COMPLEMENT FIXATION IN TUBERCULOSIS 323 



factory results. His tulase likewise confers immunity 

 when given either by the mouth or by the stomach. 



Theobald Smith x also concludes that vaccination of 

 calves with the human type of bacillus is harmless, and 

 that the procedure leads to a relatively high resistance to 

 fatal doses of the bovine bacillus. 



Clinical Examination 



I. The " complement-fixation " test was first used in tuberculosis 

 by Wassermann and Briick. The method has been further elaborated 

 by Emery. 2 He makes use of a standard emulsion of tubercle 

 bacilli in salt solution, containing about 4 per cent, by volume of 

 solid bacillary substance. This is sterilised by intermittent sterilisa- 

 tion and keeps for four to six weeks. Bacilli from various sources 

 vary somewhat, so that the emulsion should be standardised so as 

 to give an absorption-time with normal sera of about 20 minutes, 

 i.e. the complement of normal serum should be just completely 

 absorbed in about 20 minutes. A water-bath kept at a constant 

 temperature of 38 C. is used to warm all the constituents and mix- 

 tures. One part of the serum to be tested is mixed with four parts 

 of the bacillary emulsion in a small tube (e.g. a Durham's tube) in 

 the water-bath, the time of mixing being accurately noted. After 

 2^ minutes' incubation, 4 volumes of the mixture are removed by 

 means of a capillary pipette with teat (Fig. 35, p. 215), into which 

 also a single volume of sensitised corpuscles (i.e. a hsemolytic 

 system, p. 184) is taken up and the whole is expelled into a small 

 tube already standing in the water-bath. The process is repeated 

 after 5, 10, 15, and 20 minutes, and longer if necessary. By the 

 occurrence or absence of haemolysis in the various tubes, the time 

 taken for the absorption of complement is ascertained, the comple- 

 ment used being that contained in the serum itself, which therefore 

 should be fresh. A control with normal serum should always be 

 performed at the same time. With normal serum complete absorp- 

 tion should take place in about 20 minutes ; with tuberculous sera 

 it is often complete in 2| minutes. If, then, absorption of comple- 

 ment is complete in much less than the time necessary for absorption 

 with a normal serum, presumably the serum is derived from a 

 tuberculous individual. (But see Emery's paper for limitations.) 



1 Journ. Med. Research, vol. xviii, 1908, p. 451. 



2 Lancet, 1911, vol. i, p. 485. 



