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

exceptional in this case as it does not protrude. At each side it bears a pair of
stags in relief, facing each other with heads turned outward, superficially rendered,
but nevertheless comprehensible. The long handle widens imperceptably at the
top and is uncrowned. Its decoration is composed of five pairs of vulture-heads,
looking down. All of these motives are very old in the Steppe circle. In this
case their fairly good plastic quality may permit them still to belong to the
middle of the I millennium.

About the same period fits the dagger of Plate XXXIX no. 2 crowned
with an arc, double channelled and with a pronounced guard, which corresponds
to numerous examples from Minussinsk (18). As in that region, the arc ends
with two facing eagle-heads. At times these are replaced with simple antennae
in nearly every region to which this type has spread. The guard formed of two
eagle-heads, back to back, is rare.

At Minussinsk the guard is usually in the form of a heart or of a trapezium.
These two forms are also found in our region (Plate XXXIX no. 4). In this
example the separate handle has been cast in a darker alloy, and treated up to
the bird-heads as one field, decorated at each side with geometric forms derived
from the double spiral. The ornament only partly conforms to the outline.
This dagger has already reached the end of the millennium. The drawing on the
blade is exceptional. On each side it shows a long-tailed salamander.

Plate XXXIX no. 3 demonstrates with what ease Chinese elements may be
connected with those from Eurasia (unfortunately only the back which lacks
some of the ornamental motives has been reproduced). The guard with a geo-
metric design at the centre and outward-turned bird-heads at each end carelessly
cast, is derived from the Steppes. The handle which ends in an elliptical and
open ring is formed of a phoenix-head with a small spiral at the end, and is
entirely Chinese (19). Examples found in Korea belong to the III-V centuries.
This Chinese formation is entirely fantastic, as much because of the head as for the
connection with the arc. Our example has kept the original form, but has varied
it in a way unknown until now. At the end of the handle is a ram-head which
is connected with the arc by the horns. When we date this piece at the second
half of the I millennium, we are influenced more by the decorations on the guard
than by those of the handle which seem to belong to a higher art, and which
therefore are a reminder of greater antiquity.

We have mentioned that Chinese elements also survive at the northern
frontier. They are often found on a frequent Minussinsk type where there is
a vertical perforation through the top and which probably held a pin. In this
case the guard then becomes smaller in size (Plate XXXIX no. 5). This example
has a monster-mask broken up with spirals which correspond to a style called in
China, Ch'in. The confused interlacing of the design on the flat button at the
top of the handle also corresponds very well with the Chinese group. But they