^{page 157 note 1} Farber, H., Die Lyrik in der Kunsttheorie der Antike, Munich, 1936.Google Scholar

^{page 157 note 2} It is neglected, for example, in the otherwise full and excellent introduction to Weir Smyth's Greek Melic Poets.

^{page 157 note 3} Textgeschichte der griechischen Lyrik, 1900 (T.G.).Google Scholar

^{page 158 note 1} The view that the Suidas notice s.v. represents such an edition seems to me implausible. Cf. Wilamowitz, , Einkitung in die gr. Tr., 143 f., and below p. 161.Google Scholar

^{page 158 note 2} Unless Stephanus' words (ad Alcman fr. 13 D) are taken to imply the existence of a book of . by Alcman—a very doubtful implication. Corinna's books bore titles: but she was certainly not edited at this time by the Alexandrians.

^{page 158 note 3} Schol. H 76 (Pap. Ox. 1087, ii. 48) Gramm. Bekk. An. a. 783. 14 Google Scholar

^{page 158 note 4} Cf. Wilamowitz, , S.U.S. 233 fT.Google Scholar, Rose, H.J. in C.Q. xxvi (1932), 92.Google Scholar

^{page 158 note 5} Stobaeus; Schol. Theocr. 16. 36; Suidas s.v., who lists

^{page 159 note 1} Apparently between one and two thousand lines, v. Irigouin, , Histoire du texte de Pindare, p. 41.Google Scholar

^{page 159 note 2} ‘Anonymus de Lyricis Poetis’ = Schol. Epimetr. Pind. iii p. 310. 27 Dr.

^{page 159 note 3} The modern division into ‘choral’ and ‘monodic’, which appears in Diehl, derives from a passage in Plato, Laws 764 d-e, and is of no particular value. It was unknown to antiquity or the Renaissance.

^{page 159 note 4} Another idea is attributed to Apollonius the Eidographer (Et. Mag. 295. 52) who, This eccentric activity, even if historical, can hardly have any relevance to the classification of poems into books; and in fact we find him taking a perfectly rational part in the debate on the correct position of Pythian 2 (Schol. Pyth. 2 inscr. p. 31 Dr.). Wilamowitz and others who adduce this passage seem to have overlooked the second part of it.

^{page 159 note 5} Schol. Pind. i p. 3 Dr. (below, p. 161 n. I). If are regarded as the list falls neatly into two groups. Bergk's view that belong to the yevos (P.L.G. i⁴. 370, n. 2) goes against all the explicit ancient testimony.Google Scholar

^{page 160 note 1} Proclus, (p. 319 b 35) adds Google Scholar

^{page 160 note 2} Proclus, , p. 321 a 34: Google Scholar

^{page 160 note 3} A good enough reason already for Heracleides Ponticus, [Plut.] de Mus. 1134c The point is made by Wilamowitz, , T.G., P. 43.Google Scholar

^{page 160 note 4} Schol. Pind. Nem. 9 inscr. Cf. Schol. Pyth. 2 inscr.

^{page 160 note 5} The answer to the debate whether Pin dar's epinikia were in principle processional or convivial may be that there existed both kinds, but that the Alexandrians placed them all together regardless.

^{page 161 note 1} Schol. Pind. i p. 3 Dr.: Google Scholar

^{page 161 note 2}

^{page 161 note 3} Dr. i p. 9, 26 ff:

^{page 161 note 4} So Hiller (Hermes, xxi (1886), 357 ff.Google Scholar) and (for different reasons) Immisch (Rh. Mus. xliv (1889), 558 ff.Google Scholar). For another explanation see Wilamowitz, , Einleitung, 185 and n. 126.Google Scholar

^{page 161 note 5} Suidas s.v. where the cod. Ber. of Photius reads , em. Reitz. Hiller, (op. cit., p. 368, n. 2Google Scholar) believes the words to be an interpolation and reads Even if Reitzenstein's reading is accepted, the reference can hardly be to Pindar fr. 124 (cf. Reitzenstein ad loc.); Bergk (P.L.G. i⁴. 372–3), followed by Körte (Hermes, liii (1918), 138, n. 2Google Scholar), explains it as an allusion to an edition of Pindar prior to that of Aristophanes. Thus, one can either reject the reading or explain it away.

^{page 161 note 6} Cf. Körte, , Hermes, liii (1918), 138 f. Fr. 118 is quoted as an by Schol. Ol. 2. 39 a, proving that the scholiast knew of a book of Google Scholar

^{page 162 note 1} Most of the material is collected by Färber ss.vv., Reitzenstein, , Epigramm und Skolion, c. iGoogle Scholar, and Fraustadt, , Encomiorum in litteris Graecis … historia, Diss. Leipz. 1909.Google Scholar

^{page 162 note 2} Apparently followed by Artemon in Athen. 694 a and by Plut, . Qu. Symp. 615 b.Google Scholar

^{page 162 note 3} Proclus et la Chanson de Table, § 1 in Mélanges Bidez, 1934. The question is summarized by Wehrli in his commentary to Dicaearchus fr. 88 (Schule des Aristoteles, i, 1944). I do not think that it can be inferred from Schol. Ar. Nubes 1364 that D. did not use the word of the second class.Google Scholar

^{page 163 note 1} Cf. Wilamowitz, , T.G., p. 44: ‘Ganz allgemein haben sie den Namen Skolien perhorrescirt.’Google Scholar

^{page 163 note 2} The word occurs O. 2. 47, O. 10. 77, 0. 13. 29, P. 10. 53, N. 1. 7, cf. N. 4. 11, N. 8. 50. None of these passages precludes the more general interpretation ‘revel-song’, out of which the technical term doubtless grew, but at the same time none is incom patible with, and some seem positively to invite, the interpretation ‘victory-song’.

^{page 163 note 3} Later writers distinguish it from a brief commendation (eq38) e.g. Suidas s.v.

^{page 163 note 4} Ar. Eth. Eud. I.e., Rhet. 1.1367 ^b 26 ff:

^{page 163 note 5} Ar. Rhet. 2. 1388b21:

^{page 163 note 6} An example of this kind of confusion is Simonides’ poem on the dead at Thermopylae, which is referred to by Diodorus Siculus as an This phrase is surely rhetorical, not literary. Diodorus does not say and we have no other evidence that one of the books of his poems bore this title. Editors should therefore be cautious (as Bergk observes, P.L.G. iii⁴. 383) in placing this poem under the heading would surely be more appropriate. (Cf. Bowra, , Class. Phil. xxviii, 1933, 277, whose objection to is unfounded, cf. below, p. 169.) Still less can we say with Smyth (p. lxxvii) ‘The encomion was the creation of Simonides’. All that Simonides could have understood by the word, as a technical literary term, is epinikion.CrossRefGoogle Scholar

^{page 164 note 1} Why was the so named? Hephaestion (p. 50 Cons.) adduces examples of it from Alcaeus and Anacreon, which he would hardly have done if it had been called after the original It is more likely that the Alexandrians gave this name to a metre found in the numerous of Alcaeus and Anacreon, which they had by now renamed Cf. below, p. 175. meaning , still turns up occasionally in the Scholia on Pindar, e.g. Schol. Nem. 7. 113.

^{page 164 note 2} This change seems to be the point of a difficult passage in Philodemus, de Musica (p. 67 f. Kemke):

^{page 165 note 1} (700 d).

^{page 165 note 2} e.g. Ammonius p. 52 Valck.

^{page 165 note 3} For example, we find in Schol. Lond. Dion. Thrax 451, 6 Hilg. the definition,

^{page 166 note 1} Orion, p. 156, 3 ff. Sturz. Google Scholar

^{page 166 note 2} Athenaeus 631 d (probably from Aristoxenus): The sense of the context seems to me to rule out the proposed emendation, and as it stands the passage supports no conclusion weightier than that which I have drawn.

^{page 166 note 3} Et. Mag. 690, 41 has an article dividing and If this is relevant, the are the of Proclus, and it is possible that the stance of the choir should be interpreted as a comparatively restricted movement, as in the of a tragedy. Cf. Smyth, , p. xxix, n. 1. This concession in no way affects the point that a particular kind of hymn existed which was performed in a particular way.Google Scholar

^{page 166 note 4} Schol, , Epimetr. Pind. iii, p. 311 Dr.Google Scholar Cf. Färber, , i. 20–22.Google Scholar

^{page 167 note 1} Even if Snell's reconstruction, from various fragments, of Pindar's first hymn is accepted, there is still no evidence of triadic construction: indeed metrical correspon dence between the strophes is the main argu ment for grouping the fragments together. Cf. Snell, , Antike und Abendland, ii (1947), 187Google Scholar, and Turyn, , Pindari Carmina, p. 341.Google Scholar

^{page 167 note 2} (p. 342).

^{page 167 note 3} Plato refers explicitly to one such formula in Cratylus 400 e: Plato himself frequently parodies these formulae, e.g. Philebus 63 b, Prot. 358 a. Even in Homer one may detect this kind of flippancyI 96–97. For a comprehensive analysis of these formulae, cf. Zuntz, , Rh. Mus. xciv (1951), 337 ff.Google Scholar

^{page 168 note 1} One ‘hymn’ has survived which does not conform at all. Pindar fr. 29 begins a rhetorical flourish which, though it has a story attached to it (Plut. de gloria Ath. 347 f), is wildly inappro priate to what we think of as a hymnodic invocation. Yet the scholiast on Ps-Lucian (who quotes it) calls it the If this is accepted, we must eidier suppose that this is not really a hymn at all, and was only placed in its position by the Alexandrians for some adventitious reason (such as that the first verb is ); or else that Pindar already shows symptoms of the decadence diagnosed by Plato, in that he does not adhere to the strict form of the hymn. Schol. Nem. 1 a lends support to the first alternative.

^{page 168 note 2} There is not the slightest evidence for Diehl's assumption that the Danae-fragment of Simonides was a it is quoted by Dion. Hal. as A narrative of this kind is just as likely to have occurred in a dithyramb or an epinikion. Exactly the same goes for Sim. (?) 13 b D = Bacchyl. (?) 60 Sn.

^{page 168 note 3} It was this restraint which struck R. M. Rilke: ‘Erstaunte euch nicht auf attischen Stelen die Vorsicht menschlicher Geste?’ (Duino Elegies, ii.)

^{page 168 note 4} It might be maintained that of course all the fragments of Simonides' are gnomic, since they all come from Stobaeus (who was collecting just this sort of thing). But it will still have to be explained why this was the only book of Simonides in which Stobaeus found die sort of sentiment he was looking for.

^{page 169 note 1} The point is made by Reiner, E., Die rituelle Totenklage der Griechen = Tüb. Beitr. z. Altertumsw. Heft 30, 1938, pp. 8 f.Google Scholar

^{page 169 note 2} See the passages collected by Färber, , ii. 53 f.Google Scholar

^{page 169 note 3} Didymus (ap. Orion 58. 7, followed by Proclus, p. 321 a 30) offers the derivation This does not make sense unless by is understood that particular kind of consolation which consists in dwelling upon the virtues of the deceased.

^{page 170 note 1} In his commentary ad loc. Cf. also his remarks iii. 623–4. I only differ from him in making a sharper distinction between and

^{page 170 note 2} One may possibly detect the same dis tinction in Lucian, de Luctu 13–17. There the bereaved father is made to pronounce over his son: this is a typical Homeric Then the son comes momentarily to life and, with the words proceeds to dwell on the consolations of being dead.

^{page 170 note 3} The material is collected by Page, D. L. in Greek Poetry and Life, pp. 206–30.Google Scholar

^{page 170 note 4} There is the further objection that since elegy is a solo performance and the dirge a choral one, they cannot ever have been the same thing. But were never sung by a choir? Those introduced by Sacadas into Sparta at the second may well have been choral ([Plut.] de Mus. c. 9): the was predominantly a matter of choral music, Sacadas is known to have had a choir, and was a ([Plut.] c. 8). The relevant chapters in Plutarch suggest at least the possibility that this ‘Dorian elegy’ was choral.

^{page 171 note 1} 10 D. The same goes for the longer passage quoted by Stobaeus (7 D) which prob ably comes from the same poem. Cf. exactly the same sentiment in Sappho 109 D, which is quoted from a lament for her own daughter.

^{page 172 note 2} Jacoby, (Rh. Mus. Ix, 1905, 43)Google Scholar well de scribes the difference between the epic and the elegiac form: ‘die neue Form war bestimmt, nicht Stoffsondern Gedanken aufzunehmen, nicht zu erzahlen, sondern zu ermahnen und zu belehren.’

^{page 173 note 3} These were doubtless also the models for the compositions of Parthenius and others which the Alexandrians called (v. Crusius in R.E., s.v.), a form which Olympus is said by Aristoxenus to have invented, [Plut.] de Mus. c. 15. Cf. Severyns, , Mélanees Navarre, pp. 383–94.Google Scholar

^{page 172 note 4} The first lines contain typical elegiac sentiment (Andr. 103–4):

^{page 172 note 5} So Zacher, . Philol. lvii (1898), 9–22.Google ScholarSchol, . Eur. Andr. 103 is a bare statement of fact, and does not support one theory rather than another.Google Scholar

^{page 172 note 6} This origin may account for the special tone of delivery reserved for elegiacs: Dion. Thrax § 2. Like the , the archaic elegy was accompanied by the flute. Cf. Schmid- Stählin, I i, p. 354 n. 5.Google Scholar

^{page 173 note 1} Catullus 38. 8: ‘Maestius lacrimis Simonideis’ probably refers to the , but is equally inconclusive.

^{page 173 note 2} This must be the solution to the about the inventor of the dithyramb. Hero dotus (1. 23) and Aristotle (ap. Proclus, p. 320 a 31) say that it was invented by Arion: yet the dithyramb is mentioned by Archilochus (77 D) who lived long before Arion. Arion may well have been the first to make the dithyramb into a literary form: but this is not to say that before he did so no one could have sung a ‘dithyramb’ to Dionysus after a banquet. The same explanation goes for the various traditions in Pindar (Schol. 01. 13. 26). The distinction is recognized by Severyns, , Recherches sur la Chrestomathie de Proclus (1938), ii. 134.Google Scholar

^{page 173 note 3} e.g. Pindar, Paean 5, the paean of Aristonous (which is inscribed as a but is clearly a ) and the of Philodamus Scarpheus.

^{page 174 note 1} Those of lsyllus of Epidaurus and Limenius Atheniensis, and the majority of Pindar's. We cannot be sure that all those of Pindar are really paeans, but they are all closely connected with Apollo, and it is hard to know what else they might be. Still further doubt is caused by the fact that paeans could be addressed to other gods than Apollo. Cf. Smyth, , p. xxxviii, n. i.Google Scholar

^{page 174 note 2} The quotations given by Aristotle (Rhet. 3. 1409s) for the metre and suggest that it was indeed named after its use in paeans; and it is often used in conjunction with cretics. Cf. Christ, , Metrik 384–5.Google Scholar

^{page 174 note 3} The two are closely connected: both originally came from Crete, and often used the same metre. The evidence is assembled by Deubner, L. in Neue Jahrbikher, 1919, PP. 396–7.Google Scholar

^{page 174 note 4} Cf. also Ar. Rhet. 1409^b 26 ff. and Maas in R.E. s.v. Melanippides. The same development occurred with the nome. [Ar.] Probl. 9 shows that the was not antistrophic; but according to Proclus p. 320 a 32 ff. the nome developed out of choral—and therefore antistrophic—lyric.

^{page 175 note 1} In the same way the books of Bacchylides entitled and (possibly) (cf. Körte, in Hermes, liii, 1918, 137–40) probably contained poems which an earlier generation would have called, simply, .Google Scholar

^{page 175 note 2} Bergk therefore reads but his reasons (P.L.G. i. 316) are not convincing.Google Scholar

^{page 175 note 1} Cf. above, p. 164, n. 1.

^{page 175 note 2} The ‘iambelegus’ (the exact converse of the ‘encomiologicon’) is quoted by Hephaestion from an extant poem of Pindar which is in dactylo-epitrite; and the term occurs frequently in the metrical scholia to dactylo-epitrite poems of Pindar.

^{page 175 note 3} Cf. Croiset, , La poisie de Pindare, pp. 359 ff.Google Scholar and, for a recent attempt, Nierhaus, R., Strophe unci Inhalt im pindarischen Epinikion, 1936.Google ScholarSchadewalt, W., in Der Auf-bau des pindarischen Epinikion (1928), achieves a penetrating formal analysis of certain of Pindar's epinikia; but the attractive generalizations offered in his first chapter are seriously debilitated by the qualifications he is forced to introduce in his last.Google Scholar

^{page 175 note 4} I am indebted to Professors R. Pfeiffer and D. L. Page, who kindly read through earlier drafts of this article and made many valuable suggestions.