98 IMMUNITY AND ANAPHYLAXIS 



are not rapidly fatal will always be less dangerous than 

 intracellular reactions. 



We may thus affirm that the basis for the production of anti- 

 body in excess is the obligation to digest colloids under which 

 the organism finds itself in order to assimilate or eliminate 

 these colloids. We know that this digestion will produce 

 pathologic manifestations when the compound of antigen 

 with antibody causes a precipitate: but on the contrary 

 when this compound is soluble, this digestion is completely 

 harmless. 



This difference in the intimate mechanism of the reactions 

 which together result, on the one hand, in the formation of 

 a coagulum or of a precipitate and, on the other hand, of a 

 product which is soluble remains to be explained. 



For this explanation only a very limited amount of experi- 

 mental material is at hand. The basis of the reply is the 

 physicochemical constitution of colloids which is still little 

 understood. Let us also see what we know and what con- 

 clusion we can draw from the point of view which is presented 

 to us. 



From the physical point of view a colloid can be represented 

 as formed by granules or "micelles" 1 composed of a variable 

 number of molecules of a single or of several different sub- 

 stances. The structure of these "micelles" is unknown, 

 but we may assume that when they are in suspension in a 

 liquid, they have, because of surface tension, a spherical or 

 ovoid form. The volume of the granules of the simplest 

 and most homogeneous and antigenic colloid which we know, 

 arsenobenzene, is variable not only because the "micelles" 

 may contain a larger or smaller number of molecules but 

 because several "micelles" may be combined into a single 

 granule. Experiments show, moreover, that in every liquid 

 containing a colloid in suspension there is also a certain 

 number of free molecules. Thus, when we allow a solution 

 of luargol (1 to 400) in 0.5 to 0.7 per cent, salt solution, to 



1 The word ''micelle" is used to define the units of albuminoid matter 

 of every colloid in the same sense that the word "molecule" expresses the 

 unit of chemical compounds. The word "particle" is used by many 

 American authors in the same sense. 



