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

caravan track along the southern shore of
the salt-encrusted bed of the Lop sea. For
its description and its historical topography,
see the reference given above in the Notes
on Sheet No. 32. As all the springs along
this portion of the route are salt, it is used
by caravans only from about December to
early April, when the ice formed at those
springs can be used for the supply of drink-
able water. At other seasons the hill route
along the northern slopes of the Āltin-tāgh,
an easternmost extension of the K'un-lun, is
alone available. This route, too, is rendered
very difficult by the rarity of springs and the
extreme barrenness of the ground. Regard-
ing the historical topography of this route,
mentioned in early Chinese records and now
known to the Lop people as tāgh-yol, 'the
mountain route', see Serindia, i. pp. 320,
418; ii. pp. 549, 622.
The portion of the Āltin-tāgh actually
surveyed does not appear to reach the snow-
line at any point and is characterized by
extreme aridity. The wide glacis of gravel
and detritus descending from it towards the
ancient Lop sea-bed is overrun in parts by
high drift-sand and is even more barren than
the range. The occasional flood-water from
the mountains is absorbed on this glacis.
But the scanty subsoil drainage thus created
probably accounts for the few salt springs
and adjoining narrow reed-beds to be found
along the well-marked shore-line of the salt-
encrusted sea-bed at Lowaza, Kōshe-langza
and Panja (A-C. 1).
Corrections. Replace height figure 1890
at Lowaza, Camp 57, by 2037 (cf. Appen-
dix B).

Astronomically observed latitudes.
1906-08. Panja, Camp 147 (at eastern spring; C.1) ... ... 39° 59′ 31″
1913-15. Tāsh-köl, Camp 33 (C. 2) ... ... 39° 9′ 4″
Lowaza, Camp 57 (near spring; A. 1) ... ... 39° 46′ 5″

NOTES ON SHEET No. 34 (BARKUL, HĀMI)

Of the surveys recorded in this sheet,
those to the north of the Tien-shan and a
few to the south of it belong to the third
expedition, while most of the work on the
latter ground dates from the autumn of 1907.
The traverse along the main caravan road
passing through the Hāmi or Kumul oasis
had two observed latitudes and was adjusted
on the positions adopted for Pichan (see
Sheet No. 31) and An-hsi (No. 38).
The location thus derived for the town
of Hāmi (C. 3) agrees very closely in longi-
tude (circ. 93° 26′) with that shown by the
Russian Trans-frontier map, but falls some
8 minutes to the east of Mr. Clementi's
chronometric value (93° 18′ 16″). On the
other hand the position obtained in the same
way for Hoang-lung-kang, a small village
on the highroad a short march to the south-
east of Hāmi, agrees almost exactly in
longitude (93° 44′ 40″) as well as in
latitude with that observed by Mr. Clementi.
The traverse made by M. Muhammad Yakūb
from Toghucha (B. 2) to Pichan (Sheet No.
31. A. 3), when adjusted from the latter
place, was found to indicate for Toghucha
a position closely agreeing with that of the
former survey.
On the route passing along the nor-
thern foot of the Tien-shan, the position of
Barkul (B. 1) was fixed by adjustment of
the traverses to Ku-ch'êng-tzu and Turfān,
corrected to its observed latitude; a further
check for the longitude was supplied by the
location of the Barkul-dawān (D. 2) in the
1906-08 survey. The longitude thus ac-
cepted for Barkul (92° 51′ 20″) falls about
4′ 40″ to the west of the one shown by the
Russian maps. The indication of the snow-
line at about 12,000 feet is necessarily con-
jectural, since the range between the meri-
dians of Barkul and Hāmi was sighted on
both journeys at the close of October when
fresh snow had already fallen as low as the
Barkul-dawān (circ. 9,200 ft.).
A short account of my stay at Hāmi
and of the observations on archæological
excursions thence made to the small oases of
Ara-tam (D. 3) and Lapchuk-Kara-döbe
(B. 3) has been given in Desert Cathay, ii.
pp. 342 sqq. Notwithstanding the limited
extent of cultivable ground, Hāmi, owing to
its position on what has since the first
Chinese occupation of the district (A. D. 73)
remained a main line of traffic between
Kan-su and Chinese Turkistān, has played
an important part in China's political and
commercial relations with Central Asia; its