^{page 288 note 2} He is variously credited with 3,500 or 4,000 books. The figure has no doubt been tidily rounded off, but need not, as is often said, be wildly exaggerated. The papyrus itself gives us some īdea of the length of one such book; with the assistance, which Didy-mus may reasonably be supposed to have enjoyed, of trained slaves acting as research assistants, stenographers, and copyists, it ought not to have been difficult to produce two a week; such a routine, if followed for thirty-five years, would yield the required total. But more important than actual figures is Quintilian’s ‘Didymo quo nemo plura scripsit’ (1.8. 20).

^{page 288 note 3} Ed. pr. Diels u. Schubart, , Berliner Klassikertexte, i (1904)Google Scholar, cf. Didymi de Demo-sthene commenta (Leipzig, 1904).Google ScholarThe secondary literature is extensive: the following seem to me to be the most important discussions: Blass, , APF iii (1904), 284 ff.Google Scholar; Leo, , ‘Didymos’, NGG 1904, 254 ffGoogle Scholar. (═Ausgewählte, kl. Schr. ii. 387 ff.Google Scholar); Foucart, , ‘Étude sur Didymos’, Mémoires de l’Institut national de France, Académie des Inscriptions et belles lettres, xxxviii (1909), 27 ff.CrossRefGoogle Scholar; Lossau, , Untersuchungen z. antiken Demosthenesexegese (Palingenesia, Bd. ii), 1964Google Scholar; on the historical fragments Jacoby’s commentary supersedes all earlier discussions.

The papyrus itself is in the Staatliche Museen, Berlin; I re-examined it in July 1967, and should like to take this opportunity to thank Dr. Müller and his assistants for their help. It is perhaps worth noting here that nearly all the allegedly improved readings published by Crönert, (Rh. Mus. lxii (1907), 380ffGoogle Scholar., GGA 1907, 267 fr.) are illusory.

^{page 289 note 1} The meaning of the numerals has been disputed: see further Wilcken, , ‘Die Subscription des Didymus-Papyrus’, Hermes, lv (1920), 324Google Scholarf.

^{page 289 note 2} On μόρν we may also compare the note in the fragment of a lexicon to Dem. Or. 23, republished by Diels and Schubart along with the Didymus papyrus (Pack², 317).

^{page 289 note 3} A fourth item in Diels’s list is simply due to a false reference.

^{page 290 note 1} In this and subsequent quotations from the papyrus I have not indicated doubtful letters or lacunae where the context makes the reconstruction certain.

^{page 290 note 2} On(10. 15, cf. 8. 44),(10. 16, cf. 8. 45),(10. 37),(13. 4, cf. 1. 28).

^{page 291 note 1} S.v.,cf. s.v.; normally he simply cites Didymus without giving a more precise reference.

^{page 291 note 2} See LSJ and the detailed discussion of Bömer, F., ‘Der Commentarius’, Hermes, lxxxi (1953), 215 ffGoogle Scholar. Leo attached a good deal of importance to the distinction which Didymus himself makes between Aristarchus’: Sch. A on Il 2.111:; the former are evidently in this case more detailed studies. But though the two words may take on a more precise significance in a particular context, it is wrong to draw a rigid distinction between them, as is clear from Galen 15. 1, where the two terms are used interchangeably of a single work:

^{page 292 note 1} Cf. Lossau, op. cit.

^{page 293 note 1} The speech is not altogether easy to date and Didymus might have confessed himself baffled without loss of face. It most probably belongs to 353/2, cf. Cawkwell, , ‘Eubulus’, JHS lxxxiii (1963), 48 n. 9Google Scholar. The wrangle over the border territory went on for some years; it had certainly begun by 352, as we know from IG ii ². 204, a decree of late 352, which refers in 1. 55 to a previous decree. It is not clear whether Didymus can reasonably be blamed for failing to realize this. The fragments of Androtion and Philochorus relating to 350/49 give no hint of the earlier history of the affair, but we cannot safely infer from this that nothing had previously been said on the subject.

The authenticity of 13 has been questioned in modern times (see, most recently, Sealey, R., ‘Pseudo-Demosthenes XIII and XXV’, REG lxxx (1967), 250 ff.CrossRefGoogle Scholar), but there is no reason to think it was suspected in antiquity. It is true that Dionys. Hal. (ad Amm.) ignores it, but the uncertainty of its dating renders it useless for his argument, and we cannot take his silence to indicate suspicion. It would, therefore, be unreasonable to criticize Didymus for failing to raise the question of its genuineness.

^{page 293 note 2} The restoration σ]ạφές; is not altogether certain, but the sense is surely beyond doubt. The editors, who take the subject of ποιήσει to be Demosthenes, print a stop after it; there is no punctuation in the papyrus here.

^{page 294 note 1} Ed. pr. Henrichs, A., Didymos der Blinde, Kommentar zu Hiob, Bonn, 1968.Google Scholar

^{page 294 note 2} e.g. pp. 7/8 (on the identity of the author)(almost a whole line is left blank)(the latter half of this line and the first half of the next are blank): the second of the missing texts, thatis presumably Job 42. 17b; p. 111 (on Job 4. 19: five lines are then left blank; the reference is presumably to Si. 19. 3. For a list of omissions, see vol. i. 19.

^{page 294 note 3} For detailed arguments against Didymus'view, see Cawkwell, art. cit. 61–2 n. 85.

^{page 295 note 1} The supplement is Crünert's; many other suggestions have been made with equal plausibility. But there is general agreement as to the sense.

^{page 295 note 2} I find wholly unconvincing Bruce's further suggestion that the passage of Philo-chorus which Didymus here quotes really relates to the peace congress of 386, and that Didymus has somehow made a mistake in dating it: this hypothesis creates problems greater than those it is intended to solve. In any case, the mistake assumed is not easy; it is difficult to see how Didymus could have arrived at the year of Philocles at all if it is not in fact correct.

^{page 295 note 3} , the restoration of the ed. pr., is very uncertain.

^{page 295 note 4} Didymus also cites Anaximenes in his account of Hermias (col. 6, 59 ff.) and in connection with Aristomedes (col. 9, 51 f.); it is possible, though not perhaps very probable, that he found the references to Anaximenes in an earlier writer and had not himself consulted the works concerned.

^{page 295 note 5} In the Teubner text it is given thus:. Something is clearly wrong with the text: Didymus cannot have meant to say ‘Some interpreted it in a rather vulgar way’, since there is nothingabout accurate lexical comment. The editors suggest emendingto(seil.; the apparent absence of a main verb led someone to alter the text. Much has been made of, but the traces in question are very uncertain, as is clear from the Abschrift of the ed. pr.; the point may easily be verified from the reproduction of Tafel i. It is not an intrinsically plausible supplement, since it is obviously odd to give relative chronology without names, and the argument from internal, stylistic evidence was not an early, crude form of the argument from external evidence.