212 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1945 



On November 26, 236 were recorded in a single hour, but on the day 

 of the great earthquake there were very few. 



The Tanna tunnel (nearly 5 miles long) on the railway from Tokyo 

 to Kobe was under construction and the fault on which slipping oc- 

 curred had already been crossed. The west side of the tunnel where 

 it crossed the fault was shifted about 8 feet southward and 2 feet down- 

 ward. Otherwise no damage was done, though in a village directly 

 over the tunnel 55 percent of the houses were thrown down. Seismo- 

 grams showed that for periods of 1 second or less the amplitudes above 

 were two to four times that in the tunnel, while for periods of 4 or 

 5 seconds there was no difference in amplitude. 



Triangulation before and after the earthquakes showed that the 

 movements were in opposite direction on both sides of the fault by 

 similar but not equal amounts. From 1896 onward, many lines of 

 precise levels were run in the region. In the vicinity of Ito there 

 were two levelings before and two after the principal earthquake. It 

 was found difficult to find the relation between the vertical changes 

 and earthquake occurrence. 



The reports of luminous phenomena in connection with this earth- 

 quake, totaling 1,500 over an area of about 7,300 square miles, seem 

 to confirm similar reports in regard to certain Sanriku earthquakes. 

 They started a short time before the main earthquake (about 4 a. m.) 

 and continued for an hour in the epicentral region. Some lights were 

 like rays of the sun, some like searchlights, and others like fire balls. 

 They were colored from red and yellow to blue. Some of them were 

 brilliant. Although, in Japan, they have usually been attributed to 

 luminous microscopic organisms in the sea disturbed by earthquake 

 or tsunami, the explanation is not wholly satisfactory. 



Region IV. Mino-Owari, Kyoto, Tazima-Tango. — Imamura lists 14 

 shocks in the Mino-Owari region from 745 to 1918, and 18 in the Kyoto 

 region from 701 to 1830 (none thereafter) , and none before 1926 in the 

 Tazima-Tango region. Of those listed in table 1 there were 8 in 

 the entire region. The only shocks requiring detailed discussion are 

 the Mino-Owari of 1891, the Tazima of 1925, and the Tango of 1927. 

 The Mino-Owari was the greatest, though similar shocks in 745, 1185, 

 and 1586 were probably severe. 



The 1891 earthquake occurred south of the center of a long fault or 

 series of faults which extended in a northwest-southeast direction 

 across Honshu near its narrowest width. The visible fault slips ex- 

 tended for 40 and probably for 70 miles. They were generally en 

 echelon, with a maximum vertical slip of 20 feet and a maximum hori- 

 zontal slip of 13 feet. 



This was one of the greatest of Japanese earthquakes. It was too 

 early for cities of the region, such as Nagoya, to have become indus- 



