BATTI,E OF SALAMS.- RETREAT OF XERXES. 129 These measures were all taken during the night, to prevent the anticipated flight of the Greeks, and then to attack them in the narrow strait close on their own harbor the next morning. He had previously stated Phalemm as the main station of the Persian fleet: not necessarily meaning that the whole of it was there. The pas- sage which I have just transcribed intimated what the Persians did to ac complish their purpose of surrounding the Greeks in the harbor of Salamis and the first part of it, wherein he speaks of the western (more properlj northwestern) wing, presents no extraordinaiy diiSculty, though we do not know how far the western wing extended before the movement was commenced. Probably it extended to the harbor of Peirseus, and began from thence its night-movement along the Attic coast to get beyond the tovm of Salamis. But the second part of the passage is not easy to com- prehend, whei'e he states that, " those who were stationed about Keos and Kynosura also moved, and beset with their ships the whole strait as far as Munychia." What places are Keos and Kynosura, and where were they sit- uated 1 The only knowm places of those names, are the island of Keos, not far south of cape Sunium in Attica, — and the promontory Kynosura, on the northeastern coast of Attica, immediately north of the bay of Marathon. It seems hardly possible to suppose that Herodotus meant this latter prom- ontory, which would be too distant to render the movement which he de- scribes at all practicable : even the island of Keos is somewhat open to the same objection, though not in so great a degree, of being too distant. Hence Barthelemy, Kruse. Bahr, and Dr. Thirlwall, apply the names Keos and Kynosura to two promontories (the southernmost and the south- easternmost) of the island of Salamis, and luepert has realized their idea in his newly published maps. But in the first place, no authority is pro- duced for giving these names to two promontories in the island, and the critics only do it because they say it is necessary to secure a reasonable meaning to this passage of Herodotus. In the next place, if we admit their supposition, we must suppose that, before this nifjht-movement commenced, the Persian fleet v.-as already stationed in part ofi" the island of Salamis : which appears to me highly improbable. AVhatever station that fleet occupied before the night-movement, we may be very sure that it was not upon an island then possessed by the enemy : it was somewhere on the coast of Attica : and the names Keos and Kynosura must belong to some unknovu points in Attica, not in Salamis. I cannot therefore adopt the supposition of these critics, though on the other hand Larcher is not satisfactoiy in his attempt to remove the objections which apply to the supposition of Keos and Kj-nosura as commonly understood. It is difficult in this case to reconcile the statement of Herodotus with geographical considerations, and I rather suspect that on this occasion the historian has been himself misled by too great a desire to find the oracle of Bakis tiody fulfilled. It is from Bakis that he copies the name Kynosura (viii, 77). VOL. V. 6* 9oc.
