THE GENUS STREPTOCOCCUS 155 



correlation serves to fix certain groups of a fairly satisfac- 

 tory sort. The first two species, the long-chained virulent 

 form and the short-chained avirulent form, have been 

 referred to above as natural types. As to the fourth 

 species, it seems doubtful, as we have said, whether the 

 viability on potato is not too slight and fluctuating a 

 character for fixing a species. Species 5 and 6 are cer- 

 tainly only minor variants of the same liquefying type. 

 Species 7, according to our view, does not belong in this 

 genus. 



So the case rested in 1903; and v. Lingelsheim con- 

 cluded at the end of his monograph that only the future 

 could decide as to the unity or plurality of the genus. 

 In the most recent French medical treatise, Widal (1906) 

 emphatically maintained the inadequacy of any of the 

 differential methods previously proposed for the differ- 

 entiation of the various pathogenic and non-pathogenic 

 streptococci of the body. One hundred and twenty-two 

 different strains studied by him showed no constant differ- 

 ences in biological characters. Those obtained from 

 pathological conditions generally produced erysipelas in 

 rabbits, while those from the normal mouth did not; but 

 these pathogenic powers Widal recognized as wholly 

 transitory. 



Two innovations in systematic bacteriology were 

 required in order to clear up this vexed question, — a 

 detailed comparative study of fermentative powers and a 

 statistical analysis of the results obtained. The first of 



