1. Hermes 63 (1928) 81Google Scholar.

2. To adopt Mackail, J. W.'s usefully non-committal term: ‘The epilogue of the Odyssey’, Greek poetry and life: essays presented to Gilbert Murray (1936) 1–13Google Scholar.

3. Commentatio de extrema Odysseae parte inde a rhapsodiae versu ψ ccxcvii aevo recentiore orta quam Homerico (1816).

4. Thus, eg., Cohn in his RE article on Aristarchus (1895) did not think that the interpretation of the scholion called for discussion (865): ‘Den letzten Teil der Odyssee von xxiii 297 an bezeichnete er mit seinem Lehrer Aristophanes als unecht, ohne ihn deshalb ganz wegzulassen’.

5. Page, D. L., The Homeric Odyssey (1955) 101–36Google Scholar.

6. Erbse, H., Beiträge zum Verständnis der Odyssee (1972) 166–244Google Scholar.

7. See Stanford, W. B., ‘The ending of the Odyssey’, Hermathena 100 (1965) 5–20Google Scholar; Moulton, C., ‘The end of the Odyssey’, GRBS 15 (1974) 153–69Google Scholar; Stössel, H.-A., Der letzte Gesang der Odyssee: eine unitarische Gesamtinterpretation (1975)Google Scholar; Wender, D., The last scenes of the Odyssey, Mnemosyne Suppl. 52 (1978)Google Scholar.

8. Omero, Odissea vi: libri xxi-xxiv, testo e commento a cura di M. Fernandez-Galiano e A. Heubeck (1986; English ed. forthcoming); see especially the introductions to 23 and 24 and n. on 23.297. Heubeck's discussion offers a panoramic view of the controversy and an invaluable guide to the secondary literature.

9. See in particular Eisenberger, H., Studien zur Odyssee (1973) 314–26Google Scholar; Solmsen, F., ‘The conclusion of the Odyssey’ in Kirkwood, G. M. (ed.), Poetry and poetics from ancient Greece to the renaissance: studies in honor of James Hutton, Cornell studies in classical philology 38 (1975) 13–28Google Scholar (= Kleine Schriften 3 (1982) 1–16Google Scholar); Postlethwaite, N., ‘The continuation of the Odyssey: some formulaic evidence’, CPh 76 (1981) 177–87Google Scholar; van Thiel, H., ‘Aufbau und Herkunft der zweiten Nekyia’, Melanges Ed. Delebecque (1983) 435–9Google Scholar.

10. Wender (n.7) 59.

11. An awkward term to translate. Telemachus says there are many βασιλῆες in Ithaca (1.394–5): ‘noblemen, chiefs, lords’, not ‘kings’. But Odysseus appears to have enjoyed a peculiar authority in Ithaca. See further LfgrE s.v. βασιλεύς; (M. Schmidt), Hainsworth, J. B. on Odyssey 8Google Scholar (introduction).

12. Cf. Il. 18.352–3; 24.580, 587–8.

13. See further Crooke, W., ‘The wooing of Penelope’, Folklore 9 (1898) 97–133Google Scholar; Eisler, R., Weltmantel u. Himmelszelt (1910) 131–41Google Scholar.

14. On the importance of clothes in the Homeric world see Block, E., TAPhA 115 (1985) 1–11Google Scholar.

15. As is well emphasised by Wickert-Micknat, G., Archaeologia Homerica R (Die Frau) (1982) 38–50Google Scholar. For Penelope's fine work cf. 19.232–5.

16. Folktale conventions might lead us to expect that the completion of Laertes' shroud would mean his death (and the removal of an obstacle to Penelope's re-marriage?). But this association of ideas does not seem helpful here.

17. Cf. Lamer, , RE Hbbd. xxiii 427Google Scholar (s.v. Laertes).

18. 11.187–96; 14.173; 15.353–7; 16.137–53; 302.

19. Well discussed by Müller, M., Athene als göttliche Helfehn in der Odyssee (1966) 162Google Scholar.

20. ‘This suggestion reveals the embarrassment of Eustathius, who clearly felt that the simple wording of the Scholion did not admit such an evasion; nevertheless it has been repeatedly approved in modern times’ (Pfeiffer, R., History of classical scholarship I (1968) 175Google Scholar).

21. This point is well made by Stössel (n.7) 184, as part of his excellent refutation of Erbse's attempt to interpret τέλος; by reference to Aristotelian poetics; cf. Rossi, L. E., RF 96 (1968) 152Google Scholar; Schwartz, E., Die Odyssee (1924) 152Google Scholar.

22. I do not think Apollonius Rhodius should be cited as a witness in this debate. Rossi has recently revived the idea that the last line of Argonautica (4.1781, ) alludes to 23.296 (RF 96 (1968) 151–63Google Scholar; cf. Pfeiffer, (n.20) 176); his arguments get short shrift from Campbell, M., Mnemosyne ser. 4, 36 (1983) 154–5CrossRefGoogle Scholar. But even if we accept that Apollonius intended an allusion, his use of Homeric material is too subtle for its implications here to be unambiguous. See further Livrea, E., Apollonii Rhodii liber quartus (1973) 486–7Google Scholar.

23. Eustathius' amplification surely simply represents his own view.

24. Cf. his atheteses of Il. 1.366–92; 15.56–77; 18.444–56. We can only speculate about the basis for this general objection; see further Roemer, A., Aristarchs Athetesen in der Homerkritik (1912) 278–304Google Scholar.

25. See below, pp. 122–3.

26. For Aristarchus changing his mind on readings in the Odyssey see the scholia on 1.188; 3.289, 453; 5.272, 337, 346 etc. See further Ludwich, A., Aristarchs homerische Textkritik I (1884) 27–38Google Scholar. Petzl, G., Anlike Diskussionen über die beiden Nekyiai (1969) 45Google Scholar suggests that Aristarchus may have singled out 23.310–43 and 24.1–204 ‘weil sie sich als “Paradestucke” eigneten, um den unhomerischen Charakter der Schlusspartie nachzuweisen’. But their detatchability would have made them a bad choice for that purpose.

27. If we accept Porson's emendation of λ′ ( = 30) to δ′ ( = 4) in the scholion on 19.130, no other Aristarchean atheteses extending beyond 20 lines are recorded for the Odyssey. For the Iliad cf. 3.396–418; 7.443–64; 15.56–77. (Aristarchus did not share Zenodotus' suspicions of 16.432–58 and 18.483–608.)

28. On any view, it is surely inconceivable that this could ever have been literally the epic's conclusion; if what follows was added later, the Odyssey's original ending must be supposed to have been adjusted slightly. Page (n.5) 131 should not have revived Kirchhoff's suggestion that 295 might originally have ended οἱ δ᾽ἄρ᾽ἔπειτα. Friedländer, (Hermes 64 (1929) 376Google Scholar = Studien zur antiken Literatur u. Kunst (1969) 57)Google Scholar pointed out that the regular place for ftp δ᾽ἄρ᾽ἔπειτα is the second biceps and third trochee; he also noted that another simple alternative, αὐτὰρ ἔπειτα, would be suitable only if the subject remained the same as in the preceding clause (unless the change was made explicit). It may be worth considering the possibility that the second half of the antithesis was originally intended to be a reference to the circumstances of performance, to be supplied ad hoc – perhaps touching on the desirability of bringing the performance to a close (cf. 11.328–32) and/or the possibility of continuing the tale of Odysseus’ adventures on another occasion.

29. Nothing much can be inferred from the lack of any report of Zenodotus' views; we hear nothing about his text after 18.190.

30. Thus, in particular, Merkelbach, R., Untersuchungen zur Odyssee ed. 2 (1969) 143–4Google Scholar (cf. Wilamowitz-Moellendorff, U. v., Die Heimkehr des Odysseus (1927) 72Google Scholar; Stössel (n.7) 11–12). Schwartz (n.21) 150–6) also sought an explanation from rhapsodic tradition, though he envisaged simply a false attribution of what followed 23.296 to the Telegony.

31. Nor does any unidentified Ptolemaic fragment look like a remotely conceivable candidate for identification as part of a lost alternative ending.

32. Since the codex could easily accommodate several books, fragments of codices allow no inferences about relative popularity.

33. See eg. Willcock, M. M., The Iliad of Homer i-xii (1978) 295–6Google Scholar; Lesky, A., RE Suppl. xi 791–2Google Scholar (s. v. Homeros); of course, we have to assume some modifications, both to the immediately preceding narrative and to Book 10 itself. It should be noted that no ancient critic is known to have voiced doubts about this book's authenticity, nor is there any suggestion of defective MS attestation.

34. The Pisistratean recension is nowadays a staple of Homeric orthodoxy (thanks to Merkelbach's rehabilitation, RhM 95 (1952) 23–47Google Scholar = Untersuchungen 239–62), but this apparent consensus may disguise considerable differences in the underlying conception; for my view see Heubeck, A., West, S. and Hainsworth, J. B. (eds.) A commentary on Homer's Odyssey i (1988) 36–9Google Scholar. The inclusion of material not originally part of the two great epics would give substance to the tradition that Pisistratus gathered together poems previously scattered; cf. AP. 9.442.3–4 ; Paus. 7.26.13 (Pisistratus) .

35. No doubt rhapsodes could still insert complimentary flourishes ad hoc or ad hominem; but these would not affect the mainstream of the tradition.

36. Cf. Pfeiffer (n.20) 176.

37. Of course similar considerations may be deemed to lie behind the principles of selection which dictated what information was preserved in the scholia; I would not exclude the possibility that Aristarchus was originally more forthcoming.

38. See n. 3.

39. Stanford, on 23.296ff. As Solmsen well observes (n.9) 28 (= 16) n.44): ‘By the reasons collected in Stanford's commentary ad Ψ 296ff., any text, however poor and pointless, may be defended.’

40. In addition to the works already cited I have found useful Shipp, G. P.'s concise treatment of the abnormalities in this section, Studies in the language of Homer, ed. 2 (1972) 358–64CrossRefGoogle Scholar.

41. For Aristarchus’ reasoning see above, n. 24. The summary's starting-point is strange: Odysseus πτολίπορθος might be expected to tell Penelope something about his distinguished part in the Trojan War, but this recapitulation is restricted to his Odyssean adventures. If, as we may suspect, the theme of Odysseus' long delayed home-coming had once allowed for poetic elaboration on widely different lines, until at last the version given in our Odyssey achieved a kind of canonical status, it may have seemed desirable to remind the audience at this point what form Odysseus' nostos had taken.

42. Thus not too much should be made of the novelty of ἁδινός applied to persons in Σειρήνων άδινάων (326).

43. The variant ἐπιστέλλω, given by P. Rylands 53 (iii/iv) (and the ed. princ.) is probably a conjecture; ἐπιστέλλω is not otherwise attested before the fifth century. But of course we should allow for an interval between coinage and the earliest surviving example.

44. Cf. eg., οὐδόν for ὁδόν at 17.196. Stanford ingeniously argued that misinterpretation of ἐπιθύω (Il. 18.175, Od. 16.297) as ἐπι-θύω might have suggested that there was a precedent for the lengthening of i in compounds with ὲπί (see also Erbse (n.6) 192–3).

45. Eisenberger's attempt to minimise this difficulty (n.9) 316, seems to me to show too much confidence in archaic chivalry: ‘Im Anschluss an Bethe (83) kritisiert Merkelbach (149) die Erfindung des Besuchs nach der Tisis: Odysseus gebe durch den Weggang aus dem Palast Penelope und sein ganzes Haus den Angehörigen der Freier preis, die im ω nur zu dumm seien, ihre Chance zu nützen. Es ist jedoch möglich anzunehmen, dass bei der Konzeption des Besuches die Respektierung einer θέμις, die es der Sippe eines Getöten verbot, die Gattin des Taters zu rauben oder zu ermorden und sein Eigentum zu beschlagnahmen, als selbstverständlich vorausgesetzt wurde. Immerhin sind die Angehörigen der Freier adlige Achäer und gehören dem gleichen Gemeinwesen an wie Odysseus und Telemachos; ihre Duldung des sittenwidrigen Treibens ihrer Verwandten liefert kein zwingendes Argument dafür, dass sie sich über eine solche θέμις hinwegsetzen würden.’ But even if any such convention were normally recognised, it would be unrealistic to have much confidence in its effective operation when several families unite in seeking vengeance, ‘accidents’ in such circumstances being all too predictable.

46. Heubeck (n. on 1–204) sees a kind of foreshadowing at 20.351–7 (esp. 355–6), but it is hardly obvious.

47. As Eisenberger well stresses (n.9) 324.

48. (schol. on 24.1ff.); cf. Eustath. 1957, 1ff. See also Petzl (n.26) 44–66.

49. It is not helpful to compare his function as Priam's escort in Il. 24, nor to note that belief in his role as ψυχοπομπός must antedate the Homeric poems.

50. Implying, presumably, a situation in the far west, though a subterranean location is evidently indicated by 106, 204; confusion between the two conceptions is very common. The fact that Λευκάδα πέτρην is sandwiched between these fabulous localities discourages identifying it with the Leucadian rock, and thus taking it as indicating a route north-westwards towards Thesprotis (with its river Acheron). Garland, R. (The Greek way of death (1985) 49)Google Scholar well comments ‘The topography is sufficiently unfamiliar that we might be pardoned for suspecting that we had not arrived in Hades at all, but the conversation which follows the suitors' arrival is expressly stated to be taking place in that region.’

51. Though Page (n.5) 118 attached importance to it, there is much less force in Aristarchus' objection to the admission of the suitors' souls to Hades without proper funeral rites; certainly this is inconsistent with the Iliad (cf. 7. 408–10; 23.71–4), but the ghost of the unburied Elpenor comes up from Erebos with the other souls and does not appear to be under any practical disadvantage (so Wender (n.7) 32–3). There should be no difficulty in accepting some inconsistencies in current belief both about the fate of the soul after death and about the reasons for performing a duty accepted without question; divergences from the Iliad on such points do not argue a late date of composition.

52. H. van Thiel (n.9) has recently revived this view, persuasively argued by Dümmler, F., ‘Die Quellen zu Polygnots Nekyia’, RhM n.f. 45 (1890) 178–202Google Scholar, esp. 185–94. (Cf. Lord, A. B., The Singer of tales (1960) 184)Google Scholar.

53. Proclus' summary ends . Nostoi fr. 3 (Paus. 10.28.7) .

54. ‘Hier kommt der Tod des Lokrers Aias als zu episodisch nicht in Betracht, noch weniger der des Aigisthos, dagegen ist die Ankunft des Agamemnon mit seinem ganzen Gefolge wohl ein Ereigniss, welches alle Bewohner des Orkos versammeln konnte’ (Dümmler (n.52) 187).

55. ‘Zu dem Aufzug dieser Todten passen die Fragen nicht, die meisten sind waffenlos vom Mahle dahin gerafft, die Fragen setzen einen reisigen Heerhaufen voraus und Abenteuer, wie sie auf der Rückkehr vora Feldzug häufig sind, wie Odysseus Abenteuer bei den Kikonen und dem Kyklopen’ (Dümmler (n.52) 191).

56. Amphimedon has previously been mentioned simply as a brave fighter (22.242, 277, 284); he is thus a more suitable acquaintance for Agamemnon than the better-known suitors. We had better not think too hard about what age he is supposed to be.

57. There are two major differences: (1) Amphimedon puts the completion of Penelope's web immediately before Odysseus' return (146–50), whereas elsewhere (2.93–110; 19.130–61) an interval is indicated; (2) Penelope is said to have arranged the contest with the bow on her husband's instructions (and therefore must have recognised him already): but at 21.1–4 the competition is her own idea (prompted by Athena), and Odysseus does not make himself known to her till 23. Some have thought that Amphimedon is here presented reconstructing what he supposed to have passed between Penelope and Odysseus; but this kind of subtlety would hardly be appreciated by a listening audience. There are other traces in the Odyssey of a version of the story in which Penelope recognised Odysseus at a much earlier stage and accordingly acted as his accomplice. The extremely summary way in which Amphimedon reports Odysseus' fight with the suitors (178–85) should be noted.

58. Cf. the speeches of Antinous (2.85–128) and Eupeithes (24.426–37).

59. I do not intend to restrict the application of this principle to heroic epic; it obviously applies, eg., to Plautus’ Pseudolus.

60. Cf. 1.29ff., 298ff., 3.193ff., 306ff., 4.512ff., 11.409ff., 13.383ff. See further Bassett, S., ‘The second nekyia’, CJ 13 (1918) 521–6Google Scholar; D'Arms, E. F. and Hulley, K. K., ‘The Oresteia story in the Odyssey’, TAPhA 77 (1946) 207–13Google Scholar; Hommel, H., ‘Aigisthos und die Freier’, Studium Generale 8 (1958) 237–45Google Scholar; Hölscher, U., ‘Die Atridensage in der Odyssee’, Festschrift f. R. Alewyn (1967) 1–16Google Scholar; Lesky, A., ‘Die Schuld der Klytaimestra’, WS 80 (1967) 5–21Google Scholar.

61. 200–02 might be taken as a similar tribute to the Nostoi (or at least to the Ἀτρειδῶν κάθοδος).

62. Shipp (n.40) 360.

63. As Merkelbach supposes (n.30) 154.

64. Penelope might appropriately think of Dolius in the capacity in which she brought him to Odysseus' estate even if this is now only a minor part of his responsibilities. The poet needs to provide her at this point with a servant committed to her interests.

65. In practice slaves are hardly likely to find their owners scrupulous in preserving their given names when these might cause confusion, and the possibility that Odysseus might have two slaves of the same name would not obviously suggest itself in a slave-owning society.

66. From Laertes' point of view this is highly insulting; there is no call to water down κερτομίοις in translation. Odysseus'tone in 248–57 is insufferably patronising.

67. Some have argued that not Odysseus but the poet should be supposed to be guided by simple force of habit. Thus Fenik, (Studies in the Odyssey (1974) 47)Google Scholar speaks of ‘a more or less standardized format for recognition and name-giving’ assuming ‘something like a dangerous autonomy of its own’. Somewhat similarly Richardson, N. J., Papers of the Liverpool Latin Seminar 4 (1983) 229Google Scholar argues that ‘this poet… cannot let go of the crucial themes of disguise and deception which have run throughout the whole poem’.

68. . Similarly Eustathius (1959, 26ff.) . (Aristotle's authority notwithstanding (fr. 177 R.), I wonder if this is the right diagnosis of Argus' death; I would have supposed that the poet meant us to understand that he had hung onto life (‘straight-armed the death-angel’, as James Thurber wrote of another memorable dog) far beyond normal canine limits simply to see his master once more.)

69. This view seems particularly popular at present; see Heubeck on 24.315–7 and ZA 31 (1981) 73–83Google Scholar; Thornton, A., People and themes in Homer's Odyssey (1970) 118Google Scholar; Hooker, J. T., CQ n.s. 36 (1986) 35CrossRefGoogle Scholar.

70. See further Walcot, P., ‘Odysseus and the art of lying’, AncSoc 8 (1977) 1–19Google Scholar, esp. 18–9; A. B. Lord (n.52) 178.

71. 13.256–86; 14.199–359; 19.172–202.

72. In trying to determine the poet's intention what matters is popular etymology rather than philologically correct derivation; but the application of this principle does not yield clear results. Ἀλύβας (otherwise unknown) might suggest ἀλάομαι or ἀλύω. Ἀφείδας could equally well (if we take Ἀ- as privative, not intensive) be understood as ‘ruthless’ and ‘generous’. The grandfather's name, Πολυπήμων looks straightforward, ‘rich in suffering’; but many have found attractive Cobet's suggestion that Πολυπημονίδης has replaced Πολυπαμ(μ)ονίδης (cf. πολυπάμων (∼ πολυκτπήμων) Il. 4.433, Πάμμων 24.250). With Ἐπήριτο the derivation is certain; cf. Arcad. ςἐπάριτος ‘picked, selected’ (Xen. HG 7.4.33–6); but it is debateable whether its meaning would have been clear to the poet and his audience. Eustathius, perhaps drawing on much older traditions, suggests a connection with ἔρις or with ἐπίρρητος (‘infamous’); as Volksetymologie the former has appealed to some scholars, particularly because of the parallel with the derivation of Odysseus from ὀδύσσομαι (cf. 19.407). But redende Namen are so much part of traditional Homeric stock-in-trade that we might see the problem of interpretation posed by these names as another argument against authenticity. On Homeric etymology see further Risch, E., ‘Namensdeutung u. Worterklärungen bei den ältesten griech. Dichtern’, Eumusia: Festgabe f. E. Howald (1947) 72–91Google Scholar, esp. 79–89; Rank, L. P., Etymologiseerung en verwante verschijnselen bij Homerus (1951)Google Scholar.

73. It is an indication of the wealth of Menelaus' establishment that Telemachus and Pisistratus can be offered baths on arrival (4.48ff.).

74. ‘Sehr ansprechend, solange man es für sich nimmt, ohne an das übrige zu denken’, Wilamowitz, , Die Ilias u. Homer ed. 2 (1916) 486–7Google Scholar.

75. Lord (n.52) 177–83, arguing from Yugoslav songs of return, claims that the hero's recognition by a parent is an essential element in the story, and that Odysseus' reunion with Laertes is absolutely required by the inner logic of the narrative; but he does not consider whether this theme can happily coexist with the presence of the hero's grown-up son. However, it is interesting that he feels that this scene is not in the right place in the narrative, but belongs earlier.

76. 221–2 give the impression that Odysseus had been expecting to test Dolius.

77. Page, however, surely went too far when he wrote (n.5) 102, ‘that scene as a whole was composed by … a poet whose understanding of the older Epic language and versification is very imperfect’. The Iliad and Odyssey represent only a minute fraction of the older epic material which once existed.

78. Even Erbse (n.6) 210 is clearly uneasy about it: ‘Wichtiger, aber auch schwieriger ist die Frage, weshalb sich der Dichter in ω 279 mit einem gebräuchlichen Wort für ‘schön’ nicht begnügt hat … Die Vermutung liegt nahe, dass der Sprecher, der ja in der Lügenerzählung dick aufträgt, das blasse Epitheton καλαί, mit dem die körperlichen Vorzüge von Sklavinnen sonst hervorgehoben werden (vgl. I 130), überbieten will: Odysseus, so berichtet er, habe sich die Frauen ausgesucht, nicht nur weil sie sich auf ihre Handfertigkeiten verstanden, sondern auch weil sie eine Augenweide waren. Aber Sicherheit lässt sich in diesem Punkt nicht erlangen’.

79. The scale of this presentation is to be understood as a mark of respect for the recipient (277 has of course no business in the text), but it still seems overdone; I suspect a composer unfamiliar with the conventions of heroic etiquette. In view of the apparently hereditary force of ties of hospitality (cf. 1.187–9; Il. 6.215–31), we might find a little surprising Laertes' assumption that the stranger's claims to entertainment from Odysseus are nothing to do with him (283–6).

80. For another unique gen. cf. Ὀδυσσέος Il. 4.491, with Kirk's note.

81. The following items are also noteworthy: 206 τετυγμένον ‘tilled’; 210 δμῶες ἀναγκαῖοι: the sense of ἀναγκαῖοι eludes us; 252–3: a very awkward construction; cf. Erbse (n.6) 208; 295 trans. κώκυσ᾿: intrans. elsewhere in Homer; 348 ἀποψύχω intrans. aor. act. used to mean ‘faint, swoon’, whereas elsewhere in Homer it is used only in the middle, to mean ‘cool, chill’ (Il. 11.621; 21.561; 22.2); for Homeric descriptions of faints see Il. 5.696–7; 22.467; 386 ἀποψύχωἐπεχείρεον, cf. 395 σίτωι ἐπιχειρήσειν: ἐπιχειρέω here only in Homer; 388 ἐξ ἔργων μογέοντες: it is not clear what this means – ‘fatigued from their labours’? ‘hastening from the fields’? 402 οὗλε not elsewhere in Homer, though cf. H. Ap. 466.

82. Even before Spohn, the nail was hit firmly on the head by Schneider, J. G., Praef. Orph. Argonaut. (1806) 34–5Google Scholar: ‘In extremo deinde libro auctorem ingenium et spiritus place defecisse videtur: ita, ut in rerum multarum satis gravium narratione brevitate inepta, partim etiam obscura defunctus, lectoris exspectationem plane fallat.’

83. This summary treatment of subject matter rich in intrinsic interest may remind us of the disappointing compression observable in some of the longer fragments of the Epic Cycle. See further Griffin, J., JHS 97 (1977) 48–53CrossRefGoogle Scholar.

84. The imprecision of οἱ δ᾽ (415) should be noted; ὁμῶς is baffling (cf. Shipp (n.40) 364; Stanford ad loc.); we seem to be dealing with a rather maladroit adaptation of 9.401 οἱ δὲ βοῆς ἀίοντες. οἴκων (417) used like aedes of one building is unparalleled in Homer; οἴκου, given in a few MSS, looks like a conjecture. 419 involves a rather curious use of πέμπω, here practically equivalent to ‘hand over’; cf. Stanford, Heubeck ad loc.

85. This correspondence should not be used, as it has sometimes been, as an argument for the authenticity of 24.

86. There is something slightly comic in the notion of Medon and Phemius sleeping on undisturbed through the commotion which the collection of the corpses would have involved.

87. It is interesting, and surely characteristic of this composer's reluctance to elaborate, that Halitherses is not allowed to exploit what we might think the most telling argument at his disposal, the exact and punctual fulfilment of the predictions relating to Odysseus' return which he had offered at the last Ithacan assembly (2.161–76).

88. Besslich, S., Schweigen-Verschweigen-Übergehen: die Darstellung des Unausgesprochenen in der Odyssee (1966) 99Google Scholar n.24; Focke, F., Die Odyssee (1943) 381Google Scholar; Schwartz (n.21) 129 n. 1; Wilamowitz, (n.30) 83.

89. Erbse (n.6) 241; Solmsen (n.9) 8 ( = 20); Stössel (n.7) 127–8; von der Mühll, in his edition; Wilamowitz, , Homerische Untersuchungen (1884) 72Google Scholar.

90. This brief interlude on Olympus may well be intended as a counterpart to the opening divine council, but comparison with that earlier episode makes its inferior quality only too obvious, and there is little merit in so mechanical an attempt at balance.

91. Cf. Merkelbach, , Untersuchungen 148Google Scholar: ‘Das ist offenbar nicht die ursprüngliche Form eines Schiedsspruches: der eine Teil erhält vollkommen recht, die anderen sollen die erlittene Kränkung vergessen’. (Erbse (n.6) 223 n.147, who speaks of ‘ein psychologisch verständlicher Anakoluth’, has not seen the real problem.)

92. ‘A generous yet curiously utopian promise which, as far as I can see, is unparalleled in Homer’ (Solmsen (n.9) 9 ( = 21) n.25). On the ancient tendency to equate peace and prosperity see Fuchs, H., Augustin u. der antike Friedensgedanke (1926) 175–6Google Scholar.

93. For other noteworthy features of 415–9 see above, n.83.

94. ‘The Epic was very sparing of interjections … neither this expletive … nor anything else of the kind occurs elsewhere in the Homeric poems’ (Page (n.5) 110).

95. On the Telegony, Hartmann, A., Untersuchungen über die Sagen vom Tod des Odysseus (1917) 44–105Google Scholar is fundamental. We should note the obvious compliment to the royal family in the name there given to Odysseus' second son by Penelope, Arcesilaus (elsewhere (Paus. 8.12.6) Ptoliporthes). It may be significant that Eusebius assigns to the same period a substantial influx of immigrants to Cyrene.

96. Untersuchungen 147.

97. (Ἰταλίαν codd., corr. Hartmann) . Cf. [Apollod.] Epit. 7.40 .

98. Cf. 22.413–16. See also Müller (n.19) 164–7; Kearns, E., CQ 32 (1982) 7–8CrossRefGoogle Scholar.

99. I put the composition of our Odyssey, distinguished from other treatments of Odysseus' return by its extension of his nostos to the length of the Trojan War and its elaborate flashback technique, in the middle of the seventh century: see further Commentary 33–5.

100. There was a precedent in the long night of Zeus's dalliance with Alcmene. But the motif would be absurdly devalued if we were meant to suppose that this is merely the first of many nights which Odysseus and Penelope will spend together.

101. As it does for James Joyce's Ulysses (as Neil Hopkinson reminds me).

102. Mutatis mutandis, I suspect that very similar considerations apply to the curiously similar problem presented by the clearly spurious ending of Mark (16.9–20); the hypothesis that these verses have replaced the original conclusion seems to me quite untenable. See further Nineham, D., The Gospel of St Mark (1968) 439–53Google Scholar. (It is interesting to note that Wilamowitz, in his very sympathetic account of this Gospel (‘Die Verklärung Christi’, Reden u. Vorträge II (1926) 280–93)Google Scholar, did not think it worth considering the possibility that its author ever intended any ending but 16.8.)

103. That politically the Homeric poems were a serious matter is indicated by Cleisthenes' famous ban on Homeric recitations at Sicyon (Hdt. 5.67.1).

104. Its composer himself seems to lose sight of this point; Odysseus' last speech to Penelope outlines plans to restock his estate hardly consistent with imminent departure (23.356–8), and there is no suggestion of any arrangement for an interregnum by Telemachus during his father's absence, of unpredictable duration, on his pilgrimage.

105. I suspect that he also made a significant contribution to Book 11; but this is not the place to discuss that book's peculiar problems.

106. Hdt. 1.60.3–5; on this much discussed episode see most recently Connor, W. R., JHS 107 (1987) 40–7CrossRefGoogle Scholar.

* This article has evolved from a paper given at the London Institute of Classical Studies in November 1985; it has benefited greatly from discussion there, and subsequently in Los Angeles, Berkeley, Austin, Durham, Cambridge and Toruń. I take this opportunity to thank everyone to whom I owe improvements, while regretting the difficulty of acknowledging all my many debts individually.