SUBORDER RHODOBACTERIINEAE 839 



Two groups of sulfur bacteria could be distinguished, one consisting of colorless, 

 the other of red or purple organisms. The members of both groups presented an un- 

 usual morphology apart from the sulfur droplets; in all cases the individual cells 

 were considerably larger than those of the common bacteria, while many species 

 grew as distinctive colonial aggregates. Migula separated these sulfur bacteria into 

 two families, Beggiatoaceae and Rhodobacteriaceae. Even at that time, however, 

 some difficulties existed as to just what organisms should properly be considered as 

 sulfur bacteria. Miyoshi (Jour. Coll. Sci., Imp. Univ., Tokyo, 10, 1897, 143) had 

 discovered a bacterium which forms stringers, incrusted with sulfur, in sulfur springs, 

 but which does not store sulfur globules in its cells. Although physiologically this 

 organism appeared to comply with Winogradskj'^'s concept of a sulfur bacterium, the 

 absence of the typical cell inclusions made Miyoshi decide it could not be considered 

 as such. The problem was aggravated when Nathansohn, Beijerinck, and Jacobsen 

 published their studies on small, colorless, PseudoinonasAike bacteria capable of 

 oxidizing hj'drogen sulfide, sulfur, and thiosulfate, and evidently dependent upon 

 this oxidation process for their development. Morphologically these organisms 

 have little in common with the Beggiatoaceae; they were designated by Beijerinck as 

 species of Thiobacillus and have since been rightly considered as members of the order 

 Eubacteriales (see p. 78). Nevertheless, these organisms are physiologically in no 

 way different from the Beggiatoaceae, so that if physiology only is considered, a good 

 case could be made out for their incorporation in the Thiobacteriales. 



Furthermore, Molisch (Die Purpurbakterien, Jena, 1907, 95 pp.) described in some 

 detail a number of bacterial species which, in view of their characteristic pigment 

 system, appeared closely related to the Rhodobacteriaceae, but which develop only 

 in organic media and are, therefore, not sulfur bacteria in the sense of Winogradsky 

 or Migula. In stressing the importance of pigmentation Molisch combined the red 

 sulfur bacteria and the newly discovered purple bacteria into an order Rhodobacteria 

 with the two families Thiorhodaceae and Athiorhodaceae. It is this grouping that 

 has been followed in the present edition of the Manual. 



Only a very small number of typical sulfur bacteria have been studied in pure 

 cultures. As a result the descriptions of genera and species rest mainly on observa- 

 tions made with collections from natural sources or crude cultures. Most investi- 

 gators have implicitly accepted differences in cell size or in colonial appearance as 

 a sufficient justification for establishing independent species. Evidently this proce- 

 dure presupposes a considerable degree of constancy of such characteristics in the 

 organisms in question. It is true that Winogradsky 's investigations have provided 

 a reasonable basis for this belief, but later studies with pure cultures of certain purple 

 bacteria have established beyond a doubt that environmental conditions, such as 

 composition of the medium and temperature, may exert a profound influence on the 

 general morphology of these organisms. By this, it is not intended to infer that the 

 previously proposed genera and species of sulfur bacteria should be abandoned. But 

 it does follow that a cautious evaluation of the distinguishing features is necessary. 

 In the absence of carefully conducted investigations on morphological constancy and 

 variability of most of the previously recognized species of sulfur bacteria with pure 

 cultures grown under a variety of external conditions, the best approach appears to 

 be a tentative arrangement of these organisms based upon those characteristics which 

 are readily ascertainable. Experience with this group over the past twenty years 

 has shown that, while Winogradsky's fundamental work must remain the foundation 

 of present taxonomic efforts, it is advisable to simplify the much more elaborate 

 classification developed by Buchanan which was followed in previous editions of this 

 Manual. 



