154 A. Imajiuka 



light of the seismic effect on buildings in the city including the burnt 

 areas, but also in the light of seismometric comparisons of the after- 

 shocks at 32 stations in the city. This involved very troublesome 

 work. Meanwhile I may mention only a few interesting results : 

 namely, first, the intensity distribution differs according to the nature 

 of the shock, especially according to the vibration period of the earth- 

 quake waves ; secondly, soil of alluvial formation, especially made 

 ground, is shaken fnlly three times as intensely as ground of diluvial 

 formation such as that at Hongo ; and thirdly, this difference of in- 

 tensity is due not only to differences in amplitude of the earth- 

 movement but also to differences in vibration periods, which are 

 generally shorter in the alluvial soil of the lower town than in the 

 harder diluvial ground (PI. XVIII, XXXIV— XXXVIII). 



12. Conflagrations''^ As has already been stated, the loss by 

 fire was most serious. The flourishing cities of Tokyo, Yokohama and 

 Yokosuka were thus destroyed. As to the statistics concerning this 

 topic, readers are referred to the table of casualties given above. In 

 the following lines, statistics for other towns which were devastated 

 by great fires are given, the numlier in each bracket showing that of 

 the burnt houses in the corresponding town. 



In Sagami Province ; Odawara (5684), Uraga (115), 



Kamakura (562), Atugi (717), 



Hatano (928), Manaduru (882). 



In Awa Province : Hunagata (405), Tateyama (74), 



Hodyo (33). 



I am now going on to give an account of the conflagration in 

 Tokyo. As shown in Table II, the houses comx>letely burnt numbered 

 as many as 366,262 in the city alone, covering an area of 35,850,000 

 square metres. According to the investigations of the fire brigade, 

 the total number of fire-centres which caused the conflagration amounted 

 to 212, more than 40 places where the flames were successfuly ex- 

 tinguished almost as soon as they broke out being excluded from 

 the above number. Some of these fires broke out at varying intervals, 

 but 136 of them are said to have originated within 30 minutes after 

 the quake, thus causing great confusion in the minds of the people 



5) For details, refer to lieports {in Japanese) of the Imp. Eartliq. lav. Conim., 

 No. 100, E. 



