102 
MR. GROVE ON THE GAS VOLTAIC BATTERY. 
the action. In this experiment the rise was more uniform in the different tubes than 
in the last, and the action more decided. The results, although on a small scale, 
appear more definite ; thus we get the proportion as 1 : 2*3 ; and as the combining 
volumes of oxygen and carbonic oxide are as one to two, if we add the local action 
due to the oxygen of the air in solution, 1 to 2*3 is as near an approximation as can 
be expected. Though much superior to olefiant gas, the action of carbonic oxide is, 
however, very feeble when compared with that of hydrogen. 
Experiment 13. — Oxygen and chlorine. Very considerable action on the iodide 
at first, but not constant ; it abated within the first hour, and after twenty-four 
hours the action was extremely feeble, scarcely perceptible ; the water had risen 
nearly to the top of the chlorine tubes, but the level in the oxygen tubes was un- 
altered. The chlorine was negative to oxygen, or in other words, the oxygen was in 
its voltaic bearing to chlorine as hydrogen to oxygen. 
As in this experiment the water level in the oxygen tubes was unaltered, it ap- 
peared that this gas had little to do with the action, I therefore, 
Experiment 14, — Charged the alternate tubes of a battery with chlorine and dilute 
sulphuric acid; the amount of action was much the same as in experiment 13, and 
equally transitory ; a few gaseous bubbles were perceptible on the platinums in the 
oxygen cells, but not in sufficient quantity for examination. It is well known that 
chlorine of itself will slightly decompose water, forming hydrochloric acid, and evol- 
ving oxygen, and there is little doubt that the voltaic action here observed was due to 
this. There was no appearance of the platinum having been attacked in several expe- 
riments which I made with chlorine. So slight a chemical action will, however, give 
rise to voltaic effects, that the absence of any apparent corrosion is not conclusive. 
It is stated by chemists that gaseous chlorine will not attack platinum, but that it is 
only when nascent it combines with this metal ; non constat however, that in the 
gas battery the chlorine at the initiatory instant of its electro-synthesis may not be 
in a state analogous, as to its chemical energies, to that converse state called nascent, 
and therefore we cannot venture to negative the possibility of the platinum being 
slightly attacked. This circumstance, added to its extreme solubility and power of 
decomposing water, makes chlorine rather an unsatisfactory element for the class of 
actions developed by the gas battery. 
Solutions of bromine, chlorine and iodine, have been before experimented on (I be- 
lieve by Dr. Schcenbein and M. Becquerel) as to their voltaic relations, but in 
examining the voltaic relations of bodies in a gaseous state, or to express myself with 
more caution, in a state passing from gaseous to liquid, I tried, 
Experiment 15, — One set of tubes charged with gaseous chlorine, and the alter- 
nate tubes with solutions of bromine and iodine. The chlorine was negative to both, 
i. e. was to these as oxygen to hydrogen. 
I now tried hydrogen with several gases, but as it was next to impossible (I found 
