On the whole, then, it would be impossible to imagine two cases more completely parallel than are these of the effects of temperature on natural and on artificial rhythm respectively; and as it must be considered in the last degree improbable that all these coincidences are accidental, I conclude that the effects of temperature on the natural rhythm of Medusæ (and so, in all probability, on the natural rhythm of other ganglio-muscular tissues) are for the most part exerted, not on the ganglionic, but on the contractile element.

In order to test the effects of gases on the artificial rhythm, I took a severed quadrant of Aurelia, and floated it in sea-water, with its muscular surface just above the level of the water. Over the tissue I lowered an inverted beaker filled with the gas the effects of which I desired to ascertain, and by progressively forcing the rim of the beaker into the water I could submit the tissue to various pressures of the atmosphere of the gas I was using. By an appropriate arrangement the electrodes passed into the interior of the beaker, and could then be manipulated from the outside, so as to be properly adjusted on the tissue. In this way I was able to observe that different gases exerted a marked influence on the rate of the artificial rhythm.

The following table gives the ratios in the case of one experiment:—

It may here be observed that to produce these results, both carbonic acid and oxygen must be considerably diluted with air, for otherwise they have the effect of instantaneously inhibiting all response, even to the strongest stimulation. When this is the case, however, irritability returns very soon after the tissue is again exposed to air or to ordinary sea-water. But I desire it to be understood that the results of my experiments on the influence of oxygen, both on the natural and on the artificial rhythm, have proved singularly equivocal; so that as far as this gas is concerned further observations are required before the above results can be accepted as certain.

I have still one other observation of a very interesting character to describe, which is closely connected with the current views respecting ganglionic action, and may therefore be more conveniently considered here than in any other part of this treatise. I have already stated that in no case is the manubrium of a Medusa affected as to its movements by removal of the periphery of the swimming-bell; but in the case of Sarsia a very interesting change occurs in the manubrium soon after the nectocalyx has been paralyzed by excision of its margin. Unlike the manubriums of most of the other Medusæ, this organ, in the case of Sarsia, is very highly retractile. In fresh and lively specimens the appendage in question is carried in its retracted state; but when the animals become less vigorous—from the warmth or impurity of the water in which they are confined, or from any other cause—their manubriums usually become relaxed. The relaxation may show itself in various degrees in different specimens subjected to the same conditions, but in no case is the degree of relaxation so remarkable as that which may be caused by removing the periphery of the nectocalyx. For the purpose of showing this effect, it does not signify in what condition as to vigour, etc., the specimen chosen happens to be in; for whether the manubrium prior to the operation be contracted or partially relaxed, within half an hour after the operation it is sure to become lengthened to a considerable extent.

In order to show the surprising degree to which this relaxation may proceed, I insert a sketch of a specimen both before and after the operation. The sketches are of life size, and drawn to accurate measurement (Figs. 29 and 30).[29]

With regard to this remarkable effect on the manubrium of removing the margin of the nectocalyx, it is now to be observed that in it we appear to have very unexceptionable evidence of such a relation subsisting between the ganglia of the nectocalyx and the muscular fibres of the manubrium as elsewhere gives rise to what is known as muscular tonus. This interpretation of the facts cannot, I think, be disputed; and it fully explains why, in the unmutilated animal, the degree of elongation on the part of the manubrium usually exhibits an inverse proportion to the degree of locomotor activity displayed by the bell. I may here state that I have also observed indications of muscular tonus in some of the other Medusæ, but for the sake of brevity I shall now restrict myself to the consideration of this one case.

To my mind, then, it is an interesting fact that ganglionic tissue, where it can first be shown to occur in the animal kingdom, has for one of its functions the maintenance of muscular tonus; but it is not on this account that I now wish to draw prominent attention to the fact before us. Physiologists are almost unanimous in regarding muscular tonus as a kind of gentle tetanus due to a persistent ganglionic stimulation, and against this opinion it seems impossible to urge any valid objection. But, in accordance with the accepted theory of ganglionic action, physiologists further suppose that the only reason why some muscles are thrown into a state of tonus by ganglionic stimulation, while other muscles are thrown into a state of rhythmic action by the same means, is because the resistance to the passage of the stimulation from the ganglion to the muscle is less in the former than in the latter case. Here, be it remembered, we are in the domain of pure speculation: there is no experimental evidence to show that such a state of differential resistance as the theory requires actually obtains. Hence we are quite at liberty to suppose any other kind of difference to obtain, either to the exclusion of this one or in company with it. Such a supposition I now wish to suggest, and it is this—that all rhythmical action being regarded as due (at any rate in large part) to the alternate exhaustion and restoration of excitability on the part of contractile tissues, the reason why continuous ganglionic stimulation produces incipient tetanus in the case of some muscles and rhythmic action, in the case of others, is either wholly or partly because the irritability of the muscles in relation to the intensity of the stimulation is greater in the former than in the latter case. If this supposition as to differential irritability be granted, my experiments on paralyzed Aurelia prove that tetanus would result in the one case and rhythmic action in the other. For it will be remembered that in these experiments, if the continuous faradaic stimulation were of somewhat more than minimal intensity, tetanus was the result; while if such stimulation were but of minimal intensity, the result was rhythmic action. Now, that in the particular case of Sarsia the irritability of the tonically contracting manubrium is higher than that of the rhythmically contracting bell is a matter, not of supposition, but of observable fact; for not only is the manubrium more irritable than the bell in response to direct stimulation of its own substance, but it is generally more so even when the stimuli are applied anywhere over the excitable tissues of the bell. And from this it is evident that the phenomena of muscular tonus, as they occur in Sarsia, tend more in favour of the exhaustion than of the resistance theory.[30]

I will now sum up this rather lengthy discussion. The two theories of ganglionic action may be stated antithetically thus: in both theories the accumulation of energy by ganglia is supposed to be a continuous process; but while the resistance theory supposes the rhythm to be exclusively due to an intermittent and periodic discharge of this accumulated energy on the part of the ganglionic tissues, the exhaustion theory supposes that the rhythm is largely due to a periodic process of exhaustion and recovery on the part of the responding tissues. Now, I submit that my experiments have proved the former of these two theories inadequate to explain all the phenomena of rhythm as it occurs in the Medusæ; for these experiments have shown that even after the removal of the only ganglia which serve as centres of natural stimulation, the excitable tissues still continue to manifest a very perfect rhythm under the influence of any mode of artificial stimulation (except heat), which is of a constant character and of an intensity sufficiently low not to produce tetanus. And as I have proved that the rhythm thus artificially produced is almost certainly due to the alternate process of exhaustion and recovery which I have explained, there can scarcely be any doubt that in the natural rhythm this process plays an important part, particularly as we find that temperature and gases exert the same influences on the one rhythm as they do on the other. Again, as an additional reason for recognizing the part which the contractile tissues probably play in the production of rhythm, I have pointed to the fact that in the great majority of cases in which rhythmic action occurs the presence of ganglia cannot be suspected. For it is among the lower forms of life, where ganglia are certainly absent, and where the functions of stimulation and contraction appear to be blended and diffused, that rhythmic action is of the most frequent occurrence; and it in obvious with how much greater difficulty the resistance theory is here beset than is the one I now propose. Granted a diffused power of stimulation with a diffused power of response, and I see no essential difference between the rhythmic motions of the simplest organism and those of a deganglionated Medusa in acidulated water. Lastly, the facts relating to the tonus of the manubrium in Sarsia furnish very striking, and I think almost conclusive, proof of the theory which I have advanced.

2. Nitrite of Amyl.—On Sarsia the effect of this agent is much the same as that of chloroform—the description just given being quite as applicable to the affects of the nitrite as to those of chloroform. Before the loss of spontaneity supervenes, the rate of the rhythm is increased, while the strength of the pulsations is diminished.

Tiaropsis diademata, from the fact of its presenting a very regular rhythm and being but of small size, is a particularly suitable species upon which to conduct many experiments relating to the effect of poisons. On this species the nitrite in appropriate (i.e. in very small) doses first causes irregularity and enfeeblement of the contractions, together with quickening of the rhythm. After a short time, a gradual cessation of the swimming motions becomes apparent—these motions dying out more gradually, for example, than they do under the influence of chloroform. Eventually each pulsation is marked only by a slight contraction of the muscular tissue in the immediate neighbourhood of the margin. If the dose has been stronger, however, well-marked spasmodic contractions come on and obliterate such gradual working of the poison. In all cases irritability of all parts of the animal persists for a long time after entire cessation of spontaneous movements—perhaps for three or four minutes in not over-poisoned animals; but eventually it too disappears. On being now transferred to normal sea-water, the process of recovery is slower than it is after anæsthesiation by chloroform. It is interesting, moreover, to observe, that just as the power of co-ordination was the first thing to be affected by the nitrite, so it is the last thing to return during recovery.

3. Caffein.—The effects of caffein on Sarsia may be best studied by immersing the animals in a saturated sea-water solution of the substance. In such solutions the Medusæ float to the surface, in consequence of their lower specific gravity. I therefore used shallow vessels, in order that the margins of the nectocalyces might rest in the level of the water that was thoroughly saturated. The immediate effect of suddenly immersing Sarsia in such a solution is very greatly to increase the rate of the pulsations, and, at the same time, to diminish their potency. The appearance presented by the swimming motions is therefore that of a fluttering nature; and such motions are not nearly so effectual for progression as are the normal pulsations in unpoisoned water. This stage, however, only lasts for a few seconds, after which the spontaneous motions begin gradually to fade away. Soon they altogether cease, though occasionally one among a number of Sarsiæ confined in the same saturated solution will continue, even for several minutes after the first immersion, to give one or two very feeble contractions at long intervals. Eventually, however, all spontaneity ceases on the part of all the specimens, and now the latter will continue for a very long time to be sensitive to stimulation. At first several feeble locomotor contractions will be given in response to each stimulus; and as on the one hand these contractions never originate spontaneously, while, on the other hand, paralyzed Sarsiæ never respond to a single stimulus with more than a single contraction, these multiple responses must, I think, be ascribed to a state of exalted reflex irritability. After a long exposure to the poison, however, only a single response is given to each stimulus; and still later all irritability ceases. On now transferring the Sarsiæ to unpoisoned water, recovery is effected even though the previous exposure has been of immensely long duration, e.g. an hour.

An interesting point with regard to caffein-poisoning of Sarsia is, that as soon as spontaneity ceases the tentacles and manubrium lose their tonus and become relaxed to their utmost extent. This is not the case with anæsthesiation by chloroform, even when pushed to the extent of suspending irritability. If, however, Sarsiæ which have been anæsthesiated to this extent in chloroform be suddenly transferred to a solution of caffein, the tentacles and manubrium may soon be seen to relax, and eventually these organs lose their tonus as completely as if the anæsthesia had from the first been produced by the caffein. Moreover in this experiment the irritability, which had been destroyed by the chloroform, returns in the solution of caffein—provided the latter be not quite saturated—though spontaneity of course remains suspended throughout.

The effects of graduating the doses of caffein may be stated in connection with another species, viz. Tiaropsis diademata. In a weak solution the effects are a quickening of the pulsations (e.g. from 64 to 120 per minute) together with a lessening of their force. On slightly increasing the dose, the pulsations become languid, and prolonged pauses supervene. If the dose is again somewhat strengthened, the pulsations become weaker and weaker, till they eventually cease altogether. The animal, however, is now in a condition of exalted reflex irritability; for its response to a single stimulus consists not merely, as in the unpoisoned animal, of a single spasm, but also, immediately after this, of a series of convulsive movements somewhat resembling swimming movements destitute of co-ordination. If the strength of the solution be now again increased, a stage of deeper anæsthesiation may be produced, in which the Medusa will only respond to each stimulation by a single spasm. In still stronger solutions, the only response is a single feeble contraction; while in a nearly saturated solution the animal does not respond at all. But even from a saturated solution Tiaropsis diademata will recover when transferred to unpoisoned water.

4. Strychnia.—The species of covered-eyed Medusa which I shall choose for describing the action of strychnia is Cyanæa capillata, which is most admirably adapted for experiments with this and some of the other alkaloid poisons, from the fact that in water kept at a constant temperature its pulsations are as regular as are those of a heart. After Cyanæa capillata has been allowed to soak for ten minutes or so in a weak sea-water solution of strychnia, unmistakable signs of irregularity in the pulsations supervene. This irregularity then increases more and more, till at length it grows into well-marked convulsions. The convulsions manifest themselves in the form of extreme deviations from the rhythmical contractions so characteristic of Cyanæa capillata. Instead of the heart-like regularity with which systole and diastole follow one another in the unpoisoned animal, we now have periods of violent and prolonged systole resembling tonic spasm; and when the severity of this spasm is for a moment abated, it is generally renewed before the umbrella has had time again to become fully expanded. Moreover, the spasm itself is not of uniform intensity throughout the time it lasts; but while the umbrella is in a continuously contracted state, there are observable a perpetual succession of extremely irregular oscillations in the strength of the contractile influence. It is further a highly interesting fact that the convulsions are very plainly of a paroxysmal nature. After the umbrella has suffered a prolonged period of convulsive movements, it expands to its full dimensions, and in this form remains for some time in a state of absolute quiescence. Presently, however, another paroxysm supervenes, to be followed by another period of quiescence, and so on for hours. The periods of quiescence are usually shorter than are those of convulsion; for while the former seldom last more than forty seconds or so, the latter may continue uninterruptedly for five or six minutes. In short, Medusæ, when submitted to the influence of strychnia, exhibit all the symptoms of strychnia poisoning in the higher animals. Death, however, is always in the fully expanded form.

It seems desirable to supplement these remarks with a few additional ones on the effects of this poison on the naked-eyed Medusæ. In the case of Sarsia the symptoms of strychnia poisoning are not well marked, from the fact that in this species convulsions always take the form of locomotor contractions. The symptoms, however, are in some respects anomalous. They are as follows. First of all the swimming motions become considerably accelerated, periods of quiescence intervening between abnormally active bouts of swimming. By-and-by a state of continuous quiescence comes on, during which the animal is not responsive to tentacular irritation, but remains so to direct muscular irritation, giving one response to each direct stimulus. The tentacles and manubrium are much relaxed. In a sea-water solution just strong enough to taste bitter, this phase may continue for hours; in fact, till a certain opalescence of the contractile tissues—which it is a property of strychnia, as of most other reagents, to produce—has advanced so far as to place the tissues beyond recovery. If the exposure to such a solution has not been very prolonged, recovery of the animal in normal water is rapid. In a specimen exposed for two and a half hours to such a solution, recovery began in half an hour after restoration to normal water, but was never complete. In all cases, if the poisoning is allowed to pass beyond the stage at which response to direct muscular irritation ceases, the animal is dead.

On Tiaropsis indicans this poison has the effect of causing a general spasm, which would be undistinguishable from that which in this species results from general stimulation of any kind, were it not that there is a marked difference in one particular. For in the case of strychnia poisoning, the spasm, while it lasts, is not of uniform intensity over all parts of the nectocalyx; but now one part and now another part or parts are in a state of stronger contraction than other parts, so that, as a general consequence, the outline of the nectocalyx is continually changing its form. Moreover, in addition to these comparatively slow movements, there is a continual twitching observable throughout all parts of the nectocalyx. Each individual twitch only extends over a small area of the contractile tissue; but in their sum their effect is to throw the entire organ into a sort of shivering convulsion, which is superimposed on the general spasm. After a time the latter somewhat relaxes, leaving the former still in operation, which, moreover, now assumes a paroxysmal nature—the convulsions consisting of strong shudders and frequent spasms with occasional intervals of repose.

In the case of Tiaropsis diademata the action of strychnia is very similar, with the exception that there is no continuous spasm, although occasional ones occur amid the twitching convulsions. After a time, however, all convulsions cease, and the animal remains quiescent. While in this condition its reflex excitability is abnormally increased, as shown by the fact that even a gentle touch will bring on, not merely a single responsive spasm, as in the unpoisoned animal, but a whole series of successive spasms, which are often followed by a paroxysm of twitching convulsions. The condition of exalted reflex irritability is thus exceedingly well marked. Recovery in normal water at this stage is rapid, the motions being at first characterized by a want of co-ordination, which, however, soon passes off.

5. Veratrium.—In Sarsia the first effect of this poison is to increase the number and potency of the contractions; but its later effect is just the converse, there being then prolonged periods of quiescence, broken only by very short swimming bouts consisting of feeble contractions. The feebleness of the contractions gradually becomes more and more remarkable, until at last it is with great difficulty that they can be perceived at all; indeed, the progressive fading away of the contractions into absolute quiescence is so gradual that it is impossible to tell exactly when they cease. During the quiescent stage the animal is for the first time insensible both to tentacular and to direct stimulation of the contractile tissues. That the gradual dying out of the strength of the contractions is not altogether due to the progressive advance of central paralysis, would seem to be indicated by the fact that contractions in response to direct stimulation of the contractile tissues are no more powerful, at any given stage of the poisoning, than are either responses to tentacular stimulation or the spontaneous contractions. Still, as we shall immediately see, in the various species of Tiaropsis, irritability persists after cessation of the spontaneous contractions. In Sarsia the nervous connections between the tentacles and manubrium, and also between the tentacles themselves, are not impaired during the time that the bell is motionless; and even when the irritability of the bell has quite disappeared as regards any kind of stimulation, the manubrium and tentacles will continue responsive to stimuli applied either directly to themselves or to any part of the neuro-muscular sheet of the bell.

The convulsions due to the action of veratrium are well marked in the various species of the genus Tiaropsis. They consist of violent fluttering motions without any co-ordination; but there are no spasms, as in the case of strychnia poisoning. After the convulsions have lasted for some time, a quiescent stage comes on, during which the animal remains responsive to stimulation, though not abnormally so. Recovery in unpoisoned water is rapid, the movements being at first marked by an absence of co-ordination.

6. Digitalin.—The first effect of this poison on Sarsia is to quicken the swimming motions, and then to enfeeble them progressively till they degenerate into mere spasmodic twitches. The manubrium and tentacles are now strongly retracted, while the nectocalyx is drawn together so as to assume an elongated form. The latter is now no longer responsive either to tentacular or to direct stimulation; but the tentacles and manubrium both remain responsive to stimuli applied either directly to themselves or to the neuro-muscular tissue of the bell. Death always takes place in very strong systole; and as this is an exceedingly unusual thing in the case of Sarsia, there can be no doubt that, in this respect, the action of the digitalin is different on the Medusæ from what it is on the heart.

On the various species of Tiaropsis, digitalin at first causes acceleration of the swimming movements, with great irregularity and want of co-ordination. Next, strong and persistent spasms supervene, which give the outline of the nectocalyx an irregular form; and every now and then this unnatural spasm gives place to convulsive swimming motions. Evidently, however, the spasm becomes quite persistent and excessively strong. The manubrium of Tiaropsis indicans crouches to its utmost, and the animal dies in strong systole.

7. Atropin.—In the case of Sarsia atropin causes convulsive swimming motions. The systoles next become feeble, and finally cease. The nectocalyx is now somewhat drawn together in persistent systole, with the manubrium and tentacles strongly retracted. Muscular irritability remains after tentacular irritability has disappeared, but it is then decidedly enfeebled.

In the various species of Tiaropsis the convulsions are strongly pronounced. They begin as mere accelerations of the natural swimming motions, but soon grow into well-marked convulsions, consisting of furious bouts of irregular systoles following one another with the utmost rapidity, and wholly without co-ordination. Occasionally these movements are interrupted by a violent spasm, on which strong shuddering contractions are superimposed.

8. Nicotin.—On dropping Sarsia into a sea-water solution of nicotin of appropriate strength, the animal immediately goes into a violent and continuous spasm, on which a number of rapidly succeeding minute contractions are superimposed. The latter, however, rapidly die away, leaving the nectocalyx still in strong and continuous systole; tentacles and manubrium are retracted to the utmost. Shortly after cessation of spontaneity, the bell is no longer responsive to tentacular stimulation, but remains for a considerable time responsive to direct stimulation of its own substance; eventually, however, all irritability disappears, while the tentacles and manubrium relax. On transferring the animal to normal water, muscular irritability first returns, and then central, as shown by the earlier response of the bell to direct than to tentacular stimulation; but if the animal has been poisoned heavily enough to have had its muscular irritability suspended, it is a long time before central irritability returns. Soon after central irritability has returned, the animal begins to show feeble signs of spontaneity, the motions being exceedingly weak, with long intervals of repose; but the degree of such feebleness depends on the length of time during which the animal has previously been exposed to the poison; thus in a specimen which had been removed from the poison immediately after the disappearance of reflex irritability had supervened, recovery began in ten minutes after re-immersion, and was complete in half an hour.

In Tiaropsis the symptoms of nicotin poisoning are also well marked. When gradually administered, the first effect of the narcotic is a complete loss of co-ordination in the swimming motions. A slight increase of the dose brings about a tonic spasm, which differs from the natural spasm of these animals—(a) in being stronger, so that the nectocalyx becomes bell-shaped rather than square, (b) in being much more persistent, and (c) in undergoing variations in its intensity from time to time, instead of being a contraction of uniform strength; thus the spasm temporarily affects some parts of the nectocalyx more powerfully than other parts, so that the organ may assume all sorts of shapes. Such distortions proceed even further under the influence of nicotin than under that of strychnine, etc. Sometimes, for instance, one quadrant will project in the form of a pointed promontory; at other times two adjacent or opposite quadrants will thus project, and occasionally all four will do so, the animal thus becoming star-shaped. Sometimes, again, one quadrant will be less contracted than the other three, while at other times more or less slight relaxations affect numerous parts of the bell, its margin being thus rendered sinuous, though more or less violently contracted in all its parts. This state of violent spasm lasts for several minutes, when it gradually passes off, the nectocalyx relaxing into the form of a deep bowl and remaining quite passive, except that every now and then one part or another of the margin is suddenly contracted in a semilunar form. By-and-by, however, even these occasional twitches cease, and the animal is now insensible to all kinds of stimulation. Recovery in normal water is gradual, and marked in its first stage by the occasional retractions of the margin last mentioned. At about this stage also, or sometimes slightly later, the animal first becomes responsive to stimulation; and it is interesting to note that the response is performed, not by giving a general spasm as would the unpoisoned animal, but by folding in the part irritated—an action which very much resembles, on the one hand, the spontaneous convulsive movements just described, and, on the other, the response which is given to stimulation by the unpoisoned bell when gently irritated after removal of its margin. After these stages there supervenes a prolonged period of quiescence, during which the animal remains normally responsive to stimulation. Spontaneity may not return for several hours, and, after it does return, the animal is in most cases permanently enfeebled. Indeed, on all the species of Medusæ, nicotin, both during its action and in its subsequent effects, is the most deadly of all the poisons I have tried.

9. Morphia.—The anæsthesiating effects of morphia are as decided as are those of chloroform. I shall confine myself to describing the process of anæsthesiation in the case of Aurelia aurita in an extract from my notes. "A very vigorous specimen, having twelve lithocysts, was placed in a strong sea-water solution of morphia. Half a minute after being introduced commencement of torpidity ensued, shown by contractions becoming fewer and feebler. In one minute the feeble impulses emanating from the prepotent lithocyst failed to spread far through the contractile tissue, appearing to encounter a growing resistance. Eventually this resistance became so great that only a very small portion of contractile tissue in the immediate neighbourhood of the lithocyst contracted, and this in a very slow and feeble way. Two minutes after immersion even these partial contractions entirely ceased, and soon afterwards all parts of the animal were completely dead to stimulation. Recovery in normal water slower than that after chloroform, but still soon quite complete. Repeated experiment on this individual four times without injury."

10. Alcohol.—The solution must be strong to cause complete intoxication. The first effect on Sarsia is to cause a great increase in the rapidity of the swimming motions—so much so, indeed, that the bell has no time to expand properly between the occurrence of the successive systoles, which, in consequence, are rendered feeble. These motions gradually die out, leaving the animal quite motionless. The nectocalyx is now responsive to stimuli applied at the tentacles, and sometimes two or three contractions will follow such a stimulus, as if the spontaneity of the animal were slightly aroused by the irritation. Soon, however, only one contraction is given in response to every tentacular irritation, and by-and-by this also ceases—the Medusa being thus no longer responsive to central stimulation. It remains, however, for a long time responsive to stimulation of the neuro-muscular sheet; indeed, the strength of the alcohol solution must be very considerable before loss of muscular irritability supervenes. It may thus be made to do so, however; and on then transferring the animal to normal water, recovery begins in from three minutes to a quarter of an hour. The first contractions are very feeble, with long intervals of repose; but gradually the animal returns to its normal state.

The above remarks apply also to Tiaropsis. In Tiaropsis indicans the manubrium recovers in normal water sooner than the nectocalyx. Both in Sarsia and Tiaropsis the manubrium and tentacles are retracted while exposed to alcohol, and, after transference to normal sea-water, the animals float on the surface, presumably in consequence of their having imbibed some of the spirit. The period during which flotation lasts depends, (a) on the strength of the alcohol solution used, and (b) on the time of exposure to its influence. It may last for an hour or more; but in no case is recovery complete till some time after the flotation ceases.

11. Curare.—Curare had already been tried upon Medusæ, and was stated to have produced no effects; it is therefore especially desirable that I should first of all describe the method of exhibiting it which I employed.

Having placed the Medusæ to be examined in a flat-shaped beaker, I filled the latter to overflowing with sea-water. I next placed the beaker in a large basin, into which I then poured sea-water until the level was the same inside and outside the breaker, i.e. until the two bodies of water all but met over the brim. Having divided the Medusæ across its whole diameter, with the exception of a small piece of marginal tissue at one side to act as a connecting link between the two resulting halves, I transferred one of these halves to the water in the basin, leaving the other half still in the beaker—the marginal tissue which served to unite the two halves being thus supported by the rim of the beaker. Over the minute portion of the marginal tissue which was thus of necessity exposed to the air, I placed a piece of blotting-paper which dipped freely into the sea-water. Lastly, I poisoned the water in the beaker with successive doses of curare solution.

The results obtained by this method were most marked and beautiful. Previous to the administration of the poison both halves of the Medusæ were of course contracting vigorously, waves of contractile influence now running from the half in the beaker to the half in the basin, and now vice versâ. But after the half in the beaker had become effectually poisoned by the curare, all motion in it completely ceased, the other, or unpoisoned half, continuing to contract independently. I now stimulated the poisoned half by nipping a portion of its margin with the forceps. Nothing could be more decided than the result. It will be remembered that when any part of Staurophora laciniata is pinched with the forceps or otherwise irritated, the motion of the whole body which ensues is totally different from that of an ordinary locomotor contraction—all parts folding together in one very strong and long-protracted systole, after which the diastole is very much slower than usual. Well, on nipping any portion of the poisoned half of Staurophora laciniata, this half remained absolutely motionless, while the unpoisoned half, though far away from the seat of irritation, immediately ceased its normal contractions, and folded itself together in the very peculiar and distinctive manner just described. This observation was repeated a number of times, and, when once the requisite strength of the curare solution had been obtained, always with the same result. The most suitable strength I found to be 1 in 2500, in which solution the poisoned half required to soak for half an hour.

I also tried the effect of this poison on the covered-eyed Medusæ, and have fairly well satisfied myself that its peculiar influence is likewise observable in the case of this group, although not in nearly so well-marked a manner.

It has further to be stated that when the poisoned half is again restored to normal sea-water, the effects of curare pass off with the same rapidity as is observable in the case of the other poisons which I have tried. Thus, although an exposure of half an hour to the influence of curare of the strength named is requisite to destroy the motor power in the case of Staurophora laciniata, half a minute is sufficient to ensure its incipient return when the animal is again immersed in unpoisoned water.

It is also to be observed that a very slight degree of over-poisoning paralyzes the transmitting system as well as the responding one; so that if any one should repeat my observation, I must warn him against drawing erroneous conclusions from this fact. Let him use weak solutions with prolonged soaking, and by watching when the voluntary motions in the poisoned half first cease, he need experience no difficulty in obtaining results as decided as it is possible for him to desire.

12. Cyanide of Potassium.—On Sarsia the first effect is to quicken the contractions and then to enfeeble them. The animal assumes an elongated form, as already described under atropin. Spontaneity ceases very rapidly even in weak solutions; and for an exceedingly short time after it has done so, the bell continues responsive both to tentacular and to direct stimulation. For a long time after the bell ceases to respond to any kind of stimulation, the nervous connections between the tentacles and between the tentacles and manubrium remain intact, as also do the nervous connections of these organs with all parts of the bell. This interesting fact is rendered apparent, first, by stimulating a tentacle and observing that all the four tentacles and the manubrium respond; and, second, by irritating any part of the neuro-muscular sheet of the bell, and observing that while the latter does not respond both the tentacles and the manubrium retract. Recovery from this stage occupies several hours.

In the case of Tiaropsis the convulsions are, as usual, more pronounced, being marked by the occurrence of a gradually increasing spasm, which differs from a normal spasm in the respects already described under strychnia. In all the species both of Sarsia and Tiaropsis, the manubrium and tentacles are retracted during exposure to this poison.

The above comprises all the poisons which I have tried, and I think that all the observations taken together show a wonderful degree of resemblance between the actions of the various poisons on the Medusæ and on the higher animals—a general fact which is of interest, when we remember that in these nerve-poisons we possess, as it were, so many tests wherewith to ascertain whether nerve-tissue, where it first appears upon the scene of life, presents the same fundamental properties as it does in the higher animals. And these observations show that such is the case. When the physiologist bears in mind that in Sarsia we have the means of testing the comparative influence of any poison on the central, peripheral, and muscular systems respectively,[31] he will not fail to appreciate the significance of these observations. In reading over the whole list he will meet with an anomaly here and there; but, on the whole, I do not think he can fail to be satisfied with the wonderfully close adherence which is shown by these elementary nervous tissues to the rules of toxicology that are followed by nervous tissues in general. In one respect, indeed, there is a conspicuous and uniform deviation from these rules; for we have seen that in the case of every poison mentioned more or less complete recovery takes place when the influence of the poison has been removed, even though this has acted to the extent of totally suspending irritability. In other words, there is no poison in the above list which has the property, when applied to the Medusæ, of destroying life till long after it has destroyed all signs of irritability. What the cause of this uniform peculiarity may be is, of course, conjectural; but I may suggest two considerations which seem to me in some measure to mitigate the anomaly. In the first place, we must remember that in the Medusæ there are no nervous centres of such vital importance to the organism that any temporary suspension of their functions is followed by immediate death. Therefore, in these animals, the various central nerve-poisons are at liberty, so to speak, to exert their full influence on all the excitable tissues without having the course of their action interrupted by premature death of the organism, which in higher animals necessarily follows the early attack of the poison on a vital nerve-centre. Again, in the second place, we must remember that the method of administering the above-mentioned poisons to the Medusæ was very different from that which we employ when administering them to other animals; for, in the case of the Medusæ, the neuro-muscular tissue is spread out in the form of an exceedingly tenuous sheet, so that when the animal is soaking in the poisoned water every portion of the excitable tissue is equally exposed to its influence; and that the action of a poison is greatly modified by such a difference in the mode of its administration has been proved by Professor Gamgee, who found that when a frog's muscle is allowed to soak in a solution of vanadium, etc., it loses its irritability, while this is not the case if the poison is administered by means of the circulation.

I may further observe that in the case of all poisons I have tried, the time required for recovery after the animal is restored to normal water varies immensely. The variations are chiefly determined by the length of time during which the animal has been exposed to the influence of the poison, but also, in a lesser degree, by the strength of the solution employed. To take, for instance, the case of caffein or chloroform, if Sarsiæ are transferred to normal water after they first cease to move, a few seconds are enough to restore their spontaneity; whereas, if they are allowed to remain in the poisoned water for an hour, they may not move for one or two hours after their restoration to unpoisoned water. In consequence of such great variations occurring from these causes, I was not able to compare the action of one poison with that of another in respect of the time required for effects of poisoning to pass away.

I shall conclude all I have to say upon the subject of poisons by stating the interesting fact, that if any of the narcotic or anæsthesiating agents be administered to any portion of a contractile strip cut from the umbrella of Aurelia aurita in the way already described, the rate of the contraction-waves is first progressively slowed, and eventually their passage is completely blocked at the line where the poisoned water begins. Upon now restoring the poisoned portion of the contractile strip to normal sea-water the blocking is gradually overcome, and eventually every trace of it disappears.[32]

The contractile wave may be blocked by poisons in another way. A glance at Fig. 11 will show that a circumferential strip cut from the umbrella of Aurelia aurita is pervaded transversely by a number of nutrient tubes, which have all been cut through by the section. At the side of the strip, therefore, furthest from the margin there are situated a number of open ends of these nutrient tubes. Now, on injecting any of the narcotic poisons into one of these open ends, the fluid of course permeates the whole tube, and the contraction-wave becomes blocked at the transverse line occupied by the tube as effectually as if the contractile strip had been cut through at that line.

A glance at Fig. 10, again, will show that each lithocyst is surrounded by one of these nutrient tubes. Upon injecting this tube, therefore, in a contractile strip, the effect of the poison may be exerted on the lithocyst more specially than it could be by any other method of administration. In view of recent observations concerning the effects of curare on the central nervous masses of higher animals, it may be worth while to state that a discharging lithocyst of Aurelia aurita, when thus injected with curare, speedily ceases its discharges. This fact alone, however, would not warrant any very trustworthy conclusions as to the influence of curare upon discharging centres; for it is not improbable that the paralyzing effects may here be due to the influence of the poison on the surrounding contractile tissue.

It is interesting to observe that if the discharging lithocyst be injected with chloroform, or a not too strong solution of morphia, it recovers in the course of a night. With alcohol the first effects of the injection are considerably to accelerate the frequency and to augment the potency of the discharges; but the subsequent effects are a gradual diminution in the frequency and the vigour of these discharges, until eventually total quiescence supervenes. In the course of a few hours, however, the torpidity wears away, and finally the Medusæ returns to its normal state.[33]