Late Roman Unit Sizes

[Nathan Ross]

Two cohorts, each 6 centuries.

Ah, right.

Jones (LRE 2, pp681-682) reached his estimate of 3000 men for 'the rumps of the frontier legions' by assuming that each had two 1000-man (ie two cohort) vexillations drawn from it, which had turned into field army units at a subsequent point. That would make 960 men a better estimate for the comitatensis legions... although Zosimus's figure would suggest 1200...

If we took one of the estimated sizes for the so-called 'Perge legion' (1172 plus 300 cavalry), this could round up to about 1200 men, or 1500 with additional mounted troops, which might fit one of those Lejjun/Udrah/Palmyra/Boppard - sized fortresses quite well.

Meanwhile, I was looking for other late fortresses with excavated barracks or internal buildings and came across these three smaller forts:

   

All three, I think, are Diocletianic or thereabouts. El-Kab appears to have six barrack rooms in one range and three more along each side, with perhaps another pair of three rooms each at the bottom - either 12 or 18 rooms total. By size (only 0.2 hectares) this fort should house 60 men by principiate men-per-hectare standards. That would put either three or six men in each room, assuming each room was a barrack. Alternatively, with more men per room the garrison could have been larger.

Khan al Hallabat in Syria is a similar size (0.22 hectares), and was probably the Veriaraca of the ND, making it the base of the Ala Nova Diocletiana. Qasr Bashir in Jordan was only slightly larger, and had stabling for up to c.69 horses.

Abu Sha'ar (Aby Sha'ar on the plan above) is 0.5 hectares, but held 54 barrack rooms in nine groups of six, plus additional rooms around the walls. By area it could have held 150 men - but that would only be 2-3 men per room. Perhaps the men were sharing their rooms with additional 'family members', or some of the rooms were used for other purposes? Alternatively, with a full eight men per barrack room we would have 432 men in the garrison.

Dionysias fort in Egypt is 0.75 hectares, large enough to house 225 men. But again we have around the same number of barrack rooms (53 this time, it looks like), and once again they are divided, more or less, into groups of six. Four or five men in each room would make up the garrison strength. Alternatively, the garrison could have been the same size as the one at Abu Sha'ar...

So once again the results are fairly inconclusive, except to underline the relatively small sizes of late Roman frontier units: 60 or so men for an ala, perhaps. Maybe a late ala was the size of two old-style turmae, just as a legion (in some cases) was the size of two old-style cohorts? This might be disproved by the multiplication of ranks on Egyptian papyri, of course: some of the alae appear to feature numerous decurions. These plans also seem to indicate a subunit size divisible by six.