BY R. GREIG-SMITH. 265 



think, two exceptions, viz., Maze* and Moore, f the experiments 

 have been negative.! There is, therefore, reason to believe that 

 the direct fixation of nitrogen may not be the function of Nhizo- 

 bium. Since, however, nitrogen is fixed by the plant, the questions 

 arise, How is the fixation made ? and How does the microbe assist 

 in the process 1 I fear that the first question is for future 

 biochemists to determine, but in this paper I shall indicate a 

 probable answer to the second. 



In a paper that is published simultaneously with this I show 

 that Rhizubium leguminosarum, otherwise known as Bacillus 

 radicicola, is a coccoid bacterium contained within a rod-shaped 

 or branching capsule. The so-called bacteroitlal or irregular 

 forms are produced by vagaries in the growth of the capsule. 

 There is no reason to connect the fixation of nitrogen with the 

 formation of these branching forms. It would be well to bear 

 in mind that Vibrio or Mycobacterium denitrificans furnishes the 

 same branching forms,§ and instead of converting free into com- 

 bined nitrogen, it does the reverse and converts the combined 

 nitrogen of nitrate into free nitrogen gas. I shall show in the 

 eourse of the research that the gums of Rhizobium and of 

 Vibrio denitrificans are identical, and if, as there is reason to 

 believe, the bacterial capsule is the " non-diffuse " slime it will be 

 apparent that the branching is entirely due to the nature of the 

 condensed slime of which the capsule is composed. A knowledge 

 of the nature of the gum, the principal constituent of the slime, 

 is thus of some moment. 



But when we consider the matter, the importance of an investi- 

 gation of the slime becomes more apparent. In the nodule, we 

 can see the " infection threads " of Frank passing from cell to 

 cell, and it has been demonstrated that these threads consist of 



* Ann. de l'Inst. Past, xii.l, 128. 



t Bull. No.71, Bureau of Plant Industry, Washington, U.S.A. 



% In my former paper, I have shown that I could not detect a gain of 



nitrogen either in pure or in mixed culture. In the present research, it will 



be seen (p. 281) that the formation of slime is proportional to the nitrogen 



supplied. 



§ These Proceedings, 1901, 118. 



