^{page 184 note 1} Good remarks in Smith, R. E., Service in the post-Marian Roman Army (Manchester, 1958)Google Scholar, esp. chapter IV (on recruiting), and VI (on Augustus' reforms). There is more than most of us care to know in the 561 pages of Harmand, J., L'armée et le soldat à Rome de 107 à 50 avant notre ère (Paris, 1967).Google Scholar Part II, chapter II (229–322) has much information on conditions of service for the common soldier.

^{page 185 note 1} For legionary names and movements, as for many other items, an invaluable aid is Parker, H. M. D., The Roman Legions (Oxford, 1928Google Scholar, reprinted with additional bibliography, 1958). Much more extensive is the account of E. Ritterling in RE, s.v. ‘legio’ (cols. 1211–1829).

^{page 186 note 1} A valuable and entertaining book gives much information on non-military activities of the soldiers and on the camp townships (canabae): MacMullen, R., Soldier and Civilian in the Later Roman Empire (Cambridge, Mass., 1963).Google Scholar

^{page 186 note 2} All references to Dessau are to his text of, or comments on, a specific inscription in ILS. All references to Mommsen are to his comments on a specific inscription in CIL, unless otherwise noted.

^{page 186 note 3} Often the deceased is represented in the inscription as speaking to the viator.

^{page 187 note 1} Suetonius says ex Transalpinis (Iul. 24. 2)Google Scholar, and does not specify the date of enrolment, but 51 B.c. and a more extensive recruiting area are probable.

^{page 187 note 2} Parker, , 57Google Scholar; cf. 62, 66–7, 266.

^{page 187 note 3} Dessau thought it possible; J. Schmidt (CIL) merely suggested a Julio-Claudian date. Cf. Mommsen, 's comment: ‘Hic si miles non Caesaris, id quod admitti potest, …’ in EE v (1884), 214.Google Scholar

^{page 187 note 4} Cf. Schmidt, , CIL 8. S.1, p. 1428.Google Scholar

^{page 188 note 1} For the problem of the date of enrolment of legio XXGoogle Scholar, see Parker, , 79–87Google Scholar; for the cognomen Valeria from its commander Messala, Parker, , 261Google Scholar; for victrix for service in Britain under Nero, Parker, , 133Google Scholar; in general, Parker, , 271.Google Scholar

^{page 188 note 2} These were minor decorations reserved for milites gregarii, the various ratings, and centurions. For a good brief note, cf. Jones, H. S., Companion to Roman History (Oxford, 1912), 204–5.Google Scholar Fig. 36 on 205 shows the tombstone of a centurion wearing armillae on both arms, and torques suspended from his shoulders.

^{page 188 note 3} Buecheler, F., Carmina Latino epigraphica I (1895), 243.Google Scholar A trochaic septenarius.

^{page 189 note 1} CIL 3. 6824 (ILS 2237); 6828 (two other veterans of V Gallica). Cf. Ritterling, cols. 1571–2.

^{page 189 note 2} Drawn from Dessau's notes.

^{page 189 note 3} Dessau on ILS 2267 (n. 1) and 2260 (n. 1)Google Scholar suggests the first meaning. On the latter inscription he follows Mommsen's note in CIL (3. 2037).Google ScholarILS 2321Google Scholar, however, has a miles missicius veteranus. The brief section in TLL (1137. 77–1138. 17) cites 15 in scriptions (including all mentioned here), I papyrus, and some later references, but only two authors of this period. In neither of these is there any certainty of the distinction in meaning (Martial iii. 91. 1; Suet. Ner. 48. 2).Google Scholar Cf. Parker, , 213.Google Scholar

^{page 189 note 4} Parker, , 237–8.Google Scholar

^{page 190 note 1} The ablative of duration of time, unusual in literature, is regular in inscriptions, as is shown frequently by the full form or in the case of diebus by a four-letter abbreviation.

^{page 190 note 2} Alternatively hie sepultus est.

^{page 190 note 3} For legio VII see Ritterling, 1614–28. For the military situation in Dalmatia in A.d. 9–56, see Syme, R. in CAH x, 373, 804ff.Google Scholar

^{page 190 note 4} Parker, , 212–14.Google Scholar

^{page 190 note 5} MacMullen, chapter III, ‘Makers and Builders’.

^{page 192 note 1} For Reate, cf. Mommsen, , CIL 9, pp. 438–9.Google Scholar The other inscriptions are those of C. Octavius Pastor, signifer of legio VIIII Hispana (CIL 9. 4685)Google Scholar, L. Valerius Valens, veteran of the same legion (9. 4689), C. Carantius Verecundus, and C. Cominius, veterans of the cohortes praetoriae (9. 4682 and 4683).Google Scholar

^{page 193 note 1} I assume Lucius as the elder, since he has their father's praenomen. For missicius see p. 189 n. 3. Lucius as curator at the headquarters of the legion is conjectural, but probably correct.

^{page 193 note 2} Also in Barrow, R. H., A Selection of Latin Inscriptions (Oxford, 1934)Google Scholar, no. 84. Dessau and Barrow have magister (i.e. Leontius), but surely the order of the names indicates magistri. Mommsen has the plural; cf. also his ‘Die roemischen Lagerstaedte’, Hermes vii (1873), 299–326.Google Scholar

^{page 193 note 3} Mommsen put the names of the three officials in the ablative (i.e done during their magistracy), and assumed that the building was erected by the veterans and Roman citizens. I adopt Barrow's completion of the abbreviations. Also Tucca Aelius (Barrow) rather than Tuccius Aelianus (Mommsen)—reversal of nomen and cognomen is common in literature of this period.

^{page 194 note 1} These comments follow the brief but informative section in MacMullen, 119–25.

^{page 194 note 2} For Troesmis, cf. Mommsen, , CIL 3. 1Google Scholar, p. 145; 3. z, p. 999, and A. Betz in RE, s.v. ‘Troesmis’ (where this inscription is assigned to the time of Pius).

^{page 195 note 1} We do not know what the building was. MacMullen says it was a temple—an inference from the temple built near Silustria in Moesia as a dedication to Jupiter for the health of Pius and Verus by two men in the canabae of legio XI Claudia (CIL 3.Google Scholar 7474; ILS 2475).Google Scholar

^{page 195 note 2} The form Aurellius is often found in the name of Caracalla. I have followed the transcription of Kolbe, H.-G., Bonn.Jbb. clxi (1961), 88–91, pl. 18.Google Scholar

^{page 195 note 3} I consider Titus a praenomen despite its unusual place.

^{page 195 note 4} The transcription in AE of lunii no(vis) must be a misprint. Kolbe is probably correct in Iunii despite the fact that the fourth letter is I longa (this shows clearly in his pl. 18).

^{page 195 note 5} I have added II which is not given in the transcript in AE (presumably following the original publication in Russian). For the consuls of these two years, cf. Liebenam, W., Fasti consulares imperil Romani (Bonn, 1909).Google Scholar

^{page 196 note 1} For a brief but useful discussion of these officers, cf. Reynolds, P. K. B., The Vigiles of Imperial Rome (Oxford, 1926), 72–93.Google Scholar Although he is giving data for the vigiles, most items also apply to the legions. Parker's seventh chapter, ‘The Officers of the Legion’, is mostly concerned with the more important posts.

^{page 196 note 2} Cf. Ritterling, col. 1450. This inscription now places the return of II adiutrix in 217Google Scholar instead of 218 or 219.

^{page 196 note 3} The suggestion of Kolbe, 90.