268 REPORTS ON THE STATE OF SCIENCE, ETC. 
Asia, 10 3 North and Central America, with the North Atlantic, 19 ; South 
America, 4; Japan, 15 ; Pacific and Indian Oceans, 9 ; total, 57... These 
have been selected as specially well observed and capable of having the 
epicentres well determined from the data. 
The solution for P proved to be straightforward. The times already 
found by H.J. and published by the Committee were right within a second 
up to about 19°; but there is a sharp bend in the curve at this distance, 
the maximum correction needed being — 5.7s. at 30°. Beyond that dis- 
tance smaller corrections are needed, reaching + 1s. about 60°. At greater 
distances still the corrections are again negative, reaching — 3s. at 105°. 
In the case of S there was much difficulty owing to misidentifications. 
It seems that the H. J. tables are nearly right to 15°, but need a slight in- 
crease to 19°, and that there is at this distance a discontinuity similar to 
that found for P.; a correction of — 8.7s. is needed at 30°. The correction 
vanishes about 60° and is about +4s. beyond 80°. 
The waves through the core! have also been discussed and times have 
been found for PKP(= P’ =P.P-P), SKS(= ScPcS) and SKKS(=ScPcPcS). 
The forms of the curves differ very little from those given by Gutenberg, 
but, additive constant corrections are needed to adapt them to the same 
focal depth. ‘Times for the diffracted P at distances up to 143° have been 
obtained. 
The outstanding problems relate to the determination of focal depth and 
of the depths at which PP, PS and SS are reflected. So long as the focus 
is within the upper layers the effect of focal depth is simply to make S arrive 
early by about the same amount at all distances, the forms of the P and S 
curves remaining unaltered. This additive constant can be determined 
and allowed for. But it remains doubtful to what focal depth the standard 
S curve corresponds ; though several methods have been tried none seems 
satisfactory. ‘The additive constant varies between different earthquakes 
by as much as 18s. This is inconsistent with the supposition that P move- 
ment is always. generated by the original shock ; in that case the variation 
could only be about 5s. There seems to be no doubt that in some earth- 
quakes there is primitive P movement and in othersnone. All the curves 
have been made to correspond to the same focal depth, but there remains 
some doubt as to what that depth is. Some earthquakes not recorded as 
having deep foci in the I.S.S. have proved to have foci 50 to 100km. below 
the top of the lower layer. 
PP, PS and SS are frequently recorded, but the residuals are irregular and 
it is still uncertain whether any definite conclusions can be drawn from them. 
The bends in the P and S curves at 19° may be the result of either a 
rapid continuous increase of velocity or a discontinuity at a depth of about 
400km., the velocities rising by about ro per cent. when it is crossed. On 
the former alternative points of large amplitude would occur, which have 
been sought by Miss Lehmann but not found. A discontinuity on the 
other hand would give reflexions, which again have not been found, but 
their amplitudes would in any case be small. The most likely one to be 
observable may be one of PP type reflected on the inside of the discon- 
tinuity, just as SKKS is the best observed reflexion on the surface of the 
central core. 
Other waves recorded with fair frequency are PKS, P’, and SKSP. 
An attempt will be made to construct empirical tables for these also. 
Valuable supplementary information has been received in correspondence 
with other seismologists, especially Miss I. Lehmann, Miss E. F. Bellamy 
1 The letter K is used for a compressional wave through the core (Kernwelle). 
The notation is taken from Bulletins issued from Georgetown. 
