² G. Calza, ‘II santuario della Magna Mater a Ostia,’ Atti della Pontificia Accademia Romana di Archeologia, Ser. III, Memorie VI, 1947, 183–205 (hereafter Calza). See further Squarciapino, M. Floriani, I culti orientali and Ostia, Leiden, 1962, 1–18Google Scholar (hereafter Squarciapino).

³ Calza 200, no. 2.

⁴ ibid., no. 3. For the date see Degrassi, A., I fasti consolari dell'Impero romano, 1952, 57Google Scholar.

⁵ ibid., no. 5. Degrassi, o.c. 57 f.

⁶ First published by H. Bloch, Notizie degli Scavi di Antichità, 1953, 243–4, no. 7. For the date see Degrassi, o.c. 40.

⁷ It is possible that CIL XIII, 7250 from Oberolm in the same district may also refer to the hastiferi. Zangemeister originally read vican[is et hastife]ris; but after the discovery of CIL XIII, 6740^a, reading … a vicanis veteribus consistentibus Castel. Mattiac …, Zangemeister supplied …[v]ican[or(um) vici vete?]ris, which is the text given in the Corpus. The stone may best be excluded from the present discussion.

⁸ On the term plebs see Waltzing, J. P., Étude historique sur les corporations professionnelles chez les Romains (Louvain, 1895–1900) IV, 286Google Scholar; also in general Kornemann's article in PW IV, 1900, cols. 380–479, s.v. collegium. For patrons see Waltzing 1, 425–6; IV, 373–416.

⁹ ibid. 1, 367; IV, 337–40.

¹⁰ ibid. 1, 385–405; IV, 341–369.

¹¹ On the office of curator see Waltzing 1, 406–13; IV, 326–35. Occasionally a magister of one college is curator of another or may even hold both functions in the same college (cf. CIL VI, 10324). In funerary colleges the curator can be the head officer : Waltzing I, 412.

¹² On the technical uses of consistere see Kornemann in PW IV, 1900, cols 922–6; Ruggiero, De, Diz. Epig. 11, 1910, 620–3Google Scholar; Th.L.L., s.v. consistere.

¹³ Calza 198–202; Meiggs, R., Roman Ostia (Oxford, 1960), 360–5Google Scholar (hereafter Meiggs); Squarciapino 4, 7 f.

¹⁴ Snyder, W. F., ‘Public Anniversaries’, YCS VII, 1940, 225–317Google Scholar. This useful list must be used with caution since the same inscription is occasionally listed twice, once as it first appeared in l'Année Épigraphique and again after inclusion within the Corpus. See also in general Grant, M., Roman Anniversary Issues (Cambridge, 1950), 1–13Google Scholar.

¹⁵ Nock, A. D., ‘The Roman Army and the Roman Religious Year’, HThR XLV, 1952, 229–37Google Scholar.

¹⁶ On the cult of Vulcan see Wissowa in Roscher's Lexicon, s.v. Volcanus, cols. 356–69; Rose, H. J., ‘The Cult of Volcanus at Rome’, JRS xxiii, 1933, 46–63Google Scholar.

¹⁷ See the useful discussion in YCS VII, 1940, 170–1Google Scholar with bibliography; Latte, RRG 360, quoting M. P. Nilsson, Opusc. 1, 368. The long popularity of the festival is shown by its appearance in the Menologia rustica and the calendar of Philocalus.

¹⁸ In A.D. 217 Macrinus abolished the games by which the Volcanalia were celebrated in Rome, an act of hubris that was visited with a series of devastating fires; Dio Cassius LXXVIII, 25, 2; Hieronymus, Chron. 2234. The festival is apparently absent from Col. 11, line 27 of the Feriale Duranum (YCS, l.c.) of ca. A.D. 225–7, whereas the games by which it was celebrated appear in the Fasti of Philocalus (A.D. 354) : CIL I² p. 270.

¹⁹ On the development of the March festival see Lambrechts, P., ‘Les Fêtes “Phrygiennes” de Cybèle et d'Attis’, Bulletin de l'Institut Historique Belge de Rome XXVII, 1952, 141–70Google Scholar; id., Attis : Van Herdersknap Tot God (Brussels, 1962), 42 ff.; Fishwick, , ‘The Cannophori and the March Festival of Magna Mater’, TAPA 97, 1967Google Scholar, forthcoming.

²⁰ Whether this day was connected with the festival is most doubtful : Fishwick, l.c., n. 2.

²¹ The date of D is too fragmentary for analysis. Of the dates recorded on B and C., the latter coincides with no known festival in the Roman calendar; the former is the day of the old Septimontium and one of the four festivals of the Agonium : see Wissowa, RUKR ² 29 f., 399, n. 1, 439; Latte, , RRG 44, 112, 361Google Scholar, n. s. Festivals as old as this can hardly have been of any significance to the hastiferi in the early third century A.D.

²² Not. Curios. urb. reg. XIV; Platner-Ashby, Top. Dict. 325, 546–8; cf. Weiss in PW xx, 1941, col. 892; Radke-Koch in PW xv, 1955,cols 490–94, s.v. Vaticanus 1, 2.

²³ The earliest evidence for the ceremony in Rome shows that the taurobolium was already part of the Magna Mater cult : Matris deum sacra accepit et tauroboliatus est (Vit. Elag. VII, 1). The earliest taurobolic inscription from Rome is still CIL VI, 505, of A.D. 295.

²⁴ On the interpretation of the vires see K. Latte, RRG 364, n. 2; M. P. Nilsson, GGR ² 11, 646, 652–4.

²⁵ Cumont suggested that the shrine was a simple ‘tell’ built along the lines of municipal capitolia, which appear to have been artificial structures where some natural elevation was lacking as at Castel : Rev. d'Hist. et de Litt. Rel. VI, 1901, 97–110Google Scholar with earlier literature. Since the formula vetustate conlapsum restituere is used regularly, if not exclusively, of buildings of some kind (is it appropriate of an earthen ‘tell’?), especially in CIL XIII (cf. 8019, 8201, 7800, 8844, 11820), a more likely explanation is that mons Vaticanus is simply the name of an edifice called after the Vatican mount in Rome. See also Calza 192.

²⁶ Wissowa, RUKR ² 348.

²⁷ The texts are collected by Wissowa, o.c. 350, nn. 2 and 3. On the cult of Ma-Bellona see also Aust in PW III, 1897, cols. 254–7 and Hartmann in PW XIV, 1928, cols 77–91; Cumont, Or. Rel. ⁴ 50 f.; Latte, RRG 281 f. The full testimonia are conveniently grouped in Th. L.L., s.v. Bellona.

²⁸ For a survey of the evidence see Schleiermacher, W., Germania XXII, 1938, 252 f.Google Scholar, to which add RGK XL, 1959Google Scholar, no. 157. For Ma-Bellona in Spain see Bellido, A. Garcia y, Rev. de la Univ. de Madrid V, 20, 1956, 471–84Google Scholar.

²⁹ Calza, l.c.; Meiggs 356–66.

³⁰ Calza 198, no. 1a.

³¹ ibid. 199, no. 1b.

³² Meiggs 364 f.; cf. Squarciapino 6; Bloch, H., Gnomon 37 (1965), 200Google Scholar; Calza 198, with n. 41. The temple of Cybele at the western apex of the campus also dates probably from the latter pait of Hadrian's reign; cf. Meiggs, Appendix IX, p. 548. The shrine of Attis, on the other hand, while its original walling may go back as far as the Julio-Claudian period, seems to belong mainly to the time of Antoninus Pius. On the chronological significance of this see Fishwick, l.c. n. 19 above.

³³ It has been supposed, originally by Becker, , Ann. Ver. Nass. Alterthumskunde VII, 1864, 44–53Google Scholar, that the mons Vaticanus may have served the cults of both Bellona and Magna Mater simultaneously. Two arguments have been used in support of this view. (a) It has been thought that the mons Bellona played a part in the cult of Ma. The main text usually quoted here is Tertull., de Pall. IV, where an equally possible, if not preferable, reading is mentes rather than montes: cum ob diversam affectionem tenebricae vestis et tetrici supra caput velleris in Bellonae montes fugantur (… in Bellonae [sc. templo] mentes …: Th. L.L., s.v. Bellona, col. 1821, line 13). The only other evidence for a mons Bellonae is Gesta apud Zenonphilum 1 (Optat. Milev. 186, 6) : cum incursum pateremur repentinae persecutionis fugivimus in montem Bellonae. This is certainly not an artificial structure of some kind, as at Castel, and may well denote a place of asylum rather than a cult centre, (b) According to Stephanos of Byzantium (fl. fifth or sixth century A.D.) : ἐκαλεῖτο δὲ καὶ ἡ ῾Ρέα Μέα Μᾶ καὶ ταῦρος αὐτῇ ἐθύετο παρὰ Λυδοῖς, ἀφ᾿ ἧς ἡ πόλις. The statement occurs in an etymology of the Lydian Mastaura in which the town is said to owe its name to Ma, here identified with Rhea ( = Cybele). As this late passage refers only to the sacrifice of the bull, it is by itself questionable evidence for the full ceremony of the taurobolium in the worship of Ma. Cumont's claim that the rite made its way into the cult of Magna Mater from that of Ma (above, note 25) is not necessarily proved by this text. Indeed, Cumont had earlier sought its origins in the cult of the Persian Anahita : Rev. Arch, XII, 1888, 132 ffGoogle Scholar. Latte, RRG 355, n. 5, notes that the dedication of the mons Vaticanus to Virtus Bellona at Wiesbaden(!) offers no clue to the problem. On the origin of the taurobolium see further A. D. Nock, CAH XII, 424; A. Rummens, Het Taurobolium (Diss. Louvain, 1943). The rite seems to have been common to several Anatolian cults before it was imported into the cult of Magna Mater in the second half of the second century A.D.

³⁴ Calza 197 f.; Meiggs 362 f.; Squarciapino 6 f., 16.

³⁵ Mommsen's views, formulated when only the Castel inscriptions were known, are set out definitively in Westd. Korr.-bl. VIII, 1889, 19–28Google Scholar. See further Haug in PW VII, 2, 1912, cols. 2511–2512 s.v. hastiferi, with refs.; Lehner, H., ‘Orientalische Mysterienkulte im römischen Rheinland’, Bonner Jahrbücher cxxix, 1924, 45–47Google Scholar.

³⁶ Mommsen, in Hermes XXII, 1887, 558Google Scholar; cf. Lehner, K., Gesammelte Schriften (Mainz, 1836), 280–85Google Scholar.

³⁷ Cf. Maué, , Philologus XLVII, N.F. 1, 1889, 487–513Google Scholar.

³⁸ For a list of dendrophori at Ostia see Calza 190, n. 13. Many of these are dated to the same period as the hastiferi.

³⁹ Germania XL, 1962, 73–84Google Scholar.

⁴⁰ I am indebted to Professor Dr. H. Schoppa of the Städtisches Museum, Wiesbaden, for a photograph of the stone. The reading is certain.

⁴¹ A college of shepherds would be unparalleled. While the outlying districts from Cologne consisted largely of pasture lands (cf. Kisa, A., Westd. Korr.-bl. XIV, 1895, 88Google Scholar), it is impossible that the hastiferi could have been shepherds at the port of Ostia or even at Vienne, a district noted for its fine wines. Not that there is anything in the inscriptions to suggest that they were.

⁴² Inscription H gives the following laterculus : G. Meddignatius Severus cur(ator) bis, L. Levinius Qu(i)etus, T. Vitalinius Peregrinus, Co(n)stantius Marcianus, Crixsius Adnamatus, Giamillius Crescens, Titins Belatullus, Iulius Severus, I[..]cnius Co(n)sta(n)s, [S?]u[..]atius Victor, Tertinius A[b?]rosus, Macrinius (sic) Perpetu(u)s, Atregtius (sic) Cupitianus, Perrius Iustinus Pre[.]o ?, Attonius Asclepius, Ursius Maturus, Statutius Secundinus, Servandius Senu[r?]us. The entry for Perrius Iustinus in the Corpus reads : Perrius Iustinus […] tor. Thanks to the facsimile provided through the courtesy of Dr. Selzer of the Altertumsmuseum, Mainz, it is now clear that the final word reads pre[.]o ?. A possible reading is praeop(tatus); cf. Becker, Keller, Körber, Die römischen Inschriften und Skulpturen im Museum der Stadt Mainz I, no. 82 and IV, no. 40. Most of these names are of Celtic origin : cf. Holder, A., Altceltischer Sprachschatz (Leipzig, 1891)Google Scholar, s.vv. Meddignatius, etc.

⁴³ A metrical inscription from Mainz, CIL XIII, 7070, shows that Iucundus, the freedman pecuarius of M. Terentius, had only the one name. The omission of the praenomen became increasingly common under the Empire until by the third century A.D. it was almost the rule. Where it was retained it tended to be an indication of superior social status; cf. Thylander, H., Étude surl épigraphie latine (Lund, 1952), 77–81Google Scholar.

⁴⁴ Hepding, H., Attis, seine Mythen und sein Kult (Giessen, 1903), 169–72Google Scholar; Graillot, H., Le Culte de Cybèle (Bibl. des écoles franç. d'Atliènes et de Rome CVII : Paris, 1912), 278–80Google Scholar. Among those who have subscribed to Hepding's theory may be noted Hartmann, above, n. 27 and the Th.L.L., s.v. hastiferi.

⁴⁵ Crusius in Maué, above, n. 37 : ‘Zu erwägen bleibt noch die Analogie der βουκόλοι ( = pastores) in orgiastichen Kulten von Kleinasien’. On the βουκόλοι see Quandt, G., De Bachi ab Alexandri aetate in Asia Minore Culto, Dissertationes Philologicae Halenses XXI, 1913, 251–4Google Scholar, with earlier literature; Cumont, F., ‘The Bacchic Inscription in the Metropolitan Museum’, AJA XXXVII, 1933, 247–9Google Scholar; Pickard-Cambridge, A., Dithyramb, Tragedy and Comedy² (Oxford, 1962), 153, 155Google Scholar; cf. 296.

⁴⁶ The phrase is that of Rohde, E., Psyche² (Leipzig/Tübingen, 1898) 11, 15Google Scholar, n. 3.

⁴⁷ For a full list of sources see Hepding 103, n. 2; 206–10.

⁴⁸ A more likely occasion would have been the lavatio; see below p. 154.

⁴⁹ Calza, R., ‘Sculture rinvenute nel Santuario’, Atti della Pontificia Accademia Romana di Archeologia, Ser. III, Memorie VI, 1947, 217Google Scholar, fig. 18.

⁵⁰ Lucretius II, 618–23 : ‘Tympana tenta tonant palmis et cymbala circum concava, raucisonoque minantur cornua cantu et Phrygio stimulat numero cava tibia mentis, telaque praeportant, violenti signa furoris, ingratos animos atque impia pectora volgi conterrere metu quae possint numine divae.’ On the interpretation of tela praeportant see Bailey, , Commentary (Oxford, 1947)Google Scholar II, ad Lucr. 11, 621, who stresses the menacing effect of prae and suggests that tela may refer to the Asiatic ἅρπη or pruning-knife which would be used for the self-emasculation; cf. Leonard, and Smith, , De Rerum, Natura (Madison, 1942), 370–1Google Scholar, who interpret tela in the sense of flint knives.

⁵¹ On this important question see Perret, J. ‘Le Mythe de Cybele’, Rev. Études Lat. 1935, 332–57Google Scholar; P. Boyancé, ‘Une exégèse Stoicienne chez Lucrèce’, ibid. 1941, 147–66.

⁵² Hdt. 11, 168; Plato, Rep. IX, S75B; Thuc. 1, 130.

⁵³ For representations of armed praetorians see Daremberg-Saglio IV, 638–9, s.v. praetoriae cohortes. Figs. 5788, 5790. if correctly identified, are of particular interest in the present context.

⁵⁴ The nature of Herodian's sources is obscure. It was suggested by Baaz, De Herodiani Fontibus et Auctoritate (Diss. Berl., 1909), 15–64, that a common source may have existed for Dio and Herodian and that in the case of Herodian this was written in Latin as far as V, 2. Occasional passages may also have been drawn from other writers; cf. Schultz in PW VIII, 1912, cols. 954–73.

⁵⁵ Waltzing IV, 285–90.

⁵⁶ Hepdin g 169, n. 2.; Graillot 280; cf. 133.

⁵⁷ On the Κορύβαντες see Poland, F., Geschichte des griechischen Vereinswesens (Leipzig, 1909), 45Google Scholar with refs. The various names given to the Korybants at Erythrae (Εὐφρονίειοι, Θαλεῖοι, Ἀνδρεῖοι ) suggest that the direction of these groups was in the hands of individual families. Scanty though the evidence is, the hastiferi have left no traces of a similar organization.

⁵⁸ Cumont, above n. 25; Maué, above n. 37.

⁵⁹ CIL VIII, 8203 : XII, 1782; XIII, 1752; cf. Frag. Iur. Vat. 148 : is qui in Portu pro salute imperatoris sacrum facit ex vaticinatione archigalli a tutelis excusatur.

⁶⁰ Cumont maintained this interpretation explicitly in PW II, 1895, col. 484, s.v. archigallus.

⁶¹ See the lists in Graillot 159, n. 2; for addenda see Oppermann in PW IX, 2, 1934, col. 17, s.v. taurobolium.

⁶² On the connection of the archigallus with the official taurobolium for the safety of the emperor and the empire see Lambrechts, above n. 19, ‘Fêtes Phrygiennes’ 155–9; Attis 27 f.; also Latte, RRG 355.n. 4.

⁶³ Similarly none of the taurobolia performed for private intentions fall within the Magna Mater festival : Graillot 167, n. 4.

⁶⁴ Nilsson, , GGR² II, 646Google Scholar. Latte, RRG 354, n 2 takes Eutyches to be the name of the beast; cf. CIG IV, 7284. For Eutyches as a personal name see CIL VI, 9015; CIL XIV, 763; CIL XIV, 2; Meiggs, Index s.v.; Bloch (above, n. 6), no 55; RIB 143 et passim.

⁶⁵ Nilsson, , GGR² 11, 651Google Scholar, n. 7; cf. 626.

⁶⁶ CRAI 1918, 312 ff.

⁶⁷ ibid. 1919, 257; Gsell-Pflaum, , Inscriptions Latines d'Algérie (Paris, 1922) 1Google Scholar, no. 2071; cf. no. 2996.

⁶⁸ Or. Rel. 51, 226, n. 28. So also Nock, A. D., Conversion (Oxford, 1961), 82Google Scholar.

⁶⁹ Cumont (above, n. 66) 318, n. 3, quoting Strabo XII, 3, 32 (p. 557 C) : cf. XII, 2, 3 (P. 535 C).

⁷⁰ RUKR ² 350, n. 8.

⁷¹ Cf. CIL XIII, 7281 (23rd August); CIL XIII, 6385 (5th November); ILS 3806 (13th June). A variety of dates, none of other than personal significance, occurs on a series of inscriptions dedicated to Ma between A.D. 211 and 265 by freed male and female slaves at Edessa in Macedonia (first published by Papageorgiu in Ἀθηνᾶ XII, 1900, 65 ff.).

⁷² Alföldi, A., Die Troianischen Urahnen der Römer (Basel, 1957), 6Google Scholar; plate X, 9 an d 12; contra Weinstock, S., JRS XLIX, 1959, 171Google Scholar.

⁷³ See Alföldi's discussion of a denarius of Sulla's son Faustus (55 B.C.) and a denarius of Sulla's grandson L. Aemilius Buca (44 B.C.) : ‘Der Machtverheissende Traum des Sulla’, Jahrbuch des Bernischen Historischen Museums XLI/XLII, 1961–1962, 275–87Google Scholar. The earliest evidence for a servus fanaticus of Ma-Bellona in Italy is Plut., , Sulla XVIII, 6Google Scholar.

⁷⁴ Grueber, H. A., Coins of the Roman Republic in the British Museum (London, 1910) 1, 388–92Google Scholar with plate 42 : no. 1, obv.: head of Jupiter (laureate) and rev.: M. Voltei M.f., tetrastyle Doric temple of Jupiter Capitolinus; no. 2, obv.: head of young Hercules and rev.: M. Voltei M.f., Erymanthian boar running right; no. 5, obv.: head of Apollo (laureate) and rev.: M. Voltei M.f., tripod lebes, centre leg entwined with serpent.

⁷⁵ ibid. no. 3, obv.: head of Liber or Bacchus and rev.: M. Voltei M.f., Ceres in Chariot drawn by two serpents; cf. no. 4.

⁷⁶ Grueber, l.c.. In view of the minor importance of Attis at this time (cf. Lambrechts, Attis 42 ff.; JRS LV, 1965, 278 f.Google Scholar) it would be most unlikely that he should be coupled with Cybele on a coin of 81 B.C. As the remains at Ostia show, the only other possibility seems to be Bellona.

⁷⁷ Alföldi, Troianischen Urahnen 6, with plate IX, 2/9 and 1. What seems to be the same divinity appears on an original Sullan frieze now recovered from the Via della Consolazione in Rome : Picard, G. Ch., ‘Les Fouilles de la Via Del Mare’, Mélanges d'Archéol. et d'Histoire LXXXI, 1959, 263–79CrossRefGoogle Scholar with pl. 1, facing p. 272; cf. Alföldi (l.c. above, note 73) 286 with nn. 32 and 33. Here, however, she wears a helmet of similar form to that on the obverse of a triumphal gold piece of Sulla, the bust of which is closely related stylistically to the bust on the denarius of M. Volteius (both may therefore have been struck in the same year, 81 B.C.) : Alföldi, Troianischen Urahnen ⁶ with plate IX, no. 1. Alföldi recognizes Virtus-Bellona on the Sullan aureus also : ibid., with nn. 14 and 15.

⁷⁸ l.c., above n. 72.

⁷⁹ A state-temple of Bellona, aedes Bellonae Pulvinensis (CIL VI, 490, 2232–4) with a vicus Bellonae (CIL VI, 2235; cf. 30851) did not exist until the early third century A.D.; its site is unknown. See further Guarducci, M., ‘Il santuario di Bellona e il Circo Flaminio in un epigramma greco del Basso Imperio,’ Bull. Com. LXXIII, 1949–1950, 55–76Google Scholar. An aedes Bellonae Rufiliae (ILS 4181^a) was probably a private foundation; cf. Wissowa, RUKR ² 349 f.

⁸⁰ Slaves, particularly female slaves, played an important part in the highly organized cult of Ma in both Cappadocia and Pontic Comana; cf. Hartmann in PW (above, n. 27), cols. 84 f. The devotion of women to Virtus Bellona in the Rhineland is very marked; cf. CIL XIII, 3637 (from Trier); 6666 (from Mainz); 8193 (from Cologne); BRGK XXVII, 1937Google Scholar, no. 78 (from Altrip). One of the patrons of the Ostian hastiferi (C) was a woman.

⁸¹ The Corpus gives no indication of the rough date of CIL VI, 30851. Similarly, neither Dessau nor his sources provide any clue to the date of ILS 3804.

⁸² Wissowa, RUKR ² 350; Cumont CRAI, 1918, 320; Or. Rel. ⁴ 51. See also Squarciapino 8; Garcia y Bellido (above n. 28), 471 ff.

⁸³ The term lectica above in ILS 3804 (from Aquae Calidae in Mauretania) is interesting in this respect; cf. Cumont CRAI 1918, 320, n. 5, quoting Aug., Civ. Dei. II, 4 : Berecynthiae Matri omnium ante cuius lecticam die solemni lavationis talia cantitabantur; Servius, ad Aen. VI, 68 : simulacra quae portantur in lecticis … apud Aegyptios et Carthaginienses.

⁸⁴ Köves, Th., ‘Zum Empfang der Magna Mater in Rom’, Historia XII, 1963, 321–47Google Scholar.

⁸⁵ See Hartmann in PW (above, note 27), cols. 77–91.

⁸⁶ The passages are listed by Wissowa, RUKR ² 319, nn. 4 and 7.

⁸⁷ Lambrechts, ‘Fêtes Phrygiennes’ 160–161.

⁸⁸ The day of the Hilaria was certainly not in existence at the end of Hadrian's reign; cf. Arrian, Tactica XXXIII, 4 (A. G. Roos), composed A.D. 136–7; Lambrechts, l.c. 167 f.

⁸⁹ Cumont, Or. Rel ⁴ 51; cf. Wissowa, RUKR ² 350, n. 7, quoting ILS 3368, for cistiferi pedisequarii and two pedisequariae in an African inscription to Liber Pater. The relief is reproduced in Daremberg-Saglio 1, 686, s.v. Bellona. The sacred cista which he bore is at his feet; cf. Cumont o.c., plate 11, no. 1.

⁹⁰ See refs. in n. 19 above. On Lambrecht's convincing interpretation the March cycle remained a festival of mourning until the time of Antoninus Pius.

⁹¹ Th.L.L., s.v. As Ma was assimilated to Bellona, it would be appropriate for her guild to have a Latin rather than a Greek title.

⁹² Cf. Daremberg-Saglio I, 685–6, s.v. Bellona, quoting Winckelmann, Monum. ined. 1, 36, n. 29, who recognized Bellona in a fragmentary relief of an armed female figure standing on a cippus.

⁹³ That Bellona should be associated with the spear is likely enough in view of her close connection with Mars, who was certainly represented with a holy spear : Wissowa, RUKR ² 144. It may also be relevant that before the old temple of Mars Bellona stood the columna bellica over which the pater patratus with special formula threw a lance into ‘foreign territory’ to symbolize the outbreak of war : Ovid, Fasti VI, 207. The ceremony was still kept up in the time of Marcus Aurelius (Dio L, 4,5; LXXI, 33,3). Cf. Wissowa, RUKR ² 151 f; Bömer, F., P. Ovidius Naso : Die Fasten (Heidelberg, 1958), 348Google Scholarad VI, 199 ff.

⁹⁴ Th. L.L., s.v. hastifer.

⁹⁵ Breglia, L., La prima fase della coniazione romana dell'argento (Rome, 1952), 67 ffGoogle Scholar. The figure appears on the reverse of a head of Ares/Mars. Cf. Robinson, E. S. G., Num. Chron. IV, 1964, 37–64Google Scholar (at 56).

⁹⁶ For a late Hadrianic date see in general the arguments of Meiggs, 364 f. The earliest dendrophorus from the area (CIL XIV, 97) is dated A.D. 139, the same year in which we have the earliest evidence for the presence of the hastiferi. Time would surely have been very short, though not impossibly so, for Antoninus Pius to have installed a large-scale cult of Magna Mater with its various appurtenances, if both dendrophori and hastiferi were already organized in guilds as early as 139, the first year of his reign. It is true that Hadrian's enthusiasm for the metroic cult is nowhere very apparent in the West, but he was περιερϒότατος in matters of religion (Dio LXIX, 22, 1) and certainly concerned for Ostia; cf. Beaujeu, J., La Religion romaine à l' Apogée de l' Empire (Paris, 1955), 1, 270–2Google Scholar. For Hadrian's interest in the Cybele cult in Asia Minor see Graillot 148; Macridy-Bey, Th. and Picard, Ch., ‘Attis d'un Métroon (?) de Cyzique’, Bull. Corr. Hellen. XLV, 1921, 453–5Google Scholar; Walton, F. R. ‘Religious Thought in the Age of Hadrian’, Numen IV, 1957, 168 fGoogle Scholar.

⁹⁷ Meiggs 174, n. 1, with refs.

⁹⁸ Alföldi, A., ‘Hasta-Summa Imperii’, AJA LXIII, 1959Google Scholar, 5, 7, n. 75; pl. 5, nos. 4–5; pl. 9, no. 3.

⁹⁹ Inscription F is now lost, so that it is impossible to date the presence of the hastiferi at Vienne. The Corpus gives no indication of the date of E and no secure indication of date is given by the letter-forms, to judge from the photograph kindly provided by the Römisch-Germanisches Museum at Cologne. The only clue to the date of CIL VI, 2232, recording the dedication of a hasta by a devotee of Bellona (above, p. 156), is Henzen's note ad. loc.: ara marmorea litteris optimis.

¹⁰⁰ Above, n. 19. The Lavatio is first attested in the Menologium Colotianum of ca. A.D. 50; CIL VIII, 2305.

¹⁰¹ The natalicium of the dendrophori was on the same day as that of their imperial founder (1st August). The earliest epigraphical evidence for the college in Italy is A.D. 79 (CIL X, 7) and at Rome A.D. 97 (CIL VI, 641). For a dendrophorus of Julio-Claudian date from Vienne see AÉ 1956, 61; cf. above, p. 147.

¹⁰² cf. above, p. 147.

¹⁰³ While there is nothing to connect the hastiferi directly with Claudius, a link with the official institution of the March festival in his reign is possibly to be seen at Ostia in traces of a Julio-Claudian structure beside the later temple of Cybele; perhaps also in the earlier walling of Attis' shrine, which may go back to the same period; cf. Calza 203; Meiggs 364, 366 (noting that Claudius was a frequent visitor to Ostia); Squarciapino 2, 6, 17.

¹⁰⁴ Nilsson, M.P., The Dionysiac Mysteries of the Hellenistic and Roman Age (Lund, 1957), 62 ffGoogle Scholar.

¹⁰⁵ I am indebted to Professor T. B. L. Webster for an illuminating letter on the point. So also Cumont (above, n. 45), 249.

¹⁰⁶ The dance seems nevertheless to have been an important element in the cult of Magna Mater. In addition to the wild dances of the fanatici of Bellona the frenzied capering of the Galli is known to have been accompanied by clashing cymbals and music on the flute; cf. Picard, Ch., Numen IV, 1957, 21Google Scholar, n. 73. Reference may be made to the Galli in an inscription from Rome attesting sodales ballatores Cybelae : CIL VI, 2265. Squarciapino, 8, does indeed suggest that the hastiferi may originally have performed a war-dance corresponding to that of the Corybants; cf. Meiggs 360. There seems no evidence to support this.