LEARNING B Y W I N E INVERTEBRATES
49
some conflict about the interpretation of the results. Thus Jennings
(1905) found that Aiptasia, having learned to reject the paper, generally
refused real food as well, and attributed the whole phenomenon to a
difference in the degree of stimulation. Allabach (1905), using Metridium, suggested that the cessation of response could be due to accumulation of mucus on the tentacles. Neither of these, however, seems a
likely explanation of the results of Fleure and Walton (1907) who found,
with Actinia and Tealia, that the change of the response to filter paper
not only occurred in experiments at one trial per day (surely long
enough for a reversal of sensory accommodation or the dispersion of
mucus?) but remained detectable up to a week after the end of training.
Their animals learned not to pass the paper to the mouth within 3-6
days and 2 days later refused even to take hold of i t ; Tealia learned
more rapidly than Actinia. Fleure and Walton found, moreover, that
the effect of the training was limited to the tentacles immediately
involved ; other tentacles would take the filter paper once or twice when
first tested, a result that shows at least that their results are not
attributable to a steady decline in the condition of the animals. Clearly
these experiments should be repeated, under conditions designed to
show : (1) that the animals remain equally responsive to other stimuli
at the same time as the decline in response to imitation food; and (2)
that there are no alternative explanations of the change in behaviour in
terms of trial timing in relation to tidal or other rhythmicities that the
animals may exhibit. Until this is done, the experiments can only be
summarized as appearing to show learning by experience.
Like many other animals, coelenterates commonly show long-term
rhythmic behaviour when brought into the laboratory from the sea.
Diurnal and tidal rhythms sometimes persist for many days under
apparently constant conditions. Thus Cavernularia, an alcyonarian
sea-pen,” expands at “ night ” and closes up during the “ day ”
for a t least 100 days in constant laboratory conditions (Mori, 1959).
Actinia in contrast, shows tidal rhythms that die &way within 2-3 days,
or sooner, according to the markedness of tides in the places from which
they were collected. A number of other examples are known (for
references see Harker, 1958). Taken collectively, there seems to be
little doubt that environmental effects can influence rhythms in
coelenterates. Whether their influence is superimposed on innate
tendencies to cycle with a circadian or tidal periodicity is another
matter, not known for coelenterates.
The position habits of individual anemones would appear to
indicate learning by experience. Two sorts have been noted. One,
discussed by Jennings (1905), is the tendency of individual animale t o
6 6
49
some conflict about the interpretation of the results. Thus Jennings
(1905) found that Aiptasia, having learned to reject the paper, generally
refused real food as well, and attributed the whole phenomenon to a
difference in the degree of stimulation. Allabach (1905), using Metridium, suggested that the cessation of response could be due to accumulation of mucus on the tentacles. Neither of these, however, seems a
likely explanation of the results of Fleure and Walton (1907) who found,
with Actinia and Tealia, that the change of the response to filter paper
not only occurred in experiments at one trial per day (surely long
enough for a reversal of sensory accommodation or the dispersion of
mucus?) but remained detectable up to a week after the end of training.
Their animals learned not to pass the paper to the mouth within 3-6
days and 2 days later refused even to take hold of i t ; Tealia learned
more rapidly than Actinia. Fleure and Walton found, moreover, that
the effect of the training was limited to the tentacles immediately
involved ; other tentacles would take the filter paper once or twice when
first tested, a result that shows at least that their results are not
attributable to a steady decline in the condition of the animals. Clearly
these experiments should be repeated, under conditions designed to
show : (1) that the animals remain equally responsive to other stimuli
at the same time as the decline in response to imitation food; and (2)
that there are no alternative explanations of the change in behaviour in
terms of trial timing in relation to tidal or other rhythmicities that the
animals may exhibit. Until this is done, the experiments can only be
summarized as appearing to show learning by experience.
Like many other animals, coelenterates commonly show long-term
rhythmic behaviour when brought into the laboratory from the sea.
Diurnal and tidal rhythms sometimes persist for many days under
apparently constant conditions. Thus Cavernularia, an alcyonarian
sea-pen,” expands at “ night ” and closes up during the “ day ”
for a t least 100 days in constant laboratory conditions (Mori, 1959).
Actinia in contrast, shows tidal rhythms that die &way within 2-3 days,
or sooner, according to the markedness of tides in the places from which
they were collected. A number of other examples are known (for
references see Harker, 1958). Taken collectively, there seems to be
little doubt that environmental effects can influence rhythms in
coelenterates. Whether their influence is superimposed on innate
tendencies to cycle with a circadian or tidal periodicity is another
matter, not known for coelenterates.
The position habits of individual anemones would appear to
indicate learning by experience. Two sorts have been noted. One,
discussed by Jennings (1905), is the tendency of individual animale t o
6 6