191 L] Complement and Amboceptor by Means of Antigen. 279 



relatively to the complement, removed in excess. If this relationship is 

 studied more closely it is seen that the actual quantity of amboceptor 

 removed is greater than that withdrawn by the same amount of antigen 

 from a similar concentration of amboceptor afforded by active normal serum, 

 while the actual quantity of complement removed is less than would be 

 withdrawn had active normal serum been employed. 



If normal serum from which amboceptor has been removed by means of 

 antigen acting at 0° C. is added to inactivated serum (amboceptor), so that 

 a mixture is produced in which complement is in excess, it is found that 

 fractional withdrawal of complement and amboceptor proceeds in a similar 

 manner, relatively more complement than amboceptor being removed. If 

 the actual quantities withdrawn are considered, it is found that the amount 

 of complement taken up by a given amount of antigen is greater than, and 

 the amount of amboceptor less than, would have been removed had active 

 normal serum been employed to produce the same concentration of 

 amboceptor. 



The mode of fractional withdrawal of complement and amboceptor by 

 antigen is not such as would be expected if a simple stoichiometric relation 

 between these three substances existed. Whatever the ultimate nature of 

 the reaction in question may be, in the early stages of the process, at any 

 rate, no fixed relationship between the quantities interacting exists. 



When fractional removal of complement and amboceptor takes place in an 

 artificial admixture of these two substances, the relation between the com- 

 plement and amboceptor removed approximates at first to that obtaining in 

 the admixture, provided the amounts removed are small. If relatively large 

 amounts of complement and amboceptor are removed by antigen, the 

 divergence between the two, though still present, tends to become less 

 marked. In consequence, the difference between the complement and 

 amboceptor still remaining becomes less pronounced, and, as withdrawal 

 proceeds, tends to approach the relationship obtaining in normal serum. By 

 using in these experiments complement and amboceptor contained in the 

 same serum and thus avoiding the employment, of a hetero-complement (by 

 which is meant a complement which is heterologous in respect of the 

 amboceptor employed), it becomes possible to compare the relation of 

 complement and amboceptor in the liquid remaining after fractional 

 withdrawal of these substances from an admixture with that obtaining in 

 normal serum. It is seen that, as withdrawal proceeds, the liquid remaining 

 tends to become more and more closely comparable, in respect of complement 

 and amboceptor, with diluted normal serum. 



VOL. LXXXIV. — B. 



Y 



