THE
EVOLUTION THEORY

VOLUME II

THE
EVOLUTION THEORY

BY

Dr. AUGUST WEISMANN

PROFESSOR OF ZOOLOGY IN THE UNIVERSITY OF FREIBURG IN BREISGAU

TRANSLATED WITH THE AUTHOR'S CO-OPERATION

BY

J. ARTHUR THOMSON

REGIUS PROFESSOR OF NATURAL HISTORY IN THE UNIVERSITY OF ABERDEEN

AND

MARGARET R. THOMSON

ILLUSTRATED

IN TWO VOLUMES

VOL. II

LONDON
EDWARD ARNOLD

41 & 43 MADDOX STREET, BOND STREET, W.

1904

All rights reserved

CONTENTS

LIST OF ILLUSTRATIONS

LECTURE XX

REGENERATION

Budding and division—Every theory of regeneration in the meantime only provisional, a mere 'portmanteau theory'—Regeneration not a primary character—Volvox—Hydra—Vital affinities—Planarians—Heteromorphoses—Enemies of Hydroid-colonies—Regeneration in Plants—In Amphibians—In Earthworms—Different degrees of regenerative capacity according to the liability of the part to injury—Different results of longitudinal halving in Earthworms and in Planarians—Regeneration in Birds—The disappearance of the power of regeneration is very slow—Morgan's experiments on Hermit-crabs—Autotomy in Crustaceans and Insects—Regeneration of the lens in Triton.

We have endeavoured to explain the handing on of the complement of heritable qualities from one generation to another as due to a continuity of the germ-plasm, and we assumed that the germ-cells never arise except from cells in the 'germ-track'; that is, from cells which are equipped, from the fertilized egg-cell onwards, with a complete sample of slumbering germ-plasm, and are thereby enabled to become germ-cells, and, subsequently, new individuals, in which the aggregate of inherited primary constituents implied in the germ-plasm can again attain to development.

We have now to consider other cases of inheritance in relation to the same problem—the origin of their hereditary equipment.

We know, of course, that new individuals may arise apart from germ-cells, that, in many of the lower animals and in plants, they may arise by budding and fission.

For both these cases the germ-plasm theory will suffice, with a somewhat modified form of the same assumption which we made in regard to the formation of germ-cells. The origin of a new individual by budding seems often, indeed, to proceed from any set of somatic cells in the mother animal; but somatic cells, if they contain solely the determinants controlling themselves, cannot possibly give rise to a complete new individual, since this presupposes the presence of all the determinants of the species. But as these determinants cannot be formed de novo, the budding cells must contain in addition to the usual controlling somatic determinants, idioplasm in a latent, inactive state, which only becomes active under certain internal or external influences, and then gives rise to the formation of a bud. The source of this accessory idioplasm must, however, be looked for only in the egg-cell.

In plants this bud-idioplasm must be complete germ-plasm, because the budding starts only from one kind of cell, the cambium-cells; but in animals in which—as it seems—it always proceeds from at least two different kinds of cells—those of the ectoderm and those of the endoderm—the matter is more complex. In this case these two kinds of cells will contain as bud-idioplasm two different groups of determinants, which mutually complete each other and form perfect germ-plasm, and only the co-operation of these two sets will give rise to the formation of a bud. I will not, however, go further into detail in regard to these relations, for the theory can do nothing more here than formulate what has been observed; it is hardly in a position to help us to a better understanding of the facts.

The case is not much clearer in regard to the processes which lead to the replacing of lost parts. The manifold phenomena of regeneration can also be brought into harmony with the theory, if we attribute to those cells from which the replacing or entire reconstruction of the lost part arises an 'accessory-idioplasm,' which, at least, contains the determinants indispensable to the building up of the part. It is possible that the assumed accessory idioplasm frequently contains a much larger complex of determinants, and that it depends on the liberating stimuli which, and how many of these, will become active.

If we take a survey of regenerative phenomena in the animal kingdom, it strikes us at once that the capacity is very different in different species, extraordinarily great in some and very slight in others. In general it is greater in lower animals than in higher, but, nevertheless, the degree of differentiation cannot be the only factor that determines the capacity for regeneration. That unicellular organisms can completely replace lost parts, that even a piece of an infusorian can reconstruct the whole animal if only the piece contain a part of the nucleus, we have already seen when discussing the significance of the nuclear substance. In this case the nucleus must contain the complete germ-plasm, that is, the collective determinants of the species, and these induce the reconstruction of the lost part, though they do so in a way that is still entirely obscure to us. In the meantime, our interpretation will not carry us further, either here or in regard to any other order of vital phenomena. To go further would be little short of propounding a causal theory of life itself; it would mean having a complete and real 'explanation' of what 'life' is. As yet no one has been able to claim this position. We can see the different stages through which every organism passes, and that they arise one out of the other; we can even penetrate down to the succession of those delicate and marvellously complex processes which effect nuclear and cell-division; but we are still far from being able to deduce, except quite empirically, from the present state of a cell what the succeeding one will be, that is, from being able to understand the succession of events as a necessary nexus which could be predicted. How a biophor comes to develop from itself the phenomena of life is quite unknown to us; we know neither the interaction of the ultimate material particles nor the forces which bring it about; we cannot tell what moves the hordes of different kinds of biophors to range themselves together in a particular order, what molecular displacements and variations arise from this, or what influence the external world has, and so forth. We see only the visible outcome of an endless number of invisible movements—growth, division, multiplication, reconstruction, and differentiation.

As long as we are so far from an understanding of life no theory of regeneration can be anything more than a 'portmanteau theory,' as Delage once expressed himself in relation to the whole theory of inheritance, a theory which is like a portmanteau in that one can only take out of it what has previously been put in. If we wish to explain the renewal of the aboral band of cilia in a Stentor, we first pack our trunk, in this case the nucleus of the Infusorian, with the determinants of the ciliated region, and then think of these as being liberated by the stimulus of wounding, and being brought to and arranged in the proper place by unknown forces to reconstruct the ciliary region in some unknown way. No one could be more clearly aware than I am that this is not an exhaustive causal explanation of the process itself. Nevertheless, it is not quite without value, inasmuch as it allows us at least to bring the facts together in rational order—in this case the dependence of the faculty of regeneration on the presence of nuclear substance—under a formula which we can use provisionally, that is, with which we can raise new questions. As soon as we ascend higher in the series of organisms the theory gains a greater value, for, while we leave altogether out of account any answer to the ultimate question, and thus renounce for the present the attempt to find out how the determinants set to work to call to life the parts which they control, we are brought face to face with other, in a sense, preliminary questions which we can solve, and the solution of which seems to me at least not entirely without value.

The first of these questions runs thus: Is the power of regeneration a fundamental, primary character of every living being in the sense that it is present everywhere in equal strength, independently of external conditions, and thus is an inevitable outcome of the primary characters of the living substance? Or is it, though primaeval in its beginnings, a phenomenon of adaptation, which depends on a special mechanism, and does not occur everywhere in equal extent and potency?

We have already become acquainted with some facts which must incline us to the latter view. The globular Alga-colonies of Volvox ([Fig. 63]) consist of two kinds of cells, of which only one kind, the reproductive cells, possess the power of reproducing the whole, the others, the flagellate, or, as we called them, somatic cells, being only able to produce their like, but never the whole.

New investigations which have been carried out by Dr. Otto Hübner in my Institute have placed these facts beyond doubt. We may conclude that, in this case, a disintegration of the germ-plasm has taken place during ontogeny, by means of differential cell-division, so that only the reproductive cells receive the complete germ-plasm, while the somatic cells receive only the determinants necessary to their own specific differentiation, the somatic determinants.

In this case regeneration and reproduction coincide; there is no regeneration except the origin of a new individual from a reproductive cell.

Fig. 35 B (repeated). Hydra viridis,
the Green Freshwater Polyp.
Section through the body-wall,
somewhere in the direction of ov
in Fig. 35 A. Eiz, the ovum lying in
the ectoderm (ect), and including
zoochlorellæ (schl) which have immigrated
from the endoderm (ent)
through the supporting lamella
(st). After Hamann.

Let us now ascend to the lowest of the Metazoa, for instance, the freshwater polyp, Hydra ([Fig. 35 A]), and we find a high degree of regenerative capacity in the restricted sense, for, in addition to the power of producing germ-cells, that is, cells which, when two combine in amphimixis, give rise again to a new animal, almost any part of the polyp can regrow a whole animal. Not only has Hydra been cut in from two to twenty different pieces, but it has even been chopped up into innumerable fragments, and yet each of these, under favourable circumstances, was able to grow again into a complete animal. Nevertheless, we are not justified in concluding that every cell possesses the power of reproducing the whole. If, with the help of a bristle, we turn one of these polyps outside in like the finger of a glove, and then prevent it turning right again by sticking the bristle transversely through it, it does not live, but soon dies, obviously because the cells of the two layers of the body, ectoderm and endoderm, cannot mutually replace each other, and cannot mutually produce each other. The inner layer, now turned outwards, cannot resist the influence of the water, and the outer layer, now turned inwards, cannot effect digestion; in short, one cannot be transformed into the other, and we must therefore conclude that both are specialized, that they no longer contain the complete germ-plasm, but only the specific determinants of ectoderm and endoderm respectively.

The animal's high regenerative capacity must therefore depend on the fact that certain cells of the ectoderm are equipped with the complete determinant-complex of the ectoderm, in the form of an inactive accessory idioplasm, which is excited to regenerative activity by the stimulus of wounding, and that, in the same way, the cells of the endoderm are equipped with the whole determinant-complex of the endoderm. It need not be decided whether all or only many of the cells, perhaps the younger ones, are thus adapted for regeneration; in any case a great many of them must be distributed throughout the whole body, with perhaps the exception of the tentacles, which are by themselves unable to reproduce the whole animal. When the animal is mutilated, the cells of both layers, equipped with their respective determinant-aggregates, co-operate in reproducing the whole from a part.

It is true that even with these assumptions we only reach the threshold of a real explanation. For, given that all the determinants of the species must be present in a fragment, we are not in a position to show how these set about reconstructing the animal in its integrity, and the most that we can say is, that it must depend on the specific kind of stimulus to which each of the cells is exposed through its direct and more remote environment, which determinants are to be first liberated, and therefore which parts are to be reconstructed.

That there are at work regulative forces, such as we were already compelled to assume in regard to the division and regeneration of unicellular organisms, as to the nature of which we cannot yet make any definite statement, but which we may call 'polarities,' or, as I prefer to say, 'affinities,' is shown by countless experiments which have been made, particularly with the freshwater polyp. Thus Rand cut off the anterior end of the polyp with its circle of tentacles, and the excised disk of living substance lengthened in a transverse direction, so that half the tentacles came to lie to the right, the other half to the left, while the body developed between these two groups, so that they became further and further separated from each other, till finally the original transverse axis of the animal became the longitudinal axis. One group of tentacles survived and surrounded the new mouth, while the other at the opposite aboral pole, the new foot, died off. This total change of structure in the polyp, as to the arrangement of its main parts, points to unknown forces, which cannot depend on the determinants as such, but on the vital characters of the living parts, and on the interactions of these with one another.

Fig. 96. A Planarian cut transversely into nine pieces. The regeneration of seven of these into entire animals is shown. After Morgan.

The same holds true of all the lower Metazoa that have highly developed regenerative capacity, not only of polyps, but of worms such as the Planarians. Through the experiments of Loeb, Morgan, Voigt, Bickford, and others, we know that these animals respond to almost every mutilation by complete reconstruction, that they may, for instance, as is indicated in Fig. 96, be cut transversely into nine or ten pieces with the result that each of these pieces grows again to a whole animal, unless external influences are unfavourable and prevent it.

Something similar happens if the head be cut off a Tubularia-polyp, it forms a new head with proboscis and tentacles. It does so, at least, if the stalk of the polyp be left in the normal position; but if it be stuck into the sand in the reverse position a head arises at the end which is uppermost, where the roots arose previously, and the previous head-end now sends out roots. By suspending a beheaded stalk horizontally in the water a head can be caused to develop at each end of the stalk, so that we must assume that every part of the polyp is, under some circumstances, capable of developing a head, and that it must be 'circumstances'—in this case gravity, contact with earth or with water, and the mutual influence of the parts of the animal upon each other—which decide what is to be produced. Loeb, who was the first to observe this form of regeneration, called it heteromorphosis, to express the fact that particular parts of the animal might be produced at quite different places from those originally intended for them.

It would certainly be erroneous to range these cases of heteromorphosis against the determinant theory, but they certainly do not afford any special evidence of its validity as an interpretation, for all that we can say here again is that all, or at least many, cells of the animal must contain the full determinant-complex of the ectoderm, and others those of the endoderm, and that particular groups of determinants become active when they are affected by certain external or internal liberating stimuli. In regard to such animals the theory is hardly more convincing than the rival theory, that the faculty of regeneration is a general property of living substance, which does not attain to equally full expression everywhere, because it is met by ever-increasing difficulties involved in the increasing complexity of structure. The validity of the theory only begins to be seen when we deal with cases where it is demonstrable that every part cannot bring forth every other, where the power of regeneration is limited, and occurs only in definite parts in a definite degree, and can only start from particular parts. Here the assumption of a general and primary regenerative capacity fails. Any one who insists, as O. Hertwig does, that the idioplasm in all cells of the body is the same, can always plead that, in the cases in which regeneration does not occur, the fault lies, not in the regenerative capacity, but in the absence of the adequate liberating stimuli, and at first sight it does seem as if this position were unassailable. We shall find, however, that there are facts which make Hertwig's interpretation quite untenable.

My own view is that the regenerative capacity is not something primary, but rather an adaptation to the organism's susceptibility to injury, that is, a power which occurs in organisms in varying degrees, proportionate to the degree and frequency of their liability to injury. Regeneration prevents the injured animal from perishing, or from living on in a mutilated state, and in this lies an advantage for the maintenance of the species, which is the greater the more frequently injuries occur in the species, and the more they menace its life directly or indirectly. A certain degree of regenerative capacity is thus indispensable to all multicellular animals, even to the highest among them. We ourselves, for instance, could not escape the numerous dangers of infection by bacilli and other micro-organisms if our protective outer skin did not possess the faculty of regeneration, at least so far that it can close up a wound and fill up with cicatrice-tissue a place where a piece of skin has been excised. Obviously, then, the mechanism which evokes regeneration must have been preserved in some degree and in some parts at every stage of the phyletic development, and must have been strengthened or weakened according to the needs of the relevant organism, being concentrated in certain parts which were much exposed to injury and withdrawn from other rarely threatened parts. Thus the great diversity which we can now observe in the strength and localization of the regenerative capacity has been brought about. But all this can only be regarded as adaptation.

I should like to submit a few examples to show that the regenerative capacity is by no means uniformly distributed, and that, as far as we can see, it is greater or less in correspondence with the needs of the animal, both in regard to the whole and to particular parts.

It must first be pointed out that those lower Metazoa, like the Hydroid polyps in particular, which are endowed with such a high and general power of regeneration, do actually require this for their safety; they are not only soft, easily injured and torn, but they are most severely decimated by many enemies. In the beginning of May I found on the walls of the harbour at Marseilles whole forests of polyp-stocks of the genera Campanularia, Gonothyræa, and Obelia, all large and splendidly developed, with thousands of individual polyps and medusoids, but in a very short time the great majority of the polyps were eaten up by little spectre-shrimps (Caprellids) and other crustaceans, worms, and numerous other enemies, and towards the end of May it was no longer possible to find a fine well-grown colony. It must therefore be of decisive importance for these species if the stems and branches, which are spared because protected by horny tubes, possess the faculty of transforming their simple soft parts into polyp-heads, or of giving off buds which become polyps, or even of growing a new stock from the twigs which have been half-eaten and bitten loose from the stock and have fallen to the ground. If, finally, a torn-off polyp-stalk (of Tubularia) falls to the ground with the wrong side up, the end which is now the lower will send out roots, and the end now uppermost will give off a new head. This also appears to us adaptive, and does not surprise us, since we have been long accustomed to recognize that what is adapted to an end will realize this if it be possible at all. Think again of the innumerable adaptations in colour and form which we discussed in the earlier lectures. I hope later to be able to show in more detail how it comes to pass that necessity gives rise to adaptation. In regard to the case of the polyps, we can understand that, as far as a high degree of regeneration and budding was possible in these animals at all, it could not but be developed. Regeneration and budding complete each other in this case, for the former brings about in the individual 'person' what the latter does in the colony, namely, a Restitutio in integrum. It is readily intelligible that the former was not difficult to establish where the latter—the capacity of budding—was already in existence.

It seems at first sight very striking that the higher plants, which all depend upon budding, and which form plant-colonies (corms) in the same sense as the polyps form animal-colonies, only possess the faculty of true regeneration in a very low degree, although they are extremely liable to injury.

We see from this that the two capacities are not co-extensive, that germ-plasm may be contained in numerous cells of the body in a latent state, and yet that regeneration of each and every detailed defect may not be possible. This is the case in the higher plants in regard to most of their parts. A leaf in which a hole has been cut does not close the hole with new cell-material; a fern frond from which some of the pinnules have been cut off does not grow new ones, but remains mutilated. Even leaves which, if laid on damp earth, readily give off buds which grow to new plants, as the Begonias do, do not replace a piece cut out of the leaf; they are not at all adapted to regeneration.

From the standpoint of utility this is readily intelligible. It was, so to speak, not worth Nature's while to make such adaptations in the case of leaves or blossoms, partly because these are very transient structures, and partly because they are rapidly and easily replaceable by the development of others of the same kind. Moreover, the leaf in which we have cut a hole continues to function, but the polyp whose mouth and tentacles we have cut off could no longer take nourishment unless it were adapted for regeneration. But that this adaptation could have been made in the case of plants is proved by the root-tips which are formed anew when they are injured, and the closing of wounds on the stem by a 'callus.'

I shall return to plants when we are dealing with the mechanism of regeneration, but I must now direct more attention to animals, inquiring further into the question as to whether the faculty of regeneration is correlated with the degree of liability to injury to which the animal is exposed, and with the biological importance of the injured part, for this must be the case if regeneration be really regulated by adaptation.

Hardly any other vertebrate has attained such celebrity on account of its high regenerative capacity as the water-newt, species of the genus Triton. It can regrow not only its tail, but the legs and their parts if they are cut off. Spallanzani saw the legs grow six times, after he had cut them off six times. In the blind newt (Proteus) of the Krainer caves, a near relative of the common newt, the leg regenerated only after a year and a half, although the animal stands on a lower stage of organization than the newt, and thus should rather replace lost parts more easily. But Proteus lives sheltered from danger in dark, still caves, while Triton is exposed to numerous enemies which bite off pieces from its tail or legs; and the legs are its chief means of locomotion, without which it would have difficulty in procuring food. It is different with the elongated eel-like newt of the marshes of South Carolina, Siren lacertina. This animal moves by wriggling its very muscular trunk, after the manner of an eel, and in consequence of the disuse of its hind legs it has almost completely lost them. Even the fore-legs have become small and weak, and possess only two toes, and these do not regrow if they are bitten off, or only do so very slowly.

Earthworms are exposed to much persecution; not only birds, such as blackbirds and some woodpeckers, but, above all, the moles prey upon them, and Dahl has shown that moles often lay up stores of worms in winter which they have half crippled by a bite, while even Réaumur knew that moles frequently only half devoured earthworms. It was thus an obvious advantage to earthworms that a part of the animal should be able to regrow a whole, and accordingly we find a fairly well-developed regenerative capacity among them. But it varies greatly in the different species, and it would be interesting if we knew the conditions of life well enough to be able to decide whether the faculty of regeneration rises and falls in proportion to the dangers to which the species is exposed. Unfortunately we are far from this as yet; we only know that, in the common earthworms of the genera Lumbricus and Allolobophora, the faculty of regeneration is still very limited, for at most two worms, and sometimes only one, can develop from an animal cut into two pieces. Cutting into a greater number of pieces does not yield a larger number of worms, but usually only one, and often none at all.

This corresponds to the behaviour of their enemies, which may often bite off a piece or tear it away when the worm attempts to escape, but never cut it up into pieces. The regenerative capacity is more highly developed in the genus Allurus, more highly still in the worms of the genus Criodrilus which lives in the mud at the bottom of lakes, and most highly of all in the genus Lumbriculus which lives at the bottom of small ponds. Long ago Bonnet cut up a specimen of Lumbriculus into twenty-six pieces, of about two millimetres in length, and he observed most of these grow to complete worms again. His experiments have often been repeated in recent times, and have been extended and made more precise in many ways. Von Bülow was able to get whole animals from pieces consisting of from four to five somatic segments, and with eight or nine segments he almost invariably succeeded. A Lumbriculus which he had cut into fourteen pieces, one of which only measured 3.5 mm. in length, gave rise to thirteen complete worms with head and tail; only one piece perished.

These worms have little enemies with sharp jaws which may gnaw at them behind or before but cannot swallow them whole. Lyonet, famous for his analytic dissection of the wood-caterpillar (Cossus ligniperda), observed when he was feeding the larvæ of dragon-flies with these Lumbriculid worms that 'the anterior end of some whose posterior end had been gnawed away by the larvæ continued to live on the ground.' We can thus understand why a high power of regeneration is of use to these worms, and at the same time why it is advantageous to them to contract so that they break in pieces on very slight irritation, but to this we shall refer again.

The very diverse potency of the faculty of regeneration in animals belonging to the same small group, and nearly, if not quite, at the same level of organization, seems to show clearly that we have here to do with adaptation to different conditions of life, although we cannot demonstrate this in detail. It would certainly be erroneous to regard the conditions of life as uniform, since the worms in question not only live in different places—in the earth, in mud, or in water—and are thus exposed to different enemies, and since they may also be quite different in regard to size and speed, in means of defence, and possibly also of defiance, as is indeed in some measure demonstrable.

We meet with the same thing in a group of still smaller worms, Rösel's 'water-snakelets,' species of the genus Nais. These, too, behave in a variety of ways in the matter of regeneration, for while many species, such as Nais proboscidea and Nais serpentina will, if cut into two or three pieces, become two or three worms respectively, Bonnet expressly mentions an unnamed species of Nais which does not bear cutting up at all, and even dies if its head be cut off.

Thus neither the degree of organization nor the relationship alone determines the strength of the regenerative capacity. And as nearly related species may behave quite differently in this respect, so also do the different parts of one and the same animal; and here, too, the strength of the capacity seems to depend on the more frequent or rarer injury of the relevant part and on its importance in the maintenance of life. Let us take a few examples.

Parts which, in the natural life of the animal, are never injured, show in many cases no power of regeneration. This is so in regard to the internal parts of the newt, whose regenerative capacity is otherwise so high. I cut half or nearly the whole of a lung away from newts anæsthetized with ether; the wound closed, but no renewal of the organ took place. The same thing happened when a piece of the spermatic duct or of the oviduct was cut away. It is true that the kidney enlarges in higher animals when a piece has been cut out, by the proliferation of the remaining tissues, but that is a mere physiological substitution, evoked by the increased functional stimulus, due to the accumulation in the blood of the constituents of the urine. Such substitution depends on the growth of parts already existing, and it occurs in man when one kidney is removed, for the other, as is well known, may then grow to double its normal size. This is mere hypertrophy of the part that is left, it is not regeneration in the morphological sense, and it is not comparable to the re-formation of a cut-off leg in the salamander, or of a head in the worm, where the growth is not a mere increase of the remaining stump, but a new formation. It would be regeneration if a new kidney developed from the remnants of the kidney-tissue, or, in the liver, if new lobes grew in place of those which were cut off. But neither of these things happens, and, as far as I am aware, nothing of the kind has ever been observed, nothing more than new formation of liver-cells through increase of existing ones; that, however, is not regeneration in the morphological sense.

I have referred to the slight power of regeneration possessed by the blind Proteus in regard to its legs or tail, and I connected this with the absence of enemies in its thinly peopled cave-habitat. But the same animal can regenerate its gills when these are bitten off, and this is probably associated with the habit that Proteus has, in common with other newts with external gills, of nibbling at its neighbour's gills. Thus, the power of regenerating the gills was retained even when the animals migrated to the quiet caves of Krain, and were thus secured from the attacks of other enemies.

In lizards, a leg which has been cut off does not grow again, but an amputated tail does, and this has quite a definite biological reason, since the active little animal will seldom be caught by the foot by any pursuer, but may easily be caught by the tail, which is far behind. Thus the tail is adapted not only for regeneration, but also for 'autotomy' that is, for breaking off easily when it is caught hold of.

We have already seen that some segmented worms have a very high regenerative capacity; yet every part cannot produce every other, and while, in Lumbriculus, any piece of from five to nine segments is able to grow a new head or tail, neither ten nor twenty nor all the segments together, if they are halved longitudinally, can reproduce the other half, and the cause of this inability does not lie in the fact that the animal is thereby hindered from taking food, for even the transversely cut pieces do not feed until they have grown a new head and tail. The reason must lie in the fact that the primary constituents for this kind of regeneration are wanting, and they are so because a longitudinal splitting of this cylindrical and relatively thin animal never occurs under natural conditions, and thus could not be provided against by Nature[1].

[1] Morgan maintains that this statement is incorrect, and that Lumbriculus is capable of lateral regeneration. But if we look into the matter more closely we find that all he says is, that small gaps made by cutting a piece out of one side are filled up again, while the cut pieces perish. If the whole animal be halved, according to Morgan, both halves die, or if a 'very long piece' be cut out of one side, not only this piece dies, but also 'the remaining piece.' There is thus, as I have said, an essential difference between the regenerative capacity of Lumbriculus and that of Planaria.

That regeneration of this kind could have been arranged for if it had been useful we learn from the Planarians among the flat worms, in which every piece cut out of the body, large or very small, from the middle, from the left side, or from the right side of the animal, grows into a complete Planarian. The animal can be halved longitudinally, as in Fig. 97, and each half will grow to a whole. This again is quite intelligible from the biological point of view, for these flat, soft, and easily torn animals are exposed to all sorts of injuries, and are, in point of fact, frequently mutilated by enemies which are unable to swallow them whole. Von Graaf not infrequently found examples of marine Planarians (Macrostomum) which lacked 'a part of the posterior end or the whole tail region as far as the food-canal,' and of species of Monotus he found 'very often' in May specimens with the posterior end split or broken off. Probably the persecutors of these flat-worms are some species of Crustacean, but, at any rate, so much is proved, that the Planarians have abundant opportunities of making use of their faculty of regeneration, and that the species gains an advantage from it in respect to its preservation.

Fig. 97. A, a Planarian,
which has been divided into
two by a longitudinal cut.
Each half can grow into
an entire animal. B, the
left half at the beginning
of the regenerative process.
C, the same completed. After
Morgan.

In contrast to this, worms which live within other animals, and are thus secure from mutilation, such as the familiar round-worms (Nematoda), have no power of regeneration at all, and do not survive either longitudinal or transverse division.

Until recently birds were regarded as possessing a very low degree of regenerative capacity, and, as a matter of fact, they cannot replace a leg or a wing wholly or in part; but, what is otherwise unheard of among higher vertebrates, they can renew the whole anterior portion of the skeleton of the face, the bill, and can indeed almost reconstruct it with new bones and horny parts. Von Kennel communicated a case of this kind in regard to a stork, and for a long time this remained an isolated case, but a few years ago Bordage showed that, in the cocks which are used in the Island of Bourbon for the favourite sport of cock-fighting, the bill is regularly renewed when it has been broken off or shattered. Quite recently Barfurth gave an account of a case of complete renewal of a broken bill in a parrot. Yet it should not astonish us that the bill in birds has such a high regenerative power, for of all parts in a bird it is the one that is most readily injured; with it the bird defends itself against its enemies and its rivals, masters its prey, and tears it to pieces, pecks holes in trees (woodpecker), or climbs (parrot), or digs and burrows in the ground, or builds its nest, and so on. That the faculty of regeneration could be developed to so high a degree in relation to this particular part of the body, while the rest of the very important but rarely injured parts do not possess it at all, again points to the conclusion that the faculty of regeneration has an adaptive character.

It does not affect matters to discover cases in which we cannot recognize this relation between the regenerative capacity of a part and its importance or its liability to injury. Such instances do not lessen the convincingness of the positive cases, since we do not know the exact conditions which may lead to the increase of regenerative capacity in a part, and, above all, since we do not know the rate at which such an increase may take place. If adaptation in general depends upon processes of selection, these processes must also be able to give rise to an increase in the power of regeneration. On the other hand, it by no means follows that the disappearance of a faculty of regeneration which was once present in a part, but which has become superfluous in the course of time, must take place immediately through natural selection. For it is the very essence of natural selection that it only furthers what is useful, and only removes what is injurious; over what is indifferent it has no power at all. Thus it follows that the faculty of regeneration, when it has once been present in a part, cannot be set aside by natural selection (personal selection), for it is in no way injurious to its possessor. If it gradually decreases and becomes extinct notwithstanding this, when it is of no further use, as seems to be to some extent the case in regard to the legs and tail of the blind Proteus, that must depend on other processes, on those which generally bring about the gradual disappearance of disused parts or capacities. We shall attempt to probe to the roots of these processes later on; for the present let it suffice us to know that, according to our experience, they go on with exceeding slowness, and that it has taken whole geological periods to eliminate the legs of the snake-ancestors so completely as has been done from the structure of most of our modern snakes, while the Proteus which migrated into the caves of Krain as far back as the Cretaceous period is indeed blind, but still retains its eyes under the skin, though in a degenerate condition.

Since the degeneration of disused parts and capacities goes on so slowly it need not surprise us that we meet many parts which still possess regenerative capacity, although they are protected from injury. Thus Morgan found that, in the hermit-crab, the limbs which are protected within the mollusc shell were quite as ready to regrow as those which are actually used for walking, and thus are exposed to possibility of attack, but this proves nothing against the conclusion we drew from the facts cited above, according to which the faculty of regeneration comes under the law of adaptation. For the disappearance of this faculty must take place very much more slowly than its growth. For instance, the development of the tail-fin of the whale has long been an accomplished fact, while the hind-legs of this colossal mammal, which were rendered useless by the development of the tail-fin, still lie concealed in a rudimentary state within the muscles of the trunk. Yet these limbs must have lost their significance for the animal exactly at the time that the tail-fin became more powerful. Thus the retrogression must have taken place more slowly than the progressive transformation.

It is clear, then, that the faculty of regeneration is not a primary character of living beings occurring uniformly in all species of equally high organization and in all parts of an animal in the same degree; it is a power which occurs in animals of equal complexity in as varying degrees as in their parts, and which is manifestly regulated by adaptation. Between parts with the faculty of regeneration and parts without it there must be an essential difference; there must be present in the former something that is wanting in the latter, and, according to our theory, this is the equipment with regeneration-determinants, that is, with the determinants of the parts which are to be reconstructed.

If this be really so it should be capable of proof, at least in so far that we should be able to establish that the power of completing or re-forming a damaged or lost part is a limited one, localized in certain parts and cell-layers. This can be actually proved, as may be seen from numerous cases in which the faculty of regeneration is associated with autotomy, that is, with the power of breaking off or dropping off a part of the body. Even in worms we find this power, as we mentioned before in speaking of the high regenerative capacity of Lumbriculus. This worm reproduces in summer by what is called 'schizogony,' that is, by breaking into two, three, or more pieces, and it does not seem to require a very strong stimulus, such as pressure of the end of the worm by the jaws of an insect larva, to start this rupture; it often follows from quite insignificant friction on the ground. Certainly the power of regeneration is so great in this animal that it is out of the question to talk of localizing the primary constituents of regeneration; almost every broken surface is capable of regeneration.

But this localization is well illustrated in Insects and Crustaceans, which possess the power of self-amputation in their appendages, especially in their legs. As far back as 1826 MacCullock observed this remarkable power in crabs, and described the mechanism on which it depends. When the leg is irritated, for instance when it is pinched at the tip and held fast, it breaks off at a particular place. This line of breakage lies in the middle of the short second joint (Fig. 98, A and B, s), just between the insertions of the muscles (me, mf, m) which extend from this line towards the extremity of the limb and in the opposite direction towards the body-wall. Between these muscle-attachments the external skeleton is thin and brittle, and forms a suture, s, which breaks through when the animal contracts the muscles of the leg convulsively, and thus presses the lower protuberance (a) against a projection (b) of the first upper joint. Crabs require to make a very considerable muscular exertion before they can throw off the limb, and therefore they can only do it when they are in full vigour.

Fig. 98. The leg of a Crab,
adapted for self-mutilation or autotomy.
A, the first three joints of the limb, I, II, III. s, the
suture, that is, a thin area on
the second joint which is predisposed
to breakage. mf, flexor
muscle, me, extensor muscle, both
inserted at the suture. B, the
entire leg with its six joints and
with the suture (s). Slightly enlarged.
After MacCullock.

We have here a quite definite structural adaptation of the parts to a danger which often recurs—that of falling entirely into the power of an enemy which has seized the leg. By a sudden violent throwing-off of the leg the crab escapes from this danger. Quite similar adaptations are found among certain insects, such as the walking-stick insects or Phasmids, in which the mechanism is much the same, and lies at an almost exactly corresponding place, namely, at the line where the second and third joints of the leg, the 'trochanter' and the 'femur' meet. In this case the advantage of the arrangement is not merely that the animals are thus enabled to escape from enemies; it is useful in another connexion, for a knowledge of which we have to thank Bordage. This naturalist observed that the Phasmids not infrequently perished at one of their numerous moultings, by remaining partially fixed in the discarded husk. Of 100 Phasmids nine died in this way, twenty-two got free with the loss of one or more legs, and only sixty-nine survived the moult without any loss at all.

That the moulting or ecdysis of insects is often hazardous may be observed in our own country, and it is familiar to every one who has reared caterpillars. These, too, often fail to get clear of their 'cast' cuticle, and they perish unless artificial aid is given to them. I have never observed any autotomy in them, but in the Phasmids it seems to be a much-used 'device' and is therefore of great importance in the persistence of the species.

Limbs which are thus thrown off by autotomy regenerate again from the place at which they broke off, that is from the 'suture.' It had been noticed even by the earlier observers (e.g. Goodsir) that there was a jelly-like mass of cells within the joint, and that the development of the new limb started from this. It might be supposed that the regeneration-primordium is present in the rest of the leg also, but that is not the case, for the animal responds to the tearing off of one joint or of a smaller number than to the suture, not by regenerating the torn part directly, but by amputating the whole of the leg up to the suture, and then from this the regeneration of the whole leg takes place. In the Phasmids the case is similar, but with the difference that regeneration is possible from three places, from the tarsal joints, from the lower third of the tibia, and finally, from the suture between the femur and the trochanter. There is thus a regeneration-primordium (Anlage) at the beginning of the tarsal joints, another in the tibia, and a third in the 'suture' and the first must be equipped, as we should express it, with the determinants of the five tarsal joints, the second with those for the lower end of the tibia as well, and the third with all the determinants of the whole leg, from the 'suture' downwards.

In any case, regeneration is here associated with definite localized pieces of tissue, and is not a general character of all the cells of the leg, and, as it obviously runs parallel at the same time with another adaptation—that of autotomy—there can be no doubt that it too is dominated by the principle of selection, and that it can not only be increased, but that it can be concentrated at particular places and removed from others. But this is only possible if it be bound up with material particles which may be present in or absent from a tissue, and which are therefore a supplement to the ordinary essential constituents of the living cells, although they do not themselves belong to the essential organization.

I might cite many more examples of localization of regenerative capacity, but will confine myself to one other, which seems to me particularly instructive, because it was first interpreted as an indication of the existence of an adaptive principle in the organism, a principle which always creates what is useful. I refer to the regeneration of the lens in the newt's larva.

G. Wolff, an obstinate opponent of the theory of selection, attempted to solve the same problem as I had before me in my experiments on the regeneration of the internal organs of newts, that is, he tried to answer the question whether organs which are never exposed to injury or to complete removal in the conditions of natural life, and which could not therefore have been influenced in this direction by the processes of selection, are nevertheless capable of regeneration. He extirpated the lens from the eye of Triton larvæ, and saw that in a short time it was formed anew, and from this he concluded that there was here 'a new adaptiveness appearing for the first time,' and that therefore adaptive forces must be dominant within the organism. The current theory of the 'mechanical' origin of vital adjustments seemed to some to be shaken by this, and the proclamation of the old 'vital force' seemed imminent. And in truth, if the body were really able to replace, after artificial injury, parts which are never liable to injury in natural conditions, and to do so in a most beautiful and appropriate manner, then there would be nothing for it but at least to regard the faculty of regeneration as a primary power of living creatures, and to think of the organism as like a crystal, which invariably completes itself if it be damaged in any part. But we have to ask whether this is really the case.

What makes the regeneration of the lens seem particularly surprising is the fact that in the fully formed animal it must arise in a manner different from that in which it develops in the embryo, that is, it must be formed from different cell-material. In the embryo it arises by the proliferation and invagination of the epidermic layer of cells to meet the so-called 'primary' optic vesicle growing out from the brain—a mode of development which cannot of course be repeated under the altered conditions in the fully developed animal. The reconstruction of the organ must therefore take place in a different way, and if the organism were really able, the very first time the lens was removed, to react in a manner so perfectly adapted to the end, and so to inspire certain cells, which had till then had a different function, that they could put together a lens of flawless beauty and transparency, we should have reason to suspect that nearly all our previous conceptions were erroneous, and to fall back upon a belief in a spiritus rector in the organism.

But the excision of the lens in these experiments was not by any means an unprecedented occurrence! It is true enough that newts in their pools are not liable to an operation for cataract, but it does not follow that the lens is never liable to injury, and could not therefore be adapted for regeneration. It can be bitten out along with the rest of the eye by water-beetles or other enemies, and as far back as the time of Bonnet and Blumenbach (1781) it was known that the eye of the newt would renew itself if it were cut out, given that a small portion of the bulb was left. But if this were removed the possibility of regeneration was at an end. Thus, before the first artificial excision of the lens, a regeneration-mechanism must have existed, by means of which the eye with its lens was reconstructed, and this depends on the characters of the cells of the eye itself—it is localized in the eye, and without the presence of a piece of eye-tissue no regeneration can take place. Is it then so especially remarkable that the lens should be renewed when it is artificially removed without the rest of the eye? The mechanism for its renewal is there, and is roused to activity whether the lens alone or other parts of the eye also be removed. We do not need, therefore, to assume the existence of a purposeful or adaptive force; it is more to the point to inquire where the regeneration-mechanism which suggests this inference is to be found.

A definite answer to this is given in a detailed experimental work recently published by Fischel. It corroborates what Wolff had already found, that the substance of the new lens develops from cells which cover the posterior surface of the iris, that is, from cells of the retinal layer of the eye. First, the margin of the pupil begins to react to the stimulus of the injury (extraction of the lens); its cells enlarge, become clear, while previously they were filled with dark pigment, and finally they proliferate. They thus form a cell-vesicle similar to the ectoderm-vesicle from which the lens arises in the embryo, and into this the already mentioned retina-cells from the posterior wall of the iris grow, elongate, and arrange themselves to form the so-called 'lens-fibres,' on whose form, arrangement, and transparency the function of the lens depends. This is marvellous enough, but not more marvellous than that a whole foot should grow on the cut stump of a newt's leg, or that a whole eye should arise from a residual fragment. Here, again, we do not know the processes which cause the arrangement of the cells and their often manifold locally-conditioned differentiations, in short, we do not know the essential nature of regeneration. But, in the meantime, we can endeavour to find out which cell-groups regeneration is bound up with in particular cases, so as to know where the vital particles, the 'determinants,' which condition regeneration, are placed by nature.

Fig. 99. Regeneration of the lens in the Newt's eye. A, section through the iris (J); from its margin and posterior (retinal) surface the primordium of a new lens (L) has developed after the artificial removal of the old one. B, section through the eye after duplicated regeneration of the lens (L) from two areas of the iris. Gl, vitreous humour. J, iris. C, cornea. R, retina. After Fischel.

In this case there can be no doubt on that point: they are the cells on the posterior wall and the margin of the iris. And it is certainly not the absence of the lens which gives rise to its renewal, as would necessarily be the case if it were due to the dominance of an adaptive force. If the lens, instead of being excised, be simply pressed back into the vitreous humour occupying the cavity of the eye, a new lens is developed all the same from the irritated margin of the pupil. And if by chance this margin has been irritated in two places while extraction of the lens was being performed, then two small lenses will develop (Fig. 99, B). Indeed, several may begin to develop at the posterior wall of the iris, although they do not attain to full development; mechanical irritation of any part of this cell-layer is responded to by the formation of lenses. This surely disposes of the 'mystical nimbus' which would dazzle us with a new force of life, always creating what is appropriate. We have before us an adaptation to the liability of newts' eyes to injury, which, like all adaptations, is only relatively perfect, since under the usual conditions of eye injury it gives rise to a usable lens, but under unusual conditions to unsuitable structures. It is exactly the same as in the case of animal instincts, which are all 'calculated' for the ordinary conditions of life, but, under unusual conditions, may operate in a manner quite unsuited to the necessary end. The ant-lion has the instinct to bore backwards into the sand, and he makes the same backward-pressing movements when placed on a glass plate into which he cannot force the tip of the abdomen. The same is true of the mole-cricket, which makes its usual digging movements with the forelegs even on a plate of glass. The wall-bee roofs over her cell when she has laid an egg in it, but she does so even if the egg be taken out beforehand, or if a hole be made in the bottom of the cell, so that the honey which is to serve the larva for food when it emerges from the egg runs out (Fabre). Her instinct is calculated for filling the cell once with honey, and once laying an egg in it, because such disturbances as we may cause artificially do not occur or occur very rarely in natural conditions. There are countless facts of this kind, for every instinct and every adaptation can, in certain circumstances, go astray and become inappropriate. This should be considered by those who still persist in opposing the theory of selection, for herein lies one of the most convincing proofs of its correctness. Adaptations can only arise in reference to the majority of occurrences, and variations which are only useful in an individual case must, according to the principle, disappear again. Adaptation always means the establishment of what is appropriate in an average number of cases.

Therefore the inappropriate reaction of the margin of the iris to an artificial double stimulus affords additional reason for regarding regeneration as an adaptive phenomenon. If it were the outcome of an adaptive force it could never be inappropriate; and if it were the operation of a general and primary power of the organism it would be exhibited by the nearly-related frog as well as by the newt. But, in the frog, extraction of the lens gives rise only to a sac-like proliferation of the cells of the iris margin, which form no transparent lens, but an opaque cluster of cells, which destroys vision altogether. It appears, therefore, that the frog no longer requires the power its ancestors possessed of regenerating a lost lens.

LECTURE XXI

REGENERATION (continued)

Phyletic origin of the regenerative capacity—The liberating stimuli of regeneration—Production of extra heads and tails in Planarians (Voigt)—Regeneration in the Starfish—Atavistic regeneration in Insects and Crustaceans—Progressive regeneration—Regeneration has its roots in the differentiation of organisms—The nuclear substance of unicellular organisms is the first organ for regeneration—The ultimate roots of regeneration.

In the previous lecture we have considered many different forms of regeneration, and have recognized them as adaptive phenomena; we have now to inquire how such regeneration-adaptations have arisen, and this is a very difficult question even in general, while in particular cases it is often quite unanswerable at present. In regard to the case last discussed, the regeneration of the lens in the eye of Triton, our hypotheses would require to reach back to the time of the primitive vertebrates with an unpaired eye, for the lens of the paired vertebrate eye, from Mammals down to the lowest Fishes, does not arise in embryonic development from the retinal cells, but always from the corneal epithelium, as the elaborate researches of Rabl have recently shown. It is true that the unpaired parietal eye of some reptiles forms its lens from the cells of the retinal layer, but it would be difficult to demonstrate the possibility of a genetic connexion between it and paired eyes, and in the meantime we must refrain from elaborating a hypothesis as to the origin of the marvellous faculty the retinal cells possess of transforming themselves into lens-fibres.

But it is easier to form some sort of picture of the origin and adaptation of the faculty of regeneration in general.

We saw that the power of regenerating a part can be localized, and that it does not belong to all the cells of the body, but only to some of them, and we have to ask how and by what steps it has been imparted to these. The faculty depends on the possession of a regeneration-primordium (Anlage), and this again, in our mode of expression, consists of a definite complex of determinants, and as determinants are the products of an evolution, and thus are vital units which have arisen historically, they can nowhere suddenly originate anew in a species, but must be derived directly or indirectly from the sole basis which, in each species, forms the starting-point of the individual—that is to say, in the Metazoa, from the germ-plasm of the ovum. From it the determinant-complex of every regeneration-rudiment mast in the ultimate instance be derived.

We may think of the matter thus: all the determinants of the germ-plasm vary, grow slowly or quickly, and in certain circumstances may be doubled. In this way there arise what we may call 'supernumerary' determinants, which are not required in the primary building up of the body from the ovum, and which may remain in an inactive state in the nuclei of certain cells, ready to become active under certain circumstances and to produce anew the part which they control. Such regeneration-idioplasm will at first come to lie in the younger cells of the determinate organ, but it is conceivable that under the influence of selection it may be gradually shifted to other cells of a later developmental origin, or, conversely, to others in a less external position, so that, for instance, the regeneration-rudiment for the finger of a newt may be contained not merely in the cells of the hand, but in those of the fore-arm or even of the upper arm.

But all such segregation of determinant-groups cannot have taken place, as we might perhaps be inclined to think, at the periphery in the organ itself during its development; it must take place in the germ-plasm of the ovum, for otherwise it could not be transmissible, and could not be directed and modified by the processes of selection, as is actually the case, as I shall show in more detail later on.

I have already pointed out the importance of the rôle played by liberating stimuli in regeneration, and not only of extra-organismal stimuli, such as gravity, but above all of intra-organismal stimuli that is, the influences exerted in a mysterious manner by other parts of the animal on the parts which are in process of regeneration. It is a great merit of the modern tendency in evolution theory that it has demonstrated the importance of such internal influences. Although we are still far from being able to define the manner in which these influences operate, we may say so much, that it depends essentially on the nature and extent of the loss which parts are reproduced by the regenerating cells, and, also, on the position and direction of the injured surface from which the regeneration starts. The influences, still quite beyond our comprehension, which are exerted on the regenerating part by the uninjured parts constitute the liberating stimuli, which evoke the activity of one or other of the determinants contained in the regeneration-idioplasm.

Fig. 100. Regeneration of Planarians. A, an animal divided into three parts by two oblique cuts. B, the fragments(a, b, c) in process of regeneration. C, an animal with various oblique incisions in the margin of the body, which have induced the new formation of heads (k), of tails (s), and pharynx (ph). A and B after Morgan; C after Walter Voigt.

Walter Voigt has shown, by a series of most interesting experiments, that it is possible not only to cause the development of a new head in Planarians by cutting them, in which case a tail may grow from the anterior portion and a head from the posterior portion, but it is also possible in an intact animal, that is, one with both head and tail, to cause the production of a second head, or a second tail, or both at once, at any part of the body margin at will, according to the direction of the cut. If the margin of the body be cut obliquely forwards (Fig. 100, A) a supernumerary tail arises (C, s), if it be cut obliquely backwards a supernumerary head arises (C, k), and in this way several heads and several tails may be produced in the same animal. It is obvious, then, that the interaction, in the first place, of the cells of the cut surface, but probably also of the deeper-lying cells, decides which determinants are to come into action, those of the head or those of the tail, but both must be present at every part of the cut. How far below the cut surface the cells take part in this determination we cannot make out, but that it cannot be due to the co-operation of all parts is clear in this case at least, since the animal still possesses its original head and tail. The extra heads and tails thus produced prove, at any rate, that there can be no question here of the expression of an adaptive principle, a spiritus rector, or a vital force, which always creates what is good, but that it is rather a purely mechanical process, which takes its course quite independently of what is useful or disadvantageous, and that it must take this course according to the given regeneration-mechanism and the stimulus supplied in the special case. It cannot be supposed that these supernumerary heads and tails are purposeful, but who would expect an adaptive reaction from the animal in a case like this, since cuts of the kind which we make artificially, and must keep open artificially if the deformities are to develop, hardly occur in nature, and, if they did occur, would very quickly close up again? Adaptations can only develop in response to conditions which occur and recur in a majority of cases, and when they have a useful, that is, species-preserving result. The adaptiveness of the organism is blind, it does not see the individual case, it only takes into account the cases in the mass, and acts as it must after the mechanism has once been evolved. The case is the same as that of 'aberrant' or mistaken instincts, whose origin by means of selection is the more clearly proved, since we must recognize such an instinct as a pure mechanism and not as the outcome of purposeful forces.

In the regeneration of Planarians we must think of the regeneration-idioplasm as containing the full complex of the collective determinants of the three germinal layers, and possibly we must add to this cells with the complete germ-plasm for giving rise to the reproductive cells. But when the amputated tail of the newt is regenerated, or its leg, or the arm of a starfish, or the bill of a bird, we have no ground for assuming that the cells, from which regeneration starts, contain the whole germ-plasm, since the determinants of the replaceable parts suffice to explain the facts. We must even dispute the possibility of the presence of the whole germ-plasm in this case, because the faculty of regeneration of the relevant cells is really no longer a general one, but is limited to the reproduction of a particular part. This is seen in the fact that, in the starfish, whose high regenerative capacity is well known, the central disk of the body may indeed give rise to new arms[2]; but an excised arm, to which no part of the disk adheres, is in most starfishes unable to give rise to the body. Thus the arm does not contain in its cells the determinants of the disk, but the latter contains those of the arm. We are not surprised that the amputated tail of the salamander does not reproduce the whole animal, but this can only be because the impelling forces to the regeneration of the whole animal are wanting, that is, that the cut surface only contains the determinants of the tail and not the complete germ-plasm. It might be objected here that the tail-piece is too small to give rise to the whole body, but in Planaria it is only very diminutive heads and tails which grow from the artificial incisions, and the same is true of starfishes when only a single arm and a small piece of the disk have been left. Notwithstanding the small amount of living substance at their disposal, and although they are at first unable to take nourishment, they send out very small new arms (Fig. 101), close up the wounded surface, and, after reconstruction of the mouth and stomach, begin to feed anew. The new arms may then grow to the normal size.

[2] I see now that there are contradictory statements in regard to this case. Possibly these depend on the different behaviour of different species, and this on the varying frequency of mutilation. Starfishes which live on the shore between the rocks, for instance on the movable stones of a breakwater, are very frequently mutilated; in some places it is rare to find a specimen without traces of former wounds. H. D. King counted among 1,914 specimens of Asterias vulgaris 206 in the act of regenerating a part, that is, 10.76 per cent. In the case of the starfishes from deep water this cause of injury does not of course exist.

Fig. 101. A starfish arm,
growing four new arms;
the so-called 'comet-form.'
After Haeckel.

We must therefore assume that, in many cases, the regeneration-primordium consists of cells which only contain a definite complex of determinants in the form of latent regeneration-idioplasm, as, for instance, certain cells of the tail of Triton contain the determinants of the tail, certain cells of its leg the determinants of the leg, and so on. In many cases we can speak even more precisely, and determine from which cells the nerve-centres, from which the muscles, and from which the missing section of the food-canal will be formed, as was recently shown by Franz von Wagner in regard to the worm Lumbriculus, whose regenerative capacity is so extraordinarily high. We must then attribute to each of the relevant cells an equipment of regeneration-idioplasm, which includes only the relevant complex of determinants.

I need not here go further into detail, but I should still like to show that, in reality, as I assumed in regard to the regenerative capacity of a part, the root of the regeneration-idioplasm lies in the germ-plasm, that it is present there as an independent determinant-group, and, like every other bodily rudiment (Anlage), must be handed on from generation to generation. This assumption is necessary, as has been already indicated, on the ground that the faculty of regeneration is hereditary, and hereditarily variable, on the same ground, therefore, as that on which the whole determinant theory is based. The regeneration-determinants must be contained as such in the germ-plasm, otherwise a twofold phyletic development could not have occurred, as it actually has, in many parts. The tail of the lizard is adapted for autotomy; it breaks off when it is held by the tip, and this depends on a special adaptation of the vertebræ, which are very brittle in a definite plane from the seventh onwards. This is thus a very effective adaptation to persecution by enemies. The tail which has been seized remains with the pursuer, but the lizard itself escapes, and the tail grows again. But this regeneration does not take place in the same way as in the embryo; no new vertebræ are formed, but only a 'cartilaginous-tube,' a new structure, a substitute for the vertebral column; the spinal cord with its nerves is not regenerated either, and the arrangement of the scales is somewhat different.

This last point, in particular, indicates that the determinants of the regeneration-rudiment may pursue an independent phylogenetic path of their own, for this scale arrangement of the regenerated tail is an atavistic one, that is, it corresponds to a more primitive mode of scale arrangement in these Saurians. We know quite a number of cases similar to this. It not infrequently happens that cut-off parts regenerate, but that they do so not in the modern form, but in one that is in all probability phyletically older. Thus the legs of various Orthoptera, as of the cockroaches and grasshoppers, regenerate readily, but with a tarsus composed of four joints instead of five[3], and the long-fingered claws of a shrimp (Atyoida potimirim) is replaced by the older short-fingered type of claw, while in the Axolotl an atavistic five-fingered hand grows instead of the amputated four-fingered one.

[3] New investigations, specially directed to this point, by R. Godelmann, have shown that 'in the great majority of cases' the regenerated legs of a Phasmid (Bacillus rossii) exhibit a four-jointed tarsus; but the regeneration of five joints also occurs, though only after autotomy, and only in seven out of fifty cases (Archiv für Entwicklungsmechanik, Bd. xii, Heft 2, July 1901). The regeneration-rudiment in this species seems to be in process of advancing slowly to the five-jointed type.

This last case shows that it is not merely a lesser power of growth that accounts for the difference between the regenerated part and the original, for here more is regenerated than was previously present. There remains nothing for it but the assumption that the regeneration-determinants have remained at a lower phyletic level, while the determinants which direct embryogenesis have varied, and either developed further or retrogressed. It is easy to understand that the regeneration-rudiment must vary phyletically much more slowly than the parts which evolved in the ordinary way and much more slowly than the determinants of these parts, for natural selection means a selection of the fittest, and the speed with which the establishment of a variation is attained depends, ceteris paribus, on the number of individuals that are exposed to selection with respect to the varying part. If in a species of a million living at the same time nine-tenths perish by accident, there will remain only 100,000 from which to select the 1,000 which we will assume constitute the normal number of the species. The more of these 100,000 which possess the useful variation the higher will be the percentage of the normally surviving 1,000 possessing it, and the more rapidly will the useful variation increase. But when it is a question of the variation of the regeneration-primordium, the selection will take place not among all the 100,000 individuals which chance has spared, but only among those of them which have lost a limb by accident, and thus are in a position to regenerate it more or less completely. If we assume that this takes place in 10 per cent. of cases, then selection for the improvement of the regeneration-apparatus will only take place among 1,000 individuals, and thus the process of modification of the regeneration-primordium must go on very much more slowly than that of the limb itself.

I do not see how the opponents of the germ-plasm theory can explain these facts at all, for the appeal to external influences is here entirely futile, and that to internal liberating stimuli does not suffice, since these must be different after a part has been cut off from what they were when the limb developed normally, and also different from those which prevailed at the normal origin of the limb in ancestral forms. The four-jointed tarsus of the ancestors of our cockroaches did not arise as a result of amputation. We cannot therefore avoid referring the processes of regeneration to particular 'regeneration-determinants,' which are contained in the germ-plasm and are handed on in ontogeny with the other determinants from cell-division to cell-division, till ultimately they reach the cells which are to respond, or may have to respond, to the stimulus of injury by some expression of their regenerative capacity. As these determinants, as has been shown, can often only be very slightly subject to the influence of selection processes, they will, in many respects, lag behind in the phyletic development, and will tend to belong to an ancestral type of the relevant part. They will often remain for a long time at this ancestral level, and they will always adapt themselves to new requirements more slowly than the parts which arise in the normal way, and the determinants representing these in the germ. But the regeneration-determinants are variable, and, indeed, are so hereditarily, and independently of the structure of the normal parts. They thus follow their own path of phyletic development, and this one fact is enough to secure a preference for the germ-plasm theory above others that have hitherto been suggested. None of these has even attempted an explanation of this fact; the tendency has rather been to call it in question. This, however, can be done at most only in regard to the explanation of the regenerations as atavistic, certainly not in regard to the progressive variations of the regenerated part, such as have been established by Leydig and Fraisse in regard to the lizard's tail. It may be doubted whether the most primitive insects had only four tarsal joints, but there is no disputing the kainogenetic deviation of the lizard's-tail.

I have interpreted the regenerative capacity as secondary and acquired, not as a primary power of all living substance, and I should like to substantiate this in another way.

Let us go back to the simplest organism conceivable, which must have represented the beginning of life on our earth, and we see that this need not have possessed any special power of regeneration, because, for an organism without differentiation of parts, growth is equivalent to regeneration. But growth is the direct outcome of one of the primary characters of the living substance, the capacity of assimilation. This cannot be an adaptive phenomenon, nor can it have arisen through selection, because selection presupposes reproduction, and reproduction is only a periodic form of growth; but growth follows directly from assimilation. The fundamental characters of the living substance, above all the dissimilation and assimilation which condition metabolism, must have been in existence from the first when living substance arose, and must depend on its unique chemico-physical composition. But the faculty of regeneration could only be acquired when organisms became qualitatively differentiated, so that each part was no longer like every other part or like the whole. As soon as this stage was reached the faculty of regeneration would necessarily be developed, if further multiplication was to take place. For when each fragment could no longer become a whole by simply growing, some arrangement had to be made by which each fragment should receive, in the form of primary constituents, what it lacked to make up the whole. We do not know the first beginning of this adaptation, but, in its further development, it appears in the form of 'nuclear substance,' enclosed in the nucleus of the cell, and, as is well known, it is now to be found in all unicellular organisms. That the nucleus there precedes regeneration in the sense that without a piece of it the cell-soma is not able to complete itself alone, we have already seen, and the explanation of this fact has always seemed to me to be that invisibly minute vital units relating to the regeneration of the injured part leave the nucleus and evoke the development of the missing parts by laws and forces still unknown to us. Loeb has recently claimed that the nucleus is the cell's organ of oxidation; but if that be true it would still not exclude the possibility that the nucleus is also and primarily a storehouse of the material bearers of the primary constituents of a species. It must be regarded as such when we call to mind the phenomena of amphimixis in its twofold aspects as conjugation and as fertilization, and its obvious outcome among higher organisms where it implies the mingling of the parental qualities.

Thus the 'nuclear substance' of unicellular organisms is for us the first demonstrable organ of regeneration, and first of all for normal regeneration, which takes place at every reproduction, for instance, of an Infusorian. For we have already seen that, in the transverse division of a trumpet animalcule (Stentor), the anterior part must develop the posterior half anew, while the posterior half must develop the much more complex anterior half, with mouth region and spiral bands of cilia. But as soon as the arrangement for normal reproduction was elaborated, as soon as the nucleus was present, as a depôt of 'primary constituents,' this implied the possibility of regeneration in exceptional cases, that is, after injury. The mechanism was already there, and it came into operation as soon as a part of the animal was missing.

It is in the first nucleus, therefore, that we have to look for the source of all regenerative capacity, both in unicellular and multicellular organisms. But with the origin of the latter a limitation took place, either quite at the beginning or a little later, for each nucleus of the cell-colony no longer contained the whole complex of 'primary constituents' or determinants of the species, but, in many cases, only the reproductive cell possessed them. As soon as this began to develop into a whole by cell-division the determinant-complex was segregated. Thus the first cell-colonies with two kinds of cells arose, as we have seen in the case of Volvox—the reproductive cells with a complete equipment for regeneration in their nucleus, and the somatic cells with a limited equipment for regeneration in their nuclei. The somatic cell could no longer give rise anew to the whole organism, but could only reproduce itself or its like.

But as many of the lower Metazoa and Metaphyta possess the power of budding, that is, are able not only to produce a new individual from definite cells—the reproductive cells—with or without sexual differentiation, but from other cell-groups also, these must contain the whole complex of determinants appertaining to the reconstruction of the organism, and we have to ask how this is reconcilable with the differentiation of a multicellular organism, whose different kinds of cells depend, according to our interpretation, on the fact that they are controlled by different determinants.

Obviously, there is only one way out of this difficulty, and it is the one we have already indicated, that although the diffuse regenerative capacity which we have just alluded to occurs in species which exhibit gemmation, this does not exclude the control of a cell by a specific determinant; other determinants may be contained in the cell, in a state, however, in which they do not affect it, that is, in an inactive or latent state.

Thus we arrive in this way also at our earlier assumption that an inactive accessory-idioplasm is given to all, or at least to many cell-generations. Only among plants must this necessarily be complete germ-plasm, and among the lower plant-forms, as in Caulerpa among the Algæ, in Marchantia among Liverworts, it must be assumed to be present in nearly all the cells, according to the experiments in regeneration made by Reinke and Vöchting. But in multicellular animals which develop from two different germinal layers equipped with a different complex of determinants budding arises from a combination of at least two different kinds of cells, and we must only ascribe to each of these its own peculiar determinant-complex as regeneration-idioplasm. Higher plants show us that well-marked power of budding is not necessarily associated with a high regenerative capacity, the histologically specialized cells among them will contain no inactive germ-plasm, because they do not need it. But in animals the power of budding is probably always combined with high regenerative capacity, as is shown by the Polyps and Medusoids above all, and in a different way by the Ctenophores, which exhibit no budding and at the same time a very slight regenerative capacity, although they possess an organization scarcely higher than that of the Hydromedusæ. In the Ctenophores each of the first segmentation-cells, when artificially separated, yields only a half-embryo, and we may conclude from this that it contains no complete germ-plasm in an inactive state, or at least very little, and certainly not a sufficient quantity to make it readily regenerative.

Undoubtedly, however, the regenerative capacity occurs apart from the capacity for budding, yet this in no way contradicts the theory. As we have seen, a high regenerative capacity is to be found among many animals which occur only as 'persons' and not as colonies or stocks, but only in those which are readily liable to injury, and only in the manner conditioned by their injury. In the higher Metazoa the regenerative power becomes more and more limited, and in the Mammals it sinks to a mere closing up of wounds.

If we take a survey of the assumptions we have been compelled to make from the standpoint of the theory to explain the development of germ-cells, budding, and regeneration, it would seem as if it were contradictory to assume that, on the one hand, complete germ-plasm should be given to certain cell-series as inactive accessory idioplasm, and, on the other, that very numerous cells, at least in the lower Metazoa, should have received the idioplasm of budding, and still more numerous cells that of regeneration. But it is obvious that among the lower Metazoa the idioplasm of budding and the idioplasm of regeneration are equivalent; the same idioplasm, which, when liberated by stimuli unknown to us, co-operates from two or three germinal layers in the formation of a bud, effects, in response to the known stimulus of injury, the regeneration of the mutilated part. But germ-cells can never arise in the Metazoa from the partial budding-idioplasm or regeneration-idioplasm, because this is not complete germ-plasm, and because it can only give rise to budding or regeneration through the co-operation of two or more kinds of cells, while germ-cells always originate from one cell and never arise from the fusion of cells. Germ-cells can thus only arise from the cells of the germ-track, and in no other way, no matter whether the germ-track lie in the ectoderm, as in the Hydromedusæ, or in the endoderm, as in true jellyfishes (Acalephæ) and the Ctenophores, or in the mesoderm, as in many higher groups of animals. It is only apparently that these cells belong to one particular layer, for in reality they are unique in kind, and they are simply assisted in their development by one or other cell-layer, from which they not infrequently emancipate themselves, as happens so notably in the Hydromedusæ. As we have already said, it is only among plants that we must think of budding as arising from cells which contain complete germ-plasm, for here there are no 'germinal layers' corresponding to those of animal development, and the cells of 'the growing point' must be equipped with the complete germ-plasm. The plant, like the Hydroid stock and the Siphonophore colony, is saved from death, in spite of the frequent loss of its members, mainly by the fact that it is capable of producing, at almost any part above the ground, buds which develop into new shoots, with leaves and the like. This makes a power of regeneration on the part of the individual leaves and flower-parts superfluous, but at the same time it implies that an enormous number of cells must be distributed over the whole surface of the plant, each of which can in certain circumstances become the starting-point of a bud. That is to say, each must contain, in a latent state, the complete germ-plasm which is necessary for the production of an entire plant.

We must therefore assume that, in the higher colony-forming plants, germ-plasm is contained in a great many cells, perhaps in all which are not histologically differentiated, and sometimes even in those which are so, as, for instance, in the leaves of Begonias. I suppose, therefore, that in the higher plants the process of development implies a segregation of the determinant-complexes of the germ-plasm, but that this takes place at a late stage, and that in a much higher degree than among animals the individual or the 'person' carries with it germ-plasm in a latent state. To this must be attributed the fact that the plant is not only able to make good its losses in twigs and branches by sending out new shoots, but that cuttings, that is, detached shoots, are also able to take root, and in general to give rise to what is necessary to complete themselves according to the position of the part in question. In the ontogeny of animals, too, we must assume that it requires a liberating stimulus to rouse the determinants to activity, that this stimulus is to be sought for in the influence exercised by the constitution of the cell on the idioplasm contained within it, and that this constitution in its turn is subject to influences from external conditions, including the cell-soma itself. We may therefore suppose that, among plants also, the germ-plasm latent in numerous cells only becomes active in whole or in part according to the influences exerted on it by the state of the cell at the moment; but this varies with external circumstances, according to whether the cell is exposed to light or lies under ground, according as it is influenced by gravity, by moisture, chemical stimuli, and so on.

It might be objected to this that it would be simpler not to assume a segregation of the germ-plasm into determinant-complexes at all in order to explain the process of development, but rather to credit each cell with a complete equipment of germ-plasm from the beginning to the end of the ontogeny, and to attribute the differences in the cells, which condition the structure of the plant and its differentiation, solely to the different influences, external and internal, to which the cell is exposed, and which rouse some determinants to activity at one part and others at another. Perhaps the botanists would be more readily reconciled to this idea, but it seems to me that there are two points which tell against the possibility of its being correct. In the first place, it is far from being established that every cell in the higher plants is capable of giving rise, under favourable conditions, to a whole new plant; every tree and every higher plant has a multitude of cells in its leaves, its flowers, and so on, which cannot do this, which are in fact differentiated in one particular direction, that is, they contain only one kind of determinants, like the histologically differentiated cells of the tissues of the human body. Secondly, there are other organisms besides plants, and a theory of development cannot be based on the phenomena to be observed among plants alone, any more than a theory of heredity can. There are obvious differences in the processes of life among plants as contrasted with those among animals, but it is improbable that there is any thoroughly fundamental difference. It is, however, indubitable that the cells forming the tissues of higher animals, the nerve, muscle, and glandular cells, are really differentiated in one direction, and are quite incapable, under any circumstances whatever, of growing into an entire organism, and even from this alone we might conclude that they contain only one primordium or determinant. Are we then to assume that the vascular cells, epidermis-cells, wood-cells, and so on, of the higher plants, which are also differentiated in one direction, do nevertheless contain the complete germ-plasm? I do not see any ground for such an assumption.

To conclude what can be said on the subject of regeneration we must return to the question of an ultimate explanation of this marvellous phenomenon. I have declined to attempt any explanation at all, because I do not consider it possible to give a sufficient one as yet, but I should like at least to give an indication as to the direction in which we must look for it.

We assumed that there is a regeneration-idioplasm, and therefore that there are 'primary constituents' at certain positions in the body, but how does it happen that these are able to build up the lost parts in the proper situation and detail? A theoretical formula might well be thought out, according to which the determinants of successive parts would become active successively, and would thus liberate one another in an appropriate order of sequence, but there would not be much gained by this, especially as what we already know in regard to the regrowth of the legs and toes in Triton does not harmonize with such an assumption. It appears to me more important—though even here we must still be very vague as to details—to recognize that, in all vital units, there are forces at work which we do not yet know clearly, which bind the parts of each unit to one another in a particular order and relation. We were obliged to assume such forces even in regard to the lowest units, the biophors, since otherwise they could not be capable of multiplication by division, on which all organic growth depends, unless we are to assume, as Nägeli did, a continual generatio æquivoca of the specific kinds of biophors (his 'micellæ'). But we shall see later, when we come to speak of spontaneous generation, that we cannot acquiesce in such an assumption. If, then, we cannot conceive of a power of division arising from within and depending solely on growth by means of assimilation, without such attractive and repellent forces or 'vital affinities' the internal parts would necessarily fall into disorder at every division. It seems to me therefore that such 'affinities' must be operative at all stages in the life of the vital units, not only in biophors, but also in the cell, and in the 'person' as well as in determinant and id. It is true that 'persons' no longer generally possess the power of multiplying by division, but in plants and lower animals many do possess it; and the power of giving rise anew to certain parts is obviously a part of that power of doubling the whole by division. The ultimate roots of regeneration, then, must lie in these 'affinities' between the parts, which preside over their arrangement and are able to maintain it and to give rise to it anew. In this respect the organism appears to us like a crystal whose broken points always complete themselves again from the mother-lye after the same system of crystallization, obviously in this case too as a result of certain internal directive forces, polarities, which here again we are unable precisely to define. But the difference between the organism and the crystal does not—as people have been hitherto inclined to believe—lie only in the fact that the crystal requires the mother-lye to complete itself, while the vital unit itself procures the material for its further growth; it lies also in the fact that such regeneration is not possible in every organism and at every place, but that special 'primary constituents' are necessary, without which the relevant part cannot arise. The indispensableness of these primary constituents, the determinants, seems to me to depend on the fact that the new structure cannot be built up simply by procuring organic material, but that specially hewn stones, different in every case, are necessary, which can only be supplied in virtue of an historical transmission, or, to abandon the metaphor, because the vital units of which the organ is to be reconstructed possess a specific character and have a long history behind them; thus they can only arise from such vital units as have been handed on through generations, that is, from the determinants. But these primary constituents are given to the different forms of life in very varying degrees and in very unequal distribution, and as far as we can see according to their suitability to an end.

LECTURE XXII

SHARE OF THE PARENTS IN THE BUILDING UP
OF THE OFFSPRING

The ids are 'ancestral plasms'—The reducing division brings about a diversity of germ-plasm in the germ-cells—Bolles Lee's 'Neotaxis' even in the primordial germ-cells—Häcker's observations on the persistent distinctness of the maternal and paternal chromosomes—Identical twins—The individuality is determined at fertilization—Unequal share of the ids in the determination of the offspring—Preponderance of one parent in the composition of the offspring—Certain ids of the ancestors remain unchanged in the germ-plasm of the descendants—Struggle of the Biophors—Alternation of the hereditary sequences in the parts of the child—Reversion—Datura-hybrids—Zebra-striping in the horse—Three-toed horses—New experiments in hybridization among plants by Correns and De Vries—Xenia.

As far as the phenomena of regeneration and budding are concerned, we have not been able to do much more than bring them under a formula, which harmonizes with the germ-plasm theory. But the case is different with the actual phenomena of inheritance in the restricted sense, for instance, with regard to the transmission of individual peculiarities from parent to child. Here the theory really increases our insight and lets us penetrate deeper into the causes of the phenomena; it is here no longer a mere 'portmanteau-theory.'

We are well aware, especially from observation on ourselves, that is, on Man, that the children of a pair often resemble one another but are never alike, and that one child frequently resembles one parent, another the other, while a third may exhibit a mingling of both parents. How does this come about? Since the germinal substance of both parents is derived from that of the ovum, from which they themselves have arisen—and must therefore be the same in all the germ-cells to which they give rise—new determinants cannot be added, and old ones cannot be dropped out, and variation of the determinants, the possibility of which is granted, would still not directly bring about the familiar mingling of resemblances to the two parents, but would at most give rise to something new and strange.

Here the theory helps to elucidate matters. We found ourselves obliged to assume that the germ-plasm is composed of ids, that is, of equivalent portions of germ-plasm, each of which contains all the kinds of determinants appertaining to the building up of an individual, but each of these kinds in a particular individual form. I have already called these ids 'ancestral plasms,' and the term is appropriate, in so far that in every fertilization an equal number of ids from the father and from the mother are united in the ovum, so that the child is built up of the ids of his two nearest 'ancestors.' But as the ids of the parents are derived from those of the grandparents, and these again from those of the great-grandparents, the ids are in truth the idioplasm of the ancestors.

The expression, however, has been very frequently misunderstood, as if it were intended to mean that the ids retained unchanged for all time the character of their respective ancestors, and I have even been credited with supposing that our own ids still consist of the determinant-complexes of our fish-like or even Amœba-like ancestors. But in reality no id exactly or completely corresponds to the type, that is, to the whole being of any one of the ancestors in whose germ-plasm it was formerly contained, for each of the ancestors had many ids in his germ-plasm, and his entire constitution was not determined by any one of these alone, but by the co-operation of them all. The individual arising from a germ-cell must necessarily be the result of all the ids which make up his germ-plasm, but undoubtedly the share taken by some of them may be much stronger than that taken by others. It is also clear that, if we leave out of account any possible variation on the part of the ids, each of them belongs, not to one ancestor only, but to a whole series of ancestors, and must have taken part in their development, so that it is not the idioplasm of any particular ancestor, but only ancestral plasm in the general sense. In this sense we may quite well retain the designation, 'ancestral plasm,' for the id.

Thus, according to our view, the germ-plasm consists of ids, each of which contains all the determinants of the whole ontogeny, but usually in individually different quality.

Returning for a moment to the processes by which the reduction of the chromosomes, that is, of the nuclear rods of germ-plasm in the ovum and sperm-cell is brought about, we recall the fact that this happens at the last two divisions of the germ-cell, the so-called 'maturing divisions.' In these the nuclear substance, as we have seen, is divided between the two daughter-nuclei in a manner quite different from the usual one, for a longitudinal splitting of the rods, bands, or spheres in the equatorial plane of the nucleus does not take place, but half the number of rods move into the right and half into the left daughter-nucleus without previous division, so that in each daughter-nucleus the number of rods is reduced to half (Fig. 76).

Fig. 76. Diagram of the maturation divisions of the ovum. A, primitive germ-cell. B, mother-egg-cell, which has grown and has doubled the number of its chromosomes. C, first maturation division. D, immediately thereafter; Rk 1, the first directive cell or polar body. E, the second maturation spindle has been formed; the first polar body has divided into two (2 and 3); the four chromosomes remaining in the ovum lie in the second directive spindle. F, immediately after the second maturation division; 1, the mature ovum; 2, 3, and 4, the three polar cells, each of these four cells containing two chromosomes.

Although the distribution of the rods in this manner takes place twice in succession, the normal number is not, as we have already seen, reduced to a quarter, because, long before the occurrence of the first maturing division, a duplication of the rods by means of longitudinal division had taken place, and thus the first division differs from an ordinary division in that the splitting of the rods does not take place during the process of dividing but long beforehand. Only the second maturing division differs from all other nuclear divisions known to us, since it is not associated with any splitting of the rods at all, but conveys half of the existing rods into each daughter-nucleus. It is the time reducing division, through which the number of the rods is reduced to one half[4].

[4] Recent investigations have shown that the reduction of the chromosomes does not always take place exactly in accordance with the scheme here indicated, but that it differs from it in many cases. But as investigations on this point are as yet by no means complete, I need not go into the question further; the ultimate result is the same in any case.

This numerical reduction must, however, have other consequences; it must make the germ-cells of the same individual qualitatively unlike, that is, in relation to their value in inheritance. Let us assume only four chromosomes of the rod-form ('idants') as the nuclear elements of a species, two of which, A and B, come from the mother, and other two, C and D, from the father, the last maturing division may, as far as we can see, result either in removing the combination A and B from C and D, or A and C from B and D, or A and D from B and C; there is thus a possibility of one of six different combinations of rods in any one germ-cell. What is the same thing, six different kinds of germ-cells differing in their hereditary primary constituents may be developed in the same individual. As this new combination, or, as we may call it, neotaxis of the germ-plasm elements, takes place in female as well as in male individuals, there is a possibility that, in fertilization, 6 × 6 = 36 individuals with different primary constituents may arise from the germ-cells of the same two parents. Of course the number of possible combinations increases very considerably in proportion to the normal number of rods, for with eight of these it comes up to 70, and with sixteen to 12,870; the number of individuals differing in their inherited primary constituents would thus be enormous, for each of the 70 or of the 12,870 different hereditary minglings of the ovum could combine in amphimixis with 70 or 12,870 different sperm-cells, so that 70 × 70 and 12,870 × 12,870 offspring individually different in their primary constituents might arise from the same two parents. In Man there are said to be sixteen nuclear rods; so that in his case the last-mentioned number of parental hereditary minglings might occur. This may seem a disproportionately high number as compared with the small number of children of a human pair, but we must not judge from the case of Man alone, and in plants and animals, which we have already discussed, the number of descendants is very much larger, and is often enormous. We saw what significance this apparent extravagance on the part of nature has, for without it adaptation to changed conditions of life would not be possible, since, if only so many were born as could attain to reproduction, no selection of the fittest could take place. The same would be the case if all the young of a species were alike, and even if all the descendants of a single pair were alike, effective selection would be excluded, since only as many individualities could be selected as there were pairs of parents. It is easy to understand that selection works more effectively the larger the number of descendants of a species and the more they differ from each other. The chance that the best possible combination of characters will occur is thereby increased.

Although we cannot calculate how many individuals of different combinations of characters natural selection requires to work upon in order to direct the evolution of the species[5], we can understand that only as large a choice as possible can secure that the best possible adaptations of all parts and organs are brought about and maintained. Precisely in the fact that in every generation such an enormous superfluity of individuals is produced lies the possibility of such intensive processes of selection as must continually take place, if the adaptation of all parts is to be explained. For if among the thousands of descendants of a fertile species each hundred were alike among themselves, these hundreds would have, as far as natural selection was concerned, only the value of a single variant. But such an all-round adaptation as actually exists in the structure of species requires as many variants as possible; it requires that each individual should be a peculiar complex of hereditary characters; that is, that all the fertilized germ-cells of a pair should possess an individually well-marked character.

[5] For this reason I have left the number of id-combinations given above unaltered, though, according to the most recent researches into the processes of maturation, they are probably too high, since every conceivable combination does not actually occur. We are here concerned less with the exact number than with the principle.

The justification of this postulate becomes all the clearer if we take into consideration the male germ-cells as well as the female. Let us think of the enormous number of sperm-cells which are produced by many animals, and indeed by the highest of them—an almost incalculably large number which certainly goes far beyond millions. Let us assume that in Man there may be 12,870 million spermatozoa, then, with sixteen ids, and with an equally frequent occurrence of all possible combinations of germ-plasm—there would be 12,870—there would be a million of each type containing identical germ-plasm. The danger that several ova would be fertilized by identical sperm-cells would be by no means small.

It cannot, therefore, surprise us that other means have been employed by Nature to secure re-groupings of the ids. The simplest means would be, if before each division of the primitive germ-cells the nuclear rods were to divide, and if the split halves were irregularly intermingled, then at the formation of the next nuclear spindle an entirely new arrangement of the halves would result. But in animals, at least, this is certainly not the case; the processes of reduction are restricted to the maturing divisions.

Years ago Ischikawa observed that, in the conjugation of Noctiluca, the nuclei of the two animals become closely apposed, but that they do not fuse, although they behave like a single nucleus in the division which follows. In this case paternal and maternal nuclear substance remain separate ([Fig. 83, vol. i. p. 317]). The same phenomenon has since been repeatedly observed in many-celled animals, first by Häcker, then by Rückert in the Copepods, and afterwards by Conklin in the eggs of a Gastropod (Crepidula). But all these observations referred only to the earlier stages of ovum-segmentation up to twenty-nine cells, and it could not be affirmed that the distinctiveness of the paternal and maternal chromosomes lasted till farther on into the ontogeny. Professor Häcker now informs me, however, that he has been able to trace this separateness in a Copepod (Canthocamptus) not only from the beginning of segmentation on to the primitive genital cell, but also through the divisions of this up to the mother-egg-cell[6]. Thus we may now assume that the paternal and maternal hereditary bodies remain distinct, not only for a time, but throughout the whole development, a fact which confirms our assumption of the independence of the nuclear rods, notwithstanding their apparent breaking-up in the nuclear reticulum of the 'resting' nucleus. This new knowledge throws fresh light in another direction; it proves to us that the remarkable and complicated processes which go on in the nucleus during the maturing divisions have really the significance which I long ago ascribed to them[7], that of effecting the maximum diversity of intermingling of the paternal and maternal hereditary elements. For Häcker has shown that during the second maturation-division the paternal and maternal chromosomes are no longer united each in a special group, but occur scattered about in the nucleus, and subsequently come together again to form two differently combined groups.

[6] Since this was written Häcker has published his results. See Anatom. Anzeiger xv (1902), p. 440.

[7] See my essay, Amphimixis, Jena, 1891.

If this were not so, if the maternal and paternal chromosomes remained separate, then the reducing division would cause only one of these groups to reach each of the germ-cells, and thus each mature ovum or sperm-cell would contain either only paternal or only maternal hereditary bodies. But this would make a reversion to more than three generations back impossible, and as such reversions undoubtedly occur, we must conclude that manifold new combinations of the paternal and maternal chromosomes take place. This obviously happens during the maturation-divisions, at least in the Metazoa.

The more numerous the rods or the free individual ids in a species are, the more numerous are the possible combinations. Whether all the mathematically possible combinations actually occur is a different question, which I should not like to answer in the affirmative just yet; but in any case the actual number of combinations in a species with many nuclear elements will be greater than in one with few, and in this respect those species in which the ids occur as independent granules will have an advantage over those in which they are combined into rods or bands (idants). These latter, however, afford us a better possibility of deducing the new combinations of the ids, although the idants themselves are not outwardly distinguishable from each other.

I must refrain from going into these highly interesting processes in more detail just now. So much is certain, that Nature makes use of various means to bring about the re-combination, and at the same time the reduction of the ids during the two 'reducing divisions.' This is proved by the fact recently established by Montgomery, that in many animal groups reduction results from the first maturing division. Whether it operates at this stage with rings, bands, double rods, X-shaped structures, groups of four (tetrads), and so on, all this serves the same end, the more or less thoroughgoing re-arrangement of the hereditary vital units. I am convinced that new investigations into these processes, if they were undertaken from this point of view, would lead to very important results[8]. It would be important to find out how great the variations are which thus arise, for it is very probable that they differ in degree in the different animal-groups. Even the combination of the ids into rods (idants) indicates that some species may be more conservative than others in maintaining their id-combinations, and that there will be among them a greater tenacity in the hereditary combinations of characters (i.e. of the 'type' of the parents). If we should succeed in penetrating more deeply into these processes we should probably also understand why in certain human families the hereditary characters are transmitted more purely and more tenaciously than those of other families with which they have mingled, and so on. It may well be that the persistence of character is due to the fact that ids which have once combined into rods hold firmly together, for it seems to me in no way impossible that individual differences should occur even in these most delicate processes.

[8] Since this was written for the first edition observations of this process have been considerably increased, and discussions as to the exact interpretation of these are in full tide; we are surrounded by a wealth of new observations, facts, and explanations, without having attained to a consistent and unified theory. Several naturalists, such as Boveri, Häcker, Wilson, and others, have attempted interpretations, but these are in many points contradictory to one another. It is therefore impossible to enter into the question in detail here; further light from new observations must be awaited. So much we may say, however, that it is not chance alone which presides over the re-arrangement of the chromosomes during the reducing divisions; affinities play a part also; there are stronger or weaker attractions between the chromosomes, which aid in determining their relative position to one another.

But let us leave these more intimate questions out of account altogether, and turn our attention to the more obvious and less delicate phenomena, and we find that the re-arrangement of ids (Neotaxis) which we have just discussed affords a simple explanation of the generally observed phenomenon of the differences between individuals! Each individual is different from every other, not in the case of Man alone, but in all species in which we can judge of differences, and this is true not only of descendants of different parents, but even of those of the same parents.

Of course the differences between two brothers or two sisters do not depend entirely on the hereditary basis, but in part also on external conditions which have affected them from embryonic development onwards. Let us suppose that of two brothers who have sprung from identical germ-cells one becomes a sailor, the other a tailor; it would not surprise us to find them very different in their fiftieth year, one weather-beaten and tanned, the other pale; one muscular, straight, and vigorous, the other weakly and of bent carriage. The same primary constituents develop differently according to the conditions to which they are exposed. But the two brothers will still resemble each other in the features of the face, colour of hair, form of eyes, stature and proportion of limbs, perhaps even in a birthmark, more than any other human beings of their own or any other family, and this resemblance will depend upon the identity of the hereditary primary constituents, on the similar id-combination of the germ-plasm.

Man himself affords a particularly good example in favour of this interpretation in the case of so-called 'identical twins.' It is well known that there are two kinds of twins, those that are not strikingly alike, and often very different, and those that are alike to the extent of being mistakable for one another. Among the latter the resemblance may go so far that the parents find it necessary to mark the children by some outward sign, so that they may not be continually confused. We have now every reason to believe that twins of the former kind are derived from two different ova, and that those of the latter kind arise from a single ovum, which, after fertilization, has divided into two ova. This not infrequently occurs in fishes and other animals, and we can bring it about artificially in a number of species by experimentally separating the two first blastomeres.

We have here, then, a case of absolute identity of the germ-plasm in two individuals, for the id-combination of the two ova derived from the same process of fertilization must be exactly the same. That in such a case, notwithstanding the inevitable differences of external influences to which the twins are exposed from intra-uterine life onwards, such a high degree of resemblance should arise is a fact of great theoretical importance. From the basis of the germ-plasm theory we can very well understand it, for, according to the theory, only precisely similar combinations of ids can give rise to identical individuals.

But we learn more than this from the occurrence of identical twins. They prove above all that the whole future individual is determined at fertilization, or, to express it theoretically, that the id-composition of the germ-plasm is decisive for the whole ontogeny. It might have been supposed that the combination of ids could change again during development, and that a greater multiplication of some than of others might take place at certain stages of development, or through certain chance external influences. It might have been thought that there was a struggle among the ids in the sense that some of them were suppressed and set aside. All such suppositions break down in face of the fact of identical twins, which teaches us that identical germ-plasm evokes an ontogeny which runs its course as regularly as two chronometers, which are constructed and regulated alike.

But when I say that a struggle of the ids, in the sense of a material setting aside of some of them, cannot take place, I by no means intend to maintain that the influence which each individual id exerts on the course of development may not be disproportionate to that exerted by others, and, under some circumstances, very disproportionate indeed. I must refrain from entering into this subject in detail now, but I should like to give at least an indication of what I mean.

If the germ-plasm consists of ids, these ids collectively must determine the structure, the whole individuality—let us say, briefly, the 'type' of the offspring; it is the resultant of all the different impelling forces which are contained in the different ids. If these were all equally strong, and all operating in the same direction, they would necessarily all have the same share in the resultant of development, the 'type' of the child. But this is not the case.

Numerous experiments on the hybridization of two species of plant have taught us that the descendants of such hybridization usually maintain a medium between the ancestral species; but it is not always the case, for in many hybrids the character of one species, whether paternal or maternal, preponderates in the young plant.

We recognize the same thing still more clearly in Man, whose children by no means always maintain a medium between the characters of the two parents, but frequently resemble one—the father or the mother—much more strongly than the other.

How can this fact be theoretically explained? Must we ascribe to the ids of the father or of the mother a greater determining power? Without excluding such an assumption as on a priori grounds inadmissible, I am inclined to believe that we do not require it to explain this phenomenon. For, if we take our stand simply on the fact of the preponderance of one parent, it follows directly from this that not all the ids control the type of the child, let the cause of the non-co-operation of some of them be what it may. But if in this case only a portion of the ids contained in the germ-plasm controls the type, this combination of ids suffices to make the child resemble one parent, the father, for instance, and consequently half the number of ids is sufficient in some circumstances to determine the child—taking for granted that the one-sidedness of the inheritance is complete, which never actually happens. But the half number of ids can only suffice if it includes the same combinations of ids which have determined the type in the case of the father; as soon as one or more ids of this particular combination are replaced by others the paternal germ-plasm alone is not enough to call forth complete resemblance in the child.

But, at the reduction, a change of arrangement of ids takes place, and a new combination arises, and thus each germ-cell receives its particular group of ids. It may thus happen that, in one particular sperm-cell, exactly the same group of ids is contained as that which determined the type of the father, and that the same is true of a particular egg-cell in regard to the type of the mother. Let us now assume that a sperm-cell and an egg-cell meet, which contain both those groups of ids which had determined the type of the father and of the mother; if the determining power of the maternal and paternal ids were equal a child would result which would maintain the medium between father and mother.

As is well known this does happen not infrequently, although it is difficult or impossible to demonstrate it precisely. In plant-hybrids proof is easier, and it has been established that by far the greater number of hybrids maintain a medium between the characters of the two ancestral species. This proves that our assumption of equal strength of the ids of both species must be correct in general, for we know definitely in this case, as I shall show later, that the paternal and the maternal ids are equivalent as regards the characters of the species. This is the case, for instance, with the hybrids between the two species of tobacco-plant, Nicotiana rustica and N. paniculata, which were reared by Kölreuter as far back as the eighteenth century, and which then, as now, maintained a fairly exact medium between the two ancestral species, and did so in all the individuals. Both species thus strive to stamp their own character on the young plant, and in both the hereditary power is equally great; in both it is contained in the same number of ids, that is, in the half, for both kinds of sex-cell have undergone reducing division. We have here, then, strict proof that the half number of ids suffices to reproduce in the offspring the type of the species, or, more generally, of the parents.

If we apply these results to the inheritance of individual differences in Man, we may say, that those germ-cells, to which at the reducing division the same combination of ids has been handed on as that which already determined the type of the parent, will endeavour to impress this image again on the child. If a female cell of this kind combine with a male which likewise contains the facies-combination of the parent, in this case the father, the same thing will happen which we described in the case of the plant-hybrids, that is, a medium form between the type of the two parents will arise.

Not infrequently, however, there is a marked preponderance of the one parent in the type of the child, and we have to inquire whether the theory gives us any help with regard to such a case.

One might be inclined to assume a difference in the determining power of the paternal and maternal ids, but if we cannot show to what extent and for what reason this power may be different such an assumption remains rather an evasion than an explanation. Moreover, it would not always apply to the conditions in Man, for if, for instance, the ids of a particular mother were in general stronger than those of the father, all the children of the pair in question would necessarily take after the mother; but it happens not infrequently that one child resembles the father preponderantly, and another the mother. Moreover, the ids pass continually from the male to the female individual, and conversely, by virtue of the continuity of the germ-plasm, so that the idea that sex can have anything to do with the relative strength of the ids is altogether erroneous.

But, as I have already said, unilateral inheritance occurs even in the mingling of species-characters, and most clearly in the case of plant-hybrids. Thus, for instance, hybrids between the two species of pink, Dianthus barbatus and Dianthus deltoides, resemble the latter species much more closely than the former, and the hybrid between the two species of foxglove which are wild in Germany, Digitalis purpurea and Digitalis lutea, is much more like the latter than like the former.

It might be reasonably asked whether, in these crossings, the normal number of ids in one species is not greater than in the other. We know that, among animals at least, differences in the normal number of chromosomes occur even in very nearly related species. It is not impossible that this, in many cases, is really the cause of the diversity of transmitting power in different species. Nevertheless, we cannot rest satisfied with this, for, in the first place, this cause could not apply to the apparent unilateral inheritance from one parent in Man, since the normal number of ids, as far as we know, is strictly maintained in the same species, and second, this would not explain certain phenomena of inheritance in plant-hybrids.

It happens not only frequently, but usually, that the different parts of the hybrid take after one or other parent in different degree, and this is the case also with children. In the hybrid between the two species of tobacco-plant, Nicotiana rustica and N. paniculata, which I have already given as an example of a medium form between the two parents, such diversities occur regularly in all the hybrid individuals. Thus the corolla-tube of the hybrid is nearer N. paniculata in regard to its length, but nearer N. rustica in regard to its breadth. Many hybrids suggest one parent-form in the leaves, the other in the blossoms. In the same way in the child the form of eye may be that of the father, the colour of the iris that of the mother, the nose maternal, the mouth paternal—in short, the preponderance in heredity swings hither and thither from part to part, from organ to organ, from character to character, and this is even the rule though the oscillations may not always be apparent and are often invisible.

If we think of the proposition we arrived at earlier, and which was proved chiefly by the case of identical twins, that the facies or 'type' of the descendant is determined at fertilization, we may be inclined to regard such an oscillation of the hereditary tendencies as almost impossible, for it means that, with the given mingling of parental germ-plasms, the potency of inheritance from the two parents in every part of the offspring is determined once for all in advance. But the case of identical twins corroborates these oscillations, for in them, too, the father predominates in one part, the mother in another, and it proves, at the same time, that these oscillations do not depend on any chances whatever in development, but that they are exactly predetermined in the mingling of the hereditary substances in the germ-plasm of the fertilized ovum, and are strictly adhered to throughout development.

This fact can only be explained thus: the primary constituents of the different parts and characters of the body are contained in the parental germ-plasm in varying degrees of hereditary or transmissive strength, and this can be understood very well from our point of view without putting anything new into the 'portmanteau' of our theory (Delage).

But I must digress a little in order to make this plain.

When, in speaking of plant-hybrids, I said that the collective ids of the germ-plasm of a species must be equivalent in regard to the characters of the species, I did not speak quite precisely; in the majority of ids, in many cases in an overwhelming majority, this must be the case, but not actually in all, at least not on the assumption we make that the transformation of species is accomplished under the control of natural selection.

Let us recall what we have already established in regard to the evolutionary power of natural selection, namely, that the changes which it controls can never transcend the range of their utility, and it will be clear to us that, of the many ids which make up the germ-plasm of the species, only so many will be modified as are necessary to evoke the character which has varied. Just as the protective resemblance of an insect to a leaf may be raised to a very high pitch, but can never become perfect, because an imperfect resemblance is already sufficient to deceive the persecutor, and the selective process comes to a standstill because individuals which possessed a still greater resemblance to a leaf would be no better protected from destruction than the others, so in the modification of a species the whole of the ids need not at once be modified, if a majority is sufficient to stamp the great majority of individuals with the desired variation. But it may happen that, at the reduction of ids during the development of the germ-cells, an id-combination with wholly or almost wholly unchanged ids may come together in one germ-cell, and if another sperm-cell of this kind meets with an egg-cell similarly constituted, an individual of the old species must arise. But this must—on our assumption—be at a disadvantage as compared with the transformed individuals in the struggle for existence, and will perish in it, and therefore the number of unmodified ids in the germ-plasm of the species will gradually diminish. It is obvious, however, that this will take place very slowly, as we may conclude from the phenomena of reversion, of which I shall have to speak later on.

But what is true of the ids is true also of their constituent parts, the determinants, and that—if I mistake not—is fundamental in the interpretation of the alternation of hereditary succession in the parts of the child.

According to our theory, the ids do not collectively exert a controlling influence on the cells, not even on the germ-cells, whose histological differentiation into spermatic or egg-cells can only depend on control through specific sex-cell determinants. It is the different determinants of the ids that control; transformations of the species will, it is true, depend on transformations of the ids, but this need not necessarily consist in a variation of all the determinants of the id. If, for instance, two species of butterfly, Lycæna agestis in Germany and Lycæna artaxerxes in Scotland, only differ from each other in that the black spot in the middle of the wing in L. agestis is milk-white in L. artaxerxes, no other determinants in the id of the germ-plasm can be different except those which control this particular spot. In a majority of the ids in L. artaxerxes the determinants of this spot must have been modified, let us say, to the production of 'milk-white.' This majority will increase very slowly if the white colour has no pronounced advantage for the persistence of the species, but it will increase gradually, as we have already seen, though extremely slowly, through the elimination of those individuals whose germ-plasm at the reducing division has chanced to receive a majority of ids with the old, unmodified determinants, and which have therefore reverted to the ancestral form. This will happen whenever the new character has any use, however small, in maintaining the species.

But in most modifications of species quite a number of parts and characters have undergone variation either simultaneously or in rapid succession; in many cases nearly all the details of structure, and therefore almost all the determinants of the germ-plasm, must have varied. We must not, however, assume that all the equivalent determinants, for instance, all the determinants K in all the ids, have varied[9], and above all we must not take for granted that the determinants of different characters or parts of the body, for instance, the determinants L, M, or N, must all undergo variation in an equal number of ids. It will depend on two factors whether a new character is implicit, in the form of varied determinants, in a small or in a very large majority of ids: first, on the relative age of the character, and, secondly, on its value in relation to the persistence of the species. The more important a character is for the species the more frequently is it decisive for the life or death of the individual, and the more sharply will individuals not possessing it be eliminated, and the more rapidly, therefore, will those whose germ-plasm still contains a majority of the unvaried determinants of this character tend to disappear. In this way these determinants will tend to sink down from generation to generation to an ever smaller minority in the germ-plasm of those that survive.

[9] By equivalent or homologous determinants I mean the determinants of different ids which determine similar parts, e.g. the scales of that wing-spot in Lycæna agestis which is alluded to above, and to which we must refer again in more detail.

Thus in the ids of any species which has been in some way transformed—and that is as much as to say, in every species—the equivalent or homologous determinants are modified in a very varied percentage. A very modern and at the same time not very important character K´ will only be contained in a small majority of ids, while in the remainder the original homologous ancestral determinant K is contained; an older but not very much more important character M´ must have its determinants in a larger majority in the ids, while a character V´ of decisive importance for the preservation of the species, if it has been in existence long enough, must be represented in almost all the ids, so that the homologous unvaried determinants of the ancestral species V can only have persisted in an id here and there.

If this argument be correct, many phenomena of inheritance become intelligible, especially the variability in the expression of the inheritance in the parts of the offspring, which is more or less rigidly predetermined at fertilization. For the germ-plasm thus contains in advance every kind of determinant in diverse nuances, and in definite numerical proportions. In a plant N´, for instance, Ba´ may be the determinant of the modern leaf-form, and may occur in twenty-two out of twenty-four ids of the germ-plasm, while the two remaining ids still contain unmodified the old leaf-form determinants Ba, which the ancestral form N possessed. But the flower of N´ may be of still more recent origin, and contain the modern flower determinants Bl´ only in sixteen out of twenty-four ids, while in the other eight the old flower-determinant Bl of the ancestral form has persisted. Let us now suppose that another nearly related species P´ has, conversely, a recently changed leaf-form but a very ancient form of flower, so that the former is represented only in sixteen ids by the determinants of the leaf ba´ and the latter in twenty-two ids by the flower-determinants bl´: it is obvious that when the two species are crossed, notwithstanding the equal number of ids in the germ-plasm, the leaves of the hybrid will resemble more closely those of the ancestral form N, and the flowers those of the form P; it is even conceivable that in such a case the numerically preponderating leaf-determinants N, and the equally preponderating flower-determinants of P may form a close phalanx, so to speak, against the much less numerous homologous determinants of the other species, and that against this power working in a definite direction the others can make no headway and are simply condemned to inaction.

How we may or can picture this as occurring is a question which of course admits only of being answered very hypothetically, and it leads us, moreover, into the region of the fundamental phenomena of life, with the interpretation of which we are not here concerned. For the present we have assumed that life is a chemico-physical phenomenon, and we have postponed the deeper explanation of it to the remote future, that we may confine ourselves in the meantime to the solution of the problem of inheritance on the basis of the forces resident in the vital elements. But we may, nevertheless, make the supposition that a kind of struggle between the different kinds of biophors may take place within the cell, if the homologous determinants of all the ids for the control of the cell have entered into it.

In many cases this struggle will be decided by the numerical preponderance of one kind of determinant over the other, but it is certainly conceivable that dynamic differences may also have something to do with it.

Let us, however, abstain from trying to penetrate further into the obscurity of these processes, and let us content ourselves with establishing that the preponderance of one parent in some or many parts of the child may be almost if not quite complete, and that this compels us to assume that the hereditary substance of the other parent is in such cases rendered inoperative—for we know it is present—since the ids of both parents all go through the whole ontogeny, and are contained in every somatic cell.

Upon this struggle between homologous determinants depends the possibility of the entire suppression or inhibition of the influence of one parent, the whole diversity in the mingling of paternal and maternal character in the body of the offspring. It is in this that we must seek the explanation of the fact that not only whole bodily parts of the child, such as arms, legs, the nature of the skin, the form of the skull, may take after sometimes the father, sometimes the mother wholly or predominantly, but that the small separate subdivisions of a complex organ may sometimes turn out more maternal, sometimes more paternal. Thus intelligence from the mother and will from the father, musical talent from the father and a talent for drawing from the mother, may be inherited by the same child. I do not doubt that genius depends in great part on a happy combination of such mental endowments of the ancestors in one child. Of course something more is necessary, namely, the strengthening of certain of these hereditary endowments, but of this we shall speak later.

It is, however, not only the immediate ancestors, that is to say, the parents, that have to be taken account of in this mingling of hereditary contributions, but also those more remote. Not a few characters in the child do not occur in either parent, but were present in the grandparents, and their reappearance is called 'atavism' or 'reversion.' Let us consider this phenomenon in more detail, and try to find out whether and how far it can be interpreted by means of our theory.

The simplest and clearest cases are again found among plant-hybrids. It may happen, for instance, that a hybrid between two species, when dusted with its own pollen, gives rise to descendants, some of which resemble only one of the ancestral forms: thus we have reversion to one of the grandparents. The explanation of this lies in the different modes in which the reducing divisions are effected; if they take place in such a manner that all the paternal ids of the hybrid are separated from the maternal ids, then the result is germ-cells which are like those of the grandparents, that is, those of the parent species, and these, if they happen to combine in amphimixis, must give rise to a pure seedling of one or other of the two ancestral species. This case occurs less rarely than was formerly supposed, and than it could do if absolutely free combination of the idants took place at reduction. If combination were quite unrestricted, all other possible combinations would be likely to occur as frequently as these. But recent experiments have shown that, in many plant-hybrids, the germ-cells of the hybrids which are fertilized by their own pollen are either purely paternal or purely maternal. There cannot, therefore, be free combination of the idants at the reducing divisions; the idants of the two parent-forms separate from one another and do not combine. It is doubtful, however, whether the same thing occurs within a race, for instance, in the case of reproduction within a human race.

In Man reversion to a grandparent occurs not infrequently, and we may explain it thus: the id-group which controlled the type of the grandfather was also contained in the germ-cell which gave rise to the existence of the father, but it did not dominate the type in that case because a more powerful id-group was opposed to it in the germ-cell of the grandmother. When, later on, at the reducing divisions of the germ-cells of the father, this id-group again arrived in the sperm-cells of the father, it would predominately control the type of the child, that is, of the third generation, provided that the egg-cell with which it combined contained a weaker id-group.

In the case of ordinary plant-hybrids what are designated reversions can only be called so in a wide sense, for the ancestral characters are contained visibly in the parent, although mingled with those of the other parent. In human families, however, there are undoubted cases in which one or more characters of the grandparent reappear in the child which were not in any visible way expressed in the parent, and must therefore have been contained in the parent's germ-plasm in a latent state. And there are both in animals and plants reversions to ancestors lying much further back, to characters and groups of characters which have not been visible for many generations, and the occurrence of these can only be explained on the assumption that certain groups of ancestral determinants have been carried on in the germ-plasm in too small a number to be able to give rise ordinarily to the relevant character. Such isolated determinants may, however, in certain circumstances be strengthened by the amphimixis of two germ-cells both containing small groups of them, and thus augmented they may gain a controlling influence. In this case the chances of the reducing divisions have a part to play, since they bring together the old unvaried ancestral determinants which, as we have seen, may persist in the germ-plasm of any species through a long series of generations. This would, of course, only suffice to bring about a reversion if the determinants of the ancestral species were still contained in the germ-plasm in comparative abundance. If this is no longer the case, something more is necessary, and that is the relative weakness of the more modern determinants.

If two white-flowered species of thorn-apple, Datura ferox and Datura lævis, be crossed, there arises a hybrid with bluish violet flowers and brown stalks instead of green. This was interpreted by Darwin as a reversion to violet-flowering ancestral species on both sides, for there are even now a great number of species of Datura with violet flowers and brown stalks. When the two white forms are crossed the reversion takes place every time, not merely in some cases, and we may conclude from this that in both these species there is still such a strong admixture of the same unvaried ancestral ids that they always excel the ids of the two modern species crossed, in strength though certainly not in number. And this superiority must again depend on the fact that similar determinants of the same part are cumulative in their effect, while dissimilars are not.

For this reason reversions to remote ancestors occur readily when species and breeds are crossed, while they are rare in the normal inbreeding of a species. The reversions of the breeds of pigeon to their wild ancestral form, the slaty-blue rock-pigeon, never result, as Darwin showed, and as we have already noticed, from pure breeding of one race, but only when two or more breeds are repeatedly crossed with one another. Even then it occurs by no means always, but only now and again. The germ-plasm of the breeds must therefore still contain ids of the rock-pigeon, but in a small number, varying from individual to individual. If by fortunate reducing divisions and the meeting together of a sperm-cell rich in ancestral ids with a similarly endowed egg-cell the number of ancestral ids be raised to such a point that it exceeds the number of modern-breed ids contained in each one of the conjugating germ-cells, the ancestral ids control the development and reversion occurs, for the ancestral ids together have a cumulative effect while the ids of the two parent breeds are different and therefore, as far as they are so, cannot be co-operative in their influence. But it must be understood that they need not be different as far as all their determinants are concerned, but usually only as regards some groups, and thus it happens that reversion does not occur in regard to all, but only in regard to particular characters—thus in the Datura hybrids, chiefly in regard to the colour of the flowers and the stem, and in the hybrids between different breeds of pigeon mainly in regard to the colour and marking of the plumage.

The reversions of the horse and ass to striped ancestors, which Darwin has made famous, go much further back into the ancestral history of the species, for while we know the ancestral form of the domestic pigeon in the still living rock-dove (Columba livia), the ancestral form common to the horse and the ass is extinct, and we can only suppose that it was striped like a zebra, because such striping occasionally occurs in pure horses and pure asses, at least in their youth, although now only on the legs, and because this striping is often more marked in the hybrid between the horse and the ass, the mule. In Italy, where one sees hundreds of mules, the striping is not exactly frequent, but it may occur in about two per cent., while in America it is said to be much more frequent. The germ-plasm of the horse and the ass must therefore contain, in varying numbers, ids whose skin-colour determinants represent in part still unmodified ancestral characters. When two germ-cells chance to meet in fertilization, both of which have received, through a favourable reducing division, a relatively large number of such ids, a relative majority of these in the fertilized ovum is opposed to the dissimilar and therefore mutually neutralizing homologous determinants of horse and ass, and reversion to the ancestral form occurs.

These cases of reversion are enough to show us that the old unmodified ancestral determinants may persist in the germ-plasm through long series of generations. But an even deeper glimpse into the dim ancient history of our modern species of horses is afforded by the occurrence of three-toed horses, references to a small number of which the palæontologist Marsh was able to discover in literature, and one of which he was able to observe in life. Julius Caesar possessed a horse whose three-toed feet represented a reversion to the horses of Tertiary times, Mesohippus, Miohippus, and Protohippus or Hipparion; for all these genera possessed, in addition to the strong middle toe, two weaker and shorter lateral toes.

In the germ-plasm of our modern horses there must still persist in certain ids the determinants of the ancestral foot, which, after a long succession of favourable reducing divisions combined with favourable chances of fertilization, may come to be in a majority, and may thus be able to induce a reappearance of a character which has long been hidden under the surface of the present-day type of the species.

I do not propose to enter further on the discussion of the phenomena of inheritance. A more detailed investigation of the phenomena of reversion is to be found in 'The Germ-plasm' published ten years ago; and the discussion could not be resumed here without a critical consideration of a relatively large series of newly acquired facts not always harmonizing, and, as yet, not even fully available. The year 1900 has given us the investigations of three botanists, De Vries, Correns, and Tschermak, who have sought by experiments in hybridization between different sorts of peas, beans, maize, and other plants, to throw light on the phenomena of inheritance, and thus on the actual processes which occur in the germ-plasm at the reducing division. This led to the discovery that similar experiments had been published as far back as 1866 by the Abbot of Brünn, Gregor Mendel, and that these had been formulated as a law which is now called Mendel's law. Correns showed, however, that this law, though correct in certain cases, did not by any means hold good in all, and we must thus postpone the working of this new material into our theory until a very much wider basis of facts has been supplied by the botanists. There is less to be hoped for from the zoologists in regard to this problem owing to the almost insuperable difficulties in the way of a long series of experiments in hybridization in animals. I myself have repeatedly attempted experiments in this direction, and have always had to abandon them, either because the crossing succeeded too rarely, or because the hybrids did not reproduce among themselves, or did so defectively, or because the distinguishing characters of the crossed breeds proved insufficiently tenacious or diagnostic. But it would be a fine task for zoological gardens to undertake such experiments from the point of view of the germ-plasm theory, and their success would afford material for the criticism of the theory, the more valuable because it is apparent from the experiments on plants that the processes of heredity are manifold, and are far from being uniform in different domains[10].

[10] Castle and Allen have recently published the results of experiments in crossing white mice with grey, and these confirm Mendel's Law.

I have assumed for my theory that the reducing division took place according to the laws of chance, and that thus every combination of ids occurred with equal frequency. This assumption seems to be confirmed, by the experiments of the botanists I have mentioned, only in so far that in the crossing of hybrids with one another every combination of distinctive characters occurred with equal frequency. But, on the other hand, the splitting of the germ-plasm at the reducing division seems, as I said before, in many cases to take place in such a way that the id-groups of the two parents are discretely separated from one another; this was so in the stocks, peas, beans, and other hybrids. But even if this were always the case in these, we could hardly infer that it must be the same everywhere; we should rather expect that the relationship of the two parents and their ids would bring into play the finer attractions and repulsions between the ids of the germ-plasm, and would thus determine their arrangement and grouping. Further investigations may clear up this point; in the meantime we can only say that already—even among hybrids—many deviations from Mendel's Law have been established, for instance, by Bateson and Saunders (1902).

Before I conclude this lecture I should like to refer briefly to a phenomenon which Darwin was acquainted with and sought to explain through his theory of Pangenesis, but which at a later date was regarded as not sufficiently authenticated to justify any attempt at a theoretical explanation, since it seemed to contradict all our conceptions of hereditary substance and its operations. I refer to the phenomenon to which the botanists have given the pretty name of Xenia (guest-gifts), and which consists in the fact that in crosses of two different plants the characters of the male may appear not only in the young plant but even in the seed, so that a transference of paternal characters seems to take place from the pollen-tube to the mother, to the 'tissue of the maternal ovary.' In heads of yellow-grained maize (Zea) it is said that, after dusting with pollen from a blue-seeded variety, blue seeds appear among the yellow, and similar observations on other cultivated plants have been on record for more than half a century. Thus dusting the stigma of green varieties of grape with the pollen of a dark blue kind is said frequently to give rise to dark blue fruits.

Darwin accepted these observations as correct, and endeavoured to explain them as due to a migration of his 'gemmules' from the fertilized ovum into the surrounding tissue of the mother-plant. His explanation was not correct, we can say with confidence now, but he was right so far, for the phenomena of Xenia do occur; they are not illusory as most modern botanists seem inclined to believe. I myself was at first inclined to wait for further facts in proof that the phenomena of Xenia really occurred before attempting to bring them into harmony with my theory, and this will not be found fault with when it is remembered that these cases of Xenia seem to stand in direct contradiction to the fundamental postulates of the germ-plasm theory. For this depends essentially on a definite stable structure of the germ-substance, which lies within the nucleus in the form of chromosomes, and which cannot pass from one cell to another in any other way than by cell-division and division of the nucleus; how then could it pass from the fertilized ovum to the cells of the endosperm which do not derive their origin from it at all, but from other cells of the embryo-sac? In point of fact some of my opponents have cited Xenia as an actual refutation of my theory.

Fig. 82. Fertilization in the Lily (Lilium martagon), after Guignard. A, the embryo-sac before fertilization; sy, synergidæ; eiz, ovum; op and up, upper and lower 'polar nuclei'; ap, antipodal cells. B, the upper part of the embryo-sac, into which the pollen-tube (pschl) has penetrated with the male sex-nucleus (♂k) and its centrosphere; below that is the ovum-nucleus (♀k) with its (also doubled) centrosphere (csph). C, remains of the pollen-tube (pschl); the two sex-nuclei are closely apposed. Highly magnified.

That cases of Xenia really do occur is now established by the comprehensive and at the same time exceedingly careful experiments recently made by C. Correns with Zea Mais; it is only necessary to look through the beautiful figures with which his work is adorned to be convinced that heads of maize whose blossoms have been dusted with the pollen of a different kind produce more or less numerous seeds of the paternal kind, usually mingled with those of the maternal. Thus heads of the variety Zea alba resulting from fertilization with Z. cyanea exhibited a majority of white grains, but among them a smaller number of blue; and the converse experiment, of dusting Z. cyanea with the pollen of Z. alba, yielded heads in which a minority of white grains appears among a majority of blue. But it is always only the nutritive layer surrounding the embryo—the endosperm—which exhibits the character of the paternal species, and even the capsule surrounding the seed shows nothing of it, but is purely maternal. Thus the heads of different species with pale-yellow capsule, when dusted with the pollen of Z. rubra, never have red seeds like those of Z. rubra, but always seeds with a pale-yellow skin, while, in the converse experiment, dusting of the red-skinned species Z. rubra with pollen from Z. vulgata, all the seeds are red, like those of the maternal species, and the influence of the paternal species only shows when the strong red skin has been removed, so that the intense yellow colour of the endosperm, which in the pure maternal species is white, is exposed to view. Thus the mysterious influence of the pollen never goes beyond the endosperm, and the riddle of this influence is solved in the most unexpected manner, indeed was solved even before Correns had securely established the genuine occurrence of Xenia. The explanation is due to recent disclosures in regard to the processes of fertilization in flowering plants.

It had long been known that the pollen-tube contains not merely one generative nucleus but two, which arise from one by division. But what had till recently remained unknown was that not only one of these penetrated into the embryo-sac to enter into amphimixis with the egg-cell, but that the other also makes its way in, and there fuses with the two nuclei which had long been designated the upper and lower polar-nuclei ([Fig. 82, op. cit.]). Nawaschin and Guignard demonstrated that these two nuclei fuse with the second male nucleus; thus two acts of fertilization are accomplished in the embryo-sac, and one of these gives rise to the embryo, while the second becomes nothing less than the endosperm, the nutritive layer which surrounds the embryo, whose origin from the polar nuclei had been previously recognized.

Thus the riddle of Xenia is essentially solved. We understand how paternal primary constituents may find their way into the endosperm, and indeed must do so regularly; we understand also how the paternal influence never goes beyond the endosperm. The riddle is thus not only solved, but at the same time the view which assumes a fixed germ-plasm, and believes it to lie in the nuclear substance of the germ-cells, receives further confirmation, if it should need any, for if facts which are apparently contradictory to a theory can be naturally brought into harmony with it, this affords a stronger argument for the correctness of the theory than the power of explaining the facts which were used in building it up.

There is much more to be said in regard to Xenia, and I am sure that much that is of interest will be brought to light by deeper investigation; theoretical difficulties will still have to be overcome, and I have already pointed out one of these in my 'Germ-plasm,' but I must here rest satisfied with what has been already said.

We have now passed in review and attempted to fit into the theory a sufficiently large number of the phenomena of heredity for the purpose of these lectures. Although, as is natural, much of this must remain hypothetical, we may accept the following series of propositions as well founded: there is a hereditary substance, the germ-plasm; it is contained in very minimal quantity in the germ-cells, and there in the chromosomes of the nucleus; it consists of primary constituents or determinants, which in their diverse arrangement beside or upon one another form an extremely complex structure, the id. Ids and determinants are living vital unities. Each nucleus contains several, often many, ids, and the number of ids varies with the species and is constant for each. The ids of the germ-plasm of each species have had a historical development, and are derived from the germ-plasm of the preceding lineage of species; therefore ids can never arise anew but only through multiplication of already existing ids.

And now, equipped with this knowledge, let us return to the point from which we started, and inquire whether the Lamarckian principle of evolution, the inheritance of functional modifications, must be accepted or rejected.

LECTURE XXIII

EXAMINATION OF THE HYPOTHESIS OF THE
TRANSMISSIBILITY OF FUNCTIONAL MODIFICATIONS

Darwin's Pangenesis—Alleged proofs of functional inheritance—Mutilations not transmissible—Brown-Séquard's experiments on Epilepsy in guinea-pigs—Confusion of infection of the germ with inheritance, Pebrine, Syphilis, and Alcoholism—Does the interpretation of the facts require the assumption of the transmission of functional modifications?—Origin of instincts—The untaught pointer—Vom Rath's and Morgan's views—Attachment of the dog to his master—Fearlessness of sea-birds and seals on lonely islands—Flies and butterflies—Instincts exercised only once in the course of a lifetime.

As I have already said in an earlier lecture, Darwin adhered to Lamarck's assumption of the transmission of functional adaptations, and perhaps the easiest way to make clear the theoretical difficulties which stand in the way of such an assumption is to show how Darwin sought to present this principle as theoretically conceivable and possible.

Darwin was the first to think out a theory of heredity which was worthy of the name of theory, for it was not merely an idea hastily suggested, but an attempt, though only in outline, at elaborating a definite hypothesis. His theory of 'Pangenesis' assumes that cells give rise to special gemmules which are infinitesimally minute, and of which each cell brings forth countless hosts in the course of its existence. Each of these gemmules can give rise to a cell similar to the one in which it was itself produced, but it cannot do this at all times, but only under definite circumstances, namely, when it reaches 'those cells which precede in order of development' those that it has to give rise to. Darwin calls this the 'elective affinity' of each gemmule for this particular kind of cell. Thus, from the beginning of development there arises in every cell a host of gemmules, each of which virtually represents a specific cell. These gemmules, however, do not remain where they originated, but migrate from their place of origin into the blood-stream, and are carried by it in myriads to all parts of the body. Thus they reach also the ovaries and testes and the germ-cells lying within these, penetrate into them, and there accumulate, so that the germ-cells, in the course of life, come to contain gemmules from all the kinds of cells which have appeared in the organism, and, at the same time, all the variations which any part may have undergone, whether due to external or internal influences, or through use and disuse.

In this manner Darwin sought to attribute to the germ-cells the power of giving rise, in the course of their development, to the same variations as the individual had acquired during its lifetime in consequence of external conditions or functional influences.

I abstain from analysing the assumptions here made; their improbability and their contradictions to established facts are so great that it is not necessary to emphasize them; the theory shows plainly that it is necessary to have recourse to very improbable assumptions, if an attempt is to be made to find a theoretical basis for the transmission of acquired (somatogenic) characters. Even when Darwin formulated his theory of Pangenesis his assumptions were hardly reconcileable with what was known of cell-multiplication; now they are above all irreconcileable with the fact that the germ-substance never arises anew, but is always derived from the preceding generation—that is, with the continuity of the germ-plasm.

If we were now to try to think out a theoretical justification we should require to assume that the conditions of all the parts of the body at every moment, or at least at every period of life, were reflected in the corresponding primary constituents of the germ-plasm and thus in the germ-cells. But, as these primary constituents are quite different from the parts themselves, they would require to vary in quite a different way from that in which the finished parts had varied; which is very like supposing that an English telegram to China is there received in the Chinese language.

In spite of this almost insuperable theoretical obstacle, various authors have worked out the idea that the nervous system, which connects all parts of the body with the brain and thus also with each other, communicates these conditions to the reproductive organs, and that thus variations may arise in the germ-cells corresponding to those which have taken place in remote parts of the body.

Even supposing it were proved that every germ-cell in ovary or testis was associated with a nerve-fibre, what could be transmitted to it by the nerves, except a stronger or weaker nerve-current? There is no such thing as qualitative differences in the current; how then could the primary constituents of the germ be influenced by the nerve-current, either individually or in groups, in harmony with the organs and parts of the body corresponding to them, much less be caused to vary in a similar manner? Or are we to imagine that a particular nerve-path leads to every one of the countless primary constituents? Or does it make matters more intelligible if we assume that the germ-plasm is without primary constituents, and suppose that, after each functional variation of a part, telegraphic notice is sent to the germ-plasm by way of the brain as to how it has to alter its 'physico-chemical constitution,' so that the descendants may receive some benefit from the acquired improvement?

I am not of the number of those who believe that we already know all, or at least nearly all, that is essential, but am rather convinced that whole regions of phenomena are still sealed to us, and I consider it probable that the nervous system in particular is not yet exhaustively known to us, either in regard to its functioning or in regard to its finest structural architecture, although I gratefully recognize the advances in this domain that the last decades have brought about. In any case, such assumptions as I have just indicated, or similar ones, seem to me quite too improbable to furnish any foothold for progress. Yet we must always remain conscious that we cannot decide as to the possibility or impossibility of any biological process whatever from a purely theoretical standpoint, because we can only guess at, not discern, the fundamental nature of biological processes. At the close of this lecture I shall return to the question of the theoretical conceivability of an inheritance of functional adaptations; but first of all we must consider the facts and be guided by them alone. If they prove, or even make it seem probable, that such inheritance exists, then it must be possible, and our task is no longer to deny it, but to find out how it can come about.

Let us therefore investigate the question whether an inheritance of acquired characters, that is, in the first place, of functional adaptations, is demonstrable from experience. We shall speak later on of the effect of climatic and similar influences in causing variation; the case in regard to them is quite different, because they undoubtedly affect not only the parts of the body but the germ-cells as well.

When we inquire into the facts which have been brought forward by the modern adherents of the Lamarckian principle as proofs of the inheritance of acquired characters in this restricted sense, we shall find that none of them can withstand criticism.

First, there are the numerous reputed cases of the inheritance of mutilations and losses of whole parts of the body.

It is not without interest to note here how opinion in regard to this point has altered in the course of the debate.

At the beginning of the discussion they were all brought forward as evidence of undoubted value for the Lamarckian principle.

At the Naturalists' Congress in Wiesbaden in 1887, kittens with only stumps of tails were exhibited, and they were said to have inherited this peculiarity from their mother, whose tail, it was asserted, had been accidentally amputated. The newspapers reported that the case excited great interest, and biologists of the standing of Rudolf Virchow declared it to be noteworthy, and regarded it as a proof, if all the details of it were correct. From many sides similar cases were brought forward, intended to prove that the amputation of the tail in cats and dogs could give rise to hereditary degeneration of this part; even students' fencing-scars were said to have been occasionally transmitted to their sons (happily not to the daughters); a mutilated or torn ear-lobe in the mother was said to have given rise to deformity of the ear in a son; an injury to a father's eye was said to have caused complete degeneration of the eyes in his children; and deformity of a father's thumb, due to frostbite, was said to have produced misshapen thumbs in the children and grandchildren. A multitude of cases of this kind are to be found in the older textbooks of physiology by Burdach, and above all by Blumenbach, and the majority have no more than an anecdotal value, for they are not only related without any adequate guarantee, but even without the details indispensable to criticism.

As far back as the eighteenth century the great philosopher Kant, and in our own day the anatomist Wilhelm His, gave their verdict decidedly against such allegations, and absolutely denied any inheritance of mutilations; and now, after a decade or more of lively debate over the pros and cons, combined with detailed anatomical investigations, careful testing of individual cases, and experiment, we are in a position to give a decided negative and say there is no inheritance of mutilations.

Let me briefly explain how this result has been reached.

In the first place, the assertion that congenital stump-tails in dogs and cats depended on inherited mutilation proved to be unfounded. In none of the cases of stump-tails brought forward could it even be proved that the tail of the relevant parent had been torn or cut off, much less that the occurrence, in parents or grandparents, of short tails from internal causes was excluded. At the same time anatomical investigation of such stump-tails as occur in cats in the Isle of Man, and in many Japanese cats, and are frequently found in the most diverse breeds of dogs, showed that these had, in their structure, nothing in common with the remains of a tail that had been cut off, but were spontaneous degenerations of the whole tail, and are thus deformed tails, not shortened ones (Bonnet).

Experiments on mice also showed that the cutting off of the tail, even when performed on both parents, does not bring about the slightest diminution in the length of tail in the descendants. I have myself instituted experiments of this kind, and carried them out through twenty-two successive generations, without any positive result. Corroborative results of these experiments on mice have been communicated by Ritzema Bos and, independently, by Rosenthal, and a corresponding series of experiments on rats, which these two investigators carried out, yielded the same negative results.

When we remember that all the cases which have been brought forward in support of an inheritance of mutilations refer to a single injury to one parent, while, in the experiments, the same mutilation was inflicted on both parents through numerous generations, we must regard these experiments as a proof that all earlier statements were based either on a fallacy or on fortuitous coincidence. This conclusion is confirmed by all that we know otherwise of the effects of oft-repeated mutilations, as for instance the well-known mutilations and distortions which many peoples have practised for long, sometimes inconceivably long, ages on their children, especially circumcision, the breaking of the incisors, the boring of holes in lip, ear, or nose, and so forth. No child of any of these races has ever been brought into the world with one of these marks: they have to be re-impressed on every generation.

The experience of breeders agrees with this, and they therefore, as Wilckens remarks, have long regarded the non-inheritance of mutilations as an established fact. Thus there are breeds of sheep in which, for purely practical reasons, the tails have been curtailed quite regularly for about a century (Kühn); but no sheep with a stump-tail has ever been born in this breed. This is all the more important because there are other breeds of sheep (fat-rumped sheep) in which the lack of the tail is a breed character; it is thus not the case that there is anything in the intrinsic nature of the tail of the sheep to prevent it becoming rudimentary. The artificially rounded ear of fox-terriers, too, though cut for generations, never occurs hereditarily. Mr. Postans of Eastbourne informs me that the cocks which are to be used for cock-fighting are docked of their combs and wattles beforehand, and that this had been done for at least a century, but that no fighting cock without comb and wattles has been reared. In the same way various breeds of dog, such as the spaniels, have had their tails cut to half their length regularly and in both sexes for more than a century, yet in this case there is no hereditary diminution of the length of tail. Deformed stump-tails do indeed occur in most breeds of dog, but, as I said before, their anatomical character is quite different from that of artificially shortened tails, moreover they may occur in breeds whose tails have not suffered from the fashion of docking, as, for instance, in the Dachshund.

We may therefore affirm that an inheritance of artificially produced defects and mutilations is quite unproved, and in no way bears out the supposed inheritance of functional changes.

This is now admitted by the great majority of the adherents of the Lamarckian principle, and we may now regard this kind of 'proof' as disposed of.

In addition to the above, various sets of facts have been brought forward as proofs, and in particular the much discussed experiments of Brown-Séquard on guinea-pigs, from which it was inferred that epilepsy artificially induced could be transmitted. But these experiments do not really prove anything in regard to the question at issue, because epileptic-like convulsions may have very various causes, and these are, for the most part, quite unknown. Since artificial epilepsy can be induced in guinea-pigs by the most diverse injuries to the central or peripheral parts of the nervous system, this of itself points to the fact that it is not a question of the mere lesion of anatomical structure, I mean, of the breaking of the continuity of a definite part, and of its transmission. The result would, in any case, differ according to whether certain centres of the brain, or half the spinal cord, or the main nerve-trunks were cut through. There must, therefore, be something more needed to produce the appearance of epilepsy—some morbid process which may arise at different parts of the nervous system, and be continued from them to the brain-centres. This is corroborated by the fact that it takes at least fourteen days, and often from six to eight weeks, for epilepsy to develop after the operation, and that in many cases it does not develop at all. I have made the suggestion that, during or after the operation, some kind of pathogenic micro-organism might easily reach the wounded parts, and there excite inflammation, which may extend centripetally to the brain. Similar processes have been observed in connexion with lymph-vessels, and why should they not occur in connexion with nerves?

It has been objected to this that the guinea-pig's epilepsy may be produced by blows on the skull, and also by a destructive compression of the nervus ischiadicus through the skin, and that in both cases the epilepsy may reappear in the following generation; and this, it is supposed, shows that the intrusion of microbes is excluded. If this were so beyond a doubt, and if we could exclude the possibility that there were previously various microbes within the body, which could only penetrate into the nervous substance after the cutting or destruction of the neurilemma, nothing would be gained that would in any way support the Lamarckian principle. One could only say: Certain injuries to the nervous system give rise secondarily in guinea-pigs to morbid phenomena like epilepsy, and all sorts of functional disturbances of the nervous system often appear in the next generation, including in rare cases even the phenomena of epileptic convulsions. That this is a case of the transmission of an acquired anatomical modification brought about by the injury is not only unproved, but is decidedly negatived, for the injuries themselves are never transmitted. Thus what is transmitted must be quite different from what was acquired, for no one has ever detected in the offspring the lesion of the nerve-trunk which was cut through in the parent, or any other result except the disease to which the original injury gives rise. Moreover, the inheritance of these morbid phenomena has been again brought into dispute quite recently owing to the investigations of such experts in nervous diseases as Sommer and Binswanger, and the correctness of Brown-Séquard's results, which have dragged through the literature of the subject for so long, has been emphatically denied[11].

[11] See H. E. Ziegler's report in Zool. Centralblatt, 1900, Nos. 12 and 13.

Clearly formulated problems, like that of the inheritance of acquired characters, should not be confused by bringing into them phenomena whose causes are quite unknown. What do we know of the real causes of those central brain-irritations which give rise to the phenomena of epilepsy? It is certain enough that there are diseases which are acquired and are yet 'inherited,' but that has nothing to do with the Lamarckian principle, because it is a question of infection of the germ, not of a definite variation in the constitution of the germ. We know this with certainty in regard to the so-called Pebrine, the silkworm disease which wrought such devastation in its time; the germs of the pebrine organism have been demonstrated in the egg of the silk-moth; they multiply, not at once but later, in the young caterpillar, and it is the half-grown caterpillar, or even the moth, that succumbs to the disease.

Whether in this case also the disease germs are transmitted through the male sex-cells is not proved, as far as I am aware, but that this can happen is shown by the transmission of syphilis from father to child. That in this case, also, the exciting cause of the disease is a micro-organism cannot be doubted, although it has not yet been proved. Thus even the minute spermatozoon of Man can contain microbes, and transmit them to the germ of a new individual.

This discussion of scientific questions ought not to be brought down to the level of a play upon words, by bringing forward cases like the above as evidence for the inheritance of 'acquired characters,' as was done, for instance, by M. Nussbaum, who cited as a proof of this the migration of the alga-cells which live in the endoderm of the green freshwater Hydra into the ovum, which is originally colourless, and originates in the ectoderm of the animal ([Fig. 35B, p. 169, vol. i]). It seems to me better to make a precise distinction between the transmission of extraneous micro-organisms through the germ-cells and the handing on of the germ-plasm with the characters inherent in its structure. Only the latter is inheritance in the strict scientific sense, the former is infection of the germ.

Still less than the cases of inherited traumatic epilepsy can the morbid constitution of the children of drunkards be regarded as a proof of the inheritance of somatogenic characters, though this has often been maintained. I will not lay any stress on the fact that the allegation itself is, according to the most competent observers, such as Dr. Thomas Morton[12], far from being established. But even if it were quite certain that the numerous diseases of the nervous system, amounting sometimes to mania, which are frequently observed in the children of drunkards, were really caused by the drinking of the parents, it ought not to be overlooked that we have here to do not with the hereditary transmission of somatic variations, but of variations directly induced in the germ-plasm of the reproductive cells, for these are exposed to the influence of the alcohol circulating in the blood, just as any other part of the body is. That by this means variations in the germ-plasm can be brought about, and that these may lead to morbid conditions in the children cannot be denied, and ought not on a priori grounds to be called in question. For we are acquainted with many other influences—climatic, for instance—which directly affect and cause variation in the germ-plasm. Whether this is so in the case of drunkenness, and in what manner it comes about, whether through direct action of the alcohol, or through infection of the germ with some microbe, we must leave to the future to decide; the whole question is out of place here; it can in no way help us to clear up the problem with which we are now occupied.

[12] Morton, 'The Problem of Heredity in Reference to Inebriety,' Proceed. Soc. for the Study of Inebriety, No. 42, Nov. 1894.

But even if there were not a trace of proof of the transmissibility of functional modifications, that alone would not justify us in concluding that the transmission is impossible, for many things may happen that we are not in a position to prove at present. If it could be shown that there was a whole group of phenomena that could not be explained in any other way than on the hypothesis of such inheritance, then we should be obliged to assume that it really occurred, although it was not demonstrable, and, indeed, not even theoretically conceivable. This is the standpoint of the adherents of the Lamarckian principle at present.

They say there are a great number of transformations which are simply and easily explained, if we regard them as the effects of inherited use or disuse, but which admit only of a strained explanation, and sometimes of none at all, on the basis of natural selection, and these are not a few isolated cases, but whole categories of them.

I will submit a few of these, and show at the same time why I cannot regard them as convincing, even if it be the case that we are not at present in a position to explain them without the aid of the Lamarckian principle. But let me hasten to add that it is my belief that we can do this, although certainly not without first giving a somewhat extended application to the principle of selection.

It has often been maintained that the existence of animal instincts is in itself enough to prove that the Lamarckian principle is operative. In one of the earlier lectures I showed that at least the greater number of instincts must have originated in purely reflex actions, and therefore, like these actions themselves, can only be explained through natural selection. A reflex action, such as coughing, sneezing, shutting of the eyelids, and so on, differs from an instinctive action in the lesser complexity and shorter duration of the series of movements liberated by a sense-impression, and also in that it does not require to enter into consciousness at all; but no very precise boundary can be drawn between the two, and, in any case, both depend, as we have already seen, on a quite analogous anatomical basis. It is only a difference in degree whether, at the sight of a rapidly approaching object, the muscles of the eyelids contract, and by shutting the lids, protect the eye, or whether the fly, which we intend to seize with our hand, is impelled by the sight of the rapidly approaching shadow of the hand to fly quickly up. The action of the fly may be regarded as reflex, or equally well as instinctive. But there is also only a difference in degree, not in kind, between this simple action and the complex and protracted behaviour of a mason-bee, the sight of whose colony impels her to fly out and fetch clay, with it gradually to build a neat cell, to fill this with honey, to lay an egg in it, and finally to furnish the cell with a roof of clay. Since all reflex mechanisms, and all the natural instincts of animals, contribute to the maintenance of the species, and are therefore useful, their origins must be referable to natural selection, and we have only to ask whether they must be referred to it always, and to it alone.

It cannot be doubted that, in Man, and in the higher animals voluntary actions which are often repeated gradually acquire the character of instinctive actions. The individual movements pertaining to the particular action are no longer each guided by the will, but a single exercise of will is enough to liberate the whole complex action, such as writing, speaking, walking, or the playing of a whole piece of music; frequently the will-impulse may be absent altogether, and the action be set going simply by an adequate external stimulus, as in the case of sleep-walking, which is observed in fatigued children and soldiers, and in somnambulists. The external stimulus is transmitted to the proper group of muscles as unfailingly as in the case of true instincts, and this happens not only in regard to actions which, like walking, are essential to the life of the species, but also in regard to those which have arisen from chance habits or exercises. Often a short practice is sufficient to make an action in this sense instinctive, and the complexity of the instinct-mechanism gained by such practice is often astounding. Under some circumstances a person may play a piece on the piano from the score, and yet be thinking intently of other things, and be quite unconscious of what is played. In the same way it may happen that a person dominated by violent emotion, when trying to free himself from it by reading, may read a whole page, line by line, without understanding in the least what has been read. In the last case it is not directly demonstrable that the reader has made all the complex delicate eye-movements which would be liberated by the sight of the words, but in the case of playing, the listeners can perceive that the piece is correctly played, and thus that the stimulus exercised by each note on the retina of the eye is translated into the complex muscular movement of arm and finger, corresponding both to the pitch and the duration of the note, and to the simultaneousness of several notes.

In all these cases it is probably not always quite new paths which are established in the brain, but use is made of particular tracks in the innumerable nerve-paths already existing in the nerve-cells (neurons) which are 'more thoroughly trodden' by practice, so that the distribution of the nerve-current takes place more easily along them than along others[13]. This much-used metaphor does not indicate the actual structural changes which have taken place, but it serves at least to indicate that we have to do with material changes in the ultimate living elements of the nerve-substance (nerve-biophors) whether these changes be in position or in quality. Now, if such brain-structures and mechanisms acquired through exercise in the individual life could be transmitted, new instincts would certainly arise in this way, and many naturalists hold this view still.

[13] This, however, is by no means intended to cast doubt on the possibility that quite new paths may arise during the individual life, as is made probable by the recent investigations of Apáthy, Bethe, and others.

If the inheritance of acquired characters had already been proved in other ways, we could not refuse to admit that it might play a part in the higher animals in the modification and new formation of instincts. We should then have to admit that habits can be inherited, and that instincts actually are or may be, as they have often been said to be, inherited habits. But to make the converse conclusion, and to infer from the result of the brain-exercise in the individual life and its similarity to inborn instincts that the latter also depend on inherited exercise, and that there must therefore be inheritance of acquired characters, is hardly admissible.

It might be all very well if there were no other explanation! But as instincts depend on material brain-mechanisms which are variable, like every other part of the body, and as, furthermore, they are essential to the existence of the species, and, down to the minutest detail, are adapted to the circumstances of life, there is no obstacle in the way of referring their origin and transformation to processes of selection.

It has been asserted that the results of training, for instance in dogs, can be inherited, since the untaught young pointer points at the game, and the young sheep-dog runs round and barks at the flock of sheep without biting them. It is, however, often forgotten that, not only have these breeds arisen under the influence of artificial selection by Man, but that they are even now strictly selected. My colleague and friend, Dr. Otto vom Rath, who unhappily died all too soon for Science, and who was not only a capable investigator, but an experienced sportsman, told me that huntsmen distinguish very carefully between the better and the inferior young in a litter, and that by no means every whelp of a pair of pointers can be used for hunting game-birds. Lloyd Morgan points out the same thing, and he is undoubtedly a competent judge in the domain of instinct; he confirms the statement that the pointer 'often points at the quarry, it may be a lark's nest, without instruction,' but he says at the same time, that the power is inborn in very varying degrees, and that, in his opinion, selection undoubtedly plays a part.

It must not, therefore, be believed that the habit of the pointer depends on training; it is only strengthened in each individual by training, but it depends on an innate predisposition to creep up to the game, and is thus a form of the hunting instinct. Man has taken advantage of this, and has increased it, but has certainly not ingrafted it into the breed by whipping. And something similar will be found to be true in all cases of so-called inheritance of the effects of training. It must not be forgotten what astounding results can be achieved in the individual by training. The elephant is the best example of this, for it only exceptionally breeds in captivity, and all the thousands of 'domesticated' elephants in India are tamed wild elephants. Yet they are as gentle and docile as the horse, which has been domesticated for thousands of years; they perform all kinds of tasks with the greatest patience and carefulness, in many cases without being under constant superintendence. They are indeed animals of great intelligence; they understand what is required of them, and they accommodate themselves readily to new conditions of life.

The attachment of the dog to its master and to Man generally has often been cited as a proof of the origin of a new instinct by the inheritance of acquired habitude; but the dog is a sociable animal even in a wild state, and by living in co-operative association with Man it has transferred its sociable affections to him. We find exactly the same thing in the elephant which has been caught wild and tamed. It is particularly emphasized by those who have accompanied animal transports in Africa that the young elephants are wild and malicious towards the blacks who teased and maltreated them, but complaisant and harmless towards the whites who treated them kindly. The attachment of elephants to their keepers and to every one who shows them kindness is familiar enough; it does not depend on a newly acquired impulse, but on the sociable impulse inherent in the species, which, in the wild state, causes them to live in fairly large companies, and on their inoffensive, timid, and, we may almost say, affectionate disposition.

Of course it is easy enough to give an imaginative theoretical interpretation of the origin of a new instinct from a newly acquired habit. We have often heard that sailors have found the birds in distant uninhabited islands quite free from fear; they let themselves be struck down with cudgels without attempting to escape. The extermination of the Dodo three centuries ago is a well-known example of this. Chun, in his magnificent work on the German Deep Sea Expedition of 1898, has recently communicated numerous interesting examples of the indifference of birds towards Man when they have not learned what his presence means: thus the sea-birds of Kerguelen, penguins, cormorants, gulls, 'kelp-pigeons' (Chionis), and others, behaved towards Man very much like the tame geese of our poultry yards. Even enormous mammals like the 'sea-elephant,' a seal with a proboscis-like prolongation of the nose, neither attempted to escape nor showed any hostility to man, but quietly let itself be caught. Similar tales were told by Steller in 1799, after he had been obliged to pass a winter with his sailors on an island in the Behring Straits. The numerous gigantic sea-cows (Rhytina stelleri) which lived there were so confiding that they allowed the boat to come quite up to them, and the sailors were able to kill many of them from time to time, using their flesh for food. But towards the end of the winter the animals began to be shy, and, in the following winter, when other sailors to the polar regions endeavoured to hunt them too, it was very difficult to secure them; they had recognized man as an enemy, and fled from him when they saw him from afar. Thus the same individuals which had earlier carelessly allowed man to come up to them now avoided him as an enemy. This was not instinct, it was a behaviour controlled by the will and founded on experience. But it would soon become 'instinctive' if the meeting with the enemy were often repeated, just like the winding-up of a watch, which is often done at a wrong time, for instance, on changing clothes during the day, and thus without reflection. It is quite easy to conceive that if the material brain-adaptation which causes flight without reflection at the sight of man were transmissible, the flight-instinct might become a congenital instinct in the species in question. But this assumption is unfounded; for, as is shown by the case of the sea-cow, we do not require it where the animal is of sufficient intelligence to perform by its own discernment the action necessary to its existence. The action may thus become 'instinctive' through exercise and imitation in the individual life, without however attaining to transmissibility.

But in many cases this is not enough, namely, in all cases in which the degree of intelligence is not sufficiently high, or where the flight movement must follow so rapidly that it would be too late if it had to be regulated by the will, as, for instance, the shutting of the lids when the eye is threatened, or the flight of the fly or the butterfly when an enemy approaches. Both fly and butterfly would be lost in every case if they had voluntarily to set the flight-movement going after they became conscious of danger. If they had first of all to find out from whom danger threatened no individual would escape an early death, and the species would die out. But they possess an instinct which impels them to fly up with lightning speed, and in an opposite direction, whenever they have a visual perception of the rapid approach of any object of whatever nature. For this reason they are difficult to catch. I once watched the play of a cat, ordinarily very clever at catching, as she attempted to seize a peacock-butterfly (Vanessa Io), which settled several times on the ground in front of her. Quietly and slowly she crept within springing distance, but even during the spring the butterfly flew up just before her nose and escaped every time, and the cat gave it up after three attempts.

In this case the beginning of the action cannot lie in a voluntary action, for the insect cannot know what it means to be caught and killed, and the same is true of innumerable still lower animal forms, the hermit-crabs and the Serpulids, which withdraw with lightning speed into their houses, and so forth. It seems to me important theoretically, that the same action can be liberated at one time by the will, at another by the inborn instinct-mechanism. In both cases quite similar association-changes in the nerve-centres must lie at the root of the animal's action, but in the first case these are developed only in the course of the individual life by exercise, while in the second they are inborn. In the former, they are confined to the individual, and must be acquired in each generation by imitation of older individuals (tradition) and by inference from experience, in the latter they are inherited as a stable character of the species.

It has been maintained by many that the origin of instincts through processes of selection is not conceivable, because it is improbable that the appropriate variations in the nervous system, which are necessary for the selective establishment of the relevant brain-mechanism, should occur fortuitously. But this is an objection directed against the principle of selection itself, and one which points, I think, to an incompleteness in it, as it was understood by Darwin and Wallace. The same objection can be made to every adaptation of an organ through natural selection; it is always doubtful whether the useful variations will present themselves, as long as they are due solely to chance, as the discoverers of the selective principle assumed. We shall attempt later to fill up this gap in the theory, but, in the meantime, I should like to point out that the process of selection offers the only possible explanation of the origin of instincts, since their origin through modifications of voluntary actions into instinctive actions, with subsequent transmission of the instinct-mechanism due to exercise in the individual life, has been shown to be untenable.

If any one is still unconvinced of this, I can only refer to the cases we have already discussed of instincts which are only exercised once in a lifetime, since, in these, the only factor that can transform a voluntary action into an instinctive one is absent, namely, the frequent repetition of the action. In this case, if any explanation is to be attempted at all, it can only be through natural selection, and as we have assumed once for all that our world does admit of explanation, we may say, these instincts have arisen through natural selection.

Even though it may be difficult to think out in detail the process of the gradual origin of such an instinctive activity, exercised once in a lifetime, such as that, for instance, which impels the caterpillar to spin its intricate cocoon, which it makes only once, without ever having seen one, and thus without being able to imitate the actions which produce it, we must not push aside the only conceivable solution of the problem on that account, for then we should have to renounce all hope of a scientific interpretation of the phenomenon. We may ask, however, whether there is not something lacking in our present conception of natural selection, and how it comes about that useful variations always crop up and are able to increase.

But if we must explain, through natural selection, such complex series of actions as are necessary to the making of the cocoon of the silkworm or of the Saturnia moth (Saturnia carpini), what reason have we for not referring other instincts also to selection, even if they be repeated several times, or often, in the course of a lifetime? It is illogical to drag in any other factor, if this one, which has been proved to operate, is sufficient for an explanation.

Thus, as far as instincts are concerned, there is no necessity to make the assumption of an inheritance of functional changes, any more than there is in regard to any purely morphological modifications. As the instincts only exercised once show us that even very complicated impulses may arise without any inheritance of habit, that is, without inheritance of functional modification, so there are among purely morphological characters not a few which, though effective, are purely passive, which are of use to the organism only through their existence, and not through any real activity, so that they cannot be referred to exercise, and therefore cannot be due to the transmission of the results of exercise. And, if this be the case, then transformations of the most diverse parts may take place without the inheritance of acquired characters, that is, of functional modifications, and there is no reason for dragging in an unproved mode of inheritance to explain a process which can quite well be explained without it. For if any part whatever can be transformed solely through natural selection, why, since there is general variability of all parts, should this be confined to the passive organs alone, when the active ones are equally variable, and equally important in the struggle for existence?

There are, indeed, many of these passive parts among animals; I need only recall the coloration of animals, the whole set of skeletal parts, so diversely formed, of the Arthropods, the legs, wings, antennæ, spines, hairs, claws, and so on, none of which can be changed by the inherited results of exercise, because they are no longer capable of modification by exercise; they are ready before they are used; they come into use only after they have been hardened by exposure to the air, and are no longer plastic; they are at most capable of being used up or mutilated. Finally, even so convinced an advocate of the Lamarckian principle as Herbert Spencer has stated that among plants the great majority of characters and distinctive features cannot be explained by it, but only through the principle of selection; all the diverse protective arrangements of individual parts, like thorns, bristles, hairs, the felt-hairs of certain leaves, the shells of nuts, the fat and oil in seeds, the varied arrangements for the dispersal of seeds, and so on, all operate by their presence alone, not through any real activity which causes them to vary, and the results of which might be transmitted. An acacia covered all over with thorns seldom requires to use its weapons even once, and if a hungry ruminant does prick itself on the thorns it is only a few of these which are thus 'exercised,' the rest remain untouched.

But since all these parts have originated notwithstanding their passivity, there must be a principle which evokes them in relation to the necessities of the conditions of life, and this can only be natural selection, that is, the self-regulation of variations in reference to utility. And if there is this principle, we require no other to explain what is already explained.

I can quite well understand, however, that many naturalists, and especially palæontologists, find it difficult to accept this conclusion. If we think only of those parts that actively function, and thus change by reason of their function, being strengthened by use and weakened and diminished in size by disuse, and if, further, we follow these parts through the history of whole geological epochs, we may certainly get the impression that the exercise of the parts has directly caused their phyletic evolution. The direction prescribed by utility in the course of the individual life and in the phylogeny is the same, and the intra-selection, that is, the selection of tissues within the individual animal, leads towards the same improvements as the selection of 'persons.' Thus it appears as if the phyletic variations followed those of the individual life, while in reality the reverse is true; the changes arising from variations in the germ are primary, and they determine the course of phylogenesis, while the tissue-selection in the individual life only elaborates and improves, according to the demands made upon it, the material afforded by the primordial equipment of the germ.

The American palæontologist, Osborn, cites the case of the horse's feet as an example in support of his view that modification brought about by use in the individual life must be transmitted in order that the phyletic transformations may be brought about, but this example is perhaps the best that could be chosen to prove the contrary. He supposes that, in every young horse, the means of locomotion are improved at every step, so to speak, through the contact with the ground, and I am quite willing to admit that this is so. But that only proves that, even now, an elaboration and improvement of the equipment which the germ affords is indispensable, as it has been at all times and in all animals, and thus that, notwithstanding the enormous number of generations which our modern horse has behind it, the functional acquirements of the individual have not yet been impressed upon the germ. Why not? Because the horse becomes perfect without this, and there was no reason why personal selection should perfect the primary constituents of the germ still further, since the finishing touch of perfection through use is readily afforded by the conditions of each individual life.

Moreover, when Osborn, Cope, and other palæontologists emphasize that, in phyletic evolutionary series, definite paths of evolutionary change are adhered to, and are not deviated from either to right or to left, they are undoubtedly right, but the conclusion which they draw is not justifiable, whether they assume with Nägeli that there is a power of development, a principle of perfecting, or whether, as Osborn does, they assume the transmission of the modifications brought about through use in the individual life. There remains a third possibility, that the quiet and constant evolution in a definite direction is guided by selection, and as, in passively useful parts, that principle alone is admissible, I see no justification for assuming it to be inoperative in regard to those which are actively functional. All these variations which have led up, for instance, to the modern form of the horse's foot are useful; if they were not, they could not have been produced either by use or by disuse in the individual life.

At the same time, here again, we are justified in inquiring whether the assumption of 'chance' germinal variations, which we have hitherto made with Darwin and Wallace, affords a sufficient basis for selection. Osborn says very neatly in this connexion, 'We see with Weismann and Galton the element of chance; but the dice appear to be loaded, and in the long run turn "sixes" up. Here arises the question: What loads the dice?'

Until recently we might have answered, 'external conditions'; it is they that load the dice one-sidedly, and condition that the same straight path of phylogenesis is adhered to, and exactly the same direction of variations is preferred and maintained. It has to be asked, however, whether this answer, which is certainly not absolutely incorrect, is sufficient by itself, whether the dice are not falsified and one-sidedly loaded in another sense, so that they always throw a preponderating number of the useful variations. We shall attempt very soon to solve this problem, but in the meantime I must refer to another argument in favour of assuming the Lamarckian principle, perhaps the most important and it may be thought the most difficult of all to refute, the so-called co-adaptation of the parts of an organism, that is, the fitting together of many individual organs for a common purposeful functioning.

LECTURE XXIV

OBJECTIONS TO THE THESIS THAT FUNCTIONAL
MODIFICATIONS ARE NOT TRANSMITTED

Giant stag as an example of co-adaptation or 'harmonious adaptation'—This occurs even in passively functioning parts—Skeleton of Arthropods—Stridulating organ of ants and crickets—Limbs of the mole-cricket—Wing-venation—Colorations which form mimetic pictures—Harmonious adaptations in worker-bees and ants—Degeneration of their wings and ovaries—The quality of food acts as a liberating stimulus—Vom Rath's case of drones fed with royal food—Transition-forms between females and workers—Wasmann's explanation of these—The Amazon ants—Two kinds of workers—Appendix: Zehnder on the case of ants—On the skeleton of Arthropods—Hering's interpretation of Ehrlich's Ricin experiments—Hering's position in regard to the transmission of functional modifications.

It was Herbert Spencer, the English philosopher, who first brought the argument of co-adaptation into the field against my view of the non-inheritance of functionally acquired modifications. He pointed out that many, if not, indeed, most modifications of bodily parts, to be effective, implied further changes, often very numerous, in other parts, and these latter must therefore have changed simultaneously with the part which was being changed under the control of natural selection; this, however, is only conceivable as due to an inheritance of the changes caused by use, since a simultaneous alteration of so many parts through natural selection would be impossible. If, for instance, the antlers of our modern stag were to grow to the size of those of the Giant Stag of the Irish peat-bogs, which measured over ten feet across from tip to tip, this would mean—as has already been shown—a simultaneous thickening of the skull, and to bear the heavy burden, a strengthening of the ligamentum nuchæ, of the muscles of the neck and back, of the bones of the legs and their muscles, and, finally, of all the nerves supplying the muscles; and how could all this happen simultaneously with, and in exact proportion to the growth of the antlers, if it depended—as natural selection assumes—on chance variations of all these parts? What if the appropriately favourable variation in one of these organs did not occur? A harmonious variation of all the parts—bones, muscles, nerves, ligaments—which unite in a common activity, is an inadmissible assumption, because, in many cases, such co-operating groups of organs have in the course of evolution developed in opposite directions. In the giraffe, for instance, the fore-legs are longer than the hind-legs, which is the reverse of what obtains in the majority of ruminants; in the kangaroo the hind-legs, on the contrary, have developed to a disproportionate size, while the fore-legs have degenerated into relatively small grasping arms. Co-operating parts, like the fore and hind limbs, may thus follow opposite paths of evolution; their variations need not always be directed to the same end.

The difficulty presented by these so-called co-adaptations or harmonious correlations cannot be denied, and we must also admit that, if the results of exercise were inherited, the explanation of the phenomenon would, in many cases—but not, indeed, in all—be easy, because the adaptation of the secondarily varying parts in each individual life would correspond exactly to the altered function of the part, and would be transmitted to the descendants, and in them would again be subject to such a degree of variation, according to the principle of histonal selection, as might be conditioned by the further progress of the primary variation. The simplicity of the explanation is striking, if only it were at the same time correct! But there are whole series of facts, or rather of groups of facts, which prove that the causes of co-adaptation do not lie in the inheritance of functional modifications, and this must be recognized, even though we may not yet be in a position to state the causes of co-adaptation, and to say whether natural selection suffices to explain it or not.

I must first point out that co-adaptations occur not only in actively, but also in passively functioning parts. Very numerous instructive examples are to be found among the Arthropods, whose whole skeleton belongs to this category. It has been objected that this is not wholly passive, but that, like the bones of vertebrates, it is stimulated by the contraction of the muscles and incited to functional reaction, and that it thickens at places where strong muscles are inserted, and becomes or remains thin where it is not exposed to any strain from the muscles. But this is not the case, for the chitinous skeleton can only offer resistance to the muscular contractions when it is no longer soft, as it is immediately after it is secreted. As soon as it has become hard, it can no longer be altered, and can at most be worn away externally by long use. The proof of this lies in the necessity for moulting, which is indispensable to all Arthropods as long as they continue to grow, but does not occur later. Every one who has followed the growth of an insect or a crustacean knows well that the moultings or ecdyses are often accompanied by great changes, and hardly ever occur without some slight changes in the form of the body, especially of the limbs, with their teeth, bristles, spines, and so on. These new or transformed parts are formed before the throwing-off of the old chitinous shell, and under its protection, and they are brought about by an elaboration or transformation of the living soft matrix of the skeleton, the hypodermis, which consists of cells, and is the true skin. They must thus have arisen in the ancestors of our modern Arthropods in the same way, that is, not by a gradual modification during use, but by a slight sudden transformation before use. The steps in the transformation may have been very small, a bristle may have become a little longer in the second stage of life than it was in the first, or instead of five bristles a particular spot may bear six in the second or third stage of life; but the variations in the phyletic development must always be caused by germ-variations which effect from within the variation in the relevant stage of development. But the part which has varied can only function after it has become firm and immodifiable.

If these circumstances be kept clearly in mind, they furnish a quite overwhelming mass of proof against the views of the Lamarckians.

Furthermore, it is not even true that the thickest parts of the external skeleton are those at which the muscles are inserted. The wing-covers of beetles offer the best proof to the contrary, for there are no muscles at all in them, yet they are, in many species, the hardest and thickest part of the whole chitinous coat of mail. The reason is not far to seek; they protect the wings and the soft skin of the back, which lies concealed beneath them, and the muscles are inserted in this!—a relation which can be explained only by its suitability to the end, and not as due to any direct effect.

When we remember the origin—which we have just described—of the external skeleton from the soft layer of cells underneath it, the thickness of the chitinous skeleton, which is very different at different places in the same animal, but always adapted to its end, furnishes a case of co-adaptation in parts which have a purely passive function. The thickened part cannot be due to the insertion of a muscle, but it is always there in advance, from internal causes, so that the muscle finds sufficient resistance. Close to it there may lie, perhaps, the edge of a segment, and at this spot the chitinous skeleton becomes almost suddenly thinned to a joint membrane capable of being bent or folded, not because there was no pull from the muscles at this spot, but in order that the two segments may be connected movably. Thus, nowhere in the whole body of the Arthropod can the adaptation of the skeleton, in regard to thickness and power of resistance, be regulated by function itself, but only by processes of selection which imparted to each spot the thickness it required, in order to be effective in its function, whether that be offering resistance to the strain of the muscles, or giving suppleness to a joint, or affording the necessary hardness for biting the prey, or for boring into wood or earth, or merely for protecting the animal from external injuries.

Fig. 91 (repeated). Hind-leg of a
Grasshopper (Stenobothrus protorma), after
Graber. fe, femur. ti, tibia. ta, tarsal
joints. schr, the stridulating ridge.

There are, however, many individual functions of the Arthropods the exercise of which depends on the simultaneous change of several skeletal parts; as, for instance, many of the 'singing' or vocal apparatuses in insects. In quite recent times such vocal organs have been discovered in ants, in which they consist of a small striated region on the surface of the third abdominal segment, and a sharp ridge on the segment in front; the latter is rubbed against the former by the movements of the two segments. Quite a similar 'stridulating organ' has long been known in the bee-ant (Mutilla), and the whistling sound produced by it is easily heard by our ears; moreover August Forel has heard it in the large wood-ant (Camponotus ligniperdus), and has described it as an 'alarm-signal,' which the animals give each other on the approach of danger—an observation which has recently been confirmed by Wasmann and extended by Robert Wroughton in regard to Indian ants. All these arrangements for producing sound depend always on two organs, of which one resembles the bow, the other the strings of a violin; the one is of no value without the other, and they must therefore have developed simultaneously, yet they cannot have arisen through use, and the inheritance of the results of use, because they are both dead chitinous parts, which are never strengthened by rubbing against each other with the movements of the abdomen, but are rather worn away.

The same is true of the chirping organs of grasshoppers, beetles, and crickets; in all cases they consist of two different parts, which together produce a sound, and which therefore must have arisen simultaneously, and the origin of which cannot be referred to the inheritance of the results of exercise, but rather to selection. It is thus possible that co-adaptation of at least two parts may take place even when the hypothetical Lamarckian principle is altogether excluded.