National Institute of Informatics - Digital Silk Road Project
Digital Archive of Toyo Bunko Rare Books

Sec. I]   GENERAL DESCRIPTION OF THE SITE   795

The panels were almost invariably filled by elaborate compositions containing a large number of figures. Varied as the subjects were in details, two main classes could readily be distinguished among them. In one there appeared figures of Buddhas, surrounded by symmetrically grouped hosts of Bodhisattvas, saints, and other divine attendants, all shown in a carefully arranged architectural setting of pavilions, platforms, lotus tanks, etc. (see e. g. Figs. 206, 209, 210, 220, 224, 231, 235). That these panels were meant to represent scenes in Buddhist heavens could be recognized even without any claim to special iconographic knowledge. The other class of fresco panels displayed, in close juxtaposition and often bewildering variety, scenes which looked as if taken from mundane life, but often with sacred figures moving among them (Figs. 217, 218, 221, 222, 233,236). Similar scenes were sometimes found also in borders or friezes framing the large panels (Figs. 202, 2 I o). The fact that by the side of or above such scenes there appeared very often cartouche-like bands bearing short Chinese inscriptions suggested from the first that these scenes were taken from sacred Buddhist legends. But Chiang Ssû-yeh, like Chinese literati in general a stranger to all details of Buddhist mythology and iconography, was unable to interpret them adequately. So it was only after similar representations among the pictorial relics which I brought away from the Thousand Buddhas ', and which will be described further on, had been submitted to expert examination in Europe that I felt assured that these scenes in the mural paintings illustrated Buddhist Jâtaka stories.

In these legendary scenes of the panels, with their freely drawn landscape backgrounds and typically Chinese architecture, as well as in most of the decorative designs displayed by the rich floral borders and the ornate exuberance of canopies, friezes, etc., it was impossible not to be struck at once with the prevalence of distinctly Chinese style. Its penchant for bold movement and realism was just as clearly expressed in the drawing and grouping of the figures as that for graceful curves and fantastic freedom in the cloud scrolls, floral tracery, and other decorative motifs. But equally certain it was that the- representation of all the principal divine figures bore the unmistakable impress of Indian models transmitted through Central-Asian Buddhism. It was the same in the large compositions as in the groups of saints often painted by the side of the main statues (Figs. 201, 207, 208, 213), or even in that schematic multiplication of Buddhas and Bodhisattvas with which Buddhist piety, using the convenient method of stencils, had covered here as elsewhere the walls of so many shrines. Whatever difference there might be in the technique of outlines and colouring, hieratic tradition had preserved for these figures the type of face, pose, and drapery originally developed by Graeco-Buddhist art.

In spite of this strong conservative tendency there were obviously different phases of development to be distinguished among these wall paintings. Without any Sinological training or expert knowledge of the history of Chinese secular art it was impossible for me to arrive at exact conclusions as to the chronology of the various cave-temples and their pictorial remains. Yet a variety of archaeological indications, some of which will find mention hereafter, suggested that the best of the mural paintings, usually found in the cellas of the large shrines, belonged to the times of the Tang dynasty, when the sacred site, like the Tun-huang oasis itself, had enjoyed spells of prolonged prosperity, or to the period immediately following. As to the fresco work, later in style but still skilful and vigorous, which was often to be found in the antechapels and passages of these shrines, where the liability to damage had necessarily been greater than within the cellas, it appeared probable that it dated from restorations carried out during periods, such as of the Sung or the Mongol dynasties, when the old artistic traditions could still assert themselves.

In the case of the sculptural remains it seemed even more difficult to arrive at an approximately correct dating, as their character as ' idols ' and their friable material, already referred to,

Panels showing Buddhist heavens and Jätaka scenes.

Chinese style of legendary scenes.

Indian type preserved in divine figures.

Chronology of wall-paintings.

5

I 2