Sec. i] FROM TURFAN TO SINGER 725
be located,2Ö the conclusion imposes itself that in Li Tao-yüan's Mo-shan we have again a reference to the westernmost Kuruk-tâgh.
SECTION II.—TO P`O-CH`ÊNG-TZÛ AND SHINDI
On February 24th I left Singer for P`o-ch`êng-tzû, once a station on the route towards Toksun, the very name of which suggested the presence of ancient remains, of which I had also heard reports. Almost the whole of the day's march of about 27 miles led over the gravel glacis of the main range stretching west of Singer. But this fact exhibited all the more clearly the change in vegetation due to the moister climatic conditions in the hills to the south. Though our track gradually drew away from the boldly serrated range to the south known as Kizil-tâgh, scrub and tamarisk brushwood were to be found on almost every part of this Sai. Scattered groups of wild poplars were also met with until we crossed the deep-cut bed of a small stream coming from Tatlikbulak, ` the fresh spring ' (Map No. 29. A. I). From the higher level over which our route led it was easy to see that this belt of ampler vegetation extended right across to the foot of the long flat-topped hill chain on the north, which Muhammad Bagir knew as the Khawurga-tâgh. A gap in it to the north-west marked the point where the drainage from the main range breaks through it near the spring of Tunguzluk. After about 19 miles had been covered we passed a big isolated elm (kara-yagach) rising as a conspicuous route-mark. Seven miles farther on we struck the bed of a small ice-covered stream stretching down amidst luxuriant tamarisk jungle from the high massif of the Hsi-ta-shan, which was dimly visible to the south. Here we found the ruin of the Chinese station of P`o-ch`êng-tzii, abandoned years ago and now temporarily tenanted by a Turfânlik who with his three sons was working a lead pit among the foot-hills of the Hsi-ta-shan.
That the place had served in former times also for smelting operations was proved by a number of slag heaps and roughly constructed furnaces which we passed next morning when proceeding along the western edge of neglected fields to the ruins reported. They occupy the top of a small rocky ridge rising about 5o feet above the left bank of the stream and about half a mile to the north of the abandoned station. The remains are those of a small fort enclosing in the shape of an irregular polygon an area about 8o yards from east to west and half as much across. The walls, built of roughly cut slabs of clay (kisek), are about 3 feet thick and best preserved to the west and southwest, while to the east they have completely disappeared. There are remains of roughly built square towers at several of the corners and also near a gap in the northern face which seems to mark the position of a gate. Near the eastern edge of the little plateau pieces of lead ore and smelting refuse showed that the place had been occupied at a time when the lead mine in the hills to the south was being worked. No relics furnishing chronological evidence could be found ; but the general appearance of the ruined circumvallation suggested considerable age.
Though the stream of P`o-ch`êng-tzû, except at the time of rain floods, does not flow more than two or three miles farther, yet the wide bed it has cut could be made out clearly from the plateau as far as the previously mentioned gap near Tonguzluk. To the south the ravine (akin) in which the stream rises was seen to descend from the mouth of a well-marked valley, the Chong-aghiz, on the flank of the Hsi-ta-shan massif, and to be bordered by a broad belt of vegetation. A number of springs rising in this belt feed the stream, which below the ruined station carried at the time of my visit fully 22 cubic feet of water per second. This would permit of the cultivation of a considerably larger area than is irrigated at Singer. Yet the badly neglected fields of P`o-ch`êng-tzû, watered by a separate spring well above the level of the stream, cover only about three acres.
20 See below, Chap. xxi. sec. ii.