But bad as the Uxian pass had been, there remained another still worse—called the Susian or Persian gates,[413] in the mountains which surrounded the plain of Persepolis, the centre of Persis proper. Ariobarzanes, satrap of the province, held this pass; a narrow defile walled across, with mountain positions on both sides, from whence the defenders, while out of reach themselves, could shower down missiles upon an approaching enemy. After four days of march, Alexander reached on the fifth day the Susian Gates; which, inexpugnable as they seemed, he attacked on the ensuing morning. In spite of all the courage of his soldiers, however, he sustained loss without damaging his enemy, and was obliged to return to his camp. He was informed that there was no other track by which this difficult pass could be turned; but there was a long circuitous march of many days whereby it might be evaded, and another entrance found into the plain of Persepolis. To recede from any enterprise as impracticable, was a humiliation which Alexander had never yet endured. On farther inquiry, a Lykian captive, who had been for many years tending sheep as a slave on the mountains, acquainted him with the existence of a track known only to himself, whereby he might come on the flank of Ariobarzanes. Leaving Kraterus in command of the camp, with orders to attack the pass in front, when he should hear the trumpet give signal—Alexander marched forth at night at the head of a light detachment, under the guidance of the Lykian. He had to surmount incredible hardship and difficulty—the more so as it was mid-winter, and the mountain was covered with snow; yet such were the efforts of his soldiers and the rapidity of his movements, that he surprised all the Persian outposts, and came upon Ariobarzanes altogether unprepared. Attacked as they were at the same time by Kraterus also, the troops of the satrap were forced to abandon the Gates, and were for the most part cut to pieces. Many perished in their flight among the rocks and precipices; the satrap himself being one of a few that escaped.[414] [413] See the instructive notes of Mützel—on Quintus Curtius, v. 10, 3; and v. 12, 17, discussing the topography of this region, in so far as it is known from modern travellers. He supposes the Susian Gates to have been near Kala-Sefid, west of the plain of Merdasht or Persepolis. Herein he dissents from Ritter, apparently on good grounds, as far as an opinion can be formed. [414] Arrian, iii. 18, 1-14; Curtius, v. 4, 10-20; Diodor. xvii. 68.