SUPPLEMENT. 119 



ponded back. No cases of this kind are described, to my knowledge, but they 

 will very likely be found ; or at least Ave may expect tliem to appear when the 

 northern branches of the Indus cut their way backwards through the inner- 

 most range of the Himalaya, and gain possession of the drainage of the 

 plateaus beyond ; for then, as the high-level waters find a steep outlet to a 

 low-level discharge, they Avill carve out canons the like of which even Dutton 

 has not seen, and the heavy wash of waste will shut in lakes in lateral ravines 

 at many points along the lower valleys. 



In its old age, a river settles down to a quiet, easy, steady-going existence. 

 It has overcome the difficulties of its youth, it has corrected the defects that 

 arose from a period of too rapid growth, it has adjusted the contentions along 

 the boundary-lines of its several members, and has established peaceful rela- 

 tions with its neighbors : its lakes disappear, and it flows along channels that 

 meet no ascending slope on their way to the sea. 



Certain accidents to which rivers are subject are responsible for many 

 lakes. Accidents of the hot kind, as they may be called for elementary dis- 

 tinction, are seen in lava-flows, which build great dams across valleys : the 

 marshes around the edge of the Snake river lava-sheets seem to be lakes of 

 this sort, verging on extinction : crater lakes are associated with other forms 

 of eruption. Accidents of the cold kind are the glacial invasions: we are 

 perhaps disposed to overrate the general importance of these in the long his- 

 tory of the world, because the last one was so recent, and has left its numerous 

 traces so near the centers of our civilization ; but the temporary importance of 

 the last glacial accident in explaining our home geography and our human 

 history can hardly be exaggerated. During the presence of the ice, especially 

 during its retreat, short-lived lakes were common about its margin. We owe 

 many prairies to such lakes. The rivers running from the ice-front, overloaded 

 Avith sand and silt, filled up their valleys and ponded back their non-glacial 

 side-streams ; their shore-lines have been briefly described in Ohio and Wis- 

 consin, but the lakes themselves Avere drained Avhen their flood-plain barriers 

 Avere terraced ; they form an extinct species, closely allied to the existing 

 Danube and Red River type. As the ice-sheet melts aAvay, it discloses a sur- 

 face on Avhich the drift has been so irregularly accumulated that the ncAv 

 drainage is everywhere embarrassed, and lakes are for a time very numerous. 

 Moreover, the erosion accom])lished by the ice, especially near the centers of 

 glaciation, must be held responsible for many, though by no means for most, 

 of these lakes. Canada is the American type, and Finland the European, of 

 land-surface in this condition. The drainage is seen to be very immature, but 

 the immaturity is not at all of the kind that cliaracterized tlie first settlement 

 of rivers on these old lands : it is a case, not of rejuvenation, but of regenera- 

 tion ; tlie icy bajttism of tlie lands lias converted tlieir streams to a ncAV spirit 

 of lacustrine liesitation unknown before. We cannot, however, expect the 



