^{page 213 note 1} The rapidity with which this coast has advanced renders it probable that what was done in 279 B.C. would have been an easy exploit in 480, cf. Fausanias x. 21, 4.

^{page 213 note 2} Cf. vii. 175: and especially the passage in viii. 15 beginning to the end of the chapter: cf. also viii. 21.

^{page 214 note 1} Cf. the mistake made as to the defensive nature of the position at Tempe: also, ignorance of the existence of the path of Hydarnes at Thermopylae: also, the absence of all mention in Herodotus, and, inferentially, in Herodotus' sources of information, of the pass on the Lamia-Delphi road through Oeta via Cytinium, in the account of Thermopylae, and this, too, although its use by the Persians immediately after the battle is almost certainly implied (viii. 31).

^{page 215 note 1} Cf. vii. 206.

^{page 215 note 2} That the existence of the path was a matter of purely local knowledge cf. vii. 214, and that those possessing that knowledge were not present when the decision to make a stand at Thermopylae was arrived at, cf. viii. 203 especially

^{page 215 note 3} Cf. vii. 206.

^{page 216 note 1} An ingenious explanation has been put forward for what is thought by some to be the inexplicable withdrawal to Chalkis (vii. 183). It is suggested that this refers really to the sending of fifty-three Athenian vessels to guard the south part of the strait, and that these vessels returned to Artemisium after the wreck of the 200 in the Hollows of Euboea.

About this theory more need not now be said than that it demands a dislocation of the whole story. The result is a narrative consistent, indeed, with itself, but wholly at variance with Herodotus. It is only by a close examination of the Herodotean version of the story that we can judge whether so complete a reconstruction is demanded.

^{page 218 note 1} I make this suggestion of cause with the greatest diffidence, but with the conviction that Herodotus' own evidence suggests it. It seems to me that calculations from the statements of Herodotus render it improbable that the Greek fleet could have started from Artemisium before the storm broke. At the same time I do not wish to give the impression, which would be a wrong one, that I look upon the cause stated by Herodotus as being impossible. Considering the state of feeling then prevalent in a section of the fleet, it would be dangerous to assert that any excuse, however specious, might not have been seized upon as an argument for a retreat. But here, as in the long account of Plataea, Herodotus' statements of causes are not of the same credibility as his statements of facts, especially when there is present anything that might enhance or detract from the moral which he intends to convey.

^{page 218 note 2} At 4 a.m. in the first week in August it used to be pitch dark out on the Bay of Navarino. In the evening it would be bright daylight at 6.45 p.m., and quite dark at 7 p.m.

^{page 219 note 1} This fact, the fire signals on Skiathos in an earlier chapter, and the rapidity with which the news of the loss of the 200 ships in the Hollows of Euboea reached the Greek fleet point, as it would seem, to a regularly organised signalling system.

^{page 220 note 1} This promontory is, if I remember, a low one, and this perhaps tells against the identification of it with Artemisium.

^{page 221 note 1} Cf. Strabo 445

^{page 223 note 1} A recent criticism of Herodotus' story comments on the absence of ‘motive’ in Herodotus' account of the arrival of the fifty-three Athenian vessels. It suggests that the retreat to Chalkis in the early part of the narrative is to he explained as having been in reality the despatch of these fifty-three from Artemisium to that place with a view to defending the south entrance of the strait.

Does it seem probable that the Peloponnesian contingent would have consented to remain at Artemisium under such circumstances, and to entrust the re-opening of the seriously threatened line of communications to a purely Athenian squadron? for there was, it must be remembered, according to this theory, no Persian fleet as yet at Aphetae to render retreat from Artemisium dangerous.

In the absence of ‘motive’ given, the most probable which can be suggested is that the Athenians, at this time novices in the fitting out of large fleets, had not been able to make more than 147 of their vessels ready for sea in time for the despatch to Artemisium, and sent on the remaining fifty-three when they were ready.

^{page 224 note 1} Cf. after the arrival of the fifty-three.

^{page 229 note 1} Cf. viii. 15.