the soil, and it is necessary that
we should here consider the mode in which they may be obtained from
each.

_The Atmosphere as a source of the Organic Constituents of
Plants._--Atmospheric air consists of a mixture of nitrogen and oxygen
gases, watery vapour, carbonic acid, ammonia, and nitric acid. The two
first are the largest constituents, and the others, though equally
essential, are present in small, and some of them in extremely minute
quantity. When deprived of moisture and its minor constituents, 100
volumes of air are found to contain 21 of oxygen and 79 of nitrogen.
Although these gases are not chemically combined in the air, but only
mechanically mixed, their proportion is exceedingly uniform, for
analyses completely corresponding with these numbers have been made by
Humboldt, Gay-Lussac, and Dumas at Paris, by Saussure at Geneva, and by
Lewy at Copenhagen; and similar results have also been obtained from air
collected by Gay-Lussac during his ascent in a balloon at the height of
21,430 feet, and by Humboldt on the mountain of Antisano in South
America at a height of 16,640 feet. In short, under all circumstances,
and in all places, the relation subsisting between the oxygen and
nitrogen is constant; and though, no doubt, many local circumstances
exist which may tend to modify their proportions, these are so slow and
partial in their operations, and so counterbalanced by others acting in
an opposite direction, as to retain a uniform proportion between the
main constituents of the atmosphere, and to prevent the undue
accumulation of one or other of them at any one point.

No such uniformity exists in the proportion of the minor constituents.
The variation in the quantity of watery vapour is a familiar fact, the
difference between a dry and moist atmosphere being known to the most
careless observer, and the proportions of the other constituents are
also liable to considerable variations.

_Carbonic Acid._--The proportion of carbonic acid in the air has been
investigated by Saussure. From his experiments, made at the village of
Chambeisy, near Geneva, it appears that the quantity is not constant,
but varies from 3.15 to 5.75 volumes in 10,000; the mean being 4.15.
These variations are dependent on different circumstances. It was found
that the carbonic acid was always more abundant during the night than
during the day--the mean quantity in the former case being 4.32, in the
latter 3.38. The largest quantit