RESULTS OF CHANGES OF CLIMATE 357 



ture. In one place the climate may have varied from very cold to cold, 

 while in another at the same time it ranged from warm to hot ; or, again, 

 the change may have been from wet to dry in one place, and from dry to 

 very dry in another. 



A complete view of a period of climatic changes involves a study not 

 only of the types of change and of their influence on active agents, such 

 as glaciers, rivers, and lakes, but also of the results produced on land 

 forms and on life. The results of glacial action in tlie form of moraines, 

 till plains, cirques, and so forth, are well Imown. So, too, are the results 

 of the action of lakes which give rise to strands marked by beaches and 

 bluffs, and to deposits varying according to the conditions of deposition. 

 The results of the action of rivers under the influence of changes of cli- 

 mate have received less attention, although it appears probable that they 

 can be detected in a large part of the arid and semi-arid regions of the 

 earth. They take the form of terraces, and of subaerial deposits, the 

 texture and structure of which vary in response to the climatic conditions 

 under which they are laid down. On the deposits I shall not here dwell, 

 because their investigation has not as yet proceeded far. The reasons for 

 believing that many terraces are due to changes of climate have been set 

 forth in "Explorations in Turkestan" and elsewhere, but may here be 

 briefly reviewed. 



Climatic Theory of fluvial Terraces 



Throughout western and central Asia a large proportion of the valleys 

 among mountains of fairly mature topography and of arid or semi-arid 

 climate contains terraces. Where best developed, the terraces number five, 

 not counting an incipient sixth. The oldest and outermost are of large 

 size, often several hundred feet in height. They are composed of gravel 

 and silt, often more or less consolidated. Except where accidentally 

 renewed by the modern undercutting of streams, they are characterized 

 by gentle slopes and much battering and erosion, showing that they are 

 comparatively old. The other terraces are successively younger ; they are 

 composed of less consolidated material, but, nevertheless, are not so much 

 worn. Over vast areas the number, the relative size, the relative amount 

 of weathering and erosion, and the structure and form of the terraces are 

 the same. This is true not only in valleys lying close to one another, but 

 in those of remote regions one or two thousand miles apart. It appears 

 impossible to explain the terraces as due to warping of the earth's crust, 

 because of their uniformity over wide areas and because they occur in val- 

 leys which interlock in such a way that a movement which caused ter- 



