- 23 - 



This group differs from the last in the presence of a con- 

 siderable percentage of white , with sometimes a trace of yellow. 

 Here would belong the so-called "salmon pink", "coral pink", 

 "rose", or "flesh colored" cultures. 



Cultures producing "soluble" red pigment may be classed as 

 the Lactis erythrogenes group, including 

 B. lactis erythrogenes, 

 B. rutilescens (n. sp.), 

 B. rubefaciens, 

 B. lactorubefaciens. 



B. mesentericus ruber is peculiar in producing its pink 

 or red pigment only on potato, not upon agar. 



The term "Prodigiosus group" as used below, refers then to 

 the series of Group I as described. 



II. Variability in the Prodigiosus group. 

 1. Introductory. 



In the discussion of variability or variation as regards bac- 

 teria, several complicating factors are present. In the first place, 

 the facility with which bacteria, more than any other class of or- 

 ganisms, respond and adapt themselves to changes in the nature 

 of their environment is a constant character in their biology. 

 Again, a bacterial type, as we assume it, is more or less artificial, 

 developed and maintained by our methods of culture, which are so 

 arranged as to reduce or eliminate variation. And the old simple 

 distinctions between typical varieties are now complicated not only 

 by the natural variability of the organisms, but by the conception 

 of intermediary paratypes, each with its own possibilities of 

 variation. 



The variability of bacteria, whether manifested spontaneously 

 or under compulsion, seems to find its expression principally in 

 the loss of certain characters or in their return after loss. If, for 

 instance, B. anthracis be exposed to a temperature of 42, 

 B. tetani to a gradual admission of oxygen, orB. prodigiosus 

 to the action of light, the organisms will all live and develop, but 

 one will lose its power of sporulation, another its virulence, the 

 other, its power of pigment production; and unless the abnormal 

 conditions be maintained for many generations, and the process 

 fostered by artificial selection, the organisms will, upon restoration 

 to their usual environment, revert to their original "normal" type. 

 These and many similar observations indicate that bacteria have a 

 certain number of biological characters, without which they may 

 continue to live; that these characters may be lost as a result of 

 environmental modification, but are so far typical that they tend 

 strongly to return. 



This "tendency to return" finds its nearest explanation quan- 

 titatively, that is, in considering as a factor in variation not only the 

 effect of the environment upon the organism , but also the nature 

 of the compound of characters which we term a type. The aber- 



