THE CIECULATION OF NITROGEN IN NATURE. 63 
gained by the plant — the conclusion is unavoidable that the 
latter has increased its nitrogen contents at the expense of 
the air. 
This enormously difficult problem was solved by the 
investigators I have mentioned in the most admirably com- 
plete manner by means of the apparatus of which I exhibit 
photographs. 
The plants were grown in an artificial atmosphere, in 
which the proportions of nitrogen, oxygen, and carbon 
dioxide were maintained within normal limits by the 
frequent withdrawal and anal3^sis of small samples. The 
quantity of nitrogen in the soil, the seed, the water, and the 
air on the one hand, and in the full-grown plant and the 
residual atmosphere on the other, were determined with 
scrupulous exactness, check experiments showing a loss or 
gain of only 2-3 c.c. nitrogen. The leguminous plants in 
every case showed an abundant crop of tubercles. 
The results of the earlier experiments pointed to the 
following conclusions : — 
(1) A large fixation of nitrogen by the leguminous plants. 
(2) A small but distinct fixation in some other cases. 
It was noticed that in the latter cases there was a growth 
of lowly vegetable forms on the surface of the soil ; a further 
series of experiments therefore was undertaken, in which this 
growth was prevented. The subjoined table of results shows 
that no fixation of nitrogen takes place with other than 
leguminous plants, but that these do remove a very appre- 
ciable quantity from the air. It was therefore concluded 
that the higher plants of themselves do not fix atmospheric 
nitrogen, that some lower forms do, and that in the case of 
the leguminosee a very extensive assimilation takes place 
under the influence of the rhizobes. 
In the light of these and similar investigations, I think 
