^{page 191 note 1} Taylor, L. R., ‘Tiberius’ Ovatio and the Ara Numinis August’, AJPhil lviii (1937), 185–93Google Scholar

^{page 191 note 2} As instituted in c. 14–12 B.C., the official cult of the imperial genius included the sacrifice, probably of a bull, the same type of animal victim as was offered to Jupiter and most gods; in the traditional cult of the genius the offering had been simply pure wine, incense, and flowers; cf. Taylor, L. R., The Divinity of the Roman Emperor (Middle-town, 1931), 192;Google ScholarNock, A. D., Gnomon, viii (1932), 516.Google Scholar On the cult of the genius see in general Otto, W. F., RE vii, I (1910), 1155–70;Google ScholarCesano, L., Diz. Epig. iii (1922) (1962), 449–81;Google ScholarÈtienne, R., Le Culte imperial dans la péninsule ibérique d'Auguste à Dioclétien (Bibl. des écoles frang. d'Athenes et de Rome, 191; Paris, 1958), 307–9.Google Scholar

^{page 191 note 3} Though possessing numen, the emperor was not, however, himself a god like Apollo or Asklepios, to whom one turned in the hour of need; cf. Nock, loc. cit., 518; HThR xlv (1952), 237 ff.;Google ScholarJRS xlvii (1957), 115.Google Scholar For the significance of the step from genius to numen see my article ‘Genius and Numen’, HThR lxii (1969), 356–67.Google Scholar

^{page 191 note 4} On the nature of numen see Pfister, F., RE xvii, 2 (1937), 1273–91;Google ScholarRose, H. J., HThR xxviii (1935), 237–57;Google Scholar xliv (1951). 109–20; also in general Wagenvoort, H., Roman Dynamism: Studies in ancient thought, language, and custom (Oxford, 1947).Google Scholar

^{page 191 note 5} Two of the earliest examples are from Narbo, CIL xii. 4333 = ILS 112: A.D. II, and from Forum Clodii, CIL xi. 3303 = ILS 154: A.D. 18.

^{page 191 note 6} D.N.M.Q_.E. and variants first appear in the early third century; cf. Gundel, H. G., Epigraphica, xv (1953), 128–50.Google Scholar

^{page 191 note 7} Africa is unique in that the commonest form of dedication is numini (never numinibus) Augustorum: elsewhere this form is rare; cf. below, p. 197 n. 1.

^{page 191 note 8} Étienne, op. cit., 309–13.

^{page 191 note 9} On the expansion of the various abbreviations and the general nature of the cult in Britain see Fishwick, , ‘The Imperial Numen in Roman Britain’, JRS lix (1969), 76–91.Google Scholar For abbreviated forms in the Ger-manies and the Gauls see Deininger, J., ‘Numinibus Augustorum’, Germania, xliv (1966), 138–42.Google Scholar

^{page 192 note 1} This is true at least of the period after A.D. 161. Before then Augg. can certainly include past emperors on occasion: see below, p. 196. Augg. [sic] refers to Postumus on a milestone from Margam, Glamorgan: JRS xxvii (1937), 249.Google Scholar

^{page 192 note 2} For Narbonensis see Deininger, J., Die Proiinziallandtage der rÖmischen Kaiserzeit (Ves tigia: Beiträ, ge zur alten Geschichte, Band 6; Miinchen, 1965), 30;Google Scholar for Proconsularis see Fishwick, , Hermes, xcii (1964), 342–63;Google ScholarKotula, T.Eos, lii (1962), 155 ff.;Google Scholar for Baetica see Deininger, J., Madrider Mitteilungen, v (1964), 167–79;Google ScholarFishwick, ,‘The Equestrian Cursus in CIL 2, 3271’, Historia, xix (1970), 96–112.Google Scholar

^{page 192 note 3} Kornemann, E., RE iv, 1 (1900), 803 f., s.v. concilium.Google Scholar

^{page 192 note 4} Krascheninnikoff, M., ‘Ueber die Einführung des provinzialen Kaiserkultus im rom. Westen’, Philologus, liii (1894), 169, 175, 184.Google Scholar

^{page 192 note 5} Tac. Ann. 4. 37; Suet. Aug. 52; cf. Dio 51. 20. 6–7. On the cult at Lugdunum, see now Deininger, Provinziallandtage, 21–4, 99–107

^{page 192 note 6} See Fishwick, , Phoenix, xv (1961), 161–4,Google Scholar and the cogent discussion by Frere, S. S., Britannia (London, 1967), 323 f.Google Scholar

^{page 192 note 7} A cult of Roma and the living ruler is nevertheless very probable; cf. Deininger, Provinziallandtage, 24, with note 4.

^{page 193 note 1} See further below, p. 194.

^{page 193 note 2} Étienne, op. cit., 333 f. with refs. For Imperial virtues in Britain see, e.g., RIB 152, 1466, 2200 (Virtus); 1073, 1778 (Fortuna); 842–4, 1138, 1337, 1731, 1995, 2100 (Victoria); 990, 1127 f., 1978, 2092, JRS xlix (1959), 136,Google Scholar no. 6 (Disciplina). These virtues relate to the living emperor whether the abbreviation Aug. is to be expanded to Aug(usti) (cf. RIB 845: virtus Augusta [sic]) or, as is perhaps likelier in military examples, Aug(usti).

^{page 193 note 3} On the domus divina see Neumann, in RE v (1905), 1527;Google ScholarCalza, in Diz. Epig. ii, 3 (1910), 2062–7.Google Scholar

^{page 193 note 4} I have argued this in detail in HSCP lxxiv (1970), forthcoming.Google Scholar

^{page 194 note 1} Cf. Dittenberger, W., Hermes, xiii (1878), 72;Google ScholarBrandis, C. G., RE ii (1895), 480 f.;Google ScholarKornemann, E., ‘Zur Geschichte der antiken Herrscherkulte’, Klio, i (1901), 105–8.Google Scholar

^{page 194 note 2} Cf. CIL ii. 4226, 4225, 4239, 4217, with my discussion (above, p. 193 n. 4).

^{page 194 note 3} Kornemann, loc. cit. 108 f., dates the erection of the temple to about the turn of the first century. New inscriptions may yet modify the picture here.

^{page 194 note 4} The priest's title in the second century is either sacerdos ad aram Romae et Augustorum or sacerdos ad templum Romae et Augustorum. On Kornemann's very probable interpretation this means that both temple and altar served equally well the cult of Roma and the Augusti.

^{page 194 note 5} Contra Kornemann, loc. cit. 109: ‘… der Altar der Verehrung des oder der jeweils regierenden Kaiser … geweiht worden ist, wahrend der Tempel dem Kult der Roma und der Augusti, d.h. der gewesenen Kaiser gewidmet blieb.’ Yet in the second century Kornemann takes Augusti to denote both past and present rulers in the formula templum Romae et Augustorum: ibid. 109, n. 4 with refs.

^{page 195 note 1} Cf. CIL xii. 1782, 1753 f.; xiv. 41; viii. 8203 (= ILS 4130, 4133–6)

^{page 195 note 2} For pro salute domus divinae [sic] see RIB 91; CIL xiii. 5042. That domus divina denotes living members of the ruling house is even clearer in the rare formula pro salute et incolumitate d.d. (cf. CIL xiii. 520).

^{page 195 note 3} See Calza's lists, Diz. Epig. (above, p. 193 n. 3), 2063.

^{page 196 note 1} Reading statu for situ; cf. Dessau ad loc, n. 2.

^{page 196 note 2} A good example is CIL xiii. 7317 (= ILS 7095): [i]n h.d.d. n[u]min. Aug. hastiferii sive pastor(es) consistentes Kastello Mattiacorum [d]e suo posue[r]unt VIIII Kal. Apriles [I]uliano et Crispino co[s. If the erec tion of this altar on 24 March A.D. 224, reflects the vota commonly offered on this day by the archigallus for the emperor's welfare then numin(ibus) Aug(ustorum) must surely include the living emperor Alexander Severus with past emperors; see further JRS lvii (1967), 151.Google Scholar For the expansion of numin. Aug. see RIB 146 and my discussion in JRS lix (1969), 76–91.Google Scholar

^{page 196 note 3} Étienne, op. cit., 200–4.

^{page 196 note 4} See above, p. 193 n. 4. Similarly by no means im plies a cult of the divi exclusively; cf. Brandis (above, p. 194, n. 1), 481.

^{page 197 note 1} Cf. CIL viii. 958, 8808, 14395, ILAlg i. 3991; ILTun 1501 reads: num(ini) deor(um)(Augiustorum) sac(rum). This can certainly include the living emperor; cf. ILS 9495: Augusto deo cives Rotnani qui Thinissut negoti-antur … (Bir-bu-Rekba).