538 HILDA HEMPL HELLER 



In order to ensure a natural classification, characters must be 

 worked out for each group, characters that will to some extent 

 correspond, and show by such correspondence or by the lack 

 of it where lie the historical divisions and where the parallel 

 developments that have taken place independently. This alone 

 is a great labor. For the group of the anaerobic rods and for 

 many other groups nothing of the sort has as yet been seriously 

 attempted. The Winslows' classification of the Coccaceae, a 

 pioneer work in this direction, has applied several principles, 

 which may well be heeded in making future classifications. 

 These authors applied to 500 strains of cocci from various sources 

 the biometric principles in use by students of heredity, by anthro- 

 pologists, and to some extent by botanists and zoologists. Upon 

 a study of the tabulated figures based on the behavior of these 

 organisms they formulated their determination of what to call 

 a species and of how to group species into genera. They found, 

 in common with botanists and zoologists, that when abundant 

 material is at hand it is quite impossible to define as a species 

 one single type. If our methods were sufficiently refined we 

 could probably distinguish every bacterial strain from every 

 other, just as we can distinguish every human being from every 

 other. A species is finally to be determined by comparing the 

 characters of aggregates of individuals (higher plants and ani- 

 mals) or of strains (bacteria), and by selecting the types which 

 occur most frequently as the standard upon which to base specific 

 descriptions. The conclusions arrived at by Winslow and Wins- 

 low as to analysis of their data are as foUows: 



First, each center of numerical frequency, marking a group of organ- 

 isms varying about a distinct type in regard to a single definite pro- 

 perty, may be recognised as a species. Second, those species which are 

 bound together by the possession of a number of similar properties 

 may be constituted as genera, and larger groups of genera, still charac- 

 terized by some characters in common, may receive the rank of families 

 or subfamilies. 



This method of working is evidently very different from the 

 old method whereby one man described one strain and another 



