544 Journal of Agricultural Research vol. xx.No. 7 



Bacillus radiobacter was found to be peritrichic, and the same paper 

 also indicated (12, p. 592, footnote) 1 that in all probability B. radicicola 

 has the same kind of flagellation. But no faultless preparates were 

 obtained at that time. 



In the same year, 1905, G. T. Moore wrote concerning the nodule 

 bacteria (14, p. 26): 



There does not seem to be any necessity for creating a new group to include these 

 organisms, as has been done by Frank, under the name of Rhizobium, for although 

 there is a certain amount of polymorphism, it is no greater than frequently occurs in 

 the bacteria. 



With regard to the flagellation, however, Moore himself evidently 

 made no special studies, and, accepting Beijerinck's statement that the 

 "swarming bodies " (gonidia) of Bacillus radicicola are monotrichic as being 

 valid for the bacteria too, he proposed to call the nodule bacteria Pseu- 

 domonas radicicola. Numerous authors have followed this suggestion, 

 and experiments made by Harrison and Barlow (8) apparently confirmed 

 the view that the flagellation of these organisms is indeed monotrichic. 



However, these experiments are, in fact, not convincing, as has been 

 emphasized especially by Kellerman (o). This author and also G. de 

 Rossi (16,17), Zipfel (ig), and Prucha (75) secured results all of which 

 demonstrated more or less clearly that the senior author's assumption 

 was correct : Bacillus radicicola is peritrichic ; it is no " Pseudomonas." 



But this seemed again to be contradicted by certain results obtained 

 by the junior author while working with the late T. J. Burrill (6). Nu- 

 merous tests made with the bacteria isolated from cowpea, soybean, 

 Japan clover, and other plants showed clearly and invariably monotrichic 

 flagellation, and, therefore, the designation Pseudomonas radicicola was 

 restored once more. Additional results, however, indicated that there 

 are other features which differentiate the bacteria of the cowpea-soybean 

 group from those living in the roots of clover, alfalfa, pea, and vetch. 

 Especially the slime production and the speed of growth appeared to be 

 different, and the organisms studied were arranged into two groups, 

 "slow growers" and "fast growers" Both, however, were supposed to 

 be merely varieties of P. radicicola. 



This point remained to be investigated more thoroughly. In addition, 

 another "fast grower" presented itself for detailed study, which quite 

 regularly appeared on thickly sown plates of the "slow growing" groups, 

 and which, indeed, has been mistaken by several investigators as the 

 true nodule organism of cowpea, soybean, Japan clover, etc. Repeatedly 

 such cultures were sent to and tested by the junior author. They were 

 all unable to produce nodules. 



The data given on the following pages make it evident that this "fast 

 grower" is Bacillus radiobacter, which plays in this case, also, a very 



1 Reference is made by number (italic) to " Literature cited," p. 554-555. 



