272 



Y. OINOUYE 



on July 28, roaring again began, and two explosions took place in 

 sight of the writer at 11:20 a.m. (Fig. 10&). The loud roaring 

 continued till eight o'clock in the evening. On the same day there 

 was heavy rain all day accompanied by loud thunder, intense 

 lightning, loud roaring in the ground, and much dense smoke 

 hiding the mountain entirely from view. Before 9:00 a.m. on 

 July 29, judging from the numbers of smoke vents found on that 

 morning, several more explosion craters were formed (Fig. 10c). 



Fig. 16. — A group of craterlets on the second group. Taken September 4, 1910 



On August 2 two new craters were formed at the top of Nishi- 

 maruyama, a parasitic cone. For a week these two craters poured 

 forth an astonishing amount of smoke in a large black column such 

 as was not seen from any other cone (Fig. 11). The volume 

 suddenly decreased at the end of one week. Thus in the period 

 of greatest activity, from July 25 till August 2, the number of 

 craters formed amounted to 15. Thereafter, on August 7 and 8, 

 on September 3, and on October 2, small explosions occurred, so 

 that the total number of craters reached 45. Besides these craters 

 numerous crevices and faults developed on the side of the moun- 



