INTERNAL SECRETIONS 391 



and suggests that it must act directly on the liver or other 

 tissues. 



8. Duodenum. As already pointed out (p. 349), the duo- 

 denum yields an internal secretion (secretin), which acts 

 directly upon the pancreas to stimulate its secretion. 



Toxic Action and Immunity. 



It is not definitely known how these internal secretions 

 each perform its special action, but light seems to be thrown 

 upon the question by the study of the mode of action of 

 various toxic substances, and the mode of production of a 

 condition of immunity against them. As will be presently 

 shown, a process of the same nature as the production of 

 internal secretions is involved. 



Snake and Diphtheria Toxins. It may be most simply 

 explained by considering first the probable mode of action 

 of the toxin or poison of snake venom, or of that produced 

 by the diphtheria bacillus, and the way in which protection 

 against these is established by the development of anti- 

 toxins. 



By injecting under the skin of the horse increasing 

 doses of such toxins the animal becomes quite resistant 

 to the poison, and if now some of the horse's serum is 

 taken it is found that a certain quantity can neutralise a 

 definite quantity of the toxin, so that if the mixture is in- 

 jected it does no harm to the animal. Something has been 

 formed in the horse which seizes on the molecules of the 

 toxin and makes them harmless, just as when soda is added 

 to sulphuric acid it forms a neutral salt. 



The two molecules have a definite chemical affinity for one 

 another, so that the toxin is no longer free to seize upon the 

 protoplasm of the animal's body. To explain this Ehrlich 

 has suggested that the protoplasm molecule, like the proteid 

 molecule (p. 10), is to be considered as made up of a central 

 core with a number of side chains or hands which play an 

 important part in taking up nourishment of different kinds, 

 for each variety of which special side chains have a special 

 affinity (Fig. 154). He supposes that some of these side chains 

 fit the toxin molecule, and are thus capable of anchoring it 



