THE SIDE CHAIN THEORY 103 



Evidently the antitoxin acts directly upon the toxin, and in inter- 

 preting the findings in vitro, different possibilities again arise. It is 

 thus conceivable that the antitoxin may destroy the toxin. This 

 view, however, has also been disproved by a number of observations. 

 Buchner thus showed that a mixture of tetanus toxin and antitoxin 

 which was "neutral" for mice was still toxic for guinea-pigs. Roux 

 and Calmette further ascertained that a mixture of snake venom and 

 antivenin which was non-toxic for a given animal, became toxic again 

 on heating. This would evidently be out of the question if the non- 

 toxicity of the mixture had been owing to a destruction of the toxin 

 by the antitoxin. Martin and Cherry then pointed out that this 

 result is obtained only if not too long an interval has elapsed 

 after bringing toxin and antitoxin together, and that a restitution 

 of the toxic effect is no longer possible if a certain time limit has 

 been passed. They could show, as a matter of fact, that toxin and 

 antitoxin do not interact instantaneously, and that at first toxin and 

 antitoxin coexist in the free state, the velocity of reaction depending 

 very largely upon the concentration of the two solutions and the 

 temperature. 



In the light of these findings Roux and Calmette.'s original observa- 

 tions do not disprove the idea that the antitoxin destroys the toxin. 

 That this actually does not occur was, however, conclusively shown 

 by Morgenroth in Ehrlich's laboratory; for on treating a mixture of 

 Cobra toxin and antitoxin that had been kept for more than a week, 

 L e., for a period of time that was more than sufficient completely to 

 destroy any toxic effect by the antitoxin, with dilute acids, and on 

 then heating the acid mixture for some time at 100 C. the original 

 quantity of toxin was again obtained in its entirety. A destruction 

 of the toxin by the antitoxin had thus been satisfactorily disproved. 



Evidence such as this is, of course, strongly suggestive that the 

 inactivation of the toxin by the antitoxin is due to the occurrence 

 of a chemical interaction between the two, and that the specific 

 effect of the toxin disappears because the substance is chemically 

 bound by the antitoxin. This view, which was first expressed by 

 Ehrlich, is the one now generally held and forms the basis of our 

 modern conception of the production of antibodies and their specific 

 effect. 



Further studies have shown, as a matter of fact, that the same 

 principle applies to the interaction between other antigens and their 



