1898-99.] Dr H. Marshall on Polarisation Phenomena. 533 
examining it chemically was inconvenient and unsatisfactory, a 
considerable number of experiments had been carried out for the 
purpose of obtaining an idea of the time necessary for complete 
deposition with varying current strengths and different quantities 
of metal. As the result of these, the above-mentioned limit of 4 
hours was fixed as sufficient, with the other conditions as stated, 
while not unduly excessive for the smaller quantities of metal ; the 
time necessary for complete deposition is nothing like proportional 
to the amount of M present originally in the solution. 
The ammeter readings taken at intervals in the course of an 
experiment kept very constant. There were slight differences 
about the beginning, probably due to the change of temperature 
caused by the current ; thereafter the readings remained constant 
for a considerable time, sometimes until the experiment was 
stopped, though generally there was a very slight falling off at the 
end, but the total variation seldom exceeded about 0'03 ampere. 
On tabulating and comparing a series of experiments in which the 
action of the current had been continued for varying lengths of 
time, peculiarities were unexpectedly observed in the voltmeter 
readings. Here, again, there was always a certain variation at the 
beginning of an experiment, then a period during which the 
difference of potential was practically constant, but, in some cases 
only, there was a considerable rise in the final reading, amounting 
occasionally to fully 0’5 volt ; in other cases there was no evidence 
of a distinct rise in the voltmeter readings. On further comparing 
the various experiments it was found that, as a rule, those belong- 
ing to the second class were the determinations in which deposition 
had not been completed, as evidenced by a somewhat low percent- 
age result and the presence of small quantities of nickel in the 
decanted liquid. (The latter was always tested by means of 
potassium thiocarbonate.) Those cases which showed a marked 
rise of potential corresponded to good final results. 
These facts seemed to indicate that the removal of the last small 
quantities of nickel from solution corresponded to a considerable 
increase in the resistance between the electrodes. The change 
could not be caused by diminished conductivity of the solution, 
due to the removal of nickel, because of the large quantity of 
ammonium sulphate present, and the small quantity of nickel 
