No. 7] 451

104. On Ancient Musical Score of P'i•p'a ~, Discovered at Tun•huang By Kenzo HAYASI Nara GakugeiUniversity (Comm.by S. NAKAMURA,M.J.A., July 12, 1956)

1. The score in question is now kept in the Bibliotheque Nationale in Paris as Pelliot's document, No. 3808. The text consists of 25 musics, of which No. 18 and No. 25 are reproduced in Fig. 2. The author considers that the important part of these musics seems to have prevailed at the end of the T'ang dynasty (618-906) in . The p'i-p'a used at the T'ang dynasty is a stringed instrument, which is supposed to have been brought to China in the 3rd or 4th century from Iran. The most prosperous period of the T'ang dynasty in China (latter half of the 8th century) corresponds to Tenpyo' Period in Japanese history. At that time, five very fine specimens of p'i-p'a of T'ang type and also musical score of the same period have been brought to , and kept in Syoso-in _lEt in Nara, a city in central Japan. Syoso-in is a kind of repository or a museum established in the year 756, for the memory of the Emperor Syomu, keeping almost all kinds of documents, medicines, various articles used in daily life, etc. representing the cultural condition of Japan at that time. These p'i-p'a's have 4 strings and 4 frets (Fig. 1), just equal to the type of early Arabian (el') and played with . At present, in the court music and court dance of Japan, such instrument is still used and called biwa. Twenty tones emitted by biwa are denoted by symbols or letters shown in Table I, and pro- nounced as in the table.

It is fortunate for us to find that these letters were used in the Tun-huang score and also in Syoso-in score, and also in the present scores for biwa, used in the present Japanese court music and dance. These letters, of course, suffer modifications with time, but the changes are very slight, and there is no difficulty in identifying them, as shown in Table II. 452 K. HAYASI [Vol. 32,

The p'i-p'a score discovered at Tun-huang by Pelliot shall be called hereafter Score I. It contains 25 melodies as already mentioned. In order to study this score, we have at our disposal, the following 3 scores to help us. Score II. T enpyo biwa score' ' . This is a fragment of biwa score of Tenpy® Period (circa A.D. 707) discovered and kept with a pile of ancient manuscripts at Syoso-in repository in Nara. Score III. Gogenhu ELi,k . This is a score for five-stringed biwa (Chinese, wu-hsien p'i-p'a), brought to Japan at the Tenpyo Period, and is supposed to have been made at the most prosperous period in the T'ang dynasty. The text had long been buried in oblivion untill it came to be recognized in this century. It consists of 28 musics. This is at present in the possession of the Yomei Library in Kyoto, Japan. Score IV. This score is Hakuga-tekihu 1~~p alias Tyotikuhu ' p . It is a partial score edited by Minamoto-no-I3iromasa1, 1 (a noted Japanese musician in the 10th century) for horizontal in the year 906, and contains about 50 melodies in T'ang and Japanese Musics. 2. The fundamental tuning of biwa is set in such a way, as eachstring is arrangedwith an interval ofquarte (),3- to thead- o jacent string, and the positions of the frets for all strings are placed in the following manner, that the notes 0: 1: 2 : 3 : 4 are in the ratio 1 : 8 : 27 : 64: 3 . This tuning is almost the same as in the 9 32 81 4 ancient Arabian lute. When the string I is set on F, the follow- ing tuning for the rest of strings is obtained (Table III). This tuning is still used in the present court music of Bansiki-tyo X{1~1(mode of u), As regards to the way of tuning , it is known that several tens of them must have existed in the T'ang dynasty, but as time passed, some of them be- came obsolete, and about 20 kinds of them were left in both theoretical and No. 7] Ancient Musical Score of P'i-p'a ~, Discovered at Tun-huangkk 453 practical types at the time, when Huziwara-no-Moronaga (1138- 1192), a first class musician, compiled his Sango Yoroku t, the most perfect biwa score at the 12th century. The tunings practically used at the present time are 8. 3. No mention about tuning is found in Score I. It is ob- vious, however, that no score can be deciphered without knowing the tuning used in them. The first step to study any score is to con- jecture the tuning applicable to the score, and then find out the reasonable tuning for it. We made a thorough examination of all tunings used in old music books, to understand the principles of tuning, and further studied the relation between the music and the mode in the T'ang dynasty, and the mode prevailed at that time. We checked and figured various tunings as far as we could. The tone symbols used in the 25 melodies of Score I are found to be distinguishable into the following three groups, namely, Group I Nos. 1-10, Group II Nos. 11-20, and Group III Nos. 21-25, accord- ing to their different styles in calligraphy; and within the same group each music has similarity in the usage of letters. Attention was paid to the peculiar facts that these musics have in general at their ends particular ending of 3 or 4 tones, succeeding after the normal "fin" as shown in Table IV.

The fact that this ending is placed for all groups, so as to cross all four strings in quick succession implies that the ending contains the most important tones, and these tone letters have therefore great weights among 20 letters used in tunings of strings. This characteristic noticed in Score I, shows us that the musics within the same group are all based on the same tuning; and thus we reached the important conclusion that Score I has three kinds of tunings. Following this line of research further on, we at last succeeded in translating all 25 melodies in Score I into modern form of musical staff. Examples of our final results are given in Fig. 3, in which No. 18 is the whole melody of music Shui-ku-tzu 71~ T, whose text is given in Fig. 2 a, while No. 25 is the first part of the same music. The text is given in the score b in Fig. 2. 454 K. HAYASI [Vol. 32,

Fig. 2. The Tun-huang p'i-p'a-p'u, Shui-ku-tzic 7JJfl (a: No.18 b : No.25)

Minute details of our studies given in § 3, the essential part of the present work, are referred to our original paper, in the Bulletin of Nara Gakugei University, Japan, 5, No. 1, Dec., 1955.