The experiments recorded show that the skin involved in the zone of inflammation produced by the endodermal injection of virulent living erysipelas streptococci into rabbits becomes, after the lesion has healed, partially but not completely resistant to subsequent infection with the homologous organism. In the majority of experiments the lesion resulting from the first endodermal injection spread downward, i.e., ventrally, to the site of needle puncture forming an oval or elongated inflammatory zone. Subsequent injections of the same strain into the skin beyond and ventral to the apparent edge of the lesion showed that these areas in the supposed direction of the lymph flow likewise became resistant but not quite to the same degree as in the inflammatory area. The contiguous areas above, behind, and in front of the healed lesion exhibited only a very mild degree of resistance. The other side of the rabbit where no previous injections had been made reacted as did the normal skin.
Repeated injections on the same side bring about diminishing local reaction until there is almost no lesion following the injection of the standard skin dose. Thus there is a gradual spread of skin resistance on the inoculated side, whereas the non-injected side of the rabbit reacts normally. Finally, however, after many injections over relatively long periods the non-injected side becomes resistant, but at this time, there is evidence of general humoral immunity as shown by the presence of agglutinins and antitoxin in the blood. The local resistance is apparently not entirely specific, for the areas with the previous lesion and ventral to it become more resistant to another strain of beta hemolytic streptococcus, though in less degree, and to a virulent strain of Staphylococcus aureus to a still less degree. The most plausible explanation of the spread of the local immunity ventrally is that the streptococci follow the lymph channels. Indeed, in human erysipelas, the organisms are recovered by cutipuncture as far as 3 cm. beyond the advancing edge of the lesion where there is no gross evidence of inflammation. This aspect of local immunity will be considered in subsequent publications.
The skin of the rabbit involved in the inflammatory reaction following the endodermal injection of living streptococci becomes resistant to subsequent injection of the homologous strain and of other strains of erysipelas streptococci which are not immunologically identical.
The local immune areas are resistant but not to the same degree to a strain of hemolytic streptococci isolated from follicular tonsillitis and to a virulent Staphylococcus aureus.
The areas contiguous to the local lesion but outside the apparent boundary of inflammation become more resistant to subsequent injection. But the regions ventral to those areas become more resistant than those dorsal, anterior, or posterior to the inflammatory zone.
The skin on the non-injected side of the rabbit becomes resistant pari passu with the development of humoral immunity.