314 REGINALD A. DALY 



orthoclase) seems often to form an essential constituent. This 

 type is best classed as a trachydolerite of basaltic affinities. Its 

 chemical analysis is almost identical with that of a common phase 

 of the andesitic basalt, but, for some reason, alkaline feldspar 

 did not crystallize from the latter magma. This arrangement of 

 rock-species in Mauna Kea is explained as due to gravitative 

 differentiation in the normal basaltic magma. 



More pronounced splitting is registered in the highly alkaline 

 trachydolerites and phonolitic trachyte occurring in the Kohala 

 district and at Puu Anahulu and the neighboring Puu Waawaa. 

 The development of these extreme types is tentatively ascribed to 

 changes in the normal basalt by its solution of small quantities of 

 sedimentary limestone cut by the respective lava conduits. No 

 direct evidence for this hypothesis has been found; it is based 

 on facts derived from the field and chemical relations of alkaline 

 rocks throughout the world. Whether this hypothesis is correct 

 or not, there can be little doubt that the alkaline rocks of Hawaii, 

 Savaii, and other islands are as truly connected in a genetic way 

 with the normal basalt, as ordinary aplite dikes are genetically 

 connected with their respective granite batholiths. Such an origin 

 for the Hawaiian alkaline rocks is rendered all the more probable 

 because of their small relative bulk, and because of the perfect 

 chemical transition which can now b*e shown between the normal 

 basalt and the most salic of the alkaline types. 



Further, a study of the olivine-pyroxene-anorthite segregations 

 in the andesitic basalt and in the trachydolerite of Mauna Kea 

 actually illustrates a stage of the differentiation. These nodules 

 formed in the magma when its viscosity must have been enormous ; 

 else they would have rushed down into the conduit to levels where 

 no summit eruption, of the small size represented in the flows and 

 cones at the top of Mauna Kea, could have brought the nodules up 

 again. It is almost certain that the settling-out of the olivine 

 and pyroxene material (in solid or liquid phases) must have taken 

 place at slightly higher temperatures, when the viscosity was 

 much less. The residual magma must obviously become more 

 alkaline in proportion to the degree of settling-out. 



The ultra-femic material, sinking to great depth, where it 



