64 REPORTS ON THE STATE OF SCIENCE.—1914. 
muir 177°E., 51°N.; on July 9 Pulkovo gives 30°-3 E., 2°-1N., and Eskdale- 
muir 33°-9 H. and 5°-3.N.; and at greater distances still the discordance 
may be 5° or even 10°. The azimuths may still be good, though as the 
azimuthal lines do not meet so sharply, the determination becomes less 
definite ; and, moreover, it must be remembered that actual errors in 
the adjustment of the booms become of greater importance. We have 
nothing to set against the clear evidence offered in Section IV. that the 
tables for § are in error, though since the errors there found are only 
relative, we may add a constant to them all, substituting, for instance, for 
Error at 15° 35° 55° 75° 95° 115° 
m. m. m. m. m. m. 
—0:3 —0-1 +0-1 +0:3 +0-5 +0-7 
the revised values 
0-0 +0-2 +0-4 -+0-6 +0°8 +1-0 
so that the error is small near the epicentre. 
Similarly the errors for P might be written— 
Error at 15° 35° 55° 75° 95° 115° 
0-0 0-0 +0-1 +0-2 +0-4 +0-6 
if we determine to keep the error small near the epicentre. In this case 
it seems possible that the revised tables just published by the K.G. Landes- 
antalt fiir Meteorologie und Geodynamik in Zagreb (Agram) might supply 
information which would determine the unknown arbitrary constant. 
The errors of the Galitzin tables indicated by Zagreb at the above points 
are 
m. m. m m. m. m. 
+0-1 +0-1 0-0 +0-1 -+0-2 +0:3 
Difference +0-1 +01 —0-1 —0-1 —0-2 —0:3 
The differences do not, however, remain constant, even approximately. 
The present comparison indicates larger errors for values of A greater 
than 75° than the Zagreb tables admit. 
It thus appears that the moment is not yet come to suggest corrections 
to the tables which are likely to meet with general acceptance. It seems 
better to retain the old tables until a much greater mass of material has 
been discussed, and the old tables will accordingly be used for the com- 
parisons made at Shide at any rate for the observations of 1914. The 
discussion of some 100 earthquakes should provide corrections approxi- 
mating to definitive ones. Meanwhile, the best available corrections 
to the tables from the material above discussed, incorporating the in- 
formation derived from the next section, are given at the end of the next 
section. ° 
IX.—Discussion in Azimuth. 
If the receiving stations are arranged in azimuth (A) round the epi- 
centre, then 
(a) Assuming the velocity of transmission constant in all azimuths, 
any error (8) of position of the epicentre will give rise to an error 
c+ ecos (A — Ay) 
in the observed times at the stations: where A, is the azimuth in which 
the epicentre is erroneously displaced ; A is the azimuth of the receiving 
