SURFACE REFLEXION OF EARTHQUAKE WAVES. 
393 
vibration is not in the azimuth plane. The clearest movement which satisfies this 
condition I have marked provisionally as PS. It is South 4‘5 mm., West 7'0 mm., 
vertical 2‘5 down. This gives azimuth 57^° N.E. instead of 58^-° N.E. derived from 
P, which is really quite good agreement. Taking .the theoretical angles of 
impingence, we get 
— 179A— 1'60B = — 8’3 = -\ / 4 2 '5+ 7 2 '0, 
+ CV88A—P20B = 2'5, 
whence A = 3'9 mm., B = 0'8 mm., as the amplitudes of PS and SP at incidence at 
Pulkovo. 
I am, however, doubtful about the movement, for while E and Y are clear enough 
the N movement is conspicuously like the return of the seismograph to rest after an 
earlier excursion, and if so the assumption that this is PS is false. 
From examination of a number of records for different distances, both from 
Pulkovo and Eskdalemuir, I am very much disposed to think that the Wechselwellen 
are not clearly separated from S until the distance exceeds 9000 km., and that at 
greater distances one is very apt to mistake them for S itself. Indeed I am 
convinced that I made the mistake myself. If this view is correct, it raises the 
question whether the marked periodicity which one associates with S at shorter 
distances is not really due to the Wechselwellen coming earlier than the present 
time curves would suggest. 
In conclusion, it appears that quantitative analysis of reflexion theory on simple 
lines- removes a number of difficulties that have hitherto attended interpretation of 
seismograms, and suggests that a systematic examination of both times and 
magnitudes of P, S, and at least PB, SR, and PS or SP, is likely to afford important 
information about the primary time curves. It is manifest that horizontal component 
seismographs alone are quite inadequate. 
So far as the present investigation fails to resolve all the difficulties, it suggests 
the proper lines of further investigation. 
These are (l) investigation of the reflexion of waves from a variable layer of 
thickness comparable with that of the earth’s crust; (2) investigation of the 
propagation of an impulse from a disturbed centre into the earth, in order to gain 
some idea of the relative magnitude of disturbance in different directions and at 
different distances. The results even for a uniform earth would be a valuable guide 
towards the law of absorption of waves, for the determination of which material 
undoubtedly exists in the observed magnitude of reflected waves. 
PRESENTED 
i sjun.uij 
