442 THE RUINED FORT OF MIRAN CH. XXXVIII
where they were safest from erosion. But what a life it must have been, spent amongst them ! I have had occasion to acquire a rather extensive experience in clearing ancient rubbish heaps, and know how to diagnose them. But for intensity of sheer dirt and age-persisting ' smelliness ' I shall always put the rich ' castings ' of Tibetan warriors in the front rank. The recollection of these
Miran Fort perfumes was fresh enough a year afterwards to guide me rightly in the chronological determination of another site ; but that is a story to be told later.
Apart from desiccated pure filth and interspersed miscellaneous records, these rubbish layers contained remains
of implements, articles of clothing, and arms. Modest,
indeed, they all were in make, and much worn in condition. But this uniformity throughout the deposits, which must have taken a long time to grow to that height, was additional evidence that they faithfully reflected the local conditions of life. There were in abundance pieces of coarse woollen fabrics, canvas, and felt, which must have once belonged to the braves' personal outfit. The colours varied greatly, dark browns and deep shades of red prevailing. But not a rag had found its way into the rubbish until it had become hopeless for use. The scarcity of silk pieces was significant. Fragments of elaborately woven rugs had survived ; but the patterns, like the weaving of the woollen fabrics used for clothes, strikingly varied from those of the corresponding relics brought to light at the earlier sites of Niya and Lop-nor.
It is impossible here to give details of particularly curious finds, such as a small bag made up from a piece of delicate silk brocade ; a well-preserved felt pouch which might have formed part of a soldier's equipment (Fig. 138, 27) ; a quilted shoe of buff cloth stitched all over in elaborate geometrical designs. An exceedingly filthy pigtail of coarsely strung black hair did not look as if its quondam owner had troubled much about the use of the combs in wood and horn of which we recovered numerous specimens (Fig. 138, 25). Remains of nets in stout string suggested that fishing in the stream of Miran had been a more productive pursuit than at present, or else that the marshes of the Kara-koshun
IIIF Color HighRes
