A perfusion method is described whereby large discs of amphistomatous leaves are vacuum-perfused with water so that either successive fractions of perfusate may be analyzed for solutes or the infused water may be displaced and collected after equilibration with the leaf cells. With castor bean leaves, estimates of electrolyte concentration in cell wall water by the two methods were similar. Total electrolytes in leaf cell wall water of castor beans (Ricinus communis), sunflower (Helianthus annuus), and cabbage (Brassica oleracea capitata) from nonsaline cultures were about 2, 2, and 10 milliequivalents per liter, respectively, increasing to 4, 10, and 30 milliequivalents per liter under saline conditions. Electrolytes recovered in successive fractions were similar in composition, and continuous perfusion resulted in a steady release of solutes, the concentration in the perfusate varying inversely with the perfusion rate. Diffusional release of solutes from cells was less than expected at low perfusion rates, suggesting that solute reabsorption may increase as solute concentration in the perfusate increases with decreased perfusion rates. Perfusate concentration and composition were essentially unaffected by temperature (2 and 23 C) or by perfusing with 0.5 mM CaSO4 rather than with water. Electrolytes in perfusates on an equivalent basis were Ca2", 30%; Mg2+, 10%; and Na+ + K+, 60%, the proportions of sodium increasing from 10 to 50% in leaves (cabbage) that accumulated sodium under saline conditions. Salinity (added NaCl) of the root culture medium caused a 3-to 5-fold increase in total cell wall electrolyte concentration, but this amounted to an increase from less than 1 or a few per cent to no more than 7% (in cabbage) of the cell sap electrolyte concentrations. Solutes in the cell wall appear to be in dynamic equilibrium with intracellular solutes.In a tissue at water equilibrium, the water potential is the same at all points and there is no net movement of water from any point to another. The components of water potential, however, may vary. The intracellular pressure component (turgor) is absent at air interfaces of cells, and the negative potential components (solute plus matric) will, therefore, have a smaller absolute magnitude than inside the turgid cell. Average values for the potential components (12) for the regions of even a single cell are meaningless in view of the wide latitude that is possible. For any given region of the cell, the components of water potential must be specifically determined. This is very much the case for the cell walls of leaves and other aerial organs whose total water potential is directly measured whenever the water potential of the tissue is studied but whose water potential components, matric and solute, are completely unknown. Data on the relationship of matric potential to water content of killed plant tissues (3, 11) are not helpful in determining the matric potential in cell walls. Even if the water content determination were specific for the cell wall, the...