The Fall of Carthage

by Adrian Goldsworthy

  raising Gallic allies. Another blow was struck to the Roman cause when their depot at Clastidium (modern Casteggio), where supplies of grain had been massed, was betrayed to the enemy. The captured food helped to ease Hannibal's continuing supply problems. The garrison commander, a native of Brundisium called Dasius, had turned traitor for the relatively modest sum of 400 gold coins. His troops, who were most likely Latin allies as the Romans did not normally give command of their own soldiers to a non-citizen, were treated well by Hannibal, who was eager to create a reputation for clemency. The Carthaginian army was not wholly inactive whilst it observed the Romans. Shortly after Longus' arrival Hannibal became suspicious that the Gallic tribesmen on the west bank of the Trebia, who had initially welcomed him, were now negotiating with the Romans. Perhaps the Gauls were genuinely duplicitous, but it may simply be that with their loose political structures, different leaders had approached each side. A force of 2,000 infantry and 1,000 Gallic and Numidian cavalry left the camp and began to ravage the area, gathering large amounts of plunder. Some of the tribesmen now did appeal for Roman aid and Longus sent a strong cavalry force with 1,000 velites across the river to attack the enemy raiders. The Carthaginians were scattered and encumbered with loot and the Roman attack met with rapid success, chasing the enemy in disorder back to their camp. The pickets on outpost duty outside Hannibal's camp moved out to support them and in turn the pursuing Romans were driven back. The fighting rapidly escalated as each side committed more and more troops as reinforcements, all the Roman cavalry and velites eventually becoming engaged. It was a fluid fight, spreading over a large part of the open plain and not one which either general could control. Hannibal decided that he did not want to feed more troops into the action and perhaps risk it developing into a full-scale battle which he had not planned and could do little to influence. Rallying the fugitives in person, he formed a fighting line only just outside his own camp, which was most likely on the high ground west of the river. He restrained any of the reformed troops from advancing again and the Romans refused to attack an enemy who was protected by missile fire from the camp, easily reinforced by the troops within it, and probably uphill. In this way the battle ended, with the Romans inflicting higher casualties and claiming a victory. Polybius praises Hannibal for the tight control he exercised over his men and his unwillingness to let a battle occur by chance, which Polybius believed demonstrated his wisdom as a commander.10

  Longus, depicted by our sources as of an aggressive temperament, was confirmed by this success in his desire to fight a battle as soon as possible. His wounded colleague is supposed to have argued against this, predicting the disaster which did in fact occur. We must be very cautious in accepting this. This tradition may well have been spread by the Scipionic family from the immediate aftermath of the battle onwards, but it was reinforced by Polybius who was, and is, by far the most influential source for the period. It must always be remembered that the Greek historian's close association with Scipio Aemilianus led him to a favourable depiction of the latter's ancestors in his work. In this version, Scipio is supposed to have pointed out that the Roman legions had only been raised earlier in the year and would benefit from spending the winter months training, whilst Hannibal's Gallic allies might well begin to waver in their allegiance if he remained inactive, consuming their food, but winning no victories. Polybius skilfully reinforces the wisdom of these arguments by repeating them when he explains Hannibal's desire for an immediate battle. Longus did not wish to delay, for in three months' time the new consuls would take office, and before that Scipio might recover and he would have to share the glory of his anticipated victory with his colleague. Longus' boldness was typically Roman, and it is doubtful that any other Roman magistrate would have behaved differently in the circumstances. Scipio, who had led an army half the size of the force now mustered, had been just as aggressive in his pursuit of a pitched battle on the Rhone and before Ticinus, just as elated by the success of his cavalry outside Massilia, as Longus was now. It is possible that his defeat at Ticinus, and perhaps especially his own wound, had so depressed him that Scipio now doubted the likelihood of Roman victory, and probable that he hoped to postpone the battle until he could both take an active role in it and share the credit.11

  Longus' decision to seek a battle as soon as possible reflects the deep-seated confidence in their own military prowess which pervaded all classes in Roman society and contributed so much to the formidable morale of Roman armies. The Romans enjoyed a numerical superiority over the enemy, even with the addition of their new Gallic allies, and were defending their own territory. If the Roman legionaries were still inexperienced and only partly trained, then it was also true that Hannibal's best soldiers were still weary from their arduous march to reach Italy. Refusing battle when the invader was so close was an admission of weakness and it would also be difficult to maintain both consular armies concentrated in such an advanced position throughout the winter. Defeating Hannibal decisively, so soon after his arrival, would cause his new allies to defect far more quickly than a winter of inactivity and could well end all the Carthaginian's hopes. On balance the Roman willingness to seek a battle is entirely reasonable. Equally, Hannibal's invasion needed a swift major victory if it was to gather momentum. Both commanders were surely right to risk the uncertainties of battle given the potential gains, but what distinguished the two men was that Hannibal made sure that the battle would be fought on his own terms.

  In the days before the battle Hannibal and his commanders had ridden across the plain west of the Trebia, studying the ground over which he expected the battle to be fought. (The scene was paralleled when Napoleon and his Marshals examined the Pratzen Heights a few days before the Battle of Austerlitz in AD 1805; perhaps consciously so, as the Emperor was always aware of parallels from classical history.)12 The plain west of the Trebia is wide, and as flat as any land ever is naturally, until it rises sharply at its southern and south-western edges. Hannibal located a watercourse crossing the plain and running between two steep and heavily overgrown banks, in which he decided to set an ambush under the command of his brother Mago. The day before the battle a picked force of 1,000 infantry and 1,000 cavalry, apparently mostly Numidians, was formed for this task. During the night Mago led his men secretly to the ambush position, where they concealed themselves. The watercourse presumably lay behind, but to the south of where Hannibal expected the Roman army to deploy, far enough away from their line of advance to minimize the risk of premature discovery. Polybius notes that the Romans were wary of wooded terrain, where they had often in the past been ambushed by the Gauls, but did not believe an ambush possible in an open plain.

  At dawn - Polybius tells us that the day was near the winter solstice -Hannibal sent his Numidian cavalry across the River Trebia to attack the outposts stationed outside the Romans' camps and draw them into a missile fight. The Numidians had strict orders to involve the Romans in a skirmish and then steadily withdraw, luring the enemy across the river. In the meantime, Hannibal gathered his senior officers and explained his plans, encouraging them and telling them to return and prepare their soldiers for battle. The Carthaginian troops would enter the battle well fed, and physically and mentally ready for the fight.

  Longus responded just as Hannibal had hoped, sending all his cavalry out against the Numidians, closely followed by 6,000 velites. The consul then gave orders for his entire army to muster and march out against the enemy. The Numidian light horsemen continued to skirmish, but did not become closely engaged and gradually pulled back, the Romans eagerly pursuing. The heavy infantry followed more slowly, but just as enthusiastically, almost certainly formed into three columns, each probably at least 2½ miles long. In this way they forded the River Trebia, the normally shallow waters swollen by recent rain, and processed onto the flat plain beyond it, where at a given point the columns wheeled to the right and marched along what would become the main battle line, the lead unit taking up its position at the ext
reme right. This was a laborious process, each column halting whilst the tribunes fixed the position of the next maniple in the line and deployed it from marching into battle formation. Eventually the Roman and allied heavy infantry occupied a frontage of some 2 miles in length. It was a long-drawn-out technical process, requiring much activity on the part of each legion's officers. This was especially so as the army was uncommonly large and relatively inexperienced, and the two elements composing it had had little time to practice manoeuvring with each other. Under these circumstances the open country west of the Trebia was ideal terrain on which to marshal such a Roman army, as suited to its tactics and drills as it favoured Hannibal's cavalry. It must have taken several hours for the army to march the 4 or 5 miles from their camps and then deploy into battle order. By the time they had done so, the men were tired and cold, still wet from fording the river and from the sleet which fell in sporadic showers. Most were also hungry, for they had been hurried out of camp without warning and had not had time to cook a meal. Yet Longus may well have been pleased with the day so far. His cavalry appeared to be winning another victory and his strong army was formed and ready facing the enemy. Either he would get the battle he desired, or his opponent would refuse to let the action escalate into a battle, in which case the consul could assure his men that the enemy were afraid of them and know that this would make them more confident when the battle did at last occur.

  Hannibal continued his careful preparations for the battle. When the Roman heavy infantry began crossing the river and their army was fully committed, he sent out 8,000 light infantry to support the Numidians and form a screen behind which his army could deploy. Then, and only then, his main body left the camp and advanced a mile, where they formed into a battle line. The centre was formed by a single line of close order infantry, 20,000 strong. The Gallic allies, who were probably numbered about 8,000 men, seem to have formed the centre, with the Spanish and Libyan foot on the flanks. The close order Gallic and Spanish cavalry were placed on the wings, where they were soon joined by the retiring Numidians, so that about 5,000 horsemen mustered on each flank. Hannibal also divided his elephants into two bodies and seems to have placed them with the wings of the heavy infantry, although our sources disagree and are a little confused on this point.

  Longus now recalled his cavalry and formed them onto the main line. Men and horses were tired after a fruitless morning chasing agile Numidians who never stayed to meet a charge, but fled, only to rally and return to plague the Romans with a renewed shower of javelins. The Roman army had deployed in its standard formation, with the legions in the centre, the allies on their flanks and the cavalry on the wings. Longus' four legions mustered 16,000 men according to Polybius, although Livy gives the higher figure of 18,000. In both cases they have clearly assumed an average size for each unit and multiplied this by four, so that at best this provides a rough guide to the actual number of troops. If one of the legions was the one which had been commanded by the praetor Vulso and ambushed by the Boii earlier in the year, then it is likely still to have been heavily under strength. There were 20,000 allied infantry including, according to Livy, a contingent of the last Gallic tribe to remain loyal, the Cenomani. It is not clear whether this total of 38,000 infantry includes the skirmishers of the legions and alae or only the heavy infantry. It is usually assumed that the 6,000 light infantrymen mentioned earlier composed the entire skirmishing element of Longus' army, but our sources are not clear on this. Whether or not the skirmishers were included in the total, it is clear that the Romans enjoyed a significant numerical superiority in infantry. However, Longus had only 4,000 horsemen to divide between his two wings, less than half the number deployed by the enemy.

  Longus remained full of confidence and advanced his whole line, carrying this out in the proper manner of a Roman general, so that the army came on slowly and in good order. Hannibal seems to have remained where he was and let them come on, perhaps wanting to ensure that the enemy moved ahead of Mago's concealed troops. Soon the skirmishers of both armies met in front of the main lines and began to exchange missiles. The Romans fared badly in this encounter, for they were tired and had used up many of their javelins whilst supporting the cavalry against the Numidians. They may also have been outnumbered if there were only 6,000 of them, and were neither as well trained nor as experienced as their opponents. Hannibal's men included the famous Balearic slingers, and the combination of their range with the shorter-distance thrown spears of the loncbophoroi or javelinmen probably gave the Punic skirmishers greater flexibility. As the main lines closed, the skirmishers retired through the intervals between the units of close order foot, although according to Livy it was only the advance of the hastati which forced the Carthaginian light infantry to pull back, after which they moved to the flanks to support their cavalry.

  The Roman cavalry, weary and outnumbered, seems to have put up a feeble resistance as Hannibal at last ordered his horsemen forward. Livy claims that they were further weakened by sniping from the Balearic slingers now supporting the Punic horse, and also that the elephants frightened the Roman horses, but his account of the elephants' role in the battle is hopelessly confused. As the Roman cavalry broke, the Numidians and light infantry surged forward to lap around the flanks of the main Roman line, shooting at the allied soldiers. Polybius notes that they outstripped the close order cavalry, which may again be an indication that the mounts of these troops were not in good condition, although in this case they had begun the battle fresher than the Romans'. In spite of this support, the clash between the two main lines of close order foot was long and hard, its outcome uncertain. The hastati and principes alone outnumbered the enemy foot by a large margin and were also more heavily armoured. It must also be remembered that the Roman line was long and that it took a while for a reverse on the wings to affect the centre. Even when Mago's force emerged from ambush and attacked the rear of the Roman army, spreading confusion throughout the whole army, the legionaries maintained the struggle. Beset by the elephants and Punic foot to their front and skirmishers and Numidians to the rear, the wings of the Roman infantry eventually gave way. In the centre the legions managed to rout the Gauls and a unit of Libyans facing them, breaking right through the enemy line.

  Hannibal had no reserves with which to oppose this breakthrough, for his infantry had formed in a single line. Fortunately, by this time it was clear that the Roman defeat was irredeemable, with the army degenerating into a mob of fugitives. The 10,000 legionaries who had broken through the enemy line made no attempt to rejoin the fighting, but keeping in formation, they marched north, swinging round the Punic army and recrossing the Trebia opposite Placentia, where they took refuge. Hannibal made no attempt to stop them. His men were weary and his victory was already clear. The rest of the Roman army suffered heavily in the rout, but numbers of soldiers made their way as individuals or small groups back to the camps or joined the force in the colony. We do not have a figure for the Roman losses, but these must have been heavy. Our sources are similarly vague for Hannibal's casualties, although Polybius tells us that the heaviest losses were suffered by the Celts in the centre. However, in the cold spell of weather which followed the battle, many of his men and horses and all but one of the elephants died.13

  Longus at first attempted to portray the battle as an indecisive fight, in which he was deprived of victory only by the extreme weather, and it was some while before the Senate appreciated the scale of the disaster. The blow to Roman pride was probably more serious than the actual losses, for the victory persuaded those Gauls who had been wavering to embrace the Punic cause. Even so, the defeat was put down to Longus' mistakes, whilst the success of the Roman infantry in the centre seemed to confirm that the courage of their soldiers had not failed.

  Hannibal's victory gave his campaign sufficient momentum to carry it through the months of virtual inactivity forced upon him by the winter weather. His soldiers now knew that they could beat their enemy in the open field, furthe
r increasing their faith in their commander. As a general Hannibal had consistentiy outperformed both of his Roman opponents, controlling his soldiers so tightiy that a battle was only fought at a time and place of his own choice. He had been able to exploit the advantage given by his numerical superiority in cavalry, added to the flexibility derived from the mixture of light and heavy horse. At the Trebia his army had fought as a co-ordinated unit, focusing all its strength on the Roman wings. To further the anticipated success of his cavalry over the outnumbered Roman horsemen, the best of his infantry, the Libyans and Spanish, were placed on the flanks of the infantry centre, their attack given even more power by the support of the elephants. Mago's ambush had added to the confusion in the Roman ranks and probably reduced much of the forward impetus of their assault, in particular by involving the third line in combat, but the battle had already effectively been won by the success on the wings. The escape as a formed body of such a large part of the Roman infantry was regrettable, but once the Romans broke through the Gallic infantry there was little that Hannibal could do to stop them. Nevertheless the Roman eagerness to escape from the battlefield rather than renew the fight demonstrated that they had admitted defeat.

  The Battle of Lake Trasimene, c. 21 June 217 BC

  The Senate was shocked by the defeat, but began the new year grimly determined to prosecute the war with greater success. The other theatres were not ignored, but the main focus of Roman effort was to be against the enemy on their own soil and both consuls would go north against Hannibal. An air of normality was provided when Longus returned briefly to Rome to preside over the consular elections, which were won by Cnaeus Servilius Geminus and Caius Flaminius. We do not have a detailed breakdown of the citizens and allies levied in this year, but Geminus and Flaminius both seem to have been given the standard consular army of two legions and two alae, composed of a mixture of newly raised troops and the remnants of the armies defeated at Trebia. The legions may have had a stronger than usual complement and it is also possible that the armies contained a very high proportion of cavalry, perhaps as a reaction to Hannibal's superiority in this arm. Geminus' army is said to have included at least 4,000 horsemen, which was a very high proportion for a Roman army and probably consisted in the main of allies.14