Sec. iii] SURVEYS OF EARLIER REMAINS AT ENDERE . 285
walls of the small fortified post south, in single courses between intervening layers of stamped clay, these layers varying here in thickness from seven to twelve inches. Near the northern wall there were traces of a wall built of timber and wattle.
About a quarter of a mile to the south-west I found the remains of a small dwelling on a plateau
rising over a stretch of eroded open ground. Only one little room, measuring eight by ten feet, could still be made out ; its wattled walls showed a core of diagonal reed matting and survived to a height of about three feet under the protection of drift sand. There were no finds of any kind
at either ruin. Following a small depression northward for about a quarter of a mile, Mihmân took
me to a place where human bones lay exposed on the slopes of two witnesses'. At the one to the south I found a broken skull and other parts of a skeleton about six feet below the top, and near them two well-preserved earthen pots which, until disclosed by some recent ` prospecting ', had evidently rested in the ground. Both were of coarse clay, hand-made, and resembled in shape those subsequently dug up at the Kara-dong site. The larger one measured I z inches in height to the lip of the high neck, 8 inches in diameter where widest, and 4 inches across at the mouth. Its shoulder was decorated with three bands of incised double lines, separated by a plain zigzag diaper. The neck showed incised wave-lines. Though too rough for approximate dating this ornamentation looked to me distinctly old. The other pot was quite plain, measuring seven inches in height and about as much in diameter.
No ancient remains of any sort were met with on the march of about five and a half miles westwards which brought us to the grazing-ground of Korgach on the Endere River (Map No. 4o).
The way lay through an area of closely set tamarisk-cones of moderate height, which became distinctly lower as the present bed was approached. A visit paid from Korgach, on the morning of November 13, to the scanty remains which Mihmân knew of west of the river, completed my archaeological work in this neighbourhood. The' site ' of an old water-mill, which Prof. Huntington appears to have visited,° was first traced about half a mile to the north-west of camp, without difficulty, since a shallow iistang or canal, about five to six feet broad, helped to show the way. Some roughly carved Toghrak beams and the end of a dug-out trunk which had served for a conduit carrying the water to a point above the wheel, were all that was left of this ` ruin '. Neither their condition nor the appearance of the dry canal pointed to any great age.
Mihmân declared that he had not visited the other remains for the last twenty years, and the
search for them proved very troublesome and protracted owing to the closeness of the luxuriant tamarisk thickets. In this riverine jungle we first came upon a dry water channel, cut to a depth of about fifteen feet and measuring about twenty feet across (see Map No. 4o). I t was said to take off a short distance above Korgach, and its general direction to the north-west could be made out with certainty in spite of its many windings. Mihmân told me that he had followed it right through to the abandoned fort-village beyond Bilél-konghan. He, as well as the other local men with me, took the bed unhesitatingly for that of a canal. The remains, when located at last at a direct distance of about two and a half miles, proved to consist of the badly decayed débris of some dwelling built with Toghraks and showing far-advanced erosion. The clearing was not rewarded by any ' find' which might help to fix the date of the structure. But the discovery of wheat straw embedded in the mud
9 See Pulse of Asia, p. 217.