l^j^ F. OMOßl. 



the seismic frequency may p:irtly be due to those in the atmospheric 

 pressure. Especially are the curves of the annual l^aro metric and seismic 

 fluctuations very similar to each other. 



///. On the DislribiUion of Aj'ter-slioch, etc. 



§ 31. We shall lastly consider more particularly the magnitude 

 of after-shocks and their distribution. 



Earthquakes are produced when strains in the earth's crust reach 

 a certain limit, and, as a very great shock would remove a correspond- 

 ingly great underground instability, it is probable that such a shock 

 would not, for a long time, be followed Ijy anotlier of a magnitude 

 comparable to its own, in the same or a neighbouring district. When, 

 however, the initial sliock is not very great, it may be followed by 

 another like it, and, even in this case, the position of tlie origin of the 

 second shock would usually Ije quite distinct from that of the first. 



The above statements can well i)e illustrated by the f)ur recent 

 destructive earthquakes in Japan, namely, those of i\Iino-Owari, of 

 Noto, of Kumamoto, and of Kagoshima ; the three last ones were 

 nuich smaller than the tirst, which was indeed very great a.nd viokait. 



§ o2. The Kumamoto earth(piake of July !2<Sth, LSSD, was 

 followed five days later, on August 3rd, by a second shock, which 

 was, however, as we have already seen (§ 7) only one-half or one- 

 third as great as the first. All the other after-shocks were much weaker. 



The Nolo earth<p]ake of December 9th, 181)1, was fjllowed, two 

 days later, on the 11th of the same month, by a second severe earth- 

 quake. These two had ne:irly an equ;d area of disturbance, and the in- 

 tensity of motion near the epicentres was almost the same. Their origins 

 were, however^ different, the epi-centre of the first shock lacing at a 

 point, latitude 37°-!' IST, longitude 130° 40' E, or in the sea at about 



