PRACTICAL APPLICATIONS 291 



positive reactions in the first week; about 60 per cent, in the second 

 week; about 80 per cent, in the third week; about 90 per cent. in. the 

 fourth week, and about 75 per cent, in the second month of the disease." 

 In about 90 to 95 per cent, of cases in which repeated examinations are 

 made a positive reaction is to be found at some time during the patient's 

 illness. 



Occasionally the reaction appears first during the stage of convales- 

 cence, and at times it may even be absent, the diagnosis being confirmed 

 by cultivating typhoid bacilli from the blood. The possibility of a given 

 case reacting strongly one day and weakly or entirely negative a day or 

 so later has been emphasized elsewhere. 



Usually the reaction is strongest during convalescence, remains posi- 

 tive for several weeks, and then gradually returns to the normal. Occa- 

 sionally the reaction remains positive for months or even years after the 

 attack of typhoid fever many such cases are " carriers" and harbor the 

 bacilli in the gall-bladder, although the person appears to be quite well. 



Only very rarely does normal serum immediately agglutinate typhoid 

 bacilli in a dilution higher than 1:10; where a time limit of one to 

 two hours is given, a few may show some agglutination in dilutions up 

 to 1 : 30. 



If the typhoid bacillus is agglutinated by the patient's serum in a 

 dilution of 1:100, or at least 1:40, the Widal reaction may be regarded 

 as positive. It is not safe to use lower dilutions, as occasionally the 

 serum of healthy persons may agglutinate Bacillus typhosus in dilutions 

 up to 1:30. Due care must be exercised not to mistake a pseudo- 

 reaction about detritus for true agglutination. 



Positive reactions are occasionally obtained in other diseases acute 

 miliary tuberculosis, malaria, malignant endocarditis, and pneumonia. 

 It is also well to bear in mind the possibility of a patient having been 

 vaccinated against typhoid at some early date, with resulting agglutinin 

 formation. 



Owing to the similarity of symptoms an infection with Bacillus para- 

 typhosus A and B may be difficult to distinguish from typhoid fever. 

 This difficulty is increased by the confusion of the Widal reaction 

 owing to the presence of group agglutinins in the serum if proper dilu- 

 tion is not practised. Bacillus A and Bacillus B are not identical in their 

 agglutinable properties, the latter being more closely related to the 

 typhoid bacillus than the former. In this country Bacillus paratyphosus 

 is usually held responsible for paratyphoid fever. Conclusions should 

 not be drawn until tests have been made with both strains of the para- 



