BRITISH ISLES FOR THE EPOCH JANUARY 1, 1915. 
3 
think field observations are hardly good enough for this purpose. Moreover, observa¬ 
tions at two 'distances add considerably to the time required in the field witliout 
materially contributing to improved accuracy. I decided to use only one distance, 
viz., 25 cm., in the field observations, and in order to avoid personal error of setting 
from day to day I had fixed geometrical stops put on the bar. 
Special experiments were therefore made at Cambridge to obtain the distribution 
constants. Three distances 25, 30 and 35 cm. were used, and assuming the formula 
l+T/r^ + Q/r* the deduced values were 
P==8-90, Q=-2479 
and the value of log (l J-P/r^-t-Q/r*) at 25 cm. was 0’00342. 
This value was used throughout the survey. It had been calculated on the 
reputed certificate values of the distances and temperature coefficient. When the 
bar had been re-measured and the temperature coefficient determined the values were 
re-computed and gave 
P = 8-4, Q = -2245 
and 
log (l-tP/'r^-pQ/?''*) at 25 cm. = 0‘00333. 
Thus the observed values as recorded in the tables would require to be reduced by 2y 
on this account. 
It is by no means certain that one gets more accurate absolute values by using a 
two-constant formula. If we assume that the form (14- is better and that the 
differences obtained in the experiments are errors, I find that the least-square 
solution of my observations gives P' = 2’96, so that 
log (l 4-P7'^‘^) at 25 cm. = 0'00205. 
This assigns errors in the observations of — l'8y at 25 cm , 4-3’Gy at 30 cm., — l'8y at 
35 cm., which are by no means unreasonable. If this latter assumption is correct the 
observed values would have to be reduced by 25y. 
A visit was made to Greenwich in January, 1914, and while the comparison of 
Declination and Dip was satisfactory, my value of Force was considerably larger than 
that supplied by the Observatory. A re-determination of the moment of inertia of 
the Greenwich magnet was made by the Astronomer Royal and led to an increase of 
20y in the Observatory values. 
These necessary preliminary experiments took up a good deal of time, and mean¬ 
while an important arrangement was made. I have referred to Sir Arthur Rucker’s 
wish that the stations should be permanently marked and the azimuths of suitable 
marks determined once for all. The Director-General of the Ordnance Survey, South¬ 
ampton, Colonel (now Sir Charles) C. F. Close, K.B.E., C.B., was approached on 
the matter, and as a result the O.S.O. undertook to mark permanently the stations 
and determine the azimuth of suitable reference objects. The importance of this 
B 2 
