[196] Cf. the previous passages with Macrobius, _Sat._ i. 11, 32 ‘Caesar Augustus in Germania et Illyrico cohortes libertinorum complures legit, quas voluntarias appellavit’. [197] There were at least thirty-two cohortes voluntariorum, among which VI is the highest number borne by a _cohors ingenuorum_ (xiii. 8314, 8315). At about this point the supply of free-born recruits probably gave out, since cohors VIII does not bear this designation. [198] Seeck suggests that, as the Western legions were recruited mainly in Italy at the beginning of the first century, these cohorts represent the contribution of the enfranchised communities in the provinces. _Rheinisches Museum_, xlviii. 611. This, however, is not only opposed to the literary evidence, but inscriptions also show us soldiers of Italian origin. Cf. iii. 9782 (Cemenelium) and _A. E._ 1909. 130 (Placentia). [199] The original mutineers in Dalmatia seem to have been militia rather than regulars, cf. Dio, lv. 29 καί τινα καὶ σφεῖς δύναμιν πέμψαι κελευσθέντες, συνῆλθόν τε ἐπὶ τούτῳ καὶ τὴν ἡλικίαν σφῶν ἀνθοῦσαν εἶδον, but the phraseology of Velleius (ii. 110) leaves little doubt that regular auxiliaries were also implicated. [200] Tac. _Ann._ ii. 17. [201] Ib. iii. 42. [202] Ib. iv. 73. [203] Ib. xii. 27. [204] ‘Immissa cohorte Thraecum,’ Tac. _Hist._ i. 68. [205] ‘Praemissis Gallorum Lusitanorumque et Britannorum cohortibus,’ Ib. i. 70. The regiments referred to are probably the Cohors III Britannorum and the Cohorts VI and VII Lusitanorum which appear in Raetia at a later date. Cf. D. lxxiii, _I. G. R. R._ iii. 56. [206] Early inscriptions mention Cohorts VII and VIII Breucorum (xiii. 7801, 8313, 8693), IV Delmatarum (Ib. 7507, 7508, 7509), I Pannoniorum (Ib. 7510, 7511, 7582), I Ituraeorum (Ib. 7040, 7041, 7042, 7043), I Sagittariorum (Ib. 7512, 7513, 7514), and Silauciensium (Ib. 8593). The last title is unintelligible, and possibly corrupt. The soldier mentioned is a certain Tib. Iulius Sdebdas from Tyre, so the regiment clearly came from the East, and its title should perhaps read Seleuciensium—i.e. from Seleucia. [207] Tac. _Ann._ ii. 52. [208] Josephus, _Ant._ xx, 6, 1. _Bell. Iud._ ii. 12, 5. The small garrisons maintained in the _provinciae inermes_ seem also to have been of local origin; cf. the _Ligurum cohors, vetus loci auxilium_ stationed in the Alpes Maritimae, Tac. _Hist._ ii. 14. See also D. xx and xxvi for the composition of the garrison of Sardinia. [209] Tac. _Ann._ iv. 46. See above, p. 19. [210] ‘Flavus aucta stipendia, torquem et coronam aliaque militaria dona memorat,’ Tac. _Ann._ ii. 9. [211] ‘Non tulit ala Picentiana gaudium insultantis vulgi, spretisque Sancti promissis aut minis Mogontiacum abeunt,’ Tac. _Hist._ iv. 62. I follow the diplomata in using the form Picentiana—not Picentina, which Tacitus preferred. [212] D. ii. [213] iii. 3271.