1907.] Substance in Serum which influences Phagocytosis. 409 



leucocytes, bacilli, etc., and allowed to act during 15 or 20 minutes. The 

 further carrying out of the experiments was as follows : — 



(A) Normal rabbit serum, heated to 60° C. for 20 minutes, 10 volumes, + 

 Fresh normal horse serum, 5 volumes. 



(At 35°'5 C. for two hours.) 



(B) Anti-complement (horse) serum of rabbit, heated to 60° C. for 



20 minutes, 10 volumes, + 

 Fresh normal horse serum, 5 volumes. 



- (At 35°-5 C. for two hours.) 



•(C) Normal rabbit serum, heated to 60° C. for 20 minutes, 10 volumes, + 



Normal horse serum, heated to 60° C. for 20 minutes, 5 volumes. 



(At 35°-5 C. for two hours.) 



{D) Anti-complement (horse) serum of rabbit,, heated to 60° C. for 

 20 minutes, 10 volumes, + 

 Normal horse serum, heated to 60° C. for 20 minutes, 5 volumes. 

 (At 35°-5 C. for two hours.) 



The supernatant fluid from each of the above mixtures A, B, C, and D, was 

 subsequently mixed with an equal volume of human leucocytes and staphylo- 

 coccic emulsion and kept at 37° C. for 15 minutes. 



When horse serum had been injected into the rabbit to obtain the anti- 

 complement serum, it was found impossible to use horse leucocytes in the 

 experiment, on account of the rapid agglutination of the red cells and 

 -entanglement of the leucocytes. In this case, therefore, human leucocytes 

 were used, and by so doing the possible danger of a certain amount of 

 .•specific leuco-toxic action was avoided. 



The results of seven experiments are detailed in Table V (p. 410). 



It is seen that a great loss in the sensitising power has resulted in 

 Experiment 2, where the anti-complement serum was used. The higher 

 •count of Experiment 4, as compared with Experiment 3, is probably largely 

 due to the fact that the heated anti-complement serum had a higher 

 sensitising power than the normal rabbit serum employed, the ratio being 

 about 2:1. This last observation applies to all the corresponding figures in 

 the table, which summarises the results of seven such experiments, with 

 a varying proportion of anti-complement and normal sera versus fresh horse 

 serum. On comparing the second and fourth columns of that table, it will be 

 seen that fresh normal serum, after having been acted on by the anti- 

 complement serum, gives almost the same phagocytic average as is 



