326 INFECTION AND IMMUNITY 



complement by the addition of hemolytic amboceptor and red blood 

 cells. In testing for antigen, the procedure is reversed, in that the 

 serum or other substance (bacterial extract) to be tested is brought 

 into contact with an antibody specific for the antigen, in the presence 

 of complement; and at the end of an hour at suitable temperature, 

 free complement is again determined by hemolytic reaction as before. 



When dealing with bacterial antigen, it is necessary, therefore, to 

 prepare a highly potent immune serum against the bacteria which con- 

 tain the specific antigen which is sought. 



Thus in testing for typhoid-bacillus antigen in the serum of a 

 patient, the substances required are as follows : 



1. Complement: obtained from fresh guinea-pig serum. It is 

 best to titrate the complement when possible, using for the test double 

 the quantity necessary to produce complete hemolysis of 1 c.c. of a 

 five per cent emulsion of blood cells, in the presence of two units of 

 amboceptor. 



2. Hemolytic amboceptor: rabbit serum hemolytic for sheep cor- 

 puscles. Inactivated and titrated as for Wassermann test. 



3. A five-per-cent emulsion of sheep corpuscles in salt solution, 

 prepared as for Wassermann test. 



4. A highly potent typhoid antiserum obtained from an immunized 

 rabbit. In this case the smallest quantity of the immune serum which 

 will cause the fixation of complement in the presence of an emulsion 

 or extract of known typhoid bacilli is determined by experiment. 



It is best to rub up the centrifugalized bacteria with dry salt, 

 adding distilled water to isotonicity. When the extract is made, its 

 anticomplementary dose is determined and the minimum quantity 

 which in the presence of known antityphoid serum will fix complement. 

 These preliminary titrations are analogous to those described as pre- 

 liminary to the Wassermann test. When these quantities have been 

 determined, an amount of the bacillary extract (about 1-3 or 1-4 of 

 the anticomplementary dose) is chosen for the actual tests. It is well, 

 also, in such tests to determine the amount which will fix complement 

 in the presence of a known normal serum, since occasional presence of 

 antibodies against some bacteria in normal serum may otherwise con- 

 fuse the test. The quantities of antigen and complement must then be 

 chosen for the test in such proportions that no fixation will occur with 

 normal serum. 



5. Serum from the patient, inactivated at 56 C. for twenty 

 minutes. 



