272 president's address — section l. 



And so in this as in other sciences investigation of matters not 

 at first glance relevant has been fruitful in widening the under- 

 standing of the problem. On this wider basis the types of anti- 

 gens are many, in the following groups: — Bacterial cells (dead or 

 alive), bacterial exotoxins (tetano-toxin), vegetable toxins (ricin, 

 abrin, cretin, robin), snake venoms, enzymes (pepsin, trypsin), also 

 animal cells such as blood corpuscles, and foreign proteins of animal 

 or vegetable origin as casein, serum, egg albumen. These all 

 agree in being colloidal in nature and albuminoid in composition. 

 A characteristic property of albuminoids is that of forming 

 adsorption co'mplexes with other albuminoids and with many other 

 substances. The statement has been made that lipoids such as 

 those extracted from blood cells, or lecithin, could act as antigens, 

 but generally this is net accepted. It may well be, however, 

 that a lipo-protein ccmplex can act as an antigen. 



There are one or two conditions necessary for antigen action, 

 viz. : — 



(1) the antigen must penetrate into the hocit/ of the animal 



which is toi produce thei antibody, and this is the 

 reason why antigens ingested by the mouth have, as a 

 rule, no action ; 



(2) it must be foreign to the animal in which it is to act ; but 



(3) it need not be toxic. 



Now, with a diversity of antigens there is a corresponding 

 diversity of antibodies, but the tendency toi dividei them into 

 sharply limited classes widely differing from one another is not 

 justified. Our knowledge of the presence of any particular anti- 

 body is due solely to its combination with its proper antigen to 

 form a. complex with definite properties and reactions. These 

 reactions do not characterize the antibody, but the antigen-anti- 

 hody comhinafion, and although thei reactions which reveal their 

 presence may differ markedly, there is good ground for thinking 

 ngglutinins, precijMins, hcpmolysins, hacte/iolysws, and the like 

 liave a related origin and composition. 



A remarkable feature is the specific nature ol the reaction 

 following the action of an antigen. The antibody developed as a 

 result has a strong aflBnity for that antigen, and this affinity is 

 sufficiently specific or elective as to form the basis of sero-diagnosis. 

 The agglutination reaction, preicipitin test, complement fixation 

 test, and even the phenomenon of anaphylaxis, all afford evidence 

 of this characteristic. The 2weci'pitin test will suffice for illustra- 

 tion. If a series of rabbits are injected each with a different 

 serum, human, ox, pig, and dog, then after two or three injections 

 it is found that the serum of each rabbit will, if mixed with the 

 appropriate serum, namely, that one which has been used to 

 immunize it, give a turbidity and eventually a precipitate at the 

 line of union. Any other serum than the one used as antigen will 



