48 STUART : THE SRIMANGAL EARTHQUAKE OF 8TH JULY 1918. 
{3) THE SYMPATHETIC SHOCKS. 
In my Preliminary report I deduced the existence of two sym- 
pathetic shocks, one having an origin under the Bay of Bengal off 
the Madras coast, and the other under the Bay of Bengal off the 
Arakan coast near Akyab. Further information seems to confirm 
the existence of both, but has modified the position of the centrum 
of the more southerly shock. The evidence of the existence of 
this Southern Indian centrum is given by the unusual character of 
both the Kodaikanal and the Colombo seismograms and by the 
behaviour of the sidereal clock at the Madras Observatory. I will 
take them seriatim. 
The Kodaikanal seismogram begins abruptly with waves of large 
amplitude and shows no evidence of the usual preliminary tremors. 
It apparently commences with the record of long waves. This is 
the reading of the seismogram by Mr. J. Evershed, F.R.S., Directoi 
of the Madras and Kodaikanal Observatories, and I agree with him. 
In order to decide whether this is actually the correct interpretatioiij 
or w^hether it is possible that the record really commences with the 
register of preliminary tremors which so were intense that they might 
be mistaken for large waves, I visited Kodaikanal, and went care- 
fully through all the seismograms that have been recorded by the 
instrument since it was erected. The result was to confirm the 
opinion that this seismogram does commence with the record of 
large waves. In every other instance preliminary tremors on the 
usual scale have been recorded, coromencing as small vibrations 
gradually increasing in intensity. In no other case does the record 
commence abruptly, as in this earthquake, with waves registering a 
large amplitude and decreasing in intensity. The present seismo- 
gram commences with waves showing an amphtude vastly greater 
than has been recorded at the commencement of any other earth- 
quake, and I am convinced that the view that the commencement 
of the record is a register of the preUminary vibrations from Sri- 
mangal only, and nothing else, is quite untenable in the present 
instance. A further point which is against this view being tenable 
is that the Kodaikanal seismogram has no resemblance to the record 
which is of a perfectly normal type, given by the Bombay Milne 
instrument, although Bombay and Kodaikanal are situated at nearly 
the same distance from the epicentre at Srimangal. 
The Colombo record is very similar to that registered at Kodai- 
kanal with the exception that just before the abrupt start of vibrations 
