452 R. S. BREED, H. J. CONN AND J. C. BAKER 



the two groups. Jensen placed Azotohacier in one family and 

 Rhizohium in the other. In their report as originally presented 

 (Winslow, et al., 1917a), the committee on classification did the 

 same; but in their complete published report (1917b) they have 

 placed both Azotohacier and Rhizohium in the Nitrobacteriaceae. 



This change may have been unintentional; in which case it is a 

 very unfortunate mistake. DeRossi (1906), Zipfel (1911) and 

 Kellerman (1912) and others have shown conclusively that the 

 proper technic reveals peritrichous flagella on the legume 

 organism.^ The characterization of the genus given by the 

 committee on page 553 of their report is therefore incorrect. 

 That peritrichic organisms with granular structure and branch- 

 ing cells should be placed in the same family with the autotrophic 

 bacteria can be justified only if physiology is allowed to over- 

 ride morphology entirely in establishing the classification. The 

 committee distinctly state in regard to the Nitrobacteriaceae 

 (p. 551) "When motile, with polar, never peritrichous, flagella." 



The treatment of Rhizohium is important, because its relation- 

 ships probably determine the position of the Actinomyces line. 

 On the one hand its branching cells and granular structure show 

 a striking resemblance to the tubercle organism and Actinomy- 

 cetes, while on the other hand the motility of its vegetative rods 

 helps to establish a connection with the true bacteria. Jensen 

 accepted the old description of Rhizohium as a monotrichic rod 

 and therefore placed the family of Actinomycetes in the Cephalo- 

 trichinae; but as the legume organism has now been definitely 



' After finishing this paper the writers have noticed that Burrill and Hansen 

 (111. Agr. Exp. Sta., Bui. 202. 1917) have apparently observed a single flagellum 

 on this organism. As stated, however, by Hansen (who wrote the bulletin), this 

 flagellum seems to be at a corner instead of at the pole, and the figures show it at 

 times attached to the middle of the rod. This strongly suggests preparations 

 that the writers have seen of known peritrichic organisms which, because of poor 

 technic, reveal only a single flagellum on a rod. The work of such careful in- 

 vestigators as DeRossi, Kellerman, Zipfel and Prucha ought not to be dismissed 

 with the few words granted to them by Hansen, while the still more recent work 

 of Wilson (Cornell Agr. Exp. Sta., Bui. 386. 1917) apparently has not been seen 

 by Hansen. The writers have seen microscopic preparations made by Kellerman 

 which are very much better than his published photomicrographs and show peri- 

 trichous flagella without any question. 



