means of testing these numbers.  Figures in the mouth of an Oriental are vague and almost unmeaning; armies are never really counted:  there is no such thing as a fixed and definite “strength” of a division or a battalion.  Herodotus tells us that a rough attempt at numbering the infantry of the host was made on this occasion; but it was of so rude and primitive a description that little dependence can be placed on the results obtained by it.  Ten thousand men were counted, and were made to stand close together; a line was then drawn round them, and a wall built on the line to the height of a man’s waist; within the enclosure thus made all the troops in turn entered, and each time that the enclosure appeared to be full, ten thousand were supposed to be within it.  Estimated in this way, the infantry was regarded as amounting to 1,700,000.  It is clear that such mode of counting was of the roughest kind, and might lead to gross exaggeration.  Each commander would wish his troops to be thought more numerous than they really were, and would cause the enclosure to appear full when several thousands more might still have found room within it.  Nevertheless there would be limits beyond which exaggeration could not go; and if Xerxes was made to believe that the land force which he took with him into Europe amounted to nearly two millions of men, it is scarcely doubtful but that it must have exceeded one million.

The motley composition of such a host has been described in a former chapter.  Each nation was armed and equipped after its own fashion, and served in a body, often under a distinct commander.  The army marched through Asia in a single column, which was not, however, continuous, but was broken into three portions.  The first portion consisted of the baggage animals and about half of the contingents of the nations; the second was composed wholly of native Persians, who preceded and followed the emblems of religion and the king; the third was made up of the remaining national contingents.  The king himself rode alternately in a chariot and in a litter.  He was preceded immediately by ten sacred horses, and a sacred chariot drawn by eight milk-white steeds.  Round him and about him were the choicest troops of the whole army, twelve thousand horse and the same number of foot, all Persians, and those too not taken at random, but selected carefully from the whole mass of the native soldiery.  Among them seem to have been the famous “Immortals”—­a picked body of 10,000 footmen, always maintained at exactly the same number, and thence deriving their appellation.

The line of march from Sardis to Abydos was only partially along the shore.  The army probably descended the valley of the Hermus nearly to its mouth, and then struck northward into the Caicus vale, crossing which it held on its way, with Mount Kara-dagh (Cane) on the left, across the Atarnean plain, and along the coast to Adramytium (Adramyti) and Antandros, whence it again struck inland, and, crossing the ridge