282 LOCAL TLSSUE REACTIVITY 



the time when antibody 01 antigen was injeclecl chrcctly into the 

 area jjreviously prepared with a bacterial Hltratc. 



Rabbits ^vere sensitized l)y sinj^le intraxcnous injections ol 

 normal horse serinn in a dose ol 1 ( .( . per kilo ol body weight. 

 One week later they received injections ol H. typhosus "agar 

 washings' lillrate into one to three areas ol the skin ol the ab- 

 domen, i.e., upper and louver right and upper left cjtiachants, 

 respectively. Twenty-foin- hours afterAvards, the same skin sites 

 Avere reinjected with normal horse serum. From four to five 

 hotns later, the rabbits shoAved severe hemorrhagic necrosis at 

 the site of the intradermal injections. The lesions were identical 

 with those observed in the phenomenon of local skin reactivity 

 to bacterial filtrates. Histological changes ^vere also similar to 

 those previously described. The horse serimi failed to elicit the 

 reactions in areas prepared by bacterial filtrates devoid of skin- 

 preparatory potency {Streptococcus viridans culttire fdtrates) , 

 turpentine, broth and other irritants. It ^vas necessary to allow 

 a definite inctd^ation period for the preparatory effect of a potent 

 iiacterial filtrate. The lesults ^vere negative if the injections im- 

 mediately followed each other or were carried out within six 

 hoins. The interval of time successfully employed ^vas t^venty- 

 foin' hotus. Intervals of time between six and tAventy-foin- horns 

 have not been studied as yet. 



As already noted, the sensitization of rabbits to horse serum 

 was accomplished by single intravenous injections seven days 

 prior to the tests. Shorter incubation periods Avere insufficient 

 {i.e., three days) . The reaction observed was specific. Reinjecticjn 

 of prepared areas Avith sheep, guinea pig, and human serum in 

 rabbits sensitized by single injections of horse serum elicited no 

 reactions. Some of the rabbits Avere bled on the day of the pre- 

 paratory injections and the sera tested for precipitins against horse 

 serum. There Avas no apparent parallelism between the precipitin 

 titers of the \arious sera and the incidence of reactions; the latter 

 easily obtained in most of the rabbits tested in a manner de- 

 scribed. The j5ercentage of negative animals has not been ascer- 

 taiired as yet. In normal rabbits (not receiving any sensitizing 

 intravenous injection of horse serum) , injection of normal horse 

 sertnn, histamine diluted 1:1000, and sterile meat infusion broth 

 into sites prepared with potent bacterial filtrates, did not elicit 

 any hemorrhagic necrosis. 



Additional experiments Avere ( arried out in order to determine 



