JAPANESE EARTHQUAKES — HECK 207 



Region V. Japan Sea coast of Honshu, except Tazima-Tango. Earthquakes 

 generally less severe than in regions I-IV. Tsunamis small. 



Region VI. Interior of Honshu, not included in regions III and IV. Character 

 ized by landslides. 



Region VII. Kyushu Island. Shocks usually associated with volcanic activity. 



Region I. Submarine, west of Tuscarora Deep. — Only 4 of the 63 

 great earthquakes are in this region ; but this is not a complete picture. 

 Omori states that from the year 416 to about 1900 there were 23 great 

 tsunamis in regions I and II, and he found that of 257 shocks felt over 

 an area of more than 25,000 square miles, recorded from 1885 to 1905, 

 145 were in the same region. Imamura lists 15, some moderate, in 

 region I, from the year 869 to 1933. This supports his statement that 

 region I is the breeding ground of big earthquakes and that the ad- 

 jacent Sanriku coast is never free from the danger of destructive 

 tsunamis. Those of 1896 and 1933 will be described. The former 

 was the greater one, but the 1933 tsunami received the most thorough 

 scientific investigation ever given to this phenomenon. 



The 1896 earthquake was felt from Hokkaido to the Kwanto region, 

 but no important damage was caused. The shock was slow and undu- 

 lating. About 20 minutes later, a wall of water 10 to 20 feet high 

 at most places rushed in, and, as people were celebrating a holiday on 

 the beaches, thousands were swept away and drowned and many vil- 

 lages were laid flat. The waves reached a height of nearly 100 feet 

 at one place and 75 feet at a number of others. However, the damage 

 was nearly as great for a wave of 20 feet as for greater heights. Fish- 

 ermen who were well offshore failed to recognize anything unusual and 

 were greatly surprised on their return homeward to find the sea strewn 

 with house wreckage and corpses. The wave was recorded by the 

 tide gage at Sausalito, Calif. 



The great variety of height of wave was studied for the 1933 earth- 

 quake. The circumstances for the arrival of the wave were the same 

 as those in 1896, though it had less amplitude. The height over the 

 deep Pacific Ocean was estimated at 3 feet, and the wave length at 60 

 to 600 miles, which has much to do with recording at a distance. The 

 wave was recorded by tide gages in California and the oscillations were 

 noticed at Iquique, Peru, 9,000 miles from the place of origin. On the 

 coasts of Honshu and Hokkaido the usual height was 7 to 15 feet, 

 but there were maximum heights of 96 feet and 75 feet at several places. 

 In one bay the height was 10 feet at the entrance, 26 feet halfway up 

 the bay, and 75 feet at its head. Off one of these bays a motorboat 

 capable of a speed of 12 miles an hour could not make headway. The 

 explanation of the increase of height lies in the V-shaped character of 

 the bay. The wave in some places arrived as a flood, resembling a 

 very high tide, and at others it arrived as a vertical wall like a great 

 breaker — with high horizontal acceleration. Usually a withdrawal 



