latter on the south, of the southern shore of Lop-nor. But the text tells us, that »on the west of the southern shore of Lop-nor you come to Ike-ghaschon».
The lake region south of Lop-nor is distorted on the map, and the road from Sa-tscheo is incorrectly drawn in relation to the lakes, although the error is less than in the case of the two southern roads. Nevertheless I believe I have discovered the key to the puzzle. You have only to endeavour to interpret these Chinese records in immediate connection with the actual circumstances on the spot, and the matter becomes at once tolerably clear, and as I am the only traveller who has visited the region in question, the interpretation is naturally easier for me than for anybody else. But first let us examine the probable course of the northern route, as it is described in the third part of the Si yii-schuei-iao-ki.
The Wu-tschang map makes it run north of the Bulungir-gol towards the west-north-west. The same map puts Sulutu about 5o km. north-west of the Charanur; but this does not agree with the text, which says that it lies near the northern shore of the lake. The Chinese author goes on to explain that Sulutu means in Turki »Reed hut». The following names: Okhur, Ebüdük, Ike Dürbeldschin, Manitu, and Obulang belong, I feel certain, to the mountainous regions south of the Kuruktagh. And for this opinion I am able to adduce three reasons: (I) In that locality I crossed over a road running east and west, and marked by cairns of stones; this road, there can be no doubt, formerly ran to the ancient Lop-nor, the mountain route being preferred because of the greater hardness of the surface, and possibly also because of the existence of one or more springs; (2) khuduk, the word for a »well», does not occur, and it is only in the lowlands, not in the mountains, that wells are ever dug; (3) Ike Dürbeldschin, the Great Four-square, points to the shape of a valley, and Manitu, the »Mani Place», is a cairn built up of stone slabs, with the Buddhist formula On mane padme hum engraved on them. Cairns of this character, called also obo, are generally spoken of simply as mani or mane. Now a cairn of this description can hardly have been set up in any except a mountainous region, where there is an abundance of stone.* Himly translates Ma-ni-thu as meaning Place of the Mani Banner.
The name that follows next, Khuduk-ussu, or the Well Water, suggests that the road has again left the mountains; and the same inference admits of being drawn from Tschono-khuduk (the Wolf Well), Nalügei-khuduk, and Dalai-khuduk. The road between the two last-mentioned wells is said to run north of Nukitu-ling, although in reality it ought to be north-north-east or north-east of that pass, which I have already identified with Kurghan-davan. Unfortunately we are not told how far north of the pass the road ran, any more than we have been previously told how far the Tarim flows towards the east, south of the Bostan-nor. The name Dalai-khuduk is interesting, for dalai, a Mongolian word, means »sea», consequently, according to Himly, »great». But in this case »sea» is the better meaning, since, as the names of the following stations clearly prove, the road traversed the basin in which the Lop-nor was situated, and that lake was called by the
IIIF Color HighRes
