THE LOP-NOR - KOSLOFF AND THE AUTHOR. 275
from a different person, namely Kuntschekan Bek, and it carried us back to the year 172o as the date of the change in the river-bed — a date for which Richthofen rightly substituted 175o. Hence it is the same tradition upon which both Pjevtsoff's calculations and my own rest, and there exists no reason to doubt its trustworthiness. The course of the lowermost Tarim from Schirge-tschapghan, via Tschigelik-uj, Kara-buran, Jurt-tschapghan, Abdal, and right on down to Kumtschapghan is therefore a new creation, dating from the year 174o or thereabouts. Previous to that that region was, as the existing vegetation proves, a desert, or at the most a scanty steppe-land. It is possible, indeed probable — though by no means indispensable — that the Kara-koschun had even then approximately the same position as now, excepting only the more northerly lake of Utschu-kul and its generally greater extent. This is of course on the assumption that the lake was not, as it is now, undergoing a process of shrinkage, followed by one or more periods of increase and expansion. On the other hand the observations of Hallerstein and d'Espinha have nothing in the world to do with the matter, for in any and every case the Tarim did flow past the present Ajrilghan, or at any rate in its immediate vicinity.
If now the lowermost Tarim from Schirge-tschapgan downwards has occupied its existing bed for about 16o years, and yet no toghrak forest has succeeded in springing up along its banks, the question arises, how long did the river need to occupy the bed of the Tokus-tarim (Schirge-tschapghan) before the forest which stood beside it attained the maturity and development exhibited by the decayed tree-trunks ? If it is permissible to argue from analogy, one would reply — A very considerably longer period than 16o years. This river-arm originated probably in the beginning or middle of the 16th century, unless the spread of the forest to its banks was helped by exceptionally favourable circumstances. The lowermost Tarim would appear however, at all events for a time, to have been divided into several arms, for I was told that the northern lakes (Pjevtsoff's Utschu-kul) likewise derived their water from a point between Kulatscha and Nias-köl. If, when we consider the hydrographical relations of this flat region, we assume the century as our unit of time, it results that the changes effected here must be beyond the bounds of conception. I have myself seen how a single year will suffice to render a radical alteration necessary in the map. Unfortunately the historical data we possess with regard to the Lop country are extremely meagre; and yet it is inevitable that it should be so. Of the changes that do take place it is only exceptionally that the effects remain impressed upon the face of the country. In general they are relatively soon obliterated through the effects of the wind and the levelling agency of the river-sedimentation. It is only in imagination that the traveller who has traversed the country of Lop in various directions can venture to take a hurried glance backwards down the centuries, and during the last two thousand years picture to himself a fresh map at the turn of each succeeding century. It is only from the sources and down to the district of Karaul that a series of maps such as this would give us a tolerably permanent picture of the course of the Tarim; and yet even there pretty thoroughgoing changes have taken place during the last few years. Below Karaul however the maps would reveal a very restless and inconstant stream. There exists no reason to suppose
IIIFカラー高解像度
