280. THE RUINS OF DANDAN-UILIQ [Chap. IX
round the waist is slung a short, broad sword in an ornamented sheath. The most curious part, however, of the attire are the high black boots exactly resembling Hessians, which reach to the knee, and are picked out in ornamental yellow lines suggesting inlaid leather-work or stitching in gold-thread. The scarf-like Uttariya, which descends from the back of the head and is curled round both arms, is the only part of the dress that the figure shares with other representations of sacred personages in these panels. The left proper front hand holds the Vajra, while the rear one raises a spearhead ; on the right proper the front hand rests on the thigh, the one behind holding a deleted object, possibly a flower.
For the identification of this strangely attired divinity, the Vajra or mystic thunderbolt seems to afford the only clue. It is the characteristic emblem of no less than three divinities, the Dhyani-Bodhisattvas, Vajrapani, and Vajrasattva, as well as of the primordial Buddha Vajradhara 8. With none of the representations that I can trace of these at present does our figure fully agree ; but seeing that it shows neither the green colour prescribed for Vajrapani, nor the blue one of Vajradhara, the conjecture may be hazarded that it is perhaps meant for a form of Vajrasattva. This Bodhisattva is ordinarily represented white, with the Vajra in one hand, and the other resting on the hip and holding the Khania- or bells. The latter is certainly absent in our picture, yet the general pose bears some resemblance to that of the Vajrasattva seen in the miniatures which M. Foucher has published.
The real interest of the painting lies, however, not in the eventual identification of the figure, but in the application of a design of essentially Persian style to a subject of Buddhist mythology. The immediate juxtaposition of this design to the wholly Indian one shown by the obverse of the same panel necessarily increases this interest. There is nothing to suggest that the two sides of the panel were painted by different hands, however different in origin were the models which the artist in each case followed. The representation of a Buddhist divinity in forms derived from a non-Indian art has in itself nothing surprising, since we see the identical process illustrated to the largest possible extent as far as classical art is concerned in the Graeco-Buddhist sculptures of Gandhara. But of the influence of Persian art under which Buddhist iconography may have come in those extensive parts of Eastern Iran where Buddhism found a home and flourished, we know as yet practically nothing. Is it possible that this curious little painting reflects features of a ` Perso-Buddhist' art such as might have developed in Balkh, that city of many splendid Viharas, and elsewhere within ancient Ariana ?
I have left to the last the mention of the panel D. vii. t, the first discovered, because though remarkable, as Mr. Andrews observes, for its good drawing, free execution, and good composition, it has suffered far more than the other panels of this ruin in its colours, and hence cannot be properly judged from the reproduction (Plate LXVI)10. The figure is undoubtedly that of a Bodhisattva, seated on a low throne, and holding in the left proper hand a bud-shaped lotus over a long, gracefully curving stem, while the right raises a patera in front of the breast. The upper part of the body appears to be dressed in a black garment, while from behind the head descends the scarf-like Uttariya curling round the upper arms in the fashion previously noted. The flesh is coloured pink. The shape of the lotus seems to suggest the Utpala or blue lotus ; and as the latter is the characteristic emblem of Manjugri, who usually appears to hold it in his left hand and in the pose here represented, I am inclined to believe that this Bodhisattva
8 Compare Foucher, Iconographie bouddhique, i. pp. 121 sqq. ; Grünwedel, Mythologie, pp. 98, 141.
9 See Foucher, Iconographie bouddhique, i. p. 123, Plate VI. 5.