ut I have generally found it to be between one fifth and
one fourth of the whole.

Air thus diminished is not heavier, but rather lighter than common air;
and though lime-water does not become turbid when it is exposed to this
air, it is probably owing to the formation of a selenitic salt, as was
the case with the simple burning of brimstone above-mentioned. That
something proceeding from the brimstone strongly affects the water which
is confined in the same place with this mixture, is manifest from the
very strong smell that it has of the volatile spirit of vitriol.

I conclude that the diminution of air by this, process is of the same
kind with the diminution of it in the other cases, because when this
mixture is put into air which has been previously diminished, either by
the burning of candles, by respiration, or putrefaction, though it never
fails to diminish it something more, it is, however, no farther than
this process alone would have done it. If a fresh mixture be introduced
into a quantity of air which had been reduced by a former mixture, it
has little or no farther effect.

I once observed, that when a mixture of this kind was taken out of a
quantity of air in which a candle had before burned out, and in which it
had stood for several days, it was quite cold and black, as it always
becomes in a confined place; but it presently grew very hot, smoaked
copously, and smelled very offensively; and when it was cold, it was
brown, like the rust of iron.

I once put a mixture of this kind to a quantity of inflammable air, made
from iron, by which means it was diminished 1/9 or 1/10 in its bulk;
but, as far as I could judge, it was still as inflammable as ever.
Another quantity of inflammable air was also reduced in the same
proportion, by a mouse putrefying in it; but its inflammability was not
seemingly lessened.

Air diminished by this mixture of iron filings and brimstone, is
exceedingly noxious to animals, and I have not perceived that it grows
any better by keeping in water. The smell of it is very pungent and
offensive.

The quantity of this mixture which I made use of in the preceding
experiments, was from two to four ounce measures; but I did not
perceive, but that the diminution of the quantity of air (which was
generally about twenty ounce measures) was as great with the smallest,
as with the largest quantity. How small a quantity is necessary to
diminish a given quantity of air to a _maximum_, I h