38 WiNSLOW : A statistical criterion for bacteria 



modified, either experimentally or under natural conditions. 

 Among the flowering plants the study of a long series of specimens 

 from habitats separated in space or time often tends to make the 

 boundaries between species almost indistinguishable. If we could 

 modify and control the environment of these higher forms through 

 countless generations, as we do in the case of the bacteria, there 



can be little doubt but that we 



m into 



another. Yet for practical convenience, and in recognition of the 

 existing preponderance of certain types, we recognize them as con- 



stituting species. 



may 



that even mutable types in certain unstable groups are deserving of 

 systematic recognition. 



It is a curious fact that some of the most variable characters 

 among the Coccaceae, chromogenesis and reaction to the Gram 

 stain, for example, proved to be most strikingly correlated with 

 the modifications of other powers, associated respectively with the 

 parasitic and saprophytic habitats. Among the higher plants the 

 modifications due to environment are comparatively superficial. 

 Among the plastic unicellular organisms on the other hand the 

 adaptation to parasitic life, for example, may produce profound 

 changes, which warrant generic rank. Among the more complex 

 plants and animals, indeed, the same conception of purely en- 

 vironmental types Is gaining ground. It is not in connection with 

 bacteria that Jordan and Kellogg (Evolution and Animal Life, 

 New York, 1907), speak of ^' Ontogenetic species held for a num- 

 ber of generations true to a type simply because the environment, 

 the extrinsic factors in the development of all the individuals in 

 these successive generations, are the same." 



The main points which I have tried to emphasize as essential 

 to a rational natural classification of the Schizomycetes may be 

 briefly summarized as follows : 



i. 



ome 



fungi, which have differentiated along physiological rather than 

 morphological lines, differences in metabolism may have the same 

 systematic importance given to gross structural differences in other 

 groups. 



2- The characters of greatest systematic importance, whether 

 morphological or physiological, vary in each particular group of 



