Réaumur first recorded that if the leg of a crayfish or of a crab is cut off outside of the breaking-joint it is always thrown off later at the breaking-joint. Fredericq has made a careful examination of the way in which the leg is thrown off in the crab, Carcinus mænas. He found that the breaking does not take place at the weakest part of the leg; for the leg of a dead crab will support a weight of 3½ to 5 kilograms, which represents about one hundred times the weight of the crab’s body. When the weight is increased to a point at which the leg breaks, it does so between the body and the first segment or between the first and second segments. When it breaks off in this way, the edges are ragged and are left in a lacerated condition; but when the leg is thrown off by the animal at the breaking-joint, there is left a smooth surface covered over, except in the centre, by a thin cuticle. Through the opening in the centre of this cuticle a nerve and a blood vessel pass to the distal part of the leg. Very little bleeding takes place when the leg is thrown off, but if the leg is cut off or broken off at any other level the bleeding is much greater. Fredericq studied the physiological side of the process and found that it is the result of a reflex nervous act. He found that if the brain of the animal is destroyed the leg may still be thrown off, but if the ventral cord is destroyed the reflex action does not take place. The reflex is brought about ordinarily by an injury to the leg that starts a nerve impulse to the ventral nerve-cord, and from this a returning impulse is sent to the muscles of the same leg, causing the muscles in the region of the breaking-joint to contract violently, and the result of their contraction is to break off the leg. If the muscles are first injured, the leg cannot be thrown off. Andrews, who has studied the structure of the breaking-joint in the spider-crab; has found that there is a plane of separation extending inwards from the groove on the surface. This plane is made by a narrow space between two chitinous membranes that are continuous at their outer ends with the general chitinous covering of the leg (Fig. 45, C). When the leg breaks off, one-half of the double membrane is left attached to the base of the leg and the other to the part that is lost. This in-turned membrane seems to correspond to the in-turning of the surface cuticle in the region of the joints. The arrangement of the muscles at the breaking-joint is shown in Fig. 45, B. The upper muscle is the extensor muscle of the leg, and through its contraction the breaking off takes place. When it contracts the leg is brought against the side of the body, which is supposed to offer the resistance necessary to break off the leg. If the leg is held by an enemy, this may offer sufficient resistance for the muscle to bring about the breaking.

In many crabs the leg is not thrown off if simply held, but only after an injury. Even the most distal segment may be cut off and the leg remain attached, and sometimes it is not lost after the distal end of the next to the last segment is cut off. If a crab is tethered by one leg it will not throw off its leg in order to escape, unless, in the crab’s excitement, the leg is twisted or broken. How generally this holds for all crabs cannot be stated. Herrick says: “Unintentional experiments in autotomy have often been made by tethering a lobster or a crab by its large claws. The animal, of course, escapes, leaving only its leg behind. When lobsters are drawn out of the water by the claws, or when a claw is pinched by another lobster, or while they are handled in packing, especially for the winter market, they often ‘cast a claw,’ and the transportation of lobsters at this season is said to be attended by considerable loss in consequence.” The large claws of the lobster are quite heavy, the base relatively small at the breaking-joint, and it may be that simply the weight of the claw, when out of the water, may strain the leg so that it breaks off,—the leg being injured by its own weight. The lobster seems to lose its claws quite often under natural conditions. Rathburn[71] states that “out of a hundred specimens collected for natural history purposes in Narragansett Bay in 1880, fully 25 per cent had lost a claw each, and a few both claws.” Herrick[72] records that “in a total of 725 lobsters captured at Woods Holl in December and January, 1893-1894, fifty-four, or 7 per cent, had thrown off one or both claws.”

The autotomy of the arms of the starfish has been often observed.[73] The arms are thrown off very near the base in many forms. If the animal is simply held by the arm it does not break off, but if injured it constricts and falls off. The lost arm does not regenerate a new starfish in most forms, but, as stated on page 102, there are recorded some cases in which the arm seems to have this power. King has found that out of a total of 1914 starfish (Asterias vulgaris) there were 206, or 10.76 per cent, that had new arms, and all of these, with one exception, arose from the base of the arm. The growth of the new arm from the base takes place more rapidly, as shown in Fig. 38, A, than when the arm is regenerated from a more distal level; but in the latter case the arm, despite its slower growth, may complete itself before another does that originates at the same time from the base of the old arm. There is, therefore, in this respect no obvious advantage, so far as regeneration is concerned, in throwing off the injured arm nearer to the disk.

In the brittle-stars (ophiurians) the arm breaks off with greater ease and at any level. If the arm is simply held and squeezed, it will, in some forms, break off just proximal to where it is held. If the broken end is again held, another small piece breaks off, and in this way the arm may be autotomized, piece by piece, to its very base. Regeneration may take place from any region, but, as yet, no observations have been made on the relative rate of growth of the new arm at different levels.

One of the most remarkable cases of autotomy is that in holothurians, in which the Cuvierian organs, and even the entire viscera, may be ejected when the animal is disturbed. A new digestive tract is regenerated.[74]

It is known that several of the myriapods lose their legs at a definite region near the base, and that they have the power of throwing off the leg in this region if it is injured. I have often observed that the legs of Scutigera forceps are thrown off if they are held or injured, and even when the animal is thrown into a killing fluid. Amongst the insects the plasmids or walking-sticks also throw off their legs at a definite joint, as described by Scudder, and more recently by Bordage, and still later by Godelmann. New legs are regenerated from the stump of the old leg, as has long been known.[75] Other insects do not have the power of throwing off their legs, and we have only a few observations that show that the legs if lost can be regenerated. It is known in the cockroach that the tarsus can regenerate if lost or if cut off, and that fewer segments are regenerated than are present in the normal animal. Newport found that the true legs of a caterpillar are regenerated during the pupa stage if they have been previously cut off.

A further example of autotomy is found in the white ants, which break off their wings at the base after the nuptial flight. There exists a definite and pre-formed breaking-plane in this region. The true ants also lose their wings after the nuptial flight, but there does not seem to be a definite plane of breaking, so that the process can scarcely be called one of autotomy. These cases also differ from the other cases of autotomy in that the lost parts are not renewed.

It has been observed[76] that if the leg of tarantula is cut off at any other place than at the coxa, the animal bites off the wounded leg with its jaws down to the coxa. In other spiders this does not occur, although Schultz has observed that when the legs are lost under natural conditions they are found broken off in most cases at the coxa. Schultz has also found that the legs regenerate better from this region than from any other. It would be rash, I think, to conclude without further evidence that the habit of tarantula to bite off a wounded leg down to the coxa has been acquired in connection with the better regeneration of the leg at this place. It is possible that the injury may excite the animal to bite off the leg as far as possible, which might be to the coxal joint. It would certainly be very remarkable if this spider had acquired the habit in connection with the better regeneration of the leg at the base, since the leg can presumably also regenerate at any level, as in the epeirids.

In this same connection I may record that in the hermit-crab I have often observed that when a leg is cut off outside of the breaking-joint, if the leg is not thrown off at once, the claws of the first legs catch hold of the stump and, pulling at the leg, offer sufficient resistance for the leg to break off at the breaking-joint. I cannot believe that this instinct has anything to do with the better regeneration of the leg at the coxal joint, however attractive such an hypothesis may appear.

THEORIES OF AUTOTOMY

A number of writers have pointed out that under certain conditions it is an obvious advantage to the animal to be able to throw off a portion of the body and thereby escape from an enemy. It has been suggested that if a crab is seized by the leg, the animal may save its life at times at the expense of its leg; and since the crab has the power of regenerating a new leg, it is the gainer in the long run by the sacrifice. The holothurian, that ejects its viscera, has been supposed to offer a sufficient reward to its hungry enemy, and escapes paying the death penalty, at the expense of its digestive tract. Thus, having shown that the process of autotomy is a useful one, it is claimed that it must have been acquired through a process of natural selection! An equally common opinion is that autotomy is an adaptation for regeneration, since in certain cases, as in that of the crab’s leg, better conditions for subsequent regeneration occur at the breaking-joint than when the amputation takes place at any other region. Since less bleeding takes place when the crab’s leg is thrown off at the breaking-joint, and since the wound closes more quickly when the arm of the starfish is lost at the base, it is assumed that we have in both cases an adaptation to meet accidents, and that the adaptation has been acquired by natural selection.

A consideration of these questions involves us once more in a discussion of the theory of natural selection. An attempt has been made in another place (pages 108-110) to show that we are not justified in assuming that because a process is useful, therefore it must have been acquired by means of natural selection. Even if it were granted that the theory of natural selection is correct, it does not follow that all useful processes have arisen under its guidance. We may, therefore, leave the general question aside, and inquire whether the process of autotomy could have arisen through natural selection (admitting that there is such a process, for the sake of the present argument), or whether autotomy must be due to something else.

If we assume that the leg of some individual crayfishes or crabs, for example, broke off, when injured, more easily at one place than at another, and that regeneration took place as well, or even better, from this region than from any other, and if we further assume that those animals in which this happened would have had a better chance of survival than their fellows, then it might seem to follow that in time there would be more of this kind of animal that survived. But even these assumptions are not enough, for we must also assume that this particular variation was more likely to occur in the descendants of those that had it best developed, and that amongst those forms that survived, some had the same mechanism developed in a still higher degree, and, the process of selection again taking place, a further advance would be made in the direction of autotomy. This, I think, is a fair, although brief, statement of the conventional argument as to how the process of natural selection takes place. But let us look further and see if the results could be really carried out in the way imagined, shutting our eyes for the moment to the number of suppositions that it is necessary to make in order that the change may occur. It will not be difficult, I believe, to show that even on these assumptions the result could not be reached. In the first place, the crabs that are not injured in each generation are left out of account, and amongst these there will be some, it is true, that have the particular variation as well developed as the best amongst those that were injured, and others that have the average condition, but there will be still others that have the possibilities less highly developed, and the two latter classes will be, on the hypothesis, more numerous than those in the first class. The uninjured crabs will also have an advantage, so far as breeding and resisting the attacks of their enemies are concerned, as compared with those that have been injured, and in consequence they, rather than the injured ones, will be more likely to leave descendants. Even if some of those that have been injured, and have thrown off the leg at the most advantageous place, should interbreed with the uninjured crabs, still nothing, or very little, can be gained, because, on Darwinian principles, intercrossing of this sort will soon bring back the extreme variations to the average.

The process of natural selection could at best only bring about the result provided all crabs in each generation lose one or more of their legs, and amongst these only the ones survive that break off the leg at the most advantageous place; but no such wholesale injury takes place, as direct observation has shown. At any one time only a small percentage, about ten per cent, have regenerating legs, and as the time required completely to regenerate a leg, even in the summer, is quite long, this percentage must give an approximate idea of the extent of exposure to injury. It is strange that those who assert off-hand that, because autotomy is a useful process, therefore it must have been acquired by natural selection, have not taken the pains to work out how this could have come about. Had they done so, I cannot but believe they would have seen how great the difficulties are that stand in the way.

A further difficulty is met when we find that each leg of the crab has the same mechanism. If we reject as preposterous the idea that natural selection has developed in each leg the same structure, then we must suppose that a crab varies in the same direction in all its legs at the same time; and if this is true it is obvious that the principle of variation must be a far more important factor in the result than the picking out of the most extreme variations. The same laws that determine that one individual varies in a useful direction farther than do other individuals may, after all, account for the entire series of changes. If it be replied that natural selection does not take into account the causes of the differences of individual variation, this is to admit that it avowedly leaves out of account the very principles that may in themselves, and without the aid of any such supposed process as natural selection, bring about the result. The Lamarckian principle of use and disuse does not give an explanation of autotomy, since the region of the breaking-joint is not the weakest region of the leg, or the place at which the leg would be most likely to be injured.

We cannot assume autotomy to be a fundamental character of living things, since it occurs only under special conditions, and in special regions of the body. While it might be possible to trace the autotomy of the legs of the crustacea, myriapods and insects, to a common ancestral form, yet this is extremely improbable, because the process takes place in only a relatively few forms in each group. The autotomy of the wings of white ants that takes place along a preëxisting breaking-line must certainly have been independently acquired in this group. The breaking off of the end of the foot in the snail helicarion is also a special acquirement within the group of mollusca.

Bordage has suggested that the development of the breaking-joint at the base of the leg of phasmids has been acquired in connection with the process of moulting. He has observed that during this period the leg cannot, in some cases, be successfully withdrawn through the small basal region; and hence, if it could not break off, the animal would remain anchored to the old exoskeleton. It escapes at the expense of losing its leg. The animal, having acquired the means of breaking off its leg under these conditions, might also make use of the same mechanism when the leg is held or injured, and thereby escape its enemy. The fact that the crayfish has a breaking-joint only for the large first pair of legs would seem to be in favor of this interpretation, but the crab has the same mechanism for the slender walking legs, that one would suppose could be easily withdrawn from the old covering. It should also be remembered that we do not know whether the breaking-joint at the base of the leg of the crab and of the crayfish would act at the time when the leg is being withdrawn from the old exoskeleton, unless the leg were first injured outside of the joint.

Our analysis leads to the conclusion that we can neither account for the phenomenon of autotomy as due to internal causes alone in the sense of its being a general property of protoplasm, nor to an external cause, in the sense of a reaction to injury or loss from accident. There would seem then only one possibility left, namely, that it is a result of both together, or in other words, a process that the animal has acquired in connection with the conditions under which it lives, or in other words, an adaptive response of the organism to its conditions of life.

We are not, however, able at present to push these questions farther, for, however probable it may seem that animals and plants may acquire characteristics useful to them in their special conditions of life, and yet not of sufficient importance to be decisive in a life and death struggle, still we cannot, at present, state how this could have taken place in the course of evolution. For, however plausible it may appear that the useful structure has been built up through an interaction between the organism and its environment, we cannot afford to leave out of sight another possibility, viz. that the structure or action may have appeared independently of the environment, but after it appeared the organism adopted a new environment to which its new characters made it better suited. If the latter alternative is true, we should look in vain if we tried to find out how the interaction of the environment brought about the adaptation. The relation would not be a causal one, in a physical sense, but the outcome of a different sort of a relation, viz. the restriction of the organism to the environment in which it can remain in existence and leave descendants.

CHAPTER IX

GRAFTING AND REGENERATION

By uniting parts of the same or different animals, or of plants, there is given an opportunity of studying a number of important problems connected with the regeneration of the grafted parts. Trembley’s experiments in grafting pieces of hydra are amongst the earliest recorded cases of uniting portions of different animals, although in plants the process of grafting has been long known.[77] Trembley found that if a hydra is cut in two, the pieces can be reunited by their cut-surfaces, and a complete animal results. No regeneration takes place where the union has been made. He also succeeded in uniting the anterior half of one individual with the posterior half of another individual, and again produced a single individual. He failed to obtain a permanent union between different species.

More recently, Wetzel has carried out a number of different experiments in uniting pieces of hydra. He found that if two hydras are cut in two, the two anterior pieces may be united by the aboral cut-surfaces (Fig. 46, B), and the two posterior pieces may also be united by their oral cut-surfaces (Fig. 46, A). The fusion of these “like-ends” takes place as readily as when unlike ends are brought in contact, as in Trembley’s experiments. Subsequently, however, regenerative changes take place. When, for instance, two anterior pieces are united by their aboral ends, there develop after two or three days one or two outgrowths, at or near the line of union, that become new feet, and the two individuals may subsequently separate. When two posterior pieces are united by their oral surfaces, a double circle of tentacles generally develops, one on each side of the line of union. The pieces then pinch apart and produce two hydras.[78] In another experiment the head and a part of the foot were cut from a hydra, and the head was turned around and grafted by its aboral surface upon the aboral surface of the middle piece. Another animal was cut in two in the middle, and the posterior half was grafted by its oral end to the oral end of the middle piece. In this way a new, artificial individual was made, as shown in Fig. 46, C, with the middle part of the body in a reverse direction as compared with the orientation of the two end-pieces.[79] The union of the three pieces was so perfect that not even a swelling or a constriction indicated the places of fusion. After six days a normal bud appeared at the region of union of the posterior and middle pieces, that gave rise to a new hydra, which separated after a few days. The compound animal was healthy and ate many daphnias. It was kept under observation for twenty-four days, and appeared normal, giving off several more buds.

In other experiments of this same sort a foot generally developed where the two aboral surfaces came together, and the head-end separated from the rest of the piece. In another case a mouth and tentacles appeared at the place at which the oral ends had united.

In a different kind of experiment, the anterior ends of two hydras were cut off and united by their aboral surfaces; then one of the components was cut in two, just back of the circle of tentacles. After five days two short, hook-like processes appeared at the cut, oral end. They produced a foot, by means of which the animal fixed itself. In this case it will be seen that a foot developed from an oral end. The result might not in itself be considered sufficient to show whether the development of a foot at the oral end of a piece is due to the influence of the other component, or is simply a case of heteromorphosis having no connection with the presence of the other component. Since heteromorphosis has never been observed in isolated pieces of hydra, the probability is that the result is in some way connected with the presence of the other component. Peebles has made a number of experiments, in which special attention was paid to this point. Fifteen anterior pieces were united in pairs by their aboral cut-surfaces, and then one component was cut in half, leaving an exposed oral end. Out of this number five pieces formed a new head at the cut-surface, and the pieces became attached by a foot, that developed at the region of union. Two others did not regenerate at the cut-surface, but became fixed as before, and neither regenerated nor became fixed at the cut-end. Three became attached at the cut, oral surface, but none of these developed a characteristic foot. The result shows, nevertheless, that some influence was present that inhibited the development of a mouth and tentacles at the oral cut-end, since these always develop in isolated pieces. In another series of experiments posterior ends were united by their oral surfaces, and then one of the two pieces was cut in two (Fig. 46, E). A new hypostome and tentacles developed at the region of union, and a foot at the aboral cut-surface, as shown in Fig. 46, E¹. An organism, with one mouth and a circle of tentacles, and two bodies and two feet, resulted. The bodies soon began to fuse together (Fig. 46, E²) into a single one, and when the fusion had extended to the region of the feet, they also fused into a single structure (Fig. 46, E³), so that a single hydra was produced.

In another experiment, twenty-two posterior ends were united in the same way, and then one of the two components was cut in two. In five cases a single head developed on the aboral end of the smaller piece (Fig. 46, D). It is evident in this case that the union of the two pieces has been a factor in bringing about the development of an aboral head. Another of the grafts produced an aboral head, and also one in the region of union. The remaining sixteen grafts produced new heads, if they developed at all, only in the region of union. Peebles states that the regeneration of aboral heads takes place only when one component is cut off near the region of union of the two pieces.

In general, it may be stated in regard to these experiments in hydra that when pieces are united in the same direction, that is, by unlike surfaces, a single individual is formed and no regeneration takes place where the union has been made, but when like surfaces are brought together, although perfect union may result, a process of regeneration takes place later, at or near the line of union. Even the presence of cut-surfaces at one or both cut-ends of the united components does not generally affect the result, although, in a few cases, it may change it, in so far that heteromorphic regeneration may take place from one piece. This sometimes leads to a suppression of regeneration at the line of union. The experiments do not show, perhaps, conclusively whether the heteromorphosis of the smaller component is due to the polarity of the larger component effecting a change in the smaller one, or whether the closing of the oral end of the smaller component (by its union with the other) brings about the result. All things considered, it seems to me that the larger component has directly influenced the other.

King has found that if two posterior pieces of hydra are united by the oral cut-surfaces, and then after they have fused both pieces are cut off near the line of fusion, there develops from the small piece a single hydra, with a foot at one end and tentacles at the other. If only one of the pieces is cut off near the line of fusion, a new head develops from its oral surface, as Peebles had found. If two anterior ends are united by their aboral cut-surfaces, and then later both are cut off near the line of fusion, a single hydra develops from the small, double piece. If one of the components is cut off near the line of union, a foot develops from the oral cut-end. If in any of the cases the cut is made some distance from the line of union, then each cut-surface develops its typical structure. These experiments leave no doubt as to the influence of the larger piece on the smaller one. Especially interesting is the formation of one individual from two short pieces united in opposite directions. In this case we must suppose that one piece has the stronger influence on the combination (perhaps because it is a little larger), and determines the polarization of the other piece.

King finds that when two posterior pieces are united by their oral ends, regeneration of one or of two heads often takes place at the line of union (Fig. 47, B, B¹, B²), as Wetzel had found. If a dark green individual is united to a light green one, it can be seen that in many cases the new heads are formed by both components, as shown in Fig. 47, B¹. Later one of the posterior ends is absorbed, and the halves may then separate (Fig. 47, B¹, B²). If a number of pieces are united, as indicated in Fig. 47, E, a number of heads may be formed, and one or more of these may have a double origin. No evidences of separation of the pieces was observed in cases of this sort.

In one experiment two posterior pieces were united by oblique surfaces, as shown in Fig. 47, C, and one of the two was afterwards cut across, as indicated by the cross-line. The subsequent regeneration that took place is shown in Fig. 47, C¹. A head, composed of parts of both pieces, developed at the cut-surface M, and another in the region N in Fig. 47, C, composed of material of one component. In another case, shown in Fig. 47, D, a head developed only at the cut-edge, but it was made up of material from both components.

A series of grafting experiments of another sort has been made

by Rand. A part of one hydra is grafted upon the side of another one in the following way. A groove is scratched in a film of soft paraffine covering the bottom of a dish filled with water. Another groove is made at right angles to the first one, and opening into it. A hydra (the stock) is placed in the first groove, and a wound made in its side with a knife. Another hydra is cut in two, and one piece (the graft) placed in the other groove, and its cut-surface brought into contact with the wound in the side of the first individual. If the operation is successful the exposed surfaces of the two hydras quickly unite, and the combination may be taken out of the groove. If the piece grafted on the stock included about the anterior half of a hydra, a two-headed animal results, as shown in Fig. 48, B. Although the graft has been united to the side of the stock, it soon assumes an apparently terminal position (Fig. 48, B¹). This is due to the graft sharing with the anterior end of the stock the common basal portion of the stock. A slow process of separation of the two anterior ends now begins, brought about by a deepening of the angle between the halves (Fig. 48, B²). This leads ultimately to a complete separation of the two individuals (Fig. 48, B³, B⁴). Each may get a part of the original foot, or a new foot may arise on the graft as the division approaches the base.

In other experiments only a small part of the foot-end was cut from the animal that served as the graft. The long anterior piece was grafted as before upon the side of the stock. After the two had united, the graft was cut in two, leaving a part of the graft attached to the stock. The part regenerated tentacles, and in two cases subsequently separated from the stock as in the first experiment. In a third case the graft was absorbed by the stock as far as the circle of new tentacles, but its subsequent fate was not determined. In a fourth case the graft did not regenerate its tentacles, and was completely absorbed into the wall of the stock. The smaller the piece that is grafted on the stock the greater the chance that it will be absorbed, and furthermore short, broad rings are more likely to be absorbed than long, tubular pieces of the same volume.[80]

Rand’s results show in general that when hydras are grafted together they regain the typical form in one of two ways,—either by separation into two individuals, or by the absorption of the smaller into the larger component. In the former case the result is brought about in the same way as when the anterior end is partially split in two and the halves subsequently separate. When the graft is absorbed it is not clear whether the absorbed piece disappears or, as seems not improbable, forms a part of the wall of the stock.

It is important to notice the difference between lateral buds and lateral grafts. The buds separate in the course of four or five days by constricting at the base, but this never happens in lateral grafts. Rand has also made some experiments with buds. He cut off the outer oral end of a bud, and grafted it back upon the stock in a new place. It did not separate from the stock as does a bud, but by a slow process of division it was set free in the same way as are lateral grafts. The proximal end of the bud, which was left attached, developed tentacles at its free end, constricted at its base, and was set free. The separation was, however, somewhat delayed.

In another experiment a bud was split in two lengthwise, and the cut was extended so that the body of the parent was separated into two pieces. Twenty-four hours later it was found that each half-bud had closed in, and was much larger than when first cut. The half-bud, that was attached to the posterior end of the anterior piece, was constricting at its base, and subsequently it separated at its point of attachment. The other half of the bud, that had been left attached to the anterior end of the posterior piece, had swung around, so that its long axis corresponded to that of the posterior, parental piece. At first a slight constriction indicated the line of union of the two, but later this disappeared and a single hydra resulted. Whether the difference in the fate of the two half-buds is connected with their different polar relations to the parts of the parent, or is due to some other difference in the absorbing power of the anterior and posterior pieces, is not known.

Tubularia is not so well suited as hydra to show the influence of grafting on the united parts, since pieces of tubularia produce hydranths, both at the oral and aboral ends, although the latter hydranths take longer to develop. Peebles has shown, nevertheless, that grafting has an influence on the behavior of a piece. In order to show that the polarity of a small piece could be affected by a larger piece, the following experiment was carried out. After cutting off the old hydranth from the end of a stem, a short piece was then cut from the distal end of the same stem, turned around, and its oral end brought in contact with the oral end of the original piece, as indicated in Fig. 49, F. The two pieces, being held together for a few minutes, stuck together and subsequently united perfectly. From eighty-eight pieces united in this way the following results were obtained. Thirty-six formed a single hydranth at the end at which the grafting had been made. The distal row of tentacles appeared in the smaller reversed component, the proximal row in the larger piece (Fig. 49, B). The new hydranth pushed out later through the perisarc of the smaller piece (Fig. 49, C). In this experiment the smaller component was shorter than the average length of the hydranth-forming region. In two cases, in which the smaller component was larger, both circles of tentacles appeared in this piece. In six of the experiments the tips of the proximal tentacles arose from a part of the wall of the smaller piece, hence these tentacles had a double origin (Fig. 49, F). In five of the unions the smaller as well as the larger component produced a hydranth; the two were stuck together by their oral ends (Fig. 49, D, E). The remaining four unions gave somewhat different results. In three of these the smaller piece produced only a part of a hydranth that remained sticking to the end of the hydranth formed by the larger component. In the thirty-six cases in which the minor component took part in the formation of the single hydranth, the influence of the larger component was shown not only in reversing the polarity of the smaller component, although this might in part be accounted for by the closing of the oral end of the smaller piece, but also in the time of development, since the hydranth appeared sooner than does the aboral hydranth and at the same time as does the oral hydranth.

In another series of experiments, a short piece was cut from the basal end of a long piece (three to four centimetres) and brought forward and grafted in a reversed position on the anterior end of the same long piece (Fig. 49, A). Of five unions of this sort, one produced a hydranth in each component, neither being reversed. Another of the pieces produced a hydranth partly out of each component (and at the same time another at the aboral end of the large piece). The other two pieces produced a single hydranth, a part of which came from the minor component and appeared before the aboral hydranth on the aboral end of the larger piece. This last result shows that the small piece from the basal end has been affected by the oral end in such a way that it develops more rapidly than it would have done had it remained a part of the basal end.

In a third series of experiments a short piece (about a half of a millimetre) was cut from the anterior end of a long piece (one and five-tenths to two centimetres) and grafted in a reversed position on the posterior end of the same long piece (Fig. 49, G). In four cases a hydranth developed only at the oral end of the long piece and none from the aboral end or from the short piece. Eight unions produced, however, in the region of the graft, a hydranth formed partly by each component. Later another hydranth developed at the oral end of the larger piece. The latter results are not convincing, but they may show that the small piece has hastened the development of the hydranth at the aboral end.

Peebles has also made some experiments in grafting pieces of different members of the colonies of hydractinia and podocoryne. The colony of the former is made up of three different kinds of individuals: the nutritive, the reproductive, and the protective hydroids. A series of preliminary experiments showed that if these individuals are cut into a number of pieces each piece regenerates the same kind of individual as that of which it had been a part. It was also observed that if pieces of the nutritive individuals were allowed to remain quietly on the bottom of the dish they sent out branching stolons, which stuck to the bottom of the dish, and from these stolons there arose later nutritive hydranths that stood at right angles to the surface. When pieces of the same kind of individuals are grafted together, the results are essentially the same as with tubularia. If pieces of different kinds of individuals are united, the opportunity is given of testing the possible influence of one kind on the other. Peebles united a nutritive and a protective polyp by the cut, aboral ends (Fig. 46, E), and after they had grown together one of the polyps was cut off near the region of union, so that a small piece of a nutritive polyp was left attached to a protective polyp. When the piece of the nutritive polyp regenerated, it made a new nutritive polyp. The influence of the protective polyp was not apparent. If a nutritive and a reproductive polyp are united in the same way, and the latter cut in two near the line of union, a new reproductive polyp develops from the piece left attached to the nutritive polyp. Again there is shown no influence of the one on the other kind of polyp.

Hargitt has also made a number of grafting experiments on other hydroids. His most interesting results are those in which parts of two medusæ were united by holding their cut-surfaces together by means of bristles passing through the individuals. Hargitt also finds that while in certain hydroids it is possible to bring about a union of oral with oral end, or aboral with aboral, or oral with aboral end of the same species,[81] yet a permanent union between different species cannot be brought about. These results are in agreement with those of a number of writers who have recorded the difficulty or impossibility of uniting parts of different species of hydra. In a few instances it has been possible to unite temporarily a piece of a brown hydra with a piece of a green one,—as I have also seen accomplished,—yet the pieces subsequently separate. Wetzel succeeded in obtaining better results with two species of brown hydras, Hydra fusca and Hydra grisea. In one experiment the head of Hydra grisea was grafted on the body (from which the head had been cut off) of Hydra fusca. After five hours the pieces seemed to have united. Later a constriction appeared at the place of union, and the head-piece produced a foot near the line of union, and the posterior piece produced a circle of tentacles at its anterior end. Eight days later, when the animal was being killed, it fell apart into two pieces. It was observed that during the period of union a stimulus to one piece was not carried over to the other. Wetzel’s results seem to show that pieces of these two species of hydra unite at first, when brought together, as perfectly as do pieces of the same species, but the union never becomes permanent, a constriction appearing later at the line of union, and the pieces separating in this region. These results indicate, it seems to me, that the factors that bring about the first union are different from those that make the grafted pieces one organic whole. Other results indicate that the union of oral to oral end, or aboral to aboral end, while at first as perfect as between unlike surfaces, nevertheless is less permanent than when unlike surfaces are united; at least, subsequent regeneration is more likely to occur in the former than in the latter, and after this occurs the separation of the individuals often takes place. It seems, moreover, not improbable that a more permanent union results when similar regions are united by unlike surfaces, than when the union is at different levels. If, for instance, the anterior half of one hydra is united to the posterior half of another individual, the union is generally permanent; but if one or both of the pieces are longer than half the length, so that a “long animal” results, new tentacles are more often formed at the oral end of one component, and the parts subsequently separate. It may be that, at present, the data are insufficient to establish this general rule, and no doubt other modifying influences must be also taken into account; but it is important that attention should be drawn to this side of the subject.

Grafting experiments in planarians have so far been carried out in only the two cases which I have described. In one of these the anterior ends of two short pieces of Bipalium kewense were united (Fig. 50, A). Neither piece produced a head at the region of union. Later the pieces were cut apart by an oblique cut that passed across the line of union (Fig. 50, C), so that each piece retained at its most anterior end (at one side) a piece of the other individual in a reversed position. A head developed at the anterior (and lateral) end of each piece, in such a way that a part at least of the small reversed piece was contained in the new head (Fig. 50, D). In the other case two pieces of bipalium were united by their posterior cut-surfaces. Each piece produced a new head at its free end, and the pieces greatly elongated, but remained sticking together (Fig. 51).