no SCIENCE PROGRESS 



separated from it by a silent zone which may be as much as 

 fifty miles in width. 



Prof. Omori has shown that the sounds of the Sakura-jima 

 explosions were transmitted in the same abnormal manner. 

 The detonations were of course too numerous and occurred too 

 closely together to be traced over the country individually. 

 But there was no difficulty in considering the propagation of the 

 three groups of sounds. Though these groups differ greatly in 

 strength, the detonations of the second group being by far the 

 loudest, it is remarkable that the areas over which they were 

 heard are practically alike in form and do not vary greatly 

 in their dimensions. Each group of sounds, for instance, was 

 heard within two detached areas. One of the areas includes 

 Sakura-jima near its western margin, and in each case extends 

 a considerable distance to the east. Though the prevailing 

 winds in the upper atmosphere are from the west, the second 

 sound area lies to the north of the former but is elongated in 

 the east-west direction. 



The area containing Sakura-jima varies but slightly in mag- 

 nitude, the average distance of the boundary from the centre 

 of the volcano being 69, 71, and 62 miles for the three groups of 

 sounds. For the detached area, the average distance of the 

 inner boundary is 89, 94, and 108 miles, and that of the outer 

 boundary for the second and third groups 134 and 140 miles. 

 The average distances from the volcano of the central line of 

 the detached area are 121, no, and 122 miles. 



As might be expected, the width of the silent zone varies 

 with the strength of the sounds It is estimated at 25 to 30 

 miles for the strong detonations and about 67 miles for the 

 weak after-explosions. In both, however, the average distance 

 from the volcano of the central line of the silent zone is about 

 75 miles. 1 



The extreme distance to which any detonation was heard is 

 310 miles towards the north-east, though the air-vibrations 

 were strong enough to shake houses and doors to a distance of 

 364 miles in the same direction. 



1 Prof. Omori does not give any explanation of the anomalous forms of the 

 Sakura-jima sound-areas. The theory of the detached areas is considered by 

 Mr. S. Fujiwhara in a memoir, " On the Abnormal Propagation of Sound-waves 

 in the Atmosphere," second part, Bulletin of the Central Meteorological Observatory 

 of Japan, vol. ii. 1916, No. 4, pp. 1-82. A summary of this important memoir is 

 given in Nature, vol. xcviii. (1916), pp. 71-2. 



