TYPHOID FEVER. 495 



The reactive phenomena are very slightly interchange- 

 able for species of bacteria closely allied to the typhoid 

 bacillus, so that a serum with strong agglutinative 

 properties may slightly affect the colon bacillus, and a 

 serum from an animal immunized to the colon bacillus 

 may be able to affect slightly the typhoid bacillus. Far 

 from lessening the value of the test, this, as Welch points 

 out, only argues for the close relationship of the species 

 acted upon. 



The typical reaction does not occur with the other 

 closely allied members of the typhoid group of bacilli. 

 The attempts to make the colon bacillus agglutinate by 

 application of typhoid serum to their culture usually 

 fails. In cases of suspected typhoid, in which the reac- 

 tion upon the colon bacillus takes place, it is impos- 

 sible to eliminate the possibility of combined typhoid 

 and colon-bacillus infection, as suggested by Johnston 

 and McTaggart. 



Widal and Courmont found that all human sera, 

 whether normal or typhoid, have a slight action upon the 

 colon bacillus in dilutions of 1 : 10, whereas normal 

 serum, as a rule, has no effect upon the typhoid bacillus 

 in this dilution. 



The agglutinative power of the specific serum upon 

 the bacteria is so delicate that bacteriologists now make 

 use of the phenomena for the differentiation of similar 

 species. Richardson's serum-paper, which will be fully 

 described elsewhere, is a most important adjunct to our 

 means of separating the colon and typhoid bacilli. 



In a number of cases that I studied, serums which 

 reacted strongly upon the typhoid bacillus failed to pro- 

 duce any alteration in cultures of Bacillus coli communis, 

 Bacillus suispestifer, and Bacillus icteroides. 



X. The technic is of the utmost importance. Widal ' 

 suggested that a small quantity of blood be withdrawn 

 with a sterile syringe from the median cephalic vein, and 

 a few drops of it added to a fresh bouillon culture of the 



1 Semaine mid., 1896, p. 259. 



