148 F. OMom : 



It has already been pointed out by Professor J. Milne and 

 others that a distant earthquake, as observed by means of 

 delicate horizontal or vertical pendulums, begins with tremors, 

 whose duration increases with the distance of the origin of 

 disturbance from the observing station. What is here discussed 

 is based on the observations of comparatively near earthquakes 

 made in Tokyo and at the Meteorological Observatories of Kago- 

 shima, Fukuoka, Gifu and Miyako, the origins of disturbance 

 being in Japan itself or in the Ocean off its eastern coast. 



I may here note that all the earthquake diagrams, from 

 which I have measured the durations of the preliminary tremor, 

 were given, by Ewing's or by Gray and Milne's Seismographs, 

 unless otherw^ise stated. Each of these instruments, ordinarily 

 at rest, is started on the occasion of an earthquake by means 

 of an electric contact-maker, which is usually made so delicate 

 as to be sensible to a movement of about 2^0 nim. However, 

 different seismographs at one station, owing to different sensi- 

 bilities of the contact-makers of the instruments, do not always 

 give one and the same duration for the preliminary tremor of 

 an earthquake, the discrepancy amounting usually to a few 

 seconds. In these cases the longest value of the duration must 

 be regarded as nearest to the truth and is that which has 

 been adopted. 



Again, I liave excluded all the diagrams of small earth- 

 quakes, whose origins w-ere very far off from the observing 

 station, because in these cases, the beginnings of the motion, 

 being slow and extremely small, are very apt to fail to affect 

 the electric contact-maker, and, also because there often exists 

 no well-marked transition fiom the preliminary tremor to the 

 principal portion of such earthquakes. 



