In atrial wall trabeculae of Rana catesbeiana and R. ridibunda very slowly decaying membrane currents have been consistently observed in decay tails following voltage clamp depolarizing and hyperpolarizing pulses. It is not thought that these currents are carried by time-dependent conductance channels but rather result from potassium ion accumulation or depletion. 2. Since voltage clamp techniques generally impose a non-physiological situation on the membranes of excitable cells, evidence that potassium ion accumulation occurs in unclamped atrial tissue is presented. 3. When potassium ions accumulate, the reversal potentials for both atrial delayed conductance mechanisms, ixfast and ixslow, should be shifted in a positive direction, the magnitude of the shifts being a function of the charge transferred during depolarization. Experiments have been performed to test this prediction quantitatively, and as a result, a simple accumulation model is developed. 4. A second important effect of accumulation should be upon the time-independent potassium conductance, iK1. It was found that this effect produces current tails whose decay becomes exponential when the amount of accumulation is small. The time constant of this exponential is shown to be equal to the time constant of decay of accumulation, tauacc. 5. One of the most important assumptions in the accumulation model is that the iK1(Em) relations for different values of [K]O 'cross-over' one another as they do in skeletal muscle and mammalian Purkinje tissue. Experimental verification of this assumption is presented. This 'cross-over' effect allows current changes due to accumulation to show an apparent 'reversal potential' and so to appear like a conductance mechanism. 6. Potassium depletion is shown to occur during hyperpolarizing pulses. This depletion process must be allowed for in a direct kinetic analysis of the pace-maker current, ixslow, at potentials negative to the resting potential (ER).
