pH and rate of ‘dark’ events in toad retinal rods: test of a hypothesis on the molecular origin of photoreceptor noise