^{page 62 note 1} Schwyzer, 2, 352: ‘Weder bei Homer noch später sind Ind. Präs. (Perf.) … Modalpartikel anzuerkennen.’ So also Gildersleeve-Miller, 433; Kühner-Gerth, 1. 210. 2; Goodwin, , Moods & Tenses, 195Google Scholar; Stahl, 252–3.

^{page 62 note 2} For such anarthrotic substantives used with pronouns predicatively, that is to say as the equivalents of new statements, and so more or less as the equivalents of new verbs, cf. KG 1. 628. 6 and especially Harry, J. E., “The Omission of the Article with Substantives after oύtoς, őδє, kiνoς in Prose”, TAPA, 1898, pp. 48–65.Google Scholar

^{page 62 note 3} Cf. Gildersleeve-Miller, 444: ‘The optative with ἄ03BD; is often used in combination with the indicative, sometimes as a climax, giving, as it does, the warmth of personal conviction.’ This interchange of the indicative with the optative+ἄν is also mentioned in the somewhat misleading remark in KG 1. 235. 6. 1. The matter is discussed at greater length in chapter 17 of my Zürich Dissertation, ‘19 Kapitel zur Textkritik und Syntaktischen Theorie der Attischen Autoren’ (to be published in January 1971).

^{page 62 note 4} Cf. KG i. 243 f.; Gildersleeve-Miller, 458; Smyth-Messing, 1766; Krüger-Pökel, 1. 69. 7. 2: ‘Fehlen kann das zu ἄν gehörige Verbum wenn das vorhergehende (sc. Verbum) (in der da stehenden oder in einer sinngemässen Form) zu ergänzen oder ein allgemeiner Begriff wie , ciVoic zu denken ist.’

^{page 63 note 1} This better opportunity is certainly not that of adopting the suggestion of Professor Erbse, H. in his article ‘Textkritische Bemerkungen zu Xenophon’ in Rheinisches Museum, 103 (1960), pp. 157 ff.Google Scholar Erbse is a good Hellenist: he understands the context and knows that the passage is ironical in tone. He has moreover a valid point when he observes that one has the feeling that some precise intent or relationship of the collocation of pronouns γẁ αvтόs, ‘even I myself’, escapes us. But when he goes on to pronounce that we must print , that is, after all, really only a very wild stab. What Erbse seems not to understand is that it is one thing to feel insecure about the precise meaning of a turn of speech, and quite another thing simply to rewrite it arbitrarily.

^{page 63 note 2} Xénophon, Banquet—Apologie de Socrate, Texte établi et traduit par François Ollier, Paris, 1961.Google Scholar Oilier, who speaks in the ‘Notice’ of his edition (p. 35) of the ‘valeur de Q, qui avait été méconnue par les différents editeurs du Banquet’ remarks in a footnote to this statement: ‘J';ai moi-mê;me revu entièrement les manuscrits Q et R grâce aux photographes qui m'ont été procurées par l'Institut de recherche et d'histoire des textes.’ I presume that Ollier's reading at this point is taken from one of his new sources. I find it nowhere else except, curiously enough, in Krüger's Schullehre, 1. 65. 3. 4. However, since Oilier does not specifically say so, it could be his own conjecture we have here—as for that matter the reading in Krüger could be his conjecture. If it is a conjecture it is a good one. But why, once they were about it, did Krüger and Oilier not go on and remove the brackets and write ἄν-as a prepositional prefix ? Oilier surely would have mentioned the fact in his apparatus if he had simply undertaken to change the position of ἄν on his own account. On the other hand, the possibility of an identical misprint in Krüger and in Oilier seems remote. I therefore proceed on the assumption that the text which Oilier gives has some manuscript authority.

^{page 64 note 1} Not only with єύpίσkw is there confusion between ἄν the modal adverb and ἄν- the prepositional prefix. The same confusion occurs for example with other verbs at Hes. Erg. 131; Pind. Nem. 7. 68 and Plat. Leg. 712 e.

^{page 64 note 2} The reading of B νєupθο is impossible because the subjunctive without ἄν is irregular in dependent sentences in Attic (KG 2. 426. 1). The reading of W ἄν єєη is impossible because that would be a potential of the past or an irrealis and neither would give a good sense. The reading of T ἄν єєη is unlikely because the position of ἄν, which usually attaches itself to the relative, while not absolutely impossible═cf. Gildersleeve—Miller. 466—is highly irregular.

^{page 64 note 3} is unlikely because Attic avoids ἄν with the future (KG 1. 209).

^{page 64 note 4} Frid. Guil. Sturz, Lexicon Xenophonteum, Olms, Hildesheim, 1964 (Reprographischer Nachdruck der Ausgabe Leipzig, 1801 ). This feeling on the part of the scribes leads naturally also to confusion between the compound and the simple form of the verb in passages where ν- is not retained as being the modal particle ν. So we find for instance at Xen, . Cyrop. I. 6.Google Scholar 40 that only D and F have, νηύpισkoν whereas CAEGR and H have єupισkoν, In this case it may even be that the simplex is correct and the compound verb a well-intended correction. The simplex is used in the very close parallel at Xen. Mem. 3. 11. 8 and in this passage there is no variant in any manuscript.