556 



CENOZOIO TIME — MAMMALIAN AGE. 



(The underlying drift, beneath the alluvium, is not shown.) It 

 appears thus that each terrace was once part of a flood-plain of the 

 river. 



In the above explanation the terraces are supposed to correspond 

 each to a separate period of elevation. This may be the case ; and, 

 when so, the same terrace would be traceable for great distances 

 along the course of the larger rivers. But successive terraces may 

 be formed in river-valleys, either (1) during a slowly-progressing 

 elevation, or (2) in the course of the wear which may be in pro- 

 gress between periods of elevation ; and it is often difficult to dis- 

 tinguish these accidental or intermediate plains from those that 

 are distinct records of change of level. One such intermediate ter- 



Fig. 832. 



Section of a valley with its terraces completed. 



race is shown at r in fig. 832. Some of the conditions producing 

 them are the following: — (1) changes in the river-channel to one 

 side or the other of the river-valley, altering thereby the action of 

 the flood-waters during freshets, and causing them to commence 

 wear according to a new outline ; (2) resistance to wear in a por- 

 tion of the alluvium, owing to a degree of consolidation, or to some 

 obstacle ; (3) a permanent diminution in the waters of a stream, 

 arising from changes about its sources, or in some other way. 



It is important to observe also that the same terrace may dif- 

 fer in height ten to fifteen feet or more; because (1) the flood- 

 plains of rivers (the original condition of the terrace-plains) often 

 differ much in height in different parts ; (2) the rains and stream- 

 lets often wear away the soft material of the terraces, diminishing 

 their height, and sometimes obliterating the plain altogether; (3) 

 the winds carry off the light soil of the surface, and in the course of 

 centuries may produce great results. 



Again, the terraces of small tributaries at a distance from the 

 river into which they flow, are lower than those of the latter, be- 

 cause both their floods and their eroding power are less. 



Again, when there are rocks in the course of a stream, a terrace 

 above the rocky barrier may differ in height from its counterpart 



