In previous experimental studies on peritonitis, we demonstrated1 a vasodepressant toxic substance in peritoneal washings. The washings from normal animals and from animals on which an extensive gastrointestinal operation had been done, as well as washings from animals with experimental bile peritonitis, were innocuous under the same experimental conditions. Since the vasodepressant toxin occurred in the exudate after death from bile peritonitis, we suspected that the colon bacillus or one of the large gram-positive obligate anaerobic organisms might be a possible source of the substance. This association of bacteria with the production of a toxic substance has been demonstrated by many authors2 in experimental intestinal obstruction and in experimental acute pancreatic necrosis. It has likewise been known for a long time that the colon bacillus3 and allied organisms4 produce a toxic substance that brings about rapid death when administered to laboratory animals by the intravenous or