Sec. y] FROM MOJI TO THE KHOTAN OASIS 117
above for the bricks of the Mauri-Tim Stûpa. It is probable that the slightly better preservation of the north-western side of the mound is due to the same cause which we have already had occasion to discuss in connexion with the Topa-Tim ruin ; namely, accumulation of sand on the side most exposed to the prevailing winds. Whether a great débris-filled gap between the solid masonry remains on the south side (see plan) was due to excavation or to some natural cause of decay, I was unable to ascertain. As a special feature I may mention that large stones rounded by the action of water were noticed by me embedded as a regular layer in the brickwork some five feet above the ground near the south-west corner. The ruin rests on hard gravel which, however, is exposed to view only on the south and east. Elsewhere coarse drift-sand covers the ground, rising immediately to the west and north in long dunes up to a height of fifteen feet or so. This sand is absolutely sterile, forming a striking contrast to the dark line of trees of the Piâlma oasis visible far away on the horizon. The ground on the east side of the mound, being clear of sand, was found strewn with small potsherds, 'but to no great distance ; no ornamented or otherwise defined pieces were traced among them.
From Karakir-Tim to the western edge of the Piâlma oasis, a distance of close on six miles to the south-east, I could nowhere trace signs of cultivation, old or recent ; considering the character of the soil, which is hard pebble Dasht ', close up to the flood-bed of the Duwa stream west of the oasis, it is difficult to believe that cultivation was ever carried on here. The isolated position of the ruined Stûpa thus appears distinctly curious. Nor could any other remains be traced near it. Piâlma itself is a small oasis, counting, according to my information, only about one hundred households, and dependent for its water supply on a stream which comes from the mountains of Duwa. As the drainage area can scarcely include any mountains with permanent snow-beds, the supply of water available for irrigation is uncertain and often scanty2.
Pialma is now the last settlement passed by the route outside the western limit of the Khotan district, the present administrative frontier between Karghalik and . Khotan being marked by two half-decayed pillars on the road some miles from the edge of the Pialma oasis. In Hsüan-tsang's time Pialma itself must have been reckoned within the territory of Khotan ;
for there can be little doubt that the town of P`o-ch`ieh-i (Po-kia-i) dip , which, according
to the ' Life,' the pilgrim first reached after entering the frontier of the kingdom, and where he made a halt of seven days before proceeding to the capital, is to be looked for near the present Pialma 3. From the Hsi-pi-chi we learn that Hsüan-tsang placed this locality at a distance of 300 li, or three marches, to the west of the Khotan capital4, and this reckoning, according to the daily stages usual at the present day, takes us to Piâlma. The distance from Pialma to Yotkan, about 48 miles by my map, say 56 miles actual measurement, allowing for small détours and the heavy sand encountered on part of the route, agrees with this. If a resemblance in the sounds of Chinese transcriptions can be trusted, Pialma or Po-ch`ieh-i, might possibly be
meant also by P`o-hai % , which the above-quoted itinerary of the Tang Annals mentions
as the second locality to the west of Yü-t`ien b.
' For details about the Pialma oasis, see Hedin, Reisen in Z.-A., pp. 17 sq.
3 See Vie de Hiouen-Thsang, pp. 279 sqq.—My previous suggestion as to the possible location of P o-ch'ieh-i (Po-kia-i) at Moji, Ruins of Khotan, p. r 9r, was based on an inadequate consideration of the distance and of the position indicated by Hsüan-tsang's ' Life'.
4 See Memoires, ii. pp. 23o sq. ; Si-yu-ki, transl. Beal, ii. p. 314. Beal's version and that of Rémusat, Ville de Khotan, p. 45 seem to be more exact than that of Julien, who makes Hsüan-tsang arrive at P o-ch`ieh-i (Po-kza-i) from the capital, though in reality he travelled in the opposite direction.
, Comp. Chavannes, Turcs occid., p..r23, note I.