PHYSICOCHEM1CAL PROPERTIES 37 



The first stage of the transformation which the sodium 

 compounds of the arsenobenzenes undergo in the organism 

 consists then in making them change from the soluble state 

 to the insoluble state. They form precipitates or coagula 

 in the interior of vessels and these then provoke disturb- 

 ances which are more or less severe according to the in- 

 jected dose, the degree of alkalinity of the product, or what 

 amounts ultimately to the same thing, to the quantity 

 variable from one individual to another, of precipitating 

 substances which preexist in the blood: Other conditions 

 being the same, the reaction will depend also upon the rapid- 

 ity with which the injection is made. 



There is thus a primary analogy between the action of 

 the arsenobenzenes and of biologic antigens and for greater 

 convenience in the discussion which follows we may call 

 the disturbances which immediately follow the injection 

 primary intravascular disturbances and we may group the 

 total of substances in the organism which concur in the forma- 

 tion of the precipitate under the name of preexisting or normal 

 antibody (and we know that in this case, these include acids 

 and salts, sodium chloride, calcium phosphate, etc.) 



Arsenobenzene which becomes insoluble in the blood 

 plasma does not remain indefinitely in the organism in the 

 precipitated state. We may assume that only one part of 

 the injected product becomes insoluble; another part remains 

 in solution or passes to the precipitated state only after a 

 very short time, a fraction of a second. 



In fact when we inject an animal intravenously with a 

 series of small doses of luargol, we obtain after a certain 

 incubation period a serum which precipitates luargol of the 

 same alkalinity in vitro. But at the same time we know as a 

 result of seeking to define the conditions of the formation of 

 this precipitate, that the reaction is not simple; whatever 

 may be the proportions of sera and luargol in mixtures, total 

 coagulation of the latter is never obtained. It is found, further- 

 more, that in certain mixtures the precipitate already formed 

 redissolves in some minutes or some hours afterward. 



In order to explain the cause of the particular nature of 

 all these reactions, we must remember these facts: That in 



