THE GKEAT ERUPTION OF SAKUKA-JIMA IN 1914. 25 



holociystalline rock was seen, probably making dykes through the 

 basal vitroandesite (trachyandcsite) C. It contains abundantly 

 raedio-phenocrysts (4 mm.) of hypersthene and sporadic niino- 

 phenocrysts of olivine, which we frequently fail to find in many 

 slides. The rock closely resembles the trachybasalt A^ of 

 Fig. 3. The same rock makes up the flat- top of Sakkabira at the 

 defunct Seto strait, and recurs at Karasu-jima. PI. XVI. Fig. 4. 

 Besides the abundancy of phenocrystic hypersthene, which is rare 

 in the writer's experience, the hypersthene is intergrown with 

 peripheral and lamellar (100) augite. 



The light- colored trachyandesite C and the dark trachybasalt 

 -li occur in close association, although the latter is decidedly of 

 later origin. Both being widely distributed, it is an important 

 pétrographie problem in southern Kyushii to make their relation clear. 



In analogy with the neighboring plateau-land, the rock which 

 builds up the islet is of early Diluvial or late Tertiary age. 

 ^oSm " '^) Hakama-goshi. (PI. III. Fig. 1 ; PI. XII. Fig. 3)— 



The low rectangular table-hill of Hakama-goshi or Shiro-yama 

 in Geologic Map is of the size, and the same geologic formation 

 and structure as the preceding. It was originally an island rising 

 fi'om the bottom 20 fathoms deep, although it now forms a western 

 liortion of Sakura-Jlma and a landing place from the city by the 

 accumulation of talus and outwash from that mountain-slope, which 

 filled up the space between this islet and the former Sakura-jima. 

 Small crafts are said to have plied here even in historic times. 

 It was occupied by the village of Koike, now entirely de- 

 vastated and partially overflooded l:»y the deluge of the recent 

 lava-flow. 



The lava-flow rushed down the western slope of Sakura-jima 

 from the Yuno-hira vent to the sea, floodino; over the villaç^es 



