chapter one will do well to hesitate before denying the possibility of

such distant expeditions in early times. But any communication which may have existed must have been interrupted by the invasions of the Bithynoi and Treres, probably in the ninth and eighth centuries. Note may also be taken here of what is said about the Cimmerioi in Od. XI 14-19; cf. Meyer, _Geschichte des Alterthums_, II, pp. 367 f., 445 f.

[301] Cf. especially Wilamowitz-Möllendorff, _Hom. Unt._ p. 24 ff. The theory that the Ephyre of II 328 must be a different place from the Ephyre of I 259 seems to me very problematical if the author of the second book had only a vague knowledge of the geography of western Greece. Again, if Ilos Mermerides (I 259) is taken from the story of the Argo, is it really necessary that the source should be a different one from that referred to in XII 69 ff.?

[302] From this it has been argued that the 'Odyssey' known to Hesiod must have differed greatly from the poem which has come down to us. But it is to be remembered that there are quite as noticeable discrepancies between the Odyssey and the Iliad.

[303] We shall have occasion later to notice more than one point in which this passage departs from the customary Homeric standards.

[304] Except probably in XIV 9 ff., as has been ingeniously pointed out by Mr Lang (_Homer and his Age_, p. 276 ff.). In the same chapter Mr Lang shows that several features in the Doloneia which have been interpreted as marks of lateness may very well be due to the peculiar circumstances of the situation.

[305] Gemöll, _Hermes_, XV 557 ff. (cf. XVIII 308 ff.). Cf. also Shewan, _Class. Quarterly_, IV 73 ff., where this view is rejected.

[306] Mr Lang (_Homer and his Age_, p. 265 ff.) has called attention to the fact that in v. 261 ff. Odysseus is represented as wearing a cap of a type which appears to have been in use during the Mycenean age. If it could be shown that the article in question was peculiar to that period, the lateness of the book would certainly be open to serious question.

[307] Bronze is mentioned 279 times in the Iliad and 80 times in the Odyssey. In a large proportion of these cases the reference is to weapons. Cf. Helbig, _Das homerische Epos_, p. 329 ff.

[308] Iron is mentioned altogether 48 times. In nine cases it is spoken of merely as a substance--a possession or article of trade. To these we may add fifteen more in which the word is used metaphorically as a standard of hardness, etc., and one (Od. IX 393) which refers to the testing of iron in water. Iron tools or implements are mentioned thirteen times, apart from the two references to knives given above. We hear also of iron chains (Od. I 204), the iron axle-tree of a (divine) chariot-wheel (Il. V 723), and the iron door of Tartarus (_ib._ VIII 15); cf. Cauer, _Grundfragen der Homerkritik_², p. 281 ff.

[309] Cf. Macalister, _Palestine Expl. Fund, Quart. Rep._, 1903, p. 199; Lang, _Class. Rev._ XXII, p. 47.

[310] Cf. Dörpfeld, _Troja und Ilion_, p. 368.

[311] αὐτὸς γὰρ ἐφέλκεται ἄνδρα σίδηρος. Cf. Burrows, _The Discoveries in Crete_, p. 214 ff.

[312] Cf. Ridgeway, _The Early Age of Greece_, p. 294 f. Prof. Ridgeway allows the occasional use of bronze swords, e.g. in the case of Euryalos the Phaeacian (Od. VIII 403-6). But the swords of Paris, Patroclos, Achilles and Odysseus (Il. III 334 f., XVI 135 f., XIX 372 f., Od. X 261 f.) are described in very similar terms (ξίφος ἀργυρόηλον χάλκεον). Further, the tendency of bronze swords to snap off short at the hilt is well illustrated by the case of Lycon in Il. XVI 338 f.

[313] Cf. Lang, _Homer and his Age_, p. 192 f.

[314] Cf. Reichel, _Über homerische Waffen_ (Abh. d. arch.-epigr. Seminares d. Univ. Wien, Heft XI), p. 79 ff.

[315] Cf. p. 191 and Evans, _Journal of the Anthropological Institute_, XXX 213 f.

[316] Cf. Ridgeway, _op. cit._ pp. 324 f., 475.

[317] This strap seems to have been used for carrying even comparatively small shields down to a much later period. It is not found apparently in the representations of the Shardina, though they have an arm-strap as well as a handle.

[318] The Shardina on the temple of Medinet Habu and the warriors represented on the Stele hold their spears poised in their right hands, precisely at the same angle. But it is not quite clear to me whether a cast or thrust is intended.

[319] φέρων σάκος ἠΰτε πύργον (Il. XII 219, etc.). There is a reference no doubt to the hero's great stature.

[320] Very recently the history of Greek shields has been treated at length by Dr G. Lippold (_Münchener Archäologische Studien_, pp. 399-504). This work is largely taken up with a criticism of Reichel's theories, and in the course of the discussion it is pointed out that the latter are in many points insufficiently supported by evidence. Dr Lippold (pp. 406, 474) seems to have no hesitation in assigning the Warrior Vase and its congeners to the late Mycenean age--he does not distinguish between 'Mycenean' and 'sub-Mycenean'--and he also recognises (p. 461 ff.) that two kinds of shields figure in the Homeric poems. The 'tower' shield however is identified by him with the Dipylon shield, from which he believes the 'Boeotian' shield to be descended. He holds that the round shield was of Oriental origin, since it was used by the Assyrians in the ninth century, and that it was first introduced into Greece towards the end of the Mycenean period; then, after being banished for a while from the Greek mainland by the Dipylon shield, it was re-introduced, in a somewhat modified form, towards the end of the Dipylon (Geometrical) period. The Homeric poems are held to reflect the time of transition when it was re-introduced; but no date appears to be given except that it was before the eighth century (p. 468). This explanation seems to me to be open to a serious objection, namely that the Homeric shields will then have to reflect a different age from that indicated by the Homeric evidence on the use of the metals; for the latter clearly belongs to the close of the Mycenean--or rather 'sub-Mycenean'--period. So far as I can see, it is only by Dr Mackenzie's equation of the Homeric poems with the Warrior Vase and certain East Cretan graves (cf. p. 185) that we can obtain a consistent and intelligible sequence. Of course it may very well be that the round shield was banished for a time from the Greek mainland by the Dipylon type. On the other hand the suggestion that the former was of Assyrian origin surely requires evidence earlier than the ninth century; for we find it used by the Shardina, who cannot properly be regarded as Oriental, as far back as the thirteenth century. I have to thank Mr A. B. Cook for calling my attention to Dr Lippold's work.

[321] Cf. Reichel, _op. cit._ p. 53 f.

[322] Long shields were regularly used during the La Tène period by the Celtic peoples, and also by many of the Teutonic peoples probably much later. But they seem to have been of a totally different type from the Mycenean. In late times they were certainly of great length (cf. Diodorus, V 30, and the figures on the bowl of Gundestrup); but the earliest examples (e.g. the oval shields depicted on the Hallstatt sword-sheath) may really be modifications of the round shield.

[323] Cf. Ridgeway, _The Origin and Influence of the Thoroughbred Horse_, p. 510. A similar case occurs in a painting representing the victory of Rameses II over the Hittites, figured in Meyer's _Geschichte des alten Aegyptens_ (plate following p. 290).

[324] Cauer, _Grundfragen d. Homerkritik_², p. 294 ff.

[325] It is uncertain whether the sanctuaries of Apollo at Pytho (Delphoi) and of Poseidon at Scheria (Il. IX 404 f., Od. VIII 79 ff., VI 266 f.) are regarded as temples; cf. Cauer, _op. cit._ p. 301 f. In the former case however it is decidedly probable.

[326] Many well-known survivals of such usage occur in Greece itself.

[327] Adam Brem. IV 27: _corpora autem suspenduntur in lucum qui proximus est templo. is enim lucus tam sacer est gentilibus ut singulae arbores eius ex morte uel tabo immolatorum diuinae credantur._ Cf. also the (contemporary) schol. 134: _prope illud templum est arbor maxima late ramos extendens, semper uiridis in hieme et aestate, cuius illa generis sit nemo scit. ibi etiam est fons ubi sacrificia paganorum solent exerceri et homo uiuus immergi. qui dum non inuenitur ratum erit uotum populi._

[328]

Σπερχεί', ἄλλως σοί γε πατὴρ ἠρήσατο Πηλεὺς κεῖσέ με νοστήσαντα φίλην ἐς πατρίδα γαῖαν σοί τε κόμην κερέειν ῥέξειν θ' ἱερὴν ἑκατόμβην, πεντήκοντα δ' ἔνορχα παρ' αὐτόθι μῆλ' ἱερεύσειν ἐς πηγὰς ὅθι τοι τέμενος βωμός τε θυήεις.

[329] The Appendices in Miss Stawell's _Homer and the Iliad_ (pp. 238-326) suggest that many of the instances commonly cited are due to insufficient consideration.

[330] Fragm. 189, Boeckh.

[331] Cf. Mimnermos, fragm. 9. 5 f. (Bergk), Herodotus, I 150, etc. According to Pausanias (V 8. 7) Smyrna had become Ionic before the year 688; cf. Wilamowitz-Möllendorff, _S.-B. der Akad. der Wiss. zu Berlin_, 1906, p. 52, note.

[332] Acquaintance with the district round Smyrna is shown by the reference to the figure of 'Niobe' on Mt Sipylos in Il. XXIV 614 ff., although the identification of this figure is still disputed.

[333] On the problems connected with this name see Allen, _Classical Quarterly_, I 135 ff.

[334] Stephanus Byzant. s. v. Βολισσός.

[335] But not the corresponding change of _an_ to _ai_ (cf. τάς, πᾶσα).

[336] Cf. Wilamowitz-Möllendorff, _op. cit._ p. 52 f.

[337] The Ionic states of Clazomenai and Phocaia, to the west and north-west of Smyrna, seem to have been founded at a comparatively late period, though probably in the eighth century. Since the promontory of Ἄργεννον opposite Chios and to the south-west of Erythrai, has an Aeolic name, it is possible that the whole of the coast north of Teos was once occupied by Aeolians.

[338] Cf. Wilamowitz-Möllendorff, _op. cit._ p. 75 and (for a criticism) Cauer, _Grundfragen_², p. 181 ff.

[339] The Shield of Heracles contains a number of clearly Aeolic forms (ἄμμες, ὔμμι, etc.).

[340] Cf. Fick, _Die homerische Odyssee_ (1883), and _Die homerische Ilias_ (1886), where the poems are reconstructed in their original Aeolic form.

[341] E.g. the two texts of Riddle XXXVI (both printed in Sweet's _Oldest English Texts_, p. 150 f.) and the texts of Caedmon's Hymn from the Moore MS. of Bede's Eccles. History (_ib._ p. 149) and the Anglo-Saxon version (IV 24). Reference may also be made to the Dream of the Cross and the extracts given in the inscription on the Ruthwell Cross. A portion of the Old Saxon Genesis is printed, together with the Anglo-Saxon version, in Cook and Tinker's _Translations from Old English Poetry_, p. 184 f.

[342]

πέμπε δέ μιν Λυκίηνδε πόρεν δ' ὅ γε σήματα λυγρὰ, γράψας ἐν πίνακι πτυκτῷ θυμοφθόρα πολλὰ, δεῖξαι δ' ἠνώγειν ᾧ πενθερῷ ὄφρ' ἀπόλοιτο.

[343] The analysis of the Odyssey given by Prof. v. Wilamowitz-Möllendorff admittedly postulates a written text (_Hom. Untersuch._ p. 293). But I cannot assent to the proposition that the Catalogue of Ships in itself must come from a written source. This list scarcely differs in principle from the catalogues of Widsith.

[344] Further indirect evidence, of native origin, is supplied by the word _facenstafas_ (O. Norse _feiknstafir_) in v. 1018, if the original meaning of this compound was 'harmful runes' (used magically). Cf. also vv. 317, 382, 458, 1753.

[345] It may be added that we really know nothing of the Ionic of Hesiod's time. It is quite uncertain how far it had already developed those characteristics which we find in our texts.

[346] The few exceptional forms such as (τ') οὖλον (Od. XVII 343) may be due to fairly late scribes familiar with (eastern) Ionic texts.

[347] We may note also the absence of the literary Ionic forms κότε, κῶς, etc.

[348] It is true that we do not know exactly when _h-_ was lost. But before that change took place Η cannot have been used for _ē_; consequently a wholesale μεταγραμματισμός would be involved (doubtless also affecting the representation of _ō_), just as in the case of Athens.

[349] Cf. Hoffmann, _Griech. Dial._ III, p. 549 f.; Fick, _Neue Jahrbücher_, I 504 ff.

[350] Cauer, _Delectus Inscr. Graec._ 83, 84, 91. We may compare also such epic expressions as ευρυχορο, γαιας απο πατριδος, πολυμελο, κλεϝος απθιτον (_ib._ 54, 83, 445, 202), and, more particularly, an inscription on a bronze discus from Cephallenia (cf. Cook, _Class. Review_, XIII 77 f.):

Εχσοιδα μ ανεθεκε Διϝος ϙοροιν μεγαλοιο χαλκεον hοι νικασε Κεφαλανας μεγαθυμος

(cf. Hom. Hymn. XXXIII 9; Il. II 631).

[351] The same remark is true of heroic names occurring in Doric inscriptions, e.g. Ϝεκαβα, Κεβριονας, Δαιφοβος on a vase (Cauer, _op. cit._ 78) found near Caere. So also with the heroic names used by Pindar and other non-Ionic poets--not to mention the Latin forms. Yet these poets use the Ionic forms of foreign names, such as Μῆδοι (Cypr. _Ma-to-i_), which had come to them presumably through Ionic channels. On the other hand we find in inscriptions on Chalcidian vases more purely Ionic forms, e.g. Αινεες (_ib._ 545), than those preserved in our text. These seem to count against any place except Athens as the home of the final form of the 'epic dialect.'

[352] From the fact that Pindar and other non-Ionic authors use what is apparently an Ionic form--indeed, strictly speaking, a western Ionic form--in the poet's name (Ὅμηρος), while they give the names of the heroes themselves in non-Ionic form, we are justified in concluding that they had acquired the former from a different (presumably literary) source. Certainly the earliest references to the poet come from Ionic authors. Again, Thucydides (III 104) is clearly recording a generally accepted opinion when he quotes the Hymn to the Delian Apollo under Homer's name; and I can see no reason for doubting the identity of Semonides' Χῖος ἀνήρ (cf. p. 209) with the author of this poem (v. 172: τυφλὸς ἀνὴρ, οἰκεῖ δὲ Χίῳ ἔνι παιπαλοέσσῃ). The Hymn dates probably from the period when Chios was in process of becoming Ionicised. At such a time the repertoire of a Chian minstrel would have an exceptionally favourable opportunity of gaining currency (naturally under his own name) in Ionic circles--in the Cyclades probably as well as in Ionia itself.

[353] Cf. Reichel, _op. cit._ p. 59: "Das Epos schildert, wie in allen Dingen, auch hier die ältere Prachtzeit," and pp. 63, 102 f., where the first appearance of the round shield (of which the knowledge is granted, p. 55 ff.) is referred to the middle of the eighth century, and that of breast-plates to about the beginning of the seventh century.

[354] Prof. v. Wilamowitz-Möllendorff holds that the editor was a person of inferior ability and that the poem as a whole is not a success; but this view is scarcely in accordance with the generally received opinion.

[355] It is worth noting that these objections apply even to what are commonly regarded as among the latest parts of the poems. Thus in Od. XXIV 305, where Odysseus describes himself as υἱὸς Ἀφείδαντος Πολυπημονίδαο ἄνακτος, the point is entirely spoilt by the Ionic form. That the true form should be Aeolic (-παμμον-) is rendered more than probable by such names as Ἁλιθέρσης, Πολυθερσεΐδη--of which at least the second likewise belongs to the 'later' portions of the poem.