278 . president's address — section l. 



solution. Then place lightly on it a smaller triangular piece of 

 unstained filter paper. After a short time it will be found that the 

 larger piece is of a perceptibly lighter colour just over the area 

 occupied by the fresh unstained piece, which, in turn, has become 

 faintly coloured. 



What may appear contradictory to some extent to- thei last state- 

 ment is the observation that these adscrpticn complexes become 

 more firmly consolidated with time. Just as in dyeing, the colours 

 become faster with a more prolonged immersion and after a longer 

 interval, so^ here. Madsen and Walbum (1904) showed that in a 

 neutral mixture of diphtheria-toxin and antitoxin the toxin was 

 able to dilTuse through a gelatine plate more quicklv than the anti- 

 toxin. Thus, after a short interval the deeper layer of gelatine 

 contained chiefly toxin, whilst the superficial layer held antitoxin; 

 This result, however, only occurred with recently prepared mix- 

 tures. Other workers, notably Martin and Cherry, have shown 

 the same thing with other antigeai-antibody-complexes. 



If time allowed, further experiments illustrating the union of 

 antigen and antibody in the precipitin reaction, and showing it 

 to be capable of similar ex]ilanation, could be given, and so with 

 other antigen-antibody combinations. Suftice it to quote an obser- 

 vation of Biltz. Colloidal ferric hydrate is an antidote to arsenious 

 acid poisoning, and the amount of the iron salt to exactly neutralize 

 the arsenic can be determined. Now, make a mixture of the twoi in 

 which there is an excess of arsenic of several times the toxic dose for 

 a given animal, and yet when administered it will not cause death. 

 Only when a gieat excess of arsenic is present does the mixture 

 beoG'me lethal ; in other words, wei have here an experiment quite 

 analogous with the Ehrlich phenomenon. In the case of the arsenic- 

 iron compound the explanation is purely that of molecular adhesion 

 or adsorption with consequent attenuation (so far as toxic pro- 

 perties are concerned) of th^ poison. This, in Bordet's view, is 

 the explanation of the Ehrlich 's phenomenon, and supplies an 

 understanding of the antigen-antibody combinations in general. 

 So the i^roduct of a reaction between antitoxin and toxin is not 

 always the same, but varies with the exact quantities and concen- 

 trations of each factor; in fact, there is a whole gradation of 

 complexes up' to that in which the antigen is completely saturated 

 by its antibody. 



Here we may consider for a monient the question Cif 



Coin ph-mcvf and its Fixation. 

 Any student of immunity must be struck by the participation 

 of the substance known as " complement " or " alexin "as a 

 factor in many of the reactions. In clinical practice it is becoming 

 increasingly important because of the reliability of tlie diagnostic 

 method known as '' fixation of the complement," so that at least 

 a few words inust be given to a consideration of its nature and 

 mode of action. 



