1913] GODDARD—SOIL FUNGI 299 
The loss of total nitrogen in the M/ro and M/s50 cultures has 
been referred to. This may have been due, as already suggested, 
to some chemical action which gave off some compound of nitrogen 
during the sterilization. The dissolving of the calcium carbonate 
and the slight change in the color of the culture solutions suggested 
this, but no positive evidence of such nitrogen loss was obtained. 
Another possible explanation might be that nitrogen in some form 
was given off as a waste product during the metabolism of the 
growing fungus. However, no evidence of such a hypothesis was 
observed. This might raise another interesting question for further 
investigation. : 
_ DISCUSSION OF RESULTS IN RELATION TO NITROGEN-FIXATION 
As a result of this series of investigations, it may be said that 
we find no evidence whatever which seems to justify a belief in the 
Power of any of the forms investigated to fix free nitrogen under 
the conditions used. 
The question of why such conflicting results regarding nitrogen- 
fixation by fungi have been reported by different investigators stil] 
remains. This investigation was not carried on to settle this ques- 
tion, but rather because it was hoped that by investigating a large 
number of forms, and especially those in the soil, some fungi might 
be found which certainly had. the nitrogen-fixing power. There 
Seems no reason a priori why fungi, as well as bacteria, should not 
have such power, and certainly, reasoning from analogy, there seems 
much reason to expect it. However, we have been wholly disap- 
pointed in this expectation as a result of this series of investigations. 
Furthermore, while it was not designed to work over forms pre- 
viously studied, nevertheless, negative results are given here for 
one form, Hormodendron cladosporioides, for which positive results 
have been reported by others. Both Froeticu (11) and Purie- 
WITSCH (16) found appreciable nitrogen-fixation for this fungus 
on both nitrogen-free and nitrogen-poor media. We have been 
entirely unable to confirm these results. 
It seems difficult to explain the widely conflicting results. It 
may be easy, perhaps, to discard entirely such results as those 
of LarHam (20), in which, as pointed out by PENNINGTON (25), 
