146 A. Imamura 



by a lateral fovce equal to 1/10 of the weight of the construction 

 itself. This must he regarded as the maximum intensity of the earth- 

 quake motion at the site of Tokyo Imjierial University. During this 

 epoch the vertical conjponent did not exceed 1 cm. 



After that time, the earthquake motion was registered only by 

 the E. W. component. From this register and the seismogram pro- 

 duced by an aftershock which took place at 14h 22m on the same 

 day and originated very near the place of origin of the first destructive 

 shock, (PI. XVI. A) it seems that later the motion became much greater 

 but slower, and possibly there existed an enormously large vibration 

 with a range as great as 20 cm. and a period between 2.5 and 4 sec. 

 Dnring the first two minutes, the motion was most conspicuous, and 

 during the following 8 minutes, less so. It has been remarked by 

 Prof. Sida"^ that the same seismogram shows the existence of a very 

 slow earth movement, with its period as long as 110 seconds and its 

 range as large as 1.2 metres, corresponding possibly t<» the bodily 

 oscillation of the Japan Island itself. 



Before the earthquake motion was yet finished, there took place 

 many after-shocks, some of which attained to tlie category of a semi- 

 destructive nature. Consequently the end portion of the first shock 

 was often masked by after-shocks, and therefore it is impossible to 

 estimate the exact time of the final phase. Prior to the earthquake 

 of 14h 22m, very minute slow undulations were recognizable; these 

 may perhaps correspond to the latter ])art of the end portion, so that 

 I assume 2h 20m as the dia-ation of the total earthquake motion 

 of the first shock. 



6. The origin. From the data given above, I put the origin, 

 that is the place where the earthquake action began, at a point 92 

 km. distant from Hongo, Tokyo, in the direction of S26°W. Combin- 

 ing the initial horizontal dis])lacement ( = 1.5 mm.) with one half of 

 the vertical component ( = 0.25 mm.), we get 10° and 15 km. as 

 rough estimates of the dip and depth of the origin respectively. The 

 co-ordinates of the epicentre turn out as follows : — 



/I = 139° 21'.8 E, ^ = 34° 53'.6 K 



It must, however, be added that there are a few investigators who 



3) See Prof. Matuyania's Notes on the nature of the Kwanto Earthquake: 

 PitH- (ill. Bu7-eau Central Sism. Titt-y Serie A^ No. 2, 



