520 STUDIES IN IMMUNITY. 



incompletely saturated complex, can hcmolyze a large amount of 

 corpuscles but produces hemolysis very slowly, even if a very small 

 amount of the corpuscles is added. It would seem, then, as if such 

 a mixture Included a strong dose of poison, the activity of which is 

 distinctly depressed owing to the addition of a little antitoxin, and 

 the result of the experiment is in no way compatible with the idea 

 that such a mixture contains, in addition to perfectly neutralized 

 hemolysin, a certain excess of perfectly intact hemolysin. But 

 there is another way of proving that an antigen can unite with the 

 antibody in variable proportions and so form complexes of varying 

 constitution in accordance with the amounts of each substance 

 present. It may be shown that such complexes do not react in 

 the same way to certain agents such as heat. It is perfectly evident 

 from the observations of Eisenberg and Volk, that the union of the 

 agglutinin with the agglutinable substance of bacteria takes place 

 in variable proportions. Landsteiner and Jagic have confirmed this 

 idea by showing that the complexes obtained, which differ in the 

 relative proportions of each constituent, show different resistance 

 to heat.* The brilliant researches of Grassberger and Schattenfroh 

 on the toxin of symptomatic anthrax, are very instructive in this 

 connection. When we mix a certain dose of the toxin, either with 

 little or with much antitoxin, we obtain complexes of toxin-anti- 

 toxin, which vary, as is shown by their varying reaction to heat. 

 It is to be noted in the first place that the toxin in question is ther- 

 molabile, whereas the antitoxin resists heat much better. The 

 complexes containing a large amount of antitoxin enjoy the prop- 

 erties of this substance, that is to say, they resist heat much better 

 than do complexes which are weak in antitoxin; these latter mix- 

 tures are decomposed at a given temperature, and the poison is 

 destroyed so that the antitoxin may be recovered. These authors 

 have further shown themselves partisans of the idea that antitoxin 

 unites with toxin in variable proportions. They have been able to 

 prove that their poison absorbs much more antitoxin than is 

 necessary to destroy its entire toxicity and forms a stable complex 



* I noted in 1896 (Mode of action of preventive sera), that heating toward 

 50 degrees brings about a real disagglutination of the substances which have been 

 agglutinated. This is due to the fact, as Landsteiner and his collaborators have 

 shown, that the complex is dissociated. What is more, the ease of this dissociation 

 varies with the amount of agglutinin present. 



