specific heats. If we test the matter by the data now afforded, it is
best to obey the injunction, "_In medio tutissimus ibis_." In the
following table, the first column are the values obtained by Regnault;
in the second, the former values; and in the third, the mean of the two.

Gases. Reg. specific heats. Former specific heats. Mean.
Atmospheric air, .237 .267 .252
Oxygen, .218 .236 .227
Nitrogen, .244 .275 .260
Hydrogen, 3.405 3.294 3.350

The specific gravities of these gases, according to the best tables in
our possession, are:

Specific gravities. Mean. Products.
Atmospheric air, 1.0000 × .252 = .252
Oxygen, 1.1111 × .227 = .252
Nitrogen, 0.9722 × .260 = .252
Hydrogen, 0.0745 × 3.350 = .249

As might be expected, there is a greater discrepancy in the case of
hydrogen.

If we test the principle by the vapor of water, we must consider that it
is composed of two volumes of hydrogen and one volume of oxygen, and
that one volume disappears; or that one-third of the whole atomic
motion is consumed by the interference of the vibrations of the ether,
necessary to unite the atoms, and form an atom of water. We must
therefore form this product from its specific gravity and two-thirds of
its specific heat. On no one subject in chemistry has there been so much