such an antigen to produce anaphylaxis. In a similar 

 manner Bogomolez (12) used alcohol and ether extracts of 

 egg yolks to produce anaphylactic phenomena. It is 

 known, furthermore, that the solubilities of proteins 

 and other substances may be profoundly altered by the 

 presence of lipoids. The existence of a protein- free 

 antigen, therefore, was not established until the recent 

 work of Warden (to be considered further on). 



There is, however, one exception, namely the 

 protein- free antigen of Ford and Abel (9). These men, 

 studying the poisons of Amanita phalloides, found that 

 the toxic principle was an active glucoside. Their find- 

 ings have never been refuted. 



Those who favor the protein theory of the nature of 

 antigens are prone to criticise all experiments which 

 attribute fats and lipoids a role in antibody production. 

 The one chief critical point lies in the face that the 

 lipoidal antigens have not been satisfactorily proven to 

 be protein- free. On the other side, however, is it not 

 fair to inquire if the so-called protein antigens are 

 fat- free? Take practically any native antigen for use in 

 antibody production, and we must concede that there is 

 as definite an amount of lipoid as there is protein; and 

 merely because of the proportionately greater quantity 

 of protein is no reason at all to assume the role of the 

 latter as antigen. 



Numerous investigators, assuming that a given sub- 

 stance was an antigen, have endeavored to separate the 

 protein material and to determine what particular frac- 

 tion of the protein served in the capacity of antigen. 

 In the beginning they find that the substance under con- 

 sideration is a true antigen. They then, as a rule, pro- 

 ceed to separate the various protein constituents either 

 by fractional precipitation or by enzymotic digestion. 

 And finally they arrive at a point in the analysis beyond 

 which antigens are not found. In reviewing this v/ork one 

 is surprised to find in how few experiments the fat and 

 lipoid constituents have been eliminated. It is possible 

 that these lipoids were sufficiently bound to the pro- 

 teins as to be precipitated along with them; and due to 

 subsequent procedure, during the separation of the vari- 

 ous, fractions, the fats were liberated and discarded, or 

 remained in combination with higher fraction. If the 

 antigenic principle was lodged in the fatty portion, this 

 theory would, -of course, account for the results ob- 

 tained. Similarly, in analyses by means of enzymotic 

 reactions, the fats which are present may be digested by 



