30 L. A. JULIANELLE 



made that virulent types agglutinate only with serums prepared from 

 virulent strains, there must be an error somewhere. The properties of 

 virulence are obviously among the most unstable of bacterial characters. 

 Culture on laboratory mediums renders a virulent strain nonpathogenic 

 in a very short time. Yet it is scarcely conceivable that the immunity 

 reactions are as readily modified. By way of illustration: Strain LI, 

 which was distinctly pathogenic, was used for the preparation of 

 immune serum before it could have undergone avirulence ; but its serum 

 did agglutinate other strains, including A5, P3 and T9, all 3 of which 

 were proved nonpathogenic. It may be that A5, P3 and T9 were 

 pathogenic at sortie time or another, but at the time the test was made 

 they were not pathogenic. It seems clear to us that virulence does not 

 dictate the group into which a staphylococcus shall fall. 



Nor does it seem plausible that hemoly tic activity is the basis of 

 agglutination grouping. We have been unable to obtain absolutely non- 

 hemolytic cultures, and have been unable to establish this point con- 

 clusively. However, we were able to get these groups among hemolytic 

 organisms, whereas if hemoly sis were the fundamental of the grouping, 

 we should have obtained agglutination of all our strains by all our 

 serums. 



Regarding the association of pigment and agglutination, this much 

 can be said : Occasionally, there may develop on a plate streaked with 

 a pure culture, colonies varying appreciably in intensity of pigment, 

 from which, as Sullivan 60 has shown, quite distinct types may be 

 derived by selection of the extremes. Yet it does not seem probable 

 that the parent strain in such a case would vary from its successor in 

 its agglutination reactions. More relevant, however, strain Jl, which is 

 an albus, did agglutinate with serums A5, PI and LI, which were 

 prepared from antigens of varying shades of orange. An analysis of 

 the pigment and agglutination tables, with this one exception cited, 

 bears out the contention of Walker and Adkinson 58 in a general way. 

 The members of group 1 are of a light pigment either white or of a 

 light shade of yellow, which without the refined technic of Winslow and 

 Winslow 28 would easily be called a white. 



A study of the tables of the different biochemical reactions shows 

 no definite relationship between the agglutination groups and these 

 reactions. In a very general way Group 1 seems to contain the less 

 active strains, but it also contains some rather active strains. Groups 2 

 and 3 possess none of the light pigmented nor any of the less active 

 strains. 



" Jour. Med. Res., irOS, 14, p. .09. 



