HYDROLOGY. 139 



equal hardness of the strata acted upon by a powerful excavating 

 agency, that, unlike' rivers, did not cany away its rubbish as fast as 

 formed, but heaped it up promiscuously in its own track as it melted 

 backward, the inevitable result being the formation of lakes along its 

 abandoned channels. 



The number originally thus formed was vastly greater than the 

 number now in existence, the great majority having cut down their 

 own barriers and drained themselves. 



These may then be said to be glacial lakes, for there is no reason 

 to suppose that any one of those described, existed before the drift 

 period. But as the most of them were probably represented as river 

 valleys or gorges, and as their present existence is due also to the un- 

 equal powers of resistance of the rock formations, this class of lakes 

 is not entitled to be termed glacial lakes in the same sole-dependent 

 sense as the following, the most of which show little or no relation- 

 ship or dependence on the underlying strata^ but are drift lakes in an 

 exclusive sense. 



The class now to be described may be subdivided into two subordi- 

 nate ones. The first are those which lie in valleys, whose greatest length 

 is in a line with the glacial grooves, and in this respect are like the 

 last class, differing from them only in being independent of the strata 

 beneath. These were formed in drift troughs by the retreating 

 glacier, very much as the class above described were, only the troughs 

 are not excavated rock channels, but valleys, between drift ridges. 



The other kind are those whose greatest length lies at right angles 

 instead of parallel to the glacial grooves. 



These lie between the ridges that accumulated at, the foot of the 

 glacier and which were rudely parallel to each other, yet sometimes 

 joining mutually, and again separating to a considerable interval, they 

 left enclosed hollows, which on filling with water, became lakes. 



We would perhaps be justified in being even more specific than we 

 are in calling these drift lakes, by designating them moraine lakes. 



From the nature of the case, forms intermediate between all the 

 foregoing are to be expected, and are found to occur, and with the gen- 

 eral explanations given, it will be needless to designate in the case of 

 each of the numerous small lakes, the precise method of formation, 

 although it is, in most cases, sufficiently clear. 



More than one hundred of these lakes lie along or adjacent to the 

 Kettle Kange, and form an important element of its picturesque 

 scenery. Among the most noteworthy are Geneva, Delavan, Como, 

 Turtle, Bass, Roldens, Otter, Silver, Pleasant, Green and the Troy 

 lakes in Walworth county; Mary, Elizabeth, Camp and Silver lakes 



