General Characters of the Disturbance. 39
velocities of the normal vibrations in centimetres per second
are:—
Granite...... 395 x 103
Clay rock ... 348 x 103
The double shock felt at the distant stations may thus
have indicated the successive arrival of the vibrations, first
through the deep-seated hard rocks and then through the
overlying softer strata.
It is not here deemed advisable to attempt to decide
between these explanations, or to consider the possibility of
two distinct initial shocks having occurred. My own im-
pression is that the theory of relative conductivity perhaps
affords the most plausible explanation of the double shock,
at any rate at the most distant stations, inasmuch as there is
reason for believing that one of the two sets of vibrations
would completely die out before reaching these remote places.
Whether the normal or the transverse vibrations would be
the more persistent depends entirely upon the character of
the formation through which the wave is propagated.20
With reference to the sounds connected with the present
disturbance, the accounts are on the whole concordant that
the subterranean rumbling noise preceded the actual shock ;
in one or two instances only is the noise said to have accom-
panied the shock. There is no doubt that in some cases, and
especially within the area of structural damage, the sound of
the earthquake disturbance was inextricably blended with the
crash of falling masonry and the rattle and jar of the houses.
One noteworthy feature is the statement, repeated by a few of
the observers who were indoors at the time, that the noise
appeared to break in the first place overhead. This would
seem to indicate that the amplitude of vibration of the upper
parts of the buildings had become sufficient to cause
windows, furniture, &c., to rattle before the vibrations of the
lower storeys had become of sufficient amplitude to make
20 See Milne's remarks on this subject, in Trans. Seism. Soc, Japan,
vol. iii. pp. 133 and 134.