The purpose of this study has been to determine what conditions maintain the high level of osmotic pressure within liver cells present immediately after their removal from the body, to define the relation of water movement to this pressure, and to learn if intracellular osmotic pressure and water movement are changed when amino acids and related substances are transformed to produce urea. Significant information concerning the activity of liver or of kidney cells as osmometers is obtainable immediately after the tissue is removed from the body.1 With relation to a wide variety of electrolytes, liver cells acting as osmometers measure the osmotic pressure of solutions in accord with the molecular weight, valence, and ion-dissociation of each substance. Tables (Robinson and Stokes2) assembled from freezing point depression, vapor tension, electrical conductivity, and other colligative properties of electrolytes have made it possible to define the molar concentration of electrolyte solutions with the same osmotic pressure. A solution of sodium chloride isotonic for liver immediately after its removal from the body is approximately 0.34 molar and a similar solution for kidney, 0.23 molar. Water equilibrium is maintained by the tissue at these levels during 15 to 20 min. Electrolytes as varied as potassium chloride, with monovalent base, magnesium chloride, with bivalent base, and lanthanum chloride with trivalent base, are isotonic with liver or kidney tissue in solutions that have the same osmotic pressure as that of sodium chloride in solution isotonic with one or other of these tissues. The tissues, acting as osmometers, measure osmotic pressure with approximate accuracy. The plasma membrane of liver cells is evidently imperfectly semipermeable to solutions of electrolytes and with their entrance into the cell permeability to both electrolyte and water is increased. Maintenance of isotonicity in vitro is promoted by immersion in media with electrolytes similar to those of blood plasma, by temperature at 38? C and by an adequate supply of oxygen.5 Slices of liver tissue, weighing preferably from 60 to 90 mg, determined rapidly by means of a torsion balance, have been immersed in the Krebs3 modification of Ringer's solution much used for manometric measurement of oxygen consumption. The solution with electrolyte concentration of 0.154 molar has been buffered with potassium phosphate and treated with oxygen or buffered with sodium bicarbonate and gassed with oxygen, 95 per cent, and carbon dioxide, 5 per cent. The medium is hypotonic for liver tissue but when its molar concentration is increased two-fold by addition of sodium chloride isotonicity may be maintained during one or two hr.4 Water movement measured in per cent of weight before immersion varies widely with oxygen supply to the medium. With abundant oxygen under conditions that promote its diffusion throughout the medium and into the tissue, water enters rapidly during about one hr but later is lost and within 2 or 3 hr may be less than that before immersion. With anoxia caused by diminished oxygen supply or by substitution of nitrogen for oxygen, water intake during the first hr of immersion