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

of Rakas-tal.¹ He also says of Kailas and the Manasarovar: ›Both deserve a fuller
description than I can possibly give, for we were only able to reach the lake once,
and to examine the mountain from the road . . .‹ He correctly says that the Tibe-
tans usually call the lake ›Tso Rimpoché‹. Somewhat exaggerated is the statement
that ›within a radius of a few miles rise four of the greatest rivers of India — the
Indus, the Brahmaputra, the Sutlej, and the Ganges . . .‹ To the sacred lake he
gives 100 square miles and it ›is nearly square in shape‹, which indeed may be said
to be as true as to say it is round. For it keeps the medium between a square
and a circle or ellipse. It is fed by numerous small streams from the ›Mémo‹ and Kailas,
but by no river of importance, although the Tage-tsangpo must be said to be of
great importance and much greater than all the other rivers. The fact that the
›extremely salt‹ lake Gunchu-tso was frozen over from shore to shore to a depth of
several inches, whereas the fresh-water lake Manasarovar had only 100 yards of ice
round the edge, Rawling correctly explains by one of two alternatives: ›either the
lake (Manasarovar) is of excessive depth, or else it is fed by hot springs‹. The
depth is the principal cause, but there may be comparatively warm springs as well
at the bottom of the lake. He regards the channel-problem as solved by the ex-
pedition, but has nothing new to say about it, except that the Rakas-tal ›receives
the surplus water of Manasarowar Lake, and is, as far as could be seen, devoid of
islands‹. The first statement was known already by the Lama surveyors, the second
is not correct, as there are some 3 islands in the southern part of the lake. On
one of them I passed a very uncomfortable night in 1907.

Together with Ryder he visited the bed of the Satlej. ›Many have supposed
that the Sutlej runs from this lake (Rakas-tal), and as Rakas-tal and Manasarowar
are connected, it stands to reason that the Sutlej would actually originate in one of
the streams that run into Manasarowar Lake.‹ Here Rawling touches upon the
correct view, namely, that one of the feeders of Manasarovar must be regarded as the
origin of the Satlej, but he did not keep to this view as he and Ryder did not find
any signs of an outflow. He was told that there had been one 50 years ago, i. e.
about 1854, when ›the overflow took place annually‹. Now they found thick in-
crustations of salt in the bed, which ›would not have been the case had water flowed
over it during the last few years. When five miles distant from the shore, we ob-
tained from the sand-marks in the bed of the stream conclusive proof that the Sutlej
rises a few miles to the west of the lake . . .‹ This is not exactly the same view
as Ryder expressed, for he considered the source of the river to be situated
in the hills north or south of the river bed. Both Rawling and Ryder later on
abandoned their original views. Rawling says:² ›Ryder, I consider, is correct in
placing the present source in a branch flowing from the southern hills.‹ But four