The army in battle

The manipular legion was designed for fighting pitched battles. Its organization allowed it only one formation, the triplex acies with the three lines of heavy infantry supporting each other to place maximum pressure on an enemy to the front. When a Roman army was close to the enemy, the legions marched in three parallel columns, the hastati on the left, the principes in the centre and the triarii on the right. To deploy into the battle formation these columns wheeled to the right to form the triplex acies. Each maniple had to be positioned carefully in relation to its neighbours in its own and the other two lines to ensure that the legion’s front was properly and uniformly supported. Even when the army had camped only a few kilometres from the enemy it still formed itself into three columns and marched to within one and a half kilometres or less of the enemy position, and then, at the point which would form the left of the army’s position, wheeled to the right and marched along the army’s intended front to form the triplex acies. It was a time-consuming process, even in an experienced and well- drilled army, the whole column having to stop and wait as each maniple reached its appointed position and closed up from marching formation into battle formation before it could move forward again. Deploying a Roman army took hours, and required constant supervision from the tribunes. If the enemy threatened then the army’s deployment was covered by the cavalry, perhaps supported by some of the velites or extraordinarii. More often than not the enemy was too busy forming his own battle line to pose much of a threat. Hellenistic and Carthaginian armies used a similar processional system of deployment to the Romans, forming the army into a column with each unit in the order it would take in the battle line. Since they normally massed their infantry in a single deeper line they tended to use a single column rather than the Romans’ three, but the process was equally laborious.

The marching camp assisted in this process. As important as the formal and fixed positioning of the tent lines were the spaces between them containing paths and roadways. The army used these and the intervallum, the wide-open strip of land separating the tents from the ramparts, to form up into the columns which it used to deploy, the position of the tent lines automatically placing the maniples in the right order. Each of the three columns left via one of the four gates, the cavalry often using the remaining exit. If the army had deployed into battle order but not fought then it was able to retire to the camp in order, each of the three lines of the army, beginning with the triarii in the rear, forming a column and marching back to its tents.

Hastatus or Princeps

This reconstruction shows a legionary from the first or second line of a legion in the third or second century вс. He carries one of his two pila and has a gladius at his belt. A poor man, he wears a bronze pectoral instead of a mail shirt, a bronze, montefortino helmet, and carries an oval, semi-cylindrical body shield.


This veteran soldier from the third line of the third- century вс legion replaces the pilum with a long spear. He wears a mail shirt made from linked rings, a type of armour which the Romans probably copied from the Gauls. It was flexible, but heavy, and he wears a belt to relieve some of the burden which otherwise rested entirely on his shoulders.


The light infantry of the third- to second-century вс legion were recruited from the poorer citizens and those too young to join the hastati. This man carries a bundle of light javelins, a round shield and has a gladius at his belt. Velites wore pieces of animal skins, usually wolfskins, over their helmets, allowing their officers to recognize them.

The chequerboard formation used by the Roman legions has often been misunderstood and few scholars have been willing to believe that the maniples actually fought with such wide gaps between them. If they had done so then would not a charging enemy, especially a mob of howling Gauls, have swept through the gaps in the line of hastati, outflanking each of the maniples and routing them in an instant? It has conventionally been assumed that, while the Romans may have advanced with the maniples in the quincunx, they formed an unbroken line before they reached the enemy. Advancing with intervals between an army’s units is more obviously intelligible. Even on a perfectly flat parade ground it is very difficult for a unit to march in a perfectly straight line and any unevenness will drastically increase the likelihood of veering to one side or the other. If there are not significant gaps between units advancing on a parallel course then they run the risk of colliding with each other, disordering both and making it difficult for their officers to control them.


This man wears an iron, Imperial Gallic helmet and the famous lorica segmentata banded armour of soft, untempered iron plates designed to absorb a blow. Not all legionaries during the Principate were issued with this cuirass; harder to make and maintain than mail, it gave excellent protection, was flexible and slightly lighter.

Auxiliary Infantryman, FIRST CENTURY AD

This man wears what appears to have been the most common uniform for these units. He has a bronze helmet, probably a simpler version of the current legionary style, a mail cuirass, gladius, flat oval shield and wields a javelin. Such men fought in close order in much the same way as the legionaries.

Late Roman Infantryman

This fourth-century soldier wears an Intercisa pattern helmet of much cruder manufacture than his earlier predecessors. Scale or mail armour was worn where available. The spear was sometimes supplemented by darts fplumbataej fastened in the hollow of the shield. The longer spatha was now worn on the left hip.

Auxiliary cavalryman, FIRST TO SECOND CENTURY AD

The non-citizen cavalry regiments or alae of the Principate were amongst the most prestigious units in the army. Highly trained and disciplined, they provided the Imperial army with the effective cavalry force which its Republican predecessor had often lacked. The fourhorned saddle can be seen here. As the rider sat down, his weight caused the horns to close around and grip his thighs, allowing him to lean to either side without losing his seat. Cavalrymen wore bronze or iron helmets, often completely covering their ears to protect against blows from the side in a confused melee. They usually wore scale or mail armour and carried oval or hexagonal shields.

None of our ancient sources imply that the quincunx was only intended for approaching the enemy, and that the gaps between the maniples were filled just before contact. In fact for the Romans to have halted and changed formation just in front of the enemy would seem a dangerous practice. It is often forgotten that all of Rome’s opponents must have had gaps in their own lines to allow these to move. We are told explicitly that at Trebia and Magnesia the Carthaginian and Seleucid light infantry respectively were able to advance past, fight a skirmish in front of, and then retire through the gaps in their own main infantry line. Polybius tells us that Pyrrhus mixed clearly distinct units of his own pikemen and Tarentine infantry in the lines of the main phalanx, while at Cannae Hannibal’s front line alternated units of Spanish and Gallic infantry. The size of these units is unclear, but Polybius employs one of the words he also uses for maniple, and which was applied by later authors to the cohort of the Principate. This makes it likely that we are dealing with detachments of several hundred men, and almost certainly less than a thousand. The smallest independent unit in the Hellenistic military manuals was the syntagma of 256 men, which possessed its own standard, commander and musician. At Magnesia Antiochus left intervals each holding two elephants between the ten 1,000-strong, 32-rank deep blocks of his phalanx. Our sources do not suggest that the Romans uniquely formed a battle line with intervals in it, but they do imply that the gaps in a Roman line were wider than was normal. One reason why the quincunx did not risk disaster was that the maniples of the line behind covered the intervals in front. An enemy passing through the gaps between the hastati risked attack from the principes. However, the whole system becomes much more intelligible if it is considered in the light of the nature of heavy infantry combat.

Large-scale hand-to-hand combat has not been a common feature of warfare in recent centuries and it is something that is very hard for us to imagine. All too often our mental picture of a clash between sword- and spear-armed lines owes more to cinematic portrayals of spectacular battle scenes in which all the participants rush around fighting frenzied individual duels. In reality, combat seems to have been a lot more tentative. Battles usually lasted for several hours, and the battle of Pydna in 168 вс, which lasted only an hour, was considered an exceptionally brief affair. For much of this time the main lines of heavy infantry were in contact. Hand-to-hand fighting was physically very fatiguing and emotionally stressful. Actual hand-to-hand fighting can only have lasted for very short periods, and the relatively light casualties suffered in this stage of the fighting seem to support this.

Polybius’ Roman legionaries advanced noisily, yelling their war cries and banging their weapons against their shields. Most other infantry advanced in a similar fashion, the aim being to intimidate the enemy by looking and sounding threatening. Ideally, this persuaded the enemy that he had no chance of success and put him to flight long before the two lines met. Somewhere within 30 metres of the enemy, the Romans threw their pila, and the barrage of heavy missiles inflicted casualties and further reduced the enemy’s confidence. The Roman shield had a horizontal hand-grip and it was impossible for a man to use it in combat and at the same time grip a pilum in his left hand. Therefore, each Roman legionary must have thrown both of his pila before he reached the enemy. The range of these missiles was so short that a man did not have time to throw two pila and then draw his sword if he was advancing into combat. This means that either many pila were thrown outside effective range, or that the Romans usually slowed their advance or halted within 30 metres of the enemy. Such a pause in the advance is not as implausible as it may at first seem. The aim of the advance was to intimidate the enemy into an early flight. If neither side managed to gain a significant moral advantage over the other, then each may have lost confidence and been reluctant to close immediately with the enemy.

Whether or not there was an initial check, most armies with a cultural tradition of close combat seem normally to have resumed the advance and reached the enemy in the first wave of the battle. In the resulting combat the

Montefortino helmet

One of the commonest Republican designs, this bronze helmet offered good protection from a downward cut. The cheek pieces protected the face without obscuring the soldier’s eyes or ears. A Roman soldier needed good vision and unimpaired hearing if he was to function as part of a disciplined unit.

Coolus helmet

This bronze helmet, popular during the first century вс, was used well into the first century ad. The reinforcing peak gave greater strength to the helmet front, while the neckguard gave some defence against glancing blows hitting the shoulders. Roman mail cuirasses were reinforced at the shoulders to provide further protection.

Imperial Gallic helmet

These well-made iron helmets had deeper and wider neck- guards than earlier ones, that gave more protection against downward cuts. Later this pattern was developed to reinforce the helmet top with two iron ridges crossing over the centre of the bowl, a pattern seen clearly on the reliefs on Trajan’s Column and the Adamklissi Metopes.

Intercisa helmet

A common design in the later third and fourth centuries, the bowls of these crudely made helmets were composed of several pieces — in this case two halves joined in the centre. Although such designs offered poorer protection to the wearer, most of the Romans’ opponents in the west had no helmets at all.

opposing front ranks hacked, thrust and stabbed at each other, the Romans punching at the enemy with their shield bosses. When an enemy was knocked down or killed there was a chance to step into his place and attempt to fight a way into the enemy formation. More often than not the man attempting this was himself killed, but if a few men managed to work their way into the enemy ranks and break their formation then there was a real chance that the enemy would panic and flee. If neither side achieved this after a few minutes of fighting then the opposing lines drew back. Separated by perhaps as little as a few metres, the two lines then jeered and glared at each other, throwing any remaining missiles, as each attempted to build up enough energy and confidence to step forward and renew the struggle. The longer the battle went on the harder it became each time to persuade the line to close once more. Officers played a vital role in urging on their men to sustain this effort. Centurions were elected from those with a record for gallantry and the Romans took great care to praise and reward the soldiers who displayed individual boldness. Compulsion and fear of punishment also had a part to play in giving a unit the stamina to stay close to the enemy. The men in the front rank, the ones who actually fought and were in greatest danger, had to stay there as long as those in the ranks behind stayed in position, since the latter’s physical presence made escape impossible. A deeper formation gave a unit greater staying power in combat by making it hard for most of the men to flee. So did the presence of optiones, the centurions’ senior subordinates, behind the rear rank, physically pushing the men back into place. The longer a unit was close to the enemy the more its formation and cohesion dissolved. Men increasingly followed their instincts, the bravest pushing to the front, the most timid trying to slip away to the rear, while the majority remained somewhere in the middle. At any time they might follow the example of the timid and the unit dissolve into rout, and this possibility became greater the longer a unit did not advance or seem to be making progress. Most casualties on an ancient battlefield occurred when a unit fled from combat. The ones who died first were those who were slowest in turning to flee, so the men in the centre of a formation, able to see little of what was going on, were always on the verge of nervous panic.

Most armies deployed with a single main line, which as a result tended to be deep and have great staying power in combat. In the Roman triplex acies more than half of the infantry were kept in reserve and were not involved in the initial combat. Instead, the rear two lines advanced in turn to join the existing combat at later stages of the battle. Ideally, the bastati fought the main enemy line to a standstill, their discipline and the leadership of their centurions keeping them in contact with the enemy, who was probably more numerous and in a deeper formation. Then the principes advanced into the combat zone, their freshness and enthusiasm urging the whole Roman line to advance with a confidence which the enemy by that time might not be able to match. The skill of a Roman commander lay in committing his second and third lines at the right time. If it was left too long then the bastati might buckle under enemy pressure. Too soon and the value of adding a fresh contingent of troops to a combat might be lost. It was exceptionally rare for the Romans to withdraw an entire line and replace it with one from behind. Usually the troops in the rear lines were fed into the combat to support the troops already engaged.

The reconstruction of the trireme (‘three’) and its sea trials greatly increased our understanding of ancient warships. The trireme proved capable of making 8 knots in short hursts and could maintain a steady 4 knots for hours on end, despite the use of modern rowers unused to the ancient techniques and oars that were too heavy. Under sail it also reached speeds of 8 knots, but wind power was too uncertain to he used in battle, since ancient warships needed to manoeuvre quickly to close with the enemy to ram or board. By the time of the hirst Punic War, triremes had been replaced by quinqueremes ('fives') as the standard type of warship.

In this context the triplex acies offered a more effective use of an army’s numbers. The intervals between maniples were necessary to allow fresh troops to join the struggle. When combat between lines was so tentative there was little danger that an enemy would stream through the gaps and swamp the whole line. At Pydna the Macedonian phalanx advanced too far and too fast and began to break up into its constituent units. Eventually the Romans were able to exploit this, individual centurions leading parties of men to infiltrate the phalanx and attack the helpless pikemen from the flanks. This only happened after the two sides had charged each other and fought a long hand-to-hand combat. Even then it only happened through the bold leadership of individuals. The ancient battlefield was a far more open place than is often imagined.

If you find an error please notify us in the comments. Thank you!