Experimental Investigation of the Problems of Immunity 107 



added to goats' or any other milk. If, however, the rabbit had been 

 repeatedly injected with goats' milk or human milk, its serum would 

 precipitate with those milks respectively, and not with cow's milk. 

 The reaction was thus shown to be specific. 



Myers* found that the repeated intraperitoneal injection of 

 egg-albumen into rabbits caused their serum to give a dense pre- 

 cipitate when added to solutions of egg-albumen. 



Tchistowitcht found that eels' serum injected into animals 

 produced a reaction in which immunity to its poisonous action was 

 associated with the ability of their serum to produce a precipitate 

 when added to the eels' serum. 



Closely connected with these various reactions are certain others 

 variously spoken of as cytotoxic, cytolytic, hemolytic, bacteriolytic, 

 etc. The first observation bearing upon these was made by R. 

 Pfeiffer,{ who found that when guinea-pigs received frequent 

 intraperitoneal injections of cholera spirilla and became thoroughly 

 immunized, their serum behaved very pecuharly toward the bacteria 

 in the peritoneal cavity of freshly infected animals, in that it caused 

 them to become aggregated into granular masses and subsequently 

 to disappear. This became known as "Pfeiffer's phenomenon." 

 The serum of the immunized animal was devoid of action by itself, 

 the serum of the infected animal was inactive, but the combination 

 of the two brought about dissolution of the micro-organisms. Later 

 it was shown by Metchnikoff|| that the living animal was not a fac- 

 tor in the process, but that what was seen in the peritoneal cavity 

 could be reproduced in a test-tube, though not quite as well. 



Bordet§ made frequent injections of defibrinated rabbit's blood 

 into guinea-pigs, and obtained a serum that had a solvent action 

 upon the rabbit's corpuscles in viiro, and showed that the induced 

 hemolysis resembled in all points the bacteriolysis. 



EhrKch** and Morgenroth studied the hemolytic action of the 

 serum of goats that had been frequently injected with the de- 

 fibrinated blood of sheep and goats, and were able to point out the 

 mechanism of the corpuscle solution or hemolysis. It was found 

 to depend upon two associated factors, one of which, the lysin or 

 solvent; was present in normal blood, and was caUed "addiment" 

 or "complement," and another present only in the serum of the 

 reactive animals, called the "immune body" or "intermediate/body." 

 The former was labile and easily destroyed by heat, the latter 

 stabile and not affected by heat up to the point of coagulation. The 

 experiments were confirmed by von Diingern and many others. 

 It is to be observed, in passing, that this reaction differs from the 



* "Lancet," igoo, 11. . 



t "Ann. de I'lnst. Pasteur," vol. xin, 406. 

 % "Deutsche nied. Wochenschrift," 1896, No. 7. 

 § "Ann. de I'Inst. Pasteur," 1895. 

 II Ibid., 1898, XII. 

 ** "Berliner klin. Wochenschrift," 1899. 



