SURVEY OF THE BRITISH ISLES FOR THE EPOCH JANUARY 1, 1891. 
9 
observations which may hereafter be useful, and, as the Kew values of the magnetic 
elements do not appear to have any special claim to the title of “ absolute,” we have 
reduced all our observations to the Survey Standard, which we previously employed. 
We have accepted without question the tables of corrections for temperature, &c., 
supplied with the magnetometers, on the authority of the Kew Observatory. There 
is no reason to doubt their accuracy, but, if our main object had been to explain the 
discrepancies between the instruments, a full investigation of the correction tables 
would have been necessary. 
Before discussing the methods of comparison in detail, we may remark that much 
of the earlier part of this paper is taken up with an investigation of the corrections 
necessary to make all the different parts of our work strictly comparable. These are 
not always small, but we believe that the precautions we have taken have left no 
residual errors of importance. The agreement between neighbouring stations as to 
the magnitude and direction of the Disturbing Forces, even where those forces are so 
small that the results are of the same order as the error of experiment, is so remark¬ 
able that it is impossible that they can be affected with any serious inaccuracy. 
The corrections to be applied to our instruments to bring them to the common 
Survey Standard have been determined by frequent comparisons at Kew. 
Measurements have been made with them at the Observatory, and the results 
compared with the values of the elements at the time of observation as given by the 
self-recording instruments. 
These experiments proved that the differences between the Kew and survey 
magnetometers altered from time to time. This may have been due to changes in 
the standards, but we have no hesitation in adopting the view that the cause of the 
variability must rather be sought for in the instruments employed in the field. 
The indications of magnetometers 60 and 61, which were certainly constant between 
1884 and 1888, appear of late to have changed. 
This is probably due to the fact that as the corrections to these instruments w'ere 
best known, we thought it desirable that the gi'eater number of the observations 
required for the new survey should be made wdth them, and therefore assigned them 
to the assistant observers. They have thus, for three years, been in daily use 
throughout the six summer months, have travelled many hundreds of miles by train 
and by carriage, often, in the latter case, over rough country roads, and have neces¬ 
sarily been exposed to the vicissitudes of the weather. Even when all possible care 
is taken, it is probable that under such circumstances the results would be slightly 
modified, and such has apparently been the case. In this connection the following 
points deserve attention. 
The difference between the Declinations given by the survey magnetometers 60 and 
61, and the Kew instrument was about 2'‘5 up to 1887, and was less than 0''5 in 
MDOCCXCVI.—A. 0 
