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

Excavations were carried on from the 18th to the 28th of December at the site of the
ruined station of Lou-lan (L.A.), once guarding the ancient Chinese
Explorations at route to the Tārim, and at other ruins to the north-west (marked L.B.)
ancient Lou-lan. with important results. ³¹ Constant supervision on my part and the
surveyor's ailing condition precluded topographical reconnaissances beyond the immediate
vicinity of the ruins, demanded by the interest of the ground both from geographical and
archæological points of view. Thus the task of surveying the ancient delta of the Kuruk-
daryā ('Dry River') which once had brought water to this now utterly desolate region, or of
tracing the line of the ancient Chinese route where it passed across the great dried-up salt
sea eastwards, remained for my third expedition.

On completion of such exploratory work as the condition of the hard-tried men and
our limited store of ice permitted, I sent the main camp under the
Desert crossing to surveyor back to Abdal while I struck across the wholly unexplored
Tārim. desert to the south-east. Seven trying marches, almost wholly over
bare dunes, heaped up at intervals into high ridges or 'Dawāns', brought me on January
3rd, 1907, to the line of lagoons formed by the Ilek branch of the Tārim. ³² By following
them up to the small ruined site of Merdek-tim and subsequently proceeding down to Lop
where the delta of the Charchan river joins the Tārim at the final eastward bend of its course,
a useful addition was made to our surveys of the terminal depression in which the united
drainage of the Tārim basin is lost. ³³

From Charkhlik I returned to the ruins near Mirān. Their exploration under very
trying climatic conditions yielded abundant finds of interest and detained
Marco Polo's route me till February 11th. Then after needful preparations at Abdal I set
through Lop desert. out with the surveyor for the desert journey of three weeks by the
lonely track, once followed by Marco Polo but almost forgotten for centuries, and reached
Tun-huang on the westernmost marches of China proper. This route, some 330 miles long,
leads first by the southern shore of the great dried-up salt basin marking the pre-historic
Lop sea, then up a wide desert valley by the foot of the southernmost Kuruk-tāgh range,
and finally through the terminal basin, and along the lowermost course, of the Su-lo-ho
river. ³⁴ Its careful survey proved of very considerable geographical interest.

From this terminal basin onwards I traced important, and, owing to the extremely
arid climate, in many parts remarkably well-preserved, remains of an
Discovery of ancient ancient fortified border, a true Limes, which the Chinese Emperor
Chinese Limes. Wu-ti, towards the end of the 2nd century B.C., had constructed for
the protection of the earliest line of China's expansion into Central Asia. The exploration of
this ancient Limes which was subsequently traced for a total distance of over 160 miles
west of An-hsi, formed a fascinating and fruitful task for more than two months after my
arrival at Tun-huang. The ground, almost all desert, over which the wall with its watch-
towers and military posts had been built, was as interesting from a geographical point of
view as the ruins in their archæological and historical aspect. Hence all the more care was
bestowed upon an exact topographical survey of it.

The work was started on the Limes portion extending to the north-east of the
Tun-huang oasis ³⁵ and subsequently after a visit to the outlying small
Explorations along oasis of Nan-hu, the ancient 'Yang barrier', continued along the whole
ancient border line. length of the Limes westwards. This was found to run parallel to the
Su-lo-ho bed from its outlet at the western end of the Khara-nōr lake and to extend to the
southern extremity of the great marsh basin where the river terminates, fully a degree of