118 REPORT OF THE CALIFORNIA EARTHQUAKE COMMISSION. 
hodograph should pass exactly thru the records of Calcutta and Mauritius; but this is 
too uncertain to justify the extension of the hodograph to Mauritius. It does not seem 
to be possible, from the observations of this earthquake, to draw any certain inference 
regarding the velocity of propagation of the first two phases much beyond 110°. 
The beginning of the second preliminary tremors is often the most easily recognizable 
point of the seismogram. The first preliminary tremors frequently have so small an 
amplitude that their beginning can not be determined; and sometimes there is no evi- 
dence of any movement until the second preliminary tremors arrive. The latter usually 
show themselves by a definite and well-marked increase in the amplitude of the recording 
instrument. 
The records of the Kingston earthquake of January 14, 1907, offer a very instructive 
example of the influence of the magnifying power of seismographs on the times recorded. 
Washington, with a magnifying power of 25, recorded the first preliminary tremors; 
Cheltenham, with a magnifying power of 10, began its record with the second preliminary 
_ tremors; whereas Baltimore, with a magnifying power of 6, only recorded the principal 
part. These three stations are close together and practically at the same distance from 
Kingston. 
The hodograph of the second preliminary tremors, exprest in terms of the distance 
measured along the chord, shows a point of inflection at a distance of about 9,000 km. ; 
this does not indicate that the average velocity diminishes at this distance, but merely 
that it does not increase as rapidly as it does at shorter distances; this part of the curve, 
however, is quite doubtful and we are not justified in drawing any very definite conclu- 
sion from itsform. It is extremely disappointing that the observations of this earthquake 
do not lead to definite results regarding the propagation of the disturbance to very great 
distances; for the point where the earthquake occurred and the time of its occurrences 
are both known to a satisfactory degree of accuracy, and instruments recorded the 
shock at stations as far as 162° distant, that is, very nearly to the antipodes. This further 
emphasizes the importance of installing instruments recording the vertical component of 
motion, and instruments with high magnifying powers, not less than 100; for they alone 
can be expected to yield satisfactory records of the times of the arrival of the various 
phases of very distant earthquakes, regarding which our information is still very vague. 
As the earthquake originated at some distance below the surface, the surface velocity 
in the immediate neighborhood of the epicentrum would be very large; it would diminish 
rapidly as the distance increased, would reach a minimum and again increase as the 
paths of the waves to the more distant stations extended deeper into the earth. If we 
had absolutely accurate observations, the hodograph, drawn from them, would be con- 
cave upwards near the origin, would pass thru a point of inflection a little further off, and 
would then pass into the general form, concave downwards, as drawn in the plate. See- 
bach * first pointed out that the form of the curve in the neighborhood of the origin could 
be used to determine the depth of the focus; he assumed constant velocity in all upward 
directions near the origin; Prof. A. Schmidt,? assuming increasing velocity with the 
depth, modified the results; but the degree of accuracy required of the observations is 
so great that all attempts so far made to determine the depth of the focus by this means 
are unreliable; and we can not expect to apply the method successfully until the accuracy 
of our observations is far greater than it is now. ‘Table 8 shows the distances of stations 
from the centrum in kilometers, in terms of their distances from the epicentrum meas- 
ured along the surface of the earth, and of the depth of the centrum; these distances 
take into account the curvature of the earth and are accurate to a fraction of a kilometer. 
1 Das Mitteldeutsche Erdbeben von 6 Mars 1872. Leipzig, 1873. 
an CEES Rees und Erdbeben. Jahreshefte fiir Vaterlands Naturkunde in Wirtemberg, 1888, 
p. 248, 
