With the treatment of the body of Leonidas (cf. ix. 78) we may compare that of the corpse of Amasis by Cambyses (iii. 16). Artaxerxes similarly maltreated that of his brother, the younger Cyrus (Xen. Anab. i. 10. 1; iii. 1. 17), and the Parthian general or Suren, that of M. Crassus (Plut. Crass. 32). Though the story is in marked contrast with the generous treatment of Pytheas (vii. 181; viii. 92), these parallels make it unlikely that it is a Greek invention as contended by Wecklein (Ber. Bayer. Akad. (1876), p. 285) and Duncker (vii. 258).

The whole chapter is with reason regarded as an interpolation by Krger, followed by Abicht, Gomperz, Van Herwerden, and Macan. There is no proper transition to Bk. VIII, a fact which leads Stein to suspect an omission in the text. The formula introducing the [p. 234] digression is strange, but as Macan points out, this anecdote is not intended to be a resumption of the main thread of the story but to supply an omission, and the words mean I will here return to a place in the story where before I was guilty of an omission. The author is excusing himself for putting in the story here, where the only ground for its appearance is its connexion with Demaratus, instead of in ch. 220, where it was required to explain how the Spartans had early information of the intended Persian invasion. Krger also regards as suspicious the postponement of the actual story in favour of a disquisition on Demaratus' motives and the author's assertion that the motive was ill-will, and subsequent willingness to leave it an open question. Such hesitation, however, may be easily paralleled from the genuine work of H. (cf. ii. 123. 1; v. 45. 2). Krger's arguments from language are stronger. The asyndeton ἐπύθοντο is intolerable, τὸ ἐς Δελφοὺς χρηστήριον is hardly justified by (ii. 150) τὴν Σύρτιν τὴν ἐς Λιβύην; δελτίον δίπτυχον is queer Greek, as δίπτυχα in this sense is late, and elsewhere H. uses δέλτος (viii. 135); ἐπέτηξε and ἐκκνᾶν do not reappear till Aen. Tact. ch. 31, nor συμμάχεται (middle) till Xenophon, or ὁδοφύλαξ till Eustathius. It may be said that some of these strange words are quoted by Pollux (Onom. x. 58) from H., and that the story, though without names, goes back at least to (350 B. C.) Aeneas Tacticus (l. c.), but these stylistic peculiarities and late words surely betray a forger. The existence of an anonymous version of the story in Aeneas, and a variant in Trogus Pompeius (Justin, ii. 10. 12-17), in which a sister of Leonidas figures, and Demaratus' motive is patriotic, really discredit the story, as suggesting that the narrative as here given is a gradual and relatively late fabrication (Macan). It is inferior to the similar stories of Harpagus (i. 123. 3, 4) and Histiaeus (v. 35) on which it may have been modelled. Finally, the extremely unfavourable impression given of Demaratus seems un-Herodotean. It appears highly probable that some part of the text connecting Books VII and VIII was early lost, and into the gap this chapter was thrust by an interpolator. Even if it be a genuine fragment it is misplaced here.

BOOK VIII

The story of Artemisium. 1-3 The fleet and its leadership. 4, 5 Bribery of Themistocles. 6-11 First engagement. 12, 13 Storm, wreck of Persian squadron off Euboea. 14-17 Second and third days' fighting. 18-23 Retreat of the Greek and advance of the Persian fleet.

Throughout there is a close parallel and connexion between the accounts of the operations on sea and on land and of the forces at Thermopylae and Artemisium. First, we have the description of [p. 235] the double position (vii. 175-7) supplemented by a more detailed topography of Thermopylae (vii. 198-201), secondly the story of the movements of the fleets (vii. 179-95) and a brief account of the march of Xerxes' army (vii. 196-7), finally a narrative of the struggle at Thermopylae (vii. 202-33) and of the contemporary (ch. 15) seafights at Artemisium (viii. 1-23). Yet, as it stands, the opening of Book VIII is abrupt and not connected with the end of Book VII. Probably the connexion and implied contrast between the land and sea forces has been obscured by the later insertion of ch. 234-9 (Macan).

The summary of the Greek forces here given is parallel to that prefixed to the fighting at Thermopylae (vii. 202). Similar lists are given of the Greek fleets before Lade (vi. 8 n.) and before Salamis (viii. 43 f.), and of the Greek army before Plataea (ix. 28). But whereas at Plataea H. follows the line of battle from right to left (as at Lade from east to west), and at Salamis adopts a geographical order (Peloponnese, northern Greece, islands, &c.), here he arranges the states according to the number of ships furnished (ch. 2. 1), thus incidentally justifying the claim of Athens to command at sea (ch. 3).

ne/as. Triremes, line-of-battle ships, excluding not only transports (vii. 97) but even penteconters, which are not included in the total by H. (viii. 2. 1, 48), though Diodorus (xi. 12) carelessly speaks of 280 triremes.

For the Plataeans cf. vi. 108.

For the Athenian cleruchs in Chalcis cf. v. 77. 2 n.