52 REPORTS ON fHE STATE OF SCIENCE.—1913. 
but the first of them gave the direction in which we should seek for 
an origin. In the British Association Report, 1900, p. 79, I gave 
several methods which I use when mapping the position of epicentres. 
These methods were dependent on a number of observations made 
at several more or less widely separated observatories. 
As a slight addition to these I submit the following: If we have 
registers from a number of stations for large earthquakes it is usually 
easy to read the times of commencements and other phases of motion, 
together with the amplitudes. An inspection of the records which 
refer to a given earthquake shows the stations nearest to its epicentre, 
and any one of these should give us the distance of the same, and 
if we know this we can easily compute the time at which the shock 
originated. The difference between this and the arrival of the large 
wayes or the maximum motion at other stations enables us to compute 
their respective distances from the district from which they radiated. 
The intersection of ares, which I draw upon a ‘black globe,’ which 
correspond to these distances should represent the epifocal area. 
I venture to mention this simple and self-evident way of procedure 
because it is frequently of use when other methods fail. Preliminary 
tremors may have been eclipsed by air tremors or microseisms, or 
they may have died out on their journey, with the result that the 
selsmogram may Only present very small records which represent the 
large waves or maximum motion. 
V. On the Variation of Earthquake Speed with the Variation in the 
Direction of Propagation. 
In the British Association Report, 1908, p. 74, I showed thai 
megaseismic motion was propagated from its origin farther to the 
east and west than it was in the direction of a meridian. One 
explanation for this is that in the former direction the rigidity of the 
propagating medium may be greater than it is in the latter direction— 
a suggestion that falls in line with the observations of Dr. Hecker 
on the gravitational influence of the moon on the crust of our world. 
If this hypothesis is correct it might be inferred that the velocity of 
propagation of earth waves would be greatest in an east-west direction. 
To test this I took earthquakes Nos. 859, 860, 884, 1111, 1170, 
1260, 1363, and 1632 (see British Association Report, 1912, p. 71). 
I selected these particular disturbances because the positions of their 
epicentres and times of origin were known, and also because they 
had been recorded at widely separated stations. For any particular 
earthquake the only observations considered were those made at stations 
the bearings of which from the epicentre were within 30 degrees of 
east and west or within 30 degrees of north and south. The following 
tables only refer to maximum motion or large waves :— 
Earthquake No. 859, June 25, 1904, origin 160° E. 53° N. 
Per sec. 
Bombay, time to travel 48 m., distance 74°, velocity 2°85 km. 
; ) Mauritius, 5 68 m., ee bic Gee as 3°10 km. 
East-West Kedaikanal, ” 48 m., ” Miles 2 2°93 km. 
\ Calcutta, Ps 41 m., ee Gl Rate 2°75 km. 
Average . ° « 2°90 km. 
SF 
