IMMUNOLOGICAL RELATIONSHIPS 337 



loi nialinized saline, sterile egg white solution, innnune rahhit 

 serum, normal human serum, and meningc^coccirs and B. dys- 

 entoiae (Shiga) toxins. The most marked and constant inhibi- 

 tion Avas j^roduced by pituitrin. 



"Desensitization" cotdd also be obtained by clamping off a por- 

 tion of the prepared area for one hour after the intravenous in- 

 jection. The clamps themselves caused no apparent injiuy to the 

 skin. It becomes evident, therefore, that the inhibition described 

 by Btunet cannot be considered as anaphylactic desensitization. 

 It shoidd be pointed otit that Btirnet himself did not observe any 

 specificity of the inhibition. In one of his experiments, even saline 

 effected a weak desensitization. It seems more likely that the 

 agents employed produce their effects by creating a local ischemia 

 which shields the prepared skin tissue from the injurious agents 

 in the blood stream. 



H. Gross, (iggifl^ b, c) briefly described observations concerning 

 inhibition of the phenomenon. The inhibition was obtained when 

 injection of a bacterial fdtrate was given to a rabbit, shortly after 

 or simtdtaneously wdth the skin-preparatory injection. No reac- 

 tions followed the provocative intravenous injection of the same 

 filtrate twenty-four hours later. The same observation Avas made 

 independently by Ogata (1932, 1936) , in his experiments with B. 

 colt cidtiue filtrates and further studied by him in my laboratory 

 with a variety of active principles of ascertained and titrated 

 potency. The experiments showed that the phenomenon of local 

 skin reactivity can be inhibited if an additional intravenous in- 

 jection of a potent bacterial filtrate is given ^vithin a certain 

 period of time prior to or following the skin-preparatory injec- 

 tion. The inhibitory effect of the additional intravenous injec- 

 tion takes place within the limits of certain quantitative relation- 

 ships. Thus, if the skin is prepared with a large amoinit of filtrate, 

 the inhibition is absent en" incomplete. Similarly, if a large 

 amoiuit of filtrate is used for the pro\'ocati\'e injection no inhibi- 

 tion occins. The inhibition described is of a transitory nattne. 

 The additional intravenous injections given several hours before 

 or after the skin preparation ha\e no inhibitory effect. It can 

 be obtained only with filtrates capable of eliciting the phenome- 

 non itself. Bacterial filtrates of low reacting potency (Strepto- 

 coccus hemolyticus filtrate) as well as non-bacterial sid^stances 

 (0.85 per cent NaCl solution, plain broth, and normal horse 

 serum) produce no inhibition. The mechanism of the inhibition 



