THE GREAT EARTHQUAKE OF S. E. JAPAN' 163 



in flood nbove the original sea-level. 



Some of the inhabitants of Oosima observed on the western coast 

 two tunami propagated soon after the first shock, one from tlie north 

 and the other from the sonth, which met one another in front. This 

 can readily be understood by reference to the above-stated ma]) which 

 shows elevation in the north and depression in the south across the 

 strait between Idu and Oosima. 



19. The cause of the earthquake and the seismic zone. So far 

 I have described certain seismic phenomena which will probably help 

 us in determining the real cause of the earthquake in question. The 

 enormous depression and elevation of the sea-bottom in the heart of 

 Sagami Bay indicates that this area was the seat of an extraordinary 

 commotion which took place thereabout at a certain depth. Possibly 

 a fault line, as I have stated before, ma}" have been produced along 

 or across the boundary of the areas of depression and elevation, 

 containing the seismic focus in it. 



From these phenomena, it may be inferred that the commotion 

 was caused by stress which had been applied for many years in such 

 a way as to push tlie lower depressed side of the bay towards the 

 north or the northwest, and the upper elevated side towards the 

 opposite direction. Later on, I will describe the slow, gradual eleva- 

 tion of the land in the Miura Peninsula which had been continuing for 

 three years until at last it reached a critical condition on Sept. 1, 

 1923, and there suddenly occiu-red the remarkable upheaval already 

 mentioned. Indeed, uplieaval of the adjacent coast seems to be a 

 phenomenon common to every great destructive earthquake which has 

 originated on or off the coast of the Boso Peninsula. 



The focal areas of the three destructive earthquakes which took 

 idace on Dec. 31, 1703, Sept. 1. 1923, and Sept. 2, 1923, form one 

 continuous belt in themselves, and this belt is very likely a part of 

 the great seismic zone running along the Pacific coast of Japan. It 

 may not be out of place to remark here that the first two earthquakes 

 were in many respects similar to each other, especially in their 

 topographical changes, isoseismals and the tidal waves generated by 

 them, the only difference being that the centre of the recent earth- 

 quake was further to the we^-t than that of the earlier one. 



20. Secondary causes and antecedent phenomena. The secondary 

 causes of the last shock together with antecedent phenomena may 



