TOXINS AND ANTITOXINS 205 



We do know, however, from recent work, that antigen-antibody 

 complexes can dissociate, and the work of Loeb has shown that 

 reactions of proteins are, after all, in many ways strictly analogous 

 to ordinary chemical reactions between less complexly constituted 

 substances. Moreover, direct experimentation, such as that of Land- 

 steiner on agglutinated typhoid bacilli, Gay and Chickering's work 

 on the dissociation of antibody from precipitates, Huntoon's work on 

 the dissociation of antibody from sensitized pneumococcv as well 

 as observations made on the dissociation of diphtheria toxin from 

 toxin-antitoxin mixtures in connection with active immunization in 

 diphtheria, have now shown pretty definitely that dissociation of 

 these substances in the animal body and in the test tube may take 

 place. In the light of these newer researches the ideas of Arrhenius 

 and Madsen, must eventually be reexamined. 



Bordet, Landsteiner and others have brought out another point 

 of view which might explain the quantitative relations which exist 

 in toxin-antitoxin mixtures as worked out by Ehrlich. They dwell 

 particularly upon the analogy of these reactions with colloidal reac- 

 tions, assuming that when antitoxin is mixed with toxin in amounts 

 insufficient to completely neutralize, the units of antitoxin are not 

 taken up by a corresponding fraction of the total toxin present, 

 leaving a part of the toxin absolutely free, but that, on the contrary, 

 the antitoxin is equally distributed over all the toxin units present, 

 leaving all of them partially saturated. This would not sharply 

 neutralize a part of the toxin, leaving another part entirely free 

 to exert its activity, but would partially neutralize all the toxin. 

 It is more or less analogous, as Bordet brings out, to different 

 degrees of coloration which are produced when starch absorbs vary- 

 ing quantities of iodine. They compare it to an adsorption phe- 

 nomenon, rather than to a true chemical reaction. 



More or less in harmony with this view is the Dansyz effect, 22 

 which is as follows: When diphtheria toxin is added to its anti-toxin 

 in two fractions, a definite period elapsing between the addition of 

 the first and the second fraction, much more toxin remains free 

 than when the total quantities are mixed at once. This view has 

 been interpreted in Ehrlich 's sense by Von Dungern, but by a very 

 forced process of reasoning. We may assume with considerable con- 

 fidence at the present time, that, while the Ehrlich method of 



"Dansyz, Ann. <le PInst. Past., 16, 1902. 



