PROPOSAL OF ARLEUS FOR RETREAT. 35 promise of alliance, engaged also to guide the Greeks, in good faith, down to the Asiatic coast. Klearchus immediately began to ask what route he proposed to take ; whether to return by that along which they had come up, or by any other. To this Ariasus re- plied, that the road along which they had marched was impracti- cable for retreat, from the utter want of provisions through seven- teen days of desert ; but that he intended to choose another road, which, though longer, would be sufficiently productive to furnish them with provisions. There was, however, a necessity (he added), that the first two or three days' marches should be of extreme length, in order that they might get out of the reach of the king's forces, who would hardly be able to overtake them afterwards with any considerable numbers. They had now come ninety-three days' march 1 from Ephesu^ or ninety from Sardis. 2 The distance from Sardis to Kunaxa is, according to colonel Chesney, about twelve hundred and sixty-five geographical miles, or fourteen hundred and sixty-four English, miles. There had been at least ninety-six days of rest, enjoyed ait various places, so that the total of tune elapsed must have at least been one hundred and eighty-nine days, or a little more than half a year ; 3 but it was probably greater, since some intervals of rest are not specified in number of days. How to retrace their steps, was now the problem, apparently insoluble. As to the military force of Persia in the field, indeed, not merely the easy victory at Kunaxa, but still more the un- disputed march throughout so long a space, left them no serious 1 Such is the sum total stated by Xenophon himself (Anab. ii, 1,6). It is greater, by nine days, than the sum total which we should obtain by adding together the separate days' march specified by Xenophon from Sardis. But the distance from Sardis to Ephesus, as we know from Herodotus, was three days' journey (Herod, v, 55); and therefore the discrepancy is really only to the amount of six, not of nine. See Kriiger ad Anabas. p. 556 ; Koch, Zug der Z. p. 141. 2 Colonel Chesney (Euphrates and Tigris, c. ii, p. 208) calculates twelve hundred and sixty-five geographical miles from Sardis to Kunaxa or the Mounds of Mohammed. 3 For example, we are not told how long they rested at Pyla;, or opposite to Charmande. I have given some grounds (in the preceding chapter) for believing that it cannot have been less than five days. The army must bave been in the utmost need of repose, as well as of provisions.