IMPORTANCE OF ANTIBODIES IN DIAGNOSIS 391 



From nil inspection of this table, it will be observed that patient 

 R. R. S. developed aj^glutinin titers for Br. abortus quite com- 

 parable with those for P. tularensis and that the titer for Br. 

 mcUtcnais was noticeably increased. In the remaining four cases 

 of tularemia, Francis records a definite increase in titer for both 

 strains of Brucella organisms, but in no case did he observe such a 

 striking res])onse as in that of patient R. R. S. These results 

 may l)e due to similar antigenic su])stances common to the Brucella 

 group and to P. fularens{.s and also to the i)ossibility that when 

 tularemia develops in a patient who has a latent infection with 

 some member of the Brucella group agglutinins for both result. 

 Since the diagnosis of either of these diseases is usually confirmed 

 by the agglutination test, an appreciation of such phenomena 

 is of importance in interpreting laboratory reports. 



Antigenic Relationships Between S. (Iallinarum and E. 

 Typhosa. — The commercial laboratory is not infrequently called 

 upon to test the blood of chickens for the presence of agglutinins for 

 S. gallinarum or *S'. pullorum. Bushnell (1926) has recommended 

 an agglutination ])rocedure designed to determine whether active 

 infection exists in the fowl. Tliese organisms are of interest since 

 they possess remarkable antigenic similarity to E. tuphosa. Sher- 

 wood and Hoffman (1929) observed that the blood sera of several 

 typhoid patients agglutinated a suspension of S. gallinarum 

 (sanguinarum) at a higher titer (1:800) than the suspension of 

 E. typhosa used in the Widal. IMulsow (1919) working in 

 Jordan's laboratory concluded that the organisms could be dif- 

 ferentiated only by careful absorption experiments. Theobald 

 Smith (1915) has suggested that E. ti/phosa may be a variant of 

 .S'. (jalUnarum. If this should eventually i)rove to be true, it might 

 ex])lain many things at present difficult to understand in regard 

 to the epidemiology of typhoid fever. As mentioned earlier in this 

 chapter. White (1926, 1932) reports the existence of a somatic 

 antigen ("0" antigen) common to E. typhosa, S. gaUinarum, S. 

 pulloruui and S. enteritidis. Attention has already been called to 

 tiie discovery of carbohydrate-lipid complexes in E. typhosa by 

 Topley, Raistrick and their associates (1933, 1938). These sub- 

 stances are highly antigenic and incite the production of the 

 ' ' granular ' ' agglutinins. 



