The Atmosphere as a source of Nitrogen to Plants. 259 
M. Boussin^ault had, as we have already mentioned, so early 
as the year 1837, convinced himself of the fact that nitrogen is 
obtained from the air by plants, but he had formed no decision 
as to the form in which this nitrogen was furnished to them. 
His experiments, undertaken in 1851 and continued to 1853, 
were intended to clear up this point. M. Boussingault had, as 
he himself states, the choice of two methods : either to grow the 
plants in an atmosphere free from ammoniacal vapours — the plan 
actually adopted by M, Ville ; or to place them in a confined 
atmosphere which was not renewed. For various reasons, to be 
mentioned presently, he selected the second mode. 
The seeds were sown in pumice-stone in coarse powder, pre- 
viously heated to redness. To the pumice-stone he added a 
certain quantity of the ashes of farmyard manure, and sometimes 
a portion of the ash of the same kind of seed upon which he was 
experimenting ; it is obvious that in this way the plant would 
have a full supply of all the mineral ingredients it could require. 
These plants were supplied with water free from ammonia, 
and grown in cases from which the external air was rigorously 
excluded. 
M. Boussingault operated upon haricot beans, oats, cress, and 
lupins ; 12 independent experiments being made. He found 
that plants thus growing in a confined portion of air did not 
accumulate nitrogen ; that they contained in fact no more of this 
element than was present in the seed from which they were pro- 
duced. Such being the conclusion to which he arrived, we are 
not surprised that his zeal in research should induce him to study 
more fully the sources and conditions of atmospheric ammonia, 
to which he was led to attribute the supply of nitrogen to plants. 
In the spring of 1853 M. Boussingault made a number of 
experiments to determine the quantity of ammonia in the water 
of the Seine and several other streams. He found the proportion 
very much smaller than from M. Barral's analyses of rain-water 
in Paris he had been led to expect. He reasons upon the suppo- 
sition that the water of streams should contain as much ammonia 
as the rain which gives rise to them. Independently of the fact 
that the experiments of M. Boussingault, to be referred to pre- 
sently, prove that the ammonia of rain falling in the country 
is much less than that of Paris, it should hardly have seemed to 
him a matter of surprise that such a discrepancy between river 
and rain-water should exist, if the absorbent properties of the 
soil for ammonia were taken into account. My own experiments 
have shown that soils remove the ammonia falling upon them, 
and the ammonia is therefore, in great part, arrested in its pro- 
gress towards the streams. 
M. Boussingault made at the same time determinations of the 
s 2 
