1. In the experiments recorded in this paper the influence of the osmotic pressure of the blood upon absorption of fluid from the peritoneal cavity becomes apparent. Nephrectomy, removal of the adrenals, and other operations increase the osmotic pressure of the blood and increase the absorption of fluid from the peritoneal cavity. On the other hand, ether narcosis, at the period at which we tested its influence, causes neither an increase of osmotic pressure of the blood nor an increase in the absorption of fluid from the peritoneal cavity. 2. The increased osmotic pressure and increased absorption of fluid in nephrectomized animals is to a great extent not a specific effect of the removal of the kidneys, but approximately the same conditions can be observed after incisions of the skin and muscles. 3. After poisoning with uranium nitrate and in cases of peritonitis, complicating factors come into play, and under such conditions the absorption from the peritoneal cavity is not increased, notwithstanding the higher osmotic pressure of the blood. 4. In conditions in which the osmotic pressure of the blood is very high before the injection of sodium chloride solution into the peritoneal cavity (nephrectomized rabbits or rabbits injected with uranium nitrate three days previously), adrenalin causes no increase, or only a very slight one, in the absorption of peritoneal fluid. On the other hand, one day after the injection of uranium nitrate the osmotic pressure of the blood is only slightly increased before the injection of the sodium chloride solution into the peritoneal cavity, and here adrenalin causes a marked increase in absorption of fluid from the peritoneal cavity. 5. In animals injected with uranium nitrate the retention of sodium chloride and other osmotically active substances in the blood is not entirely due to interference with the functions of the kidney. This retention may be explained either by an inability of the tissues to bind the sodium chloride and other osmotically active substances or to a diminished permeability of the blood vessels for such substances. 6. While in nephrectornized animals the elimination of sodium chloride from the peritoneal cavity and also from the blood is increased, in animals injected with uranium nitrate such an elimination is diminished. This increase in the sodium chloride content of the peritoneal fluid in animals treated with uranium nitrate is accompanied by a decrease in the diffusion of other osmotically active substances into the peritoneal cavity. 7. While in nephrectomized animals and in animals injected with uranium nitrate one day previously, adrenalin causes a diminution of the fluid retained in the blood-vessels similar to the diminution noted in normal animals, adrenalin no longer exerts such an effect at a later stage of the uranium nitrate poisoning. At this period after the administration of uranium nitrate, the retention of fluid in the blood vessels is apparently equal in experiments with and without the injection of adrenalin, and following the absorption of fluid from the peritoneal cavity, the retention of fluid in the blood vessels in the uranium nitrate animals is increased comparatively to a greater extent than in normal animals. 8. Our experiments show a marked difference in the distribution of fluid and of osmotically active substances in nephrectomized animals and in animals injected with uranium nitrate. This difference may explain the much greater liability to the development of edema in animals injected with uranium nitrate.