LElLRNINC? BY MARINE INVERTEBRATES
3
what are the mechanisms concerned. Most of the recent work has been
done with Octopus vulgaris Lamarck.
Since Boycott and Young’s first publication in 1950, there has been
a rapid increase in the literature, so that at the present time something
like ten or a dozen papers on learning in cephalopods appear every
year. Two reviews are available, Wells (1962a), and Young (1961).
Recent research has approached the subject by means of experiments
on sensory integration, and experiments involving brain lesions in
attempts to discover which parts are concerned in memory storage.
It is convenient to subdivide the present account on this basis.
A. Experiments on sensory integration
Here the concern has been firstly, “ What events in their external
and internal worlds can octopuses detect?” And following this “ What
happens to sensory information as it passes from the sense organs
towards those parts of the brain ultimately concerned with the accumulation of experience?” At the first level it is now clear that octopuses
can be taught to make a wide variety of visual and tactile discriminations. They can taste things that they touch and they are sensitive to
low-frequency vibrations, which they probably detect tactilely. They
do not respond to sonic frequencies (Hubbard, 1960), although electrophysiological evidence (Maturana and Sperling, 1963) indicates that
the statocysts can detect these, and they do not appear to have organs
of electrical sense or anything comparable to the lateral line pressure
receptors of vertebrates. They learn to recognize many patterns of
stimuli very rapidly and the performance of individuals, despite their
being taken from wild populations, is sufficiently consistent for the
relative difficulty of discriminations to be compared from the number
of trials required to attain given standards of accuracy of response.
1. Detectable attributes of things seen
Octopus vulgaris Lamarck of 200-800 g will live, feed, grow and
even breed in aquaria, given a brisk circulation and some means of
preventing the animals from climbing out. Experimentally, asbestos
tanks 100 x 60 x 40 cm, with a “ home ” of bricks at one end, have
proved convenient. The animals generally sit in the home, looking
down the tank. Boycott and Young (1950) showed that it was possible
to train octopuses kept in this way to distinguish between crabs
shown alone and crabs shown together with plastic or metal shapes.
The animals were allowed to seize and eat the crabs in the one situation
and given small electric shocks (6-9 V A.c.) for attacking in the other.
The result of a typical training experiment, with the two situations
3
what are the mechanisms concerned. Most of the recent work has been
done with Octopus vulgaris Lamarck.
Since Boycott and Young’s first publication in 1950, there has been
a rapid increase in the literature, so that at the present time something
like ten or a dozen papers on learning in cephalopods appear every
year. Two reviews are available, Wells (1962a), and Young (1961).
Recent research has approached the subject by means of experiments
on sensory integration, and experiments involving brain lesions in
attempts to discover which parts are concerned in memory storage.
It is convenient to subdivide the present account on this basis.
A. Experiments on sensory integration
Here the concern has been firstly, “ What events in their external
and internal worlds can octopuses detect?” And following this “ What
happens to sensory information as it passes from the sense organs
towards those parts of the brain ultimately concerned with the accumulation of experience?” At the first level it is now clear that octopuses
can be taught to make a wide variety of visual and tactile discriminations. They can taste things that they touch and they are sensitive to
low-frequency vibrations, which they probably detect tactilely. They
do not respond to sonic frequencies (Hubbard, 1960), although electrophysiological evidence (Maturana and Sperling, 1963) indicates that
the statocysts can detect these, and they do not appear to have organs
of electrical sense or anything comparable to the lateral line pressure
receptors of vertebrates. They learn to recognize many patterns of
stimuli very rapidly and the performance of individuals, despite their
being taken from wild populations, is sufficiently consistent for the
relative difficulty of discriminations to be compared from the number
of trials required to attain given standards of accuracy of response.
1. Detectable attributes of things seen
Octopus vulgaris Lamarck of 200-800 g will live, feed, grow and
even breed in aquaria, given a brisk circulation and some means of
preventing the animals from climbing out. Experimentally, asbestos
tanks 100 x 60 x 40 cm, with a “ home ” of bricks at one end, have
proved convenient. The animals generally sit in the home, looking
down the tank. Boycott and Young (1950) showed that it was possible
to train octopuses kept in this way to distinguish between crabs
shown alone and crabs shown together with plastic or metal shapes.
The animals were allowed to seize and eat the crabs in the one situation
and given small electric shocks (6-9 V A.c.) for attacking in the other.
The result of a typical training experiment, with the two situations