Sec. iv] THE DADO OF `ANGELS IN CELLA NI. III 511
The evidence of the Gandhara relievos just discussed seems sufficient to warrant the conclusion that these winged figures of the Miran dado must be traced back to the classical god of love as their original iconographic prototype. But there are indications, too, warning us that this descent may well have been affected at intermediate stages by the influence of Oriental conceptions. In the figures before us, with their youthful but not childlike looks, their low-cut plain garments and quasi-sexless features, there is something vaguely suggestive of representations of angels such as we might have expected to meet with rather in some Early Christian church of the East than in a Buddhist shrine. I am unable to secure either time or materials for the researches which would be needed to test and eventually to explain this impression. There may be reasons, chronological or other, to put aside altogether the possibility of influence exercised by early Christian iconography. But it should be remembered that the idea of angels as winged celestial messengers was familiar to more than one religious system of Western Asia long before Christianity developed its iconography, and that the Zoroastrian doctrine of Fravashis had specially prepared the ground for it in those wide regions of ancient Iran through which both the influence of classical art and Buddhist cult must have passed before reaching the Tarim Basin. No graphic representations of angels appear to have survived in the Hellenistic East from a sufficiently early period to help us in clearing up the question where and when the Cupids of classical mythology underwent transformation into that type of winged figures of which the painter of the dado in M. 111 seems to have made use for the decoration of a Buddhist shrine.' The unmistakable presence of Semitic traits in most of these faces makes our thoughts turn instinctively to regions like Mesopotamia and Western Iran as likely ground for such an adaptation.
However this may be, it is certain that the appearance of such strange figures, unconnected with Buddhism, in the fresco decoration of a Buddhist place of worship need cause us no surprise. The carved friezes of Gandhara Stupa bases previously referred to, and an abundance of other relievos, show us how familiar a procedure it was for Graeco-Buddhist art on Indian soil to use, for the decoration of Buddhist shrines, figures and whole scenes entirely unconnected with the cult or sacred tradition of Buddhism. That this decorative practice was inherited by the early Buddhist art of Central Asia and carried to the very confines of true China was conclusively demonstrated when, on excavating the neighbouring shrine M. v, of exactly the same type, I discovered that the interior walls of its cella, under a painted frieze with pious scenes from a well-known Buddhist legend, were decorated with a dado displaying figures of an altogether secular and frankly Western character. Finally, it should be remembered that if ever a Central-Asian Herodotus had visited this temple of Miran, and had cared to inquire from the priest holding charge about the significance of the winged beings so strangely reminiscent of figures he might have seen before in regions where Buddhism had never effected a footing, the local guardian would scarcely have been at a loss for a name and might well have called them Gandharvas. Though in reality not needed, it would have been an acceptable label ; for there is abundant evidence to show that this class of celestial attendants was as popular in the Buddhism of Central Asia and the Far East as their representation was varied.
I am aware that the angel figures which meet us in Byzantine art are based upon the type of the classical Nike; cf. Diehl, Manuel de l'art byzantin, p. 8 ; Strzygowski, Orient oder Rom, p. 26. Yet the early instances where angels -
are presented as youths of distinctly male type, and still more the representation of Cherubim, point to another source of inspiration, the winged Eros. I am unable to follow up the question further here.