1893.] Root Tubereles of Leguminose. 789 
and allows the nucleoplasm to escape and mix more or less 
with the cell protoplasm (cytoplasm). The nucleoli retain 
their normal size and form; some of them retain a position 
next to the nuclear wall, others escape into the “ mycoplasm ” 
of Frank. I have not noticed such extensive nuclear changes 
in any tubercles except those of Phaseolus vulgaris. 
The question as to what becomes of the contents of the infec- 
ted cells might be briefly considered. Formerly I expressed it 
as an opinion that the plant at the close of the vegetative 
period and also at other suitable times absorbed the protoplas- 
mic contents of the Rhizobia. I do not now think that that is 
the rule if it occurs at all. I believe that normally the tuber- 
cles are destroyed by the ordinary process of decay and the 
Rhizobia thus liberated. I have found the usual number of 
tubercles on Trifolium pratense and Melilotus alba as late as 
December 10th when I was compelled to chop them out of the 
frozen ground. The tubercles were filled with Rhizobium muta- 
bile, apparently in normal condition. I have found well filled 
tubercles on dead, matured plants of Phaseolus vulgaris. The 
Rhizobia were apparently in a resting stage, waiting to be lib- 
erated. It is true in some cases the tubercles were emptied 
during or before the close of the vegetative period. Even in 
such cases I do not believe that the contents were taken up by 
the host but rather that the tubercles died and decayed because 
it did not receive enough nourishment from the plant and soil. 
The question as to what position the tubercles hold, morpho- 
logically considered, is of some interest. The absence of a root- 
cap, of root-hairs and the fact that it develops exogenously 
would show that it does not resemble a root-branch. In posi- 
tion it shows neither positive nor negative geotropism though 
it occurs most frequently on roots near the surface. That may 
be, as Frank maintains, because the tubercles require a coat- 
ing of air to keep them from direct contact with moisture. But 
since the tubercle is nourished by the plant and the coating of 
air escapes from the tubercle itself it is difficult to see why 
they should not develop in deep soil as well. It may be that 
there is a tendency toward negative geotropism. I have never 
been able to find tubercles on any other part of the plant than 
