82 REMAINS OF KASHGAR, OASES OF YARKAND AND KARGHALIK [Chap. IV
approach to correct orientation. The same arrangement will be observed in the Strips of Endere and Rawak described below, while in the case of the only other two Stupas sufficiently well preserved for an exact restoration of the ground-plan, those of Topa-Tim and the Niya Site, the sides of the square base instead of the corners were found to be orientated 7. This difference in arrangement is of interest, but in view of the small number of examples it would scarcely be safe to draw from it any chronological or other argument.
The whole of the Stûpa was originally coated with a fairly hard plaster of whitish colour, which still adheres on the south-east face, as Plate I shows, to some portions of the base where projecting mouldings have offered protection. On the same side, which evidently had suffered least from ` weathering ', I also found in perfect preservation some of the woodwork employed to support the bold mouldings of plaster at the top and bottom of the drum. The tamarisk sticks used for this purpose were in numerous places still firmly embedded in the brickwork, and often had retained their covering of bark. Where the outer surfaces had lost their plaster coating, but were otherwise in good preservation, they displayed solid masonry of sun-dried bricks in regular horizontal layers, each separated from the other by a setting of clay about two inches in thickness. The bricks measured i 6 to i 7 inches square, with a thickness of 31i inches. The deep cuttings into the base and dome made on different levels from the north-west and south-west faces enabled me to ascertain that the same materials and the same manner of construction had been used throughout the Stûpa.
The fact that the stability of the whole structure has scarcely as yet been impaired by the great cutting effected on the south-west side, evidently long ago, is a proof of solid construction s. This cutting has laid bare a small shaft, 2 feet 8 inches square, running vertically through the centre of the Stûpa dome and drum. It may have continued below the latter into the base, but the wide cutting made there has destroyed all trace of it. Above, within the dome, this shaft widens into a little chamber, four feet square, which extends to the top of the extant masonry of the dome, and thus shows a height of 5â feet. No indication is left of the original height of this little chamber, nor of the manner in which it may have been closed from above.
It can scarcely be doubted that this central shaft or chamber was originally built to hold a relic deposit. It was found in the Stupas of the Niya Site, Endere, and Rawak, that is, in all of the structures of this class seen by me in Chinese Turkestan which were sufficiently well preserved to permit of detailed examination. Invariably I found this place of deposit in Stupas laid open by cuttings made in search of treasure. Nor can it surprise us that such conspicuous structures were thus systematically opened and ` explored ', probably at an early date, in view of the wholesale spoliation of ancient remains which has gone on in Eastern Turkestan during Muhammadan times, as attested for various localities, from Kashgar to Khotan, by curious records of Mirza Haidar s. At the Mauri-Tim Stûpa, as elsewhere, we can see that the hope of treasure had prompted operations even after the centre of the Stûpa had been reached, and
7 See Plates XIX, XXIX, XXXVII, XL. It will be noticed that, in the case of the Stupas of Mauri-Tim, Endere, Rawak, and the Niya Site, the plans of which were prepared on the spot and with the help of a properly orientated plane-table, the north to south line indicated by the bearings of the magnetic compass (a reliable instrument in the use of which proper allowance was made for the ascertained degree of magnetic variation) falls slightly to the west of the orientation line indicated by the plans of the structures themselves. I am at present unable to suggest a satisfactory explanation
for this deviation, which by its regular recurrence invites attention. It cannot be due to magnetic variation, as the latter was at the time of my journey for Eastern Turkestan about 4° 3o' to the East of the true North.
8 This cutting is clearly seen in the view reproduced in Fig. 13.
a See Tarikh-i-Rashidi, pp. 255 sqq., for the extensive excavations for treasure carried on under Mirza Aba Bakr's orders at ancient sites of his dominions ; also below chap. VIII. sec. 5.