166 MR. A. NORMAN SHAW: A DETERMINATION OF THE ELECTROMOTIVE 
The following values for the other positions were obtained in the same manner as 
those in Table IV. :— 
(B) 8317-9 
(F) 8306-8 
(B) 8318-9 
(G) 8299-3 
(B) 8320-3 
(G) 8298-7 
(B) 8319-4 
(H) 8329-0 
(C) 8312-0 
(I) 8334-6 
(C) 8310-0 
(J) 8318-6 
(C) 8311-2 
(K) 8310-1 
(D) 8341-9 
(L) 8341-8 
(D) 8342-1 
(M) 8325-6 
(D) 8346-0 
(N) 8307-8 
(D) 8345-3 
(O) 8298-8 
(E) 8322-5 
(P) 8334-1 
(F) 8305-0 
In column (13) the values for d, half the mean distance between the planes of the 
large coils, are given. Each of these determinations was obtained from a full set of 
observations similar to those outlined in Section III. ( b ). 
It is necessary to express the ratio of the currents in terms of d where d would 
equal 12'4448 cm. at 20° C., the figure used for the mean as explained in the section on 
the main deflection observations. The slide reading must therefore be reduced to its 
value for d = 12‘4448 {l+a! (t — 20)} where “a'” is the co-efficient of expansion, 
taken as 0'000018. It can easily be shown that for small differences, an increase of x 
per cent, in d would mean a decrease of 0'60a; per cent, in the ratio of the currents, 
and hence an increase of 0'56x per cent, in the Kelvin-Varley reading. This 
correction is given in column (14). 
Columns (15) and (16) contain the corrected means. The readings at the upper 
end of the Kelvin-Varley scale are subtracted from 100,000. (The slide had been 
calibrated in terms of the exact value of 100,000 for the whole slide. See Appendix.) 
The sixteen positions include two equally reliable groups of eight, namely, positions 
A, F, I, N, C, H, K, P and B, E, J, M, D, G, L, O. It will also be seen that there 
are a number of suitably chosen groups of four whose means should approximate very 
closely to the final mean. 
Table V.—Summary of Mean Values. 
The mean value for positions A, B, C, D, E, F, G, H, I, J, K, L, M, N, O, P, is 8321'0 4 . 
Position A gives 8334-8 
Position B gives 8319 "1 
„ F „ 8305-9 
„ E „ 8322-5 
„ I „ 8334-6 
„ J „ 8318-6 
„ N „ 8307-8 
„ M „ 8325-6 
„ C „ 8311-1 
„ D „ 8343-8 
„ H „ 8329-0 
„ G „ 8299-0 
„ K „ 8310-1 
„ L „ 8341-8 
„ P „ 8334-1 
„ O „ 8298-8 
Mean = 8320-9 
Mean = 8321 • 1 
