24 
MR. GEORGE W. WALKER ON THE MAGNETIC RE-SURVEY OF THE 
The scheme is as follows :— 
District. 
Latitude. 
Longitude. 
Stations. 
Observed. 
Still to Observe. 
I. 
56°-59° N. 
2-5° W. 
22 
0 
II. 
56°-59° N. 
.5‘’-8° W. 
7 
10 
HI. 
54°-56° N. 
0°-5° W. 
18 
3 
IV. 
N. 
5°-10° W. 
16 
2 
V. 
.52°-54° N. 
■ 2 ° E.-r w. 
23 
0 
VI. 
52“-.54“ N. 
r-5° w. • 
24 
0 
VII. 
5r-54° N. 
6-11° W. 
31 
0 
VIII. 
50''-52° N. 
1 
o 
18 
0 
IX. 
50°-52° N. 
l°-6° W. 
22 
0 
In this scheme Cavan, strictly belonging to VII., was moved to IV., and King’s 
Sutton, strictly in VI., was moved to VIII. They are seen to lie practically on the 
corresponding boundaries and were moved to secure a more equable number of stations 
per district. 
Since District II. is far from complete, it was not included in deducing the mean 
isomagnetics. This procedure has been deliberately adopted with a view to including 
District II. in the general scheme in the simplest manner when more peaceful times 
obtain, and the observations have been completed. 
Each district was reduced separately. I examined very carefully whether the 
method of least squares would be of real advantage, and I came to the conclusion that 
Rucker and Thorpe’s method of “equations of conditions” (‘Phil. Trans.,’ vol. 181 
(1890), p. 235) was the best and simplest procedure, and that the method of least 
squares gave no gain in accuracy commensurate with the extra labour involved. 
But an important difference in procedure was adopted. It was evident that at 
some stage the geographical components would have to be calculated ; and I therefore 
decided to convert the values of H, D and I straight away to N, W and V before 
reducing. This procedure is generally recognised to be rational, and it possesses the 
important advantage, that it enables one to deduce with little trouble the best 
“ Potential Solution ” for the district. 
A slight digression from the main argument must now be introduced in order to 
explain this matter. It may be premised that, except close to the earth’s magnetic 
poles, the magnetic state of any small region of the earth’s surface may be regarded 
as due to magnetic forces from (l) causes at a considerable distance from the region, 
and (2) causes within the region. An analysis of a magnetic survey aims at separating 
these two causes. 
If the first cause existed alone, the magnetic state of the region would be 
characterised by a potential function, and if the region is small enough the isomagnetic 
