Sec. i] EXCAVATION OF RUINED DWELLING, L.A. I. 371
Stûpa, and their results held out encouraging hopes from the start. At about 6o yards' distance to the south-south-east of the Stûpa the top of a terrace, rising with steep banks above the wind-eroded ground, bore what manifestly was but the remnant of a well-built house once much larger. As the photographs, Figs. 93, 94, clearly show, heavy timber débris covered the slopes of the terrace, particularly to the east and south, thus indicating the position of rooms which had completely disappeared through erosion of the underlying soil. On the south the deepest portion of the immediately adjoining ground lay fully 18 feet below the original surface level, as marked by the foundation beams still in situ. Other such beams are seen in Fig. 94, either fallen over the slopes or else overhanging them. Here, as at all other ruins of this site, these foundation beams, as well as all posts, etc., forming the timber framework of the walls, were invariably of wild poplar wood. In the ruin L.A. I the foundation beams were of remarkable solidity, showing a thickness of nearly one foot. Below them a layer of tamarisk fascines, as subsequently observed also under the walls of L.A. II and the two Stûpa ruins, served as a flooring. The constructive features of the walls resembled very closely those observed at the ruined dwellings of the Niya Site and also of Khadalik. Set in sockets in the foundation beams were rows of square posts, as seen in Fig. 93, the thicker ones meant to carry the roof beams and the smaller dividing the space between them at regular short intervals. To the latter were fastened horizontal bundles of reeds, which formed the wattle of the walls and were covered outside with mud plaster.6 The walls in this, as in most other structures of L.A., were found to be carefully orientated with regard to the prevailing direction of the wind, one side of the rectangle always facing east-north-east.
In Plate 24 the ground-play' of the surviving portions of this house has been reproduced. To judge from the size of the large room, L.A. 1. i, which measured 31 by 13 feet, and that adjoining it on the south-east, which though destroyed for the most part still retained a sitting platform of considerable width, the ruin might have been the residence of an official or some person of position. Though the drift-sand covering the floor of the extant rooms was nowhere more than two feet deep, and in most places even less, it had sufficed to protect a number of interesting relics which soon lent support to that conjecture. Three narrow slips of wood, L.A. 1. i. 1; ii. 1; iii. 1 (Nos. 886-888 in Chavannes, Documents, Plate xxvii), about half an inch wide and each bearing one vertical line of Chinese characters on its obverse and reverse, I at once recognized as identical in shape, and probably also in character, with the Chinese official records on wood which my excavations of 1901 had first brought to light at the Niya Site.G The official nature of their contents is now proved by M. Chavannes' translation, which also has shown one of the slips, L.A. 1. iii. 1, to be accurately
dated in A.D. 330.' As this date, from a variety of reasons to be mentioned below, may be assumed to have been soon followed by the final abandonment of the site, it is highly probable that L.A. I continued to be occupied by some official till the end.
The close resemblance in type and internal arrangement between this ruin and the ancient dwellings, so familiar to me at the Niya Site, made me feel at the time less surprise than I might have felt at the discovery immediately following of two oblong tablets in wood, L.A. I. ii. 1, 2, of which each bore four or five lines of faint but still legible writing in Kharosthi. Yet it was
a most gratifying find, which thus at once gave assurance that the use of this ancient Indian script and language had extended so far away east, to the very end of the Târim Basin. When the
miscellaneous rubbish which had found refuge under the fallen pieces of timber marking the position of an eroded apartment to the north-east (L.A. 1. iv) came to be searched, there were