io8o A CRITIQUE OF THE EHRLICH THEORY 



failure is usually attributed to an assumed impossibility of obtaining antisera of suf- 

 ficiently high specific titre, or to an assumed antagonism between the human body and 

 foreign antibodies. It is ecjually logical, however, to assume that the failure is a logical 

 result of an error in our accepted conceptions as to the nature and function of specific 

 antibodies. 



That humoral antibodies do not always represent the sole or essential specific de- 

 fense of the body is readily shown by massive immune blood transfusions. Tubercle 

 bacilli, for example, injected into the peritoneal cavity of a tuberculous guinea pig 

 are rapidly destroyed by a process of extra-cellular lysis. Total blood replacement of 

 a normal guinea pig by transfusion from a tuberculous donor does not confer upon 

 the normal guinea pig the power to produce this intraperitoneal lysis.' This failure 

 cannot be attributed to inadequate dosage, nor to an assumed antagonism to foreign 

 antibodies. 



My conviction that there is an error in our currently accepted basic assumptions 

 was strengthened by several years' experience in the study of the quantitative charac- 

 teristics of serum reactions.^ This study led to numerous contradictory and paradoxi- 

 cal results impossible to harmonize with the accepted theory without the necessity 

 of numerous minor hypotheses.-' For example, if a relatively large amount of hemo- 

 lytic amboceptor (heat-inactivated, hemolytic serum) is exposed to red blood corpus- 

 cles until sensitization is complete, titration of the residual amboceptor in the super- 

 natant fluid often shows, in place of the expected decrease in amboceptor, a distinct 

 or even marked apparent increase in amboceptor," This finding can be harmonized 

 with the Ehrlich theory by the assumption that heat-inactivated hemolytic serum 

 contains in addition to hemolytic amboceptor one or more non-hemolytic antibodies 

 capable of binding complement, the non-hemolytic antibodies having a relatively high 

 affinity for the red blood corpuscles. On exposure to corpuscles the relatively greater 

 absorption of the non-hemolytic antibodies would remove from the supernatant 

 fluid a complement-deviating anti-hemolytic agent. This removal might give an ap- 

 parent increase in hemolytic titre, in spite of a real decrease in amboceptor. 



• My conviction that there is a radical error in our accepted theory was further 

 strengthened by a study of the so-called "lecithin" activation of cobra venom. Both 

 cobra venom and lecithin are non-hemolytic for certain erythrocytes when tested in 

 ordinary doses. Added to each other, however, they constitute a powerful hemolytic 

 agent. According to the Ehrlich theory, lecithin and cobra venom could co-operate 

 in hemolytic action only by their direct conjugation to form a "cobra-lecithid." 

 Chemical analysis, however, shows that this conjugation does not take place. The 

 active component of cobra venom is a powerful lipase, capal^le of very rapidly split- 

 ting off one fatty acid radical from the lecithin molecule. The resulting "mono-fatty- 

 acid-lecithin" is a powerful hemolytic agent, and accounts for the total hemolytic 

 action of a lecithin-venom mixture.^ 



' /. Exper. Med., 18, 601. 1913. (AH references are to author's own works.) 



^ J . Infect. Dls., 4, 219. 1907. 



3 Ibid., p. 67. 1907. 



^Ibid., 2, 485. 1905. 



^Zlschr.f. ImmimUiitsforsch. ti. exper. Tlicrap., 6, 513. 1910. 



