134 EAETHQUAKES. 



likely to be shaken down, as in most of the previous ex- 

 amples. 



The results of my own observations on this subject 

 point as much in one direction as in the other. In Tokio, 

 from instrumental observations upon the slopes and tops 

 of hills, the disturbance appears to be very much less 

 than it is in the plains. Thus, at my house, situated on 

 the slope of a hill about 1 00 feet in height, for the earth- 

 quake of March 11, 1882, I obtained a maximum ampli- 

 tude of motion of from three to four millimetres only, whilst 

 Professor Ewing, with a similar instrument, situated on the 

 level ground at about a mile distant, found a motion of fully 

 seven millimetres. This calculation has been confirmed 

 by observations on other earthquakes. Thus, for instance, 

 in the destructive earthquake of 1855, when a large por- 

 tion of Tokio was devastated, it was a fact, remarked by 

 many, that the disturbance was most severe on the low 

 ground and in the valleys, whilst on the hills the shock 

 had been comparatively weak. As another illustration, I 

 may mention that within three-quarters of a mile from 

 my house in Tokio there is a prince's residence which 

 has so great a reputation for the severity of the shakings 

 it receives that its marketable value has been consider- 

 ably depreciated, and it is now untenanted. 



In Hakodadi, which is a town situated very similarly 

 to Gribraltar, partly built on the slope of a high rocky 

 mountain and partly on a level plain, from which the 

 mountain rises, the rule is similar to that for Tokio, 

 namely, that the low, flat ground is shaken more severely 

 than the high ground. At Yokohama, sixteen miles south- 

 west from Tokio, the rule is reversed, as w^as very clearly 

 demonstrated by the earthquake of February 1880, when 

 almost every house upon the high ground lost its chimney, 

 whilst on the low ground there was scarcely any damage 



