452 THE ETSIN-GOL DELTA AND THE RUINS OF KHARA-KHOTO [Chap. XIII
in A.D. 1226 and the subsequent destruction of Hsi-hsia rule, the first quarter of the thirteenth
century may be accepted as the terminus ad quern after which such a deposit could not have been
made. On the other hand, the prevalence of texts in that Hsi-hsia script which was only invented
and introduced by the Tangut ruler Li Yuan-hao about A. D. 1032 precludes an earlier date than the second half of the eleventh century. It must be hoped that the discovery of dated texts in the Petrograd collection will make it possible still further to reduce the chronological range between these two limits.13
Of the small ruins shown by the plan in P1. 22 quite close to K.K. II, that to the east consisted only of a brick-built platform about 12 feet square, with the scanty remains of a completely wrecked superstructure whose character was no longer recognizable. A low mound to the south of K.K. 11, when excavated, disclosed only the badly decayed brick walls of a rectangle measuring 192 by 2I2 feet. The walls were only i foot in thickness, and the bricks measured 12" X 6" x 3". No find of any sort was made within the interior, which was filled with coarse sand and gravel. The structure may have served as quarters for attendant priests of the shrine.
The only structure remaining to be described in the immediate vicinity of the town walls is the domed building, K.K. v1, seen in Figs. 251, 252. It stands at a distance of about 3o yards to the south-west of the south-western corner bastion and rises still to its original height of close on 23 feet, though part of the dome on the north side has fallen in. Apart from this and some minor damage to the side walls, the structure has suffered but little, and still retains most of its original plastering inside and portions of it also on the outside eastwards. The chief features, as shown by the plan in Pl. 21, are a domed hall, 182 feet square, within, and a large vaulted porch projecting on the east. By means of a succession of projecting niches in the four corners, as seen in Fig. 272, each terminating in a pointed arch, the square of the hall is reduced to an octagon, which in turn by a similar device of niches is made to bear the circular drum of the high dome. A corresponding architectural arrangement is applied to the vaulted portion of the porch. The front of the latter is given a very massive appearance by the outward slant of its flanking walls. The outer corners of the walls enclosing the square hall are rounded off into semicircular buttresses, also slanting outwards, thus creating for the whole an impression of great strength. For other architectural details the reader may be referred to the photographs and the elevation and sections shown in Pl. 2 I . The masonry consists of sun-dried bricks, about 16" x 8" x 4" in size, and its horizontal courses distinguish it entirely from that used in other buildings of the site.
Muham- The structure in its plan, purpose, and architecture is wholly Western and unmistakably meant
madan for a Muhammadan tomb or ` Gumbaz '. Its style is plainly Saracenic, but it must be left to
of structure. others more competent than myself to determine whether the details of style offer any indications
as to date of construction. The interior was found completely empty ; but divers holes in the plaster flooring and walls showed that ` treasure-seekers ' had been at work here too. We know from Marco Polo's account of the ` Province of Tangut ' that towards the close of the thirteenth century, when Mongol dominion over these parts as well as over the rest of China was fully established, Islam had its adherents among the population of these north-western marches, though, no doubt, Buddhism vastly prevailed. It is to this later phase of Khara-khoto's history, when Marco Polo knew it as the ` City of Etzina ', that I am inclined to ascribe the construction of the tomb. Whose mortal remains it was meant to guard, whether those of some local Muhammadan notable or official, or
Small ruins near
Domed structure K.K. VI.
13 No conclusion as to the relative chronology of K.K. u by comparison with the other shrines of the site can be drawn from the fact that the miniature votive Stüpas found here, too, in great quantity, K.K. u. 0109-1r, 121-3, are of
the type of K.K. 0090-6, &c., represented at most of the Stûpas of Khara-khoto ; for the form of these small votive offerings was evidently stereotyped throughout the occupation of the site.