‘ 
METHODS OF STUDYING EARTHQUAKES 307 
two parts of the shock were superposed, showing that the 
impulses were not simultaneous, and that the focus within the 
concave part of the hyperbola was last inaction. In the Cornish 
earthquake of 1898, the interval between the two parts of the 
shock was so great that the first and weaker part was felt all 
over its disturbed area before the second was felt at its epicenter. 
Consequently, the broken line in Fig. 1, which surrounds all the 
places where the double shock was observed, constitutes the 
boundary of the disturbed area of the earlier portion. 
In studying an earthquake, there will be found on almost every 
point considerable conflict in the evidence collected. Much of it 
is no doubt due to inaccurate observation, part to a misunder- 
standing as to the information desired. It is in the records of the 
sound phenomena that the greatest diversity exists, a diversity 
which can hardly be ascribed to inattention or defective observa- 
tion, and which can only be explained completely on the suppo- 
sition that the sound is so deep that some persons are incapable 
of hearing it. Near the epicenter, the strength of the sound- 
vibrations is so great that they are audible to nearly every person, 
but the percentage of observers who hear the sound decreases 
rapidly towards the boundary of the sound-area. The variation 
in audibility throughout the sound-area may be illustrated by 
means of isacoustic lines. The percentage of auditors of the 
sound among those observers within a given area who felt the 
shock is taken to correspond to the center of the area in question, 
the lines joining adjacent centers are divided so as to give points 
where the perccntage would, on the hypothesis of uniform varia- 
tion, have certain definite values, say 90, 80, 70, etc., and lines 
are drawn through all points where the percentage marked is the 
same. The isacoustic lines for the Hereford earthquake of 1896 
are represented in Fig. 2. The axis of the isoseismal lines runs 
almost exactly northwest and southeast, and the points of great- 
est extension of the isacoustic lines lie on a curve (broken in the 
figure) which coincides almost exactly with the hyperbolic 
band referred to above. The explanation of the peculiar distor- 
tions of the isacoustic lines is that, along this band, the 
