168 BACTERIOLOGICAL EXAMINATION OF WATER. 



no effect, but exceptionally it showed an agglutinative action in 

 a dilution of 1-5. Biberstein and Stern examined the sera of 

 persons who were not suffering from typhoid fever. The following 

 results were obtained : 



BIBERSTEIN. STERN-SKL.OWER. 



Dilution B. Typhosus culture. Dilution B. Typhosus culture. 



of serum. Agglutinated in of serum. Agglutinated in 



1-10 24 per cent, of the cases 1-10 25 per cent, of the cases 



1-20 6 ,. 1-20 8 ' 



1-30 2 1-30 2 



1-40 1-40 1 



In my own experiments with normal serum from a horse I 

 found that the B. typhosus was always completely agglutinated 

 by the serum diluted 1-20, and a marked reaction was obtained 

 when the serum was diluted 1-80. 



From these studies it is evident that, when an unknown serum 

 is being tested for a specific reaction with a known bacillus 

 (Widal's reaction), the errors introduced by the action of the 

 agglutinins present in normal blood may be avoided by working 

 with the unknown serum in a dilution of 1-50 ; but when 

 attempts are made to diagnose an unknown bacillus by the 

 action of a known specific serum upon it, the problem is more 

 difficult. Much recent work on this subject has shown that at 

 least three variable quantities enter into the reaction, viz. : 

 (1) the nature of the bacillus ; (2) the strength of the specific 

 serum ; and (3) the time during which the serum is allowed to 

 act on the bacillus. It will be necessary to consider these three 

 points in detail : 



(1) The Nature or Species of Bacillus operated on by the 

 Specific Serum. In the early days of serum diagnosis it was 

 thought that a specific serum only agglutinated the bacillus 

 which had caused the serum to assume its specific characters. 

 But a very little study soon showed that a specific serum, such 

 as an anti-typhoid serum, would agglutinate other bacilli besides 

 the B. typhosus ; and of late years an enormous amount of 

 work has been done in order to find out the essential points 

 required to establish a really specific reaction. The B. coli and 

 its varieties and B. enteritidis of Gartner being closely allied to 

 the B. typhosus in their cultural characteristics, have necessarily 

 received much attention, and many experiments have been made 



