THE AGGLUTINATION REACTION 357 



up to 1 in 3 or 4, but occasionally is more active. Dead 

 bacilli may be used. The reaction is not obtained before 

 the sixth or seventh day of fever, occasionally not until 

 much later. Very rarely the reaction seems to be inter- 

 mittent. The blood may retain its agglutinating power 

 for years after an attack, and inoculation with anti- 

 typhoid vaccine also confers agglutinative properties. 

 Cases do occur in which agglutination is absent throughout, 

 but they are rare and often tend to be severe and to ter- 

 minate fatally. Usually, if the blood during the course 

 of an attack fails to give a reaction when tested on three 

 occasions at intervals of three to four days, it is improbable 

 that the case is one of typhoid fever. Moreover, cases 

 occur, simulating typhoid closely, due to infection with 

 the so-called para-colon or para-typhoid bacilli. These 

 " para " bacilli belong properly to the Gartner group of 

 organisms (see p. 371). If a positive reaction be obtained, yet 

 the case does not seem to be one of typhoid, a previous attack 

 or inoculation with typhoid vaccine must be excluded. The 

 previous injection of a typhoid anti-serum into the patient 

 might induce a non-typhoid infection to give the reaction. 



Gwyn 1 found that out of 265 cases diagnosed as typhoid 

 and accurately studied, only one persistently failed to 

 give the reaction. The blood of this case, however, reacted 

 typically with a Gartner-like organism obtained from the 

 blood (a case, therefore, of para- typhoid infection). 



Johnson and McTaggart 2 found that typhoid blood dried 

 for sixty days still gave a typical agglutination reaction. 

 An incomplete reaction was occasionally obtained as early 

 as the end of the second day, and the complete reaction 

 was rarely delayed beyond the fifth day. They also 

 noticed that the blood of the horse often produced clump- 

 ing, etc., of typhoid bacilli, indistinguishable from an 



1 Johns Hopkins Hosp. Bull., vol. viii, 1900, p. 387. 



2 Brit. Med. Journ., 1896, vol. ii, p. 629. 



