328 R. A. DALY PETROGRAPHY OP THE PACIFIC ISLANDS 



advanced gravitative differentiation of common olivine basalt, the most 

 abundant of Pacific lavas. The grounds for this belief will not here be 

 repeated. The chemistry and mineralogy of basalt, pyroxene andesite, 

 dunite, wehrlite, lherzolite, and the corresponding transitional types sug- 

 gest that the splitting has been, in a sense, spontaneous, and not induced 

 by the incorporation of any foreign material except, perhaps, vadose 

 water in a few cases. The natural condition for such splitting won Id be 

 a comparatively low temperature for the basaltic magma. If, at such a 

 temperature, a certain percentage of the phenocrystic material were to 

 segregate and then settle out, the remaining liquid would have the com- 

 position of molten pyroxene andesite. The mother liquor might freeze 

 in the volcanic vent, giving a dioritic rock, or freeze on the surface after 

 its extrusion as a lava flow, giving the andesitic type. 



The temperature appropriate for gravitative differentiation may be 

 temporarily established in a basaltic vent at any epoch in its history, but 

 probably most often toward the close of the life of a great basaltic volcano, 

 when the feeding magmatic chamber approaches the temperature of final 

 solidification. Even then, however, flows of primitive basalt may often 

 be expected to alternate with flows of andesite, and these with flows of 

 picrite or picritic basalt. Such alternations are common among the 

 superficial, and therefore younger, lava flows of the gigantic Hawaiian 

 volcanoes. 5 In general, the field relations of the Pacific island lavas seem 

 to agree with the hypothesis of gravitative splitting for the andesites, 

 etcetera; yet the hypothesis is emphasized more to indicate the need of 

 special, detailed field-work in the islands and to serve as one of the guides 

 in that research than to suggest any finality for the explanation. 



Alkaline Eocks 



The twenty-six species listed in Class 7 are called "alkaline," after the 

 Eosenbusch tradition. The word is, of course, used metaphorically. 

 Thus in Class 7 are placed several species which are not rich in alkalies, 

 but the world over are regularly, and doubtless genetically, associated with 

 alkali-rich species. 



Alkaline rocks have been found in thirty-five of the islands, repre- 

 senting thirteen major archipelagos and two isolated islands. In thirty 



e In Professional Paper 88 of the U. S. Geological Survey (p. 92), Cross states that 

 the writer believes all of "the upper 6,000 feet of Manna Kea" to be composed of ande- 

 sitic and closely allied, trachydoleritic, rocks. This statement is based on a complete 

 misapprehension. Throughout his published account of the lavas on the upper slopes of 

 Mauna Kea (Journal of Geology, vol. 19, 1911, pp. 297, 311, 313), the writer refers only 

 to the surface rocks on the eastern side of the volcano. It is entirely possible that 

 typical olivine basalt occurs at the surface in "the upper 6,000 feet" of Mauna Kea and 

 very probably is there much more abundant below the surface. 



