SERUM DIAGNOSIS. 345 



Such is what occurs in the case of a typical reaction. 

 There are several details, however, which require attention, 

 and on which the value of the method as a means of 

 diagnosis largely depends. The race of typhoid bacillus 

 employed is important. All races do not give uniformly 

 the same results, though it is not known on what this 

 difference of susceptibility depends. The bacteriologist 

 must, therefore, apply a process of selection to the races at 

 his disposal, with a view to obtaining one which gives the 

 best result in the greatest number of undoubted cases of 

 typhoid fever, and which gives as little reaction as possible 

 with normal sera or sera derived from other diseases. This 

 latter point is important, as some races react very readily to 

 non-typhoid sera. Again, care must be taken as to the 

 state of the culture used. The suitability of a culture may 

 be impaired by varying the conditions of its growth. Con- 

 tinued growth of a race in surroundings very favourable to 

 vegetative activity makes it less suitable for use in the test, 

 as the bacilli tend naturally to adhere in clumps, which may 

 be mistaken for those produced by the reaction. Wyatt 

 Johnson recommends that the stock culture should be kept 

 growing on agar at room temperature and maintained by 

 agar sub-cultures made once a month. For use in applying 

 the test, bouillon sub-cultures are made and incubated for 

 twenty-four hours at 37 C. As the reaction of the medium 

 has also an important effect on the sensitiveness of a culture 

 he recommends that such bouillon should first be made 

 neutral to phenol-phthalein, and then have added to it three 

 or four per cent of normal hydrochloric acid. When these 

 precautions are taken a growth occurs which only gives a 

 uniform turbidity in the bouillon without any adhesion of 

 the bacilli in masses. The relation of the dilution of the 

 serum to the occurrence of clumping is most important. 

 It has been found that if the degree of dilution be too 

 small a non-typhoid serum may cause clumping. If 

 possible, observations should always be made with dilu- 

 tions of i : 10, i : 30, i : 60, i : 100. To speak generally, 

 the more dilute the serum the longer time is necessary for 



