102 ME. E. D. OLDHAM ON A SEASONAL YAEIATIOK [vol. lxxiv„ 



II. DlURNAL DISTRIBUTION OF EARTHQUAKES (ItALV, 1891-1910). 



Sours. 



yumber of Shocks. 



Ratio to Mean. 











0- 2 



Year. 



Summer. Winter. 



Year. 



Summer. Winter. 



1001 



500 



501 



132 



129 T37 



2- 4 



1094 



556 



538 



1-45 



1-43 1-47 



4- 6 



903 



448 



455 



1-20 



1-15 1-24 



6- 8 



735 



386 



349 



•97 



•99 -95 



8-10 



582 



316 



266 



•76 



•81 73 



10-12 



623 



329 



294 



•83 



•85 -80 



12-14 



603 



335 



268 



•80 



•86 -73 



14-16 



666 



351 



315 



•88 



•90 -86 



16-18 



619 



340 



279 



•82 



•87 -76 



18-20 



627 



312 



315 



•83 



•80 -86 



20-22 



755 



361 394 



1-00 



•93 1-08 



22- 



858 



438 420 



1-14 



1-12 1-15 



Totals ... 



9066 



4672 4394 



12-00 



12-00 12-00 













III. Diurnal Distribution" 



op Earthquakes 



^Japan, 1885-1892;. 



Hours. 



Number of Shocks. 



Ratio to Mean. 



Year. Summer. 



Winter. 



Yea r. 



Summer. Wimter. 



0-3 



3- 6 



6- 9 



9-12 



12-15 



15-18 



18-21 



21- 



1086 378 



1016 365 



975 396 



926 355 



1049 425 



922 346 



1079 353 



1265 441 



708 

 651 

 579 

 571 



624 



- 576 



726 



824 



1-04 

 •98 

 •94 

 •89 



1-01 

 •89 



1-04 



1-21 



•99 1'08 

 •95 -99 



1-04 1 -88 

 •93 '87 



1-11 -95 

 •91 '88 

 •92 1-10 



115 1-25 



Totals ... 



8318 3059 



5259 



8-00 



8-00 | 8-00 



I 





1 



tabulation of the Shillong seismograph, though the number of 

 shocks dealt with in this case is too small to give it much weight. 

 For the present this must remain a mere suggestion ; all that can 

 be asserted is that, whatever may be the actual ratio of day to 

 night shocks in any district and any period, it will be rather 

 greater in summer, and rather less in winter, than the average of" 

 the whole year and period. 



In this paper no attempt has been made to apply the method 

 of harmonic analysis because the applicability to earthquake fre- 

 quency seems doubtful, for reasons which cannot be discussed in 

 detail. As, however, something of the kind seems expected, and as 

 the Italian record, by reason of its completeness and of the fact 

 that the period covered is nearly coincident with the lunar cycle 

 of 19 years, is less unsuited to this method of treatment than the 



