288 LOCAL ILSSLE REACTIVITY 



paratoiy pcjlcncy ol B. lyjjliusus was aj)j)r<).\imatc'l) li\c limes 

 higher than that of meningococcus lihrate. In contrast, tlie react- 

 ing potency ol B. I\})h().sus filtrate was tinee and one-hall times 

 lower tlian that of meningococcus. Nexertheless, irrespective of 

 the concentration ol reacting units, the dilutions employed were 

 effective only provided they contained a certain minimal amount 

 of skin-preparatory factors. Thus, for example, the dilution 1:70 

 of meningococcus filtrate containing approximately 15 skin-pre- 

 paratory units and 50 reacting tniits produced no effect. In order 

 to obtain the reactions it was necessary to employ a dilution of 

 1:45 which had about 2'^ skin-preparatory iniits and 75 reacting 

 iniits. On the contrary, the dilution 1:200 of B. typhosus filtrate 

 containing 25 skin-preparatory imits and only 5 reacting iniits 

 was sufficient to elicit strong reactions. It is obvious, therefore, 

 that the ability of the bacterial hltrates to elicit reactions in 

 antigen-injected sites is strictly conditioned by their skin-pre- 

 paratory potency and seems to be independent of the concentra- 

 tion of reacting units (Slnvartzman, i9.^^()c) . 



EFFECT OF INTRAVENOUS INJECTION OF ANTIGEN -(- ANTIBODY COM- 

 PLEXES UPON SKIN SITES TESTED WITH HORSE SERUM IN RABBITS 

 SENSITIZED TO HORSE SERUM 



As already emphasized, potent provocative factors, but dexoid 

 of preparatory potency, may be obtained from combinations of 

 animal protein antigens with homologous antibodies. Advantage 

 was taken of this in order to substantiate further the above con- 

 clusion that the ability of bacterial filtrates injected intravenously 

 to produce reactions at sites xvith normal horse serum in sen- 

 sitized rabbits is strictly conditioned by its skin-preparatory 

 potency. Antigen -)-antibody complexes employed xvere mixtures 

 of anti-human horse serum Avith human serum made in propor- 

 tions yielding potent reacting factors. After ascertaining the re- 

 acting potency of the antigen-^antibody complexes, it was at- 

 tempted to determine the effect of these complexes upon skin 

 sites injected xvith horse serum in rabbits sensitized to horse 

 serum six days pre\iously. As xvas noted, the intravenous injec- 

 tion of potent bacterial filtrates preceded by one hour or fol- 

 lowing after twenty-four hours the intradermal tests Avitli horse 

 serum, invariably resulted in j^rompt reactions at sites of the test 

 with horse serum in rabbits sensitized to horse serum. In the 



