Sec. i] SCULPTURED REMAINS OF RUIN M. n 489
way by more or less parallel double grooves. The tendency towards this representation of folds by shallow grooves can be traced also itr certain of the Rawak statues and elsewhere.72 The outside
layers of the stucco, where exposed in these torsos, contained a considerable admixture of vegetable fibres or hair, and in iii and iv portions of the facing plaster still retained that red colour wash which regularly appeared in the Rawak relievos.
Remembering that at the Rawak Stûpa court I had found the wall-spaces between the colossal standing Buddhas utilized for smaller images, I cleared the interstices between the big statue bases
with special care. But only between i and ii did I come upon the base of a small image, with remains of feet and drapery which might have belonged to a relievo figure about three feet high. The space left free between ii and iii had, however, proved useful by affording safe shelter to the colossal head which had once belonged to the third statue. It was found, as the photograph in Fig. 122 shows, firmly wedged between its own base and that of its right-hand neighbour, still upright and with the front portion almost uninjured. In expression and certain details of modelling, e.g. the divided chin, it differed from the other colossal heads, which, by their place of discovery, had to be assigned to statues ii, iv, and v respectively.
The colossal head of iv, which had fallen face upwards just in front of its torso, was, as already stated, found in a badly-battered condition. But, curiously enough, on the left knee of this statue we came upon another stucco head of a Buddha, M. II. oo6, of life size, which, except for the peeled-off surface, had survived in fair preservation. Judging from its size, it may safely be supposed to have belonged to one of the stucco images that formerly filled the niches on the opposite passage wall. Similar observations made in the course of my excavations at the Rawak Stûpa suggest the explanation that the head had been removed to this position of safety by the hand of one of the last worshippers when it threatened to fall off .13 Two other life-size heads, one found in the north-east passage and the other in the south-east, are also likely to have dropped from images in the niches, but were too much injured to merit removal. The origin of the two sculptured fragments, M. II. OOI2-13, belonging to statues of small size, could not be determined. Nor is it possible to guess the use of the two wooden half-balusters, M. II. 009-0010, which turned up in the débris of the north-east passage together with some timber pieces of uncertain character. The absence in the ruined shrines of any signs of destruction by fire deserves notice here, as well as the indication furnished by the small fresco fragment, M. Ii. 004, that its walls were once decorated in colours.
The most important of the small ` finds' in the ruin was made when, on clearing the débris in front of the image base iv, there were discovered a mutilated folio and a fragment of another from a palm-leaf Pothi, M. II. ooI I (Plate CXLIII), written in Sanskrit. Its Brâhmi characters were at the time recognized by me as of an early Gupta type. The mutilated folio, measuring about 6 by 2 inches, and showing five lines on either side, lay quite close to the image base, embedded in stucco débris which had evidently fallen from the crumbling statue above. The position in which it was found pointed clearly to its having been deposited on the base as a votive offering after the fashion which I had first observed in the ruined shrines of Dandân-oilik and the Endere fort, and which so many manuscript finds at the Khâdalik temples had since illustrated.14 The Pôthi, according to Dr. Hoernle's statement,"a seems to have contained a grammatical text, and the material made it certain that it had been written in India. Judging from the palaeographic features of the writing, which make it appear somewhat older than the Bower Manuscript from Kuchâ, Dr. Hoernle was inclined to place its date about A. D. 400. There is no special reason to believe