MODE OF PRODUCTION OF ANTITOXINS 473 



more ammonia or more boracic acid were added, part would 

 remain free while part would combine. Accordingly, if toxin 

 and antitoxin behaved in a similar manner an explanation of the 

 Ehrlich phenomenon would be afforded. Madsen and Arrhenius 

 have worked out the question in the case of a great many toxins 

 and find that the graphic representation of neutralisation is in 

 every case a curve which can be represented by a formula. It 

 should be noted in connection with this controversy that there 

 are two questions which may be independent of each other, viz. 

 (1) does the "toxin" in any particular case represent a single 

 substance or several 1 (2) What is the nature of the combination 

 of any one constituent substance and its anti-substance is it 

 reversible or is it not 1 It may be said that it is practically 

 impossible to explain the facts with regard to diphtheria toxin 

 on the hypothesis of a single substance even if this should have 

 its combining and toxic actions equally weakened ; " toxoids " in 

 Ehrlich's sense must in our opinion be supposed. Then there is 

 an important fact established by Danysz and by v. Dungern, 

 namely that the amount of toxin neutralisable by a given amount 

 of antitoxin is different according as the toxin is added in several 

 moieties or all at once in the latter case the amount of toxin 

 neutralisable is greater. There seems no explanation of this 

 according to the view of Madsen and Arrhenius as the same state 

 of equilibrium ought to be reached in the two cases, that is, the 

 amounts of toxin neutralised should be the same. On the other 

 hand we have instances of the combination of a substance and its 

 anti-substance being reversible the example of a hsemolytic 

 immune-body may be cited (p. 481) and there is no doubt that 

 there are varying degrees of firmness of the union. It is quite 

 evident that if there should be several toxic bodies in a "toxin," 

 and that if the union of some of these with antitoxin should be 

 reversible, the problem becomes one of extreme complexity. 



There has recently been a tendency on the part of some 

 authorities to consider that the union of toxin-antitoxin does not 

 correspond to what takes place in ordinary chemical union, but 

 is a physical interaction of bodies in a colloidal state, the action 

 being one of the so-called absorption phenomena. The smaller 

 toxin molecule becomes entangled, as it were, in the larger 

 antitoxin one, very much as a dye becomes attached to the 

 structure of a thread. Bordet has long maintained a theory of 

 this nature and gives reasons for believing that there is no 

 definite quantitative relationship in the combination of the 

 molecules of the two substances, different amounts of antitoxin 

 affecting in varying degree all the molecules of a given amount 



