Phases of Activity in the Sea-Anemone, Metridium Senile (L.), and Their Relation to External Stimuli

We have shown in a previous paper (Batham & Pantin, 1950b) that the sea-anemone Metridium senile exhibits continual very slow muscular activity. It is most evident in the changes of shape of the column caused by the reciprocal contractions of the circular muscle sheet, and of the longitudinal parietal musculature which runs at the junction of each mesentery with the body wall. This inherent activity varies, so that in the same individual it presents a different pattern on different occasions. The pattern may remain constant for hours on end, and may then change over a comparatively short period. Thus a moderately filled anemone giving fairly regular and substantial contractions may pass into a condition with a larger volume and much less apparent activity, or it may pass into a condition of prolonged and almost complete contraction. These patterns of activity often endure for long periods—hours or even days. It is convenient to refer to such patterns as ‘phases’, and these phases form the subject of this paper.

Our conclusions are based on direct observation, quick-motion cinematography and kymographic records. These methods have been discussed fully in the previous paper (Batham & Pantin, 1950b). For kymograph tracings, long light isotonic levers were used. When continuous tracings for days or weeks were required, slow clockwork drums were employed, and the writing levers were arranged to swing at right-angles to the surfaces of the smoked drums. For all the tracings figured in this paper, the recording thread ran vertically from the sphincter region of the anemone to the lever. Hence the records represent chiefly the vertical activity of the body wall (Fig. 1 a). All tracings have been photographed and arranged in the figures so that they read from left to right, as indicated by arrows, and so that contractions are registered downwards. To imprôve reproduction, the long fine tracings of the records were broadened by being thrown slightly out of focus and photographed on panchromatic process plates through a red filter. Throughout the work described in this paper the specimens of Metridium used were of medium size, the diameters of their pedal disks lying between 2 and 4·5 cm. They were collected from low spring tide level at Mill Bay and Salcombe, Devon, a few days or weeks before they were used. Very prolonged recordings were made under controlled conditions described in a subsequent section.

The phase of activity of an animal may sometimes change, as we shall see, without any evident external stimulus. In other cases the new phase may be clearly engendered by a particular stimulus to which its character bears a significant relation. This happens in the response of the animal to the presence of food. Fig. 2 records an instance of this. At first the animal was in a somewhat contracted state showing fairly regular alternations of parietal contraction, and re-extension by contraction of the circular muscle. At the point marked ‘Fed’, about 1 g. of tissue of Mytilus was placed on the disk. The mouth opened by contraction of the radial muscles of the perfect mesenteries (Batham & Pantin, 1950a), and the food was completely swallowed within 10 min. (Fig. 1 b). The immediate mechanism of appropriation of food by anemones and its ingestion has been discussed by Parker (1919) and others (Pantin & Pantin, 1943). We need not add to these accounts.

The sequence of events following ingestion in the experiment recorded in Fig. 2 was typical. The first effect of ingestion was relaxation of the sphincter and the flowering out of the capitulum, the disk and the tentacles. Thirteen minutes later contractions of the circular muscle appeared in the column (Fig. 1,c). Quick-motion cinematograph films of this stage show active, though very slow peristalsis. By pressure on the coelenteric fluid, this peristalsis causes elongation of the column. Elongation in the case illustrated in Fig. 2 was almost complete in an hour, during which time the volume of the animal had increased as well as its height.

This change of height and volume resulted from a change of the animal’s phase of activity. Examination of such cases as this shows that during elongation, the number of alternating contractile events in the antagonistic muscle system of the column which take place in a given time does not undergo great change. But there is marked change in their relative extent. The parietal contractions are now much smaller, and may do no more than check extension, whilst the circular contractions are more extensive and more powerful.

In Fig. 2 the animal remained in this phase of activity for nearly 3 hr., when a new phase supervened. At first unilateral (X, Y, Z) bending movements took place. In more marked instances of this, a tall Metridium very slowly sweeps or sways over and round to one side after another (Fig. 1 d, e). These movements are sometimes just fast enough to be appreciable to the eye. They seem to be caused in part by unilateral contractions of the parietal and the parieto-basilar muscles. Cinematograph records also make it clear that they are in part due to loss of circular tone, as a result of which the coelenteric turgor is unable to support the animal and it collapses under its own weight. When in this state each wave of circular contraction causes a re-erection of the column followed by a fresh collapse.

In the experiment of Fig. 2, large parietal contractions began to occur after the initial swaying, so that the elongated animal was now very active. These contractions resulted in some shortening and widening of the shape.

Three to six hours after a meal, a Metridium has increased greatly in volume and settled down to a phase of only slight apparent activity (cine film). This seeming lack of activity, however, may be partly of mechanical origin. When the animal is greatly distended, muscular contraction of the body wall causes a rise in pressure against which the muscles cannot easily shorten (Batham & Pantin, 1950a).

In due course, further changes of phase take place in connexion with defaecation. The ingested food is sometimes extruded only partially digested. In one series of observations on twenty-six animals this occurred in ten of them between 24 and 40 hr. after ingestion. More usually faecal pellets surrounded by thick mucus are extruded within 48 hr. Immediately after defaecation ‘shrivelling’ takes place (Batham & Pantin, 1950a). This is a rapid and generalized contraction of the whole musculature with loss of coelenteric water so that the anemone becomes a shrivelled caricature of its normal state (Fig. 1,i). Shrivelling may be complete in less than 3 min., after which the animal refills itself; a process taking about 2 hr. in instances observed. Subsequent acts of defaecation occur during the fortnight following a meal ; the extruded faeces becoming smaller and darker. A yellow excretory ring of mucus and uric acid (Fox & Pantin, 1941) develops rapidly round the column 1 to 2 days after a meal. Similar rings develop in starved animals but much more slowly (Fig. 1 k).

Defaecation occupies a phase of great activity. Fig. 3 shows the extensive contractions and re-extensions of the body wall which lead up to the process. These are accompanied by great activity of the circular muscle. Deep constrictions may remain localized or travel up or down the column. About half a dozen antiperistaltic waves, each occupying several minutes, may be succeeded by some peristaltic waves. Then further antiperistaltic activity may be shown. A succession of these events leads to defaecation and the sudden loss of water by generalized contraction (‘shrivelling’), which is followed by re-inflation (Fig. 1 f-j).

We see from these observations that ingestion of food is followed by a sequence of phases. There is expansion of the disk, often elongation of the column, sometimes swaying of the column, followed by inflation with some shortening, defaecation, shrivelling and a return to normal. Each of these phases has its own pattern of activity and most may last an hour or more.

This sequence is set in motion by a stimulus : the receipt of food. But the relation between stimulus and response is not that of a simple reflex, like the response of the retractor muscles following strong stimulation ; it may involve very different neuromuscular machinery. To begin with, the response is not a simple contraction, but a prolonged and irregular pattern or sequence of patterns of activity. Secondly, these phases or patterns of activity are not simply built up from specific responses to a specific sequence of stimuli. We may contrast them with the response sequence involved in the capture of food by the tentacles (Pantin & Pantin, 1943). In that case, contact of a piece of solid food with a tentacle produces a direct local response. This in turn leads to a combined chemical and mechanical stimulation of other nearby tentacles which transfer the food over the disk towards the mouth. This engenders new local stimuli of the disk and finally of the mouth, which result in ingestion. Each event of the sequence in this case is conditioned by the appropriate local chemical and mechanical stimuli. If the relation of the mechanical and chemical components of the stimulus is artificially changed, the orderly character of the response is upset. If the stimulus (food object) is removed, the response sequence is brought to a premature end (Pantin & Pantin, 1943).

In contrast with this, while the phases of activity we are now considering may be initiated by specific stimuli, they can on occasion be initiated by stimuli of other kinds. Indeed they may appear in the absence of any apparent external stimulus. Shrivelling may be initiated by other stimuli than the presence of food in the last stages of digestion ; strong and prolonged stimuli such as repeated intense illumination may initiate it. Swaying, shrivelling and other phasic activities may appear in unstimulated animals, or to stimuli not previously effective. Thus, in Fig. 11 a starved Metridium contracting and expanding to daily illumination, in a manner to be described later, abruptly passed into the shrivelled state 1 hr. after the beginning of illumination on the third recorded day.

The patterns of these phasic activities are in some way implicit in the ‘excitoreffector ‘machinery of the animal. They do not depend on a chain of specific external stimuli. Even when an external stimulus initiates a phase, it merely releases a pattern of activity which the animal is quite capable of executing in its absence. The stimulus, as it were, merely pulls a ‘hair-trigger’.

Notwithstanding this limited connexion between the external stimulus and the pattern of phasic activity, the phases take place in an orderly sequence, and, as quick-motion cinematography of the food reaction makes clear, the sequence of phases is often evidently purposive and functionally significant. What ensures that each phase is followed by the next in the series remains to be determined. We can only point out that individual phases of the food-response series can on other occasions appear by themselves or in a different sequence: as in the alternating phases of expansion and retraction in relation to light and dark to be discussed below.

This relation of stimulus to the phase initiated is made clear by examining the effect, not of solid food, but of food solutions. It has long been recognized that the presence of dissolved substances from food could cause expansion of the disk in sea-anemones (Pollock, 1883; Parker, 1919). This effect is frequently, though by no means always, to be observed.

In the following experiments the stimulating solution was obtained from soft parts of Mytilus edulis, or of Helix aspersa, or the epithelium of Loligo. These were ground with sand and with sea water and the resulting fluid filtered through cloth. The sensitivity of different Metridium to such solutions varied greatly, but the addition of sufficient extract to raise the concentration of dry organic matter in the medium to about 1 part in 10⁶ was generally sufficient.

Fig. 4 shows the effect of Mytilus extract, at about 1 in 10⁶ dry weight in the medium, on a specimen of Metridium. The animal was at first in clean sea water and had assumed a phase of moderate contraction and activity. Within 5 min. of addition of the extract the disk and tentacles began to expand. The tentacles showed some twitching movements. During the next 2 hr. the column of the anemone greatly elongated. This took place in just the same way as in animals fed on solid food. The frequency of contractile activity remained approximately unchanged, but the waves of circular muscular contraction increased in intensity whilst the extent of the reciprocal parietal contractions was greatly diminished.

Two hours after the addition of the food extract.the water was completely changed three times. The state of expansion was in no way diminished, and after an hour or so a phase of strong parietal contractions began to develop, just as occurs at the corresponding time after a meal of solid food. And, as in that case, elongation was followed by swaying movements. Swaying is often seen in such experiments, and it is noteworthy that whereas it seems to have little functional significance when engendered by a meal of solid food, it appears strikingly purposive as a response to dissolved food material. By sweeping the ground with its tentacles the anemone by such movements increases the chance of ‘finding’ the prospective meal the dissolved food might signify.

The first part of the record ends hr. after the removal of the food extract. By this time the anemone ceased to be elongated, though now moderately expanded. The record begins again about 2 hr. later, and at the point indicated Mytilus extract was once more added to the medium to a concentration of 1 in 10⁶. There now resulted an even more rapid elongation of the column than on the first occasion. The experiment was then discontinued.

There are some important features to note in this. In the first place, the sequence of phases almost precisely resembles that of the first few hours after the acceptance of solid food. The succession of phases and their nature does not therefore depend upon the stimulus of a solid food object in the actinopharynx or coelenteron.

Not only can the succession of phase changes be caused by a filtered food extract as well as by solid food, but it will be seen from Fig. 4 that the change of phase continues for hours after the removal of the food extract from the external medium. The ability of a stimulus to cause a change of phase which endures long after its termination is seen in Fig. 5 in a case of another kind. The animal in this experiment was moderately open and decidedly active, as the frequent large parietal contractions signify. During the period marked ‘Shocks’, condenser shocks were administered to the column at a frequency of six a minute for 20 min. This caused an increased state of parietal contraction which slowly passed into extension. For several hours after the stimulus, the pattern of activity remained decidedly different from what it was before stimulation. Parietal contractions were much less frequent.

The various phases of activity are thus initiated rather than maintained by the stimulus. Whether a stimulus will initiate a phase, however, varies according to the state of the animal. The complete sequence of phases in response to food substances only takes place in some individuals and at certain times. These differences may in part depend on the temporary shape of the anemone. There is some evidence that suggests that it is primarily the tentacles that receive the stimulus leading to phasic expansion as a response to food substances. Anemones that are completely contracted down (Fig. 1 l) show the feeding response much less frequently and more slowly than those in which the disk is exposed. For instance, in one experiment twenty-three previously starved Metridium were exposed to food extract whilst under prolonged observation in a large aquarium. Of these twenty-three, sixteen had their sphincters contracted, thereby covering the tentacles more or less completely at the time when the extract was added. During the next 6 hr. the other seven anemones whose tentacles had been exposed each showed a greater or lesser degree of expansion. Of the sixteen closed specimens, however, four opened very slightly and the other twelve all remained completely closed. In other tests it was found that the injection of food extract through the body wall into the coelenteron was less effective in producing phasic expansion than its addition to the external water bathing the surface of the tentacles.

But the threshold of phase changes to food extracts does not simply depend on the degree of exposure of the tentacles. It is true that, as in the above experiment, individuals whose tentacles are covered by the contracted marginal sphincter usually fail to respond to food extract, while those animals whose tentacles are exposed all give some response. But in animals with exposed tentacles and of the same apparent shape there are great differences in the extent of expansion of the disk and even greater differences in the incidence of other phases of the food sequence. Whilst all may give the expansion of the disk and tentacles, not all give the subsequent elongation of the column. Others again may omit other phases of the sequence, such as the swaying of the column. The threshold for the various phases may vary greatly.

The threshold for a reflex response may vary owing to sensory adaptation. Repeated application of a stimulus may thereby temporarily reduce its effectiveness. This can be seen in the tentacular responses to food (Parker, 1919; Pantin & Pantin, 1943). But the changes of threshold of phasic activity are not solely to be ascribed to this factor. When the animal is in a suitable state, a phase may be initiated repeatedly. We have seen in Fig. 4 that the elongation reaction could be induced with ease twice in succession at an interval of 10 hr. The reaction was again strongly evoked in the same anemone 5 days later. Only after a meal and previous prolonged exposure to food extract (on the 6th day) did a repetition of the experiment elicit expansion of the disk alone, and fail to cause elongation of the column. In contrast with this there were many other cases where elongation of the column was never induced, even though there had been no recent previous exposure to food extract. Such variations of threshold cannot simply be attributed to sensory adaptation. There must be, in addition, some other, internal, factor which governs the ease with which a phase is initiated, and which varies. We must suppose the existence of some such factor when we find the phases of elongation and swaying occurring spontaneously in apparently unstimulated animals. Similarly shrivelling, a phase normally following defaecation, has been several times observed in animals maintained in clean sea water under constant conditions of light and temperature.

We thus come to consider these phases as patterns of activity potentially present in the animal independently of specific external stimuli. The stimuli merely release them. The threshold for their release varies, and may become so low that the phase occurs spontaneously. Moreover, the same phase may be initiated by more than one kind of stimulus.

One essential factor which governs the threshold of initiation of a phase depends upon what we can only at present call the ‘state’ of the animal. Starved animals are certainly in a different physiological state from well-fed ones, and on the whole starvation progressively lowers the threshold for initiation of the phases associated with feeding, and vice versa. How it comes about that the neuromuscular system can assume different physiological patterns remains to be seen. But we may note some degree of analogy to the changes of pattern in visceral activity in the Vertebrata which follow changes in endocrine balance.

In concluding this section we would say that in our experience Metridium will on occasions readily take large pieces of solid food. That it will not always do so is, we believe, satisfactorily accounted for by the variation of its state from time to time. It is not alone amongst carnivores in only being ‘on the feed’ at intervals. Its behaviour on receiving solid food both when witnessed by the eye and by cinematography gives no grounds for the supposition that ingestion of large food masses is an ‘abnormal ‘process. It may well be that it commonly ingests planktonic organisms but we see no reason to consider it purely a specialized planktonic feeder (cf. Elmhirst, 1925).

A phasic activity of a different kind from that which we have just discussed is locomotion. As McClendon (1906) and Parker (1917) state, this takes place through exceedingly slow waves of contraction passing across the foot from behind forwards in the direction of locomotion. This is vividly shown by accelerated cine-photography of walking specimens. Parker finds that the direction of locomotion bears no relation to the siphonoglyph axis of the animal. He also remarks upon the independence between this activity of the foot and the presence of the rest of the animal ; animals with the oral disk removed walked as readily as entire ones.

We shall discuss the mechanism of locomotion in a later paper based on the analysis of cinematograph records. Here we shall confine ourselves to consideration of the conditions under which locomotion occurs. In an aquarium, the position of the foot of Metridium may remain unchanged for weeks. In the field, the relative positions of members of an asexual clone of one colour suggests that they remain on one site for even longer periods. As Gosse (i860) remarks, some older naturalists actually supposed that the animals were permanently fixed to the substratum. In such cases, sections show the foot to be attached to the substratum by a layer of hardened protein material. Nevertheless, Metridium kept in aquaria not infrequently show intermittent locomotor phases (Flatteley & Walton, 1922).

Our own observations suggest that locomotion may be preceded by a condition of pedal instability. Though remaining on the same general site, the foot then shows irregular extensions and retractions in various directions, so that its outline progressively changes over a period of hours. If locomotion supervenes, this localized activity becomes co-ordinated and directed.

Though locomotion shows co-ordination and direction, it may begin under conditions of constant temperature and light intensity and the absence of any evident external stimulus. Thus a Metridium was placed without food in a dimly lit aquarium 8 days after it had been fed. It attached itself and remained stationary until the night of the 13th day, during which it walked. It walked again on the 34th and 35th days. On the 36th day it shrivelled and recovered. On the 37th day it again walked but thereafter remained stationary till the 46th day.

Occasionally locomotion in Metridium seems directly initiated in response to a prolonged adverse stimulus. It was often observed that an anemone subjected to repeated electrical or mechanical stimulation or to prolonged very powerful illumination or to the sewing-in of a recording thread would begin walking during the following night. Similarly, a freshly collected animal whose foot has pieces of the substratum still adhering to it often walks for some days until the pedal disk is cleared of debris.

Locomotion initiated by a stimulus may at times bear no directional relation to the stimulus, but it is commonly away from it. In certain cases, stimuli which had nothing to do with initiating locomotion may come to control its direction. Thus there is a frequent tendency for locomotory animals that reach the sides of an aquarium to travel upwards vertically till they reach the surface of the water, as though gravity can in some way act as a directing force. Possibly the slight extra tension on the uppermost muscles is here the directing stimulus.

The threshold for the initiation of locomotion is influenced by some general factors. A large meal is often followed by locomotion. Thus eight anemones from one clone were starved for a week. They were then placed in separate bowls of sea water in a cellar under constant conditions. Four were at this juncture given half a Mytilus and the rest left unfed. All were observed daily at 9 a.m. and 9 p.m. for 3 weeks except upon 3 days. During these 12 hr. intervals the four fed animals walked on two, three, five and six occasions respectively. Of the four unfed ones, one walked twice and the rest remained stationary.

Stimuli may not only initiate a phase ; they may inhibit it. Apart from the tendency for a strong local stimulus by light to initiate locomotion, general illumination tends to inhibit this activity. It was repeatedly observed that locomotion, when it occurred, almost always did so at night. Thus a starved anemone in an aquarium exposed to daylight began to move on 11 August 1946. It remained stationary during the daytime, but travelled 11·0, 11·5 and 6·5 cm. on three successive nights, by which time it had reached the surface-level of the water. After that no further locomotion occurred during the next 5 days and nights.

The locomotory phase is an activity pattern of considerable complexity. Notwithstanding Parker’s observations of isolated pedal disks, cinematograph and other records show that the muscular waves across the pedal disk are accompanied by co-ordinated constrictions running down the column and extensive vertical activity. The latter is shown in Fig. 6, a tracing of the vertical activity of an anemone exposed to alternating 12 hr. periods of light and dark. During the first period of darkness figured, a locomotor phase began and the animal moved across the stone to which it had attached itself. The greater frequency of the contractions of the column during this phase is conspicuous.

As in the phase of activity of the column associated with feeding, we have in locomotion a co-ordinated pattern of activity which can appear in the apparent absence of any stimulus. It may follow a stimulus, but the phase may endure for hours after its cessation. The stimulus initiates the phase, but its continuation is not necessary for the phase to be maintained. The threshold of the phase varies and can be affected by a number of factors, such as food which lowers the threshold and light which raises it. These features are characteristic of phasic activity.

That many sea-anemones exhibit alternating phases of prolonged general expansion and retraction is well known (Parker, 1919). This alternation has sometimes been correlated with the sequence of night and day, whilst other workers have stressed its relation to the tide. Parker has shown that there is a well-marked rhythm frequently to be observed in Metridium marginatum, and has related this to the daily rhythm of light and darkness. As he points out, Metridium is undoubtedly sensitive to light, and we can fully confirm this, though, as Cotte (1922) remarks, the responsiveness of individual actinians to light is very variable.

The direct action of light on a Metridium which has been in darkness causes a slow contraction of the parietal musculature. This response is local if the illumination is local, resulting in contraction of that sector of the parietal wall on which the light falls. It thus differs from the generalized parietal contraction caused by electrical stimulation (Hall & Pantin, 1937). If the illumination is general and prolonged, the parietal contraction is followed in unfed anemones by a withdrawal of the tentacles, closing of the sphincter, and further shortening of the column ; the animal then maintaining a phase of prolonged contraction during which inherent activity takes place with the anemone in a much greater state of average tone.

In the following experiments we are primarily concerned with the influence of light and darkness on rhythmic phases of contraction and of expansion of the order of 12 hr. duration. Hence environmental conditions had to be controlled as far as was practicable for periods of days or even weeks. Freshly collected anemones were first kept for a week without food. Then each, after it had become attached to a stone, was placed separately in a glass aquarium tank holding about 15 1. of clean sea-water. This was kept gently circulating and aerated by compressed air. The glass tanks were placed in a cellar hewn out of rock, where extraneous mechanical disturbances were minimal, illumination could be controlled, and temperature variations were slight (varying by only 1-2⁰ C. over periods up to 6 weeks). A stimulus that could not be avoided was the attachment of the recording thread to the anemone. This was done by a fine glass hook. The hook was very gradually extruded from the tissue by a process of local destruction of the tissue on one side and regeneration on the other, so that over long periods re-hooking was necessary. However, a single ‘hooking’ might last for a week or more.

Sometimes hooking initiated locomotion on the succeeding night. But in general it caused no more than a temporary contraction, and effects seen in the records could all be confirmed in normal animals without hooks attached. Breaks in the tracings shown in Figs. 8, 9 and 11 are periods not recorded because the hook had been sloughed off. In nearly every instance the anemone remained in an apparently healthy condition, and the water remained clean throughout the period of the record.

Illumination was effected by a 150 W. clear glass light bulb at distances ranging from 160 to 225 cm. from the anemones. During illumination, the recorded contractions could be related to the state of the animal by direct observation. During continuous darkness this naturally could not be done. In some cases the animals were examined occasionally during darkness with the aid of a red hand-light rendered as dim as could be used by a dark-adapted observer. In one case, and only one, was there any indication that this extremely weak illumination affected the animal in any way. One animal showed a slight contraction on two occasions when exposed to the dim hand-light.

Many, but not all, individuals of Metridium show a marked diurnal phasic rhythm, not only in their natural habitat but also when placed in an aquarium exposed to daylight. Fig. 7 shows a typical case. During the night the continual inherent activity of the animal took place about a phase of expansion, and by day about a phase of retraction. The aquarium was in this case set up exposed to daylight in a north-facing window of the Marine Biological Laboratory, Plymouth. Such a clear rhythm as this is moderately common, though by no means universal.

The phasic rhythm of animals similar to that shown in the experiment illustrated in Fig. 7 is controlled by light. Fig. 8 shows part of the tracing of the activity of a starved Metridium, recorded over a period of 48 days in all. The temperatures recorded twice daily during this very long experiment ranged from 16·2 to 16·8° C. The aquarium could be illuminated when desired by a 150 W. electric bulb 185 cm. from the anemone.

For the first fortnight the last 4 days of which are recorded in Fig. 8, the light was switched on at 9 a.m. and off at 9 p.m. (G.M.T.). There was thus alternate light and darkness approximately corresponding to the outside day and night. There is an evident light-controlled diurnal rhythm of contraction and expansion. The anemone was then left in darkness for 22 days and nights. Finally, alternating 12 hr. periods of light and darkness were recommenced, but now with illumination during the night and darkness during the day. The corresponding reversal of the rhythm during this last phase is clearly shown and indicates the directness of the relation of the alternations of phase to the state of illumination. Such results show that Metridium may possess a rhythmic alternation of its phases of extension and retraction that can be controlled by light, with all other environmental factors maintained constant.

But the experiment illustrated in Fig. 8 shows another feature of great interest. During the intermediate period, the anemone was maintained in complete darkness and the external conditions were varied not at all. Nevertheless, the general activity was as great as during the periods of daily illumination, and moreover changes of phase are still evident. The phases of contraction and expansion are at first somewhat irregular, but they presently develop a rough regularity with a period somewhat greater than 24 hr. The rhythm is by no means exact, and the phases soon get quite out of step with the previous periodic illumination or, for that matter, with the sequence of external day and night.

Inherent changes of phase in Metridium under constant external conditions are common. But they do not always show so clear an approximation to a regular rhythm as is to be seen in Fig. 8. Usually they are very variable and are often of short duration. Sometimes for long periods there is no evident change of phase. Fig. 9 shows a record of a Metridium which, after exposure to alternating periods of light and darkness, was then kept in continuous darkness for several days. During the first 30 hr. of darkness this individual showed no phase changes. After this the activity alters its character and irregular alternations of phase occur; but it is far from a regular rhythm.

Though inherent changes of phase in constant conditions are usually irregular, phase changes are very commonly set into a rhythm by periodic stimuli. In Figs. 7 and 8 we see how periodic illumination can control rhythmic phase change. Other stimuli as well as light may be effective in controlling phase changes. Parker (1919) showed in M. marginatum that water movement could cause expansion, and that through this factor the tides might act as a pace-maker for expansion and contraction, as well as light. Fig. 4 also makes it clear that even periodic exposure to appropriate chemical stimuli might enforce periodic expansion.

But the connexion between stimulus and change of phase is not simple. Regular stimuli may not always succeed in controlling phase change. We may contrast the experiment shown in Fig. 8 with that of Fig. 10. In the latter the animal shows irregular alternations of phase although exposed to regular alternate 12 hr. periods of light and darkness. At the onset of each light period the animal gave some degree of contraction, but its extent was variable and the contracted phase was not maintained. The responses to light are marked by dots on the tracing. In this case the light stimuli, though effective in causing immediate contraction, did not succeed in controlling phase change. The accessibility of phase change to stimulus is as variable in the case of periodic light stimuli as we have seen it to be in the case of food stimuli ; and not only do different individuals vary in this respect, but the same individual varies on different occasions.

The phases of contraction during periodic illumination are not simply direct responses of a passive animal to light. The stimulus is not exciting an inert system. In the absence of stimulation, phase changes of an irregular or semi-rhythmic character take place inherently, and a periodic stimulus appears to act by controlling this inherent rhythm. The effectiveness of the stimulus in initiating a contraction phase varies with the state of the animal and the illuminating stimulus may or may not be effective in controlling rhythmic expansion and contraction. But when it is effective it appears to act by setting the pace of the inherently occurring phase. Examination of Fig. 8 shows that even when a good rhythm is established, the inherent rhythm may from time to time break free of the pace-setting stimulus. The fact that the periodic illumination appears to be setting the pace for an inherent alternation of phase is also to be seen in a curious effect not infrequently observed with well-established light rhythms : the phase change may anticipate the recurring stimulus, so that the animal passes into the contracted phase just before the daily illumination commences. This recalls various claims that Actinia equina may ‘anticipate’ tidal stimuli (Parker, 1919).

Fig. 8 plainly shows that on occasion Metridium can exhibit inherent periodic changes of phase which may approximate to a rhythm in the absence of any rhythmic external stimulation. Even well-established inherent phasic rhythms vary widely in frequency, and though sometimes exhibiting a period of the general order of 24 hr. there is no close approximation to this period. But the existence of such more or less rhythmic changes of inherent origin may perhaps account for recorded cases of apparent persistent night-and-day and ‘tidal’ rhythms, such as the tidal rhythm recorded by Bohn (1906, 1909, 1910) in Actinia equina, which was said to be maintained in phase with the tides for 2-3 days, or even longer, after removal of the animal to constant conditions.

Many have doubted the existence of such rhythms, notably Piéron (1908) on Actinia, Parker (1919) on Sagartia luciae, and Gee (1913) on Cribrina. We can confirm the absence of any exact rhythm persisting in phase with either the tidal or the daylight cycle. Metridium observed under constant conditions in the laboratory immediately after their removal from the intertidal zone showed no relation in their subsequent phasic activity to the external tidal rhythm. Similarly, fifteen specimens of Actinia equina, which were collected from their upper intertidal limit and immediately placed under constant conditions in a dimly lit cellar, showed no trace during the following 3 days of a persistent tidal rhythm in the absence of the direct influence of the tide.

Remarkably accurate persistent rhythms have been reported from organisms widely separated in the animal kingdom. Gunn (1940) records persistent and exact diurnal rhythmic activity to be maintained in the cockroach Blatta orientalis for upwards of a week in constant conditions. Among Protista, Bracher (1937) records the rhythmic migration of the estuarine flagellate Euglina limosa. A tidal rhythm is maintained for 3 days under constant laboratory conditions. Likewise, Fauré-Fremiet (1948) records the remarkable tidal excystments of the ciliate Strombidium oculatum, which are maintained for 2-3 days in constant laboratory conditions. Metridium and Actinia certainly do not possess ‘internal clocks’ capable of maintaining such accurate rhythms : nor are they in some mysterious way en rapport with rhythmic events in the outside world from which they are experimentally isolated. But the fact that Actinians removed to constant environmental conditions may sometimes develop inherent phase changes of the general order of 24 hr. might well lead occasionally to the impression of a persistent rhythm.

Parker (1919) questioned the existence of any persistent rhythms in phase with the tides or with previous daily illumination. But he advanced as evidence against the existence of such rhythms the fact that specimens of Metridium he kept in darkness remained continuously expanded for some 36 hr., and he implied that phase change only occurred as the direct result of a stimulus. With this last implication we cannot agree. Inherent phase change under constant external conditions certainly takes place, though the extent of its development is very variable in different individuals, and 36 hr. is often too short a period for its unequivocal manifestation.

From our experiments we conclude that the phases of contraction and expansion accompanying diurnal illumination resemble the phases we have discussed in connexion with feeding and locomotion. The threshold for the initiation of the phase varies in different animals. In many, a phase of contraction is easily initiated by the stimulus of illumination, in others it is not. The threshold may be so low that the phase is initiated spontaneously under constant external conditions. The new and interesting feature of these phases of expansion and contraction is that in those anemones in which they occur spontaneously there may on occasion develop a roughly rhythmic alternation of the two phases with a frequency which is sometimes of the general order of that of night and day, though not necessarily in phase with it. Under these conditions spontaneous phase change can often be set in step with a regular external stimulus such as that of daylight and darkness to give a well-marked diurnal rhythm.

Though the pattern of activity which constitutes a phase may primarily concern one particular system of effectors, such as the muscles of the pedal disk, all parts of the animal are to some extent involved in them. Consequently, two separate phases may reinforce or may conflict with one another.

The most striking example of conflict of phases is seen in the effect of feeding on the light-controlled rhythm. If a Metridium which has been regularly closing and remaining contracted during light and expanded during darkness is given a meal (e.g. half a Mytilus), its behaviour is modified for several days afterwards. When illuminated after feeding, it still gives initially the usual parietal shortening. But then, instead of assuming an enduring phase of contraction, the column elongates again notwithstanding the light, and its disk usually remains expanded throughout. The food has in some way temporarily abolished the capacity to retain a state of phasic contraction during prolonged illumination. The inhibitory effect of the meal persists for about a week, after which phasic retraction to periodic illumination gradually returns.

Not only may phasic retraction to light be inhibited for several days after a meal but occasionally some of its features even appeared to be reversed. In some specimens for the first day or so the sphincter, which is closed in complete retraction, actually opened more fully when first illuminated.

This inhibition by other phases, of phasic retraction to periodic illumination is shown in Fig. n. For the first 3 days of the record the animal shows a clear light controlled phasic rhythm. At the beginning of the fourth day the animal passed spontaneously into the shrivelled state, and thereafter remained contracted for 36 hr. The failure to expand during the fourth night may possibly be due to repeated examination with dim red light. At the beginning of the fifth night, however, the animal expanded. During that night it accepted a meal. Thereafter the animal remained for 48 hr. in a predominant phase of expansion. The three periods of illumination succeeding the meal began 80 min., 24I and 48 hr. after it. On each of these three occasions, the switching on of the light caused some contraction ; but this was not maintained, nor did the sphincter maintain closure.

About 36 hr. after the meal the expanded animal passed into a phase of great activity associated with defaecation (Defaec.). Finally after defaecation, soon after the beginning of the last day recorded, the animal shrivelled (S) and afterwards rapidly re-expanded in a greatly inflated condition. The immediate sequence of events following the meal is essentially the same as that seen in Figs. 2 and 3. The predominant expansion associated with the feeding phase in Fig. 11 has obliterated the earlier light-controlled rhythm and has reduced the response to illumination to a variable temporary contraction. Phases may therefore block each other, much as we see in the case of incompatible reflexes.

A complete extension phase induced by food is incompatible with a retraction phase to light. But the causes of inhibition of a phase are not always obvious. There is no clear reason why light should inhibit locomotion; though we may note that light causes contraction whilst food, which may excite locomotion, causes expansion. There may be some correlation between the phase of locomotion and the phase of expansion. Why this should be is not clear, though peristalsis of the column, which plays a significant part in locomotion in Metridium, is somewhat more evident during phases of expansion.

Parker (1919) came to the conclusion that an Actinian was an animal whose internal state approximated to one of general uniformity and that the changing environment calls forth various responses without seriously disturbing the internal equilibrium. He contrasts this with the condition in higher animals in which response is novel and diverse, and in which much of the behaviour has that character which leads us to describe it as spontaneous. Our observations show that this view of Actinian behaviour is too simple.

We showed in the previous paper (Batham & Pantin, 1950b) that Metridium is in a state of continual and varying activity which is not a series of direct responses to external stimuli. The activity is a character of the unstimulated animal, and we have termed it inherent. The activity may be truly spontaneous, or it may involve internal stimuli though, if so, these must be more complex than those of a simple chain reflex. We see in the present paper that the pattern of this activity can change through a series of phases. Like the activity itself, a phase is an inherent character which may appear in the externally unstimulated animal; the animal will change from one phase to another under apparently constant environmental conditions. Whether the change of phase is truly ‘spontaneous’ or whether it depends upon internal stimuli is not yet certain. It is well to bear in mind that in the end this distinction may merely depend on whether or not there is an internal mechanical link in the cycle of events leading to apparently spontaneous activity. We shall return to this question in a later work.

Like simple reflex responses, phase changes may be induced by external stimuli. But as we have seen, the relation of a phase to a stimulus is much more complex and in many ways different in character. In particular, the stimulus initiates the phase rather than maintains it; and the whole pattern of activity of the phase is independent of the stimulus. It may be initiated by various external stimuli or apparently initiate itself. While some of the responses of Metridium to stimuli have the character of simple reflexes as in the retractor response, others, such as those involved in phasic activity, are more complex, more variable and depend on the state of the animal. These are the very characteristics which Parker so truly noted in the behaviour of the higher animals but supposed to be relatively insignificant in Actinians. It was precisely this variability of behaviour and the dependence of response upon the state of the organism which was stressed by Jennings (1915). The figures in this paper, in fact, give objective records of the existence in Metridium of ‘physiological states’ such as Jennings showed to be characteristic of all organisms from Protista upwards including Coelenterates,

Inherent phasic activity is an integral part of the behaviour patterns of Metridium. In the first place, muscular action due to inherent activity balanced against the coelenteric pressure is responsible for the shape of the animal (Batham & Pantin, 1950a). Secondly, the activity exhibited during a phase is behaviouristically significant. The elongation and swaying movements which follow temporary exposure to food solution seem quite evidently purposive in that they increase the chance of discovering a food object, the presence of which might be signified by the original stimulus. We see here inherent activity playing a significant part in a behaviour pattern, and moreover it does so in a peculiar manner. Actions such as swaying are behaviouristically relevant not so much to the original stimulus as to a future event, the capture of a solid food object, the probability of which is increased by sweeping the ground in the neighbourhood. Such activities seem so clearly purposive, and seem so clearly directed towards a goal in a manner we ourselves can understand, that earlier workers such as Gosse (1860) inferred the existence of consciousness and will in these animals. Such an assumption is unnecessary; the properties of phasic activity are completely open to mechanistic analysis. But it is idle to neglect the fact that the mechanisms involved in phasic activity are of a much higher order of complexity than simple and direct reflex responses such as have been demonstrated by earlier workers (Parker, 1919; Pantin, 1935). Some responses, such as the retractor response and the transfer of food by the tentacles to the mouth are undoubtedly simple and direct, and their physiological analysis is in parts tolerably complete. The fact that these activities are rapid makes them easily perceived, whereas the more complex phasic activities of Metridium are so slow that they are commonly overlooked. Consequently, unless prolonged observation is made we may easily receive the false impression that Actinian behaviour consists of simple direct reflex responses to external stimuli by a passive animal.

But even the direct reflex responses of Actinians are not quite so simple as would at first appear. Simple patterns of nerve impulses can be set up by electrical stimuli which call up feeding, retraction and so on. But when we consider these responses as they are initiated by natural stimuli, it is evident that there is some degree of complication on the sensory side, which is ‘short-circuited’ by direct electrical stimuli. The threshold of chemical stimulus required to produce reflex capture of food by the tentacles varies greatly with the state of the animal (Pantin & Pantin, 1943). After a meal the animal is much less sensitive to food than before (Parker, 1919). This may be due to sensory adaptation. But a meal also appears to raise the threshold for mechanical stimulation (Pantin, 1935). And even Actinians that have been starved together in a tank show great and varying changes of thresholds for food capture by the tentacles. Thus, in an experiment, a piece of mussel flesh was placed on the tentacles of each of twenty-four Metridium previously starved for at least 3 weeks. In thirteen the food was accepted at the first offer, in ten of the others it was accepted between the second and fifth offers, whilst one refused to feed. Similarly, the time taken for food to be ingested varies greatly. In a series of eleven observations food placed on the oral disk required in different cases from 3 min. to 1 hr. 45 min. to be passed into the coelenteron.

Even for direct responses such as food capture and retraction there are factors as yet unidentified which vary the sensory threshold of the response with the state of the animal. It is to be noted that one of the anemones which failed to accept food did so because the threshold for mechanical stimulation had temporarily become so low that even the light contact of a food object with the tentacles brought about retraction.

Nor are the direct responses of Actinians influenced only by changes in sensory threshold. Chemical changes in the medium can greatly modify the extent of the sphincter response in Calliactis following electrical stimuli. Ross & Pantin (1940) showed that excess magnesium depressed the response, whilst excess potassium augmented it. Chemical factors may influence facilitation and other properties of the excitation system. In Vertebrates, Bozler (1941) showed that the excitability of mammalian uterus varies greatly according to endocrine conditions. Excitability is very low during anoestrus and reaches its height during oestrus. The previous injection of oestrogenic substances causes the muscle to pass from a condition of low excitability and local response to one of high excitability and total response. There are many similar cases of endocrine influence upon response. These examples suggest the possibility that in Actinians both direct responses and inherent activity may conceivably be influenced by endocrine or other chemical changes in the tissues. Such changes might indeed be primarily responsible for changes of phase. But on this point we have at present no evidence.

In conclusion we would draw attention to the very considerable complication of response mechanism in Actinians, notwithstanding their low grade of morphological organization. We are, in fact, led to consider the direct reflex responses as secondary simplifications to meet specific purposes, rather than as examples of an elementary unit of all behaviour patterns. As Jennings (1915) showed, many supposed characteristics of behaviour of the higher animals are already present in the Protozoa. The varied phases of Metridium call to mind not only the more complex behaviour of the higher animals, but also the physiological states described by Jennings in Stentor. There is similar variation of character and threshold. Moreover, we already have in the Protista clear evidence of inherent rhythmic activity which may vary in frequency from ciliary or flagellar action to the diurnal and tidal rhythms of Euglena and Strombidium already referred to. Inherent activity and phasic change are characteristic of all grades of organism and, since they are found in Protozoa as well as Metazoa, it is evident that their mechanisms can be achieved in more than one way. If we find on occasion that they appear to be absent, and that behaviour seems to consist of simple reflexes, we should perhaps rather look for some evolutionary reason for the suppression of inherent activity than consider that we are dealing in such cases with primitive nervous machinery.

1. In constant environmental conditions Metridium senile exhibits continual slow inherent activity. The pattern of this activity varies in character from time to time. These different patterns of activity have been termed ‘phases’. A particular phase may endure for hours on end and then rather quickly give place to another phase.

2. A change of phase may be initiated by certain stimuli. Ingestion of food initiates a sequence of phasic changes involving expansion of the disk and elongation, followed by swaying, parietal contraction, distension, defaecation and ‘shrivelling’. Each of these phases has its own pattern of activity. Not all may be exhibited by any one animal, or by the same animal at different times. A similar sequence of phases to those induced by solid food may be initiated by mere temporary exposure to filtered food solution.

3. In contrast with the direct responses, such as the retraction reflex, the stimulus does not directly maintain a phasic response; it merely initiates a new phase, and the activity pattern of this new phase is maintained long after (even hours after) the initiating stimulus. This relation of phase change to stimulus may be evident after electrical stimulation as well as after exposure to food or other stimuli.

4. Phase changes also differ from simple reflexes in that the threshold for their initiation varies enormously in different animals and in the same animal at different times. Also the threshold may sometimes be so low that the phase occurs spontaneously in the absence of evident external stimuli. Changes of threshold cannot be adequately accounted for by sensory adaptation. The stimulus apparently acts by releasing a complex activity pattern (the phase) which is so far independent of the stimulus that it may appear spontaneously.

5. Locomotion is another phasic activity. It is a complex co-ordinated activity pattern. It may be initiated by various stimuli. The threshold varies at different times and in different animals. It may take place spontaneously in the absence of evident external stimuli. The threshold of this phase is lowered after feeding and raised by illumination.

6. Alternating phases of expansion and contraction frequently occur in Metridium. Their relation to diurnal and other rhythms is discussed. Daily illumination can often initiate and control regular daily phases of contraction. This is true both of daylight and of periodic exposure to artificial light. As with other phase changes, the threshold varies greatly. In some cases each periodic illumination may only induce a brief temporary contraction and fail to control phase change.

7. In complete darkness and constant environmental conditions, alternating phases of expansion and contraction may still take place. These may be irregular, but sometimes they may assume a very rough rhythm. When present, this rhythm does not keep in step with previous daily stimuli, nor with current external changes of day or night, nor with other environmental rhythms. The unstimulated animal thus possesses an inherent tendency to alternating phase change which may approach a rhythm. It appears that periodic stimuli, such as daily illumination, act by ‘setting the pace’ of this inherent alternating phase change.

8. Different phases which involve the same groups of effectors may reinforce or may conflict with one another.

9. Our experiments show that continual and varying patterns of inherent activity play an important part in the behaviour of Metridium. The behaviour is not simply a succession of direct reflexes to stimuli acting on a passive animal. Such direct responses are, however, more easily observed because phasic activity is extremely slow. Phasic activities play an essential part in behaviour patterns such as food capture. They are often behaviouristically relevant to a future possible event rather than to a past stimulus; as when sweeping and swaying movements increase the chance of finding food.

The relation of phases to the ‘physiological states’ of Jennings is noted.

Much of this work was done at the Marine Biological Laboratory, Plymouth, to the Director and staff of which we are most grateful for the many facilities given us. Part of the work was done during the tenure by one of us (E.J.B.) of a Shirtcliffe Fellowship of the University of New Zealand. We wish to thank the Department of Scientific and Industrial Research for a grant for the development of a special research which enabled this work to be concluded.