488 BACTEEIOLOGY. 



to neutralize the toxin in much the same way that an acid 

 and alkali unite and neutralize one another. While this is 

 a convenient way of explaining the action of antitoxin^ it i& 

 probably no more correct than it would be to assume that 

 the anti-infectious serum is germicidal. Antitoxic serum,- 

 like anti-infectious serum, does not act directly but it stimu- 

 lates the cellular defenses of the body to activity. The cells 

 of the body by preliminary training, such as repeated 

 poisoning with a toxin, can take up such poison or give off 

 products which will destroy it. If, as in the cholera ex- 

 periment above, living cells are injected then the phagocytes 

 are brought into action either directly, or by their germi- 

 cidal products. In either case immunity always depends, 

 directly or indirectly, upon the cellular elements of the 

 body. 



Pfeiffer's phenomenon is not limited to the cholera 

 vibrio but will be met with, under like conditions, when thie 

 Eberth bacillus or other organism is employed for immuniza:- 

 tion. It is of great value as a means of differentiating the 

 cholera vibrio from other similar vibrios; or the Eberth 

 bacillus from the colon and other Eberth-like bacilli. The 

 reaction is specific. That is to say, the anti-infectious cholera 

 serum will cause the destruction of cholera germs inside 

 the peritoneal cavity of an animal. It will not protect 

 against other vibrios or against the Eberth bacillus. Simi- 

 larly, the anti-infectious typhoid serum will protect against 

 the Eberth bacillus but not against the colon or pseudo- 

 typhoid bacteria. 



In employing this test it should be remembered that 

 normal serum may protect, in 1 c.c. dose, against the 

 minimal fatal dose of the cholera or typhoid bacteria. 

 Hence, several times the fatal dose should be employed and 

 a control test should be made with a like amount of normal 

 serum. The serum of convalescents from typhoid fever is 

 not antitoxic but is anti-infectious and can, therefore, be 



