276 president's address — section l. 



more reasonable view is to imagine all the toxin molecules sur- 

 rounded or clothed with a " film " of antitoxin, so that all of 

 them are propoitionately attenuated, even when the amo^unt of 

 antitoxin is insufficient to make them inert. 



Another .explanation of antitoxin-toxin combination, and one 

 providing at the same time an explanation of Ehrlich's 

 phenomenon, was that of Arrhenius and Madsen. Time forbids 

 more than a brief statement. They supposed the combination was 

 ■similar to that between a weak acid and a weak base, or to that 

 between acetic acid and alcohol . Here the reaction is complicated 

 by the effect of electrclytic dissociation and of mass action, and the 

 reaction is really reversible. For example: — 



Acetic acid -j- alcohol '^ ethyl acetate -j- water, 



and all four constituents are always present and in a state of 

 equilibrium. Alter the proportion of any one of them by addition 

 or subtraction, and you alter the proportions cl each of the others 

 as well. Such a theory can explain Ehrlich's phenomenon up to a 

 point, but, as Nernst has pointed out, since the equation is ejc 

 liypothesi a reversible one.* there must be suiiu- toxin free all the 

 time. If the mixture were injected into an animal this free toxin 

 would combine with its proper susceptible cells, but, being in too 

 small a dose, would fail to poison them. But the abstraction of 

 this free toxin would upset the equation, and niore free toxin 

 Mould be liberated, which in turn would unite with the susceptible 

 cells, and so' by a cumulative action the animal would be poisoned. 

 In practice this does not occur, and this disproves Arrhenius and 

 Madsen's hypothesis as to the mode of union. 



We now come to' Bordet's Adsorption Theory, sometimes called 

 the physical theory. Bordet had observed that agglutination of 

 bacteria or corpuscles by a suitable serum containing agglutinins 

 would not take place in the absence of salt. The same is true for 

 the precipitation of colloidal solutions. He further discovered 

 strong analogies between the fixation of hsemolysin and coinple- 

 ment by red corpuscles, and the adsorption of dyes by a fabric. 

 If a piece of filter paper of a certain size is introduced into' a watch 

 glass containing a dilute staining solution, the paper will take up 

 the dye and assume ai uniform pale tint. But if a similar piece 

 of paper is cut up into a number oi small pieces, and these are 

 introduced into the staining solution one by one with an interval 

 between each, then the first adsorbs a considerable proportion of 

 the stain, the second a little less, and so on, until the last is 

 hardly stained at all, and the intermediate ones show gradatio^ns 

 of colour. Now, compare this with the combination of red cor- 

 puscles and the haemolysins of an immune serum. It will be found 

 that a considerably larger dose of corpuscles will be haemolyzed if 

 the|y aire added at once than will be the case if small repeated 

 doses are added at intervals. 



