12 GLACIAL AND SUEFACE GEOLOGY OF THE PACIFIC COAST. 



All floating masses of ice haviiiQ- fraorments of rock frozen into them are 



O ij CD 



able to striate and smooth the projecting surfaces of the rocks with which 

 they may be brought in contact. The work which they thus clo is closely 

 analogous with that of the true glacier ; but a comparison of all the con- 

 ditions of any locality where such striations occur can hardly foil — in most 

 cases at least — to settle the question of the nature of the agent involved. 



The occurrence of work done by icebergs of course necessitates the former 

 presence of the ocean, since the glacier must descend to the sea-level before 

 it can give birth to an icebero-. That larsje bodies of fresh water mio-ht, 

 however, to a certain extent play the part of the ocean as floaters of ice 

 masses, can hardly be denied. 



The abundance of lakes in certain regions which have been formerly 

 covered with glaciers is looked upon by some geologists as a proof of the 

 powerfid erosive action of ice ; and there are those who consider that all or 

 nearly all lakes owe their origin to this cause. This seems to the writer an 

 entire misconception of the nature of the work done by glacial agencies, and 

 in support of tliis opinion the following considerations are offered. In the 

 first place, however, it may be mentioned that there is one class of lakes in 

 resjrard to whose glacial ori2:iu there can ]je no mistake. These are the so- 

 called " moraine lakes," or collections of water formed behind a terminal 

 moraine, left on the retreat of the glacier, and barring the valley more or 

 less perfectly. The tendency of the streams issuing from such lakes is to 

 wear away the opposing detrital mass, and thus to allow the superfluous 

 water to escape ; hence moraine lakes usually disappear in time, after the 

 glacier has retreated sufficiently fiir, or disappeared altogether. 



The existence of lakes depends, in most cases, on a somewhat complex set 

 of causes. The principal conditions influencing their formation are such as 

 are cormected with the orography and the climatology of the region where 

 they occur; but the nature of the rocks also plays an important part in their 

 formation. If tlie orographic conditions are favorable, and the rainfall is not 

 overbalanced by the evaporation, lakes will occur in greater or less abun- 

 dance, and of greater and smaller size, in strict accordance with the character 

 of the topography of the locality. Central Africa is an excellent illustration 

 of the rationale of the formation of large lakes in a region of excessive rain- 

 fall, favored by suitable orographic conditions. The number of those great 

 bodies of water is not yet known, nor have those which are known been 

 much explored ; but it is clear that within an area occupying a length of 



