of burden, and Indian dogs, in indefinite multitude, increasing the consumption of the regular army. [53] The army which Darius had conducted against Scythia is said to have been counted by divisions of ten thousand each, but the process is not described in detail (Herodot. iv, 87). [54] Herodot. vii, 60, 87, 184. This same rude mode of enumeration was employed by Darius Codomannus a century and a half afterwards, before he marched his army to the field of Issus (Quintus Curtius, iii, 2, 3, p. 24, Mutzel). [55] Herodot. vii, 89-97. [56] Herodot. vii, 185-186. ἐπάγων πάντα τὸν ἠῷον στρατὸν ἐκ τῆς Ἀσίης (vii, 157). “Vires Orientis et ultima secum Bactra ferens,” to use the language of Virgil about Antony at Actium. To admit this overwhelming total, or anything near to it, is obviously impossible: yet the disparaging remarks which it has drawn down upon Herodotus are noway merited.[57] He takes pains to distinguish that which informants told him, from that which he merely guessed. His description of the review at Doriskus is so detailed, that he had evidently conversed with persons who were present at it, and had learned the separate totals promulgated by the enumerators,—infantry, cavalry, and ships of war, great and small. As to the number of triremes, his statement seems beneath the truth, as we may judge from the contemporary authority of Æschylus, who in the “Persæ” gives the exact number of twelve hundred and seven Persian ships as having fought at Salamis: but between Doriskus and Salamis, Herodotus[58] has himself enumerated six hundred and forty-seven ships as lost or destroyed, and only one hundred and twenty as added. No exaggeration, therefore, can well be suspected in this statement, which would imply about two hundred and seventy-six thousand as the number of the crews, though there is here a confusion or omission in the narrative which we cannot clear up. But the aggregate of three thousand smaller ships, and still more, that of one million seven hundred thousand infantry, are far less trustworthy. There would be little or no motive for the enumerators to be exact, and every motive for them to exaggerate,—an immense nominal total would be no less pleasing to the army than to the monarch himself,—so that the military total of land-force and ships’ crews, which Herodotus gives as two million six hundred and forty-one thousand on the arrival at Thermopylæ, may be dismissed as unwarranted and incredible. And the computation whereby he determines the amount of non-military persons present, as equal or more than equal to the military, is founded upon suppositions noway admissible; for though in a Grecian well-appointed army it was customary to reckon one light-armed soldier, or attendant, for every hoplite, no such estimate can be applied to the Persian host. A few grandees and leaders might be richly provided with attendants of various kinds, but the great mass of the army would have none at all.