A study of the composition of pericardial fluid, with special reference to the probable mechanism of fluid formation.

The composition of pericardial fluid and simultaneously withdrawn plasma have been measured in rabbits and greyhounds. 1. Sodium and chloride distributions were found to be not markedly different from the ratio predicted for a passive distribution. The small deviation found in greyhounds could be largely corrected by the in vitro dialysis of plasma against pericardial fluid. 2. Calcium and magnesium were distributed in a manner expected from a passive ultrafiltrate of plasma. 3. Pericardial fluid was found to contain between one quarter and one third of the protein of plasma. 4. Separation of the protein constituents demonstrated a far higher proportion of albumin to other proteins in the pericardial fluid. 5. The osmolality of plasma was slightly higher than that of pericardial fluid, as would be expected from a plasma ultrafiltrate. 6. The potassium concentration of pericardial fluid was higher than the plasma concentration in all animals studied. This difference could be abolished, and an expected distribution obtained in the samples from greyhounds, by the in vitro dialysis of plasma against pericardial fluid. This observation for potassium cannot be attributed to haemolysis of blood in pericardial fluid samples or to the use of any inappropriate references. It is suggested that the elevated potassium concentration of pericardial fluid may reflect the lability of the cardiac intracellular potassium during cardiac contraction. 7. The results obtained in this study do not support the concept of an active secretion of pericardial fluid as has been claimed by others. The distribution of ions would appear to be passive and to follow the values predicted by the Gibbs-Donnan relationship.