216 



ANNUAL REPORT SMITHSONIAN INSTITUTION, 1945 



1858 there was a similar occurrence. The detritus dammed the river 

 Zyogwanzi; the dam held for 49 days and then broke, flooding the 

 lower valley for 70 miles and drowning 148 persons. Other shocks 

 listed for region VI have no known features of special interest. 



Region VII. Kyushu Island. — The list includes the earthquakes of 

 1596, 1792, 1889, 1922, and 1914. In the earthquake of 1596 the island 

 of Urya-zima near Oita, with an area of V/ 2 square miles, sank beneath 

 the sea to a depth of 30 to 40 fathoms, two-thirds at once and the rest 

 later. In 1792 there was a great eruption of the volcano Unzen-dake. 

 This began in February. On May 1, possibly as a result of an earth- 

 quake, there was a great avalanche which produced great water dis- 

 turbance with heavy loss of life. It is possible that this should not 

 be listed as a severe earthquake. In 1914 there was a great eruption 

 of the volcano Sakurijima. The strongest earthquake, on January 

 12, was probably of volcanic origin though unusually severe for such 

 a shock. Buildings of soft material and fence walls were thrown 

 down. The other shocks listed for region VII were probably of 



volcanic origin. 



FORMOSA EARTHQUAKES 



Formosa has suffered severely from earthquakes, largely because 

 of the type of buildings, heavy but weak. Omori has listed 18 shocks 

 including 2 to 5 of the list from 1655 to 1906. In the shock of 1862 

 more than 1,000 persons were killed. In 18G7 a wave swept the harbor 

 of Keelung, northeast of Formosa; great landslides occurred. In 

 the earthquake of March 17, 1906. there was fault slipping 6 feet ver- 

 tical and 6 feet horizontal. (See pi. 3, fig. 2.) There were sand 

 eruptions and cracking of the ground. The shock of April 14, center- 

 ing 10 miles from the March earthquake, was not accompanied by 

 surface slipping. 



Table 2 . — Destructive earthquakes in Formosa, 1771-1943 



For explanation of terms, see table 1, p. 202. 



SOME LESSONS FROM JAPANESE EARTHQUAKES 



The Japanese have long wished to predict earthquakes, probably 

 never more than now. They have studied crustal movements by re- 



