128 THE SIDE-CHAIN THEORY 



correct a still more plausible explanation of the specificity of the 

 antibodies would thus be afforded. 



Before concluding the present chapter one more point may yet 

 be appropriately considered, viz., the question why those poisons 

 w r hich we can prepare in pure form in the chemical laboratory, and 

 whose structural composition is known, such as the various alka- 

 loids, glucosids, alcohols, etc., do not give rise to antibody formation. 

 The fundamental reason for this differing behavior according to 

 Ehrlich lies in the fact that the true antigens are chemically bound, 

 and that chemical interaction between antigen and cell receptor 

 takes place because the bodies in question are structurally closely 

 allied to the true foodstuffs. The majority of poisons of the chemi- 

 cal laboratory, on the other hand, are not taken up by the cells in 

 virtue of the existence of a special chemical affinity, but merely in 

 consequence of physical influences. 



This is well shown in the following experiment. After it had been 

 discovered by Ehrlich and Overton that the injection of various 

 anilin dyes leads to their storage in certain tissues of the body, and 

 that this storage is due to the presence in these tissues of certain 

 lipoids which act as solvents for the pigments in question, Hans 

 Meyer and Overton could demonstrate that the strength of various 

 narcotics is not dependent upon their chemical composition, but 

 upon their coefficient of distribution which regulates their distri- 

 bution between the blood-plasma and the lipoids of the brain. This 

 is well shown in the table on page 129, which is taken from Baum. 

 The first column of figures represents the coefficient of distribution 

 of the various narcotics, as calculated for water on the one hand, 

 and fat on the other (calculated for olive oil), while the figures of 

 the second column indicate the amount of the substances per liter, 

 expressed in fractions of the corresponding normal solutions, which 

 are just sufficient to produce narcosis in the test animal (usually 

 frog larvse); this is termed the threshold of action. By comparing 

 the two columns it will be noted that notwithstanding the wide 

 variations in the chemical structure of the different narcotics, their 

 effect is evidently dependent upon purely physical conditions, viz., 

 the coefficient of distribution. 



