The Northern Frontier

I created this map many years ago to see how the different types of Roman army unit listed in the Notitia Dignatatum were deployed along Hadrian's Wall. Although there is much discussion and debate about the exact location of certain Roman place names, the map is intended to give a general overview of the northern frontier at the end of the 4th century. It shows that:

  • The alae were deployed on the lower ground to the east and west of the Cheviot Hills

  • The cohortes mainly occupied the higher ground between the coastal lowlands. They also garrisoned a number of forts along the Irish Sea

  • The numeri were located at key strategic points in the hinterland

  • The equites and the only legion were held in reserve, some distance behind the frontier.

This deployment raises a number of interesting questions.

  • Did the alae patrol beyond the Wall? Their positioning, so close to the frontier, suggests that they did.

  • Is the location of the cohortes a legacy of an earlier deployment? These established regiments, like the alae, seem to be based predominantly along the line of the Wall or at sites overlooking the Solway Firth and the Irish Sea.

  • How large were the units of numeri? Did the size of each group vary? These units seem to have been stationed at key strategic points behind the frontier. Thus, they guarded the navigable stretch of the Wear (as far as its tidal reach at Chester-le-Street), the Stainmore pass, and the navigable waters of the Derwent (down to the coast at Workington). If their role was mainly defensive then their complements may have been correspondingly small.

  • What was expected of the single legion and the cavalry regiments which were located eighty or more miles behind the frontier?

What is clear is that the strategic deployment of the Roman army along the norther frontier was closely aligned to the local topography. This relationship with the landscape probably enabled the Roman high command to minimise the number of soldiers required to defend the frontier.

The use of numeri provided flexibility and may have been a way of mitigating the problems caused by a lack of manpower. For example, the Pennine forts were probably garrisoned by small bands of soldiers. That numeri were used for this purpose my indicate that the cohortes were now much smaller in size than in the earlier empire when they may have provided the detachments needed to undertake such duties. By controlling the main communication routes the numeri could obstruct any advance and delay any retreat, a function which was also provided by the Wall and its vallum.

The luxury the Romans had in the north was that they could defend in depth. There was no significant city beyond York, and less to defend than in the agriculturally prosperous south (see Smith). In this context they appear to have constructed three layers of defence. First, the alae and cohortes provided a frontier presence along the Wall to deal with and/or deter any minor incursions. Next, the numeri guarded the major estuaries and controlled trans Pennine communication by occupying a number of key strategic points. Finally, eighty miles to the rear, the main Roman strength lay in wait, ready if necessary to engage an opponent who would hopefully be worn down and vulnerable after days travelling through hostile territory.

The deployment also throws some insight onto the nature of the threat. AHM Jones (P679-686) speculates that a border legion in Late Antiquity might muster three thousand men, whilst a cavalry unit was comprised of five hundred riders. Treadgold argues that border legions and those attached to the field army were probably about the same size - roughly one thousand men (P49). Southern and Dixon seem to take a more nuanced view, suggesting that unit size cannot be judged with certainty but may have been a consequence of history and practicality. In this context, two cavalry regiments were located at or near to York and a third (not shown on the map) was based just a little further south. The proposition is that together with the legion, these regiments represented the main Roman strength and unlike those deployed elsewhere, were probably maintained at something approaching their full complement. Depending upon whose hypothesis you follow, the Romans may have been able to assemble an army of 2,500 - 4,500 men, but even the lower figure may be too high when the difficulty of recruiting and the problems elsewhere in the empire are taken into account.

Distance bought time and would have enabled the regiments in the third tier to deploy effectively, first to block any further incursion and then to destroy or mop up the invading force. The preponderance of cavalry suggests that the Romans needed to counter the mobility of their opponents and their possible dispersal over a wide area.

The only documented account we have of a successful barbarian incursion, is Ammianus' description of the invasion which occurred in 367 AD. It seems that the invaders were well aware of the Roman defensive strategy, something which they had gleaned from the scouts sent to patrol north of the Wall. They also appear to have struck at the most disadvantageous time, probably in the spring when many soldiers were still on a winter furlough. During the hostilities which followed it seems that the unprepared Roman forces in the north were scattered and their commander Fullofaudes marooned, possibly at York. Meanwhile, Nectaridus the commander of the coastal region was killed, although it is unclear whether he died fighting against the Picts or the Saxons. Order was only restored when a number of elite palatine regiments were sent from the continent to reinforce the British army. This force was similar in size, a few thousand men, to the third layer of defence evident in the Notitia. Clearly an enormous army wasn't required to defeat the invaders. Perhaps the three-layered system of defence evident in the Notitia, was a consequence of the disastrous experience of 367 AD.

To conclude, the system as described was self- sufficient and didn't require the support of a Comes Britanniarum. Reinforcement, if needed, could come from the continent, as occurred on a number of occasions. The deployment represents a minimalist approach and is a testament to the skilful planning of the Roman high command. No unit would have been larger than it needed to be, and each had a particular role.

These are my thoughts. What do you think? Are there alternative interpretations?


A.H.M Jones; The Later Roman Empire 284-602; Blackwell and Mott Ltd 1964

Alex Smith; The Rural Settlement of Roman England;

Ammianus Marcellinus; The Later Roman Empire; Penguin Books 1986

Southern and Dixon; The Late Roman Army; Routledge 2000

Warren Treadgold; Byzantium and its Army; Stanford University Press; 1995