^{page 91 note 1} The trade in table grapes (‘escariae’) would be near towns, cf. Colum. III. ii. init. ‘ad escam…promum uendere.’

^{page 91 note 2} An exact equivalent for the phrase does not seem to be possible in English as it is in German, where Apelt renders ‘die sogenannte Edeltraube’.

^{page 91 note 3} ἐπαφῶσιν is from ἐπαφίημι—not from ἐπαφάω For the use cf. Theophr. H.P. VIII. 4, τὰ πρóβατα ἐπαφιᾶσιν, they let loose the cattle (in the cornfields).

^{page 91 note 4} And to the substance of the expression we should add another century. Galen (131–201 A.D.) περὶ τροφῶν δυνάμεως, Kühn VI., p. 576, says ὀνομάζουσι δὲ εὐγενεῖσ τὰσ τοιαύτας σταφυλάς, the grapes, that is to say, that have rather less juice (τὴν ὑγρὰν οὐσίαν) and more pulp (σάρξ). (He proceeds to talk of their storage, but it is wet storage, as described also by other writers, e.g. Pliny, Athenaeus.) Presently, 578 fin.-9, he says that the σαρκώδεις are the best for eating, whether from the vine or after hanging. Galen uses the word εὐγενής instead of γενναῖος in conformity with his regular antiarchaism; cf. e.g., p. 579, med.

^{page 91 note 5} Plin. N.H. XV. 17 (18), § 64; id. XV. 21 (23), § 85: Col. iii., ch. 2. By the expression uitis generosa Columella is not referring to vines which produced ‘generous’ wines, as in such cases he says generositate uini (e.g. commendantur).

^{page 91 note 6} Met. XIII. 817–8 contrasts pruna … liuentia with generosa nouasque imitantia ceras (cf. Verg. E. II. 53 cerea pruna) in a passage imitated from Theocr. XI. (where, although nothing is said of plums, Galatea is promised, 46, ἂμπελος ἁ γλυκύκαρπος, evidently the γενναία); but ibid. XV. 710 et Surrentino generosos palmite colles alludes to the fine Surrentine wines.

^{page 92 note 1} Cf. e.g. Theocr. XXL, and see Tarn, , Hellenistic Civilization, pp. 103, 164Google Scholar.

^{page 92 note 2} The scene as described in 625B is almost Theocritean! I note even κυπαρίττων … ὓψη.

^{page 92 note 3} I say ‘even’ because, as I understand the passage, the casual picking of other people's γενναῖαι is in general permitted to the citizen; the stipulation μὴ πείσας in 844E amounting to no more than ‘if the owner objects.’

^{page 92 note 4} But I have heard of plum-orchards being thrown open (locally) in England at the height of a glut. Cimon's throwing open his orchards to the public (Theopompus ap. Athen. XII., P 533A) was a bid for popularity.

^{page 92 note 5} As Mr. A. S. F. Gow produces no evidence for his portent of a bread-eating fox (C.R. XLIV. 9–10), I will cite the only evidence there is, viz. Babrius 86 = Aesop 30 (Chambry); and this passage was all the more relevant as it would appear to be the only other passage where a fox eats out of a πήρα. With some questions raised no by this fable I hope to deal on another occasion; in the meantime all that need be said is that the farinaceous element in this fox's diet obviously comes, with the substance of the fable itself, from our manuscript tradition of Hor. Epist. I. vii. 29.

^{page 93 note 1} Hom. Od. XVII. 357–8; Ar. Clouds 921–4.

^{page 93 note 2} Od. IX. 212–3; cf. V. 266–7.

^{page 93 note 3} 304 Chambry (ed. mai.). For πήρα as receptacle for articles other than food cf. e.g. Menander, Epitr. 114 (a characteristic function. of this utensil, cf. e.g. Longus I. 6)

^{page 94 note 1} The present note does not exhaust such references; I am reserving others for my note on Ar. Wasps 1452 (as to which see p. 95, n. 4).

^{page 94 note 2} For figs cf. also Ath. III. 79A δυνάμεις δ' ἔχει [sc. τὰ σῦκα] τῶν ὑγρῶν.

^{page 94 note 3} Cf. the opinion of Democritus in Athenaeus II. 46F fin. (… μέλιτι βρέχοι).

^{page 94 note 4} Of five mistakes made by Rivaud (Budé translation) Professor A. E. Taylor—to whose writings I am of course, like all students of Plato, deeply indebted—corrects two (p. 118, n. 1, of his translation oí Timaeus and Critias, Methuen, 1929), but himself commits the other three, (i.) At 114E 8 he renders τά τε αὗ περὶ τὰ ζῳα as if it had been ζῷά τε. (ii.) At 115a 7, not knowing (as it is evident that Theocritean scholars also have not known) the meaning of ξηρóς as applied to produce, he takes τóν τε ξηρóν as balanced by καὶ ὃσοις in the following line, whereas it is, of course, balanced by καὶ τòν ὃσος ξύλινοσ. (iii.) At 115b 1 he translates ξύλινοσ ‘the woodland kind,’ and in his note is puzzled by the thus implied exclusion from Atlantis of revine and olive; but observe (a) ξύλινος, which Jowett rendered ‘having a hard rind,’ and Rivaud now renders ‘ligneux,’ is a regular and frequent word for fruit-trees or fruits as opposed to cereals; e.g. Strabo V. 4. 2 (240 fin.) βελτω δὲ τοῖς ξυλίνοις καρποῖς ἢ τοῖς σιτικοῖς; id. XV. 1, 20 (p. 693); Diod. Sic. III. 63 τὰς οἰνποιίας καὶ συγκομιδὰς τῶν ξυλίνων καλουμένων καρπῶν (cf. below τῶν σύκων καὶ τῶν ἄλλων ἀκροδρύων); same phrase in Athenaeus III. 78D; C.I. 93, 19 (opposed to Δημήτριος); cf. ξύλινον καρπóν τῶν δένδρων Hesychius; and similarly ξύλικóς in Artemidorus 2, 37, p. 215 (… μάλιστα ἀμπέλους). (b) ξύλινος of vines is correct, cf. Eur. Cycl. 572 σοφóν γε τò ξύλον (tree) τῆς ἀμπέλου (c) παιδιᾶς ἕν δυσθησαύρ shows that vines are particularly referred to, cf. Laws 844d 6: (d) Plato's ἀκρóδρυα is used by Theophrastus in H.P. IV. iv. 11 to include vine and olive. Professor Taylor here draws the conknowing elusion that ‘the Critias has all the marks of being an unrevised as well as an unfinished sketch,’ but to me this part seems written in Plato's best manner. I may, however, add that, while the sense is clear from the text as it stands, for δς in 115A 6 I do not doubt that we should read ὃσα; throughout this passage (from 114D 7) it is by ὅσα … καὶ ὅσα (which occurs revine peatedly) that all such subdivisions are made, and this present ὃς … καὶ ὅσοις is the only exception (the ὃς in 115B 2 being a determinative relative).

^{page 95 note 1} Editors of the Laws rightly resist Ast's γενναίαν for παιδιάν at 844d 6, but do not appear to know that it is definitely disproved by Critias 115b 6–7, as I go on to show.

^{page 95 note 2} Probably (or at least partly) because the ξύλινος, though in the main ὑγρóς, necessarily includes (what would strictly have come within the ξηρóς) dried fruits; for to them the term μεταδóρπια, distinguished (by ὅσα τε) from the δυσθησαιριστος, would naturally point. For Plato's use of θησαυρ. cf. Soph. Fr. 398 (p), 2.

^{page 95 note 3} This is the answer to Professor Taylor's difficulty, p. 118, n. 3.

^{page 95 note 4} Ar. Wasps 1452, where for ξηρῶν τρóπων, which is neither Greek nor appropriate sense, I read ξηρῶν τροφῶν, which is both; but this will have to be the subject of a separate note. In that note I shall exhibit many more examples of ξηρóσ (and ὑγρóς) in our sense in association with τροφή.

^{page 96 note 1} Cf. Galen, quoted below, on this page.

^{page 96 note 2} Cf. Alciphron II. (Epistulae Rusticae) 19, where a slave writes that he had been ordered by his master to guard the vineyard against foxes, and that having (through no fault of his own) disastrously failed to do so, he so dreads the consequences that he is resolved to run away.

^{page 96 note 3} The v.l. ἀκριδοθήκην is common to both passages; but you must catch your locust (Longus III. 24 ἐθήρων ἀκρίδας, Aesop 293 (Ch.) παῖσ πρò τοῦ τείχους ἀκρίδας ἐθήρευε) before you can make a pet of him (λάλον ἀκρίδα θεραπεύειν, Longus I. 14); and you catch him near vines, Theocr. v. 108–9. Ar, Birds 588.

^{page 96 note 4} For grape-guarding as a regular thing cf.Catullus xvii. 16 asseruanda nigerrimis diligentius uuis, and Anth. Pal. vi. 22.

^{page 96 note 5} Except that a small boy is, I take it, only a stop-gap guard.

^{page 97 note 1} I can find no parallel, and I have, of course, not overlooked Lobel Sappho, p. lxii. I agree with Mr. Lobel that ‘it would be interesting to observe statistically the relative frequency of these peculiarities’ (of synecphonesis) ‘in various kinds of literary styles,’ and all my own impressions incline me to accept his suggestion that ‘they are not so much metrical expedients as vernacular licences, which versification admits or excludes in proportion to its elevation.’

^{page 97 note 2} Perhaps (in Homer at least) ἤ less frequently than ἤ, but for various reasons (e.g. disputed accentuation in places) distinction would be unprofitable or deceptive.

^{page 97 note 3} Jebb thought of ἤ κείμεσθ', but did not read it, and it is in fact ruled out by κειμένῳ in 82.

^{page 97 note 4} Of the instances of hiatus after the second syllable of a spondee ‘a considerable number,’ says A. Shewan in C.Q. XVII. 18, are ‘concerned with ἤ or ἢ.’

^{page 97 note 5} Except that XVIII. 27 is false, and XXIV. 88 should be 90. V. 148 should be added (cf. Wilamowitz's ed.), and there may be others.

^{page 97 note 6} Fritzsche had reason to think a line had fallen out between ἢ εἰ and χρυσóν. I have this further reason to suspect something wrong, and I have another reason too; but the whole question is one with which I cannot deal here.

^{page 97 note 7} Thus Hesiod Scut. 407 ἢ ἀγροτέρης, hiatus.

^{page 97 note 8} Fritzsche parallels this πρὶν ἤ with ‘e. gr.’ Iliad V. 288–9. But those words (while they occur twice, XXII. 266–7) are the only instance in Homer of πρὶν ἤ, and the ἤ is required to fill metre. πρίν … πρίν, on the other hand, is very frequent in Homer, Goodwin M.T. 657. πρίν with subjunctive occurs in Homer, not so πρίν ἤ which indeed does not seem to be found with subjunctive except in Herodotus, with whom it is frequent enough. Theocritus uses πρίν ἤ only once, and that is where ἤ is required by the metre, V. 148.

^{page 98 note 1} For ‘an active verbal’ say ‘a verbal,’ and the only exception is the (ἁπ λεγ.) Homeric ληïστóς. I am of course (but in i. a only) ignoring privatives and other compounds (e.g. δυσλóγιστος, φιλοίκτιστος cf. φιλóθυτος) which are notoriously formed with considerable freedom, and are in fact in a different category; I am here concerned only with parts of verbs. σπανιστóς is from σπανίζω; it is active (neuter), and at that rate is (I have found) itself unique; unless one is to couple with it ὑβριστóς, against ancient authority but with some modern, cf. e.g. Bishop in Amer. J. Phil. XIII. 332; the fact of being comparable is no bar, cf. καταρατóτατος Soph. O.T. 1344, μακαριστóτατος Xen. Mem. II. I, 33.

^{page 98 note 2} Who for instance would believe for a moment in ἀκροβολιστóς ‘skirmishant’?

^{page 98 note 3} I see no ground for Radermacher's statement some on σπανιστóσ at O.C. 4 ‘schwerlich hat Sophokles dies Adjektiv geschaffen,’ and he himself has to follow it with (my italics) ‘obwohl es in der älteren Sprache sonst nicht vorzukommen scheint und erst nieder bei Strabon auftaucht.’

^{page 98 note 4} σπανιστóς is a nice example, cf. n. I, and, L. and S.'s references.

^{page 98 note 5} Here, very oddly, Greek contrasts with Latin, which just has actives here although so seldom otherwise (apart of course from deponents); pransus, cenatus, potus. But these words refer to past time; ἀκράτιστος here would have to mean ‘breakfasting.’

^{page 98 note 6} καθίζειν ἀκράτιστον ‘to set a-breakfasting’ has to be on the model of the phrase κλαίοντας καθίζειν. And as to that, Cholmeley's gay ‘κλ. καθ., etc.’ (ed. 2, p. 385) must not mislead us. There is no ‘etc.’; the phrase is always that, except that Plato once (Ion 535E) coins beside it its antithesis γελῶντας καθ. And καθίζειν is not the Greek for ‘to start some thing.’ This is an entertainer's phrase, and means ‘to make them sit weeping.’

^{page 99 note 1} Itself, I think, no exception; the cake being problematical, the salt becomes the basis.

^{page 99 note 2} But only as a pis aller; Legrand is entirely unsatisfied with current renderings of either of our two cruces.

^{page 99 note 3} C.R. loc. cit.

^{page 99 note 4} Both of these conjectures originated in mis-interpretation of a scholium, which is itself transparently a desperate attempt to explain our corruption.

^{page 99 note 5} Cf. Pearson on Eur. Phoen. 1485–6. Jebb's ‘partitive genitive’ at Soph. Ant. 1265 is absurd. In Soph. Ant. 1209 and O.C. 923 the abstractions ‘indistinctness’ and ‘supplication’ are used artificially, the expression being highly poetical; or the genitive may be of definition or ‘material’ in the latter passage.

^{page 99 note 6} There is another just after this; but I do not know of a third.

^{page 99 note 7} Even things like τὰ πρῶτα τῆς μοχθηρίας (Ar. Frogs 421), though in rather different case (cf. Eur. Med. 917), seem to be very rare. In Plato Charm. 158A τὰ ὁρώμενα τῆς ίδέας, the genitive is defining.

^{page 100 note 1} See C.R. XXVI. 241.

^{page 100 note 2} ‘The correction is easy enough,’ writes Mr. Gow of -ισμῶ. To me it seems most difficult and rash.

^{page 100 note 3} Cf. L. and S.⁹s.v.; Heyschius ἄριστον τὺν πρωινòν ἀκρατισμóν. ò ἡμεῖς λέγομεν ᾶριστον τοῦτο ὁ ποιητὴς καλεῖ δεῖπνον, τò δὲ δεῖπνον δóρπον; and see Athenaeus I. 11 B–E, who is right in his rule but wrong in his two Comic exceptions; e.g. the Cantharus (Kock i. 766) passage obviously means ‘don't let's breakfast here as we shall be lunching at the Isthmus.’ Unless you were an invalid or a glutton you did not necessarily take both breakfast and lunch; in the Antiphanes (Kock ii. 126) passage the host suggests that they should have a little snack as the luncheon is still only in process of being cooked. This sense of ἀκρατίσασθαι is given in the gloss of Photius s.v.: τò μικρòν ἐμφαγεῖν πρò τοῦ ἀρίστον.

^{page 100 note 4} Or the two variants may have originated together (as alternative solutions of a puzzling tradition) and been both so labelled then.

^{page 101 note 1} I have not been able to trace it further back and so to find out on what authority Voss attributed it to cod. Urs. This would evidently require a much wider and more regular access to books than is available to me.

^{page 101 note 2} According to the edition of Ahrens, which is here manifestly (from the words which follow) correct as against Wendel's.

^{page 102 note 1} Cf. the words quoted on p. 91.