ORIGIN OF AUGITE AN DBS IT E 417 



eruptions. King's observations warrant the belief that some differ- 

 entiation by fractional crystallization occurs in the lava after it escapes 

 from the calderas of Hawaii. However, the Hawaiian rocks are 

 more allied to those of an Idaho or Oregon lava-flood than to the 

 typical andesites of Ecuador. The poverty of Etna lava in pheno- 

 crysts may be explained as due to the partial differentiation of the 

 primary olivine basalt.' 



5. Relation of augite andesite to other andesites. — The hypothesis has 

 been framed on the supposition that the primary basaltic magma has not 

 been essentially affected by its solution or assimilation of chemically 

 contrasted rock, material with which the lava makes contact as it 

 rises or lies in the volcanic conduit. It is, however, most probable 

 that the primary lava may in many cases absorb a certain amount of 

 foreign material, either rock or fluid, and that the products of fractional 

 crystallization must then vary notably from the few types so far 

 mentioned. The process itself may be aided or retarded by such 

 absorption, according as the absorption affects the temperatures 

 of crystalhzation or the viscosity. Phenocrysts and ground-mass 

 must change in chemical composition. Their separation by gravity 

 would, then, produce magmatic materials not directly corresponding 

 to augite andesite nor to any of the ultra-basic rocks formed directly 

 from olivine basalt. 



Hornblende andesites, mica andesites, dacites, etc., may, on this 

 view, be derivatives of the primary basalt which has been modified 

 by the assimilation of foreign rock-substance; while certain of the 

 peridotites may represent some of the correlative differentiates of the 

 compound lava. In a similar manner, one might possibly consider 

 many lamprophyres and aplites as due to the analogous differentiation 

 of their respective parent-magmas by the settling of basic minerals 

 from the latter in its magmatic condition. It is not here the writer's 

 purpose to discuss these possibilities in detail; they are noted rather 

 to point to his belief that the commonly observed field associations 

 and chemical relationships of augite andesite and other andesites 

 are facts not opposed to the hypothesis outlined. It is, however, 

 by no means intended to express the belief that all andesites or all 



I Cf . H. Rosenbusch, Mikros. Physiographic der massigen Gesteine, 3d ed. (Leipzig, 

 1896), p. ion. 



