281 Arrian, Tactica, 44.
282 The Κελτοὶ ἱππῆς, however, which are mentioned in the Ectaxis, 2, are probably, as Ritterling (Wiener Studien, xxiv. 127-40) suggests, cavalry of the Cohors Germanorum M. E.
283 The nearest case I know of is xiii. 3177, where we have the order signifer-centurio-tribunus, but this is in one of the Cohortes civium Romanorum, which occupy an exceptional position.
284 Rangordnung, pp. 112-15, 122-30.
285 ix. 996, x. 4862.
286 Rangordnung, pp. 57, 72. He also considers that at this date the centurions and decurions of the auxiliary regiments were drawn from the ranks of the legions, a suggestion which has already been discussed. See above, p. 38.
288 Tac. Hist. iv. 12 ‘(Batavorum) cohortibus quas vetere instituto nobilissimi popularium regebant’.
289 Ib. iv. 18. Its praefectus Claudius Labeo, ‘oppidano certamine aemulus Civili,’ was clearly a Batavian.
290 Ib. iv. 16.
291 Ib. iii. 35; ii. 14; iv. 55. To the same class of officers belonged ‘Chumstinctius et Avectius tribuni ex civitate Nerviorum’, who played an important part in one of the campaigns of Drusus. Epit. Livy, cxxxxi.
292 v. 7003 is an example of a career of this kind which dated from the reign of Claudius.
293 ix. 2564; A. E. 1902. 41.
294 Cf. iii. 1918 ‘I. O. M. Sulpicius Calvio c(enturio) leg(ionis) I Min(erviae) praepositus coh(ortis) I Belgarum’.
295 Cf. viii. 18007 ‘… M. Annius Valens leg(ionis) III Aug(ustae) praepositus n(umeri) Palmyrenorum’; xiii. 6526 ‘… M. Octavius Severus (centurio) leg(ionis) VIII Aug(ustae) Praeposit(us) Brit(tonum)’. The office of praefectus numeri does, however, occur; iii. 1149. See above, p. 87.
296 xi. 5669 ‘C. Camurio C. f. Lem(onia) Clementi … praef(ecto) coh(ortis) VII Raet(orum) equit(atae), trib(uno) mil(itum) coh(ortis) II Ulpiae Petraeor(um) miliar(iae) equit(atae), praef(ecto) alae Petrianae …’.
297 D. xlvii, li, lix, lx, lxi, lxvi, lxx.
298 In a list of over two hundred and fifty praefecti whose place of origin is known I have not come across one from either of these provinces. But it is of course impossible to be sure that such a list is even as complete as the existing evidence permits.
299 For the military qualities of the Gauls in the fourth century cf. Ammianus Marcellinus, xv. 12, xix. 6.
300 A. E. 1899. 177.
301 Such men probably stood a better chance than the Greeks. See my article on the Caristanii of Pisidian Antioch in J. R. S. iii.
302 Possibly curator, the Greek being ἐπιμελητής.
303 A. E. 1911. 161. A son, or other relative, who erected the inscription was praefectus cohortis II Hispanorum equitatae, C. R., tribunus cohortis III Ulpiae Petraeorum.
304 It is true that we do not know where the Cohors III Thracum Syriaca was stationed, but the other units in the series of four bearing this title all appear in the East. The Cohors II Hispanorum, in which his son served, is probably that mentioned on an inscription from Ancyra. iii. 6760.
305 D. xlviii and cviii.
306 Cf. Tacitus’s remarks in the Annals, xiii. 35, with the account in Dio Cassius, lxxv. 11-13, of the siege of Hatra by Septimius Severus, especially τῶν μὲν Εὐρωπαίων, τῶν δυναμένων τι κατεργάσασθαι and the promise of one of the officers ἐάν γε αὐτῷ δώσῃ πεντακοσίους καὶ πεντήκοντα μόνους τῶν Εὐρωπαίων στρατιωτῶν, ἄνευ τοῦ τῶν ἄλλων κινδύνου τὴν πόλιν ἐξαιρήσειν.
307 This point has been well made by Dessau in Hermes, 1910. The evidence does, however, suggest that an unusually large proportion of Africans obtained commands during the reign of Septimius Severus.
308 This is put very strongly on pages 133 and 134 of the Rangordnung, ‘die Italiker und die Weströmer sind von der militia equestris ausgeschlossen.’
309 The first name is that of the province in which the praefectus was stationed. His place of origin is placed last.
310 Its best justification is the solidarity of the Empire in the fourth century, which appears so markedly in the pages of Ammianus, and exercised so powerful an influence over the minds of the barbarian invaders.