¹ For the importance of the group or θίασος cf. Cl. Rev. 1924, p. 105 ff.; but we must not under-estimate the significance of the individual believer.

² Cf. an inscription of Prusias, I.G. Rom. iii. 63Google Scholar (Dittenb. O.G.I. 528), 1. 10, where the μυστήρια probably belong to Caesar-worship. The divine personification of the city Side is called ΜΥΣΤΙΣ, B.M.C. Lycia, p. 163Google Scholar, n. 126, Pl. XXIX. 3. μυστήριον in itself means no more than ‘secret rite’ (Boisacq, , Dict. Etym., p. 654Google Scholar, Farnell, , Cults, iii. p. 129 ff.Google Scholar): for the increase in such under the Empire cf. Lécrivain, , Dar.-Sagl. iii. p. 2137.Google Scholar

³ Cf. I.G. Rom. i. 662, 1. 19 (Dionysopolis), [1st cent. B.C.], Ramsay, , C.B. i. p. 290Google Scholar, n. 127 (dedications by the for city, etc.), I.G. Rom. iii. 162 (with Ramsay's, W. M. notes, J.R.S. xii. p. 165 ff.Google Scholar) for the association of a hierophant with the dedication by the Koinon of Galatia, , C.I.L. ii. 5521Google Scholar (Corduba: 238 A.D.), ex iussu Matris deum pro salute imperii taurobolium fecit, and the ritual of the Kenoi Tekmoreioi discussed by Ramsay, W. M., Studies, p. 305 ff.Google Scholar

⁴ Cf. Carm. lat. epig. 25, 4, inde cuncti didicimus, and Cl. Rev. 1924, p. 59₁₀. On the individual aspect of Hellenistic belief, cf. Lehmann, E., Lehrbuch der Religionsgeschichte ¹, II., p. 265.Google Scholar

⁵ We are here putting on one side the harmless agrarian or weather-controlling magic often practised by or for a community, as for instance the ascent of young men clad in skins to the cave of Chiron at the rising of the Dogstar (Heraclid. Descr. Graec. ii. 8, in Geogr. gr. min. i. p. 107), or the rain-making in Arcadia (Pausan. viii. 38, 4) and at Crannon in Thessaly (Antig. Hist. mir. 15).

⁶ A magical process is described as τελετή (P. Leid. W. vi. 18) in Dieterich, , Abraxas p. 187Google Scholar, 12, p. 200, 17, p. 200, 23, P. Par. 26, 1596, 2205; for the Byzantine use of the word, cf. Dawkins, R. M., Folklore, 1924, p. 226Google Scholar; τετελεσμένον comes to mean ‘enchanted,’ since many objects used magically needed some preliminary operation to animate and excite their hidden powers; cf. P. Par. 2967 ff. for this treatment as applied to a herb, the Cyranides passim for its application to stones, also Eitrem, S., Lina Laukar (offprinted from Festskrift til Bibliothekar A. Kjaer: Kristiania, 1924), p. 1, 5 ff.Google Scholar; for μυστήριον of alchemic processes, etc. cf. Röhr, J., Philol. Suppl. XVII. p. 84 f.Google Scholar, its recipient as μύστης (cf. P. Par. 744, Hopfner, Th., Griechisch-Ägyptischer Offenbarungszauber, ii. p. 14, § 28Google Scholar); it is supposed to be handed down from father to son (cf. Dieterich, , Mithrasliturgie ³, p. 52Google Scholar, with Weinreich's note, p. 230, cf. Cyran. III. φ. 28, p. 43). In magic we find also the same desire for intimate union with the deity addressed which is regular in orgiastic cults (cf. Dieterich, op. cit. p. 97 ff.), and the notion of militia sacra (P. Par. 1. 193, cf. Reitzenstein, , Hellenistische Mysterienreligionen ², p. 73 ff.Google Scholar)

⁷ We can hardly agree with Gruppe's, view (B. ph. W. 1911, p. 931 f.)Google Scholar, that the character of the earlier part of the work militates against the view that the end represents genuine conviction in Apuleius when writing. Lucius tells first of his sins, then of his conversion and the new life it involved.

⁸ I should wish to express my thanks to Mr. A. B. Cook for advice and help, and to Mr. C. T. Seltman for useful suggestions and corrections. The following abbreviations are here used, in addition to those customary in the Journal: B. ph. W. = Berliner Philologischer Wochenschrift, K.P. i. ii. iii. = J. Keil und A. Von Premerstein: Bericht über eine erste, zweite, dritte Reise in Lydien in Denkschr. Ak. Wiss. Wien, phil. hist. Kl. LIII. ii., LIV. ii., and LVII. i., respectively, I.G. Rom. = Inscriptiones graecae ad res Romanas pertinentes, Dess. = Dessau, Inscriptiones latinae selectae, P. Par. for the great Paris magical papyrus (Suppl. gr. 574) as edited by Wessely, K., Denkschr. Wien, XXXVI, ii. (1888)Google Scholar, (other papyrological abbreviations are as in general use), Le Bas = Le Bas-Waddington, Voyage Archéologique en Grèce et en Asie Mineure: Inscriptions, V. (1858–1870)Google Scholar, Cumont T.M. = Cumont, Textes et monuments, Not. = Notizie degli scavi di antichità, Mitteis-Wilcken = Mitteis und Wilcken, Grundzüge und Chrestomathie d. Papyruskunde, A.R.W. = Archiv für Religionswissenschaft, J.T.S. = Journal of Theological Studies, Ramsay, C.B. = Ramsay, Cities and Bishoprics, Dittenb. O.G.I. = Dittenberger, Orientis graeci inscriptiones selectae.

⁹ Orat. VIII. i. p. 88, Dindorf = ii. p. 356, § 15, Keil.

¹⁰ So Weinreich, O., Neue Urkunden zur Sarapisreligion (1919), p. 4₂.Google Scholar

¹¹ Dess. 4192 a.

¹² Le Bas, 1764 b.

¹³ Dess. 3995. Virtus Bellona (as Dess. 3803–5) may be the goddess conceived as the embodiment of divine power; she is called simply Virtus in an inscription at Madaura, (C.R. Acad. Inscr. 1919, p. 257).Google Scholar

¹⁴ B.C.H. vi. p. 515. Ramsay, , C.B. ii. p. 700Google Scholar, suspects Christian influence, perhaps wrongly (cf. C.B. i. p. 152, inscr. 49, and Steinleitner, F., Das Beicht, p. 54 f.)Google Scholar: for the Saines cf. Le Bas, 784, 1499, 1683, 1704. We may note also J.H.S. iv. p. 385, and the epithet compos of Hercules, (Dess. 3438)Google Scholar, compotens of Diana (ib. 3249).

¹⁵ Kaibel, Epigr. gr. 1139.

¹⁶ As P. Par. 1. 1275, 1024 1190 (for the identity of name and thing in magic language cf. Journ. Theol. Stud. XXVI., p. 177₂), P. Parthey, i. 344, P. Lond. 124, 11, 20, Wessely, μέγας καὶ ἰσχυρὸς θεύς ὸσχυρός of a δαίμων, P. Lond. 46, 130 Wess. = 129, Kenyon (cf. 148 W. = 147 K.), also P. Parthey, i. 164, P. Par. 1537, 2032, and [name] 251. If the deity is strong (cf. also P. Lond. 121, 664 W. = 601 K.), so also is an amulet conferring his protection, cf. P. Lond. 121, 592 W. = 582 K., cf. P. Par. 1653 [prayer for it], or a spell constraining him to act, cf. P. Par. 1331, where one is called 1872, (?) P. Lond. 121, 986 W. = 918 K., P. Par. 3170 For δύναμις in a Gnostic text cf. Acta Iohannis ed. Bonnet, p. 163, 28,

¹⁷ S. Athanasius is at pains to distinguish ὁ θεὁς from (Orat. I. contra Arianes, § 11 (xxvi. p. 36 A, Migne): this suggests that some confusion of thought existed. The emphasis on δύναμα here discussed is naturally not peculiar to Graeco-Roman paganism: thus for its analogue in the North cf. Mogh, E., Streitberg-Festgabe, 1924, p. 278 ff.Google Scholar

¹⁸ Acts xix. 34 (the Bezan ἡ Ἀ. supported by Ramsay, W. M., Church in R.E. p. 142Google Scholar, is perhaps right): Aristid. i. p. 467, Dindorf: Ramsay, , C.B. i. p. 151Google Scholar, inscr. 49: K.P. ii. p. 103, n. 204: ib. p. 105: J.H.S. xviii, p. 319, n. 32: ib. x. p. 226, n. 21: Bull. soc. archéol. Alex. iv. p. 98, n. 72: cf. C.Q. 1924, p. 188, and a Christian example, (Mart. s. Ariadnes, 18, in Studi e testi, vi. p. 133). μέγας means ‘powerful,’ cf. Soph. O.T. 872. We are here concerned with its use as a predicate, not as a cult-epithet (for which Müller, Br. has collected material, Dissert, phil. Hal. XXI. iii.Google Scholar). Ramsay, notes also B.C.H. vii. p. 322Google Scholar, of Zeus at Venasa (where we should punctuate after ).

¹⁹ Somn. Nectanebi, col. ii. 17 (in Lavagnini, E., Erot. frag. pap. p. 39Google Scholar).

²⁰ I.G. Rom. iv. 1529, 10 ff.

²¹ Corp. Herm. i. 31.

²² Cf. Graillot, H., Le culte de Cybèle, p. 157 f.Google Scholar

²³ Irenaeus, S., Adu. haer. i. 14, p. 181Google Scholar, Harvey: (sc. )

²⁴ Cf. Corp. Herm. i. 27, ἐγὼ [the man Poimandres has been instructing] ib. § 32 [the final prayer], implies a conception of this divine power as light very commonly found in religious texts of our period: cf. G. P. Wetter, ΦΩΣ, Skrifter utgifna af K. Humanistiska Vetenskaps-Samfundet i Uppsala, XVII. i. (1915)Google Scholar, and ch. iii. of the Prolegomena to my forthcoming edition of Sallustius.

²⁵ Cf. P. Par. 1665, [prayer to Sun] 1616, 197, δύνάμωσον, P. Lond. 121, 648 W. = p. 115 K. [give me] ( Keny.), 991 W. = 925 K., P. Parthey, i. 273, P. Leid. V. ii. 13, vii. 18, W. Dieterich, vii. 13 (Abraxas, J. 5, p. 190, 6), vii. 16, and for a claim to possess it, P. Par. 1828, (sc. ) we may note also the phrase, P. Par. 477,

²⁶ Cf. P. Par. 518,

²⁷ Actae Sanctorum Sept. vii. p. 204, 205. In this passage I have inserted οὐδείς (rather than the <τις> of Baluzius after δυνήσεται) and substituted for is a common word in magical texts, as P. Par. 160, 1718, P. Mimaut, 289, cf. P. Parthey, i. 194, ). Hippol. Refut. v. 7, 22, p. 84, Wendland, preserves a Naassene tradition that the Egyptians taught all men the ἰδέαι and of the gods; cf. Corp. Herm. xvi. 13, p. 352, 23, Reitzenstein = 270, 1, Scott, Dittenb. O.G.I. 262, 4, of Zeus of Baitokaike.

²⁸ Cf. P. Par. 216, The name is naturally of supreme importance in invocations, cf. Connolly, R. H., J.T.S. xxv. p. 354 f.Google Scholar, also P. Parthey, ii. 127, and the treatise of J. Röhr quoted in the next note, p. 12 f. (ib. p. 15 ff. on ).

²⁹ As P. Par. 2967 ff., e.g. 2998, (cf. for λαμβάνω a spell Terram teneo, herbam lego; in nomine Christi prosit ad quod te colligo given by Marcellus Empiricus, xxv. 13), Anthol. lat. ed Riese², i. p. 28, n. 6, 1. 7, uos hue adeste cum uestris uirtutibus, a Greek treatise, De herbis, edited by Haupt, W., Opusc. ii. p. 475 ff.Google Scholar, 1. 76, Cf. in general Röhr's, J. work, Der okkulte Kraftbegriff im Altertum (Philol. Suppl. XVII. i. 1923)Google Scholar, in which abundant illustrations of this kind of belief will be found. Pfister, suggests (Pauly-Wissowa, xi. p. 2155, 21 ff.)Google Scholar that the of a word is firstly its magic power, secondly its meaning.

³⁰ So ἀρεταί of works of healing (cf. Weinreich ap. Ditt. Syll. ³ 1172, 8), and of the miracles of wonder-working images (Le Bas, 519), ἀρετή in the singular of divine punishment for sin compelling the offender to make public confession (Buckler, , B.S.A. xxi. p. 180Google Scholar): δυνάμεις of miraculous cures (Grimm-Thayer, , Lex. N.T. ² p. 159Google Scholarb, P. Oxy. 1381, 1. 41, 90, 216, Aristid. Λαλιά εἰς Ἀσκληπιόν, § 4, i. p. 64, Dindorf), of divine visitations (K.P. ii. n. 204, Le Bas, 668), and more generally, as Le Bas, 2683, P. Leid, V. xi. 25, while is inscribed under an engraving of a thunderbolt at Thyateira in Lydia, (B.C.H. x. p. 401Google Scholar, n. 4, cf. K.P. ii. p. 17). For uirtutes cf. Passio S. Perpetuae, i. p. 62, 14, Gebhardt, xxi. p. 95, 1, also Ter. Ad. 535: potentia occurs in a concrete sense in Apul. Met. xi. 13.

³¹ Cf. Wetter, G. P., ΦΩΣ, p. 53.Google Scholar

³² Orat. XXIV, § 36 (ii. p. 530, 4, Förster); cf. Eunap. fr. 23 (F.H.G. iv. p. 23)

³³ Carm. lat epigr. ed. Buecheler, , 111, 25.Google ScholarHecates trina secreta refers perhaps to mysteries celebrated at Aegina, cf. Pausan. ii. 30; Lobeck, , Aglaophamus, p. 242.Google Scholar

³⁴ Cf. Cumont, , T.M. i. p. 338.Google Scholar

³⁵ Silvanus was not excluded (T.M. ii. p. 103, inscr. 54, p. 467, inscr. 54 a); a dedication to the native deity Cissonius was found in the Mithraeum recently discovered at Königshofen, (Ann. épigr. 1920, 128).Google Scholar

For the relations of Cybele worship and Mithraism, , cf. T.M. i. p. 333.Google Scholar In view of Cumont's denial that the taurobolium ever formed part of the Mithraic, liturgy (T.M. i. p. 334₅)Google Scholar, it may be noted that a pit and hypogaeum almost certainly intended to serve this purpose exist in the Mithraeum in the Baths of Caracalla (Lanciani, , La Zone monumentale di Roma, pt. A, p. 58Google Scholar); here at least the taurobolium is closely connected with rites peculiar to Mithraism, and it was probably borrowed from the cult of Cybele.

³⁶ Dess. 3949.

³⁷ ib. 3197.

³⁸ C.I.L. ix. 3146.

³⁹ C.I.L. vi. 490.

⁴⁰ Dess. 4396.

⁴¹ Dess. 4413: pater probably refers to the cult of Mithras, ierofanta to that of Hecate (Dessau ad. loc). Cf. Carm. lat. epigr. 264, 1529 A, Ramsay, , C.B. i. p. 246Google Scholar, inscr. 88, Kaibel, , Epigr. gr. 588.Google Scholar

⁴² K.P. ii. p. 107, n. 210.

⁴³ P. Oxy. 1380, 6.

⁴⁴ Cf. Weinreich, , Neue Urkunden, p. 24 ff.Google Scholar, for illustrations of this, as also of (cf. Le Bas, 2313, ), P. Par. 1814,

⁴⁵ L. Canet ap. Cumont, C.R. Acad. Inscr. 1919, p. 313 ff.Google Scholar On the other side of the slab (which Paribeni, , Le Terme di Diocleziano ⁴, p. 125Google Scholar, n. 205, calls an ‘aretta’) is inscribed (Ghislanzoni, , Not. 1912, p. 323Google Scholar).

⁴⁶ I.G. Rom. i. 1063.

⁴⁷ Cf. a lamp inscribed (Bull. Soc. Alex. iv. p. 98, n. 71), a Cnidian dedication (Lebas, 511), and the profession of faith we have from one Sansnos, (Mitteis-Wilcken, I. ii. p. 147Google Scholar, n. 116), which, while enjoining special reverence to Isis and Serapis says, 1. 2,

⁴⁸ P. Oxy. 1242 (according to von Premerstein, , Philol. Suppl. xvi. 2 (1923), p. 64 ff.Google Scholar part of a larger work, which he dates in the time of Caracalla). Religious antagonism was not, it seems, the only motive for this anti-Semitism, on which cf. Wilcken, U., Abh. sächs. Ges. Wiss. xxvii. (1909), p. 827Google Scholar, Bell, H. I., Jews and Christians in Egypt, p. 10 f.Google Scholar

⁴⁹ For such monuments cf. Cumont, , Dar.-Sagl. iv. p. 314 f.Google Scholar, also a relief at Neontichos in Lydia combining attributes of Aphrodite and Cybele in one figure (K.P. i. p. 94, Abb. 95), many instances of the investing of Isis with diverse attributes in R. Weisshäupl's useful paper Jahresh. xiii. (1910), p. 176 H. (ib. p. 193 ff. cases of Venus Panthea), a coin of Alexandria struck by Hadrian representing Zeus with attributes of Helios, , Serapis, , Khnemu, , Ammon, , Neilos, , Poseidon, , and Asklepios, (B.M.C. Alexandria, p. 88Google Scholar, n. 744, Pl. XV., cf. Cook, A. B., Zeus, i. p. 361₄Google Scholar), Tyche Panthea on a coin of Tarsus, (B.M.C. Phrygia, p. 298Google Scholar, n. 126, Pl. XXXVI. 6). For this view of making a god be such as you represent him cf. Cook, A. B., Folklore, 1903, p. 270 f.Google Scholar

⁵⁰ Cf. Cumont in the article quoted in n. 41. We may note, P. Oxy. 1766, 18, Ditt. Syll. ³ 1153,

⁵¹ Dess. 4362 (at Capua).

⁵² Apul. Met. xi. 5.

⁵³ P. Oxy. 1380, Kaibel, , Epigr. gr. 1028Google Scholar (= I.G. xii. 5, 739), cf. Apul. Met. xi. 2, 5; Isis is μυριώνυμος (O.G.I. 695, 2, Dess. 4361, 4376 a), (Kaibel, 1029, 5, P. Oxy. 1380, 97). Cf. an invocation of Selene – Artemis – Hecate – Kypris – Deo at Hierapolis in Cilicia, (I.G. Rom. iii. 903)Google Scholar, the Naassene hymn to Attis (ap. Hippol. Ref. v. 7, 22, p. 83, 22, Wendland), an invocation of Apollo–Titan–Osiris–Mithras in Stat. Theb. i. 716 ff., and P. Leid. v. 8, 65 f., and (Cramer, , Anecdota Parisina, i. p. 170, 28Google Scholar), § 39, where the governor of Egypt is made to say (so ἐὰν is emended)

⁵⁴ Relatio pro ara Pacis, § 10.

⁵⁵ Carm. lat. epig. 111, 15.

⁵⁶ Dess. 4303, supplemented by Cumont, , Rev. phil. 1902, p. 7 f.Google Scholar, Études Syriennes, p. 197. Cumont compares S. Aug. Ciu. Dei, iv. 11, and the Philonian δυνάμεις. For the notion of effluences cf. P. Leid. J. 384, viii. 7, Sallustius, ch. ii., makes the gods functions of the First Cause. The rhetorician Alexander remarks that philosophical teaching makes the Divine have neither beginning nor end in time, and that Plato harmonised this with the popular view by making the ordinary gods proceed from the first god (ix. p. 336, 19 ff., Walz).

⁵⁷ Epp. 16 [48] = Migne, xxxiii, p. 82; Saturn, i. 17, 4 (on which cf. Cumont, , La théologie solaire du paganisme romain, Mém. Acad. Inscr. XII. ii. p. 454 f.Google Scholar), Lactant. Placid, , ad. Stat. Theb. iv. 516.Google Scholar

⁵⁸ As that represented by a bas-relief at Wiesbaden, on which Hygieia and Asklepios are made to draw their healing power from Zeus (Jahn, O., Die Heilgötter, Taf i.Google Scholar, an offprint from Annalen des Vereins für Nassauischen Altertumskunde und Geschichte, VI), and by the Εὶς Δία of Aristides, who says (i. p. 6, Dindorf) and again (p. 9) A Clarian oracle, quoted by Lactant. Inst. Diu. i. 7, describes the ordinary gods as angels and as a small portion of the Supreme Being (cf. Picard, , Éphèse et Claros, p. 716 ff.Google Scholar on the policy of Claros).

⁵⁹ Epist. 89a, p. 127, 3, Bidez-Cumont, , In Christianos, p. 179Google Scholar, 7, Neumann, cf. Symmach. Rel. iii. 8. Mr. C. T. Seltman draws my attention to the frequency of DEO SERAPIDI and of other Egyptian types on Julian's, coinage, Cohen ², viii. p. 42–73.Google Scholar

⁶⁰ Cf. Otto, W. F., Rh. Mus. lxiv. (1909), p. 449 ff.Google Scholar, and note in the East K.P. ii. 203, (with note, p. 104, for parallel ).

⁶¹ These instances are all to be found in Dessau's index viii. (iii. p. 516 ff.), except Frontonianus, H. (C.I.L. xi. 4669).Google Scholar For meus cf. Dess. 4778, 4790, suus 4811, 9321, 9322, domeslicus 4779 (Matres), 9321–2 (Suleuae), 3442 (Hercules), 3558–9. Jahresh. xii. (1909) Beibl. p. 172, n. 34, (Siluanus), Dess. 3705 (Fortuna Bona), C.I.L. xiii. 8718 (Juppiter Optimus Maximus). An interesting example of this appropriation of a deity is the application of the epithet Purpurio to Juppiter Optimus Maximus in a dedication at Rome, (Dess. 3040)Google Scholar by three women, one of whom, presumably the chief giver, was named Purpuris (for another explanation, which I find hard to accept, cf. Cook, A. B., Zeus, i. p. 58Google Scholar₁).

⁶² Dess. 3254. So in a wallpainting from Ostia the children adore a child Artemis (Dieterich, , Kl. Schr. p. 324Google Scholar, figures this: a sketch is given by Reinach, , Rép. peintures G. et R. p. 235Google Scholar, 1).

⁶³ This, of course, is not new. As examples of it we may give Le Bas, 2209, Ann. Epig. 1913, 88, Dess. 4500, Dianae Aug. Maurorum, 3178, Veneri uictrici Hyblensi, and the Hercules Gaditanus attested by Cohen. ² ii. p. 174, n. 814. We may note that on the coins of Gerasa Artemis appears as Tyche of the city, with the inscription (B.M.C. Arabia, p. xxxi₂, 31 f. n. 1–9, Pl. V. 4–6), and that on an alliance coin of Mytilene and Perga in Pamphylia the Tyche of Mytilene is confronted by the Artemis, not the Tyche, of Perga, (B.M.C. Troas, p. 215Google Scholar, n. 235, Pl. XLIII., 4: this Artemis is called in I.G. Rom. iii. 797, an inscription quoted by B. Pace in his useful study of the goddess in Anatolian Studies, p. 297 ff.): cf. Le Bas, 382 (a Cretan city allows Mylasa ).

⁶⁴ Apart from the di militares treated by Domaszewski, Von, Westd. Zeitschr. xiv. (1895), p. 1 ff.Google Scholar, we may note Iuppiter cohortalis (Thes. ling. Lat. iii. p. 1560, 58 ff.: Not. 1920, p. 102, is decisive for his nature). For sodalities two instances may suffice, Dess. 3540, Siluano dendrophoro (cf. Peter, R., Roscher, iv. p. 866, 15 ff.Google Scholar), 4105, Matri magnae cognationis.

⁶⁵ Dess. 4125–6, 4130–7.

⁶⁶ C. R. Acad. Inscr. 1919, p. 443.

⁶⁷ Le Bas, 2481.

⁶⁸ ib. 2562.

⁶⁹ ib. 2332; 2526 (Eirene), 2527 (Isis) 2528 a (unnamed).

⁷⁰ ib. 19, 1620 d; I.G. Rom. iv. 316; θέαι Le Bas 1723.

⁷¹ ib. 1700, 2480.

⁷² ib. 2296, 2301.

⁷³ For the use of Augustus as an epithet abundant illustration is given in Thes. ling. Lat. ii. p. 1393 ff. Augustus means ‘sacrosanct,’ and a dedication Augustis Laribus at Betriacum dating from 49 B.C. is known to us (C.I.L. v. 4087). In the great majority of cases, however, it must refer to the Emperor. Herculi Augusto cannot be separated from Herculi Augusti (C.I.L. iii. 3305, 5531, 10406), or Mercurio Augusto from Mercurio Augusti (ib. xiii. 364); the genitive and the adjective must have meant the same thing (cf. also Siluano Augusto and Siluano Flauiorum [C.I.L. vi. 644]. Otto, , Pauly-Wissowa, vii. p. 36, 1. 58 ff.Google Scholar remarks that the genitive makes the reference even clearer. The Greek equivalent σεβαστός seems not to be used as religious epithet before its application to the Emperor, as equivalent to Augustus. We find at Alabanda a statue base inscribed Ἀπόλ. λωνος Ἐλευθεερου Σεβαστοῦ (Le Bas, 549), near Smyrna (I.G. Rom. iv. 1488) and (ib. 1490), at Cos Rhea is Σεβαστή (Paton-Hicks, , Inscriptions of Cos, 119, p. 147Google Scholar), as is perhaps Hephaestus at Mylasa, (Lebas, 363)Google Scholar; an Anaitis Sebaste at Philadelphia is probably postulated by (Le Bas, 645). A Thasian inscription published by Picard, Ch. (Xenia, Athens, 1912)Google Scholar mentions a priest of Turning to the evidence of coins, we find at Tralleis (B.M.C. Lydia, p. 340 f. n. 94–6, Pl. XXXV. 7), at Amastris (Imhoof-Blumer, , Griechische Münzen, p. 62Google Scholar = [586] where it is assigned to the time of Trajan, , Recueil Asie Min. i. p. 139Google Scholar, Pl. XVIII. 38, Pl. V. 11; Imhoof-Blumer, op. cit., p. 72 = [596] interprets as [the city] though it may well be a nominative; in Crete, (Svoronos, , Numism. de la Crète, i. p. 343Google Scholar, Pl. 33, 17), time of Domitian. This list gives all clear examples I could find of σεβαστός applied to a deity in no sense an abstraction (I.G. iii. 460, looks like being the equivalent of Salus Augusta); we may perhaps conclude that it is much rarer in this sense than is Augustus in Latin.

Wissowa, , Religion und Kultus der Römer ², p. 85Google Scholar, explains this use of Augustus as meaning ‘dass man die betreffende Gottheit in demselben Sinn verehre, wie es der Kaiser in seinem Hauskulte tue.’ His view that it spread from Lares Augusti and Vesta Augusta hardly agrees with the dated texts, whose testimony is summarised by Otto, , Thes. I. L. ii. p. 1393, 1. 52 ff.Google Scholar (a most serviceable article).

⁷⁴ C.I.L. viii. 6353.

⁷⁵ Eph. Epig. vii. 759.

⁷⁶ C.I.L. viii. 895.

⁷⁷ Dess. 633.

⁷⁸ ib. 632.

⁷⁹ From which CERES AVGVSTA can hardly be distinguished: cf. B.M.C. Rom. Emp. i. p. 183, 191, 332, 375, etc.: but at times, as Mr. Seltman informs me, this must mean ‘Augusta who is Ceres,’ the Empress identified with the deity. But it should be noted that the formula AUG. Hardly occurs elsewhere till the third century; we may notice IVNO AVGVSTAE for Mammaea as an instance of the longer form (Cohen², iv. p. 493, Mammaea, n. 33 ff.)

⁸⁰ C.I.G. 4332: ib. 3886: Le Bas, 754: Ramsay, , C.B. ii. p. 637Google Scholar, inscr. 530: I.G. Rom. iv. 1593: Svoronos, , Num. Crète, i. p. 343Google Scholar, Pl. 33, 15: B.M.C. Alexandria, p. 36, n. 292. Pl. XXII.: B.M.C. Lycaonia, p. xxii: B.M.C. Alexandria, p. 36, n. 291, Pl. VIII.

⁸¹ So Ruggiero, De, Diz. Epig. di antich. Rom. i. p. 927.Google Scholar Cf. Mattingly, H., B.M.C. Rom. Emp. i. p. lxxiii f.Google Scholar; I doubt if his distinction between Pax Augusta as ‘Imperial Peace,’ and Pax Augusti as ‘the Peaceableness of the reigning Emperor,’ reasonable as it is in theory, held in practice. In any case we must allow for much vague ness of thought on these matters: perhaps many who used these epithets had no clearcut views of their precise meaning. It should be noted that Pax Augusta was in the first century of our era very closely connected with the memory of Augustus himself. Claudius declined a statue to Claudiana Pax Augusta (Bell, H. I., Jews and Christians in Egypt, p. 24Google Scholar, where is printed P. Lond. 1912, 1. 34 f.

⁸² As Dess. 368, 434 and coins indexed in Cohen ², viii. p. 370 s.v. Cf. on the coins of Amastris, (B.M.C. Pontus, p. 7Google Scholar, n. 4).

⁸³ As Dess. 3786, Cohen ², viii. p. 369 f.

⁸⁴ We may perhaps include among the results of this mental attitude the popularity of belief in young and approachable incarnations of the old gods, invested with full vigour, sang the Athenians in honour of Demetrius Poliorcetes (carm. ap. Athenae. vi. 253 ff.). If we cannot see Asklepios (except occasionally by special grace), we can behold Glycon, (Lucian, , Alex. 43Google Scholar, well discussed by Weinreich, O., Neue Jahrb. 1921, p. 145Google Scholar).

So the Ptolemies and the Attalids were honoured with the title of (cf. von Prott, H., Ath. Mitth. XXVI. 1902, p. 187Google Scholar), Arsinoe, (Bull. Soc. Alex. i. p. 40Google Scholar, n. 6) as Cleopatra as (B.M.C. Galatia, p. 158, n. 53–6, Pl. XIX. 3 [Antioch], B.M.C. Phoenicia, p. 53, n. 14, Pl. VII. 9 [Berytus], probably as Ἀφροδίτη (P. Oxy. 1628, 8, 1629, 7, 1644, 8, mentioning an ), Mithridates the Great as (Posidon ap. Athen. V., p. 212 D, cf. Dittenb. O.G.I. 370), Antioehus I. of Commagene as (O.G.I. 383, 61). Rome followed Hellenistic precedent. Not merely the triumvir Antonius (Plut. Anton. 60) but also Trajan (Prott, l.c. p. 183, inscr. 265), Hadrian, (Le Bas, 1619)Google Scholar, and Verus, (K.P. ii. 125)Google Scholar were honoured as Caligula, (I.G. Rom. iv. 145)Google Scholar, Nero, (A.P. ix. 178, etc.)Google Scholar, Hadrian, (I.G. Rom. iv. 1551)Google Scholar, Caracalla and Geta jointly (B.M.C. Ionia, p. 89, n. 292 [Ephesus]), Septimius, Severus (J.H.S. iv. p. 424)Google Scholar as Nero as (B.M.C. Alexandria, p. 20, n. 171, Pl. XXVI). On a contorniate figured by Cook, A. B., Zeus, ii. p. 1128Google Scholar, fig. 956 (app. M.) he is at one and the same time and ), as also Antinous (cf. Ganschinietz, , Pauly-Wissowa Suppl. iii. p. 47Google Scholar); Antinous also as (B.M.C. Lycaonia, p. 189, n. 159 [Tarsus]; cf. B.M.C. Mysia, p. 4, n. [Adramyteum]; Hadrian as (I.G. Rom. iv. 341). Caesar, Gaius (I.G. Rom. iv. 1064, 1094)Google Scholar and Germanicus (ib. 74, 75) receive the bare title of Of the women of the Imperial house Julia was νέα Ἀφροδίτη (ib. 319), as was also Drusilla (ib. 78 b, 145), and (ib. 464), Livia (ib. 180), Sabina (ib. 1492), Plautina, (Le Bas, 1703Google Scholar, and B.M.C. Caria, p. 12, 19 [Alinda]), and Domina, Julia (I.G. Rom. iii. 856Google Scholar, iv. 851) as earlier again Livia (ib. 319).

Accordingly when Horace speaks of Augustus as Mercury in human form (Carm. i. 2, 41, cf. Six, J. P., Rev. Arch. (1916), iv. p. 257 ff.Google Scholar) he is not uttering the casual flattery of a Court poet, but rather what would in the Greek East be a commonplace.

Such divinity was not the privilege of kings and princes alone. A benefactor of Teos is called (Le Bas, 108), a benefactor of Dorylaeum (I.G. Rom. iv. 527, 2): cf. also two inscriptions quoted by Weinreich, , A.R.W., xviii. p. 23Google Scholar, Jahresh. xiii. Beibl. p. 42 [Erythrae] (unless he is a local deity recognised as a reincarnation of Asklepios), and I.G. V. i. 493 [Sparta] But we must separate from these men to whom the title is applied after death; it means little if any more than ‘lately departed’ (as Le Bas, 793, 1723b, I.G. Rom. iv. 453, and on the coins of Mytilene B.M.C. Troas, p. 199, n. 164, Pl. XXXIX. 5). Special, again, is the application of to the dead (cf. Cl. Rev. 1924, p. 108₁₀: Lesbonax on the coin just quoted is represented in the character of Dionysos).

À propos of Nero as it may here be noted that he was the first emperor to be represented in his lifetime on coinage with the radiate crown symbolical of the Sun, (B.M.C. Rom. Emp. i. p. lxiv, clxxiii, 217Google Scholar, and Pl. XLIII., 4, etc.), given to Augustus only after his death (ib. p. 128, 145, etc.), but to Ptolemy III. Euergetes I. of Egypt [Head, , H.N. ² p. 853Google Scholar] and to Antioehus VI. of Syria (ib. p. 766) while alive. For its significance cf. Stat. Theb. i. 27 ff.

⁸⁵ Gallandi, , Bibliotheca veterum patrum, iii. (1767) p. 449Google Scholar = Migne xlvi. p. 913 D—( is there p. 917 A).

⁸⁶ What has been here attempted is an account of this aspect of pagan belief in a particular period. The larger question of the relation of such faith in divine power to mythology, the growth and decline of the latter, and the extent of its real penetration of popular circles requires and deserves a special study.

⁸⁷ So ex oraculo in Dess. 3982–4, secundum interpretationem oraculi Clarii Apollinis (ib. 3230 a, b, also K.P. i. 16, cf. Picard, , Éphèse et Claros, p. 716 ff.Google Scholar), in B.C.H. vii. 276 (it may be noted that παλλακεύσασα implies that the woman was regarded as the god's consort; παλλακή is not a synonym of ).

⁸⁸ Lairmenos, Helios Apollo, C.B. i. p. 147Google Scholar, n. 38, Hermes, ib. p. 181, n. 68, Parthenos, , Inscr. Oris. Pont. iv. 85Google Scholar, Hagne, Thea, Stud. Pont. iii. p. 74Google Scholar, n. 65, (probably the rider-god) S.E.G. ii. 405, and in Latin texts Juppiter, Dess. 3005, Castor, ib. 3392, Pluto and Proserpine, ib. 3972, Silvanus, ib. 3523, Nemesis, ib. 3738, Fata diuina, ib. 3757, Sol, ib. 3942, Mithras, ib. 4203, Isis and Serapis, ib. 4375, Virtus Bellona, ib. 3803, Mars, Jahresh. xii. (1909) Beibl. p. 192, n. 35, Caelestis, Eph. Epig. ix. 436, Iuppiter Dolichenus, Dess. 4309, Iuppiter Heliopolitanus, ib. 4286, Iuppiter Sabazius, ib. 9277, Veror, ib. 4507.

On commands given to sick people sleeping in temples to be cured, cf. Deubner, De incubatione. Appearances of deities in sleep would be the commonest form of revelations (cf. Le Bas, 186 [three apparitions of the Thesmophoroi] and on appearances of Asklepios, , Origen, , In Cels. iii. 24Google Scholar): we may also note P. Oxy. 1381, 102 ff. 138, where the writer's mother is said to have seen with waking eyes what he simultaneously saw in a dream. Belief in divine epiphanies was widespread: cf. Max. Tyr. ix. 7 (claims to have seen Dioscuri, Asklepios Herakles), Le Bas, 137, 4 (Ephesian Artemis), Inacr. Oris. Pont. (Latysehev), i. 184 (the of the Chersonese), and in general Pfister, , Pauly-Wissowa Suppl. iv. p. 277 ff.Google Scholar

⁸⁹ A dedication at Nicopolis to the (I.G. Rom. i. 568).

⁹⁰ Apul. Met. xi. 22, cf. 29, et ecce post paululum tempus inopinatis et usquequaque mirificis imperiis deum rursus interpeller et cogor tertium quoque teletam sustinere, Julian, , Caesares, p. 336 CGoogle Scholar,

⁹¹ x. 32, 13: cf. Juv. vi. 526, si candida iusserit Io ibit ad Aegypti finem … 530, credit enim ipsius dominae se uoce moneri.

⁹² Met. xi. 15: cf. P. Oxy. 1381, 205, [Asklepios-Imouthes) Le Bas, 1890, (for κυρίας cf. n. 93), Somnium Nectanebi (Erst. frag. pap. p. 40), iii. 5, (of Ares) Man is P. Leid. W. col. ii. 32, P. Lond. 121 l, 812 W. = 746 K., B.C.H. xxi. p. 60, also I.G. Rom. i. 101, (for the wider sense of the word cf. Hepding, H., Pauly-Wissowa, viii. p. 1467Google Scholar, and in connexion with Serapis, , Ramsay, , C.B. i. p. 147Google Scholar₂).

⁹³ Used of Zeus in Syria, Marnas, Helios, the Ephesian Artemis, Isis, Serapis, the Dea Syria, etc. (cf. Williger, , Pauly-Wissowa, xii. p. 176 ff.Google Scholar). A Greek use is in the cult of Poseidon at Troezen (Pausan, ii. 30, 6).

⁹⁴ So Helios, (Le Bas, 2393Google Scholar, cf. P. Parthey, i. 297), Glykon (Lucian, , Alex, 43Google Scholar): this title, like κύριος, occurs in genuine Greek contexts as a predicate, not it seems as a fixed cult epithet (Williger, l.c. p. 180, 30 ff.).

⁹⁵ So Men constantly (as Le Bas, 668, 685), Attis (in the dedications Attidi Meno tyranno, as Dess. 4146, 4148): cf. Herodas, V. 77, with Rose's, H. J. note, C.Q. 1923, p. 32 f.Google Scholar, P. Parthey, ii. 88, and Preisendanz, , Roscher, v. p. 1455 ff.Google Scholar

⁹⁶ Cf. K.P. ii. 204, (apparently the village was temple property). Juppiter Dolichenus in Imperial times was made to wear the dress of a Koman Emperor as earlier of an Oriental king (Cumont, , Études Syriennes, p. 188Google Scholar, Syria, i. p. 183 ff., ib. p. 185₂, on similar representations of Iahribol, Hadad and Men). For the use of a Royal Court as an analogy for heaven cf. [Aristot.] p. 398 a, 11 ff. (comparison of God with Xerxes; parallels are given by Capelle, W., Neue Jahrb. 1905, p. 556Google Scholar₁).

Cybele and Bellona had hastiferi, δορυφόροι probably originally the bodyguard of the priest-king in processions (cf. Cumont, , C.R. Acad. Inscr. 1918, p. 317 ff.Google Scholar); Apollo Archagetes at Hierapolis in Phrygia had (Hogarth, D. G., Journ. Phil. xix. p. 80Google Scholar, n. 2, 1. 5 = Ramsay, , C.B. i. p. 115Google Scholar, n. 19 = Judeich, , Altertümer von Hierapolis [Jahrb. Erg. H. iv.] p. 119, n. 153)Google Scholar, who doubtless carried standards in processions in his honour (so Hogarth, Judeich). Such an institution makes militia sacra (on which cf. Reitzenstein, , Hell. Mysterienreligionen ₂, p. 71Google Scholar ff.) much more real. It should be noted that this κύριος-reverence of the Hellenistic East has affected even dedications to the native deities of the West.

⁹⁷ Cf. earlier Pind., P. iii. 85, Callim. H. i. 70, and in the Hellenistic age above all the Nemroud-Dagh reliefs, showing Antiochus I. of Commagene, νέα Τύχη being greeted in friendly fashion by Mithras, by Zeus Oromasdes, and by Herakles (Reinach, , Rép. Rel. i. p. 195Google Scholar; cf. Cook, A. B., Zeus, i. p. 742 ff.Google Scholar). Rostovtzeff interprets a. Karagodeuashkh rhyton and other objects as showing Mithras conferring divine right on a monarch (Rev. Ét. gr. 1919, p. 476, Iranians and Greeks in South Russia, p. 104 ff.¹⁵⁹. and Minns, E. H., Scythians and Greeks, p. xxxixGoogle Scholar). Certainly divine right is at home in Imperial Mithraism: cf. Cumont, , T.M. i. p. 328Google Scholar, ii. p. 462, 13, C.I.L. iii. 4413.

⁹⁸ Sallustius, iv, cf. Cook, A. B., Zeus, i. p. 742₁Google Scholar: Cumont, , Syria, ii. p. 40 ff.Google Scholar: T.M. ii. p. 240, n. 82, p. 450, n. 10, fig. 403. Cf. for the association of the zodiac with Mithras, , T.M. ii. p. 393Google Scholar, n. 73, with the Mithraic Kronos, ib. p. 325, n. 220, with Bacchus, , Cumont, , Dar.-Sagl. v. p. 1057Google Scholar₃.

⁹⁹ Dess. 4316: ib. 2998: ib. 4333.

¹⁰⁰ Apul. Met. xi. 25.

¹⁰¹ V. 2, p. 220, 27, vii. 3, p. 271, 28, both quoted by Cumont. For as a Stoic word cf. Schenkl's index to his Epictetus, p. 680 b, s.v., also Epict. iii. 24, 34, For militia cf. also Kaibel, Epigr. 650, 12 (of the blessed dead contrasted with the unblest dead), says the departed man, now like the Dioscuri (1. 3): the θεός would normally be Hermes (though Mithras also performed this function, and was in consequence identified with Hermes, cf. Cumont, , Mystères de Mithra ³, p. 146Google Scholar, ib.n. 2).

¹⁰² Cumont, O., Mystères ³, p. 157Google Scholar; on the passage after death, p. 145, cf. also, Bousset, , A.R.W. iv. p. 136 ff., 229 ff.Google Scholar, Dieterich, , Mithrasliturgie, p. 177 ff.Google Scholar: for the belief as held by Gnostics, , Bousset, , Pauly-Wissowa, vii. p. 1522Google Scholar, as in the Oracula Chaldaica, Kroll, , De oraculis Chaldaicis (Bresl, phil. Abh. VII. i.), p. 63Google Scholar, as in Hermetism, , Corp. Herm. i. 25Google Scholar, cf. iv. 8.

¹⁰³ Origen, , In Celsum, vi. 22Google Scholar (xix. p. 336 ff. Lommatzsch), tells us of ladders composed of seven metals corresponding to the seven planets and symbolical of the soul's journey, cf. Kroll op. cit. p. 63 (Chaldaic ). [Ladders, possibly Mithraic, have been found in Roman graves in the Rhineland, , Cumont, , T.M. ii. p. 525.Google Scholar The ladder is commonly associated with after-life: it is associated with Hades, Serapis, and the griffin in a bas-relief at Myra published by Benndorf, , Reisen in Kilikien, ii. p. 42Google Scholar, fig. 31 (= Reinach, , Rèp. Rel. ii. p. 107Google Scholar, 4), with death in Egypt (cf. Cumont, , Rev. Arch. 1917, v. p. 101Google Scholar₁), buried with the dead in Nepal and in Russia (Frazer, J. G., Folklore in the Old Testament, ii. p. 57Google Scholar: in Russia a seven-runged ladder is used), used as a religious symbol in apotropaic plaques (Jahn, O., Ber. Sächs. Ges. Wiss. 1855Google Scholar, Der böse Blick, Taf. v. 2): a ladder from earth to heaven occurs in Passio S. Perpetuae, iv, § 4, p. 67, 14, Gebhardt; though 70 steps to hell in Visio Beati Esdrae, Mercati l.c. p. 70, et dati sunt ei VII angeli gui porlauerunt eum in infernum super LXX gradus. Doctor Faustus saw a ladder reaching from hell to heaven, Early English Prose Romances,² ed. Thoms, W. J., iii. p. 194.Google Scholar] The seven half-circles in the mosaic floor of the chief Ostian Mithraeum (shown in Fig. 1 from a drawing kindly made for me by Mr. S. Walsh, Rome Scholar in Architecture) are undoubtedly connected with the seven planets (cf. T.M. i. p. 63), possibly also with the seven grades: were they the places where stood or knelt those being admitted to the various grades? Planetary symbolism appears again in the seven aediculae of the ex-voto scaenae frons of Augentius, Tamesius (T.M. ii. p. 354, n. 1)Google Scholar, in the seven altars in bas-reliefs (as T.M. ii. p. 277, n. 135, fig. 20), the seven candles burning before the sacred relief in the Mithraeum on the Esquilme (ib. p. 200, n. 15), etc.

¹⁰⁴ Cf. Boll, , Neue Jahrb. 1913, p. 112 ff.Google Scholar, 121, also P. Par. 835 ff.

¹⁰⁵ De antro nympharum, ch. vi.; cf. uita Commodi, 9, and Frazer, J. G., Golden Bough ³, xi. p. 277Google Scholar; ib. p. 225 ff., Belief in Immortality, i. p. 254, on mimic deaths (commonly followed by feigned resurrection as animal, G.B. ³ xi. p. 270: can we compare T.M. ii. p. 8, and the bas-relief from Konjica, (Mystères, p. 164Google Scholar, fig. 16), showing the mumming of the corax and the leo?

¹⁰⁶ iv. On rebirth cf. Deiterieh, , Mithrasliturgie, p. 159 ff.Google Scholar

¹⁰⁷ Cf. Usener, , Kl. Schr. iv. p. 398 ff.Google Scholar, Dieterich, op. cit. p. 171, 199, Kittel, R., Die hellenistischen Mysterienreligionen und das alte Testament (1924), p. 12 f.Google Scholar Among the Akikuyu in East Africa each boy when ritually reborn ‘lives on milk for some days afterwards’ (Hollis, A. C.ap. Golden Bough ³, xi. p. 262Google Scholar). A curious magical parallel can be quoted for rebirth as a little child. In the Orphic Lithica, 366 ff., the prophetic λίθος σιδηρίτης which is being treated for use (cf. supra Introd. n. 6) is washed in water from a spring, and then ‘nurtured in clean raiment like a babe (1. 370, ): cf. Eitrem, S., Lina Laukar, p. 6.Google Scholar

¹⁰⁸ Dionysiaca, ix. 111 ff.: the infant Dionysus in a cult scene is perhaps correctly recognised on an Attic vase reproduced by Pfuhl, E., Malerei und Zeichnung der Griechen, iii.Google ScholarAbb. 571, cf. ii. p. 572.

¹⁰⁹ Thanks to the kindness of Prof. A. Minto I visited this find on April 10th, 1924. A brief description, with important observations, has been published by Cumont, , C.R. Acad. Inscr. 1924, p. 113 ff.Google Scholar

¹¹⁰ The force of καθιέρωσαν is that the tomb is put under the deity's protection, cf. Mordtmann, , Ath. Mitth. x. p. 17.Google Scholar

¹¹¹ The subjects of these texts are not kings or horses (for their τιμή from the gods cf. Hom. Il. i. 505, xi. 43, Hes. Theog. 81 etc.), and their τιμή cannot well be as vague as that implied in an inscription at Philae (Kaibel, , Epigr. gr. 980Google Scholar) and in a Ciarian oracle ap. K.P. i. p. 17, addressing the men of Trocetta as nor, again, can here mean ‘buried at the expense of the temple funds of …’, as in K.P. i. n. 192, p. 89

¹¹² Cf. Ramsay, , C.B. i. p. 135, 147.Google Scholar

¹¹³ Dess. 4167–4170: Apul. Met. xi. 13, 16: cf. the Hermetic use in [Apul.] Ascl. 22.

¹¹⁴ Dess. 4210–1, 4262, 4142: cf. the sacratae of Caelestis, ib. 4438.

¹¹⁵ Dess. 4181–2: Juv. iv. 123.

¹¹⁶ Ramsay, , C.B. i. p. 147.Google Scholar

¹¹⁷ Cf. Ganschinietz, , Pauly-Wissowa, x. p. 2534.Google ScholarC.I.G. 3163, P. Lond. 44 (i. p. 34.18) afford strong evidence for their religious character, though the term κάτοχος clearly covered a category of persons who had taken refuge in the temple as an asylum. von Woess, F., Das Asylwesen Ägyptens in der Ptolemäerzeit (= Münchener Beiträge zur Papyrusforschung, v. 1923, p. 146 ff., 237 ff.)Google Scholar, has urged with force that the term is used purely in the latter sense. But there remain the astrological texts, treated by Kroll, W., Klio, xviii, p. 219 f.Google Scholar; these are quite inconsistent with such an explanation, and κάτοχος in these means ‘possessed’ (cf. also Bell, H. I., Journ. Eg. Arch. x. p. 154Google Scholar). The κάτοχοι at Baetocaece are clearly religious (Ditt. O.G.I. 262, 26, interprets k. as ‘tenants,’ but cf. Steph. Thes. iv. p. 1369 B, and the use of κατοχή in Iambl. De myst. iii. 9, p. 119, Parthey); it may be noted that the high priest was supposed to be directly appointed by the god (the inscription, 1. 12), and that the shrine was famous for special manifestations of divine power (1. 4).

¹¹⁸ Le Bas, 519, 9; cf. Waddington's note.

¹¹⁹ Ditt. O.G.I. 441, 55.

¹²⁰ Dess. 4436.