1 1. ELECTROFECEPTION
539
It is tempting to ascribe the two kinds of external response to the two
kinds of receptor cell. The responses at onset of anodal and termination
of cathodal stimuli probably represent either delayed rectification or
graded regenerative responses of ordinary polarity in the inner faces of
one kind of receptor cell. The responses at onset of cathodal and termination of anodal stimuli are unusual in that their latency changes little as
stimulus strength is increased. They do not appear to be anode break
responses and may represent increases in resistance during hyperpolarization similar to those observed in a number of other cells (Bennett and
Grundfest, 1966). The brevity of these responses as compared to the
others suggests that the blocking capacity in series with the generating
membrane is different for the two response types. This observation is
consistent with the responses being generated in the two kinds of receptor
cell and with location of the blocking capacity in the outer faces of the
cells. The briefer response would be most likely to arise in the outer receptor cells, the external surface of which is smaller and presumably of
lower capacity. Nerve fibers that innervate both classes of receptor would
then respond to either direction of change in stimulating voltage; those
innervating only the inner receptors would respond only to onset of
anodal and termination of cathodal stimuli. The relatively small outer
face of the outer receptor cells would presumably not be important because of the low sensitivity of these cells.
The lack of correspondence between external response and afferent
impulses suggests that transmission from receptor cell to innervating
nerve is chemically mediated as at the phasic receptors of gymnotids.
This inference is supported by the fine structure of the receptor synapses
(Barets and Szabo, 1964) and by the presence of a synaptic delay of
about 1 msec (Steinbach and Bennett, 1971).
A limited amount of pharmacological data has been obtained from
medium receptors ( Steinbach and Bennett, 1971). High magnesium
solutions reduce transmitter secretion at neuromuscular junctions and
interneuronal synapses, presumably by competing with Ca2+ at some stage
of the release process (cf. Katz, 1969). These solutions greatly increase
the stimulus strength required to evoke afferent impulses from medium
receptors. Nerve responses are not eliminated completely; at least a single
impulse always remains. A tentative explanation is that Mg2+ can, to a
small extent, substitute for Ca2+ in the release process.
Transmission at medium receptors is not cholinergic; curare, atropine,
and acetycholine have no effect. The levorotatory form of glutamate has
been implicated as a transmitter at a number of synapses (Beranek and
Miller, 1968; Usherwood and Machili, 1968). The medium receptors
are excited by L-glutamate and by the related L-aspartate, but they are
not affected by D-glutamate. In procedures to date the concentrations
539
It is tempting to ascribe the two kinds of external response to the two
kinds of receptor cell. The responses at onset of anodal and termination
of cathodal stimuli probably represent either delayed rectification or
graded regenerative responses of ordinary polarity in the inner faces of
one kind of receptor cell. The responses at onset of cathodal and termination of anodal stimuli are unusual in that their latency changes little as
stimulus strength is increased. They do not appear to be anode break
responses and may represent increases in resistance during hyperpolarization similar to those observed in a number of other cells (Bennett and
Grundfest, 1966). The brevity of these responses as compared to the
others suggests that the blocking capacity in series with the generating
membrane is different for the two response types. This observation is
consistent with the responses being generated in the two kinds of receptor
cell and with location of the blocking capacity in the outer faces of the
cells. The briefer response would be most likely to arise in the outer receptor cells, the external surface of which is smaller and presumably of
lower capacity. Nerve fibers that innervate both classes of receptor would
then respond to either direction of change in stimulating voltage; those
innervating only the inner receptors would respond only to onset of
anodal and termination of cathodal stimuli. The relatively small outer
face of the outer receptor cells would presumably not be important because of the low sensitivity of these cells.
The lack of correspondence between external response and afferent
impulses suggests that transmission from receptor cell to innervating
nerve is chemically mediated as at the phasic receptors of gymnotids.
This inference is supported by the fine structure of the receptor synapses
(Barets and Szabo, 1964) and by the presence of a synaptic delay of
about 1 msec (Steinbach and Bennett, 1971).
A limited amount of pharmacological data has been obtained from
medium receptors ( Steinbach and Bennett, 1971). High magnesium
solutions reduce transmitter secretion at neuromuscular junctions and
interneuronal synapses, presumably by competing with Ca2+ at some stage
of the release process (cf. Katz, 1969). These solutions greatly increase
the stimulus strength required to evoke afferent impulses from medium
receptors. Nerve responses are not eliminated completely; at least a single
impulse always remains. A tentative explanation is that Mg2+ can, to a
small extent, substitute for Ca2+ in the release process.
Transmission at medium receptors is not cholinergic; curare, atropine,
and acetycholine have no effect. The levorotatory form of glutamate has
been implicated as a transmitter at a number of synapses (Beranek and
Miller, 1968; Usherwood and Machili, 1968). The medium receptors
are excited by L-glutamate and by the related L-aspartate, but they are
not affected by D-glutamate. In procedures to date the concentrations