ies REPORT— 1885. 
maximum amplitudes, maximum velocities, and maximum accelerations 
obtained in the pit with those obtained at about thirty feet distance at 
station J, they are for one particular earthquake respectively in the ratios 
of 1 ; 43,1: 52, and 1 : 82 (see diagram for March 20, 1885).!_ In most 
earthquakes the extent of motion has been too small to admit of measure- 
ment, and that there had been any movement could only be detected by 
holding the plate on which the record was written up to the light and 
glancing along it lengthways. This investigation tends to confirm the 
view which I have previously put forward, that an earthquake at a short 
distance from its epicentrum is practically a surface disturbance, prin- 
cipally consisting of horizontal movements. The vertical motion is small, 
and is best seen in the preliminary tremors either of an actual earthquake 
or of a dynamite explosion. From a practical point of view these results 
must be of the greatest importance to those who have to erect heavy 
structures in earthquake districts. Ata future time I hope to continue 
these experiments by comparing together the motion—Ist, of a foundation 
unattached to the sides of the excavation where it is built; 2nd, of a 
similar foundation connected to the sides by brushwood and covered with 
a layer of earth ; 3rd, of a similar foundation put in the ground in the 
ordinary method ; 4th, of a piece of ground isolated with trenches ; and 
5th, of the ordinary ground. 
Buildings in Earthquake Countries. 
As during the last few years so much destruction both to life and 
property has taken place in various parts of Europe, it seems that an 
epitome of the results of observations and experiments carried on in 
Japan relative to construction in seismic districts might not only be 
interesting, but possibly it might also be of practical value. When 
erecting a building it appears that we ought first to reduce as far as 
possible the quantity of motion which ordinary buildings receive; and, 
second, to construct a building that it will resist that portion of the 
momentum which we are unable to keep out. To reduce the momentum 
which usually reaches a building the following may be done :— 
1. Institute a seismic survey of the district or area in which it is 
intended to build, and select a site where experiment shows that the 
motion is relatively small. 
2. For heavy buildings adopt deep foundations (perhaps with lateral 
freedom), or at least let the building be founded on the hardest and most 
solid ground. It is perhaps because the tops of the hills in Tokio are 
harder than the plains that they have relatively the least motion. To 
what extent a building may be isolated by trenches or natural valleys is 
not yet known, but it must be remembered that a building only partially 
isolated may be exceedingly dangerous from the fact that motion entering 
in the unprotected side will make the excavations (cuttings, valleys, &c.) 
upon the opposite side into free surfaces which will swing forward 
through a range greater than they would have swung had the excavations 
not existed. 
3. For light buildings, especially if erected on soft ground, where the 
range of motion is always great, if the structure rests on layers of fine 
cast-iron shot, it cannot possibly receive the same momentum as a build. ° 
? Tam not yet prepared to state whether these ratios will hold for all earthquakes. 
