adopted by the lyric and tragic poets. Poetic fancy and an ethical purpose transferred, as early as the Nékvia of Homer, the tortures of Tantalos, Sisyphos, etc., to the nether world. The earliest form of the legend appears in Athen. 7. 281 B (from the Cyclic Return of the Atreidai'), according to which Tantalos, who lived in heaven with the gods, had a rock suspended over his head by Zeus, who had pledged himself in advance to grant any request that his son might make; but who was filled with wrath when Tantalos petitioned that his appetites be gratified, and that he live in the same manner as the gods. This scene is laid in heaven, and the suspended rock not merely robs the conviva deorum of his power to enjoy the divine nectar and ambrosia, but is an added torture because of his immortality (μετὰ τριῶν τέταρτον Tovov Pind. Ol. 1. 60). Pindar does not certainly localize Tantalos in Hades. Cf. Comparetti Philol. 32. 230. On the view that Alkman keeps to the original story, év doμévolow of the MSS. is among the blissful,' 'the well pleased gods.' The rock of terror was explained by Welcker R. M. 10. 242 as merely the creation of the distressed mind of the living sufferer. For such phantasms, cf. those of Io, Orestes (Aisch. Choeph. 1051), Pentheus (Eur. Bacch. 918, Verg. Aen. 4. 469). Hecker's ȧpuévolov is taken to mean either the bound' captives, or in bonds,' and transfers the scene to the nether world, thus making the poet follow Homer rather than the Cyclic epic. The rock is then a reality, and all the more awful because invisible. I doubt whether ἀρμένοισιν can have either of the above meanings. It should mean 'amid pleasures,' 'good cheer'; cf. Hes. W. D. 407, Shield 84, Theogn. 275, and such expressions as ȧyalà пáνта, äþlovа πάντα. Hecker supplied πᾶσιν before ἐν (Pind. Nem. 3. 58 év áρμévoιoi mâσi; cf. schol.).

2. θάκω : the Ms. θάκα = θᾶκος might be defended by the pair σκοπός σκοπή, etc. Welcker took κάτα (sic) with ἧστο. Hermann wrote θάκοις κάτω. OTо presupposes motion, so there is no difficulty about the κará. There is no need to take it with ὁρέων, or to regard πέτρας οὐδέν as = οὐδεμίαν πέτραν.

3. Eust. Od. 1701. 23 has opéοvтI and doкéovтi, whence Bergk doKéovтi d' (colкús).-Metre: uncertain, probably logaoedic.

XXXII. Oxyrhynchus Papyri 1. No. viii. Attributed to Alkman by Blass. From a partheneion. It is possible that the poem is a happy imitation of Alkman's manner by some Alexandrian.

1. ἤνθομεν : cf. ἐνθοίσα iv. 73. The ending -μεν is either epic or Aiolic, and seems to be used, as the editors remark,

in order to avoid the sigmatism of -μes ès. In iv. 12 παρήσoμes. μeɣáλas: this epithet of Dem. does not recur until Kallim. 6. 121 μεγάλα θεὸς εὐρυάνασσα; μεγαλαῖσι θεαῖσι οἱ mother and daughter, Anth. Pal. app. epigr. 1. 59. 3 (Cougny). ἐάσσαι = ἐοῦσαι. Dor. ἔασσα <* ἐστια reappears in Philolaos and the Pseudo-pythagoreans. In Messen. and Argive we find ἔασα, in Cretan ἴαττα. In 64 Alkm. has παρέντων.

2. παρθενικαί: see on vii. 1. Cf. η 20 παρθενικῇ . . . κάλπιν ἐχούσῃ.

3. Kāλá: for the variation with SONGS i. In xii. Alkm. has κaλws. the editors write.

κăλá, 1. 2, see FOLKöpμws: not opμovs, as

4. πριστῶ ἐλέφαντος: cf. σ 196 λευκοτέρην δ ̓ ἄρα μιν (Penelope) θῆκε πριστοῦ ἐλέφαντος, τ 563 πρ. έλεφ. of one of the dream-gates, θ 404 κολεὸν νεοπρίστου ἐλέφαντος. αἴγλῃ (Blass) is paleographically uncertain. Blass suggests that the next verse began Xeukotáтas xióvos.-Metre: dact.-hexameter. Whether the Frag. is connected with vii. is uncertain.

ARION.

THE only early account of Arion's rescue that is extant is Herodotos 1. 24, where it is introduced as an anecdote, based on Korinthian and Lesbian sources, in connection with the mention of Periander (625-585). Most of the later recounters of the tale-e.g. Hyginus 194, Dio Chrys. 37, p. 455, Plutarch Sept. Sap. Conv. 18, Fronto 262 and the numerous writers who allude to it, depend ultimately on Herodotos, though some may have derived the legend from some Hellenistic poet. The story was greatly embellished in later times, Plutarch, for example, throwing moonlight on the scene, and making the story subserve his reverence for nature and his piety. The fragment is cited by Aelian Hist. An. 12. 45 to show the dolphin's love of music.

Herodotos says that at Tainaron there was a bronze statuette of Arion, a man riding on a dolphin. Aelian gives the epigram on the votive offering: abavárшV πομπαῖσιν 'Αρίονα Κύκλονος υἱόν, | ἐκ Σικελοῦ πελάγους σῶσεν ὄχημα τόδε. This inscription may have been added after the time of Herodotos. That the legend wandered from Sparta to her colony, Thera, was concluded from an epigram found there (Kaibel 1086), but now shown to

be worthless evidence (Athen. Mittheilungen 21. 253). Neither the inscription nor the figure of Arion on coins of Methymna is proof of the existence of a poet of this

name.

The legend of Arion's romantic rescue is due to a misinterpretation of the figure at Tainaron. The statuette was either that of a god or of some hero originally identical with the god, but in course of time individualized and dissociated from him. The rider has been identified with the Korinthian Melikertes-Palaimon (cf. Ant. Denkm. d. Arch. Inst. 1. 7. 26, Inscr. Sicil. et Ital. 2519 C); or with Taras, the son of Poseidon, who rode from Tainaron to Tarentum on a dolphin's back. Studniczka, Kyrene 181, has, however, shown that the rider was not Taras, but Phalanthos, who, at first a form of Poseidon, gradually became an historical person connected with the emigration of the partheniai. Hartung thought the rider was Orpheus. Most probably it was either Poseidon or Apollo, with whose cults the dolphin is intimately associated. In Lakonia there was a goddess 'Apiovтía (I. G. A. 79), in whose honour horse-races were established. Mr. Paton (Class. Rev. 4. 134) thinks that she corresponds to Demeter Erinys of Thelpusa, the mother of the mythical horse Arion, whose father was Poseidon. Now both horse and dolphin are symbolical manifestations of the god of waters, and it is noteworthy that the only places mentioned in the story of Arion-Methymna on the island of Lesbos and Korinth of the double sea, the birthplace of the poet, and the place where he is said to have practised the dithyramb-are the seats of legends of grateful dolphins.

It is, in fact, probable that the poet Arion is one and the same with the mythical horse, the manifestation of Poseidon. Exactly how the invention of the cyclic chorus and of the rpayıкòs трóжоs came to be attributed to him we cannot say; doubtless Lesbian legends are here at work, just as they created Phaon, the mythical lover of Sappho. It may be noticed that, apart from the steed of Adrastos (346, Hes. Shield 120), the name Arion occurs nowhere in early literature. It is possible to derive it from dpi-Ftwr,' very swift' (Maass Ind. Forsch. 1. 166), though Fick-Bechtel (Personennamen 433) propose to connect it

and the Arkadian form 'Epíwv (coin of Thelpusa S. G. D.-I. 1253) with 'Epivís, rivalis. Κυκλεύς and Κύκλων, the names of the father of the supposed poet, are inventions made to account for the belief that his son first set up the Kúκλos xopos, an institution which is involved in obscurity, though it is supposed that a circular chorus of fifty members took the place of the older rectangular arrangement in ranks and files. Some suppose that Arion first made the dithyramb choral, it having been monodic up to his time, and that the Tрayɩkòs тpóños, which he invented, alludes to the 'fashion' of satyrs, who, clothed as goats, spoke in verse, thus forming the beginning of the tragic' drama. Others think the 'goat-fashion' is the pathetic fashion in contradistinction to that of the nome; others refer it to the introduction of the tales of heroes. Many theories, little certainty. Despite the statement in Herodotos that Arion was the founder of the dithyramb, some Hellenistic critics seem to have doubted his existence, and given that honour to Lasos, Pindar's teacher. Most of the statements in Suidas may be an expansion of Herodotos' account, or based on some book on the Korinthian festivals. Arion is strangely enough called the scholar of Alkman.

The authenticity of the poem was first disputed by Van der Hardt in 1723. Hermann regarded it as an example of ornatus qui varietate et venustate constat. Welcker (Kl. Schr. 1. 89 ff.) was inclined to regard it as old, if not by Arion himself. But considerations of style, metre, and dialect show that it must be later than the lyric age. It cannot be a forgery by an author of the quality of Aelian, as Lehrs supposed (Popul. Aufsätze 204), nor indeed the composition of a nomic writer of an early period (Boeckh Berl. Acad. 1836. 74), but is rather the production of an Athenian dithyrambic poet of the last period of Euripides, or later. The style, despite its partial smoothness, recalls the fulsomeness, the veneer, of the later dithyramb; the metre is ornate, with its many resolutions, syncopated feet, and anacruses, and shows its late authorship by its frequent instances of positio debilis. The dialect is Attic diluted with Doric, a mixture that became common in the fifth century. Rossbach conjectures that the poem is either the work of a

scholar of the dithyrambic poet Phrynis, or of the master himself. Aelian quoted the poem in good faith, but originally it was put into the mouth of Arion, without intent to deceive, in order to serve as an exaltation of the power of music, a theme that was popular with the later dithyrambic poets, to judge from the Argo of Telestes, a fragment of which is akin to the hymn of Arion in the frequency of anacruses. Just so Kallimachos made Simonides himself tell of his miraculous rescue by the Dioskuroi; and so the story of Sappho's leap from the Leukadian cliff and of her love for Phaon arose from her mention of this resort of hapless lovers and her story of the ferryman of Aphrodite.

The poem falls into two parts: (1) 1-11, invocation of Poseidon, around whom the dolphins dance; and (2) 12-18, the rescue of the poet. The mention of the dolphins is withheld till v. 9, Oĥpes standing in epexegetical apposition. Throughout we have a series of pictures produced by ornamental and characteristic' epithets. The poet is prodigal in his use of colours.

2. χρυσοτρίαινε may be defended by χρυσηλάκατος, χρυσοκέφαλος. Hermann and Bergk read -Tpíaiva from a nom. in -ης, with which cf. χρυσοχαίτης, χρυσομίτρης, χρυσοκόμης (Xpvoeокóμα Sim. 26 B; Alk. iii. Xpuσoкóμas). Alk. xiii. has μελλιχόμειδε, though we have φιλομμειδής; in Anakr. xxviii. there is warrant for evéleɩpe and evébeɩpa; Sim. xxvii. has φυγόμαχος; Sa. xxii. καλλίκομος. Pindar's Όρσοτρίαινα, Αγλαοτρίαιναν, Εὐτρίαιναν, are sometimes explained as Boiotisms (-a for -ns). Aristoph. Equit. 559 & xpvooтpiaiv', ŵ|deXpivwv μεδέων (where the schol. has χρυσοτρίαινα), is not necessarily either a parody or an imitation of this poem. Cf. M. 27 Εννοσίγαιος, ἔχων χείρεσσι τρίαιναν. Πόσειδον : the Attic form.

3. γαιάοχος is the earth mover” (γαίης κινητήρ), as ἐννοσίγαιος, ἐνοσίχθων, ἐλασίχθων. Cf. Lakon. γαιάFoxos, Pamphyl. Fexérw, veho. (F)oxos was later confused with -(o)oxos (in πολιάοχος, ῥαβδούχος etc.), and Artemis is called γαιάοχος, Soph. O. T. 160. ȧv' is due to Hermann, who connected ἐγκ. ἀν' ἅλμαν with the following.

4. ẞpáyxio: if correct, is a neologism. Hermann read βραγχίοις περὶ δή ; Buchholz περί σέ γε ; Sitzler ἐν κύμασι πάλμυ βρυχίοις. Cf. Ν 27 βῆ δ' ἐλάαν (Ποσ.) ἐπὶ κῦμα· ἄταλλε δὲ κήτε ὑπ' αὐτοῦ | πάντοθεν ἐκ κευθμῶν, οὐδ ̓ ἠγνοίησεν ἄνακτα.