¹ Correspondence of the late Gilbert Wakefield, B.A., with the late right honourable Charles James Fox, in the years 1796–1801, chiefly on subjects of classical literature (London 1813)Google Scholar = Memorials and Correspondence of Charles James Fox, ed. Russell, Lord John, iv (London 1857) 296–440Google Scholar, Letter xxxi; the following quotations come from Letters xxxix (27 May 1800) and xlvi (26 Jan. 1801). The parts of this correspondence relating to Lycophron were translated into German and published by Doederlein under the title ‘Über Lycophron’ in RhM iii (1829) 465–73Google Scholar.

² We cannot claim Fox as the first modern reader known to have been troubled by the problem, since he appears to have been anticipated by an unidentified scholar whose comments on 1229 and 1281, recorded ‘manu satis antiqua’ in a copy of Potter's first edition in the Leipzig Ratsbibliothek, are reported by Bachmann.

³ AJP xxii (1901) 344Google Scholar, = Selections from the Brief Mention of B.L. Gildersleeve (Baltimore 1930) 69Google Scholar.

⁴ Niebuhr had been anticipated by another aristocratic amateur, Viscount Royston, in the introduction to his spirited blank-verse rendering, published when he was only twenty-two: Cassandra, translated from the original Greek of Lycophron and illustrated with notes (Cambridge 1806)Google Scholar.

⁵ The controversy is lucidly surveyed by Ziegler (2354 ff.), whose account is brought up-to-date (to 1968) by Josifović. The following contributions should also be noted: Fraser 1972 ii 1065–7 n. 331 (analyst), Hurst, A., ‘Sur la date de Lycophron’, Mélanges P. Collart (Lausanne/Paris 1976) 231–5Google Scholar (conservative unitarian), Fraser 1979 (radical unitarian).

⁶ I shall use this phrase as a collective for 1226–80, 1446–50; though some have doubted whether the second passage in fact relates to Roman affairs, recent discussions indicate that this point is no longer regarded as controversial.

⁷ Hermes xxvi (1891) 579Google Scholar, cf. Wilamowitz 1924 152.

⁸ Holzinger's remarks on the poet's fading enthusiasm should be compared (Einl. 26).

⁹ P. M. Fraser has recently argued (Fraser 1979) that the Alexandra is substantially indebted to Eratosthenes for information on Cyprus; since Eratosthenes' Geography was most probably composed between 240 and 210, and Lycophron the tragedian is unlikely to have been less than seventy (if not already dead) in 240, this conclusion would provide valuable support for the radical unitarian case. The argument (to which I cannot do justice in a footnote) depends heavily on what seems to me a very questionable interpretation of the opening sentence of the scholium to 447 (ε ´δέ φησιν εἰς Κύπρον ἀπενεχθη̑ναι Τευ̑κρον, Πράξανδρον καὶ Κηφέα), taking the subject of ϕησιν to be not Lycophron but Eratosthenes, who is mentioned in the last sentence of the immediately preceding note on Μάγαρσος (444) (μέμνηται δὲ αὐτου̑ καὶ ᾿Ερατοσθένης ). Even if this reading of the scholium is possible, the move from the sources of the scholia to the poet's sources is tricky. Since the Alexandra does not display very extensive knowledge of Cyprus, it seems easier to suppose that its Cypriot details derive from the accumulated store of a magpie mind rather than from consultation of an up-to-date work of reference; thus Lycophron may be supposed to have known about the Satrachus, as most of us do, because of its associations with Myrrha/Smyrna (Catul. 95), whose incestuous love-affair was related by Panyassis (fr. 25 K) and Antimachus (fr. 102 Wyss).

¹⁰ Momigliano 58.

¹¹ εὐδοκίμει δὲ τότε ὁ Λυκόφρων οὐ τοσου̑τον διὰ τὴν ποίησιν ὅσον διὰ τὸ λέγειν ἀναγραμματισμοὺς οἱ̑ ον ὅτι Πτολεμαι̑ος ἀπὸ μέλιτος λέγει μετα- γραμματιζόμενον, ᾿Αρσινόης δὲ ἴον ῾᾿Ηρας καὶ ἔτερα τοιαυ̑τα τούτοις ὅμοια (Tzetzes). The fact that this detail was remembered suggests there might be more to it than meets the eye: could there be a connection between the second anagram and Callimachus' use of the epithet ἰόζωνος for Arsinoe in the Coma Berenices (fr. 110.54)?

¹² ‘Si respicias quot res et vocabula Lycophro et Callimachus ex iisdem fontibus prompserint, Alexandram potius tertio saeculo tribuas quam altero’ wrote Pfeiffer, , Callimachus ii (Oxford 1953) xliiiGoogle Scholar. The passages selected for inclusion in his index s.v. ‘Lycophro’ represent only a fraction of the relevant material as revealed by his notes; see also History of Classical Scholarship i (Oxford 1968) 120Google Scholar.

¹³ οὐκ ἄν <οὕτω>? Scheer.

¹⁴ Against Ziegler who claims him as a radical Unitarian (2355), see Fraser 1979 341–2; neither interpretation is quite satisfactory, and the note has evidently suffered some garbling in the course of transmission. As it plainly represents inference, not independent testimony, its ambiguities do not directly affect my argument. But since the younger Lycophron conjured up here has been taken very seriously by the radical unitarians, it should be stressed that such hypothetical homonyms are a regular expedient of ancient scholarship. Perhaps the best known example is Didymus' second Antiphon (ap. Hermog. de ideis ii 11.7 DK 87 A 2); one of the earliest instances is Herodotus' attempt to avoid certain difficulties arising from the over-generous application of interpretatio Graeca by postulating a second Heracles (ii 43). See further Heitsch, E., AAWM 1972, xi 6–16Google Scholar.

¹⁵ Wilamowitz 1883 10.

¹⁶ I have concentrated on Momigliano's presentation of the conservative case because of the influence which his exposition has enjoyed.

¹⁷ As Fox and the anonymous critic ap. Bachmann (n. 2 above) noted (though the scholiast fails to remark on it).

¹⁸ Thus (though they take very different views of the relevant historical circumstances) Sudhaus, , RhM lxiii (1908) 487Google Scholar n. 2, Christ-Schmid, , Gesch. gr. Lit.⁶ (Munich 1920) 170Google Scholar.

¹⁹ Scheer suggested that something was missing (though he held 1226–80 to be interpolated); this hypothesis has not, so far as I know, appealed to any unitarians. But certainly this section of Lycophron's nostoi (1090–1280), dealing with those who returned to domestic tragedy, is short of candidates. We might wonder why Lycophron did not include Neoptolemus, whose childless marriage to Hermione sent him on the journey to Delphi which ended in his death. Cassandra does not pass over his dreadful part in the sack of Troy (335 ff.), and his death beside Apollo's altar might well be regarded as divine retribution for his slaughter of Priam at the altar of Zeus Herkeios (cf. Pind. Pae. 6. 112–20); this exemplary punishment of an appalling act of sacrilege would have made an impressive end to the tale of Greek misfortune. But what inference we should draw from this omission is another matter.

²⁰ Fortunately we do not depend on the Alexandra alone for the story: cf. X. Cyn. i 15 (probably spurious: see RE s.v. 1913–14), D.S. vii 4, Varr. Res Hum. ap. Serv. Dan. ad Aen. ii 636, Historiae 2 ap. Schol. Ver. ad Aen. ii 717Google Scholar, Ael. VH iii 22. Here two separate actions are conflated: the rescue of Anchises preceded, and made possible, the rescue of the Penates. When the Greeks offered Aeneas, like the other surviving Trojans, the chance to take away something from the sacked city, he chose to save his father; the Greeks were so impressed by this evidence of εὐσέβεια. that they granted him alone (τούτῳ μόνῳ 1268) a second choice, which enabled him to rescue his household gods; he was then allowed to take what he would and leave unmolested. Holzinger well observes ‘Lykophron erzählt weder, warum die erste Wahl gestattet wurde, noch auch, dass die Rettung des Vaters und der Penaten zwei Acte waren, noch auch, dass dem Aineias ein drittes Zugestandnis gemacht wurde, in Folge dessen er seine ganze werthvolle Habe mitnahm’.

²¹ ‘Stilistisch zeigen sie keinerlei Abweichung gegen das Übrige, ebensowenig metrisch’ (2365); rather oddly he offers in support of this claim some statistics bearing on the distribution of third- and fourth-foot caesuras, a feature of the tragic trimeter not normally thought to be of particular significance.

²² Not only the analysts Scheer and Schmidt, but also the unitarian Holzinger, (Einl. 26)Google Scholar.

²³ It also appears to me to offer a relatively large number of phrases and patterns of line occurring earlier in the poem (1234 ∼ 403, 1244 ∼ 823, 1249 ∼ 804, 1252 ∼ 978, 1257 ∼ 967), but this impression may merely reflect a greater alertness on my part to such phenomena when they relate to this passage.

²⁴ See further Perret 345–66.

²⁵ What we are here told about Tarchon and Tyrsenus (1245–9) is hard to reconcile with what is later said about Etruscan origins (1351–61), following Herodotus (i 94) and Timaeus (FGrH 566 F 62).

²⁶ See further Perret 637–41.

²⁷ Not the least noteworthy feature of this passage the absence of any reference to Timaeus in the scholia.

²⁸ The only evidence that Stesichorus told of Aeneas' journey to the West comes from the Tabula Iliaca Capitolina (PMG fr. 205; c. 15 BC), which shews Aeneas σὺν τοῖς ἰδίοις ἀπαίρων εἰς τὴν ῾Εσπερίαν , alleged following the ᾿Ιλίου Πέρσις κατὰ Στησίχορον. Some scholars have found this hard to believe: see further Horsfall, N., ‘Stesichorus at Bovillae?’, JHS xcix (1979) 26–48CrossRefGoogle Scholar, who well says (43) ‘There is, moreover, an argumentum ex silentio, well-worn but still powerful, to reinforce scepticism. Dionysius of Halicarnassus knew his Stesichorus, scoured the sources for references, however obscure, to the legend of Aeneas in the West, and did not come up with Stesichorus' Iliou Persis. Had he not known the poet and had he not read so widely in both prose and poetry for the material in Antiquitates Romanae i, then the argument might seem feeble; as it is his silence commands our respect and attention.’ See also Castagnoli, F., Studi Romani xxx (1982) 7–8Google Scholar. At all events it cannot be supposed that the story would have been well known to Lycophron's original readers.

²⁹ Besides Perret, see Horsfall, N., ‘Some problems in the Aeneas legend’, CQ xxix (1979) 372–90CrossRefGoogle Scholar.

³⁰ ἐμός τις σύγγονος 1232 ~ αὐθαίμων ἐμός 1446; γῆς καὶ θαλάσσης σκῆπτρα καὶ μοναρχίαν /λαβόντες 1229–30 ~συμβαλὼν ἀλκὴν δορὸς/πόντου τε καὶ γῆς 1446–7; τῷ καὶ παρ᾿ ἐχθροῖς εὐσεβέστατος κριθείς 1270 ~ πρέσβιστος ἐν φίλοισιν ὑμνηθήσεται 1449. These resemblances do not of course argue a single author for both passages, though almost all scholars have supposed that they come from the same hand.

³¹ Wilamowitz 1883 suggested that the generations might, in effect, be reckoned backwards: taking μεθ᾿ ἓκτην γένναν closely with αὐθαίμων ἐμός, ‘mini post sex generationes cognatus’, he identifies Cassandra's kinsman as Artabazus, her fifth cousin about thirty times removed (their common ancestor being Zeus, father of both Dardanus and Perseus) and grandfather of the Heracles to whose death Lycophron earlier alludes (801–3). This interpretation allows Cassandra's predictions to terminate with Alexander, but is more ingenious than plausible; not surprisingly Wilamowitz himself thought better of it.

³² See further Roscher, W. H., Die Sieben- u. Neunzahl im Kultus u. Mythus der Griechen, ASAW liii (1906) 1Google Scholar, Fehling, D., Die Quellenangaben bei Herodot (Berlin 1971) 154–67CrossRefGoogle Scholar.

³³ Momigliano does not ignore this difficulty, but, I think, underestimates its seriousness (59): ‘What the sixth ‘γέννα’ means, I do not know: neither does anyone else. It may refer to the Macedonians, and in this case allude to the six kings after Alexander (Alexander, Philippus Arrhidaeus, Cassander, his sons, Demetrius Poliorcetes, Pyrrhus), or it may refer to the Romans … Timaeus may have spoken of the Roman-Etruscan reckoning of human life-spans, and Lycophron may have been tempted to use it more or less exactly. Where much is obscure, and the possibilities are many, further discussion is useless’. But what did Lycophron expect his readers to make of the phrase?

³⁴ Many readers must have been struck by the thought that the poem would have been incomprehensible if it had not been provided by its author with some explanatory notes, and certainly we ought to wonder about the sources for much of the curious information confidently retailed by the scholia. We may note that in POxy 2528 Euphorion is perhaps the interpreter of his own poems: see further Pfeiffer, , Hist. Cl. Schol. i 150Google Scholar n. 5.

³⁵ Derived mainly from Welcker, (Die griech. Tragödien iii [Bonn 1841] 1259 ff.Google Scholar) and Cauer, (RhM xli [1886] 396–7Google Scholar).

³⁶ On 1253.

³⁷ Cassandra's historical knowledge is another aspect of her prophetic powers, hindsight as well as foresight being part of the ancient conception of a prophet; thus Calchas (Il. i 70) ᾔδη τά τ᾿ ἐόντα τά τ᾿ ἐσσόμενα πρό τ᾿ ἐόντα, and Cassandra's Aeschylean counterpart establishes her reliability by alluding to the earlier misfortunes of the house of Atreus before turning to the future (Ag. 1087 ff.).

³⁸ On Wilamowitz's interpretation see above, n. 31.

³⁹ Wilamowitz 1924 146. This reading of the passage, earlier suggested by Cessi, (SIFC xix [1912] 75)Google Scholar and commended by Lesky as the least unsatisfactory of the available options (Gesch. gr. Lit.³ [Bern 1971] 835Google Scholar), comes very near to allowing Lycophron to predict, in rather Old Testament terms, the coming of Christ; if the same idea had occurred to any of Lycophron's Byzantine admirers, the poet might have been added to the small band of pagan authors traditionally deemed to have been granted some prevision of the Incarnation.

⁴⁰ We expect an expression meaning ‘by sea’ in 1436, but though there have been several ingenious conjectures, none is immediately convincing.

⁴¹ Contrast these out-of-the-way periphrases with the plain phraseology of 1447–8.

⁴² Strab. 593, Arr. An. i 12.7–8.

⁴³ Timaeus would surely not have let such an opportunity slip: compare his laboured onomastic fantasy of a link between the mutilation of the Hermae and Hermocrates, son of Hermon, the agent of the Athenians' destruction (FGrH 566 A 102a).

⁴⁴ This designation is obviously modelled on Paris' alias, but the explanation given in the scholia (on 30) παρὰ τὸ ἀλέξειν τοὺς ἄνδρας surely has something in it. The name is perhaps also to be connected with the curious Laconian identification of Cassandra with the mysterious Alexandra worshipped at Amyclae: Paus. iii 19.6, cf. 26.5; Hsch. Κασσάνδρα ᾿Αλεξάνδρα ἐν Λακεδαιμονίᾳ Bull. Epigr. 1968, no. 264; see further Stiglitz, R., ‘Alexandra von Amyklai’, JÖAI xl (1953) 72–83Google Scholar (who, however, questions whether this is relevant to Lycophron). Wilamowitz, (‘Lesefrüchte’, Hermes xliv [1919] 60 f.Google Scholar, no. 160,= Kl. Schr. iv 298–9Google Scholar) saw an allusion to the Spartan cult in 1126–7, but I am not altogether convinced; it is clear from Pausanias that the identification was unfamiliar in his day, and the couplet goes so easily with what follows (taking αὖθι as ‘hereafter’, like αὖθις, not as ‘there’) that such an allusion would be all too easily overlooked. At all events, though knowledge of the Amyclaean Alexandra may have influenced Lycophron's choice of nomenclature, he can hardly have expected this association to spring to the minds of his readers.

⁴⁵ The scholia offer the following note on ᾿Ακταίων: ᾿Ακταῖοι δὲ οἱ ᾿Αθηναῖοι. ἀπώκησαν δὲ μετὰ τῆς Μηδείας εἰς Πέρσας τινὲς τῶν ᾿Αθηναίων. νῦν οὖν ᾿Αθηναίους λέγει τοὺς Πέρσας διὰ τὴν ἀποικίαν. This sounds more confident, but is no more convincing, than Wilamowitz's explanation of Ἀργείων (Wilamowitz 1924 145): ‘Dass unter die Argeiern die Perser, von Perseus und Andromeda her (Herodot vii 150), nicht die Argeaden zu verstehen sind, ist allerdings toll genug, bringt aber zum Ausdruck, dass sich in den Herrschern von Ost und West das Blut bereits gemischt hat’. Scheer conjectured Ἀρταίων (cf. Hdt. vii 61).

⁴⁶ On the various legends about Odysseus' death see Hartmann, A., Untersuchungen über die Sagen vom Tode des Odysseus (Munich 1917), esp. 145 ff.Google Scholar, Meuli, K., Hermes lxx (1935) 167 fGoogle Scholar. ( = Gesammelte Schriften ii 868–70Google Scholar); on Odysseus' Italian adventures see Phillips, E. D., ‘Odysseus in Italy’, JHS lxxiii (1953) 53–67CrossRefGoogle Scholar.

⁴⁷ ᾿Αριστοτέλης ‹δέ› φησιν ἐν ᾿Ιθακησίων πολιτείᾳ (fr. 508 Rose) Εὐρυτᾶνας ἔθνος εἶναι τῆς Αἰτωλίας ὀνομασθὲν ἀπὸ Εὐρύτου, παρ᾿ οἶς εἶναι μαντεῖον ᾿Οδυσσέως.

⁴⁸ As Geffcken well emphasises, GGA clviii (1896) 122Google Scholar.

⁴⁹ Sch. 800: Τράμπυια πόλις ᾿Ηπείρου ἔνθα μετὰ τὸν νόστον ᾿Οδυσσεὺς ἀπῆλθε, καθὰ καὶ ῾῾Ομηρος ἱστορεῖ ῾εἰσόκε τοὺς ἀφίκηαι, οῖ οὐκ ἴσασι θάλασσαν᾿ (Od. xi 122) ἔνθα καὶ τιμᾶται ὁ ᾿Οδυσσεύς Steph. Byz. p. 182 Mein. s.v. Βούνειμα πόλις ᾿Ηπείρου, οὐδετέρως, κτίσμα ᾿Οδυσσέως, ῆν ἔκτισε πλήσιον Τραμπύας, λαβὼν χρησμὸν ἐλθεῖν πρὸς ἄνδρας, ῾οἰ᾿ οὐκ ἴσασι θάλασσαν᾿. βοῦν οὖν θύσας ἔκτισε: p. 631 Mein. Τραμπύα, πόλις τῆς ᾿Ηπεἰρου πλησίον Βουνίμων.

⁵⁰ Holzinger's note on 800 is surely right: ἐδέθλιον ist nicht ein Tempel (wie in vs. 987), sondern überhaupt ein ἐρυμνὸν κτίσμα (vs. 78)’.

⁵¹ τὸν Αἰακοῦ τε κἀπὸ Περσέως σπορᾶς 803 ~ ἀπ᾿ Αἰακοῦ τε κἀπὸ Δαρδάνου δεδώς 1440.

⁵² Wilamowitz, in his valuable discussion of this passage, Homerische Unlersuchungen (Berlin 1884) 189–90Google Scholar, argued that the lines do not in fact refer to Etruria: ‘Bei diesem rätselpoeten ist nicht grade wahrscheinlich, dass die Tyrsener, die es ziemlich überall gibt, bedeuten was am nächsten liegt. Nun wohnen nach Herodot Tyrsener bekanntlich zwischen Axios und Strymon, im innern Makedonien, und am mittleren Axios liegt eine Stadt Gortynia oder Gordynia.’ This is ingenious, but some hint that the reader should look towards Macedonia would be needed if the trick credited to Lycophron were to work. But even Wilamowitz's interpretation does not avoid the apparent bilocation of Odysseus' remains.

⁵³ He would have been unmoved by the attempts of Holzinger and Ciaceri to refute him.

⁵⁴ As in the Telegony, where, however, Odysseus was made immortal (Τηλέγονος δ᾿ ἐπιγνοὺς τὴν ἁμαρτίαν τό τε τοῦ πατρὸς σῶμα καὶ τὀν Τηλέμαχον καὶ τὴν Πηνελόπην πρὸς τὴν μητέρα μεθίστησι. ἡ δὲ αὐτοὺς ἀθανάτους ποιεῖ, Proclus).

⁵⁵ Op. cit. (n. 48) 123 n. 1.

⁵⁶ Sch. 806: Θεόπομπός φησιν (FGrH 115 F 354) ὅτι παραγενόμενος ὁ ᾿Οδυσσεὺς καὶ τὰ περὶ τὴν Πηνελόπην ἐγνωκὼς ἀπῆρεν εἰς Τυρσηνίαν καὶ ἐλθὼν ᾤκησε τὴν Γορτυναίαν, ἔνθα καὶ τελευτᾷ ὑπ᾿ αὐτῶν μεγάλως τιμώμενος . According to Aristotle (fr. 507 Rose; from the Ithacesion Politeia) Odysseus, having been exiled after his massacre of the suitors, εἰς Ἰταλίαν μετέστη; fr. 508 should be compared (see n. 47), but there is no reason why Aristotle should not have recorded in the same work various traditions about what happened to Odysseus after the Odyssey.

⁵⁷ As in Herodotus (ii 112 ff.) Helen is detained in Egypt with Proteus, without any extenuation of her guilt, though Lycophron retains the Stesichorean eidolon invented to exculpate her; see further West, S., ‘Proteus in Stesichorus' Palinode’, ZPE xlvii (1982) 6 ff.Google Scholar

⁵⁸ This involves taking σῦϕαρ as a substantive, as indeed it usually is.

⁵⁹ The problems of kindred and affinity are sorted out in the scholia; there is much to recommend Fox's conjecture in 808, κάσις for πόσις, implied, as he notes, by Scaliger's translation ‘frater’.

⁶⁰ I need not discuss the awkward question of a possible connection between Odysseus and Hellanicus' Nanas, the Pelasgian founder ofCortona (FGrH 4 F4), though I would not object to the idea that some such thought suggested νᾶνος at 1244; see further Horsfall (n. 29) 380–81.

⁶¹ On this section see Walter 8–47.

⁶² Cf. Strab. 244: ἐμύθευον δ᾿ οἱ πρὸ ἡμῶν ἐν τῷ ᾿Αόρνῳ τὰ περὶ τὴν νέκυιαν τὴν ῾Ομηρικήν καὶ δὴ καὶ νεκυομαντεῖον ἱστοροῦσιν ἐνταῦθα γενέσθαι καὶ ᾿Οδυσσέα εἰς τοῦτ᾿ ἀφικέσθαι. Wilamowitz was surely right in locating Aeschylus' Psychagogoi here (Aischylos. Interpretationen [Berlin 1914] 246Google Scholar n. 1). The question has attracted renewed interest with the publication of the Cologne fragment of that play (PKöln iii 125Google Scholar): see, most recently, Rusten, J. S., ‘The Aeschylean Avernus’, ZPE xlv (1982) 33–8Google Scholar.

⁶³ Thus Holzinger (on 710): ‘Odysseus wird nach glücklicher Vollendung seines Ganges in die unterweltlichen Gefilde der Persephone und ihrem Gemahle Hades ein Anathem darbringen.’

⁶⁴ Editors usually punctuate with a high stop at the end of 709, but Holzinger's comma and dash make matters clearer for the reader.

⁶⁵ For a sensible discussion of the rather sketchy geography of this passage see Phillips (n. 46) 59.

⁶⁶ The text is uncertain in 664: see app. crit. The reading generally adopted by modern editors, εὔτορνα σκάϕη, fails to convey the essential restriction of πάντα to ships in the harbour; the scholia and periphrases preserve this by retaining the MS reading εὐτόρνῳ, though with some disagreement as to whether it should be taken substantially (ἐν τῷ εὐτόρνῳ καὶ περιφερεῖ-λείτει λίμενι) or with σκάϕει, read by some MSS and understood as meaning ‘basin’. Walter, who draws attention to this difficulty (which is ignored by Holzinger and Ciaceri), commends Schaefer's conjecture ἐν τόρνῳ, comparing the description of the Laestrygonian harbour at Od. x 87–90: ‘etiamsi genuinam lectionem non restituerit, in eo certe a vero non deerravit, quod portus notioncm in hac voce inesse voluit’. However, ἐν τόρνῳ is not in itself convincing, and the general effect seems somewhat bald (for Lycophron) if we simply assume that some more recherché term for a harbour lurks in this part of the line; moreover, εὔτορνα σκάϕη is rather pleasing as an iambic equivalent for νέες ἀμϕιέλισσαι. I suspect a line has dropped out after 664.

⁶⁷ We may note that the Sirens' necrology contains what is generally supposed to be the poem's only anapaest (Παρθενόπην 720); but see further n. 71.

⁶⁸ On 664 see above n. 66. On 738–40 Tzetzes justifiably comments κακῶς δὲ καὶ συγκεχυμένως καὶ ἀδιαρθρώτως ὁ Λυκόφρων τὴν περὶ ᾿Οδυσσέως ἱστορίαν λέγει . 738–9 take us back to a much earlier stage in the narrative, Odysseus' dealings with Aeolus (Od. x 1–75), and lack logical connection with what follows, the disastrous storm in which Odysseus' comrades perished (Od. xii 405–19); συμϕλεχθήσεται ought to mean ‘will be burnt to ashes’, and hardly suits Odysseus, who escaped the thunder-bolt unscathed. This incoherence must be connected with the absence of any counterpart to the crucial Thrinacian episode (Od. xii 260 ff.).

⁶⁹ ‘Pluris fecit Lycophro eruditionem geographicam et historicam quam quid Alexandrae conveniret; saepius enim narrantem earn facit quae cum bello Troiano aut omnino non coniuncta sunt aut certe per filum tenuissimum, velut in Egestae Siridis Tylesii Terinae antiquitatibus’ (Wilamowitz 1883 5).

⁷⁰ On this legend see Perret 257–83, 631–6. He convincingly argues that the story so obviously has the makings of a tragic plot that we are justified in assuming it formed the basis of a lost Laomedon or Hesione; his attempt to connect this with the Athenian alliance with Segesta in 428 is obviously rather speculative, but harmless in itself provided it is not treated as a starting point for further hypothesis.

⁷¹ I am reluctant to credit even an interpolator with the extraordinary lengthening Φοινοδάμαντος (953); we have the licence of an anapaest to admit a proper name at 720, and Φοινοδάμας could have been accommodated similarly. Sir Charles Willink (per epist.) has suggested reading Φοινοδαμαντείους (cf. Δυμαντείου 1388), which is much better than my own idea that κόρας had replaced a cretic like ἐκγόνους (West 119).

⁷² The complicated story is unravelled by Jacoby on Timaeus (FGrH 566 F 51); see also RE s.v. ‘Siris’, iiiA 1.309 ff. (Philipp).

⁷³ A second reference to ocular reaction by Athena's statue in the presence of sacrilege (988–90) diminishes the dreadful effectiveness of 361–2; this may not be much of an argument against the authenticity of the passage, but should reduce any lingering regrets at its removal.

⁷⁴ He returns to this problem in his note on 1047. Since I have more than once appealed to Tzetzes as a witness to inconsistency sufficiently blatant to worry a reasonably careful reader, I should note that he also quite unjustly accuses Lycophron of self-contradiction over Laodice's death (see his notes on 314, 447, 497).

⁷⁵ Geffcken (n. 48) 123 rightly rejects Holzinger's attempt to remove the inconsistency by taking the name Calchas to be used ‘proleptisch als den des Vertreters des ganzen Gattung’, to designate an unknown Italian seer. The passage presents a further problem arising from the scholiasts' confidence that the seer's death was brought about by Heracles; thus Tzetzes complains at some length about anachronism: νῦν δὲ φλυαρῶν ὑφ᾿ ῾Ηρακλέος φησὶν αὐτὸν ἀναιρεθῆναι . . .πρὸ γὰρ τῆς στρατιᾶς ᾿Αγαμέμνονς πολλῷ πρότερον ἐν Οἴτῃ ῾Ηρακλῆς ἐτελεύτησεν ὅμως δὲ ὁ Λυκόφρων μηδ᾿ ὅλως τῆς ἀληθείας ἤ κἄν τοῦ πιθανοῦ πεφροντικὼς οὕτω φησὶ τὴν ἱστορίαν . This criticism looks quite unfair, since there is no discernible reference to Heracles in the text: I wonder whether something has dropped out.

⁷⁶ Josifović (905) takes this group to be Colchians, but they must surely be Greeks (so Holzinger, Ciaceri).

⁷⁷ The scholia here are obviously defective and offer no help: see Scheer's apparatus. The problem has been recently discussed by Bonanno, A., ‘Lycophron and Malta’, Φιλίας χάριν: miscellanea di studi dassici in onore di Eugenio Manni i (Rome 1980) 271–6Google Scholar.

⁷⁸ Holzinger rightly observes that the repetition of Ὀθρωνός at so short an interval is ‘auffallend’.

⁷⁹ Mein. p. 484 ᾿Οθρωνός πόλις , οἱ δὲ νῆσον πρὸς νότον Σικελίας ἄλλοι δὲ μελιτηνῆς.

⁸⁰ Not everyone has accepted this evidence, and there have been other attempts to meet the difficulty, Scheer, , RhM xxxiv (1879) 452Google Scholar, thought Lycophron had somehow muddled the two Melitae; on the ancient tendency to confuse them see Pfeiffer on Call fr. 579. Ciaceri proposed re-punctuation, putting commas after νῆσον and πλαγκτοί (so Mascialino), ‘dopo esser sbalzati sin presso ad Otrono’, but this is artificial and no-one could be expected to understand the text like this without modern punctuation. Holzinger's attempt to deal with the problem (n. on 1042) is far-fetched.

⁸¹ Odysseus' visit to Cape Pachynus is mentioned again, rather more horrifically, at 1181–8.

⁸² Sch. ad 1083, Μέμβλης ποταμὀς ᾿Ιταλίας πλησίον τῶν Λευκανῶν οἵτινες εἰσιν ἔθνος ᾿Ιταλίας : this looks like a guess.

⁸³ The starting-point for ἐκπεπλωκότες is very oddly defined; ἐν Λαμητίαις δίναισιν must mean ‘beside Lametian waters’, but is hardly a natural way to convey this simple notion.

⁸⁴ I hope this will not be taken to imply an excessively naive view of Lycophron's debt to Timaeus; on this much-discussed question see Fraser 1972 i 763 ff., ii 1065–7 nn. 330, 331.

⁸⁵ I must admi t to some unease about the second, Italian, half of Menelaus' nostos (852–76). The narrative is not sufficiently elaborated: we are not told why Menelaus makes for Italy, nor is it clear whether 871 ff. refer to permanent settlement—Elba as an alternative to Elysium (Od. iv 563 ff.)?—or to the turning point in his peregrinations. There are various peculiarities of expression (in particular at 855, 857, 869, 875–6) and the topographical details of 868–70 are problematic: see further Platt, A., JPh xx (1892) 113 ff.Google Scholar, Manni, E., Geografia fisica e politica della Sicilia antica (Rome 1981) 33–5Google Scholar. 851 would undeniably form a more effective close to Menelaus' adventures than 876; Cassandra would then create the impression of a disastrous nostos by simply suppressing the happy sequel to Menelaus' Egyptian adventures. (With similar effect Vergil's Diomedes, cataloguing Greek misfortunes after the fall of Troy and understandably not quite au courant with the latest developments, speaks of Menelaus as exiled ‘Protei … adusque columnas’, A. xi 262–3.)

⁸⁶ Thus Welcker (n. 35) 1261: ‘Wenn irgendwo Interpolation nicht unerwartet ist, so muss es in einer langen Orakelpoesie seyn: und wenn irgend ein Gegenstand zur Fortführung derselben auffordern konnte, so war es die Morgenröthe einer Weltherrschaft’.

⁸⁷ Secondary Epic (Collected Poems of W. H. Auden, ed. Mendelson, E. [London 1976] 455–6)Google Scholar.

⁸⁸ Pace Wilamowitz 1924 148–9.

⁸⁹ See further Dihle, A., ‘Der Prolog der “Bacchen”’, SHAW 1981 2Google Scholar, and with reference to the Alexandrian theatre Fraser 1972 i 620–1.

⁹⁰ Plut., Vit. Lycurg. 15Google Scholar (Mor. 841a); see further Page, D. L., Actors' Interpolations in Greek Tragedy (Oxford 1934)Google Scholar, Barrett, W. S., Euripides' Hippolytos (Oxford 1964) 46Google Scholar.

⁹¹ Op. cit. (n. 89).

⁹² FGrH 566 F 98.

⁹³ I cannot altogether repress a suspicion that Cassandra's mysterious relative may be Augustus, who could, as Vergil shews, credibly be presented out-trumping Alexander's achievements (A. vi 791 ff.; cf. Norden, E., ‘Ein Pancgyrikus auf Augustus in Vergils Aeneis’, RhM liv [1899] 466–82Google Scholar, = Kl. Schr. 422–36). πρέσβιστος (1449) nicely combines the connotations of Augustus and princeps, and the line as a whole well suits Augustus' preference for titles without clear-cut official connotations. I have argued for this interpretation elsewhere (West 127 ff.) and suggested that the immediately preceding lines (1442–5) were intended to refer to the Second Macedonian War but have been rendered virtually unintelligible by accidental omission before 1442. I now think this corollary may be over-ingenious; the interpolator may simply have made rather a poor job of adapting whatever Lycophron said about Alexander so that it would take an extension. Admittedly, μεθ᾽ ἕκτην γένναν may not seem adequate for an interval of about three centuries, but the attractions of seven as a symbolic number would account for some poetic licence.

⁹⁴ Steph. Byz. p. 50 Mein.: Αἵνεια, τόπος Θρᾴκης , ὡς Αἵπεια Ζέλεια, ἀπὸ Αἰνείου. Θέων δ᾿ Αἰνειάδας ταύτην καλεῖ, ὑπομνηματίζων τὀν Λυκόφρονα Αἰνείας δὲ μετὰ τὴν ᾿Ιλίου πόρθησιν εἰς Θρᾴκην παρεγένετο καὶ ἔκτισε πόλιν Αἰνειάδας, ὅπου τὸν πάτερα ἔθαψε.

⁹⁵ See further Guhl, C., Die Fragmente des alexandrinischen Grammatikers Theon (Diss. Hamburg 1969) 1–3Google Scholar.

⁹⁶ ὑπόμνημα is applied to a variety of literary productions ranging from rough jottings to the history of Polybius (see Bömer, F., ‘Der Commentarius’, Hermes lxxxi [1953] 215–50Google Scholar), and we cannot tell whether Theon attempted a systematic commentary or just wrote about passages which interested him.

⁹⁷ Of course, even external evidence, unless it consists of the author's autograph, does not settle such arguments beyond peradventure, as may be observed from the debate surrounding the opening lines of Euripides' Phoenissae, against which there is now considerable papyrological testimony: see further Haslam, M. W., GRBS xvi (1975) 147–74Google Scholar, van der Valk, M., GRBS xxiii (1982) 235–40Google Scholar. Stylometric criteria, it need hardly be said, are no use in dealing with short passages.

⁹⁸ Interpolations which actually enhance a work are theoretically conceivable, but it is difficult to suggest probable examples, and critics seem generally to discount the possibility.

⁹⁹ Op. cit. (n. i) Letter xxix, 27 Nov. 1799.

¹⁰⁰ The poem's most sympathetic interpreter is the radical Unitarian Josifović; but though his approach to the poem is in many ways attractive, the argument on which it is based seems to me quite unconvincing.

¹⁰¹ See above, n. 44.

¹⁰² Paris' seduction of Helen is sacrilegious because it is an offence against Zeus Xenius.

¹⁰³ On its popularity with Hellenistic poets see further Hensel, L., Weissagungen in der alexandrinischen Poesie (Giessen 1908)Google Scholar.

¹⁰⁴ Op. cit. (n. i), Letter xxxii, 13 March 1800.

¹⁰⁵ We are presumably to understand that Apollo simply does not shew Cassandra what happens after the Palladium's miraculous reaction, so that she is left to infer that Ajax will achieve his deplorable purpose (cf. 411–12, 1089, 1142–3, 1151). If this inference were correct, it would be hard to understand her subsequent selection as Agamemnon's concubine (1108 ff.); while Agamemnon might be satisfied with a second-hand concubine when, as in the Iliad, only a rather restricted selection was available, it is almost incredible that with first refusal of every girl in Troy, he should choose one who had been violated by an unruly subordinate only hours earlier. Euripides, who like Lycophron combines both aspects of Cassandra's tale, expressly assures us that she was rescued from Ajax before the worst occurred (Tro. 69–71, 453); Athena could, after all, save her suppliant quite simply, by sending other Greeks after Ajax. Lycophron has incurred some criticism for combining incompatible erotic motifs here, but I suspect he intended us to supply a rescue-party from our general knowledge of the story; we are handicapped here by the loss of the Cyclic Iliou Persis. Cassandra's second sight is restricted to what the god cares to reveal, and it is not his purpose to save her unnecessary anxiety.

¹⁰⁶ See further Fontenrose, J., The Delphic Oracle (Berkeley 1978) 131–7Google Scholar, Graf, F., ‘Die lokrischen Mädchen’, SSR ii (1978) 61–79Google Scholar (where references to earlier discussions may be found).

¹⁰⁷ The text of 1157 is clearly corrupt: I would emend ὅταν to ὅτων (with έκβράσσει in 1158), and transpose 1159 to follow 1173. Though Lycophron appears not to say how many girls were involved, our other sources, including the Vitrinitsa inscription (IG ix. 1²(3) 706), leave no doubt that two girls served (until, perhaps, the Tribute was reduced to one). Lycophron could be made to supply this detail, which seems quite important, if ταῖς θανουμέναις (1160) were changed to τοῖν θανουμέναιν: see further West 119–21.

¹⁰⁸ On these aspects of the archaic world-view see further Dodds, E. R., The Greeks and the Irrational (Berkeley/L.A. 1951) 28–63Google Scholar, esp. 33 ff.; on hereditable guilt see also Parker, R., Miasma (Oxford 1983) 199–206Google Scholar.

¹⁰⁹ On Hellenistic interest in Herodotus see Murray, O., ‘Herodotus and Hellenistic culture’, CQ xxii (1972) 200–13CrossRefGoogle Scholar (though Lycophron is unfortunately passed over).

¹¹⁰ There is a nice verbal reminiscence of Herodotus' ταῦτα μὲν δὴ ἴσα πρὸς ἴσα σφι γενέσθαι (i 2.1) 1302–3, οὐδ᾿ οἴ γ᾿ ἀπηρκέσθησαν ἀντ᾿ ἴσων ἴσα λαβόντες, introducing non-Herodotean material.

¹¹¹ Aristophanes is generally supposed to have found this section rather comic (Ach. 524–39). Even if we are wrong in seeing an allusion to Herodotus here (see Fornara, C. W., ‘Evidence for the date of Herodotus' publication’, JHS lxxxxi [1971] 25–34CrossRefGoogle Scholar), the idea that there might be a connection must have struck many readers independently and might be expected to have occurred to Lycophron in connection with his work on comedy.

¹¹² Eg. i 32 (Solon), 71 (Sandanis), iv 83 (Artabanus), v 36 (Hecataeus), viii 68 (Artemisia); see further Lattimore, R., ‘The wise adviser in Herodotus’, CPh xxxiv (1939) 24–35Google Scholar.

¹¹³ The symbolic beasts of her prophecies of course remind us of the animal imagery of the Oresteia and the Supplices, but are a conventional feature of oracular language.

¹¹⁴ See further Garvie, A. F., Aeschylus' Supplices: Play and Trilogy (Cambridge 1969) 163–233Google Scholar. The Danaid theme of cousins in lustful pursuit of cousins (cf. A. PV 855 ff.) is echoed in Lycophron's narrative of the combat between the Dioscuri and Apharidae (546–7).

¹¹⁵ Cf. Cho. 71–4, where the loss of virginity is used to illustrate and underline the indelible pollution of murder.

¹¹⁶ The question of the authenticity of the Prometheus-plays is irrelevant here, since no-one in antiquity, so far as we know, doubted their ascription to Aeschylus.

¹¹⁷ Fr. 326 Mette (=199 Nauck).

¹¹⁸ Cf. Ar. Ra. 923 ff.

¹¹⁹ Fr. 191. 1–4.

¹²⁰ LGS 138 (SLG S 262, fr. 298 Voigt). It is surely more likely than not that the well-read Lycophron knew this poem, which was sufficiently popular in Roman Egypt for two copies (PKöln 59, POxy 2303) to have been identified, as well as a fragment of a commentary on it (POxy 2506, frr. 84 + 108).

¹²¹ Cf. Thgn. 1231–4.

¹²² Approximately this period is indicated by Lycophron's reference to the duration of the Locrian Tribute as τὸν χιλίωρον χρόνον (1153), which implies a fairly high date for the Trojan War. Duris had contrived to date the fall of Troy to 1334/3 (FGrH 76 F 41); in view of the obvious poetic appeal of a round millennium from the Trojan War to Alexander's invasion of Asia, Lycophron might be expected to prefer this computation to more sober estimates such as those of Herodotus (c. 1250, ii 145.4) and Timaeus (1193, FGrH 566 F 125).

¹²³ We of course know from the Vitrinitsa inscription (n. 107) that c. 280 Locrian girls were again (?) serving Athena, and a tradition about the resumption of the Tribute can be pieced together from the literary sources; but that is another story.

¹²⁴ The scholia are surely right to interpret Μεμϕίτῃ πρόμῳ (1294) as a reference to Osiris; for the identification of Io with Isis cf. Call. Epigr. 57.1, fr. 383. 12 ff.

¹²⁵ On Herodotus' Egyptophilia see further Lüddeckens, E., ‘Herodot und Ägypten’, ZDMG civ (NF xxix) (1954) 330–46Google Scholar, = Herodot. Eine Auswahl aus der neueren Forschung, ed. Marg, W. (Darmstadt 1962) 434–53Google Scholar, and Froidefond, C., Le Mirage égyptien dans la Littérature grecque d'Homère à Aristote (Gap 1971) 115–207Google Scholar.