vi PREFACE 



earth, in the laboratory of the Massachusetts Institute 

 of Technology. All were submitted to eleven definite 

 and, in most cases, quantitative tests; and the results 

 were analyzed with a view to the centers about which 

 each character varied in the series as a whole, and to 

 the general correlation between different characters. 

 While this work was under way it was unexpectedly 

 extended, and its principles confirmed, by the publication 

 of a systematic study of the streptococci by Andrewes 

 and Horder. These observers, working in England 

 on the finer divisions of one group of cocci, have inde- 

 pendently arrived at the same conclusion reached by 

 us in regard to the general systematic relationships 

 of the bacteria: that bacterial groups can be defined, 

 and can only be defined, by a study of the numerical 

 frequency of various characters in a large series of 

 cultures. 



The results of the statistical study of over five hundred 

 cultures of cocci, considered in connection with Andrewes 

 and Border's investigation of the streptococci, and 

 supplemented by a careful review of previous publica- 

 tions on the group as a whole, furnish a reasonable con- 

 ception of the relationships of the family, Coccaceae, on 

 this basis of numerical frequency types. Two main 

 series, or subfamilies, may be distinguished: one pri- 

 marily parasitic and the other saprophytic. The groups 

 differ in morphology, staining reactions, cultural charac- 

 ters, and biochemical powers. Within these two sub- 



