Construction of auxiliary forts in ad 287-8 (Diocletian’s third year) at el - Kantarah (Sile) on the Suez Canal and Deir el-Gebrawi (Hierakon) in Upper Egypt, for ala I Thracum Mauretana and cohors I Augusta praetoriana Lusitanorum respectively, is attested by inscriptions from those sites ( CIL III 13578 and CIL III 22- see CIL III 6626 n.22), perhaps reflecting a general reform of Egypt’s defences in or about that year. In addition, there are more substantial archaeological remains from other sites, notably Dionysias (Qasr Qarun) in the Fayum, and Abu Shaar on the Red Sea coast of the Eastern Desert. The former has a largely mud-brick rectangular wall circuit c.94 m by c.80 m, with projecting square corner towers and semi-circular bastions at gateways and on curtain walls (Schwartz and Wild 1950; Schwartz et al. 1969). The internal plan, with barracks and a central command/administrative block, has been recovered, although its chronological development is uncertain, and the excavators’ views that it represented a single construction of the reign of Diocletian have been challenged by Carrie (1974: 839-840) who (not entirely convincingly) redates it to the period of Palmyrene occupation of Egypt in ad 270-2, on the basis of architectural parallels. It is very likely that this fort was the one occupied for much of the fourth century by the cavalry unit the ala VPraelectorum, known very well, along with its commander Abinnaeus, from an archive of documents found at Philadelphia (discussed below - Bell et al. 1962; Schwartz et al. 1969: 2).
The fort at Abu Shaar was excavated in 1990-1 by a group from the University of Delaware, directed by Steven Sidebotham (Sidebotham 1994). It had a roughly rectangular wall circuit c.77.5 m by 64 m, constructed primarily of local cobbles, with baked - and unbaked brick and gypsum employed in the towers and gates. Like the other late Roman fortifications discussed, Abu Shaar had numerous projecting towers, at the corners and on the curtain walls, and catapult balls were found in one of them, emphasizing their importance as artillery platforms at this period. The internal arrangements were broadly similar to those of Dionysias. A fragmentary inscription found in the course of excavation suggests a tetrarchic date (c. ad 309-11) for the primary military occupation of the site and provides evidence for the garrison as an ala] nova Maximi[ana, an auxiliary cavalry unit, and Sidebotham characterizes the installation as a satellite of the legionary fortress at Luxor (Sidebotham 1994: 143, 157-8; Bagnall and Sheridan 1994: 159-63). Neither the fort nor its garrison can be identified in the Notitia Dignitatum.
World History
