204 



a ,. fringe", and of different sizes etc.). Subcultures were made 

 from one colony of each kind. These subcultures were denoted 

 by the addition of (a) and (b) to the original number of the strain 

 in question (I 1, I 22, I 57, I 91, P 15, H 19, H 152). In new 

 subcultures some months later from (a) and (b) cultures there was 

 no longer any difference however, in the appearance of the colonies 

 of 4 of the strains, nor as regards agglutination. The (a) and (b) 

 subcultures of the remaining strains (I 91, H 19, and H 152) were 

 still different as regards form of colony or agglutination, or both. 

 There was no difference in the indol test between the two cultures 

 It was further remarked that in several strains the culture 

 consisted of ligbt and dark portions, arranged in sectors, and separa- 

 ted from one another by sharp boundary lines. The granulation 

 structure extended quite unaffected through the light and dark sec- 

 tors. From one such culture, plate cultures were made; a mixture 

 of light and dark colonies appeared on the plates. No difference 

 could be made out in agglutination tests combined with absorption 

 between cultures from the light and dark colonies. 



We have now arrived at the end of the analysis of the 

 bacterial group which — hitherto as a working hypothesis — 

 we have included under the general heading „Pfeiffer's bacil- 

 lus". We are confirmed in the opinion that all these strains 

 ought to be placed in the same group by the existence of a 

 series of common characters which bind them 1 together, but 

 which does not exclude the group being somewhat hetero- 

 geneous in several directions. But we have only observed a 

 distinct discontinuity in one property, namely the indol test. 

 This discontinuity cannot however disrupt the group because 

 the necessary correlation with other characters is wanting. 

 The group can be split up in a more rational manner, into the 

 two sub-groups, typical and atypical Pfeiffer's bacilli, which 

 however are associated with transitions forms. 



It is principally the „inner structure" of the Pfeiffer's 

 bacillus group that we have examined so far. We will now 

 regard the group „from without", by comparing it with three 

 different related bacterial types, the haemolytic haemoglo- 

 binophilic bacteria, Bordet's whooping-cough bacillus, and Ba- 

 cillus haemoglobinophilus canis. 



