VolcrtDO Oshima, Llzu. 103^ 



in the magma, Bowen'^ made a thorough consideration, based on 

 the results of recent experiments with artificial silicate mixtures 

 which are comparable to the natural magma. 



A 



It was seen in the rocks of Oshima that the observed facts 

 correspond, at least in a general sense, to what might be expected 

 from Bowen's theory. 



Before a detailed petrologic discussion is given, the writer will 

 here briefly outline the conclusion to which he has been led. 



^ A 



The original magma from which the rocks of Oshima were 

 derived was basaltic ; from it the crystals of olivine, hypersthene 

 (partly clino-hypersthene), augite, and plagioclase separated in the 

 course of cooling before being extruded as lavas. Of these minerals, 

 the former three, being much heavier tlian the magma, sank in 

 the magma basin as they crystallized out ; while the last mentioned, 

 matching the magma closely in density, remained practically 

 suspended in it. Thus, the heavy crystals relatively concentrated 

 toward the bottom. The observed rock- types represent various 

 parts of the magma basin differentiated as above, or different stages 

 of the subsidence of the heavy crystals, but represent only one 

 stao:e of the coohng of the mao:ma. 



Probable High Fluidity of the Magma of Oshima. 



The facility of the movement of crystals in the magma must 

 be greatly influenced by the viscosity of the latter. 



The data at hand suggest a probable high fluidity of the magma 



A 



of Oshima. 



In the first place, it is known of various melts that their 



1) "TJie Later Stages of the Evolution of the Igntoiis Kocks," Jour. Geol. Vol. XXIII., 

 Supplement to No. 8, 1915. 



