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



earthquake frequency, and in a laborious paper, 1 on the annual and semi- 

 annual periodicity of earthquakes, came to the conclusion that, treating 

 each region separately, there was a distinct variation in frequency, which 

 was in excess of that which might be expected if the occurrence of 

 earthquakes was in no way connected with the seasons. 



From this brief review it will be seen that the question, of whether 

 earthquakes are at all affected by extra-terrestrial influences, is at present 

 an open one, and for this reason I made every effort, after the great 

 earthquake of 1897, to obtain the fullest possible record of the extreme- 

 ly numerous after-shocks, thinking that if there was any external cause 

 at work it should be especially easy to trace at a time when, and in a 

 region where, the earth's crust was evidently in an extremely unstable 

 condition. The discussion of these records is not complete but iu the 

 case of one of them it has been completed, so far as one particular phase 

 of the frequency is concerned, and the results obtained appear to be of 

 sufficient interest to justify some notice of them. 



In July of 1897, Mr. T. D. LaTouche, who was then in Shillong 

 reporting on the results of the earthquake, constructed a seismograph 

 on the duplex pendulum system, which was set up by the Executive 

 Engineer, and from which continuous records have been taken ever 

 since. The instrument, like all seismographs, is far from a perfect one, 

 it does not record many shocks which can be distinctly felt, and it does 

 not record the time, yet the records are of great value. In the first place 

 we know that every shock recorded attained a certain standard of range 

 of motion of the wave particle and of violence, if such a word may be 

 applied to what in many cases are merely slight shocks, and that all the 

 shocks exceeding this standard are recorded. The absence of automatic 

 time record is more serious, but as the time of the shock was, in every 

 case, recorded by the observer we may take it that there is no very serious 

 error or omission in this respect. Every shock recorded represents one 

 at approximately the time given, and the only cause likely to affect 

 the periodicity is a possible error in the case of the night shocks: it 

 is possible that the instrument may at times have registered a shock 

 while the observer was asleep, and the record afterwards referred to 

 one, felt when he was awake, which did not affect the instrument. The 

 uncertainty due to this cause is, however, slight, as the gentlest shock 

 registered by the instrument is sufficiently strong to usually awake 

 a sleeper. 



From this instrument we have received records from August 1897, 

 but those discussed as yet only extend to the end of 1901 5 so far 

 they have only been examined with a view to the hourly variation in 



1 Phil. Trans, clxxxiv, A, 1107 (1893). 



