Sec. SCULPTURED REMAINS OF RUIN M. II
Upper portion of base.
Remains of upper story.
North-west Frontier.' In the second place, an indication of distinct archaeological value is to be found in the shape of the double bracket which surmounts the pilasters in the Mirân ruin.ba This accurately reproduces that simplified form of the double bracket, which in a number of Gandhdra relievos replaces the original Persepolitan model made up of a pair of animal figures, usually bulls, crouching back to back.° The turned-down volute-like ends, which are such a striking feature of this double bracket in the Mirân pilasters, appear with equal . clearness in the Gandhara representations, and we have actual examples of them in wooden double brackets which have come to light at Lou-lan sites. Two specimens found at the ancient Chinese station L.A. are seen in Fig. 99. Two others which I discovered in 1914 among the ruins of the sites L.K. and L.M., belonging to the same period, show those voluted ends even more fully developed (see illust. 13, p. 486). We shall have occasion to discuss further on how these similar features of Gandhdra and Lou-lan sculpture may help us towards an approximate dating of the Mirân shrine.
No trace had survived of the stucco decoration which the faces of the base must have borne in their upper portion. But from the size of what was left of the standing figure in a niche of the north-east face, it is certain that the niches must have extended upwards. Taking into account the available height, about five feet, it seems probable that the projecting parts of the wall flanking the niches bore a second row of pilasters in stucco. Such an arrangement would agree with the superimposed rows of pilasters seen in the extant Stuipa bases and the walls of Vihâras in Gandhara.7 But I am at present unable to mention any instance there in which these repeated rows do not also flank separate rows of niches. It is far more difficult to guess what the badly ruined structure of the second story may have been. Its oblong ground-plan and its position on one side of the base do not lend support to the suggestion that it may have been a Stûpa base, nor does the fact that it retains its rectangular shape up to a considerable height. More probably it may have served as a platform, against the walls of which the principal stucco images of the Vihâra were built after a fashion illustrated by the Endere temple and certain of the cave-shrines at the ` Thousand Buddhas '.8 Apart from a wooden beam (seen in Fig. 120), which must have been originally inserted in the masonry of this superstructure and which was lying loose among the débris produced by the treasure-seekers' burrowing, nothing was found in clearing the upper story.
Fortunately we fared better in clearing the débris of broken clay and plaster which lay heaped up against the north-east face of the base, and the remains there brought to light left no possible doubt about the Vihâra character of the ruin and its early date. When on my first visit I had probed the débris on that side near the centre, we had come upon a colossal head in soft stucco. Badly damaged as it was, it showed in modelling and proportions the influence of Graeco-Buddhist
Cf. e.g. Foucher, L'art du Gandhdra, i. Fig. 8o, 81, showing the wall of a Vihâra at Takht-i-bâhi and the base of a Stûpa at Ali-masjid ; Stein, Excavations al Sahri-bahlôl, in Archaeol. Survey Annual, 1911-12, Pl. XLIV, L.
as See illustration A, p. 486, prepared from measured drawings.
e Cf. Foucher, L'art du Gandhdra, i. Figs. Jo', 103, 18o; pp. 227 sq. For examples where the original form of two crouching oxen is retained by the sculptors of the Gandhara relievos, see ibid. Figs. Ioz, 76,.77; Stein, Archaeol. Survey Annual, 191 1-1 2, P1. XLVII.
It appears to me probable that the volute-like pendants at the ends are a conventionalized development of the animal's forelegs, which hang downwards in the representations of such Indo-Persian columns, as seen in M. Foucher's Figs. 77, 102. In Grünwedel-Burgess, Buddhist Art in India, Fig. 81,
we find both the original and the conventionalized forms of this Indo-Persian double bracket represented in the architectural design of the same relievo.
7 Cf. e.g. Foucher, L'art du Gandhdra, i. Figs. 8o (Vihâra wall at Takht-i-bâhi), 8r (base of Stupa at Ali-masjid); Stein, Archaeol. Survy Annual, r9I I—I2, Pl. L (platform of Vihâra, Sahri-bahlôl).
s See Ancient Kho/an, i. p. 423, Fig. 49 ; below, chap. xxv. sec. i. The ruin E excavated by me in 1912 at Sahribahlôl shows a curious resemblance in the oblong ground-plan of its two-storied platform, which originally seems to have belonged to a Buddhist Vihâra; see Archaeol. Survey Annual, I 911—it 2, pp. II 4 sqq. ; Pl. XXXIV. But the ruin has suffered too much by later adaptations and otherwise to offer any safe indication.
Colossal heads in stucco.