The rate of tritium loss from ring-labelled resorcinol to the solvent has been measured for aqueous solutions of resorcinol, containing either added hydrochloric acid or sodium hydroxide, at 13·5 °C. The experimental technique involved isotopic assays of samples of water vapour in equilibrium with the solution, i.e., without disturbance of the acid–base equilibria or concentrations of solutes.

The measured rate constants represent fairly closely tritium loss from the equivalent 4- and 6-positions.

In the presence of acid, the first-order exchange rate constant is proportional to the concentration of hydrogen ion, indicating that exchange involves attack of hydrogen ion on the resorcinol molecule. In the presence of alkali, the first-order constant depends both on the concentration of free hydroxide ion in solution and on the concentration of free resorcinol, in an apparently complex fashion. This behaviour is quantitatively accounted for by the assumption that the following pairs of catalyst acid and substrate species make detectable contributions to the observed exchange at various concentrations (H₂R resorcinol): H₃O⁺ and H₂R, H₂R and R^2–, HR^– and R^2–, H₂O and R^2–, and probably H₃O⁺ and HR^–. The corresponding second-order rate constants have been evaluated by calculations based in part on a multiple regression of the first-order exchange rate constant on the stoicheiometric concentration of resorcinol and various powers of the concentration of free hydroxide ion. The required values of the first and second dissociation constants of resorcinol at 13·5 °C have been approximately evaluated (pK₁= 9·45, pK₂= 11·71) the former by potentiometric titration and the latter by measurement of the catalytic effect of the disodium salt of resorcinol on the depolymerisation of diacetone alcohol.

The analysis implies the observation of general acid catalysis in a system in which the substrate alone provides the components of the buffer.