1278 . RUINED SITES EAST AND NORTH OF KHOTAN [Chap. XXXI
facing the spectator, with eyes straight and fully open. The dress consisted of a tight-fitting pink jacket or coat reaching up to the neck and well down over the hips. Below it there was seen in front a blue vest between the turned-back lapels of the coat, leaving but little of décolletage below the triple folds of the neck. The flat bust and low waist gave a curious Elizabethan look to the strait-laced figure. The body of the jacket and the tight long sleeves showed elaborate patterns mostly worked in small whitish dots, evidently meant for embroidery and perhaps also strings of pearls. The head displayed well-proportioned youthful features, which but for the elongated lobes of the ears had no hieratic look. Over the hair, dressed smoothly on the forehead and hanging straight down behind, and by the side of the ears there appeared a yellow diadem closely resembling that worn by the lady on the left of the Dandân-oilik panel D. x. 4 to be presently mentioned. The portion of the diadem on the left proper was broken off; for the same reason the object held up by the left upper arm and suggesting a flower could not be determined with certainty. The right upper hand carried a round or foreshortened oval object of whitish-yellow with fine lines in light pink radiating from the centre. The two lower arms were hanging down stiffly to below the hips, where the painting was mostly effaced and the hands could no longer be made out. On the right shoulder the deified lady carried four well-defined narrow and long leaves rising from what looked like a small elliptical basket ; between the first and second leaf to the left there appeared a smaller oval object which might have been meant for another curled-up leaf, but might be interpreted otherwise.
It was in the first place the cognizance held in the upper right hand, suggesting a cocoon by its shape, and the green leaves so strangely adorning the right shoulder which made me at the time recognize in the figure a representation of the Chinese princess to whom the old Khotan legend reproduced by Hsüan-tsang ascribed the introduction of sericulture into the kingdom.12 The renewed examination of the figure, made possible by the photographs taken before this portion of the wall collapsed, has confirmed my belief in the correctness of this identification. The painted panel D. x. 4, found in one of the Dandân-oilik shrines where it had been deposited as a votive offering,13 conclusively proves the popularity of the legend, and also that the princess to whom Khotan owed the introduction of its silk industry, so important to the present day, was honoured with worship. Considering the local character of this worship and the non-Indian origin of its recipient, it seems particularly appropriate that the deified lady, to whom Khotanese cultivators must have prayed for protection of their silkworm crops, should be represented by our fresco in secular costume, wholly different from the conventions of the Buddhist Pantheon, and that she should have been given her place by the side of another local divinity, the king of the sacred rats."
Beyond the frescoes now described, the walls of the shrine Ta. i were found completely perished. To search for further structural remains, which the high ridge to the north and east may well hide,
" See Ancient Kholan, i. pp. 229 sq.; Julien, III/moires, ii. pp. 237 sqq. ; Beal, Siyu-lie, ii. pp. 318 sqq. ; Rétnusat, Ville de Kholan, pp. 55 sq.
18 Cf. Ancient Kholan, i. pp. 259 sq. ; ii. Pl. LXIII. 14 It would be possible to think that our fresco panel represented, not the beneficent princess herself, but the Goddess of the Silkworms' to whom, according to Julien's translation of a passage in Hsüan-tsang'sHsi yii-chi (Mémoires,
ii. p. 239), she had dedicated the Lu-shê ) t4 convent
near the Khotan capital. Rémusat's and Beal's translations (Ville de Kholan, p. 55; Si yu-ki, ii. p. 319) make no mention of the goddess. [Dr. Giles confirms Julien's rendering :
` A is the usual designation of the Goddess of Seri-
culture, Lei Tsu.'] In any case, if there had been an officially recognized cult of the Goddess of the Silkworms' in Khotan, popular worship was bound soon to mix up the goddess and the princess in one ' Lady of the Silkworms'.
Mr. Watters (Yuan Chwang, ii. p. 302) does not give a translation of the passage ; but he points out that one version of the text gives the name of the convent as Mo-shë
At, which agrees better with the form 16Madza found in the Tibetan record of the legend. [' The Kyoto edition of
the Tripitaka has the reading Ma-shê, which agrees
better still with the Tibetan.'—L. Giles.]
Painting of four-armed lady.