Jan. 3 , 1921 Nodule Bacteria of Leguminous Plants 551 



solution, mannite-nitrate solution as used for studying the life cycle of 

 Azotobacter, tap water plus 0.5 per cent beef broth, and 2 per cent salt 

 agar furnished the most satisfactory results, especially with regard to a 

 more complete knowledge of the cell morphology of the organisms. For 

 diagnostic purposes, however, these substrates are of minor importance, 

 as they do not bring out anything which is not already to be seen on 

 the standard media. Nevertheless, it should be pointed out that cultures 

 of the nodule bacteria in soil are to be recommended for two reasons. 

 First, they are useful in keeping the organisms in a normal state of 

 virility for a long time, and, in the second place, they demonstrate very 

 clearly, when studied microscopically, that it is erroneous to believe — 

 though numerous authors have promoted such hypotheses — that the 

 nodule bacteria behave very differently in soil and could, therefore, not 

 be isolated in their typical form from their natural habitat. Our results 

 are in complete agreement with those recently obtained by Barthel (j) 

 concerning the growth of bacteria in sterilized soil. 



Tap water containing 0.5 per cent beef broth gave also very good 

 development and proved repeatedly helpful in reviving old, weakened 

 strains which refused to grow on solid substrates. 



DISCUSSION 



Our experimental results leave no doubt that the nodule bacteria of 

 the leguminous plants are to be divided at least into two distinct groups, 

 differing morphologically as well as culturally. It is equally beyond 

 dispute that these differences are so marked and constant that one might 

 be inclined to establish the nodule organism of the cowpea-soybean group 

 as a new species. On account of its behavior in the inoculation test O. 

 Kirchner has considered the soybean organism a distinct species, which 

 he named in 1895 Rhizob deter ium japonicum(io). According to the 

 rules of priority, this species name would have to be given preference, 

 despite the fact that different behavior in the inoculation test generally 

 can not be accepted as a distinguishing mark of such quality as to vali- 

 date the creation of a new species. The generic name Rhizobacterium, 

 used by Kirchner, is, of course, equally as untenable as the generic name 

 Rhizobium. According to the two tmost requently used modes of clas- 

 sifying the bacteria, one might name the cephalotrichic non-sporulating 

 rod of the cowpea-soybean group Pseudomonas japonica or Bacterium 

 japonicum, while the name Bacterium or Bacillus radicicola would have 

 to be retained for the peritrichic organisms to be found with clover, 

 alfalfa, sweet clover, vetch, pea, etc. 



However, we do not advocate such a procedure. We are firmly of 

 the opinion that much more must be known of the complete life history 

 of a bacterium than is obtainable along the standardized lines of cus- 

 tomary bacteriological research, before one can safely proceed to estab- 

 lish a genuine species on a truly scientific basis. Undoubtedly many if 



