1. The tension generated by isolated frog atrial trabecules, during exposure to solutions containing a high
potassium concentration, is not maintained but spontaneously relaxes. The final part of this relaxation can be fitted by a single exponential function.2. The recovery of the tension generating mechanisms following the spontaneous relaxation of a
potassium contracture depends on the preceding membrane potential and the time since the last
contracture.3. The rate of the exponential phase of the spontaneous relaxation is independent of the [K](o) and hence the membrane potential, the [Ca](o); and when the [Ca](o)/[Na](o) (2) ratio is maintained it is also independent of the [Na](o). This relaxation is not influenced by
atropine or
pronethalol.4. When
sodium is totally excluded from the bathing medium the rate of relaxation of a later
potassium contracture is much increased. It is argued that this change is due to a fall in the intracellular
sodium concentration.5. The consequences of these results are discussed, and the hypothesis that is favoured would require that contraction is induced by a transient release of
calcium into the sarcoplasm, probably triggered by a potential dependent, and probably also transient, influx of
calcium through the cell membrane. Relaxation is supposed to occur when this activator-
calcium is then removed by an intracellular relaxing system that resembles the sarcoplasmic reticulum of other muscles. What this intracellular structure might be, is also discussed.