and North American continents were so changed that no great 

 rivers were to flow into the Arctic Ocean, because the ice from 

 the Siberian and North American rivers greatly lowers the 

 summer temperature there. 



It is worth remarking, that throughout all geological time 

 the northern hemisphere must have been comparatively conti- 

 nental with a continental climate, and the southern hemisphere 

 oceanic with an oceanic climate. The arrangement of land 

 and water on the earth's surface is such, that a hemisphere of 

 which England is the centre contains a greater proportion of 

 land than any other hemisphere that can be drawn, and the 

 opposite hemisphere a greater proportion of water. This is 

 obviously due to a slight excentricity of the earth's centre of 

 gravity as compared with her centre of figure, drawing the 

 water of the ocean to the farther side from us ; and no 

 geological revolution can have materially altered this, which 

 must have been fixed at the original condensation of the 

 earth. 



To return to the subject of the former greater heat of the 

 sun ; — we have seen that a much colder sun would not produce 

 glaciation ; but on the other hand it is evident that a very hot 

 sun would prevent glaciation, by melting the snow either before 

 it fell or soon after. Nordenskiold states that there is a total 

 absence of any evidence of glaciation having occurred in the 

 Arctic regions during the Tertiary period, although such evi- 

 dence, consisting chiefly of ice-borne boulders, could not be 

 destroyed except by such denudation as would have removed 

 the strata altogether. It appears safe to attribute this to the 

 greater heat of the sun during the Tertiary period, which is 

 also attested, as we have seen, by the fossil vegetation. 



There is however some evidence of glaciation in Eocene and 

 Miocene times, not regularly recurrent, but only occasional as 

 to place and time. Those cases which appear to be most clearly 

 made out are in Switzerland and Piedmont, near the Alps ;* 

 and there is no difficulty in seeing how vast glaciers might 

 proceed from the Alps at a time when none were formed on 



* Croll's Climate and Time, pp. 305, 306. 



