266 N. S. SHALER — SPACING OP RIVERS 



level areas being effected mainly b}' solutional processes until the retreat- 

 ing head escarpments of the larger streams work back across the surface 

 of the u])land. Notwithstanding these differences between the small and 

 the great erosive work, the exhibitions of it which are afforded in the min- 

 iature stream systems throw much light upon true river action. 



EFFECT OF RAIN ON SMOOTH, SLOPING SURFACES 



In the first stage of rain work on a moderately smooth and gently 

 sloping surface of bared earth we find the very numerous closely set little 

 valleys before noted, such as may be produced in the course of half an 

 hour of moderate rainfall. If we watch the further steps of the process 

 we observe that as these channels cut deeper the valleys of some of them 

 are widened, so that in an area where there was at first a dozen distinct 

 grooves there is now perhaps but one. A change of this sort may at 

 times be traced in the course of a few hours of continuous rain. It is, 

 however, better seen during a season's changes which deepens valleys 

 from the average depth of half an inch to that of a foot or more, with 

 proportionate reduction in the number per unit of area. 



Inspection of the process of change in the swiftly developing valleys 

 of bared earth at once shows that the increase in the size of the channel 

 is brought about by a process in which certain of the streams cut down 

 and laterally with greater rapidity than the others. This process of de- 

 velopment is in its nature and methods one of selection ; it more clearly 

 resembles the principle of the survival of the fittest than an}'- other known 

 to me in the inorganic realm. The reason for this likeness is to be found 

 in the fact that in the mechanical changes of the streamlet, as in those of 

 the living form, success is determined hy the adjustment of the action 

 to exceedingly varied conditions. In each we have something like a 

 continuity of endeavor with the limitation which circumstances put upon 

 it. In the institution of a very small drainage system, such as we are 

 considering, the first action is determined by the currents which set up 

 in the sheet of water before it begins to cut. The initial channels are in 

 effect a map of these original streams. As soon as scouring begins those 

 guiding features arising from the diversities of the surface and of the initial 

 cutting power of the" streams at once gives some of them the mastery. 

 To perceive the value of these differences we should note that the cutting 

 power of the currents increases in a very high ratio to their speed, cer- 

 tainly as the square thereof, and perhaps in a much higher function. 

 The result is that, given a very slight preponderance in size of one stream 

 among many, this advantage will cause its channel to cut down more 

 ra[)idly than do those of its competitors. .As it extends its drainage 

 slopes from its rapidly lowered base it inevitably captures the valleys 



