VolciTio Oshiraa, Idzn. 2T 



If the former were correct wo should naturally expect the 

 presence of explosion products above the uppermost layer of lava, 

 exposed on the riiig-wall ench'cling the caldera, bub on exarahia- 

 tion no such products are to be found. 



Upper nuxrffin 

 of the ririo-wfilt 



! y72o\^ 

 — ^— ^o '''"'"^:;;7r7;s^^ Southwestern , Knjiamihfita 



630^. O^— ^ ff^P 660 ^ ^^„ ., , 



Sur-facB of ^^^^^-~-^^i>''^'~5Sif~~^^ ^oo^yiio^-^SSS. i?9.__ Nortfteir stern 



the caldera ~~^ ^^ X^ri -en' ^^"^ •'"^' 



4SÖ 



Fig. IS. Elevation o£ the crest line around the ring- wall. 



On tlie other hand, the latter explanation seems to be pref- 

 erable as it is highly probable that a hollow space would bo 

 formed below consequent on the withdrawal of magma from below 

 the area, and that this hollow would induce the depression of the^ 

 apical part of the volcano. 



Moreover it would seem that the depression must have taken 

 place with its cantre somewhat to the west of the very top of the 

 mountain, which supposition agrees with the observed features that 



(1) tlio r'ng-wall is higher on the east than on the west, aiid that. 



(2) the distance from tlie margin of the ring-wall to the sea shore 

 is longer on the east than on the west, the main volcano being . 

 considered as symmetrical. 



There is a great gap in the ring- wall of the somma ou its 

 northeastern part, and the north side of this part is hemmed in 

 by an arc-shaped wall. 



Sato and Fukuchi^^ consider this great gap of the somma wall 

 to be the result of an explosion, which formed, at the same time, 

 that part of the wall which extends from due north of the central 

 cone, Mihara,'^ northeastwards.' They based this supposition c a 



1) '■ Geological Notes on Ushima. l(^7A-^." Jour. Geo'jr. Tokyo, Vol. XIV., ]91'2 (in Japanese). 



2) ~W. 



