敦煌における変文と絵画―再び牢度叉闘聖変（降魔変）を中心に―=Pien-wen and Paintings at Tun-huang: Again with Special Reference to the “Magic Competition between Sariputra and Raudraksa (Chiang-mo-pien)”
In No. 187 of this Journal I published a study of the scroll-painting of “Raudrāksa's fight with the saint” in the collection of La Bibliothèque Nationale in Paris where it is labelled as Manuscrit Chinois P. 4524. It illustrates a story about a magic competition between “heretics”(nonBuddhists), with Raudrākṣa as their representative, who try to prevent the construction of Jatavana by the Buddha, and Śāriputra who is one of the Buddha's disciples. It is an incomplete scroll retaining five of the six magic bouts. I noted in that study that the versified parts of the text of the story, written near the ends of respective scenes on the back side of the paper, have been derived from the Ching-mo Pien-wên, one of the Pien-wên (narrative stories) among the manuscripts from Tun-huang (Pl. III). It appears that when priests of the Tang Dynasty told legendary stories about Buddhism to visitors to their temples, they showed them the pictures and read the prose parts aloud, singing the versified parts at the ends of respective segments of the stories. I have discovered fifteen examples on similar subjects from among the murals in the three groups of cave-temples at Tun-huang, namely the Chien-lo-tung, Hsi (west) Chfien-fo-tung and Yu-lin-ku, and arranged them in chronological order. (The cave numbers are as follows, the numbers crowned with A being those of Ch‘ien-fo-tung according to the system of the Institute of Tun-huang Research. Hsi Ch‘ien-fo-tung9, A335, A9, A196, A146 A85, A72, Yü-lin-k'u C10, A98, A108; Yülin-k'u C12 A55, A25, A454; Yü-lin-k'u C22.) Concerning A146, in particular, based upon the investigation (unpublished) by Paul Pelliot in 1908, I pointed out that among the texts and inscrptions within cartouches in its mural there were more than ten examples which are the same as corresponding of the Ching-mo Pien-wên. These researches have established a link between pien-wên and Tun-huang murals which had earlier been a subject of discussions. Thereafter I had another opportunity of studying ex- amples and reference materials concerning this problem in Europe. In China, Prof. Chin Wei-no of Peking Art College studied this subject and published a valuable article in Nos. 10 and 11, 1958 of the journal Wen-wu San-k'ao Tzu-liao. Based upon these new sources I should like to take up this subject again and discuss about the interre lation between pien-wên and wall-paintings. The stories concerning “Rāudrakṣa's fight with the saint” found among the Tun-huang manuscripts are as listed below and can be roughly classified by their contents into two groups. Group A. (1) Collection of Mr. Hu Shih: When the missing first part of this copy is supplemented by Stein S5511, this forms a complete text of a pien-wên entitled Chiang-mo Pien-wên. This is judged by the calligraphic style to be a copy made during the tenth century. (2) S3498v: This is a segment of the first part of the same Ching-mo Pien-wên. From the date of the official letter on its recto this is known to date from a period a little later than AD949. (3) Collection of Mr. Lo Chên-yü: A segment corresponding to about the middle part of the same pien-wên. Recorded in the publication Tun-huang Ling-shih. Its age has not been identified. (4) P4615r (Fig. 1～3): A peculiar text consisting of extracts from four parts of the Ching-mo Pien-wên, with brief headings and arranged in an order disregarding that of the original text. The text of the epitaph for Li Ming-chên inscribed after it proves that this was copied in the end of the ninth century. Group B. This is not in the advanced form like pien-wên. consisting of alternation of prose and verse parts, and its contents are also simpler with only four of the magic bouts. It is in a construction even earlier than that of the Hsien-yü-ching, the sutra from which the Ching-mo Pien-wên was derived. There are two texts of this kind in La Bibliothèque Nationale in Paris, namely P 3784 (Figs. 5 & 6) and P 2344 (Fig. 4), both dating fr
