560
M. V. L. BENNETT
mV/cm (Moller, 1970). These gradients would give a receptor stimulus
of about 0.5 mV in a 20-cm long fish which is above threshold for most
large receptors ( Bennett, 1965, 1967). Thresholds for locally applied
stimuli to evoke an acceleration of organ discharge are similar to the
thresholds of large receptors (Harder et al., 1967).
The jamming avoidance response of gymnotids has a sensitivity of 3100 pV/cm (Watanabe and Takeda, 1963). In these experiments the
receptor stimulus would have been about five times the applied voltage
per centimeter. The larger values easily exceed the absolute thresholds
of the receptors measured directly ( Bennett, 1967).
The relevant parameter in the jamming avoidance response is the
incremental threshold of the receptors when the applied stimulus is
superimposed on the stimulus provided by the organ discharge. This
property has not been studied in fish emitting discharges at a fairly
constant frequency. In Gyrnnotus incremental thresholds are not much
different from absolute thresholds ( Suga, 1967a). However, in phasic
receptors of Hypopornus, the stimulus required to evoke a single spike
is much greater than the increase in stimulus required to add a spike
to a train evoked by a moderately suprathreshold stimulus (Hagiwara
and Morita, 1963). Correspondingly, the animal responds to a weaker
applied pulse when it sums with the electric organ discharge than when
it is presented between discharges (Bullock, 1970; Larimer and MacDonald, 1968). In the mormyrid Gnathonemus the behavioral responses
to electric pulses have a threshold at least 10-fold higher when the pulses
are superimposed on the organ discharge compared to when they are
given between discharges ( Moller, 1970). Although for some medium
receptors incremental thresholds are lower than absolute thresholds
(Szabo and Hagiwara, 1967), the large receptors no doubt mediate the
lower threshold behavioral responses, and their sensitivity to changes
in the organ discharge is not yet clear. The “significance” to the fish
may also differ in that pulses between organ discharges could represent
another electric fish while pulses superimposed on organ discharges could
represent resistance changes.
The conclusion at this time is that there is a factor of approximately
ten between observed behavioral and receptor thresholds. Direct measurements of the stimuli applied to the receptors under the conditions of
the behavioral experiments are still required and may alter this factor
somewhat. At least part of the discrepancy must lie in the difficulties
in electrophysiological experiments of finding the lowest threshold
receptors and of maintaining the animal in optimal condition.
A further aspect of the discrepancy may be correlation by the central
nervous system of inputs from a number of receptors. In the presence of
M. V. L. BENNETT
mV/cm (Moller, 1970). These gradients would give a receptor stimulus
of about 0.5 mV in a 20-cm long fish which is above threshold for most
large receptors ( Bennett, 1965, 1967). Thresholds for locally applied
stimuli to evoke an acceleration of organ discharge are similar to the
thresholds of large receptors (Harder et al., 1967).
The jamming avoidance response of gymnotids has a sensitivity of 3100 pV/cm (Watanabe and Takeda, 1963). In these experiments the
receptor stimulus would have been about five times the applied voltage
per centimeter. The larger values easily exceed the absolute thresholds
of the receptors measured directly ( Bennett, 1967).
The relevant parameter in the jamming avoidance response is the
incremental threshold of the receptors when the applied stimulus is
superimposed on the stimulus provided by the organ discharge. This
property has not been studied in fish emitting discharges at a fairly
constant frequency. In Gyrnnotus incremental thresholds are not much
different from absolute thresholds ( Suga, 1967a). However, in phasic
receptors of Hypopornus, the stimulus required to evoke a single spike
is much greater than the increase in stimulus required to add a spike
to a train evoked by a moderately suprathreshold stimulus (Hagiwara
and Morita, 1963). Correspondingly, the animal responds to a weaker
applied pulse when it sums with the electric organ discharge than when
it is presented between discharges (Bullock, 1970; Larimer and MacDonald, 1968). In the mormyrid Gnathonemus the behavioral responses
to electric pulses have a threshold at least 10-fold higher when the pulses
are superimposed on the organ discharge compared to when they are
given between discharges ( Moller, 1970). Although for some medium
receptors incremental thresholds are lower than absolute thresholds
(Szabo and Hagiwara, 1967), the large receptors no doubt mediate the
lower threshold behavioral responses, and their sensitivity to changes
in the organ discharge is not yet clear. The “significance” to the fish
may also differ in that pulses between organ discharges could represent
another electric fish while pulses superimposed on organ discharges could
represent resistance changes.
The conclusion at this time is that there is a factor of approximately
ten between observed behavioral and receptor thresholds. Direct measurements of the stimuli applied to the receptors under the conditions of
the behavioral experiments are still required and may alter this factor
somewhat. At least part of the discrepancy must lie in the difficulties
in electrophysiological experiments of finding the lowest threshold
receptors and of maintaining the animal in optimal condition.
A further aspect of the discrepancy may be correlation by the central
nervous system of inputs from a number of receptors. In the presence of