国立情報学研究所 - ディジタル・シルクロード・プロジェクト
『東洋文庫所蔵』貴重書デジタルアーカイブ

•

Sec. vi]   LOKAPALAS AND VAJRAPANIS   873

The characteristics of the other group of Lokapâla figures, distinguished by Miss Lorimer as ` Chinese', have been fully indicated in the Descriptive List and can be easily recognized in the available reproductions.20 Chief among them are the three-quarter profile and sweeping curve of pose ; the body thrown out to waist ; the freedom and movement imparted to' the drawing mainly by the treatment of the flowing drapery ; peculiarities in armour and dress, which latter always includes sandals or string-shoes. The invariably oblique cut of the eyes and a distinct tendency towards the grotesque and occasionally even the monstrous are significant for this ` Chinese group.21 The character of the style leaves no doubt that this type was developed from the former, or ` Central-Asian', under the impulse of Chinese art feeling ; and, considering that all our paintings were the work of Chinese hands, it is only natural that we should find the artistically best Lokapâla representations among this ` Chinese ' group.

The rich armour and dress with which the Four Kings are always depicted, and the manifold and carefully indicated variations in their details, obviously afford abundant and valuable materials for the study of antiquarian questions such as Dr. Laufer has discussed with much learning and painstaking thoroughness in a recent work, rightly described in its sub-title as Prolegomena on the history of defensive armour.22 But this cannot be the place for any attempt at a general examination of them. I must content myself with referring to Miss Lorimer's descriptive notes on the paintings 23 and with the briefest mention of a few points of special archaeological interest. Attention may be drawn to the valuable indications yielded by the comparison of the elaborate scale-armour of the Four Kings with the actual remains of leather mail from the sites of Niya and Mirân,24 or with that represented in stucco relievos found at other Turkestân sites.2b Without going into any details, such as the different arrangements of overlapping and lacing, I may point out the interesting main fact that the scales shown on the skirt portion of the mail coat are almost invariably oblong and those on the upper part round-edged.26 I have before had occasion, with reference to the stucco image of Vai§ravat}a brought to light in one of the Dandân-oilik shrines, to observe that this distinctive arrangement of the scales is found in the armour which two soldiers of Mâra's army wear in a well-known Gandhâra relievo.27 In this connexion it may be noted that chain mail is represented only in a single one of our Lokapâla paintings.28

That the armour and other equipment of the Lokapalas may be safely assumed to reproduce with more or less accuracy those actually in use about the period when the type was evolved seems a priori probable. Nevertheless it is a welcome confirmation to find that both the plain sandals

' Chinese' type of Lokapâla.

Armour and dress of Lokapâla figures.

Details of foot-gear.

oo6, xxxvii. ooz. Owing to the intermingling of types observed in certain banners, we meet with these last-named features occasionally also in the other group; see e.g. *Ch. 0035 (green iris); lv. 0018, 0046. They have their ethnographic significance and may point to the surviving influence of Central-Asian racial prototypes, probably of the Homo Alpinus kind.

20 See below Descriptive List, *Ch. 0035 ; also 0040 (Pl. LXXXV; liv. 003 (Thousand B., PI. XXVIII); xxvi. a. ooz (PI. LXXXV); lv. 0017, oo18 (Pl. LXXXV), 0020 (Pl. LXXXIV), 0046 (ibid.).

21 This tendency is well brought out by the demonic head in Ch. 0098 (Thousand B., Pl. XLViIi), a fr. from a large painting ; see also lv. 0018 (Pl. LxxXV).

22 See Laufer, Chinese Clay Figures, Part I (1914), especially Chap. III—v, pp. 201-305, with many fine illustrations (see e.g. Pl. LXI for the marble plaque of a Lokapâla in armour, closely resembling the figures on our banners and rightly described p. 300 as Virnpâksa and of the Tang period).

1374

23 See below Descriptive List, particularly *Ch. ooro, where the different portions of the armour and accoutrement are systematically described.

24 See above, pp. 246, 463 sqq. ; also Ancient Khotan, i. pp. xvi, 411.

25 For the latter, cf. below, chap. xxIx. sec. iii, iv, and Ancient Khotan, i. pp. 252 sq.

26 See for illustrations particularly Pl. LXXXIV, LXXXV.

27 Cf. Ancient Kholan, i. p. 252, Fig. 3o ; ii. Pl. II; for a reproduction of the Gandhâra sculpture see, e. g., Grünwedel Burgess, Buddhist Art, Fig. 48. For oblong scales throughout, see above, p. 465.

28 See Ch. liv. 003 (Thousand B., Pl. XXVIII), a fine fragment of a large picture, probably representing VaiSravana. Dr. Laufer, Chinese Clay Figures, I. pp. 237 sqq., assigns Persian origin to chain mail. It is first referred to by Chinese records at the beginning of the seventh century A.D. among tribute from Samarkand ; cf. ibid. p. 247.

5T