e top of the vessel to the bottom. At another
time, upon presenting a lighted candle to the mouth of the same vessel,
filled with the same kind of air, the yellowish flame ascended two
inches higher than the flame of the candle. The electric spark taken in
alkaline air is red, as it is in common inflammable air.

Though alkaline air be inflammable, it appeared, by the following
experiment, to be heavier than the common inflammable air, as well as to
contract no union with it. Into a vessel containing a quantity of
inflammable air, I put half as much alkaline air, and then about the
same quantity of acid air. These immediately formed a white cloud, but
it did not rise within the space that was occupied by the inflammable
air; so that this latter had kept its place above the alkaline air, and
had not mixed with it.

That alkaline air is lighter than acid air is evident from the
appearances that attend the mixture, which are indeed very beautiful.
When acid air is introduced into a vessel containing alkaline air, the
white cloud which they form appears at the bottom only, and ascends
gradually. But when the alkaline air is put to the acid, the whole
becomes immediately cloudy, quite to the top of the vessel.

In the last place, I shall observe that alkaline air, as well as acid,
dissolves _ice_ as fast as a hot fire can do it. This was tried when
both the kinds of air, and every instrument made use of in the
experiment, had been exposed to a pretty intense frost several hours. In
both cases, also, the water into which the ice was melted dissolved more
ice, to a considerable quantity.

SECTION II.

_Of COMMON AIR diminished and made noxious by various processes._

It will have been observed that, in the first publication of my papers,
I confined myself chiefly to the narration of the new _facts_ which I
had discovered, barely mentioning any _hypotheses_ that occurred to me,
and never seeming to lay much stress upon them. The reason why I was so
much upon my guard in this respect was, left, in consequence of
attaching myself to any hypothesis too soon, the success of my future
inquiries might be obstructed. But subsequent experiments having thrown
great light upon the preceding ones and having confirmed the few
conjectures I then advanced, I may now venture to speak of my hypotheses
with a little less diffidence. Still, however, I shall be ready to
relinquish any notions I may now entertain, if new facts should