320 
Proceedings of the Royal Society 
ance of which is to he determined, is an electrolyte, the electromotive 
force acting along the circuit is only partly employed in producing 
the current. So that we obtain the equation 
where E is the electromotive force, R the resistance, x the current, 
and P the reverse electromotive force of polarisation. Since, for a 
considerable stretch of values of the current, P seems to be propor- 
tional to it, we may write 
where p is a constant for the given values of x. That is to say, 
if the ordinary method of measuring resistance could be used, w r hat 
would be measured is the actual resistance plus an unknown quan- 
tity. Hence, in most methods of experimenting, alternating cur- 
rents, or non-polarisable electrodes, are used so as to prevent polarisa- 
tion. In Horsford’s method, p is eliminated by giving R different 
values, while x remains constant. The difficulty here consists in 
the fact that polarisation may not be the same in both cases, for the 
nature of the solution at the electrodes may be different because of 
the liberated ions. The object of this paper is to describe a method 
of measuring the resistance w r hich is free from this objection. The 
while the other is joined to an electrode at A, which in turn is 
joined to a similar electrode at C by a metallic conductor of resist- 
ance r. The current may be supposed to flow in the directions 
indicated by the arrow-heads. Let A, B, C represent the potentials 
at these points, and let x be the current flowing from B to A, and 
y be the current flowing from B to C, while R is the resistance 
between A and B, and R + p is that between B and C. We have 
then 
E = Ra; + P , 
E = (R +p)x , 
annexed diagram will show 
the arrangement. ABC re- 
presents an electrolytic conduc- 
tor, with one pole of a battery 
( b ) joined to an electrode at B ? 
B - A = (R +p)x , 
B-C = (R+p + p)y, 
C - A =' ry . 
