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

of a map the men could not have struggled through to Kaurük-bulak, the nearest of the salt springs
in the Eastern Kuruk-tāgh, in less than three more days, and their animals must certainly have
perished before that. Nor could ice have been found by them in November at Kaurük-bulak
or at any of the other salt springs farther north (Map No. 32. A, B. 2) which Lāl Singh explored
in the following winter.

Probable It was thus highly probable that the ill-fated party of thieves had met in this inexorable waste
doom of with an end involving far harder retribution than human justice would have inflicted for their
rogues. misdeed. Abdurrahīm, when accompanying Lāl Singh in January, 1914, to Kaurük-bulak and
the northern springs previously known only to a few daring hunters of wild camels, could find no
trace of their passage. But we came upon their footprints again on the marches of the next two
days, and followed them right through to the point where their track emerged on ' land ' after
crossing the great bay of the salt-encrusted sea-bed. At two places we found traces of their having
camped or halted, and wondered whether the poor rogues had been troubled by any misgivings
as to the fate which was awaiting them.

OBJECTS FOUND AT FOOT AND ON SLOPES OF MESA, SIX MILES EAST OF CAMP C. CI

C. ci. 01. Iron dagger blade and tang, with broken C. ci. 04. Solid copper ball, pierced with rectang. hole.
cross-piece at top of tang. Edges blunt ; point sharp ; Well preserved. Diam. ⅜″, hole 1⁄16″ × ⅛″. Pl. XXIII.
cross-piece prob. part of guard. Rusted but strong.
Length of whole 9¼″, gr. width of blade ⅝″. Length of C. ci. 05. Copper hook of graceful curve, spatulate at
tang 3″, thickness 5⁄16″ × ¼″. Pl. XXIII. hookless lower end, on back of which is stud. Recurved
end narrower, but slightly thickened. Cast. Corroded
C. ci. 02. Link of iron snaffle-bit (broken), with fr. of through at one edge, or perhaps air-hole in casting ; but
other link rusted on to it. Ring at end corroded away condition generally good. Cf. Anc. Khotan, i. p. 464,
for nearly half its circumference. Section roughly sq. D.K. 001 ; ii. Pl. LI. 2″ × ⅜″ (broad end) 3⁄16″ (narrow
Corroded but hard. Cf. L.A. 034 (Pl. XXI) ; T. XXII. f. 01 end). Pl. XXIII.
(Pl. XLVII). Length 3½″, thickness of rod c. 3⁄16″ ; diam.
of outer ring c. 1½″. Pl. XXIII. C. ci. 06. Small fr. of iron, corroded. ⅜″ × 3⁄16″.

C. ci. 03. Iron skewer with ring handle, as Ser. ii. C. ci. 07. Fr. of marble (?), roughly cuboid, veined grey
p. 775, T. XII. a. 0026 (Pl. LIV) ; see also ibid., T.W. 005, in dark pink and buff. ⅝″ × ⅜″ × ⅜″.
007 ; T. XVIII. ii. 9. b ; XXVIII. 0019. Broken in two,
much corroded. Length 2⅜″, thickness c. 3⁄16″ ; diam. of C. ci. 08. Glass bead, translucent pale green, roughly
loop 1⅛″. Pl. XXIII. rectang. with irregularly chamfered sides. ⅜⅜″ × ⅜″ × ¼″.
Pl. XXIII.

SECTION IV.—THE 'WHITE DRAGON MOUNDS'

Probable When we reached the open plain of clay at the foot of a gentle gravel glacis where our Camp
continua- civ was pitched on the evening of March 2nd, I knew that the most difficult portion of our journey
tion of in search of the ancient ' Lou-lan route ' had been left behind us. We had come upon definite
ancient indications of ancient traffic on ground where that route was likely to have emerged from the salt
route. wastes of the dried-up sea-bed. The configuration of the land before us to the south and east left
little doubt as to the line it had subsequently pursued towards its goal, the western end of the
Tun-huang Limes. From what the plane-table and our mapping of 1907 showed, it seemed certain
that the early Chinese pioneers of that route had only to keep to the eastern shore of the dried-up
sea, as it stretched away to the south in order to reach the mouth of the wide valley-like depression
leading towards Bēsh-toghrak and the terminal basin of the Su-lo-ho (Map No. 32. C, D. 4). How-
ever far into this depression the great eastern bay of the salt-encrusted sea-bed might prove to extend
—our survey of 1907 along the caravan track towards Tun-huang furnished no definite indication
on this point—it was clear, from what I had then observed along the southern edge of the depression,
that desert vegetation sufficient for grazing en route, and possibly water, might be expected also
along the northern edge. It was there, skirting the foot of the barren hill range which from the north