MAGMATIC DIFFERENTIATION IN HAWAII 309 



differentiation, but the carbon dioxide introduced from the sedi- 

 ment carries the alkalies with it as the gas rises through the magma. 

 An instructive series of experiments by Giorgis and Gallo bear 

 on this suggestion. They mixed the powders of various recent 

 Vesuvian lavas with water, and passed a current of carbon dioxide 

 through each mixture for a period of two months. Analyses 

 showed that the powders lost, on the average, 37 per cent of the 

 soda originally contained, the remaining constituents being but 

 little altered in amount. 1 At high temperatures the upward 

 transfer of the alkali would presumably be much more rapid. 



In the case of the Puu Anahulu trachyte or in that of the 

 Waimea "andesite," the principal factor in the differentiation 

 may have been the solution of coral or foraminiferal limestones, 

 such as are known to be interbedded with the older lavas of the 

 archipelago. That the normal magma of the archipelago has 

 been locally affected by such solution is suggested by the occur- 

 rence of melilite and nephelite in the basalts of Maui and Oahu, 

 the melilite indicating an excess of lime and the nephelite showing 

 such desilication of the salic part of the magma as is expected 

 when it dissolves foreign lime. The carbon dioxide entering the 

 primary basaltic magma because of this solution of sedimentary 

 rock would belong in the "resurgent" class of emanations. A 

 special concentration of juvenile carbon dioxide in a basaltic vent 

 would, in an analogous way, tend to concentrate the alkalies of 

 the basalt at the top of the vent. 



This hypothesis, that the decidedly alkaline rocks of Hawaii 

 have been derived from the normal, subalkaline basalt through 

 gravitative differentiation in the volcanic conduits, is supported 

 by the intimate field-association of the two classes of rocks, and 

 by the fact that the alkaline bodies are all of very small volume 

 as compared with the known mass of normal basalt in Hawaii. 

 The first-mentioned relation is obvious; the second is already 

 clear, even though the island has been covered only by recon- 

 naissance journeys. It is practically certain that the trachyte 

 of Puu Anahulu and vicinity, the most salic type and one very 

 conspicuous in the field, can be exposed at but very few and small 



1 G. Giorgis and G. Gallo, Gazetta (1906) [I], 137. 



