122 ^'ol- XLIII., Art. 6. — S. Tsuboi : 



Anotlior fact to corroborate this inference is that it is closely 

 related to rocks of Kazahaya which are believed to have been 

 extruded from the lower portion of the magma basin. 



(5) Augite-micro-diorite. — This is an ejected block, about 3 cm. 

 across, found in an agglomerate bed at the northern foot of 

 Kagamihata^^ (Ic). It is a holocrystalline mass consisting of zonal 

 structured plagioclase ranging from Ab^gAugs to Abj^An^s, uncommon 

 augite which is greenish and slightly pleochroic, and acidic inter- 

 stitial matter with micrographie fabric, with magnetite and apatite 

 as accessory constituents (pp. 90-101). 



This is a mere block showing no relation to any lavas, but 

 its pétrographie characters are just what might be expected to 

 form at the upper portion of the magma reservoir if the cooling 

 of the magma proceeded at a moderate rate. 



As already stated (pp. 117-118) the magma of Oshima is 

 believed to have reached the stage at which it was in the middle of 

 the crystallization of plagioclase and augite, and the temperature 

 was that at which bytownite crystals Ab^sAug^ were in equilibrium 

 with the residual liquid. This state of the magma has been main- 

 tained throughout the whole volcanic history so far as it is now 

 traceable. 



If the cooling of the magma proceeds further from this state 

 at a very slow rate the results would be as follows :— 



Plagioclase and augite, which were crystallizing till the above- 

 mentioned state of the magma was reached, would continue to 

 crystallize, and magnetite would soon Itegin to crystallize. The 

 plagioclase and augite crystals in separating would become richer 

 in the lower-melting components as time goes on. Even the crystals 

 already separated, if there is a sufficient amount of the available 



1) m^ 



