1264 RUINED SITES EAST AND NORTH OF KHOTAN [Chap. XXXI
Numismatic already been of old date when Sung Yün visited Han-mo. While among the coins actually picked up
evidence of there on my own visit or R. B. Lai Singh's preceding reconnaissance one is a Sung dynasty piece of early occu-
pation. A. D. 11o2-7 and another a Muhammadan coin ascribed to the fourteenth century, there were
secured also some much-worn Wu-chu pieces and one coin which seems to be of the Sino-Kharosthi class.6 These finds of early coins are of special interest because they bear out Sung Yün's statement with regard to the antiquity of the shrine with the miracle-working Buddha statue which he visited 15 li to the south of Han-mo, i. e. at the site now marked by the pilgrimage place of Ulûgh-mazâr. He mentions that of the multitude of votive banners he saw there about half showed dates of the Wei dynasty (A. D. 386-534), while one was of the period A. D.
During my halt of one day at Ultigh-mazâr I was able to make an excursion northward to the nearest portion of the Uzuti-tati débris area, and in the course of it found the remains of a roughly built structure, undoubtedly of Muhammadan origin and probably a mosque, about half a mile to the north of the ruined fort described in Ancient Khotan.7 As other traces of later occupation were found in the immediate vicinity, the view previously expressed as to the late origin of the fort has been strengthened. It may well date from an attempt made to reoccupy this ground long after Marco Polo's ` Pein ' was abandoned to the desert.
Ulitgh-mazâr was the last place which saw my whole party reunited at the close of our winter campaign (Fig. 320). On March 22 I moved south to the oasis of Chira, and from there Naik Ram Singh, my ` handy-man ', started eastwards in the company of Ibrahim Bag on the long journey which was to take him back to the temples of Miran—and from which he returned blind three months later.
Chira, reckoned at no less than 3,500 households, is a large and flourishing tract with cultivation dependent mainly on the abundant ak-su which its river brings down straight from the slopes of the great glacier-crowned massif of the Muz-tagh Peak. Here, too, the extension of the irrigated area was proceeding rapidly, the annual increase being reported at fully a thousand Chinese mou. Of the way in which this recent reclamation of land, centuries ago abandoned to the desert, is bound to affect any old remains there surviving, I had a good illustration on visiting the ruins known as Rawak, about a mile to the west of Oghrelik, where my camp stood. There, in the midst of new fields, I found the walls of a relatively large structure, built of stamped clay and now crumbling under the effect of the heavy flooding which the ground all round had received from a recent canal extension. The extant portions of the structure comprised a hall measuring 48 feet by 37, with three rooms of considerable size adjoining at its back. There was no clear indication
Ruined fort north of Uliughmazar.
6 See below, Appendix B. I may add that of five coins which were given to me at Kenya as having been brought from the ' Tatis ' near Uldgh-mazâr two are Tang pieces (K'ai yüan), while the rest are of Sung dynasty issues, ranging from A.D. 1017-I107. Of nine more coins received from Domoko in 1 go6 five are Wu-chu ' cash', tentatively ascribed to the fifth century, three Tang pieces of A.D. 758-9, and one probably a mediaeval Muhammadan coin.
6 Cf. Chavannes, Voyage de Song Yun, pp. 14 sq. ; above, p. 841.
7 Cf. Ancient Kholan, i. p. 462. In Map No. 27. D. 4 the position of the fort has by an error of compilation been shown to the north-north-east, instead of east-north-east of Ulûgh-mazar. The position of the débris area marked north of the fort must be shifted accordingly. The direction of the dry river-bed reached further on and connected with
the present Ponak-akin needs also correction, as it has a more northerly bearing.
[The correct positions of these features relative to Ulughmazâr, Old Domoko, etc., will be found duly indicated in Sheet No. 14 of the final Atlas containing all surveys made on my three journeys, 1900-15.]
This dry bed is very clearly marked with an average width of about 6o yards, while the ridges of close-set tamarisk-cones on either side leave clear a depression fully three times as wide. Judging from the appearance of the dead kötek lining the banks and the ' Yardangs ' within the bed, the latter must be of considerable antiquity. The possibility of its having once carried the waters from the united streams between Domoko and Chira towards Dandân-oilik, as first suggested by Professor Huntington, deserves careful consideration.