50 STUART : THE SRIMANGAL EARTHQUAKE OF 8TH JULY 1918. 
before they would reach Colombo. If therefore a circle is des- 
cribed having Madras as centre and a distance of 195 miles as 
radius, and another circle is described having Colombo as centre 
and a distance of 162 miles as radius>, then the long waves emanat- 
ing from the subsidiary sympathetic centrum must have reached 
the circumferences of these two circles and also Kodaikanal at the 
same moment. A centrum a few miles south of Madura would be 
equidistant from Kodaikanal and the circumferences of these two 
circles and therefore would appear to be about the position of this 
subsidiary sympathetic centrum. 
If the rate of travel of the long waves from this Subsidiary 
centrum be supposed to be 3-5 km. per second however, then the 
position of the centrum would be a few miles further east. In 
either case the position coincides closely with the centrum of a 
previous earthquake recorded by M. de Ballore in Mem. Geol. Sur. 
Ind., Vol. XXXV, page 28, and map. 
If we assume a rate of propagation for the large waves from this 
centrum similar to that deduced for the rate of propagation for the 
long waves from the Srimangal centrum, then it would appear that 
the time of the subsidiary shock at the Madura centrum was shortly 
after 10 hours 26 minutes, a time which agrees almost exactly Avith 
the calculated time at which the prehminary vibrations from the 
Srimangal centrum would reach Madura. Plate 12 shows the 
position of the Madura subsidiary centrum and also shows the 
position reached by the long waves from both the Srimangal centrum 
and the Madura subsidiary centrum at 10 hours 27 minutes 58 
seconds, the moment when the Madras clock commenced to be 
affected. From this map it will be seen that one would expect 
the arrival of the preliminary tremors from the Srimangal centrum, 
and the arrival of all the waves from the Madura subsidiary centrum 
to have been approximately simultaneous, giving a seismogram 
commencing abruptly with the register of large weaves which is 
actually how the Kodaikanal seismogram does commence. On the 
other hand one would expect the preliminary tremors from the 
Srimangal centrum to have arrived at Colombo earlier than the 
waves from the Madura subsidiary centrum giving a seismogram 
showing a slight thickening of the line, due to the prehminary 
tremors from the Srimangal centrum, slightly previous to its dis- 
placement by the waves from the Madura subsidiary centrum, which 
