oe 
ON SEISMOLOGICAL INVESTIGATIONS. 1] 
The average velocity, added in the last line, falls off very definitely, 
and the hypothesis of a ‘ polychordal’ P wave transmitted by numerous 
reflections with approximately uniform arcual velocity was therefore 
withdrawn. But in the course of the examination of Dr. G. W. Walker’s 
hypothesis of deep-seated focus a modification of the hypothesis suggested 
itself which seems worth attention. Repeating the subjoined figure from 
the 20th Report, 
we shall assume that cC is an arc of 17°-1 so that a P wave starting from 
a deep-seated focus E along EC will be reflected at C so as to strike the 
Bumtace again at C, C, C,... where CC; =C,C, =0,03. . . = 171; 
and after five such reflections will reach a distance A from K (the epicentre) 
equal to 5-5 x 17°-1 =94°. The whole time required would be, with 
adopted tables, 
5-5 X 246s. = 1353s., 
but if we adopt the corrections of Table VI. as in the column ‘ New’ 
we should get 
5-5 & 253s. = 1391s. 
The value extrapolated from the above-quoted values for Y would 
be still greater, viz. 1410s., but we see at once that a comparatively 
slight increase to the tables would give this exact value: viz. the time 
for an arc of 17°-1 must be 256s., indicating a correction to tables of + 10s. 
near A = 17° instead of + 7s. 
For paths slightly differing from EC, such as Ea or EB, the chords 
are greater, and five reflections will bring us beyond 94° in either case ; in 
