^{page 149 note 1} Winter's assertion that, in the subscribed title, ‘ 'αλκιδ⋯μαντος is a certain restoration’, is perhaps not beyond dispute, although it is probably true. No other author whose name was composed of 3 or 4 letters before δαμας is known to have written on Homer; but as Körte observes, there is no other evidence for a special treatise by Alcidamas entitled περ⋯ 'Oμἠρου and the lines from the Certamen are referred by Stobaeus to his Mουςεῖον. It may be that this work was the source of the papyrus, and that the copyist, writing half a millennium later than Alcidamas, made the same sort of mistake as those doxographers who gave the title περ⋯π⋯σεως to any early philosophical work. The character of the Mουσεῖον is a matter for dispute, but it was clearly a heterogeneous work which dealt not only with Homer.

^{page 149 note 2} The most recent editions of the Lives are by T. W. Allen, in the fifth volume of the Oxford text of Homer, , and Wilamowitz, , in Vitae Homeri et Hesiodi, Bonn, 1916Google Scholar; of these the second is much the sounder, although it omits both Tzetzes' accounts of Homer's death; references are to it unless otherwise specified. The Certamen is printed in full in both these editions, as in Rzach's edition of Hesiod, Teubner, 1908. In the edition of Wilamowitz the Lives are given Westermann's original numbering: (I) Ps.-Herodotus. (II) Ps.-Plutarch, two versions. (III) Proclus. (IV) The anonymous Escurial Life, first version. (V) The anonymous Escurial Life, second and fuller version. (VI) The anonymous Muretan or Roman Life. (VII) The Life in Suidas, taken in part from Hesychius. (VIII) The Certamen. To these should be added the version preserved in two accounts of Tzetzes: Chil. xiii, 626 ff. Kiessling, and Exeg. in Iliadem, p. 37Google Scholar Hermann, the former being in Allen's edition.

^{page 150 note 1} The relevant passage is as follows, according to the edition of Paton and Wegehaupt, Teubner, 1925, and to Wilamowitz: ⋯πε⋯…ἥ τε δ⋯ξα τ⋯ν ⋯γωνιστ⋯ν [Oμἠρου κα⋯ 'Hσι⋯δου] πολλ⋯ν ⋯πορ⋯αν μετ⋯ α⋯δος τοῖς κρ⋯νουσι παρεῖχεν, ⋯τρ⋯ποντο πρ⋯ς τοια⋯τας ⋯ρωτἠσεις κα⋯ προραλε μ⋯ν, ὥς ϕασι, ∧⋯σχης Mοσ⋯ μοι ἔννεπ' ⋯κεῖνα, τ⋯ μἠτ' ⋯γ⋯νοντο π⋯ροιθεν | μἠτ' ἔσται μετ⋯πισθεν ⋯πεκρ⋯νατο δ⋯ 'Hσ⋯οδος κτλ. Allen read ὥς ϕησι ∧⋯σχης, after Wyttenbach: but O (a reasonably good emended version of the archetype) is alone in giving φησι. ϕασι is the reading of QhJ nwB, of which Q is one of the two soundest MSS. QB also have προ⋯βαλε μ⋯ν: O has προυβ⋯λομεν which supplies the authority for Allen's προ⋯βαλ' ⋯ μ⋯ν. The reading πασι involves the omission of 'Oμἠρου κα⋯ 'Hσι⋯δου as a gloss: so Wilamowitz; cf. Il. u.Homer, 405. Thus Lesches becomes a fellow competitor; this, so far from being a difficulty, accords better with the introductory sentence of Periander, who is speaking: ⋯κο⋯ομεν ἅτι κα⋯ πρ⋯ς τ⋯ς 'Aμ=ιδ⋯μαντος τα=⋯ς ε⋯ς Xαλκ⋯δα τ⋯ν τ⋯τε σο=⋯ν οἱ δοκιμώτατοι ποιητα⋯ συν⋯λθον. The last four words suggest to me more than two com petitors. According to Clement, Strom. 1. 131. 6, Lesches had a contest with Arctinus: Plutarch had in any case doubted the authenticity of the Amphidamas-passage at Erga 654 ff., and would not be particularly concerned over the accuracy of Periander's Story.

^{page 150 note 2} The composition of this Life is dated by Allen, , op. cit. 16 ff.Google Scholar, to about the time of Hadrian; Wilamowitz, , however, Ilias u. Homer, 415 f.Google Scholar, suggested the end of the Hellenistic period. His two chief reasons for this date are ‘die lebendige hellenistische Rede’ and a reference to the Athenian archon-list in the last chapter of the Life, which, he holds, would not have been made by anyone writing in Imperia l times, when chronology was determined by Olympiads. The first reason is very subjective, and the second, frail as it is, entirely leaves out of account the deliberate archaization of ps.-Herodotus. On the whole the Hadrianic period is much the more likely for the composition of this Life: thus there is no reason so far for dating the immediate common source of all the Lives (see below, p. 155, n. 1) earlier than the beginning of the Christian era.

^{page 153 note 1} Of course the restoration ταυτῃ[ν] is not certain; ταυτῃ[σ] is an alternative, and may be slightly preferable in the present case.

^{page 155 note 1} This is the period plausibly suggested by Allen, (flrigins, 33)Google Scholar for the composition of the ὑπ⋯μνημα or commentary which is the basis of all the Lives. Aristarchus, Crates, Dionysius of Thrace, are the latest authorities named in the Lives except the special sources used by Hesy chius; they worked in the second century B.C. In the first century A.D. there was a revival both of work on Homer and of the Ionic dialect, in which the Herodotean Life is written.

^{page 156 note 1} ‘Giving to him’ is not entirely clear: it is probably to Homer as imagined author of the moral pronouncements in the non-griphic partsof the Agon that gratitude is given. The Agon consists of four sections different in character:(i) two straightforward questions by Hesiod, τ⋯ π⋯ρτατ⋯ν ⋯στι βροτοῖσιν; and τ⋯ θνητοῖσιν ἄριστον οἴεαι ⋯ν =ρεσ⋯ν εἶναι; with replies by Homer, the second of which isOd. 9. 5–11. (ii) The griphic section, of which the collection of delayed predicate paradoxes forms the main part, (iii) A rather weak interrogation on ethical subjects. (iv) The recital by each poet of what he considers to be his best passage. These sections may not all have been compiled at the same date: Wilamowitz, , Il. u. Homer, 403Google Scholar, plausibly suggested that (iii) is an addition.

^{page 156 note 2} The argument that the impropriety of two of the groups of lines made such a collection unsuitable for school use, revived by Allen, was disposed of by Wilamowitz, , Il. u. Homer, 400Google Scholar.

^{page 157 note 1} The contention of Certamen, § 9, ad init., that the first line in each case was actually a Hesiodic line, and the next line or lines Homeric,need not be taken seriously; in fact ἔστιν οὖν ⋯ μ⋯ν πρ⋯τος [sc. στ⋯χος] 'Hσι⋯δον, ⋯ δ⋯ ⋯ξ⋯ς 'Oμἠρον κτλ.looks like a gloss.

^{page 157 note 2} Comprehensive treatments of Greek riddles by W. Schultz in R.E., s.v. Rätsel, and K Ohlert, Rätsel u. Gesellschaftsspiele der alten Griecken².

^{page 157 note 3} It is possibly of interest that myfirstclose acquaintance with the Lives of Homer—for would scarcely read them for their literary merit—was due to a similar doubt to Bywater's of the genuineness of Heraclitus 56. One's instinct is to put all such fictions about great figures in, or after, the Alexandrian period. That such doubts cannot be lightly dismissed is shown by the fact that they are held, not only by Bywater, but also by Professor D. S. Robertson and Mr. W. Hamilton.

^{page 158 note 1} Diels, , SBBerlin (phil.-hist.kl.) ix, 1901, 190 f.Google Scholar, held that Heraclitus was particularly instrumental in the formation of verbal substantives in -σις. Snell, , Hermes, lxi. 377Google Scholar, observed that such substantives in H. are all words of perception or learning.

^{page 159 note 1} Although Reinhardt, K., for example (Parmenides u. die Geschichie der griechischen Philosophie, 206, 214; Hermes, lxxvii. 27)Google Scholar, and Calogero, G. (Giornale Criiica delta Filosqfia Italiana, xvii, 1936, 222)Google Scholar are right in making the riddle analogous with the riddle in the whole of nature, i.e. that all things are one from certain points of view.

^{page 160 note 1} Heraclitus presumably did not himself invent either the lice-riddle or its connexion with Homer. There may be a genuine clue to the origin of this connexion in the common account represented by Certamen, ps.-Plut. 1, Proclus, Tzetzes, according to which the poet at some stage in his wanderings consulted the Delphic oracle about his birthplace (or περ⋯ ⋯σϕαλε⋯ας in Proclus). The oracle accepted or invented the tale that Homer's mother came from Ios, and may have conceded that Homer died there, Delphi seems to have been a storehouse of riddles, and the lice-riddle may have been a convenient means of changing the subject from a delicate and confusing question, that of Homer's native city.

^{page 160 note 2} Heraclitus talked of παῖδες, without specifying that they were fisher-boys. He of course was primarily interested in the fact that Homer's interlocutors were mere children, and so failed to describe them any further. None the less the fact that the boys had been out fishing is probably an original and essential part of the riddle: it made the riddle much more difficult, because the victim naturally thought that the things which were caught or not caught were fish, and so was distracted from the true solution. The same motive would apply in the case of hunters, if θηρ⋯τορες is accepted: see below.

^{page 163 note 1} Other references to the story: (i) a wallpainting at Pompeii, with the descriptions OMHPOS and AAEIS, and a version of the fishermen's riddle (Kaibel, , Epigr. Graeca, no. 1105)Google Scholar —this is the earliest dateable evidence for the story after Heraclitus, and might point to a terminus ante quem for the ⋯π⋯μνημα. (ii) Valerius Maximus,9.12.3. (iii) Tzetze's version of Homer's question and the reply (i.e. with ⋯λι⋯τορες) recurs in the Anth. Pal. ix. 448.

^{page 166 note 1} τ⋯ν δεξ⋯αν with πλενρ⋯ν occurs only in Tzetzes a, but may be simply an addition by Tzetzes for both graphic and rhythmic effect.

^{page 167 note 1} I am greatly indebted to Professor D. S. Robertson for his guidance and criticism on all aspects of this paper. Mr. C. H. Roberts, Mr. W. Hamilton and Mr. H. Lloyd-Jones have been kind enough to advise me on special points.