THE ANTIOEN-ANTIBODY REACTIONS 193 



involved in the phenomenon of bacteriolysis. One of these is present in any normal 

 serum, is inactivated by heating to 55° C. for 1 hour, and is not increased in 

 amount as the result of immunization. The other, if present in a normal serum, 

 is seldom found in any considerable amount. It appears, or increases greatly in 

 amount, in response to immunization with a particular bacterium, and is not in- 

 activated by heating at 55° C. He noted in the same paper that, when cholera 

 vibrios are subjected to the action of an immune serum, they often aggregate into 

 clumps before lysis occurs, a phenomenon which had, indeed, been casually recorded 

 by previous observers. 



In 1896 Gruber and Durham published the first detailed studies of this aggre- 

 gation, or agglutination, of bacteria. 



In 1897 Kraus made the observation that the addition of filtrates of cultures 

 of the plague bacillus, or of the cholera vibrio, to the serum of an immunized animal 

 led to the formation of a precipitate, and that this reaction, like the reactions of 

 bacteriolysis and agglutination, was specific ; that is, the filtrate of a cholera 

 culture reacted with its own, or homologous antiserum, but not with the plague 

 antiserum, and vice versa. 



In 1898 Bordet described the appearance of lytic antibodies in the serum of 

 animals which had been injected with the blood of an animal of some other species, 

 and showed that the lysis of the red corpuscles, like the lysis of bacteria, depended 

 upon the interaction of two distinct substances, the thermolabile, ferment-like body, 

 present in normal serum and a specific thermostable substance, developed as the 

 result of immunization, which he referred to as the sensitizing substance. The 

 thermolabile component, which Buchner named alexine, is now generally known 

 as complement. Bordet's paper had far-reaching results, for it substituted the 

 simple technical procedure of observing the lysis of red cells in the test-tube for 

 the more difficult process of determining the death of bacteria. The natural result 

 was that many investigators in this field turned their attention to the phenomenon 

 of heemolysis, often with the tacit assumption that the data obtained could be 

 transferred by analogy to the interaction of any serum with any bacterium, an 

 assumption which now appears to be largely unwarranted. 



In 1901 Bordet and Gengou showed that the absorption of complement by 

 sensitized bacteria could be demonstrated without relying on the observation of lysis, 

 or other change, in the bacteria themselves. Using many species of bacteria, they 

 found that, if a suspension of a given organism were mixed with the corresponding 

 immune serum in the presence of complement, the removal of complement from the 

 fluid could be demonstrated by the addition of sensitized red cells, which failed to 

 undergo haemolysis. In the following year Gengou (1902) showed that the presence 

 of organized material was unnecessary, and that the interaction of soluble proteins 

 with antisera prepared against them was associated with a similar absorption of 

 complement. 



Denys and Leclef, in the course of experiments on antistreptococcal immunity, 

 which were recorded in 1895, showed that the phagocytosis of bacteria by leucocytes 

 was promoted by an action of the blood serum exerted on these bacteria, and 

 Mennes (1897) confirmed and extended their observations. For some cause, which 

 is difi&cult to understand in view of the inherent importance of these observations, 

 little attention was paid to them by other workers during the years which immedi- 

 ately followed. In 1902 Leishman again, and independently, demonstrated the 

 efiect of serum in stimulating phagocytosis, but did not analyse the mechanism 

 P.B. H 



