^{page 1 note 1} This is not the place to say how much I am indebted to Professor Schubart both for my acquaintance with papyrology and for much else besides. Here I will only thank him for checking my readings of the text. I may add that I am indebted also to Professor Friedländer of Halle and to Professor D. S. Robertson of Cambridge, who very kindly read my manuscript and made some valuable comments.

^{page 4 note 2} After Σηρή in 1. 4 a similar mark seems to have been omitted by the copyist.

^{page 4 note 3} According to the principle of lectio rarior I do not suppose Σηρή 1. 4 to be a scribal error for Σειρήν (although the latter form occurs in Pap. Ox. 659=Partheneia fg. 104d 33 Schr.); the Alkman papyrus gives Σηρηνίδων (fg. 1. 96D) and σήριον (ib. 62; cf. Herodian II 573L σηρά for σειρά). Hence Crusius changed σειρήν to σηήν in Alkman fg. 10D. Solmsen (Beiträge z. griech. Wortforschung I 1909, p. 129) remarks: ‘Σηρηνίδες beweist, dass σειρήν gedehntes e in der Wurzelsilbe hatte.’ So we may infer that the Dorian form was usual in choric poetry.

^{page 5 note 1} Cf. also I. II 14 c. schol.: Ξενοκράτει … αὐτῷ (less striking, indeed, as there the participle ὀπάσαις relates to the substantive, and πέμπεν to the pronoun). All these places have been assailed by conjectures.

^{page 5 note 2} Apparently there was a critical sign in the text.

^{page 5 note 3} See also schol. Ol. III 67e, schol. Eur. Phoen. 497.

^{page 5 note 4} See schol. Arist. Aves 1293 with White's note; cf. schol. Ar. Nub. 23. The idea that χελιδών might mean the part of a horse's hoof so called, or even a horse's name, does not help to interpret the text.

^{page 5 note 5} He might have found support for his hypothesis in the use of πτέρνγες or πτερύγια for certain parts of the λόγχη or ξιφος: see Pollux V 21, Hesychius s.v. πτερύγια (Professor Friedländer); perhaps also in the double meaning of the Latin bipennis.

^{page 5 note 6} Professor Zahn of the Berlin Museum was kind enough to show me a convincing explanation of this mark : on Dipylon-vases there are pictures of horses above whose backs great double axes are painted. It seems, then, to be originally a sacred symbol, as in the Cretan cult; cf. the κηρυκε ῖον and the snake, which are found on vases as marks branded on horses, indicating originally that the horse was consecrated to Hermes or to Hades.

^{page 5 note 7} For I. II 6 is of another kind.

^{page 5 note 8} Like Soph. O.T. 1378. In Hom. Od. XXII 222 the first of three οὐδέ takes the place of οὐ.

^{page 5 note 9} Ad Ol. II. 15/29.

^{page 5 note 10} Ib. 46/82.

^{page 5 note 11} Ad Nem. IV 3/5.

^{page 5 note 12} Ad Ol. X 70/83.

^{page 5 note 13} See schol. Ol. XIII 20/27, Ol. X 70/83.

^{page 5 note 14} A reference to the Latin bipennis (see above, note 5) would also be natural to a grammarian who lived at Rome and wrote de latinitate.

Here is one less weighty argument: from the wretched remains of the left column of the papyrus only 1. 9 ]βικα seems to be of some significance. I see four possible supplements :Ἀραβικά, βέμβικα, ἄμβικα, κίμβικες. The first two may be disregarded. Ἄμβικες in Athenaeus XI 480d are said to be Ἄργεῖαι κύλικες, something like which might have been written by Pindar. Now the ‘cup lexicon’ in Athenaeus is almost certainly derived from Pamphilus (cf. L. Cohn, Jahrb. f. klass. Phil. Suppl. 12, 1881, 324 and Rudolf, Leipz. Stud. 7, 1884, 125); and frequent citations of Didymus by name make it very probable that Pamphilus here borrowed from Didymus (the Σύμμικτα ᾒ Συμτοσιακά, I suppose), all the more so as the explanation of λεπαστή Athen. 485a (finally derived from Aristophanes Byz. fg. 77 N.) occurs also in Hesychius s.v. and schol. Arist. Pax 916, and that of κυμβίον Athen. XI 477 f. in Hesychius and Harpocration. Ἄμβιξ is found only in this passage of Athenaeus; if it appeared in our papyrus, it must have been Didymus who was quoting here from one of his lexica, as he did so often.—Finally κίμβικα : Xenophanes in the Σίλλοι used this word of Simonides (see schol. Aristoph. Pax 656, Wilamowitz, Sitz. Ber. Berl. Ak. 190, 1303). In the very learned scholion to Aristophanes—Didymean, I suppose—the fact is mentioned as evidence for Simonides' φιλαργυρία, which was attacked by Pindar in I. II and elsewhere and therefore is often commented on in our scholia to Pindar. As the author of our papyrus commentary was interested in Pindar's relations with Simonides, I hold the supplement κίμ]βικα to be the most likely, and its author to be sought in the common source of the extant Pindar scholia, the Aristophanes scholion, and Athenaeus, namely Didymus, who himself derived this information from Chamaeleon μερὶ Σιμωνίδον (see Athenaeus 656c, d and 456c.)

I am well aware that this is a highly uncertain calculation; I have only written it down to spare someone else the same trouble.

^{page 6 note 1} Ἵππος of course may be an addition of the commentator.

^{page 6 note 2} A δημηγορία of Peisistratos is mentioned by Aristotle, Ath. Pol. 15, 4.

^{page 6 note 2} As a variation of the prosaic δορυφόροι ?—πέλεκυς = executioner's axe : fg. trag. adesp. 412 N.

^{page 6 note 4} κεῖνόν τέ might refer to this opponent; or the object might be supplied : τὸν] οὔτε πελέκεις …; as we do not know whether in P's text our second lemma followed the first immediately.

^{page 6 note 5} See schol. Pyth. VII 17 = fg. 137b Schr.; cf. Wilamowitz Pindaros 155.

^{page 6 note 6} Notwithstanding that hitherto we have not known of a Didymus-commentary on the threnoi. If this ascription is correct, the view of Boeckh and Wilamowitz, that fg. 137a Schr. Belongs to this threnos, must be rejected on metrical grounds; while fg. inc. 141 Schr., which is in metrical correspondence with our fragment, might be claimed for it (cf. P. Maas Grieck. Metrik § 56, 3 on the ‘rhythmus Anaxiphorminges’).

^{page 6 note 7} Simonides a praiser of tyrants : see Plato Protag. 346b 5.