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

portion is visible in Fig. 295, projected about 6 inches, and was moulded of stucco put over a frame-
work of wood and reed wattle. Its height was about 2 feet. It had several rows of holes left by
the wooden pegs which once fastened the individual small relievos of the frieze to the cornice.
The evidence thus gained as to the technicalities of the frieze does not help us to reconstruct the
relievo decoration as a whole from the disjecta membra recovered. But the number of interesting
and relatively well-preserved pieces among them is great enough to justify a separate notice here.
The value of the small relievos from xii is much increased by the preservation in most cases of
their original bright colouring.

It is impossible to guess what scene or scenes, sacred or quasi-profane, the frieze as a whole Warrior
may have represented. But certainly the proportion of pieces belonging to warrior figures is figures of
striking. Plate CXXXV shows one of these almost completely reconstructed from detached pieces frieze.
(Mi. xii. 008, 0010, 0015. 0017). We see there the warrior's head and body protected by scale armour
of a type closely resembling that actually illustrated by the lacquered leather scales excavated in
the Tibetan fort of Mirān.³ A particularly interesting feature of the armour is the shield, which
is represented in several varieties. One, evidently meant for leather, is round and fitted with five
bosses (Mi. xii. 0018, Plate CXXXV). Another, plainly intended as of metal, has its centre occupied Shield with
by an elaborate Gorgoneion relief (Mi. xii. 0020, Plate CXXXV). That this goes back to a classical Gorgo-
prototype is quite certain. What adds to the interest of this shield is that it definitely proves the neion.
corresponding grotesque head, which is found so frequently as an appliqué ornament on terra-cotta
vases from Yōtkan and other Khotan sites, to be directly derived from the model of the classical
Gorgon's head.⁴

Of other human figures may be mentioned a Bodhisattva, Mi. xii. 001 (Plate CXXXIV), modelled Methods of
in pure Gandhāra style, and a cross-legged male excellently posed, Mi. xii. 007⁵(Plate CXXXV). production
The head Mi. xii. 006 (Plate CXXXII) has an interest of its own because it well illustrates both the from
method by which all this stucco statuary was produced from moulds and the clever way in which moulds.
this method could be used for introducing variations. The face is undoubtedly produced from
a mould which served for one type of head in the warrior figures; it still retains the bar of the
helmet passing down the forehead. But instead of the helmet the head bears a mass of hair
brushed up straight in an almost grotesque fashion. A very clever piece of modelling is the head
Mi. xii. 005 (Plate CXXXII), which combines a grotesque weeping face of an unmistakably 'Mongolian'
type with an elaborate Indian head-dress familiar from Gandhāra statues. The naturalistic skill Naturalistic
with which animal figures could be handled by the artists who produced the moulds for this wealth animal
of relievo decorations is attested by pieces like the excellently observed camel's head, Mi. xii. 0025, figures.
and the forelegs of a prancing horse, Mi. xii. 0023–4 (both in Plate CXXXVI). Fragments like
Mi. xii. 0028, 0030, 0031 (Plate CXXXVII) show architectural details which are constantly met with
in Gandhāra relievos and are there plainly recognizable as borrowed or developed from classical art.
In the wooden finial of a small model Stūpa, Mi. xii. 0027 (Plate CXXXVIII), we find reproduced with
equal care all the details with which we are familiar in stone from miniature representations of
Stūpas as they used to be built on the Indian North-West Frontier.

Were we to look at these relievo remains without any knowledge as to where they were Date of
found and as to the methods by which they were reproduced, we might well feel inclined to relievos.
postulate for them an origin centuries nearer to the period that created their models in the Graeco-
Buddhist sculpture of Gandhāra. So it is well to lay stress on the conclusive numismatic evidence