ace of acid or alkali from
the sulphate of soda appeared.

977. From these experiments it may, I think, be concluded, that a solution
of sulphate of soda can conduct a current of electricity, which is unable
to decompose the neutral salt present; that this salt in the state of
solution, like water, requires a certain electrolytic intensity for its
decomposition; and that the necessary intensity is much higher for this
substance than for the iodide of potassium in a similar state of solution.

978. I then experimented on bodies rendered decomposable by fusion, and
first on _chloride of lead_. The current was excited by dilute sulphuric
acid without any nitric acid between zinc and platina plates, fig. 87, and
was then made to traverse a little chloride of lead fused upon glass at
_a_, a paper moistened in solution of iodide of potassium at _b_, and a
galvanometer at _g_. The metallic terminations at _a_ and _b_ were of
platina. Being thus arranged, the decomposition at _b_ and the deflection
at _g_ showed that an electric current was passing, but there was no
appearance of decomposition at _a_, not even after a _metallic_
communication at _b_ was established. The experiment was repeated several
times, and I am led to conclude that in this case the current has not
intensity sufficient to cause the decomposition of the chloride of lead;
and further, that, like water (974.), fused chloride of lead can conduct an
electric current having an intensity below that required to effect
decomposition.

979. _Chloride of silver_ was then placed at _a_, fig. 87, instead of
chloride of lead. There was a very ready decomposition of the solution of
iodide of potassium at _b_, and when metallic contact was made there, very
considerable deflection of the galvanometer needle at _g_. Platina also
appeared to be dissolved at the anode of the fused chloride at _a_, and
there was every appearance of a decomposition having been effected there.

980. A further proof of decomposition was obtained in the following manner.
The platina wires in the fused chloride at _a_ were brought very near
together (metallic contact having been established at _b_), and left so;
the deflection at the galvanometer indicated the passage of a current,
feeble in its force, but constant. After a minute or two, however, the
needle would suddenly be violently affected, and indicate a current as
strong as if metallic contact had taken place at _a_. This I actually found