276 REMAINS OF ANCIENT LOU-LAN [Chap. VII
intelligent Khotanlik of Lai Singh's small party, showed me a ` house ' he had noticed on his previous passage. It proved to be a kind of underground chamber cut into the hard clay, about twenty feet square and approximately orientated. Its roof was formed by closely adjoining trunks of young Toghraks resting on two roughly cut rafters of Toghrak wood. Above this roofing there was spread a kind of very coarse matting made up of bundles of a thorny scrub (now known as yiken) laid crosswise. This matting, again, was covered with wheat-straw and a layer, about one foot thick, of fairly hard clay.
Wind-erosion had carried off the clay over the eastern portion of the roofing. This allowed me to see at once that the interior, about 4 or 5 feet deep, was occupied by three coffins. All proved
to be roughly made up of Toghrak trunks hollowed out and closed at their ends by cross-boards.
The first one opened, near the north side of the chamber, was found to be filled only with solid clay. Was this a coffin, ready stored for burial but never used, which fine loess dust blown in by
the winds had ultimately filled completely ? Or had the body it once held completely disintegrated
under the influence of moisture ? There was nothing to give a definite answer. In the next coffin, in the middle, were discovered decayed human bones and a mass of very brittle rags, L.H. i. oi,
difficult to open but clearly containing numerous fragments of silk, evidently from worn-out garments. The third coffin near the southern side of the chamber was not opened by us for fear of losing more time than we could afford. It was for the greater part embedded in a mass of clay, which may have accumulated from fine dust blown into the chamber and subsequently hardened under the influence of moisture.
Scanty as werethe finds made here, this rough burial-chamber itself afforded instructive evidence. Its construction, entirely different from that of the graves in which the indigenous Lou-lan people
of L.F. had been laid to rest, suggested a method of burial that would account for miscellaneous
remains such as those I had found thrown together in the grave-pits of L.C. The presence of silk rags, from worn-out clothes, distinctly supported this inference. Finally the presence of
wheat-straw afforded a clear indication that cultivation had been carried on at some point not very distant from the burial-place. It was moreover safe to conclude that the position of the latter had been chosen with a view to securing the protection of the bodies both from moisture and from wind-erosion.
These inferences were confirmed when I proceeded to examine the small cemetery, L.H., reported by Lai Singh and situated about one mile away to the north-west.18 It was reached about
2 p.m., and, in view of the tramp still before us and the risk of missing camp in the dark if we were
belated, our examination had to be hurried. In a shallow dip of the gently sloping Sai I found a row of four wooden coffins placed close together from north to south and half-exposed on the
gravel-covered surface (Fig. 169). A single Toghrak post standing upright still supported on its
gabled end a big beam, on which a roof similar to that observed at the burial-place farther south had no doubt once rested. Wind-erosion had carried off the roof as well as the wall on which it
is likely to have rested, leaving only loose remains of much-splintered and shrivelled timber. The coffins, too, had all been injured to a greater or lesser extent by the corroding force of wind and sand. A big coffin, at the southern end of the row, was found completely broken and its contents dispersed. It was made of massive Toghrak planks joined into an oblong ; in the other coffins, hollowed out Toghrak trunks served for the sides. None of them retained their covers.
The much-decayed northernmost coffin measured fully nine feet in length and was evidently intended to hold other deposits besides the body. Such bones as survived were found wrapped in
18 In Map No. 29. D. 3 the ` site-mark ' L.11. has by above mentioned, instead of against the entry ` Ancient
mistake been shown against the site of the ruined dwelling cemetery '.