Abstract
1 End-plate potentials (e.p.p.) have been recorded from the neuromuscular junctions of frog sartorius and extensor longus dig. IV muscles, using intracellular micropipettes. Either curare or MgCl2 were present in the Ringer solution, to keep the e.p.p. amplitude below the threshold for a muscle action potential and contraction. 2 It has been shown that saxitoxin (paralytic shellfish poison) usually caused a progressive reduction in the amplitude of the e.p.p. Occasionally, when it was applied in the presence of MgCl2, the e.p.p. disappeared abruptly. 3 Tetrodotoxin usually caused the e.p.p. to disappear abruptly. Occasionally, when applied in the presence of curare, the e.p.p. declined progressively for a short time before disappearing abruptly. 4 It is concluded that at the frog neuromuscular junction the preferential site of action of saxitoxin is at the nerve terminals, but tetrodotoxin preferentially blocks nerve conduction at a site proximal to the junction. 5 It is suggested that this preparation would be a convenient and reliable test object for distinguishing saxitoxin from tetrodotoxin.