autions by which it becomes an exact measure were of course not
then known.--_Dec. 1838._

P vi. _On the primary or secondary character of the bodies evolved at the
Electrodes._

742. Before the _volta-electrometer_ could be employed in determining, as a
_general law_, the constancy of electro-decomposition, it became necessary
to examine a distinction, already recognised among scientific men, relative
to the products of that action, namely, their primary or secondary
character; and, if possible, by some general rule or principle, to decide
when they were of the one or the other kind. It will appear hereafter that
great mistakes inspecting electro-chemical action and its consequences have
arisen from confounding these two classes of results together.

743. When a substance under decomposition yields at the electrodes those
bodies uncombined and unaltered which the electric current has separated,
then they may be considered as primary results, even though themselves
compounds. Thus the oxygen and hydrogen from water are primary results; and
so also are the acid and alkali (themselves compound bodies) evolved from
sulphate of soda. But when the substances separated by the current are
changed at the electrodes before their appearance, then they give rise to
secondary results, although in many cases the bodies evolved are
elementary.

744. These secondary results occur in two ways, being sometimes due to the
mutual action of the evolved substance and the matter of the electrode, and
sometimes to its action upon the substances contained in the body itself
under decomposition. Thus, when carbon is made the positive electrode in
dilute sulphuric acid, carbonic oxide and carbonic acid occasionally appear
there instead of oxygen; for the latter, acting upon the matter of the
electrode, produces these secondary results. Or if the positive electrode,
in a solution of nitrate or acetate of lead, be platina, then peroxide of
lead appears there, equally a secondary result with the former, but now
depending upon an action of the oxygen on a substance in the solution.
Again, when ammonia is decomposed by platina electrodes, nitrogen appears
at the _anode_[A]; but though an _elementary_ body, it is a _secondary_
result in this case, being derived from the chemical action of the oxygen
electrically evolved there, upon the ammonia in the surrounding solution
(554.). In the same manner when aqueous solutions of metallic salts are
de