1398 THE ENTEROIDEA GROUP OF TROPICAL FEVERS 



agglutinations; 55 per cent, reacted with Gaertner, or Brion and 

 Kayser; 41 per cent, with Schottmiiller B; 33 per cent, with Aer- 

 trycke; and 12 per cent, with Schottmiiller A. 



The greater the quantity of typhoid agglutinin, the greater, as a 

 rule, the subsidiary agglutination. 



To distinguish between primary and secondary agglutinins, and 

 to facilitate the diagnosis of closely allied organisms and mixed 

 infections, the absorption test, discovered by Castellani in 1902, is 

 useful. This test has been further studied, and the technique 

 improved, by Boycott, Bainbridge, O'Brien, Cummins and Gumming, 

 Alcock, and others. An excellent monograph on the absorption 

 test is the very recent one by Frank E. Taylor. 



Castellani found out that in rabbits immunized for typhoid 

 only, whose serum agglutinated besides the typhoid bacillus, also 

 certain ' coli ' germs, the saturation with an excess of typhoid 

 bacilli would remove not only the primary (specific, homologous) 

 typhoid agglutinin, but also the secondary (heterologous, non- 

 specific) coli agglutinin; while in a serum derived from rabbits 

 immunized both for typhoid and coli bacilli, neither saturation 

 with typhoid alone nor coli alone, but only both together, simul- 

 taneously or successively, would remove the whole of the agglu- 

 tinins present in the serum. 



He experimented with various other germs, and applied the 

 method to the differentiation of closely allied bacilli and mixed in- 

 fections in man. For practical clinical purposes it may be said 

 that if the blood of a typhoid patient presents, besides agglutination 

 for the B. typhosus, also agglutination for, say, B. paratyphosus B 

 or /4, or a coli-like germ, and if this agglutination for the B. para- 

 typhosus B or A, or coli-like, persists to a great extent after satura- 

 tion with B. typhosus, the case is ver^^ probably one of true mixed 

 infection: typhoid + paratyphoid B, or paratyphoid A, or coli- 

 like, as the case may be. 



Of course, as with ever^/ other biological test, the results obtained 

 cannot be accepted as absolute, but are only of relative value. 



To remove the typhoid agglutinin from the typhoid serum, this is diluted, 

 and an excess of typhoid bacilli from young agar cultures added and kept in 

 contact for two hours. The serum is then tested for agglutination for the 

 various germs. A similar technique is used for removing the specific para- 

 typhoid B, paratyphoid A agglutinins, etc. ; by adding to the serum an excess 

 of paratyphoid B or paratyphoid A, etc., bacilli. For technical details one 

 should consult books on advanced bacteriology, or Taylor's excellent mono- 

 graph on the absorption test. 



Prognosis. — The typhoid mortality is, in the tropics, about 20 to 

 25 per cent, for ordinary hospital practice, and rather lower for 

 private practice. The mortality from paratyphoid A and para- 

 typhoid B is lower. Unfavourable symptoms are the early appear- 

 ance of haemorrhage, severe nervous symptoms, considerable 

 meteorism, and severe diarrhoea. A bad sign, pointing to a probable 

 relapse, is the temperature falling while the pulse remains rapid. 



