a mixture of about
four times as much fixed air, that afterwards mice lived in it
exceedingly well, and in some cases almost as long as in common air. I
found it, indeed, to be more difficult to restore _old_ putrid air by
this means; but I hardly ever failed to do it, when the two kinds of air
had stood a long time together; by which I mean about a fortnight or
three weeks.

The reason why I do not absolutely conclude that the restoration of air
in these cases was the effect of fixed air, is that, when I made a trial
of the mixture, I sometimes agitated the two kinds of air pretty
strongly together, in a trough of water, or at least passed it several
times through water, from one jar to another, that the superfluous fixed
air might be absorbed, not suspecting at that time that the agitation
could have any other effect. But having since found that very violent,
and especially long-continued agitation in water, without any mixture of
fixed air, never failed to render any kind of noxious air in some
measure fit for respiration (and in one particular instance the mere
transferring of the air from one vessel to another through the water,
though for a much longer time than I ever used for the mixtures of air,
was of considerable use for the same purpose) I began to entertain some
doubt of the efficacy of fixed air in this case. In some cases also the
mixture of fixed air had by no means so much effect on the putrid air
as, from the generality of my observations, I should have expected.

I was always aware, indeed, that it might be said, that, the residuum of
fixed air not being very noxious, such an addition must contribute to
mend the putrid air; but, in order to obviate this objection, I once
mixed the residuum of as much fixed air as I had found, by a variety of
trials, to be sufficient to restore a given quantity of putrid air, with
an equal quantity of that air, without making any sensible melioration
of it.

Upon the whole, I am inclined to think that this process could hardly
have succeeded so well as it did with me, and in so great a number of
trials, unless fixed air have some tendency to correct air tainted with
respiration or putrefaction; and it is perfectly agreeable to the
analogy of Dr. Macbride's discoveries, and may naturally be expected
from them, that it should have such an effect.

By a mixture of fixed air I have made wholesome the residuum of air
generated by putrefaction only, from mice plunged in