38 THROUGH CHITRAL AND MASTOJ [Chap. II
It can scarcely be accidental that, whereas the introduction of square bases, in place of the single round one of the ancient Stûpa type, is illustrated by the majority of the Stûpa ruins now extant on the Indian North-West Frontier and in the Kabul Valley, yet the regularity of the triple number, so clearly prescribed by Hsüan-tsang's tradition, is in them conspicuously absent.13 On the other hand, practically all the ruined Stûpas in Eastern Turkestan I have so far been able accurately to survey, show this threefold arrangement of base with striking regularity. The explanation seems obvious, that the tradition recorded by Hsüan-tsang applied specifically to a type of Stûpa construction which had developed in the Buddhism of ancient Bactria, and that this, with much else of art, culture, and religious literature, had found its way to the east of the Pamirs as well as into the Chitral Valleys.
The successively diminishing height of the bases seen in the Pakhtoridini rock-carving was not an essential feature of the type, as is shown by a comparison of its pendant at Charrun (Plate 2). On the other hand, the narrow projecting frieze which separates the drum from the dome is repeated at Charrun, though merely in the form of a dividing line, and is clearly seen again in the Mauri-Tim Stûpa near Kashgar.'4 That the height of the dome is in excess of the original hemispherical shape is a feature shared by all the Stupa ruins I know in the Tarim Basin, and common to the majority of those found on the Indian North-West Frontier. The design intervening between the top of the dome and the spire composed of successive umbrellas is too coarse to permit of a very definite interpretation. But it can scarcely be doubted that surfaces slanting outwards are intended. These may have been meant to represent either the faces of a gradually projecting pedestal, such as most Stûpa models show in a corresponding position,1' or else figurative supports leaning outward, such as are seen below the bottom umbrella in the finely carved fragment of a miniature Stûpa in soapstone (Yo. 00121) I obtained from Yotkan (Plate vI).
Finally we have the ` clocheton d'ombrelles ', as M. Foucher graphically calls it, surmounting the whole edifice. Quaint as its drawing is, this too represents points of interest. M. Foucher has justly insisted upon the important part which the crowning spire of umbrellas must have played in the architectural effect of all Stûpas.'6 It has survived on the North-west Frontier only in a few Stûpa models of small size, and in relievo representations. In these it always absorbs at least one-third of the whole height of the structure." Now a reference to the carving shows that this proportion is there duly observed, the spire of umbrellas with its pedestal measuring 17 inches out of a total height of close on 50 inches. It is true that the number of the umbrellas or discs represented, which I take to be seven, exceeds the number of five which appears to be normal in the extant specimens of Gandhara and the Kabul Valley. It is known, however, that this was by no means the limit ; for the Chinese pilgrims attest for Kaniska's great Stûpa at Peshawar at least thirteen umbrellas, and the same number is still seen on the Stûpas of Nepal and on a rock-carving near Dras.18 And in order to give still further assurance on the point, both the small Stüpa models carved in wood which I discovered at the Lop-nor Site (LB. ii. 0033, 0034 ; Plate XxxII) show
" Thus among the numerous Stupas, miniature Stûpas and relievo representations of Stûpas, which M. Foucher reproduces in illustration of chap.l of his L' Art du Gandhdra, dealing with the Stüpa, I can only find one model which may be assumed to show a Stüpa with three square bases (Fig. 72, i. p. 185), and even this is doubtful (see p. 182). Yet other numbers of square bases are met with in plenty, from a single one (e. g. Fig. 71) to five (Fig. 19).
14 See Ancient Khotan, ii. Pl. 1, xxiI. It also appears in the wooden Stüpa models from the Lop-nor Site, LB. 1r.
0033, 0034, reproduced in Pl. XxxII ; and in that from the Niya Site, N. xvi. o01 (see below, chap. vr, sec. vi).
10 Cf., e. g., Foucher, L'Art du Gandhdra, i. Figs. 20-3, 70, 71. This pedestal appears, however, in the Charrun carving (Plate 2) represented in a different and better recognizable fashion.
16 See L'Art du Gandhdra, i. pp. 74 sqq.
. 17 Cf. ibidem, p. 76.
18 Cf. Foucher, L'Art du Gandhdra, i. pp. 77 sq.