No CrossRef data available.
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 22 July 2025
Scholars have regularly debated the question of how the Galli, priests of Magna Mater/Cybele, fit into the Roman social milieu. Several have argued (citing Domitian’s legislation) that membership of the Galli was restricted to foreign citizens, whilst others have argued that the chief priests—the Archigalli—were Roman citizens, while the ‘lower’ Galli were non-citizens, thus separating the Galli and the Archigalli within the Cybele cult. These views remain prevalent in modern discussions on the cult, and have not undergone significant scrutiny or analysis. By assessing these views as well as the existing material and literary evidence for the Galli, this article argues that the Archigalli and the Galli were indistinct in terms of behaviour and affiliation. Moreover, this article uses archaeological and literary evidence to suggest that the Galli most likely included Roman citizens among their members, contrasting with the prevailing view of them as foreign residents in Rome.
1 J. North, ‘Gender and cult in the Roman West: Mithras, Isis, Attis’, in E. Hemelrijk and G. Woolf (edd.), Women and the Roman City in the Latin West (Leiden and Boston, 2013), 109–27, at 110–11.
2 A.T. Fear, ‘Cybele and Carmona: a reassessment’, AEA 63 (1996), 95–108, at 97.
3 L.E. Roller, In Search of God the Mother: The Cult of Anatolian Cybele (Berkeley, 1999), 301.
4 W. Roscoe, ‘Priests of the goddess: gender transgression in ancient religion’, HR 35 (1996), 195–230, at 201–2.
5 C.O. Tommasi, ‘Cross-dressing as discourse and symbol in late antique religion and literature’, in D. Campanile, F. Carlà-Uhink, M. Facella (edd.), TransAntiquity: Cross-Dressing and Transgender Dynamics in the Ancient World (London and New York, 2017), 121–34, at 124.
6 J.A. Latham, ‘Roman rhetoric, Metroac representation: texts, artifacts, and the cult of Magna Mater in Rome and Ostia’, MAAR 59/60 (2015), 51–80, at 54.
7 S. Hales, ‘Looking for eunuchs: the Galli and Attis in Roman art’, in S. Tougher (ed.), Eunuchs in Antiquity and Beyond (London / Oakville, CT / Swansea, 2002), 87–102, at 89.
8 W. Spickermann, ‘Women and the cult of Magna Mater in the western provinces’, in E. Hemelrijk and G. Woolf (edd.), Women and the Roman City in the Latin West (Leiden and Boston, 2013), 147–68, at 149.
9 C. Mowat, ‘Don’t be a drag, just be a priest: the clothing and identity of the Galli of Cybele in the Roman Republic and Empire’, Gender & History 33 (2021), 296–313, at 299.
10 M.L. Meijer, Identifying Borrowings between Eastern Mediterranean Cults: A Methodology Based on a Comparison of Cultic Practices for Ištar and Meter (Amsterdam, 2021), 271–2.
11 Meijer (n. 10), 249–50, 253.
12 Hales (n. 7), 88.
13 J.F. Ubina, ‘Magna Mater, Cybele and Attis in Roman Spain’, in E.N. Lane (ed.), Cybele, Attis and Related Cults: Essays in Memory of M.J. Vermaseren (Leiden / New York / Cologne, 1996), 405–33, at 423–4.
14 K. Summers, ‘Lucretius’ Roman Cybele’, in E.N. Lane (ed.), Cybele, Attis and Related Cults: Essays in Memory of M.J. Vermaseren (Leiden / New York / Cologne, 1996), 337–65, at 356–7 suggests that, with regard to Varro’s Eumenides (Varro, Sat. Men. 150 Astbury; when the protagonist encounters a group of Galli in the temple), the Latin may read aedituus (‘priest’) instead of aedilis, since Roman aediles could not, as Roman citizens could, be cultists—perhaps also implying the same for the Galli?
15 Dom. 7.
16 67.2.3.
17 18.4.5.
18 For example Fear (n. 2), 47; Spickermann (n. 8), 157; W. Roscoe, ‘Precursors of Islamic male homosexualities’, in W. Roscoe and S.O. Murray (edd.), Islamic Homosexualities: Culture, History, and Literature (New York and London, 1997), 55–86, at 69–70.
19 Julius Obsequens, Prodigiorum liber 44a; Roller (n. 3), 292.
20 Val. Max. 7.7.6.
21 Transl. E. Cary, The Roman Antiquities of Dionysius of Halicarnassus, 7 vols. (Cambridge, MA and London, 1950), 1.367.
22 Apul. Met. 8.26.3.
23 Apul. Met. 8.26.3 sed postquam non ceruam pro uirgine sed asinum pro homine succidaneum uidere, nare detorta magistrum suum uarie cauillantur.
24 Transl. P.G. Walsh, Apuleius: The Golden Ass (Oxford, 1995), 154.
25 Transl. F.R. Walton, Diodorus of Sicily: The Library of History. Books XXXIII–XL (Cambridge, MA and London, 1967); cf. Plut. Vit. Mar. 17.5–6.
26 E.g. Apul. Met. 8.27.1–3; Varro, Sat. Men. 121 Astbury.
27 Tert. Apol. 25.5. Transl. T.R. Glover, Tertullian: Apology. De spectaculis (Cambridge, MA and London, 1931), 137.
28 Firm. Mat. Math. 3.5.24. Translation is my own.
29 Firm. Mat. Math. 3.6.22. Translation is my own.
30 Serv. on Verg. Aen. 9.115. Translation is my own.
31 As seen in numerous literary texts such as Anth. Pal. 6.234 and Apul. Met. 8.27.4–6.
32 Serv. on Verg. Aen. 9.115 et effecit ut cultores sui uiriles sibi partes amputarent, qui archigalli appellantur; Firm. Mat. Math. 3.5.24, 3.6.22 Archigallos facient et qui uirilia propriis sibi amputent minibus…faciet eunuchos aut abscisos archigallos aut hermafroditos.
33 Noted by Meijer (n. 10), 271, quoting L. Dubosson-Sbriglione, Le culte de la Mère des dieux dans l’Empire romain (Stuttgart, 2018), 177 n. 212; many of the above are also noted by her as well.
34 Indeed, Meijer (n. 10), 255 argues that the later Christian authors likely confused the Galli and the Archigalli given their iconographic parallels.
35 Many examples of potential Galli exist throughout the Graeco-Roman world; however, since several of these have no direct evidence proclaiming them as such (e.g. no overt symbol of office, i.e. the occabus, and/or epigraphic inscription declaring themselves as Galli), we will refrain from using them here. Good case studies include two silver figurines from Antalya that may depict eunuchs: see Roller (n. 3), 105, 253; T. Şare, ‘An archaic ivory figurine from a tumulus near Elmali: cultural hybridization and a new Anatolian style’, Hesperia 79 (2010), 53–78, at 58–9, 65–71.
36 M.J. Vermaseren, Corpus Cultus Cybelae Attidisque (CCCA), I–Asia Minor (Leiden / New York / Copenhagen / Cologne, 1987), 94–5; M.J. Vermaseren, Cybele and Attis: The Myth and the Cult (London, 1977), 29; Meijer (n. 10), 229–30; B. Obrador-Cursach and I.X. Adiego, ‘A Greek reading of the “Pisidian” inscription N 30’, Kadmos 56 (2017), 173–6, at 175; L.E. Roller, ‘The ideology of the eunuch priest’, Gender & History 9 (1997), 542–99, at 544–5; Roller (n. 3), 332–3; F. van Straten, ‘Images of gods and men in a changing society: self-identity in Hellenistic religion’, in A. Bullock, E.S. Gruen, A.A. Long and A. Stewart (edd.), Images and Ideologies: Self-Definition in the Hellenistic World (Berkeley / Los Angeles / Oxford, 1994), 248–65, at 255–6; E. Metropoulou, ‘The goddess Cybele in funerary banquets and with an equestrian hero’, in E.N. Lane (ed.), Cybele, Attis and Related Cults: Essays in Memory of M.J. Vermaseren (Leiden / New York / Cologne, 1996), 135–65, at 153; Roscoe (n. 4), 203.
37 Vermaseren (n. 36 [1987]), 228; Meijer (n. 10), 229–30, 262; Obrador-Cursach and Adiego (n. 36), 174–5; L.E. Roller, ‘The priests of the Mother: gender and place’, in C.C. Mattusch, A.A. Donohue and A. Brauer (ed.), Common Ground Archaeology, Art, Science, and Humanities. Proceedings of the XVIth International Congress of Classical Archaeology (Oxford, 2003), 52–5, at 53–4.
38 Obrador-Cursach and Adiego (n. 36), 173–6; Meijer (n. 10), 229–30, 262; A. Brixhe, T. Drew-Bear and D. Kaya, ‘Nouveaux monuments de Pisidie’, Kadmos 26 (1987), 122–70.
39 The relief of Menneas is fragmented, but follows the same iconographic format as the other two.
40 For example, Brixhe, Drew-Bear and Kaya (n. 38), 147 interpret the figure on the Papas relief as a woman.
41 See Meijer (n. 10), 261–2; Vermaseren (n. 36 [1977]), 29; Vermaseren (n. 36 [1987]), 94; Obrador-Cursach and Adiego (n. 36), 175; Roller (n. 3), 544; Brixhe, Drew-Bear and Kaya (n. 38), 147–8; and C.G. Cavanagh, ‘“ἡμιγύναικα”: the Galli of Cybele, Latian/Anatolian archaeological evidence and comparative analyses’ (Diss., Queens University Belfast, 2022), 88–9.
42 However, as noted in Cavanagh (n. 41), 75–6, the pyxis may also be the folds of their long robes draped over their extended hand.
43 A. Klockner, ‘Tertium genus? Representations of religious practitioners in the cult of Magna Mater: religious entrepreneurs and innovators in the Roman empire’, in R.L. Gordon, G. Petridou and J. Rupke (edd.), Beyond Priesthood: Religious Entrepreneurs and Innovators in the Roman Empire (Berlin and Boston, 2017), 343–84, at 352–3; Klockner is circumspect and merely refers to this individual as ‘priest’; however, the various accoutrements, the lack of facial hair and the hair imply that they may have been a Gallus/Archigallus—unfortunately, the lack of an occabus frustrates definitive interpretation.
44 On the occabus, see A.S.F. Gow, ‘The Gallus and the lion’, JHS 80 (1960), 88–93, at 90; Cavanagh (n. 41), 334; the occabus is an elaborate bracelet and is often viewed as a common indicator of the Archigalli, featuring as a badge or symbol of rank for these priests.
45 For the lid, see Gow (n. 44), 90; Latham (n. 6), 70–2; Klockner (n. 43), 350–1, 363–4; Roller (n. 37), 53; M.J. Vermaseren, Corpus Cultus Cybelae Attidisque (CCCA). III. Italia—Latium (Leiden and Boston, 1977). For Panel A, see Latham (n. 6), 70–2; Vermaseren (this note), 141–2; Klockner (n. 43), 351–2, 364; Roller (n. 37), 53. For Panel B, see Latham (n. 6), 70–2; Vermaseren (this note), 142; Klockner (n. 43), 351–2, 364; Roller (n. 37), 53.
46 See Klockner (n. 43), 346–50, 353–4, 357, 359, 362, 367–8; Meijer (n. 10), 275–7. See also F. van Haeperen, ‘Rappresentazioni dei ministri della Mater Magna a Roma e nelle province occidentali dell’Impero’, in F. Fontana, E. Murgia (edd.), Sacrum facere. Atti del IV Seminario di Archeologia del Sacro (Trieste, 2018), 241–62.
47 E.g. Anth. Pal. 6.234; Lucian, Syr. D. 50; see also Apul. Met. 8.27.4–6, Mart. 11.84 and Lucian, Syr. D. 50–1 for instances of Galli self-mutilation.
48 For example the ‘Lanuvium Gallus’, noted in Klockner (n. 43), 350, 354–5; Meijer (n. 10), 275–7; Gow (n. 44), 89; Latham (n. 6), 67–8; Vermaseren (n. 45), 152–3; Roller (n. 37), 53; R. Bell, ‘Revisiting the pediment of the Palatine Metroön: a Vergilian interpretation’, PBSR 77 (2009), 65–99, at 77 and the ‘Rome Archigallus’ noted in Klockner (n. 43), 353–4, 365; Gow (n. 44), 89–90; Roller (n. 3), 294–5; Latham (n. 6), 69; Vermaseren (n. 45), 64–5; Roller (n. 37), 52, as well as a headless Gallus bust noted in Klockner (n. 43), 355, 365, all of which are from Latium. See also Cavanagh (n. 41), 325–6, 338–9, 342.
49 For comparison, several ancient depictions of maenads also feature enlarged eyes on occasion, such as a skyphos from Paestum depicting a maenad with her head swinging back, mid-dance (British Museum inv. no. 1867,0508.1171), or an Attic altar depicting a maenad with enlarged eyes in a similar pose (British Museum inv. no. 1816,0610.330).
50 For example, Klockner (n. 43), 357–8 suggests that the wide eyes seen on several of the Latium case studies may be read as a reference to prophetic power; Meijer (n. 10), 275–7 argues that the large eyes of Bassus represent divination and possibly prophetic power.
51 Vermaseren (n. 45), 123–4; Hales (n. 7), 92–7; Klockner (n. 43), 345; Cavanagh (n. 41), 330.
52 For example the ‘bejewelled Gallus’, the ‘Lanuvium Gallus’, the ‘Fyrmus stele’ and the ‘Galli procession’ (among others); see Cavanagh (n. 41), 326–8, 332–3, 338–9.
53 See Roller (n. 3), 253 and Roller (n. 36), 544.
54 These iconographic parallels are noted by other scholars: Klockner (n. 43), 349 remarks that the clothing of Bassus is different from the norm, and that Bassus is beardless, contrary to contemporary styles, while Meijer (n. 10), 253 and 255 notes that the Archigalli had an appearance similar to the Galli and ‘evidently had no consequences for doing this’ in society; Mowat (n. 9), 2 notes how representations of Galli in the archaeological record are similar to the literary sources: extravagant pectoral ornamentation, headdresses and hand-held objects that ‘tap into the Roman visual vocabulary of foreignness/Easternness’, beardlessness, as well as both masculine and feminine clothing that ‘points to the gender incongruity so keenly noted by the (ancient?) writers’.
55 E.g. a sarcophagus from Termessus in Anatolia references the occupant, a Hierogallus (ἱϵρόγαλλος) named Aurelius Protoctetus (TAM III, 1 740); also see Vermaseren (n. 36 [1987]), 226; Cavanagh (n. 41), 308–9.
56 CIL 12.1782; CIL 13.1752 ex uaticinatione Pusoni Iuliani Archigalli, also noted by Spickermann (n. 8), 152; Meijer (n. 10), 271–2; and A.D. Nock, ‘A feature of Roman religion’, HThR 32 (1939), 83–96, at 86.
57 CIL 8.8203.
58 E.g. Firm. Mat. Err. prof. rel. 4.2; Polyb. 21.37; Livy 38.18; August. De ciu. Dei. 7.26; Plut. Vit. Cleom. 36.4; Clem. Al. Protr. 7; Ath. Deipn. 12.58.
59 Meijer (n. 10), 271–2 argues that the Archigallus also appears to have had prophetic roles as evidenced by the words ex uaticinatione Archigalli in some inscriptions, notably three from Lyon and Tain in Gaul (a.d. 184–90) as well as Milevum in Africa (a.d. 222–35); however, rather than them being leaders of Galli, she argues that they held prophetic/divinatory roles in keeping with the cult of the goddess in Anatolia. Yet this argument does not appear to take into account the literary attestations of prophetic powers attributed to the ‘common Galli’ priests.
60 TAM III, 1 578, discussed by Cavanagh (n. 41), 305–6; see also Vermaseren (n. 36 [1987]), 225; Obrador-Cursach and Adiego (n. 36), 175; P.K. Takalić, ‘Vrijeme uvođenja i uloga Arhigala u svjetlu natpisa L. Publicija Sintropa iz Kopra / Period of introduction and role of Archigalli in context of the inscription of L. Publicius Syntrophus from Koper’, Archaeologia Adriatica 6 (2012), 87–105, at 97.
61 This is evident from the archaeological record elsewhere, since priests can hold multiple jurisdictions; for a Gallus example, see the monument of Lucius Fyrmus, priest of Isis in Ostia and Magna Mater ‘across the Tiber’, in Latham (n. 6), 69; Vermaseren (n. 36 [1987]), 133–4; Klockner (n. 43), 346–9, 357–9, 366–7; Roller (n. 37), 53; Bell (n. 48), 77–8; Cavanagh (n. 41), 332–3.
62 A marble plaque of an Archigallus contains the abbreviation H.M.H.E.N.S. (hoc monumentum heredem externum non sequitur, ‘this monument does not pass to any external heir’), which might be indicative of a son or a daughter, but whether this is a ‘blood’ child or an adoption is unclear, especially given the Latin externum, implying a ‘non-intimate’ family member. See, further, Vermaseren (n. 36 [1987]), 69–70; Klockner (n. 43), 345; and Cavanagh (n. 41), 328.
63 For instance the Mahu of Tahiti and the Crow Bote in North America, both of whom are attested as engaging in formal adoptive practices of young children: see further Cavanagh (n. 41), 238–40.
64 E.g. the male sex slave of the Galli in Apul. Met. 8.26.5–6 or the liaisons implied between Galli and women in Mart. 3.81 and Hor. Sat. 1.2.115–23.
65 See S. Treggiari, ‘Contubernales in CIL 6’, Phoenix 35 (1981), 42–69, at 42–4.
66 Treggiari (n. 65), 59. This argument is also supported by van Haeperen (n. 46), 251; Meijer (n. 10), 253; and Klockner (n. 43), 362.
67 Vermaseren (n. 45), 150.
68 Meijer (n. 10), 243–4.
69 For more information on this law, see M. Perry, ‘The Lex Scantinia and the public response to stuprum’, Eugesta 13 (2023), 31–54; M. Johnson and T. Ryan, Sexuality in Greek and Roman Literature and Society: A Sourcebook (Abingdon and New York, 2005), 8; A. Richlin, The Garden of Priapus: Sexuality and Aggression in Roman Humor (New York, 1992), 86, 224. One of the earliest references to the law can be found in Cic. Fam. 8.12.
70 Alt.v.Hierapolis 33; Roller (n. 37), 54; Vermaseren (n. 36 [1987]), 30; Cavanagh (n. 41), 316–17.
71 CIL 14.34; Vermaseren (n. 45), 125–6; Klockner (n. 43), 345; Cavanagh (n. 41), 331–2.
72 CIL 14.385; Vermaseren (n. 45), 123–4; Hales (n. 7), 92–7; Klockner (n. 43), 345; Cavanagh (n. 41), 330.
73 CIL 6.2183; Vermaseren (n. 45), 69–70; Klockner (n. 43), 345; Cavanagh (n. 41), 328.
74 CIL 6.19875/6.32466; Klockner (n. 43), 346–50, 353–4, 357, 359, 362, 367–8; Meijer (n. 10), 275–7; Cavanagh (n. 41), 340–1.
75 CIL 10.3810.
76 CIL 5.488; P.K. Takalić, ‘Presence of the Archigalli on the eastern Adriatic coast. Examination of their role in the cult of Magna Mater and Attis’, in A. Nikoloska, S. Muskens (edd.), Romanising Oriental Gods? Religious Transformations in the Balkan Provinces in the Roman Period. New Finds and Novel Perspectives (Leiden and Skopje, 2015), 371–90.
77 See H. Solin, ‘Zur Entwicklung des römischen Namensystems’, Politische Kultur und soziale Struktur der römischen Republik. Bilanzen und Perspektiven. Akten der internationalen Tagung anlässlich des 70. Todestages von Friedrich Münzer (Stuttgart, 2017), 135–53 for a more detailed analysis of Roman nomenclature over time; see also B. Salway, ‘What’s in a name? A survey of Roman onomastic practice from c. 700 b.c. to a.d. 700’, JRS 84 (1994), 124–45. The reader should note that Roman onomastic culture was complex, and that it varied over time, from a simple nomen in the earliest iteration of the Roman state to the tria nomina as indicator of status, shifting again to a convoluted scheme of mixed binominal/polynominal system by the Later Imperial period, after the Edict of Caracalla. Despite this, according to Salway (this note), 133, up until the Edict users of the Roman naming system comprised ‘native Italians, their descendants in the provinces, and foreign enfranchisees who had embraced the Roman name with their citizenship’.
78 In fact, this argument is supported by Klockner (n. 43), 349, who argues that two such archaeological case studies from Latium—Bassus and Fyrmus—were Roman citizens owing to their having three names or tria nomina, while Meijer (n. 10), 253 argues that the Archigalli were mostly Roman citizens, some possibly even belonging to the aristocracy with their tria nomina.
79 Cavanagh (n. 41), 332–3; Latham (n. 6), 69; Vermaseren (n. 45), 131–2; Klockner (n. 43), 346–9, 357–9, 366–7; Roller (n. 37), 53; Bell (n. 48), 77–8. Interestingly, Vermaseren (n. 45), 131–2 claims that they are ‘unquestionably a Gallus’, while Klockner (n. 43), 348–9 identifies them as a sacerdos.
80 Meijer (n. 10), 243–4; Cavanagh (n. 41), 343–4; L. Vidman, Sylloge inscriptionum religionis Isiacae et Sarapiacae (Berlin, 1969), 190.
81 Cass. Dio 78.9; Dig. 1.5.17. For a detailed insight into the Edict, see A. Imrie, The Antonine Constitution: An Edict for the Caracallan Empire (Leiden and Boston, 2018). In addition, see M. Lavan, ‘The spread of Roman citizenship, 14–212 ce: quantification in the face of high uncertainty’, Past & Present 230 (2016), 3–46; J.A. Latham, ‘“Fabulous clap-trap”: Roman masculinity, the cult of Magna Mater, and literary constructions of the Galli at Rome from the Late Republic to Late Antiquity’, JR 92 (2012), 84–122, at 116; and Salway (n. 77), 133.
82 Vermaseren (n. 45), 140–2; Cavanagh (n. 41), 334–7.
83 CIL 6.32462; Cavanagh (n. 41), 343–4; Vidman (n. 80), 190; Meijer (n. 10), 243–4.
84 RECAM II 206; Vermaseren (n. 36 [1987]), 15; Takalić (n. 60), 97; Cavanagh (n. 41), 310–11.
85 Smyrna 136; Vermaseren (n. 36 [1987]), 164; Cavanagh (n. 41), 312.
86 Alt.v.Hierapolis 33; Vermaseren (n. 36 [1987]), 30; Cavanagh (n. 41), 316–17.
87 TAM III, 1 619; TAM III, 1 578; TAM III, 1 740; Vermaseren (n. 36 [1987]), 225–6; Cavanagh (n. 41), 305–9; Obrador-Cursach and Adiego (n. 36), 175; Takalić (n. 60), 97.
88 RECAM II 206; Vermaseren (n. 36 [1987]), 15; Takalić (n. 60), 97; Cavanagh (n. 41), 310–11.
89 Also see Klockner (n. 43), 350–1.
90 Gow (n. 44), 90.
91 Hales (n. 7), 93–4.
92 Klockner (n. 43), 351.
93 Klockner (n. 43), 352.
94 Klockner (n. 43), 351.
95 Diod. Sic. 36.13; Varro, Sat. Men. 121 Astbury.
96 CIL 14.429; Klockner (n. 43), 348.
97 Cavanagh (n. 41), 332–3; Latham (n. 6), 69; Roller (n. 37), 53; Bell (n. 48), 77–8.
98 Vermaseren (n. 45), 133–4.
99 IGRR 1.92; CIL 6.32462; Vermaseren (n. 45), 150; Meijer (n. 10), 243–4; Vidman (n. 80), 170.
100 Roller (n. 3), 278–9, 299–300.
101 CIL 14.429; CIL 14.4627; CIL 6.496; Vermaseren (n. 45), 9–10, 132, 134.
102 E.g. Lucr. 2.581; Catullus 63; Verg. Aen. 9.616–17; Mart. 1.35, 3.91, 7.95, 9.2, 11.84.
103 Val. Max. 7.7.6; Julius Obsequens, Prodigiorum liber 44a.
104 Plut. Vit. Mar. 17.5.
105 With the exception of the aforementioned examples of Esopos and Abdedera, who could have been non-local/foreign, thus mitigating local social stigma. See, further, Cavanagh (n. 41), 205–6.
106 Dion. Hal. Ant. Rom. 2.19.
107 Roller (n. 3), 315–16; D.M. Cosi, ‘Salvatore e salvezza nei misteri di Attis’, Aevum 50 (1976), 42–71, at 51–2; Roscoe (n. 4), 201–2; Summers (n. 14), 355–6; Fear (n. 2), 47.
108 Lucr. 2.581; Diod. Sic. 36.13; Roller (n. 3), 283 notes that, although Cybele was a respected Roman deity, her foreign roots were always remembered; for example, her games were the only ones not given a Latin name, and her hymns were always sung in Greek, never in Latin.