230 ^- OMORI. 



occurred when the frequency was going to reach a maximum, as with 

 the earthquakes of January 10th, 1894, and of September 7th, 181)2 

 (see Figs. 5 and 26). In the case of the latter earthquake, a maxi- 

 mum frequency took place indeed at Gifii, which was quite near the 

 origin, in the same month (September); but, at Mitake, a town about 

 7 ri from the origin, greater numbers of shocks were recorded in the 

 following two months. The three other maxima of frequency for 

 Gifu which took place on April, 1891, and April and September, 1898, 

 were accompanied by no particularly severe shocks. Similarly with 

 the maxima in the monthly earthquake frequency for Kumanioto. 



In the case of the Mino-Owari earthquake, some of the after- 

 shocks are doubtless to be reiifarded as " fore-shocks " of the stron"-er 

 ones which followed. 



§ 24. The Diurnal Fluctuation of tlic Aftcr-slioch Freqiiencij. 



The diurnal and annual fluctuations of the earthquake frequency 

 have been discussed by various European investigators. The results 

 obtained by Perry, Mallet, and others, however, are more or less 

 doubtful, as they were chiefly based on stiitistics of vulgar records. 

 In the present instance, after-shocks have 1jeen recorded at meteoro- 

 logical stations provided with seismographs, and the results here 

 deduced should have therefore fu- ij-reater weiii'ht than those hitherto 

 obtained. 



The curves of six-hourly earthquake frequencies for Gifu and 

 Nagoya (Figs. 13 and II) indicate the daily fluctuation very clear]3^ 



The Mino-Owari earthquake took place on the morning of 

 October 28th, 1891, and during the next thirteen complete days, from 

 October 29th to November 10th, there were 1258 shocks at Gifu arid 

 572 shocks at J^agoya. The distribution of tliese shocks in the 

 twenty -four hours of the day are shewn in Fig. 19, (1) and (2) (see 

 Tables XIII and XIV). 



