Slow Contraction and Its Relation to Spontaneous Activity in the Sea-Anemone Metridium Senile (L.)

But in Calliactis stimulation at frequencies which are too low to cause the sphincter response nevertheless succeeds in calling certain other very slow systems of muscles into action. The mesenteric retractors, the longitudinal parietal muscles, and the circular muscle of the body wall seem to be excited by successively lower frequencies (Pantin, 1935 a). Hall & Pantin (1937) found a rather similar facilitation sequence in Metridium senile, though in this case it was the mesenteric retractors which responded most rapidly and to the highest frequency of stimulation.

It was at first supposed that the responses of these slow muscle systems were essentially facilitated contractions differing from those of the sphincter of Calliactis only in their much longer time scale. But, as we shall show in this paper, there are in fact important differences between the contraction of the slow parietal and circular muscle systems on the one hand, and that of typically facilitated fast muscles on the other. Moreover, these slow muscles can show prolonged complex co-ordinated activity which seems to be spontaneous; that is, it is inherent in the animal and not the direct consequence of individual external stimuli (Batham & Pantin, 1950b, c). Further, their response to stimulation has more in common with this spontaneous activity than with a simple and direct contractile response to a stimulus. Accordingly, it was necessary to re-examine these phenomena. We choose particularly the slow responses of M. senile for this purpose.

Anemones for experiments were collected carefully without rupturing the pedal disks, and placed in tanks on small stones or squares of ground glass which had for some hours previously been soaked in sea water. Plastic or celluloid collars helped keep the anemones from drifting off stones or glass before their attachment, which may take a day or so. Anaesthetization with equal parts of MgCl₂.6H₂O and sea water for about half an hour usually preceded the sewing in of recording threads—each a double loop of thick cotton. The anemones were then left to recover in sea water for a few hours or overnight. The specimens should be starved for at least 3 days prior to low-frequency stimulation tests. Otherwise, the peristaltic activity associated with ejection of food remains may modify reactions in a way subsequently to be described.

The actinians were stimulated by condenser shocks applied at various frequencies through non-polarizable Ag/AgCl electrodes in the manner described by Hall & Pantin (1937). Movements were recorded kymographically with long light isotonic levers attached to the animal by threads sewn into it.

Experiments were made not only on intact animals but also on various preparations of isolated parts. In making such preparations it is essential that all operations should be quickly and deftly carried out; otherwise there is great and prolonged contracture and secretion of mucus, and the preparation becomes useless. The following plan proved best. Large specimens in good condition were placed in bowls of aerated sea water overnight. Next morning an equal amount of MgCl₂. 6H₂O was added to whichever bowl contained the most expanded anemone. 30−60 min. later about 50 ml. of the MgCl₂ solution was also pipetted into the coelenteron via the mouth. This was repeated several times in the next hour. At intervals the margin of the pedal disk was gently pinched to see if the retractor response was still obtainable. Half an hour after it vanished the anemone was lifted out and the operation performed. The prepared piece of tissue was at first gently stretched with the fingers. One end of the strip or loop was then attached by a soft string and pins to a weighted cork base in a large bowl of aerated sea water. A thread from the other end of the preparation hung over the edge of the bowl and was attached to a plasticine weight heavy enough to raise the tissue in the water.

The tissue was then left for 3 or 4 hr. to recover, though the threshold for electrical stimuli might remain rather high and erratic for a further hour or two. By this time, in good preparations, the threshold and the response were about the same as in intact anemones.

Both the slow responses of the parietal muscles and the rapid facilitated contraction of the retractor can be elicited by electrical stimulation of any part of the column. In both, the response is not locally confined to the region stimulated but takes place round the whole column of the anemone; that is, there is through-conduction. In both also there is a threshold below which no response can be elicited ; and the threshold for the response in both systems of muscle is precisely the same. It therefore appears that each system of muscles, the parietal and the retractor, can be excited by way of the same through-conduction system.

But these common features do not imply identity of the mechanism of excitation of the muscles themselves in the two systems, and further inspection shows essential differences. The parietal contraction is much slower than that of the retractor. It is also quite smooth and sigmoid, showing none of the sharp step-like increments corresponding to each stimulus, which are characteristic of the retractor’s facilitated response. At first sight this might perhaps be attributed to the slower mechanical properties of the muscle. A higher ‘viscosity’ might smooth out the step-like increments. But that will not account for the extraordinarily long delay between the stimulus and the moment when these slow responses begin. If a quiet and intact Metridium is given a regular series of shocks, the ensuing parietal response regularly begins at about the seventh shock (Fig. 1 b). This holds over a wide range of frequencies, from about 1 stimulus every 2 sec. to 1 every 15 sec.

If the stimulus series be short, another important feature is apparent. There is an enormous latent period; so that the response not only shows no direct relation to the individual shocks, but may even take place long after the stimulus series has ceased.

These peculiarities of the parietal contraction are shown even more clearly by the very slow circular muscle of the column. Fig. 2a shows the responses of a ring preparation of circular muscle, 4 cm. in diameter. The ring was left attached to a longitudinal strip of body wall, on which the electrodes were placed 1·5 cm. from the ring. Stimulation was thus indirect. The sigmoid character of the contractions makes their exact beginning hard to determine, but in this case it is evident that the latent period exceeds 100 sec. In contrast, the latent period of each step of the retractor response (Fig. 1 a) is only some 50−80 msec. ; and even this is made up chiefly of conduction time in the through-conduction system. It is the enormous latent periods which differentiate the slow contractions most completely from facilitated ones.

These long latent periods of the parietal and circular muscles are not due to slow conduction up the column of the anemone. The through-conduction system which excites these muscles conducts at a high speed, as can be seen at once by the promptness of the facilitated response of the retractor if the frequency of stimulation is raised. Hence the delay in the slow muscles must be peripheral. In the intact animal, different sectors of the parietal muscle system respond with periods of delay which differ in an arbitrary way from sector to sector and bear no relation to the site of stimulation. We shall discuss these parietal delays later.

The peripheral character of these long latent periods is even more clearly shown in preparations of the circular muscle. A preparation is made in which the column of the anemone is cut horizontally at various levels, to give a number of rings of the column wall. These rings are left in contact with a longitudinal strip of the column. We thus have a preparation consisting of a connected series of rings of the circular muscle sheet of the column, from the region of the marginal sphincter to near the pedal edge (as in Fig. 3).

Electrical stimulation of such-a preparation may cause all the muscle rings to contract. If the frequency of stimulation is high, as in Fig. 3, a rapid facilitated response can be evoked from any muscle rings which include a portion of the marginal sphincter. The promptness of this response, whatever the point of stimulation, shows the high speed of conduction in the preparation.

But, in addition, every muscle ring shows a slow response with its characteristic delay. In the first record of Fig. 3, the first ring to respond, after the facilitated contraction of the sphincter, is the pedal loop (15 sec. after the last stimulus). The next is the mid-column ring (26 sec.), whilst the slow main contraction of the subsphincter begins some 29 sec. after the last shock; though the preceding small facilitated response makes the exact onset hard to determine.

As in the responses of different sectors of the parietal muscle system, the latent periods of these rings of circular muscle vary, and the order in which the rings contract bears no relation to the site of the stimulus. In both the first two records of Fig. 3 the electrodes were placed at the oral end of the preparation (near the subsphincter) ; but the order of contraction of the rings is not the same. And though it so happens that in the second record the muscle loops contract in an orderly succession from the electrodes, nevertheless when those were placed at the opposite end near the pedal loop, as in the third record, the first loop to receive the excitation was actually the last to give its slow response.

Evidently, in both circular and parietal muscles, excitation by way of the through-conduction system is followed by prolonged delay ; a delay about a thousand times that in facilitated responses, and one which occurs locally in the neighbourhood of the responding muscle rather than in the through-conduction system.

In animals or preparations showing little spontaneous activity it is possible to study the relation of the size of the slow responses to the number and frequency of stimuli. Here also the relation differs from that of simple facilitation. Preparations have an advantage over the intact animal because the response is not complicated by the contractions of several different muscle systems ; and consequently the response to a wide range of frequencies can be clearly studied.

In a typical experiment, a ring of quiescent circular muscle cut from just below the marginal sphincter was stimulated by electrodes placed on an attached tongue of body-wall tissue. As in the step-like facilitated contraction of the retractor muscle, the size of the smooth slow response of this ‘subsphincter’ ring of circular muscle was more or less constant over a certain range of frequencies; diminishing when these were still higher (cf. Hall & Pantin, 1937). The range of frequencies over which the subsphincter gave a fairly constant response was, however, surprisingly large—from i stimulus in sec. to 1 in 20 sec. And below this, instead of the regular progressive decline characteristic of facilitation, the response merely became irregular. At frequencies of 1 stimulus in 40, 60 or 80 sec. there was a complex succession of contractions and relaxations which could not be related to individual stimuli, and which might have been no more than an increase in ‘spontaneous activity’.

The number of stimuli required to produce a detectable response in these slow muscles is generally much larger than that needed for a facilitated contraction. Though the number needed for a parietal response varies from one animal to another, at least 3 or 4 stimuli are usually required. Beyond this the response tends to increase with the number of stimuli, but above 15 or 20 stimuli it gets no bigger, and long continued stimulation leads to irregular contractions and relaxations; whereas in a facilitated contraction each stimulus of a series adds a contraction step till there is a maintained tetanus.

A further, and important, feature of the slow contractions is their frequent variability, which contrasts with the machine-like regularity of facilitated contractions. It is most evident if the animal, or tissue preparation, is undergoing much spontaneous activity. Fig. 26 shows the parietal response from two sides of an intact animal which, unlike that used for Fig. 1 b, showed great spontaneous activity. Here the spontaneous activity so greatly modifies the response that occasionally it is hard to tell whether contractions which follow stimulation are in fact caused by it. Even in an animal giving regular parietal responses, these may suddenly diminish and even fail after long experimentation. This is not through failure to excite the nervous system, for a rapid pair of stimuli will still evoke a normal facilitated response of the retractor muscle.

The slow contractions thus differ from typically facilitated ones. In the latter, the response to a number of shocks is chiefly a mechanical summation of unit steps. The contraction of the parietal and circular muscle systems is not of this kind. Here, the successive shocks seem to cause a state of excitation in the muscle which, if it reaches a sufficient intensity, is followed after a considerable delay by a smooth contraction. Unlike a facilitated one, the size of a slow contraction does not decrease in a simple way with decreasing frequency of stimulation, and, moreover, often shows considerable variability which seems to be related to spontaneous activity in the responding muscle.

The existence of two distinct kinds of contraction in actinians raises the question of whether they possess muscles of more than one sort. They show no histological evidence of two clearly distinct kinds of muscle fibre (Batham & Pantin, 1951). All are unstriated; though the fibres in different parts of the body do vary considerably in length, from 20 to 50 p in the circular muscle sheet and in the radial muscle of the mesenteries to over 1 mm. in the mesenteric retractors of Metridium. But even the retractors are organized in the same way as the general muscle sheets : they are simply hypertrophied regions of a widespread muscular network. All we can say is that specific effectors, like the retractors of Metridium and the sphincter of Calliactis, which give powerful facilitated contractions, tend to possess large, long fibres. We cannot even say that quick facilitated responses are to be found only in such anatomically differentiated effectors, for, as we shall see, they can be obtained from the radial musculature of the oral disk.

While there is no evidence that the muscle fibres of actinians fall into two distinct kinds, with different kinds of contraction, there is evidence that in some muscles the self-same muscle can contract in both the quick and the slow way. The mesenteric retractors in Metridium give a rapid facilitated response to rapid stimuli. But they can also show typical slow responses. Fig. 5 shows such a smooth delayed mesenteric contraction (second lever down) in the response following a low-frequency stimulus ; the contraction begins well after the stimulus is over. In Calliactis such a slow delayed contraction is the most obvious element of the mesenteric retractor response.

If the retractor of Metridium could execute only quick facilitated contractions, it should remain extended or become bent and buckled when an extended animal slowly shortens in this way. This does not happen. And when whole mesenteries are removed from the animal and observed under the microscope it is clear that all the fibres of the muscle sheet, including the retractor, can slowly and smoothly contract. It is hard to account for these phenomena except by supposing that the same fibres can contract in both ways.

The mesenteric retractor is not the only muscle which seems to contract in more than one way. Hall & Pantin (1937), working on intact Metridium, considered the marginal sphincter to be a very slow muscle, which however was facilitated by stimuli of about the same range of frequency as the rapid retractor. But our present work has thrown a different light on the matter. Owing to the proximity of the powerful retractors, the behaviour of the sphincter is not easy to analyse in the intact animal. But this difficulty can be overcome by using preparations like those already described in which the sphincter and various parts of the circular musculature were partially separated as isolated rings of tissue. These rings of tissue remained connected along one side by a strip of body wall on which the electrodes could be placed.

Fig. 4 shows the anatomical relations of the sphincter muscle, the circular muscle sheet and the physiologically specialized region of it, the ‘subsphincter’. It is from the subsphincter that peristaltic waves usually begin.

If a ring of tissue is isolated, including the sphincter, the response to rapid stimuli is distinctly double. This can be seen in Fig. 6 a for shocks at frequencies of 1·7, 1·85 and 2 sec. There is one component which clearly facilitates with step-like contractions, as shown in the faster record of Fig. 6,b. This resembles facilitation of the retractor of Metridium or of the sphincter itself in Calliactis. The only obvious difference is that the facilitated sphincter responses of Metridium sometimes require a number of preliminary stimuli before the contraction steps begin (in Fig. 3 it begins at the 5th shock), whilst at other times they may show some degree of permanent facilitation, so that a single stimulus causes a response. In records from the intact Metridium this facilitated sphincter response tends to be lost in the disturbance due to the great facilitated retractor responses at the same range of frequency.

As in the retractor, the facilitated contraction of the sphincter of Metridium vanishes as frequencies of one shock in 3 sec. are approached (at 16−17°C.). But on approaching this frequency, the second component of the response ceases to be masked by the first, and becomes increasingly evident. In contrast with the facilitated contraction, this component has all the characteristics of the slow, sigmoid type with long latent period.

The interpretation of this double contractile response of the sphincter region is not easy. The marginal sphincter itself is anatomically differentiated within the mesogloea from the circular muscle layer from which it is derived. It is not possible to make a preparation including the sphincter but completely excluding the underlying circular muscle layer of the body wall. In ring preparations of the body wall, all parts can give the slow delayed contraction. But only those which include part of the marginal sphincter can also give a clear facilitated response (Fig. 3): all other parts give the slow contraction alone. We can thus attribute the fast, facilitated responses to the sphincter. The slow contractions of ring preparations which include the sphincter may partly be due to that portion of the general circular muscle layer inevitably included in the preparation. Nevertheless, there is some evidence that the sphincter itself contributes to the slow delayed contraction. Direct observation shows that when the sphincter region slowly contracts, all parts are involved in the process ; we do not find an extended inactive sphincter bending and buckling as the inmost region of the preparation contracts.

Quick facilitated contractions can also be obtained from the radial muscle system of the oral disk. A sector of the disk of a large Metridium was cut out; its tentacular edge was then pinned down, and electrodes placed upon it. The radial contractions were recorded from a thread attached to the pharyngeal edge of the cut sector. Such preparations showed small but typical responses to every shock after the first of the series, provided the stimulation frequency was of the same range as that effective for the retractor (less than 1 per 2 sec.). Pantin (1935 a) noted that in Calliactis even a single stimulus could cause a twitch of the tentacles owing to excitation of radial musculature. Directly facilitated contractions of the radial muscle can be obtained in this way from stimulation of either the oral region or the distal edge of the disk. In fact, in radial directions the disk shows through-conduction. This radial through-conduction contrasts strongly with the localized conduction laterally round the circumference of the disk. Early experiments showed (Pantin, 1935 a) that in this direction the response of the disk to stimuli shows an apparent decrement as it is transmitted round the circumference; owing to what was termed ‘interneural facilitation’.

It is important to notice that this localization of the response is restricted to circumferential conduction, and that radially there is ‘linear through-conduction’.

This may be contrasted with the total through-conduction in every direction of the retractor system both over the whole of each mesentery and from one mesentery to another. But radially, the facilitated disk responses are quite comparable to facilitated responses of the retractor. And like those, in addition to simple facilitation, slow delayed responses were also obtained to stimuli of low frequency, between 1 in 5 and 1 in 20 sec. Unfortunately we cannot tell certainly whether the quick facilitated contraction and the slow responses were due to two different muscle systems or to one alone : for there are not only powerful radial muscles in the ectoderm of the disk, but also some endodermal radial fibres along the insertion of the mesenteries. But again we may note that slow contraction of the disk does not cause buckling as would be expected if the ectodermal muscle took no part in it.

All parts of the actinian muscle system can give the slow type of response. Some parts, such as the undifferentiated circular muscle of the column and the parietal muscles, appear to give none other. There is no evidence that any part can give rapid facilitated contractions alone. But some parts of the animal, like the radial muscle of the disk and the anatomically differentiated retractors and marginal sphincter, show both the quick facilitated contractions and the slow type as well. There is no evidence of two sharply distinct histological types of muscle fibre to account for this. On the other hand, the direct observation of the retractor and of the sphincter indicates that when these slowly shorten all parts contract and none remains passively extended. Though nothing but direct observation of the contractions of individual living muscle fibres can be conclusive in such a case, it is hard to explain the phenomena unless the same muscle fibres can contract in both fashions. The phenomena recall in some ways the two kinds of contraction which can be elicited from some Crustacean muscles (Wiersma, 1952): though in the anemones, there is at present no evidence of multiple innervation of the muscles.

An important feature of the slow response is that in the intact animal it is generally not a simple contraction of a single muscular system, but a co-ordinated sequence of activities. Fig. 1 b shows two successive responses of an intact Metridium to a series of 10 shocks at 5 sec. interval. In this anemone, each such series of shocks caused a parietal contraction with great regularity; but this was followed by prolonged activity.

An initial parietal contraction is usually followed by a slow contraction of the marginal sphincter, and this in turn by a peristaltic wave. This can be seen in Fig. 5, which shows the response of an intact Metridium to a series of 10 shocks at 3 sec. interval. It is recorded from two points on the parietal wall, from the mesenteries and laterally from the marginal sphincter. The parietals lead, and are followed by the mesenteries : after an interval the marginal sphincter contracts, and this initiates a peristaltic wave. Such sequences are also characteristic of spontaneous activity (Batham & Pantin, 1950b ; Pantin, 1952). A spontaneous movement of the column is recorded in Fig. 5 previous to the response to stimulation. The spontaneous contraction and the response to stimulation are essentially the same.

The similarity between spontaneous contractions and responses to low frequency stimulation extends even to details. When a parietal contraction takes place during spontaneous activity, the different sectors of the parietal wall do not all contract simultaneously and to the same extent. One side acts as ‘leader’ ; and this may be at any point round the disk, and is as often seen between two of the recording strings as immediately beneath one of them. The remaining sectors begin their contractions in some cases as long as a minute after the leader and do so in an irregular order ; though this order may be repeated with each contraction over long periods (Batham & Pantin, 1950 a). The parietal response to electrical stimulation is of exactly the same peculiar character. Moreover, the responses to stimulation tend to call into activity the same leading sector, and the other sectors respond in the same arbitrary sequence that is seen in a spontaneous response in the same animal. Thus, a specimen was from time to time giving spontaneous parietal contractions; and occasionally between these, stimuli of 10 shocks at 5 sec. interval were given. The parietal contraction was recorded from three different sectors of the body wall. The subsequent contraction of the marginal sphincter was also recorded. Table 1 shows the sequences observed. The parietal sectors are arbitrarily lettered A, B and C, and the sphincter Sp. The delay of each sector after the leading parietal is shown in seconds. The responses to stimuli are marked ‘Stim.’ : and in these the time between the first shock and the response of the leading parietal is also shown in seconds. Contraction sequences not marked ‘Stim.’ are spontaneous.

It will be seen that parietal sector B leads throughout, and that the sequence of the other parts and their delay times follow essentially the same pattern whether we are concerned with a spontaneous contraction or response to stimulation. We might in fact almost say that a stimulus seems to unloose a pattern of ‘spontaneous’ activity in the parietal system rather than cause a simple contraction of a single muscle. Such a system is not only distinct from simple facilitation, but also presents an obvious parallel to autonomic excitation of the semi-autonomous viscera and vascular systems of the vertebrates.

The complexity of slow responses implies interaction between the muscular components of the column. Occasionally, though by no means invariably, there is clear evidence that a contraction present in one set of muscles actually inhibits the response to stimulation of the opposing set. In the column, the action of the parietal muscles is mechanically opposed by that of the circulars. Now it is possible directly to excite a local contraction of the circular muscle by local mechanical stimulation of the column. Fig. 7 shows the influence of such local contractions of the circular muscle on the parietal response. The anemone was giving strong and fairly regular parietal responses to series of 10 condenser shocks at 5 sec. interval. The first contraction of the series was given by the resting anemone when stimulated. During the second series of stimuli, a constriction of the circular muscle was induced by gently stroking the body wall with a rounded glass point: the contraction of the parietal is very greatly diminished. During the third series of shocks a large spontaneous circular muscle contraction developed just above the pedal disk: the parietal contraction is almost abolished. The fourth series of shocks was unaccompanied by any circular activity; and the parietal response is very large. After the record illustrated, a fifth series of shocks again showed the effect of a mechanically induced circular contraction in reducing the parietal response ; and a last series of shocks again showed the large uninhibited parietal contraction.

In this experiment, the circular muscle contractions, whether spontaneous or induced, extended right round the column so that every sector was affected. But it sometimes happens that mechanical stimulation of the column wall induces only a local contraction of the circular muscle which does not extend right round. When this happens, electrical stimulation may fail to elicit any parietal response in the sector with the localized contraction of the circular muscle, but it still causes a full parietal response in sectors on the opposite side : the animal thus bends away from the mechanical stimulus. This localization of parietal inhibition shows clearly that the failure to respond cannot be a simple mechanical effect, such as might follow a general rise of coelenteric pressure consequent on circular contraction.

Just as there is evidence of inhibition of the parietals by the circulars, so also there is occasional evidence of the converse. Sometimes electrical excitation of the parietal will cause a concurrently induced circular contraction to fade away; the parietal response apparently inhibiting the circular contraction. In one experiment a series of shocks at 10 sec. interval induced a prolonged state of parietal contraction which endured for some time after the stimulus. During this period the ‘spontaneous’ contractions at the circular muscle were almost completely inhibited.

Fairly often there appears to be a reciprocal relation between parietal and circular responses, in that those animals which give strong parietal contractions often give weak circular ones, and vice versa. But such a relation is not invariable, and sometimes a stimulus leads to simultaneous contraction of both parietal and circular muscles with consequent expulsion of sea water; as in the ‘withering’ reaction (Batham & Pantin, 1950a). Nevertheless, we may conclude that reciprocal inhibition is a factor, and perhaps a very important one, in the slow parietal-circular sequence.

It is evident that, both in the response to low-frequency electrical stimulation and in the essentially similar spontaneous movements of the column, the parietal-circular sequence involves co-ordination. The first phase of these movements is a contraction of the whole parietal system of muscle. This itself requires co-ordination, for the parietals are also capable of purely local contractions. Local parietal contractions are produced by certain specific stimuli. Light, even in the pure white variety of Metridium, causes them. Fig. 8 shows the response of such an animal to lateral illumination of the middle of the column for 30 sec. at about 5000 metre-candles. The animal was in a dark vessel for 2 hr. or more before the experiment. Recording threads were taken from the lighted and unlit sides. The local parietal contraction to such illumination starts some 15−25 sec. after illumination begins. It never reaches maximal extent. After prolonged exposure (5-10 min.) there may occur a secondary, greater, and more general contraction involving retraction of the capitulum and constriction of the sphincter. Apart from the possible general excitation through spread of light during prolonged illumination, this secondary response differs from the primary local one in that it can be antagonized by recent feeding, particularly the sphincter contraction.

In Metridium, local parietal contraction cannot be initiated by mechanical stimuli. In some other species, local mechanical stimulation can cause striking local parietal responses. In a Brazilian Bunodactis (sp. ?), electrical stimulation causes slow symmetrical activation of the parietal musculature, but local gentle mechanical stimuli cause remarkably rapid and localized parietal contractions (Pantin & Vianna Dias, 1952).

The parietals are separate cords of muscle, not connected by a common muscle sheet. The above experiments show that these can be excited locally and independently by specific stimuli. When electrical stimulation causes all the parietal muscles to contract, it is exciting them by a different path, the through-conduction system. It is, in fact, the excitation of this system which co-ordinates the response of the separate parietal elements.

Excitation of the through-conduction system thus plays an essential part in initiating the parietal-circular co-ordinated response to electrical stimulation. Now we have seen that there is a fundamental similarity between this response to electrical stimulation and the spontaneous co-ordinated parietal-circular sequence (Fig. 5 and Table 1). This similarity suggests that, just as in electrical stimulation, these spontaneous contractions are initiated by occasional low-frequency nervous impulses, in this case of natural origin, traversing the through-conduction system. By analogy with the results of electrical stimulation, we should expect the larger spontaneous contractions to coincide with bursts of several impulses at a frequency of the order of one impulse per 5 or 10 sec.

There is no doubt that impulses do pass occasionally over the through-conduction system in apparently unstimulated animals. Very occasionally even unstimulated intact animals will be seen to give a ‘spontaneous’ contraction of the retractors; a response which will only occur if two impulses pass over the through-conduction system within about 2 sec. of each other. Spontaneous contractions of the retractor are very rare, and the hypothesis we are considering would not demand the presence of impulses in the through-conduction system at the relatively high frequency required to produce them. However, the presence of spontaneous impulses even at very low frequency should be brought to light in another way. If a single electric stimulus happens to be sent into the through-conduction system within 2 sec. before or after a spontaneous impulse, the resulting pair of impulses (one generated spontaneously, the other set up by the single stimulus) will be sufficiently close together to cause a quick contraction of the retractor. Such an effect can in fact be observed. It is possible to gauge the natural frequency of spontaneous impulses, by comparing the number of times which a single electric shock happens to call forth a retractor response, with the number of times it happens not to. Since a retractor response will occur if a stimulus happens to fall within about ± 2 sec. of a ‘spontaneous’ impulse, it may be said that each stimulus ‘investigates’ the presence of such an impulse over a 4 sec. period. It follows that if 1 in n shocks succeeds in eliciting a response, there is about 1 spontaneous impulse occurring every 4n sec.

In a total of 3241 single shocks, given to twelve different Metridium, 42 caused a retractor response. This corresponds to an average of about 1 spontaneous impulse for every 80 shocks: and, if each shock would reveal a spontaneous impulse within ± 2 sec., this implies an average frequency of 1 spontaneous impulse every 5 min. Individual animals vary greatly in the apparent frequency of ‘spontaneous ‘impulses. In those we have studied the average frequency varied from 1 such impulse every 2 min. to 1 in 12 min. Ross (1952) has published figures (his table 1) which would indicate from 1 impulse every 7 min. to 1 every 70 min. These values are probably too low for normal resting anemones, for Ross was continuously stimulating his animals for 24 hr. at frequencies of 1 shock in 5 sec. to 1 in 30 sec. ; and continual low-frequency stimulation depresses spontaneous activity (cf. Batham & Pantin, 1950b, Fig. 5). Our anemones, also, were subjected to fairly frequent stimulation, though not continuously as in Ross’s experiments. Even so, these experiments show that the frequency of spontaneous impulses is very low; moreover, these may include some ‘after discharge ‘impulses generated by the stimulus itself.

Nevertheless, the animals chosen for stimulation experiments tend to be naturally quiet ones, and only a limited number of them show the co-ordinated parietal sequence. Moreover, such co-ordinated contractions only occur about every 10 min., so that occasional outbursts of impulses at long intervals are all that the phenomenon requires. The experiments certainly show that from time to time some spontaneous impulses traverse the through-conduction system.

Since any impulses causing parietal contractions must of necessity precede those contractions, it is not easy to devise a method of detecting their presence directly. However, highly suggestive phenomena have been seen in some preparations of circular muscle attached to a strip of body wall in the way described earlier. Fig. 9 shows an outburst of spontaneous activity of the very slow partly isolated subsphincter muscle. In this preparation the sphincter itself gave a small facilitated response even to a single shock. It will be seen from the record that the outbursts of slow activity in the subsphincter seem to be preceded by the appearance of occasional impulses registered by facilitated contractions in the partly isolated sphincter.

In the parietal-circular sequence, the slow parietal contraction is followed by a contraction of the subsphincter; and a wave of peristalsis then slowly passes down the column, taking several minutes to do so. This can be seen both in stimulated and in spontaneously active animals (cf. Batham & Pantin, 19506, fig. 2). Such activity involves a co-ordinated sequence, not only between the parietal and circular muscle systems but also between the different parts of the circular muscle layer (Fig. 5).

Apart from the folding of the circular muscle sheet to give the marginal sphincter, the histological and anatomical organization of the whole circular muscle system of the column is of uniform character. But the various regions differ physiologically. Fig. 10 shows the spontaneous activity of partly separated rings of the sphincter, the subsphincter, the middle of the column and the pedal edge. These rings retained contact with a strip of parietal body wall as in the figure. There is a gradient of spontaneous activity, which is least in the region of the subsphincter and greatest near the pedal edge. This activity is mostly unco-ordinated ; but as Fig. 10 shows, there are occasional contractions in which all four muscle strips act more or less together. These co-ordinated contractions are often preceded by a twitch of the sphincter ring (Fig. 10), indicating that they are associated with occasional spontaneous impulses in the through-conduction system.

The co-ordination which is apparent when the four muscle rings contract together in this way is limited : for their precise sequence of contraction is neither orderly nor constant, though all contract within about i min. of each other. This is quite unlike what happens in the intact animal. In that case there is a higher degree of co-ordination. Whilst occasional circular contractions may start in various regions of the column, usually a contraction is followed by a peristaltic, or at times an anti-peristaltic, wave which brings about contraction of the parts of the circular muscle layer in a single, orderly succession. This peristaltic wave takes several minutes to traverse the column, so that the interval between a contraction of the subsphincter and that of the lowest parts of the circular muscle field reached by the peristaltic wave is far longer than the interval between the contractions of the partly separated muscle strips of Fig. 10. It is as though in the intact animal the peristaltic wave ensures a co-ordinated sequence of contraction down the whole muscle sheet, so that the contraction of each successive part is in some way held back until the peristaltic wave slowly reaches it in due order.

This same contrast can be seen between the responses to stimulation in ring preparations and in intact animals. We have already illustrated those of such a ring preparation in Fig. 3. Here again, the rings of tissue do not contract in regular sequence ; and despite the long latent period, the interval between their contractions is far less than the time required for the passage of the co-ordinating peristaltic wave in the intact animal. In the preparation illustrated in the first record of Fig. 3 the pedal circular muscle is responding only 15 sec. after the stimulus, whereas in the intact animal a peristaltic wave will not reach this region for several minutes. Secondly, individual loops, particularly near the pedal edge, may each show considerable spontaneous activity in which the other loops do not partake. Moreover, though all the muscle loops may respond to the stimulus they do not do so in due order. As in spontaneous activity, it seems as though in the intact animal the response of the circular muscle is temporarily inhibited and controlled; first so that the subsphincter follows the initial parietal contraction in an orderly way, and then so that the contraction of each part of the circular muscle sheet of the column is held back till the peristaltic wave reaches it.

Though a response of the muscle loops can be invoked by stimulation of the through-conduction system, the passage of the peristaltic wave in the intact animal cannot be due to transmission down that system: the conduction of the wave is much too slow and it shows an ordered sequence absent in the response of muscle loops. Whatever the system conducting peristalsis, it seems closely associated with the muscle sheet. Indeed this conducting system may perhaps be the muscle sheet itself, and not nervous. The circular muscle field of the column certainly provides a connected network of muscle fibres (Batham & Pantin, 1951). It is significant that, as Parker (1919) showed, even after Mg anaesthetization gentle mechanical stimulation of the column in Metridium can still give rise to a localized contraction of the circular muscle layer. And during Mg anaesthetization peristalsis of the circular muscle sheet is the last activity to disappear, failing after the through-conduction reflexes seem to have vanished. Moreover, peristaltic conduction is sensitive to deformation of the muscle layer. We have noted elsewhere (Batham & Pantin, 19506) that slight local damage to the column of Metridium by the insertion of a thread may result in a local ring of diminished tone in the circular muscle. This region of low muscular tone acts as a barrier to the transmission of peristaltic waves. These facts do not permit us to conclude that peristaltic conduction is certainly muscular; though at present that is the simplest interpretation of them. It is, however, certain that peristalsis is not transmitted along the fast through-conduction system.

Our experiments show that when electrically stimulated at a very low frequency, the muscles of Metridium regularly respond, though in a manner different from rapid, facilitated responses. Though extensive, the response is very slow, with a very long latent period, and is variable. In the intact animal, it is not a single response of one muscle but a co-ordinated sequence of reciprocal antagonists. These complex responses to low-frequency stimulation are precisely similar to the rhythmic spontaneous parietal-circular sequences of contraction which occur naturally from time to time in the intact animal.

Analysis of the slow responses to stimulation can thus throw light on the mechanism of these spontaneous contractions. We have seen that the parietal muscle bands are excitable independently : though we do not yet know the pathway of this local excitation. We know also that these same muscle bands often execute local spontaneous activity independently of neighbouring parietals. Their activity may become co-ordinated, in the parietal-circular sequence. When it does so, one part of the parietal system acts as leader, and is followed, after some delay, by the other parts in an arbitrary order ; and followed still later by a slow contraction of the sphincter region of the circular muscle, and this by a peristaltic wave. This complex and peculiar co-ordination of the parietals with the circular muscle system seen in spontaneous activity is of the same character as the response to low-frequency excitation of the through-conduction system.

We have also seen that there are occasional spontaneous impulses to be detected in the through-conduction system ; and that in circular muscle preparations impulses may be detected during spontaneous outbursts of slow contractions. It is reasonable therefore to suppose that the rhythmic spontaneous parietal-circular sequence is initiated by the more or less periodic discharge in the through-conduction system of a few impulses ; these serving to co-ordinate the natural independent spontaneous activity of the separate parietal sectors. We do not know from whence these impulses arise; though they may come from the leading parietal sector itself.

After the co-ordinated parietal contraction there follows the slow contraction of the marginal sphincter, and this engenders a slow peristaltic wave. The conduction of this wave and the co-ordination of the successive parts of the circular muscle system which it entails are too slow to be due to transmission down the through-conduction system itself. There is some evidence suggesting that peristalsis is directly conducted in the muscle sheet.

1. The very slow responses of the body wall of the actinian Metridium senile have been studied, both in intact animals and in partially isolated tissue preparations.

2. The slow responses to electrical stimulation differ from the rapid facilitated responses of the retractor muscle. There is an enormous latent period of peripheral origin. The contraction is sigmoid, and does not show the step-like character of the retractor response. The relation to frequency and number of electric stimuli also differs. The responses vary with the state of the animal.

3. Some tissues, such as the marginal sphincter region, give two distinct kinds of contraction : a quick facilitated response and a slow delayed response. There is no striking histological differentiation into two kinds of muscle fibres, though there is some anatomical differentiation into two regions.

4. In contrast with the total through-conduction to the sphincter and the retractor, the radial muscle of the disk shows radial ‘linear through-conduction’, the response remaining localized circumferentially. Nevertheless, in the disk also there are both quick and slow responses.

5. The slow responses in the intact animal are not simple contractions. They consist of a co-ordinated sequence in several muscles which may continue for many minutes. There is evidence of reciprocal inhibition between the circular and parietal muscles. The responses to the same stimulus vary greatly at different times.

6. The slow responses to electrical excitation show a detailed resemblance to the spontaneous contractions of the same muscle systems. The muscle systems excited are not passive, but spontaneously active systems.

7. The parietal muscle system can be excited to contract locally by specific stimuli. Electrical excitation excites always the whole parietal system. The coordinating system in the latter case is considered to be the through-conduction system. The complete identity of the spontaneous parietal-circular sequence with that resulting from electrical excitation indicates that the through-conduction system co-ordinates the spontaneous sequence as well as the response to the stimulus. There is evidence of occasional impulses in the through-conduction system during spontaneous activity.

8. Partly isolated rings of the circular muscle in such a preparation of the body wall, connected by a strip of tissue, respond to electrical excitation. In contrast, in the intact animal the contractions of the parts of the circular muscle system are co-ordinated to give an exceedingly slow peristaltic or antiperistaltic wave.

9. The excitation system of the slow muscle resembles the visceral neuromuscular system of vertebrates rather than that of skeletal muscles such as the limb muscles.