PROSPECTS FOR A NATURAL SYSTEM OF CLASSIFICATION OF BACTERIA 



grouped with the obligatory oxidative saprophytic corynebacteria. 



For analogous reasons, but now because of the scantiness of mor- 

 phological data, the position of the genera Nitrobacter, Didymohelix, 

 Sideromonas, Kurthia, and Methano bacterium must be deemed privisional. 



The preliminary nature of the table manifests itself also by the 

 circumstance that in a restricted number of cases it was considered 

 desirable to maintain a number of genera which according to the 

 principle adopted should have been amalgamated. 



Because we have felt that there possibly exist sufficiently essential 

 differences between such genera as Azotobacter, Acetobacter, Kurthia, and 

 the genera with which they are grouped, we have refrained from 

 rashly abolishing them, although the morphological differentiation 

 which is at the basis of our system does not suffice to keep them sepa- 

 rated. In some instances a separation is already fully justified on the 

 basis of the outcome of the Gram-stain. In this connexion we refer to 

 the genera Kurthia- Alcaligenes, and Listerella-Pseudomonas, the species 

 of the first-mentioned genus of each couple being Gram-positive, those 

 of the second one Gram-negative. We may trust, moreover, that future 

 investigations will provide the means for a more adequate generic 

 differentiation in the cases under discussion. 



When we now proceed to an examination of the table as a whole, we 

 must first of all realize that its two-dimensional nature evidently does 

 not permit of a true representation of the natural relationships of the 

 various genera. The postulated many-sided morphological evolution 

 starting from the group of spherical organisms in reality asks for a 

 grouping of the three families of rod-shaped organisms around that 

 of the Micrococcaceae. 



Notwithstanding this shortcoming of the table it is obvious that in 

 many of the physiological groups distinguished the distribution of the 

 genera is far from random, as a glance at the upper half of the table 

 shows convincingly. Apparently there is a close correlation between 

 the phototrophic and chemo-autotrophic modes of life and morpholo- 

 gy. In this domain it is therefore rather inessential whether a physiol- 

 ogical or a morphological evolution is assumed to have been primary. 

 The mutual relationships remain clearly expressed in either case. 

 Another physiological group, viz., that of the true lactic acid bac- 

 teria, also shows a definite correlation between physiological and 

 morphological properties. 



313 



