I . Experiments were done to find whether the rat liver can be maintained in a satisfactory condition when perfused with oxygenated Krebs-Ringer bicarbonate buffer without added protein or red cells.2. The condition and performance of the liver in this system were assessed from measurements made to ascertain its general condition or viability, its basal characteristics and its response to added substrates.3. It was found that the rapid flow-rate of the medium through the livers and the efficient oxygenation of the medium ensured that enough oxygen was available for the livers to deal with large quantities of added lactate.4. The potassium concentrations in the livers and the rates of leakage of alanine aminotransferase (EC 2.6. I . 2 ) from the cells during perfusion, and the water content after perfusion showed that the livers were not grossly damaged and that they did not deteriorate measurably for up to 3 h of perfusion. 5 . Liver oxygen consumption, ATP concentrations, lactate and pyruvate concentrations and ratios, and rates of urea and glucose synthesis and bile secretion, all in perfusions without added substrate, were either similar to measurements by other workers from livers perfused withmedia containingred cells and protein or were reasonable extrapolations from availabledata.6. The rates of glucose production from lactate, and urea and glucose output from amino acids indicated that the liver responds adequately to added substrates.
7.Measurements of amino acid concentrations in perfusate indicated that the livers of rats starved for 18-20 h regulated the amino acids to characteristic levels, by overall output or uptake, except for valine, leucine and isoleucine which were continuously given out into the medium. l ' h e results suggest that in vivo there is a general flow of most of the amino acids from extrahepatic tissues to the liver during fasting, while valine, leucine and isoleucine flow from liver to extrahepatic tissues. 8. When pentobarbitone sodium (Nembutal) was used as the anaesthetic for removal of the liver from the donor rat, the rates of urea and glucose output in perfusions without added substrates were lower than when halothane (Fluothane) was used, indicating that pentobarbitone has an inhibitory effect on these measures of liver function during the subsequent perfusion.Perfusion of the isolated rat liver with blood leads to difficulties of interpretation because of the chemical reactions which take place in the cells and plasma of the blood. These reactions cannot easily be distinguished from changes caused by the liver by conducting simple 'blank' experiments without the liver because of the influence of one of these tissues on the metabolism of the other. Other difficulties connected with the use of blood, such as its variable and relatively unknown composition, the danger of haemolysis and the possible effects of heparin on the metabolism of the liver (Coon & Willis, 1966), have been overcome to some extent by using simplified media containing protein or dextran, washed red cell...