² ANRW II. 18. i (1986). 3–112. at 34f.

³ Summarized in Britannia ii (1971). 313. The original paper, given at a session of the epigraphy seminar in the Department of Archaeology. University of Durham, seems not to have been published.

⁴ As argued by Fishwick, D.. JRS lix (1969). 76f.Google Scholar

⁵ Weinstock, S.. JRS xxxix (1949). 167Google Scholar ; cf. Pfister, in RE XVII, 2 (1937). 1273–91Google Scholar s.v.

⁶ Rose, H.J., HThR xliv (1951), 109–20Google Scholar ; cf. Wagenvoort, H., ‘Wesenszüge altrömischer Religion’ in ANRW I, 2 (1972), 348–376Google Scholar , at 352–56.

⁷ Rose, H.J., HThR xxviii (1935), 237–257Google Scholar , at 245ff. See further Fishwick, D., HThR lxii (1969), 356–367Google Scholar with ref s.

⁸ Oxford Latin Dictionary, 1202 s.v. 4; cf. 2, 3, 5, 6.

⁹ Cf. RIB 1142 referring to the praesentissimum numen of the god (?Dolichenus).

¹⁰ Degrasssi, Inscrit XIII, 2, p. 115, cf. 401. For the year see Alföldi, A., Die zwei Lorbeerbäume des Augustus, Antiquitas 14 (Bonn, 1973), 42–44.Google Scholar

¹¹ Fishwick, D., The Imperial Cult in the Latin West. Studies in the Ruler Cult of the Western Roman Empire, EPRO 108 (Leiden, 1987), I, 86f.Google Scholar

¹² For further, probably early examples see CIL iv, 3882 (Pompeii); CIL xiii, 389 (Bigerriones, Aquitania); ILS 158 (Gortyn, Crete).

¹³ For the distinction between concomitant and immanent deified concepts see D. Fishwick, ‘Augustan Blessings and Virtues’, ICLW (above, note 11) Vol. II (forthcoming). A basic survey of the subject is provided by Fears, J.R., ‘The Cult of Virtues and Roman Imperial Ideology’, ANRW XII, 17, 2 (1981), 827–948.Google Scholar

¹⁴ I am much indebted for help and advice to Dr P. Flury of the Direktion, Thesaurus Linguae Latinae, also to the wissenschaftlicher Secretar, Herr Eder. No Thesaurus article on numen yet exists, but comprehensive card indices have been compiled of the material relevant to the unpublished part of the alphabet, including complete entries for texts down to about A.D. 200 and extensive excerpts from later authors and the inscriptions. On the basis of the material he has researched, documented on well over a thousand cards, Herr Eder confirms that the attribution of the plural numina to one deity or personality is almost entirely a poetical usage, dating from the time of Virgil, Horace, Tibullus and Ovid; most of the passages are cited by Pfister in RE, op. cit. (note 5). In prose, examination of all references down to Apuleius, along with later references where the entry on the index card provides a context, reveals the following exceptions (the ThLL cannot guarantee that all relevant passages have been caught):

Livy VIII, 6, 1: … vocem Anni spernentis numina Iovis Romani auditam.

For Livy's use of plural for singular here see Löfstedt, E., Syntactica I (Oxford, 1942), 52.Google Scholar n.5. He takes this to be a literary affectation, influenced by contemporary poetical usage. For numen in the singular see Livy, viii 6,5: est caeleste numen; es, magne luppiter.

Apuleius, Met. III, 26: Sed pro luppiter hospitalis et Fidei secreta numina.

This is clearly an example of the intensifying use of the plural in a rhetorical passage; cf. Löfstedt, op. cit. 38fr.

Apuleius, Met. IV, 29: Puellae supplicatur, et in humanis vultibus deae tantae numina placantur …

Here numina evidently matches vultibus, where the plural is used for the singular (cf. V, 11): see Löfstedt 30, 38ff.; further Oxford Latin Dictionary 2123 s.v. 2. The variant reading numen for nomen would attest the regular use of numen in the singular later in the same sentence: … victimis et epulis Veneris absentis nomen (numen Colv) propitiatur …

Servius, Aen. I, 8: namque luno habet numina.

He then goes on to explain that she is Curitis (a Sabine cult title), Lucina, Regina; sunt et alia eius numina (nomina C). Here Servius plainly refers to the different guises or manifestations of luno, under each of which she has a separate title; hence Aeneas' perplexity: quo numine laeso. While Juno has a plurality of numina, each is a different aspect of the goddess identified by a distinct name.

August., De civ. dei VII, 24, p. 305, 11.3–5: unius sint ista multa numina, non tam deae multae quam nomina.

Quite apart from the fact that multa nomina I nomina multa are variant readings, the context shows that numina refers to different manifestations or spheres of activity of Tellus, very much as with luno in the Servius passage above. The editors of the Teubner text understand numina in the sense of facultates vel curae divinae. However, every activity of the goddess has a separate name, each of which men have thought to be a separate deity: Ops, Mater, Magna, Proserpina, Vesta; cf. p. 304, 11.26ff.: deinde adiungit et dicit, Tellurem matrem et nominibus pluribus et cognominibus quod nominarunt, deos existimatos esse complures. This is very different from attributing multiple numina to a single deity.

¹⁵ In addition to Löfstedt, op. cit. (note 14), 38–65 see in general Bell, A.-J., The Latin Dual and Poetic Diction (London, 1923)Google Scholar; Cunningham, M.P., CPh xliv (1949), 1–14.Google Scholar

¹⁶ Fishwick, op. cit. (note 4), 87.

¹⁷ See D. Fishwick, ‘Domus Divina’ in ICLW, op. cit., (note 11) II, forthcoming.

¹⁸ A possible late example is CIL viii, 1357 (= ILS 679: Testur, AD. 316/7: … col(onia) Bisica Lucana devota numinibus | maiestatique eius. The inscription is known only from a nineteenth-century copy of dubious quality: Hoffmann, M., Index gramm. ad Africae…titulos Lat. (Strasbourg, 1878), 105Google Scholar; cf. Dessau's remark ‘… saeculo superiore semel nisi tailor, nee fortasse sine menais, descripta.’ Whether the joint rule of Constantine and Licinius at this date could have influenced the plural numinibus (if this reading is in fact correct) can hardly be said.

¹⁹ Fishwick, op. cit. (note 4), 77.

²⁰ So Birley, op. cit. (note 2), 35, n. 117, inferring the plural numinibus Aug(usti) in RIB 707:81, n. 417, reading n[u]minibus Aug(usti) in RIB 1130; cf. p. 18 ‘… in association with Numina Augusti or Numina Augustorum …’

²¹ For an attempt see Fishwick, op. cit. (note 4), 78–85.