188 F. OMORI AND K. HIRATA : 



Propagation along the coast is, however, evidently much favoured 

 by the geological formations, which generally run parallel to 

 the coast. On the other hand, it is also very probable that 

 the epifoci of most of the earthquakes under consideration were 

 not points, but were linear and parallel to the axis of the 

 Island. Thus, the forms of the isoseismal lines in figs. 4 and 

 5, suggest very naturally that the earthquakes, Nos. 20 and 8 

 were probably caused by long fissures or faults in the direction 

 of nearly NEN— SWS. 



37. The position of the earthquake origin. 



Of the 25 earthquakes above considered, six originated at 

 distances of 100 to 130 km. to the W or WSA¥ of Miyako, 

 while the remaining nineteen originated under the sea at dis- 

 tances between 40 and 210 km. to the ENE, E, or ESE of the 

 same station. The positions of the nineteen suboceanic origins 

 are included within a circle of about 2° radius, the centre 

 being in long. 143° E, lat. 39° N. This point, which is at about 

 1° or 110 km. to the E of Miyako, is thus to be regarded as 

 the most active centre of earthquakes disturbing the eastern 

 part of Northern Japan. 



As the majority of earthquakes disturbing Äliyako, or more 

 generally the eastern coast of Northern Japan, originate under 

 the Pacific, it is to be expected that the shocks will sometimes, 

 like No. 1 and No. 12, be accompanied by sea-waves. It may 

 be that the disturbance of the sea occurs more or less whenever 

 a sufficiently strong shock takes place, either at the sea-bottom 

 itself or at a very small depth below it, so as to give commo- 

 tion to the water. Of course, the same phenomena may be 

 produced, when a comparatively sudden elevation or depression 

 of the sea-bottom takes place. 



