Gulakhma the two men, Turdi and Kurban of Domoko, from whom he had obtained his share of the find, subsequently passed on to Badruddin Khan, and they readily guided me to the spot. It proved to be a patch of wind-eroded ground about 150 yards across, covered with pottery debris. It was situated amidst closely set tamarisk-cones and about a quarter of a mile to the north of the track that leads from the northern edge of the Domoko oasis to the village of Achma eastwards. Reference to our survey of 1906 indicated that this small Tati, called Kuduk-köl by my guides, lay a little over a mile to the east of Darabzan-dong and about the same distance to the south of Kök-jigda, both small ruined sites that I had visited in 1906 from Khadalik.$
A low shapeless mound on the eastern edge of this area and close to the foot of a tamarisk-cone MS. leaves rising steeply to about 3o feet in height (Fig. 96) marked the ruin from which Turdi and Kurbau from
while collecting fuel last winter, had extracted a packet of manuscript leaves, evidently those of a Pôthi. The complete destruction of the shrine, for such it soon proved to have been, was attributed to the burrowing which Mullah Khwaja had carried on some months later in the hope of securing more khais for his Amban at Keriya. On clearing the small mound composed of sand and debris we reached the plastered floor at a depth of about four feet. What had survived of the square walls, built as at Khadalik of timber and plaster, had been levelled to the ground by Mullah Khwaja's operations, except on the east where the sand sliding down from the tamarisk-cone had afforded some protection. Enough remained there of the foot of the wall to show that the cella of the shrine, or possibly its enclosing passage, had measured 21 feet square.
The numerous fragments of plaster painted in tempera which were found amidst the debris Remains of
left no doubt that the walls had been decorated with compositions in the style of the mural paintings shrine of which I had recovered remains in 1906 from the ruin of the main temple of Khadalik.9 The fragments brought away (Kuduk-köl. oI I-17, 020-8), though most of them have suffered badly by exposure, show workmanship in design and colour-treatment quite equal, if not superior, to that of the Khadalik frescoes. Among them I may single out for mention the representation of a miniature Stiipa (025) with pennons, and the head of a Trimûrti figure (014) closely recalling that on the Dandan-oilik panel D. vii. 6.10 That diapers of small Buddha figures seated in rows prevailed in the decorative scheme, just as at Khadalik and Dandan-oilik, is proved by the fragments 024, 026. Among other relics recovered are the fragments of a finely carved wooden halo showing portions of small Buddha figures (oi, Pl. IX) ; two elaborately turned wooden finials (03, 04) ; some stucco relief fragments (05-7, Pl. IX), and the painted panel, o8, badly effaced but still showing the figures of three horsemen riding to the right proper. The pose and dress of the middle one suggest that the same personage is depicted who appears in the Dandan-oilik panel D. vii. 5.11 Several small Brâhmi manuscript fragments (Kuduk-köl. 029-30) recovered in clearing the site served to confirm the statement that the Pbthi leaves had come into Mullah Khwaja's possession from here.
Scanty as the remains are that have here survived prolonged exposure to erosion and the still Likelihood more destructive burrowing by men, they amply suffice to prove that the shrine belongs, like the ruins le remains traced at Darabzan-dong and Kök- i da, to the same period as the temples of the Khadalik site tamarisk-
g J g P P tamarisk-
situated only some three miles farther north, and is likely to have been abandoned like them about cones.
the end of the eighth century A. D.12 At the same time the way in which this small ruin had for so long escaped discovery, notwithstanding the search made in this neighbourhood during my former stay at Khadalik, made it appear very likely that more remains of the same character might still remain hidden in the maze of tamarisk-cones that cover the ground to the east of the Domoko-yar.
8 See Serindia, i. pp. 199 sq. ; for the relative position of the three sites, see Map No. 14 c. 2.
9 See ibid., i. p. 166 ; iv. Pl. XI, XII.
1° See Ancient Kliotan, ii. Pl. LX, for D. vii. 6.