142 fflSTORY OF GREECE. horse as well as foot, together with select detachments of the remaining contingents : making in all, according to Herodotus, three hundred thousand men. But as it was now the beginning of September, and as sixty thousand out of his forces, under Artabazus, were destined to escort Xerxes himself to the Helle- spont, Mardonius proposed to winter in Thessaly, and to postpone farther military operations until the ensuing spring.! Having left most of these troops under the orders of Mardonius in Thessaly, Xerxes marched away with the rest to the Helle- spont, by the same road as he had taken in his advance a few months before. Respecting his retreat, a plentiful stock of stories were circulated,^ — inconsistent with each other, fanciful and » Herodot. viii, 114-126. '■^ The account given by ^schylus of this retiring march appeai-s to me exaggerated, and in several points incredible (Persse, 482-513). That they suffered greatly during the march from want of provisions, is doubtless tnie, and that many of them died of hunger. But we must consider in deduction : 1. That this march took place in the months of October and November, therefore not very long after the hai-vest. 2. That Mardonius maintained a large ai'my in Thessaly all the winter, and brought them out in fighting condition in the spring. 3. That Artabazus also, ■with another large division, was in military operation in Thrace all the winter, after having escorted Xerxes into safety. When we consider these facts, it will seem that the statements of ^schy- lus, even as to the sufferings by famine, must be taken with great allow- ance. But his statement about the passage of the Strymon appears to me incredible, and I regret to find myself on this point differing from Dr. Thirlwall, who considers it an undoubted fact. (Hist. Greece, ch. xv, p. 351, 2d ed.) •' The river had been frozen in the night hard enough to bear those who arrived first. But the ice suddenly gave way under the morning sun, and numbers perished in the waters," — so Dr. Thirlwall states, after -.Eschylus, — adding, in a note, " It is a little surprising that Herodotus, when he is describing the miseries of the retreat, does not notice this disas- ter, which is so prominent in the narrative of the Persian messenger in ^schylus. There can, however, be no doubt as to the fact : and perhaps il may furnish a useful warning, not to lay too much stress on the silence of Herodotus, as a ground for rejecting even important and interesting facts which are only mentioned by later writers," etc. That a large river, such as the Strj'mon, near its mouth (180 yards broad, and in latitude about N. 40° 50'), at a period which could not have been later than the beginning of November, should have been frozen over in one night so hardly and firmly as to admit of a portion of the army marching over it at daybreak, before the sun became warm, — is a statement which
