Spontaneous electrical activity recorded from solitary Tubularia solitaria Warren consisted of long bursts of potentials regularly alternating with series of short bursts. Hydranth pulse (HP) systems in isolated hydranths produce a pulse pattern similar to that in intact animals though at a much lower frequency. Stalk pulse (SP) systems in isolated stalks produced extremely regular bursts of pulses resembling bursting patterns in molluscan single neurones. The bursts were associated with strong bending movements. In contrast with hierarchies in other hydroids, T. solitaria’s HP system dominates the SP system.

Excised parts of the stalk produced bursting patterns. Interburst-and interpulse intervals were longer in these parts than in the whole stalk. Numbers of pulses per burst and overall pulse frequency were highest in the most distal part of the stalk, and lowest in the most proximal part.

Spontaneous, rhythmically occurring behavioural events are a characteristic feature of some hydroids. Examples are: the concerted oral movements of the tentacles preceding peristaltis of the proboscis in Tubularia (Josephson & Mackie, 1965); hydranth contraction in Obelia (Morin & Cooke, 1971); contraction of the body column and tentacles in Hydra (Passano & McCullough, 1963 ; Rushforth & Burke, 1971); and zooid contraction in the hydrocoral Millepora (de Kruijf, 1975, 1976a, b). In these hydrozoans the periodic movements are associated with electrical potentials, and are based on the electrical activity of one or more integrated pacemaker systems. The term ‘pacemaker’, as used here and defined by Josephson & Mackie (1965), refers to those elements which spontaneously and repeatedly (although not necessarily rhythmically) initiate measurable events such as electrical potentials or muscle contractions.

Josephson & Mackie (1965) showed the presence of pacemaker systems in colonial Tubularia and how they may interact. In subsequent studies their schematic representation was refined and extended (e.g. Josephson & Uhrich, 1969; Josephson & Rush-forth, 1973 ; Josephson, 1974a). There are two master pacemaker systems in colonial Tubularia polyps, the neck pulse (NP) system in the neck region and the hydranth pulse (HP) system in the more distal hydranth. Although there is much variability in pattern, the most common output from the NP system is a series of single electrical pulses (neck pulses) interrupted at intervals by bursts of pulses. This output pattern appears to be inherent to the NP system. Furthermore the neck pulses are not directly linked to any known overt behaviour (Josephson, 1974b, c). The HP system produces single pulses, short bursts of pulses, and long pulse bursts which coincide with the NP bursts. The loose hierarchy of the pacemaker systems most often seen in Tubularia is, in descending order : NP system to HP system to tentacles and gonophores. Ball (1973) presented a scheme of the interaction of the pacemakers and effector systems in Corymorpha palma which is a solitary relative of Tubularia. He showed that in Corymorpha there is an HP system quite similar to, if not homologous to, the HP system of Tubularia and a stalk pulse system (SP system) which is in many respects similar to Tubularia’s NP system. Some important differences between the NP system of Tubularia and the SP system of Corymorpha are: (1) The SP system is found throughout the stalk, and pacemaker activity may arise at any point in the stalk, whereas the NP system is limited to the neck region. (2) Stalks of Corymorpha contract either on the first pulse of a stalk burst or throughout the burst whereas Tubularia NP’s are not related to any overt behaviour. The NP pulses produced in hydranthless stolons show basically the same pattern as in the intact animal. (3) In Tubularia there is tight coupling of muscle groups and pacemakers in distal tentacles and gonophore peduncles whereas in Corymorpha there is no visible response associated with electrical activity in these structures. Furthermore, Ball (1973) suggests that there may be triggering of the SP system by the HP system, but this is not yet firmly established. The HP system can trigger the NP system as has been shown by Josephson and colleagues (Josephson, 1974c; Josephson & Rushforth, 1973).

The cellular basis of most types of electrical activity recorded from hydroids is unknown. Some potentials of non-nervous origin have been found in the Cordylophora stolon (Josephson, 1961) and in Hydra (Passano & McCullough, 1965). Because of the size of the potentials and the minute size of the nerves Josephson (1974b) argued that stalk pulses in Corymorpha (Ball & Case, 1973) may be of epithelial origin too. Neuronal conducting systems have been unequivocally demonstrated once in Hydrozoa, namely a pair of giant nerve fibres in the stem of the siphonophore, Nanomia (Mackie, 1973). Nerve as well as neuroid conduction was also clearly demonstrated in the anthomedusa Stomotoca (Mackie, 1975). However, as yet there is no test or technique available to decide whether potentials recorded from hydrozoans are of non-nervous, neuronal, or myoid origin (Spencer, 1974) although circumstantial evidence may suggest one or another.

In this paper I describe the presence of a bursting pacemaker system throughout the stalk of a solitary Tubularia sp., and some basic characteristics of this system. An explanation of pattern and pace in this system is also given.

Tubularia solitaria is a solitary hydroid reported rarely for the Americas. Some doubt about the name still exists although the specimens closely fit the description of Warren (1908). In the bay just before the Caribbean Marine Biological Institute it occurs from just below the water line down to about 3 m on piles, docks, and a shark net. All specimens observed were living in or on various species of encrusting sponges. They were collected once or twice a week and could easily survive for a week or more in tanks of running sea water. A piece of sponge containing several full-grown animals was transferred to a small experimental tank. All experiments were finished within 4 h. Only species of sponges were used which did not show excessive mucus secretion during transfer and in the experimental tank, and which did not grow over and cover the perisarc of the hydroid. The polyps used in all experiments were full-grown and usually had gonophore peduncles in many stages. The size of a fully extended polyp was up to 5–8 mm for the stolon and about 3–4 mm for the hydranth. The immobile basal part of the polyp was embedded in or lying on the sponge and its size varied from 2 to 10 mm. Electrical recording was done using suction electrodes made by drawing out flexible ‘Tygon’ tubing to produce a tip orifice of 40–100 μm. The in-different electrode was an Ag-AgCl wire in the bath. Electrical activity was amplified with an AC-amplifier, monitored on a Philips 5200 oscilloscope and recorded on a Brush and Gould recorder. Minimum time-course of the pulses was above 8 ms and could thus be easily followed by the recording system.

Two small incisions were made : one on the stalk near the neck and one in the basal or proximal part of the stolon. In the distal incision an electrode was carefully inserted and by gentle suction attached to the inner wall, the endoderm. Electrodes attached to the outside stolon wall, which is covered by a thin layer of chitin (the perisarc), were usually loosened within a few minutes, and recordings from the electrodes were very poor compared to recordings from the endoderm.

Comparison of the electrical activity recorded from outside and inside showed no difference with regard to the pattern and the firing rate. A small excision was made between the neck and the distal edge of the basal part. The damage done by the incisions could have altered the output patterns, but visual observations as well as recordings did not show any differences between these animals and undisturbed ones.

The spontaneously active pulse systems are named according to the part of the animal in which the pulses appear to originate, in the manner of Josephson & Mackie (1965). A stalk pulse system has not been described for Tubularia, but has been found in the solitary Corymorpha palma. As will be shown below, the solitary Tubularia used in this study has a stalk pulse system (SP system).

Electrical activity in intact animals

Recordings were made from the hydranth between the neck and the proximal tentacles, and from the endoderm of the stalk. Duration and amplitude of pulses recorded from the hydranth (hydranth pulses) and the stalk (stalk pulses) were similar to the corresponding pulses in Corymorpha palma (Ball, 1973). The pattern of activity in hydranth and stalk is typically a mixture of single pulses and long and short bursts of pulses (Figs. 1, 2a, b, 3a). Characteristics of the burst patterns are summarized in Table 1. The HP burst patterns recorded from the base of the hydranth could indicate that there is but one system (Fig. 2), but the HP burst patterns recorded from the area between the proximal and distal tentacles are not one to one with the stalk burst pattern (Fig. 1), thus showing the existence of two different but closely related systems.

Fig. 1.

Electrical activity in an intact animal, simultaneously recorded from the hydranth and the stalk. This animal showed bursts alternating with single pulses. Vertical scale, 0·5 mV. Time mark, 10 s.

Fig. 1.

Electrical activity in an intact animal, simultaneously recorded from the hydranth and the stalk. This animal showed bursts alternating with single pulses. Vertical scale, 0·5 mV. Time mark, 10 s.

Fig. 2.

Electrical activity in the intact solitary Tubularia and in isolated parts, (a, b) The activity in the intact animal, recorded from the hydranth and the stalk respectively, (c) The record of the hydranth with only half of the stalk. Note the increase in number of pulses in the small bursts, (d) The activity in the lower part of the stalk, (e) Activity in the hydranth; few pulses are recorded between the long bursts. (f) The bursting activity in the upper part of the stalk. Vertical scale, 0·5 mV (in (d), 0·2 mV). Time scale, 10 s.

Fig. 2.

Electrical activity in the intact solitary Tubularia and in isolated parts, (a, b) The activity in the intact animal, recorded from the hydranth and the stalk respectively, (c) The record of the hydranth with only half of the stalk. Note the increase in number of pulses in the small bursts, (d) The activity in the lower part of the stalk, (e) Activity in the hydranth; few pulses are recorded between the long bursts. (f) The bursting activity in the upper part of the stalk. Vertical scale, 0·5 mV (in (d), 0·2 mV). Time scale, 10 s.

Fig. 3.

Electrical activity in the intact solitary Tubularia (recorded from the stalk) and in isolated parts. The sequence of the excisions is different from that in Fig. 2. The firing rate in the isolated hydranth is markedly lower than in the whole animal (a). In (c) the electrical activity of the stalk is regular. Note the distribution of the interpulse-interval within the burst, (d) and (e) show the activity in the two halves of the same stalk. Vertical scale, 0·5 mV. Time scale, toe.

Fig. 3.

Electrical activity in the intact solitary Tubularia (recorded from the stalk) and in isolated parts. The sequence of the excisions is different from that in Fig. 2. The firing rate in the isolated hydranth is markedly lower than in the whole animal (a). In (c) the electrical activity of the stalk is regular. Note the distribution of the interpulse-interval within the burst, (d) and (e) show the activity in the two halves of the same stalk. Vertical scale, 0·5 mV. Time scale, toe.

The number of pulses in long bursts varies from animal to animal but is usually around 19 pulses per burst (Table 1). The interval between these long bursts varies widely from 87 to 253 s and within an individual there is much variation also (coefficients of variation range from 10 to 76%). The short bursts, including single pulses, consist of 1·7–3·6 pulses with little variation. The number of short bursts in the interval between long bursts is quite variable between the animals, but within individuals can be quite consistent.

Firing rates for stalk pulses in intact animals vary from 0·37 to 0·67/s (Table 1) as calculated from 30 min records of 11 animals.

The various parameters in Table 1 were tested for any significant relation with each other and with the firing rates. The single significant correlation (P < 0·01) was between the average number of bursts and the average interval length between long bursts. There is no significant relationship between the number of pulses in long bursts and both the average number of pulses in short bursts and the number of short bursts. This may be interpreted as an adjustment to a more or less constant firing rate of the HP-SP system.

Although there is quite some variation in burst length, interval length and firing rate, the general pattern is that of long bursts alternating with a variable number of very short bursts except for one animal which showed rather long short bursts only.

In terms of behaviour this pattern is seen as large movements of the whole animal during long bursts, alternated with short bending movements of the hydranth and the upper part of the stalk during short bursts.

Electrical activity in isolated hydranth and stalk

Excision of the hydranth or any part of the stalk disrupted the burst patterns completely, but after 5 to 10 min the pattern of electrical activity regained stability. However, the patterns in isolated parts may be very different from the original pattern in the intact animal.

In Fig. 2 the spontaneous electrical activity in the intact polyp is shown (2a, b), then the hydranth plus a part of the stalk was cut from the rest of the stalk (Fig. 2c, d), and finally the hydranth was separated from the upper part of the stalk (Fig. 2e, f).

All parts of the records in Fig. 2 are from one animal and taken after at least 10 min of adaptation after the excisions. Characteristics of the burst patterns are shown in Table 1. The stalk-hydranth section has an electrical activity quite similar to the intact animal except for the firing rate which is 15–25% higher in the mutilated polyps (compare Fig. 2c with 2a and b). The increase in firing rate is the final outcome of the increase in mean number of pulses in the short bursts (2·5 v. 3·1), a lowering of the duration between long bursts (148·5 v. 133·5) and a tendency, though not present in all animals, to produce more pulses per long burst (Table 1). But on the other hand, the number of short bursts in all experiments is lower in the excised stalk-hydranth part (17·8) than in the intact polyp (23·7). When the hydranth is separated from the rest of the stalk, the firing rate drops markedly in the excised hydranth (Figs. 2e, 3b). As can be seen in the figures, this is mainly due to a decrease in the number of short bursts and a decrease in the number of pulses in these bursts. The electrical activity in the proximal part, after the first excision, is completely different from the pulse patterns in the intact animal, the hydranth-stalk section and the single hydranth (Fig. 2d). The stalk pulses are produced in very regular bursts at a rate which is usually lower than in the stalk-hydranth section and, in general, also lower than the rate in the intact polyp (Table 1). The SP pattern in the distal stalk part after the second excision has essentially the same pattern as the proximal part : extremely regular bursts, but at a different rate (Fig. 2f). The firing rate (0·78) is higher than in any other intact or partly intact part of the polyp (Table 1). The number of pulses in the bursts is higher in the distal part of the stalk than in the proximal part.

In a second series of experiments the hydranth was excised from the entire stalk (Fig. 3). Activity in the separated stalk was very regular. Activity in the hydranth had a pattern similar to that in the intact animal, but at a very different rate. When the stalk was cut in half, a regular pattern of pulses was found in each half, but fewer, shorter bursts were found in the basal or proximal part than in the distal part.

There is one particular characteristic of the SP bursts which should be mentioned. The interpulse intervals tend first to shorten and then to lengthen again, as can be seen in Fig. 4, where interval duration is plotted against serial number within a burst. It can also be seen that the shorter the burst the longer the intervals.

Fig. 4.

The distribution of interval durations in bursts of 6 pulses (•), 7 pulses (▽), 8 pulses (▾) and 9 pulses (▄). Each line is based on 25 bursts of two or three animals.

Fig. 4.

The distribution of interval durations in bursts of 6 pulses (•), 7 pulses (▽), 8 pulses (▾) and 9 pulses (▄). Each line is based on 25 bursts of two or three animals.

Electrical activity in parts of the stalk

To locate the area of origin of the pacemaking activity, stalks were cut into three, four or five pieces. After each excision electrical activity was monitored from the parts. The pulse patterns in the separated parts regained stability after 5–10 minutes. The various parts were not of equal size and the excision was not exactly perpendicular to the longitudinal axis of the stalk (due to contraction and bending of the stalk as soon as it was touched by the scissors).

Examples of recordings from the intact stalk as well as from parts thereof are shown in Fig. 3, 5, 6 and 8. Several characteristics of the pulse patterns in these experiments are graphically summarized in Fig. 7. The pulse frequency varied from 0·31 to 0·60 in seven experiments (Fig. 7d). In one experiment the frequency was much higher (0·97) mainly due to a very high number of pulses per burst (S 7, Fig. 7a). The bursting pattern was maintained when the stalk was cut into two pieces, though the firing rate in each of the parts could be below as well as above the rate observed in the intact stalk (Fig. 7d). If the distal part was cut into two the firing rate was highest in the most distal part, i.e. the part nearest to the hydranth, compared to all other pieces of the stalk in seven experiments. In one experiment the pulse frequency in that most distal part was just below the highest frequency in the next part (S 5). The most proximal part of the polyp, i.e. the part farthest away from the hydranth, showed by far the lowest frequency in all experiments. It appeared not to be important for the frequency whether the cutting was done by cutting each piece into two (Figs. 3, 5, 6) or that the stalk was sliced from one end to the other (Fig. 8). There was a clear gradient from a very high pulse frequency in the most distal part (higher in most experiments than in the intact stalk) to the lowest frequency in the most proximal part (Fig. 7d). Rarely, if ever, was a frequency the arithmetic mean of the frequencies of the parts of which it was built.

Fig. 5.

Similar to Fig. 3 except for the number and sequence of the excisions, (a) Whole stalk; (b) about two-thirds of the stalk; (c) lowest part of the stalk; (d) part of the stalk nearest to the hydranth ; (e) centre part of the stalk. Note the nearly continuous firing in the uppermost part of the stalk (d) and the sequence of the interpulse intervals. Vertical scale, 0·2 mV (in (e), 0·5 mV). Time scale, 10 s.

Fig. 5.

Similar to Fig. 3 except for the number and sequence of the excisions, (a) Whole stalk; (b) about two-thirds of the stalk; (c) lowest part of the stalk; (d) part of the stalk nearest to the hydranth ; (e) centre part of the stalk. Note the nearly continuous firing in the uppermost part of the stalk (d) and the sequence of the interpulse intervals. Vertical scale, 0·2 mV (in (e), 0·5 mV). Time scale, 10 s.

Fig. 6.

Electrical activity in the stalk and parts of the stalk, (a) Whole stalk; (b) upper half of the stalk; (c) the lower half of the stalk; (d) uppermost part of the stalk of part (b) nearest to the hydranth; (e) lower part of (b); (f) upper part of (c);(g) most basal part of the stalk. Vertical bar indicates 0·5 mV, except in (d) and (f) where it indicates 0·2 mV. Time scale, 10 s.

Fig. 6.

Electrical activity in the stalk and parts of the stalk, (a) Whole stalk; (b) upper half of the stalk; (c) the lower half of the stalk; (d) uppermost part of the stalk of part (b) nearest to the hydranth; (e) lower part of (b); (f) upper part of (c);(g) most basal part of the stalk. Vertical bar indicates 0·5 mV, except in (d) and (f) where it indicates 0·2 mV. Time scale, 10 s.

Fig. 7.

Various characteristics of the burst patterns in the whole stalk and parts of the stalk. The value of the parameter for the whole stalk is indicated at the left side of each graph. The numbers on the horizontal axis indicate the place of that particular part in the stalk, from the part nearest to the hydranth (most distal, 1) to the most distant part (proximal, 2, 3, 4, or 5). The code for the experiments is indicated in (a), (a) Mean number of pulses per burst; (b) mean interval duration between bursts ; (c) mean shortest interval duration within bursts ; (d) number of pulses per second (pulse frequency).

Fig. 7.

Various characteristics of the burst patterns in the whole stalk and parts of the stalk. The value of the parameter for the whole stalk is indicated at the left side of each graph. The numbers on the horizontal axis indicate the place of that particular part in the stalk, from the part nearest to the hydranth (most distal, 1) to the most distant part (proximal, 2, 3, 4, or 5). The code for the experiments is indicated in (a), (a) Mean number of pulses per burst; (b) mean interval duration between bursts ; (c) mean shortest interval duration within bursts ; (d) number of pulses per second (pulse frequency).

Fig. 8.

Electrical activity in four pieces of the stalk cut in a different sequence to that in Fig. 6. (a) Intact animal ; (b) whole stalk ; (c) most distal part of the stalk; (d) part remaining after removal of (c); (e) next part of the stalk; (f) part remaining after removal of (e); (g) next part of the stalk; (h) part remaining after removal of (g); (i) next part of the stalk; (j) most basal part of the stalk in the sequence indicated. Continuous firing in (c) shows variation in interval length between the pulses. Vertical scale for (ad, f, h) 0·5 mV; for (e, g, i,j) 0·2 mV. Time scale, 10 s.

Fig. 8.

Electrical activity in four pieces of the stalk cut in a different sequence to that in Fig. 6. (a) Intact animal ; (b) whole stalk ; (c) most distal part of the stalk; (d) part remaining after removal of (c); (e) next part of the stalk; (f) part remaining after removal of (e); (g) next part of the stalk; (h) part remaining after removal of (g); (i) next part of the stalk; (j) most basal part of the stalk in the sequence indicated. Continuous firing in (c) shows variation in interval length between the pulses. Vertical scale for (ad, f, h) 0·5 mV; for (e, g, i,j) 0·2 mV. Time scale, 10 s.

In four experiments the most distal part fired continuously so that no interburst intervals could be measured. The pattern of this continuous firing was cyclical, as can be observed in Figs. 3d and 8c. The intervals between the bursts in the other small parts of the stalk did not show any consistent pattern except that they were usually higher than the intervals in the whole stalks (Fig. 7b). However, the number of pulses per burst showed a gradient similar to the firing rate (Fig. 7a). The lowest number of pulses was found in the bursts produced by the most proximal parts of the stalk, the highest in the most distal parts.

The shortest intervals within bursts were measured for the whole stalk as well as for the parts (Fig. 7c). These intervals were smallest in whole stalk preparations except S 8. In addition there was a slight tendency for longer interval durations in more proximal parts.

Each of the parts of the stalk has the capability to spontaneously produce pulses in definite patterns. Some of the characteristics of the pulse pattern showed a consistent pattern in their distribution in the stalk parts.

The solitary Tubularia, the colonial Tubularia and the solitary Corymorpha

The pattern of electrical activity of the intact Tubularia solitaria is very similar to the pulse pattern found in a colonial Tubularia (Josephson, 1965, 1974c). As would be expected due to the systematic relationship, the activity also closely resembles that found in the solitary Corymorphapalma (Ball, 1973). The general pattern found in each of these species is that of conducted movements accompanied by long bursts of pulses alternated with short bursts or single pulses related to movements involving parts of the polyp. The stalk pulse system described here for a solitary Tubularia is apparently homologous to that in Corymorpha. It seems likely that it is equivalent to the neck pulse system of colonial Tubularia, but extended so that pacemaker activity occurs over a longer length of stalk

The results indicate that in the solitary Tubularia there is a tight coupling between the muscle groups and pacemakers in distal tentacles and gonophore peduncles, as in colonial Tubularia but unlike in Corymorpha. Neck pulses in colonial Tubularia are not directly related to known overt behaviour: they are cryptic (Josephson, 1974c). In contrast, stalk pulses in Corymorpha almost always occur simultaneously with contraction of the longitudinal muscles of the stalk, although one or two pulses sometimes occur without any apparent behavioural correlation (Ball, 1973). Much the same situation was found in the solitary Tubularia. The relation between stalk pulses in bursts and contraction of longitudinal muscles was clearly demonstrated in hydranthless preparations. Each pulse preceded one contraction of muscles with a delay long enough to be easily seen by the human observer.

The typical pattern of single pulses and bursts that is inherent to the NP system can be recorded from stalks from which the hydranth is removed in colonial Tubularia (Josephson, 1974c). This procedure isolated the NP system from all other known pacemaker systems. Along with other arguments the results indicated a dominance of the NP system over the HP system. In similar experiments with Corymorpha palma, it has been impossible to attribute many changes unequivocally to excision or separation of pacemaker systems (Ball, 1973 ; Ball & Case, 1973). The various experiments on Corymorpha palma led Ball to the conclusion that the usual triggering path is that of the SP system towards the HP system and that the HP system less commonly triggers the SP system. The pattern of single pulses and small bursts interrupted by longer bursts is inherent to the HP system in the solitary Tubularia as is shown in the experiments. This pattern, however, is also found in the intact animal, thus indicating that the HP system in the polyp is the dominant system. The difference in firing rate between the excised hydranth and the intact animal – the latter firing at a much faster rate – further indicates that the SP system in the solitary Tubularia is responsible for the rate at which the pacemaker activity is produced but not for its pattern. Therefore the HP system may be considered to be dominant over the SP system in contrast with the pacemaker hierarchies in the colonial Tubularia and Corymorpha.

Differences in the pattern of electrical activity observed in this study and those observed in other animals (with the exception of a study of Corymorpha by Ball & Case (1973), employing isolated stalks) could have been due to the use of a different recording technique ; cutting the animal up and recording from the inside or the edge of the stalk instead of the outside. This is not the case, however, for recordings from the outside were essentially the same as those reported here, the only difference being a much poorer spike-to-noise ratio.

Coordination in the stalk pacemaker system

The stalk pulse system in Tubularia solitaria apparently fits the criteria of Josephson & Mackie (1965) of being a pacemaker system. The most prominent characteristic of this system is the regularity of the bursting pattern of stalk pulses as indicated by small variations in interburst interval duration, by consistent patterns in the spacing of the interpulse intervals, by small variations in number of pulses per burst in any given piece of the stalk, and simply by visual examination.

Horridge (1959) showed that segments containing a marginal ganglion from the scyphomedusa Aurelia give swimming pulsations with approximately normal activity patterns: Needier & Ross (1958) observed normal activity patterns in rings from the column of sea anemones although at a different rate; Hofman & Rushforth (1967) found that pieces of tentacles of a colonial Tubularia produce electrical pulses in a pattern similar to that of a whole tentacle. The surgical experiments with the stalk of Tubularia solitaria revealed that small parts of the stalk produce pulses in patterns essentially similar to that of the whole stalk. The patterns are produced at different rates but these differences are quite consistent in that the most distal part has the highest pulse frequency, the most proximal part the lowest frequency. The number of pulses per burst also showed a similar gradient (Fig. 7).

From these experiments it is concluded that the stalk pulse system must contain a number of potential pacemaker elements and, of course, a conduction system to connect the elements. The number of elements as well as the location of these elements are unknown except that they are present all over the stalk. These particulars agree with those found in other coelenterates with the exception of the gradient which to my knowledge has not been shown yet in another coelenterate.

Horridge (1959) points out that the output pattern from a set of coupled pace-makers, each with similar properties like the marginal ganglia in Aurelia will be more regular and the mean frequency somewhat faster than the output of any one pacemaker considered alone. This is because the rhythm becomes dominated by short intervals from any of the pacemakers, and the long intervals produced by an isolated ganglion do not appear in the output of the coupled group. Lerner et al. (1971) found, however, that while activity in whole animals is usually more regular than that of isolated ganglia, segments of Aurelia with a single ganglion can beat for periods as rapidly and regularly as intact animals. They argue that the regularity of whole animals may result both from coupling of ganglia and from more regularity in individual ganglia due to increased excitatory input. The spontaneous patterns of the intact Tubularia are certainly more complex than the Aurelia swimming pulsations But in the stalk of the solitary Tubularia there is a simple regular pattern clearly being produced by one pacemaker system.

A review of the essential characteristics which define the bursting pattern of the whole stalk as well as the parts shows that: (1) the interburst intervals in the isolated parts vary without any specific pattern but the intervals in the whole preparations are shorter in nearly all experiments; (2) the interpulse intervals also vary within the parts without a specific pattern but here too the whole stalk preparation produces pulses in a burst faster than in any of the parts; (3) the number of pulses per burst shows a gradient whereas the whole stalk produces bursts with a number of pulses somewhere in between the two extremes of the isolated parts. The burst frequency follows closely the prediction of Horridge (1959) and follows the pulse frequencies within bursts. As the pulse production follows another course I expect that a different mechanism, in addition to the integration of pacemakers in the way Horridge suggested, must be involved.

The initiation of bursts

Bursting pacemakers whose activity is similar in gross wave form and time-course to the events recorded under some conditions in Tubularia are found in some other organisms, e.g. crustacean motoneurone preparations (Gillory & Kennedy, 1969; Davis, 1969) and in particular a number of molluscan neurones. Endogenous bursting pacemakers were found in Aplysia californica (Strumwasser, 1967, 1973; Barker & Gainer, 1975), in Tritonia (Dorsett, 1974), in Helisoma (Kater & Kaneko, 1972), and others. Explanation of these bursting patterns involves an oscillating membrane potential which underlies alternating periods of action potential bursts and interburst hyperpolarizations (Strumwasser, 1967, 1973; Barker & Gainer, 1975). A similar oscillating membrane potential as well as some other mechanism, e.g. postinhibitory rebound (Perkel & Mulloney, 1974), may underlie the Tubularia bursting pattern. The pattern would require the presence of two mechanisms, of different oscillatory period, one in each of two conducting systems. Two mechanisms appear present in the stalk of Corymorpha, one being neuronal and the other apparently epithelial (Ball & Case, 1973).

I wish to thank Dr Robert K. Josephson for critical reading of the manuscript and Dr Norman B. Rushforth for comments. Dr W. Vervoort, Leiden, kindly identified the hydroid. I am also grateful to the anonymous referee whose comments improved this manuscript.

Ball
,
E. E.
(
1973
).
Electrical activity and behaviour in the solitary hydroid Corymorpha palma. I. Spontaneous activity in whole animals and isolated parts
.
Biol. Bull. mar. biol. Lab., Woods Hole
145
,
233
43
.
Ball
,
E. E.
&
Case
,
J. F.
(
1973
).
Electrical activity and behaviour in the solitary hydroid, Corymorpha palma. II. Conducting systems
.
Biol. Bull. mar. biol. Lab., Woods Hole
148
,
243
64
.
Barker
,
J. L.
&
Gainer
,
H.
(
1975
).
Studies on bursting pacemaker potential activity in molluscan neurons. I. Membrane properties and ionic contributions
.
Brain Res
.
84
,
461
77
.
Davis
,
W. J.
(
1969
).
The neural control of swimmeret beating in the lobster
.
J. exp. Biol
.
50
,
99
117
.
Dorsett
,
D. A.
(
1974
).
Neuronal homologies and the control of branchial tuft movements in two species of Tritonia
.
J. exp. Biol
.
61
,
639
54
.
Gillory
,
H. L.
&
Kennedy
,
D.
(
1969
).
Pattern generation in a crustacean motoneuron
.
J. Neurophysiol
.
32
,
595
606
.
Hoeman
,
F.
&
Rushforth
,
N. B.
(
1967
).
Electrical activity of isolated parts of Tubularia
.
Biol. Bull, mar. biol. Lab., Woods Hole
133
,
469
.
Horridce
,
G. A.
(
1959
).
The nerves and muscles of medusae. VI. The rhythm
.
J. exp. Biol
.
36
,
72
91
.
Josephson
,
R. K.
(
1961
).
Repetitive potentials following brief electric stimuli in a hydroid
.
J. exp. Biol
.
38
.
579
93
.
Josephson
,
R. K.
(
1965
).
Three parallel conducting systems in the stalk of a hydroid
.
J. exp. Biol
.
42
,
139
53
.
Josephson
,
R. K.
(
1974a
).
Factors affecting muscle activation in the hydroid Tubularia
.
Biol. Bull, mar. biol. Lab., Woods Hole
147
,
594
607
.
Josephson
,
R. K.
(
1974b
).
Cnidarian neurobiology
.
In Coelenterate Biology: Reviews and Perspectives
(ed.
L.
Muscatine
and
H. M.
Lenhoff
), pp.
245
80
.
New York
:
Academic Press
.
Josephson
,
R. K.
(
1974c
).
The strategies of behavioral control in a coelenterate
.
Am. Zool
.
14
,
905
15
.
Josephson
,
R. K.
&
Mackie
,
G. O.
(
1965
).
Multiple pacemakers and the behaviour of the hydroid Tubularia
.
J. exp. Biol
.
43
,
293
32
.
Josephson
,
R. K.
&
Rushforth
,
N. B.
(
1973
).
The time course of pacemaker inhibition in the hydroid Tubularia
.
J. exp. Biol
.
59
,
305
14
.
Josephson
,
R. K.
&
Uhrich
,
J.
(
1969
).
Inhibition of pacemaker systems in the hydroid Tubularia
.
J. exp. Biol
.
50
,
1
14
.
Kater
,
S. B.
&
Kaneko
,
C. R. S.
(
1972
).
An endogenously bursting neuron in the gastropod mollusc, Helisoma trivolvis. Characterization in vivo
.
J. comp. Physiol
.
79
,
1
14
.
Kruijf
,
H. A. M. De
(
1975
).
General morphology and behaviour of gastrozoids and dactylozoids in two species of Millepora (Milleporina, Coelenterata)
.
Mar. Behav. Physiol
.
3
,
181
92
.
Kruijf
,
H. A. M. De
(
1976a
).
Spontaneous electrical activity and colonial organization in the hydrocoral Millepora (Milleporina, Coelenterata)
.
Mar. Behav. Physiol
.
4
,
137
159
.
Kruijf
,
H. A. M. DE
(
1976b
).
The effect of electrical stimulation and some drugs on conducting systems of the hydrocoral Millepora complana ta
.
In Coelenterate Ecology and Behaviour
(ed.
G. O.
Mackie
), pp.
661
70
.
New York
:
Plenum
.
Lerner
,
J.
,
Mellen
,
S. A.
,
Waldron
,
I.
&
Factor
,
R. M.
(
1971
).
Neural redundancy and regularity of swimming beats in scyphozoan medusae
.
J. exp. Biol
.
55
,
177
84
.
Mackie
,
G. O.
(
1973
).
Report on giant nerve fibres in Nanomia
.
Publ. Seto Mar. Biol. Lab
.
20
,
745
56
.
Mackie
,
G. O.
(
1975
).
Neurobiology of Stomotoca. II. Pacemakers and conduction pathways
.
J. Neurobiol
.
6
,
357
78
.
Morin
,
J. G.
&
Cooke
,
I. M.
(
1971
).
Behavioural physiology of the colonial hydroid Obelia. I. Spontaneous movements and correlated electrical activity
.
J. exp. Biol
.
54
,
689
706
.
Needier
,
M.
&
Ross
,
D. M.
(
1958
).
Neuromuscular activity in sea anemone Calliactis parasitica (Conch)
.
J. mar. Biol. Ass. U.K
37
,
789
805
.
Paesano
,
L. M.
&
Mccullough
,
C. B.
(
1963
).
Pacemaker hierarchies controlling the behaviour of Hydra
.
Nature, Lond
.
199
,
1174
5
.
Paesano
,
L. M.
&
Mccullough
,
C. B.
, (
1965
).
Coordinating systems and behaviour in Hydra. II. The rhythmic potential system
.
J. exp. Biol
.
42
,
205
31
.
Perkel
,
D. H.
&
Mulloney
,
B.
(
1974
).
Motor pattern prediction in reciprocally inhibitory neurons exhibiting post inhibitory rebound
.
Science, N.Y
.
185
,
181
3
.
Rushforth
,
N. B.
&
Burke
,
D. S.
(
1971
).
Behavioral and electrophysiological studies of Hydra. II. Pacemaker activity of isolated tentacles
.
Biol. Bull. mar. biol. Lab., Woods Hole
140
,
503
19
.
Spencer
,
A. N.
(
1974
).
Non-nervous conduction in invertebrates and embryos
.
Am. Zool
.
14
,
917
29
.
Strumwasser
,
F.
(
1967
).
Types of information stored in single neurons
.
In Invertebrate Nervous Systems
(ed.
C. A. G.
Wiersma
), pp.
291
319
.
Chicago
:
University of Chicago Press
.
Strumwasser
,
F.
(
1973
).
Neural and humoral factors in the temporal organization of behaviour
.
Physiologist
16
,
9
42
.
Warren
,
E.
(
1908
).
On Tubularia solitaria sp.nov., a hydroid from the Natal Coast
.
Ann. Natal Mus
.
1
,
83
96
.