ON SEISMOLOGICAL INVESTIGATIONS, 89 
degrees. These measurements are deduced from information published 
in the British Association circulars, which contain records from the 
particular kind of instrument they have adopted. ‘The earthquakes 
20 40 60° 80 100° 120 140° 169 
Relation of Amplitude (in seconds of arc) to distance from an origin for six 
large earthquakes. 
considered, which I number according to their order in the accom- 
panying diagram, were as follow :— 
I. California, April 18, 1906: observations from 30 stations, 17 good. 
II. Mexico, March 26, 1908 :’ 25 observations, 12 good. 
III. Mexico, July 30, 1909 : 20 observations, 11 good. 
IV. Chile, June 8, 1909: 19 observations, 9 good. 
V. Japan, November 10, 1909: 22 observations, 11 good. 
VI. Messina, December 28, 1908: 23 observations, 12 good. 
The observations indicated as ‘ good’ fall on or near to the curve 
to which they refer ; the remainder are far removed and erratically placed 
with regard to the same. The curves as they stand can therefore only 
be regarded as rough approximations to the truth. They indicate that 
up to about 80° from an origin amplitude decreases uniformly. 
From 80° to 120° it decreases less rapidly, and beyond this distance the 
