convert it back into fluid water; but I have lowered the temperature of
the gas which I have collected in this jar, by passing it through water
after it has traversed the iron barrel, and still it does not change back
into water. I will take another test and apply to this gas. (I hold the
jar in an inverted position, or my substance would escape.) If I now apply
a light to the mouth of the jar, it ignites with a slight noise. That
tells you that it is not steam. Steam puts out a fire--it does not burn;
but you saw that what I had in that jar burnt. We may obtain this
substance equally from water produced from the candle-flame as from any
other source. When it is obtained by the action of the iron upon the
aqueous vapour, it leaves the iron in a state very similar to that in
which these filings were after they were burnt. It makes the iron heavier
than it was before. So long as the iron remains in the tube and is heated,
and is cooled again without the access of air or water, it does not change
in its weight; but after having had this current of steam passed over it,
it then comes out heavier that it was before, having taken something out
of the steam, and having allowed something else to pass forth, which we
see here. And now, as we have another jar full, I will shew you something
most interesting. It is a combustible gas; and I might at once take this
jar and set fire to the contents, and shew you that it is combustible; but
I intend to shew you more if I can. It is also a very light substance.
Steam will condense: this body will rise in the air, and not condense.

[Illustration: Fig. 15]

Suppose I take another glass jar, empty of all but air: if I examine it
with a taper, I shall find that it contains nothing but air. I will now
take this jar full of the gas that I am speaking of, and deal with it as
though it were a light body. I will hold both upside-down, and turn the
one up under the other; and that which did contain the gas procured from