National Institute of Informatics - Digital Silk Road Project
Digital Archive of Toyo Bunko Rare Books

Position of royal' residence.

Difference of levels.

Question of earlier origin of Chinese labels.

220   THE NIYA SITE   [Chap. VI

Khotan.20 But it is certain that the ancient site representing its capital during Han and Chin times must by then have been abandoned for centuries. We have also seen that Hsüan-tsang knew the town which he passed in a corresponding position on the Niya River, a few years before Tang rule was extended to the Târim Basin, by the name of Ni jang, identical with the present Niya, and that it served then as the frontier station of the Khotan kingdom.21 Hence it seems safe to assume that Ching-chüeh in the Tang Annals is merely an archaic designation for what is now the Niya Oasis. With this the distance of seven hundred li to the east of Khotan there indicated agrees remarkably well, seven daily marches being still the usual reckoning between Khotan and Niya.22

The presence of these Chinese tablets in the dustbin of N. xiv. iii, can readily be accounted for only on the assumption that the local chief's residence had stood near this great refuse deposit. Was it from the large hall, N. xiv. i, adjoining that these quaint relics of a royalty as yet unrecorded had found their way into the dustbin ? Or did this hall, before it was covered up by stable refuse, belong to a more imposing structure of an earlier period which had disappeared without leaving a trace on the surface, even before the site was abandoned towards the end of the third century A.D. ? Had I known at the time the significance of these little Chinese tablets, I should certainly have endeavoured to give an answer to these questions by excavating below the ground level of the ruins N. xiv. i, ii, and by any other practicable operations.

The evidence otherwise available does not suffice for a definite answer, but yet merits careful record. In the first place the fact must once more be noted that the floor of the boarded enclosure, in which the tablets and other ` sweepings ' of archaeological interest were found, lay fully seven feet below the ground level occupied by the ruins and eroded remains of timber and plaster walls which the plan (Plate 9) shows along the north-east and north-west sides of the refuse deposits. This difference of levels is readily explicable only in two ways. Either there had originally been a natural depression by the side of the buildings which we may presume to have been part of the royal residence, and this was suitably utilized as a place for stabling and depositing refuse until the surface level rose to that of the adjoining ground westwards. Then it may have been built over by some structure to which the walls referred to belonged. Or else we should have to assume that the ground level was originally the same, both at the area occupied by i and ii, and at that of the refuse heap iii, and that just as at the latter the level had gradually risen by successive rubbish deposits, so it had been raised at the former, too, as the result of long-continued occupation by periodically renewed buildings. Which of these two assumptions is more likely I am unable to decide now. But it is noteworthy that in the south-western portion of the refuse heap I came upon what looked like remains of an encircling wall built of sun-dried bricks, about three feet thick. Amidst its débris there lay pieces of clay which seemed to have been fired accidentally. These remains were found well above the natural ground, resting between refuse layers both below and above.

There remains thus the possibility open that the miscellaneous articles found within the boarded enclosure and mostly in its lower portion may date from a time not inconsiderably earlier than the period immediately preceding the abandonment of the site. This justifies a closer scrutiny of these articles with a view to any chronological indications they may, perhaps, furnish. Looking at the Chinese tablets in the first place, I cannot help being struck by the difference their writing presents

Y0 See Chavannes, Turcs occid., p. 127.

21 Cf. Ancient Khotan, i. p. 31 I ; above, p. 212.

n The difference between this Tang estimate of 700 li and that of 850 li in the Former Han Annals (see Wylie in J. Anthrop. Inst. x. p. 29; Herrmann, Seidenstrassen, i. p. 96) deserves notice. It is fully accounted for by the fact that the direct distance from Khotan to the Niya Site (the

Ching-chüeh of Han times) is distinctly greater than that from Khotan to the Niya oasis. In addition it must be remembered that the direct route to the former between the Keriya and Niya Rivers crosses what already in ancient times must have been a troublesome belt of sandy desert, whereas the modern route to Niya avoids this and after Keriya passes only over far easier gravel desert.