1002.] R. D, Oldham — Tidal Periodicity in Earthquakes of Assam, 151 



The fact is that in both these cases the total number of shocks 

 considered is too small to get an approach to a true average, and, in this 

 small number of shocks, accidental variations of distribution may pro- 

 duce an irregularity of the curve which exceeds its normal variation. 



To some extent this difficulty may be overcome. If we refer to the 

 tables I and II, we will see that when the sun is within 9° of the 

 equator, there is not a very great variation in the times of passage of 

 the tidal circles as compared with the times of passage during the day 

 when the sun is north, and during the night when it is south, of the 

 equator. On the other hand the night when the sun is north, and the 

 day when the sun is south, of the equator, show a much greater range of 

 time in the passage of the circles and not only is the range of time 

 greater and the effect consequently less conspicuous, but during part of 

 the time the maximum of horizontal force is not felt at all, and during 

 the rest of the time the passage is so oblique that the rate of change 

 is slow and the tidal forces probably less effective. 



Excluding these shocks we may add together the two groups of 

 shocks already considered and so obtain a larger one, in which the tidal 

 effect is tolerably uniform. The result is given in the table, and shown 

 graphically in Fig. 2. Here it will be seen that the two maxima preced- 

 ing and following the meridian passage are both distinct, and exceed 

 those obtained from the total number of shocks. 



Shillong Seismograph 1897-1901. Semidiurnal curve of frequency, 



ni i ii in mi v vi vii vin DC x xt m 



fSun>9 N. Day- 

 All Shoclo X* - - -*{Sun 9'N. 9'S. All 

 (Sun>9'S.Ni£ht 



( fSua>ST&Nigil 

 lSun>3'S.Da; 



Fig. 2. Semidiurnal curves of freqnenoy. 



