Wellington
02-06-2003, 02:13
Thought I'd post this as a seperate thread in order to elicit some discussion ... just my own opinion of Cannae.
The Background - Roman legions
------------------------------
At the time of the 2nd Punic war a Roman legion consisted of 4,000 infantry and 200 cavalry. Each legion was always supplemented by allies who provided at least the same number of infantry and considerably more cavalry (3:1). Therefore a "Legion" can be considered to be around 10,000 men, 10% being cavaly and around 40-45% being Roman citizens. Many ancient sources when refering to a legion of this period use the term to include both Roman and allied components.
Romes legions comprised not only citizens of Rome but also soldiers levied from several Latin cities that had been granted Roman citizenship. Romes allies were generally other Italian City States that, whilst not being granted Citizenship, paid tribute to Rome (not always volentarily). Rome specified the numbers and composition of troops that such allies were expected to provide. Such allies may have been equipped and fought in a similar manner to the Roman legions.
A Roman legion consisted of Velites (light troops/skirmishers), Hastatii, Principes and Triarii (the Veteran troops armed with long spears instead of pilum). The proportions were roughly equal - perhaps the Velites comprised 30% of a legion. The cavalry, both Roman and allied, consisted of the weathier classes - people who could afford to keep and equip a horse.
A Roman legion generally fought with the Velites out in front and 3 lines of heavy infantry - a 1st line of Hastatii, a 2nd line of Principes and a 3rd line of Triarii. The Triarii were only ever comitted to battle if the Hatatii and Principes had been defeated.
Roman forces generally operated in pairs of legions. A standard Roman consular army of this period comprised 4 legions which, together with allies, consisted of around 40,000 men at full strength. The standard deployment of a Roman army was Roman legions in the centre, allies on either flank and the cavalry on the wings.
Hannibal arrives in Italy
-------------------------
In 218BC Hannibal crossed the Alps and was confronted by a Roman army under Scipio/Sempronius (4 legions 1st/2nd/3rd/4th). A small cavalry engagement was fought by the Tacinus river (November 218BC) that Hannibal won. Polybius tells us the Roman cavalry "lost many men".
One month later the battle of the Trebia was fought. Rome fielded 16,000 men and 20,000 allies, were obviously weak in cavalry as a result of the Tacinus engagement, and were utterely defeated. Only 10,000 men escaped. Roman dead numbered 15,000 with around 10,000 prisoners being taken. Hannibal sold the Roman prisoners into slavery and released all allied prisoners after telling them "My fight is with Rome - not her allies".
A few months later, April 217BC, a second Roman army under Flaminius, 30,000 men consisting of 4 legions (2 full strength, 10th/11th, and the remains of the 3rd/4th) and allies, was ambushed by Hannibal at Lake Trasimene. Very few escaped. Rome lost 15,000 dead and around the same number prisoners. Hannibal sold 7,000 Romans to Greek slave traders and again released the allied prisoners.
Whilst Flaminius army was being slaughtered at Lake Trasimene the consul Geminius was marching with a further 4 legions (12th/13th and the remains of the 1st/2nd) to join him. Knowing Flaminius was weak in cavalry Geminius sent all 4,000 cavalry attached to his army ahead to join Flaminius. 4 days after Lake Trasimene was fought Hannibal sent Maharbal to ambush this cavalry on the Via Flaminia. Both Polybius and Livy say all this cavalry was lost (Polybius tells us 2,000 were killed and 2,000 captured). Geminius, upon hearing this, immediately returned to Rome with his 4 legions - minus the cavalry contingent.
All Roman prisoners taken in these battles were offered back to Rome for ransom. Rome refused, and Hannibal sold them into slavery.
Within 6 months Hannibal had totally destroyed 6 Roman legions (and allies), destroyed almost all the cavalry attached to 8 Roman legions (and allies), and had undoubtedly caused questions to be asked by Romes allies.
Hannibal re-equiped his army with all the captured Roman weapons/armour, moved into Central/Southern Italy and spent the next 12 months eliciting support from discontented Italian states.
The run up to Cannae
--------------------
Following these disasters the Senate appointed Fabius as Dictator for 6 months, as was the custom in times of danger.
Fabius immediately raised 4 new legions (14th/15th/16th/17th) and leaving the 16th and 17th to train in Rome marched to Geminius with the 14th/15th. Taking command of Geminius army Fabius now had a field army of 47,000 infantry and around 2,500 cavalry. This army kept in close contact with Hannibal but only a few minor skirmishes occured. Both Fabius and Minucius (who assumed command after Fabius was recalled to Rome) were reluctant to engage Hannibal. It is quite likely the morale of the Roman army was poor, especially considering the events of the last year.
After Fabius 6 month dictorship expired the Senate appointed new Consuls and a further levy was raised in order to increase the strength of the legions and create even more new ones This new levy was used to create 4 new legions (18th/19th/20th/21st) and also to increase the standard size of a legion from 4,000 to 5,000 men, and to increase the cavalry contingent from 200 to 300. Two of these new legions (18th/19th) were sent to the Po valley to cut off Hannibals Celtic reinforcements. The other 2 (20th/21st) remained in Rome to guard the city.
As both Fabius and Minucius had been reluctant to engage Hannibal with their current, somewhat raw, field army (1st/2nd/12th/13th/14th/15th legions), this levy was probably deemed necessary not only to replace all losses but also as a show of strength. Additionally it was probably also intended to increase the armies morale and to 'encourage' Romes allies.
As a result, Rome now had 16 legions (160,000 men including allies) in Italy, but over half of the manpower were new recruits, quite probably underarmed, and as yet 'untested'.
Here we have 3 points of interest. First, it is most unlikely (considering the scale of the levies over the last 12 months) that Rome had the means to thoroughly equip all these new soldiers. It is quite probable that the percentage of light troops (Velites) in the new 5,000 strength legions was greater than the standard 30% Second, it also unlikely that the cavalry attached to these legions were up to strength. Considering the time it takes to train cavalry as opposed to infantry it is reasonable to conclude they were of poor quality. Third, in order to create and train 8 new legions in the space of 12 months each legion must have been comprised of a cadre of experienced soldiers and officers. Such a cadre could only have been provided from the existing 8 legions. In other words it's fair to assume all 16 legions in Italy comprised both veteran troops and a fair percentage of the new recruits - as opposed to 8 'veteran' legions and 8 new legions.
The Senate had by now decided that Hannibal must be brought to battle and defeated and consequently the consuls Paullus and Varro were sent, together with 2 additional legions (the 16th and 17th that had been training in Rome), to take command of Geminius army and to engage Hannibal. Thus the Roman field army now consisted of 2 consular armies (8 legions plus allies). This was the 1st time in Romes history that such a huge force had been assembled and was double the size of all previous consular armies (4 legions). Polybius informs us that this huge army numbered 80,000 infantry and slightly over 6,000 cavalry.
One interesting point. No Roman had ever had experience of commanding such a huge force in the field. In fact, there was no established command structure for such a force and the Roman practice of allotting command of this army to each consul on alternate days must only have added to such command and control problems.
By August 216BC Paullus and Varro had caught up with Hannibal by the river Ofanto where Hannibals army had been encamped for some weeks. The consuls established two camps, one large one on the same side of the river (Paullus camp) as Hannibal (Northern bank) and a smaller one on the opposite bank (Varros camp). The intention was obviously to prevent Hannibal foraging.
The following day Hannibal deployed his forces on the Northern bank, where there was ample room on the plain for cavalry manouvers. The Romans refused to engage him. This is quite revealing as it infers Hannibal also had a plan for fighting on the Northern bank had the Roman army offered battle.
The following morning Paullus forded the river with most of his force (leaving 10,000 men to guard the larger camp) and joined Varro on the Southern bank. This bank was quite narrow, only 2 kilometres wide, and the 2 consuls deployed the vast Roman army between the river and the nearby hills ... upon which lay the small town of Cannae.
Seeing this Hannibal also crossed the river and deployed his troops.
The scene was now set for Romes greatest military disaster.
Carthaginian Forces
-------------------
Most ancient and contempory sources agree on the general composition of Hannibals army at Cannae, 40,000 infantry and 10,000 cavalry. Whilst many of his African and Spanish troops had been lost crossing the Alps he had been joined by a substantial number of Celts. By this time all of his Elephants had been lost.
Hannibals infantry comprised 8,000 African extra heavy infantry (Pikemen), 6,000 Spanish heavy/medium infantry and around 14,000 Celtic medium infantry. The remaining 12,000 would have been African/Spanish/Celtic skirmishers.
One interesting point. Most contemporary and ancient sources consider Spanish infantry of this period to be medium infantry. Polybius specifically refers to them, at Cannae, as heavy. Is this because they had been rearmed with the Roman armour captured at Trebia/Trasimene?
The cavalry comprised 4,000 Numidian light cavalry, 2,000 Spanish medium cavaly and 4,000 Celtic medium cavalry.
Over the last 18 months Hannibals army had won 2 major engagements and 2 minor, had accumulated a considerable amount of booty over the last year and had seen the army of Fabius reluctant to engage them in open battle. We can safely assume the morale, experience and confidence of Hannibals army would be very high. This would be a major factor at Cannae.
Roman Forces
------------
The composition of the Roman forces at Cannae deserve a closer look. Polybius and other ancient sources are in close agreement on the strength (80,000 infantry and 6,000+ cavalry). However, the composition of the 8 legions present at Cannae must have comprised a good percentage of new recruits. Livy tells us "almost two thirds were raw recruits", but how were these recruits alotted amongst the legions present.
As the levies of the previous year had increased the strength of a Roman legion to 5000 Infantry and 300 Cavalry this tells us that the old legions that had not yet seen action (12th/13th) now comprised at least 20% new recruits. Also the 1st and 2nd legions that had been decimated at the Trebia had by now been brought back up to strength. If we consider these legions to have been at half strength (2,000?) they now presumably comprised 60% new recruits. Furthermore, the remaining 4 legions were new ones comprised of levies; ie: new recruits that had had no more than 1 years training.
Even if we are to be generous and consider the likelyhood of cadres being removed from existing legions in order to form the new ones (thus giving the new legions say 1,000 experienced troops), we can still conclude that the composition of the Roman legions at Cannae would be roughly -
4 'new' legions -
14th - 5,000 Infantry, 300 cavalry (4,300 new recruits)
15th - 5,000 Infantry, 300 cavalry (4,300 new recruits)
16th - 5,000 Infantry, 300 cavalry (4,300 new recruits)
17th - 5,000 Infantry, 300 cavalry (4,300 new recruits)
4 'old' legions -
1st - 5,000 Infantry, 300 cavalry (3,300? new recruits)
2nd - 5,000 Infantry, 300 cavalry (3,300? new recruits)
12th - 5,000 Infantry, 300 cavalry (1,100 new recruits)
13th - 5,000 Infantry, 300 cavalry (1,100 new recruits)
This gives us 42,400 Roman troops of which around 26,000 were new recruits.
Hence, of 42,400 Roman troops only around 16,400 (40%) had any experience. This is consistent with Livy. If we further consider the minimum 30% Velites this leaves us with just 12,000 experienced heavy infantry - less than 30% of the total fielded. Ok, the figures may be open to question but the percentages must be close.
Polybius figures leave us with Romes allies fielding a further 40,000 infantry and 4,000 cavalry. How reliable these troops were is open to question. Considering Romes recent defeats, the inexperience of the Roman army, the presence of Hannibal's army in their own territory and the fact they knew Hannibal would release all none-Roman prisoners it is highly unlikely they had any real incentive to sacrifice their lives for Rome.
We must also remember that of this total of 86,000 around 10,000 were still on the Northern bank and had instructions to attack Hannibals camp. As it is highly unlikely Paullus would have allotted a task to just allies we can safely presume this force was 1 Roman legion together with a similar number of allies (but see the comments later). This even further reduces the number of experienced Roman troops that formed up on the Southern bank.
Therefore, 70,000 Roman infantry and 6,000 cavalry, the vast majority being either unreliable or inexperienced, faced Hannibals veteran army of 50,000.
The Deployment
--------------
The Roman army deployed as normal. Legions in the center flanked by the allies, allied cavalry on the left wing (between the legions and the hills) and Roman cavalry on the right wing (between the legions and the river).
As for the legions, Polybius states "the maniples were grouped more closely than usual and the depth was several times greater than the width." As a maniple consisted of around 120 soldiers this tells us two things. First the infantry was 'packed' closer together than normal. Considering the number of soldiers present and the maximum 2 kilometres width of the Southern bank, this sounds logical. Second, the maniples were drawn up in depth. As the standard maniple deployment was 20x6 (width/depth) this different formation must have been more like 6x20.
If we draw a rough analogy with Napoleonic formations the standard Roman practice was for the legions to adapt the "British line". At Cannae they formed up more like a "French column".
Why draw the maniples up in such depth? Maybe the reason lies in the lack in experienced troops. By adopting such a formation this ensured the front ranks comprised the veterans which was presumably necessary if the Roman tactic was to batter through the Carthaginean centre as soon as possible. The unfortunate side effect of such a formation would be the lack of experienced soldiers on either the flanks or the rear.
As for the Carthaginians. Hannibal arrayed his Spanish and Celtic infantry in the centre, interspaced in units of 100-500, and his African infantry on both flanks.The Numidian cavalry were deployed on the right wing (facing the Roman allied cavalry), Spanish/Celtic cavalry on the left wing (facing the Roman cavalry).
The Battle
----------
I have no intention of describing the battle in detail, most of you will know that. Just the salient points that many people tend to overlook.
Livy tells us that just before the battle started 500 Numidian cavalry approached the Roman lines, threw their weapons on the ground and appeared to have defected. They were sent to the Roman rear. Whenthe battle strted they picked up their weapons and attcked the romans from rear causing chaos. Whilst this would appear to be a typical Hannibal tactic I doubt this ouccred. Livy was prone to exageration in an attempt 'excuse' Roman failure. Also Polybius does not mention this, which leaves the incident suspect.
Once the battle on the Southern bank started the 10,000 Roman troops on the Northern bank then attacked Hannibals camp. It appears quite likely that this was a ruse by Paullus/Varro intended to draw some of Hannibals troops back to the camp and hence reduce his numbers on the Southern bank. It did'nt work. Hannibal had left only 2/3,000 troops (probably skirmishers) to guard the camp and he totally ignored this engagement, even though the Romans were getting the upper hand there.
Most of the Roman and allied cavaly on the wings chose to fight on foot. This is not only confirmed by Polybius but is consistent with Polybius telling us that less that 400 cavalry escaped with their lives. This preference to fight as infantry appears to confirm the poor quality and training of Romes cavalry at Cannae. Once the Roman cavalry on the wings had been defeated the Numidian cavalry pursued the routers, and probably the Velites too who (now missile less) would have retired to the rear. The Spanish and Celtic cavalry then charged the Roman legions rear.
By this time the Spanish/Celts had lost much ground to the legions 'battering ram' with the unfortunate effect that their flanks were now exposed to the fresh African infantry. Considering the depth of the Roman formation the Africans now flanked the legions frontage. Hannibal had obviously planned this from the outset.
Hannibals Africans now turned and attacked the Roman flanks. It is probable the Africans came into contact with Romes allies on the flanks. It's more than likely both Roman flanks routed quite quickly and considering the closely packed infantry the Roman legions in the centre had no opportunity to come to grips with the enemy. Panic quickly spreads in an army, and consider the lack of experience in the centre it's same to assume the whole mass of infantry collapsed and routed. However, considering the closely packed formations, both flanks routing inwards, the unusual depth of the formation and the arrival of Hannibals cavalry in the rear, they would have been falling over each other in their haste to escape. Perhaps this explains the high casualty figures at Cannae.
One interesting point is the failure of the Roman rear line, normally the veteran Triarii, to turn face and confront the cavalry attacking them. Maybe they also panicked. Peter Connelly ("Rome and Greece at War") suggest that it was the Triarii who comprised the 10,000 on the Northern bank as it was the custom for the Triarii to guard the Roman camp. Personally I find this hard to believe. It's unlikely Varro or Paullus would have detached some of their best troops for the purposes of a ruse. It appears more probable that the Triarii were actually in the front ranks as opposed to the rear. This would be consistent both with the battering ram tactic employed and the collapse of the rear when confronted by Hannibals cavalry.
Once victory had been secured on the Southern bank, Hannibal led a force across the river killing 2,000 and capturing the other 8,000 of Romans who had been attacking the Carthaginian camp. Considering the slaughter these troops had witnessed on the far bank it's unlikely they put up much of a fight.
What of the consuls? Paullus died attemping to rally the centre, together with Geminius. Varro fled/routed along with a few allied cavalry and returned to Rome.
Casualties
----------
Hannibal took around 15,000 prisoners. The allies were released and the 7,000 Roman prisoners sold to Greek slave traders (the Senate again refusing to pay a ransom for them!http://www.totalwar.org/forum/non-cgi/emoticons/wink.gif. Around 10,000 Romans escaped and these were later formed into 2 new legions and sent to Sicily as a punishment for defeat. Ancient sources put the dead at between 50,000 and 60,000 and most contemporary sources now agree with this figure. It seems likely that a number of allies also escaped which probably explains the discrepancy if you choose to accept the lower casualty figure.
Hannibal losses are quoted as being 6000; mainly Celt's.
The Aftermath
-------------
In three battles Rome had lost around 100,000 men. Another Dictator was named who promptly raised another levy of all men over 17 years of age and formed 2 more legions (22nd/23rd). Three further legions were formed. Two from 8,000 slaves and another from 6,000 debtors and criminals. After Cannae Roman manpower was stretched to it's limits.
For the next 14 years Hannibal remained in Italy, fostering the Carthaginian cause, and was never again seriously challenged by a Roman army on Italian soil.
Whilst Cannae was a devastating military blow to Rome, the Political ramifications were even worse. The whole of the Southern Appennies defected to Hannibal thus depriving Rome of 20-25% of it's traditional allied manpower and ensuring a steady flow of Italian reinforcements for Hannibals forces in Italy. Many Italian City States overthrew their Oligarchies and adapted the Democratic model of Hannibal and Carthage (Carthago had been originally founded by Greeks - hence the term Liby-Phoenitian used to describe Carthaginians).
The Question of command
-----------------------
One final question remains to be answered about Cannae. Which consul was really in command on the day?
Both Livy and Polybius state it was Varro, but considering Polybius close association with the Paulli family and the probability that Livy was merely repeating the words of Polybius, are we safe to believe this.
Peter Connelly raises a very good point when he reminds us that "the two positions of status" in a Roman field army was command of the legions and command of the Roman cavalry. At Cannae it was Paullus who commanded the Roman cavaly whilst Varro commanded the allied cavalry. The legions remained under the command of Geminius.
This tells us two things. As Geminius had been in command of the army for the last 6/9 months it was wise, considering his experience, to leave him commanding the centre - the 1st position of prestige. It was Paullus who assumed the 2nd position of prestige - not Varrus. Also, Paullus "choose to die" attempting to rally the centre. Considering he had been wounded early in the battle, but had not retired, this may suggest Paullus felt personally responsible for his failure and prefered death to dishonour.
Upon Varrus return to Rome he was welcomed by the Senate and remained in office throughout the reminder of the 2nd Punic War. He was not disgraced.
Conclusion
----------
Rome lost Cannae long before the battle was fought. Romes pride in refusing to pay Hannibal for the return of the prisoners taken at the Trebia and Lake Trasimene deprived Rome of the services of around 15,000 experienced troops. Hoping to replace such experience with numbers was foolhardy.
This left Paullus and Varro with an army that 'on paper' appeared impressive, but in reality was sadly deficient in morale, experience and cavalry. If we consider the magnitude of the levies it was probably also seriously deficient in armour and heavy infantry, and quite probably comprised a higher than normal percentage of Velites. Romes allies were also suspect. The serious losses of cavalry, for which they provided 75% of the Roman total, prior to Cannae must have hurt, and such allies would be well aware of the true nature of the Roman legions that accompanied them.
The Senates insistance that this army engage Hannibal is a classic example of politics interfering with the realities of the strategical military situation. Neither Varro nor Paullus could have devised successful tactics at Cannae with the means at their disposal.
Considering the lack of experience in the Roman army, increasing the depth of the maniples to ensure the experienced men were in the 1st few ranks, and hoping to batter through the center of the Carthaginian line using these experienced troops as a spearhead was the only sensible option.
Hannibal was obviously aware of the composition of the Roman army and his brilliance is shown by his refusal to leave a strong force to guard his camp, the ruse (if it really occured) with the 500 Numidian 'deserters', his eagerness to engage the Romans on poor terrain (whilst knowing the wind would be blowing in their faces), the maneovering of his African heavy infantry and his obvious command and control abilities that ensured his African infantry and Cavalry commanders knew the plan and adhered to it.
Hannibal had also devised his plans for an engagement on either the Northern or Southern bank long before the Romans even arrived. He KNEW he would win, regardless of the location of the forthcoming battle, regardless of the numbers involved and regardless of the Roman commander on the day.
Varro has taken much criticism from Polybius, less from Livy. Considering the magnitude of the defeat the Historians had to hold someone responsible. Then again, was Varro really in command on that fateful day, or was it Paullus? Perhaps we shall never know.
Welly
The Background - Roman legions
------------------------------
At the time of the 2nd Punic war a Roman legion consisted of 4,000 infantry and 200 cavalry. Each legion was always supplemented by allies who provided at least the same number of infantry and considerably more cavalry (3:1). Therefore a "Legion" can be considered to be around 10,000 men, 10% being cavaly and around 40-45% being Roman citizens. Many ancient sources when refering to a legion of this period use the term to include both Roman and allied components.
Romes legions comprised not only citizens of Rome but also soldiers levied from several Latin cities that had been granted Roman citizenship. Romes allies were generally other Italian City States that, whilst not being granted Citizenship, paid tribute to Rome (not always volentarily). Rome specified the numbers and composition of troops that such allies were expected to provide. Such allies may have been equipped and fought in a similar manner to the Roman legions.
A Roman legion consisted of Velites (light troops/skirmishers), Hastatii, Principes and Triarii (the Veteran troops armed with long spears instead of pilum). The proportions were roughly equal - perhaps the Velites comprised 30% of a legion. The cavalry, both Roman and allied, consisted of the weathier classes - people who could afford to keep and equip a horse.
A Roman legion generally fought with the Velites out in front and 3 lines of heavy infantry - a 1st line of Hastatii, a 2nd line of Principes and a 3rd line of Triarii. The Triarii were only ever comitted to battle if the Hatatii and Principes had been defeated.
Roman forces generally operated in pairs of legions. A standard Roman consular army of this period comprised 4 legions which, together with allies, consisted of around 40,000 men at full strength. The standard deployment of a Roman army was Roman legions in the centre, allies on either flank and the cavalry on the wings.
Hannibal arrives in Italy
-------------------------
In 218BC Hannibal crossed the Alps and was confronted by a Roman army under Scipio/Sempronius (4 legions 1st/2nd/3rd/4th). A small cavalry engagement was fought by the Tacinus river (November 218BC) that Hannibal won. Polybius tells us the Roman cavalry "lost many men".
One month later the battle of the Trebia was fought. Rome fielded 16,000 men and 20,000 allies, were obviously weak in cavalry as a result of the Tacinus engagement, and were utterely defeated. Only 10,000 men escaped. Roman dead numbered 15,000 with around 10,000 prisoners being taken. Hannibal sold the Roman prisoners into slavery and released all allied prisoners after telling them "My fight is with Rome - not her allies".
A few months later, April 217BC, a second Roman army under Flaminius, 30,000 men consisting of 4 legions (2 full strength, 10th/11th, and the remains of the 3rd/4th) and allies, was ambushed by Hannibal at Lake Trasimene. Very few escaped. Rome lost 15,000 dead and around the same number prisoners. Hannibal sold 7,000 Romans to Greek slave traders and again released the allied prisoners.
Whilst Flaminius army was being slaughtered at Lake Trasimene the consul Geminius was marching with a further 4 legions (12th/13th and the remains of the 1st/2nd) to join him. Knowing Flaminius was weak in cavalry Geminius sent all 4,000 cavalry attached to his army ahead to join Flaminius. 4 days after Lake Trasimene was fought Hannibal sent Maharbal to ambush this cavalry on the Via Flaminia. Both Polybius and Livy say all this cavalry was lost (Polybius tells us 2,000 were killed and 2,000 captured). Geminius, upon hearing this, immediately returned to Rome with his 4 legions - minus the cavalry contingent.
All Roman prisoners taken in these battles were offered back to Rome for ransom. Rome refused, and Hannibal sold them into slavery.
Within 6 months Hannibal had totally destroyed 6 Roman legions (and allies), destroyed almost all the cavalry attached to 8 Roman legions (and allies), and had undoubtedly caused questions to be asked by Romes allies.
Hannibal re-equiped his army with all the captured Roman weapons/armour, moved into Central/Southern Italy and spent the next 12 months eliciting support from discontented Italian states.
The run up to Cannae
--------------------
Following these disasters the Senate appointed Fabius as Dictator for 6 months, as was the custom in times of danger.
Fabius immediately raised 4 new legions (14th/15th/16th/17th) and leaving the 16th and 17th to train in Rome marched to Geminius with the 14th/15th. Taking command of Geminius army Fabius now had a field army of 47,000 infantry and around 2,500 cavalry. This army kept in close contact with Hannibal but only a few minor skirmishes occured. Both Fabius and Minucius (who assumed command after Fabius was recalled to Rome) were reluctant to engage Hannibal. It is quite likely the morale of the Roman army was poor, especially considering the events of the last year.
After Fabius 6 month dictorship expired the Senate appointed new Consuls and a further levy was raised in order to increase the strength of the legions and create even more new ones This new levy was used to create 4 new legions (18th/19th/20th/21st) and also to increase the standard size of a legion from 4,000 to 5,000 men, and to increase the cavalry contingent from 200 to 300. Two of these new legions (18th/19th) were sent to the Po valley to cut off Hannibals Celtic reinforcements. The other 2 (20th/21st) remained in Rome to guard the city.
As both Fabius and Minucius had been reluctant to engage Hannibal with their current, somewhat raw, field army (1st/2nd/12th/13th/14th/15th legions), this levy was probably deemed necessary not only to replace all losses but also as a show of strength. Additionally it was probably also intended to increase the armies morale and to 'encourage' Romes allies.
As a result, Rome now had 16 legions (160,000 men including allies) in Italy, but over half of the manpower were new recruits, quite probably underarmed, and as yet 'untested'.
Here we have 3 points of interest. First, it is most unlikely (considering the scale of the levies over the last 12 months) that Rome had the means to thoroughly equip all these new soldiers. It is quite probable that the percentage of light troops (Velites) in the new 5,000 strength legions was greater than the standard 30% Second, it also unlikely that the cavalry attached to these legions were up to strength. Considering the time it takes to train cavalry as opposed to infantry it is reasonable to conclude they were of poor quality. Third, in order to create and train 8 new legions in the space of 12 months each legion must have been comprised of a cadre of experienced soldiers and officers. Such a cadre could only have been provided from the existing 8 legions. In other words it's fair to assume all 16 legions in Italy comprised both veteran troops and a fair percentage of the new recruits - as opposed to 8 'veteran' legions and 8 new legions.
The Senate had by now decided that Hannibal must be brought to battle and defeated and consequently the consuls Paullus and Varro were sent, together with 2 additional legions (the 16th and 17th that had been training in Rome), to take command of Geminius army and to engage Hannibal. Thus the Roman field army now consisted of 2 consular armies (8 legions plus allies). This was the 1st time in Romes history that such a huge force had been assembled and was double the size of all previous consular armies (4 legions). Polybius informs us that this huge army numbered 80,000 infantry and slightly over 6,000 cavalry.
One interesting point. No Roman had ever had experience of commanding such a huge force in the field. In fact, there was no established command structure for such a force and the Roman practice of allotting command of this army to each consul on alternate days must only have added to such command and control problems.
By August 216BC Paullus and Varro had caught up with Hannibal by the river Ofanto where Hannibals army had been encamped for some weeks. The consuls established two camps, one large one on the same side of the river (Paullus camp) as Hannibal (Northern bank) and a smaller one on the opposite bank (Varros camp). The intention was obviously to prevent Hannibal foraging.
The following day Hannibal deployed his forces on the Northern bank, where there was ample room on the plain for cavalry manouvers. The Romans refused to engage him. This is quite revealing as it infers Hannibal also had a plan for fighting on the Northern bank had the Roman army offered battle.
The following morning Paullus forded the river with most of his force (leaving 10,000 men to guard the larger camp) and joined Varro on the Southern bank. This bank was quite narrow, only 2 kilometres wide, and the 2 consuls deployed the vast Roman army between the river and the nearby hills ... upon which lay the small town of Cannae.
Seeing this Hannibal also crossed the river and deployed his troops.
The scene was now set for Romes greatest military disaster.
Carthaginian Forces
-------------------
Most ancient and contempory sources agree on the general composition of Hannibals army at Cannae, 40,000 infantry and 10,000 cavalry. Whilst many of his African and Spanish troops had been lost crossing the Alps he had been joined by a substantial number of Celts. By this time all of his Elephants had been lost.
Hannibals infantry comprised 8,000 African extra heavy infantry (Pikemen), 6,000 Spanish heavy/medium infantry and around 14,000 Celtic medium infantry. The remaining 12,000 would have been African/Spanish/Celtic skirmishers.
One interesting point. Most contemporary and ancient sources consider Spanish infantry of this period to be medium infantry. Polybius specifically refers to them, at Cannae, as heavy. Is this because they had been rearmed with the Roman armour captured at Trebia/Trasimene?
The cavalry comprised 4,000 Numidian light cavalry, 2,000 Spanish medium cavaly and 4,000 Celtic medium cavalry.
Over the last 18 months Hannibals army had won 2 major engagements and 2 minor, had accumulated a considerable amount of booty over the last year and had seen the army of Fabius reluctant to engage them in open battle. We can safely assume the morale, experience and confidence of Hannibals army would be very high. This would be a major factor at Cannae.
Roman Forces
------------
The composition of the Roman forces at Cannae deserve a closer look. Polybius and other ancient sources are in close agreement on the strength (80,000 infantry and 6,000+ cavalry). However, the composition of the 8 legions present at Cannae must have comprised a good percentage of new recruits. Livy tells us "almost two thirds were raw recruits", but how were these recruits alotted amongst the legions present.
As the levies of the previous year had increased the strength of a Roman legion to 5000 Infantry and 300 Cavalry this tells us that the old legions that had not yet seen action (12th/13th) now comprised at least 20% new recruits. Also the 1st and 2nd legions that had been decimated at the Trebia had by now been brought back up to strength. If we consider these legions to have been at half strength (2,000?) they now presumably comprised 60% new recruits. Furthermore, the remaining 4 legions were new ones comprised of levies; ie: new recruits that had had no more than 1 years training.
Even if we are to be generous and consider the likelyhood of cadres being removed from existing legions in order to form the new ones (thus giving the new legions say 1,000 experienced troops), we can still conclude that the composition of the Roman legions at Cannae would be roughly -
4 'new' legions -
14th - 5,000 Infantry, 300 cavalry (4,300 new recruits)
15th - 5,000 Infantry, 300 cavalry (4,300 new recruits)
16th - 5,000 Infantry, 300 cavalry (4,300 new recruits)
17th - 5,000 Infantry, 300 cavalry (4,300 new recruits)
4 'old' legions -
1st - 5,000 Infantry, 300 cavalry (3,300? new recruits)
2nd - 5,000 Infantry, 300 cavalry (3,300? new recruits)
12th - 5,000 Infantry, 300 cavalry (1,100 new recruits)
13th - 5,000 Infantry, 300 cavalry (1,100 new recruits)
This gives us 42,400 Roman troops of which around 26,000 were new recruits.
Hence, of 42,400 Roman troops only around 16,400 (40%) had any experience. This is consistent with Livy. If we further consider the minimum 30% Velites this leaves us with just 12,000 experienced heavy infantry - less than 30% of the total fielded. Ok, the figures may be open to question but the percentages must be close.
Polybius figures leave us with Romes allies fielding a further 40,000 infantry and 4,000 cavalry. How reliable these troops were is open to question. Considering Romes recent defeats, the inexperience of the Roman army, the presence of Hannibal's army in their own territory and the fact they knew Hannibal would release all none-Roman prisoners it is highly unlikely they had any real incentive to sacrifice their lives for Rome.
We must also remember that of this total of 86,000 around 10,000 were still on the Northern bank and had instructions to attack Hannibals camp. As it is highly unlikely Paullus would have allotted a task to just allies we can safely presume this force was 1 Roman legion together with a similar number of allies (but see the comments later). This even further reduces the number of experienced Roman troops that formed up on the Southern bank.
Therefore, 70,000 Roman infantry and 6,000 cavalry, the vast majority being either unreliable or inexperienced, faced Hannibals veteran army of 50,000.
The Deployment
--------------
The Roman army deployed as normal. Legions in the center flanked by the allies, allied cavalry on the left wing (between the legions and the hills) and Roman cavalry on the right wing (between the legions and the river).
As for the legions, Polybius states "the maniples were grouped more closely than usual and the depth was several times greater than the width." As a maniple consisted of around 120 soldiers this tells us two things. First the infantry was 'packed' closer together than normal. Considering the number of soldiers present and the maximum 2 kilometres width of the Southern bank, this sounds logical. Second, the maniples were drawn up in depth. As the standard maniple deployment was 20x6 (width/depth) this different formation must have been more like 6x20.
If we draw a rough analogy with Napoleonic formations the standard Roman practice was for the legions to adapt the "British line". At Cannae they formed up more like a "French column".
Why draw the maniples up in such depth? Maybe the reason lies in the lack in experienced troops. By adopting such a formation this ensured the front ranks comprised the veterans which was presumably necessary if the Roman tactic was to batter through the Carthaginean centre as soon as possible. The unfortunate side effect of such a formation would be the lack of experienced soldiers on either the flanks or the rear.
As for the Carthaginians. Hannibal arrayed his Spanish and Celtic infantry in the centre, interspaced in units of 100-500, and his African infantry on both flanks.The Numidian cavalry were deployed on the right wing (facing the Roman allied cavalry), Spanish/Celtic cavalry on the left wing (facing the Roman cavalry).
The Battle
----------
I have no intention of describing the battle in detail, most of you will know that. Just the salient points that many people tend to overlook.
Livy tells us that just before the battle started 500 Numidian cavalry approached the Roman lines, threw their weapons on the ground and appeared to have defected. They were sent to the Roman rear. Whenthe battle strted they picked up their weapons and attcked the romans from rear causing chaos. Whilst this would appear to be a typical Hannibal tactic I doubt this ouccred. Livy was prone to exageration in an attempt 'excuse' Roman failure. Also Polybius does not mention this, which leaves the incident suspect.
Once the battle on the Southern bank started the 10,000 Roman troops on the Northern bank then attacked Hannibals camp. It appears quite likely that this was a ruse by Paullus/Varro intended to draw some of Hannibals troops back to the camp and hence reduce his numbers on the Southern bank. It did'nt work. Hannibal had left only 2/3,000 troops (probably skirmishers) to guard the camp and he totally ignored this engagement, even though the Romans were getting the upper hand there.
Most of the Roman and allied cavaly on the wings chose to fight on foot. This is not only confirmed by Polybius but is consistent with Polybius telling us that less that 400 cavalry escaped with their lives. This preference to fight as infantry appears to confirm the poor quality and training of Romes cavalry at Cannae. Once the Roman cavalry on the wings had been defeated the Numidian cavalry pursued the routers, and probably the Velites too who (now missile less) would have retired to the rear. The Spanish and Celtic cavalry then charged the Roman legions rear.
By this time the Spanish/Celts had lost much ground to the legions 'battering ram' with the unfortunate effect that their flanks were now exposed to the fresh African infantry. Considering the depth of the Roman formation the Africans now flanked the legions frontage. Hannibal had obviously planned this from the outset.
Hannibals Africans now turned and attacked the Roman flanks. It is probable the Africans came into contact with Romes allies on the flanks. It's more than likely both Roman flanks routed quite quickly and considering the closely packed infantry the Roman legions in the centre had no opportunity to come to grips with the enemy. Panic quickly spreads in an army, and consider the lack of experience in the centre it's same to assume the whole mass of infantry collapsed and routed. However, considering the closely packed formations, both flanks routing inwards, the unusual depth of the formation and the arrival of Hannibals cavalry in the rear, they would have been falling over each other in their haste to escape. Perhaps this explains the high casualty figures at Cannae.
One interesting point is the failure of the Roman rear line, normally the veteran Triarii, to turn face and confront the cavalry attacking them. Maybe they also panicked. Peter Connelly ("Rome and Greece at War") suggest that it was the Triarii who comprised the 10,000 on the Northern bank as it was the custom for the Triarii to guard the Roman camp. Personally I find this hard to believe. It's unlikely Varro or Paullus would have detached some of their best troops for the purposes of a ruse. It appears more probable that the Triarii were actually in the front ranks as opposed to the rear. This would be consistent both with the battering ram tactic employed and the collapse of the rear when confronted by Hannibals cavalry.
Once victory had been secured on the Southern bank, Hannibal led a force across the river killing 2,000 and capturing the other 8,000 of Romans who had been attacking the Carthaginian camp. Considering the slaughter these troops had witnessed on the far bank it's unlikely they put up much of a fight.
What of the consuls? Paullus died attemping to rally the centre, together with Geminius. Varro fled/routed along with a few allied cavalry and returned to Rome.
Casualties
----------
Hannibal took around 15,000 prisoners. The allies were released and the 7,000 Roman prisoners sold to Greek slave traders (the Senate again refusing to pay a ransom for them!http://www.totalwar.org/forum/non-cgi/emoticons/wink.gif. Around 10,000 Romans escaped and these were later formed into 2 new legions and sent to Sicily as a punishment for defeat. Ancient sources put the dead at between 50,000 and 60,000 and most contemporary sources now agree with this figure. It seems likely that a number of allies also escaped which probably explains the discrepancy if you choose to accept the lower casualty figure.
Hannibal losses are quoted as being 6000; mainly Celt's.
The Aftermath
-------------
In three battles Rome had lost around 100,000 men. Another Dictator was named who promptly raised another levy of all men over 17 years of age and formed 2 more legions (22nd/23rd). Three further legions were formed. Two from 8,000 slaves and another from 6,000 debtors and criminals. After Cannae Roman manpower was stretched to it's limits.
For the next 14 years Hannibal remained in Italy, fostering the Carthaginian cause, and was never again seriously challenged by a Roman army on Italian soil.
Whilst Cannae was a devastating military blow to Rome, the Political ramifications were even worse. The whole of the Southern Appennies defected to Hannibal thus depriving Rome of 20-25% of it's traditional allied manpower and ensuring a steady flow of Italian reinforcements for Hannibals forces in Italy. Many Italian City States overthrew their Oligarchies and adapted the Democratic model of Hannibal and Carthage (Carthago had been originally founded by Greeks - hence the term Liby-Phoenitian used to describe Carthaginians).
The Question of command
-----------------------
One final question remains to be answered about Cannae. Which consul was really in command on the day?
Both Livy and Polybius state it was Varro, but considering Polybius close association with the Paulli family and the probability that Livy was merely repeating the words of Polybius, are we safe to believe this.
Peter Connelly raises a very good point when he reminds us that "the two positions of status" in a Roman field army was command of the legions and command of the Roman cavalry. At Cannae it was Paullus who commanded the Roman cavaly whilst Varro commanded the allied cavalry. The legions remained under the command of Geminius.
This tells us two things. As Geminius had been in command of the army for the last 6/9 months it was wise, considering his experience, to leave him commanding the centre - the 1st position of prestige. It was Paullus who assumed the 2nd position of prestige - not Varrus. Also, Paullus "choose to die" attempting to rally the centre. Considering he had been wounded early in the battle, but had not retired, this may suggest Paullus felt personally responsible for his failure and prefered death to dishonour.
Upon Varrus return to Rome he was welcomed by the Senate and remained in office throughout the reminder of the 2nd Punic War. He was not disgraced.
Conclusion
----------
Rome lost Cannae long before the battle was fought. Romes pride in refusing to pay Hannibal for the return of the prisoners taken at the Trebia and Lake Trasimene deprived Rome of the services of around 15,000 experienced troops. Hoping to replace such experience with numbers was foolhardy.
This left Paullus and Varro with an army that 'on paper' appeared impressive, but in reality was sadly deficient in morale, experience and cavalry. If we consider the magnitude of the levies it was probably also seriously deficient in armour and heavy infantry, and quite probably comprised a higher than normal percentage of Velites. Romes allies were also suspect. The serious losses of cavalry, for which they provided 75% of the Roman total, prior to Cannae must have hurt, and such allies would be well aware of the true nature of the Roman legions that accompanied them.
The Senates insistance that this army engage Hannibal is a classic example of politics interfering with the realities of the strategical military situation. Neither Varro nor Paullus could have devised successful tactics at Cannae with the means at their disposal.
Considering the lack of experience in the Roman army, increasing the depth of the maniples to ensure the experienced men were in the 1st few ranks, and hoping to batter through the center of the Carthaginian line using these experienced troops as a spearhead was the only sensible option.
Hannibal was obviously aware of the composition of the Roman army and his brilliance is shown by his refusal to leave a strong force to guard his camp, the ruse (if it really occured) with the 500 Numidian 'deserters', his eagerness to engage the Romans on poor terrain (whilst knowing the wind would be blowing in their faces), the maneovering of his African heavy infantry and his obvious command and control abilities that ensured his African infantry and Cavalry commanders knew the plan and adhered to it.
Hannibal had also devised his plans for an engagement on either the Northern or Southern bank long before the Romans even arrived. He KNEW he would win, regardless of the location of the forthcoming battle, regardless of the numbers involved and regardless of the Roman commander on the day.
Varro has taken much criticism from Polybius, less from Livy. Considering the magnitude of the defeat the Historians had to hold someone responsible. Then again, was Varro really in command on that fateful day, or was it Paullus? Perhaps we shall never know.
Welly