258 INFECTION AND IMMUNITY. 



exact dissolving dose for the given volume of cor- 

 puscles may be determined. Although the term 

 hemolysis is a perfectly proper one, we are to un- 

 derstand that serums cause solution of the hemo- 

 globin, but not solution of the whole cell ; we speak 

 loosely of solution of the corpuscles. 

 similarity After Bordet had shown the analogy between 



Between n T -, n /. 



Bactericidal oactericidal and hemolytic serums, and after the 

 iyt!c Action" phenomena of inactivation and reactivation had 

 been developed by Bordet and Metchnikoff, Ehrlich 

 and Morgenroth undertook the study of ambocep- 

 tors and complements as they occur in hemolytic 

 serums. The facts ascertained by them and the 

 methods of research which they devised have pro- 

 vided many investigators with a starting point for 

 work of the highest importance concerning the 

 bactericidal serums and antibacterial immunity, 

 and their interpretations, moreover, served to ex- 

 tend the side-chain theory of immunity to its pres- 

 ent comprehensive limits. 



For the sake of convenience one may speak of a 

 heated immune serum, or one in which the com- 

 plement has become inactive from age, as a solu- 

 tion of amboceptors, disregarding temporarily the 

 agglutinins, precipitins and perhaps other bodies 

 which the serum contains. Also, since fresh nor- 

 mal serums are rich in complements and usually 

 contain but a small amount of any one amboceptor, 

 they may conveniently be considered as solutions 

 of complements; yet normal serums may not be 

 considered as pure complement and used as such 

 in unlimited quantities for actual experiments, be- 

 cause of the bacteriolysins and hemolysins which 

 many contain. Only a quantity of the normal 

 serum which in itself is not toxic for the cell 



