took the sparks,
the redness extended twice as far on the lower side as on the upper.

The most important, though the least expected observation, however, was
that, in proportion as the liquor became red, it advanced nearer to the
wire, so that the space of air in which the sparks were taken was
diminished; and at length I found that the diminution was about one
fifth of the whole space; after which more electrifying produced no
sensible effect.

To determine whether the cause of the change of colour was in the _air_,
or in the _electric matter_, I expanded the air which had been
diminished in the tube by means of an air-pump, till it expelled all the
liquor, and admitted fresh blue liquor into its place; but after that,
electricity produced no sensible effect, either on the air, or on the
liquor; so that it was evident that the electric matter had decomposed
the air, and had made it deposite something that was of an acid nature.

In order to determine whether the _wire_ had contributed any thing to
this effect, I used wires of different metals, iron, copper, brass, and
silver; but the result was the very same with them all.

It was also the same when, by means of a bent glass tube, I made the
electric spark without any wire at all, in the following manner. Each
leg of the tube, fig. 19. stood in a bason of quicksilver; which, by
means of an air-pump, was made to ascend as high as _a, a_, in each leg,
while the space between _a_ and _b_ in each contained the blue liquor,
and the space between _b_ and _b_ contained common air. Things being
thus disposed, I made the electric spark perform the circuit from one
leg to the other, passing from the liquor in one leg of the tube to the
liquor in the other leg, through the space of air. The effect was, that
the liquor, in both the legs, became red, and the space of air between
them was contracted, as before.

Air thus diminished by electricity makes no effervescence with, and is
no farther diminished by a mixture of nitrous air; so that it must have
been in the highest degree noxious, exactly like air diminished by any
other process.

In order to determine what the _acid_ was, which was deposited by the
air, and which changed the colour of the blue liquor, I exposed a small
quantity of the liquor so changed to the common air, and found that it
recovered its blue colour, exactly as water, tinged with the same blue,
and impregnated with fixed air, will do. But the following