476 THE TECHNIC OF COMPLEMENT-FIXATION REACTIONS 



a much stronger reaction than one would expect in testing human serums. 

 Further than this, the antibody content in antiserums varies considerably 

 so that the antigenic unit fluctuates according to the particular serum 

 used in making the titration. In general, therefore, it is sufficient to 

 determine the anticomplementary dose and to use half or quarter this 

 amount in performing the main test. 



After antigens are prepared they may require further dilution with 

 saline solution. This can be determined only by experience and as the 

 result of a trial titration. 



As watery extracts are prone to deteriorate, it should be made a rule 

 that the anticomplementary dose be determined each time before the main 

 test is conducted. 



It has quite generally been proved that alcoholic extracts of bacteria 

 do not yield satisfactory antigens, despite the advantage to be gained 

 because of their stability. 



Principles of Complement Fixation with Bacterial Antigens. The 

 principles of complement fixation in general should be thoroughly under- 

 stood. 



After making considerable comparative studies with various methods, 

 I am convinced that, in the final analysis, a simple technic is best. I 

 use a relatively small but safe dose of complement, 1 c.c. of a 1 : 20 

 dilution ( = 0.05 c.c. undiluted serum), and titrate the hemolytic 

 amboceptor with this constant dose or unit of complement and the 

 corpuscle suspension. This titration is made each time the reactions 

 are performed, and with each and every complement serum and cor- 

 puscle suspension. In this way differences in the activity of different 

 guinea-pig serums are readily detected and adjusted in the amboceptor 

 titration. If exactly one unit of amboceptor is used in conducting the 

 main test, the controls are not likely to be completely hemolyzed unless 

 the serum contains natural antisheep amboceptor, for the serum alone 

 will probably be slightly anticomplementary. For this reason I am 

 accustomed to use 1J^ doses of amboceptor, and I secure reactions that 

 are very delicate and yet sharp and clear cut. When testing an old 

 serum, one that is likely to contain considerable thermostabile anti- 

 complementary bodies, I use two hemolytic doses of amboceptor. If 

 one wishes to use exactly one unit of complement and one unit of ambo- 

 ceptor, the serum and antigen alone should be controlled, as in the fourth 

 method of performing the Wassermann reaction (p. 446). I frequently 

 use this technic in the gonococcus-fixation test, but as a rule I have 

 found the simpler technic herein given equally sensitive and reliable. 



