3 o8 



REGINALD A. DALY 



of Puu Anahulu and the "andesite" of Waimea are consanguine- 

 ous and that transitional types connect them with the distinctly 

 subalkaline, normal basalt of the island. The steady decrease 

 of the iron oxides, magnesia, and lime, and the corresponding 

 increase in silica, alumina, and alkalies are as systematic as could 

 be expected in a syngenetic series derived by the process of differ- 

 entiation already in part described. 



The norms calculated for the analyzed types tell the same story. 

 They are summarized in the form here given : 



Calculation shows that the alkalic members of the series were 

 probably not formed by a mere subtraction of ferromagnesian and 

 cafemic, phenocrystic material from the normal basaltic magma. 

 On the other hand, a positive addition of the alkalies seems to have 

 occurred when the more silicious types were developed. 



The concentration of the alkalies in the upper part of an ini- 

 tially basaltic vent may conceivably be due to two principal causes. 

 In the first place the feldspathic or feldspathoid matter of the 

 basalt might individualize in liquid phases or as plastic or rigid 

 crystals, and, because of the low density of any of these phases, 

 rise in the magma column, just as the individualized olivine or 

 augite (in liquid or solid phases) must sink. Or, secondly, the 

 volcanic vent may become temporarily enriched in emanating 

 gases, which, as they rise, bring the alkalies with them in loose 

 combination. 



Of these two possible causes the partial control by rising volatile 

 substances seems to be specially clear in intrusive bodies. The 

 writer has suggested that most of the alkaline rock types have 

 been derived from subalkaline magma through the solution of 

 limestone or other carbonate-bearing sediments. 1 Thereby the 

 subalkaline magma is not only fluxed and so prepared for drastic 



1 Bulletin Geological Society of America, XXI (1910), 87-118. 



