DIFFEKENTIATION OF UNKNOWN BACILLI 393 



of these residuary agglutinins either show no change or a slight 

 rise with a fall to their initial levels, or a marked rise, syn- 

 chronous or slightly earlier than that of the curve of the infect- 

 ing organism : the confusion introduced by the last occurrence 

 may be dispelled by the fact that the initial content of the 

 serum in inoculation agglutinins may be higher than that in 

 infection agglutinins. The explanation of the state of the 

 inoculation curves is that after, say, antityphoid inoculation, 

 the curve against the b. typhosus rises rapidly to a maximum 

 just as during infection, but falls very slowly over a period of, 

 it may be, many months. It is an interesting fact that an 

 acute paratyphoid B infection seems to have the effect of 

 stimulating a fresh formation of primary agglutinins against the 

 b. typhosus if these be already present in the serum. 



In trying to differentiate between primary and secondary 

 agglutinins the absorption method (p. 119) may also be used. 

 Castellani studied this method in experimental infections of 

 rabbits. He found that, when an animal had been immunised 

 with the b. typhosus, this organism would in vitro remove from 

 the serum not only the primary typhoid agglutinins but also the 

 secondary agglutinins, which might act on, say, the b. coli. If, 

 however, the animal had been immunised with both the 

 b. typhosus and the b. coli, then the b. typhosus could not absorb 

 from the serum the (primary) b. coli agglutinins. Castellani 

 therefore put forward the view that by this means primary 

 could be differentiated from secondary agglutinins, and conse- 

 quently pure could be distinguished from mixed infections. 

 The method has been applied in natural human infections and 

 may provide data of value especially when these are correlated 

 with the results of the other methods described. 



From what has been said it will be readily gathered that 

 agglutination reactions are of great value in the diagnosis of 

 intestinal infections and may enable a positive opinion to be 

 given in cases where attempts to isolate the causal organism by 

 culture fail. This is especially true of infections which have 

 reached a chronic stage in which cultures are often unsuccessful 

 in, it may be, 50 per cent, of the cases investigated. 



Differentiation of Unknown Bacilli through Agglutination 

 Reactions. — If the relation of a bacillus to its anti-serum is 

 specific it is obvious that the properties of such a serum can be 

 utilised for the recognition of bacilli of unknown species. Thus 

 if a serum originated by the b. typhosus agglutinates a bacillus 

 with the cultural characters of this organism the unknown strain 

 is almost certainly the typhoid bacillus ; this is especially likely 



