Sec. in FIRST EXCAVATION OF BUDDHIST SHRINES 245
Paris. This accounts for the hardness of almost all the pieces from D. I and the relatively good preservation of their surface, including the colouring. The stucco of the walls, on the other hand, being soft in comparison and friable, proves to consist of loess containing a considerable proportion of true clay. A special distemper was probably applied to prepare its surface for frescoes. The fact that the stucco of the large reliefs and images at Dandân-Uiliq, as well as at the other sites, was invariably made up of loess like that of the walls, explains why, of the large image or images which the shrine D. 1 must be supposed to have once contained, nothing turned up in the débris but a life-size finger in red clay (D. I. 14). It is probable, however, that the colossal hand grasping a billet (D. T. o 12), made of the same friable stucco, which Turdi had brought from his previous visit, was also obtained in this ruin.
Turning to individual pieces recovered from the wall decorations of D. 1. we may specially notice the small relief figure D. 1. I1 (Plate LV), represented in a series of replicas, of a Buddha standing with the R. hand raised in the Abhaya-mudrâ. attitude, but brought across the centre of the breast. D. 1. 99 (see Plate LV), one of several replicas, shows a small Bodhisattva seated on a lotus and surrounded by an elaborate vesica, and D. T. 05 on the same plate a more carefully modelled head of a Buddha or Bodhisattva. The graceful figure of a garland-holding woman rising from a lotus (D. T. 02, D. 1. II o, D. I. 014) which Plate LVI illustrates, is probably meant for a Gandharvi, and must have been a favourite decorative theme with the local artists ; for her figure reappears in a variety of poses and sizes among the stucco reliefs from the walls of other shrines (see Plate LVI). The female head reproduced in Plate LIV from D. 1. 90, one of several fragmentary specimens (comp. D. 1. 18, 43, 89) must also, in view of the curious hairdress, be attributed to a Gandharvi, though of a somewhat different type.
Among the fragments of relief ornaments, D. 1. 42 (Plate LIV) serves to illustrate the frequently recurring half-round border of a vesica, showing rows of overlapping lotus-petals originally in alternate colours (comp. D. n. 34 in same Plate). D. I. 87 (Plate LV I I) is a specimen of a curious ornament representing tongues of flames in varied colours, which seems to have regularly adjoined the border just noticed (comp. D. ii. 55, 89, Plate LV). Peculiar to D. 1, and distinctly graceful in design, are the floral ornaments (see Plate LVI I) showing vine scrolls and grapes (D. 1. 114 ; also D. 1. 53 in Plate LIV), lotus leaves springing from a moulding of beads (D. 1. 6o, 74), lotus petals (D. T. oII) or conventional chrysanthemum leaves (D. i. 012). Jewel-shaped ornaments are illustrated by D. i. 10 (Plate LVII), D. I. 44 (Plate LV).
The clearing of this single small ruin not only yielded a large number of sculptural fragments showing unmistakable affinity to the Buddhist art of Gandhâra, but supplied me with the indications needed in order to start the systematic excavation of structures more deeply buried in the sand. So when on the next day I proceeded to the group of ruins situated half a mile to the south of my camp, where on my first rapid survey I had noticed a number of small buildings well covered by the sand and thus giving promise of better preservation, I was able with some assurance to gauge their construction and character, though only the broken and bleached ends of posts were visible above the sand. The arrangement of these, in the case of the easternmost structure D. II, is shown by the accompanying photograph, Fig. 28, as it appeared before excavation. Together with a bit of fresco-covered wall laid bare by some recent burrowing of Turdi's party, it suggested two small temple cellas. The excavations commenced here soon proved the surmise to be true, and in addition revealed unexpectedly- rich relics of the sculptures and frescoes which had once adorned the cellas. Their constructive features and adornment proved typical of those observed in other shrines subsequently excavated at this site, and may therefore be described in some detail.
| || || || || || || || |