LEARNING BY MARINE INVERTEBRATES
45
achieves the result it gets, and selection of one set from a whole range
of appropriate movements may be a matter of chance.
Attempts to show that performance improves with practice have
mainly centred round the animals’ righting reactions and their capacity
to escape from obstacles restricting their movements. I n most cases the
time taken to turn over or escape was measured. Thus Preyer (1886)
showed that ophiuroids could extract arms from rubber tubes placed
around them. He interpreted the changes in the movements that the
animals made as a sign that the animals were in some way aware of the
problem and intelligently struggling to free themselves. Neither he nor
Glaser (1907) who repeated the experiments (but disagreed with the
interpretation) could detect any improvement in performance with
practice. I n a further series of experiments (for references see Thorpe,
1963) Preyer and others studied the attempts of ophiuroids and
asteroids to escape when confined by means of pegs driven in around
them, between the rays. The latest and most thorough of these works
is Ven’s (1921) series of experiments with Asterias. He showed first
that these animals most commonly lead with the arms on either side
of the madreporite (an observation since confirmed by Kerkut, 1955).
By trapping these two arms under hoops on a board (see Fig. 24), Ven
obliged his animals to lead with other arms in order to escape. Trials
were at l-S/day ; as tests continued “ the animals fatigued less easily ”
and the larger number of trials became possible. Considerable improvements in the time taken to escape were found, both for trials beginning
with a move in one of the right directions and for those starting with a
move in the direction of one of the trapped arms (Fig. 24).
It is a little difficult to interpret these experiments. Ven himself
states that the improved escape time was due : (1) to less time being
spent on false starts ; (2) to more activity on the part of the untrapped
arms ; and (3) to an increased suppleness, enabling the arms to bend up
over the pegs more readily. Taken together with the statement that
the animals fatigued less easily as tests continued, statements 2 and 3
make one wonder whether (for example) a starving starfish moves more
vigorously than a fed one. Of Ven’s six animals two died after showing
some improvement in performance and one failed to learn, “ pour lui
la puniton et la recompense Btaient trop faibles pour vaincre sa nature
trop indolente.” Are the other results explicable as changes in the
general condition of the animals? Against this conclusion are various
observations. First, there are occasions when training was resumed
after a 48-h break; performance after the breaks in the single experiment that Ven plots out in full tended to be unusually poor : one would
not expect this result if improvement in performance was just a
45
achieves the result it gets, and selection of one set from a whole range
of appropriate movements may be a matter of chance.
Attempts to show that performance improves with practice have
mainly centred round the animals’ righting reactions and their capacity
to escape from obstacles restricting their movements. I n most cases the
time taken to turn over or escape was measured. Thus Preyer (1886)
showed that ophiuroids could extract arms from rubber tubes placed
around them. He interpreted the changes in the movements that the
animals made as a sign that the animals were in some way aware of the
problem and intelligently struggling to free themselves. Neither he nor
Glaser (1907) who repeated the experiments (but disagreed with the
interpretation) could detect any improvement in performance with
practice. I n a further series of experiments (for references see Thorpe,
1963) Preyer and others studied the attempts of ophiuroids and
asteroids to escape when confined by means of pegs driven in around
them, between the rays. The latest and most thorough of these works
is Ven’s (1921) series of experiments with Asterias. He showed first
that these animals most commonly lead with the arms on either side
of the madreporite (an observation since confirmed by Kerkut, 1955).
By trapping these two arms under hoops on a board (see Fig. 24), Ven
obliged his animals to lead with other arms in order to escape. Trials
were at l-S/day ; as tests continued “ the animals fatigued less easily ”
and the larger number of trials became possible. Considerable improvements in the time taken to escape were found, both for trials beginning
with a move in one of the right directions and for those starting with a
move in the direction of one of the trapped arms (Fig. 24).
It is a little difficult to interpret these experiments. Ven himself
states that the improved escape time was due : (1) to less time being
spent on false starts ; (2) to more activity on the part of the untrapped
arms ; and (3) to an increased suppleness, enabling the arms to bend up
over the pegs more readily. Taken together with the statement that
the animals fatigued less easily as tests continued, statements 2 and 3
make one wonder whether (for example) a starving starfish moves more
vigorously than a fed one. Of Ven’s six animals two died after showing
some improvement in performance and one failed to learn, “ pour lui
la puniton et la recompense Btaient trop faibles pour vaincre sa nature
trop indolente.” Are the other results explicable as changes in the
general condition of the animals? Against this conclusion are various
observations. First, there are occasions when training was resumed
after a 48-h break; performance after the breaks in the single experiment that Ven plots out in full tended to be unusually poor : one would
not expect this result if improvement in performance was just a