68 Elastic Constants of Rochs and Velocity of Seismic Waoes. 



geneous substance, as Dr. Wiechert assumes it to be, it is 

 quite probable that the velocity of the longitudinal wave is 

 13 kilometres, and that of the transversal wave is about 

 7 kilometres, as found by Prof. Omori. These velocities are 

 nearly twice as great as those observed in the oldest rocks 

 here examined. 



Elastic waves travel with slow velocity in surface rocks. 

 If the principal shocks in the seismometer-record be taken 

 into account, the velocity turns out to be very small and 

 about 3*3 kilometres. This evidently is about the mean 

 velocity of propagation in most of the surface rocks, and 

 shows that waves of large amplitude creep along the surface. 

 It is not wonderful that with distant earthquakes, the duration 

 sometimes extends over several hours, as the disturbance 

 travels through strata of different elastic constants, and the 

 waves modified in various ways will appear all blended to- 

 gether on the seismograph. Although 3 kilometres may be 

 a mean velocity, there are certain surface rocks in which the 

 velocity is less than a kilometre. The shock at the epicentre 

 may last only for a short time, but the duration at a distance 

 will be lengthened, as the range of velocity is very wide. The 

 disturbance coming from the strata of greatest rate of propaga- 

 tion will first make its appearance at the beginning of the 

 preliminary tremor, followed by waves travelling with slower 

 velocity till the principal shock arrives as surface waves. It 

 will be followed by waves travelling with still slower velocity 

 leaving faint record on the seismograph, till they at length 

 fade away. Neglecting the time of passage from the stratum 

 above mentioned to the surface, it is natural to expect that 

 the duration of the so-called preliminary tremor preceding 

 the earthquake shock increases linearly with the distance of the 

 epicentre from the place of observation. The above relation 

 was established from various earthquakes which happened in 

 Japan, recorded by Prof. Omori. 



With great earthquakes which are perceptible on a seismo- 

 graph at very great distances, the duration will continually 

 increase with distance ; the disturbance may sometimes pro- 

 pagate still unabated in one or other direction round the earth. 

 If the last-mentioned case actually takes place, the tremor will 

 probably last even for days. As such records have sometimes 

 been obtained by seismologists, it may not be out of place 

 here to notice the possibility for such undulatory movement 

 of the ground. 



In conclusion, I wish to express my thanks to Prof. Koto 

 and Mr. Fukuchi for valuable information concerning the 

 geological and petrological character of rocks examined in 

 the present experiment. 



