TRANSITION RESISTANCE—MACGREGOR. 51 
insertion across the cell increased slightly the depth of the 
liquid and displaced a certain quantity of liquid by a better 
conductor, the change produced in the resistance of the cell must 
have been very small. The specific resistance of the solutions 
I used was about 28 ohms, and the ends of the tube connecting 
the compartments were about 6 cm. from the electrodes. The 
total resistance of the cell was about 3000 ohms. With these 
data it may be easily shown that the change in the resist- 
ance of the cell produced by the insertion of the plates could 
not be more than between ;5h55 and g5455 of the resist- 
ance of the cell itself. The fact that the connecting tube 
was completely surrounded by the electrolyte and the large 
volume of the box rendered it easy to prevent any variation 
of the resistance of the cell due to change of temperature 
during the short time necessary to make a measurement. The 
electrodes and the four plates described above, were of zinc, 
carefully amalgamated. The resistance of the cell was determined 
by using WHEATSTONE’s bridge, in the same way as when the 
resistance of a metallic conductor is measured. In the galvano- 
meter branch of the bridge, I used one of Sir Wm. THomson’s 
very sensitive galvanometers, whose resistance was so propor- 
tioned to the resistances in circuit as to give it the greatest 
possible sensitiveness. The solutions which I used were of zine 
sulphate, bought as pure and refined by erystallization. Their 
density was about that of minimum specific resistance. In order 
to prevent the condensation of dissolved gases on the surface of 
the electrodes and plates, the solutions were boiled before each 
set of experiments, and the electrodes and plates were kept for 
some time in a_ boiling solution .of approximately the same 
density as that in the cell, before being used. I found great 
difficulty in getting the electrodes and plates into such a state 
that they would neither originate a current themselves when 
dipped in the solution, nor become polarized when a current was 
sent through the circuit. In some cases, however, I succeeded, 
and was therefore able to make the necessary measurements. 
In these cases the observations were, of course, very simple. 
They consisted in the measurement of the resistance of the 
electrolytic cell, first without the plates, then with the plates, 
