国立情報学研究所 - ディジタル・シルクロード・プロジェクト
『東洋文庫所蔵』貴重書デジタルアーカイブ

506   THE ANCIENT BUDDHIST SHRINES OF MIRAN   [Chap. XIII

0055, 0062 prove that animals, too, figured in the legend depicted, but the remnants are too small to permit of certain identification. The architectural background seen in M. III. 0058 is of special interest on account of the details we meet with in it, such as a bell-shaped capital and pillar and a panelled pilaster decorated with rosettes and scale imbrications. They usefully supplement the information which the scanty remains actually brought to light at the Mirân temples give us of the locally prevailing style of architectural ornament, and help to demonstrate still more clearly its dependence on Graeco-Buddhist and purely Hellenistic models."

Sr.C'rION IV.—THE DADO OF `ANGELS' IN CELLA M. IrI

Winged figures of dado.

The preceding analysis of the fragments which have survived of the frescoed friezes of M. ni will, I hope, make it easier for us to appreciate fully the artistic interest presented by the fine wingcd figures of the dado and to interpret correctly their iconographic significance. The fascination which, at their first appearance, they exercised for my archaeologist's eyes has in no way diminished since I could view them in safety and under less trying conditions. I have explained above how these figures of ` angels ' were disposed in the painted dado of the rotunda walls, groups of six occupying each of the four segments or arcs into which the wall was divided by the entrance and windows.'

In the north-eastern and south-eastern segments the lunettes containing these figures were all extant, at least in parts. But in those immediately adjoining the east window, vi and vii, as well as in the lunettes x–xii, the heads had been either completely destroyed by falling masonry or so badly effaced that only portions of the wings and shoulders remained to indicate their position. The same was the case with the lunettes xxiii and xxiv on that portion of the north-west segment which had retained some of its plaster surface. It was due to this that I was not able to rescue more than seven of the dado figures. Of these, i–v form a continuous series, ii being shown in colour in Plate XL and the rest in Plate XLI. Of the two lunette panels from the south-east, viii is reproduced in colour (Plate XL) and ix in monotone (Plate XLI). The seven panels recovered withstood the risks of their long and difficult journey remarkably well. Practically all the damage visible in the reproductions was suffered by these dado panels while still occupying their positions on the wall. The colour plate IV of Desert Cathay shows two of the panels, viii and ix, in the condition in which they had reached the British Museum and before the friable clay and straw backing had been replaced by plaster of Paris. A comparison of the earlier reproduction of panel viii with the one now presented in Plate XL will illustrate the care with which the re-backing of the painted surface was effected, and will show how well the cracks it had suffered, partly when still on the wall, have closed up in the process.

If we examine the cycle of angel-like figures presented by Plates XL and XLI as a whole, two general features must strike us at the outset as determining the artistic result. On the one hand, we realize clearly that, in keeping with the decorative purpose of the dado, the aim in all externals, such as the type of head, the wings, and the simple but graceful dress, is manifestly a homogeneous effect befitting a heavenly fraternity. On the other, it is equally obvious that the painter wished to introduce a pleasing variety into his cycle, and secured it by making a distinct individual element prevail in the faces. It will be convenient to review first those details in which the aim at concordant

Preservation of fresco panels.

Homogeneous effect of cycle.

10 With the scale imbrications may be compared those shown by Strzygowski, Amida, Fig. 78, Pl. IX, from the ancient façade of the Great Mosque of Diarbekr ; Figs. 8r, 84, from Early Christian churches of Bawit and Sakara in

Egypt (probably early fifth century n.n.); for the large overlapping rosettes, also represented in M. ni. 007, cf. ibid., Fig. 77.

1 See Pl. 32, and above, p. 497.