Facilitation in Sea Anemones: III. Quick Responses to Single Stimuli in Metridium Senile

Quick protective closing movements are a feature of the behaviour of the sea anemones Metridium senile and Calliactis parasitica. By studying the way in which these movements can be elicited by electrical stimulation, Pantin (1935 a–d) has shown that the response is controlled by a process of facilitation, developed here to a unique degree.

A single electrical stimulus applied anywhere on the column of the animal completely fails to bring about a quick response. A series of stimuli at intervals of about 3 sec. or longer is similarly ineffective, though such a series may cause slow movements of various kinds after some delay. If the interval between stimuli is shortened, a point is reached where tiny quick-closing movements occur to every stimulus after the first. Further shortening of the interval has the effect of increasing the size of the separate quick-closing movements which occur on the second and subsequent stimuli. Apparently the only effect of a single electrical stimulus is to set up a state of facilitation which falls off over a period of 2 or 3 sec. A subsequent stimulus is only effective if it is given before the state of facilitation has decayed, and the size of the response to such a stimulus will depend on how far the decay of the facilitated state has gone.

Published accounts of the nervous system and muscles of actinians do not throw much light on where and how this process of facilitation takes place. However, by showing that the ineffective stimulus throws the whole nerve net into the refractory state, Pantin (1935 a) established that the facilitation process occurs at or near the neuro-muscular junctions. He therefore called it neuro-muscular facilitation.

Some of the physiological properties of neuro-muscular facilitation emerged from investigations on the influence of various treatments upon it. Hall & Pantin (1937) showed that the process was affected by temperature changes; the facilitated state could be prolonged by cooling and shortened by warming the sea water bathing the anemone. Subsequent papers described the changes in the quick response induced by various ionic conditions and drug treatments (Ross & Pantin, 1940 ; Ross, 1945 a). In general, the character of the response was not altered by any of these treatments. Excess K+ and Ca++ increased, and excess Mg++ diminished, the size of the response at any given frequency, but without altering either the duration of the facilitated state, or the rule that every stimulus except the first is effective within the appropriate frequency range.

Most of the drug effects followed the same pattern. Some drugs, especially tyramine, 933 F and cocaine, increased, others, including ergotoxine and trimethylamine, decreased the size of the response, but again without affecting the duration of facilitation. Yet some of these treatments did occasionally alter the rule that only stimuli after the first were effective. After tyramine and 933 F particularly, and with several other biological amines to a lesser extent, Calliactis showed a tendency to respond to single shocks during certain phases of the treatment. This evidence that a few substances having effects at adrenergic nerve-endings elsewhere had facilitating effects on the anemones was consistent with the view that facilitation was a neuro-humoural process carried out by a ‘facilitator’. Attempts were made, therefore, to see whether substances with facilitating properties, i.e. inducing responses to single shocks and prolonging the facilitated state, could be detected in material extracted from the anemones themselves. Preliminary results of such tests have been described (Ross, 1945 b), and these showed that responses to single stimuli did occur frequently following treatment with extracts of both Calliactis and Metridium. These responses to single stimuli, unlike those with such drugs as tyramine, occurred very soon after the extract was introduced (1–5 min.). Furthermore, they were generally not accompanied by changes in the size of the response, and the rate of decay of facilitation was not affected. Thus, notwithstanding the responses to single stimuli obtained by this treatment, it could not be maintained that the active material in the extract possessed all the properties that would be expected of a ‘facilitator’.

This paper deals with further tests undertaken to clarify the situation in which the earlier work had left the problem. In the main, the experimental work consists of a new description of the effects of anemone extract on Metridium. This description includes data on the frequency of the occurrence of responses to single stimuli, on the potency of various extracts and on temporal aspects of the effects. But it was found early in the work that responses to single stimuli occurred with certain other treatments, and so a survey of the effects of a variety of substances and conditions forms part of the account presented here. It might be as well to state in advance that the work has led to the exclusion of the possibility that the extract effects are due to the activity of a ‘facilitator’. In spite of this, the effects are still important because from their nature it is possible to learn a good deal about the quick response and the processes of excitation which initiate it in the normal animal.

The effects of the extracts were studied in essentially the same way as in the earlier work. Minced anemones, usually five or ten at a time, were extracted with cold ethyl alcohol, and the extract so obtained was concentrated under reduced pressure to about one-twentieth of its original volume. Tests were carried out by introducing small quantities of this concentrate (diluted in sea water) directly into the sea water bathing the test animal.

The effects of the extract on a sea anemone were observed in the earlier work by giving single stimuli at irregular intervals from a stimulator delivering condenser discharges through fluid electrodes. In the work now reported it seemed desirable, using the same stimulator, to send in shocks at regular intervals well outside the frequency at which quick responses normally occur. The frequency most commonly employed was one stimulus every 10 sec. It was expected that such experiments would show when and how often the responses to single stimuli occur, and how long the effect lasts. Moreover, the use of a standard repeatable procedure should allow quantitative comparisons of the potency of different extracts and treatments to be made.

Some of the tests were recorded on smoked paper, and these give a continuous graphical record of the results. But even very tiny quick responses are easily detected by eye, so where data on the number of responses and the times at which they occurred were all that was required, it was the usual practice to follow the experiment visually and record these data for controls and treatments alike.

Preliminary experiments confirmed the earlier observation that following the introduction of small quantities of Metridium extract into the sea water, Metridium frequently responds to single stimuli. These preliminary tests showed that if this effect occurred at all, it could usually be detected within the first 3 min. of the treatment. It therefore became the standard practice to continue the tests only for 3 min. after the introduction of the sample of extract.

Figs, 1 and 2 show two typical smoked records taken during treatments of Metridium with Metridium extract and bring out clearly the character of the effects induced. The control series shows the typical absence of quick responses in an untreated animal stimulated at a frequency of one shock, every 10 sec. and the response to a pair of shocks 1 sec. apart at the end. Following the introduction of the extract, responses occur occasionally to isolated stimuli in the series, to shocks at 10 and 20 sec. in one example and 1 min. 50 sec. in the other. Clearly the effect of the extract is to pave the way for quick responses only to occasional stimuli, not to every stimulus over an extended period. This sporadic aspect of the effect was not so evident in the earlier work, and its significance will need to be carefully considered. Fig. 3 shows a similar result with stimuli every 15 sec. in which quick responses occurred 15 sec. and 2 min. 15 sec. after the beginning of the treatment. Fig. 4 shows the same type of record with stimuli every 5 sec. At this frequency slow contractions occur which are reflected in the lowering of the lever after five stimuli in the control series. Following the treatment quick contractions occur at 15 sec. and 1 min. 55 sec., then to three successive stimuli beginning at 2 min. 30 sec. and to three more beginning at 3 min. 10 sec. In some other cases responses occurred to successive stimuli in this way, but in general such results are exceptional, as will be seen from statistics to be presented later.

The size of the responses to single stimuli deserves some comment. That the size of the individual contractions is variable is perfectly clear from the records, and this has been the experience of the unrecorded tests as well. No maximal contractions have been recorded. In the unrecorded tests a preponderance of powerful over small contractions has been indicated, but, of course, in the absence of records it is not possible to be certain of this. What is clear is that the upper limit in the size of the responses to single stimuli seems to lie within the range given by pairs of stimuli within the facilitation period. The implications of this are important and will be considered later on.

These records illustrate one noteworthy feature of the results obtained from both recorded and unrecorded tests. This is the comparative infrequency of quick responses elicited directly by the extract. Such direct effects would take the form of quick closing movements when the extract was introduced or between the stimuli in the series. Figures to be presented later will show that such spontaneous movements are absent in the vast majority of tests, a fact of some significance in determining the origin and significance of the effects.

There are other direct effects that occur more frequently during the treatments, but which do not show up on the smoked records. Thus some twitching and withdrawal of the tentacles usually occurs as soon as the extract is introduced. With weak doses this ceases between 30 sec. and 1 min. With stronger doses the tentacles may remain withdrawn and a slow retraction of the whole animal may occur producing a state of partial closure. As a rule this condition is not maintained longer than 2 or 3 min., and unless closure occurs in response to the electrical stimuli, the anemone is found in the expanded condition at the end of the treatment. The records show also that at the end of the treatment the response to a pair of stimuli 1sec. apart is virtually unchanged in size and form, and it can be established that there is no effect on the duration of the facilitated state.

For most practical purposes it can be assumed that at ordinary temperatures Metridium and Calliactis never give quick responses to single stimuli or to individual stimuli in a series more than about 3 sec. apart. This is not strictly true. Untreated animals occasionally do respond to single stimuli or to occasional stimuli in a low-frequency series of the kind employed in these tests. A simple explanation for such responses is that the effective single stimulus follows closely on or precedes an impulse set up in the nerve net by ordinary sensory stimulation. Because the effect of the extract usually takes the form of occasional quick responses in such a series, it is desirable to establish exactly how often quick contractions occur in untreated animals and to compare this with the frequency of such responses in animals treated with extracts.

Data on the frequency of quick responses in untreated animals can be obtained from the controls carried out in the tests. Early in the work a series of ninety-six tests were carried out in which five Metridium extracts were tested on Metridium over a moderate dosage range, and from this series many of the quantitative data about the responses to single stimuli have been derived. In the controls of these ninety-six tests a total of eight quick contractions occurred to 1400 stimuli, i.e. a frequency of 1 in 175. A further 179 controls were carried out in connexion with other work, and in these eleven quick contractions occurred to 3136 stimuli, i.e. a frequency of 1 in 285.

Information on this point comes also from experiments in which Metridium was stimulated for long periods at frequencies of 5, 10 and 30 sec. A typical record of such an experiment is shown in Fig. 5, where a stimulus was given every 10 sec. for 24 hr. The quick responses can be recognized as sudden downward movements of the lever, all the smaller movements being slow and almost imperceptible movements of the animal. Not more than thirteen of the movements on this record can be typical quick contractions. Assuming that all these are in response to thirteen out of the 8640 electrical stimuli delivered in 24 hr., this experiment shows a frequency of one quick response for 665 stimuli in an untreated animal. Data from this and other experiments of the same type are summarized in Table 1. Although there is considerable variation in the frequency of such responses in the different experiments, it never rises above 1 per 100 stimuli and the average figure is 1 in 400.

There is a fair measure of agreement between both these sets of results, from the controls of the tests and from long-continued stimulation. Considering all these data together one can conclude that a normal untreated Metridium requires of the order of hundreds of single or low-frequency stimuli for each quick response that occurs.

The same ninety-six tests whose controls were referred to in the preceding section provide the essential data for estimating the frequency of quick responses following the treatment. These tests were all carried out with moderate doses, and the extracts used were all ordinary crude extracts tested in a routine manner. To the 1400 stimuli given to animals in the course of these ninety-six tests, 155 quick responses were obtained. This is a frequency of 1 in 9 compared with 1 in 175 in the controls of the same tests. Thus the effect of the extract may be expressed statistically as a 20-fold increase in the probability of a quick contraction in the course of low-frequency stimulation lasting 3 min. It is useful also to consider the data from these ninety-six tests from the standpoint of the number of quick responses observed per control and per treatment. Table 2 sets out the results in this way, and shows that whereas quick responses occurred in only one-thirteenth of the controls, they occurred in five out of every six treatments. Exactly two-thirds of the treatments gave only one or two responses and only the remaining one-sixth (16) gave three or more. The highest number of responses observed in the course of a single treatment was ten.

The extracts were prepared from five or ten moderately large Metridium or Calliactis. Volumetric samples introduced in the tests therefore represent fractions of the material extracted from these animals. For want of a more precise quantity, an arbitrary unit representing the material extracted from a single anemone will be used in comparing the effects of different doses and these will be termed M (Metridium) and C (Calliactis) units. These tests were usually done in 50 or 100 c.c. containers using small Metridium as the test animals, a measured sample of concentrate being run in rapidly to ensure thorough mixing with the sea water bathing the animal. It is convenient to express doses as M or C units per 100 c.c.

Preliminary work confirmed the rough observation reported in the earlier work (Ross, 1945b) that the introduction of samples containing less than one-tenth of the material extracted from a single anemone was followed by the occurrence of responses to single stimuli on occasion. It was important in comparing the potency of different extracts to find out what is the effect of varying the dose and especially to determine the minimum effective dose as accurately as possible. The number of responses occurring to stimuli at a frequency of one in 10 sec. in ten tests lasting 3 min. was taken as a measure of the effectiveness or potency of a given treatment.

Fig. 6 summarizes in graphical form the number of responses occurring in ten such tests with three different extracts of Metridium over a dosage range of 0 · 1–0 · 0045 M units in 100 c.c. of sea water. Thus each point on the graph represents the number of responses obtained when 180 stimuli are given. There is clearly a close relationship between the quantity of extract introduced and the number of responses obtained and a fair measure of agreement in the dose-potency curves given by the three different extracts. Thus all three extracts agree in showing over twenty quick responses per ten treatments with 0-1 M units and less than five responses below 0 · 005 M units. This lower figure approaches the frequency of responses in untreated animals (1in 175 in the controls of the ninety-six tests) and must therefore be considered as close to the minimum effective dose.

A noteworthy feature of the potency-dose curve is represented in Fig. 6 by the tests with extract 4 in which doses were reduced by small amounts over the range 0 · 0075–0 · 004 M units. Here there is evidence of a sharp falling off in the frequency of quick responses, a result that is highly suggestive of the approach to a threshold dose.

The absolute quantities represented by the doses of extract used in these tests are small. The dry weight of the solid matter contained in samples of extract 4 was determined and has given a figure of 680 mg. as the weight of an M unit in this case.

Thus the quantities of extracted substances contained in doses from 0 ·1 down to 0 · 004 M units range from 68 to 2 · 7 mg. per too c.c. These are equivalent to dilutions of solid substance in sea water of 1 : 1500 to 1 : 37,000. If as seems likely, the substance (or substances) responsible for inducing these responses to single stimuli forms only a small proportion of the extracted matter, the active principle in the extract is probably exerting its effects over a dilution range of hundreds of thousands to millions of parts in sea water. Clearly it is a sensitive reaction, though less so than the direct response of certain isolated effectors in other animals to substances of great activity like acetylocholine.

The extracts whose potency we have just considered were obtained from animals stimulated intensely by the handling and cutting involved. Attempts were made to prepare extracts from unstimulated animals for comparison with the ordinary extracts. One method used was to chill the animals to–2 ° C. and anaesthetize with MgCl₂ (0 · 4 m), treatments which enable animals to be handled and cut up with little or no muscular response. Another method was to * quick-freeze ‘the anemones with liquid air, a treatment which fixes them in the expanded condition and must kill them almost instantaneously.

Fig. 7 sets out the results of tests of such extracts as potency-dose graphs, with one ordinary extract from Fig. 6 for comparison. These results show no significant difference between the activity of extracts from ‘stimulated’ and ‘unstimulated’ animals; the number of responses per ten treatments is of the same order over the range of doses examined. The results of tests on ‘quick-frozen’ animals are less conclusive, but at the near-threshold doses, to which perhaps more importance should be attached, such extracts appear to display the normal degree of activity.

Some tests on extracts of Calliactis were also carried out. It was confirmed that these caused responses to single stimuli exactly like the Metridium extracts. The potency-dose curves were similar in form, but the minimum effective dose is slightly lower, between 0-003 o-ooi C units.

In both Metridium and Calliactis separate extracts were prepared from animals divided into ‘sphincter-disk’ (includes tentacles, disk and marginal sphincter) and subsphincter’ fractions, and the activity of the two fractions compared. Table 3 presents the results and shows that the extracts from the sphincter-disk fraction are consistently more effective in producing responses to single stimuli. The activity of the sphincter-disk extracts is relatively even greater than these figures indicate because this fraction contains only about one-quarter of the total solid in Metridium and one-third in Calliactis (estimates based on wet weights of residues after filtering the crude extract). It is clear that whatever the nature of the presensitizing substance and its role in the living anemone, it is present in much larger quantities in the disk, tentacles and sphincter than in the rest of the anemone. Perhaps it is worth noting that this division would probably correspond with a division based on nematocyst distribution.

The distribution of the quick responses over the 3 min. period following the introduction of the extract in the ninety-six standard tests is shown in Fig. 8. There is a pronounced tendency for the responses to occur early in the treatment, almost one-half of the responses occurring in the first minute and less than one-fifth in the third minute of the treatment. Further breakdown of these figures has shown that a large proportion (30 out of 155) occurred in the quarter-minute between 15 and 30 sec. after the extract sample was introduced. This underlines the rapidity with which the effect makes its appearance, a feature which was not so evident in the earlier results and a point of some importance in considering the nature of the effects.

Further work has shown that there is some relation between the time of appearance of the quick responses and the amount of extract introduced. Table 4 brings out this fact, showing that as one passes down the dosage range from 0 · 5 to 0 · 005 M units, the proportion of the responses occurring early in the 3 min. tests falls off, while the proportion in the later stages of the treatment rises sharply. This is perhaps what one would expect on a priori grounds assuming that the effect reflects the rise of a substance in solution to a threshold concentration, the threshold being reached earlier the higher the dose introduced.

Perhaps the most characteristic temporal feature of the effect is its sporadic and discontinuous character. Thus it is unusual for responses to occur to successive stimuli in a series. If we examine the data from the ninety-six standard tests, we find that out of the 155 quick responses obtained in these tests, 129 occurred to isolated stimuli. Only 7 pairs of responses to two successive stimuli and only four groups of three responses to three successive stimuli were observed. Clearly the condition set up by the extract, which enables a single stimulus to cause a response, is generally of short duration, in most cases not lasting longer than 10 sec. if one takes the data from these tests at their face value. Four of the other tests with extracts, not included in the ninety-six standard series, have produced responses to four successive stimuli ; on two other occasions responses to five successive stimuli were observed. Yet the overall picture is the same, suggesting a condition that comes and goes during the treatment, and perhaps does not outlast the ordinary decay period of facilitation. This aspect of the results must assume considerable importance in interpreting the effects.

The observation that extracts of Metridium and CaUiactis enable responses to occur to occasional shocks in a low-frequency series so soon after being introduced, led to the re-examination of the effects of other treatments studied previously over a longer time scale. This matter has not been exhaustively investigated, but enough has been done to estabfish the fact that a variety of treatments can cause responses which are apparently similar to the responses to single stimuli which follow the extract treatments.

Fig. 9 illustrates the effect on Metridium of suddenly increasing the acidity of the sea water to pH 3 · 0 and the occurrence of a response at 20 sec. Fig. 10 shows records taken before and after treatment with sodium taurocholate (1 : 200) in which several responses to single stimuli may be seen. Table 5 summarizes the results of these and other treatments according to the number of responses obtained per ten tests. It seems that most treatments have some effects in this respect, but usually only at high concentrations. In certain cases the effect is very slight. This is particularly so with some of the drugs which have the most pronounced neuro-muscular effects elsewhere, acetylcholine, histamine, etc. No substances of this type approach the activity of the extracts which gave upwards of twenty responses per ten tests with higher doses. The ineffectiveness of tyramine and the relatively slight action of 933 F are noteworthy here, in view of the effects of these substances in increasing the size of the quick responses to pairs of stimuli and causing responses to single stimuli in Calliactis after 1–3 hr. treatment (Ross, 1945 a).

The reagents which have the most pronounced effects here are strong acid and alkali, potassium chloride (K+ × 8) and especially bile salt and saponin. These are drastic treatments which might be expected to have widespread non-specific effects on a whole animal. Yet if one compares this list with Pantin’s (1942) list of sub stances causing the nematocysts in the tentacle of Anemona to discharge, it becomes apparent that treatments of the same character are involved in both cases. In this respect the special activity of bile salt must be regarded as highly suggestive.

The effects of these treatments do not differ in any obvious way from the effects observed with extracts. The responses tend to occur more frequently in the first minute of the treatment. They generally occur to isolated rather than successive stimuli ; in other words, the effect is usually sporadic and of short duration. Nevertheless, a number of remarkable results were observed in some of the tests with the more effective treatments, especially bile salts, saponin and KC1. These results showed that it is possible with these reagents, as with extracts on occasion, to obtain quick responses from as many as four successive stimuli 10 sec. apart, so the effect is not invariably short-lived.

The effects of these treatments are not accompanied by an appreciable number of spontaneous contractions, even at the relatively high doses here employed. If much higher doses than these are used, the introduction of the reagent often causes an immediate and complete closure of the animal and re-opening may not occur. It is noteworthy that, as with extracts, there seems to be no transitional phase in which a high proportion of responses to single stimuli grades into increasing degrees of spontaneous activity. The latter has a distinct all-or-nothing character which is very striking indeed.

With the exception of K+ × 8, bile salt and saponin, the activity of these reagents is generally far below that of the extract treatments. There is certainly no possibility that the extract effects can be attributed to any of the substances here tested. Tétraméthylammonium hydroxide, identified in extracts of Actinia equina by Ackermann, Holtz & Reinwein (1923) and suggested by them as the poison of the nematocysts, is only active at concentrations far above any it could possibly have in the extract treatments. Moreover, the doses employed to induce the effects are too strong for it to be suggested that any of these substances could be involved in the physiological processes of the quick response in the normal animal. There is a contrast here with the extract effects, for we have seen already that any active principle in the extract is probably diluted hundreds of thousands of times in the treatments over the moderate dosage range.

It seemed from the earlier work with extracts (Ross, 1945b) that the presence of a substance in extracts of sea anemones which enabled single stimuli to cause responses lent considerable support to the notion that facilitation is a chemical process and the active substance in the extract a ‘facilitator’ in the anemone itself. This view may have reflected an unduly strong influence of the concept of chemical transmission on the interpretation of those results. At any rate, for a number of reasons, the view that the effects of the extracts can be attributed to a ‘facilitator’ now seems unaceptable.

The demonstration that such effects can occur with other treatments besides the extracts would not by itself be inconsistent with a chemical interpretation of facilitation. Indeed, if facilitation were a chemical process, it would be surprising if certain substances with widespread effects on living tissues did not have a facilitating action, especially when employed in powerful doses. Nevertheless, the fact that similar effects can be produced by so many different treatments makes it necessary to be cautious before regarding the extract effect as anything more than a general response to substances not normally present in the sea water.

The results of the potency tests on extracts from ‘unstimulated’ animals are relevant in this connexion. If the responses to single stimuli were due to a ‘facilitator’ it would seem reasonable to expect that the activity of such extracts, even if not completely abolished, would at least be substantially depressed. The results, however, provided no evidence that the extracts from anaesthetized or quick-frozen animals were less potent than extracts prepared in the ordinary way. It therefore seems unlikely that there is any connexion between the active substance (or substances) in the extracts and the processes of stimulation and response in the normal animal.

Finally, the consistently early appearance of the effects, as early as 10 sec. after the introduction of the extract, has some bearing on this matter. It is hardly credible that in such a short space of time the active substance could enter the tissues of the animal and reach the junctions where the facilitation process is apparently carried out. This feature of the effect is much more in keeping with some superficial action by which the extract sets up facilitation indirectly and not by itself acting as a ‘facilitator’.

If the responses to single stimuli are not due to a direct facilitating action, then we must look for their origin in excitation conducted to the muscles through the nerve net. There is overwhelming evidence in Pantin’s papers (1935 a-d) that general excitation in the nerve net takes the form of discrete impulses in the through-conduction tracts, impulses whose frequency determines which muscles will be brought into operation. We can apply this fact to interpret the effects described in the foregoing account.

It will be obvious that a response will occur whenever a single electrical stimulus happens to be preceded or followed by an adventitious impulse in the through-conduction tracts. Of course such a response would not be a response to a single impulse at all but a response to the second of a pair of impulses. This is how one tends to explain the few quick responses which occur in untreated animals (p. 239). Is it possible that the effects of the treatments simply reflect the passage of occasional impulses in the nerve net? Is it possible in fact that in these results we see the effect of increasing the frequency of an event which is going on all the time in the normal animal at a very low frequency indeed?

The evidence does not yet allow an unqualified answer to be given to either of these questions, but there are certain important features of the results that are consistent with affirmative answers to both. Thus one cannot fail to be impressed with the sporadicity of the responses to single stimuli observed in these tests. Figures cited above have shown that the majority of these responses occurred in response to isolated and not to successive stimuli. Although up to five responses to successive stimuli have been observed, it remains true that in any given test the responses seem to occur in a random and discontinuous fashion. Both these features of the effects are in keeping with the conception that the treatments are setting up occasional impulses which now and then are close enough to some of the electrical stimuli for apparent responses to occur to the latter.

Again one must note the variability in the size of the responses to single electrical stimuli that have been observed in the tests. We have seen that the size of the individual responses generally varies between the range given by pairs of stimuli in the normal animal. This, too, is consistent with the view that the response to a single electrical stimulus occurs when it is accompanied by an adventitious impulse within the normal facilitation period.

There are several other features of the observations that are, however, less consistent with this view, at any rate in this simple form. If the effect of the treatments is to set up impulses in the general nerve net similar to those set up by electrical or mechanical stimuli it would seem reasonable to expect a good deal of spontaneous activity. After all, whenever two or more impulses are close enough together (2 sec. or less), spontaneous quick responses should occur in the absence of electrical stimuli altogether. However, such effects are rare. In the ninety-six tests which yielded 155 responses to single electrical stimuli only ten spontaneous contractions were observed, most of them (seven) when the extract was being run in, or immediately after, when any response might equally well be due to mechanical stimulation as to direct chemical effects. Thus if the responses to single stimuli are to be ascribed to adventitious impulses set up by the treatments, it must be assumed that these impulses are generally too far apart to cause quick contractions by themselves.

There is indeed some evidence that treatments of this kind do set up low frequency excitation waves in the nerve net. Pantin (1935b, 1950) has shown that certain of the slow reflex movements, which can be elicited by food substances and other chemicals, can be obtained by electrical stimuli at frequencies outside the range which brings in the quick response. Yet if low-frequency impulses of this kind are responsible for the effects observed in these experiments, it is puzzling that food stimulants like molluscan juices, which presumably set up low-frequency excitation, do not cause responses to single stimuli (Table 5). Evidently one must bear in mind that there are qualitative differences in the excitation set up by different kinds of stimuli as well as differences in the frequency of impulses set up in the general nerve net.

The results show a second curious feature. If the responses to single stimuli are due to adventitious impulses which happen to occur close to some of the electrical stimuli, one would expect such adventitious impulses to follow the electrical stimuli just as often as they precede them. Thus, in a large number of experiments, one should get about equal numbers of responses ‘on the stimulus ‘and responses which follow the stimulus only after a delay of anything up to 2–3 sec. This expectation is not fulfilled. Delayed responses do occur from time to time (cf. Fig. 10 for an example), but they are remarkably infrequent. In the ninety-six standard tests 155 responses were obtained, all of them on or very close to the stimuli. In the same tests only five delayed responses were observed. In the other tests the number of visibly delayed responses was equally insignificant. If adventitious or low-frequency impulses are indeed responsible for the responses to single stimuli, it is surprising that they should so often precede, and so seldom follow, the electrical stimuli. The matter can be explained, of course, by ascribing exceptional properties of refractoriness to those parts of the nerve net in which the adventitious impulses arise, or by attributing inhibitory effects to the electrical stimuli. Yet to suggest that is to imply that the responses to single stimuli are associated with a special part or activity of the nerve net, a part whose properties differ from the through-conducting system. In that case one cannot assume that the excitation set up by such an activity is identical in scope or character with the general excitation of the through-conduction tracts. To conclude. It can be stated that the general character of the responses to single stimuli is in keeping with the insertion of extra impulses into the series set up by the electrical stimuli employed in the tests. At the same time, the capacity of these impulses to set up facilitation contrasts sharply with their incapacity to call forth the quick response, judging from the infrequency of spontaneous and delayed contractions. This might be due to the low frequency of such impulses and to special refractory or inhibitory properties in the sensory or nervous elements involved. Alternatively, it might reflect a genuine incapacity to initiate the contractile phase of the response on the part of impulses set up by the treatments. That would imply separate processes or channels of excitation, responsible separately for the facilitating and excitant phases of the response. It is worth recalling in this connexion that earlier work, especially on ions and drugs, emphasized that the two processes involved in the quick response—facilitation and excitation—are essentially independent (Pantin, 1937; Ross & Pantin, 1940; Ross, 1945a). Perhaps the possibility of distinct facilitating and exciting modes of local excitation in the neuro-muscular units taking part in this response is one that should be borne in mind in future work.

It is a pleasure to acknowledge a special obligation to Mr Graham Hoyle who, during the final year of an Honours Course, assisted me by carrying out some hundreds of tests on anemone extracts and other substances. Data from these tests provided a major part of the information on the potency of extracts from stimulated, unstimulated and divided animals, and the discussions we had together were of real value to me. I am also much obliged to the Director and Staff of the Plymouth Laboratory for facilities enjoyed in August and September 1948, to Mr G. P. Wells, Dr Bernhard Katz and Mr D. R. Newth, who read the paper and offered many valuable criticisms and suggestions, and to Dr C. F. A. Pantin, F.R.S., for the privilege of consulting him from time to time.