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

hundred years do we possess a few European observations. The two first years are
unreliable, and the earliest reliable information dates from the Lama surveyors 200 years
ago. The years of observation are too irregularly represented to allow us to draw
regular curves. In the 18th century there is a gap of nearly 50 years without re-
cords, and there may possibly be a depression in the Rakas-tal curve, or in both. In
the first half of the 19th century there is a gap of 25 years, and in the latter half
there is one of 32 years, of which nothing is known. From the last 10 years, on
the other hand, there are no fewer than seven observations. This uneven distribution
of the years over the period in question shows how difficult it is to extract really
reliable laws from the material existing.

One thing seems to be evident: a proceeding desiccation during the last 200 years.
For, disregarding the more numerous observations of later years, the signs become
more numerous as we approach our own time. Do these 200 years represent a
fraction of the general post-pluvial desiccation, or do they only represent the acceler-
ated change of climate during a depression belonging to a period of lower order?
I believe this latter supposition is the right one, and under such conditions we should
expect a return to a maximum during the next 200 or 300 years. But the former
supposition is not excluded, if we admit that the lakes are such a sensitive instru-
ment, such a delicate self-registering apparatus, that it would permit us to read
climatic changes, which are hardly noticeable on ordinary instruments. Such an instru-
ment we find in the movements of the glaciers. But nobody would pretend that the
general post-glacial climatic changes could be controlled by readings within 200 years.
This would at least not be possible in Asiatic regions. For we know of passes, as
the Mus-tagh, which were used in former times, and now are impracticable. The
glacial researches in the mountains north of India deliver abundant arguments for
general retreat in post-glacial time. But nothing is known about the speed, and the
last 200 years are too short a space of time to prove anything. We can only
follow the changes in the period of the lowest degree, the advances and retreats
during some 50 years.

The oscillations of the Manasarovar and Rakas-tal, on the other hand, should
permit us, if we had the possibility of reading them constantly, to follow extremely
closely the curves of at least three different periods. The lowest falls within the
course of a year, a rise of the level from the beginning of the rainy season, and a
fall from its maximum towards the minimum of the dry season.

The next is the one which may be called the Manasarovar period. It will be
well understood if we consider the five years from 1905 to 1910; 1905 was regarded
as extremely dry; 1907 was very dry with only very occasionally short rains; in 1908
there was a good deal of rain, specially abundant in July and August, and making
all the water-courses, coming down from the north to the upper Satlej, grow into rivers,
difficult to ford. In 1905 the level of the Manasarovar must have been unusually low;
in 1907 it stood, even during the rainy season, 2.263m. below the highest point in