THE SYMPATHETIC SHOCKS. 
of lar^e amplitude there is a small thickening of the line equivalent 
to that indicating the commencement of preliminary vibrations in 
normal records. 
Finally there is the case of the Madras clock. There can be 
Httle doubt that the time of the earthquake at Srimangal was within 
12 seconds of 10 hours 21 minutes 50 seconds (i.M.T., as indicated 
by the recorded seismograms of observatories situated all over the 
world. The rate of travel of the long waves in this earthquake 
seems to have been high and approximately 3 7 km. per second. 
Even granting this rate, the long waves from the Srimangal centrum 
could not have reached Madras until more than a minute had 
elapsed after the disturbance had occurred which affected the Madras 
sidereal clock. It is reasonable to assume therefore that the dis- 
turbance of the Madras clock was not due to the long waves from the 
Srimangal centrum but to something else. 
In the Preliminary Report already published I have already 
indicated the probability of a subsidiary sympathetic shock having 
occurred not very far from Madras and it seems probable that the 
disturbance of the Madras clock was due to this subsidiary shock 
and not to the Srimangal shock. The Madras clock commenced 
to record irregularly at IC hours 27 minutes 58 seconds G.M.T., 
the Colombo seismogram commences with the record of long waves 
at 10 hours 27 minutes 30 seconds, and the Kodaikanal seismogram 
with the record of long waves at 10 hours 26 minutes 24 seconds. 
If these three observations are due to long waves from the subsi- 
diary sympathetic shock (they are all too early to be due to long 
waves from the Srimangal centrum), then the long waves from this 
subsidiary sympathetic centrum reached Kodaikanal first, the^ 
reached Colombo 1 minute 6 seconds later than they reached 
Kodaikanal, and they reached Madras 18 sees. later than they reachsd 
Colombo. That is to say, that the centrum was nearer to Kodaikanal 
than to Colombo, and nearer to Colombo than to Madras. The 
above times of arrival of the long waves at three different places 
approximately fix the position of the centrum. If the long waves 
from this centrum travelled at the same rate as the long waves 
emanating from the Srimangal centrum, namely 3-7 km. per second 
or 139 miles per minute, then av the moment when the long 
waves from this subsidiary centrum commenced to register on the 
Kodaikanal seismograph they had still approximately 195 miles to 
travel before they would reach Madras, and 152 miles to travel 
I 
