292 THE LOP-NOR PROBLEM.
Chinese »sea» in various compound forms. The Great Well is therefore an improbable name; it ought to be Deep Well, or something similar.
The text then goes on to say, that the road continues »to Temen-ghaschon south of Lop-nor». But on the Wu-tschang map, on which the western half of the road, as we have 'seen, must be shifted a little to the south, we have Temmenai-khuduk. Upon this Himly makes the following comment, »Mongolian lemon = »camel»; noir = »sleep» ?; khuduk = »well»; that is to say, the Well of the Camel's Sleep, or the Sleeping Camel's Well.» Against this interpretation there is no objection to make, for that is precisely the region in which wild-camel are found, and it is easy to suppose that one was on some occasion seen near a well. But the text appears to me to give much the likelier version when it goes on to add, »towards the west to Temen-ghaschon south of Lop-nor, and farther southwards to Bagha-ghaschon». Ike-ghaschon is not mentioned at all. Why is that? Simply because Temmenaikhuduk and Ike-ghaschon are one and the same, namely Temen-ghaschon, a salt lake on the shore of which there possibly was also a well, and this has given rise to the confusion.* The transcription of Temmenai-khuduk adopted by the Chinese map is, according to Himly, Thö-mön-nai-hu-tu-k'ö. By analogy therefore the Temenghaschon of the text ought properly to be Tömön-ghaschon or Tömön-koschun; the words tömön, l men, lömenki mean in Turki the »lower», the »nether». ** Hence it is some Lower Lake which we have to consider; and it was clearly called by the Mongols Ike-ghaschon, or the Big Salt Lake, by way of contrast to Bagha-ghaschon. My survey proved that Kara-koschun, which in my opinion is certainly identical with the Ike-ghaschon and Temen-ghaschon, lay south of the ancient Lop-nor. Consequently in this case the text is right, but the map is wrong, that is to say with regard to the road and the small lakes.
In the third part of the Chinese document we are told, that the Bagha-ghaschon lies south of the Tömen-ghaschon, but in the parts dealing with the Lop-nor that the lke-ghaschon lies west of the southern shore of the Lop-nor, and that the Baghaghaschon lies west of the Ike-ghaschon. This contradiction is not however so serious as several others which have been pointed out above, for at the present day the western parts of the Kara-koschun lie south-west of at all events the eastern part of the old Lop-nor, and Chinese itineraries use scarcely any others except the four principal quarters of the compass, so that in this particular the writer may have been in doubt whether he should say south or west.
I have already stated (p. 286 and 287) that in my opinion the author of the itinerary, in describing the southern route, via Khas-nur, has been guilty of interchanging two routes, or at all events of an anachronism, in that, after bringing us to Nukituschan-k'ou, he goes on to say that, when we are on the southern shore of Lop-nor, we have the Ike-ghaschon to the west. Lop-nor and Ike-ghaschon both belong to the northern route, which is more distinct and easy to follow. It is true, the two routes may conceivably have united at Ike-ghaschon, so that the last portion, west from that point, was common to the two, as is indeed the case at the present time, when
IIIFカラー高解像度
