Abstract
End-plate potentials produced by brief trains of action potentials (5-7 at 50-100 Hz) were recorded at toad [Bufo marinus] sciatic-sartorius neuromuscular junctions. When transmitter secretion was depressed in solutions containing Mg2+, the increase in amplitude (growth pattern) of successive end-plate potentials was greater than could be accounted for by arithmetic summation of facilitation (arithmetic model) as proposed by Mallart and Martin. With EPP of normal quantal content or in solutions in which the Ca2+ concentration was lowered, growth patterns were occasionally reasonably close to those predicted by the arithmetic model but there was always some degree of disparity. A simple, 2-step, kinetic model was described which was more consistent with the varied growth patterns of end-plate potentials that were recorded. The model could predict growth patterns of EPP with high or low quantal content.