VARIABILITY AMONG BACTERIA. 31 



of methods by which growths of a single kind of bacteria without 

 admixture of any other variety can be obtained. Undoubtedly many 

 of the earlier observations were made on mixtures of different organisms. 

 Thus the upholders of the occurrence of great variability founded their 

 view very largely on observations of a group of organisms closely 

 allied to the bacteria which, from a peculiar pigment called " purpurin " 

 in their protoplasm, have been called the purpurin bacteria (German, 

 Purpurbakterien). From a more recent study of the group by 

 Winogradski, however, these seem to be a mixture of many different 

 species, each of which maintains during multiplication its characteristic 

 form. Practically no one at the present day holds that an organism 

 can appear now as a bacillus, now as a coccus, now as a spirillum. 

 Nor is such a view in any way supported by the occurrence in isolated 

 cases among the higher bacteria of coccus-like and bacillus-like seg- 

 mentation, such as we have seen to take place in the streptothrix 

 group. 



With regard to the bacteria as a whole we may say that 

 each variety tends to conform to the definite type of structure 

 and function which is peculiar to it. On the other hand, 

 slight variations from such type can occur in each. The 

 size may vary a little with the medium in which the organism 

 is growing, and under certain similar conditions the adhesion 

 of microbes to each other may also vary. Thus cocci, which 

 are ordinarily seen in short chains, may grow in long chains. 

 The capacity to form spores may be altered, and such 

 properties as the elaboration of certain ferments or of certain 

 pigments may be impaired. Also the characters of the 

 growths on various media may undergo variations. As has 

 been remarked, variation as observed consists largely in a 

 tendency in a bacterium to lose properties ordinarily 

 possessed. Practically no case of a bacterium acquiring 

 new properties has been observed, and all attempts to 

 transform one bacterium into an apparently closely allied 

 variety (such as the B. coli into the B. typhosus) have failed. 

 This of course does not preclude the possibility of one 

 species having been originally derived from another or both 

 having descended from a common ancestor, but we can say 

 that only variations of an unimportant order have been 

 observed to take place, and here it must be remembered 

 that in many cases we can often have forty-eight or more 



