ON THE EARTHQUAKE PHENOMENA OF JAPAN. 



245 



deflected, is set swinging by an automatic arrangement in my honse. Its 

 first swing relieves the catches and sets the plates in motion. By means 

 of the time ticks it is easy to compare the occurrence of any special 

 vibration taken at the various stations within one hundredth part of a 

 second. At the corner of my triangle, at Station Number I., the ground 

 is moderately hard. Station IT. is situated on a small promontory leading 

 out into a marsh and near a shallow pond. Station III., where the 

 ground is moderately hard, is behind a heavy brick building whicb stands 

 very near to the almost perpendicular face of a deep moat. The results 

 which have hitherto been obtained, are briefly as follows : — 



1. The diagram extending over the longest period of time and showing 

 the largest waves is always obtained from Station II. in the vicinity of 



'the marshy ground,- -the diagrams at the other two stations being much 

 smaller. The smallest record is invariably that at Station III. near the 

 deep moat. 



2. At Statious I. and II. waves which may be the same can oc- 

 casionally be identified, but the identification of a wave at III., which is 

 common to I. and II., is not only rare, but it is accompanied by great un- 

 certainty. 



3. In a given earthquake we find that the frequency of waves at the 

 different stations in given intervals of time is different. For example, the 

 number of complete east and west vibrations during the first twenty 

 seconds of time at the different stations during five earthquakes was as 

 follows : — 



Frequency of Waves. 

 Number of Waves in twenty seconds 



From the above table it is evident that the average period must be 

 different at different stations. The small number of waves observed at 

 Station III. in March 31 and April 6 is probably due to the smallness in 

 amplitude of many waves, which, because the period of the earthquakes 

 was long, have coalesced in the diagram to form a straight line. Speaking 

 generally we may say that the average period is longest at Station II. 

 near the marsh. 



At any given station, however, the period varies considerably during 

 the same disturbance. Thus, in March 31, the period of the north and 

 south motion near the commencement of the disturbance was '26 second. 

 A few seconds later it was -4 second. 



A similar result is obtained by the analysis of the diagram taken at 

 fetation II. Selecting the largest waves from the diagrams of the different 



