Chap. IV] NOTES ON SHEET No. 38 95
shown for Tun-huang, 94° 47', is itself affected by the erroneous adjustment of
R. B. Lâl Singh's triangulation which, as
explained above under Sheet No. 30, brought the easternmost triangulated point on the
Mirân—Tun-huang route by 5' 10" too far to the east, it is possible that the longitude value adopted for An-hsi is also somewhat overestimated. This suspicion is borne out by Mr. Clementi's chronometric longitude value for An-hsi, which is 95° 47' 20.6", and by Dr. Vaillant's values for Tun-huang and Hung-liu-yüan (D. 2), N.W. of An-hsi, each derived from observation of two lunar occultations. These are 94° 36',5 and 95° 23',7 against cire. 94° 47' and 95° 33' respectively in our sheet. '24
Roborovsky's observations at Tun-huang (Sha-chou), based on chronometric values and taken with special care, indicate for his station (a short distance from the town and almost due N. of it) the longitude of 94° 42' 24". 24a This agrees very closely with the position derived from our surveys and shown in this sheet, 94° 47', if allowance is made for the correction of — 5' 10" which, as just mentioned, has to be made in the longitude of the easternmost triangulated point near the Miran—Tun-huang route.
It deserves to be further noted that the plotting of available traverses from the Tibet side brings An-hsi to a longitude of about 95° 52', and that the Russian Trans-frontier map shows one of approximately 95° 58'. The various routes radiating from Tun-huang and An-hsi are controlled by the comparatively large number of latitude observations recorded below.
Regarding the historical topography of the An-hsi—Hami `highroad', see the references given above for Sheet No. 37. The geographical features of the lower Su-lo-ho basin have been fully discussed in Serindia, ii. pp. 578 sqq., with special reference to the natural line of defence offered by the Su-loho for the earliest Chinese road into the Tarim basin past Lou-lan. There, too, I
24 Cf. La Géographie, xxxv (1921). p. 499. The uncertainty inherent to all astronomical longitude observations under ordinary travel conditions is illustrated by the fact that the same observer's list shows for Ch`ion-fo-tung (B.4) a longitude of 96° 6' which is quite irreconcilable with the position of this place relative to Tnn-huang, the longitude difference indicated being at least 16', if not more, in excess of the
have indicated the importance of the large oasis of Tun-huang with regard to this road and the irrigation facilities it derives from its situation on the alluvial fan of the Tang-ho, the largest tributary of the Su-to-ho.
For descriptions of the desert belt, both marsh-edged and gravel `Sai', extending
along the Su-lo-ho from the Khara-nôr lake towards the cultivated area of An-hsi, and followed by the line of the ancient Chinese Limes, see Desert Cathay, ii. pp. 5 sqq., 40 sqq., 131 sqq.; for the oases of Tun-huang and An-hsi, and the important sacred site of the `Thousand Buddhas' near the fotmer, see ibid, ii. pp. 10 sqq., 159 sqq., 235 sqq. The topography and remains of the Limes sections falling within this sheet are treated in Serindia, ii. pp. 585 sqq., 717 sqq.; iii. pp. 1089 sqq.
Within the area of the sheet we distinguish three well-marked zones. In the north the surveyed route towards Hami, first opened by the Chinese in A.D. 73 and since then a main line for China's Central-Asian expansion, crosses in succession the much-decayed hill ranges of the Central Pei-shan. Those shown on the map between the stations Shach`üan-tzu (.B.1) and Pi-ting-tzu (D.3) are manifestly connected with the five Pei-shan ranges distinguished by Professor Futterer's very careful topographical and geological survey along his more easterly route from near Mu-t`ou-ching (No. 37. D. 4) to near the Su-lo-ho bend (No. 40.B.4). 25 On both routes water and scanty grazing can be found only in a few isolated depressions of the broad desert valleys separating those ranges.
The low southernmost range of the Peishan towards the west seems to merge in an
outlier of the Kuruk-tâgh (A,B.3). Be-
tween it and the foothills of the Nan-shan in the south extends the trough of the lows: Su-lo-ho valley. The portion lying to the west of Tun-huang entirely shares the character of the delta and terminal basin of the Su-to-ho, as shown in Sheet No. 35. C,D.4.
The abundant supply of water provided
24• bee Scientific, Results of Roborovsky's Ezpedi• lion (Russian), Astronomical Observations, p. 7; also his map, scale 20 versts to 1 inch.
25 See Fetterer, Geograph. Skizze der Wüste Gobi, in Petermann's Mittheilungen, Ergänzungsheft No. 139, pp. 11-22 and map.