paratively weak affinity of water for it,
and in the other to the stronger one of soda. In the latter case the
quantity transferred is from two and a half to three times what it is in
the former; and it appears therefore very evident that the transfer is
greatly dependent upon the mutual action of the particles of the
decomposing bodies[A].

[A] See the note to (675.),--_Dec. 1838._

531. In some of the experiments the acid from the vessels _a_ and _b_ was
neutralized by ammonia, then evaporated to dryness, heated to redness, and
the residue examined for sulphates. In these cases more sulphate was always
obtained from _a_ than from _b_; showing that it had been impossible to
exclude saline bases (derived from the asbestus, the glass, or perhaps
impurities originally in the acid,) and that they had helped in
transferring the acid into _b_. But the quantity was small, and the acid
was principally transferred by relation to the water present.

532. I endeavoured to arrange certain experiments by which saline solutions
should be decomposed against surfaces of water; and at first worked with
the electric machine upon a piece of bibulous paper, or asbestus moistened
in the solution, and in contact at its two extremities with pointed pieces
of paper moistened in pure water, which served to carry the electric
current to and from the solution in the middle piece. But I found numerous
interfering difficulties. Thus, the water and solutions in the pieces of
paper could not be prevented from mingling at the point where they touched.
Again, sufficient acid could be derived from the paper connected with the
discharging train, or it may be even from the air itself, under the
influence of electric action, to neutralize the alkali developed at the
positive extremity of the decomposing solution, and so not merely prevent
its appearance, but actually transfer it on to the metal termination: and,
in fact, when the paper points were not allowed to touch there, and the
machine was worked until alkali was evolved at the delivering or positive
end of the turmeric paper, containing the sulphate of soda solution, it was
merely necessary to place the opposite receiving point of the paper
connected with the discharging train, which had been moistened by distilled
water, upon the brown turmeric point and press them together, when the
alkaline effect immediately disappeared.

533. The experiment with sulphate of magnesia already described (495.)