136 



F. ÖWORI. 



Tlie mean curve (drawn in red), Fig. 2o, seem« to indicate the 

 existence of a period' whose length is some 12 years. 



One interestino- fact is that the three severest of all tlie earth- 

 quakes recorded at Tokyo during tlie eighteen years, namely, those of 

 February 2nd, 18<S0, of October 15th, 18S4, and of February 18th, 

 18S1), happened respectively in the years which mark, exactly or 

 nearlv, the epochs of successive maximum frequency. Ft may likewise 

 Ix' the case that the maxima of the longer (12 years) period may often 

 be marked l)y great earthcpiakes. 



It may herebv 1)2 remarked that the greatest monthly earthquake 

 number, namely 45, was recorded in October, LSDl. Of these, 2S 

 happened within four days immediately after the great Mino-Owari 

 earth(piake (which took place on the 28tli of the same month), and 

 tlie remaining 17 before the latter, so that the ])roper number of 

 earthcpiakes in Tokyo may be assumed to have been 17x-|f, or 19.- 

 Of the other 225 monthly numbers the next greatest namely 18, 

 occurred in April, LSS!). 



A good exam])l ' of the «')ccurrence of destructive earthquakes at 

 rather regular intervals is found in the seismic fre([uency for Kansu, a 

 Department in the north-western corner of China, where the intervals 

 between successive paroxysmal epochs range from 15 to (U years, 

 giving the mean length of -îo years.'' 



To discover periodicity in seismic frequency Ave must treat records 

 of earthquakes for différent localities separately. Ijy way of example, 

 in Fig. 2o, (J), (i>), (C) and (7)), are shewn the curves of yearly seismic 

 activity during 11) years, 18(55 to 1883, in Swdtzerland, the Vesuvian 



1 If the sensibilities of the seismofçraphs )jy which the record of shocks has been made have 

 been different in the earlier and later years, the chief consequence will have been to affect 

 slightly the amplitude. 



2 At Giiu only one earthquake had been recorded in October, liefere the 28th. 



3 See the present author's paper, •' On Chinese Eartliquakes," Seis. Jour. Vol. T. 



