GENERAL CONCLUSIONS. 289 



has not sufficed even to produce new races ; and the 

 inevitable conclusion is that any possible derivation of 

 one species from another is pushed back indefinitely, that 

 the origin of specific types is quite distinct from varietal 

 modification, and that the latter attains to a maximum in 

 a comparatively short time, and then runs on unchanged, 

 except in so far as geological vicissitudes may change the 

 localities of certain varieties. This is precisely the same 

 conclusion at which I have elsewhere arrived from a 

 similar comparison of the fossil floras of the Devonian 

 and Carboniferous periods in America. 



A second leading point to which I would direct atten- 

 tion is the relative value of land ice and water-borne ice 

 as causes of geological change in the Pleistocene. On 

 this subject I have constantly maintained that moderate 

 view which was that of Sir Koderick Mnrchison and Sir 

 Charles Lyell, that the Pleistocene subsidence and refrig- 

 eration produced a state of our continents in which the 

 lower levels, and at certain periods even the tops of the 

 higher hills, were submerged, under water filled every 

 season with heavy field-ice formed on the surface of the 

 sea, as at present in Smith's Sound, and also with abundant 

 ice-bergs derived from glaciers descending from unsub- 

 merged mountain districts. These conclusions have been 

 reinforced by the recent establishment of the fact of 

 differential elevation and submergence, whereby the moun- 

 tain ridges retained their elevation even when plains and 

 table-lands were submerged. I need not reiterate the 

 arguments for these conclusions, but may content myself 

 with a reference to the changes of opinion on the subject. 

 The glacier theory of Agassiz and others may be said to 

 have grown till, like the imaginary glaciers themselves, it 

 overspread the earth. All northern Europe and America 



20 



