510 THE ANCIENT BUDDHIST SHRINES OF MIRAN [Chap. XIII
and it could be illustrated by hundreds of examples.3 In the great majority of these carved friezes the spaces left above the downward droops of the festoon are filled by youthful figures, generally, but not always, conceived as winged dancers or musicians and shown down to the breast.4
It is obvious that the composition of the dado in the circular passage of M. v is an exact illustration of the decorative scheme which, as we must assume, was used in the Graeco-Buddhist pictorial art now lost. It is equally easy to see that the arrangement of the lunettes in M. III is directly derived from the same scheme by omitting the carriers of the festoon and retaining the winged figures above its downward droops. That the broad semicircular bands which edge the lunettes below, and form a kind of frame for the ` angels ', are only conventionalized substitutes for the descending curves of the festoon is made particularly clear by a comparison of the remnant of the dado once adorning the square outside passage of M. v, seen in Fig. 133 and Plate XLV. There, too, we see the bust of a winged figure rising from the hollow of a lunette-shaped band, and the width of the latter and its decoration with bold scroll-work leave no doubt that it was originally meant to represent part of a festoon. I am at present unable to find an example among GraecoBuddhist relievos of a similar development of this frieze motif in which the topmost parts of the festoon, together with the amorini supporting them, are omitted and only the lunettes left. But I may refer to the frieze fragment from Gandhâra, shown in the right bottom corner of Plate XLI, as illustrating.a modification of the scheme in the reverse direction. Here the downward droop of the festoon is so much reduced in width that there is no longer space left for a bust, but merely for a small floral ornament, which, however, in its shape still curiously recalls the wings of the figure it replaces.
The close connexion which the preceding observations have established between the designs used for the decorative dados of the Mirân temples and the festoon friezes of the Gandhâra relievos helps us to trace the true iconographic descent of the winged figures appearing on the walls of M. III. They correspond too closely to the youthful figures with wings which we see rising from the hollows of the festoons in so many of the Gandhâra friezes to allow any other direct origin to be claimed for them. The smallness of these carved winged figures, and still more the much-reduced scale of the reproductions, make it often difficult to ascertain whether boys or girls are intended .5 But almost invariably their forms are childlike, and this, combined with the constant male representation of the festoon-carrying putti which flank them, makes it highly probable that the Gandhâra- sculptors, in accordance with their regular wont using a classical type which was ready at hand, modelled them after the youthful winged Eros of Greek mythology. How accustomed these sculptors were to draw upon the classical Cupids, whether with or without wings, for their decorative personnel, and how closely the type presented conformed to classical tradition, M. Foucher has lucidly demonstrated.° Nor is it difficult to discover why they preferred the winged form for insertion in the hollows above the festoons. No ornamental device could have been artistically better suited for filling the tapering sides of the lunettes thus created than the graceful ends of the wings.
s Cf. Foucher, L'art du Gandhâra, i. pp. 24o sq. ; Figs. 116-18, also Figs. 72, 75, 76, 136. See also Pl. XLI below, where the available space has been utilized to reproduce three relievo fragments of this type now in London.
' For winged figures thus appearing above the garland cf. Foucher, L'art du Gandhdra, i. Figs. 117-18 ; also British Museum relievo in centre of Pl. XLI, below. The relievo of Dr. T. W. Arnold (probably from Sahri-bahlôl), shown in the same plate on the left, is a good example of the type with
5 In the friezes reproduced in Figs. 117, 118 of Foucher, L'art du Gandhdra, the winged musicians look like boys, as they may be also in the relievos of which much-reduced reproductions are given in Figs. 75, 76, 116, 136. On the other hand, the winged figure in the British Museum fragment shown in the centre of Pl. XLI, below, is undoubtedly female. The distinction is of interest, but cannot be followed up further here.
e Cf. L'art du Gandhdra, i. p. 241.