1894. ] Anatomical and Physiological Researches. 291 
ism or the plant itself or is a partnership act is by no means 
settled. 
The most important contributions to this phase of the ques- 
tion that have appeared in the past year are the joint papers 
of Nobbe and Hiltner.2? In their several papers, covering 
experiments since 1890, they show concordant results. They 
hold that the assimilation of nitrogen by legumes bears a di- 
rect relation to the formation of bacteroids. In numerous 
cases they found that plants (peas) growing in good soil and 
well supplied with nodules were unable to make much gain. 
When inoculated with pure cultures of B. radicicola, some 
plants would gain largely in nitrogen while others would ap- 
parently suffer from nitrogen hunger. When the tubercles 
on these plants were carefully examined they noted that the 
nodule-producing organisms were unchanged in those plants 
that hungered for nitrogen, while in the thriftier ones, the 
bacteria were changed into bacteroids. They conclude that 
(I) tubercles in which bacteroid formation does not occur 
are injurious instead of beneficial to the host plant, (the un- 
changed bacteria are then merely parasites); (2) the un-. 
changed bacteria present in tubercles seem to have no rela- 
tion to the nitrogen fixation by legumes; (3) the more 
In some experiments with Robinia they obtained striking 
results. The plants gained more nitrogen in the end when 
cultivated in nitrogen-free soil than in soil containing nitro- 
gen. This was because there was a more complete conver- 
Sion of the bacteria into bacteroids in non-nitrogenous soil 
than where nitrates were present. Manuring with nitrates 
Causes a more rapid development of the plants at first, and 
| with this a more rapid growth of smaller nodules, but these 
| were of less benefit than the larger nodules noted in nitrogen- 
| free soil, the bacteria of which. were entirely changed into 
bacteroids. 
The formation of bacteroids in the light of this view will 
ve then an increased interest. 
Nobbe and Hiltner claim that the bacteroids are formed by 
repeated division of the tubercle germ without the separation 
into isolated individuals. This continued division usually 
2 
* Sachs. landw. Zts. 76: 165. 1893. Landw. Vers. Stat. 42: 459. 1893. 
| 
ha 
