^{page 411 note 1} JRAS. 1911, p. 1007, n. 3.

^{page 411 note 2} The Origin of Tragedy, Cambridge, 1910Google Scholar. It is almost needless to say that ProfessorMurray, , Rise of the Greek Epic ², p. 290Google Scholar, accepts the theory, but adds to it a reference to Schurtz, Altersklassen und Männerbunde, which already has inspired von Schroeder in his conception of the origin of drama; see Keith, , JRAS. 1909, pp. 204, 205Google Scholar.

^{page 411 note 3} c. 4. ⋯πò τ⋯ν ⋯ξαρχóντων τòν διθ⋯ραμβον and ⋯κ μικρ⋯ν μ⋯θων κα⋯ λ⋯ξεως γελο⋯ας δι⋯ τ⋯ ⋯κ σατυρικο⋯ μεταβαλεῖν ⋯ψ⋯ ⋯πεσεμν⋯νθη.

^{page 411 note 4} See references in Farnell, , Cults of the Greek States, v, 233Google Scholar; cf. also Dieterich, , Archiv für Religionswissenschaft, 1908, pp. 168Google Scholar, 169.

^{page 412 note 1} See Usener, , Archiv für Religionswissenschaft, 1904, pp. 303 seqGoogle Scholar.

^{page 412 note 2} Schol. Plat. Symp. 208 D; Sohol. Arist. Acharn. 146.

^{page 412 note 3} See Dawkins, , Journ. Hell. Stud. 1906, pp. 191–206Google Scholar, and see also Wace in Ridgeway, pp. 16–24.

^{page 412 note 4} Farnell, op. cit. v, 230–6.

^{page 412 note 5} Farnell, v, 161 seqq.

^{page 412 note 6} Op. cit. pp. 1–108. The criticism of Dr. Farnell is at pp. 73–93.

^{page 413 note 1} Herodotos, v, 67. That ⋯π⋯δωκε has been rendered “restored” is perfectly natural and very possibly what Herodotos meant. It is absurd to say, as does Professor Ridgeway (p. 28, n. 1), that ⋯πςδ⋯δωμι means “assign”; Liddell & Scott are perfectly correct in taking the normal sense as “to render what is due”, which gives the sense of “restore”, and the zeugma is one of the least difficult possible.

^{page 413 note 2} That there was anything but a dance is not certain; χςρóς has no necessary allusion to more, and τραγικóς may refer to the dancer's dress.

^{page 414 note 1} Paus. viii, 23; see Farnell, v, 231.

^{page 414 note 2} See Dieterich, op. cit. pp. 168, 169. The idea is that of MissHarrison, , Prolegomena, p. 421Google Scholar; cf. Farnell, v, 232, n.

^{page 414 note 3} Op. cit. pp. 94, 95.

^{page 414 note 4} Op. cit. pp. 95–100.

^{page 414 note 5} Op. cit., pp. 100–2.

^{page 415 note 1} The Origin of Tragedy, pp. 102–6.

^{page 415 note 2} Ibid. p. 106.

^{page 415 note 3} Ibid. p. 84.

^{page 415 note 4} Clearly this can hardly be taken seriously.

^{page 415 note 5} He knows only a lecture of May, 1909, before the Hellenic Society; see Journ. Hell. Stud. 1909, p. xlvii.

^{page 416 note 1} Archiv für Religionswissenschaft, 1904, pp. 303 seqq. The importance of this paper for the origin of drama in Greece and in India was first (so far as I am aware) pointed out by me in JRAS. 1908, p. 172, and in 1909 by Dr. Farnell, v, 235; Dietericb, op. cit., ignores it.

^{page 416 note 2} The preface is dated August 6, 1910.

^{page 416 note 3} ZDMG. lxiv, 534 seqq.; JRAS. 1908, p. 172; 1911, p. 1008; Classical Quarterly, iv, 283, 284.

^{page 416 note 4} Cf. my notes in JRAS. 1908, pp. 169 seqq. The human character of Kṛṣṇa is not older than the divine: it is an essential characteristic of vegetation spirits that they take temporary embodiments in man or animal; compare the extremely human character of the Greek Dionysos.

^{page 416 note 5} Established beyond all doubt by Frazer in The Golden Bough.

^{page 416 note 6} Journ. Hell. Stud. 1894.

^{page 416 note 7} Op. cit. v, 168.

^{page 417 note 1} Cults of the Greek States, v, 78 seq.

^{page 417 note 2} e.g. Servius ad Æn. viii, 343: “caper quæ est hostia Libero propria,” which is of special value as showing the intimate connexion; see also Farnell, v, 303.

^{page 417 note 3} The assertions of Arnobius, (adv. Nat. v, 19)Google Scholar and Lactantius Placidus (ad Stat. Theb. v, 159) are supported by the inscription at Mykonos in Ditten, b. Syll. 373Google Scholar, 27, which seems to refer to a sacramental meal.

^{page 417 note 4} Paus. ix, 8. 1, 2. Dionysos was Melanaigis at Eleutherai and Hermione and in the Apatouria.

^{page 417 note 6} Op. cit, p. 81.

^{page 418 note 1} Farnell, , Cults of the Greek States, ii, 434 seqq.Google Scholar; , Arist.Lys. 645Google Scholar.

^{page 418 note 2} Farnell, op. cit. iv, 26; v, 233; Athen. 425 E.

^{page 418 note 3} Hesychius, s.v.

^{page 418 note 4} Farnell, op. cit. v, 233, 328; Paus. ii, 23. 1.

^{page 418 note 5} Op. cit. v, 233.

^{page 418 note 6} Op. cit. p. 108.

^{page 418 note 7} te 'pi hi teṣāṃ (i.e. Kṛṣṇa, Kaṃsa, and their followers) utpattiprabhṛty ā vināśād buddhīr vyācakṣāṇāḥ sato buddhiviṣayān prakāśayanti; ataś ca sataḥ vyāmiśrā dṛśiyante kecid Kaṃsabhāktā bhavanti kecid Vāsudeva bhaktāh; varṇānyatvaṃ khalv api puṣyanti, kecid kālamukhā bhavanti, kecid raktamukhāḥ; see Weber, , Ind. Stud. xiii, 354 seqq., 488 seqqGoogle Scholar.

^{page 419 note 1} The Origin of Tragedy, pp. 80 seqq.

^{page 419 note 2} Fragm. 72. According to Bywater, , Aristotle on the Art of Poetry, pp. 134Google Scholar, 135, the sense is “compose and teach to the chorus” as in Herodotus, i, 23, where διδ⋯σκειν is used. This interpretation makes no difference to the argument. It should be noted that Archilochos' dithyramb may have been a literary form, not sung by satyrs. But this we do not know.

^{page 419 note 3} Op. cit. p. 38.

^{page 419 note 4} Suidas, s.v. Arion; Solon in his elegies is said to have referred to him as introducing τ⋯ς τραγῳδ⋯ας δρ⋯μα; see Rabe, , Rhein. Mus. lxiii, 150Google Scholar; Dieterich, op. cit. p. 170.

^{page 419 note 5} Fragm. 207.

^{page 419 note 6} Cyclops, 74–81. On the other hand, satyrs as opposed to actors in the satyric drama were horse-shaped; see Ridgeway, p. 72.

^{page 420 note 1} Ol. xiii, 18, 19; the reference is clearly to Arion.

^{page 420 note 2} Simonides(556–467 b.c.) is said to have composed dithyrambs called Europa and Memnon; see Fragm. 27, 28; Strabo, p. 619, 43. But even Lasos is not said to have composed non-Dionysiac dithyrambs, though Ridgeway, pp. 8, 9, assumes that he did. Both are too late to be evidence of the early dithyramb. The nineteenth (eighteenth) of Bacchylides' Odes was held by Kenyon (p. 185) to be a dithyramb because of the introduction of an allusion to Dionysos' birth; of. Plato, , Legg. 700 BGoogle Scholar, where the dithyramb is associated with the birth of Dionysos. Jebb, pp. 38 seqq., classifies xiv–xix as dithyrambs, xiv and xviii being really connected with Dionysos, xv and xvii perhaps performed by a chorus (xvii is in dialogue), and the other two being merely formally so called.

^{page 420 note 3} Arist. Poet. c. 4; cf. Archilochos, loc. cit.; Pollux, iv, 123, who tells us that even before Thespis some one got upon an ⋯λ⋯ομ or ancient table, and held a dialogue with the members of the chorus. Aristotle, as reported by Themistios (Or. xxvi, p. 382), ascribed the introduction of ῥ⋯σις to Thespis.

^{page 421 note 1} This is seen in the Bakchai, with its triumph of Dionysos and the agony of Pentheus. The two sides are adumbrated in the tale of Lykourgos, , Il. vi, 132 seqqGoogle Scholar.

^{page 421 note 2} τραγικο⋯ τρóπου εὑρετ⋯ς; cf. Mahaffy, , Greek Classical Literature, I, i, 221, 222Google Scholar. The question of the tragic character of tragedy is explained by Dieterich, op. cit. pp. 163–96, as due to the fact that in addition to the Dionysiac side of tragedy there was the element of threnoi as seen at the Anthesteria, a public mourning for the dead; the masked dancers are the spirits of the dead. Dionysos is surrounded by the souls of the dead; in spring the earth becomes fruitful anew and the souls of the forefathers arise; Dionysos, the god of the fruitfulness, of the new life, is god of the dead; the phallic demons of fruitfulness and the soul demons are one, a view which Murray accepts as well as Ridgeway's theory. This theory is, indeed, really more probable than Ridgeway's but is not so simple, and it is fair to add that Dieterich (pp. 181–6) adds as a probable influence the dromena at Eleusis, the tale of Persephone and Demeter. But he definitely refuses to admit that there was any trace of the death of a god, whether in goat or bull form (p. 175); the epiphania of Dionysos shows him entering the city on his ship (borne in a car), for he has come from afar to bring new life. But this theory has the great disadvantage of ignoring the peculiar tradition of the god, seen clearly in the Bakchai and in the prevalence of Pentheus as a title of drama, and it ignores also the overwhelming evidence adduced by Frazer for the killing of the spirit of vegetation and the assumption by the spirit of vegetation of animal shapes. Nor is its view of the dancers on the whole supported by adequate evidence; it remains a hypothesis, and not a very attractive one.

^{page 422 note 1} Cf. Ridgeway, , The Origin of Tragedy, pp. 58 seqGoogle Scholar. The attribution of dramas to Thespis is uncertain because of the tradition that Herakleides Pontikos wrote tragedies and ascribed them to him. Ridgeway thinks that he divorced the chorus and dithyramb tragedy from connexion with a single tomb and carted it round (cf. Horace, , A.P. 275, 276)Google Scholar to fairs and markets as a pastime. But this is pure guesswork, and if A. P. l.c. does represent facts Thespis seems not to have attained the creation of tragedy proper, as the reference is rather to comedy (cf. Christ, , Gesch. der Griech. Lit. p. 175)Google Scholar. Dieterich, op. cit. p. 174, makes the attractive suggestion that the plaustra of Horace is an allusion to the currus navalis of the epiphania of the god. More probably Thespis' merit lay in a development of dialogue between his chorus and single actor, in accordance with Aristotle's statement that he invented πρóλογóν τε κα⋯ ῥ⋯σιν to add to the simple songs of the chorus; above, p. 420, n. 3.

^{422 note 2} Weber, , Ind. Stud. xiii, 309–19Google Scholar; Keith, , Aitareya Āraṇyaka, p. 23Google Scholar.

^{page 423 n ote 1} It does not definitely appear whether the śaubhikas actually acted and spoke their parts, but the Mahābhāṣya knows of naṭas, “actors,” who speak and sing (naṭasya śṛṇoti, naṭasya śroṣyāmaḥ, i, 4. 29; agāsīn naṭaḥ, ii, 4. 77), and it is difficult to suppose that the combination of action and speech was not in use.

^{page 423 note 2} Cf. Macdonell, , Sanskrit Literature, pp. 347 seqq.Google Scholar, 414 seqq.; my note, JRAS. 1909, p. 208.

^{page 423 note 3} e.g. the vast ritual of the horse sacrifice with its great animation, the Rājasūya, the Vājapeya, and others.

^{page 423 note 4} Kāṭhaka Saṃhitā, xxxiv, 5; Pañcaviṃśa Brāhmaṇa, v, 5, 14 seqq. (in the comm. on 14 read parimaṇḍale śvete); Āpastamba Śrauta Sūtra, xxi, 19. 9–12; Keith, , Śāṅkhāyana Āraṇyaka, p. 78Google Scholar.

^{page 424 note 1} Cults of the Greek States, v, 237.

^{page 424 note 2} Epigenes of Sikyon is invoked as a producer of tragedy which did not deal with Dionysos and as developing the tragic choruses (Ridgeway, pp. 58, 67, 68). But we know all but absolutely nothing about him, and what we do know (Zenob. v, 4; Suidas, s.v. ςὐδ⋯ν πρòς Διóνυσον) is based on the assumption that tragedy in his day dealt with Dionysos and not with the deaths of heroes. Moreover, it is extremely doubtful if he did more than write dithyrambs; so Haigh, , Tragic Drama of the Greeks, pp. 22, 25Google Scholar; Mahaffy, , Greek Classical Literature, I, i, 223Google Scholar. The point which Ridgeway ignores is that the dithyramb had already been extended to other topics before tragedy first arose, and therefore had no need to stick to Dionysiac topics. His account of the orthodox origin of drama (p. 2) is really quite unfair. It is easy to overthrow an imaginary opponent.

^{page 424 note 3} Ridgeway, pp. 75–7.

^{page 424 note 4} The exact points in the development must be uncertain; Haigh, op. cit. pp. 19–21, takes the view that Arion introduced conversations between the chorus and the leader (cf. Pollux, iv, 123), but that depends on pressing the word λ⋯γοντας in Suidas, s.v. Arion, and this development seems rather, as held by Aristotle, to be due to Thespis. According, however, to Rabe, , Rhein. Mus. lxiii, 150Google Scholar, a fragment mentions Solon in his elegies as ascribing to Arion some connexion with tragedy (τ⋯ς δ⋯ τραγῳδ⋯ας πρ⋯τον δρ⋯μα . . . εἰσήγαγεν). But this need not mean more than that he was τραγικο⋯ τρ⋯που εὐρετής, as Suidas tells us. Comedy proper in Greece has a different origin in ritual cathartic cursing; see Farnell, v, 211, 212. For such ribaldry we have a curious parallel in the Vedie αἰσϰρολογ⋯α in the horse sacrifice (Taittiríya Saṃhitā, vii, 4. 19; Kāṭhaka Saṃhitā, Aśuamedha, iv, 8; Maitrāyaṇī Saṃhitā, iii, 12, 20; Vājasaneyi Saṃhitā, xxiii, 18–32) and the exchange of abuse by a Brahmacārin and a hetaira at the Mahāvrata (Keith, , Śāṅkhāyana Āraṇyaka, p. 79)Google Scholar. Traces of this form of ritual may be seen in the farces of the later Sanskrit stage. Dieterich, op. cit. p. 167, thinks that satyric drama and comedy have one origin, but this seems only true if the ritual of the worship of Dionysos is regarded as a whole; the different sides of that worship produce different literary forms.

^{page 425 note 1} The Origin of Tragedy, pp. 7, 57.

^{page 425 note 2} c. 4.

^{page 425 note 3} Ibid.

^{page 425 note 4} τεμάϰη τ⋯ν 'Ομήρου δε⋯πνων, Athen. 347 E.

^{page 426 note 1} It is perhaps unnecessary to say that the legend about Peisistratos yields little or no support to the theory of the debt of the present form of the Homeric poems to his action (see Lang, , The World of Homer, pp. 281–8Google Scholar, and cf. Murray, pp. 212 seqq.). But there is the evidence of Isokrates, (Paneg. p. 74)Google Scholar and Lykourgos, (adv. Leokr. 102)Google Scholar for the attention paid at Athens to the epic.

^{page 426 note 2} Some minor points may be added in this note. (1) The contrast drawn between the Attic Xanthos of Boiotia and the dark-haired aboriginal Neleid Melanthos is only justifiable by the theory held by Professor Ridgeway that the aborigines of Greece were the dark-haired Mediterranean race akin to the Lycians, that they spoke Greek, and that the Achaians and other invaders were Celtic. But there is much more probability that there is a further stage, namely, before the latest invaders the Dorians (who were, no doubt, closely connected in race with the earlier Greek invaders, and like them, being Āryans, allied to the Celto-Teutonic races, though Ridgeway believes that the Dorians were dark Thracians or Illyrians, matriarchal in character), earlier Greek invaders, the Ionians, and other tribes who brought the Greek tongue and imposed it on the more primitive race, in which case the Neleid cannot be assumed to have been black-haired. It seems that the earlier civilization is that of the Mediterranean race; the reading of the Cretan discoveries will decide if they spoke Greek or not. Cf. Ridgeway, Early Age of Greece; Who were the Dorians? Burrows, , Discoveries in Crete, pp. 146–62, 163 seqq., 196seqq., 202seqq.Google Scholar; Hall, , Oldest Civilisation of Greece, pp. 94 seqq.Google Scholar; Journ. Hell. Stud. xxv, 324; Murray, , Rise of the Greek Epic ², pp. 61 seqqGoogle Scholar. It is asserted by Ridgeway (p. 120) that Orestes and his sister recognized each other by their blond hair, being Achaians from the north (so Tucker, , Chæphori, pp. lx seqq.)Google Scholar; Achaians are not Dorians in the tradition, but are opposed to the Dorians, so that we seem reduced to realize even on his own theory that there were different waves of people from the north, and the theory of a gradual penetration of Greece by Āryan peoples bringing the Greek tongue appears most probable; cf. Hall, , Annals of British School at Athens, xi, 222Google Scholar.

(2) This consideration applies with great force to Professor Ridgeway's ingenious attempt to prove that in the Supplices and the Eumenides Aischylos as a reformer is defending the recognition of the binding character of the marriage tie, the change from exogamy to endogamy, and the abolition of the matriarchy, coupled with the supremacy of the religion of Zeus and Apollo, introduced mainly through the influence of Homer. As proof of the early prevalence of matriarchy in Athens he cites the case of the Lycians (Herod, i, 173), who were, he says, allied with the Greeks in blood, a statement which he does not attempt to prove, and which is true only if we take the oldest strata of population, who were, in the ordinary acceptation of Greeks as those speaking an Āryan tongue, pre-Greek, for Lycian is generally held not to be an Indo-European speech (Kretschmer, , Einleitung in die Gesch. der Griech. Sprache, pp. 289 seqq.)Google Scholar. The Spaniards of Cantabria (Strabo, p. 137, 30) were apparently in like case to the Lycians, and there is some, though not decisive, evidence of matriarchy among parts of the Mediterranean race, though Murray, op. cit. pp. 67, n. 1, 98, with wonted lightheartedness, accepts it sans phrase, and Myres, , Anthropology and the Classics, pp. 153 seqq.Google Scholar, seems to accept it. Then he adds that in Athens in Isaios' time endogamy was usual, as shown by the rule that an “heiress” was expected to marry the next-of-kin, and indeed was no more than a burden on the estate (⋯πίκληρος) the term “heiress” is misleading in this regard. But endogamy had once been the rule under matriarchy, and a survival is seen in the rule that half-brothers and half-sisters by the same father could marry, not, however, those by the same mother. This is, however, clearly no argument at all: Attic law (like Egyptian law) to preserve the estate developed the system of allowing an intermarriage, though it never allowed (as did Egyptian law) full brother and sister to marry. It therefore allowed those with one father to marry, but not those with one mother, natural relationship prevailing over the dictates of property. But that exogamy ever was the rule in Athens is not even hinted at in any authority. In the Supplices the maidens who flee from marriage with a cousin are evidently hard for the king to understand, for he argues with them in the best Attic style, and enuntiates a principle of private international law of great interest. “If the sons of Aigyptos,” he says, “have power over thee by the law of the city, claiming it on the base of kinship, who would care to oppose them? Therefore must thou defend thyself according to the laws of thy house, on the ground that they have no power over thee,” a clear assertion that domicile is the rule regarding personal relations. When it is remembered that the king compares them to Egyptians, Libyans, or Amazons, it can hardly be denied that Aischylos is not discussing a question of burning moment at Athens. Still less does the Eumenides avail. The guilt of slaying a mother is a grave one, and the prosecution makes a strong case in favour of the close connexion of son and mother, but the triumph of the opposite view is complete and doubtless in harmony with Athenian feeling, which sympathized, despite its love to succour the oppressed, with the splendid falsehood of Hypermnestra. Nothing can be made of the argument used in the Mumenides, 201 seqq., that a woman is not kindred to her husband, for the Attic law even in the fourth century recognized that a woman by marriage did not cease to be of her father's family; the father could, if he liked, divorce her from her husband, and on her husband's death she could return to her father's family if she wished (see Gardner-Jevons, , Greek Antiquities, pp. 553 seqq.)Google Scholar. The laws of Gortyn also recognize fully the male predominance, despite their proximity to Lycia, and their innovations in favour of the female are no doubt rightly attributed by Jevons to advanced ideas, not to survivals of matriarchy. Aischylos was a reformer, no doubt, but his reforms lay not in these matters, but in his spiritual conception of God, and it is curious that Professor Ridgeway should still cling (p. 204) to the improbable theory of his defence of the Areopagos (of. Haigh, op. cit. pp. 56, 57, with Jevons, , Greek Literature, p. 196Google Scholar; the defence of the Argive alliance in Eumenides, 721 seqq., is not really consistent with a defence of the Areopagos, which clearly must have preferred the Spartan alliance).

(3) The view that the Bacchants were merely Thracian maidens is no doubt correct, but it is difficult to say if the same explanation (pp. 11, 12) applies to the Satyrs, Sileni, Hermenoi, Sauadai, and Deuadai, and like companies, who were, it seems, all real Thracian aboriginal tribes, addicted to tattooing and lax morality; von Schroeder, (Mysterium und Mimus, pp. 476 seqq.)Google Scholar argues that the spirits of the dead and the mimetic dances in imitation of them, with their power of evoking reproduction, lie at the bottom of these legends, but it is certainly probable that the explanation of these curious figures of myth is simply the dances for vegetative magic, in which the worshippers assumed the semblance of the god, the vegetative spirit in some animal form. This lies at the bottom of the modern Thracian rites, and the legend of satyrs and their variants could easily be derived thence. This view saves us from the necessity of seeing in the satyrs any distinct tribes of aborigines; any performers of the rites could give rise to the myth.

(4) The prominence of the worship of the dead, their tombs, ghosts, the kommoi and threnoi found in Aischylos, and also to some degree in Sophokles and Euripides, cannot be used (pp. 131, 162) to prove the origin of drama. Drama as in Aischylos is really in kind different from the primitive material from which it emerged, and it deals with the great questions of the day and the religious feelings of the time, in which the care of the dead undoubtedly played a great part, as it did in Vedic India and in Persia. Nor must Aischylos' connexion with Eleuses and the probable influence of the mysteries be ignored; cf. Arist., Ran. 886 seq.Google Scholar: δ⋯μητɛρ ⋯ θρ⋯Ψασα τἢν ⋯μἢν φρ⋯να, ɛἶναί μɛ τ⋯ν σ⋯ν ἄξιον μνστηρίων. In this regard Professor Ridgeway seems to lay too much stress on the aboriginal character of reverence of the dead, and on the fact that burning was the only Homeric custom, as showing that the Homeric conception was totally opposed to reverence of the dead. Homer is dealing with the exceptional circumstances of foreign wars (cf. Dörpfeld, , Mélanges Nicole, pp. 95 seqq.)Google Scholar, and the care for the dead is in his poems extremely well marked, e.g. as regards Patroklos; he may even represent a stratum of thought rather than a complete racial change; cf. Lang, , Homer and his Age, pp. 101 seqq.Google Scholar; The World of Homer, pp. 105–12.

(5) The acceptance by Professor Ridgeway (p. 164) of the legend that Themistokles sacrificed before Salamis three Persian captives to Dionysos Omestes seems unfair to that great man, even if we accept the version of Plutarch, in Them. 13Google Scholar, that he did so of necessity at the demand of those around him on the bidding of the seer Euphrantides. But it rests only on the evidence of Phanias of Lesbos, writing two hundred years later, it is not noticed in Herodotos, and it can safely be put down as a lie (cf. Murray, , Rise of the Greek Epic ², p. 34Google Scholar, and see Grote's, criticism, History, iv, 227, n.)Google Scholar. It is, however, true that here and there human sacrifices long survived in Greece (cf. Farnell, v, 303, and see Murray, pp. 326–31), though again it is doubtful if in Pausanias' time human sacrifice was made on the Lycæan mount in Arcadia. Pausanias (viii, 38. 7) hints at it, but this may be merely a case where he quotes his authority without vouching for his own time.

(6) Professor Ridgeway takes great pains to show that the Dorians were not the originators of tragedy, and declares (p. 2) that “it has been universally assumed that the Dorians were the inventors of tragedy”. But I cannot find any evidence of such assumption; see e.g. Haigh, op. cit. p. 25, and Jevons, op. cit. p. 190, who see matters in a truer perspective.

(7) The view that the Anthesteria was properly and solely a great festival of the dead, put forward by MissHarrison, , Prolegomena, pp. 34 seqq.Google Scholar, and accepted by Professor Ridgeway (p. 50), is relied upon to prove the evidence of pre-Dionysiac choruses in honour of the dead; it is sufficiently refuted by Farnell, v, 219 seqq.

(8) It is difficult to follow Professor Ridgeway's elaborate investigation of the meaning of thymele (pp. 39–48) and his conclusion that as there were two altars there were two cults. The θυμ⋯λη, as he himself admits, is the place of the offering of burnt-sacrifice to the god, round which the chorus naturally danced; that a chorus could dance round a tomb is perfectly true, but a tomb is not called and could not be called θυμ⋯λη, for the dead receive no burnt-offerings (cf. his own clever rendering (p. 137) of ⋯μπύρους τ'⋯ρθοστάτας in , Eur.Hel. 574)Google Scholar, and the theory that an altar replaced a tomb is gratuitous. There was also on the stage as a rule (Pollux, iv, 123; , Aisch.Ag. 1080)Google Scholar a βωμός of Apollo Aguieus, but it is a mere conjecture that this was really an old gravestone, and the author's argument rests on the supposed substitution of a θυμ⋯λη for a tomb. On the other hand, he seems right in correcting Haigh's, view (Attic Theatre ², pp. 106, 107)Google Scholar that the table which may have stood beside the θυμ⋯λη was used by the leader of the chorus as a place to stand upon and converse with the chorus. That is called ⋯λεός in Pollux, iv, 123, and was presumably an ordinary table, and the notice in Et. Mayn. s.v. θυμ⋯λη is apparently confused; it does not yield the sense desired by Haigh in θυμ⋯λη=table on which the choir leader stood, nor is Ridgeway's version satisfactory.

^{page 429 note 1} Gött. Nach. 1911, pp. 441–68.

^{page 430 note 1} JRAS. 1911, pp. 979–95.

^{page 430 note 2} Neither Hertel, VOJ. xxiv, 122 seqq., nor I have maintained that none of the citations are composed by the author of the prose. The earlier style, in my view, is where they are not so composed (op. cit. p. 986, n. 1).

^{page 430 note 3} Op. cit. p. 451.

^{page 430 note 4} Op. cit. p. 464.

^{page 430 note 5} Kāṭhaka Saṃhitā, xxx, 1; Kapiṣṭhala, xlvi, 4.

^{page 431 note 1} SBE. xliv, 95. 1 am not sure if Eggeling is not right, but the point is indifferent to the inquiry.

^{page 431 note 2} See my note ad loc.

^{page 431 note 3} The parallel with the Māra legend is proof of that (cf. Macdonell, , Sanskrit Literature, p. 225)Google Scholar. But the Upaniṣad is not the legend.

^{page 432 note 1} Professor Oldenberg would no doubt admit the priority of these Upaniṣads to the Kaṭha. The evidence for it is given by Deussen, , Philosophy of the Upanishads, pp. 22 seqq.Google Scholar; ef. Keith, , Aitareya Āraṇyaka, pp. 41 seqqGoogle Scholar.

^{page 432 note 2} See e.g. Macdonell, , Sanskrit Literature, p. 205Google Scholar; Wackernagel, , Altind. Gram. I, xxxGoogle Scholar; Keith, , Aitareya Āraṇyaka, pp. 30–3Google Scholar.

^{page 432 note 3} See Vedic Index, ii, 193.

^{page 433 note 1} See e.g. ZDMG. xxxvii, 79. Contrast Sieg, , Die Sagenstoffe des Ṛgveda, p. 46Google Scholar.

^{page 434 note 1} Gött. Nach. 1911, p. 461.

^{page 434 note 2} VOJ. xxiii, 273 seqq.

^{page 434 note 3} Op. cit. pp. 466–8.

^{page 434 note 4} vii, 18. 12, 13.

^{page 435 note 1} Göft. Nach. 1911, pp. 457, 458.

^{page 435 note 2} ZDMG. lxiii, 13.

^{page 436 note 1} Gött. Nach. 1911, p. 452, n. 4.

^{page 436 note 2} But I did not so use it; see op. cit. p. 986, n. 1. I only used it to show that the relation of prose and verse was not, according to a recent and careful student of the texts, as Professor Oldenberg thinks, always one of priority.

^{page 436 note 3} See Keith, , JRAS. 1900, pp. 127–31Google Scholar; Sieg, op. cit. pp. 37 seqq.

^{page 436 note 4} Op. cit. pp. 444, n. 3, 450, n. 1.

^{page 436 note 5} Gött. Nach. 1897, p. 128, n. 1. Oldenberg here abandons him, p. 449, n. 3.

^{page 436 note 6} Op. cit. p. 456.

^{page 437 note 1} The argument at p. 453 I do not follow. The Jātaka collection treats, of course, the verses as the really important part, because in that collection they are so, the prose being clearly subsidiary. But that tells nothing of their original condition.

^{page 438 note 1} Gött. Nach. 1911, p. 466, n. 1.

^{page 438 note 2} Op. cit. p. 488. I point out the distinction between the prose of the assumed Ākhyāna and the much less closely connected prose of the Saṃhitās, a fact which might have prevented the accusation. But I know how easy it is to misunderstand, and I cannot hope to have avoided the same fault with regard to others; see e. g. the complaint of Speyer in ZDMG. lxiv, 319, 320, though I still think he does regard the phenomenon-discussed by me in ZDMG. lxiii, 346, in a somewhat different light than I do.