206 REPORT — 1882. 



Examples of the records obtained from these instruments, which 

 records are written on plates of smoked glass that are subsequently var- 

 nished and photographed, are given with the description of the instru- 

 ment. The greater number of earthquakes which have been recorded 

 in Tokio have also been recorded in Yokohama, Chiba, and Kisaradzn. 

 Now and then the same earthquake has also been recorded in Kamaishi, 

 and now and then even in Hakodate and Saporo. In no instance during 

 my seven years' i-esidence in Japan have earthquakes, even when they 

 have originated so near to Yokohama as to destroy chimneys and to 

 completely unroof houses, been propagated so far sonth as Kobe. So 

 far as ordinary earthquakes are concerned Kobe belongs to a special 

 seismic area. Nagasaki in a similar manner is independent of all the 

 other districts where seismographs have been placed. The reason why 

 an ordinary earthquake in North-Eastern Japan is unable to spread far 

 beyond Yokohama towards the south-west apparently depends upon the 

 fact that in going southwards it is intercepted by many high and 

 broad tracts of mountainous country. For a similar reason these same 

 earthquakes are unable to cross the central backbone of the island, and so 

 disturb the inhabitants upon the eastern shores. These facts have been 

 illustrated in a very remarkable manner during the last year by an 

 analysis of some hundreds of communications which I have received from 

 various parts of North Japan. 



These communications were obtained by sending to the Government 

 offices at all the important towns within a radius of from 60 to 100 

 miles of Yedo bundles of post-cards, with a request that every week one 

 of these cards should be returned with a statement of the earthquakes 

 which had been felt. The result obtained by the examination of these 

 communications showed that a great number of earthquakes came from 

 the north-east, and hardly ever passed the ranges of mountains to the west 

 and south-west. Subsequently the boundary of the post-card area was 

 extended farther to the north, and the result which was obtained showed 

 that many earthquakes came from the sea on the east._ The general 

 results obtained by this system of investigation are exhibited in a seis- 

 mological atlas which is being prepared for North Japan, in which the 

 area "shaken by every shock has been tinted, the dark tints indicating 

 where the shock was most severely felt, or where there were a number 

 of shocks at short intervals corresponding to the single shock felt at 

 more distant localities. The information to be derived from this series 



of maps is — i • i • 



1st. An approximate origin for many of the shocks, which is of great 



value as a check upon the records of instruments and an assistance to 



their interpretation. 



'2nd. In North-Eastern Japan there appear to be several^ seismic 



centres, and from these centres ordinary disturbances only radiate into 



surrounding districts, the boundaries of which are sharply defined by 



certain ranges of mountains. 



3rd. The greater number of seismic centres appear to be beneath the 



ocean. • • j. j 



4th. In the case of heavy shocks, and shocks which have originated 



some distance out at sea, the disturbance may be felt over the whole ora 



great portion of North-Eastern Japan, from Saporo and even Nemuro in 



the north to Yokohama in the south. 



5th. Although disturbances may be felt for several hundreds of miles 



