AT THE turn of the century it was generally assumed that high pressure in the portal system was the cause of the ascites which occurred with cirrhosis of the liver; the doctor hoped that development of collateral circulation would lead to disappearance of the fluid, and much attention was paid to the anatomy of possible channels. Herrick1 tackled the problem experimentally and concluded that "the communication of the arterial pressure to the portal pressure is an important factor in an explantion of the increased portal pressure in portal cirrhosis" and that compression of the venous radicles by fibrous tissue was perhaps of less consequences. However, even fifty years ago, those giving careful thought to the subject were not sure that portal obstruction was the entire answer to the problem of "cirrhotic" ascites. In his classical monograph, Rolleston,2 for example, raised the question of whether poisons developing during the