¹ This paper was originally written as a contribution to a series of classes on early Sparta conducted by Professor A. Andrewes and Miss L. H. Jeffery. I am greatly indebted to Mr Oswyn Murray for a number of valuable suggestions, and to others present at the class (in particular Mr W. S. Barrett) for rescuing me from several errors; to Mr M. L. West, who is also preparing a paper (see now ‘Alemanica, I: The date of Alcman’, CQ n.s. xv [1965] 188–94) on this papyrus, for some additional references, for criticism, and for allowing me to see a draft of his article before publication; to Mr N. G. Wilson; to Mr W. G. Forrest; and to Mr E. Lobel, who was good enough to answer an inquiry on a point of detail. The responsibility for faults which remain is my own.

The papyrus was first published by Lobel, E. in Oxyrhynchus Papyri xxivGoogle Scholar (text, 52–3; commentary, 54–5; photograph, plate viii). It is discussed in reviews of that volume by Page, D. L. in CR n.s. ix (1959) 20–1Google Scholar, and by Barrett, W. S. in Gnomon xxxiii (1961) 688–9.Google Scholar See also Janni, P., ‘Alcmane: problemi di cronologia’, Studi Urbinati xxxiii (1959) at 168–71Google Scholar; and Davison, J. A., ‘Notes on Alcman’, in Proceedings of the ninth international Congress of Papyrology, Oslo, 1958 (Norwegian Universities Press, 1961) 32–4.Google ScholarPage, D. L., Poetae Melici Graeci (Oxford, 1962) 22–3Google Scholar (Alcman 5 fr. 2), adds nothing to his review. For the later part of the second poem, which I do not discuss, see Bowra, C. M., Greek Lyric Poetry ² (Oxford, 1961) 25–6Google Scholar; Fränkel, H., Dichtung und Philosophie des frühen Griechentums ² (Munich, 1962) 184–5, 290–1Google Scholar; Burkert, W., review of Fränkel, , Gnomon xxxv (1963) 827–8Google Scholar; and West, M. L., ‘Three Presocratic Cosmologies’, CQ, n.s. xiii (1963) 154–6.CrossRefGoogle Scholar I cite these by the author's name only; ‘Page’ refers to the CR review.

² For the trouble over καὶ τρίτος σκότος (col. iii, lines 21 ff.) see Lobel, 55, Page, 20 and Barrett, 689; contra West, 156.

³ Davison writes: ‘It appears that Alcman spoke of him (i.e. Leotychidas) as δαίμων, i.e. he was already dead when Alcman was writing’. Despite the evidence of Xen. Lac. Pol. 15.9, this remains conjectural.

⁴ Daly, L. W. (A. J. Phil. lxxxi [1960] 91)Google Scholar wrongly says that Leotychidas appears in the lemma. See, however, n. 3 supra.

⁵ As Huxley, G. L., Early Sparta (London, 1962) 117 n. 253Google Scholar, implies.

⁶ A procedure rightly condemned by Beloch, , Griechische Geschichte ² (Strassburg, 1913) i 2.179.Google Scholar

⁷ Op. cit. 61.

⁸ He elides elsewhere: col. ii, 6, 12; col. iii, 25. But in col. iii, 21 we have πρέσγ[υς] δέ ἀντὶ (Lobel's reconstruction; the scribe himself abbreviated to δ′): should this be πρέσγ[υς] δὲ ἀντὶ?

⁹ One possibility is πότερον; but if so, I cannot reconstruct the passage as a whole, πότερον τούτου θ]νγάτηρ ἤ (instead of ή; we have seen that the commentator dispenses with articles with names, above) Τιμασιμβρότα has been suggested; but then we would surely require etc., which would be too long. West informs me that he proposes

¹⁰ ‘The writing is very irregular, and for the larger gaps in particular the estimates of letters lost are more than usually approximate,’ Barrett 688, n. 2. For this reason I have not indicated the length of gaps, for which see Lobel.

¹¹ This suggestion, which is central to my argument, is one of several for which I am indebted to Mr Oswyn Murray.

¹² There is no need to follow Barrett's tentative suggestion (689 n. 1) that the scribe may have written Λεωτυχίοα by mistake for Ἀναξάνδρον (on the grounds that the name Leotychidas, which has just occurred five lines before, might have been in his mind, and that all would then be plain sailing). This proposal should be rejected because (a) the papyrus can be made to yield perfectly good sense as it stands; (b) the scribe does not seem prone to error (but see p. 62 supra). If Barrett is right, and Λεωτυχίδα is a slip of the pen, then we cannot know what the scribe should have written.

Some think the repetition of Λακεδαιμονίων in lines 15 and 20 strange, and suggest ‘the Eurypontid king’. I regard the repetition as recapitulatory, and therefore unobjectionable. West favours τοῦ [προειρημένου] βασιλέως.

¹³ Is it unnatural to use the phrase ‘Polydoros, the fair-haired son of Eurykrates’, if the comparison is between brother and sister? I think not. The probability that Timasimbrota is the subject might be increased if we knew more about the poem as a whole. (Lines 9–10 show that the poem is not concerned exclusively with girls.)

¹⁴ Something indicating that a conclusion is being drawn would be preferable to the bald ἐστι; but there is no room for a particle, and is out of the question.

¹⁵ Davison, whose paper appeared before Page's and Barrett's reviews were available, says (34) that Timasimbrota may be Leotychidas' daughter, and that Polydoros may be his son. The last letters of line 20 are sufficient to rule this out. (He notes that this would imply that the name Polydoros had changed families. There is nothing wrong with this—Anaxandridas appears on both the Agiad and Eurypontid family trees; but his guess must be rejected on other grounds.)

I agree, however, with Davison, 34, and with Janni, 169, that the present fragment makes the suggestion that the Agido of Alcman fr. 1 Diehl is an Agiad highly probable.

¹⁶ Evidence for the genealogies:

(i) For Agiad line: Hdt. vii 204.

(ii) For Eurypontid line: Hdt. viii 131.

(iii) For the family of Anaxandridas the Agiad: Hdt. v 41.1, 3.

(iv) For making the Eurypontid line from Anaxandridas to Hippokratidas kings, but not Hegesilaos and Menares: Hdt. viii 131 unemended (see p. 64).

(v) For making Leotychidas I king: Plut. Mor. 224 C–D; Plut. Lycurg. 13.7; Rhianus ap. Paus. iv 15.2; P.Oxy. 2390.

(vi) For making Agasikles and Ariston kings: Hdt. i 65.1; i 67.1.

(vii) For making Archidamos, Zeuxidamos, Anaxidamos and Archidamos kings: Paus, iii 7; for rejecting them, see Huxley, op. cit. 117, n. 252.

(viii) For splitting the Eurypontid family tree after Hippokratidas: see Huxley, op. cit. 117–18, n. 253, where another possible reconstruction (Beloch, op. cit. i 2.189) is also given.

(ix) For synchronisms: (a) Alkamenes and Theopompos: Plut. Mor. 239 F. (b) Polydoros and Theopompos: Plut. Lycurg. 6. (c) Anaxandros and Leotychidas I: P. Oxy. 2390? (d) Leon and Agasikles: Hdt. i 65.1. (e) Croesus, Anaxandridas and Ariston: Hdt. i 67.1.

¹⁷ So far, this is m agreement with Page.

^* See Addendum, p. 73.

¹⁸ But see n. 3.

¹⁹ See Page, D. L., Alcman, The Partheneion (Oxford, 1951) 164–6.Google Scholar Page and Davison extract conflicting dates from Eusebius. The years I give are from the edition of R. Helm, Leipzig, 1913.

²⁰ This conclusion as to the date of Alcman has already been drawn by Janni, 171. The new date does not conflict with Page's general conclusion, op. cit. 166. The latest discussion of Alcman's date is by Davison, who believes (33–4) (i) that Leotychidas was dead when Alcman wrote the poem commented on by our papyrus (see n. 3 supra); (ii) that Eusebius' first date, 659/8, is nearer the truth than that in the Suda, 672/68; and (iii) that Eusebius' alternative date can safely be eliminated from the discussion. But (i) and (ii) seem hardly compatible, for it would follow that the reigns of kings Hippokratidas and Agasikles between them covered the greater part of the century between the death of Leotychidas (before 659/8, according to Davison) and the reign of Ariston, which according to Herodotus (i 67.1) corresponds with that of Croesus (c. 560–546 [P-W. suppl. v, 457]). This is unlikely.

²¹ The scribe does, however, abbreviate fairly frequently: / for ἐστι in fr. 2 col. ii, 24; fr. 30, line 2; but sometimes ἐστι in full, fr. 2 col. ii, 19; and other abbreviations in fr. 1 (c), line 2; fr. 2 col. ii, 14, col. iii, 14, 21, 23; fr. 33 (c), lines 2–3; fr. 49 col. ii, 13. Alternatively, the dislocation of the papyrus (see above) might perhaps allow the longer restoration.

²² φνσικός is Aristotle's regular word for the pre-Socratics in the Physics and elsewhere; see L.-S.-J. s.v. φυσικός Π 2.

²³ ἐκφορά can mean ‘expression, enunciation’ of ideas (L.-S.-J. sense V). On the analogy of such phrases as (Dion. Hal. de comp. verb. 8) the genitive must refer to the thing expressed; so if ἐκφοράς is read, τῶν λοιπῶν cannot mean ‘the other commentators’. There are no doubt other possibilities besides πεί]ρας and ἐκφο]ράς.

²⁴ Whether -ατρος is lemma or comment, it is certain that τῆς in line 24 must be comment. The lack of a connective after τῆς tells us nothing: cf. lines 14–15.

²⁵ Evidence in P-W. xvii, 1694 (V. Ehrenberg, s.v. Obai).

²⁶ Pitana: P.Oxy. 2389 col. i, 3, 7, 8, 11, possibly 10. Dymainai: col. i, 5, 7.

²⁷ The entry in Hesychius is rejected by Ehrenberg (P-W. xvii, 1695–6 [s.v. Obai]); it would be easy, but not helpful, to multiply references.

The chief objection to Dyme as an obe is that it would be perverse for anyone naming a set of new tribes to think up a name so similar to one of the old. But perhaps his choice was not free: these are place-names and local divisions, and if there were a place called Dyme, he would have no option. The place could have received this name because Dymanes settled there in early days. A parallel of sorts might be Pamphylia, which Huxley (op. cit., 15) thinks was settled by the same Pamphyloi as at Sparta. However, no one has succeeded in finding a Dyme in Laconia.

The feminine form Δύμαινα is confirmed by Stephanus of Byzantium s.v. Δυμᾶνες, as well as by P.Oxy. 2389.

²⁸ Barrett rightly separates the words -ντιδων and φυλή.

There are other ways out of the impasse. One is to read Or φυλή could mean something other than a ‘tribe’ (as I have understood the word throughout this discussion): then we might have a But such a usage would be more probable in poetry than in a late commentator. (Hdt. iii 26.1 is not a safe parallel.) He could, of course, have got the phrase from the poem; or he may have been ignorant enough to think that the Eurypontids really were a φυλή.

²⁹ L.-S.-J. cites Pindar, Pyth. viii 38Google Scholar, cf. Nem. vi 36, viii 46, and I.G. v (2) 495 (Megalopolis, second century A.D.); cf. also Andrewes, A., ‘The Patrai of Kamiros’, BSA lii (1957) 30–7.Google Scholar