FORMATION AND EXTENT OF DESERTS. 105 
a decrease in the recorded size of the lake, even though the towns of Lulan were 
being abandoned and their water was being set free to reinforce the lake. Then, 
in the Middle Ages, there was an expansion of the lake, which can not have been 
due to diminished use of the rivers for irrigation, for the population of the Lop 
Basin at that time was greater than now, though not equal to that of the flourishing 
Buddhist times, a thousand or more years earlier. Finally, during the last few 
hundred years there has been a decrease both in the size of the lake and the popu- 
lation about it. If Lop-Nor alone is considered, this sequence of events is not 
proved by compulsory evidence in all particulars; but it fits the facts better than 
any other theory as yet suggested. And, more than this, it agrees with all the data 
which I gathered from the whole of the 1,500 miles of longitude and 400 of latitude 
of the Lop basin and from Kashmir. All the facts are explicable on the theory of 
a secular change of climate from moister to drier conditions, with a rapid intensifi- 
cation in the early part of the Christian era and a slight reversal in the Middle Ages.* 
Hedin, who has carried on extensive explorations, holds views quite 
different from those of Huntington and sees no evidence of climatic 
change in the extensive disturbances of the population and pursuits of 
the inhabitants which he notes. 
Evidence upon a matter of this kind is not easily sifted, and a spirit 
of controversy prevails in most of the writings upon the subject. The 
short-period oscillations of climate, which swing back and forward with 
average amplitude; the movements of nomads and half-civilized people 
in response to these changes, and the abandonment of schemes for divi- 
sion and use of water for agricultural purposes, may in some instances 
give remains simulating the effects of increasing desiccation. Exact 
meteorological observations extend over so brief a period that they throw 
no light on the main question. (See R. DeC. Ward, Changes in Climate, 
Popular Sctence Monthly, 1906, vol. 69, p. 458.) Much has been written 
upon the subject by explorers and investigators of archeological remains 
in the deserts of northern and southern Africa. A review of these writ- 
ings would be out of place in the present volume. 
INFLUENCE OF THE DESERT ON LIFE. 
The geological and climatic changes by which a region varies from a 
cool, moist, equable climate, with well-leached soils and complete drain- 
age, to an arid basin, with saline or alkaline soils, imperfect drainage, 
great evaporation, and wide diurnal and annual changes in temperature, 
entail, of course, sweeping modifications in the composition of the flora, 
and consequently the fauna. Such movements would put vegetation in 
a state of continued high stress. Forms in existence at any given time 
would gradually perish as their habitats were narrowed and the limits 
of their power of accommodation and acclimatization were reached. 
New forms which arose by a supposed gradual modification and selection 


*E. Huntington: The Rivers of Chinese Turkestan and the Desiccation of Asia. 
Geographical Journal, October 28, 1906, p. 352; and Lop-Nor, A Chinese Lake, Bul- 
letin American Geographical Society, February and March, 1907; see p. 146. 
