The etiology and pathology of gastric and duodenal ulcer have attracted the interest of investigators and clinical observers since the time of Cruveilhier, the more because of the fact that no satisfactory explanation has appeared. Many theories have been advanced, based on experimental investigation and clinical observation, and as many counter-theories and refutations, based on equally good material, have been called up in answer. The literature, therefore, abounds in probable theories of the causation, but everywhere one is impressed by the diversity of opinion and the variety of the avenues of approach employed by the investigators. Our interest in the question was aroused by the observations reported from this laboratory by Ellis,1 who found, in a few cases, hemorrhagic erosions and even perforating ulcers in the stomach and duodenum of dogs after the intravenous injection of toxins isolated from animals with a high intestinal obstruction. Erosions and ulcers were