1. Adrenalin injected intraperitoneally increases the rapidity of absorption of fluid from the peritoneal cavity, independently of whether the solution to be absorbed is hypotonic or hypertonic or is approximately isotonic with the blood serum. The intravenous injection of adrenalin also increases the absorption of fluid, but not so markedly as does the intraperitoneal injection. 2. Adrenalin injected either intraperitoneally or intravenously increases the quantity of sodium chloride absorbed. The relative absorption of sodium chloride-the movement from the peritoneal cavity of sodium chloride, as compared with the movement of water-is slightly increased when 0.85 per cent. of sodium chloride solution and adrenalin are injected intraperitoneally; but it is diminished when adrenalin is injected intravenously, or when 1.5 per cent. sodium chloride solution and adrenalin are injected. When distilled water has been injected intraperitoneally, adrenalin decreases the relative amount of sodium chloride in the peritoneal fluid-a fact that is evidently related to the increased elimination of sodium chloride through the kidneys under the influence of adrenalin. 3. When 0.85 per cent. sodium chloride solution is injected into the peritoneal cavity, the blood becomes diluted after two hours and a half. When adrenalin is also injected, the dilution of the blood is less marked, in spite of the increased absorption under the influence of adrenalin. When distilled water is injected into the peritoneal cavity, the blood is diluted equally in control and adrenalin experiments. When 1.5 per cent. sodium chloride solution is injected, the dilution of the blood is very slight, and in adrenalin experiments it is the same as in control experiments or very slightly greater than in control experiments. 4. The increase of absorption from the peritoneal cavity caused by the injection of adrenalin is not due to the increased diuresis caused by the injection of this substance. 5. The injection of adrenalin causes a temporary increase in the osmotic pressure of the blood, which gradually returns to normal. Under certain conditions, after the injection of adrenalin, there is a tendency toward maintaining the higher osmotic pressure of the blood serum, even up to the end of the experiment. We have reason to believe that this increase in the osmotic pressure of the blood is the main factor in increasing the absorption of fluid from the peritoneal cavity. 6. In experiments in which 0.85 per cent. sodium chloride solution has been injected intraperitoneally, either with or without adrenalin, there exists a tendency of the peritoneal fluid to attain a greater osmotic pressure than the blood serum, in spite of the fact that the injected fluid is slightly hypotonic as compared with the blood serum. We note a similar condition in cases of general edema in man, in which the osmotic pressure of the ascitic fluid is greater than that of the other edematous fluids, or even that of the blood serum. There exists, therefore, a mechanism that causes the passage of osmotically active substances from the blood or from the tissues into the peritoneal cavity, and that causes the osmotic pressure of the peritoneal fluid to become higher than that of the blood. It follows from our experiments that this mechanism, which causes the ascites in edematous persons to have such a high osmotic pressure, is not dependent upon certain pathological changes in the lining membranes or upon other pathological conditions, but exists already in normal animals. 7. The addition of 1.22 per cent. calcium chloride solution to 0.83 per cent. sodium chloride solution, in such proportions as we used in our infusion experiments, in which we determined the transudation into the peritoneal cavity, delays the absorption of fluid from the peritoneal cavity but very slightly. Therefore, calcium chloride increases directly the transudation into the peritoneal cavity and does not cause an increase in the amount of fluid in the peritoneal cavity merely by inhibiting the absorption. 8. It follows that adrenalin does not increase the amount of peritoneal transudate found after the intravenous infusion of large quantities of sodium chloride solution, to which adrenalin has been added, by delaying the absorption from the peritoneal cavity. The increased amounts of peritoneal fluid must be due to increased transudation into the peritoneal cavity; and the adrenalin, in view of its marked effect on absorption from the peritoneal cavity, must increase the movement of fluid into the peritoneal cavity much more strongly than could be assumed from the figures obtained in the infusion experiments.