DEVIATION OF COMPLEMENT 147 



is no distinct deposit. It will be seen from Table 3 that 

 with this anti-serum, which produces no precipitate on 

 addition to the homologous serum, the phenomenon of 

 deviation is produced by exceedingly small amounts of the 

 latter, viz. 0-000,01 c.c. This shows that the injection of 

 the guinea-pig's serum into the rabbit gives rise to anti- 

 substances, although the latter cannot be demonstrated by 

 the phenomenon of precipitation. We also found in the 

 course of immunizing a rabbit against human serum that 

 the deviating power appeared before any precipitating 

 action was detectable. We may state, however, that when 

 a precipitate is formed we have always found that it had 

 the property of fixing complement the test of course being 

 made after the precipitate had been washed and centri- 

 fugalized several times. In fact the use of a separated 

 precipitate supplies a very convenient method for freeing 

 a serum of complement. We have, for example, been able 

 by this means to deprive guinea-pig's complement of practi- 

 cally all haemolytic power for ox's corpuscles treated with 

 immune-body, though the hsemolytic dose of such a com- 

 plement for 1 c.c. of suspension of red corpuscles may be 

 as low as 0-01 c.c. That the precipitate when formed, or 

 rather molecules in the precipitate, have the power of fixing 

 complement, there is therefore no doubt. Are there any 

 molecules with this property in the separated fluid ? To 

 each of a series of tubes 0-025 c.c. anti-serum rabbit v. ox 

 and 0-001 c.c. of ox's serum were added, the mixture being 

 made up to 1 c.c. in each tube with salt solution. Another 

 similar series was prepared with 0-01 c.c. ox's serum instead 

 of 0-001 c.c. The tubes were incubated for two hours and 

 then placed in the refrigerator till next morning. The 

 tubes were then centrifugalized and the supernatant fluid 

 was carefully pipetted off from the precipitate in each tube. 

 To the fluid thus obtained different doses of guinea-pig's 

 complement were added and after incubation deviation was 



L 2 



