COMPLEMENTOPHILE GROUPS OF THE AMBOCEPTORS. 229 



present in normal guinea-pig serum. This serum reactivates two 

 immune bodies, of which one, immune body A, was obtained by treat- 

 ing rabbits with ox blood, and the other, immune body B, by treat- 

 ing a goat with sheep blood. These immune bodies, naturally, acted 

 respectively on ox blood-cells and sheep blood-cells. This anti- 

 complement is strongly active in case A, while it is entirely without 

 effect in case B. From this we may conclude that the complements 

 concerned in these two cases, and which we may designate as x and 

 6, are unlike. 



A further question was whether immune body A binds other com- 

 plements in guinea-pig serum besides its own dominant complement. 

 In order to determine this the following experiment was made : First, 

 ox blood-cells and sheep blood-cells were saturated with their re- 

 spective amboceptors A and B, and then to each cubic centimeter of 

 the 5% blood suspension varying amounts of guinea-pig serum were 

 added as complement. In the first case 0.0075 cc. guinea-pig serum 

 sufficed to cause complete solution; in the second case 0.005 cc. 

 was required. 



Thereupon another test was made exactly like the preceding with 

 ox blood and immune body A. After the mixture had remained in 

 the thermostat at 37 C. for H hours and haemolysis was practically 

 completed, the same quantity of ox blood-cells laden with immune 

 body (0.05 cc. ox blood freed from serum and made up to the original 

 volume) was added anew and the mixtures kept in the thermostat 

 for two hours longer. The haemolysis which had then taken place, 

 observed by allowing the mixture to sediment in a refrigerator, indi- 

 cated the amount of complement x left after the first haemolysis and 

 available for the case A. 



At the same time a similar experiment was made in which, 

 after the first haemolysis, sheep blood-cells saturated with ambo- 

 ceptor were used in the place of the ox blood-cells. In this case, 

 after determining the amount of complement originally present, 

 that of complement 6, left after the first haemolysis, could also be 

 found. 



In this a considerable loss of complement is observed for both 

 cases; for it now requires 0.075 cc. of the complementing guinea- 

 pig serum to cause complete solution for case A and 0.025 cc. for 

 case B, so that 1/10 and 1/5 respectively of the original complement 

 are still preserved. This shows that the binding of complement a, 

 dominant for case A, is accompanied by a binding of complement/?, 



