HISTORICAL NOTICES. 9 



the winds, depending on the earth's temperature and 

 rotation. The courses of the currents are also constant, 

 except in so far as modified by coasts and banks ; and the 

 direction of the drift-scratches and transport of boulders 

 in the Pleistocene both of Europe and America, show that 

 the arctic currents at least have remained unchanged. 

 But the distribution of land and water is a variable 

 element, since we know that in the period in question 

 nearly all northern Europe, Asia and America were at 

 one time or another under the waters of the sea, and it is 

 consequently to this cause that we must mainly look for 

 the changes which have occurred. 



" Such changes of level must, as has been long since 

 shown by Sir Charles Lyell, modify and change climate. 

 Every diminution of the land in arctic America must 

 tend to render its climate less severe. Every diminution 

 of land in the temperate regions must tend to reduce 

 the mean temperature. Every diminution of land any- 

 where must tend to diminisli the extremes of heat 

 and cold ; and the condition of the southern hemisphere 

 at present shows tliat the submergence of the great 

 continental masses would lower the mean temperature, 

 but render the climate much less extreme. Glaciers 

 might then exist in latitudes where now the summer heat 

 would suffice to melt them, as Darwin has shown that in 

 South America glaciers extend to tlie sea level in latitude 

 46° 50' ; and at the same time the ice would melt more 

 slowly and be drifted farther to the southward. [In the 

 southern hemisphere, indeed, a glacial period of a peculiar 

 kind exists at present, since there is an ice-bound 

 antarctic continent 2,000 miles in diameter and boulder- 

 drift extending from it half-way to the equator.] Any 

 change that tended to divert the arctic currents from our 



