SEISMOLOGICAL INVESTIGATIONS 227 
buted to Z effect or high focus in the superficial layers of the earth’s crust, 
but it is difficult to believe that a difference of focal depth in so very shallow 
a stratum should produce so great an effect on the travel time. The two 
shocks from the same epicentre make it impossible to attribute the dis- 
crepancy to peculiarities of internal structure, at least in this case.‘ 
The Seismological Bulletin of the Central Meteorological Bureau of Tokyo, 
for 1931, contains readings from many stations in Japan which would not 
otherwise be received by the J.S.S. Several of these readings appertain to 
large distant earthquakes, but mostly they are records of local shocks 
obtained with seismographs with short periods. Wherever possible, the 
epicentre determined at Tokyo is adopted for the Summary, but in a few 
cases it is necessary to modify the position in the light of information from 
other sources. As one example of the excellence of these readings, it may 
be mentioned that the table for the shock of 1930 November 25d. 19h. with 
epicentre 35: 1° N.139-0° E. contains records of thirty-nine Japanese stations, 
all with A less than 10°. 
Further valuable data have begun to appear in 1931 from the group of 
stations in California, whose headquarters is at Pasadena. ‘They are 
Haiwee, La Jolla, Mt. Wilson, Riverside, Santa Barbara, and Tinemaha. 
All their observations are extremely good and their position most fortunate, 
for, although they exist primarily for carrying out experiments on short wave 
transmission, they are able to record shocks in those distant parts of the 
Pacific too far west to give satisfactory readings at places near the Atlantic 
coast and for which observations to the east and north-east would otherwise 
be scanty. Pasadena itself, with its large instrumental equipment, often 
gives several readings for each phase, but when these only differ by one or 
two seconds the earliest is considered sufficiently representative. 
The earthquakes of 1930 July 23d. oh. and November 21d. 2h. from 
41‘1° N. 15-4° E. and 40-0° N. 19°5° E. are notable for lists of nearly 
seventy stations with A not greater than 22°. The epicentres are both given 
by De Bilt, the latter being quoted from the Bulletin of the Russian stations, 
and is of much greater accuracy than the round figures suggest. Here 
again we have large positive residuals for S, though in the nearest stations 
S* or Sg may be recorded instead of S. 
1930 December 3d. 18h. epicentre 18:2° N. 96°4° E. is strangely 
unsatisfactory. Although there are two large groups of observing stations 
in azimuths round 55° and 320°, consisting of nearly all the best data from 
Japan and Europe, yet the residuals are scattered indiscriminately over the 
wide interval between -+8s., which at 70° of A corresponds to a range of 
apparent errors in distance of about +1°5°. ‘There has been here some 
general difficulty in reading the phases, and examination of the traces for 
Oxford instruments shows marked but not exceptional disturbance by 
microseisms which in the case of instruments of high magnification might 
make it impossible to distinguish a P of very small amplitude. As contrast to 
this, the shocks of 1930 July 2d. and 1930 September 22d. 14h., with origins 
in the same neighbourhood, show as good a fit as can be wished, both for 
Japanese and European records. For these dates microseisms are almost 
entirely absent from the Oxford films and it remains a question whether 
stations in Japan as well as in Europe were troubled by the microseisms in 
making their readings of the December 3d. films. 
_* It may be that in the first of these two earthquakes no appreciable compres- 
sion waves were generated at the focus so that the phase recorded as P was really 
the phase sP; the latter phase is generated from shearing waves which reach 
the'surface near the epicentre. The interval between sP and S is less than that 
between P and S—F. J. W. W. 
