before the coil of the galvanometer has had time to take up 
the position of equilibrium due to it, i.e., in less than fths 
of a second. 
To show that gas is occluded by the electrodes during 
polarising, and soaks out during depolarising, the cell was 
polarised up to a fixed point, and then allowed to depolarise 
down to a second fixed point. On breaking circuit for 
a period, and then making again, a defiection was observed 
which was greater than the one before breaking circuit, and 
which increased with the time of insulation. 
The analogy between an electrolytic cell and a condenser, 
seems extended by the similarity of the depolarisation cur- 
rent to the oscillatory discharge of a condenser. 
The experiments are being continued with the object of 
finding the relation between the actions of the occluded and 
superficial gas during depolarisation, as referred to by Helm- 
holtz.^ 
“ On the delicacy of spectroscopic reaction in gases,” by 
T. W. Best. Communicated by Dr. Arthur Schuster, F.R.S. 
Seeing that very minute quantities of a great many bodies 
can be detected with ease in the spectroscope; at the 
request of Sir Henry Boscoe the following experiments 
were undertaken with the view of testing the efficiency of 
the spectroscope as a means for ascertaining the purity of 
gases. 
The three gases, hydrogen, nitrogen, and oxygen only 
were experimented upon ; thus the smallest amount of 
hydrogen, whose spectrum could be observed in that of 
nitrogen, and vice versa, was first determined ; and then the 
smallest amount of nitrogen in oxygen, and oxygen in 
nitrogen respectively, at the atmospheric pressure. 
Apparatus used. 
1. Spectroscope. 2. Induction coil. 3. Eudiometer. 
* See Wissenschaftliche Abhandlungen, of Helmholtz, p. 823. 
