1. A single shock applied through wick electrodes to the isolated radial nerve cord of a sea urchin produces a recordable potential in the cord. The potential is conducted along the cord at a velocity of between 14 and 20 cm./sec.

  2. The potential is complex and graded. Two components of the potential can be identified and have different thresholds to stimulation, conduction velocities and amplitudes. They are believed to represent two classes of fibres.

  3. The potential is conducted decrementally along the cord and normally cannot be recorded at distances greater than 60 mm. from the stimulus. The amplitude of the potential decays logarithmically falling to half after 7 mm. spread. There is no facilitation of amplitude or distance of spread.

  4. Potentials initiated simultaneously at either end of the isolated nerve cord collide and partially occlude each other.

  5. Stimulation of a side branch of the nerve cord evokes potentials recordable from only ipsilateral neighbouring side branches and the whole cord. However, contractions of the contralateral ampullae following stimulation of lateral branches reveal spread of the excitation beyond the region of recordable potentials.

  6. A single shock to a cord still attached to the test causes contraction of the associated ampullae. One ampulla will contract several times after a single shock, a period of relaxation following each contraction.

  7. Electrical activity recorded from the ampullae, and lasting many seconds after the single shock, corresponds with their contractions. The activity is believed to be muscle action potentials.

  8. Evidence of a feedback from damaged tube feet to the cord, suppressing ampulla response to cord stimulation, was found.

Electrical activity in the radial nerve cord of the sea urchin Diadema setosum has recently been reported by Takahashi (1964), who recorded nerve impulses following direct photic stimulation of the isolated nerve cord. No other electrical activity has been recorded from sea-urchin nervous systems. The following is a report on electrical phenomena in the isolated radial nerve cord and ampullae of sea urchins, induced by electrical stimulation.

The experiments were performed mainly on Strongylocentrotus franciscanus in Los Angeles; similar phenomena were observed in Tripneustes sp. and Toxopneustes sp. studied in Lower California, near La Paz. The test sizes of the experimental animals ranged from 7 to 15 cm. in diameter.

The radial nerve cord was stripped out of the test after cutting its lateral branches and removing the overlying radial water-vascular canal. The isolated cord was then supported across recording and stimulating wick electrodes. To record from an ampulla, a portion of the test bearing the radial nerve cord and the water-vascular canal was separated from the animal. A lobe of an ampulla was supported on one recording electrode (the second recording electrode was grounded with the preparation), and a short length of the radial cord, freed from the test, was placed on the stimulating electrodes. Recordings from the side branches of the nerve cord were made using similar relatively intact preparations. The lateral branches of the nerve cord were cut peripherally and dissected away from the test ; one was then lifted out of the grounded solution covering the radial nerve cord, and placed over stimulating or recording electrodes.

The stimulus was provided by a Grass SD 5 stimulator and the electrical impulses from the nerve cord were recorded with conventional AC-coupled amplifiers.

(1) The radial nerve cord

A single shock produces a complex potential in the nerve cord after a latency which is proportional to the distance between the stimulating and recording electrodes. The conduction velocity of the initial deflexion, measured with two recording electrodes along the length of the nerve cord, is between 14 and 20 cm./sec. and the duration of the entire potential produced by maximal stimulus is of the order of 200 msec. Stimulus shocks are more effective if applied with the cathode proximal to the recording site. The reverse polarity causes a longer latency and a decrease in the amplitude of the response. Crushing the tissue between the stimulating and recording electrodes abolishes the response.

Alteration of the intensity or duration of the single stimulus shock alters the amplitude of the response (Fig. 1). Two main components of the complex potential can be recognized if the stimulus intensity or duration is gradually increased from a sub-threshold to a maximal value. The amplitude of the first component increases evenly and reaches its maximum when a 2 V., 2 msec, stimulus pulse is applied. At this point, marked by an irregularity in the curve of Fig. 2, the second and slower component appears and soon masks the smaller initial potential. The second component increases evenly from this point with the increase in the intensity of the stimulus until its maximum amplitude is reached. The area beneath the potentials may be regarded as a measure of their size though it does not distinguish between the first and second components and may exaggerate the irregularity of the transition in Fig. 2.

Fig. 1.

The graded properties of the response. Superimposed oscillograph traces show the stepped change in the amplitude of the response produced by a stepped alteration of the intensity and duration of the stimulus. In A the intensity was changed from 1 to 8 V., the duration being 2 msec. In B the duration of the shock ranged from 2 to 16 msec., the intensity being maintained at 1 V. The complex nature of the potential is apparent in both A and B. The records also suggest an equivalence of effect in the increase of the stimulus intensity and duration within the ranges tested.

Fig. 1.

The graded properties of the response. Superimposed oscillograph traces show the stepped change in the amplitude of the response produced by a stepped alteration of the intensity and duration of the stimulus. In A the intensity was changed from 1 to 8 V., the duration being 2 msec. In B the duration of the shock ranged from 2 to 16 msec., the intensity being maintained at 1 V. The complex nature of the potential is apparent in both A and B. The records also suggest an equivalence of effect in the increase of the stimulus intensity and duration within the ranges tested.

Fig. 2.

The development of the two components of the potential. The area under the potential was taken as a measure of its size and has been plotted on the ordinate against the intensity of stimulation on the abscissa. The duration of the stimulus was maintained at 2 msec. The size of the first component increases evenly to point A. It then becomes masked by the large second component, the development of which extends smoothly from B to C. The irregularity of the transition between the first and second components (A to B) is probably exaggerated by the method of estimating the size of the response.

Fig. 2.

The development of the two components of the potential. The area under the potential was taken as a measure of its size and has been plotted on the ordinate against the intensity of stimulation on the abscissa. The duration of the stimulus was maintained at 2 msec. The size of the first component increases evenly to point A. It then becomes masked by the large second component, the development of which extends smoothly from B to C. The irregularity of the transition between the first and second components (A to B) is probably exaggerated by the method of estimating the size of the response.

Decremental spread of the potential is revealed by two recording electrodes placed at different distances from the point of stimulation. Potentials lose amplitude over a few millimetres and no response can be recorded 5 or 6 cm. from the stimulating electrodes. The nature of the decrement was determined by sampling the isolated cord with four pairs of recording electrodes placed at 5 mm. intervals (Fig. 3). The decay of the potential is almost purely logarithmic, falling to half amplitude in 7 mm. Potentials travelling in the oral-aboral direction show the same loss of amplitude as those travelling in the aboral-oral direction.

Fig. 3.

Decremental spread of the potential. Recordings from four places along the isolated nerve cord. The top trace shows the potential recorded by the electrode furthest from the stimulus. The two parts of the potential become separated as they progress along the cord, indicating the difference in their conduction velocities. The more distal electrodes record irregularities in the potential not apparent in the potential recorded near the stimulus. Stimulus : single shock of 8 V. and 2 msec, duration.

Fig. 3.

Decremental spread of the potential. Recordings from four places along the isolated nerve cord. The top trace shows the potential recorded by the electrode furthest from the stimulus. The two parts of the potential become separated as they progress along the cord, indicating the difference in their conduction velocities. The more distal electrodes record irregularities in the potential not apparent in the potential recorded near the stimulus. Stimulus : single shock of 8 V. and 2 msec, duration.

Pairs or bursts of stimuli produce no facilitation of the amplitude or distance of spread of the wave. A relative refractory period of 400 msec, follows each wave. During this time stimulus shocks evoke potentials ranging from zero amplitude immediately after the initial wave to full amplitude after 400 msec.

The nerve cord can be split lengthwise and each half of the nerve cord responds to stimulus intensities of the same order as the whole cord. The same pattern of decremental spread of the potential prevails (Fig. 4). Sometimes the amplitude of the potential from half the cord was greater and more irregular than that of the whole cord, due possibly to the relatively smaller tissue shunt between the recording electrodes. In the whole cord the potential form recorded at a proximal electrode is often smooth and shows little complexity, whereas the potential at a distal electrode is usually irregular, and the first and second stages of the response become more widely separated. The two components of the response could not be separated by splitting the cord. Transverse cuts in the nerve cord show that the entire recordable response is confined to one of the several tracts visible in unfixed preparations of the isolated cord. If the active tract alone is sectioned no response can be evoked from the cord.

Fig. 4.

Decremental spread of the potential. The normalized height of the potential is plotted on the ordinate against the distance of spread on the abscissa. Each point represents the average of nine preparations and the amount of scatter is indicated. Values obtained from half the cord are no different from those of the whole cord. The linearity of this plot indicates the almost perfect logarithmic decay of the potential.

Fig. 4.

Decremental spread of the potential. The normalized height of the potential is plotted on the ordinate against the distance of spread on the abscissa. Each point represents the average of nine preparations and the amount of scatter is indicated. Values obtained from half the cord are no different from those of the whole cord. The linearity of this plot indicates the almost perfect logarithmic decay of the potential.

Mechanical and photic stimulation of the spines and tube feet of relatively intact preparations produce no recordable electrical discharge in the radial nerve cord. No potential change was recorded during direct photic stimulation of the isolated nerve cord of either the shadow-sensitive or non-shadow-sensitive species of sea urchin.

Collision of the potentials was brought about by stimulating both ends of the cord simultaneously. The recording electrodes were placed in the centre. The stimuli were timed so that the effects of the collision could be observed at the recording electrode (Fig. 5). Partial occlusion of the one wave by the other was observed. There was no greater occlusion of the potentials from the aboral pole by those from the oral pole or vice versa. In one experiment the nerve cord was transected through half its width in two places, the cuts being made from opposite sides so that both longitudinal halves of the cord would be interrupted. The preparation was arranged so that the cuts lay on opposite sides of the recording electrodes. Impulses initiated at either end of the cord now pass each other and the central recording electrode without losing amplitude by occlusion, suggesting that the waves of excitation are confined to their respective longitudinal halves of the cord. Occlusion here would have indicated a lateral spread of the potentials.

Fig. 5.

The collision of the potentials travelling in opposite directions in the isolated nerve cord. A, the potential initiated at the oral end of the preparation and recorded in the centre. B, The potential initiated at the aboral end. C, Simultaneous stimulation of both ends of the preparation timed so that the collision occurs at the recording electrodes. The amplitude of the potential is diminished but not entirely occluded. Stimulus: single shock of maximal intensity.

Fig. 5.

The collision of the potentials travelling in opposite directions in the isolated nerve cord. A, the potential initiated at the oral end of the preparation and recorded in the centre. B, The potential initiated at the aboral end. C, Simultaneous stimulation of both ends of the preparation timed so that the collision occurs at the recording electrodes. The amplitude of the potential is diminished but not entirely occluded. Stimulus: single shock of maximal intensity.

Central connexions between the adjacent lateral branches of the nerve cord can be shown if the central end of a peripherally cut lateral branch is stimulated. A recordable potential is evoked only in the neighbouring ipsilateral branches. Peripheral pathways connecting adjacent lateral branches of the same side could not be shown, nor central connexions across the cord between lateral branches. However, a stimulus applied to the whole cord elicits activity in all the lateral branches on both sides and similarly a stimulus to any lateral branch produces a small potential in the whole cord.

(2) The ampullae

Discrete contractions of the ampullae are produced by a single shock to the oral or aboral end of the radial nerve cord of relatively intact preparations. The ampullae nearest the stimulating electrodes contract first but those further from the stimulus source contract with the same apparent intensity. Latencies are of the order of 1 sec. Up to four or five separate contractions of one ampulla can occur after a single shock to the nerve cord, each contraction in the series lasting for approximately 6 sec. and separated from the subsequent contraction by a relaxation period of 1 or 2 secs. A short burst of shocks which are of subthreshold intensity if applied singly, causes the ampulla to contract. Facilitation lasts for about 5 sec., during which time a single, previously subthreshold, shock causes contraction.

Stimulation of half of a partially split nerve cord produces contractions of the ampullae on both sides of the cord beyond the split. This is unchanged by cutting between the rows of tube feet on the outside of the test to destroy any peripheral connexions. Similarly, stimulation of the central end of a lateral branch causes contractions of the ampullae on both sides of the cord and not only those innervated by the adjacent lateral branches. This contradicts the results obtained with electrical recording, and shows a spread of the effect of stimulation beyond the region of recordable potentials. Contractions of the ampullae normally brought about by electrical stimulation of the cord are prevented by the removal of the whole, or a portion of, the associated tube feet. However, these ampullae will still contract if mechanically stimulated.

A burst of electrical activity can be recorded from the lobe of an ampulla during its contraction (Fig. 6). The duration of the burst corresponds precisely with the contractions of the ampulla. The potentials within the burst are of various amplitudes and are difficult to resolve into units ; however, single small deflexions occur irregularly during the periods of relaxation when no movement of the ampulla is apparent. These single waves have a duration of approximately 100 msec, and vary in amplitude from 70 to 95 μV. The correspondence between ampulla contraction and the electrical discharge demonstrates in Fig. 6B the facilitating effect of a burst of shocks which are subthreshold if applied singly. Simultaneous records from the ampulla and radial nerve cord (Fig. 6C) show the latency for the ampulla contraction to be very much longer than that of the potentials in the nerve cord (i.e. 1 sec. compared with 30 msec.). Artifacts deliberately caused by relatively violent movements of the electrode supporting the ampulla lobe were characteristically different and small. Stimulation of the nerve cord of intact animals will cause movement of the lantern and occasionally synchronized contractions of the ampullae in a neighbouring ambulacrum. The latencies for these reactions are of the order of 1 sec.

Fig. 6.

Action potentials from the lobe of an ampulla. A, Activity before the application of the stimulus, and (lower trace) after a single maximal stimulus to the radial nerve cord. The first two contractions of a series are shown, separated by a period of relaxation. B, Facilitation of the ampulla contraction. The first trace shows the result of applying a single subthreshold shock to the radial nerve cord. A burst of shocks (lower trace) of the same intensity and duration produces the contraction. Stimulus: 3 V.,1 msec, duration. C, Simultaneous recordings from the radial nerve cord (upper trace) and the ampulla (lower trace). The electrode recording from the ampulla was approximately 5 mm. from the stimulus site and the electrode on the nerve cord 15 mm. away from the stimulus. The amplification of the upper trace is less than that of the lower trace. Stimulus: a single maximal shock.

Fig. 6.

Action potentials from the lobe of an ampulla. A, Activity before the application of the stimulus, and (lower trace) after a single maximal stimulus to the radial nerve cord. The first two contractions of a series are shown, separated by a period of relaxation. B, Facilitation of the ampulla contraction. The first trace shows the result of applying a single subthreshold shock to the radial nerve cord. A burst of shocks (lower trace) of the same intensity and duration produces the contraction. Stimulus: 3 V.,1 msec, duration. C, Simultaneous recordings from the radial nerve cord (upper trace) and the ampulla (lower trace). The electrode recording from the ampulla was approximately 5 mm. from the stimulus site and the electrode on the nerve cord 15 mm. away from the stimulus. The amplification of the upper trace is less than that of the lower trace. Stimulus: a single maximal shock.

The responses in the radial nerve cord described here are different from those reported by Takahashi (1964) whose recordings with micro-electrodes show discrete nerve spikes of an all-or-none kind. The potentials caused by electrical stimulation are compound, graded and long-lasting.

Two main possibilities as to the origin of the responses caused by electrical stimulation may be examined. The first, that they are due to muscular activity, is tentatively excluded because not even microscopically observable contraction accompanied the most vigorous stimulation of the nerve cord. Besides, there is no anatomical evidence of muscular tissue associated with the nerve cord in sea urchins (Laverack, personal communication). The second, that the activity is the summed electrical potential change of a number of small nerve fibres, is supported by some evidence. Movements of the ampullae and lantern follow electrical stimulation of the nerve cord after a respectable latency. These contractions must be the result of activity initiated in the nerve cord by the single shock and transmitted to the site of action as nerve impulses. Anatomically the cord is known to consist of large numbers of fibres of less than i/i in diameter (Hamann, 1887, Laverack, 1965). The primarily complex potential is made more uneven by thinning down the nerve cord or recording from a greater distance from the site of stimulation. These results are consistent with the hypothesis that the potential is caused by nervous activity, for in both cases the number of active fibres from which the recordings are being taken is less and thus less smoothing should occur.

The presence of two major components in the complex potential in the nerve cord is suggestive of conducting elements of two kinds having different latencies, conduction velocities and thresholds. Attempts to separate the components by splitting the nerve cord were unsuccessful and revealed only that both components must be confined to a single ‘tract’. It may be assumed that the two types of element are equally distributed throughout this tract. Attempts to stain this and other tracts histologically have so far been unsuccessful.

Decremental spread of the potential along the nerve cord attests to the absence of a through-conducting system. The decrement is the same in both directions, and it is concluded that the numbers of fibres in the portion of nerve cord from which the recordings were taken did not decrease considerably towards the aboral pole, in spite of the many side branches extending from the cord. The collision experiments also support this view.

The distance over which the potential spreads in the isolated nerve cord is unchanged by applying multiple shocks. This absence of facilitation in the nerve cord contrasts with the clear facilitation of the ampulla contractions. Facilitation may be taking place within the cord but in relatively few fibres, in which case the effect would not be recorded. The logarithmic decay of the potential suggests a purely electrotonic spread of the excitation along the nerve but the marked increase in latency with distance of spread indicates the contrary. More information on the electrical activity of single units within the cord is needed before predictions can be made about the nature of possible synaptic junctions and the transmission of nervous excitation in the cord.

The collision experiments, in showing partial occlusion of one deflexion by another from the opposite end of the preparation, suggest the existence of pathways capable of carrying information in both directions along the same route. Separate pathways either not extending the full length, or incapable of conducting impulses in both directions, could explain the incomplete occlusion at the collision point. A cord consisting of uniformly staggered short fibres would fit the experimental results.

Central or peripheral pathways connecting the lateral branches of opposite sides of the radial nerve cord could not be demonstrated by electrical stimulation and recording. That central transverse connexions do exist, however, is clearly shown by the contractions of ampullae following electrical stimulation of the contralateral side branches. The failure of the electrical recording technique to show these connexions may be due to the nervous pathways being too fine to conduct enough impulses to give a recordable response.

The potentials recorded from the ampullae are deemed to be muscle action potentials. They are slow (100 msec.) and, when recorded singly, are discrete potentials. They are always accompanied by contractions of the ampulla, whether the contractions are brought about by electrical or locally applied mechanical stimuli. The possibility that these impulses are from sensory cells in the walls of the ampulla, signalling deformation of the tissue, may be excluded because gentle stretching of the ampulla produces no discharge. Comparison of ampulla contraction with the potential in the nerve cord caused by the same shock reveals a large difference in their latencies. The relation between the two phenomena is difficult to establish, for the repetitive electrical discharge associated with the ampulla contractions is long-lasting and shows cyclic activity after all traces of recordable electrical activity have ceased in the radial nerve cord. The potential in the cord may trigger a local reaction which can be detected only in the region of the cord adjacent to the active ampulla. The ampulla contractions are undoubtedly caused by nervous excitation in the cord, but whether this excitation is represented by the complex potential cannot be definitely shown.

The ampulla contractions normally produced by stimulation of the nerve cord are prevented by damaging the associated tube foot but the contractile ability of the ampulla is not affected, as shown by its immediate response to a locally applied mechanical stimulus. This may suggest that the necessary nervous pathways to the ampulla run from the nerve cord via the tube foot, and the illustrations of Hamann (1887) would support this view. However, in the starfish (Smith, 1946) a three-neuron arc links the sensory-motor system of the tube foot, ampulla and central nervous system. If a neural loop of this kind exists in the sea urchin, removal of the tube foot would destroy the nervous continuity of the system without necessarily interrupting the pathways from the radial cord to the ampulla. Hydraulically the tube foot-ampulla system of the starfish is described as a closed system (Smith, 1946) and contraction of the ampulla results in the extension of the tube foot. A valve between the ampulla and the radial canal prevents fluid from flowing out of the tube foot-ampulla complex back into the radial canal. The pressure of the fluid in the radial canals is not considered to contribute significantly to the actual extension of the tube foot (Smith, 1946). In sea urchins a slightly different situation exists. Extension of the tube foot is not accompanied by one large contraction of the ampulla but by a number of smaller contractions, separated by periods of relaxation. This pattern of contraction is the same as that induced by electrical stimulation of the cord. It is inferred that the ampullae are pumping water from the radial canals into the tube feet during their extension. In such a system it is likely that neural feedback loops link the tube foot and its ampulla, signalling the relative distension of the different parts of the system. Removal of the tube foot may therefore not interfere directly with any neural pathway at all, but the immediate change in hydraulic pressure due to the removal of the tube foot could induce a block to the normal incoming commands from the central nervous system. This type of control would have functional significance to an animal in which the tube foot was damaged, for continued attempts to extend the damaged foot would cause an unnecessary loss of fluid from the water-vascular system. The prompt contraction of the ‘inhibited’ ampulla to locally applied mechanical stimulation can probably be explained by reflex activity, involving only the nerve and muscle cells of the ampulla.

The nervous systems of the echinoderms have remained little understood in spite of substantial study (Smith, 1965). The whole phylum presents the same peculiar problems for the neurophysiologist in that the nervous elements are very small and apparently hard to show histologically. However, good results have been obtained with the electron microscope and the presence of recordable electrical activity in the nervous system is encouraging. The animals also afford good preparations requiring little or no care with regard to perfusion, and parts of the animal appear to behave in much the same way as the whole animal. The electrical recording technique reported in this paper is not as satisfactory as could be desired in that behavioural reactions are not always correlated with electrical potentials. Refinements of the technique are undoubtedly possible. The preparation which may prove to be the most interesting is that of the tube foot and ampulla, for here some anatomical details are known, at least for the asteroids.

The author would like to thank Dr T. H. Bullock for his help in providing research facilities at the University of California, Los Angeles, and also for his criticism of the manuscript.

Hamann
,
O.
(
1887
).
Beitrâgezur Histologie der Echinodermen. III
.
Anatomie undHistologie der Echiniden und Spatangiden
.
Jena
:
Fischer
.
Laverack
,
M. S.
(
1965
).
Fine structure of the radial nerve cord of Echinus esadentus. (unpublished)
.
Smith
,
J. E.
(
1946
).
The mechanism and innervation of the star-fish tube foot-ampulla system
.
Phil. Trans. B
,
232
,
279
310
.
Smith
,
J. E.
(
1965
).
Echinodermata
.
In Structure and Function in the Nervous Systems of Invertebrates
,
T. H.
Bullock
and
G. A.
Horridge
.
San Francisco
:
W. H. Freeman and Company
.
Takahashi
,
K.
(
1964
).
Electrical responses to light stimuli in the isolated radial nerve cord of the sea urchin Diadema setosum (Leske)
.
Nature, Lond
.,
301
,
1343
14
.