452 KARA-DONG AND THE SEARCH
K. 002. j. Lip or foot of glass vessel, fragment. Lustreless. Edge, which is intact, rolled over and tubular. 21" xi". See Pl. LI.
K. 002. k. Thin yellowish glass ; small fragment.
K. 002. 1. Terra-cotta whorl. See Pl. LI.
K. 002. m. Bronze ball, pierced. Perhaps used as a whorl. Diam. g" nearly.
K. 002. n. Iron arrow-head. Conical with long thin tang from centre of base. Length of head i", diam. ". Length of tang 2/", diam. 1". See Pl. LI.
K. 002. o. Sandal-wood comb ; parabolic-shaped. The teeth, which are rather fine, extend from square end to
FOR HSÜAN-TSANG'S P`I-MO [Chap. XIII
a length of about r1". They are well formed and retain considerable elasticity. The pattern of this comb is the same as that now common in Northern India. Length 3" ; width at square end 21"; thickness at rounded end I".
Handle of terra-cotta vessel ; fragment. Upper surface ornamented, three rows of small impressed rings. Clay very impure. Width râ", length IA".
a. Fragment of coarse terra-cotta vessel, having a rapidly-drawn wavy or nebule pattern upon it.
K. 004. b. Handle of coarse terra-cotta vessel ; fragment, ornamented with three rows of small incised rings. 11" x r i".
SECTION II.=HSÜAN-TSANG'S NOTICE OF PI-MO AND MARCO POLO'S PEIN
My survey and excavations at Kara-dong were completed on March t7 in the midst of a. sandstorm such as had greeted our arrival. Though the force of the wind, this time from the south-west, was somewhat less, the driving sand made the conclusion of the work decidedly trying both to my men and to myself. Next morning I left this desolate spot, just as I had reached it, in an atmosphere thick with dust, and oppressive by its haziness. My eyes were now turned to the south again, where a number of archaeological tasks still awaited me in the vicinity of the inhabited area.
First among them was a search for the ancient town of Pei-mo. Hsüan-tsang had visited it on his way from Khotan to Ni fang or Niya, and its probable identity with Marco Polo's Pein, first suggested by Sir Henry Yule, made me all the more anxious to determine, if possible, its position. The distance and direction which the Chinese pilgrim's narrative records for Pei-mo, and which, together with other archaeological indications we shall presently have occasion to discuss in detail, had long before made me look out for the place somewhere to the north of the small oases extending from Chira to Keriya. I had been anxious to search for it when leaving Dandan-Uiliq by marching due south through the desert. But the absence of local information and practical considerations connected with the condition of my men and animals prevented the execution of this plan. I was hence much pleased when, on my second visit to Keriya, I heard direct from Huang-Daloi, the kindly Amban, of two ` kbne-shahrs' which had been reported to him to exist in the desert beyond Gulakhma, an oasis on the Khotan road some forty miles to the north-west of Keriya. Ram Singh, too, when marching along this route in December, had heard of old remains in that direction. In order to save time I now decided to reach them, if possible, direct from the Keriya river by striking across the desert south-westwards.
During the four days which saw us returning along the Keriya river as fast as camels and ponies could be got to move, I was surprised to notice that the water in the river had fallen by some 3 ft. as compared with the level of the preceding week. Thus the first spring flood had passed by ; yet in the vegetation of the riverine jungle I still looked in vain for any sign of approaching spring. Passing again on March 22 the familiar shrine of Burhanuddin-Padshahim, I picked up en route the two men who, under the Amban's orders, had been sent by the Bag of Gulakhma to guide me to the sites. They looked unusually reticent and stupid, but not till
Location of Hsüantsang's Fl-mo.
March from Keriya river to Shivul Darya,