IIo RECLAIMING LOST NITROGEN 
with the Connecticut soil, subsequently planting the soy bean. 
The result was an excellent growth of the soy bean and the develop- 
ment of tubercles. Afterward these particular plots of land were 
capable of producing large luxuriant crops of the soy bean, with 
abundant root tubercles and a large fixation of atmospheric nitro- 
gen. Evidently Connecticut soil does not contain the bacteria 
adapted for producing the tubercles in the soy bean, although those 
which produce tubercles on the pea and the clover are abundant 
enough. 
Similar experiments have been repeated elsewhere until it has 
become evident that the root tubercle bacteria are not all alike. 
Varieties adapted to one species of legume may be unable to pro- 
duce tubercles upon a second species; in some cases one type of 
bacteria may be able to grow in the roots of several allied legumes 
but not in others. For example, the tubercle organism of sweet 
clover will do well with alfalfa. All of these facts have suggested 
that there are different types of leguminous bacteria, each adapted 
to different species of legumes. 
To what extent this conclusion is true it is by no means easy to 
determine. It is certainly true that some varieties of legume will 
grow in soils with an abundant production of tubercles, while 
other varieties of closely related legumes are unable to produce 
an abundant crop of tubercles in the same soil. This is evidently 
due to the lack of microérganisms especially appropriate to the 
legume in question, since inoculation with proper soil infusion pro- 
duces tubercles at once. But just what this means is not so evi- 
dent. It certainly means that different legumes demand different 
varieties of tubercle bacteria. Whether these different varieties 
are distinct species is, of course, a fruitless question inasmuch as 
we do not know what we mean by a species among bacteria. But 
it is of importance to know whether these types are quite distinct 
or whether they are simply physiological varieties of the same 
general species. If the former is true we should expect them to 
remain distinct, but if the latter is true we might expect the soil 
