[184] The crucial passage is of course Tac. _Ann._ xvi. 13 ‘eodem anno dilectus per Galliam Narbonensem Africamque et Asiam habiti sunt supplendis Illyrici legionibus’, which appears to come from the _acta senatus_. But the evidence for imperial control is very strong, and the Senate may merely have been consulted as a matter of courtesy. Tiberius used to bring military questions before the Senate in the same way—‘de legendo vel exauctorando milite ac legionum et auxiliorum descriptione’, Suet. _Tib._ 30—without giving up his prerogative. [185] Later, of course, the franchise became as widely spread in Africa as in Spain. In the first half of the first century, however, this was not yet the case, and the example of Tacfarinas (‘natione Numida, in castris Romanis auxiliaria stipendia meritus’, Tac. _Ann._ ii. 52) shows that auxilia were recruited in this province while it was still completely under senatorial control. [186] xiii. 6860, 6864. Dio, lxxiv. 2, brackets Italy, Spain, Macedonia, and Noricum together, as the ‘civilized’ provinces from which the Praetorians were recruited before the reforms of Severus. [187] If the _cohortes voluntariorum_ be excluded from the reckoning. [188] Strabo, p. 196 κρείττους δ’ ἱππόται ἢ πεζοί, καὶ ἔστι Ῥωμαίοις τῆς ἱππείας ἀρίστη παρὰ τούτων. [189] _Tactica_, 33. [190] See the boastful words of the Gascon Antonius Primus. Tac. _Hist._ iii. 2 ‘Duae tunc Pannonicae ac Moesicae alae perrupere hostem: nunc sedecim alarum coniuncta signa pulsu sonituque et nube ipsa operient ac superfundent oblitos proeliorum equites equosque’. Cf. also Tac. _Ann._ xv. 10 ‘alaris Pannonios, robur equitatus’. [191] Arrian, loc. cit. [192] The list shows that the majority of the oriental regiments were not raised until after 70. [193] Two cohorts numbered III and VII bear this title, for which I can find no explanation. To be distinguished from these is the Cohors I Campanorum voluntaria (vi. 3520), which was stationed first in Dalmatia, then in Pannonia. Apparently it really was originally a regiment of Campanians, since a soldier gives Suessa as his birthplace (iii. 14246¹). The statement of Cichorius that the Dalmatian Cohors I Campanorum is identical with the Pannonian Cohors I Campestris is misleading. On no Pannonian inscription does the title occur otherwise than in the abbreviated form ‘Camp.’ On the other hand, the Roman inscription cited above speaks plainly of the _coh(ortis) primae voluntariae Campanorum in Pannonia Inferiore_. [194] Tac. _Ann._ i. 8. [195] Dio, lv. 31, lvi. 23; Velleius, ii. 111; Suet. _Aug._ 25. Similar regiments may have been raised at a later date, e.g. the cohorts I and II Italica c. R., which seem to form a fresh series and appear only in the East. Can they represent the remainder of the 4,000 oriental freedmen whom Tiberius enrolled to put down the brigands in Sardinia (Tac. _Ann._ ii. 85)? If any survived, the Eastern provinces would have been the natural place to send them to.