400 



NATURE 



[September 15, 1923 



Ministry of Education — are destroyed. The most 

 serious loss is that of the lofty steel-brick buildings 

 recently erected. It was supposed that they would 

 resist a shock of the utmost violence, and if their 

 destruction was, as is probable, due to the earthquake 

 and not to the fire, it may be necessary to prohibit 

 their erection in the future, and this will greatly restrict 

 the manufacturing power of the country. Estimates 

 of the total loss of life vary widely. Some place it as 

 high as half a million, and in Tokyo inquests have 

 already been held on more than 32,000 bodies. There 

 can be little doubt that the work of a few minutes has 

 been more costly in life and treasure to Japan than a 

 great and long-continued war. 



There appear to have been no fore-shocks strong 

 enough to give warning of the first and greatest earth- 

 quake. Among the crowd of after-shocks that followed, 

 one was strong enough to be felt at Osaka at 2.25 p.m. 

 on September i. Mr. J. J. Shaw at West Bromwich 

 recorded a second earthquake at 9 a.m. on the same 

 day (6 p.m. Japanese time). On September 2, almost 

 exactly twenty-four hours after the principal shock, 

 seismographs in Great Britain revealed the occurrence of 

 another earthquake, almost as powerful as the first, 

 with an origin at about the same distance and in nearly 

 the same direction as the first. No mention is made of 

 this earthquake in the Japanese reports, unless it is 

 the shock which on the morning of September 2 is 

 said to have destroyed 6000 houses in the town of 

 Kawaguchi. But its origin may have been situated 

 more to the south and possibly near the Bonin 

 Islands. 



The number of after-shocks was unusually great. 

 According to the Tokyo Central Observatory, 1039 were 

 recorded between noon on September i and 6 a.m. on 

 September 6, the numbers being 356 on September i 

 and 2, 289 on September 3, 173 on September 4, 148 on 

 September 5, and 63 during the first quarter of 

 September 6, the usual decline in frequency being thus 

 manifest. In the two months following the great 

 earthquake of 1854, the number of after-shocks actually 

 felt was 443. During the five days after the Mino- 

 Owari earthquake of 1891, 808 shocks were recorded 

 at Gifu. The number of after-shocks, however, seems 

 to depend on the magnitude of the vertical, rather than 

 of the horizontal, displacement ; and thus, the large 

 number following the recent earthquake may imply 

 that the movement which caused it possessed a notice- 

 able vertical component. 



Other evidence of vertical displacement at the epi- 

 centre is provided by the arrival of the sea-waves soon 

 after the earthquake. Little is known about these waves. 

 They appear to have swamped the reclaimed portions 

 of Yokohama and Tokyo and to have caused much 

 damage along the numerous creeks and canals. Manv 

 villages along the coast of the peninsula south of 

 Yokohama were washed away. The naval base at 

 Yokosuka (about 10 miles south of Yokohama) was 

 destroyed, partly by the earthquake, partly by the 

 sea-waves. There is no evidence, however, that the 

 waves were of great height like those of the Sanriku 

 earthquake of 1896. And it is important to notice 

 that, of the three cables leading to Tokyo, only one 

 was fractured by the earthquake, the others continuing 

 to work normallv. 



NO. 



281 I, VOL. I 12] 



With regard to the position of the epicentre, we have 

 some, though not much, evidence. The earthnnalc 

 was evidently stronger at Yokohama than at ']■ 

 16 miles farther north. The sea-waves may hav( 

 caused by submarine landslips, but they were pn/ 

 due to a vertical displacement of the ocean-bed. i n,. 

 the movement at the surface, at any rate in Tok\' 

 Bay, was not very considerable seems to lie indirattd 

 by the preservation of two of the three lines of cable. 

 The apparent lowness of the sea-waves may have been 

 due to the smallness of the vertical movements, but 

 it may have resulted from a restricted area of submarine 

 displacement, such as would be provided by an epi 

 central area crossing land on one or both sides of Sagani 

 Bay, the inlet leading up to Tokyo Bay. Not mu< !, 

 trust can be placed on the reported disappearance of 

 the island of Oshima, which seems to be near thi 

 epicentral district, but it may have taken part in .; 

 general movement of subsidence and thus be 01 

 diminished area. 



P'or our knowledge of the earthquakes of the Tokyo 

 district, we are chiefly indebted to the labours of Prof. 

 Omori. In two recent numbers of Seismological Note 

 (No. 2, 1922, pp. 1-21, and No. 3, 1922, pp. 1-30) he ha 

 described the semi-destructive earthquakes of December 

 8, 1921, and April 26, 1922, and the distribution of 

 earthquake-origins in the neighbourhood of Tokyo. 

 A glance at the map of Japan will show that the inlet 

 consisting of Sagami Bay and Tokyo Bay runs in a 

 northerly direction up to Tokyo, the entrance to tin 

 latter bay being known as the Uraga channel. On tin 

 west side, the inlet is bounded by the Sagami-I/ 

 peninsula, and on the east side, by the Awa-Kazus. 

 peninsula. During the eight years 1914-1921, kj 

 earthquakes originated in the country- round Tokyi 

 and, with few exceptions, in four seismic zones, on' 

 off the east coast of the Main Island, the second in th- 

 neighbourhood of Mount Tsukuba about forty mih 

 north-east of Tokyo, the third in and near the Awa- 

 Kazusa peninsula, and the fourth round Hakone at 

 the northern end of the Sagami-Izu peninsula. In 

 other words, during these years, the immediate neigh- 

 bourhood of Tokyo was seismically quiet, while th- 

 three mountainous regions surrounding the city at .^ 

 distance of about forty miles, gave rise to " ver) 

 frequent occurrences of earthquakes, which, thougli 

 often sharply felt in the city, are harmless, as tl ' 

 districts in question do not belong to a great seism !■ 

 zone." Then follows this remarkable prediction. 

 " In the course of time, however, the seismic districts " 

 referred to above " will become gradually quiet, whil- 

 the Musashi plain and the Tokyo bay may, as a com- 

 pensation, recommence its seismic activity, and may 

 result in the production of a strong earthquake, prol - 

 ably just after a year of marked minimum of seismi 

 frequency." 



The last strong earthquake (that of April 26, 1922 

 originated, according to Prof. Omori, off the west 

 coast of the pro\ance of Awa in the Uraga channel : 

 and, he concludes, " the Awa-Kazusa peninsula and 

 the Sagami earthquake regions, at present so active, 

 form obviously one continuous system sejiarated b\- 

 the Uraga channel of small seismic frequency, and it 

 was exactly at the latter locality that the . . . strong 

 earthquake [of April 26, 1922] took place. It seems 



