Complement Fixation in Tuberculosis. 135 



of Medical Sciences, the writers described work on complement- 

 fixation in tuberculosis, carried out with a very simple antigen 

 which had yielded and is still yielding results more satisfactory 

 than those hitherto reported by other workers who had used other 

 antigens. The work followed a study of culture-filtrate antigens, 

 such as those devised by Besredka and by Petroff , and the special 

 modifications of the Besredka medium employed by Bronfen- 

 Brenner and by Craig. These antigens did not in our hands react 

 with the regularity which we thought should attend a reaction of 

 specific diagnostic value. Owing to irregularities perhaps due to 

 constituents of the media, it was thought wise to return to the 

 bacillary substances themselves, work along this line having been 

 attended by considerable success within recent years — notably in 

 the hands of Radcliffe, Dudgeon, Weir, and Stimson. It should not 

 be forgotten that the same direction of investigation was followed 

 in the earlier work of Wassermann and Citron and in that of Cal- 

 mette. 



The method employed is in general identical with that which 

 we have been using in this laboratory for the extraction of Tre- 

 ponema pallidum, typhoid bacilli and streptococci, and differs in 

 no essential particular from the so-called "endotoxin" extraction 

 method employed by Besredka in 1906 with organisms of the 

 typhoid-colon group. Since we feel that the procedure at present 

 in use in the Columbia laboratory should be thoroughly rein- 

 vestigated by other workers, we believe that it is proper to give 

 in great detail the method by which the antigen is made. 



The bacilli which, so far, have been used for the production 

 of the antigen have been of the human type, some of them isolated 

 by Miller, some of them obtained from Professor Theobald Smith, 

 some from the laboratory of Professor William H. Park, and some 

 from the laboratory of Parke Davis & Co. They have been grown 

 mainly on the gentian-violet medium of Petroff and on Miller's 

 modification of this medium; also on Petroff's potato broth. It is 

 at present the impression of the writers that the medium on which 

 the bacilli are grown plays no great part in determining the use- 

 fulness of the antigen. It seems, however, to be important that a 

 number of different strains should be used — that is, that the 

 antigen should be polyvalent — and the use of relatively young 



