RELATION OF SERUM TO PHAGOCYTOSIS. 317 



ance. Examination shows an enormous degree of 

 phagocytosis. When, on the other hand, non-im- 

 mune rabbits are submitted to similar inocula- 

 tions, the fluid which accumulates locally is of a 

 clear serous character, contains few leucocytes, and 

 no phagocytosis is observable ; the animals die of a 

 rapidly developing septicemia. From the results 

 one may well suspect that the immunity is related 

 to and perhaps coextensive with the acquired pha- 

 gocytic power. 



But is the serum of no influence ? It has often Phagocytes 



Ttike Uil 



been held that phagocytes take up bacteria only virulent 

 after the latter have been injured or killed by the 

 serum or plasma. Metchnikoff answers this objec- 

 tion experimentally by inoculating an immune rab- 

 bit with anthrax, withdrawing some of the exu- 

 date at a time when phagocytosis is complete, and 

 injecting it into a non-immune rabbit. The sec- 

 ond animal dies. Since none but phagocytized 

 bacilli were injected into the non-immune rabbit 

 ( !), and since the latter succumbs to anthrax, it 

 seems not only unnecessary, but unjustifiable, to 

 assume that the bacteria must be attenuated by 

 the serum before they can be taken up by the leu- 

 cocytes. May the serum, nevertheless, have some The influence 

 obscure action which may not be included under 

 such terms as bactericidal and attenuating? It 

 seems fairly well established that anti-anthrax 

 serum, at least from certain animals, may exert a 

 protective influence when injected into other ani- 

 mals in conjunction with or in advance of the cul- 

 ture; yet Metchnikoff discredits the importance of 

 such protection and says that "those properties of 

 the body fluids, as the bactericidal, preventive and 



