LEARNING BY MARIN'E INVERTEBRATES
23
in activity with cycle lengths varying from a few minutes to an hour or
more ; the worms are never inactive and their activities include spontaneous withdrawal into their tubes. Most of the training procedures
used have involved training long enough to permit interference from
cyclic fluctuations in the tendency to withdraw spontaneously. Thus,
for example, Yerkes (1906), working on Hydroides dianthw (Verrill),
comments that her worms, normally withdrawing for 26 sec or so after
a tactile stimulus, periodically stayed inside their tubes for 4-12 min
instead. She gives a table showing a typical series of results (it shows,
incidentally, that Hydroiderr eventually habituates to touch as well as
to light stimuli) and from this it appears that the unusually long periods
of withdrawal occurred at intervals of 16-20 min, i.e. with a period comparable with the cycles since shown in other worms. Yerkes interpreted
these results as periodic changes to alternative behaviour in the face of
apparently unavoidable repeated stimuli.
2. The ekctrophy8iology of habituatbn
The giant fibre responses of worms disappear when the evoking
stimuli are repeated. It is relatively easy to record from giant fibres,
and so discover whether habituation is a result of changes occurring
before or after this stage in the action system. The results vary with
the species. Thus it seems that habituation in nereids could be the
result of non-transmission between the sensory elements and the giant
fibres or between the latter and the segmental motor fibres (Horridge,
1969), while in tubiculous polychaetes such as Branchiomma and
Myxicola it is certainly a property of parts of the nervous system involved before the giant fibres themselves. In Hyxicoh, where the giant
fibres branch and run direct to the longitudinal muscles, more than
100 shocks to these nerves at 1 per sec are needed to produce an
appreciable decline in the muscular response. With a longer time
between shocks the number needed is further increased ; clearly this is
far too many for habituation to be accountable as a neuromuscular
phenomenon (Roberts, 1962). In a further analysis of the situation in
Branchiomma, Krasne ( 1966) has likewise ruled out sensory accommodation as an explanation of the failure to respond when repeatedly
stimulated. Responses, having once been eliminated, are not restored
when a Werent part of the crown is touched: the labile part of the
system is after the sense organs themselves, and prior to the giant
fibres. It seems that in Branchiomma withdrawal can be sudden,
governed by the giant fibre system, or more gradual, the organization
of the contraction involving other nervous pathways. With repeated
stimulation the giant fibre response fails first (Fig. 14).
23
in activity with cycle lengths varying from a few minutes to an hour or
more ; the worms are never inactive and their activities include spontaneous withdrawal into their tubes. Most of the training procedures
used have involved training long enough to permit interference from
cyclic fluctuations in the tendency to withdraw spontaneously. Thus,
for example, Yerkes (1906), working on Hydroides dianthw (Verrill),
comments that her worms, normally withdrawing for 26 sec or so after
a tactile stimulus, periodically stayed inside their tubes for 4-12 min
instead. She gives a table showing a typical series of results (it shows,
incidentally, that Hydroiderr eventually habituates to touch as well as
to light stimuli) and from this it appears that the unusually long periods
of withdrawal occurred at intervals of 16-20 min, i.e. with a period comparable with the cycles since shown in other worms. Yerkes interpreted
these results as periodic changes to alternative behaviour in the face of
apparently unavoidable repeated stimuli.
2. The ekctrophy8iology of habituatbn
The giant fibre responses of worms disappear when the evoking
stimuli are repeated. It is relatively easy to record from giant fibres,
and so discover whether habituation is a result of changes occurring
before or after this stage in the action system. The results vary with
the species. Thus it seems that habituation in nereids could be the
result of non-transmission between the sensory elements and the giant
fibres or between the latter and the segmental motor fibres (Horridge,
1969), while in tubiculous polychaetes such as Branchiomma and
Myxicola it is certainly a property of parts of the nervous system involved before the giant fibres themselves. In Hyxicoh, where the giant
fibres branch and run direct to the longitudinal muscles, more than
100 shocks to these nerves at 1 per sec are needed to produce an
appreciable decline in the muscular response. With a longer time
between shocks the number needed is further increased ; clearly this is
far too many for habituation to be accountable as a neuromuscular
phenomenon (Roberts, 1962). In a further analysis of the situation in
Branchiomma, Krasne ( 1966) has likewise ruled out sensory accommodation as an explanation of the failure to respond when repeatedly
stimulated. Responses, having once been eliminated, are not restored
when a Werent part of the crown is touched: the labile part of the
system is after the sense organs themselves, and prior to the giant
fibres. It seems that in Branchiomma withdrawal can be sudden,
governed by the giant fibre system, or more gradual, the organization
of the contraction involving other nervous pathways. With repeated
stimulation the giant fibre response fails first (Fig. 14).