SERUM DIAGNOSIS 341 



with which clumping is obtained. There is a point in this 

 connection regarding which further light is required. Many 

 races of b. coli in use have been isolated from typhoid cases, 

 and we as yet do not know what effect a sojourn in such circum- 

 stances may have on its subsequent sensitiveness to agglutination 

 by typhoid serum. Again, Christophers has pointed out that a 

 large proportion of sera from normal persons or from those 

 suffering from diseases other than typhoid will clump the b. coli 

 in dilutions of from 1 : 20 to 1 : 200, and no doubt many of the 

 reactions shown by typhoid sera towards b. coli are due to the 

 pre-existence in the individuals of an agglutinative property 

 towards the latter bacillus. 



With regard to the value of the serum reaction there is little 

 doubt. In nearly 95 per cent of cases of typhoid it can be 

 obtained in such a form that no difficulty is experienced if the 

 precautions detailed above are observed. The causes of possible 

 error may be summarised as follows : the serum of the person 

 may naturally have the capacity of clumping typhoid bacilli ; 

 there may have been an attack of typhoid fever previously with 

 persistence of agglutinative capacity ; the case may be one of 

 disease caused by an allied bacillus ; the disease may have a 

 quite different cause, and yet the serum may react with typhoid 

 bacilli ; the disease may be typhoid fever and yet no reaction 

 may occur. The most important of these sources of error is 

 that .with which diseases caused by allied organisms are 

 concerned, as it is probable that all the forms which these take 

 in men have not been recognised. The very wide application 

 of the reaction has elicited the fact that it is given in many 

 cases of slight, transient, and ill -denned febriculse, which occur 

 especially when typhoid fever is prevalent. Some of these may 

 be aborted typhoid, some may be paratyphoid. There is no doubt 

 that, if all the facts are taken into account, the cases where the 

 reaction gives undoubtedly correct information so far outnumber 

 those in which an error may be made that it must be looked on 

 as a most valuable aid to diagnosis. In conclusion here we may 

 say that the fact of a typhoid serum clumping allied bacilli in 

 no way, so far as our present knowledge goes, justifies doubt 

 being cast on the specific relation of the typhoid bacillus to 

 typhoid fever. 



In connection with the phenomenon that a serum either from 

 a normal person or a typhoid patient may clump several varieties 

 of bacteria, some points arise. The theoretical consideration of 

 agglutination is reserved for the chapter on Immunity, but here 

 it may be said that agglutinating properties may be present 



