part 2] VARIATION' IX THE FBEQUENCY OF EARTHQUAKES. 99 



'S. A Seasonal Variation in the Fbequency of Earthquakes. 

 ByRlCHABB Dixon Oldham, F.K.S., F.G.S. (Head Feb- 

 ruary 6th, 1918.) 



[Plate XII.] 



When investigating the aftershocks of the great Indian earth- 

 quake of 1917, I found that the ratio of shocks occurring during 

 the day half of the twenty-four hours, between 6 a.m. and 6. P.M., 

 to those occurring during the night half, was somewhat greater 

 than the general average during the summer months, when the 

 sun is in north declination, and somewhat less during the winter 

 months, when the sun is south of the equator. As this relation 

 was found not only in the whole record, but also in the natural 

 groups of observations, it seemed probable that the variation was 

 a real one, and not fortuitous ; while the fact that the record 

 of the Shillong seismograph, when tabulated by lunar times and 

 seasons, gave a similar variation suggested that the cause was not 

 climatic, but in some way connected with the tidal stresses set up 

 by the attraction of the sun and the moon. 1 



At a later date I made a tabulation (which has not been pub- 

 lished until now) of Milne's catalogue of Japanese earthquakes, 2 

 and a^ain found a similar relation between the ratio of day to 

 night shocks in the two halves of the year ; and a similar tabulation 

 of the records from a number of observatories, 3 mostly independent 

 of and covering a later period than Milne's catalogue, again gave 

 the same result. 



Recently there has appeared an abstract tabulation of twenty 

 vears' records of observations, carried on bv the Italian seismologrical 

 service, which enables the subject to be more completely studied 

 than was possible with the more limited and imperfect records 

 previously available. 4 In this, once more, the relatively greater 

 frequency of earthquakes during the day in summer and the night 

 in winter is shown by a tabulation of the two halves of the year, 

 and the completeness of the record enables a further test to be 

 applied, based on the consideration that, if the observed variation 

 is not accidental but due to some cause, whether climatic or 

 physically dependent on the variation in the position of the sun. it 

 should be greater at the height of each season than the average of 

 the whole of each half-year. The months June and July were, 

 therefore, taken as representing midsummer conditions, and the 

 comparison of day and night frequency showed a divergence from 

 the average of the year in the same direction as that of the season 

 as a whole, but greater in amount. Similarly, the months of 



1 Mem. Geol. Surv. Ind. vol. xxxv, pt. 2 (1903) pp. 117-45. 



2 Seismol. Journ. Japan, vol. iv (1895). 



3 Publ. Earthq. Invest. Committee, No. 8, 1902. 



4 Boll. Soc. Seismol. Ital. vol. xx (1916) pp. 9-31. 



Q. J. G. S. No. 294. i 



