Inhibition and Rhythmic Activity of the Circular Muscles of Calliactis Parasitica (Couch)

While much is known of excitatory phenomena in coelenterates, relatively little information is as yet available about inhibitory effects in these animals. Batham & Pantin (1954) have shown the existence, in Metridium senile (L.), of reciprocal inhibition between the parietal and circular muscles of the column and further postulate the presence of some inhibitory mechanism controlling the contractions of the circular muscles to permit peristaltic waves to pass down or up the column, although they also suggest that peristaltic wave formation may depend upon a slow conduction between muscle cells rather than on nervous activity. Among the Hydromedusae Horridge (1955) has shown that, following stimulation of its margin, the radial muscles of Aequorea forskalea P. & L. contract and that associated with this contraction is an inhibition of the activity of the circular muscle fibres responsible for the rhythmic contractions of the bell. A similar effect was demonstrated with representatives of many other genera of Hydromedusae.

Ross (1957) has shown the value of isolated rings of tissue in analysing the behaviour of the sphincter muscle of Calliactis parasitica (Couch), while Needler & Ross (1958) have demonstrated that the method is also applicable to the study of the activity of circular muscle from different levels of the column. The experiments described below have used this technique almost exclusively.

Specimens of Calliactis were narcotized in a mixture of approximately equal volumes of sea water and isotonic magnesium chloride for at least 6 hr. before preparation. The column was then divided into a series of rings and mesenteries were trimmed down with fine scissors. These rings can be broadly characterized as subsphincter, mid-column, low mid-column and pedal rings. The ring including the sphincter muscle was discarded and a circular incision made through the base of the foot to cut away the greater part of the pedal disk. The division between low mid-column and pedal rings corresponds approximately to the level of the cinclides. After dissection the rings were allowed to recover for 36-48 hr. in sea water which was changed at intervals; during this period they lay passively and were not stretched.

The electrode assembly described by Ross (1957) was used to stimulate the rings with short condenser discharges from the type of apparatus devised by Hall & Pantin (1937); the intensity of stimulation was generally about twice threshold. Recording was with lightly loaded gimbal levers, the preparation always being held at a slight tension. Latent periods were determined by stop watch. All experiments were conducted at room temperature which varied from 14 to 19° C. according to the time of year.

The effect of continuous low-frequency stimulation maintained for long periods depends upon the level from which the ring has been taken and also upon the age of the preparation. When a fresh subsphincter, mid-column or low mid-column ring is set up it will soon start to contract spontaneously, frequently in an almost regular rhythm. Each contraction lasts for about 3 or 4 min., the time interval between contractions varying from 7 to 25 min. Such a preparation, if left undisturbed, will continue to show spontaneous activity for long periods of time. If, after about 12 hr. of such spontaneous activity, the preparation is continuously stimulated at a frequency of one shock/2.75 sec.,^* it will respond, after a latency of 30-40 sec., by the development of a strong contraction which is maintained, with minor alterations in length, for the duration of the stimulation (Fig. 1 a). On cessation of stimulation tone is gradually lost during a period of apparent inactivity, spontaneous activity recommencing after an interval varying in different preparations from 5 to 20 min.

If, however, a fresh pedal ring is stimulated, a very different result is obtained (Fig. 1 b). In this case either all activity ceases for the duration of the stimulation or only very slight activity is found. When stimulation ceases activity starts again after a period varying in different cases from 3 to 20 min. This type of behaviour is also typical of subsphincter, mid-column and low mid-column rings when they are first used. However, as these preparations age the inhibitory effect of stimulation becomes less marked; thus Fig. 1,c and d show records of the response to continuous stimulation of the same low mid-column preparation, 90 min. and 12 hr. respectively, from the time it was originally set up. Pedal rings may continue to show the type of behaviour illustrated in Fig. 1 b until they cease to be responsive. In some cases, however, an aged pedal ring preparation will show a different picture: initially there is a period varying from 10 to 30 min. during which no response to stimulation is recorded, but then a tetanic contraction follows which may or may not be held for the duration of the stimulation.

That electrical stimulation is excitatory to mid-column rings has already been reported by Needler & Ross (1958). It also has an excitatory effect on pedal rings. This is most easily seen by using a brief series of shocks. After a long but fairly constant latent period varying in different preparations from 2 to 5 min. there is a strong slow contraction which takes from 90 sec. to 3 min. to reach maximal shortening. Relaxation is equally slow. That these slow contractions, despite their long latency, are related to the stimulation and are not simply ¹ spontaneous ‘contractions which occurred subsequently is demonstrated in Fig. 2 where a series of ten shocks at a shock-interval of 2.75 sec. was applied at varying times from the onset of the previous contraction. It can be seen from Table 1 that the latencies of these responses are not markedly affected by the interval between stimuli, while inspection of the record shows clearly that the contractions relate to the pattern of stimulation, the natural rhythm of activity of the preparation being far slower.

That brief stimulation of the pedal ring results also in an inhibition may be demonstrated in the following manner. A series of identical stimuli is given in regularly spaced intervals, timed from the onset of the previous response. If now, just as one of these contractions starts to develop, the preparation is again stimulated, the full development of the contraction is inhibited (Fig. 3).

If the number of shocks given is increased the latent period is increased (Table 2) while the magnitude of the contraction following is decreased. Prolonged series of shocks may be followed by only tiny contractions (Fig. 4).

While an increasing number of shocks invariably causes an increase in the value of the latency, the extent of this effect is variable. As can be seen from Table 2 its magnitude appears to be correlated with the age of the preparation. The readings with AZ were taken on the third day of use ; similar results had been obtained from this preparation 2 days earlier, while 2 days later the results were similar to that of the preparation BF, itself in the fourth day of use. Preparation BH was also 4 days old and completely unresponsive the following day.

The increase in the value of the latent period with the number of shocks may be regarded as being due, in some manner, to an enhancement of inhibition. Making this assumption, it is possible to use an increase in the value of the latency to determine the minimal number of shocks which will produce an inhibitory effect. A series of shocks, say ten, is given and then, 2 min. later and before the onset of a response, another series. The effect upon the latent period of varying the number of shocks in the second series can be observed. Results from experiments such as these are summarized in Table 3. It should be emphasized that in any of these preparations, although the value of the latent period to a constant stimulus is reasonably consistent over a short period, it may change fairly abruptly. As a result in many experiments concerned with latent period a standard stimulus of ten shocks was used at regular intervals as a check upon such variation ; further it was not usually possible to obtain large numbers of readings from a single preparation. An examination of Table 3 shows that three intercalated shocks markedly lengthen the latency, as do two. A single shock also, on the average, lengthens the latency but in this case there is considerable overlap with the control values; the difference is just significant at the 0.05 level in the first experiment, but not in the second. However, the number of observations is too small to allow any value to be attached to this statistical result.

The problem may be approached in another way which is best described by taking a specific example. A pedal ring showing frequent steady rhythmical contractions is selected; in the immediate case these contractions occurred every 7 min. Five minutes after the start of such a contraction a single shock was applied to the preparation; this resulted in a strong contraction and there followed a pause of 7 min. before another spontaneous contraction occurred. The preparation was allowed to make two further spontaneous contractions and 5 min. later two shocks, separated by an interval of 2 min. 33 sec., were given. A further spontaneous contraction did not occur until after an interval of 13 min. Again after two further spontaneous contractions three shocks at the same interval as before were applied and the pause before the next spontaneous contraction lengthened to 25 min. It is clear from this type of result that single shocks can be inhibitory in effect.

The effect of stimulation frequency upon latent period has also been examined. The results of a series of such experiments are shown in Table 4. As with other determinations of latency, the number of observations which can be made on any one preparation is small, but the figures suggest that there is no effect of stimulation frequency upon latency, or if any effect does exist it is very small compared with the influence of the number of shocks.

The inhibitory effect of stimulation may finally be demonstrated by its action upon the rate of relaxation from a slow contraction. If a preparation is stimulated by a short battery of shocks and then, as maximal shortening is reached, the preparation is again stimulated, the rate of relaxation is far more rapid than in a preparation which is not stimulated a second time. Such an effect is shown in Fig. 5 a and b. While any accurate assessment is impossible, an estimate of latency of the inhibitory effect can be obtained from these experiments. It is relatively brief, of the order of 15-20 sec.

In sum, the foregoing experiments demonstrate that electrical shocks applied to pedal rings of Calliactis have a dual effect, stimulating both an excitatory and an inhibitory system which control the activity of the circular muscles ; and further that such inhibition can also exist in more adoral regions of the column though the extent of its expression varies with the age of the preparation. A more detailed discussion of the results follows later.

If the latency of a pedal ring is compared with that of a mid-column ring which has been set up and active for many hours, they are seen to be markedly different. Such mid-column preparations show latencies of 30-40 sec., while values of the order of 120 sec. or more are typical of pedal rings. The latency of response of fresh mid-column preparations is, however, also of the order of 120 sec., but very soon a small initial contraction appears with a far shorter latency and, as time passes,this contraction becomes dominant. Fig. 6 shows a series of contractions from the same preparation taken at hourly intervals, while Table 5 records the latencies of these responses. It will be seen that initially there is a clear-cut double response, but that with the growth of the first contraction the two gradually merge. The latency of the onset of the second contraction does not, however, significantly alter. This same pattern of transformation has been observed in a considerable number of preparations.

A number of pedal ring preparations show double responses similar to the contraction recorded after 2 hr. in Fig. 6. The latency of the initial response in these cases is about 50 sec. but unlike a mid-column preparation a pedal ring preparation shows almost no variation in the pattern of response with time.

As has already been stated, rings of tissue from regions below the sphincter show spontaneous activity which is often of a fairly regular and rhythmical character. The experiments now to be described attempt some analysis of the genesis of this rhythm.

The regular rhythmic contractions of these circular muscles recall, albeit on a far slower time scale, the beat of a medusa bell. The rhythmic beat of a medusa such as Aurelia has been shown to depend upon the presence of pace-makers in the ganglion masses of the tentaculocysts, although the muscles may also show rhythmic spontaneous activity when the tentaculocysts have been cut away (Pantin & Vianna Dias, 1952). Clearly there is no anatomically defined pace-maker in the circular muscle of Calliactis, but an attempt may be made to answer the question as to whether or not pace-maker function lies within whatever system is responding to electrical stimulation. As Pantin & Vianna Dias (1952) have elaborated, if the pace-maker is an integral part of the excited system, an intercalated stimulus will either reset the rhythmic activity or possibly result in a missed beat so that there follows a ‘compensatory pause’. If, however, the pace-maker is not integral with the excited system an intercalated shock may have no effect or be followed by a compensatory pause.

Experiments of this type have been made upon a considerable number of midcolumn preparations and a wide variety of effects encountered. Examples are shown in Fig. 7, details of the experiments illustrated being given in Table 6. Intercalated stimuli may have no effect (Fig. 7 a); they may cause the rhythm to reset from the time of the intercalated contraction (Fig. 7b); an effect which could be interpreted as a compensatory pause is frequently found (Fig. 7 c), while pauses longer than compensatory pauses may also occur (Fig. 7d). All four of these events appear to occur with about equal frequency. In twenty-four preparations five showed no alteration in rhythm, five showed a simple reset, ten showed what could be interpreted as a compensatory pause and four showed longer delays. It seems likely that, as in Aurelia, many of the cases which have been classified as compensatory pauses are not so in any simple way. Thus in Fig. 7c a small contraction E developed after the intercalated stimulus at a time corresponding to a simple reset of the rhythm. In Fig. 7d the two long pauses can be regarded as due to two and three missed beats, respectively, which would have taken their rhythm from the intercalated contractions; that is, in both cases the rhythm has been reset by the intercalated stimulus.

Despite numerous trials it has not been possible to demonstrate any consistent effect of the number or frequency of the shocks in an intercalated stimulus or of the time of application of the stimulus in relation to the normal rhythm.

It is clear that, even if the cases of compensatory pauses are disregarded as spurious, cases occur both of a reset of the rhythm and also of the stimulation not altering the rhythmic pattern of the contractions. This combination of events is not expected in terms of either hypothesis concerning the possible site of the pace-maker. It seems probable that the reason for this lies in the mistaken initial assumption that the effect of stimulation is purely excitatory to these preparations of Calliactis muscle. As we have seen in the preceding sections, electrical stimulation has both an excitatory and an inhibitory effect; the precise balance of these varies in mid-column preparations with time. As a result, if the stimulus is strongly excitatory it may have no effect upon the rhythm, while if it is at least partly inhibitory in action a greater or less delay may follow depending upon the exact balance of excitation and inhibition produced by the intercalated stimulation. If this interpretation is correct, it follows that when the stimulus is purely excitatory the effect of an intercalated stimulus is without effect on the rhythm, and this carries the implication that the pace-maker mechanism is not an integral part of the excited system. Such an interpretation is in keeping with a further analysis of the origin of the rhythmicity arising from a more detailed study of the behaviour of mid-column rings.

So far detailed consideration of inhibitory phenomena has been limited to the behaviour of pedal rings and the question arises as to whether inhibitory effects may be demonstrated in mid-column preparations which have been active for many hours and show only a single response of short latency to electrical stimulation. We have seen above that when stimuli are applied in a regular programme to a pedal ring the magnitude of the shortening decreases and the latency increases with an increase in the number of shocks. This is partly true of mid-column preparations. Fig. 8 a shows the responses of such a ring to a series of ten and to a series of thirty shocks. It will be seen that while the response is less for the greater number of shocks, the latency is unaffected by the number of shocks (Table 7). It can further be seen that following stimulation with thirty shocks the next response to a further thirty shocks is less and that the subsequent responses to groups of ten shocks gradually grow. As with the pedal rings the inhibitory effect of stimulation is clearly of long duration and it may be taken that the magnitude of a response is a reflexion of the balance between excitation and inhibition—an assumption which cannot be made with the pedal rings owing to the varying duration of the latency.

Granted this assumption, it is possible to use a test battery of stimuli to investigate the level of inhibition in a preparation following a standard stimulus. Such an experiment is shown in Fig. 8 b where, following standard stimuli of ten shocks at a shock interval of 2.75 sec. test batteries of five shocks were applied at varying intervals after the onset of the contraction in response to the standard stimulus. It will be seen that as this interval increases the response to the test stimulus also increases in magnitude. The recorded latencies (Table 7) of the responses to the test stimuli are in general longer than those to the standard stimulus. This is almost certainly due to the fact that the former were measured to the time of first upward movement of the recording lever, while the true latency would be the time to the checking of the lever’s downward movement during the relaxation from the standard stimulus. Lastly, it should be noted that the record ends with a final response to the test stimulus alone and that this is greater than the penultimate response which followed a standard stimulus by $412$ min. It is thus clear that even after this period the response to the test stimulus is still partially inhibited.

Fig. 8 c shows a continuation of this type of experiment on the same preparation 2 hr. later with still longer intervals between standard and test stimulus. Details of stimuli and latencies are shown in Table 7. By this time the response to the test stimulus of five shocks 5 min. after the standard stimulus B was no greater than that 4 min. after the standard stimulus A. However, the response to a test stimulus of four shocks 5 min. after the standard stimulus C was markedly less. When the interval from the standard stimulus D to a test stimulus of four shocks was increased to 6 min., the response was greater and of the same magnitude as that shown to a test stimulus of three shocks 7 min. after the standard stimulus E. The response to two shocks 9 min. after the standard stimulus G is greater than that 8 min. after the standard stimulus F. Finally, the response to a single shock 10 min. after the standard stimulus H is almost maximal. That this latter response relates to the stimulating shock and is not simply a spontaneous contraction is shown by the fact that its latency is similar to that for responses to larger numbers of shocks. Fig. 9 shows high-speed recordings of responses from a different preparation to groups of shocks of varying number ; it will be seen that the characteristics of the response are almost unaffected by the number of shocks applied.

It becomes clear from these results that the responses of a mid-column preparation are not absolute but will depend upon the previous stimulation applied to a preparation. As a result it is not possible to make a useful analysis of responses to different frequencies and numbers of shocks. What is, however, of prime importance in an understanding of the genesis of the rhythmic activity of these preparations is the demonstration of a slow variation in excitability.

The preceding experiments have been concerned with two different phenomena in the circular muscles of the column and pedal ring of Calliactis, namely, inhibition and rhythmicity of spontaneous activity. These must be considered separately although, as will be suggested later, they may in fact be interrelated.

It is clear that electrical stimulation is not usually simply excitatory but has a dual action whose expression may change with the activity of the tissue. In a pedal ring we may visualize the long latencies recorded as being due to this dual action. On stimulation inhibitory effects dominate and then gradually die away until the balance of excitation to inhibition is reversed and a contraction follows. The duration of latent period thus becomes a measure of the relative magnitude of the inhibitory effect of stimulation. Evidence has been produced which shows that a single stimulus to a pedal ring can have an inhibitory action, and that the latency of the response is affected by the number of shocks applied (Table 2) but not by their frequency (Table 4). These facts suggest that each stimulus produces a quantum of inhibitory effect and that where many stimuli are given the latency is prolonged by the accumulation and subsequent dissipation of this inhibition.

Such an interpretation implies that the excitatory effects of stimulation persist for prolonged periods of time after the end of stimulation, dying away gradually. This die-away may be regarded as reflected in the fact that the longer the latent period the smaller the mechanical response (Fig. 4). The nature of the excitation in the coelenterate neuro-muscular system is as yet unclear, as is indeed the precise character of the events which are involved in such slow contractions as are developed by the circular muscle of Calliactis. Two views seem possible: one is that the slow contraction is the expression of a mechanically slow contraction of the individual muscle fibres, the other that these contractions are in the nature of prolonged tetani with the further implication that excitatory stimuli arrive constantly at the myoneural junctions during the course of the contraction. There appears to be no direct evidence for the former view. Horridge (1956) has, however, made direct observations of slow contractions in the radial muscle of Cyanea and reports that the ‘sustained weak contraction of the whole muscle appears to be a combination of many irregular local contractions out of phase with one another’, which would suggest that, in this case at least, the contraction is tetanic in character. Further, in the circular muscle of the pedal disk of Calliactis stimulation can accelerate relaxation (Fig. 5); this effect is most readily interpreted upon the assumptions that the contraction depends upon excitatory impulses continuously arriving at the myoneural junction and that their activity is inhibited by the release, following stimulation, of an inhibitory transmitter at the junction. Certainly there is no evidence that the muscle must already be in a state of contraction before inhibitory events can be effective for it has been shown (i) that intercalated stimuli, given before the development of a contraction, will prolong the latent period (Table 3); (ii) that stimulation applied just as a contraction is starting to develop will prevent its further development (Fig. 3); and (iii) that continuous stimulation will inhibit contractions over long periods of time (Fig. 1 b).

It may then be suggested as a provisional hypothesis that, with the circular muscles of the pedal ring, electrical stimulation liberates a slowly destroyed inhibitory transmitter at the myoneural junction and produces at the same time a state of prolonged excitation within the nerve net of that region which is maintained by some self-excitatory mechanism such as that postulated by Young (1938) in Sepia. The interaction of these two influences is responsible for the remarkably long latent periods found in these preparations.

Experiments using pedal rings designed to see whether the thresholds for excitation and inhibition were distinct gave negative results. This is hardly surprising as it seems probable from experiments with Metridium (Batham & Pantin, 1954) that electrical excitation stimulates the specialized through-conducting system, either directly or by way of sensory structures, while what may, by analogy with medusae (Horridge, 1956), be referred to as the ‘diffuse nerve net* is only indirectly stimulated. The properties which have been here ascribed to the ‘diffuse nerve net’ differ markedly from those already well attested for the through-conducting system; the observations of Batham (1956) show that in Metridium canum (Stuckey) these two systems are probably histologically distinct, and until detailed information on the structure of the ‘diffuse’ system is available there seems no profit in considering further the material basis of the phenomena which have been described above.

In regions oral to the foot the behaviour of the circular muscles is different. When an unstretched preparation is first used it behaves in a manner akin to a pedal ring, showing long latencies, but fairly rapidly the contractions become double (Fig. 6). The fact that the time to the development of the second part of the contraction remains approximately constant suggests that the changing form of the response is due mainly to the growing excitability of the preparation rather than to a diminishing inhibitory effect ; if the latter occurred the latency of the second contraction would, on the basis of previous assumptions, be expected to fall. Although this second latency is very variable, is showed no definite downward trend in a series of fourteen experiments in which it was observed. The persistence of inhibitory action is further demonstrated by experiments such as that shown in Fig. 8 a.

The reason for the growth of excitability with activity shown by adoral rings is not clear. It might possibly be due to an effect of stretching upon the muscle: the contractions recorded in Fig. 6 have been orientated so that the gradual stretching of the preparation may be seen. However, pedal rings also stretch in a similar manner, and although a double response may sometimes be obtained the magnitude of the first response is always small. The effect certainly cannot be due to the slow washing out of narcotic, for experiments by Dr D. M. Ross (personal communication) show that recovery of spontaneous activity in the circular muscle of Calliactis after a period of magnesium narcosis takes only a few hours ; moreover, if a fresh mid-column preparation is permitted to be active for some hours until a clear-cut double response is obtained and then allowed to lie unstretched for 12 hr. in sea water, the response of the preparation is almost unchanged when it is again set up for recording and stimulated.

It is suggested that the difference in latency between pedal and mid-column rings lies in the relatively lower level of excitation in the former. But there is some evidence that the normally observed latencies of the order of 35 sec. which characterize mid-column preparations are also prolonged owing to the presence of an initial inhibitory effect. In a limited number of mid-column preparations initial latencies which lay between 6 and 10 sec. were found. In these cases there was a small initial contraction followed by a pause and then, after the more usual latency of about 35 sec., the shortening continued. The behaviour of these preparations recalled that of a mid-column preparation giving a typical double response, albeit the present event was on a shorter time scale. Unfortunately, lack of time made it impossible to study this phenomenon further, and it cannot be excluded that the first response of brief latency may be due to a direct effect of the stimulus upon the muscle tissue.

The second problem to be considered is the origin of the rhythmical spontaneous activity shown by these preparations of circular muscle. This property is not dependent upon the intactness of a ring, being shown as clearly by rings cut across and ligatured at both ends, and also by rings cut obliquely across the column so that no complete circle of muscle fibres remains.

The experiments illustrated in Figs. 8 b and c show that following a contraction the response to a test stimulus grows as the interval between standard and test stimuli lengthens, and that if this period is sufficiently prolonged an all but maximal shortening of the ring may follow a single stimulus. It has been suggested that this effect is due to the presence of an inhibition whose slow decay is reflected in the gradual development of excitability in the preparation. Clearly such a mechanism could produce a rhythmical series of contractions if it is assumed that a naturally occurring spontaneous contraction arises from excitation within the nerve net, but that, in its occurrence, it produces a prolonged inhibition which prevents the development of a further contraction until it has died away. Such an explanation receives support from the results shown in Fig. 7 a, where it will be seen that the extent of the shortening to intercalated stimuli increases as the time between the start of a spontaneous contraction and the stimulus is increased. Indeed there is a very striking similarity between Figs. 7 a and 8 b , albeit one shows spontaneous activity and the other an artificially imposed rhythm. The growth of relative excitability with time can also be seen in Fig. 8 c where small spontaneous contractions occur when the time interval between standard and test stimulus is long.

It is therefore suggested that the rhythmical contractions arise from an inhibitory feed-back generated in some manner by the contraction of the circular muscles themselves. Such an interpretation is in keeping with the results obtained when studying the effect of intercalated stimuli upon the spontaneous activity rhythm (Fig. 7). It was in this case suggested that where the intercalated stimulus is purely excitatory in action the normal rhythm of activity is not upset, from which it would follow that the pace-maker is not integral with the system excited by electrical stimulation. If indeed the rhythm is controlled by a prolonged inhibition at the myoneural junctions normally initiated in some manner by the spontaneous contractions themselves, this is the type of result which would be expected to follow.

There are various ways in which this feed-back could originate. The first is that there is a simple reciprocity between the circular and parietal muscles of the column. A few preliminary experiments were made with preparations consisting of a ring of subsphincter or of low mid-column to which a tail of the column remained attached. Recordings of the activity of the ring and of the tail simultaneously gave no indication of any reciprocity in the activity of circular and parietal muscles. The second possibility is that there is some type of proprioceptor in the column which, stimulated by a contraction of the circular muscle, excites the inhibitory mechanism in the anemone. Evidence in favour of such an assumption cannot be obtained with the methods used in these experiments, for it would depend upon being able to stimulate the muscles to contract without also stimulating the inhibitory mechanism ; this cannot be achieved with existing techniques of electrical stimulation. The final possibility is that the excitatory influences which cause the circular muscles to contract also excite the inhibitory mechanism so as to bring the contraction to a stop. Again no arguments for or against such an hypothesis can be advanced upon the basis of the present evidence and technique.

Mention must be made of the great variation in character of the pattern of activity of different preparations. An analysis of the performance of thirty-three preparations has been made; these have been divided into three classes—those showing complete inhibition of activity when continuously stimulated, those showing a partial but definite contraction and those showing a strong and sustained contraction. These three categories correspond to Fig. 1 b, c and a , respectively. In nine out of twelve preparations assigned to the first group the spontaneous pattern was irregular, while in the three cases in which there was a regular rhythm it was slow with one contraction every 15-20 min. The second and third categories both included a few preparations showing an irregular pattern ; four out of eleven in the second and three out of ten in the third. When the rhythm was regular its frequency was relatively high with a mean interval of 11 min. in both categories. It is clear that those preparations which more readily display inhibition also show a slower rhythm—a result which would follow from the hypothesis already elaborated.

In some preparations it is clear that complex rhythms arise because the circular muscles of the preparation are not all contracting together; the muscle has fragmented into a series of independent regions which may each have a slightly different rhythm and their resulting activity will be a series of contractions of varying extent and without any pattern obvious to casual inspection.

The hypothesis which has been elaborated to account for the origin of rhythmical spontaneous activity in the circular muscles of Calliactis is broadly similar to that which has been suggested as an explanation of similar events in the pharyngeal retractor muscle of Cucumaria (Pople & Ewer, 1958). In both cases it has been assumed that the slow contraction, which has a very similar time course in both preparations, is a tetanic contraction; it is possible to show in Cucumaria that stretching of the pharyngeal longitudinal muscle, which produces an inhibition, results in an acceleration of the rate of relaxation of the retractor muscle following a spontaneous contraction, just as electrical stimulation will accelerate relaxation in Calliactis circular muscle (Fig. 5). In both cases it is suggested that an inhibitory feed-back mechanism of prolonged activity is involved and in both it is suggested that this inhibition acts in opposition to a general excitatory input. In both cases evidence for inhibitory phenomena and for the existence of balances between excitatory and inhibitory influences has been presented. But while in Cucumaria it has been suggested that the inhibitory event lies within the motor complex proper to the pharyngeal retractor muscle, in Calliactis the evidence suggests that this may lie at the level of the myoneural junction. However, the results presented here cannot exclude the possibility that the spontaneous activity of the circular muscles of Calliactis is myogenic in origin and that electrical excitation stimulates extrinsic excitatory and inhibitory nerves just as the nerve supply to the myogenic heart of a vertebrate can control its rhythm.

These experiments were made in the Department of Zoology, University College, London. My thanks are due to Prof. P. B. Medawar, F.R.S., for his kindness in allowing me to work in his Department and to Dr D. M. Ross not only for putting at my disposal both his equipment, and his extensive knowledge of Calliactis but for many invaluable critical discussions and comments. My thanks are also due to Mr L. Sutton for his skilled technical assistance.