166 A MANUAL OF BACTERIOLOGY 



Whereas on Ehrlich's views the combination of toxin and anti- 

 toxin would be represented by a straight line, and the crude toxin 

 seems to be composed of a whole series of different toxins and sub- 

 stances having an avidity for antitoxin, on this hypothesis, although 

 the greater part of the toxicity of toxin is removed by the anti- 

 toxin, the latter must be added in large excess before the toxicity 

 completely disappears, and the course of neutralisation would be 

 represented by a hyperbolic curve. In fact, as the antitoxin is 

 added, the amount of free toxin diminishes but never completely 

 disappears. There comes a point, of course, when the amount of 

 free toxin is so small as to be negligible and cannot be recognised 

 by the ordinary indicators (blood-corpuscles, animal tests, etc.). 

 This hypothesis would explain the fact that while a certain amount, 

 V, of a mixture of toxin and antitoxin is innocuous to an animal, a 

 multiple of the dose, n V, of the same mixture may be toxic ; it 

 would also explain Buchner's experiments alluded to above (p. 161), 

 and Roux's experiments in which a toxin-antitoxin mixture in- 

 nocuous to normal guinea-pigs was toxic to guinea-pigs whose 

 resistance had been reduced by injections of the Massowah vibrio. 



Nernst has questioned from the mathematical standpoint the 

 validity of the views of Arrhenius, and so has Craw from much 

 experimental work on agglutination and on the interaction between 

 megateriolysin and anti-megateriolysin ; Craw also considers that 

 there is some doubt attaching to Arrhenius's calculations. Accord- 

 ing to Craw, the two substances most thoroughly investigated by 

 Arrhenius and Madsen, diphtheria toxin and tetanolysin, do not 

 admit of sufficiently exact determination, the former because of the 

 uncertainty attaching to animal experiments, the latter because 

 tetanolysin is a most unstable body. Working with a more stable 

 substance, megateriolysin, he holds that the Arrhenius and Madsen 

 equation does not apply. Again, on the addition of a small amount 

 of antitoxin to toxin there is no decrease in toxicity (as noted by 

 Ehrlich and attributed by him to the presence of toxoid) as there 

 should be, and Arrhenius was thus forced to the conclusion that a 

 second substance, epitoxonoid, is present with the toxin in diphtheria 

 toxin. Craw denies that the toxin-antitoxin reaction is reversible, 

 believes that antitoxin must be regarded as a colloid (and is not in 

 true solution), that the mixture therefore is heterogeneous, not 

 homogeneous, and that the chemical law of mass action is not 

 applicable. 



On the other hand, Craw maintains that the phenomena of the 

 toxin-antitoxin reaction, including the Danysz effect, have their 

 counterpart in adsorption phenomena, such as occur in the staining 



