Dumont and Mulholland1 2 in 1962 were the first to describe gross contamination of the thoracic-duct lymph with blood in man. They made the observation in 20 patients with Laennec's cirrhosis and postulated that this phenomenon resulted from an increased hepatic sinusoidal pressure. In this condition, red blood cells are presumed to lead from the sinusoids into the hepatic lymphatics. However, this conclusion is not substantiated by a considerable number of experimental studies in which inferior vena caval or hepatic venous constriction has been used in the production of ascites in dogs.3-5 Such animals exhibit a similar increase in thoracic-duct flow rate to that described by Dumont and Mulholland in man, but no instance of hematochyli has been described.
We became interested in this area when animals with chronic abdominal arteriovenous fistulas, being studied for unrelated purposes, were found to have bloody chyle. An experimental study, therefore, was