36 president's address. 



freshwater lakes, and hence not available for return to the air, 

 the form in which it is stored is of no consequence so far as 

 glacial changes are concerned. It is in the consumption of 

 carbon dioxide in the weathering of rocks that the great primary 

 source of permanent loss lies. The rocks constituting the exposed 

 surface are largely composed of silicates and by the action of 

 carbon dioxide and moisture these are decomposed, the bases 

 combining with the carbon dioxide and ultimately finding their 

 way into the ocean. Practically all the carbon dioxide so fixed 

 may be considered as permanently lost. 



We have seen that depletion of atmospheric carbon dioxide 

 induces a cold surface condition while enrichment results in the 

 opposite effect. When the land surface is at work removing 

 carbon dioxide and thereby bringing about cooling, and the ocean, 

 responding to the change, aids in the withdrawal, all the con- 

 ditions necessary for the inauguration of a glacial epoch are 

 present, and, accordingly, when the rate of removal of carbon 

 dioxide exceeds that of supply, it is only a matter of time for the 

 change to occur. With the advent of an icy covering the land 

 would be effectually shielded from the action of atmospheric 

 carbon dioxide, and the loss through weathering being stopped, 

 or at any rate greatly reduced, a time of steady accumulation 

 would set in, resulting in the dawn of a genial period; the 

 encroachment of ice would be stopped, the line of glaciation 

 driven back towards the poles, and the rocky surface again 

 exposed. On the completion of one cycle there would be a 

 gradual swing in the opposite direction, and so the continued 

 succession of glacial and warm periods, of which we have evidence, 

 would be accounted for. 



There are so many modifying influences, such as variation in 

 the relationship of water and land, which would tend in one way 

 or another to affect the rate and intensity of climatic change as 

 well as the time of duration of both conditions, that nothing in 

 the way of regular periodicity is to be expected in glacial epochs, 

 and that there was no such periodicity seems to be the trend of 

 the evidence. Partial retreats and advances of ice, and greatly 



