244 MEDICAL BACTERIOLOGY 



than four times the mean antigenic quantity it must be rejected as unfit for 

 use. 



Titration shows that different quantities of an extract are antigenic, e.g., 

 from o.oi to i.o cc. or from o.i to 0.5 cc., that is, in the first case, any quantity 

 of the extract from o.oi to i.o cc. would cause complement fixation with a syphi- 

 litic serum and have no effect on the complement with non-syphilitic sera; in 

 the second case only quantities of extract from o.i to 0.5 cc. would do this. 

 In the first case the mean antigenic quantity would be 0.5 cc.; in the second case 

 the mean antigenic quantity would be 0.3 cc. 



The unit of antigen, the amount used in making Wassermann tests, is the 

 mean antigenic quantity. It is determined for each new antigen before use, 

 and when in use antigen should be restandardized at least every month. The 

 method of standardization will be described later. 



CONTROLS 



The end result and accuracy of any complement fixation test depends upon 

 the purity and potency of the different substances used, their proper combina- 

 tion and quantities, the time and temperature of incubation. With so many 

 sources of possible error it is imperative that each be controlled. Nine control 

 tubes should be carried through with every Wassermann test and these tubes 

 should always have the same positions in the test-tube rack. The author 

 places them as illustrated in Fig. i . Three of these tubes are used for making a 

 Wassermann test, using a human serum that was obtained from a patient, who, 

 so far as medical skill could determine, never had syphilis, a serum that pre- 

 viously showed a negative Wassermann reaction, a serum that does not in any 

 way act on guinea-pig serum, rabbit serum, syphilitic antigen or red cells. 



Three other controls are used for making a Wassermann test, using human 

 serum that was obtained from a patient that presented typical positive clinical 

 signs of syphilis, a serum that previously showed a positive Wassermann reac- 

 tion, a serum that does not in any way act on guinea-pig serum, rabbit serum, 

 syphilitic antigen or red cells, except that it fixes the complement in guinea-pig 

 serum together with syphilitic antigen. 



When a known positive serum gives a positive reaction, and a known nega- 

 tive serum gives a negative reaction, with the same reagents, in the^ame quan- 

 tities, and under the same conditions as the unknown sera are subjected to, it 

 is the strongest evidence that the tests have been properly performed and the 

 results reliable for this no other controls are necessary but sometimes a test 

 goes wrong, and these controls are not always sufficient to indicate the source 

 of error; for this purpose three other controls are made: one for complement, one 

 for hemolytic system and one for syphilitic antigen (Fig. i). 



Known positive and known negative sera, procured under aseptic conditions 

 and kept sterile, if stored in a refrigerator, in all but very exceptional cases re- 

 main fit for use as controls for several weeks or months. They must be inacti- 

 vated immediately before each time they are used; if this is not done they are 

 not reliable. 



