SEISMOLOGICAL INVESTIGATIONS 229 
out 5 with some of the data that were published by W. Rohrbach. The 
velocities derived for Love waves across Eurasia fit very closely the theory 
of the propagation of Love waves in a double superficial layer, and corre- 
spond to a continental structure in which about 15 km. of granite rest on 
some 30 km. of tachylyte, with dunite as the underlying material. The 
observations of Love waves which have travelled across the floor of the 
Pacific are consistent with the presence of a layer of granite 10 km. thick, 
with dunite as the subjacent material, or alternatively with the presence of 
about 16 km. of diorite over the dunite. 
One further difficulty arises in separating the two types of surface wave. 
Either the transition from ocean to continent, or variations of thickness of 
the upper layers within a continent or below an ocean floor, will cause, in 
addition to a scattering of waves, regular refraction and reflection. It is, 
therefore, not correct to assume that in general these waves travel by a great 
circle path from epicentre to recording station. This point is further 
discussed in the paper above cited. 
It is desirable to call attention to the intrinsic differences that may occur 
between different earthquakes. Prof. H. H. Turner found that, for the 
majority of the earthquakes that he investigated, the observations of L, the 
onset of the long wave phase, indicated transmission times corresponding 
to 0°477 minutes per degree, or about 3:9 km./sec. This corresponds to 
the onset of the Rayleigh waves and may be distinguished as LR. The 
corresponding onset of the Love waves (or ‘ Querwellen ’) should correspond 
to about 4°4 km./sec., or 0°42 min./deg. Actually, Turner ® found that 
for one earthquake, 1926 October 3d. 19h. the observations of L corre- 
sponded to 0-41 min./deg., so that in this earthquake the onset of the Love 
waves, which may be denoted by LQ, was specially prominent, although for 
European stations the arrival was in an easterly azimuth, and the E. com- 
ponents, on which the J.S.S. was preferentially based, would have been 
expected to record mainly the Rayleigh waves. Some stations record both 
LQ and LR; Simla E., for instance, gives in that earthquake LQ, whereas 
Simla N. records LR. Evidently, then, the primitive S movement in this 
earthquake was predominantly of an SH type, and such earthquakes may 
not be common ; at any rate, the point is worth further investigation, and 
information may be forthcoming from a study of the body waves. The 
fact that one of the two L onsets may be found among the ‘ Additional 
Readings’ is a justification, if any were needed, for 'Turner’s policy of 
printing as many readings in the J.S.S. as space and financial considerations 
would allow. 
Those who have had experience of the heavy analysis and computation 
involved in the problem of Rayleigh waves in a heterogeneous medium will 
welcome a paper’ in which Dr. Jeffreys applies Rayleigh’s Principle to find 
the velocity of Rayleigh waves. Even in the unfavourable case in which 
the whole change of properties occurs in a single abrupt transition, the 
velocity of a wave of any given period can be calculated within a few 
per cent. ; and the saving of labour is enormous. 
REAPPOINTMENT. 
The Committee asks for reappointment, for the continuation of the 
grant of {100 from the Caird Fund, and for a special grant of £50 for the 
maintenance of the International Seismological Summary. 
5 R. Ast. Soc., Mon. Not. Geophy. Suppt., III, 262 (1935). 
8 International Seismological Summary, 1926. 
7 Geophysical Supplement, III, 253.(1935). 
