I wish to thank three referees, τῶν ἐγὼ εἰδὼς τὰ ὀνόματα οὐ γράφω,, for their expert comments and criticisms.

¹ To some extent I shall be using the same canvas as in an earlier essay, ‘Greek poetry 2000-700 BC’, CQ xxiii (1973) 179–92Google Scholar.

² H. M. and Chadwick, N. K., The Growth of literature (Cambridge 1932-1940) ii 12 fGoogle Scholar; Bowra, C. M., Heroic poetry (London 1952) 372 ffGoogle Scholar.

³ Snodgrass, A. M., The Dark Age of Greece (Edinburgh 1971) 192–4, 397 f.Google Scholar; Archaic Greece (London 1980) 38-40, 68–78Google Scholar; Coldstream, J. N., JHS xcvi (1976) 8–17CrossRefGoogle Scholar; Geometric Greece (London 1977) 341–57Google Scholar.

⁴ I wonder, however, whether this hexameter fragment should not be attributed rather to the Kirka of Alexander Aetolus (Powell, Collectanea Alexandria 122).

⁵ Fr. 44.6-8 μάτε[ρ' έξονομ]άσδων έκαλη νά[ïδ' ύπερτάταν] | νύμφ]αν ένν]αλίαν ἆ δέ γόνων [άψαμένα Δίος] | ίκέτέυ[' άγαπά]τω τέκεος μᾶνιν [άπυστρέφην.] (end of poem). This not only corresponds to events in Iliad i but implies the whole framework of the epic: Zeus’ response to Thetis’ prayer, granting the Trojans increasing success in battle until the Greeks are forced to plead with Achilles for his assistance. I draw attention to this because it is in my opinion the earliest text that clearly attests knowledge of our Iliad (as distinct from unidentified epic about the Trojan War). Corresponding evidence from art does not go back before about 630 (Johansen, K. Friis, The Iliad in early Greek art [Copenhagen 1967]Google Scholar).

⁶ See Schmitt, Rüdiger, Dichtung und Dichtersprache in indogermanischer Zeit (Wiesbaden 1967) 6 ffGoogle Scholar., and in the introduction to the collection of articles edited by him under the title Indogermanische Dichtersprache (Darmstadt 1968) 2 fGoogle Scholar.

⁷ Besides the works of Schmitt just cited see Durante, M., Sulla preistoria delta tradizione poetica greca (Rome 1976)Google Scholar; Campanile, E., Ricerche di cultura poetica indoeuropea (Pisa 1977)Google Scholar. (Cited hereafter as ‘Durante’, ‘Campanile’.)

⁸ Meillet, , Les Origines indo-européennes des mètres grecs (Paris 1923)Google Scholar, cf. his Aperçu d'une histoire de la langue grecque⁷ (Paris 1965) 145–52Google Scholar; Jakobson, R., Oxford Slavonic Papers iii (1952) 21–66Google Scholar. ≍ Selected Writings iv (1966) 414–63Google Scholar; Watkins, C., Celtica vi (1963) 194–249Google Scholar; Pighi, G. B., Studi di ritmica e metrica (Torino 1970)Google Scholar; West, M. L., Glotta li (1973) 161–87Google Scholar, cf. my Greek Metre (Oxford 1982) 1–4Google Scholar; Nagy, G., Comparative studies in in Greek and Indie meter (Cambridge Mass. 1974)CrossRefGoogle Scholar.

⁹ See e.g. Childe, V. Gordon, The Aryans (London 1926)Google Scholar; Prehistoric migrations in Europe (Oslo 1950) 146–51Google Scholar; Bosch-Gimpera, P., El problema indoeuropeo (Mexico 1960)Google Scholar; Devoto, G., Origini indeuropee (Florence 1962) 71–156Google Scholar; Gimbutas, M., American Anthropologist lxv (1963) 815–36CrossRefGoogle Scholar; Piggott, S., Ancient Europe (Edinburgh poetica 1965) 78–97Google Scholar; various papers in Enrich, R. W. (ed.), Chronologies in Old World archaeology (Chicago 1965)Google Scholar and in Cardona, G. et al. (ed.), Indo-European and Indo-Europeans (Philadelphia 1970)CrossRefGoogle Scholar and in Crossland, R. A. and Birchall, A. (ed.), Bronze Age migrations in the Aegean (London 1973)Google Scholar; Crossland, in CAH3³ i(2) 868–73Google Scholar.

¹⁰ Caskey, J. L., Hesperia xxix (1960) 302Google Scholar and CAH³ ii(i) 139Google Scholar; Vermeule, E., Greece in the Bronze Age (Chicago and London 1964) 66–72CrossRefGoogle Scholar; S. S. Weinberg in Ehrich (as n. 9) 305; R. J. Howell in Crossland and Birchall (as n. 9) 73-99; Schachermayr, F., Griechische Frühgeschichte (Wien 1984) 27Google Scholar, modifying his earlier views.

¹¹ Hammond, N. G. L., BSA lxii (1967) 77 f.Google Scholar; Epirus (Oxford 1967) 341 f.Google Scholar; A History of Macedonia i (Oxford 1972) 252–76Google Scholar, and in Crossland and Birchall (as n. 9) 189-97; R.J. Howell (as n. 10).

¹² These peoples together with the Greeks make up a distinct ‘eastern’ group within the IE family. Cf. Kretschmer, P., Einleitung in die Geschichte der griechischen Sprache (Göttingen 1896) 169Google Scholar; Schwyzer, E., Griech. Grammatik i (München 1939) 56 fGoogle Scholar; Porzig, W., Die Gliederung des indogermanischen Sprachgebiets (Heidelberg 1954)Google Scholar; Harmatta, J., Annales Universitatis Scientiarum Budapestiniensis, sect, classica iii (1975) 9Google Scholar; Durante ii 16-87.

¹³ Presumably pre-Homeric, though not mentioned in the Iliad or Odyssey (Dawn's horses: Od. xxiii 244-6); it is in several of the Hymns, Mimnermus, other archaic poets, and in art perhaps from the first half of the seventh century.

¹⁴ Wiesner, J., Archaeologia Homerica I F: Fahren und Reiten (Göttingen 1968) 79 fGoogle Scholar; Crossland, R. A., CAH³ i(2) 873 f.Google Scholar; Crouwel, J. H., Chariots and other means of land transport in Bronze Age Greece (Amsterdam 1981) 148Google Scholar It may be worth recalling that the first chariot mechanic in Greek myth, Myrsilos, has a Hittite name and comes to Greece from the east with Pelops.

¹⁵ I refer to the Trundholm horse and sun-disc group discovered in 1902. See Gelling, P. & Davidson, H. E., The chariot of the sun (London 1969) 14 ffGoogle Scholar; Glob, P. V., The Mound People (London 1974) 99–103Google Scholar, who states that the hollow-cast bronze horse has contemporary parallels only in Mycenaean Greece.

¹⁶ Cf. Schmitt, R., Indogermanische Dichtersprache und Namengebung (Innsbruck 1973)Google Scholar; Duranteii 103; Campabergnile 82.

¹⁷ As Ibycus frankly explained to Polycrates: κλέος ἄφθιτον έξείς ὠς κατ' άοιδάν και έμόν κλέος, PMG 282(a) 47 f.; cf. Pind. P. 3.110-5, Od. xxiv 196-8, H. Ap. 171—5, Thgn. 237-52; Durante ii 180 f., who compares some Vedic and Germanic texts.

¹⁸ Cf. Campanile 37-47.

¹⁹ One might also refer to the ancient Roman carmina dé clárórum uirórum laudibus that according to Cato and Varro were once sung at banquets; but these are under suspicion of being a literary fiction. Cf. H. Dahlmann, Abh. d. Ak. d. Wiss. u. d. Lit. in Mainz, Geistes- u. sozialw. Kl. 1950, 1191-202.

²⁰ Old Irish oirgnech, oirgnid, Gk. άνδροφόνος; Campanile 118 f. In the Graeco-Aryan tradition the epithet was also applicable to martial gods: Vedic nrhán ( = άνδροφόνος) of Indra, Gk. 'Eνυαλιωι άνδρειφόντηι.

²¹ Irish, Greek (δουρικλυτός etc.), Campanile 122. Again transferable to divinities: 'Aπόλλωνι κλυτοτόξωι. Cf. Vedic śrutáratha- ‘with famous chariot’.

²² Irish, Welsh, cf. Il. xii 131-5; Campanile 119. The motif occurs in the Russian byliny in the form that a hero, surprised that a mighty blow has no effect on his adversary, wonders if his strength is failing and tries smiting a tree, which falls. (I am indebted to Dr Paul Foote for this item.)

²³ Irish, Welsh, cf. Il. xiii 330 φλογί εἵκελον άλκήν, xvii 565, xviii 154; Campanile 121.

²⁴ As with CúChulainn in the Táin Bó Cúalnge, quoted by Onians, R. B., The Origins of European thought (Cambridge 1951) 157 f.Google Scholar, in a wider context; Il. v 4-7, xviii 205-6.

²⁵ Bíowulf 1521, The Wanderer 100, cf. Il. iv 126, xi 574, xv 542-3, xxi 70, 168; Durante ii 145.

²⁶ Bíowulf 610, 1832, al., foles hyrde; Rgv. iii 43.5 jánasya gopāh (cowherd); Gk. ποιμένα λαῶν; also Irish (Campanile 25, 120, and Études celtiques xv [1976/7] 18) and Near Eastern (Gadd, C. J., Ideas of divine rule in the ancient East [London 1948] 38 f.Google Scholar). Cf. Gonda, J., Zeitschr. f. vgl. Sprachforschung lxxiii (1956) 152Google Scholar; Durante ii III.

²⁷ Bíowulf 428, 663 eodor Scyldinga, 1044 eodor Ingwina; Fáfnismál 36.3 hers iaðarr, Helgaqviða Hundingsbana ii 42.3 folcs iaðarr, Locasenna 35.6 ása iaðarr; ἔρκος 'Aχαιῶν in Homer only of Ajax (but adapted to Achilles with different syntax, Il. i 284). Old Irish fál ‘en- closure’=‘king’ (Campanile, Studi e saggi linguistici xiv [1974] 207). For Vedic analogues—less exact—see Durante ii 114.

²⁸ The Norns, Helgaqviða Hundingsbana i 2-4; Aisa or Moira, Il. xx 127 f., xxiv 209 f; the (Kata)Klothes, Od. vii 197 f.; Durante ii 112, who also cites an Anglo-Saxon echo of the concept.

²⁹ Durante ii 121. In the Vedic passages it is a god who thus fells his enemies.

³⁰ On all this see Durante ii 114-9.

³¹ After Homer we find άθάνατον κλέος (Bacch. 13.65), κλέος άγηρατον (Eur. IA 567, cf. Pind. P. 2.52 κῦδος άγηραον), corresponding to Vedic śrávo… amŕtam ajuryám (iii 53.15).

³² On all this, and more, cf. Durante ii 93-7. I grant that not all of what I have mentioned need be evidence for heroic as distinct from e.g. hymnic poetry, and I pass over a number of other parallel phrases of which the same is true: immortal gods, immortal and unaging, god(s) giver(s) of blessings, broad heaven, broad earth, boundless earth, Zeus’ rain, the sun's wheel, the sun that sees all, good disposition (μένος/mánah), bold disposition, etc.

³³ Zeitschr. d. Deutschen Morgenländ. Gesellsch. N.F. xxviii (1953) 133 =R. Schmitt (ed.), Indog. Dichtersprache 163 f.

³⁴ Cf. Il. xvi 459 (raindrops of blood); 805 f. τὸν δ' ἄτη φρένάς εΓλε, λύθεν δ' ὔπο φαίδιμα γυῖα, στῆ δέ ταφὠν; xvii 426 ff. (Achilles’ horses weep on learning of Patroklos’ death); Pisani 130 f. = 159 f. On heroic and lachrymose horses see Bowra, Heroic poetry 157-70.

³⁵ The comparison was made by Köhler, R., Rh. Mus. xiii (1858) 316 fGoogle Scholar. Cf. Pisani 127 f = 156 f.; W. Kullmann, Philologus xcix (1955) 186Google Scholar.

³⁶ See W. Burkert, Die orientalisierende Epoche in der griech. Religion u. Literatur (Sitz.-Ber. d. Heidelb. Ak., Phil.-hist. Kl. 1984.1) 95-8.

³⁷ Behaghel, , Indog. Forsch. xxv (1909) 110 ffGoogle Scholar. Cf. Hirt, H., Indog. Grammatik i (Heidelberg 1927) 126Google Scholar. Morgan, G., Folklore xciv (1983) 44–56CrossRefGoogle Scholar, has identified a similar triadic pattern as a widespread feature of popular verse, from ancient Irish and Welsh to modern American.

³⁸ Durante ii 151 f.

³⁹ Cf. Nilsson, M. P., Homer and Mycenae (London 1933) 158Google Scholar; Gray, D. in Platnauer, M. (ed.), Fifty years of classical scholarship (Oxford 1954) 28 fGoogle Scholar; Luce, J. V., Homer and the Heroic Age (London 1975) 101–7, 119Google Scholar.

⁴⁰ Lorimer, H. L., Homer and the monuments (London 1950) 152 f., 273 f.Google Scholar; Page, D. L., History and the Homeric Iliad (Berkeley & Los Angeles 1959) 232–5Google Scholar; Borchhardt, H., Archaeologia Homerica I E: Kriegswesen (Göttingen 1977-1980) 25 fGoogle Scholar.; S. Foltiny, ibid. 237.

⁴¹ Horrocks, G. C., Space and time in Homer (New York 1981)Google Scholar. Horrocks shows that the Homeric use, though freer than the Mycenaean, is subject to limitations, only certain types of positioning being admitted. On pp. 153-61 he points out the advantages to bards in preserving the ancient freedom of arrangement Luce, in the flexible-formula system which they used.

⁴² Wackernagel, J., Kleine Schriften (Göttingen 1953) 116n. 1, 1170n. 1Google Scholar. On the substitution of ℓδέ for καί see below, n. 54.

⁴³ Ruijgh, C.J., Études sur legrammaire et le vocabulaire du grec mycénien (Amsterdam 1967) 69 f.Google Scholar; Wathelet, P., Les traits éoliens dans la langue de l'épopée grecque (Rome 1970) 171 f.Google Scholar; Durante i (1970) 89; Horrocks (as n. 41) 162; Mühlestein, H., Homerische Namenstudien (Frankfurt am Main 1987) 186 fGoogle Scholar. For a dissenting view see A. Heubeck, Acta Mycenaea ii ( = Minos xii) (1972) 55-79.

⁴⁴ So Helbig, W., Das homerische Epos aus den Denkmälern erläutert² (Leipzig 1887) 315 n. 2Google Scholar; Lorimer (as n. 40) 189; Gray, D. H. F., CQ lxi (1947) 120Google Scholar; Durante i 90.

⁴⁵ Cf. Crespo, E., Elementos antiguos y modernos en prosodía homérica (Minos Suppl. vii, Salamanca 1977) 24.Google Scholar

⁴⁶ πότνια 'Hβη Il. iv 2 and 'Hως H. Aphr. 223, vocabulaire 230, were no doubt modelled on π. 'Hρη after the hiatus became accepted, though in principle π. 'Hβη (< *yēg^wā) could be as old as π. 'Hρη. Cf. Ruijgh in A.Morpurgo Davies & Y.Duhoux (ed.), Linear B: a 1984 survey (Louvain-la-neuve 1985) 155.

⁴⁷ Durante i 82 f.

⁴⁸ Ruijgh (as n. 43) 53. Linear B still for the most part uses -e (-ει) as the dative ending, though -i (originally locative) has begun to make inroads.

⁴⁹ Durante i 85; Crespo (as n. 45) 23.

⁵⁰ Ruijgh (as n. 43) 53 and (as n. 46) 158.

⁵¹ Szemerényi, O., Studi micenei ed egeo–anatolici xx (1980) 207–11Google Scholar.

⁵² Schwyzer, Griech. Gramm. i 219 f.

⁵³ This assumes that the Homeric licence of leng- thening a final syllable such as -ος in the princeps position before a vowel was a post-Mycenaean development, or at any rate that it would not have been used in a formula.

⁵⁴ The substitution of idé for kai (Ruijgh [as n. 46 163) avoids the anomalous correption before y\h.

⁵⁵ Hence the periphrases g^wíā Hērakkweheíā, Wīphikleweheíā had to be created to stand for unmetrical Hērakléwēs, Wīphikléwes: Durante i 117–9. But when did Klytaimestra get into the tradition?

⁵⁶ In this connection a further observation may be made. Ruijgh, L'élément achéen dans la langue épique (Assen 1957) 74-89 (cf. Wathelet [as n. 43] 307-10; aliter M. Peters in o-o-pe-ro-si [Festschr. E. Risch, Berlin & New York 1986] 308 n. 20), has made it probable that -ξω -ξα in the future and aorist of -ʒω verbs with dental stems is neither Aeolic nor Ionic but goes back to the ‘Achaean’ phase of the epic language. The άμβροτάξομεν which I have mentioned is one example. The other verbs are (έξ)άλάπάʒω, πτολεμίʒω, πελεμίʒω, (έξ)εναρίʒω, δαίʒω, κτερείʒω, δνοπαλίʒω, έγγυαλίʒω, στυφελίʒω, έλελίʒω, μερμηρίʒω—nearly all verbs of war. The last in the list suggests that the recurrent Homeric motif of pondering alternatives is an early feature.

⁵⁷ So e.g. Cauer, P., Grundfragen der Homerkritik³ (Leipzig 1921-1923) 263Google Scholar; Page (as n. 40) 234-8; Webster, T. B. L., From Mycenae to Homer (London 1958) 101, 115Google Scholar; Durante i 115 f.

⁵⁸ This may also be connected with his status as a cult hero; see Mühll, P. Von der, Der grosse Aias (Basel 1930)Google Scholar.

⁵⁹ Wathelet, P., L'Antiquité classique xxxi (1962) 13Google Scholar. The figures for Aἵαντα and Aἵαντε are as follows: Aἵαντα, 17 occurrences, of which 4 scan –––, but all before πρ- (3) or μ- (1, Aἵαντα μεγαλητορα adapted from the dative formula?). Aἵαντε, 17 occurrences, of which 4 scan –––, but all before πρ- (2) or σφ- or ν-.

⁶⁰ Cf. also Webster (as n. 57) 101, 117 f.; Durante i 94.

⁶¹ Palmer, L. R., Serta philologica Aenipontana iii (1979) 260Google Scholar f., derives it from an old prepositional prefix ó- and the obsolete root *deuk/duk=ducere; cf. Poly- deukes, and Idomeneus' father Deukalion. He takes no account of the variant forms Olysseus etc. (common outside the epic tradition), Pol-lux, Leukarion (Epicharmus p. 112 Kaibel); the d/1 variation may point rather to a non-IE provenance. Cf. Gusmani, R., Studi micenei ed egeo-anatolici vi (1968) 16.Google Scholar

⁶² Also to Pylaimenes and Automedon (once each), and predicatively, not as a name-formula, to onsets of Patroklos and Hektor. Hektor is often claimed as another pre-Trojan hero, transferred from a Greek setting: Dümmler, F., Kleine Schriften ii (Leipzig 1901) 240–9Google Scholar, and others.

⁶³ Bowra, Heroic poetry 519–36.

⁶⁴ The sackings may have been done by freebooters who left no visiting cards. Cf. F. J. Tritsch in Crossland & Birchall (as n. 9) 233-8.

⁶⁵ Il. vii 133-56; xi 670-762; xxiii 629-42; cf. Paus. vi 25.2.

⁶⁶ Ps.-Hes. fr. 211; Pind. N. 3.34 μόνος ἄνευ στρατιἄς cannot be literally true, but could be an exaggerated version of a true coup. Apollodorus iii 173 (Pherec. FGrH 3 F 62) does speak of an army.

⁶⁷ Od. xi 235-57, ps.-Hes. frr. 30.31–33a. They were conceived in the river Enipeus, and there is some confusion about whether it was the Thessalian Enipeus or the tributary of the Alpheios. See West, M. L., The Hesiodic Catalogue of Women (Oxford 1985) 142Google Scholar f.

⁶⁸ Il. i 260-73, 743. Od. xxi 296-304. We now see that Nestor's participation in the event is not an extravagant autoschediasm of Homer's (Mülder, D., Die Ilias und ihre Quellen [Berlin 1910] 47Google Scholar and others), but a reflex of the Pylian hero's reception into Thessalian poetry. I do not know if it is so significant that at the start of the Trojan War he appears at Peleus’ house as a recruiting officer together with Odysseus (Il. xi 765 ff.).

⁶⁹ Desborough, V. R. d'A., The Greek Dark Ages (London 1972) 100Google Scholar; cf. CAH ³ ii(2) 666.

⁷⁰ There is a small amount of LH IIIc pottery in Troy Vila (Mee, C. B. in Foxhall, L. and Davies, J. K. [ed.] The Trojan War. Its historicity and context [Bristol 1984] 48Google Scholar). I Cyprus, LH IIIc pottery appears between two destruction levels conjectured to correspond to the Sea Peoples' two attacks on Egypt, which are nowadays dated c. 1209/8 and 1176/5; at Ugarit, on the other hand, where the destruction is associated with the second of these dates, the latest Mycenaean ware is LH IIIb (J. Mellaart in the same volume, 63–6).

⁷¹ J. B. Hainsworth in Foxhall and Davies 114; cf. Durante i 48. An early start is no guarantee of historical reliability. A survey published in the Sunday Times for 25 May 1980 revealed that, 35 years after the end of the Second World War, 34 per cent of the public thought that Dunkirk was a British victory; 18 per cent thought that the atom bomb was dropped by Britain, Russia, Germany, or Japan; Montgomery was identified by some as the British Prime Minister or an American soldier, Eisenhower as President of France or a Russian leader, Mussolini as a Russian general, a Czech communist leader, or a Jew.

⁷² They have been studied by Page (as n. 40) 251 f.; Bowra, , JHS lxxx (1960) 16–23CrossRefGoogle Scholar; Blegen, C. W., Troy and the Trojans (London 1963) 16–18Google Scholar; J. Pinsent in Foxhall and Davies (as n. 70) 137-62.

⁷³ Sch. A.R. i 139-448. Pherecydes reconciled the variants by making Thestor Idmon's son; Chamaileon fr. 17 Giordano did so by making Idmon a nickname of Thestor.

⁷⁴ See Chantraine, P., Grammaire homérique i³ (Paris 1958) 509–11Google Scholar; Hoekstra, A., Homeric modifications of formulaic prototypes (Amsterdam 1965) 147Google Scholar; P. Wathelet (as n. 43) 63-369; Durante i 24-38.

⁷⁵ Hooker, J., The language and text of the Lesbian poets (Innsbruck 1977) 70–82Google Scholar; Miller, D. G., Homer and the Ionian epic tradition (Innsbruck 1982) 23 f., 75 fF., 147 f.Google Scholar

⁷⁶ Lorimer (as n. 40) 458 f.; Page (as n. 40) 220 with 265 n. 6; Webster (as n. 57) 140 f., 153-5, 159-62; Miller (as n. 75) 22.

⁷⁷ Cauer (as n. 57) 158 f.; Chadwick, J., Greece & Rome iii (1956) 47Google Scholar; Ruijgh, C. J., Lingua xviii (1967) 96 f.Google Scholar; Wathclet (as n. 43) 104, 180 f., 290; Durante i 28, 34, 54; Janko, R., Homer, Hesiod and the Hymns (Cambridge 1982) 89 f.Google Scholar

⁷⁸ La Roche, J., Homerische Untersuchungen ii (Leipzig 1893) 164Google Scholar f.

⁷⁹ Wackernagel's derivation of the suffix -τειρα from Lesbian -τερρα < -τρια is uncertain; and if this change had been accomplished during the Aeolic phase of epic, we should expect to see Πέρραμος or (to preserve the original metre) Πέραμος, as in Sappho and Alcaeus, for how could Ionian poets, without an epic tradition of their own, have restored Πρίαμος? At most it is conceivable that ‘Eκτορι Πριαμίδηι, with Πρ failing to make position, reflects the initial stage of the Lesbian shift, i.e. “Eκτορι Πριαμίδάι or -μίωι (with metrical lengthening of r). Ionians would probably have interpreted Πριαμος as Πρίαμος. But more likely “Eκτορι Πριαμίδηι is an Ionian formula.

⁸⁰ Often explained as a Lesbian innovation, a conflation of -μεν with Ionic -ναι (Porzig). But (a) there are traces of the same ending in Doric areas, (b) the Lesbian restriction to athematic forms with monosylla bic roots agrees with the profile of Vedic -mane (Wackernagel).

⁸¹ Buck, C. D., The Greek Dialects (Chicago 1955) 102Google Scholar § 129.2, mentions a Boeotian ὅττω, which I cannot trace. Corinna PMG 655 i 6 has όττι, for what that is worth.

⁸² J. M. Cook, CAH ³ ii(2) 778.

⁸³ Troy VIIb was destroyed (by fire, like its predecessor) sometime between about 1080 and 1020, to go by the excavators' estimate of the duration of the settlement (Blegen, C. W. et al. , Troy iv (1) [Princeton 1958] 143Google Scholar) in conjunction with the revised date for the end of Troy VIIa (above, n. 70).

⁸⁴ Durante i 55 f. He might have mentioned that the specialized sense borne here by έξίημι is paralleled in Sappho 94.23 έξίης πόθον.

⁸⁵ Contrast the adaptation Tροίηι έν εύρείηι Il. xxiv 256 = 494, which is untraditional also in not placing τροι-(<τρωï-) in the biceps.

⁸⁶ I have discussed the problem of what this was in JHS ci (1981) 113-25.

⁸⁷ The massive three-stringed lyre shown on an early Geometric Cypriot amphora (Karageorghis, V. and des Gagniers, J., La Céramique chypriote de style figuré [Rome 1974] 97Google Scholar; Paquette, D., L'Instrument de musique dans la céramique de la Grèce antique [Paris 1984] 104Google Scholar f.) has more the structure of a cithara.

⁸⁸ See Wegner, M., Archaeologia Homerica III U: Musik und Tanz (Göttingen 1968) 25–7Google Scholar.

⁸⁹ Chr. Tsountas, 'Eφημ. 'Aρχ. iii (1892) 14 and pl. 3 no. 5; Lorimer (as n. 40) 456; Vermeule (as n. 10) 308, 312 fig. 49a; Wegner (as n. 88) 15 fig. 31, 27, 82 no. 139. It is not the number of strings that is significant so much as the structure of the frame and sound-box.—On the antiquity and affinities of the word φόρμιγξ cf. Durante i 159.

⁹⁰ It is true that Clement quotes Il. parva fr. 12 Allen in the form νύξ μὲν έην μεσάτα, λαμπρά δ' έπέτελλε σελάνα, but this conflicts with all other evidence for the fragment and for the poem. It is not convincing Aeolic in any case.

⁹¹ Chantraine (as n. 74) i 161-3; Wathelet (as n. 43) 154-7 and L'Antiquité classique 1 (1981) 819-33.

⁹² West, M. L., Studies in Greek elegy and iambus (Berlin and New York 1974) 113CrossRefGoogle Scholar f. In the later sixth and fifth centuries Attic correption becomes more frequent on both sides of the Aegean.

⁹³ Nothing can be argued from the fact that Euboea, or a part of it, had ρρ and ττ for ρσ and σσ in the classical period (Thumb, A.—Scherer, A., Handbuch der griechischen Dialekte ii [Heidelberg 1959] 264Google Scholar f.). The first is an innovation shared with Attic, the second is shared with Attic and Boeotian; neither need have been tenth- or ninth-century Euboean.

⁹⁴ Desborough (as n. 69) 206-15, 346 f.; Cold- stream, Geometric Greece 43, 45; Popham, M. R. and Sackett, L. H., Lejkandi i (London 1980) 360, 362.Google Scholar The statement about Lesbos is based on hearsay and I have not yet been able to obtain confirmation of it.

⁹⁵ Popham, M. R., Touloupa, E. and Sackett, L. H., Antiquity lvi (1982) 169–74CrossRefGoogle Scholar; Arch. Reports 1981/2. 15-17 and 1983/4. 12-15; first half of the tenth century.

⁹⁶ Desborough, V. R. d'A., Nicholls, R. V. and Popham, M. R., BSA lxv (1970) 21–30Google Scholar; Lefkandi i 168-70, 215-6, 344-5, 362, pi. 169 and frontispiece. Chiron's knee wound: Apollod. ii 85.

⁹⁷ We must infer, however, that references to warriors' burials were not such a firmly entrenched feature of older epic tradition as references to bronze weapons.

⁹⁸ For an (oriental) mythical component in the story of the Seven against Thebes cf. Burkert, W. in Brillante, C. et al. (ed.), I poemi epici rapsodici non omerici e la tradizione orale (Padova 1981) 29–51Google Scholar; Die orientalisierende Epoche (as n. 36) 99-106.

⁹⁹ Cf. Strunk, K., Die sogenannten Äolismen der homerischen Sprache (Diss. Köln 1957) 111Google Scholar, ‘ein Zusammenhang des Westgriechischen mit dem epischen έσσεïται ist noch nie behauptet worden; und eine solche direkte Beziehung ist wohl zu Recht nicht in Betracht gezogen worden, denn ein dorischer Einfluss auf das Epos ist nicht gut denkbar’; Wathelet (as n. 43) 51 n. 62, 287, 322, 368, al. I must except W. F. Wyatt, 'Eπιστημονικη 'Eπετηρίς Φιλοσοφ. ∑χολῆς τοῦ 'Aριστοτ. Πανεπιστ. Φεσσαλονίκης xiv (1975) 143, ‘scholars have not looked for Doric forms, and one or two do exist’.

¹⁰⁰ τέος and τεοῦς in Sophron, τεοῦ and τεῦς in Epicharmus, τίω and τίως in Rhinthon, τεῦς and τεοῦς in Theocritus, τεῦς and τεοῦς and probably τείν in Corinna.

¹⁰¹ Wackernagel, Sprachl. Unters. zu Homer 50-2, as he does not consider the possibility of W. Greek, is forced to assume that the text originally had the Aeolic forms. So too Chantraine, Gramm. hom. i 272.

¹⁰² Schwyzer, Griech. Gramm. i 786 f; Buck, Greek Dialects 115; Thumb, A.—Kieckers, E., Handbuch der griech. Dialekte i (Heidelberg 1932) 72Google Scholar f.—An unscrupulous person might add to the above list the adverbs χαμάδις, ἄμνδις and ἄλλνδις, as the suffix -δις is otherwise found only in οίκαδις Epicharm. 35.13, Ar. Ach. 742, 779 (the Megarian), and four other words cited as Doric by Herodian. But the υ of ἄμνδις and ἄλλνδις suggests an old Achaean origin; see Strunk (as n. 99) 121-4. Peloponnesian Doric -δις may be inherited from the Achaean substrate.

¹⁰³ M. L. West (as n. 67) 155-62.

¹⁰⁴ Ibid. 145 f., 149 f., 153.

¹⁰⁵ In reviewing Burkert (as n. 36), JHS cvi (1986) 234.

¹⁰⁶ See Stella, L. A., Il poema di Ulisse (Florence 1955) 107-23, 134-47, 157-68, 195–248Google Scholar; Tradizione micenea e poesia dell' Iliade (Rome 1978) 88, 362-8, 374-91; Webster (as n. 57) 69-76; Burkert (as n. 36) 106-10.

¹⁰⁷ See (e.g.) Gordon, C. H., AJA lvi (1952) 93CrossRefGoogle Scholar f.; Webster 82-8; McNeill, I., Anatol. Stud. xiii (1963) 238Google Scholar f.; Ullendorff, E., Bull. John Rylands Library xlvi (1963/1964) 242–247Google Scholar; Considine, P., Studi micenei ed egeoanatolici viii (1969) 85–159Google Scholar; Maróth, M., Ada antiqua xxiii (1975) 65–76Google Scholar; Stella, Trad. mic. e poesia dell'Il. 96, 122 f.; Miller (as n. 75) 16-21; Burkert, in Hägg, R. (ed.), The Greek Renaissance of the eighth century BC (Stockholm 1983) 51–6Google Scholar, and (as n. 36) 85-99.

¹⁰⁸ McNeill (as n. 107) 238; cf. Gordon, , AJA lvi (1952) 93CrossRefGoogle Scholar; Webster 74; Stella, Trad. mic. e poesia dell'Il. 363.

¹⁰⁹ Ullendorff (as n. 107) 243 f.; Brock, S. P., Vetus Testamentum xviii (1968) 395–7CrossRefGoogle Scholar; Maróth (as n. 107) 68.

¹¹⁰ McNeill (as n. 107) 239; Ullendorff (as n. 107) 246; Burkert (as n. 36) 107. Óðin is ‘father of all gods and men’ in Snorri, Gylfaginning 9, cf. 20.

¹¹¹ Watkins, C. in Mellink, M. J. (ed.), Troy and The Trojan War (Bryn Mawr 1986) 58–62Google Scholar.

¹¹² J. N. Coldstream, Geometric Greece 56, 66, 80, 100, and in Niemeyer, H. G. (ed.), Phönizier im Westen ( = Madrider Beiträge viii, Mainz 1982) 261–72Google Scholar.

¹¹³ Hesiod, Theogony (Oxford 1966) 28Google Scholar f., cf. Early Greek philosophy and the orient (Oxford 1971) 205Google Scholar; Hesiod, Works and Days (Oxford 1978) 27-30, 177Google Scholar; cf. Actes du X^e congrès de l'assoc. Guillaume Budé, Toulouse 8-12 avril 1978 (Paris 1980) 117 f.Google Scholar

¹¹⁴ Burkert in Hägg (as n. 107) 53-5, and (as n. 36) 86-92, 98.

¹¹⁵ Probably in its Assyrian form, where the narrative that included the ghost episode, originally a separate Sumerian poem, is incorporated with the main epic as Tablet xii.

¹¹⁶ Considine (as n. 107) 88—91, 147; Gresseth, G. K., CJ lxx (1975) 14 n. 24Google Scholar; Burkert (as n. 36) 92-5. Aphrodite, ‘the Cyprian’, is herself an oriental immigrant in the Homeric pantheon, probably post-Mycenaean. Cf. Burkert, , Greek Religion (Cambridge Mass. 1985) 152–4.Google Scholar Hers is one of the few Homeric names that involves Attic correption.

¹¹⁷ Burkert in Hägg (as n. 107) 53 f., and (as n. 36) 87-9.

¹¹⁸ In Niemeyer (as n. 112) 268. On the same page he refers to a gold diadem from the Teke treasure found in a tomb at Knossos, ‘perhaps made locally, but wholly oriental in style and theme: two Mesopotamian heroes, perhaps Gilgamesh and Enkidu, stand back to back as each one slays a lioness’ (cf. his pi. 26 f).