The most comprehensive source for the late Roman army in Egypt, or, indeed, anywhere in the Roman empire at that time, is the Notitia Dignitatum. This document, preserved in the medieval manuscript tradition, lists military units (and Roman officials) throughout the empire on a region-by-region and province-by-province basis. In the case of military units it indicates the type and title of the unit and its place of deployment. The document as a whole was probably compiled c. ad 429, with the eastern (Oriens) sections, including Egypt, being finalized somewhat earlier, probably c. ad 395 (Carrie 1986: 451). However, the changes which saw the emergence of the ‘‘late Roman army’’ began amid the military and political crises of the mid-third century AD, nearly two centuries before the final compilation of the Notitia, and intensified in the reigns of emperors like Diocletian and Constantine. In order to view the changes that took place through that period, it is necessary to extrapolate backwards from the Notitia Dignitatum with the aid of contemporary (but less comprehensive) documents such as the Diocletianic papyri from Panopolis (P. Beatty Panop. - Skeat 1964). The Notitia divides the garrison of Egypt into two sections, troops under the command of the comes limitis Aegypti in Aegyptus (essentially Lower Egypt - Or. XXVII) and those under the dux Thebaidos (Upper Egypt - Or. XXXI). In addition to the regular garrison set out in these sections of the document, other units, from regional and central field armies (comitatus), might be deployed in Egypt for a particular campaign or on a longer term basis. For example, documentary sources attest to the presence of comitatenses (troops from the field armies) at Oxyrhynchos and Antinoe (Carrie 1986: 458-9).
Some of the unit types appear familiar from the Principate, for example, there are units termed legions and lower status units called alae and cohorts, clearly the notional (at least) descendants of the legionary and auxiliary units of earlier Roman history. However, there are also new unit types, notably those titled equites, high-status cavalry units, reflecting the increased importance of that element of the Roman army from the third century ad. However, it is clear from careful reading of the documentary evidence as well as from the sizes of military bases that most, if not all, late Roman units were much smaller than their namesakes of the Principate. Legionary units typically may have had 1000 men at most, and legionary detachments (vexillations) recorded in the Diocletianic Panopolis papyri probably were about 500 strong. Duncan-Jones’ (1990) interpretation of the same document reveals auxiliary cohorts and alae of c.160 and 120 men respectively. Thus, while the Notitia lists significantly more units in Egypt in late antiquity than in the Principate (12 ‘‘legions,’’ nine equites, 32 alae, 19 cohorts and three other units, compared to three legions and 12 auxiliary units under Augustus, or a single legion with 16 auxiliary units in the mid-second century AD), it is quite possible that the late Roman army of Egypt was of similar size to that of the Principate, perhaps c.20,000 men strong. This estimate, based on the work of Duncan-Jones (1990), contrasts with the views of Skeat (1964, the principal edition of the Panopolis papyri) and A. H. M. Jones (1964: 679-84), who assumed that late antique units were comparable in size to those of the Principate and came up with an estimate of c.64,000 late Roman soldiers in Egypt.
As noted above, the Notitia Dignitatum provides us with a picture of the late Roman army as it evolved over a century and more. Some units have imperial titles such as Theodosiana, Arcadiana, Valentiniana and (equites) Honoriani that show they were raised in the later fourth century ad. However, it is clear that the reign of Diocletian (ad 284-305) was a time of significant reform, when, in many respects, the late Roman army of Egypt was ‘‘created.’’ Not only do we have literary evidence for Diocletian’s presence in Egypt and his personal interest in its defence, along with archaeological evidence of military construction in, or close to, his reign (both discussed below), but also the titles of several of the units in Egypt in the Notitia Dignitatum suggest tetrarchic origins. For example, there are multiple elements of the two legions III Diocletiana (ND Or. XXXI.37) and I Maximiana (ND Or. XXVIn.18; XXXI.31, 33, 38), named after Diocletian and his tetrarchic colleague, and some auxiliary units titled Herculia (ND Or. XXXI 50, 54), after Maximian’s patron deity Hercules.
With regard to the general character of the late Roman army in Egypt, it is clear from the unit titles that infantry, both legionary and auxiliary, remained an important component. However, cavalry units were more numerous than in the Principate, and, despite the relatively small size of these late-antique cavalry formations, it seems likely that overall numbers were somewhat greater too. There is also some evidence of local specialization, especially in Upper Egypt. Of particular note is the large number of units described as equites sagittarii indigenae on both sides of the Nile on and around the Koptos bend, at sites such as Tentyra (Dendera), Koptos itself, Diospolis Parva (Hiw), Latopolis (Esna), and Maximianopolis/Kaine (Qena). Presumably such troops would have been particularly well suited to controlling nomadic peoples such as Blemmyes (but see Barnard 2005 on the problem of Blemmye identity) and may well have been recruited from them. There are also a number of specialized camel units (alae dromedariorum), including one at Qena, again, of particular value in policing desert areas. There is only one unit of heavy armored cavalry ( catafractarii) attested in the Notitia, in the Thebaid, (although a papyrus also refers to such a unit of the comitatus at Antinoe - Carrie 1986: 458-9), probably reflecting the lack of suitable enemies for such troops in Egypt, in contrast to the Persian frontier.
Many auxiliary units bear ethnic names, although typically (again) these reflected the units’ origins rather than necessarily continued recruitment from those peoples. A number of titles reveal at least a nominal connection with units already in Egypt in the Principate. These include the ala Apriana (ND Or. XXVHI.32) ala veterana Gallorum (XXVHI.28), cohors II Ituraeorum (XXVHI.44), cohors II Thracum (XXVni.45) in Aegyptus and the cohors scutata civium Romanorum (XXX.59) and cohors I Apamenorum (XXXI.60) in Upper Egypt. Other ethnic titles suggest recruitment from among Germanic peoples including the cohors IX Alamannorum (ND Or. XXVHI.63), cohors VIIFrancorum (XXVHI.67) and cohors IVIuthugorum (XXXI.43). Probably these were raised and moved east during the reigns of fourth-century emperors like Constantine and Constantius II, although presumably they began recruiting more locally after their arrival and eventually lost any ethnic distinctiveness they once may have possessed.
On the surface, the deployment of the Roman garrison of Egypt in late antiquity appears more balanced and dispersed than that seen through much of the Principate, with the latter’s heavy concentration of troops (legionary and auxiliary) in and around Alexandria. The Notitia Dignitatum reveals a substantial concentration of all troop types on the southern frontier, a significant presence in the strategically important Cairo (Babylon-Memphis) area, and dispersed bases throughout the Nile Delta, along the major route to Syro-Palestine, down the Nile valley and in outlying locations such as the Fayum and the western oases. This pattern of dispersal is confirmed by the evidence of excavated late-antique military bases (discussed below). However, this apparent change from the earlier situation may be more illusory than real. As noted above, the army of the Principate was actually more dispersed than the purely literary evidence suggests, and the later imperial situation depicted in the Notitia Dignitatum is probably a formal recognition and crystallization of the de facto dispersed deployment of the Principate (I am grateful to Valerie Maxfield for this suggestion). In some cases sites named as military bases in the Notitia Dignitatum also provide evidence for military deployments in the Principate, even beyond the major concentrations of troops like Alexandria, Babylon, and the Thebaid. For example, Latopolis (Esna) is named as the base of a unit of horse archers in the Notitia ( Or. XXXI.28), but there is also evidence for the presence of a part-mounted detachment there in the second century ad (Bagnall 1975).
World History
