8TREPTOCOCC1C SERUMS. 533 



example, not all strains produce hemolysin to the 

 same degree, and they differ greatly in their sus- 

 ceptibility to the action of an agglutinating serum. 

 We have also to remember that pathogenicity for 

 animals is not a reliable index of pathogenicity for 

 man. From these confusing conditions we can 

 only regard the question of unity or multiplicity 

 of streptococci as an open one, which may be de- 

 cided by future investigations. 



The serums of Marmorek and Aronson are uni- univaient and 



, , . , . Polyvalent 



valent serums, a single strain being used for im- serums. 

 munization. Certain investigators, believing in 

 the multiplicity of streptococci, utilize several 

 strains in immunization. The serum of Denys is 

 obtained by immunizing with several strains the 

 virulence of which has been artificially increased. 

 Such a serum would, theoretically, have a wider 

 range of action than a univalent serum ; it is poly- 

 valent. Having in mind the fact that passing a cul- 

 ture through rabbits increases the virulence of 

 the organism for the rabbit, but alters its virulence 

 for the original host (man), Tavel, Moser and 

 Menzer prepare serums on a different basis. 

 Tavel employs several strains of streptococci cul- 

 tivated from pathological processes in man, avoid- 

 ing such alterations in virulence as would be caused 

 by passing the cultures through animals. On the 

 assumption that scarlet fever is a streptococcus 

 disease, Moser immunizes horses with strains 

 (about twenty) which are cultivated from cases 

 of scarlet fever. In a similar manner, Menzer, 

 supposing that rheumatic fever is a streptococcus 

 infection, immunizes with a number of strains cul- 

 tivated from the tonsils of cases of rheumatism. 

 Both Moser and Menzer avoid passage in order to 



