MAGMATIC DIFFERENTIATION IN HAWAII 315 



mixes with the hot, normal basalt, is subject to extrusion through 

 lateral fissures which bring the deeper levels of the conduit into 

 communication with the surface. Such is the preferred explana- 

 tion for the ultra-femic olivine-basalt which emanated from Mauna 

 Loa in 1852, and for the wehrlitic porphyry composing the Uweka- 

 huna laccolith. 



An explanation is offered for the apparently contradictory fact 

 that gravitative differentiation is little evident in the thoroughly 

 basaltic summit rocks of Mauna Loa and Hulalai, or in the material 

 of the active vent at Kilauea. 



As a result of studies in this and other fields, and in the general 

 literature of petrology, the writer is inclined to the belief that all 

 late pre-Cambrian and younger " alkaline" rocks are the result 

 of differentiation within primary basaltic magma or within syn- 

 tectic magmas formed by the solution of solid, generally sedi- 

 mentary, rock in the primary basalt. The marvelously uniform 

 composition of the basaltic magma issuing from countless fissures 

 in every ocean basin as in every continental plateau, seems capable 

 of explanation only on the premise that it forms the material of a 

 continuous, world-circling substratum. The facts of geology 

 suggest that this substratum was formed by an ancient liquation 

 which took place when the globe was molten at its surface. This 

 general conception became gradually clear to the writer during 

 the genetic study of many intrusive bodies; it had been visualized 

 in much the same form by that extraordinarily suggestive observer 

 of the Hawaiian volcanoes, W. L. Green, whose Vestiges of the 

 Molten Globe, Part II, first became known to the present writer 

 in 1909. Not a single one of the myriad facts recorded in general 

 petrography and geology definitely opposes this hypothesis, which, 

 to the writer, seems to be the best working premise for a general 

 philosophy of the igneous rocks. 



Lastly, it appears from the accumulating results of geological 

 work that the division of igneous rocks into "Atlantic" and 

 " Pacific" races or groups is not warranted by the facts of distri- 

 bution, nor by the requirements of sound petrogenic theory, nor 

 by the needs of systematic petrography. In the heart of the 

 Pacific basin, as in many regions along its borders, rocks of foyaitic, 



