The Numidians 300 BC–AD 300

In Military History, Featured

On the blog today, William Horsted, author of the recently released MAA 537: The Numidians 300 BC–AD 300, discusses the appearance of the Numidians in their various accounts and depictions.

 Numidian Cover


This is my first book, and the first piece of ‘history’ I have done since leaving university more than fifteen years ago.  I greatly enjoyed learning about the whole process of bringing a book to publication, of which writing is only the beginning; so much more work goes in to parts of the book the reader cannot see, such as producing references for the artists, obtaining permissions for the reproduction of images, proofreading and editing.  I am enormously grateful to my editor at Osprey, Nick Reynolds, and his team for helping me throughout.  However, what I most enjoy is research, particularly the collation and evaluation of primary source material.  In the following extract from my book, I explore the evidence available to historians for the appearance and clothing, and weapons and equipment, of the Numidians.

The clothing and appearance of Numidian troops can be reconstructed from a range of literary and non-literary sources, including epigraphy (inscriptions and images carved into stone monuments and steles), numismatic evidence (images on coinage), sculpture and archaeology (surviving artefacts).

The works of Greek and Roman writers frequently mention the Numidians, and other North African peoples. Unfortunately, despite numerous accounts of the exploits of the famous Numidian cavalry during the three Punic Wars (246–164 BC) and later conflicts in North Africa, few include detailed descriptions of the Numidians’ appearance. Strabo, however, tells us (Geography, 17.3.7) how North African people enjoyed wearing gold jewellery, growing beards and dressing their hair in braids. They took such care over their hairstyles that they avoided touching one another when they walked, in case their hair should be disturbed. They wore animal skins, including those of bears, lions and leopards, which they also slept in, and dressed in unbelted tunics decorated with wide borders. Strabo also records the North Africans carried animal skins for protection, made small shields out of raw hide, and were armed with javelins with wide heads.

Writing at a similar time, Livy described (History of Rome, 35.11) the unimpressive appearance of Numidian auxiliaries: the horsemen were small and slender, just like their mounts, and wore no armour. Their only weapons were javelins. These accounts are frustratingly brief, so we must look to other sources for evidence of the Numidians’ clothing and appearance.

A number of stone steles have been found in northern Algeria and Tunisia on which images of Numidian chieftains and other figures are carved. Details of clothing, hairstyles and beards are visible in many cases. As these are often ‘royal portraits’, however, it cannot be assumed that they represent the appearance of ordinary Numidians. Of course, this is not the case for the most famous artistic portrayal of Numidian soldiers: the unit of 14 mounted cavalrymen shown in such wonderful detail on Trajan’s Column. Completed in AD 113 to commemorate the emperor Trajan’s victory over the Dacians in his wars of AD 101–106, Trajan’s Column is decorated with a 200m-long frieze showing the progress of the campaign, scene by scene. It is the best visual source for the appearance, weapons and equipment of the Roman Army during this period. This includes allied troops such as the Numidian cavalry who accompanied the Romans as auxiliaries.

In Scene 64 of Trajan’s Column, the Numidian cavalry are shown attacking the Dacians, and much detail is visible. It is possible, for example, to see how the riders’ short tunics were tied over one shoulder; how the cavalrymen grasped their distinctive small, round shields; to see the Numidians’ ringleted hair and beards; and the woven collars with which they guided their horses in place of bridles. The relief carving has been surprisingly well preserved, though the polychrome paint and metal detailing has not survived. This means the javelins and other weapons that the Numidian horsemen may have originally carried have been lost, and we have no clue as to whether the tunics were dyed a particular colour.

Coin portraits are our best source for the appearance of Numidian rulers such as Syphax, Masinissa and Juba I, especially the way their hair and beards were styled. Coins show us only how each king wished to be portrayed, however, both to a domestic audience and to foreigners in neighbouring states into which the coins would travel through trade. On some coins, Numidian kings adopted a ‘Hellenistic’ appearance reminiscent of coin portraits of monarchs from the eastern Mediterranean, probably to make themselves and their kingdom appear to be part of wider Greek culture. Portraits of Numidian kings on their coins do not, therefore, necessarily show us how the kings, or their subjects, usually appeared. The reverse of Numidian coins – i.e. the side opposite to the obverse, or ‘head’ – can also be useful: several show images of mounted horsemen in which their tunics and billowing cloaks can be seen.

There is a terracotta statuette of a wounded Numidian ‘chieftain’ in the Louvre in Paris. This small mounted figurine was originally part of the decoration of a Greek water jug, or askos, and was found in a grave in Canosa, Italy. It has been very well preserved, and the clothing, equipment and physical appearance are crisply portrayed. There was another, similar figurine representing a Numidian horseman in the Victoria and Albert Museum in London, but it was sold in the 1950s. Fortunately, a photograph and description survive in a 1946 article by M. Rostovtzeff. These statuettes, which likely date from the late 3rd to the early 2nd century BC, were probably copied from reliefs on a monumental frieze depicting a battle of the Second Punic War, when Numidian horsemen were conspicuous in their role as the main light cavalry of the army of Hannibal. They provide the best sculptural source for the appearance of Numidian troops.

Little clothing survives from the ancient world – organic materials such as wool or leather soon decompose unless, in rare circumstances, they are preserved. In 1947, however, a fragment of a fringed leather tunic similar to those worn by figures on a number of Numidian steles was found in southern Algeria. Carbon dating suggests it was made around AD 600, and had survived in the very dry conditions found on the edge of the Sahara desert.

Examples of Numidian arms and armour are very scarce in modern collections. Most Numidian troops wore no armour, and carried few weapons other than javelins. Only the iron heads of these have survived, deposited in graves alongside the cremated ashes of their owners. Unfortunately, graves containing weapons are themselves rare, and they are concentrated in the western region corresponding to ancient Mauretania and the north-western edge of Numidia. A few leaf-shaped iron javelin heads have been found in several graves near Oran in north-western Algeria, an area controlled during the second half of the 3rd century BC by the Masaesyli, whose most famous king, Syphax, ruled partly from nearby Siga. The Punic necropolis on the small, volcanic island of Rachgoun, which lies about 2km off the coast and about 5km from Siga, contained some well-preserved examples dating from the 6th and 5th centuries BC.

In the eastern part of Numidia, inhabited by the rival Massyli, graves containing weapons are almost unknown. It is not clear why there should be such a difference in funerary practice across Numidia but the inclusion of weapons in Masaesyli burials may be due to influence from the nearby Iberian Peninsula, where during the 4th century BC the custom of burying warriors with their weapons was widespread. Many examples of Iberian pottery have been found in the necropolis on Rachgoun and at other nearby coastal sites, probably brought by the Phoenicians, who had been trading on both sides of the Alboran Sea since the 9th century BC, transmitting cultural and religious practice as well as material goods.

The important exception in eastern Numidia is the royal tomb known as the Soumaâ, situated in El Khroub, close to the ancient Numidian capital of Cirta. This once magnificent tomb, now ruined, is believed to contain the cremated remains of the Numidian king Micipsa, who died in 118 BC, and possibly also his son Hiempsal, who was murdered by Jugurtha in the same year. The tomb was excavated in 1915 by a team led by the archaeologist François Bonnell. In the burial chamber, surrounding the twice-burned bones of the two men, they found a number of expensive possessions: a silver drinking horn; a silver bowl; a bronze couch; two gold medallions and, crucially, an elite panoply: a sword with a hilt decorated with gold and copper in a sheath made of sandalwood and riveted leather; a bow; a conical iron helmet embossed with ears; a mail tunic; an oval, thyreos-type shield; and a collection of spears and javelins.

Apart from the spear and javelin heads from the area around Siga, these are the only examples of weapons and armour known to have been found in the whole of Numidia. Their use in the reconstruction of the armament of Numidian cavalry and infantry is problematic, however. This is a manifestly elite panoply, likely to be that of a king, who – or whose ancestors – may have received the individual items as gifts from the leaders of other Mediterranean states. We know, for example, from Livy about the many gifts given to Micipsa’s father, Masinissa, by Scipio and the Roman senate after the defeat of Syphax, which included two complete cavalry armours (Livy, History of Rome, 30.17); a tunic decorated with palm branches (tunica palmata) and a purple toga enriched with golden thread (toga picta), which were normally worn by Roman generals when celebrating a triumph; and the traditional symbols of consular rank: an ivory staff, a ‘curule’ chair and a gold wreath (30.15).

The elite warrior or king entombed in the Soumaâ could also have been given the weapons or armour by his own troops, however, taken from their defeated enemies as war trophies. The Soumaâ panoply could therefore not reflect the equipment of ordinary Numidians at all. This was certainly the opinion of the eminent French historian Gabriel Camps, who asserted that the pointed helmet, oval shield, spear and sword appeared to be the equipment of a warrior from Carthage or the eastern Mediterranean, and not that of a Numidian (Camps 1960: 263). Camps is correct but only to the extent that the Soumaâ panoply does not represent the ‘usual equipment’ of a Numidian, by which we assume he meant the famous light cavalryman, armed only with a hide shield and a clutch of javelins, as he appears on Trajan’s Column or in the description of Livy (35.11.7); or his similarly outfitted infantry counterpart; but to dismiss the Soumaâ panoply as being unrepresentative of any Numidian troops at all is to deny the wealth of epigraphic evidence from Numidia, which shows oval shields, cuirasses, swords, and even perhaps a pointed helmet. Another French historian, André Berthier (1981: 169), has pointed out that the oval shield, sword, spears and helmet found in the tomb are very similar to those represented on a stele excavated from the El Hofra sanctuary in Cirta and now held in the Musée national Cirta de Constantine (inv. 3.C. p. 654). Three other similar steles, all of which show shields and other weapons, have been found at the same location.

Camps’ assertion that the Soumaâ panoply is not Numidian is also contradicted by the highly detailed weapons friezes which ornamented the Numidian ‘royal monuments’ or ‘altars’ situated near Simmithus (modern-day Chimtou) and Kbor Klib in what is now Tunisia. As can be seen in the reconstruction of the Chimtou monument in the nearby Chimtou Museum, these monuments, which dominated their landscapes from hilltops, were built in the style of a Greek or Roman temple. They were symmetrical, rectangular and set upon stepped bases, and punctuated with rows of classical columns. Friezes of highly detailed, veristic, relief representations of cuirasses and shields lined each elevation, possibly accompanied by real helmets and weapons.

The purpose of these monuments is not fully understood, however, nor is the precise date of their construction known. Scholars also disagree about who actually built them. Though they are very large (the Chimtou monument is 12m long and 9m high), they are not temples and do not appear to have had any specific religious or ritual purpose; their function was probably more about the commemoration of recent victory and the assertion of military power. This power was once assumed to be Roman, and therefore imperialistic: they were perhaps built by Caesar after his triumph over Juba I and the absorption of the Numidian kingdom into the Roman Empire. More recently it has been proposed (Ross 2005) that the Kbor Klib monument at least was built to honour Scipio’s defeat of Hannibal at the battle of Zama in 202 BC. Historians have tended, therefore, to define the shields and cuirasses represented on the monuments as ‘war booty’ – trophies collected from a defeated army by the monument’s builder – which in the case of the Kbor Klib monument featured Carthaginian panoplies stripped from Hannibal’s dead. There is certainly contemporary precedent for this: the Sanctuary of Athena at Pergamon (built c.180 BC) was decorated with a frieze depicting captured armour, shields and weapons tumbled together in a heap.

The art historian Ann Kuttner (2013: 228–48) asserts, however, that the Chimtou and Kbor Klib monuments were built by the Numidian kings in the aftermath of the Punic Wars to proclaim the martial prowess of the Numidian army, and the cultural sophistication of an increasingly confident kingdom, to the Numidian people. Of course, visitors would have been equally impressed by the realistic images of Numidian panoplies, and the explicitly ‘Hellenistic’ architectural style. The monuments served as a kind of ‘static victory parade’, like those still held in Moscow’s Red Square, but here, rather than columns of marching troops, tanks and missile launchers, the Numidian kings permanently displayed the best of their own military equipment – cuirasses and shields – formed up in orderly ranks for all to see. Therefore, we can safely assume the armour depicted on the ‘royal monuments’ is Numidian, and in the case of the shields uniquely so, as their flat profile has been found nowhere else.

Kuttner proposes a date of the mid-2nd century BC for the Chimtou and Kbor Klib monuments. They were built at some point from the latter part of the reign of Masinissa (203–148 BC) to early in the reign of his son Micipsa (148–118 BC), and their construction was therefore most immediately intended to commemorate the Numidian role in the total defeat of Carthage in 146 BC. The massive size, hilltop location and almost brazen confidence of these monuments suggests, however, that their true function was to exult in the rise of the Numidian state itself. During the long period of stability under the reigns of Masinissa and Micipsa, Numidia developed economically, culturally and militarily, and became an important ally of Rome. These monuments were built as much to celebrate these aspects of Numidian society as the military victory over Carthage.

Another ancient visual source could provide clues about Numidian arms and armour. In 1939, a substantial stone monument base was excavated near the medieval church of San Omobono in Rome. The statue it once supported has been lost but the base, which now resides in the Musei Capitolini in Rome, carries a very similar array of relief carvings of armour and shields to those found on the Numidian ‘royal monuments’. Yet here, images of greaves and horse armour accompany the round shields and cuirasses. Kuttner (2013: 248–67) suggests that this monument could have been commissioned by the Numidian kings Masinissa or Micipsa to mark their close relationship with Rome, and to promote the Numidian military and Numidian culture in the city of Rome itself.

The writings of ancient authors contain many references to the Numidians; but those same authors were often not concerned with recording the details of soldiers’ arms and appearance. Polybius is, of course, an exception: his description (Histories, 6.19–42) of the organization of the Roman Army and the weapons and equipment carried by Roman legionaries is invaluable, although he unfortunately does not do the same for the Numidians. Also, ancient authors were often not precise about the terms they used for particular pieces of equipment, such as specific shield or sword types, which can be frustrating for modern historians trying to reconstruct the appearance of soldiers from this period.



Berthier, A. (1981). La Numidie: Rome et le Maghreb [Numidia: Rome and the Maghreb]. Paris: Éditions Picard.

Camps, G. (1960). ‘Aux origins de la Bérberie: Massinissa ou les débuts de l’histoire’ [The origins of the Berbers: Massinissa or the beginnings of history], Libyca archéologie-épigraphie VIII: 1–320.

Kuttner, A. (2013). ‘Representing Hellenistic Numidia, in Africa and at Rome’, in Prag, J. & Quinn, J., eds, The Hellenistic West: Rethinking the Ancient Mediterranean, Cambridge: Cambridge University Press: pp. 216–72.

Ross, D. (2005). Kbor Klib and the Battle of Zama. BAR International Series 1399. Oxford: BAR Publishing.

Rostovtzeff, M. (1946). ‘Numidian Horsemen on Canosa Vases’, American Journal of Archaeology 50: 263–67.


Order your copy of MAA 537: The Numidians 300 BC–AD 300 from the website now!

Post Comments

There are no comments on this post yet.

Submit a Comment

You must be logged in as a Bronze, Silver or Gold Osprey member to comment on this post.

Click here to log in.