A phenomenon of local skin reactivity to B. typhosus culture filtrates is described in this report. The reactivity was induced by skin injections of the filtrate followed 24 hours later by an intravenous injection of the same filtrate. The local response consisted of severe hemorrhagic necrosis and was fully developed 4 to 5 hours after the second injection. About 78 to 79 per cent of the rabbits employed were susceptible to this phenomenon. Different areas of the skin of the abdomen, when similarly treated, responded with equal severity to the intravenous injection. There were variations in the size of different areas in the same animals. The intensity and size of the local hemorrhagic reactions were not related to the intensity of the erythema produced by the preparatory skin injections. Following intravenous injection very severe hemorrhagic reactions were obtained, in those areas which reacted negatively in this respect to the preparatory skin injections. Evidently, the local trauma produced by the preparatory skin injections was not responsible for the localization of the toxic factors introduced by the intravenous route. It was necessary to allow a short interval of time between the skin preparatory injections and the intravenous injection, for the reproduction of the phenomenon. An incubation period of 2 hours was insufficient. An interval of 24 hours was invariably sufficient. The ability to react disappeared in 48 hours after the preliminary skin injections. Repeated direct injections of the filtrate into the same areas of the skin, with an interval of 24 hours between the injections, did not result in reactions similar to the above described hemorrhagic necrosis. The second skin injection was followed by reddening, some swelling and a local accumulation of polymorphonuclear neutrophil leucocytes which showed no signs of necrobiosis. There was no rupture of blood vessels. Skin injections followed after a suitable interval by intravenous injection, were necessary for the reproduction of the severe local hemorrhagic response. Skin reactivity to B. typhosus culture filtrate injected intravenously was not induced by turpentine in various dilutions, sterile tryptic digest broth, culture filtrates of 4 strains of streptococci or by the Streptococcus erysipelatis toxin. It was possible to titrate the skin preparatory factors. Dilutions of the filtrate up to 1:64 were able to induce the local skin reactivity. But, whereas dilutions up to 1:4 invariably prepared the skin so that very severe hemorrhagic reactions followed the second injection in susceptible animals, dilutions from 1:8 to 1:64 were uncertain and their preparatory effect varied in different animals. The skin preparatory factors showed considerable heat resistance. The heat resistance varied with the strains employed. One strain produced factors totally resistant to heating in the autoclave for 1 hour. However, there was definite and unquestionable inactivation of these factors as derived from other strains when the filtrate was diluted 1:2 and heated in the autoclave. Various hydrogen ion concentrations in the range from 9.0 to 4.0 had no effect upon the skin preparatory factors. Heat resistance was not modified by the various pH within this range. The mechanism of the phenomenon described has not been fully studied as yet. An experimental comparison of it with the manifestations of bacterial allergy of the skin is necessary. There are certain features, however, which considered together, distinguish this phenomenon from the known phenomena of bacterial hypersusceptibility. These features are: local reactivity; the short incubation period necessary to induce the local reactivity; the short duration of the state of reactivity; the ability to induce local reactivity by a single skin injection; the severity of the reaction; and the necessity to make the second injection of the toxic agent by the intravenous route. Studies on the relation of specific antisera to the phenomenon described are under way.