46 STUART : THB SBIMANGAL EARTHQUAKE OF 8TH JULY 1918. 
the Srimangal centrum) mentioned by Mr. Evans is estimated by 
him to represent 12 seconds and s'o the observed time — the calcu- 
lated time of arrival of P at Colombo falls from +52 to +40 seconds. 
The case of Dehra Dun is at present inexphcable. The suggestion 
of an error in the clock does not seem tenable because S and L on 
the record seem to be approximately in accord with the tunes 
registered at other observatories. The early arrival of P at Dehra 
Dun giving abnormal values for S — P and L — P, must therefore 
remain inexplicable. In the remaining cases quoted in the table 
between the observed and the calculated time. The difference for 
the arrival of P are exceedingly small and appear to indicate that 
the assumed time of 10 hours 21 minute^ 50 seconds G.M.T. for the 
time of origin is practically correct. This time indicated by the 
Indian and surrounding observatories differs by only 12 seconds 
from that indicated by the times of arrival of P at the British 
Observatories, and seems to be the preferable of the two. 
To advance the time suggested by the British records from 10 
hours 22 minutes 2 seconds to 10 hours 21 minutes 50 seconds for 
the time of occurrence at orighi would mean that the observed 
times of arrival of P at the British Observatories were a few seconds 
later than one would have been expected, whereas a retardation of 
the time suggested by the Indian and surrounding Observatories, 
from 10 hours 21 minutes 50 seconds to 10 hours 22 minutes 2 
seconds would mean that the observed times of arrival of P are 
several seconds earUer than they should be according to calculation. 
The former alternative appears to be the more acceptable, and I there- 
fore suggest 10 hours 21 minutes 50 seconds G. M. T. as the Time 
of the earthquake at its origm. 
RATE OF PROPAGATION OF THE LARGE WAVES. 
Accepting 10 hours 21 minutes 50 seconds G.M.T. as the time 
of the earthquake at its origin then the speed of the long waves 
becomes 3"7 km. per second. This speed seems to be indicated by 
the records from Dehra Dun, Simla, Bombay, Kodaikanal, Colombo, 
Manila and Barcelona, and there can be little doubt that it is the 
correct speed to assume for this earthquake. 
{2) DEPTH OF THE FOCUS. 
In attempting to determine, as far as possible, the depth below 
the surface at which the focus lay, only one method, of the many 
