THEORIES OF IMMUNITY 27 



destroys the antibody and does not destroy the ricin, it is 

 possible to recover the ricin which the precipitate contains. 



Ricin presents another peculiarity which permits us to 

 study with a little more precision what it has not been possible 

 to study with diphtheria and with tetanus toxin. Ricin 

 dissolves red blood corpuscles in the test-tube. This pecu- 

 liarity is very precious because it permits us to make a large 

 number of tests under identical conditions and which are in 

 consequence exactly comparable; for example, to estimate 

 the pathogenic dose of diphtheria toxin or, in other words, 

 to establish a "unit of measure," it is necessary to inject 

 different quantities into a very large number of animals. 

 Each animal can be injected only once and as there are very 

 appreciable individual differences between the animals from 

 the point of view of their sensitiveness to the action of toxin, 

 even when they are of the same age, of the same weight and 

 apparently of the same species, one can obtain units of meas- 

 ure of only an approximate value, sufficiently exact for the 

 practical use of antitoxic serum but insufficient for the study 

 of the mechanism of the reactions. 



With ricin the same solution (relatively more stable than 

 that of bacterial toxins) can be made to act in the test-tube on 

 the same blood cells of the same animal and these tests can 

 be multiplied almost indefinitely so that " units of measure" 

 can be much more exactly obtained. 



By studying the reaction of ricin and antiricin in the pres- 

 ence of blood cells it has been made possible to discover what 

 one may call the phenomenon of surcharge, that is to say, the 

 property of antigens to combine with their antibodies in 

 variable proportions and not according to the immutable 

 law of equivalents established for chemical combinations 

 as such. Thus, for example, when we mix one hundred toxic 

 units of ricin with one hundred antitoxic units, presumably 

 well titrated, a neutral mixture is obtained; but when the 

 hundred toxic units are added in fractions of five or of ten 

 units to one hundred units of antitoxin, a mixture is obtained 

 in which there will be a certain number (five to fifty accord- 

 ing to circumstances) of free toxic units. This fact has been 

 verified for other toxins and antitoxins and the conclusion 



