Principate Army – Organization


First Battle of Bedriacum or Cremona, 69 AD


A great body of information on the unit size and organization of the Principate army has been amassed by the patient work of several generations of scholars. The literary sources are often obscure or contradictory on the details of unit structures, but we are fortunate in that much information has been derived from epigraphic, numismatic and papyrological record as well as that of archaeology. Here contemporary evidence, if not overabundant, is explicit and reliable. As a result a fairly coherent picture of the army’s structure has emerged and what follows, then, is the briefest of sketches of the army as it existed in Neronian times.

As an instrument of war the Principate army presented a powerful picture, and there is certainly little about it that a modern infantry soldier would fail to recognize. The professional standing force of a modern size, conscription, military training, institutionalized discipline, weapons factories, administrative and combat staffs, military maps, roads, logistics systems, military hospitals, intelligence services, communications, strategy and tactics, efficient killing technologies, siege machines, rank structures, scheduled promotions, permanent records, personnel files, uniforms, regular pay, and even military pension schemes – to name but a few – had already become part of every day, military life.


Men had a thousand reasons for joining the army, but mainly they were escaping from poor local conditions or looking for what they hoped would provide a regular source of food and income. Unsurprisingly, therefore, the army seems to have been most attractive as a definite career to the poorest citizens. For such men, long underfed and ill-clothed, the legions offered a roof over their head, food in their bellies, and a regular income in coin. Basic military pay was not the road to riches, but there was always the chance of bounties and donatives, and the certainty of a discharge bonus, a rich contrast to civilian unemployment. Army pay certainly did not depend on the weather, taxation, rent, interest payments or fluctuating prices. Overall, a soldier’s life was more secure than that of an itinerant labourer (an unpaid labourer would starve; an unpaid soldier still ate), and he enjoyed a superior status too.

Of course, we must remember the harsher side of such a career. The rewards of army life may have been greater, but so were the risks. A soldier ran the risk of being killed or crippled by battle or disease, but also on an everyday basis was subject to the army’s brutal discipline. And then there was maltreatment, which did not include the routine harshness or the standard Spartan quality of military life. The dividing line between discipline and maltreatment was crossed when officers treated their men with unnecessary severity, when they paid no attention to their welfare, and when they expected fear rather than respect from their men. Such officers firmly believed that you got more out of men by using brutality, than by treating them with patience tempered by firmness. Most of us are familiar with the martinet centurion Give-me-Another, nicknamed because of his habit of beating a soldier’s back until his gnarled vitis – the twisted vine-stick that was his badge of rank – snapped and then shouting for a second and a third.

Such a bully and a beast was common in the army, the general assumption being that soldiers had to be treated roughly so as to toughen them up for fighting, yet to many people in the empire who struggled to survive at subsistence level, the well-fed soldier with his ordered existence in his well-built and clean camp must have seemed comfortably off. Soldiers also shared a comradeship with their fellow soldiers, which was often warm and comforting. And so the legions became permanent units with their own numbers and titles and many were to remain in existence for centuries to come.

Most prominent in the life of the empire was the army, the organization of which began, and almost ended, with the legion. Yet from Augustus onwards the emperor commanded no more than twenty-five legions in total (twenty-eight before the Varian disaster of AD 9), which seems paltry considering the extent of the empire. Legions were probably in the order of 5,000 men strong (all ranks) and composed of Roman citizens, though sickness and death could quickly pare away at this figure. Legionaries were mostly volunteers, drawn initially from Italy (especially the north), but increasingly from the provinces. As the first century progressed, many recruits in the west were coming from the Iberian provinces, Gallia Narbonensis, and Noricum, and in the east from the Greek cities of Macedonia and Asia. Thus, by the end of the century the number of Italians serving in the legions was small. Statistics based on nomenclature and the origins of individuals show that of all the legionaries serving in the period from Augustus to Caius Caligula, some 65 per cent were Italians, while in the period from Claudius to Nero this figure was 48.7 per cent, dropping even further to 21.4 per cent in the period from Vespasianus to Trajan. Thereafter, the contribution of Italians to the manpower of the legions (but not of the Praetorian Guard naturally) was negligible. It must be emphasized, however, that these statistics represent all legionaries in the empire. In reality, there was a dichotomy in recruitment patterns between the western and eastern provinces, with legions in the west drawing upon Gaul, Iberia, and northern Italy, while those stationed in the east very quickly harnessed the local resources of manpower.

Ordinarily a legion consisted of ten cohorts (cohortes), with six centuries (centuriae) of eighty men in each cohort, apart from the first cohort (cohors prima), which from AD 70 or thereabouts was double strength, that is five centuries of 160 men. Commanded by a centurion (centurio) and his second in command (optio),4 a standard-size century (centuria) was divided into ten eight-man subunits (contubernia), each contubernium, mess-group, sharing a tent on campaign and pair of rooms in a barrack block, eating, sleeping and fighting together. Much like small units in today’s regular armies, this state of affairs tended to foster a tight bond between ‘messmates’. There would have been a strong esprit de corps among men built upon the deep concern each had for everyone. In the pressure cooker environment of small combat units where soldiers are forced into close contact with one another, they worked together, they fought together, they shared discomfort and death and victory. This was man-to-man friendship, a gutsy bond. A spirit of military brotherhood would explain why many soldiers (milites) preferred to serve their entire military career in the ranks despite the opportunities for secondment to specialized tasks and for promotion. Nonetheless, a soldier (miles) who performed a special function was excused fatigues, which made him an immunis, although he did not receive any extra pay.

Finally, there was a small force of 120 horsemen (equites legionis) recruited from among the legionaries themselves. These equites acted as messengers, escorts and scouts, and were allocated to specific centuries rather than belonging to a formation of their own. Thus, the inscription on a tombstone from Chester-Deva describes an eques of legio II Adiutrix pia fidelis as belonging to the centuria of Petronius Fidus. Citizen cavalry had probably disappeared after Marius’ reforms, and certainly was not in evidence in Caesar’s legions. However, apart from a distinct reference to 120 cavalry of the legion in Josephus, the equites seem to have been revived as part of the Augustan reforms.


When territory was added to the empire, a garrison had to be put together to serve in its defence. New legions were sometimes raised, but normally these green units were not themselves intended for service in the new province. So when an invasion and permanent occupation of Britannia became a hard possibility under Caius Caligula, two new legions, XV Primigenia and XXII Primigenia, were formed in advance. Their intended rôle was as replacements for experienced legions earmarked to join the invasion force: XV Primigenia to release legio XX from Neuss-Novaesium, and XXII Primigenia to release XIIII Gemina from Mainz-Mogontiacum. The invasion force that eventually sailed for southern Britannia in the summer of AD 43 consisted of XX and XIIII Gemina, along with II Augusta, which had been at Strasbourg-Argentoratum, this camp was now left vacant, and VIIII Hispana from Sisak-Siscia in Pannonia, which may have accompanied the outgoing legate governor, Aulus Plautius, on his journey to take up his new post as the expeditionary commander. It must be said, however, that only II Augusta and XX are actually attested as taking part in the invasion itself, though all four legions are recorded very early in Britannia.

Nevertheless, transfers of legions to different parts of the empire could leave long stretches of frontier virtually undefended, and wholesale transfers became unpopular as legions acquired local links. An extreme case must be that of II Augusta. Part of the invasion army of AD 43, its legatus legionis at the time was in fact the future emperor Vespasianus, this legion was to be stationed in the province for the whole time Britannia was part of the empire. An inscription from near Alexandria, dated AD 194, is of particular interest to us as it records the names of forty-six veterans of legio II Traiana fortis who had just received their honourable discharge and had begun their military service in AD 168. Of the forty-one whose origins are mentioned, thirty-two came from Egypt itself and twenty-four of these state the camp as their place of birth, or more precisely origo castris, ‘of the camp’. It is likely that most of them were illegitimate sons born to soldiers from local women living in the nearby canabae legonis, that is, the extramural civilian settlement associated with the garrison. So it seems that many recruits were the sons of serving soldiers or veterans, and in time these soldiers’ sons became a fertile source of recruits, particularly so as soldiers’ sons did not have to make a major adjustment from a civilian to a military world. With bastard sons following their soldier fathers into the army, the custom developed of sending not an entire legion to deal with emergencies, but detachments drawn from the various legions of a province. As we have seen, in the year AD 69 legionary detachments played a major rôle in the formation of the Vitellian and Flavian armies.

Detachments from legions operating independently or with other detachments were known as vexillationes, named from the square flag, vexillum, which identified them. Until the creation of field armies in the late empire, these vexillationes were the method of providing temporary reinforcements to frontier armies for major campaigns. And so it was that Cnaeus Domitius Corbulo received a vexillatio from legio X Fretensis, then stationed at the Euphrates crossing at Zeugma, during his operations in Armenia. Later he was to take three vexillationes of a thousand men (i.e. two cohorts) from each of his three Syrian legions (III Gallica, VI Ferrata, X Fretensis) to the succour of Caesennius Paetus, whose army was retreating posthaste out of Armenia. Likewise, despite the disaster to legio VIIII Hispana during the Boudican rebellion, no new legion was despatched to Britannia, but a vexillatio of 2,000 legionaries gathered from the Rhine legions.


Under Augustus the rather heterogeneous collection of auxiliary units, auxilia, serving Rome was completely reorganized and given regular status within the new standing army. Trained to the same standards of discipline as the legions, the men were long-service professionals like the legionaries and served in units that were equally permanent. Recruited from a wide range of warlike peoples who lived just within or on the periphery of Roman control, with Gauls, Thracians and Germans in heavy preponderance, the auxilia were freeborn non-citizens (peregrini) who, at least from the time of Claudius, received full Roman citizenship on honourable discharge after completion of their twenty-five years under arms.

Tacitus tells us that the Batavi, on the lower Rhine, paid no taxes at all, but ‘reserved for battle, they are like weapons and armour, only to be used in war’. The Batavi made capital stuff for a soldier, and from Tacitus we hear of eight cohortes and one ala, nearly 5,000 warriors from the tiny region of Batavia serving Rome at any one time. He also remarks of a cohors Sugambrorum under Tiberius, as ‘savage as the enemy in its chanting and clashing of arms’, although fighting far from its Germanic homeland in Thrace. Further information concerning these tribal levies comes from Tacitus’ account of the ruinous civil war. In April AD 69, when Vitellius marched triumphantly into Rome as its new emperor, his army also included thirty-four cohortes ‘grouped according to nationality and type of equipment’.

Take the members of cohors II Tungrorum for instance, who had been originally raised from among the Tungri who inhabited the northeastern fringes of the Arduenna Silva (Ardennes Forest) in Gallia Belgica. Under the Iulio-Claudian emperors it was quite common for such units to be stationed in or near the province where they were first raised. However, the events of the AD 69, with the mutiny of a large proportion of the auxilia serving on the Rhine, would lead to a change in this policy. After that date, though the Roman high command did not abandon local recruiting, it did stop the practice of keeping units with so continuous an ethnic identity close to their homelands.

As expected, by the late first century, units were being kept up to strength by supplements from the province where they were now serving or areas adjacent to it. Such units retained their ethnic identities and names, even if they enlisted new recruits from where they were stationed. The epitaph of Sextus Valerius Genialis tells us that he was a trooper in ala I Thracum, and his three-part name indicated he was a Roman citizen. But it adds that he was a ‘Frisian tribesman’. So, Genialis came from the lower Rhine, served in a Thracian cavalry unit stationed in Britannia and styled himself a Roman. So after the military anarchy of AD 69, auxiliary cohorts were plausibly made up of a great diversity of individuals of all kinds of nationalities. Nonetheless, despite such conflicting backgrounds and cultures, the Roman military system forged these foreign cohorts into cohesive, aggressive fighting units.

Auxiliary cohorts were either 480 strong (quingenaria, ‘five hundred strong’) or, from around AD 70, 800 strong (milliaria, ‘one-thousand strong’). Known as cohortes peditata, these infantry units had six centuries with eighty soldiers to each if they were quingenaria, or if milliaria had ten centuries of eighty soldiers each. As in the legions, a centurion and an optio commanded a century, which was likewise divided in to ten contubernia.

Now to turn to matters concerning mounted auxilia. Cavalry units known as alae (‘wings’, it originally denoted the Latin-Italian allies, the socii, posted on the flanks of a consular army of the Republic) are thought to have consisted of sixteen turmae, each with thirty troopers commanded by a decurio and his second-in-command the duplicarius, if they were quingenaria (512 total), or if milliaria twenty-four turmae (768 total). The later units were rare; Britannia, to cite a single example, had only one in its garrison. Drawn from peoples nurtured in the saddle – Gauls, Germans, Iberians and Thracians were preferred – every horseman of the alae was well mounted, knew how to ride, and was strong enough and skilful enough to make lethal use of his long straight sword, the spatha. The alae provided a fighting arm in which the Romans were not so adept.

Additionally there were mixed foot/horse units, the cohortes equitatae. Their organization is less clear, but usually assumed, following Hyginus, to have six centuries of eighty men and four turmae of thirty troopers if cohors equitata quingenaria (608 total), or ten centuries of eighty men and eight turmae of thirty troopers if cohors equitata milliaria (1,056 total). An inscription, dated to the reign of Tiberius, mentions a praefectus cohortis Ubiorum peditum et equitum, ‘prefect of a cohort of Ubii, foot and horse’, which is probably the earliest example of this type of unit. It may be worth noting here that this Tiberian unit was recruited from the Ubii, a Germanic tribe distinguished for its loyalty to Rome. In Gaul Caesar had employed Germanic horse warriors who could fight in conjunction with foot warriors, operating in pairs.

Organized, disciplined and well trained, the pride of the Roman cavalry were obviously the horsemen of the alae, but more numerous were the horsemen of the cohortes equitatae. Having served for some time as infantrymen before being upgraded and trained as cavalrymen, these troopers were not as highly paid, or as well mounted as their brothers of the alae, but they performed much of the day-today patrolling, policing and escort duties.


In addition, as in earlier times, there were specialists fulfilling roles in which Roman citizens, better utilized as legionaries, were traditionally unskilled. The best-known of these specialists were archers from Syria and slingers from the Baleares, weapon preferences that were solidly rooted in cultural, social and economic differences.

Among the Romans the bow seems never to have been held in much favour, though after the time of Marius it was introduced by Cretans serving Rome. During our period, however, archers were being recruited from amongst experienced people of the eastern provinces. Like slingers, it is possible they were equipped as regular auxiliaries rather than their exotic appearance on Trajan’s Column would indicate (e.g. scene lxx depicts them with high cheekbones and aquiline noses, wearing voluminous flowing skirts that swing round their ankles). Certainly first-century tombstones show archers in the usual off-duty uniform of tunic with sword and dagger belts, cinguli, crossed ‘cowboy’ fashion.

Also likely is the possibility that individual soldiers within any given unit acquired the necessary ability to use bows, rather than simply relying on specialist units. In his military treatise, among other matters, Vegetius includes a recommendation that at least a quarter of all recruits should be trained as bowmen. Yet, despite his sound advice here, Vegetius says it is the self-bow that will be used in training soldiers in the art of archery. It is assumed, therefore, that the standard of archery was obviously not expected to be the same as that provided by specialists units such as cohors I Hamiorum Sagittariorum, who were trained and experienced in precisely the sort of warfare in which the Principate army was decidedly deficient.

Leave a Reply