¹ Class. Rev. 1967, p. 23.

² Unfortunately, the poetess' elegance was lost on the critics: for instance, Gow-Page (ad loc.) go as far as to say that the epithet ῥαδινήν ‘may be corrupt’!

³ On all this cf. ‘L'epigramma ellenistico’, in Introd. alla Cultura Classica, Milano 1972, p. 127.

⁴ The insect is ‘not dead, but captive’: so, correctly, Gow-Page in their introduction to the epigram.

⁵ Eust. 1817, 44 =Pollux VII, 23 Leonidas' (‘overcooked’) and (liter, ‘non coctus’, cf. Thes., s.v.) constitute a beautiful oxymoron. For a similar erotic metaphor involving cf. Luc., Dial. Deor. XIX, 1 (). That Leonidas should have used , an acceptation otherwise preserved by a lecicographer (Pausanias) is typical of the epigrammatists' diction: cf. e.g. Quad. Urbin. 1973, p. 19 f., 29; words or acceptations attested in the epigrammatists and otherwise only in lexicographers such as Hesychius are legion (cf. e.g. REG 1972, p. 62, n. 2).

⁶ For similar metaphors, in the mouth of a female speaker (as is the case with the speaker in Leonidas' epigram) cf. Herond. VI, 97 λαιμάττει (where the subject is Metro's ὗς, as I have indicated in Class. Rev. 1974, p. 35); the lady in Leonidas' epigram is ‘cooked’ by the same πολύ πῡρ καιόμενον mentioned by the girl in Fragm. Grenf, line 15 f. (cf. line 24, κατακαίομαι).

⁷ In Leonidas' sentence (‘I carry a fig which is those ’) and in similar sentences studied by Wifstrand (loc. cit.: e.g. A.P. V, 151, 4 ‘eat flesh which is my limbs’, Pol. I, 89, 1 ‘it eats flesh which is men’, Philo, De Migr. Abr. 144 carry a burden which consists in labours’) the accusative governed by the compound verb in -έω can be in the plural in that it is a predicate. Cf. A.P. V, 20, 3–4 (= Gow-Page, Garl. Phil. 2400 ff.), where means (cf. Waltz, ad loc.) ‘ripe beauty is fruits of season ( is predicate to on = ‘fruits of season’ cf. Gow-Page, Hell. Epigr., on line 1987) for Aphrodite's bed’. In Leonidas A.P. IX, 329, 4, = Gow-Page, Hell. Epigr. 1987 is exactly parallel to (i.e. is an example of the construction studied by Wifstrand, loc. cit. and not known to Gow-Page), and means ‘carries a present which is, consists in, the fruits of the season’. The plural which Buchheit (art. cit., p. 213 f.) has difficulty in explaining, is obviously part of Leonidas' metaphor: only the plural meant ‘crumbs of overcooked bread’.

⁸ On the ‘thème du rival’ in Hellenistic epigrams cf. Rev. Et. Gr. 1968, p. 51, n. 3.

⁹ Cf. also A.P. IX, 229, 5 f. (Gow-Page, Garl. Phil. 1427 ff.), where ἂμικτος is referred to a bottle containing pure wine and to a virgin bride. At Gow-Page, Hell. Epigr. 1839, since the μίτρη was traditionally called ἂχήρατος Hedylus jocularly refers to the μίτρε of a bibulous girl the epithet ξωρή which is a synonym of ἂχραντος, ἀκήρατος and which at the same time denounces the girl's propensity to drunkenness: a beautiful case of humorous metalepsis (Hedylus' ξωραīς μίτρησι is a poetic plural, cf. μίτραις in A.P. IX, 602, 8 = Gow-Page, Garl. Phil. 2317).

¹⁰ It is in character with Leonidas that the obscene humour of his epigram is based—until the final dénouement—on ambiguity: for his ‘jocularly ambiguous’ in A.P. VII, 657 (a mordant aside at the being notoriously ) cf. Class Rev. 1967, p. 22. Leonidas' humorous and ambiguous epigram is entirely based on traditional ingredients. The reader thinks at first that he is faced with a speaking fig-tree: speaking trees are common in the epigrammatic genre (cf. Gow-Page, Garl. Phil., vol. II, p. 103); vegetable metaphors applied to ladies (such as in Leonidas' epigram) are usual in epigrams (cf. A.P. V, 20); parody of sepulchral poetry (we have already noted that Leonidas' epigram opens as a parody of serious epitaphs) is not unknown to epigrammatists (cf. e.g. Gow-Page, Hell. Epigr., vol. II, p. 639: both Meleager's and Leonidas' parodies of sepulchral poetry are obscene).

¹¹ Cf. Rebmann, , Die Sprachl. Neuer, in den Kyneg. Oppians, p. 134 ffGoogle Scholar.

¹² For Archias cf. Reinach, , De Archia poeta, p. 38 ffGoogle Scholar. Cf. Quad. Urbin. 1973, p. 11; the most common form of cumulatio is of two epithets, cf. e.g. Gow-Page, Garl. Phil., lines 366, 380, 621 f., etc.

¹³ Cf. Reinach, op. cit., p. 39. It must be re membered that epigrammatists, in their constant search for the new, eagerly employed motion in order to obtain new adjectival forms. The motion -ής / -ος (cf. Quad. Urbin. 1973, p. 97, n. 41) gives birth e.g. to βαρύπενθος, βαθύκλεος, arbitrarily altered in Gow-Page, Garl. Phil., 2819, 3013. At A.P. IX, 551, 4 (= Gow-Page, Garl. Phil. 841 ff.: for this epigram cf. Quad. Urbin. 1973, p. 19 ff.) the adjective τεναγīτον, I should like now to add, is not to be altered into τεναγīτιν: the form τεναγīτον is the result of motion (for this type of motion cf. Quad. Urbin. 1973, p. 77) and a feminine cf. e.g. the feminine αἱμύλον in Garl. Phil. 1874).

¹⁴ The statue, that is, is standing at some point on one of the two halves of the breakwater, which latter is regarded by the poet as one single wall made up of two parts. The reader who is familiar with the ‘common theme’ of Priapus and Pan standing on the breakwater (cf. Gow-Page, in their introduction to Archias XXVII)will instantly recognise, and admire, the words ἱερῆς ἐπὶ δισσάδος αἰγιαλίτης (‘humid stone-wall consisting of two parts, on the shore’) as Archias' ingenious description of the breakwater, a description typical of Archias' ‘inventio’, achieved ‘in aenigmatis modum’ (on this trait of Archias' style cf. Reinach, op. cit., pp. 35–40; the epigrams A.P. X, 7 and 8 were indeed recognised to be a case of ‘Selbstvariation’ in Archias, as shown by the fact that they follow one another in the Anthology). I have already noted that supplying one noun (δισσάς) with two epithets (ἱερή and αἰγιαλίτη) is a characteristic feature of epigrammatic art.

¹⁵ Cf. e.g. Gow-Page, Garl. Phil., on lines 2601 or 3846.

¹⁶ Cf. Philost., H.E. Migne 65, 589B, Damasc. Pr. 400: it must be remembered that epigrammatists often used words in meanings which are for us otherwise attested in late prose (cf. below, note 22).

¹⁷ I need hardly add that Jacobs' emendation is palaeographically impeccable: confusion between λλ and λμ, -ων and -ov (abbreviated as supralinear̃ and \) is common in the minuscule, and it is well known that most corruptions in the Anthology are reading errors presupposing ‘an exemplar in minu scule’ (cf. e.g. Gow-Page, Garl. Phil., on line 1163). On genitive of appurtenance in epigrammatic poetry cf. Quad. Urbin. 1973, p. 21; αἰγιαλοῦ is genitive of appurtenance in A.P. X, 10, 4 =Garl. Phil. 3769.

¹⁸ It is interesting to see that Bast, loc. cit., quotes an example of confusion between ανταand -ακτα, analogous to the confusion between αντον and ακτον indicated by me.

¹⁹ Cf. e.g. Gow-Page, Garl. Phil., Index, s.v. ‘Death, caused by…’; Fohlen, Les circonstances de la mort dans les épitaphes grecques métriques, Mélanges Magnien, Toulouse 1949, p. 29 ff.

²⁰ Cf. ὒπερθε τάφου A.p. IX, 117, 2 ( = Gow-Page, Garl. Phil. 3828: also by Flaccus); ὐπερ τúμβου A.P. IX, 272, 2 (= Gow-Page, Garl. Phil. 1702).

²¹ I have underlined this point in my review of Gow-Page, The Garland of Philip, forthcoming in Class. Rev. The form χαιτάεις occurs in A.P. VI, 234, 1 =Gow-Page, Garl. Phil. 2256.

²² Often such prosaic meanings are attested for us in late prose. Cf. e.g. Gow-Page, Garl. Phil., on lines 26, 475, 749, etc.; a particularly instructive example in Theocritus XIV, 15 f. (ληνός = πίθος) I have indicated in Antiq. Class. 1968, p. 506, n. 37. Formulae of ἀποπομπή and ἐπιπομπή ‘attestate in epoca imperiale’ already occur in Leonidas, as was shown by Weinreich (cf. Gigante, , L'edera di Leonida, p. 50)Google Scholar.

²³ I.e. to employ the ‘unique and bold’ (Gow-Page, Garl. Phil., vol. II, p. 242), the ‘very rare’ (ibid., pp. 139, 175).

²⁴ On Crinagoras' employment of epic forms cf. Rubensohn, Crinagoras, p. 24. Epigrammatists, as is well known, often employ forms which are attested in later epic: e.g. Antiphilus' πολύτμητος (A.P. XI, 66, 1 =Gow-Page, Garl. Phil. 1095) occurs again only in Oppian, Philip's ἀρτίφυτος (A.P. IV, 2, 14 = Gow-Page, Garl. Phil. 2641) reappears in Nonnus; Antipater's κάθετος (‘fishing-line’, A.P. VII, 637, 2 = Gow-Page, Garl. Phil. 402) is found in Oppian, and Flaccus' παντοπαθής (A.P. V, 5, 4 = Gow-Page, Garl. Phil. 3799) occurs in Manetho; the form ἔκρυφεν is attested in A.P. VII, 700, 1 (=Gow-Page, Garl. Phil. 2148) and again in Quintus and Nonnus; κητοφόνος occurs in A.P. VI, 38, 3 (= Gow-Page, Garl. Phil. 2694) and elsewhere only in Oppian; χρεμέθω (A.P. IX, 295, 3 =Gow-Page, Garl. Phil. 1721) reappears only in Oppian; χλαινόω (A.P. IX, 293, 2 Gow-Page, Garl. Phil. 2960) is found elsewhere only in Nonnus.

²⁵ On this meaning of χθαμαλός cf. below, note 35. Just as Crinagoras says that Corinth, razed to ground, is γαίη, so Barboukallos states that Berytos, razed to the ground by an earthquake, is mere κύνις, i.e. flat soil (A.P. IX, 425, 4: κόνις is, in epigrams, a common synonym of γαίη, cf. Gow-Page, Garl. Phil. on line 2391; κόνις is χθαμαλή, cf. A.P. VII, 629, 1).

²⁶ No ‘higher than the levels on which they were built’ (Gow-Page, ad loc.).

²⁷ Fourth century A.D.: once more, a meaning attested in an epigrammatist re-emerges in late prose. Cf. Gow-Page, Garl. Phil., on line 1052 (ἐναυτολόγει).

²⁸ On oxymora cf. Waltz, , Antip., p. 47 f.Google Scholar

²⁹ Cf. ‘Gli epigrammi alessandrini come arte allusiva’, in Quad. Urbin. 1973, p. 7 ff., for a methodological treatment of this literary feature.

³⁰ The object of ἔβλυσαν πάρα is ὕδωρ, mentioned in the previous line. Παραβλύζω is otherwise attested in late prose: for such cases (frequent in epigram matic poetry) cf. e.g. Gow-Page, Garl. Phil., on lines 2719 (ζοφάω), 1052 (ναυτολογέω).

³¹ The correction suggested by Hecker (ἐξερέων) is singularly elegant, but in view of the excessive love of epigrammatists for ‘partiziplose Konstruktionen’ (cf. lastly Quad. Urbin. 1973, p. 31) I hesitate to alter the mss. reading ἐξ ὀρέων: if the reading is correct, the sense is ‘when he espied the animal (as he roamed) down ridges and bushy ravines (which proceed) down from mountains, whilst it was cooling its hooves and flanks on a river bank’.

³² No doubt because the shade afforded by trees protected the celebrants from the fierce southern European sun, which scorches and glares. In Anacr. LVII, 14 ff. Bergk means ‘who has spread her body on a bed of leaves protected by the shade’, as I have explained in detail in ‘On the Text of the Anacreontea’, forthcoming in Quad. Urbin.

³³ On the meaning under discussion of the adjective αὐτός cf. Gow-Page, Hell. Epigr., on line 2459 = A.P. VII, 731, 1). I take this opportunity of explaining αὐτῷ in A.P. VII, 731, 1, which has been misunderstood by Gow-Page. The old man who speaks in the epigram means that until recently he was able to obtain firm support (στηρίζομαι) from his legs (πόδεσσι Il. XXI, 241 f.) combined with his stick, i.e. his legs were still firm enough to offer reliable support if aided by the stick, but now his legs have become so weak that his stick alone, without his legs which had hitherto been the companions of the stick in creating firm support, can afford him any of the firm support he needs.

³⁴ This type of ‘Weglassung’ of the personal pronoun (in this case Weglassung of the dative αὐτῷ, ‘him’) is of course ‘sehr gewöhnlich’ (Kühner-Gerth, II, p. 562), especially in epigrams.

³⁵ On χθαμαλός = ‘terrae aequalis’, ‘complanatus’, ἰσόπεδον τῇ πεδιάδι cf. Thes., s.v. χθαμαλός.

³⁶ Gow-Page (ad loc.) note that the poet has used ἅρη (line 2) as a synonym of σίδαρον (line 3). It may be added that the employment of synonyms is frequent in Hellenistic poetry (cf. Class. Rev. 1971, p. 355): the feature reached its greatest development in Nonnus (Class. Rev., loc. cit.; Wifstrand, Von Kallim. zu Nonn., p. 154, n. 1). In A.P. IX, 343 (=Garl. Phil. 3738 ff.) Archias uses the three synonyms νεφέλη, θῶμιγξ and πάγη. Cf. Ouvré, Méléagre, p. 178.