ON SEISMOLOGICAL INVESTIGATIONS. 



101 



earthquake occurs in countries sufficiently civilised to allow of an accurate 

 determination by direct observation, more frequently we have one or two 

 instrumental records at distances of a hundred to a thousand miles from 

 the origin, and the results obtained from the study of iadividual earth- 

 quakes have so far been too discordant to enable these to be used with 

 certainty. Recently Professor Imamura has published a memoir ' On the 

 Transit Velocity of the Earthquake Motion originating at a Near Dis- 

 tance,' ^ which should be of assistance, but, owing to the mode of dis- 

 cussion adopted, his results are not directly applicable for the purpose in 

 hand. As, however, all the data are published in detail, I ha»ve been able 

 to plot them on squared paper and obtain two time-curves from which the 

 intervals tabulated below were measured. 



Table shoioing the intervals taken by earthquake loaves to travel from their origin 

 to distances nj) to 10° (llll'l kilometres) deduced from tlie records of ticenty- 

 four Japanese earthquakes. 



Two groups of times were dealt with, P, and P3 of Japanese seismo- 

 logists. The first of these represents the commencement of the preliminary 

 tremors, and two columns are devoted to it in the table : the first repre- 

 sents the measurements of a curve drawn so as to average the records at 

 each distance most closely. It is convex upwards, indicating a slower rate 

 of propagation as the epicentre is neared, but the irregularity of the 

 diflferences of time for equal difierences of distance suggests that it 

 departs from the true time-curve. Another curve was accordingly drawn, 

 representing the average of the observations almost as well as the first, 

 while the more regular decrease in the difierences, as seen in the second 

 column, indicates that it is a closer approximation to the true curve than the 

 other. Both columns give unduly high rates of transmission at the longer 

 distances, the curves being uncertain, beyond 1°, from paucity of records. 



' Publications of the Earthquake Investigation Committee in Foreig'" Languaqes, 

 No. 18, Tokyo, 1904. 



