THERE has been an interest in venous fistulas since 1877, when Eck1 first suggested a technic for producing a portacaval anastomosis. Since that time many studies of the physiologic changes produced in animals by an Eck fistula have been reported.
Blalock,2 Blakemore3 and Whipple4 have reported that a side to side portacaval fistula without obstruction of the portal vein will close, but to date no specific report of experimental work has been found in the literature. A similar fistula will remain patent in the typical animal with an Eck fistula, in which the portal vein is occluded.5
A few clinical cases have been reported in which there were recurrent gastrointestinal hemorrhages after the establishment of a portacaval fistula. It seems possible that this may be due to closure of the fistula in some instances. Many factors contribute to the maintenance of the patency of a