72 SCIENCE PROGRESS 



elements shows that if gabbro, as we know it, existed through 

 and below a depth of 30 kilometres, the total radioactivity- 

 would be so great that the earth should be growing hotter. 

 We are therefore restricted to a shell of granite not more than 

 three kilometres in thickness, on an average. If, after the 

 early Pre-Cambrian intrusions, there was less granite than this 

 amount in existence, then some of the basaltic magma below 

 might still be able to produce more. Otherwise refusion by 

 stoping would be the only mechanism whereby future batho- 

 liths could be formed. 



These alternatives may be tested by an appeal to the 

 field associations of alkali rocks. Phonolites and basalts are 

 commonly found together, but according to Bowen 's scheme 

 rhyolites ought to fall between them in the course of genesis. 

 Yet they are rarely found in alkaline volcanic fields, unless 

 they themselves are strongly alkaline. Why is it that the 

 volcanic islands of the Central Pacific within the andesite 

 border persistently fail to provide rhyolitic lavas? The 

 reason is probably that the basalts are incapable of supplying 

 the necessary silica. If 90 per cent, of the composition of 

 average gabbro be subtracted from 100 percent, of that of 

 average basalt, the 10 per cent, difference gives a great excess 

 of alumina, alkalies, and iron oxides over that required by 

 granite, but a marked deficiency in silica. It would seem, 

 then, that the salic differentiates of basalt ought to be markedly 

 alkaline. The granite or rhyolite stage should, in general, be 

 passed over. This is fully in accordance with the facts, and 

 suggests an additional reason why alkali rocks seem to 

 have become more abundant in the later periods of geological 

 history. During early periods, basalt may have had poten- 

 tialities for the differentiation of granite. Later, having lost 

 those potentialities, it has attained a greater chance of gener- 

 ating alkali rocks than formerly. It is very likely that 

 Bowen has provided the raison d'etre of the earth's original 

 granitic shell, but not necessarily that of later granites. 



The second difficulty supplements the first. While not 

 accepting magmatic stoping as a chief factor in the mechanism 

 of batholithic intrusion, Bowen fails to suggest an alternative 

 process. Where are the rocks that the batholiths have dis- 

 placed ? The oldest known sediments, even in their meta- 

 morphosed condition, imply a granitic origin. The oldest 



