598 
Froceeclings of the Boyal Society 
liave it unusually strong. If one compares the intensity of the 
residual current with that which would have been produced in a 
metallic conductor of the same resistance (600 ohms), it was about 
0'025 of the latter with an electromotive force of 0-8 Daniell, 0*125 
with 1*0 Daniell, 0*4 with 2*0 Daniells. I shall not try to give 
you more exact numbers, because I hope to get them still more accu- 
rate than they are at present. But there was not the slightest trace 
of evolution of gas. And you see that even with two Daniells the 
current which was kept up required the force of 0*8 Daniells to 
overcome the resistance of the circuit. Therefore, only 1 *2 Daniells 
remained for the decomposition of the water, which are insufficient 
to develop the two gases under atmospheric pressure. By reducing 
the resistance of the circuit to 300 ohms I could get a decomposing 
force of 1*36 Daniells. But also this was not sufficient for visible 
decomposition. The arrangement of the apparatus used hitherto 
did not admit of going farther. Bor these experiments a very con- 
stant electromotive force is needed, which is steady through months, 
and of which well-measured parts can he derived to pass through 
the electrolyte, and I had not yet had the time to introduce those 
modifications of the apparatus which are necessary for the employ- 
ment of higher electromotive force. 
These experiments show that in this case a current of about 0*002 
ampere could pass contantly through acidulated water without 
developing any visible trace of oxygen and hydrogen. 
I don’t think, nevertheless, that the electrolytic law of Faraday 
is violated in this case. I suppose that really oxygen and hydrogen 
exist separately at the electrodes, only they don’t bubble off, hut 
remain dissolved in the fluid, where they exist electrically neu- 
tralised, being no longer subject to electric attraction, and therefore 
free to migrate through the fluid by diffusion. But when electri- 
cally neutral oxygen reaches the cathode, where positive electricity 
is subject to the attraction of the negative electricity of the metal, 
it will yield its + E far easier to the cathoda than does hydrogen. 
And the same will happen at the anode. Neutral hydrogen, carried 
over by diffusion, will yield its - E easier to the positive metal 
than the anion oxygen will do. This, as you see, produces only 
a convective current of electricity. At the cathode diffused 
