SURVEY OE THE BRITISH ISLES FOR THE EPOCH JANUARY I, 1891. 
11 
In the case of the Declinations and Horizonta] Forces the more satisfactory result 
was reached that our earlier conclusions would not have been substantially altered if 
the plan now adopted had then been used. 
The only occasion on which any noteworthy difference would have been produced 
in the Declinations, was in the case of two observations made with magnetometer 60 
on September 30, 1887. 
It will be noticed, on referring to our previous Table (“ Mem.,” 90, p. 58) that 
the two values of the difference between Kew and No. 60 found on that day were 
4'’3 and 3''9, which do not agree well with the mean, viz., 2'‘5. If the correction for 
the month is applied, these figures are much improved, but our confidence in the 
improvement is shaken b^^- the fact that just at that time the position of the base¬ 
line of the Kew curves was altering very rapidly. Thus the number to be added 
to the readings of the self-registering instruments to reduce them to absolute 
measure was ;— 
— 74''9 in September, 1887 
— 72''4 „ October, ,, 
— 69''3 „ November, ,, 
It is therefore very doubtful what value ought to be assumed for September 30. 
In spite, however, of these rapid changes, the fact that we made six comparisons, 
of which two were in September and four in October, enables us to regard the mean 
of all the results as trustworthy. If we use (l) the monthly values for the base-line 
given above, or (2) the mean for the Summer half-year for the September 
observations, and that for the Winter half for October, the final mean results will 
only differ by — 0''6 or -f- 0''2 from the value obtained by using, in all cases, the 
mean base-line for the whole year. 
On the whole then, we come to the conclusion that there is no evidence that our 
results are affected by the slight modifications in the method of reduction, and that 
—as is 'primd facie most probable—the changes which have taken place must be 
ascribed to the survey instruments. 
We now give the details of the comparisons of the survey instruments with the 
Kew Standards. 
Here and throughout the paper, where initials are used to distinguish between the 
observers, R., T., G., B., and W. indicate Professor Rucker, Dr. Thorpe, and Messrs. 
Gray, Briscoe, and Watson, respectively. When Professor Reinold observed with 
Professor Rucker {cf, p. 28) the initials R. and R. are employed. 
The observations with the survey instruments at Kew were made in the Kew 
Magnetic House. 
