. Sec. i] SEVENTH-CENTURY TOMBS IN GROUP i 643
one side. This entrance, however, was not indicated by short walls symbolizing a gate as at Tun-huang and Nan-hu. The line of raised gravel was merely broken in the middle of that side, and its ends on either hand carried outwards for a short distance, varying more or less in proportion to the size of the enclosures. These themselves varied greatly in dimension, from a square of i 50 yards down to others scarcely more than io yards square. The enclosures were always rectangular, those of oblong shape prevailing. No definite proportion was traceable between the size of individual enclosures and the number of tombs within them. Nor was there a fixed bearing for the entrance, though the sides were in most cases roughly orientated.
Within each enclosure low mounds of modest size and rarely over 5 to 6 feet in height served to mark the position of the tomb chamber, which was cut in the hard clay soil beneath them. These tumuli were constructed of gravel with intervening layers of thorny scrub, and the larger ones were often decked with rough stones. They were generally much dilapidated, but appeared usually to have had the shape of a truncated pyramid resting on a square base. The orientation of individual tumuli seemed always to correspond to that of the enclosures, where these existed. But their grouping within these was by no means uniformly regular. Subsequent observations showed that the larger mounds, up to 3o feet square, were generally found to lie over tombs whose chambers were more elaborately arranged. From the middle of that side of the mound which faced the front of the enclosure as marked by the entrance, there always extended originally a low handle-like mound of gravel, sometimes bordered with rough stones, running at right angles towards that side. This mound marked the position of the deep-cut narrow trench which formed the approach to the sepulchral chamber. As almost all the trenches had been dug up and the tombs searched and plundered at one time or another, little remained of these ` handles ' beyond that portion lying close to the tumulus where the trench ended in a short tunnel-like passage giving access to the tomb chamber or its anteroom. With the surface remains of the Astâna cemetery area, I may here also mention a large tower-like ruin, badly decayed and much burrowed into by ` treasure-seekers '. It rises not far from the above-mentioned canal and near the middle of the area where tombs are few and scattered. It is built partly of stamped clay and partly of rough lumps of clay, and contains some small half-underground rooms which, of course, had been cleared out long ago.
Before I proceed to describe the results of the work which kept us busy for a fortnight at this great cemetery of Astâna, it will be convenient to record briefly what I ascertained at the outset, from local information and ocular evidence, as to the ravages that its tombs had undergone during comparatively recent periods. It was easy to realize from the condition in which the passages of approach to the tombs were found that most, if not all, of the latter had at one time or another been opened and searched, whether for valuables or in later years for antiques. Instead of the lines of embanked gravel and stones that covered these narrow trenches after they had been originally filled in on completion of the burials in the tomb chambers, the surface showed furrow-like depressions, sometimes just perceptible to the eye, sometimes well marked. They contained that fine drift-sand which the strong north-westerly winds prevailing through spring and summer sweep in masses across the Turfân basin and which quickly accumulates in any excavation or similarly sheltered place. Where the walls of the passages cut into the hard ground were still partly exposed, the excavation had evidently been of very recent date. Local opinion in Astâna and Kara-khôja was uniformly to the effect that all the tombs, or at least their vast majority, had been plundered by Tungans during the times of the last great Muhammadan rising and Yâqûb Beg's régime, mainly in search of valuables deposited with the dead. But, as our investigations soon showed, the wood of the solid old coffins must also have been prized as a very useful by-product. It apparently compensated for the labour involved in these operations, even where these yielded no profit in the