LEARNINQ BY MARINE INVERTEBRATES
9
Sepia, which strike prey with two long tentacles, rarely do so at an
" incorrect " range ; if too close to the prey, they back away before
striking. There is evidence that the capacity to range prey and strike
correctly is innate, since in decapods a t least it does not measurably
improve with practice, being as accurate immediately after hatching as
subsequently (Wells, 1962b).
4. Discrimination of wavelength and plane of vibration of light
A number of experiments have been made on colour vision in
cephalopods, and in some of these octopuses have been trained to
respond to particular colours. The experiments, none of them recent,
have been reviewed in Wells (1962a). They may be summarized as
showing that cephalopods can, in all probability, learn to distinguish
the wavelength of light reaching the eye ; but none of the controls were
sufficiently stringent to be certain of this.
A number of more recent experiments have been concerned with
the capacity to distinguish the plane of vibration of light reaching the
retina. Octopuses, tit least, can learn to respond to differences in the
plane of polarization of light sources. As in shape discrimination
experiments, they most readily learn to respond to horizontal and
vertical differences ; discriminations between obliquely polarized
sources are learned less readily, a finding that appears to be related to
the predominantly horizontal and vertical orientation of the individual
retinal elements (Moody, 1962 ; Moody and Parriss, 1961). Further
evidence that polarized light discrimination is based on intraocular
detection of vibration plane rather than on patterns of reflexion in the
experimental tanks comes from training experiments following statocysts removal. As in shape discrimination, recognition of orientation
fails when retinal orientation is disrupted by the operation. Discrimination can be maintained only if the plane of the electric vector is matched
to the position of the retina at each trial (Rowel1 and Wells, 1961). It
is not known whether Octopus normally orients itself with respect to
polarized light, but some squids do (Jander et al., 1963) and it is at
least possible that this is one of the means by which octopuses find their
way about.
5. Detour experiments
Octopuses can be taught to make a detour in order to reach prey
that they can see but cannot approach directly. In the apparatus
shown in Fig. 6 , crabs could be seen but not touched by octopuses
confined to the home compartment. In order to get a crab, the animals
had to go out of sight of their prey, into the corridor and along, making
B *
9
Sepia, which strike prey with two long tentacles, rarely do so at an
" incorrect " range ; if too close to the prey, they back away before
striking. There is evidence that the capacity to range prey and strike
correctly is innate, since in decapods a t least it does not measurably
improve with practice, being as accurate immediately after hatching as
subsequently (Wells, 1962b).
4. Discrimination of wavelength and plane of vibration of light
A number of experiments have been made on colour vision in
cephalopods, and in some of these octopuses have been trained to
respond to particular colours. The experiments, none of them recent,
have been reviewed in Wells (1962a). They may be summarized as
showing that cephalopods can, in all probability, learn to distinguish
the wavelength of light reaching the eye ; but none of the controls were
sufficiently stringent to be certain of this.
A number of more recent experiments have been concerned with
the capacity to distinguish the plane of vibration of light reaching the
retina. Octopuses, tit least, can learn to respond to differences in the
plane of polarization of light sources. As in shape discrimination
experiments, they most readily learn to respond to horizontal and
vertical differences ; discriminations between obliquely polarized
sources are learned less readily, a finding that appears to be related to
the predominantly horizontal and vertical orientation of the individual
retinal elements (Moody, 1962 ; Moody and Parriss, 1961). Further
evidence that polarized light discrimination is based on intraocular
detection of vibration plane rather than on patterns of reflexion in the
experimental tanks comes from training experiments following statocysts removal. As in shape discrimination, recognition of orientation
fails when retinal orientation is disrupted by the operation. Discrimination can be maintained only if the plane of the electric vector is matched
to the position of the retina at each trial (Rowel1 and Wells, 1961). It
is not known whether Octopus normally orients itself with respect to
polarized light, but some squids do (Jander et al., 1963) and it is at
least possible that this is one of the means by which octopuses find their
way about.
5. Detour experiments
Octopuses can be taught to make a detour in order to reach prey
that they can see but cannot approach directly. In the apparatus
shown in Fig. 6 , crabs could be seen but not touched by octopuses
confined to the home compartment. In order to get a crab, the animals
had to go out of sight of their prey, into the corridor and along, making
B *