ake sufficient care to pull out all the old and rotten
leaves. The plants, however, had grown, and looked so well upon the
whole, that I had no doubt but that the air must constantly have been in
a mending state; when I was exceedingly surprized to find, on the 24th
of that month, that though the air in one of the jars had not grown
worse, it was no better; and that the air in the other jar was so much
worse than it had been, that a mouse would have died in it in a few
seconds. It also made no effervescence with nitrous air, as it had done
before.

Suspecting that the same plant might be capable of restoring putrid air
to a certain degree only, or that plants might have a contrary tendency
in some stages of their growth, I withdrew the old plant, and put a
fresh one in its place; and found that, after seven days, the air was
restored to its former wholesome state. This fact I consider as a very
remarkable one, and well deserving of a farther investigation, as it may
throw more light upon the principles of vegetation. It is not, however,
a single fact; for I had several instances of the same kind in the
preceding year; but it seemed so very extraordinary, that air should
grow worse by the continuance of the same treatment by which it had
grown better, that, whenever I observed it, I concluded that I had not
taken sufficient care to satisfy myself of its previous restoration.

That plants are capable of perfectly restoring air injured by
respiration, may, I think, be inferred with certainty from the perfect
restoration, by this means, of air which had passed through my lungs, so
that a candle would burn in it again, though it had extinguished flame
before, and apart of the same original quantity of air still continued
to do so. Of this one instance occurred in the year 1771, a sprig of
mint having grown in a jar of this kind of air, from the 25th of July to
the 17th of August following; and another trial I made, with the same
success, the 7th of July 1772, the plant having grown in it from the
29th of June preceding. In this case also I found that the effect was
not owing to any virtue in the leaves of mint; for I kept them
constantly changed in a quantity of this kind of air, for a considerable
time, without making any sensible alteration in it.

These proofs of a partial restoration of air by plants in a state of
vegetation, though in a confined and unnatural situation, cannot but
render it highly probable, that the i