408 F. oMOEi : 



of 172 occurred in Spring, and the minimum of 134 in Sum- 

 mer, the ratio of the two numbers being as 1.3:1. Finally, 

 according to Table VIIT, tbe number of earthquakes in the 

 warmer months of the year was greater in the ratio of 1.1:1 

 than that in the colder months. The results contained in Tables 

 VI and YII are illustrated in figs. 2 and 4. 



8. Comparing figs. 1 and 3 with ligs. 2 and 4, the annual 

 variation of the frequency of ordiiiarij small shocks, whose 

 minimum occurs in the months of June, July, August and 

 September, will be observed to be just the reverse of that of 

 destructive earthquakes, whose maximum occurs in the months 

 of July and August.^ 



To explain this fact, we may remark that the geotechtonic 

 disturbances, which caused the destructive earthquakes given 

 in Table T, were most probably greater tlian those of the 

 majority of the ordinary earthquakes included in Table 

 YI. Xow, as an earthquake implies the removal of a weak 

 point in the earth's crust, and the consequent settling of the 

 latter to a stable equilibrium, the constant occurrence of small 

 earthquakes is, for unstable countries like Japan, to be regard- 

 ed as maintaining the region concerned in a normal or safe 

 condition, thereby preventing any abnormal accumulation 

 of stress in the earth's crust. Their non-occurrence, or an 

 unusually low seismic frequency may, on the other hand, 

 cause an accumulation of stress in the earth's crust, thereby 

 facilitating the occurrence of great or destructive seismic distur- 

 bances. 



1. Dr. C. Davison found from Mallet's Catalogue a similar reversal of the maximum 

 epoch in the annual variation of des'rudive and slifjht earthquakes. See Trans. Euy. Soc. 

 Lond., VuL 184 [IS'-Xô), j). lUG. 



