Ancient History & Civilisation


As on many military tombstones, this man, Marcus Julius Sabinianus, is depicted in undress uniform, without helmet or cuirass, but carrying shield and weapons. He was a sailor in the Italian fleet stationed at Misenum.

Legionary Equipment


The tunic: In the Greco-Roman world the principal garment of a male civilian was a short-sleeved, knee-length tunic, since trousers were considered a barbarian fashion. In the old militia army, soldiers provided their own clothes as well as their own armour, making it unlikely that the legions of this period displayed much uniformity in dress or equipment. Gradually, as the army became a professional force, the state began to issue clothing, armour and weapons to the troops, and by the early Principate regular deductions were made from a soldier’s salary to cover the cost of each of these items. A document survives from Egypt which mentions the army ordering large consignments of clothing from civilian suppliers and insisting on a standard design and quality. A version of the civilian tunic remained the normal dress of the Roman soldier until the beginning of the 3rd century ad.

The military tunic was somewhat longer than the type normally worn by civilians, stretching halfway down a man’s calves. Normally, however, it was gathered up by a belt so that it hung above the knee. There seems to have been a fashion at some periods for gathering the sides of the tunic higher than the front and rear, so that its bottom edge forms a curve. The tunic’s design was simple, consisting essentially of two matching squares of material - usually wool or linen - sewn together at the sides and shoulders, whilst leaving openings for the arms and neck. Some tunics had sleeves, usually fairly short, although some depictions of cavalrymen suggest that their tunics were long-sleeved. At least some infantry tunics could be worn with the right shoulder and arm left bare, which appears to have been more comfortable for heavy labour. Tunics of this sort had a slit down from the neck opening at the centre of the back. Normally this was tied into a knot by a leather thong, or occasionally perhaps held by a brooch, and only unfastened when the soldier wanted to free his right arm. Modern re-enactors have found that this knot can be uncomfortable if this type of tunic is worn with armour. Evidence is limited, but it is possible that each soldier possessed more than one pattern of tunic, different patterns being worn for specific activities.

Cloaks and capes: There were two basic patterns of cloak worn by ordinary soldiers. The first was the sagum, a simple rectangle of heavy wool, although sometimes its fringe was decorated. The two sides of material were held together by a brooch on the right shoulder in a way that left a man’s right side, and his sword-arm, free. The alternative to the sagutn was the paenula, worn more like a poncho.

In this scene from Trajan's Column, the Emperor and a group of senior officers are shown on campaign, with cloaks over their ornate cuirasses. The cloak was the sagum, a simple rectangle of material fastened at one shoulder with a brooch. These could be worn in various styles.

It was probably oval in shape, with an opening for the head, and sometimes an attached hood. The front of the cape fastened with a row of buttons and toggles, so that it could be left partially open. The neck opening tended to be loose-fitting, so that a scarf was essential in cold or wet weather. Senior officers, including the Emperor Trajan, are sometimes depicted wearing a version of the sagum, although it seems more than likely that such garments were of superior quality to the ones worn by the ordinary soldiers. On other occasions, officers from the centurionate upwards are shown wearing the more formal paludamentum, which could be worn draped over the left arm, rather like the formal civilian toga.

Belts: The military tunic required a belt to be worn properly, so much so that Augustus was known to have symbolically punished centurions by making them stand at attention outside his tent without wearing their belts. A soldier’s belt, from which his sword and - at least at some periods - dagger were suspended, was also an important symbol of his identity. Even in undress uniform, without armour or helmet, the belted tunic marked a man out as a soldier. This was reflected in the ornate plates and buckles which covered the functional leather belts. In the early 1st century at least two belts were normally worn, one for the sword, whose scabbard was fastened to it by cords from the four scabbard rings, and the other for the dagger whose scabbard hooked onto a frog on one of the belt plates. The two belts were worn in a criss-cross pattern, rather like the double holster of the gunfighters depicted in many Western movies. By the end of the 1st century, fashion changed and it was more common to wear a single, broader belt supporting both sword and dagger. Even so, some men appear to have continued to wear multiple belts and one soldier on Trajan’s Column is shown with no less than four.

During the 1st and 2nd centuries AD, an apron was often attached to a soldier’s belt. This might consist of from one to nine straps, usually studded and with decorative metal terminals. Four to six such straps are most common. These may have served a defensive function, even if the protection offered to a soldier’s groin was more psychological than real. However, the experience of modern re-enactors has suggested that they could be a hindrance to a running man if allowed to swing freely. The aprons certainly made the soldier’s belt even more decorative and also jingled as he moved, adding to a soldier’s presence.

Boots: Apart from the tinkling of apron fittings and the noise produced by any metal armour, the sole of the military boot (caliga) was heavily studded. Josephus tells us that the centurion Julianus was pursuing the enemy across the Temple Court in Jerusalem in AD 70 when the metal studs of his boots skidded on the flagstones and he fell, only to be surrounded and killed by the enemy. This was a possible danger, but in most circumstances and on less smooth surfaces, the studs gave the boots better grip. The studs tended to wear out and needed to be replaced. On one occasion Vespasian received a demand from sailors of the Italian fleet for a larger allowance of boot-money (caldarium) because of the wear suffered on frequent long marches from Puteoli or Ostia to Rome. In reply the Emperor ordered them to march barefoot, a practice which appears to have continued at least to the end of the century.

The tombstone, front Mainz, of Publius Flavoleius Cordus, a soldier in Legio XIV Gemina. The subject is once again shown wearing only a tunic. It clearly shows his sword on his right side and dagger on his left, while he holds a pilum in his right hand and has an oval shield on his back. He holds a scroll in his left hand which may indicate that he held a clerical post. Having served for 23 years, Cordus was not too far away from discharge when he died aged 43, in the early 1st century AD.

Although the open appearance of caligae makes them look rather like sandals, their construction was considerably sturdier. Made in three parts - a sole, insole and upper - the straps could be tightened to fit more closely. The apparent lack of protection against the weather is also deceptive, since it seems clear that caligae were normally worn with socks, and one monument depicting praetorian guardsmen actually shows an open-toed and open-heeled sock being worn on parade. During the 2nd century ad, other types of footwear with more enclosed uppers appear to have become increasingly common, and may well have supplanted the traditional pattern of caligae altogether. Most, if not all, footwear worn by the army continued to have hobnailed soles.

Other clothes: One of the most famous letters from Vindolanda records a soldier being sent some ‘pairs of socks (udones) from Sattua, two pairs of sandals/slippers (soleae) and two pairs of underpants (subligares)...'. This is one example of the ample evidence which is gradually eroding the persistent myth that Roman soldiers spent their time in northern Britain wearing garb more suited for the Mediterranean. Instead, troops adapted to the local climate, however extreme. Breeches and, especially amongst cavalry, longer trousers were worn beneath the tunic, as were socks and perhaps leggings of various sorts.

Later changes: Both civilian and military fashion began to change again in the 3rd century ad. From this period a long-sleeved tunic became normal. Depictions in wall-paintings and mosaics suggest that these often had a different coloured border and sometimes round or lozenge-shaped decorative patches. Sleeves were usually fairly tight to the wrist. Examples from Dura Europus suggest that these woollen garments were woven in one piece, with a slit for the neck and sometimes also on each hip. Trousers, usually tight fitting, also became increasingly common. In many respects these styles endured until the collapse of the Western Empire and for much longer in the East.

The problem of colour

Fabrics such as wool, linen and leather survive only under exceptional circumstances in the archaeological record. Much of our evidence for military costume comes from the depictions of soldiers on great monuments such as Trajan’s Column and on the funerary memorials left by soldiers. Originally these reliefs were brightly painted, a colour often being used to indicate detail which was difficult to carve, such as mail armour or socks, but these pigments have not survived in any useful way. Some Roman sites have produced colourful mosaics or wall-paintings but in the main these have rarely featured extensive scenes involving soldiers in the uniforms of the day. Our literary sources rarely talk in much detail about clothing or equipment either, and on the rare occasions when colour is mentioned it is hard to be too specific about shade. Therefore, although we can with some confidence reconstruct most aspects of the appearance of Roman soldiers, we can be far less certain when it comes to the fundamental question of colour. Reconstructions of military scenes, whether by artists or re-enactment groups, inevitably involve a fair degree of conjecture in this respect.

The tombstone of an optio from Chester shows him wearing a paenula cloak and holding his hastile staff. The paint is modem, although traces of paint were found on this tombstone and it is probable that most were originally painted. The inscription reads: 'To the spirits of the departed: Caecilius Avitus, from Emerita Augusta [modem day Merida in Spain], optio of Legio XX Valeria Victrix, of 15 years service, died aged 34; his heir had this erected ’ The texts of most military tombstones are extremely formulaic.

It is often assumed that all soldiers wore tunics of a standard colour. In fact there is no positive evidence for this and it is possible that different colours as well as different patterns of tunic were employed for different orders of dress. It is equally possible - though there is no positive evidence for this - that at some periods some units wore distinctively coloured tunics, either through choice or because of local availability. There is some direct evidence for soldiers wearing white, or off-white tunics, and a little, though far less, for red. A papyrus from Egypt records a unit ordering pure white tunics for its use, and white is the most common colour for soldiers appearing in paintings or on mosaics. Undyed tunics, which would therefore be anything from white to light grey to light brown in shade, were probably also the most common form of civilian dress. Soldiers may well have worn a purer white woollen tunic than could normally be afforded by poor civilians. Tacitus’ description of the triumphal march of an army into Rome during the civil war following the death of Nero speaks of camp prefects, tribunes and senior centurions in dazzling white uniforms, which suggests that the higher ranks had brighter, better-quality tunics than ordinary soldiers. This is similar to the system in the 19th-century British army, when the soldier’s redcoat was a duller shade than that worn by a sergeant, which in turn was less bright than, and a different cut from, an officer’s jacket. An alternative suggestion is that centurions were marked out from ordinary soldiers by wearing red as opposed to white tunics. Although this is possible, it does seem to go against the evidence of Tacitus.

Cloaks are normally shown in paintings as a dull yellowish brown. They were essentially practical items, intended to guard against the wind and weather, so such an unimpressive colour is less surprising. However, several funerary portraits from Egypt dating most probably to the 2nd century ad show bearded men wearing sword belts. These men may well have been veterans or, since the quality of their burial suggests a degree of wealth, former officers. All wear white tunics, but cloaks range in colour from dark blue to dark olive green. It is therefore possible that individuals or particular ranks may have had cloaks in differing shades and that the apparent uniformity of yellow-brown is deceptive. It is also known that Roman generals normally wore a red cloak as a mark of rank, and in the 1st century BC Crassus caused a stir when he appeared wearing a black cloak, since the colour was considered to be unlucky.

Legionary helmets

The study of Roman helmets, as with many other types of military equipment, has often been hindered by emphasis on external fittings which have more to do with decoration than either design, manufacture or function. Most British scholars follow Russell Robinson’s categorization of helmets according to type and pattern, whereas those on the continent employ a completely different labelling system associated with find spots. Therefore the same helmet would be either an Imperial Gallic Type A according to Robinson's system or a helmet of the Weisenau/Nijmegen type under the continental system. However, in most respects, there is broad agreement between the two schools over the development of Roman infantry helmets.

In the last century of the Republic the most commonly used helmets were all derived originally from Gallic designs. At first, the most frequently used pattern was the Montefortino which had been in use since at least the 3rd century BC. Over time the neck- guard on this pattern gradually became larger. The Coolus helmet was very similar in appearance and by the end of the 1st century вс had supplanted the other type in popularity. Most, though by no means all, Coolus helmets had a crest knob or spike on the top of the bowl. They also tended to have wider neckguards than Montefortino helmets and from the middle of the 1st century вс most had a reinforcing peak mid way up the front of the bowl. By this time both patterns of helmets had acquired broader cheek-pieces offering more protection for the face. These seem to have developed from patterns of Gallic iron helmets, such as the Agen type.

The Soldier of the 1st Century ad

This scene shows two legionaries from Legio XIV Gemina on campaign in Britain in ad 60 during the suppression of Boudicca s rebellion. This legion played a distinguished role in Rome’s eventual victory and was awarded the titles Martia Victrix by a delighted Emperor Nero. Both men wear segmented cuirasses (lorica segmentata) and helmets of the type known as Imperial Gallic. The conditions of active

service are likely to have resulted in a somewhat different appearance from the classic image of Roman soldiers. These men have slightly ragged and patched tunics, and have adapted their uniform to make it more suitable to a north European climate. Some of the details shown remain conjectural, but all are based on interpretations of the surviving evidence.

Both the Montefortino and Coolus helmets, as well as the later Imperial types, had the helmet bowl and the neckguard made from the same piece of metal. There were two principal methods of manufacture, forging, which involved beating the metal into shape over a former, and spinning, where the metal was shaped using a revolving former (a shaped piece of wood or stone). Iron helmets could only be beaten into shape, because the iron available to the Romans was not sufficiently pure to undergo the spinning process without cracking. However, the alloy - a mixture mostly of copper with about a quarter zinc - used in bronze helmets could be spun, and this production method became common for Montefortino and Coolus helmets in the 1st century BC. The metal in the bowl of a spun helmet was not as well hardened as in a forged helmet and proved brittle, surviving examples often showing damage to this area. It may be that the addition of a peak to the Coolus pattern was at first an attempt to counteract this weakness, although it clearly also conformed with the longer-term trends in Roman helmet design.

This example of a late-pattern (1 st century вс) Montefortino helmet has wider cheek pieces and a broader neckguard than earlier types. It offered reasonable protection from a blow to the top of the head.

In the Imperial Gallic helmet (middle to late 1 st century ad), the trends evident in the Montefortino and Coolus helmets are taken further. The neckguard is not only broader, but it is now lower, ribbed for greater strength and angled to deflect a blade.

This iron helmet found at Heddernheim in Germany was probably used by a cavalryman, but shows many of the same trends apparent in infantry helmets.

The Coolus-pattern helmet (early 1 st century ad) was in many ways similar to the Montefortino, but tended to have even wider cheek pieces and neckguard. It also added a reinforcing peak to the front of the helmet to ward off an attack from this direction.

The Imperial Italic helmet (early 2nd century ad) differed only slightly from the Gallic types in shape, although they are usually less ornate. Many later examples have reinforcing cross pieces over the top.

The Intercisa-type helmet differed radically from earlier helmets in design. The bowl was composed of two separate sections joined by the ridge in the middle, small neckguard and no reinforcement to the front and top of the helmet.

Later-pattern Coolus helmets had increasingly wide cheek-pieces and neckguards. These trends, combined with others taken from the tradition of iron Gallic helmets, eventually produced the Imperial Gallic and Imperial Italic helmets, the vast majority of which were made of iron. The earliest types of Imperial helmets have cut-outs for the top of a soldier’s ears, which soon developed into ear guards. Neckguards soon ceased to project straight out from the back of the helmet and became increasingly lower and deeper, the guard itself at the same time becoming wider. In time these offered some protection not just for the back of the neck, but also for the shoulders. Rows of ribbing usually strengthened the back of the helmet. The front of the bowl was protected by a peak, much like the Coolus patterns, but increasingly thick. Imperial helmets might be made of either bronze or iron, although the latter is far more common for the Imperial Gallic type. The principal differences between Gallic and Italic helmets is stylistic, although the quality of the latter’s finish is also generally superior. Distinctive features of Imperial Gallic helmets are the heavily stylized raised eyebrows on the front of the helmet bowl, the brass piping edging much of the helmet and the brass or enamelled bosses used as decoration. Neither these, nor Italic patterns, tend to have crest knobs, and instead crests were slotted into forked crest holders. In the late 1st century ad, further efforts were made to strengthen the front and top of the helmet. Two reinforcing bands were attached, in some cases to existing helmets as well as in new models, forming a cross on the top of the bowl, so that sometimes this pattern is referred to as the ‘hot-cross bun’ helmet.

Whilst it would probably be wrong to depict the development of Roman helmets as a steady evolution, the long-term trends do reveal the preoccupations of the makers. Protection to the top of the head was always the highest priority. This area was exposed to the enemy, especially an enemy armed with a slashing weapon such as a sword and able to cut down onto the top and front of the bowl. The peaks on the front of the helmet, as well as the thick reinforcing bands, were all intended to protect against such blows. Equally interesting is the development of deeper and broader neckguards. An enemy in front aiming a blow at the top of the helmet could very easily miss or find his blow deflected down. Neck guards not only protected the neck itself, but also gave some defence to shoulders. Another notable feature of all Roman infantry helmets before the 3rd century AD was that they left the face and ears uncovered. It was important for a soldier to see what was going on and to be able to hear orders, and both vision and hearing were seriously impaired by closed helmets such as the Corinthian helmet associated with Greek hoplites or the Great Helms of the Middle Ages. However, as much protection as possible was offered by the large cheek-pieces and earguards on Roman helmets.

Helmets, crests and ranks

According to Vegetius, centurions were distinguished from ordinary soldiers by wearing a wide transverse crest. Although depicted on only two monuments, and not yet confirmed by any extant example of such a helmet, it is generally accepted that this applied throughout the Late Republic and Principate. Standard-bearers wore animal skins over their helmets, and it is just possible that this practice was also followed by some auxiliary units. In Caesar’s day it was clearly normal practice for soldiers to mount crests or plumes in their helmets during a battle, for he mentions an occasion when the suddenness of an enemy attack denied the soldiers time to do this, as well as remove the leather covers from their shields. Many helmets until the end of the 1st century ad were fitted with the mountings to take a crest. Some helmets, especially of the Coolus pattern, additionally have tubes on either side of the bowl which presumably supported tall plumes. It is unclear whether these represented a badge of rank, perhaps for optiones, or were the insignia of a particular legion or legions, or even of a sub-unit within the legion. Julius Caesar formed a legion from Gallic recruits to whom he subsequently gave Roman citizenship. There is an attractive, if unsubstantiated, suggestion that this unit, Legio V Alaudae or ‘the Larks’, may have worn larks’ feathers either side of their helmet in this fashion. On Trajan’s Column, and other contemporary monuments, crests are rare, associated only with parades. The reinforcing cross bars on the tops of helmets made it impossible to mount a crest in any of the traditional ways, but whether this, or simply fashion, was the reason for this change in practice is unclear.

Linings and headgear

Some surviving Roman helmets, for instance an Imperial Gallic helmet from Brigetio, show traces of a lining inside the bowl. This ensured that the helmet was a more comfortable fit and was most likely padded so that it helped to cushion any blow delivered to the helmet. Such cushioning was essential if the wearer was not to be at least stunned, and possibly more seriously injured by non-penetrating strikes to the helmet. If the helmet lacked an integral liner, then the soldier would certainly have worn some sort of headgear beneath it to perform this function.

In Caesar’s day it had been normal to wear crests and plumes in battle, but this practice appears to have become rare in the army of the Principate However, many helmets could be fitted with crests and such ornamentation was still common for parades and ceremonies A group of re-enactors display a range of different decorations to their Imperial Gallic helmets In the centre the centurion’s transverse crest (which was probably worn even in battle) marks him out as an officer. It is not known whether other ranks, such as optionee, were also distinguished by crests and plumes of specific shapes or colours, as is suggested here, but it is certainly plausible that they were.

Later developments in helmet design

The 3rd century saw significant changes in many aspects of the Roman army’s equipment. Whilst some Imperial-pattern helmets continued to be used in the first decades of the century, a new type of infantry helmet, heavily influenced by the design of cavalry helmets, seems to have become common. A good example of this type was found at Heddernheim in Germany. Unlike earlier infantry helmets, the ears are almost entirely covered and the neck- guard is steeply angled down. There were a few elements of continuity, notably in the continued manufacture of bowl and neckguard from the same piece of metal, but this tradition was soon to be broken by the adoption of several types of helmets made in sections which supersede most other types by the later 3rd century ad. In simple designs such as the ‘ridge’ type found at Intercisa in Hungary and elsewhere, the top of the helmet consisted of two halves joined by a ridged strip across the top. Several examples show traces of a lining. Although generally cheaper and quicker to make, some of these helmets are decorated with false eyeholes. Many of the Heddernheim-type helmets were exceptionally ornate, and this trend continued with some later designs, most notably the two examples from Berkasovo. The cheaper Intercisa-type ridge helmets or the equally crude spangenhelms - the bowls made from four pieces of iron - were probably infinitely more common in everyday usage, especially by ordinary soldiers. The spangenhelm may well owe its origin to helmets employed by nomadic races from beyond the Danube, such as the Sarmatians and the related Alans. Such simple helmets were not of the quality of earlier types. Nevertheless, the Roman army’s ability to provide all, or virtually all of its soldiers with such gear, continued to give them an advantage over barbarian peoples where helmets were the preserve of only the wealthiest warriors.

Body armour

Mail (lorica hamata): Mail armour was in regular use by the Roman army throughout the period covered by this book. In the 1st century вс the vast majority of legionaries wore a mail cuirass, and at least some continued to do so after the famous segmented plate armour came into use. Although some shirts may have been made of copper alloy, the vast majority consisted of iron rings, on average about 1 mm thick and 7 mm in diameter. Each ring was normally connected to four others, the ring being either riveted or welded whole. It was common for legionary mail shirts in particular to have shoulder doubling and extend down onto the hips. A mail

A close-up of an auxiliary soldier from Trajan's Column showing him wearing mail armour, helmet, tunic and breeches. The tunic is a little unusual in that its lower fringe is decorated, while his sleeves are covered in a row of leather strips Like all auxiliaries on the Column, he carries a flat oval shield.

cuirass is flexible and essentially shapeless, fitting more closely to the wearer’s body than other types of armour. In this respect it is comfortable, whilst the wearing of a belt helps to spread its considerable weight of 10-15 kg (22-33 lb) which would otherwise be carried entirely by the shoulders. Mail offered reasonable protection, but could be penetrated by a strong thrust or an arrow fired at effective range.

Scale (lorica squamata): Scale armour was sometimes employed by legionaries, and is shown several times on the Adamklissi metopes in Romania. Unlike mail, which was relatively easy to maintain and repair, scale was more prone to damage so that individual scales appear relatively frequently in the archaeological record. Scales were usually of copper alloy or occasionally of iron. The former might be tinned to present a silvery finish and this type of armour could be polished to a high sheen. The size of the individual scales varied from shirt to shirt, although most were relatively small, no more than a few centimetres long and narrower than they were deep. Rows of scales were wired together and then sewn onto a fabric backing. The scales themselves were thin, and the main strength of this protection came from the overlap of scale to scale, which helped to spread the force of the blow, or from the smooth surface deflecting a hit in the first place.

(Above) Pieces of scale armour found at Newstead in Scotland date to the end of the 1st century AD. Scales can be of iron or bronze and vary in size, although their basic design is broadly consistent. Broken scales were less easy to repair than rings of mail, so finds of broken scales are more common.

(Above) A diagram showing the construction of lorica segmentata based upon the finds at Corbridge, south of Hadrians Wall. Leather straps support the articulation of the iron plates The many bronze fittings from these cuirasses were very prone to breakage and are comparatively common finds as a result.

Plate (lorica segmentata): The segmented plate cuirass is the type of body armour most immediately associated with the Romans, although it seems only to have been in use under the Principate. Part of an early version of this armour was discovered in the excavations at Kalkriese, the probable site of Varus’ military disaster in At) 9, pushing back the assumed date of its adoption by about half a century. Its use became less frequent in the 3rd century AD and the type was soon altogether abandoned. Although depicted on Trajan’s Column, it was not until Russell Robinson successfully reconstructed the fragments of three cuirasses found in the hoard of military equipment at Corbridge, south of Hadrian’s Wall, that its design became intelligible. The cuirass consists of a series of iron plates articulated by leather straps running underneath. Especially good protection was provided for the shoulders. The iron plates appear not to have been hardened by forging, probably so that the ‘soft’ metal absorbed a blow, spreading its force. Modern tests have confirmed the efficacy of segmented armour, suggesting that it would deflect or stop most strikes from arrows or spears. Weighing in at some 9 kg (20 lb), a segmented cuirass was a little lighter than mail armour, but its shape made it less comfortable to wear and could restrict deep breathing. The complexity of its design, with the numerous plates linked by copper-alloy buckles, hinges and hooks and the leather harness underneath presented many maintenance problems. Chemical reactions between the bronze fittings and the iron plates fostered corrosion, whilst many of the fittings broke or fell off far too easily. Fittings from segmented cuirasses are fairly common finds on Roman military sites. Although simpler versions of the armour were developed, for instance the ‘Newstead pattern’ named after the site in southern Scotland where the example was found, these problems were never adequately solved and, combined with the technological skill required to construct the armour in the first place, probably explain its eventual fall from favour in the difficult conditions of the 3rd century AD.

All types of armour at all periods would have been worn over some sort of padded garment and not directly on top of the tunic. Little is known about this item, since it was certainly made of perishable textiles and no example or unambiguous depiction has survived. Apart from making the wearer more comfortable, this extra layer complemented the protective values of each type of armour, and helped to absorb the shock of any blow striking the outer armour. Some of these jerkins may well have ended in a row of leather straps hanging down almost to the edge of the tunic and providing a measure of protection for the lower body and thighs. A very late, and generally unreliable source, refers to such a padded garment as a thoramachus, whilst another History speaks of a subartnalis. The second term probably appears on a text from Vindolanda describing some sort of textile, although a roughly contemporary document from Carlisle has confused the issue by speaking of a type of javelin known as a subarmalis. In some cases an additional, waterproof layer, most probably of leather, may have been worn either between the thoramachus and the cuirass or on top of the armour to guard against the weather.

The legionary shield (scutum)

In many respects the legionary shield changed little from the Mid Republic through to the early 3rd century AD. It remained a long, body shield, semicylindrical in shape and of plywood construction. Before the Marian reform legionaries carried an oval shield, both the tops and sides curving, like the example from Kasr el-Harit in Egypt. Oval shields of much the same pattern remained in use well into the Principate, both with the legions and, perhaps especially, the praetorian guard. Yet by the beginning of the 1st century ad the vast majority of legionaries carried rectangular shields. These were a little shorter, though of similar width to the earlier pattern. Most were flat-topped, with curved or straight sides, but a leather shield cover from Caerleon was intended for a shield with a curved top and straight sides so there was evidently some variation. The evidence for the shape of legionary shields in the Late Republic is exceedingly poor and it is difficult to know when and why this change occurred. The shorter rectangular shield was a little lighter than the old oval type, and was perhaps more attractive to the professional legionaries in the post-Marian army who had to carry their weapons and equipment on the march. This is conjecture, but there is equally little basis for the common assumption that the rectangular shield did not come into use until the end of the 1st century вс.

No example of a legionary scutum survives from the 1st to 2nd centuries ad (although one 1st- century shield may now have turned up at Masada), but one well-preserved rectangular, semi-cylindrical shield was found with the 3rd-century ad remains at Dura Europus. The shield was 1.02 m (3 ft 3 in) long and 0.83 m (2 ft 8 in) wide. It was made of three layers of strips of wood glued together, the back and front set laid at right angles to the longitudinal middle layer. Its thickness was about 5 cm (2 in), but unlike the Kasr el-Harit shield it was not thicker in the centre than at the edges. A rectangular gap in the centre of the shield had been prepared to take the shield boss, although in fact this had not been fitted. The back of the shield was reinforced with a framework of wooden strips glued or pegged into place. The shield had a horizontal handgrip. Both back and front were covered with a thin sheet of leather, over which were stitched reinforcing leather pieces for the corners and a wide binding to protect the edge. This appears to have been a later, cheaper alternative to the brass binding which was normally used on these shields. Examples of this metal shield trim are common finds, suggesting that they were subject to frequent damage. A reconstruction of the Dura Europus shield with an added iron boss and bronze binding weighed in at 5.5 kg (12 lb). A shield of the same dimensions but with the increased thickness of wood in the centre would have weighed something like 7.5 kg (16.5 lb), which was still lighter than the reconstructed Kasr el- Harit shield at 10 kg (22 lb).

When not in use, shields were protected from the weather by leather shield covers, fragments of which have survived from several sites. The front of the shield was normally decorated with the unit’s insignia - usually in paint - and some of these symbols were highly elaborate. The sculptors of Trajan’s Column took great care to carve devices onto shields, but it is now impossible to identify these with specific units, and only a handful of unit symbols are clearly shown on tombstones. It is not clear whether the entire legion shared a common shield device, or whether each cohort was distinguished in some way, perhaps using colour, but it is clear that some sort of system existed. Tacitus recounts an incident during the civil war after Nero’s death when two soldiers picked up shields from the enemy dead and, using this disguise, were able to infiltrate their positions and put a large catapult out of action. The Dura shield was painted a pinkish red on both sides and the front decorated with a highly detailed geometric pattern, as well as the figures of a lion, eagle and twin victories. This was most probably the insignia of one of the legions in garrison. Such delicate detail can only have been subject to rapid wear, especially when the shield was employed in battle, and it is quite possible that on campaign more basic decoration would have been employed on replacements for damaged shields. Yet unit insignia were obviously important, and even the leather covers which protected shields on campaign sometimes included decoration. One mid-1st-century ad rectangular shield found in Holland belonging to a member of Cohors XV Voluntariorum civium Romanorum was decorated with two capricorns, the symbols of Augustus, who originally raised the unit. Even without heavy use, a single wooden and leather shield was unlikely to last a soldier for his entire 25 years service, unlike some metal gear, notably helmets which have sometimes shown signs of being owned by more than one soldier.

(Left) A curved oval scutum based on the Kasr el- Harit find in Egypt with a design taken from the Arch of Orange, southern France (in use from at least the 3rd century вс).

(Centre) A curved rectangular scutum based on the example found at Dura Europus on the River Euphrates (1st-early 3rd century AD).

(Right) A shield with straight sides and curved ends, based on a number of surviving shield covers. It is impossible to tell from these whether the shield was curved or flat (1st-2nd century AD).

(Left) The dimensions of this fiat oval shield are based on a shield cover excavated at Valkenburg in Holland. It has a design taken from an auxiliary's shield on Trajan's Column (lst-3rd centuries AD).

(Centre) A flat shield based on an example found at Doncaster. Unlike most Roman shields it luis a vertical handgrip (1st century AD).

(Right) An oval shield based on examples found at Dura Europus. On the original the detail of the figure’s right hand is lost and this reconstruction is conjectural (3rd-4th centuries AD).

Other armour

Although various forms of muscled cuirass - most probably metal - are depicted as being worn by senior officers in Roman art, no example has survived from this period to confirm that this reflected actual practice and was not simply artistic convention. Similarly, it is now difficult to know whether or not the Romans employed any types of leather armour.

The Adamklissi metopes show that Roman legionaries in the late 1st or early 2nd centuries AD sometimes felt the need for greater protection than that offered by shield, helmet and cuirass. In all the scenes showing legionaries in combat these men are depicted as wearing greaves to protect their calves, and a segmented armguard or vambrace on their right arm. Many of their barbarian opponents carry two-handed, scythe-like fakes, long weapons capable of reaching round the shield to strike at the right arm or lower leg, and the extra protection can be seen as a means of guarding against this specific threat. The armguard was constructed of articulated iron plates with leather harness underneath, in many respects similar to the segmented plate cuirass. Although plate armour is not shown on the Adamklissi metopes, where all legionaries wear scale or mail, an example of such a cuirass has been found associated with an armguard at Colonia Sarmizegethusa Ulpia, the capital of Roman Dacia. Yet finds from Newstead and more recently Carlisle in northern Britain have shown that this additional armour was not confined to the Danubian front but also employed in other areas. No example of Roman greaves clearly dated to this period has been found, but it is probable that, like earlier Roman examples, these were lined with leather and tied into place rather than clipping to the leg like those worn by Greek hoplites.

A detail from one of the metopes at Adamklissi showing a legionary using the boss of his scutum to punch an enemy in the face and then stabbing him in the stomach with his gladius. This is the only clear depiction of a soldier using this well-attested fighting technique.

Offensive weapons

The pilum: The pilum or throwing-spear was used by most legionaries until it gradually fell from use in the 3rd century AD, its design changing only in small details from the weapons in use under the Mid Republic. Pilum heads survive from a good number of sites, whilst several examples from Oberaden in Germany were excavated with parts of their wooden shaft remaining. There were two methods of attaching the long iron head to the wooden shaft.

Many of the legionaries on the Adamklissi metopes are shown carrying pila with a circular object attached just behind the top of the staff. This is most probably a weight intended to increase its penetrative power when thrown.

The legionaries from a relief found at Croy Hill in Scotland each carry an ordinary, unweighted pilum and wear paenula cloaks. They are bareheaded, their helmets hanging either from the top of their shields or across their body. Two of the soldiers appear to have decorative pteruges (aprons) over their tunics

Some of the iron heads ended in a socket, the joint reinforced by an iron collet fitting over the top of the wooden shaft, but the majority had a wide rectangular tang which slid into a groove in the wood and was fastened into place by two rivets. One relief clearly shows a pilum with a butt-spike, but it is not known if this was always a feature of this weapon. Other sculptural evidence, most notably the Cancelleria relief showing praetorian guardsmen, and the Adamklissi metopes, show a ball-shaped object just below the wider top of the wooden shaft. This was most probably a weight, perhaps in lead, intended to increase the penetrative power of the pilum by focusing even more power behind its small head. One 3rd-century relief shows two of these weights. A pilum’s effective range was about 15 m (50 ft), but such modifications may well have reduced this. This weapon was intended first and foremost to kill or wound, punching through an enemy’s shield and armour, but it required a high standard of discipline to refrain from using it until the enemy came within effective range. A single early tombstone shows a legionary carrying two pila, but, whilst it is possible that two were carried on campaign, it is extremely unlikely that a soldier took more than one into battle.

Spears and javelins: Legionaries may have sometimes employed other shafted weapons. The early 2nd-century ad Roman commander Arrian left an account of the formation of his army for an anticipated battle with the Alans, a nomadic people who relied predominantly on heavily armoured cavalry. His legionaries were formed up very deep to meet the enemy charge. The first four ranks carried pila, but the fifth to eighth ranks seem to have had the lancea, a type of javelin, and were ordered to throw these over the heads of their comrades once the enemy closed into contact. Ultimately the use of different types of throwing-spear can have required little re-training, certainly not in comparison to a modern infantryman being issued with a new rifle. Therefore it seems likely that legionaries were sometimes equipped with javelins other than the pilum if the particular situation required it. A series of 3rd-century legionary tombstones from Apamea in Syria suggest that Legio II Parthica had men skilled in a wide range of weapons, including light javelins (lanceae). Whether these specialists were always equipped in this way, or only when required, is impossible to say.

Spear and javelin heads are relatively common finds from Roman military sites. They vary considerably in size and shape, and the few scholarly attempts to classify them have rarely employed useful criteria. It is usually impossible to estimate the length of the shaft fitted to any of these weapons. Those used for fighting would doubtless have been long, perhaps just under 3 m (9 ft), whilst those solely designed for use as missiles were much shorter at a little more than a metre (3-4 ft).

(Left) A scale drawing comparing the shapes of, on the left, a Mainz-type gladius, in the centre the Pompeii type, and on the right a longer spatha Individual examples of each type vary to some degree in dimensions.

(Right) The blade of a spatha found at Newstead in Scotland.

The gladius: Although the evidence for Republican swords is limited, far more is known about the side- arms of the Imperial army. In the early 1st century AD the dominant type was the ‘Mainz’ pattern. This has a slightly tapering blade and an exceptionally long point. The length of blade on surviving examples varies from 400 mm (16 in) to 550 mm (22 in), and width from 54 to 74 mm (2.1-2.9 in) at the top to 48 to 60 mm (1.8-2.3 in) before the point. A shaped handgrip of bone was protected by a guard and pommel, usually of wood. The pommel also served as a counterweight and helped the balance of the weapon. Although especially suited to thrusting, with the long point - sometimes as much as 200 mm (7.87 in) - intended to penetrate armour, the Mainz- pattern sword was also an effective slashing weapon. Later in the 1st century ad, Mainz-pattern blades appear to have been largely supplanted by the ‘Pompeii’ type. This was a straight-bladed weapon with a much shorter point. Blade lengths vary between 420 and 500 mm (16.5-20 in) and widths between 42 and 45 mm (1.6-2.2 in). Even more than the Mainz pattern, the Pompeii-type gladius was a supremely well balanced and effective weapon for both cutting and thrusting. In the latter part of the 2nd century a similarly shaped pattern was introduced. The main difference was in the handle and fittings, for instead of the familiar pommel, the tang of the blade was extended into a ring of iron.

The gladius of whatever pattern was invariably worn on the right side, save by centurions, and perhaps other senior officers, who wore their swords on the left. Although awkward to modern eyes, experiments have repeatedly demonstrated that these swords can easily be drawn when slung on the right side, and that this avoids any entanglement with the shield. Analysis of several examples of Roman sword blades has shown that some were of very high quality, consisting of both low-carbon iron and carburized steel. The better weapons had been quenched to harden the metal and tempered. Other examples were less sophisticated, but in general the quality of the army’s weaponry was good.

The pugio: The military dagger (pugio) varied in blade length from 250 to 350 mm (9-14 in) and provided a stout backup to the sword for both legionaries and auxiliaries. Their scabbards were often richly decorated, adding to the splendour of a man’s weapons’ belts. Daggers remained in use for most of the Principate, although they do appear to have become significantly less common in the 2nd century ad and are not shown on Trajan’s Column. The dagger was worn on the opposite side to the sword - on the left for ordinary soldiers and one the right for centurions who wore their sword on the left.


The eagle (aquila): The eagle standard was an object of massive reverence. Given this importance, it is not surprising that no example has survived in the archaeological record and that we must rely largely upon sculptural evidence in reconstructing its appearance. Marius is said to have issued each legion with a silver eagle, but by the Principate the top of the standard appears to have been gold or gold plated. Decoration was usually fairly simple and the eagle’s staff virtually bare, although a figure on the breastplate of the Prima Porta statue of Augustus does show a row of discs like those normally associated with signa.

The signum (pl. signa): The tradition of each century in a legion having its own standard (signum) appears to have continued throughout the Principate. Certainly, each century continued to have a standard-bearer (signifer) amongst its principales, and it is normally assumed that these men actually carried a standard. Signa appear to have been topped either by an ornamental spearhead or an upraised hand. Their shafts were heavily decorated with cross-pieces, wreaths, and from two to six large discs. The actual significance of any of these items is unknown, though it does seem probable that together they provided a system for identifying the particular century. The upraised hand may originally have been the symbol of the maniple, for the word probably derived from the Latin for hand, manus, meaning a small group or handful of men.

(Above) Many legions had a particular symbol or symbols which often appear on sculpted stones or moulded on to tiles Here we see the boar symbol of Legio XX Valeria Victrix.

The vexillum (pl. vexilla): A range of flags were carried by different units of the army. Traditionally a distinctive vexillum or flag, usually in red, marked out the commander’s position in camp before a battle and on the battlefield. Vexilla also provided the main standard for detachments of troops serving away from their parent unit, so that in time such detachments became known as vexillations (vexillationes). Roman flags were suspended from a cross-bar to hang down in front of the main shaft. One example of such a banner has been found in Egypt and carries the figure of a Victory on a red background.

The imagines: Under the Principate each unit also included a series of images of the emperor and his close family which were mounted on poles and kept with the standards. These served as a reminder to the soldiers of their oath and loyalty. Mutinous troops, especially those incited to support their own commander in a bid for the throne, usually began by tearing down the imagines.

The draco: Auxiliary standards seem in general to have followed the patterns of those used by legionaries, but by the early 2nd century AD an additional type had been adopted by some cavalry units for use in parades and perhaps at other times. This was the dragon or draco, a bronze animal head with an open mouth and neck to which was attached a multi-coloured tube of material. When the standard-bearer moved quickly, the tube of material acted like a wind-sock, streaming behind the head and making a whistling sound. These standards seem to have been copied from some of Rome’s opponents on the Danubian frontier, most notably the nomadic Sarmatians, and are depicted on Trajan’s Column flying over the enemy armies.

A detail from a gravestone showing two standards, the square flag or vexillum and a heavily decorated century standard or signum. The standards of the praetorian cohorts were at times so heavy with ornaments that the Emperor Caligula gave his guardsmen permission to carry their signa on pack animals during a long march.


On campaign a Roman legionary had to carry more than simply his armour weapons, personal belongings and food. Josephus, perhaps unconsciously echoing the old joke about Marius’ mules, compared the legionary to a pack animal, saying that each man carried a saw, a basket, a pick and an axe, as well as a strap, a bill-hook and a chain. Whilst it is unlikely that every man was equipped with all of these items, it probably gives a fair impression of the range of tools available to a contubemium of eight men. As one modern commentator pointed out, the professional legionary was as much combat engineer as infantryman. Gnaeus Domitius Corbulo, one of the most famous generals of the 1st century ad, is even reported to have declared that ‘battles were won with the pick-axe’. The Roman military pick-axe, or dolabra, was a well-designed implement, combining a cutting blade with a spike which could be used to break up the earth or undermine an enemy wall. Another type of tool blade is very similar in shape to a modern turf cutter, and is usually assumed to have performed a similar role, although its suitability for the task has been disputed. Controversy also surrounds the precise use of the so-called wooden ‘palisade stakes’. Traditionally these were equated with the two stakes reportedly carried by each soldier and planted on top of an earth rampart to form a palisade. Recently, one scholar has suggested an alternative in which three of these stakes were tied together to form a self-standing barrier. Either interpretation is possible, but it must also be admitted that we may have wrongly identified these items in the first place. In addition to the tools and equipment required in some quantities, more specialized items were required for the men overseeing any project. Chief amongst these was the Roman surveying tool or groma, which consisted of an upright pole, mounting a cruciform piece with a lead weight hanging from the end of each section.

Auxiliary Equipment

A tombstone from Housesteads fort on Hadrians Wall showing an auxiliary archer. He has an unusual pointed helmet, probably mail armour, a quiver over his shoulder and carries a short axe in addition to his re-curved composite bow. Apart from having an ordinary tunic instead of long robes, he is not too dissimilar from the eastern archers shown on Trajan's Column.

It is in fact surprisingly rare that an excavated piece of military equipment can be incontestably associated with a particular unit or branch of the service, since only a tiny fraction of surviving finds carry such detailed identity markings. The distinctions in dress and equipment between legionaries and auxiliaries have as a result tended to be based primarily upon the depiction of soldiers on commemorative and funerary monuments. On Trajan’s Column the clear distinction between the citizen legionaries and the native auxiliaries is maintained throughout. Legionaries wear the segmented cuirass and carry curved rectangular shields, whilst auxiliaries wear longer mail shirts, often wear breeches and have flat oval shields. Such a depiction is clearly stylized, and the contemporary Adamklissi metopes demonstrate that many legionaries in fact wore mail or scale armour. In spite of several attempts to argue that there were no distinctions in equipment between the two halves of the army, it does seem possible to state a few points. There is absolutely no evidence that auxiliaries ever wore lorica segmentata, used the pilurn or, with the possible exception of a handful of units called scutata, carried semi-cylindrical shields. Instead they appear to have carried spears or javelins, worn mail and probably scale, and had flat shields. The last may sometimes have been rectangular or hexagonal instead of oval. Scholars have frequently associated simpler, less well-made or decorated items - most notably helmets - as auxiliary equipment. Although this is plausible enough, it must be emphasized that it does not rest on any solid evidence.

(Above) The tombstone of Maris, a horse archer, found at Maim. Notice the servant standing behind him and holding a sheath of arrows Servants appear on a number of cavalry tombstones

The majority of auxiliary infantrymen wore a helmet and body armour, carried a shield and were equipped with one or more spears/javelins and a gladius. Although often described by modern commentators as light infantry who fought in a looser order than the legions, this view must be treated with caution. Apart from the shield, which may have been lighter than the scutum carried by legionaries, auxiliary equipment was of similar weight to that borne by their citizen counterparts. Our ancient sources do refer to light armed units who operated as skirmishers in open order, but the bulk of auxiliary infantry appear to have fought in much the same way as legionaries. Whether the actual light infantry were organized into separate units or were sub-units of ordinary cohorts is impossible to say. Nor is it clear precisely how these men were equipped. Slingers appear on Trajan’s Column, but no unit of slingers (funditores) is known. At Lambaesis in North Africa Hadrian complimented the cavalry of a cohort on their prowess with slings, so it is more than possible that some units trained some or all of their men in the use of this weapon. Moulded lead sling bullets, sometimes with obscene messages on them, have been found at many sites, especially those associated with the civil wars of the Late Republic. Stone pebbles may well also have been used as ammunition.

Archers are depicted on Trajan’s Column, some dressed in a caricature of eastern costume with long flowing robes and distinctive helmets. Many auxiliary units, both horse and foot, were formally designated as archers (sagitarii), although even here the evidence is a little unclear, since a few of these units appear to have carried other weapons whilst some units without this title employed bows. The Romans used a sophisticated recurved composite bow (i.e. made of more than one type of wood). Since wood rarely survives we have no example of such a weapon, but bone laths, which were fitted to the grip and ‘ears’ or ends of the bowstave, along with arrowheads, occur with some frequency at military sites. The vast majority of Roman archers employed what is known as the Mediterranean release, where the string is held in two fingers and drawn back to the chin. A leather bracer was worn on the inside of the left arm beneath the elbow to protect this from bruising as the released string whisked past. There is a little evidence to suggest that the alternative method, known as the Mongolian release, which used a thumb ring to grip the bowstring, was practised by some units in the 3rd century ad.

Cavalry Equipment

Helmets and armour

Cavalry helmets differ from those used by the infantry in several important respects. In the first place the soldier’s ears were almost invariably covered by extensions from the cheek-pieces. This can only have impaired the wearer’s hearing, and some helmets have small holes drilled in these cheek-pieces to counter this. In a whirling cavalry melee, when two squadrons interpenetrated each other, a horsemen could easily be attacked from the side or rear and so the extra protection to the face was clearly considered to be more important than some loss of hearing. The other most obvious difference to the patterns of helmet worn by Roman foot soldiers is the much deeper but comparatively narrow neckguards on cavalry helmets. A wide projecting neckguard could easily break a man’s neck if he was to fall backwards from his horse. However, in many other respects, including methods of construction, Roman cavalry helmets mirror those of the infantry and, as we have seen, by the 3rd century the two styles drew even more closely together. Some types of decoration do appear unique to the mounted arm, for instance the sculpting of stylized hair on the helmet bowl. In a few examples, real animal or human hair was attached to the helmet instead of simply being depicted on it.

Auxiliary cavalrymen used flat shields in a similar variety of shapes to their infantry counterparts. The majority appear to have been oval in shape, but rectangular shields were used by some units, whilst others, including the cavalry which supported the praetorian guard, had hexagonal shields. A flat shield was found in a 1st-century ad Roman fort at Doncaster in Britain and was probably used by an auxiliary, and quite possibly a cavalryman. This example was large, measuring over 125 cm (4 ft) in length and 64 cm (2 ft) in width with straight sides and a curved top and bottom. Construction seems to have been very similar to the legionary shields discussed above, with three layers of plywood covered with thin leather, but unlike other Roman shields it had a vertical handgrip. A reconstruction proved almost as heavy as the Fayum shield, at 9 kg (20 lb), but was well balanced. The boss was slightly above the centre, which tended to make the shields lower half angle back towards the legs, which may have been useful for a man on horseback. Oval shields of far simpler construction, made from a single layer of wooden strips glued together, were found at Dura and similar types may already have been in use by auxiliaries before the 3rd century.

A cavalry 'parade helmet' from Nijmegen in Holland presenting a variation on the common style of hatring the top of the helmet shaped to represent hair. In this case real animal hair has been fixed to the helmet instead. This may have been a fashion most common amongst Batavian units.

(Above) The Bridgeness stone from the Antonine Wall portrays an auxiliary cavalryman in a pose reminiscent of many cavalry tombstones as he tramples over naked barbarians.

(Above) Two spatha blades found at Newstead in Scotland, giving a good idea of the longer, slimmer swords used by cavalrymen.

Roman cavalrymen wore cuirasses of mail or scale armour. One relief from Belgium appears to show a combination of the shoulder sections of lorica segmentata with a mail shirt, but this is the only evidence for such a composite cuirass. In the 2nd century AD, Hadrian raised the first known ala of Roman cataphracts, Ala I Gallorum et Pannoniorum cataphracta, in which both horse and rider were heavily armoured. Amongst the armour found at Dura Europus was a set of scale horse armour.

The spatha

Roman cavalry employed a longer, slimmer sword known as the spatha, for a cavalryman required a weapon with a longer reach, especially if he was to strike at an opponent on foot. Blades range in length from c. 65 to 91.5 cm (26 to 36 in) with a width usually of under 4.4 cm (1.75 in). Pommel, guard and hand-grip were generally similar to gladius types. Like the gladius, these longer swords were usually also worn on the right side.

Spears and javelins

A range of shafted weapons were employed by Roman cavalrymen. The longest was the contus, a spear some 3.65 m (12 ft) in length and held in both hands by a shieldless rider. This appears to have been first adopted in the 2nd century ad and only ever equipped a small number of specialist alae. It was a weapon for shock action, and could not have been thrown with any great effect. Most cavalry carried a shorter, one-handed fighting spear, and usually several smaller javelins for throwing. A text recording the inspection of the weapons of a cavalry ala in northern Britain in the late 1st century ad speaks of fighting spears (lancias pugnatorias) and the smaller javelins (minores subarmales). Each man was supposed to have one fighting spear and two subarmales, as well as a sword. This may have been the regulation equipment for this specific ala, for Josephus talks of Roman auxiliary cavalry with one long spear and three or more short throwing javelins carried in a quiver.

(Above) Another cavalry tombstone, from Gloucester, this time of a horseman from a cohors equitata and dating to the middle of the 1st century ad. The inscription reads 'Rufus Sita, cavalryman in Cohors VI Thracum, 40 years old and of 22 years’ service, lies here His heirs erected this in accordance with his will’.

(Above) Gains Romanius Capito was a trooper in the Ala Noricorum. Notice once again a servant standing behind his master’s horse and carrying extra spears or javelins, and also the decorated horse harness

(Above) A drawing showing the construction and shape of the four-horned saddle employed by the Romans

The saddle and horse harness

An idea frequently encountered in older books on ancient warfare is the belief that lack of stirrups prevented ancient cavalrymen from delivering any form of effective mounted charge and restricted them to tentative harassing action. This view was not reflected in our ancient sources, but these were ignored, for it was believed that without stirrups a horseman could not have had a stable seat. It is only in recent years that the reconstruction of the Roman saddle, pioneered by Peter Connolly, and its subsequent testing has finally demonstrated that this was simply untrue.

The four-horned saddle was employed by the Romans and also the Gauls, Parthians, Sassanid Persians and the Sarmatians, as well perhaps as other races. It is not known who invented it, although the Gauls must be prime candidates, and it is more than likely that the Romans copied the design from their enemies, as they did with so many other types of equipment. When the rider’s weight is lowered onto this type of saddle, the four horns close around and grip his thighs. This provides considerable support, allowing him to throw or thrust a spear and wield a sword effectively, even leaning to one side and recover.

'Parade armour' and the cavalry games

Although the vast majority of Roman cavalrymen were not citizens, this arm of the auxilia enjoyed considerable prestige. Cavalry alae, most of all the small number of milliary alae, were expensive and prestigious, presenting an imposing sight on parade or in battle. Under the Principate the cavalry sports developed as spectacular public displays of the splendour and skill of the army’s horsemen. For these events both horses and men were provided with highly ornate equipment. Most conspicuous in the archaeological record are the ‘parade helmets’. Similar in basic shape to ordinary cavalry helmets, these were often silvered and always far more elaborately decorated and had a shaped face mask with eye holes permitting vision. Representations of hair were even more common and a significant number of surviving helmets have strong female characteristics, presumably intended to allow their wearers to represent Amazons. Arrian, the only source to describe these ceremonies in any detail, says that the helmets were normally topped by a streaming yellow crest, although there does not seem any obvious mounting for such a decoration on any surviving helmets. Horses wore leather-studded chamfrons (decorative and protective head-pieces), an example of which was found remarkably well-preserved at Vindolanda, with dome-shaped grills to protect their eyes but still permit vision. The horse’s harness and saddle cloth, as well as the man’s clothing, was brightly coloured and additional spectacle was added by the use of many standards, including the tube-like dracones.

Arrian appears to describe a standardized version of the games as set down by the Emperor Hadrian, but claims that much of the display had very ancient traditions. It began with a mixture of manoeuvring and charges, and this was followed by individual and group exercises. Divided into two sides, the cavalrymen lobbed blunt-headed spears at each other, aiming to strike the target’s shield. Normal weapons were thrown at targets at various speeds and from different directions. During a straight charge across the display area a good soldier was capable of hurling 15 light javelins at the target, whilst a few of the very best managed 20. Though stylized, these exercises reflected many of the skills required in actual combat.

Helmets categorized by archaeologists as ‘parade helmets’ are common, scarcely less so than those thought to have been worn in battle. Although the cavalry games are well attested and were clearly an important phenomenon, we have perhaps been over-ready to ascribe any highly decorated piece of equipment to these sports. Roman military equipment often combined high levels of ornamentation with practical utility and it is at least possible that some helmets with face masks were worn in battle. Such a mask has been found amongst the equipment thought to have been lost by the Roman column during the Varian disaster in AD 9, suggesting that it was at least taken on campaign. An admittedly badly preserved tombstone from Germany appears to show a legionary signifer wearing a masked helmet, so such items may not have been restricted to the cavalry. The fixed features of a metal mask may have been very intimidating, hiding the fact that the wearer was simply an ordinary man. Perhaps under some circumstances this, and the added protection, were considered more important than unimpaired visibility.

(Above) A richly decorated leather chamfron (horse’s head-piece) is reconstructed from an example found at Vindolanda dating to the late 1st or early 2nd century AD.

These modern re-enactors give a good impression of the splendour of the Roman cavalry. The standard-bearer wears a helmet with a face mask, but in other respects the men s equipment is of the type worn in action. When taking part in parades or games Roman cavalrymen were often dressed even more elaborately.

You can support our site by clicking on this link and watching the advertisement.

If you find an error or have any questions, please email us at Thank you!