38 ReEporRT OF DEPARTMENT OF BACTERIOLOGY OF THE 
_ bacteria and that only a very small proportion of those getting upon 
the seed would survive until the time of sowing. 
A more probable source of this natural infection is the germs 
from the other species of legumes. Reference has already been 
made to investigations which indicate that the bacteria associated 
with sweet clover and with burr clover readily associate with alfalfa. 
The number of species of wild legumes which grow naturally and with 
an abundant formation of nodules in our fields is large and it would 
be strange if the two forms already known were the only ones from 
which the germs pass readily to the alfalfa. It has been found in 
a number of investigations that when legumes are grown in steri- 
lized soil and are heavily inoculated with cultures derived from 
other species of legumes nodules are produced in many instances. 
NEED OF ARTIFICIAL INOCULATION. 
When we consider that only. 15 of the 67 check plats produced suf- 
ficient hay to be considered a profitable crop although they were prac- 
tically all at least slightly inoculated it is readily seen that something 
more than the presence of a scattering inoculation is necessary in 
order to make alfalfa growing profitable. That many plats were a 
failure simply because of a. lack of sufficient inoculation will be 
seen from the results obtained upon adjoining plats where arti- 
ficial inoculation was supplied. 
The question which is important agriculturally is whether there 
is sufficient inoculation in order to insure a successful crop, so far 
as it is affected by that factor. The experimental fields presented 
all gradations between no inoculation and a very abundant one. In 
some fields it was easy to see that there was not enough inocula- 
tion and in others that there was sufficient, but in many cases one 
could not decide from an observation of the roots as to the real situa- 
tion. There seemed to be no better way than to wait until the 
second season, when the plants should be fully established, and let 
the verdict turn on whether the crop produced was sufficient to be 
considered a profitable one. On this account we are reporting at 
this time only upon the experiments begun in 1905 and 1906. The 
results of the experiments begun in 1907 will probably be given in 
a bulletin of rgo9. 
While 67 fields are too few in comparison with the number of 
fields in the State to form a satisfactory basis for generalization, 
so far as they go the results from them indicate that less than 
one-quarter of the fields are in condition to produce alfalfa suc- 
