11. ELECTRORECEPTION
547
meaningful measure of stimulus strength than is the potential gradient,
that is, voltage per unit length along the fish. Because of accessory structures, an applied voltage gradient multiplied by the diameter of a receptor cell gives a very great underestimate of the actual potential across
the receptors.
The simplest example is provided by the skate, the skin of which is
of relatively low resistance. When the fish is placed in a uniform field,
the field is not greatly distorted within the fish, although the body is of
somewhat higher resistivity than seawater (Murray, 1967). As noted
above, canals of the ampullae of Lorenzini have walls of very high resistance, and the space constants are quite long compared to the lengths
( Waltman, 1966). As far as dc potentials are concerned, the potential in
the ampulla itself must be very close to that at the external opening of
the canal. The receptor thus detects the difference in potential between
the canal opening and the serosal side of the receptor cell. The receptor
is like a voltmeter with its terminals at these points. If one assumes a uniform field, the stimulus to the receptor is given by the field (in volts
per unit length) times the distance between receptor cells and canal opening times the cosine of the angle between the field and the axis of the
canal. The stimulus is greatest if the canal is oriented in the direction of
the field; the stimulus is zero if the canal is oriented perpendicular to the
field.
The fish obtains directional information by having many receptors
with canals oriented in different directions (Fig. 6 ) . Canals are of different lengths, and receptors with long canals are sensitive to smaller uniform gradients but provide less spatial resolution. Receptors with long
canals are also less sensitive to higher frequencies because of the capacity
of the canal walls (Waltman, 1966). There may be a sacrifice of frequency response for sensitivity in these receptors, which could represent
a useful adaptation to overcome noise. However, response of the receptor
cells drops off at high frequencies in any case (Murray, 1965) and the
more important limiting factor is undetermined.
When uniform gradients are applied to mormyrids and gymnotids
the situation is more complex. The skin is of high resistance and the inside of the fish is of much lower resistance than the external medium.
The voltage gradients within the fish are thus much smaller than those
in the external medium, although they have not been determined in detail.
The individual receptors detect the difference between the external and
the internal potential, where the latter is some average of the potential
over the whole fish. In a uniform gradient the stimulus to the receptors
depends on the maximum voltage across the animal. If the skin resistance is uniformly high, about one-half the maximum voltage across the
animal is developed across the skin at either end (or side) in the direc-
547
meaningful measure of stimulus strength than is the potential gradient,
that is, voltage per unit length along the fish. Because of accessory structures, an applied voltage gradient multiplied by the diameter of a receptor cell gives a very great underestimate of the actual potential across
the receptors.
The simplest example is provided by the skate, the skin of which is
of relatively low resistance. When the fish is placed in a uniform field,
the field is not greatly distorted within the fish, although the body is of
somewhat higher resistivity than seawater (Murray, 1967). As noted
above, canals of the ampullae of Lorenzini have walls of very high resistance, and the space constants are quite long compared to the lengths
( Waltman, 1966). As far as dc potentials are concerned, the potential in
the ampulla itself must be very close to that at the external opening of
the canal. The receptor thus detects the difference in potential between
the canal opening and the serosal side of the receptor cell. The receptor
is like a voltmeter with its terminals at these points. If one assumes a uniform field, the stimulus to the receptor is given by the field (in volts
per unit length) times the distance between receptor cells and canal opening times the cosine of the angle between the field and the axis of the
canal. The stimulus is greatest if the canal is oriented in the direction of
the field; the stimulus is zero if the canal is oriented perpendicular to the
field.
The fish obtains directional information by having many receptors
with canals oriented in different directions (Fig. 6 ) . Canals are of different lengths, and receptors with long canals are sensitive to smaller uniform gradients but provide less spatial resolution. Receptors with long
canals are also less sensitive to higher frequencies because of the capacity
of the canal walls (Waltman, 1966). There may be a sacrifice of frequency response for sensitivity in these receptors, which could represent
a useful adaptation to overcome noise. However, response of the receptor
cells drops off at high frequencies in any case (Murray, 1965) and the
more important limiting factor is undetermined.
When uniform gradients are applied to mormyrids and gymnotids
the situation is more complex. The skin is of high resistance and the inside of the fish is of much lower resistance than the external medium.
The voltage gradients within the fish are thus much smaller than those
in the external medium, although they have not been determined in detail.
The individual receptors detect the difference between the external and
the internal potential, where the latter is some average of the potential
over the whole fish. In a uniform gradient the stimulus to the receptors
depends on the maximum voltage across the animal. If the skin resistance is uniformly high, about one-half the maximum voltage across the
animal is developed across the skin at either end (or side) in the direc-