Ascites due to cirrhosis has been treated by producing adhesions about the liver and other viscera, as first described by Drummond and Morison in 18961 and by Talma in 1898.2 This type of operation came to be used widely, but then was abandoned because of the unpredictable results. More recently, Madden et al3 described a method of hepatopexy, in which magnesium trisilicate powder was applied to abraded adjacent surfaces of the liver and the diaphragm. The technique, coupled with renewed emphasis on lymph exudation from the surface of the liver as a factor in the pathogenesis of ascites,4 suggested a new evaluation of Morison's operation.
In a previous report,5 we demonstrated that experimental ascites can be prevented by enclosing the entire liver in dense adhesions created by abrasion and application of magnesium trisilicate powder. However, the technique produced perihepatic adhesions of variable extent and density, often