6x8 AT RUINED SITES OF TURFAN [Chap. XVIII
03. a (Pl. LXXXII), block-printed with a floral pattern of good naturalistic design, deserves to be specially mentioned.
Frs. of wall- Brief reference still remains to be made to a small cave-shrine on the right bank of the stream,
paintings v (see plan, Pl. 25). It is approached from above over a very steep rock slope at a point about 300 from cave-
shrine yards below Toy. iv and, owing to the difficulty of access, I was unable to reach it with my injured
Toy. v. leg. Attention was called to it by some fresco pieces (Toy. 067-8) brought for sale by a local man.
On examination by my assistants, the walls of the passage around the central rock-carved pillar, which still retained an image base in its front recess, were found completely stripped of their paintings. But numerous small broken pieces of painted plaster littered the sand-covered floor, sad evidence of the vandal destruction which had been wrought here, probably for the sake of selling a few carelessly extracted panels.
Cave-shrine The same fate had been suffered by the paintings that once decorated the walls of a rock-cut
Toy. vi. shrine, Toyuk vi, situated above the left bank of the Toyuk stream at a distance of about six
hundred yards above the Maz5.r of Yetti-kalandar and close to where the road leading up the gorge makes its steep drop down to the stream. This shrine, as the sketch-plan in Pl. 28 shows, consisted of an antechapel, a little over 20 feet wide and probably about as long, and a cella, measuring 10 feet by II feet 4 inches, approached from the former by a passage 6 feet wide and about 51- feet deep. The antechapel had its sides faced with brick walls which once carried a vaulted roof ; it was found completely ruined and contained but little debris. But the floor of the cella retained a layer of sand about 3 feet deep, which had preserved, not only the lowest portion of a stuccoed image base, but also a considerable number of fine fresco fragments as well as a few remains of stucco relievos. According to statements made by the villagers the plastered surfaces of the walls bearing paintings were torn down a long time ago by men searching the cave for timber. Some support for this came to light in the shape of a rough splinter of wood, about two feet long, evidently reduced from a larger piece of timber and retaining on its smooth surface remains of Uigur writing in two columns. There were also visible in the side walls of the cella shallow groove-like recesses cut into the rock, which evidently had served for the insertion of a wooden framework probably meant to strengthen the plastering ; such support may have been all the more needed as the cella walls were not cut vertically from the rock but, as the section in Pl. 28 shows, sloped slightly inwards as they rose.
Frs. of wall- Among the fresco pieces recovered, only the smaller fragments have so far been set up and
paintings examined, and these, though in many instances showing interesting details, do not furnish adequate from
Toy. vi. clues to the general character of the decorative scheme. Referring for details to the Descriptive
List, I may confine myself here to mentioning Toy. vi. 03,051 from a representation of the goddess Hâriti with her babes ; 02, 015, 032, 035, o68 as showing robes figured with lions, birds, &c. ; o6, o66, 073 with figures of suppliants. Besides fragmentary inscriptions in Uigur (029, 033, 071), we have others in Brâhmi script (031, 039, 049, 091). Of special interest is the treatment of ` high lights ' observable in 052, 065, where an originally white or other light colour appears to have oxidized to black. A number of the fragments show signs of having suffered from smoke, no doubt at some time when the cave-shrine, conveniently near to the road, was used as a habitation. Within the cella was found also a well-carved stone block, i6 inches square and 7 inches high, bearing on its top the roughly carved representation of an open lotus in high relief. The original purpose served by it is not clear. Small fragments of Chinese manuscript rolls, some with Uigur or Brahmi writing on the reverse, were also discovered when the cella was cleared.
But the chief object of interest that had survived in this shrine is the painted ceiling of the flat-shaped dome over the cella (Fig. 3 i 3). This showed, as far as preserved, two rows of carefully