102 IMMUNITY AND ANAPHYLAXIS 



arsenobenzene we can find all intermediary substances. It 

 would thus be very interesting to know whether all the 

 different amino-acids which are found in casein, in serum, 

 or in bacterial antigens are united in different proportions 

 in each granule or whether each of them forms different 

 granules. Biologic chemistry will probably undertake some 

 day to solve these problems which we cannot take up now. 

 What appears certain is that all the colloids of egg-white or 

 of serum or of bacterial bodies when introduced into the 

 interior of the organism do not become antigenic, nor do 

 all the colloidal granules of beterologous origin take part 

 in the same degree in the formation of specific antibodies. 

 Levaditi and Mutermilch 1 have shown by their studies on 

 the production of anti-nagana antibodies in the guinea-pig, 

 rabbit and rat on the one hand and in the hen on the other 

 hand, that antibodies produced by the same antigens in 

 animals of different species are not identical. 



Different antibodies can be produced only by different 

 antigens, which means that in the particular case brought 

 forward by Levaditi and Mutermilch, products of bacterio- 

 lysis of the trypanosome do not form a uniform antigen 

 but these products contain a mixture of colloidal granules 

 whose composition and affinities are different. In con- 

 sequence the colloidal granules which are antigenic for the 

 organism of the rabbit, guinea-pig and rat on the one hand 

 and of the hen on the other hand are not compounds of the 

 same amino-acids or at least are not grouped in the same 

 way. We have also seen that in the case of Bacillus typhi 

 murium the substance virulent for mice is not the same 

 which gives this bacteria its virulence for rats. These two 

 examples allow us to conclude that in complex colloids formed 

 of every sort of amino-acid these latter are not uniformly 

 distributed among all the granules but constitute groups of 

 granules of different chemical compositions. 



And if this is so, we should necessarily conclude that every 

 colloidal granule of an albumin or of a complex colloid which 

 is not antigenic ought to be immediately digested and trans- 



1 Antibody and Animal Species, Ann. de 1'Inst. Pasteur, 1913, xxvii, 924. 



