1. Prepotentials and action potentials were recorded from amphibian striated muscle fibers. Intracellular electrodes were used for stimulating and recording. The resting potential was varied from 55 to 120 mv. by alterations of the KCl concentration of the Ringer's fluid. The magnitude of the prepotential at the initiation of the spike potential was measured and compared to the resting potential and the latent period (time between stimulus "make" and excitation). The magnitude of this prepotential varied with the resting potential.
2. A large prepotential or cathodal depolarization was required to excite a fiber with a high resting potential. If a fiber with a high resting potential fired late (long latency), the adequate prepotential was larger than if the fiber fired early. Fibers with low resting potentials had smaller adequate prepotentials. Also, the adequate prepotential was independent of the latent period, in these depolarized fibers.
3. If the concentration of Ca++ was increased tenfold, the adequate prepotential of depolarized fibers became strongly dependent upon the latency.
4. Fibers with large or normal resting potentials were prone to respond repetitively during the passage of long duration shock, whereas depolarized and Ca++-treated fibers were not.
5. The so-called critical membrane potential (which is defined as the transmembrane potential at the point of excitation) was not independent of the resting potential.