² In my edition of the Prometheus, p. 179.

³ Aesch. Theb. 1, 606, Cho. 580, Eum. 277–8, fr. 208, Eur. fr. 413, cf. Aesch. Ag. 1301. In Aesch. Eum. 277 Herwerden's πολλοῖσι καιροὺς for πολλοὺς καθαρμοὺς is tempting for two reasons: (1) the mention of καθαρμούς seems out of place; (2) of the two interpretations recorded in the scholium, the second seems to have been written to καιρούς: .

⁴ Theogn. 813 , Strattis 67 (Kock) , cf. Hesych. .

⁵ Stob. Fl. 34. 7 , Philostr., Apoll. I. 1Google Scholar, A.P. x. 46, xiv. 1, xvi. 325–6, Greg. Naz. Or. 23. 535 c.

⁶ Soph. O.C. 1052–3 . Jebb remarks that κλής ‘cannot well be rendered key here … there is no evidence for the Eleusinian hierophant putting a key on the lips of the initiated.’ No, nor an ox either. His difficulty disappears when it is realised that these are religious symbols, not realities. In this passage the metaphor is mixed: βοῦς βέβηκεν and κλῂς ἐπίκειται have coalesced.

⁷ Hence ἀθυρόγλωττος as an epithet of an indiscreet speaker, like Niobe: A.P. xvi. 132 , cf. Eur. Or. 903, Ar. Ran. 838 ἀπύλωτον στόμα.

⁸ Cf. Heliod. vi. 15 .

⁹ For the mystic vow of silence see Alcaeus 27. 4, Philostr., Apoll. i. 15, 17Google Scholar, Lucian. i. 458, Ach. Tat. i. 10. 5, 11. 16, Plut. M. 10 F, etc.

¹⁰ Cho. 941–2: ἐπολολυξάτω corr. Seidler.

¹¹ Eum. 82–3, cf. Cho. 1057 where I would read , see my note in CQ. 1934, pp. 77–8.

¹² Firmicus de err. prof. rell. xxii.

¹³ For further references see Lobeck, , Aglaophamus, pp. 69–74Google Scholar. In the Phaedo Plato is probably thinking primarily of Orphic beliefs, but what he says of them makes it clear that the Orphic view of the fate of the soul in the other world was substantially the same as the Eleusinian; for in the Frogs, where the same view is expounded, there is no doubt that Aristophanes was thinking of the Eleusinian Mysteries (see Schol. 159, 314, 320, 343, 395, 401, 408). The close connexion between the two cults in this respect does not seem to have been fully appreciated by archaeologists, who tend to regard as peculiarly Orphic what was really common to both.

¹⁴ Theon Sm. Math. i. 10 . Theon does not expressly say that he is describing the Mysteries of Eleusis, but every detail of his description shows that he is: see Apollod. ii. 5. 12, Diod. iv. 14, Hdt. viii. 63, and see n. 23. In particular, his θεοφιλές καὶ θεοῖς συνδίαιτον is clearly the θεοφιλοῦς ἑορτῆς of Aristophanes, a doctrine which Plato criticised in Rep. 363 c (i.e. Eumolpus) .

¹⁵ 348 δυσδαίμονες corr. Stanley: for the error cf. Eum. 911 τῶν δ εὐσεβούντων Heath for τῶν δυσσεβούντων.

¹⁶ Cf. 532, 535, 919–20, Hdt. i. 32, Soph. O.T. 1528–30, Eur. Tro. 509–10, 268, Andr. 100–2, Or. 1338 (malicious as here), F.Ph.G. i. pp. 212, 214, Dio Chr. 473 R.

¹⁷ Ag. 1269, 1579, Cho. 1, 487, 581, 689, 983, 1061, Eum. 220, 224.

¹⁸ Harpocr. , cf. Ar. Ran. 745 and Schol. . Theon of Smyrna (loc. cit.) counts ἐποπτεία as the third grade, having counted κάθαρσις as the first instead of regarding it as preliminary.

¹⁹ Cf. Heliod. ix. 9 . The Eleusinian temple was called τὸ ἀνάκτορον, Ath. iv. 167 f, and for further references see Lobeck, pp. 188–9.

²⁰ Numerous examples are collected by Lobeck, pp. 127–31.

²¹ 582 ὀρθώσαντι corr. Pearson. Tucker takes τούτῳ to refer to Hermes, an image of whom he supposes to stand upon the stage. That there was such an image is very probable; but it is hard to believe that Orestes would refer to the God as τούτῳ, and it is necessary for us to know what part in the plot is to be assigned to Pylades.

²² It is possible that in ἀγῶνας we have an allusion to the mystical πόνος τε καὶ ἀγών: see Plat., Phaedrus 247Google Scholar a quoted below.

²³ Plut. de anima fr. vi. 4, Headlam, , Choephoroe, p. 50Google Scholar (p. 266 in the collected edition). The reference to στέφανος in this and other passages quoted in this article (Plat. Rep. 363 c, Ar. Ran. 156, Theon Sm. i. 10) should be sufficient to fix the meaning of the word in the inscriptions on the Orphic plates. For the βόρβορος as an Eleusinian conception see n. 62. Here it is combined with ἐποπτεία (ἐφορῶν).

²⁴ Ar. Ran. 154–5 , Lucian i. 644 ; In the first passage the sights and sounds of the other world are described with reference to the corresponding features in Eleusinian ritual; in the second, where the scene is laid in Hades, the features of Eleusinian ritual are recognised in the sights and sounds of the other world. Thus the other-world pageantry of Eleusis was as old as Aristophanes. And now add Plat. Phaedo 108 a νῦν δὲ ἔοικε (sc. ἡ πορεία) (or ὁσίων) . Surely it is clear that here too the reference is to Eleusinian ritual, and not to the Orphic tablets (Guthrie, W. K. C., Orpheus and Greek Religion, p. 176Google Scholar).

²⁵ Cf. Rep. 615 d (these are τὰ δεινὰ of Plutarch), 365 a , Plut. M. 81 e : so, with the same allusive language, 36 e , 47 a ,943 c , Aristid. i. 256 ;

²⁶ Cf. Plat. Rep. 614 e, Ar. Ran. 326, 344, 352, 372–4.

²⁷ Throughout the play Orestes, as καθαρτὴς of the house, is likened to Heracles, καθαρτὴς of the earth: see Headlam on 159–63.

²⁸ τριβᾶς corr. Stanley.

²⁹ Headlam ad. loc. and cf. Eum. 390 δυσοδοπαίπαλα.

³⁰ 960 οἴκων corr. Rossbach.

³¹ CR. 1900, p. 113.

³² Cook, A. B., Zeus, i. p. 219 f.; 425 f.Google Scholar, Harrison, J. E., Prolegomena, pp. 229–30Google Scholar, and cf. Dem. de cor. 259 . The candidate for initiation was also veiled, Harrison loc. cit., p. 547; hence the allusion in Cho. 805–7 : cf. Plat. Phaedr. 250 b , Theaet. 155 e. The same conception has influenced the language of Ag. 1177–9

³³ Aesch. P.V. 693 ἄφετον ἀλᾶσθαι, Horn. Il. vi. 308–9 , Od. iii. 383 .

³⁴ Firmicus de err. prof. rell. xxii. His nocte quadam is clearly an allusion to the celebrated νύκτες Ἐλευσίνιαι: see Lobeck, p. 651, and add the passage from the Φιλοσοφούμενα quoted by Harrison, J. E., Prolegomena, pp. 548–9Google Scholar.

³⁵ 965 and 967 corr. Headlam, the latter being a common error discussed by him in CR. xv. p. 17 f. Thus restored the phrase provides δυσοίμον τύχας (944) with a balance and a contrast (just as balances ). Wilamowitz reads , which is rhythmically impossible.

³⁶ Cf. also Φιλοσοφούμενα loc. cit. .

³⁷ Headlam, CR. 1900, p. 113.

³⁸ It was already nightfall at 656–8. The play began before dawn, and the day has now passed into the darkness of another night. Torches would be effective, even in daylight, on a stage which faced north.

³⁹ JHS. 1906, pp. 268–77Google Scholar.

⁴⁰ 1476 νῦν δ΄ corr. Headlam. With these and other passages may be compared Phaedr. 244 d–c, where Plato is clearly thinking of the Oresteia: . The recurrent δὴ marks the allusions to familiar ideas: cf. Burnet, on Phaedo 107 d 7Google Scholar.

⁴¹ 1574 corr. Headlam.

⁴² J. W. Mackail in CR. xix. p. 197.

⁴³ In his note on Soph. El. 448 Jebb quotes several passages in which σὺ δέ marks a n antithesis of clauses, not of persons; but here there is not even an implied antithesis between the two clauses.

⁴⁴ Aristid. i. 182 , cf. Plut. M. 795 e, 378 a.

⁴⁵ Liban. 356 , cf. Hdt. viii. 63, Theon Sm. i. 10 (supra n. 14). The two qualifications mentioned in the last passage are parodied in Ar. Ran. 354–5 . For further references see Lobeck, pp. 15–16, 190.

⁴⁶ Lys. Andoc. 51.

⁴⁷ Abel, Orph. pp. 153, 148, Gnom. p. 128, Brunck (cited by Headlam in a MS. note), Diog. L. viii. 7; cf. Plut. M. 636 d , .

⁴⁸ I am here following two MS. notes in the fullest and most important of Headlam's three interleaved Wecklein's (in King's College Library) which was apparently unknown to Pearson. On Cho. 450 he writes: ‘Eleusinian like Orphic, cf. Aglaoph. 382, 15, 39 [quoted above], Ag. 1035–6 .’ And on the later passage: ‘τέτραινε μῦθον Cho. 448 n. from Mysteries W. H.’ The addition of his initials indicates, in accordance with his regular practice, that he regarded the allusion as certain. Further, Dr. J. T. Sheppard informs me that he took the following note from Headlam in 1904: ‘W. H. thinks that there is a slight allusion to the Mysteries: the mystic must not be ’. It seems clear, therefore, that Headlam perceived the allusion in this passage, and that Pearson missed the point of his published note on 1034.

⁴⁹ The exact meaning of 1044–5 I take to be as follows: ‘If, being a barbarian, you do not understand Greek—then do you tell her, not with (Greekspeaking) voice, but with barbarian-speaking hand.’ καρβάνῳ χερί is contrasted with : for, as Headlam remarks on 1034, βάρβαρος and ἀξύνετος are virtually equivalent.

⁵⁰ Cf. 1033, 1036, 1038, 1055, and my note in CQ. 1934, pp. 73–4.

⁵¹ For these reasons at 810 I would accept Hermann's θυηλαὶ for θύελλαι. Ἄτη had nothing to do with whirlwinds, but the dramatic significance of ἄτης θυηλαὶ is clear: 1434 , Cho. 966 .

⁵² 268 corr. Schütz, corr. Casaubon.

⁵³ Charges of impiety (ἀσέβεια) were heard before the Eumolpidae: Dem. 601.

⁵⁴ The words exactly correspond to , because was used technically of the promulgation of a law, as in the extract from a law of Solon quoted in Plut. Sol. 19 .

⁵⁵ Cf. Dem. 790 , Eur. fr. 852 .

⁵⁶ Cf. Pseudo-Phocyl. 8, Iambi. Vit. Pyth. 37. According to Diog. L. viii. 1, Pythagoras instructed his followers , and see further Plat. Legg. 717 a–b, 724 a, 730 b, 854 e, Lycurg. Leocr. 15, 94, 97.

⁵⁷ Cf. Plat., Legg. 729 e–730Google Scholar a , .

⁵⁸ Hygin. Fab. 158.

⁵⁹ Paus. x. 28–31. Pausanias connected the painting with the Eleusinian Hades, for of the figures representing the uninitiated he says (31.11), . See further Plat., Axioch. 371Google Scholar e, Plut. V. 761 a, Diog. L. viii. 32.

⁶⁰ ibid. x. 28. 5.

⁶¹ According to Plut. M. 564 e the fate of sinners of this class was that . The doctrine of Purgatory underlies Cho. 59–63 (restored by Headlam) . Add Plat. Rep. 615 c.

⁶² This passage is itself sufficient to shew that the was an Eleusinian, as well as an Orphic, conception (see n. 13), and cf. Aristid. i. 259 (p. 22), Plut. de anima iv. 6 (p. 24). It is therefore not quite accurate to describe it as ‘a particularly Orphic form of punishment’ (Guthrie, p. 160).

⁶³ In this part of my argument I have been anticipated at several points by Headlam, whose MS. notes on Eum. 269–73 shew that he had identified what he calls the ‘three Greek Commandments,’ and had connected them with the Mysteries, though not apparently with the . Neither of these points was understood by R. Hirzel in his long and learned article on (Sachs. Abh. Philol.-Hist. Kl. xx, 1–98).

⁶⁴ Aesch. PV. 530–4 n. The point is brought out by the Furies themselves in Eum. 335–6, 395, cf. 173, 725–6.

⁶⁵ Eum. 548–52.

⁶⁶ Cf. Hes. Op. 801–2.

⁶⁷ Cf. Il. ix. 454.

⁶⁸ Lucian. i. 471, a passage in which there is no mention of Eleusis, but cf. i. 643–5 (supra n. 24), where again a Fury appears acting under the orders of Rhadamanthys. This evidence is late, but it is confirmed by the earlier. The conception of a judgment after death is clearly implied in several of the passages I have quoted in illustration of the , which, as I have shewn, were expounded at Eleusis. Guthrie, discussing Plat. Rep. 330 d–e (Orpheus, p. 153), writes: ‘Perhaps the belief in a judgment seat was an “Orphic myth.” If that is so, I have no doubt that it was not the invention of the Orphics, but something taken over by them from vague popular belief.’ It was taken over, I suggest, from the teaching of Eleusis.