellaunus would prove quite adept at effecting, resorting to a guerilla-style campaign against the legionaries, to which Caesar commended (The Gallic War, Book 5.16). But diplomacy won over most of SE Britian for Caesar, compelling Casivellaunus to come to terms.

So, in my loose opinion, Caesar's five legions, possibly numbering 20,000 (probably less) foot and 2,000 horse, possibly sailed aboard 300 transports - 2/3 of what Scipio sailed to Africa with in a two day trip 150 years before, and with 2/3 the amount of troop transports.
L C Cinna:

...Someone already mentioned Scipio's "losses" before the battle. Even if you get the numbers for troops transported to Africa right it still doesn't tell us anything about how many men Scipio had at Zama.

With "losses before the battle" I mean all troops which couldn't participate. Don't go and underestimate this.

The numbers for peole injured and ill due to fighting or other things like diseases and so on are always pretty high, especially in such an operation overseas in a slightly different climate. When you look at such army strenght reports through all of history you notice that a lot of people will be missing sooner or later. I just remember my own time in the army when we stayed outside for combat practice in winter after 1 week nearly a 5th of the regiment were whether ill, injured or not fully "usable" anymore and that without any enemy around.

The next thing is that you need a lot of troops to do other duties. Strategically important places need to be garrisoned, the supply lines secured, the camps secured to have a place to retreat to and so on.

I'm not an expert on such matters, I just speak of my own experiences, maybe someone knows how to calculate an average of troops which would not have been able to participate. If we remember that Scipio had several major and minor engagements before and needed to secure several strategically important places I'd say that at least 1/4th of his troops would not be present. So if he had over 30.000 paperstrenght before the campaign there would only be maybe 25.000 present at Zama as a maximum I'd say.

Just a guess.

That hardly strains credibility, Michael, and I'm no expert, but, as I've been intimating, if any substantial amount of troops (1/4 of 30,000 is still a large amount, from my view, when giving enumerations) were afflicted with a plague, or large strongholds needed to be garrisoned, and if supply-lines needed to be guarded, we would have been told, if not by the ancients, by modern experts in military matters, such as B.H. Liddell Hart or Bevin Alexander etc. Each event has to be scrutinized, and each campaign is different (as I'm sure you know, more than as you were in the army). Scipio did secure places, but found them abandoend, such as Tynes (as I mentioned a little earlier). His casualities were proportionatley very minimal; of the two major engagements he fought before Zama, in one he burned the enemy forces at night, the other was a textbook display of economizing one's forces.

Upon Hannibal's approach, there were no standing enemy forces whatsoever to Scipio's western flank, and Massinissa had won back his kingdom. Scipio underook a supremely shrewd but bold and indirect strategic move. Instead of trying to interpose between Hannibal and Carthage, or simply stand his ground until Massinissa arrived, he marched away from Hannibal and Carthage, SW along the Bagradas River - Carthage's only lifeline, whether food or levies of manpower, as Roman naval power had by now cut her from the Mediterranean. Scipio assaulted every town, refusing surrender. He appropriated all the supplies and sold everyone still alive into slavery. This was a paradigm of striking at his enemy's COG (Center of Gravity); Scipio was depriving Carthage of its ability to resist, and Hannibal would have to move before he wanted to, as he wasn't ready with the raw army he had (2/3 of it, actually). Moreover, every move Scipio made west closed the disatnce between he and Massinissa, who were force-marching to junction with him. Any town in their path had no chance either. Once he faced Hannibal on the ground of his choosing, now with a qualitative advantage in almost every facet (albeit he had smaller numbers, something Hannibal utilized smartly with a sound plan) of his arms, there was very little that needed to be garrisoned or secured, thus it would have possibly entailed a negligible amount of men: ie, the number I come up with at Zama would 100s less (maybe even 1,000) allocating for the elemnents you're mentioning. of course soldiers became ill, particularly in that time, but if Scipio had lost thousands from a virulent disease, I think we would have read about it. Some fresh soldiers perhaps joined him with the supplies we read from Livy about; but if so, they would have numbered very little.

But that's just my guess.

Thanks, James

[Aryaman2]

James
That was a long and full of details post, I enjoyed reading your effort. I think we have had several discussions on the Paradox forums, including some on XVII century tactics. About logistics in the XVII century I would recomend you the works of John A Lynn. Looking into the great pains it took then to supply 40.000 men, you would wonder how could be they supplied in ancient times, remember that Scipio moved inland, away from the coast, to fight Zama.
I will not enter into more details about armies transported by sea, I think there are too many missing details to be specific, my point was that in the sources the lower numbers could be supported as well as the higher numbers.
As for replacements, I think you put too much confidence in the ex silentio argument, in the II Punic War legions received replacements regularly for sure, however we are told of this only occasionally, remember than Livy or Polybius were writing literature, not military history, and even in modern books about modern wars those details are many times missing or buried in document lists, the unglamorous intrincacies of replacement and supply systems have been regularly relegated in history.
Anyway, it has been a very interesting discussion.
Regards
Iñaki