ENDLICn ON EROSION IN COLORADO. 833 



to the foregoing, but on a small scale. Water entering fissures and seams, 

 or saturating porous rocks, severely affects them by expansion incident 

 upon freezing. Within certain classes of rocks, this process, prepara- 

 tory to the final removal of material, is one of great importance. IS^ot 

 only are those that may be regarded as mechanical deposits thereby 

 affected, but also the crystalline aggregates. Water entering minute 

 openings between the cleavage-planes of crystals will gradually pro- 

 duce a separation so great as to render the original position of the 

 crystal no longer tenable. This mode of separation is analogous, in its 

 results, to the effects produced by growing roots. For tlowiug and 

 precipitated water is reserved the ultimate transportation of such 

 loosened material from its original place of occurrence. 



Wind, finally, is the last of the important agents of erosion. By its 

 force, small, loosened particles are removed and are carried away. Sand 

 carried before the wind is capable of producing very marked results. 

 By the frequent repetition and violence of the concussions caused by 

 grains of sand striking against some fixed obstacle, a type of erosion is 

 ])roduced that may be regarded as unique in its detail characteristics. 

 While the cutting action of the sand detaches fragments of the rock, 

 the wind rapidly carries them off', thus ever offering fresh surfaces to the 

 attacks of the rapidly abrading material. The comparatively small 

 amount of work that is apparently accomplished by this powerful factor 

 of erosive agents may be due to the fact that peculiar positions of the 

 eroded material are required. Unless these conditions be complied with, 

 the sand will speed harmlessly upon its way, or produce such results as 

 furnish no adequate examples of its j)ower. 



Eeviewing, briefly, the characteristics resulting from the various 

 methods of erosion, we observe that certain analogous physical causes 

 produce essentially the same forms. Water acts as a solvent agent upon 

 many of the minerals constituting rocks. Althougli the quantity of min- 

 eral matter taken into solution by pure water is, as a rule, indefinitely 

 small, the jDresence of carbonic acid gas makes a great difference in its 

 solvent power. Frequently exposures of limestones may be seen, exhibit- 

 ing a minutely corrugated surface. Gypsum is affected in the same way 

 by chemical aqueous erosion. Admixtures of silex and clay in either 

 limestones or gypsum produce definite results, which lead to a recogni- 

 tion of their presence. Although the chemical erosion caused by grow- 

 ing vegetation in the aggregate will show extensive results, its direct 

 evidence is not very manifest. Owing to the distribution of minute 

 root-fibers, their chemical action is spread so uniformly that it can be 

 recognized as such only in rare instances. 



Perhaps the most universally observed products of erosion are those 

 shaped by flowing water. Channels are worn into yielding rocks, rough 

 places are smoothed, soft inclosures in hard rocks are removed, and, 

 throughout, the outlines are modified. These results are, in a great 

 measure, dependent upon the quantity and quality of the material which 



