National Institute of Informatics - Digital Silk Road Project
Digital Archive of Toyo Bunko Rare Books

observed latitudes in the azimuth derived
from this setting. All surveys were done
in May-June, 1914, when the atmospheric
conditions of the season seriously interfered
with both astronomical observations and dis-
tant views.

The terminal course of the Etsin-gol
and the drainageless basin into which it
carries all the moisture brought down from
the ranges of the Central Nan-shan, are of
distinct geographical interest, both in their
physical aspects and with regard to histori-
cal topography. The route leading along
the river has always possessed importance as
a great natural highway from the Mongolian
steppes into westernmost Kan-su and thus in-
to China. This has been briefly explained
by me in Geograph. Journal, 1916, xlviii. pp.
196 sqq., with special reference to Marco
Polo's notice of the 'City of Etzina', marked
by the ruins of Khara-khoto (45.C.1), and to
Chingiz Khan's conquest of Kan-su. There,
too, attention has been drawn to the strik-
ing parallel presented by this route to that
once leading from Tun-huang, past the
ancient Lop sea bed, to Lou-lan and the
Kuruk-daryâ delta. The evidence of desic-
cation noticed along the Etsin-gol aptly illus-
trates conditions such as are likely to have
prevailed in the Lou-lan area before its final
abandonment.

The riverine area comprised in our sur-
veys falls into three distinct sections. From
below Mao-mei to the outlying rocky spur
of Bayin-bogdo (45.B.2), the river is con-
fined to a single wide bed, lying in a trough
which is flanked on the west by the steep
gravel glacis of the Pei-shan and on the east
by one sloping down more gently from the
westernmost hill chain of the Ala-shan.

At the southern end of the Bayin-bogdo

spur the river spreads out into a steadily
widening delta. Among a number of trace-
able branches but few ever receive water now,
and these, too, only during the short season
of summer floods. For the greater part of
the year water can be obtained in them only
from wells dug in deep hollows at rare inter-
vals. Yet in the narrow belts of riverine
jungle flanking the beds, wild poplars are
found living right down to the two terminal
salt lakes, the Gashûn-nôr and Sogo-nôr,
(No. 44. C.3,4).

The presence of these two lakes separat-
ed by a well-marked ridge and occupying
different levels—the Sogo-nôr lies about 200
feet higher than the Gashûn-nôr ²¹—strik-
ingly illustrates the relations which our sur-
veys at the end of the Su-lo-ho delta have
led me to suppose between that river's present
terminal marsh bed and the ancient lacustrine
basin found at the head of the Bêsh-toghrak
depression. ²²

There is definite evidence, as the map
shows, of the shrinkage which the Sogo-nôr
must have undergone in comparatively quite
recent times, and which is obviously connect-
ed with the drying-up of the eastern beds
of the delta. Having been prevented by
my excavations at the Khara-khoto site from
visiting the terminal depression myself, I am
unable to judge what indications of older
shore-lines, etc., may also be traceable in the
case of the larger lake. In the north the
whole basin is fringed by low hills, evidently
outliers of the great Altai system of Mongolia
proper.

Corrections. 44. C.4. Owang-gol should
be printed in blue.

45. B. 1. For Kôk-zilgda read Kôk-
zigda.

Astronomically observed latitudes.
1913-15. Buk-tokhai, Camp 142 (on right bank of Etsin-gol, near ruined
tower; 45. B.3) ... ... ... 40° 58′ 32″
Atik-tsagan, Camp 145 (on right bank of Etsin-gol; 45.C.1) ... 41° 30′ 27″
Ulân-bürük, Camp 151 (on right bank of western river branch; 44.B.4) 42° 3′ 0″

NOTES ON SHEET No. 46 (KAN-CHOU)

The surveys recorded in this sheet re-
present the easternmost extension of our
work. They were made mainly in June-
August, 1914; but some of the routes west