DEVELOPMENT OE SEISMOLOGY liV JAPAN 27/ 



the susceftihility and conductivity of seismic waves, by means of which 

 he succeeded in accounting for the anomalies in some actual examples. 

 Nakannu-a found from his investigations on the relation between the 

 intensity of an earthquake and the epicentral distance that the geo- 

 graphical distribution of the ' active ' regions, i.e., the regions wliich 

 show an abnormally large intensity compared with the average for 

 the same distance, reveals some regularity or symmetry about the 

 epicentre. This latter result seems worth further studies with respect 

 to more abundant examples, especially under the light of the recent 

 theoretical research of Nakano and Matuzavva above quoted. Naka- 

 mura also assumed a formula for the law of the decay of the intensity 

 with the distance and remarks that the coeflScient of absorption is 

 larger for local earthquakes than for shocks of wide extent. Suda 

 investigated the rate of decrease of the intensity with the epicentral 

 distance, up to 1500 km. with respect to 28 conspicuous, i.e strong or 

 ivecdc, earthquakes. Diagrams showing the relation between the dis- 

 tance and log A-/T^, where A is the maximum amplitude and T the 

 period, were classified into two principal types, i.e. the ordinary type 

 and the extraordinary type, according as the intensity falls off 

 continuously with the distance, or shows a maximum at a certain 

 distance. The relative frequency of the two types were 70^ and 30^ 

 respectively. The former type is attributed to the case when the 

 displacement at the focus is nearly vertical, while the latter, i.e., the 

 extraordinary type is considered to correspond to the case when the 

 original disturbance is directed horizontally or obliquely. He also 

 remarks that the ordinary type is manifested mostly in weak earth- 

 quakes, whereas the extraordinary one is characteristic of the strong 

 or destructive shocks. 



The absurdity of regarding the seismic origin as a mere geo- 

 metrical point has been repeatedly emphasized by most seismologists, 

 especially with respect to cases of strong earthquakes observed near 

 the origins. Hirano"^^^ remarked in the case of the Kwanto Earthquake 

 that it becomes necessary to assume a i-egion with an infinite velocity 

 of propagation, if we apply the ordinary linear formula, giving the 

 epicentral distance as the function of the duration of the preliminary 

 tremor, for coordinating the data of observations obtained at numerous 



(1) J.M.S., [ii] 2 (1924), 112. 



