832 COMPLEMENT FIXATION 



of immunized animals, the complement fixation test was easily and readily demon- 

 strated, but certain difficulties arose in the precise standardization of the test, es- 

 pecially when it was applied to the identification of species or the diagnosis of infec- 

 tion. In the identification of species, cross-reactions were obtained with closely re- 

 lated and, less frequently, quite different species of bacteria, and it was evident that 

 these crude antigens possessed non-specific as well as specific activities which must be 

 differentiated if the test was to possess practical, or even academic, significance. 

 When applied to the diagnosis of infection, the reactions occurred with apparently 

 healthy individuals and, not infrequently, failed to detect infection known to exist. 

 Altogether, the results of the early work were not encouraging; in fact, so discouraging 

 that neither Bordet nor Wassermann continued their work with the bacterial anti- 

 gens. Shortly before his death, Wassermann' again tried to make practical the test 

 in tuberculosis; how successfully is not even yet quite clear. Thus, the work has been 

 developed very slowly along two slightly different but related lines. Observers who 

 have been interested in the taxonomic study of the identification of species and types 

 of species by delicate, specific immune reactions have turned, in some instances, to 

 the complement fixation reaction. This work covers a large field, including many, or 

 nearly all, of the known common pathogenic bacterial species. Others, more deeply 

 interested in the practical diagnosis of infection, have studied the more common 

 diseases, the clinical diagnosis and prognosis of which might be greatly aided by 

 reliable laboratory tests of the immune reactions that develop during various stages 

 of the infectious processes, either for the purpose of diagnosis or prognosis or to de- 

 termine cure; for example, such infections as tuberculosis and gonorrhea, on the one 

 hand, and, on the other, meningitis and pneumonia in which the different strains of 

 the species are so important. In determining the etiological relationship of his bacillus 

 to pertussis, Bordet^ relied largely upon the complement fixation reaction. 



Thus there has been a real interest in the study of complement fixation with 

 bacterial antigens in general, but, on account of the pitfalls with which the study has 

 been and is still beset, attempts to solve the problems have proved inconclusive and 

 have yielded results not satisfactory or convincing. 



It is not within the scope of this brief summary to include a review of all earlier 

 work. Reference has been made to the early work of Bordet^ and Wassermann. ^ 

 Widal and Le Sourd^ and Camus and Pagniez^ might also be mentioned. 



Simple suspensions of the bacterial cell in physiological salt solution and crude 

 concentrations of the whole culture, such as tuberculin, were early abandoned for 

 autolyzed culture material or various extractions or residues after extraction. 01m- 

 stead and Luttinger^ found the most reliable antigen of the pertussis bacillus to be 

 an autolysate of an aqueous emulsion. Forty per cent of their cases reacted when in- 



' Wassermann, A. P.: Deutsche med. Wchnschr., 49, 303. 1923. 



2 Bordet, J., and Gengou, O.: Ann. de I'lnsL Pasteur, 20, 731. 1906. 



3 Bordet, J., and Gengou, O.: ibid., 15, 289. 1901; Compt. rend. Acad, de sc, 137, 351. 1903. 

 ''Wassermann, A. P., and Bruck, C: Deutsche med. Wchnschr., 32, 449. 1906. 



5 Widal, F., and Le Sourd, L.: Soc. med. dcshop., 18, 624. 1901. 



* Camus, J., and Pagniez, P.: Compt. rend. Soc. de biol., 53, 734. 1901. 



'Olmstead, M., and Luttinger, P.: Arch. Int. Med., 16, 67. 1915. 



