mechanically combined,--one of the things about which the learned are
not fully agreed,--it is found to be chemically the same in its
constituents, all over the world, whether collected on mountains or on
plains, on the sea or on the land, whether obtained by aëronauts miles
above the earth or by miners in their deepest excavations. On the
theory of its mechanical combination, however, as by volume, and that
each constituent acts freely for itself and according to its own laws,
important speculations (conclusions, indeed) have arisen, both as
regards temperature and climatic differences. It should be observed,
that volume, as we have used the word, is the apparent space occupied,
and differs from mass, which is the _effective_ space occupied, or the
real bulk of matter, while density is the relation of mass to volume,
or the quotient resulting from the division of the one by the other.
Those empty spaces which render the volume larger than the mass are
technically called its pores.

Has the composition of the atmosphere changed in the lapse of years?
On this point both French and German philosophers have largely
speculated. It is computed that it contains about two millions of
cubic geographical miles of oxygen, and that 12,500 cubic geographical
miles of carbonic acid have been breathed out into the air or
otherwise given out in the course of five thousand years. The
inference, then, should be, that the latter exists in the air in the
proportion of 1 to 160, whereas we find but 4 parts in 10,000. Dumas
and Bossingault decided that no change had taken place, verifying
their conclusion by experiments founded on observations for more than
thirty-five years. No _chemical_ combination of oxygen and nitrogen
has ever been detected in the atmosphere, and it is presumed none will
be.