^{page 109 note 1} e.g. in N 803 (cf. 157) Hector's shield is called π⋯ντοσ’ ⋯ἰσην; since Reichel thought it behaved like an indubitably Minoan shield, he forced the epithet to mean ‘überall angemessen, passend’ (Homerische Wqffen, 2nd ed. (1901), p. 17). In H 219 the shield of Aias is ἠὐτε π⋯ργον and in 267 it is struck μ⋯σσον ⋯πομϕ⋯λιον; Kunze, Emil (Kretische Bronzereliefs (1931), p. 60)Google Scholar therefore searches for examples of a central ⋯μφαγ⋯ς on a shield which is not round. The tacit assumption is general, cf. Persson, A. W. in his discussion of the all-bronze helmet (New Tombs at Dendra (1942), p. 124)Google Scholar.

^{page 109 note 2} Short discussion with reference to main diverworks by Chantraine, P. in Mazon's, p.Introduction à l'lliade (1942), pp. 115 ffGoogle Scholar. For this note the works of Milman Parry are most important. For authors with divergent views on dating cf. Nilsson, M. P., ‘Der homerische Dichter in der homerischen Welt (Die Antike,xiv (1938), pp. 22 ff.)Google Scholar and Carpenter, Rhys, Folk Tale, Fiction and Saga in the Homeric Epics (1946), pp. 6 ffGoogle Scholar. See also Lorimer, H. L., ‘Homer's Use of the Past’ (J.H.S. xlix (1929), pp. 145 ff.)Google Scholar: but this is the least part of my debt to Miss Lorimer.

^{page 109 note 3} πολεφλο⋯σβοιο: A 34, B 209, Z 347, I 182, N 798, ψ 59, v 85, 220. πολι⋯ν (-⋯ς): Δ 248, ζ 272, λ 75, χ 385.εὐρυπ⋯ροιο: O 381, δ 432, ν 2. ἠχ⋯εσσα: A 157. γλαυκ⋯: π 34. ⋯τρυγ⋯τ¿ιο: Ξ 204. In ε 413 ⋯γχιβαθ⋯ς and in η) 273 ⋯θ⋯σφατον are predicative. All references are to the Oxford Classical Text, and since a small variation in the number of examples would not affect the argument no account has been taken of divergent readings.

^{page 109 note 4} Ξ 16, γ 179, 321.

^{page 109 note 5} εὐρ⋯Ϊ (-α): Z 291, I 72, α 197,β 295, δ 498, 552, μ 293, 401, ω 118. ⋯πε⋯ρονα: A 350, δ 510. ¿ἲνοπι (-α): B 613, E 771, H 88, ψ 143, 316, α 183, β 421, γ 286, δ 474, ε 132, 221, 349, ζ 170, η 250, μ 388, τ 172, 274. ἠεροειδ⋯Ϊ (-α):β 263, γ 105, 294, δ 482, ε 164, 281, θ 568, μ 285, ν 150, 176. ἲοειδ⋯ος (-α): Λ 298, ε 56, λ 107. με⋯λανι: Ω 79. πολυκλ⋯στῳ: δ 354, ζ 204, τ 277. κυμα⋯νοντα: Ξ 229, δ 425, 510, 570, ε 352, λ 253. μεγακ⋯τεα: γ 158. ἰχθυ⋯εντι (-α): I 4, I 746, π 378, δ 381, 390, 424, 470, 516, ε 420, ι 83, κ 458, 540, ψ317. ⋯τρὐγετον: O 27, β 370, ε 84, 140, 158, η 79, ν 419; ρ 289. ⋯πε⋯ρτος in κ 195 is predicative.

^{page 110 note 1} πολιήν (-οίο, etc.): A 350, 359, M 284, N 352, 682, ξ 31, O 190, 619, T 267, Y 229, Ф 59, φ 374, β 261, ς 405, 580, ε 410, ι 104, 132, 180, 472, 564, μ 147, 180, φ 236. μαρμαρέην: ξ 273. ποϕυυρ⋯ην Π 391. βαθεῖαν (-ης): A 532, N 44. πολυβενθέος: δ 406. ⋯τρυγ⋯τοιο: A 316, 327, Ω 752, α 72, ε 52, ς 226, θ 49, κ 179. δίαν: A 141, B 152, ξ 76, O 161,177, 223, Ф 219, y 153, δ 577, ε 261, θ 34, λ 2.

^{page 110 note 2} ε 101.

^{page 111 note 1} οῖνοφ is distinguished from εὐρύς by the digamma and àλμυρ⋯ν ύδωρ is needed when water swallowed or exgurgitated.

^{page 112 note 1} But not with a conventional effect. Sometimes a traditional phrase receives new life from an untraditional context. àλ⋯ς ⋯τρυγ⋯τοιο is a regular ending. Yet in the simile of the cor-morant (ε 51–3) noun and adjective might have been carefully selected and combined for the first time in their history. The name ‘Homer’ is used without prejudice for the poet or poets who gave the poems approximately their final order.

^{page 112 note 2} The 9 other forms are found 27 times in all.

^{page 117 note 1} One traditional association of epithet and noun need mean no more than that the epithet was at some time appropriate to the whole category, as the traditional description of Peleus as ίππήλατα does not imply that the other heroes were unskilled charioteers but that skill in driving chariots was characteristic of all heroes.

^{page 117 note 2} The verbal association does not mean that Homer envisaged the κυν⋯η as a particular type of helmet: he clearly did not. The argument is therefore not affected by the appearance of ϕ⋯λαρ' εῢποίηθ' on a πήληξ in Π 106 or by the phrase κ⋯ρυθος ϕ⋯λον.

^{page 118 note 1} The epithet is appropriate to (a) the unique bronze helmet from Dendra, dated by Persson not later than the first half of II, L.H. (op. cit., pp. 119–26)Google Scholar, or (b) the ‘Corinthian’ helmet(Lorimer, H. L., ‘Defensive Armour in Homer’, Liverpool Annals of Archaeology and Anthropology, xv (1928), p. 123)Google Scholar. Reminiscences of objects which disappeared before the end of the Mycenaean period are not impossible, but it seems more probable that an epithet was formed by analogy to describe the characteristic feature of a new type of helmet.

^{page 118 note 2} Lorimer, H. L., op. cit., pp. 118 ff.Google Scholar; Wace, A. J. B., ‘Chamber Tombs at Mycenae’, Atchaeo-logia, vol. lxxxii (1932), pp. 212 ff.Google Scholar; Persson, , op. cit., pp. 126 ffGoogle Scholar.

^{page 119 note 1} σ⋯κος θ⋯το τετραθ⋯λυμνον, though listed among the ‘individual’ epithets, is almost certainly an obsolescent traditional phrase, προθελύμνω is odd; elsewhere it means ‘by the roots’, and there is no real evidence for the meaning ‘overlapping’ in N 130, in a couplet which contains several linguistic and tactical peculiarities. Per-haps it was used in error simply because its form suggested that it was a doublet of τετραθ⋯λυμνος, on the analogy of ⋯μϕίϕαλος and τετρ⋯ϕαλος. The simple adjectives μ⋯γα and δειν⋯ν, found six times and never with ⋯σπίς, may possibly have had a verbal association in the tradition with σ⋯κος.

^{page 119 note 2} χ 184–6. In the Doloneia, where the exceptional κυν⋯αι are described, Diomedes receives his shield in a succinct καί σχκος, Odysseus has none, and Dolon wraps himself in a wolf-skin.

^{page 119 note 3} Λ 32–40: the ⋯σπίς of Agamemnon has ten concentric circles of bronze, a central boss, and twenty subsidiary bosses; also a gorgoneion (cf.Payne, Humfry, Necrocorinthia, p. 79)Google Scholar surrounded by Δείμος and Фóβος. Θ 192: the ⋯σπίς of Nestor is famous for being all of gold. M 294–7: the ⋯σπίς of Idomeneus has a face of hammered bronze and, within, layers of hide sewn together with circular rows of gold stitches or rivets, N 406–7: the ⋯σπίς of Idomeneus is fashioned out of hides and bronze so as to be δινωτήν (with a circular face or with circular plates or with a spiral pattern?). N 803: Hector's—quoted in the text. The shield of Aias, which alone Achilles could consider borrowing (Σ 193), and the shield which Hephaistos made for Achilles have a real The individuality which is part of the story; but they resemble the others in structure and epithets except that, as they are never called ⋯σπίδες, they never have the epithets which go with dxrnts only. H 219–23: the shield of Aias is made of seven layers of hide faced with an eighth of bronze. Σ 478 ff.: the shield of Achilles has five layers (according to Y 269–72, athetized by Aristarchos, they are of bronze, gold, and tin with the gold in the middle). On the metal face is an inlaid pattern in concentric circles. It has a shining triple rim. It would be a curious understatement to describe such shields as these simply as strong or made of layers of hide or circular or bossed.

^{page 120 note 1} In M 402 and Υ 281 -ōū; in arsis could stand before a vowel. See e.g. Σ 390, 499 and -η, -Υ in formulae in E 856 and Y 474, 480. But it is possible that ⋯μϕιβρότης and εῢκῢκλου came into the tradition as a convenient metrical doublet.

^{page 120 note 2} The alternative that Homer invented an episode which happened to be precisely what might have occurred several centuries earlier does not seem probable.

^{page 121 note 1} The consistent grouping of epithets round ⋯σπίς and σάκος is against assuming a single type which combined the attributes of both.

^{page 121 note 2} It is discussed by Helbig, in ‘Ein homerischer Rundschild mit einem Bgel’ (Jhrh. des Oesterr. Arch. Inst., Bd. xii (1909), Heft i)Google Scholar, and by H. L. Lorimer and Emil Kunze in the works already quoted.