THE GIŒAT EKUPTION OF SAKUEA-JIMA IX 1914. 15 



inul liornbleude ciystals. TJie fossils found by the ^vriter ;ire Dcntalinm 

 sp. Hud il nin\alv(\ This oürrespouds to the fossil hori/on at Kaikata 

 iininediately to be reforred to. The two beds A and B lie conformably 

 with <>TndnMl transitions. 



C. — The bottom exposed here is of a gre}', soft and soaking 

 mass of liai-isld or hypersthene-trachj'andesite, already briefly outlined 

 (p. 13). This is a sul)mariue lava-flow wliicli gave birth to the lapilli. 

 PI. XYI. Fig. 1. 



Following the railway from Kagoshima northward along the shore, 

 the effiisive C becomes sphenüitic with pjT-'oxenes at Jigami Station, i.e., 

 the r tyi)e (p. 13). PI. XYI. Fig. 2. At Shigetomi the lapilli plateau, 

 500 m. high, is capped with trachybasalt A^. PL XVI. Fig. 4. Farther 

 northward, at the foot of Kirishima, the cap-rock is ji hornblende-bearing 

 pyi'oxene-andesite A-^, which apparently attains a thickness of more than 

 200 III. {see Fig. 3). 



The extensive lied of l)lock-mud capped with liornblende-p}TOxeue- 

 andesite in northern Kj^ûshû, especially in the Yabakei region in Prov. 

 Bungo, seems to the writer to be the equivalent of (jur lapilli bed ^i and 

 the cap-rock ^L. The Yalmkei region is well known for its canyon 

 scener3^ The lapilli bed or its equivalent is, therefore, regional, and by 

 no means confined to a corner of southern Kyfishû. 



The stretch, 100 to 500 ni. high, between Kagoshima and Shigetomi 

 just referred to, and all the pla.teau-lands (400 m.) on the southern foot 

 of Kirishima di^scend precipitoush', a/., :d Kokubu Statiou,^^ to the 

 shore-flat, or directly to tlie water's (;dge (PL lY. Fig. 1) — a special 



1) After having finished the manuscript, the writer came across the passages in -which 

 the lite Y. Pdchthofen has not only given an exact written profile of hornblende-bearing 

 l^umiceous complex capped with trachytie rock, accompanied with pétrographie details at the 

 north of Kokubu, but also the canyon-like erosion valleys in the jilateau land, which characterize 

 the scenery of this region. (' Geomorphologische Studien', III., S. 16.) The writer was greatly 

 surprised at the author's keen observation many decades ago, \\hen that part of Japan was still 

 entirely unknown ; but his comparison of the Bay of Kagoshima and the Sakura-jima volcano with 

 the Laguna de Taal in Luzon seems, however, not a fortunate one. The writer considers the 

 bay as a rift valley and not a caldron depression ; but it nmst be remembered that a partial 

 nature of Icottle depres.sion is not yet fundamentally confuted. 



