jan. 2.1917 Diagnosis of Tuberculosis by Complement Fixation 5 
negative reactions, of which 48 were found positive and 48 negative on 
post-mortem examination. Exception was taken to his findings by 
Bierbaum and Berdel (5), who concluded that alcoholic or aceton extracts 
of tuberculous tissues as antigens do not confirm the results of Hammer. 
Of all antigens the greatest support was given to the one prepared by 
Besredka (3, 4). It is essentially a tuberculin prepared in glycerin and 
peptone-free egg medium. Many investigators reported favorably on the 
action of this antigen, and the only objection against it was its nonspecific 
reaction obtained in cases of nontuberculous syphilitic individuals. 
Bronfenbrenner and Rockman (8) concluded with regard to Besredka’s 
antigen that different samples of his tuberculin, though prepared appar¬ 
ently in an identical manner, may differ in their specific values. Thus, 
different samples were found to vary in the amount of lipins they con¬ 
tained. It is necessary to free the samples of tuberculin entirely from 
the lipins present when used for the complement deviation test. Bes- 
redka’s antigen seems to give the best results in the diagnosis of tuber¬ 
culosis by means of complement fixation, and though the test is positive 
in a certain number of nontuberculous cases the reaction seems to be 
specific. However, some investigators experienced difficulties in prepar¬ 
ing the antigen according to the directions given by Prof. Besredka. 
More recently Stimson (35) carried on extensive investigations on 
this subject and employed an alcoholic precipitate of the tubercle bacillus 
emulsion filtrate as antigen, obtaining promising results. Favorable 
results were further reported by Craig (14) with an antigen prepared from 
a liquid egg-medium growth material, and by Miller and Zinsser (28) with 
a bacterial emulsion in hypotonic salt solution. These authors claim 
good results, especially for differentiation between active and nonactive 
infections. 
The foregoing historical outline furnishes ample evidence that strong 
efforts were made toward devising a method by which the complement 
fixation might be utilized for the diagnosis of tuberculosis. 
ANTIGENS 
In the foregoing literature mention was made of the various antigens 
which were used for the diagnosis of tuberculosis by means of comple¬ 
ment fixation. These antigens may be classified into three groups, 
according to their contents: 
1. Tuberculin. 
2. Emulsion of tubercle bacilli. 
3. Extracts of tuberculous tissues. 
In consideration of the failures of former investigators with various 
antigens containing the specific products of the germ and of the disease, 
it suggested itself that an antigen containing the toxic products together 
with bacteria would probably combine the active elements for the phe¬ 
nomena of the reaction. Therefore experiments were undertaken with 
