Published online by Cambridge University Press: 16 May 2013
Recent research reveals that in the so-called Mithraic tauroctony, the god is, in fact, wounding a bull, not killing it. I argue that the scene combines the overall design of evil-eye amulets with the pose of the goddess Nike performing a military sphagion and I suggest that the scene must have been understood by its creator and by some viewers, at least, to offer protective power in this world, as well as salvific assurance about the next, a dual focus that seems to have been especially strong in Mithraism.
I am grateful to Jás Elsner, Thomas Carpenter, Richard Gordon, Ken Lapatin and François Lissarague for their various reactions to and comments on earlier written versions of this essay and for their good advice, even when I did not take it. Enormously helpful, too, were comments, questions and reactions to different versions presented at the University of Geneva in September 2010, at the Central European University in Budapest in October 2010, at the Workshop on Ancient Societies of the University of Chicago in February 2011, at Yale University in January 2012, and at the University of Puget Sound in October 2012. I am thankful to my hosts at each venue.
1 See, e.g., Gordon, R., ‘Authority, salvation and mystery in the mysteries of Mithras’, in Huskinson, J., Beard, M. and Reynolds, J. (eds), Image and Mystery in the Roman World (1988), 45–80Google Scholar, at 49; Zwirn, S. R., ‘The intention of biographical narration on Mithraic cult images’, Word and Image 5 (1989), 2–18, at 2CrossRefGoogle Scholar; and Elsner, J., Art and the Roman Viewer: The Transformation of Art from the Pagan World to Christianity (1995), 211–12Google Scholar. The following abbreviations are used throughout: CIMRM = M. J. Vermaseren, Corpus inscriptionum et monumentorum religionis mithriacae (1956–60), 2 vols; LIMC = Lexicon Iconographicum Mythologiae Classicae; SGG = A. Mastrocinque (ed.), Sylloge gemmarum gnosticarum, Bollettino di Numismatica Monografia 8.2.1 and 2 (2003 and 2008), 2 vols; SMA = C. Bonner, Studies in Magical Amulets, chiefly Graeco-Egyptian, University of Michigan Studies, Humanistic Series 4 (1950); TMM = F. Cumont, Textes et monuments figurés relatifs aux mystères de Mithra (1896–99), 2 vols.
2 H 56 cm, W 86 cm. Fig. 1 is after TMM 2.195, no. 7, fig. 19 = CIMRM 417.
3 H 1.8 m, W 1.76 m. Fig. 2 is after TMM 2 after p. 364 = CIMRM 1083.
4 The small dimensions, narrow entrances and underground locations of most mithraea would, in fact, have also prevented the sacrifice of such a large animal; see Elsner, op. cit. (n. 1), 212 n. 44, and Gordon, R., ‘Small and miniature reproductions of the Mithraic icon: reliefs, pottery, ornaments and gems’, in Martens, M. and de Boe, G. (eds), Roman Mithraism: The Evidence of the Small Finds (2004), 259–83, at 259Google Scholar.
5 See e.g. Beck, R., The Religion of the Mithras Cult in the Roman Empire: Mysteries of the Unconquered Sun (2006), 190–239Google Scholar, who reviews most of the older scholarship. Gordon, R., ‘Panelled complications’, Journal of Mithraic Studies 3 (1980), 200–27Google Scholar, sees an elaborate ‘sacred geography’ in the communal versions that alludes to the planets, the stars and the seasons and seems based on the idea that the Mithraic cave was a model for the whole cosmos.
6 Zwirn, op. cit. (n. 1), 8–10, and Clauss, M., The Roman Cult of Mithras: The God and his Mysteries, trans. Gordon, R. (2001), 74–9Google Scholar, give a good review of this approach. For the feasting scene, see Clauss, ibid., 110–11, and Appendix A to this article.
7 Summarized with bibliography by Clauss, op. cit. (n. 6), 79–85.
8 LSJ offer only the verb ταυροκτονέω (‘to kill a bull’) and the adjective ταυροκτόνος (‘bull-slaying’), both of which are poetic words that show up in Aeschylus and Sophocles and refer to the standard sacrificial slaughter of bulls. The word ταυροβόλιον, on the other hand, offers a good contrast, as it and related words show up frequently in both Greek and Latin inscriptions connected with the worship of the Magna Mater and clearly refer to the ritual killing of a bull by stabbing it in the chest or ribs with a large spear.
9 Palmer, G., ‘Why the shoulder?: A study of the placement of the wound in the Mithraic tauroctony’, in Casadio, G. and Johnston, P. A. (eds), Mystic Cults in Magna Graeca (2009), 314–23Google Scholar, at 314–16 with fig. 17.2, shows how the bull's shoulder blade blocks any access from the shoulder region to the heart and major blood vessels, and he argues that an attack on the shoulder with a short sword or dagger would more likely enrage the animal than kill it. Palmer connects the shoulder wound with one of the mythological vignettes in the side-scenes, in which Mithras holds the foreleg of a bull as he challenges the authority of Sol, and he argues for Egyptian mythological and astronomical influence (the foreleg = the Little Dipper).
10 Palmer, op. cit. (n. 9), 314.
11 See Appendix A for further arguments against the sacrificial interpretation.
12 W. Burkert, Ancient Mystery Cults (1987), 16–17, discussing the votive inscriptions, stresses the practical benefits that Mithraic worshippers thought they received in this world: ‘Mithras is the deus invictus, and soldiers, who knew what victory means, were prominent among his followers’ (p. 17). This combination of protection in this world and salvation in the next seems to have been a regular feature of mystery religions: see Burkert, ibid., 12–29 for a general discussion; C. A. Faraone discusses evidence dating as early as the late Classical period for this feature of the Samothracian mysteries (‘Twisting and turning in the prayer of the Samothracian initiates (Aristophanes Peace 276–79)’, Museum Helveticum 61 (2004), 30–50Google Scholar) and the Orphic (‘Mystery cults and incantations: evidence for Orphic charms in Euripides' Cyclops 646–48?’, Rheinisches Museum 151 (2008), 127–42Google Scholar; and ‘A Socratic leaf-charm for headache (Charmides 155b–157c), Orphic gold leaves and the ancient Greek tradition of leaf amulets’, in Dijkstra, J., Kroesen, J. and Kuiper, Y. (eds), Myths, Martyrs, and Modernity. Studies in the History of Religions in Honour of Jan N. Bremmer (2010), 249–70Google Scholar).
13 It is 7.5 cm in diameter. My discussion follows Gordon, op. cit. (n. 4), 273–5, who illustrates the piece in his fig. 14.
14 ibid., 265–7; the quote is from p. 267.
15 The photograph in Fig. 3 is used with the permission of the Ashmolean Museum.
16 CIMRM 2353, for which see the discussions of Delatte, A., ‘Études sur la magie grecque III: Amulettes mithriaques’, Le Musée Belge 18 (1914), 5–20, at 8–9Google Scholar; and Mastrocinque, A., Studi sul mitraismo: Il mitraismo e la magia (1998), 25–7Google Scholar. Fig. 4 is after Cook, A. B., Zeus: A Study in Ancient Religion 2.1 (1925), 512 fig. 390Google Scholar.
17 There are, however, some eastern features: Zeus holds a wilted ankh-sign in his left hand and Athena holds or supports with her left hand a tall ribbed rhyton. See Delatte, op. cit. (n. 16), 8–9, and Mastrocinque, op. cit. (n. 16), 26–7.
18 Bakazichuch translates the Egyptian phrase ‘son (or “soul”) of darkness’, even though it is often used, somewhat paradoxically, to describe solar deities, here presumably Mithras; see Delatte, op. cit. (n. 16), 10. Mastrocinque, op. cit. (n. 16), 26–7, explains the appropriateness of the name ‘son of darkness’: since in Egypt the sun travels through the underworld every night, it too is associated with darkness. The second word Papapheiris has yet to be fully deciphered. Delatte, op. cit. (n. 16), 10, and Mastrocinque, op. cit. (n. 16), 26–7, review the debate and try to connect it with other solar names of Egyptian orientation, like Phiri and Paphieti.
19 C. A. Faraone, ‘Inscribed Greek thunderstones as house- and body-amulets in Roman imperial times’, Kernos 27 (2014) forthcoming, discusses the widespread tradition of using thunderstones (prehistoric axes) to protect houses and people from lightning strikes and gives a detailed analysis of similarly inscribed axe-heads from Ephesus, Pergamum, Smyrna, Thessaly and Herculaneum.
20 SGG vol. 1 no. 256 (= SGD vol. 2 no. Fl 52). Fig. 5 is after Maffei, A., Gemme antichi figurate 2 (1707), pl. 217.2Google Scholar.
21 Mastrocinque, op. cit. (n. 16), 25–7.
22 Mastrocinque, op. cit. (n. 16), 25–7; Gordon, op. cit. (n. 4), 275; and Ezquerra, J. A., ‘Mithraism and magic’, in Gordon, R. and Simón, F. Marco (eds), Magical Practice in the Latin West: Papers from the International Conference held at the University of Zaragoza, 30 Sept.–1 Oct. 2005 (2009), 519–49CrossRefGoogle Scholar.
23 Appendix B Nos 1, 4, 6–7, 11, 15 and 17 show only Mithras and the animal(s). Nos 1–3, 5–6, 12, 15 and 17 lack the scorpion. On Nos 1–3, 6, 9–12, 14 and 16–17 the animals do not lick the blood. Those with magical inscriptions include Nos 1–2, 4–6, 9–10 and 13.
24 Gordon, op. cit. (n. 4), calls many of these gems ‘amulets’. Mastrocinque, op. cit. (n. 16), 25–7, has suggested that all of them (inscribed and uninscribed alike) were magical and that this Mithraic practice goes back to much older Persian traditions. Ezquerra, op. cit. (n. 22), argues that the inscribed ones are magical and the rest not.
25 Gordon, op. cit. (n. 4), 275.
26 ibid., 278.
27 Dunbabin, K. M. D. and Dickie, M. W., ‘Invidia rumpantur pectora: the iconography of Phthonos/Invidia in Graeco-Roman art’, Jahrbuch für Antike und Christentum 26 (1983), 7–37Google Scholar.
28 The Greek phrase first appears in ch. 18 of the Testament of Solomon, now thought to date to the first or second century a.d. In the text the demon who casts the evil eye (baskania) on everyone says that he is prevented from doing so when the ‘all-suffering eye’ is engraved (18.39). For discussion of the passage see Giannobile, S., ‘Inscrizione su un medaglione di Gela’, JAC 5 (2002), 189–91, at 179–80Google Scholar.
29 Fig. 6 is after Schlumberger, G., ‘Amulettes byzantins anciens destinés a combattre les maléfices et malades’, REG 5 (1892), 73–93CrossRefGoogle Scholar, at 75–6 no. 1. For a recent survey of these Byzantine amulets, see Matantséva, T., ‘Les amulettes byzantines contre le mauvais oeil du Cabinet des Médailles’, JAC 37 (1994), 110–21Google Scholar.
30 Schlumberger, op. cit. (n. 29), no. 10; Engemann, J., ‘Zur Verbreitung magischer Übelabwehr in der nicht-christlichen und christlichen Spätantike’, JAC 18 (1975), 22–48Google Scholar, at 26, discusses the descent over the eye of Zeus' thunderbolt and Heracles' club.
31 SGG 2 no. RoC 4.
32 SGG 1 no. 290. A slightly garbled Homeric verse on the back of this gem suggests that another deity was imagined as joining in the attack (Iliad 5.291): ‘[Athena directed the missile (βέλος)] to the eye near the nose and pierced the white teeth.’
33 Fig. 7 is after Compte du Mesnil du Buisson, Les peintures de la synagogue de Doura-Europos (1939), fig. 96 no. 1 = E. R. Goodenough, Jewish Symbols in the Greco-Roman Period (1953), no. 1066.
34 Fig. 8 is after Millingen, J., ‘Some observations on an antique bas-relief, on which the evil eye, or fascinum, is represented’, Archaeologia 19 (1821), 70–4CrossRefGoogle Scholar, at 74 pl. 6. The stone is roughly 2 ft square, i.e. as large as some of the Mithraic icons.
35 I give Millingen's description (ibid., 70).
36 Fig. 9 is after Levi, D., ‘The evil eye and the lucky hunchback’, in Stillwell, R. (ed.), Antioch on-the-Orontes III: The Excavations 1937–39 (1941), 220–32Google Scholar, at pl. 56 no. 121.
37 For similarly placed leaping dogs on other evil-eye amulets, see Engemann, op. cit. (n. 30), fig. 2 and pls 11c and 12c–d.
38 The Greek text καὶ σύ, ‘the same to you’, is common on evil-eye amulets and aims to turn the jealous and destructive glance back upon its owner.
39 For other scenes of phalli threatening the eye, see Bandinelli, R. B. (ed.), The Buried City at Leptis Magna (1966)Google Scholar, pls 196 (a centaur attacks the top of an eye with his trident and the bottom with his phallus) and 197 (a large phallus with equine haunches attacks the eye).
40 Hinnells, J. R., ‘Reflections on the bull-slaying scene’, in idem (ed.), Mithraic Studies (1975), 2. 290–312Google Scholar, at 298, points out with regard to the Mithraic scene that except for the Sidon relief (CIMRM 1400; illustrated in Clauss, op. cit. (n. 6), 90 fig. 55), the scorpion always turns its tail away from the bull and never aims to inject it with its deadly venom.
41 This seems to be a feature of many ancient Greek protective rituals: to remove or chase away disease and danger, for example, by pharmakos-rituals, fumigation or exorcism, but not to destroy them entirely; see Faraone, C. A., ‘Hipponax Frag. 128W: epic parody or expulsive incantation?’, Classical Antiquity 23 (2004), 209–45CrossRefGoogle Scholar.
42 Gordon, op. cit. (n. 1), 65–6 and Ulansey, D., The Origins of the Mithraic Mysteries: Cosmology and Salvation in the Ancient World (1989), 30–1Google Scholar. For the general evolution of the scene, see LIMC s.v. ‘Nike’ nos 169–72. The Megarian mirror case (14.5 cm in diameter) is in the British Museum; Fig. 10 is after Smith, C., ‘Nike sacrificing a bull’, JHS 7 (1886), 275–85CrossRefGoogle Scholar, pl. D.
43 Clauss, op. cit. (n. 6), 79.
44 Smith, op. cit. (n. 42), and Jameson, M. H., ‘The ritual of the Athena Nike parapet’, in Osborne, R. and Hornblower, S. (eds), Ritual, Finance, Politics. Athenian Democratic Accounts Presented to David Lewis (1994), 307–19Google Scholar, at 320–4.
45 Jameson, M. H., ‘Sacrifice before battle’, in Hanson, V. D. (ed.), Hoplites: The Classical Greek Battle Experience (1991), 197–227CrossRefGoogle Scholar. Most of the discussion has focused on the famous Athenian frieze. Hölscher, T., Griechische Historienbilder des 5. und 4. Jhs. v. Chr. (1973), 68–9Google Scholar, for example, suggested that Nike is sacrificing the bull to the war-dead, and Simon, E., ‘La decorazione architettonica del tempietto di Atena Nike sull'acropoli di Atene’, Museum Patavianum 3 (1985), 271–88Google Scholar, argued that the hero Theseus was the recipient. In a later study Simon, ‘An interpretation of the Nike temple parapet’, in D. Buitron-Oliver (ed.), The Interpretation of Architectural Sculpture in Greece and Rome, Studies in the History of Art 49 (1997), 127–43, at 139, claimed that in a parallel scene on an Apulian volute krater in Ruvo, Iris — in a posture like Nike's — is offering a chthonic sacrifice to ‘the three heroes who surround the offering’ and stressed that ‘it is clear that they are heroic receivers of this chthonic offering’. But the chthonic status of the three figures is not at all clear from the scene on the vase. Hölscher, T., ‘Ritual und Bildsprache zur Deutung der Reliefs and der Brüstung um das Heiligtum der Athena Nike in Athen’, MDAI(A) 112 (1997), 143–65Google Scholar, reviews the various interpretations of the ritual and revives an old interpretation of Stengel, P., Opferbraüche der Griechen (1910), 113–16Google Scholar. Jameson's interpretation, however, makes the best sense of the imagined locale of Nike's actions. Perhaps because they were thinking of the Parthenon frieze as a parallel, both Simon (op. cit. (1997), 140: ‘nobody could imagine … anywhere on the parapet the beginnings of battles’) and Hölscher (op. cit. (1997), 153: ‘Nike ist in der Tat nicht eine Göttin des Kampfes’) insisted that the balustrade scenes are a model of or for celebrations in Athens, but Jameson is right to stress that a parallel scene in the same frieze (the erection of trophies) always takes place on the battlefield. The creative leap of the balustrade sculptor was, then, (i) to imagine that the goddess Nike and not a simple soldier performed these rituals; and (ii) that in the case of the sphagion, she performed it most heroically on the largest and wildest sacrificial animal in the Greek repertoire, an uncastrated bull. Jameson stresses the fact that on nearly all sphagia scenes the animal seems to be uncastrated.
46 The best illustration is a fragment of a late Classical cup in the Cleveland Museum of Art (CVA 26.242), which shows a helmeted soldier pressing his knee into the back of an uncastrated ram and piercing its throat with his military sword. Jameson, op. cit. (n. 45), 218–19, commenting on this cup and two other versions of the scene, notes that the absence of the altar, fire and the usual officiants reveals that this is not a normal sacrifice to be followed by a communal meal, but rather a sphagion. He also points out that ‘the sacrificer may be a mantis, but he is also a member of the fighting force, since he wears a helmet and carries a sword … which he will shortly use in battle’. Flowers, M. A., The Seer in Ancient Greece (2008), 163–4CrossRefGoogle Scholar, points out in addition that the soldier here is beardless (i.e. a younger man) and wears no other armour except his helmet, but he agrees that ‘this scene possibly depicts a seer in the moment of performing sphagion before battle’. There is one other interesting detail: the man is barefoot, a detail that suggests he is performing an important ritual.
47 Palmer, op. cit. (n. 9), 315. See, e.g. LIMC s.v. ‘Nike’, nos 170–1 and 714–18; for the Roman period, see ibid. s.v. ‘Victoria’, nos 258a, 259 and 281, and a frieze from the Basilica Ulpia illustrated in Jameson, op. cit. (n. 44), 323 fig. 18.10.
48 See, e.g. LIMC s.v. ‘Nike’, nos 257 and 258b.
49 See, e.g., Ulansey, op. cit. (n. 42), 30.
50 Saxl, F., Mithras: Typengeschichtliche Untersuchungen (1931)Google Scholar, 14 and Ulansey, op. cit. (n. 42), 30. The latter sees close mythological and astrological similarities between the Greek hero Perseus and Mithras.
51 The blood was over-painted red on the Cleveland cup (see n. 46) to emphasize the bleeding, from which seers sometimes took omens. The animal was not, however, butchered, cooked or eaten; see Jameson, op. cit. (n. 45).
52 See, e.g., Clauss, op. cit. (n. 6), 80–1, who summarizes the interpretation as follows: ‘We can explain why the dog, the serpent and the scorpion are so eagerly pushing their way towards the bull … by assuming that the dying beast is emitting some sort of magical force … imagined to reside in the animal's blood, hide and seed, often too in the tail, as is shown by the corn ears that shoot up from it.’
53 See, e.g., Strabo 1.3.21; Plutarch, Flamininus 20.5; Moralia 168 (= De superstitione 8); cf. idem, Themistocles 31.5. I am grateful to Philippe Borgeaud for reminding me of this fact and providing the references.
54 See e.g. Millingen, op. cit. (n. 34), or Bienkowski, P., ‘Malocchio’, Eranos Vindobonensis (1893), 285–303, at 293–4Google Scholar. Bonner, C. in SMA (1950), 98 n. 39Google Scholar, noted Bienkowski's remarks, but was not convinced.
55 Some scholars think that a statue found in Ostia may have been carved in the first century and then repaired in the second (V 230 = LIMC 98, where Simon is given credit for the date), but in this case Mithras has a different pose — he holds the knife aloft, the bull has not (yet) been stabbed and the only animal present is a snake. Similar in gesture and date are five terracotta figurines from Panticapaeum, in which Mithras is assimilated to Attis (see Gordon apud Clauss, op. cit. (n. 6), 157 in the caption of fig. 114).
56 Elsner, op. cit. (n. 1), 220.
57 The members of this internal audience vary considerably; most scholars, for example, refer to the torchbearers as being a standard part of the composition, but as Hinnells, J. R., ‘The iconography of Cautes and Cautopates’, Journal of Mithraic Studies 2 (1976), 36–67Google Scholar, at 38 shows, there are important regional variations: the torchbearers are standard in the monuments from Central and Eastern Europe, commonly omitted in Italian ones and a feature of a minority of images from Rome itself — and when they do appear on the Italian peninsula it is with Cautes on the left and Cautopates on the right, i.e. the reverse of the position found in Central and Eastern Europe.
58 See the summary in Elsner, op. cit. (n. 1), 220–1.
59 Elsner, op. cit. (n. 1), 220.
60 Gordon, op. cit. (n. 1), 48. For a case study, see ibid., 65–6, where he detects such bricolage already in the earliest of the bull-slaying scenes.
61 Gordon, op. cit. (n. 1), 60–4: ‘such stories were the raw materials from which the Mithras figure was created.’
62 Zwirn, op. cit. (n. 1), 4-5.
63 For example, out of roughly 460 extant examples of the icon (Gordon, op. cit. (n. 4), 261) only eight include the banquet scene = ‘feast scene’ in the table in Gordon, op. cit. (n. 5), 226.
64 Gordon, op. cit. (n. 5), 220 and op. cit. (n. 1), 49, proposes a schematic orientation of episodes to the right and left according to astronomical or astrological codes, while Zwirn, op. cit. (n. 1), 2–4, defends a vague sense of ‘biographical intention’ according to which ‘the sense of sequence occurs, if not an absolute order’, giving as examples several scenes involving Mithras and Sol. He rejects as unlikely, however, the idea of a pre-existing full biography, invoking Kermode's notion of a ‘narrative kernel’ as a source for further and further elaboration.
65 Zwirn, op. cit. (n. 1), 9, drawing on the work of Saxl, op. cit. (n. 50), describes the relationship a bit differently as between ‘theophany (centre) and temporal dimensionality (frame)’, and he suggests that this ‘juxtaposition and contrast of visual modalities reveal the inherent and inseparable “dualism” of the divine nature of Mithras. The relative scale of the central image removes it from the narrative matrix’. Gordon, op. cit. (n. 5), 219–20, also demonstrates that the side images were not understood to be part of a fixed sequence or continuous narrative about Mithras, but rather that they were generally concerned with ritual and the organization of ritual space.
66 Gordon, op. cit. (n. 4), 260.
67 Elsner, op. cit. (n. 1), 216; see also Gordon, op. cit. (n. 1), 69.
68 Elsner, op. cit. (n. 1), 212.
69 ibid.
70 Gordon, op. cit. (n. 1), 60–4.
71 Vermaseren, M. J., Mithras, the Secret God (1963), 102–3Google Scholar, sums up the sacrificial theory as follows: the initiates ate bull's flesh and drank bull's blood, so they could be reborn ‘just as life itself had been created anew from the bull's blood’. This inscription is sometimes quoted in heavily restored form as et nos servasti eternali sanguine fuso (‘and you have saved us by spilling eternal blood’), see e.g. Elsner, op. cit. (n. 1), 359 n. 40, who quotes others; but S. Panciera, ‘Il materiale epigrafico dallo scavo del Mitreo di S. Stefano Rotondo (con un addendum sul verso terminante … sanguine fuso)’, in Bianchi, U. (ed.), Mysteria Mithrae: Atti del Seminario Internazionale su ‘La Specificità Storico-Religiosa dei Misteri di Mitra, con Particolare Riferimento alle Fonti Documentarie di Roma e Ostia’, EPRO 80 (1979), 87–127Google Scholar, at 103, has shown that the first four words are not visible and probably never were.
72 See, e.g., Hinnells, op. cit. (n. 40), 304–5: ‘the clear link between the bull slaying and ritual meal scenes’, but he can only cite a handful of banquet scenes in which the bull-skin appears. See also Zwirn, op. cit. (n. 1), 8–10.
73 Clauss, op. cit. (n. 6), 52, mentions ‘a few’ double-faced cult-images, including the massive one from Heddernheim (depicted above in Fig. 2), which has the feast of Sol and Mithras on the reverse; the scene depicted in his fig. 72, however, shows a bull collapsed on its belly, on which he remarks: ‘the significance of the bull's death is suggested by representing it unflayed.’ It is true that the four other figures in the scene have implements for the feast, but it is unclear, in fact, whether the bull is dead or asleep! His only other example of double-sided relief which includes the bull skin (his fig. 71) is the badly defaced example from Ruckingen; see Kane, J. P., ‘The Mithraic cult-meal in its Greek and Roman environment’, in Hinnells, J. R. (ed.), Mithraic Studies (1975), 2.313–51, at 344–51Google Scholar, for a more detailed discussion.