STREPTOCOCCUS PYOGENES 363 



tinate the strain of streptococci used for immunization. 

 It may agglutinate a certain number of the other hemo- 

 lytic varieties but it is unlikely that it will agglutinate 

 all. We would assume then that etiologically all those 

 agglutinated by that serum were of one type, while all 

 others of our group were of another type or probably types, 

 and so we might continue throughout the whole group 

 originally selected and arrange them into types or classes 

 the members of each of which would react specifically with 

 its homologous serum and not with other sera. 



The object of such grouping is more far-reaching than, 

 that of simply identifying variations; it has a most practical 

 bearing on all efforts to produce antisera that may be used 

 in preventing infection or in curing it when once under way. 



We have already enough evidence to justify the general 

 statement that for any immune serum to possess protective 

 or curative properties for a bacterial infection the animal 

 from which it is obtained must have been immunized with 

 either bacteria direct from the disease against which the 

 serum is to be used or with types closely allied to them in 

 the fundamental infective and immunologic characteristics. 

 (Does this hold for such infections as tetanus and diphtheria?) 



This being the case the desirability of establishing groups 

 or types of streptococci, the members of which are closely 

 allied in these particulars, becomes evident: for it is not 

 practicable in efforts to treat infections along these lines 

 to always immunize animals from which the curative serum 

 is to be obtained with the organism specifically concerned 

 in the case under consideration. 



Efforts have been made to overcome this difficulty by the 

 production of "polyvalent" antisera, i. e., serum from ani- 

 mals immunized not by the use of a single strain of strepto- 



