M. V. L. BENNETT
514
voltage is developed across the series resistances of the ampullary canal,
the receptor cell cytoplasm, and the animal's interior. The impulse frequency evoked by a step excitatory stimulus often accelerates gradually
over several impulses, which could represent gradual charging of the
capacity of the inner faces. The decreased responsiveness of tonic receptors at high frequencies accounts for the fact that these receptors are unaffected by the electric organ in the many fish in which the organ discharge contains little or no low frequency component (Bennett, 1970;
Bullock and Chichibu, 1965; Roth, 1967; Suga, 1967a; Szabo, 1970).
Transmitter released by the receptor cells presumably causes inward
current to flow through the subsynaptic membranes of the nerve fiber
and thereby initiates nerve impulses in nearby spike-generating membrane. The nerve impulses do not appear to involve the terminals themselves since they are monophasic and positive outside the receptor opening and have a duration similar to that of the intracellularly recorded
action potentials in nerve fibers (Fig. 15; Bennett, 1965). These observations suggest that the terminal lacks spike-generating membrane and that
its chemosensitive membrane is electrically inexcitable like that at most
other synapses ( Bennett, 1964; Grundfest, 1966). This conclusion is supported by the observation that the nerve fiber can be directly stimulated
by an anodal, but not by a cathodal, stimulus applied externally (Fig.
15; Bennett, 1967).
The frequency of nerve impulses should provide a reasonable measure
of the transmitter concentration, although there may be some distortion
because of accommodative processes such as refractoriness of the spikegenerating membrane and desensitization of the subsynaptic membrane.
One can conclude that in spite of distortions the relationship between presynaptic potential and transmitter release has the same shape as the graph
of Fig. 14. There is a steady release of transmitter that is accelerated or
retarded by stimuli of opposite polarity, and for small signals the response is linearly related to the stimulus strength. This excitation-secretion relation is strikingly different from that at the squid giant synapse
( Katz and Miledi, 1967b), the neuromuscular junction ( Katz and Miledi,
1967a), and the Mauthner fiber, giant fiber synapse in the hatchetfish
(Auerbach and Bennett, 1969). In these synapses there is little transmitter release for small stimuli, and at some approximate threshold the
rate of release rises sharply.
Another important characteristic of the input-output relationship of
the receptor synapses is sensitivity. A very small range of stimulation,
about 2 mV in Fig. 14, encompasses most of the range of transmitter
secretion. The corresponding range of presynaptic potential for the squid
synapse is about 50 mV. It is also likely that the postsynaptic potential
514
voltage is developed across the series resistances of the ampullary canal,
the receptor cell cytoplasm, and the animal's interior. The impulse frequency evoked by a step excitatory stimulus often accelerates gradually
over several impulses, which could represent gradual charging of the
capacity of the inner faces. The decreased responsiveness of tonic receptors at high frequencies accounts for the fact that these receptors are unaffected by the electric organ in the many fish in which the organ discharge contains little or no low frequency component (Bennett, 1970;
Bullock and Chichibu, 1965; Roth, 1967; Suga, 1967a; Szabo, 1970).
Transmitter released by the receptor cells presumably causes inward
current to flow through the subsynaptic membranes of the nerve fiber
and thereby initiates nerve impulses in nearby spike-generating membrane. The nerve impulses do not appear to involve the terminals themselves since they are monophasic and positive outside the receptor opening and have a duration similar to that of the intracellularly recorded
action potentials in nerve fibers (Fig. 15; Bennett, 1965). These observations suggest that the terminal lacks spike-generating membrane and that
its chemosensitive membrane is electrically inexcitable like that at most
other synapses ( Bennett, 1964; Grundfest, 1966). This conclusion is supported by the observation that the nerve fiber can be directly stimulated
by an anodal, but not by a cathodal, stimulus applied externally (Fig.
15; Bennett, 1967).
The frequency of nerve impulses should provide a reasonable measure
of the transmitter concentration, although there may be some distortion
because of accommodative processes such as refractoriness of the spikegenerating membrane and desensitization of the subsynaptic membrane.
One can conclude that in spite of distortions the relationship between presynaptic potential and transmitter release has the same shape as the graph
of Fig. 14. There is a steady release of transmitter that is accelerated or
retarded by stimuli of opposite polarity, and for small signals the response is linearly related to the stimulus strength. This excitation-secretion relation is strikingly different from that at the squid giant synapse
( Katz and Miledi, 1967b), the neuromuscular junction ( Katz and Miledi,
1967a), and the Mauthner fiber, giant fiber synapse in the hatchetfish
(Auerbach and Bennett, 1969). In these synapses there is little transmitter release for small stimuli, and at some approximate threshold the
rate of release rises sharply.
Another important characteristic of the input-output relationship of
the receptor synapses is sensitivity. A very small range of stimulation,
about 2 mV in Fig. 14, encompasses most of the range of transmitter
secretion. The corresponding range of presynaptic potential for the squid
synapse is about 50 mV. It is also likely that the postsynaptic potential