¹ The following works are cited by the author's name alone: Müller, C. G., Isaaci et Joh. T'zetzae scholia in Lycophronem, cum notis variorum et scholiis minoribus nondum editis I–III (Leipzig, 1811)Google Scholar

Bachmann, L., Lycophronis Alexandra (Leipzig, 1830)Google Scholar

Scheer, E., Lycophronis Alexandra I–II (Berlin, 1881–1908)Google Scholar

Holzinger, C. von, Lykophron's Alexandra (Leipzig, 1895)Google Scholar

Mascialino, L., Lycophronis Alexandra (Leipzig, 1964)Google Scholar

(I have not been able to consult the edition of E. Ciaceri (1901).)

As Mascialino's edition neglects the papyri, which cover some of the passages to be discussed, a list may be convenient; all dates are A.D.:

P. Oxy. 17. 2094 (Pack² 1285; ii): 586–92, 924–39, 1345–79

P. Monac. inv. 156 (Pack² 1286; ed. Hartmann, pr., Philologus 76 (1920), 228–33CrossRefGoogle Scholar; i): 1108–28, 1156–63

PSI6. 724 (Pack² 1287; re-published by Vitelli, , Aegyptvs 3 (1922), 141 f.Google Scholar; iii?); commentary on 743–7

?P. Oxy. 27. 2463 (Pack² 2861: ii or iii): perhaps from a commentary on 326 ff.

The first two are discussed by Criscuolo, U., ‘Per la tradizione papiracea dell' Alexandra di Licofrone', Dioniso 44 (1970), 72Google Scholar ff. (with good photographs).

² The following characterization of Aeschylus' style brings out very well the features which clearly appealed to Lycophron, and indeed, mutatis mutandis, would serve very well as a description of Lycophron's own style: ‘For ancient and modern critics alike the most outstanding characteristic of Aeschylus’ style is undoubtedly its ἄγκοఁ, the luxuriance, boldness, exuberance, and grandiloquence, and often the apparent unintelligibility, that is effected partly by the diction itself - long compound words, many evidently of Aeschylus' own invention - and partly by the use of circumlocution and redundance' (Garvie, A. F., Aeschylus' Supplices: Play and Trilogy (Cambridge, 1969), p. 57)Google Scholar.

³ See further pp. 125 ff.

⁴ I have not been able to consult Walter, G., De Lycophrone Homeri imitatore (Basle, 1903)Google Scholar.

⁵ ‘Coiner of legends’ LSJ; the translations of Holzinger, Mair and Mooney agree in taking this to imply something more than a tendency to improve the tale in the telling (‘Und bricht in Jammer aus, der seinen Lügen hilft’, ‘Shall yelp out his fictitious tale of woe’, ‘And whining shall recount fictitious woes’).

⁶ Homerische Untersuchungen (Berlin, 1884), p. 193 n.Google Scholar; ‘Noteson Alexandra and Lithica’, JPh 20 (1892), 113 ffGoogle Scholar.

⁷ Holzinger's note ‘Der Vers ist als Uebergang zu einer neuen Partie aufzufassen; daher ist die Umstellung…abzulehnen’ explains nothing. Mascialino ignores the suggestion.

⁸ ‘Transposition in the Ibis of Ovid’, JPh 34 (1918) 222 ffGoogle Scholar. (= Classical Papers III pp. 969 ff.). He was not at a loss for an explanation: ‘One of the causes why any proposal to correct a verse or sentence alarms and distresses the natural man is that it makes an unusual demand upon his intellect and entails the weary work of reading and considering the context. That form of correction which consists of transferring a verse or sentence from one place to another is in consequence doubly discomposing, because the mental fatigue which it involves is twice as heavy. There are two contexts to be read and considered, not only one.’

⁹ RhM 34 (1879) 464: ‘Man hat die Spielerei der Stelle nicht verstanden: nicht von Poseidon, sondern von Odysseus ist die Rede. Mit ὕπεπ⋯οఁ λἠθηఁ deutet der Alexandriner auf δ⋯ τ⋯τε γ' ⋯τρ⋯μαఁ εὗδε λελαఁμ⋯νοఁ ὧν ⋯πεπ⋯νθει. M⋯λανθοఁ – dabei ist das homerische Mελανθε⋯;ఁ verwendet - ist ein passender Räthselname für de, von dem es ζ 231 heisst κ⋯δ δ⋯ κ⋯ρηιοఁ ⋯κε κ⋯μαఁ, ὑακινθ⋯ῳ ἄνθει ⋯μο⋯αఁ…ἱππηγ⋯τηఁ endlich konnte zu ähnlichem Zwecke sehr wohl der genannt werden, won dem es θ 494 heisst ἵττου…ὅν ποτ' ⋯ఁ ⋯κρ⋯πολιν δ⋯λῳ ἤγαγε δῖοఁ ‘Oδυఁఁε⋯ఁ vgl. λ 522’.

¹⁰ ‘Möglicherweise deutet aber Lycophr. gerade durch ⋯γκλ⋯νεఁθαι(= κ⋯πτεఁθαι), vgl. Hom. h. xxiii, 3: ⋯γκλιδ⋯ν ⋯ζομ⋯νῃ, auf die Erschaffung des Rosses Skyphios durch den schlafenden Poseidon. Vgl. Schol. Pind. Pyth. iv, 246’. The reference to h. Horn. 23. 3 does not support this interpretation.

¹¹ This detail perhaps throws a strange new light on Menelaus' Egyptian experiences (848–9).

¹² 113–43. Stesichorus' invention of the εἴδωλον was intended to exculpate Helen; Lycophron reduces the guilty couple's mutual enjoyment to a minimum, without any diminution of Helen's guilt. For a lucid account of the various versions of the legend see Dale, A. M., Euripides' Helen (Oxford, 1967), pp. xvii ffGoogle Scholar. ‘Lykophron zu gern kontaminiert’ observes Wilamowitz, (Hellenistische Dichtung II (Berlin, 1924), p. 160)Google Scholar; this is fair warning to those who would like to use the Alexandra to reconstruct lost works, but rather unfair as literary criticism.

¹³ Something also seems to be wrong with the transmitted text at 738–40: Tzetzes (on 740) rightly protests: κακ⋯ఁ δ⋯ κα⋯ ఁυγκεχυμ⋯νωఁ κα⋯ ⋯διαρθρώτωఁ ⋯ ∧υκ⋯πρων τ⋯ν περ⋯ Oδυఁఁ⋯ω ⋯ఁτορ⋯αν λ⋯γει. If Cassandra is supposed to be following the Odyssean order, 738 belongs before 662; there might be a connection between its apparent displacement and the disquieting combination of carelessness (π⋯ντα) and textual uncertainty in 664. The difficulty is not altogether met by supposing that Cassandra is rearranging the order of events, since 738 as it stands is simply an isolated detail, without any proper connection with the following narrative. Moreover, ఁυμπλεχθἠఁεται (740) is used in a very strange sense if it merely means that Odysseus was mildly scorched by the thunderbolt which destroyed his comrades: we expect it to mean that he shared their fate.

¹⁴ 892 αὐδ⋯ξει] ηὕδαξεν Wilamowitz (op. cit. II, p. 173 n. 1); 1350 καθιερώఁει] καθι⋯ρωఁε Scheer; Holzinger (on 845 and 1350) adds 890 ⋯νἠఁε7iota; to the list, but this is in order after δε⋯ξαντι, representing the tense which Triton used in giving his directions.

¹⁵ So LSJ, though I do not understand how the substantive was extracted from Scheer's text.

¹⁶ As, for instance, at 128, where the return to the main clause, after a long relative clause describing Proteus' unsatisfactory family-life, is marked by the resumptive κεῖν⋯ఁ ఁε.

¹⁷ On intrusive proper names supplied where the author used a circumlocution see Merkelbach, R., ‘Interpolierte Eigennamen’, ZPE 1 (1967), 100–2Google Scholar.

¹⁸ Problems of course arise when we try to relate it to the other evidence: see further Graf, F., ‘Die lokrischen Mädchen’, Studi Storico-Religiosi 2 (1978), 61 ff.Google Scholar, and Fontenrose, J., The Delphic oracle (1978), pp. 131–7Google Scholar.

¹⁹ T⋯μαιοఁ (FGrHist 566 F 146) ⋯ఁτορεῖ ὅτι αἱ παραγεν⋯μεναι παρθ⋯νοι ⋯δο⋯λευον ⋯ν τῷ ⋯ερῷ τ⋯ఁ ' Aθην⋯ఁ, β οὖఁαι. εἰ δ⋯ τιఁ ⋯ποθ⋯νaοι, ⋯τ⋯ραν παραγ⋯νεఁθαι ⋯νaτ' αὐτ⋯ఁ ⋯κε⋯νην δ⋯ οὐ θ⋯πτεఁθαι παρ⋯ τ⋯ν Tρώων, ⋯λλ⋯ κα⋯εఁθαι ⋯γρ⋯οιఁ ξ⋯λοιఁ, κα⋯ τ⋯ ⋯ఁτ⋯ αὐτ⋯ఁ ῥ⋯πτεఁθαι εἰఁ θ⋯λαఁఁαν.

²⁰ The Munich papyrus preserves the beginnings of 1156–63, but only the first letter of 1157, survives.

²¹ ‘Ebensowenig duldet der Stil πυτοῑఁ und ὅταν zu vertauschen’ noted Wilamowitz, (Die Ilias u. Homer² (Berlin, 1920), p. 387)Google Scholar. He himself, without much confidence, suggested something, like βαι⋯ν δ'.

²² There is probably another instance at Xen. Oec. 3. 2.

²³ Müller, whose notes reveal a good sense of Lycophron's style, interestingly observes that ‘hoc vers. plane omisso sensum reliquorum verborum sibi beneconstare, servato autem omnia turbari atque obscurari’. His feeling that the line would be better removed is significant, though with ὅτανa the clause lacks a genitive an d one seems desirable.

²⁴ Graf (loc. cit. 66) compares the Agrionia at Orchomenus, in which the priest of Dionysus, armed with a sword, pursued a group of women and might kill any he caught; but when a priest actually did kill someone, signs of divine displeasure followed (Plut, . Quaest. Gr. 38Google Scholar, 299e-f).

²⁵ ఁυνa;⋯βη μιᾷ ఁτελλομ⋯νῃ ⋯κ οκρ⋯δοఁ διαπθαρ⋯νaαι ⋯π⋯ τιοఁ λཹπου τ⋯ Tρο⋯αఁ καλουμ⋯νου Tρ⋯ρωνοఁ εἶτα τοὺఁ ∧οκρο⋯ఁ τ⋯ν μ⋯ν θ⋯ψαι, ఁιωπ⋯ఁαι, δ⋯ κα⋯ μηκ⋯τι π⋯μπεινa τ⋯ఁ παῖδαఁ π⋯ఁκονταఁ πεπληρ⋯ఁθαι τ⋯ν τ⋯ν χρ⋯νον ⋯καρπ⋯αఁ δ⋯ καταఁχο⋯τοὺఁ π⋯λιν π⋯μπειν οὐκ⋯τι δ⋯ο, ⋯λλ⋯, ⋯ρκοఁαν εἶνaαι δοκονταఁ τ⋯ τιμωρ⋯αν. ⋯ δ⋯ χρηఁμ⋯ఁ οὐκ εἶχεν ὡριఁμ⋯νον χρ⋯νον.

²⁶ The dual is used of the Maidens in the Vitrinitsa inscription; the conditions of service presupposed in it are less nightmarish than in Lycophron, and the girls go, and return, as a pair; JG xi. l. 706: 9/10 τοῖν]/κ⋯ραιν; 23 δικ⋯ζειν τοῑν κ⋯ραιν, ⋯πιδικηఁαι τοῖν πρ⋯ఁηεν.

²⁷ A. 11.252 ff. All that concerns me here is a similar line of thought, and it makes no difference to my argument whether Vergil had Lycophron in mind or not (though I believe he had). On reflections of the Alexandra in the Aeneid see appendix.

²⁸ Paean 6. 112 ff. (cf. the proverbial Nεοπτολ༑μειοఁ τ⋯ఁఁ (Paus. 4. 17. 4)); a different version in N. 7, better suited to Aeginetan patrons. Cf. Tryph. 640 ff.

²⁹ For a concise and lucid exposition of the problem see Fraser, P. M., Ptolemaic Alexandria II (Oxford, 1972), pp. 1065 ff.Google Scholar, n. 331. Ziegler's assertion (RE 13. 2365) that stylistically and metrically the passage is indistinguishable from the rest of the Alexandra needs some qualification. That it is metrically indistinguishable proves nothing: the austere monotony of Lycophron's hyper-aeschylean iambics is unmistakable and would greatly simplify an interpolator's task. Assessment of the style must be subjective; the passage is too short to allow any inference from quantifiable phenomena. That the passage seems at first reading easier than the rest of the poem might be accounted for in various ways, but is not simply explained by the fact that the general outline of the story is so familiar to us. The passage shews a relatively high proportion of phrases and patterns of line which occur earlier in the poem: 1234 ∼ 403; 1244 ∼ 823; 1249 ∼ 804; 1252 ∼ 978; 1257 ∼ 967; though Lycophron has other repeated hemistichs (e.g. ఁιθ⋯νοఁ ε⋯ఁ θυγατρ⋯ఁ (583, 1161), λοιβαῖఁι κυδανοఁι 929, 1213) and patterns of line, the concentration here seems suggestive, and might be thought a sign of an interpolator playing safe rather than of flagging invention on the poet's part. 1263–9 is a very clumsy sentence, and would be hard to follow with the rather economical punctuation normal when Lycophron composed; this kind of difficulty is not characteristic of his work in general. It may, incidentally, surprise us that Aeneas was not introduced earlier in the poem if he was an important element in Lycophron's plan. His departure from Troy was a well-known episode in the tale, and his universally acknowledged piety would have formed an effective contrast to the sacrilege of Neoptolemus (335 ff.) and Ajax (348 ff.); an earlier mention would have avoided the need for the clumsy recapitulation of 1263 ff. The impracticability of settling the authenticity of short passages on stylistic grounds alone is well illustrated by the controversy surrounding the Helen, -episode in Aeneid 2 (567–88)Google Scholar; since the magisterial discussion by G. P. Goold (HSCP 74 (1970), 101 ff.) it is hard to believe that anyone might ever again argue that the passage could be by Vergil, but Goold's demonstration rests on a painstaking examination of the circumstances of its transmission, not on consideration of those stylistic niceties which had left those best qualified to judge deadlocked. Here we can probably say with confidence that the passage is not up to the standard of Lycophron's best work, and since, if we are dealing with an interpolator, he was clearly a competent composer, this is about the most the circumstances allow.

³⁰ λε⋯πει δ⋯ ⋯ κα⋯ ఁ⋯νδεఁμοఁ, ἵν' η “κα⋯ κ⋯δροι”. δηλοῖ δ⋯ τοὺఁ ' Aθηνα⋯ουఁ ⋯π⋯ κ⋯δρου βαఁ⋯λεωఁ 'Aθν⋯ν. μ⋯τεఁχον δ⋯ 'Aθηναῖοι τ⋯ఁ ⋯ποικ⋯ఁ τ⋯ν 'Hρακλειδ⋯ν.

³¹ cf. Eust. ad Od. 1. 58, p. 1391, 37.

³² op. cit. II 145.

³³ The very full discussions by Ziegler, , RE 13. 2354 ffGoogle Scholar, and Josifović, , RE Suppl. 11. 888 ff.Google Scholar, summarize earlier literature; see also Hurst, A., ‘Sur la date de Lycophron’, Mélanges P. Collart (1976), pp. 231–5Google Scholar.

³⁴ We might expect Lycophron to follow Herodotus' declared principle of reckoning three generations to a century (2. 142), though it must be admitted that Herodotus himself is inconsistent in his practice: he allows 505 years for 22 generations of Lydian kings (1. 7), i.e. a generation of 23 years. No ancient historian appears to have used a generation shorter than 23 years: see further Samuel, A. E., Greek and Roman Chronology (Munich, 1972), pp. 241 ffGoogle Scholar.

³⁵ ἄντιτοఁ, παλ⋯ντιτοఁ, and the lexicographically rather dubious τιτ⋯ఁ (as in Il. 24. 213 τ⋯τ'ἂν τιτ⋯ ἔργα γ⋯νοιτο) would be particularly useful. Josifović's attempt to press Tιτοఁ (941) into service in this way (op. cit. 928) is most unconvincing, given the difference in quantity and the very unsuitable context.

³⁶ op. cit. II 146; he had advocated a different (but over-complicated) approach forty years earlier (De Lycophronis Alexandra commentatiuncula (Greifswald, 1883) = KS 2. 12 ff.)Google Scholar. Wilamowitz's theory has often been dismissed as a curious aberration, but is revived as the least unsatisfactory explanation, with a marked lack of enthusiasm, by Lesky, Albin (Gesch. griech. Lit.³ (Bern and Munich, 1971), p. 835)Google Scholar.

³⁷ ‘Zu bedenken ist auch das orphische Rätsel ἕκτηι δ' ⋯ν γενε⋯ι καταπα⋯ఁατε οἷμον ⋯⋯οɩδс (Plato Phlb. 66c [fr. 14 Kern])'. It is of course true, and relevant, that seven is a typical number such as storytellers and prophets favour; when such numbers occur in reality (or when reality may be interpreted according to such a pattern) they impress themselves on the imagination. See further Fehling, D., Die Quellenangaben bei Herodot (Berlin, 1971),pp. 154ffCrossRefGoogle Scholar., Blom, J. W. S., De typische getallen bij Homeros en Herodotos (Nimwegen, 1936),Google ScholarDreizehnter, A., Die rhetorische Zahl (Zetemata 73, Munich, 1978)Google Scholar.

³⁸ If indeed it is sound; Ziegler (op. cit. 2370–1) rightly draws attention to its oddity and questions the conventional translation of εἶс as ‘peerless, unique’.

³⁹ e.g. Livy 1.19. 3; ResGestae 13; Suet. Aug. 22. See more fully Momigliano, ‘Terra marique’, JRS 32 (1942), 53–64 (esp. 62 ff.).

⁴⁰ Hor. Sal. 2. 5. 62–4.

⁴¹ 31. 1. 9; 5. 5; 7. 6; 14. 6ff., etc.

⁴² 31. 14. 6: contraxerant autem sibi cum Philippo bellum Athenienses haudquaquam digna causa, dum ex vetere fortuna nihil praeter animos servant; 44. 9: Athenienses quidem litteris verbisque, quibus solis valent, bellum adversus Philippum gerebant.

⁴³ It is a long shot, but 1345 ⋯λκῇ νέανδροс ⋯κπρεπέϲτατοϲ уένουϲ looks tailor-made for Alexander the Great; in its present position it describes a not certainly identifiable Trojan king. The coinage νέανδροϲ seems to hint at something, but no-one appears to know what. If it stood after 1441, ⋯λκῇ νέανδροϲ would seem to point unmistakably to νέοϲ ' Αλέξανδροϲ, a new and superior version of Paris. Such a play on names is eminently Lycophronic, the outstanding example being Cassandra's own designation as Alexandra; there was good Aeschylean precedent: see further Kranz, W., Stasimon (Berlin, 1933), pp. 287 ff.Google Scholar, Schmid-Stählin, , Gesch. griech. Lit. i 2 (Berlin, 1934), p. 297 n. 3Google Scholar. Accidental displacement over so great a distance is unlikely, though not impossible. Perhaps our interpolator thought the line too good to sacrifice completely and moved it to a position where it was harmless but deprived of its real point.

⁴⁴ 1446 ~ 1229, 1449 ~ 1270–1, cf. 1265; see also Ziegler, op. cit. 2371.

⁴⁵ It seems certain that Vergil had read Lycophron: see below, appendix. There is perhaps an allusion in Propertius (4. 1.51 f.); Theon's commentary belongs to this period. But the most telling piece of evidence comes from Statius' elegy on his schoolmaster father (Silv. 5. 3. 156–8): ‘tu pandere doctus / carmina Battiadae latebrasque Lycophronis atri / Sophronaque implicitum tenuisque arcana Corinnae’. This sounds like occasional fare for particularly able boys, as opposed to the core curriculum of Homer, Hesiod, Theocritus and the lyric poets (148–56), but, at all events, Lycophron had by now attained the status of a set book; the pairing with Callimachus is significant. Plainly, we are lucky to have testimony as explicit as this, and a sceptic might wonder whether the absence of evidence of interest in Lycophron during the century before Vergil really indicates that his work was neglected. The lack of Ptolemaic papyri proves nothing, since the chora was slow to acquire a taste for Hellenistic poetry. But Catullus would surely have found the Alexandra interesting; he liked difficult poetry, and we might expect to detect some allusion in 64 (which is, in some ways, very similar in spirit) or in 68, had he known Lycophron's work. The long servitude of the Locrian Maidens would have provided Lucretius with a terrible illustration of the folly and suffering engendered by conventional religious belief. This survey of Republican poetry which might have been otherwise does not prove very much, but it seems safe to infer at least that Lycophron was not known at Rome to Catullus' generation.

⁴⁶ See further Speyer, W., Die literarische Fälschung im heidnischen und christlichen Alter turn (Munich, 1971), especially pp. 32 ff., 111 ffGoogle Scholar.

⁴⁷ The editor was Edward Wells, and his text, first published in 1704, extends to a generous 1459 lines (against the conventional 1187); since his work achieved a sixth edition (1761), he was presumably right in supposing it to be needed.

⁴⁸ Another line of defence against awkward questions about Lycophron's prophetic gifts is suggested by the poem's reference (807 ff.) to the ancient belief that dying men have second sight (cf. II. 22. 356 ff., Genesis 49. 1–27; Stith Thompson, Motif-Index of Folk-Literature, D. 1715; 1812. 2. 4); perhaps in articulo mortis Lycophron foresaw the glories of the pax Augusta and his dying words were faithfully recorded by his sorrowing relatives (for a second edition?).

⁴⁹ Suet, . Aug. 31Google Scholar.

⁵⁰ Similarly a Sibylline oracle turned up conveniently when Augustus was preparing to celebrate ludi saeculares in 17 B.C.; see Fraenkel, , Horace, pp. 365 ff.Google Scholar, text in Kiessling-Heinze on Carmen Saeculare.

⁵¹ On Theon see Guhl, C., Die Fragmente des alexandrinischen Grammatikers Theon (Diss. Hamburg, 1969)Google Scholar.

⁵² Two examples should suffice. (1) Aristarchus' commentary on Herodotus 1 (P. Amherst 2. 12 = Pack² 483; reprinted by Paap, A. H. R. E., De Herodoti reliquiis inpapyris et membranis Aegyptiis servatis (Leiden, 1948), pp. 37 ffGoogle Scholar.; third century A.D.): the subscription is preserved 'Αρɩϲτάρϰου / 'Ηροδότου / ᾱ / ὑπόμνημα. The editors comment: ‘Unless the papyrus gives only a series of excerpts from Aristarchus’ commentary, which is not very likely, that work must have been extremely brief, for in the second column, which is fairly well preserved but has only sixteen lines, as many as twenty-two chapters are disposed of, there being no notes on chapters 195–214'. Before the discovery of the papyrus there was no reason to suppose that Aristarchus had commented on Herodotus at all, and the meagreness of the material preserved here suggests that the work simply comprised observations on passages which happened to interest him without any attempt at systematic coverage; ὑπόμνημα covers a wide range of literary production, from rough jottings to the history of Polybius, and such an interpretation would not imply misleading labelling. (Pfeiffer, (History of Classical Scholarship (1968), pp. 224–5)Google Scholar takes a different view, but there is no more reason now than there was sixty years ago to suppose that the papyrus merely contains extracts or was based on a defective copy.) (2) Didymus on Demosthenes' Philippics (BKT 1 = Pack² 339; second century A.D.): the subscription runs Δɩδύμου περì Δημοϲθένουс κη Φɩλɩππɩκ⋯ν Ῡ. This appears slapdash and ill digested, and its deficiencies cannot be explained away by the hypothesis that the papyrus merely preserves extracts or that this was a monograph designed to supplement a proper commentary; particularly worrying is the cursory discussion of the authenticity of Dem. 11 (col. 11, 7 ff.). See further CQ n.s. 20 (1970), 288 ff.

⁵³ Sch. ad 1226: ⋯ντε⋯θεν περ⋯ 'Ρωμαίων λέуεɩ κα⋯ Λυκόϕρονοϲ ⋯τέρου νομɩсτέον εἶναɩ τ⋯ ποίημα, οὐ το⋯ уράΨαντοс τ⋯ν τραуῳδίαν сυνήθηϲ у⋯ρ ὢν τῷ Φɩλαδέλϕῳ οὐκ ἂν περì 'Ρωμαίων δɩελέуετο. To suggest at this point that the poem as a whole might be by someone other than the famous third-century Lycophron seems extraordinarily unmethodical. This possibility, if it is worth taking seriously, ought to have been mentioned in the introduction. The reference to a homonym implies a rather blinkered view of the possibilities of false attribution. Interpolation would be the obvious explanation for the difficulty, and an ancient scholar might have been expected to suggest athetesis, or at least to explain why he thought such a remedy unsatisfactory here. But this note cannot possibly be so construed. The hypothesis of a homonym is irrelevant to interpolation, and it is normal, and necessary, in discussing interpolation to specify the length of the suspect passage. These two difficulties might, I suppose, be met by (1) excising Λυκόϕρονοϲ as an unintelligent insertion, (2) taking the commentator to mean that everything after 1226 (⋯ντε⋯θεν) is spurious; τò ποίημα would then refer to the last 250 lines, and τ⋯ν τραуῳδίαν would mean the Alexandra (which could no doubt be called a tragedy in much the same way as we might call Watership Down an epic). (1) is easy enough, but not (2); ancient critics were often drastic in athetesis, but I cannot imagine how anyone could have thought 1412–34 or 1451–74 were by any hand but Lycophron's. Whatever point was originally made here has surely been muddled in transmission. It should be noted that Tzetzes seems to have misunderstood το⋯ уράψαντοϲ τ⋯ν τραуῳδίαν(‘the writer of tragedy’ as opposed to ‘another Lycophron’), since he protests το⋯το δ' οὐ δ⋯ναμαɩ νο⋯сαɩ π⋯с οὐκ ἔϲτɩ το⋯ уράψαντοϲ αὐτό οῠτω у⋯ρ ⋯ϕεɩλεν εἰπεῖν εἰπεῖν οὐκ ἔϲτɩ το⋯ λεуομ⋯νου уράϕων αὐτ⋯ αὐτ⋯ ⋯λλ' ⋯τ⋯ρου. His misconception suggests a possible source of confusion for his predecessors. This scholium could probably be made to support any view of the passage we care to take, but does not provide a satisfactory basis for further argument. The note may be made to look slightly more sensible by the insertion of οὕτω after οὐκ ἂν as Scheer suggested, but superficial improvement is probably mistaken kindness.

⁵⁴ ‘Zur Kenntniss Lykophrons’, Hermes 26 (1891), 567 ffGoogle Scholar. (a remarkably perceptive article, though apparently animated by an altruistic determination to prevent others wasting time on this author). I quote his concluding paragraph: ‘Von dem Geiste der Melancholic der noch einem Wakefield aus den eintönigen Versen der Alexandra herausklang, empfinden wir nichts mehr. Eher glaubt man, wenn man dem Dichter durch seine verschlungen Irrwege gefolgt ist, glücklich den Sinn erkannt zu haben wähnt und schliesslich sich doch so oft genarrt sieht, das schadenfrohe Lachen eines neckischen Koboldes zu hören’.

⁵⁵ There is a useful, though necessarily brief, discussion of Vergil's debt to Lycophron by Trencsènyi-Waldapfel, I. in ‘Das Bild der Zukunft in der Aeneis’, Studii Clasice 3 (1961), 281–304Google Scholar.

⁵⁶ op. cit. 922 ff.

⁵⁷ See further Hensel, L., Weissagungen in der alexandrinischen Poesie (Giessen, 1908)Google Scholar.

⁵⁸ Scaliger, whose remarkable translation of the Alexandra into iambic senarii may most conveniently be found in Bachmann's edition, actually uses the Vergilian phrase here.

⁵⁹ The parallel is noted by Norden ad loc. as ‘auffallend ähnlich in der formellen Ausdrucks weise’; he draws attention to other parallels in A. 6, though without committing himself as to r their significance (e.g. 88 Dorica castra ~ Al. 284 Δωρɩεὺϲ ϲτρατόс 94 exterbu thalami ~ A1 60 λέκτρων θ' ἕκατɩ τ⋯ν τ'⋯πεɩсάκτων уάμων). There may be a reflection of Lycophron in the Sibyl's presentation of the impending war in Latium as a kind of replay of the Trojan War, i with Aeneas cast for the uncomfortable role of Paris (88 ff.); as with Lycophron, proper names; serve for mystification (‘non Simois tibi nee Xanthus nee Dorica castra / defuerint; alius Latioiam partus Achilles’). But it is difficult to be sure whether all this is intended to evoke Lycophron in particular or the language of oracles in general.

⁶⁰ On Neoptolemus see above pp. 121 ff. ‘Libycone habitantis litore Locros’ (265) was found puzzling in antiquity, to judge from Servius' comments; it is perhaps worth considering whether it was suggested by Lycophron's account of Thessalian settlement in North Africa (Al. 876–908).

⁶¹ His accurate ornithological observation is commended byBenton, S., ‘Two notes’, CQ n.s. 10 (1960), 110–11CrossRefGoogle Scholar, who identifies the birds as shearwaters (Procellaria diomedea).

⁶² FGrHist 566F 56. We learn from Strabo (6. 3: 9) that there was a similar practice at the neighbouring (associated?) shrine of Calchas: δείκνυταɩ δ⋯ τ⋯ϲ Δαυνίαϲ περì λόϕον, ᾧ ⋯νομα Δρíον, ⋯ρῷα, τò Κάλϰαντοϲ ⋯π' ἄκρᾳ τῇ κορυϕῇ ⋯ναуίζουϲɩ δ' αὐτῷ μέλανα κρɩòν οἱ μαντενόμενοɩ, ⋯уκοɩμώμενοɩ ⋯ν τῷ δέρματɩ τ⋯ δ⋯Ποδαλεɩρίου κάτω πρ⋯ϲ τῇ ῥίζῃ, δɩέϰον τ⋯ϲ τ⋯ϲθαλάτηϲ ⋯ϲον ϲταδίονϲ ⋯κατόν. The same ritual was observed at the sanctuary of Amphiaraus (Paus. 1. 34. 5).

⁶³ Fordyce ad loc.

⁶⁴ But not necessarily towards the parallel passage. I have argued above that the section on Aeneas is almost certainly an Augustan interpolation, and it may not have been known to Vergil. This uncertainty does not affect my argument.

⁶⁵ Dionysius (1. 55. 4) says that the prophecy was variously ascribed to Dodona and to the Erythraean Sibyl.

⁶⁶ cf. Conington's note on 7. 123.

⁶⁷ Josifovic (922) sees in this passage a specific allusion to the Alexandra (i.e. to 1226 ff.): ‘Wo hatte jedoch Kassandra ausführlicher die Irrfahrten des Aeneas und die zukünftige Grösse der Römer prophezeit als in der Alexandra?’ But 183 does not encourage this interpretation; the Alexandra is a report to Priam by his servant, not a confidential communication to Anchises.

⁶⁸ The other references in the Aeneid to prophecy by Cassandra are these: 2. 246–7: she warns the Trojans against the Wooden Horse: a conventional motif underlining the critical moment in Troy's history; 5. 636 ff.: Juno, disguised as the Trojan Beroe, claims that Cassandra in a dream urged her to burn the ships; this bogus appeal to Cassandra's authority enjoys a success quite unprecedented in her earthly career; 10. 67–8: Juno again unscrupulously exploits Cassandra's reputation, to cast aspersions on Aeneas' common sense: ‘Italiam petiit fatis auctoribus (esto) / Cassandrae impulsus furiis’; she overlooks the fact that Cassandra's predictions must pass unheeded until it is too late for them to matter.