278 Dr. J. O. W. Barratt. Fractional Withdrawal of [May 1 5, 



of the serum. The latter was then compared with the hemolytic power 

 of a number of admixtures of complement and amboceptor, in which the 

 content of complement was the same as in the serum tested, but the 

 amounts of amboceptor present formed an increasing series, the comparison 

 being made under similar conditions of experiment. The quantity of 

 amboceptor is represented by the amount contained in that admixture 

 which exhibited the same hsemolytic power as the serum under con- 

 sideration. 



Before determining the extent to which partial withdrawal of complement, 

 and amboceptor by subminimal amounts of antigen occurs, it was necessary 

 to study the effect of variations in the amounts of complement and 

 amboceptor in respect of hsemolytic power. An investigation involving 

 this inquiry has been carried out with the aid of a heteroamboceptor by 

 Kiss,* who determined the varying amounts of complement and amboceptor 

 which were just capable of producing complete haemolysis of the same 

 amount of red blood cells. This author observed that as the complement, 

 was increased in amount the quantity of amboceptor required became 

 diminished. In our own experiments, in which the employment of a 

 heteroamboceptor was avoided, both complement and amboceptor being; 

 derived, as already mentioned, from the same source, a, similar relationship 

 is also exhibited in a quantitative form. When a normal serum is. 

 employed the extent to which a disproportion of complement and 

 amboceptor can be produced is limited by the relative feebleness of the 

 serum. If, however, an immune serum is used, wider divergences in the 

 amount of complement and amboceptor employed may be obtained. 



If, in normal serum, fractional withdrawal of complement and amboceptor 

 by means of antigen is carried out, it is found that these two substances are 

 withdrawn in the proportion in which they naturally exist in the serum, so 

 that the liquid remaining after partial withdrawal has been effected is 

 comparable, as far as complement and amboceptor are concerned, to diluted 

 normal serum. 



If, instead of employing normal serum alone, a mixture of normal active 

 serum (complement + amboceptor) with inactivated normal serum (amboceptor), 

 is used, so that a liquid is obtained in which the concentration of amboceptor 

 is increased, while, at the same time, the concentration of complement i& 

 reduced, it is found that complement and amboceptor are at first removed by 

 subminimal amounts of antigen in approximately the same relative propor- 

 tion in which they exist in the mixed sera, the amount of amboceptor being,. 



* " Unterauchungen iiber die Fermentnatur des Komplementes," ' Zeitsohr. f. Immu- 

 nitatsforschung,' 1909, vol. 3, p. 558. 



