Voltaic Electricity on Alcohol, &c. 349 
very conclusive evidence before it could be adopted as a univer- 
sal law. 
Independently of the consideration, that, according to almost 
all the continental chemists, the principal haloid salts are not 
composed of a like number of atoms, there is at least one sub- 
stance, the oxide of antimony, which Mr Farapay found to be 
decomposed, in its fused state, by voltaic action, and which has 
been hitherto held, by the universal opinion both of British and 
continental chemists, to be a sesqui-oxide.* The oxide of bismuth, 
it is also extremely likely, is in a similar situation, according to 
the views on the Continent, as to its atomic constitution, which 
the specific heat of the metal renders very probable. 
Neither does it necessarily follow, because a substance is not 
decomposed in solution, that it will not give way in the dry and 
fused state ; because, if water is in its own nature more suscepti- 
ble of voltaic agency than the dissolved body, the action may be 
limited to the solvent when the solution is acted on. Assuming 
potash to suffer direct decomposition in its fused state, we have 
no reason to suppose that it gives way in solution. Our views as 
to the oxyacids will therefore be taken, from experiments made 
on them in the anhydrous and fused state, which, however, it is 
not always very easy to accomplish. 
Arsenic acid I could not succeed in freeing from water with- 
out decomposing it. Of anhydrous sulphuric acid I possessed too 
little to attempt any experiments with it. 
I made some experiments on dry and fused iodic acid, which, 
from the slight affinity of its constituents, I thought likely to 
throw light on the subject. A difficulty, however, presented it- 
* The attempts to show that it contains traces of a protoxide are evidently too 
imperfect to be entitled to regard, in their present state. 
+ The iodic acid was prepared in the usual way by the agency of nitric acid, 
and was several times alternately dissolved in water and evaporated to dryness, to 
free it from nitric acid. 
yy2 
