148 R. D. Oldham — Tidal Periodicity in Earthquakes of Assam. [No. 3, 



Here we again see that the day shocks are proportionately more 

 numerous when the sun has declination of more than 9°N. than when 

 the sun is more than 9°S. of the equator, and that in the latter case the 

 night shocks are proportionately more numerous than in the former. 

 It is also evident, from the irregularity of distribution from hour to 

 hour, that the number of shocks is not enough to give a near approach to 

 the true curve, when plotted directly, and a process of smoothing has 

 to be adopted. This has been done by adding together the number of 

 shocks recorded during each group of three successive hoars and, by re- 

 garding them as grouped round the centre of the middle hour, obtain- 

 ing a fresh series of hourly means, from which a great deal of the irregu- 

 larity of the curve has disappeared. The result is represented graphi- 

 cally in Fig. 1, so far as the shocks which occurred when the sun was 

 more than 9° north and south of the equator respectively. 



From this curve it will be seen that as regards the shocks occurring 

 about two hours before midnight there is little difference, but that 

 for the rest of the twenty-four hours the curve for south declination is 

 steadily above that for north declination throughout the twelve hours of 

 the night, and below it for the day. Moreover there is a distinct maxi- 

 mum in the earthquakes recorded round three hours after and two hours 

 before midnight, while the earthquakes recorded near midnight are 

 much more frequent than when the sun was more than 9° north of the 

 equator. Turning to the shocks recorded when the sun was north of 

 the equator, not only are they proportionately more numerous, than 

 when it was south but there is again a distinct pair of maxima, shortly 

 before and three hours after midday. Among the shocks recorded 

 when the sun was within 9° of the equator we have maxima distinctly 

 marked at about 5 hours after midnight and midday, another at about 

 2 hours before midnight and a less marked one at about 2 hours before 

 midday. 



There is consequently an approach to what might be expected if 

 the tide-producing forces caused by the attraction of the sun had their 

 effect in determining the time of origin of earthquakes, but it is also 

 evident that, if these forces have any effect, it is so small and so 

 complicated by other causes, giving rise to a greater variation in 

 frequency than they do, that it is necessary to adopt some method of 

 discussion, which will more or less completely eliminate the effects of 

 variation, other than those due to the tide-producing forces. 



The most obvious of these would be the conversion of the solar 

 into lunar times. The moon moves through the heavens at a rate 

 which brings it on the average about 50 minutes in advance of the sun 

 for each day. If, then, we consider the interval between the two 



