IS^OTE.S OX JAPANE>SE EAllïHQUAKE.S. 405 



According to Table III, the maximum monthly number 

 of 32 occurred in August and the minimum of 10 in January, 

 the ratio of these two numbers being as o.2 : 1. 



According to Table IV, the maximum seasonal number of 

 74 occurred in Summer, and the minimum of 4Ö in Winter, the 

 ratio of these two numbers being as l.G:l. The numbers for 

 Spring, Autumn and AVintei', do not, however, differ much 

 among themselves. 



Further, according to Table V, the earthquake number in 

 the warmer months of the vear was «rreater in the ratio of l.o :1 

 than that in the colder months. 



The results contained in Tables III and IV" are illustrated 

 respectively in iigs. 1 and 3, the curves being drawn, not always 

 through representation points themselves, but by free-hand 

 through the positions of the arithmetical means of two succes- 

 sive points. The same method of curve-drawing has been em- 

 ployed in all the other similar figures. 



7. For the sake of comparing the annual variation of the 

 frequency of destructive earthquakes with that of ordinary 

 small ones, or those which arc not the after-shocks of great 

 earthquakes, I give in Table VI the monthly numbers of earth- 

 quakes observed in whole of Japan^ during ten years between 

 1885 and 18U4. 



The enormous rise in earthquake numbers after November 

 181)1 is due to the occurrence of the Mino-Owari earthquake 

 on the 28th October, 181*1, of the Kagoshima earthquake on the 

 7th September, 18^3, of the Hokkaido earthquake on the 20th 

 March, 1894, and several other strong earthquakes, each of 

 which was followed by a great number of after-shocks. Hence, 



1) Foriuusa and Liu-kiu Isltuids excepted. 



