Abstract
1. The histology of the heart muscle of Squilla mantis is briefly described. Vital staining with methylene blue revealed only a sparse distribution of nerve endings on the muscle network. 2. Intracellular electrodes recorded from the muscle a multi-peaked junction potential at each heart beat. Each peak followed an impulse in the ganglionic nerve trunk burst. All the peaks were approximately the same height and none more than about three-fifths the height of the resting potential (average values for 10 hearts: resting potential, 51.5 mV; junction potential, 27.6 mV). 3. Inverted (negative-going) signals were never recorded just outside the muscle membrane suggesting that at all the points which were searched the membrane was acting passively. 4. Driving Squilla heart muscle via its nerve supply at 100 stimuli per second did not depolarize it by more than about 35 mV, nor would depolarizing pulses given directly to a fibre through an intracellular electrode set up any sort of current-generating activity in the membrane. 5. The magnitude of muscle contraction, measured locally using a microelectrode transducer, depended on the absolute level of the potential across the membrane, rather than on change of potential. 6. A directly applied electrotonus, similar in magnitude and duration to the nerve-induced junction potential, caused a local contraction of similar magnitude. 7. The recorded junction potential is therefore interpreted as the composite record of the electrotonus spreading within the muscle network from current initiated at relatively infrequent active points on the muscle membrane (the nerve endings) which passively depolarizes the rest of the membrane. 8. The junction potentials showed facilitation when the intervals between them were below 4 sec. At intervals less than 630 msec, they summed.