iometer to the
Royal Society and detailed observations he had made to determine whether
or not the atmosphere is constant in composition; after testing the air
on nearly 60 different days in 1781 he could find in the proportion of
oxygen no difference of which he could be sure, nor could he detect any
sensible variation at different places. Two papers on "Experiments with
Airs," printed in the _Phil. Trans._ for 1784 and 1785, contain his
great discoveries of the compound nature of water and the composition of
nitric acid. Starting from an experiment, narrated by Priestley, in
which John Warltire fired a mixture of common air and hydrogen by
electricity, with the result that there was a diminution of volume and a
deposition of moisture, Cavendish burnt about two parts of hydrogen
with five of common air, and noticed that almost all the hydrogen and
about one-fifth of the common air lost their elasticity and were
condensed into a dew which lined the inside of the vessel employed. This
dew he judged to be pure water. In another experiment he fired, by the
electric spark, a mixture of hydrogen and oxygen (dephlogisticated air),
and found that the resulting water contained nitric acid, which he
argued must be due to the nitrogen present as an impurity in the oxygen
("phlogisticated air with which it [the dephlogisticated air] is
debased"). In the 1785 paper he proved the correctness of this
supposition by showing that when electric sparks are passed through
common air there is a shrinkage of volume owing to the nitrogen uniting
with the oxygen to form nitric acid. Further, remarking that little was
known of the phlogisticated part of our atmosphere, and thinking it
might fairly be doubted "whether there are not in reality many different
substances confounded together by us under the name of phlogisticated
air," he made an experiment to determine whether the whole of a given
portion of nitrogen (phlogisticated air) of the atmosphere could be
reduced to nitric acid. He found that a small fraction, not more than
1/120th part, resisted the change, and in this residue he doubtless had
a sample of the inert gas argon which was only recognized as a distinct
entity more than a hundred years later. His last chemical paper,
published in 1788, on the "Conversion of a mixture of dephlogisticated
and phlogisticated air into nitrous acid by the electric spark,"
describes measures he took to authenticate the truth of the experiment
describ