REACTING POTENCY 277 



actively acquired homologous antibodies is \vell su})ported by 

 observations on formation of reacting factors in passi\ely sensi- 

 tized rabbits. In these experiments rabbits were prepared by a 

 single intradermal injection of B. typhosus Tl "agar washings" 

 filtrate and twenty-three and one-half and twenty-four hours later 

 injected intravenously, first with the antigen, and secondly with 

 the antibody, respectively. In some experiments the order of in- 

 travenous injections \vas reversed; in others, the interval of time 

 ^vas varied. Parallel experiments ^vere done in which mixtures 

 of the antigen with the antibody made /// vitro were injected in- 

 travenously into prepared rabbits. 



Tavo separate intravenous injections of precipitinogen-contain- 

 ing seriun (himian seriun) and precipitin-containing seriun (i.e., 

 anti-human horse seriun) elicited severe hemorrhagic necrosis 

 in skin sites prej)ared with bacterial filtrates. No reactions Avere 

 obtained when each of the sera was injected alone in the same 

 dose; Avhen either the anti-himian horse serum or the normal 

 human serum Avas reinjected twice, or when there were made 

 t^vo separate injections of normal horse and normal human sera. 

 The experiments demonstrated, therefore, formation of reacting 

 factors in vivo through the interaction of passively accjuired anti- 

 gen and antibody. 



Positive results were obtained when the interval of time be- 

 tween the injections of seriun-precipitinogen and precipitating 

 antiserum was one-half hoiu', and also, in one experiment, when 

 it was t^venty-three and three-quarter hoins. A forty-eight hoiu" 

 interval gave negative results. 



In the majority of experiments the antigen Avas injected first 

 and the antibody second. No reactions ^vere elicited wiien the 

 order of intravenous injections Avas reversed. Quantitative studies 

 on reacting potency of mixtures of animal protein antigens ^vith 

 antibodies showed tiiat there were required larger amoiuits of 

 antibody than antigen for elicitation of the reaction. Thus, a mix- 

 tiue of human serum diluted 1:100 with undiluted horse anti- 

 human serum ^vas active, ^vliilst the mixture of horse anti-human 

 serum diluted 1:40 with undiluted human serum was devoid of 

 potency. Inasmuch as unquestionably a part of the injected serum 

 rapidly disappears from the blood stream, it is conceivable that 

 in the experiments in which antibody Avas injected first and the 

 antigen second, the reaction did not take place because an in- 



