The slow reactions between phosphate and soil were studied by incubating, at a range of temperatures, soils to which phosphate had been added in solution. After various intervals, solutions which were 0.01 M with respect to calcium chloride, and which contained phosphate, were shaken briefly with the soils and the concentration in equilibrium with the soils was interpolated. The rate of change of phosphate concentration was described by combining the following equation with an adsorption isotherm: (Equation present) where (1 − α) is the proportion of the added phosphate which remains in a form which equilibrates with the solution, k varies with temperature, and n is a coefficient. The effects of temperature were described by the Arrhenius equation and the value of its coefficients were such that a 10° increase in temperature caused an approximately threefold increase in rate. When the temperature was changed during incubation, the rate of reaction changed to a value appropriate to the new temperature. When repeated applications of phosphate were made, the concentration of phosphate was found to be a little higher than predicted on the assumption that the effects were additive. This was attributed to blocking of some of the adsorption sites by the previous additions. While high temperatures accelerated the changes which followed the initial adsorption and thus favored low solution concentration, the effects of temperature on the equilibrium between adsorbed and solution phosphate was found to be in the opposite direction. High temperatures favored high solution concentration—that is, the adsorption was exothermic. When soil was incubated with phosphate it was possible to separate these opposing effects of temperature, but when soil was shaken with phsophate solutions, both effects of temperature were present. The effects of temperature on the changes in solution concentration were therefore smaller when soil was shaken with phosphate than under incubation conditions. On the other hand, the decrease in concentration tended to be faster at a given temperature. This was attributed to effects of shaking in breaking down soil particles.