120 THE SIDE-CHAIN THEORY 



meets by throwing off the unnecessary number of receptors as such. 

 These cast-off receptors will, of course, have the same combining 

 groups as the sessile ones, which had originally anchored the toxin 

 molecule, and it stands to reason that if the toxin molecule and the 

 corresponding free receptor are brought together either within or 

 outside of the body the two will unite, the result being indicated by 

 absence of toxicity on the part of the mixture. As this is exactly 

 what happens when toxin and the serum of a correspondingly 

 immunized animal are brought together, Ehrlich very properly 

 concludes that the antitoxic properties of an immune serum are due 

 to the presence of free receptors which are "tuned" to the toxin in 

 question; in other words, that the antitoxin is not newly formed 

 in the body, but is identical with those receptors of the cell which 

 render the attack of the toxin upon the cell possible. 



While this conception of the nature, production, and mode of 

 action of the antitoxins originally had reference to these only, 

 subsequent observations led Ehrlich to extend his theory to the other 

 antibodies as well. But in accordance with the facts observed it is 

 necessary to assume that the structure of the other antibodies, viz., 

 those receptors which enter into relationship with such antigens as 

 the agglutinable substance of bacteria, the precipitable complex 

 of albumins and those cellular constituents w r hich give rise to lysin 

 formation, must be different from that of the antitoxic receptors. 

 For whereas the antitoxic antibodies merely combine with the toxins 

 to form non-toxic components, the other antibodies not only fix the 

 corresponding antigens, but bring about further changes. 



Ehrlich very appropriately remarks that the mere fixation of 

 certain food molecules would not suffice to render them available for 

 purposes of nutrition, but that with molecules of large size their 

 destruction must precede assimilation. This could be effected, if 

 the receptor in question had not only a haptophoric group "tuned" 

 to the combining group of the food molecule, but, in addition, either 

 a second group of ferment character as part and parcel of the same 

 receptor, or a second haptophoric group which might anchor ferment 

 molecules, normally occurring free in the blood, when the other 

 combining group is occupied by a food molecule of a certain structure. 



Experimental investigation has shown that receptors of both types 

 actually exist, and we may accordingly conclude that the antigens 

 in question, like the toxins, do not represent true foodstuffs, but 



