THE SViMPATn[7r[(' SHOCKS 
5i 
is how the Colombo seii^iuogram does coinniencc. The Colombo 
«eismograin gives further evidence. The Colombo instrument is a 
Milne seismograph with horizontal pendulum which is set north- 
south, so that the line on the seismogram is first displaced towards 
the direction from which the long waves arrive. If the long waves 
first registered on the Colombo instrument came from the Srimaugal 
centrum, therefore, the hne on the seismogram should first have 
been disj^laced towards the east. This is not what actually happened. 
On the Colombo seismogram the first displacement of the line, after 
the slight thickening due to the preliminary tremors from the 
Srimangal centrum, is towards the west indicating that the long 
waves causing the displacement came from a centrum situated west 
of Longitude 79° 55', E., and not from the Srimangal centrum. 
Furthermore there is the record from Rocca di Papa near Rome. 
There on three different records the waves of a second earthquake 
of small intensity are recognized superposed on the long waves 
registered from the Srimangal centrum, and these Smaller waves 
commence practically exactly where one would expect the long 
waves from the Madura centrum to commence recording. 
The assumption of a sympathetic shock from a centrum situated 
near Madura seems to me the simplest explanation of the observed 
facts. Its occurrence is not improbable because a previous earth- 
quake has been recorded from nearly the same spot, and its occur- 
rence would be an explanation of the Kodaikanal and Colombo 
seismograms, which otherwise are extraordinary, of the disturbance to 
the Madras clock which otherwise seems inexplicable, and would 
explain the Rocca di Papa seismograms which indicate a second 
shock from a centrum other than the Srimangal centrum. 
I therefore think that the occurrence of a sympathetic shock 
with a centrum situated near Madura has been reasonably estabhshed. 
The other sympathetic shock described in my Preliminary report of 
the earthquake, is sufficiently substantiated by the earthquake 
reports and the unusual course followed by the isoseists. There is, 
however, one point of interest about it which I have not yet men- 
tioned, and that is its effect upon the Batavia seismogram. 
Batavia records M. at 10 hours 38 minutes which if due to the 
long waves from the Srimangal centrum necessitates the assumption 
that the rate of propagation of the long waves between Srimangal 
