Dr. C. Davison on Earthquake- Sounds. 51 



on the nature of the rock beneath which it originated, the 

 percentage of audibility should not vary perceptibly at dif- 

 ferent times ; and this is very far from being the case. 

 Taking the district 7 as a whole, the percentage in November 

 1891 (immediately after the Mino-Owari earthquake) was 

 17*9, and during the next five months it lay between 10*5 

 and 12*1 ; then it suddenly rose to 38'8 in May 1892, and 

 during the following seven months never fell below 32"4, 

 while it maintained an average of 41*8. In the smaller 

 rectangles we find the same variety ; in one (that numbered 

 1457). for instance, the percentage was 7*7 during the three 

 months November 1891- January 1892, and 39*0 during the 

 next eleven months; in another (1402) the percentages during 

 the same intervals were 10'3 and 55*4 respectively. 



Thus whatever the nature of the relation between the 

 audibility of Japanese earthquakes and the geological structure 

 may be, it seems to be often masked by other conditions of 

 greater importance, one of which will now be considered. 



Equal- Audibility Curves in the Mino-Owari District 

 (Japan) . 



The district represented in the accompanying map (fig. 2) 

 is that part of Japan in which the Mino-Owari earthquake of 

 1891 and the majority of the very numerous after-shocks 

 were most prominently felt. It is bounded by the parallels 

 34° 40' and 35° 50 / of north latitude and by the meridians 

 2° 10' and 3° 30' west of Tokio. During the eight years 

 1885-1892 the total number of earthquakes felt in the district 

 was 3014, and of these 604, or 20 per cent., were accom- 

 panied by sound. The percentage, however, varies very 

 widely in different parts of the area in question, and the con- 

 tinuous curves in fig. 2 represent this variation. Dividing 

 the whole area into 56 rectangles, the sides of which are one-, 

 sixth of a degree of latitude and longitude in length, the 

 percentage of audible earthquakes originating within each 

 rectangle is supposed to correspond to the centre of that 

 rectangle, and the curves are drawn in the same way as 

 isacoustic lines. The meaning of any curve, say that marked 

 40, is as follows : — If any point on the curve be regarded as 

 the centre of a small area, then 40 per cent, of the earth- 

 quakes whose centres are included within this area were 

 accompanied by sound*. 



* In constructing- these curves the percentage was regarded as zero in 

 the rectangles where there are no epicentres, or in isolated rectangles 

 where the number of epicentres is very small. 



E 2 



