VAN NIEL: SYSTEMATICS OF THE BACTERIA AND BLUEGREEN ALGAE 105 



This diagram shows that the new situation called for a decision as to the rela- 

 tive importance of the characters that can be used to link the different groups. 

 Obviously, a combination of morphological and physiological properties, once 

 justified because "intermediate" gi'oups were not known, was no longer ade- 

 quate. The formulation of a diagnosis of separate orders had, from now on, to 

 be based on either morphological or physiological features. Even this could 

 not provide a fully satisfactory solution to the problem of establishing larger 

 systematic units, however. For, when "morphology" was given preference, there 

 would still be the question whether the occurrence of sulfur globules, the indi- 

 vidual cell size, or the presence or absence of the special pigment system was 

 considered the most significant, while preferential use of physiological charac- 

 ters would imply the need for "grading" the respective values of sulfide oxi- 

 dation and pigment formation. 



Of course, the very admission of physiological characters in bacterial sys- 

 tematics might be blamed for the confused situation here discussed. Would it 

 not have been better if such criteria had been left out altogether in the crea- 

 tion of the two orders? In that event the filamentous colorless sulfur bacteria 

 could have been neatly segregated from the Thiohacillus group and from the sul- 

 fur and nonsulfur purple bacteria, regarding the latter assemblage as members 

 of the order Bubacteriales. While this may be considered a great improvement, 

 it nevertheless serves merely to shift the basic problem to the question of how 

 families should be defined. It can still be maintained that there would be ample 

 justification for the creation of a large systematic group of all the purple bac- 

 teria, especially because it is now known that the pigment system of these or- 

 ganisms confers upon them the ability to carry out an "aberrant" photosynthetic 

 mode of life (Molisch, 1907; Buder, 1919; van Niel, 1931, 1941, 1952). And 

 many arguments could be advanced to defend the thesis that such a unit, which 

 would also accommodate the green sulfur bacteria, has considerably greater 

 phylogenetic significance than, for example, groups comprising all Gram nega- 

 tive, nonsporeforming, polarly flagellated rod-shaped bacteria, regardless of 

 their physiological properties. 



The preceding discussion of the systematic status of the sulfur- and purple 

 bacteria may have served to illustrate the difficulties inherent in attempts to 

 accomplish primary divisions in the realm of the bacteria. Similar difficulties 

 are encountered at lower levels, and here, too, the problem must be faced 

 whether physiological characters are admissible. In some circles the idea that 

 they are not still prevails; on the other hand, the large number of generic 

 names with definite physiological connotations {Thiohacillus, Acetohacter, Lac- 

 tohacillus, Projnonihacterium, Hydrogenomonas, Nitrohacter, Methanococcus, 

 Photohacterium, etc.) testifies that this attitude is not universal, 



Manj^ of these names were introduced by Beijerinck and Winogradsky, and 

 it is clear that the ecological-physiological approach to general microbiology of 

 these two masters was largely responsible for the practice. The discovery that 

 a particular tj^pe of metabolism (sulfur oxidation, acetic acid production, lactic 

 or propionic acid formation, hydrogen or nitrite oxidation, methane production, 

 or ability to luminesce) seemed to be closely associated with certain types of 

 bacteria that were both easily procurable and readily distinguishable, compris- 

 ing relatively small groups of organisms with many common morphological 



