62 



wlieu supplied with air deprived of ammonia aud nitric acid. Ho 

 repeated the experiments of Boussingault in confined air with the excep- 

 tion that he inocuUited his plants. This difference changed the result 

 from negative to positive, and showed tliat with the aid of the microbes, 

 legumes did obtain nitrogen from the air when ammonia and nitric acid 

 were absent. 



Admitting this conclusion, Prazmowski sees three possible methods 

 by which the organisms i)roduec tlieir ettect on the phint: (1) The bac- 

 teria may produce certain substances (ferments ?) by the aid of wliich 

 the i)lant is able to fix free nitrogen in a compound which the phint 

 incori)orates into its own substance; {2) the bacteria themselves may 

 seize upon the nitrogen and incorporate it in coini)ouuds iii their own 

 bodies and then the plant may malce use of the bacteria for food; (.'^) 

 the assimilation of nitrogen may be a function of the combined Hfe of 

 the phint and the microbe acting together, the necessary energy being 

 the result of their symbiosis. Of these three possibilities Prazmowski 

 inclines to the second, thinking that the bacteria, after fixing the atmos- 

 pheric nitrogen in their own bodies, under the intluenceof the i)lant tis- 

 sue degenerate into bacteroids and are finally absorbed by the jihint. 



It will be seen that there are several points of ditterence between 

 the views of Frank and Prazmowski. Tlie most important is the claim 

 made by Frank that the tuberck'S are of vahie to the plants only when 

 growing in poor soil, while tliose i)lants which grow in rich soil gain 

 an equal amount of nitrogen whether or not they are supplied with 

 tubercles. Other more recent exj)eriments do not confirm this conclu- 

 sion. Atwater and Woods* have found that the tubercles do have a 

 very important relation to the fixation of atmospheric nitrogen, even 

 when nitrogen is supplied to the roots in the form of nitrates. Wil- 

 farth t in a recent address criticises Frank's experiments as inexact antl 

 containing numerous sources of error. He thinks that the experiments 

 by which Frank undertook to prove that other plants beside legumes 

 can fix nitrogen, are unsatisfactory, the amount of nitrogen ac(]uired by 

 the plants being within the limits of error. Wilfarth also mentions 

 further experiments of his own which directly contradict the claim of 

 Frank, that the tubercles are of no value in rich soils, and promises 

 soon a more complete description of these investigations. 



Ureal J n)ade a study of the root tubercles — their composition, their 

 formation through the agency of bacteria, the methods of inoculation, 

 and the gain of nitrogen by legumes, peas, beans, and alfalfa in water, 

 artificial soil and otherwise. He observed that inoculation couhl be 

 readily brought about in the roots of legumes growing in water or m 

 soil by ])iercing a root and inserting a portion of another root or other 

 material containing bacteria ; that cultures of the bacteria could be 



• Am. Chem. Jour., vols, xii aud xiii. 



t Laiulw.Vers. Stat. , :<!^, p. :V22. 

 t Anu. Agrou., ir» (is^it), p. 529. 



