34 TOXINES AND ANTITOXIN ES. 



used at the beginning, it is extremely probable that symptoms 

 of poisoning will appear on multiplying the respective doses. For 

 it is quite possible, when a single lethal dose is neutralised, for 

 a slight excess of poison to remain unnoticed in the mixture, 

 since it may not even reach the minimum amount necessary to 

 cause illness ; but if now the relative quantities of poison and 

 antitoxine be increased tenfold, this excess of poison is also 

 multiplied by ten, and the poisonous action of the mixture is 

 manifested. Ehrlich's theory cannot be upset by such proofs 

 as these. 



Taking all things into consideration, we are justified in 

 assuming, on practical and theoretical grounds, that the action 

 of an antitoxine upon a toxine consists essentially of a reciprocal 

 combination of two groups endowed with specific affinity. 



This at once leads to the fundamental conclusion that the 

 reciprocal action of the two substances must follow the laws 

 that hold good for the reciprocal saturation of two simple 

 chemical substances possessing atomic groupings specifically 

 adapted to each other i.e., that they must combine in fixed 

 quantitative proportions. Just as the same amount of pure 

 sodium hydroxide invariably requires the same amount of 

 pure hydrochloric acid for neutralisation, so must the propor- 

 tion between a definite dose of toxine and the amount of 

 antitoxine that exactly " neutralises " it be an absolute constant. 

 A given quantity of pure toxine must invariably require the 

 same dose of pure antitoxine for its activity to be exactly 

 neutralised, provided that the combination is stable, and does 

 not lead to a state of dissociated equilibrium, a question with 

 which we shall deal later. 



The difficulty of establishing this extremely important fact 

 is enormously increased by the circumstances. In the first 

 place, neither toxines nor antitoxines are known in the free 

 state. We are not dealing here with substances that can 

 be isolated as chemical entities, and to which the balance can 

 be applied in determining whether x grms. of diphtheria anti- 

 toxine invariably neutralise y grms. of diphtheria toxine. No ; 

 the only measurement that is applicable to these poisonous sub- 

 stances is the physiological one the determination of the "single 

 lethal dose" which we are compelled to adopt as the fundamental 

 unit of measurement for toxines, or, in the case of hcemolysines, 

 the measurement of the amount of solvent action. 



This drawback, however, would not be so very serious if we 

 could only establish a constant relationship between every toxine 

 solution of definite strength and a given antitoxine solution, so 



