Chapter VI: Part 6

He considers that the results of use (that is, increased function) are generally transmitted, because the increase in the organ which is functioning more strongly changes the composition of the blood, by withdrawing from it in a greater degree the specific substances which the organ in question--a muscle, for instance--requires for its activity. All parts of the animal are thereby affected and modified, but especially those smallest vital units or 'fistellæ' (corresponding to biophors) which preside over digestion, and of which there are several sorts. Among them those work most arduously which have to produce the specific substances which serve for the nutrition of the muscles with increased function, because these are needed in larger quantities. This kind of digestive 'fistella' therefore multiplies, while other kinds, whose products are not required and therefore not used up, cease to be so active, diminish in number, and in course of time disappear. In this way the composition of the blood is altered, and with it to a greater or less degree all the characters of the whole organism. Of course the reproductive cells are also under the influence of this change in the composition of the blood, because the different nutritive substances are accumulated within them in an altered proportion corresponding to the changed composition of the blood, the nutritive substances for the muscles with increased function being contained in it in a larger quantity, and thus the greater development of the muscle will repeat itself in the progeny, that is to say, _the acquired character is transmitted_.

It is obvious that this is precisely the same line of argument as that used in reference to the origin of the worker and soldier ants. The different kinds of 'digestive fistellæ' correspond to the different food-carrying workers, and the blood to the assumed storehouse from which soldiers and workers select the food suitable for their respective needs, while the sex-cells in the one case, the sexual animals in the other, partake of all kinds exactly in the proportion in which they are stored, and thus the organ which functions most vigorously must be stronger in the offspring.

How the minute quantity of nutritive material contained in the ovum, still less in the sperm, is to effect the strengthening of the particular muscles in the descendants is not stated; moreover, such minimal quantities of food must soon be exhausted, and cannot possibly increase. It would seem as if the muscles could not even begin by being stronger, much less that they should remain so, if they were not exercised equally vigorously by the descendants. If the specific nutritive stuffs were 'fistellæ,' that is to say, were living units capable of multiplication, one could understand it. But there can, of course, be no possibility of a production of living units through digestion; that can only give rise to digested substances. Or if the alteration in the composition of the blood produced in the determinant system of the germ-plasm just those variations requisite to bring about a strengthening of the muscular system, it would remain to be shown how this could happen, for the gist of the problem lies in this. For muscles do not lie in the germ-plasm as miniature models of the subsequent muscular system, and even if they did, would not all the muscles, and not merely those which were no longer exercised, decrease hereditarily when a particular group, like the muscles of the ear in man, degenerates? Zehnder replies to this with the hypothesis that the muscles are not all chemically alike, but that each possesses a particular chemical formula, though they may all be very similar, and that, therefore, the nutritive materials required by each must be slightly different. In that case there would require to be, in the ovum and sperm of man, in order that functional modifications might be transmitted, as many special nutritive substances as there are muscles, and, in addition to these, innumerable hosts of other kinds of specific nutritive substances for all the other parts of the body, since all of them can be strengthened by exercise and weakened by disuse. And even if we suppose that all these millions of specific nutritive substances are accommodated within the germ-cells, as Zehnder's theory requires, they could not perform what Zehnder ascribes to them, for, as we have already said, they cannot multiply in the manner of living units, and so control the growing organism. The different specific nutritive materials contained in the blood are just as powerless to perform the task ascribed to them by Zehnder as the specific kinds of food in the hypothetical storehouse of the ants are to give rise to the different persons of the ant-colony.

Zehnder also attempts to overthrow the arguments against the Lamarckian principle which I based on the skeleton of Arthropods.

It does not seem to him probable that the chitinous coat of mail can be an absolutely dead structure, and he supposes that very delicate nerve-fibrils penetrate into all its most minute parts, and so are stimulated by 'every pressure and every strain' exerted on the chitinous skeleton. They 'work' when they are stimulated, and in doing so they use up 'their specific food-stuffs.' At places which are frequently stimulated the corresponding nerves develop more than elsewhere. The necessary specific food-stuffs for these particular nerves therefore increase proportionately within the body, and also in the reproductive cells. Accordingly, in the germ-cells there is an increase of the aforesaid nervous substances, which in the offspring become associated with the relevant part of the chitinous covering, and induce in development the secretion of chitin at this part. At this particular spot, then, the chitin will be specially thick.

This clearly implies that each particular part of the skin has its specific nutritive substances, necessitated by the nerves which traverse it! Thus there must be as many nerve-nutritive substances as there are skin-nerves, specific chemical combinations for every part of the body which is capable of heritable variation. This is so extraordinarily improbable that I need say nothing more about it. If the Lamarckian principle requires this kind of hypothesis to bolster it up, it is undoubtedly doomed.

If we disregard altogether the positive aspect of Zehnder's hypothesis, and assume that the skin-nerves are really stimulated through the chitin by every strain and pressure to which a spot of skin is exposed, and that they cause a correspondingly greater secretion of chitin, which would then, according to the Lamarckian principle, be hereditary, does this harmonize with what actually occurs in the development of the skeleton as we know it in the case of Insects and Crustaceans? Not at all! Can we suppose that the carapace of a crab or the enormously hard wing-covers of a water-beetle are exposed to a continual pounding and pressing and pushing? Exactly the contrary is the case. Every assailant takes care not to grasp the animal where it is so well protected, and seeks out the most vulnerable parts for its attacks. It may be answered that, while this is certainly the case now, the animals were badly protected when the ancestral forms were evolving. But that they could not have become hard by dint of being frequently bitten or otherwise wounded should be obvious from the fact that the whole of the wing-covers and the whole of the carapace is uniformly covered with thick chitin, while each wound would only stimulate particular spots; and we should also have to admit that, since these parts of the skin which are now so well protected are no longer seized and stimulated, they would long ago have become thin again, according to the principle of the degeneration of parts no longer used, or, in this case, no longer stimulated. But there is no need for wasting time over such quibbles, since there is a fact which absolutely contradicts Zehnder's hypothesis. I mean the degeneration of the chitinous skeleton in those Crustaceans and Insects which protect the abdomen within a shelter like the hermit-crabs, the caddis-flies (Phryganidæ), (Fig. 107) and the sack-carrying caterpillars of the Psychidæ among Lepidoptera. The hermit-crabs, as is well known, squeeze their abdomen into a usually spirally-coiled Gasteropod shell, and they always choose houses which are wide enough to conceal the whole body up to the hard claws when necessity arises. In this case there is surely a continual pressure on the abdomen, which, being soft, must be squeezed very tightly every time the animal retreats into its shell. One of my opponents has described the disappearance of the tough integumentary skeleton from the abdomen of these animals as an inherited result of this pressure, and another regards it as the inherited result of the degeneration of the muscles in this part of the body. But, according to Zehnder, this continuous pressure, and the frequent rubbing up and down of the abdomen on the inner surface of the Gasteropod's shell, would undoubtedly have a stimulating effect on the skin-nerves, and would therefore bring about a thickening of the chitinous cuticle. In regard to the larval Phryganidæ and Psychidæ, the case would be the same, though perhaps hardly to the same degree, for while these larvæ make their own houses, and will therefore at least make them big enough to begin with, the pressure and friction must increase with the growth of the animal.

If the regulation of the strength of the integumentary skeleton be referred to selection, we see at once why carapace and wing-covers should be of equal thickness throughout their whole extent, and why they do not disappear, although they do not function actively, and are less stimulated than any other parts of the skeleton; and we also understand why the abdomen of hermit-crabs and of larval Phryganidæ and Psychidæ has become soft, whether it be exposed to pressure or friction in a greater or a less degree. It no longer requires to be hard, because it is protected by the house, and in the case of the Pagurids it must not be hard because it could not then be readily squeezed into the hard-walled and narrow recesses of the Gasteropod shell; in this case there has therefore been positive selection. I have not yet referred to the fact that the chitinous covering is certainly not living, though it is not exactly dead; it is a secretion of the epidermic cells, not a tissue, and we cannot suppose that there are any nerve-endings in it. It is almost superfluous to say that the fact that the skin is cast is in itself enough to make such an assumption untenable, for the whole of the assumed delicate nervous network would be shed at every moult and torn away from the nerves which lead to it. As far as my knowledge goes, nothing of this kind occurs anywhere in the whole range of the animal kingdom.

Even if we assume, for the benefit of Zehnder's hypothesis, that although there are no nerves in the chitin itself yet irritations affecting the chitinous coat may be transmitted through it to the delicate nerve-endings lying beneath it, this should take place in a greater degree at the thin places of the skeleton than _at the thick parts_! But this interpretation is again fallacious, for we see that the tactile organs of Arthropods always break through the chitinous cuticle and protrude beyond it in the form of setæ.

Of the many other opponents of my views in regard to the transmissibility of acquired functional modifications, I need only deal in detail with Oscar Hertwig.

He seeks for direct proofs of an inheritance of acquired characters, and believes that he has found these in the hereditary transmission of acquired immunity from certain diseases. He reminds us of Ehrlich's well-known experiments on mice with ricin and abrin.

Even small doses of these two poisons kill mice, but they are tolerated in very minute doses, and if their administration be continued for some time in such minute doses, the animals gradually acquire a high degree of insensitiveness to these poisons; they become immune to ricin and abrin.

This immunity is transmitted from mother to young, but it only lasts for a short time, about six to eight weeks after birth. Yet this is regarded by Hertwig as an illustration of the transmission of an acquired character, as an acquired modification of the cells of the body, for he explains the immunity on the assumption that all the cells of the body undergo a particular variation due to the influence of the poison, and are thus, to a certain extent, modified in their nature, and that the ovum also undergoes this variation and transmits it to the young animal. The immunization might certainly come under the category of functional modifications, and it might be thought that we have here a case of transmission of such an acquired character.

Against this, however, we have to put the fact that _the acquired immunity is not transmitted from the father to the descendants_. Hertwig attempts to explain this by saying that the short duration of the experiments has only allowed the poison to affect the cell-substance (cytoplasm) and not the nucleus, that is, the hereditary substance of the sperm-cells, an assumption which has little probability considering the intimate nutritive relations between the cell-nucleus and the cytoplasm. I should be rather inclined to conclude from the difference in the transmitting power of the sperm and of the ovum, that this 'inheritance of immunity' does not depend, as Hertwig supposes, on a modification of the cells to 'ricin immunity,' but, as Ehrlich and the bacteriologists believe, on the production of so-called 'anti-toxins,' and that these anti-toxins are handed on to the embryo not by the ovum itself, but by the interchange of blood between mother and offspring which lasts throughout the whole embryonic period. It is then self-evident why no transmission of immunity through the father occurs.

But it would lead me too far were I to attempt to refute all the attempts that have been made to interpret individual cases as due to the inheritance of acquired characters. I should, however, like to say something as to the theoretical possibility of such an assumption.

When we try to conceive how experiences and their consequences can be entailed, how new acquirements of the 'personal part can have representative effects on the germinal part,' we find ourselves confronted with almost, if not entirely, insuperable difficulties. How could it happen that the constant exercise of memory throughout a lifetime, as, for instance, in the case of an actor, could influence the germ-cells in such a way that in the offspring the same brain-cells which preside over memory will likewise be more highly developed--that is, capable of greater functional activity? We know what Zehnder's answer to such a question would be; he would make the blood the intermediary between the brain-cells and the germ-cells, but we have seen that specific food-stuffs for each specific cell-group cannot be assumed, and that, even if they could, they would not meet the necessities of the case. Yet every one who does not regard the germ-plasm as composed of determinants is constrained to make some such assumption. But if we take our stand upon the theory of determinants, it would be necessary to a transmission of acquired strength of memory that the states of these brain-cells should be communicated by the telegraphic path of the nerve-cells to the germ-cells, and should there modify only the determinants of the brain-cells, and should do so in such a way that, in the subsequent development of an embryo from the germ-cell, the corresponding brain-cells should turn out to be capable of increased functional activity. But as the determinants are not miniature brain-cells, but only groups of biophors of unknown constitution, and are assuredly different from those cells; as they are not 'seed-grains' of the brain-cells, but only living germ-units which, in co-operation with the rest exercise a decisive influence on the memory-cells of the brain, I can only compare the assumption of the transmission of the results of memory-exercise to the telegraphing of a poem, which is handed in in German, but at the place of arrival appears on the paper translated into Chinese.

Nevertheless, as I have said before, I do not disagree with those who say, with Oscar Hertwig, that the impossibility of forming a conception of the physiological nexus involved in the assumed transmission does not _ipso facto_ constrain us to conclude that the transmission does not occur. I cannot, however, agree with Hertwig that the case is exactly the same as in the 'converse process,' that is, 'in the development of the given invisible primary constituents in the inheritance of the cell into the visible characters of the personal part.' Certainly no one can state with any definiteness how the germ goes to work, so that from it there arises an eye or a brain with its millionfold intricacies of nerve-paths, but although the process cannot be understood in detail, it can in principle, and this is just what is impossible in regard to the communication of functional modifications to the germ. Moreover, in addition to this, there is the very important difference that, in the one case, we know with certainty that the process actually takes place, although we cannot understand its mechanical sequence in detail, while in the other we cannot even prove that the supposed process is a real one at all. From the fact that we are unable to form clear conceptions of a hypothetical process, we are not justified, it seems to me, in assuming it to be real, even though we are aware of many other processes in nature which we are unable to understand.

Nor does Hertwig take up this position, for he is at pains to show the mechanical possibility of the process of inheritance which he assumes, and he bases this upon the suggestions made by Hering in his famous work _Ueber das Gedächtniss als eine allgemeine Function der organisirten Materie_ [_On memory as a general function of organized matter_], 1870. As this essay probably contains the best that can be said in favour of a transmission of functional modifications, and as it also includes some indisputable truths, we may consider it in some detail.

Hering is undoubtedly right in regarding 'the phenomena of consciousness as functions of the material changes of organic substance, and conversely.' That is, he believes that every sensation, every perception, every act of will arises from material changes in the relevant nerve-substances. But we know that 'whole groups of impressions, which our brain has received through the sense-organs, are stored up in it, as if resting, and below the margin of consciousness, to be reproduced when occasion arises, in correct order of space and time, and with such vividness, that we may be deceived into regarding as a present reality what has long ceased to be present.' There must therefore remain in the nerve-substance a 'material impact,' a modification of the molecular or atomic structure, which enables it 'to ring out to-day the note that it gave forth yesterday if only it be rightly struck.'

Hering attributes a similar power of memory and reproduction to the germ-substance; he believes that he is justified in making the assumption that acquired characters can be inherited, although he admits that it 'appears to him puzzling in the highest degree' how characters which developed in the most diverse organs of the mother-being can exert any influence on the germ. That he may be able to assume this he points to the interconnexion of all organs by means of the nervous system; it is this that makes it possible that 'the fate of one reverberates in the other, and that, when excitement takes place at any point, some echo of it, however dull, penetrates to the remotest parts.' To the delicate-winged communication by means of the nervous system, which unites all parts among themselves, must be added the general communication by means of the circulation of the fluids of the body. According to Hering's view, the germ experiences, in some degree, in itself all that befalls the rest of the organs and parts of the organism, and these experiences stamp themselves more or less upon its substance, just as sense-impressions or perceptions stamp themselves upon the nerve-substance of the brain, and these experiences are reproduced during the development of the germ, just as the brain brings memory-pictures back to consciousness. He says, 'If something in the mother-organism has so changed its nature, through long habit or exercise repeated a thousand times, that the germ-cell resting in it is also penetrated by it in however weakened a fashion, when the latter begins a new existence, expands, and increases to a new being whose individual parts are still itself and flesh of its flesh, it reproduces what it experienced as part of a great whole. This is just as wonderful as when an old man suddenly remembers his earliest childhood, but it is not more wonderful than this.'

But I think it is more wonderful. There exist demonstrably in the brain thousands upon thousands of nerve-elements, whose activity is a definite and limited one, because each particular visual impression, for instance, only excites to activity certain definite nerve-elements, and can leave memory-pictures in these alone. According to my conception of it, the germ-plasm is quite as complex in its composition, and does not consist of homogeneous elements, but of innumerable different kinds, which are not related to the parts of the complete organism indiscriminately, but only to particular parts. But is it allowable to assume that there are invisible nerve-connexions, not only to every germ-cell, but also within the germ-plasm, to every determinant, like the nerve-paths which lead from the eye to the nerve-cells in the optic-area of the brain? For if it were otherwise, how could we conceive of the modification of an organ--as, for instance, the ear-muscles in Man--communicating itself to the precise determinants of these muscles in the germ-plasm? I have often been met with the reproach that my conception of the composition of the germ-plasm is much too complex--but the complexity of Hering's suggestion seems to me to go a long way beyond mine.

Hering's ideas, which are not only ingenious but very stimulating, might be accepted as the first indication of an understanding of the assumed inheritance of functional modifications, if it could be proved that such inheritance is a fact; but, as we have seen, that is not the case. The assumption might be permitted, perhaps, if it could be shown that certain groups of phenomena left no other possibility of explanation open except this assumption, but that also, as far as I can see, is not the case. Of course, others hold a different opinion, but chiefly because they have rejected without much reflection the sole explanation which presents itself for numerous phenomena--I mean the processes which we are about to study under the name of 'germinal selection.' But, in any case, Hering's ideas seem to me very valuable, because they make it apparent that, however much we know of the organism, we only know it in a general way, and that numberless delicate processes go on in it which leave no trace for our microscope, and that we can only recognize the final results of numerous invisible and often, in their subtlety, also unimaginable factors. This ought to be taken to heart, especially by all those who speak of simplicity in reference to the germ-plasm. So much at least is certain: If there were any inheritance of functional modifications, we should have another proof that the germ-plasm is composed of determinants, for without them there could be no possibility that the 'experiences' of an individual organ would be transmitted to the germ in the way that the Lamarckian principle implies. Something, and that something material, must be modified in the germ-plasm if the vigorous use of a group of muscles, or of a gland, or of a nerve-cell, is to be communicated to the germ, and not to the whole germ-plasm, but only to so much of it as is necessary to cause variation in the corresponding group of cells in the child. It may perhaps be said that this still does not necessitate the assumption of special determinants for these cell-groups, and that one might, with Herbert Spencer, conceive of the germ-plasm as consisting of homogeneous units which vary in the development in accordance with the diverse regularly alternating influences to which it is exposed from step to step, and that, therefore, in each of these units of very complex structure only a single molecule, or perhaps only a single atom, would need to vary in order that, in the course of development, the resulting cell-group should appear in the rudiment in somewhat altered strength.

But I do not believe that a chemical molecule, still less an atom, is sufficient for this, for reasons which I have already stated--yet we need not go into this now, but rather deduce the consequences of this admission. It follows that the 'unit' is made up of numerous 'molecules' or 'atoms,' of which each, by dint of changes it has undergone, causes particular parts of the body to vary in a definite manner; in other words, we have here again a theory of determinants, only they are on a much smaller scale, since each invisible little vital particle or 'unit' contains all the determinants within itself, while in my theory it is only the id, that is, the visible chromosome, which includes the determinant complex. Such a theory would be far from a simplification of mine, it would rather complicate it enormously, and that without anything being gained. At most it would be made more evident how inconceivably complex the nerve-paths must be which lead from the part that has been modified by exercise to the germ-plasm, and must also lead to all the innumerable 'molecules or atoms' of the individual 'units.' But even on my theory of the composition of the ids as aggregates of living determinants, such nervous transmission of qualities would be a monstrosity which no one would accept, and I think on this account that my argument as to the impossibility of conceiving of the transmission of the modifications of the personal part to the germinal part retains its force, notwithstanding Hering's interesting analogy.

If the transmission of functional modifications were an indisputable fact, I repeat, we should have to give in, and then we might regard the 'memory of organized material' as affording a hint of the possibility of the unimaginable process. But as long as the occurrence of this transmission cannot be proved either directly or indirectly, such a vague possibility of explanation need not induce us to assume an unproved process.

LECTURE XXV

GERMINAL SELECTION

On what does disappearance after disuse depend, if not on the Lamarckian principle?--Panmixia--Romanes--Fluctuations in the determinant-system of the germ-plasm due to unequal nutrition--Persistence of germinal variations in a definite direction--The disappearance of non-functioning parts--Preponderance of minus germinal variations--Law of the retrogression of useless parts--Variation in an upward direction--Artificial selection--Influence of the multiplicity of ids and of sexual reproduction--Personal selection depends on the removal of certain id-variants--Range of influence of germinal selection--Self-regulation of the germ-plasm, which is striving towards stability--Ascending variation-tendencies may persist to excess--Origin of secondary sexual characters--Significance of purely morphological characters--The markings of butterflies.

Now that we have recognized that the assumption of a transmission of functional modifications is not justifiable, let us discuss some of the many phenomena to explain which many people believe the Lamarckian principle to be indispensable, and let us inquire whether we are in a position to give any other explanation of these. How has it come about that the effects of use and disuse _appear_ to be inherited? Can we find a sufficient explanation in the principle of selection, and in the natural selection of Darwin and Wallace?

The answer to these two questions will be most quickly found if we begin by seeking for an explanation of the disappearance of a part when it ceases to be exercised.

That this cannot lie in the Lamarckian principle we have already learnt from the fact that passively functioning parts, such as superfluous wing-veins, also disappear, and that the loss of the wings and degeneration of the ovaries has taken place in worker ants, which can transmit nothing because they do not reproduce.

We might be inclined to regard this gradual disappearance and ultimate elimination of a disused organ as a direct gain, on the ground that the economy of material and space thus effected may be of decided advantage to the individual animal and thereby also for the maintenance of the species, and that those animals would have an advantage in the struggle for existence in which the superfluous organ was reduced to the smallest expression. But that would be far from supplying us with a sufficient explanation of the phenomenon; the individual variations in the size of an organ which is in process of degenerating are, even in extreme cases, far too slight to have any selection-value, and I cannot call to mind a single case in which the contrary could be assumed with any degree of probability. What advantage can a newt or a crustacean living in darkness derive from the fact that its eye is smaller and more degenerate by one degree of variation than those of its co-partners in the struggle for existence? Or, to use Herbert Spencer's striking illustration, how could the balance between life and death, in the case of a colossus like the Greenland whale, be turned one way or another by the difference of a few inches in the length of the hind-leg, as compared with his fellows, in whom the reduction of the hind-limb may not have gone quite so far? Such a slight economy of material is as nothing compared with the thousands of hundredweights the animal weighs. As long as the limbs protrude beyond the surface of the trunk they may prove an obstacle to rapid swimming, although that could hardly make much difference, but as soon as the phyletic evolution had proceeded so far that they were reduced to the extent of sinking beneath the surface, they would no longer be a hindrance in swimming, and their further reduction to their modern state of great degeneration and absolute concealment within the flesh of the animal cannot be referred even to negative selection.

Years ago I endeavoured to explain the degeneration of disused parts in terms of a process which I called Panmixia. Natural selection not only effects adaptations, it also maintains the organ at the pitch of perfection it has reached by a continual elimination of those individuals in which the organ in question is less perfect. The longer this conservative process of selection continues, the greater must be the constancy of the organ produced by it, and deviations from the perfect organ will be of less and less frequent occurrence as time goes on.

Now if this conservative action of natural selection secures the maintenance of the parts and organs of a species at their maximum of perfection, it follows that these will _fall below this maximum as soon as the selection ceases to operate_. And it does cease as soon as an organ ceases to be of use to its species, like the eye to the species of crustacean which descends into the dark depths of our lakes, or to the abyssal zones of the ocean, or into a subterranean cave-system. In this case all selection of individuals ceases as far as the eye is concerned; it has no importance in deciding survival in the struggle for existence, because no individual is at a disadvantage through its inferior eyes, for instance, by being in any way hindered in procuring its food. Those with inferior organs of vision will, _ceteris paribus_, produce as good offspring as those with better eyes, and the consequence of this must be that there will be a general deterioration of eyes, because the bad ones can be transmitted as well as the good, and thus the selection of good eyes is made impossible.

The mixture thus arising may be compared to a fine wine to which a litre of vinegar has been added; the whole cask is ruined because the vinegar mingles with every drop of the wine. As deviations from the normal are always occurring in every part of every species, and among them some that lessen the value of the organ, rarely perhaps at first, but after a time in every generation, a sinking of the organ from the highest point of possible perfection becomes inevitable as soon as the organ becomes superfluous. The functional uselessness of the organ must go on increasing the longer it is disused, as will readily be admitted if it be remembered that only the most perfect adaptation of all the separate parts of an organ can maintain its functional capacity, that all the parts of an organ are subject to variation, and that every deviation from the optimum implies a further deterioration of the whole. An eye, for instance, can no longer vary in the direction of 'better' if it has already reached the highest possible point of perfection; every further variation must deteriorate it.

Romanes gave expression to this idea, that the cessation of natural selection alone must cause the degeneration of a part, a decade before I did, but neither he nor the scientific world of his time attached great importance to it, and it was forgotten again. This was intelligible enough, for, at that time, the validity of the Lamarckian principle had not been called in question, and therefore the need for some other principle to explain the disappearance of disused parts had not begun to be felt.

I found myself in quite a different position. As my doubts regarding the Lamarckian principle grew greater and greater, I was obliged to seek for some other factor in modification, which should be sufficient to effect the degeneration of a disused part, and for a time I thought I had found this in panmixia, that is, in the mingling of all together, well and less well equipped alike. This factor does certainly operate, but the more I thought over it the clearer it became to me that there must be some other factor at work as well, for while panmixia might explain the deterioration of an organ, it could not explain its decrease in size, its gradual wearing away, and ultimate total disappearance. Yet this is the path followed, slowly indeed, but quite surely, by all organs which have become useless. If panmixia alone guided the deterioration of the organ, and it was thus only chance variations which were inherited through panmixia and gradually diffused over the whole species, how could it come about that all the variations were in the direction of smaller size? Yet this is obviously the case. Why should no variations in the direction of larger size occur? And if this were so, why should a useless organ not be maintained at its original size, if it be admitted that an increase in size would be prevented by natural selection? But this never occurs, and diminution in size is so absolutely the rule that the idea of a 'vestigial or rudimentary' organ suggests a 'small organ' almost more than an 'imperfect' one.

There must then be something else at work which causes the minus-variations in a disused organ to preponderate persistently and permanently over the plus-variations, and this something can lie nowhere else than where the roots of all hereditary variations are to be found--in the germ-plasm. This train of thought leads us to the discovery of a process which we must call selection between the elements of the germ-plasm, or, as I have named it shortly, _Germinal Selection_.

If the substance of the germ-plasm is--as we assumed--composed of heterogeneous living particles, which have dissimilar rôles in the building up of the organism, there must of necessity be among them a definite labile state of equilibrium, which cannot be disturbed without modifying in some way the structure of the organism itself which arises from the germ-plasm. But if our further view be correct, that these individual and different living units of the germ-plasm are 'determinants,' that is, are the primary constituents of particular parts of the organism, in the sense that these parts could not arise if their determinants were absent from the germ-plasm, and that they would be different if the determinants were differently composed, we can draw far-reaching deductions.

It is true that we cannot learn _anything directly_ in regard to the intimate structure of the germ-plasm, and even in regard to the vital processes going on within it we can only guess a very little, but so much we may say--that its living parts are nourished, and that they multiply. But it follows from this that nourishment in a dissolved state must penetrate between its vital particles, and that whether the determinants grow, and at what rate they do so, depends mainly on the amount of nourishment which reaches them. As long as the germ-cells multiply by division the determinants have no other function but to grow; a part of their substance undergoes oxidation and thereby yields the supply of energy necessary to assimilation, that is, to the formation of new living substance.

If each kind of determinant always secured the same quantity of nourishment, all would grow in the same degree, that is, in exact proportion to their power of assimilation. But we know that in less minute conditions which we can observe more directly, there is nowhere absolute equality, that all vital processes are subject to fluctuations; any little obstacles in the current of the nutritive fluid, or in its composition, may cause poorer nutrition of one part, better of another. We may therefore assume that there are similar irregularities and differences in the minute and unobservable conditions of the germ-plasm likewise, and the result must be a slight shifting of the position of equilibrium as regards size and strength in the determinant system; for the less well-nourished determinants will grow more slowly, will fail to attain to the size and strength of their neighbours, and will multiply more slowly.

But the vigour of growth does not depend only on the influence of nourishment; one cell grows quickly, another slowly in the same nutritive fluid; it depends in great part on the cell's power of assimilation. In the same way the assimilating power of the determinants and their affinity for nourishment will vary with their constitution, and a weaker determinant will remain smaller than a stronger one, even when the stream of nourishment is the same.

It seems to me that it is upon the unequal nutrition of the determinants conditioned by the chances of the food-supply that individual hereditary variability ultimately depends. If, for instance, the determinant _A_ receives poorer nourishment at a particular time than the determinant _B_, it will grow more slowly, remain weaker, and then, when the germ-cells develop into an animal, the part to which it gives rise will be weaker than it usually is in other individuals.

These primary inequalities in the equipment of the determinants which are caused by a passing inequality in the food-stream are, of course, so slight that we are unable to observe their consequences. They must persist for a considerable time before they become observable, but they may persist for a long time, and their effect must then mount up, because every diminution in the strength of the determinant also signifies a lessened power of assimilation, and growth becomes slower for the twofold reason that passive and active nutrition decrease at the same time. In the less minute conditions observable in the histological elements of the body we know that function strengthens the organ, and that disuse weakens it, and we are justified in applying this proposition also to these more intimate conditions and minuter vital units. Thus, in the course of the multiplication of the germ-cells, the less vigorously working determinant, _A_, will gradually, but very slowly, become weaker, that is, of diminished power of assimilation, presupposing of course that the intra-germinal food-stream does not become stronger again at the same place--a possibility to which I shall subsequently refer. But while one determinant may be slowly becoming weaker, its neighbour, on the other hand, may be varying on an ascending scale, just because the former is, on account of its diminished power of assimilation, no longer able to exhaust completely the food-stream which flows to it.

The determinants are thus in constant motion, here ascending, there descending, and it is in these fluctuations of the equilibrium of the determinant-system that I see the roots of all hereditary variation, while in the fact that the variation-directions of particular determinants must continue the same without limit as long as they meet with no obstacle lies the possibility of the adaptation of the organism to changing conditions, the increase and transformation of one part, the degeneration and disappearance of another, in short, the processes of natural selection. The reason why such variation movements must continue until they meet some resistance is that every chance upward or downward movement--due, that is, to mere passive fluctuation in the food-supply, at the same time strengthens or weakens the determinant, and makes it either more or less capable of attracting nourishment to itself; in the former case an increasingly strong stream of food will be directed towards it, in the latter more and more of the available food-supply will be withdrawn from it by its neighbour-determinants on all sides; in the former the determinant will go on increasing in strength as long as it can go on attracting more nourishment, in the latter it will continue to become weaker until it disappears altogether. To the ascending progression, as is evident, there are limits set, not only by the amount of food which can circulate through the whole id, but also by the neighbour determinants, which will sooner or later resist the withdrawal of nourishment from them; but for the descending progression there are no limits except total disappearance, and this is actually reached in all cases in which the determinants are related to a part which has become useless. But both these movements, the upward and the downward alike, are quite independent of natural selection, i.e. of personal selection; they are processes of a unique kind which run their course purely in accordance with intra-germinal laws. Whether a determinant 'ascends' or 'descends' depends solely upon the play of forces within the germ-plasm, not upon whether the direction of the variation in question is useful or prejudicial, or on whether the organ in question, the determinate, is of value or otherwise. In this fact lies the great importance of this play of forces within the germ-plasm, that it gives rise to variations quite independently of the relations of the organism to the external world. In many cases, of course, personal selection intervenes, but even then it cannot directly effect the rising or falling of _the individual_ determinants--these are processes quite outside of its influence--but it can, by eliminating the bearers of unfavourably varying determinants, set a limit to further advance in such directions. This we shall consider in more detail later on. Personal selection operates by removing unfavourably varying individuals from the genealogical tree of the species, but at the same time the determinants which are varying unfavourably are also removed, and their variation is thus put a stop to for all time.

I have called these processes which are ceaselessly going on within the germ-plasm, Germinal Selection, because they are analogous to those processes of selection which we already know in connexion with the larger vital units, cells, cell-groups and persons. If the germ-plasm be a system of determinants, then the same laws of struggle for existence in regard to food and multiplication must hold sway among its parts which hold sway between all systems of vital units--among the biophors which form the protoplasm of the cell-body, among the cells of a tissue, among the tissues of an organ, among the organs themselves, as well as among the individuals of a species and between species which compete with one another.

If this be the case, we have here ready to hand the explanation of every heritable variation of a part, ascending and descending alike. Let us consider for a little the latter category--that is, the disappearance of functionless _or useless organs_. It is clear that, from the moment in the life of a species that an organ, _N_, becomes useless, natural selection withdraws her hand from it; individuals with better or worse organs _N_ are now equally capable of life and struggle, the state of panmixia is entered upon, and the organ _N_ of necessity falls somewhat below its previously attained degree of perfection.

That this must be so will be admitted when it is remembered that each organ of a species is only maintained at its highest level because personal selection keeps ceaseless watch over it, and sets aside all the less favourable variations by eliminating the individuals which exhibit them. But this is no longer the case with a useless organ. When a weaker variant of a disused organ arises through the intra-germinal fluctuations of nutrition, this is transmitted to the descendants just as well as the normally developed organ, and in the course of generations will be inherited by a greater and greater number of individuals, and must ultimately be inherited by all in some degree or other. The objection has been urged from many sides that variations upwards would be quite as likely to arise as those downwards, but this is an error. Even if, at the beginning, the minus-variations were rarer than the plus-variations, in the course of generations the minus ones would preponderate because ascending variations of disused organs are not indifferent for the organism but injurious to it. Perhaps an increase in the size of the organ itself would do no harm, but in that of its determinant it certainly would, because an ascending determinant requires more nourishment than previously, and withdraws it from its surroundings, and thus from the determinants in its immediate neighbourhood; but these are those of functioning and indispensable parts. Individuals in whose germ-plasm the determinants of disused organs ascend, and thereby depress the determinants of organs which are still active, are subject to personal selection, and are eliminated. There thus remain only those with descending determinants; in other words, the chance of variants in the direction of weakness in useless determinants far outweighs that of variants in the direction of increased strength; the latter will soon cease to occur at all, for as soon as a determinant has fallen a little below its normal level, it finds itself upon an inclined plane, along which it glides very slowly but steadily downwards. This might be disputed if it could be maintained that, at every stage of the descent, a change of direction was possible. But this probably takes place rarely and only in the case of individual ids, and will therefore not be permanent because in general the stronger neighbour determinants will possess themselves of the superfluous nourishment, and a lasting ascent will thus be impossible to the weakened determinant. This is precisely what I have called Germinal Selection. The determinant whose assimilating power is weakened by ever so little is continually being robbed by its neighbours of a part of the nourishment which flows towards it, and must consequently become further weakened. As no more help will be given to it by natural selection, since the organ is no longer of any value to the species, the better among the weakened determinants of _N_ are never selected out, and they must gradually give way in the struggle with the neighbouring determinants which are necessary to the species, becoming gradually weaker and ultimately disappearing.

This process can, of course, no more be proved mathematically than any other biological processes. No one who is unwilling to accept germinal selection can be compelled to do so, as he might be to accept the Pythagorean propositions. It is not built up from beneath upon axioms, but is an attempt at an explanation of a fact established by observation--the disappearance of disused parts. But when once the inheritance of functional modifications has been demonstrated to be a fallacy, and when it has been shown that, even with the assumption of such inheritance, the disappearance of parts which are only _passively_ useful, and of any parts whatever in sterile animal forms, remains unexplained, he who rejects germinal selection must renounce all attempt at explanation. It is the same as in the case of personal selection. No one can demonstrate mathematically that any variation possesses selection value, but whoever rejects personal selection gives up hope of explaining adaptations, for these cannot be referred to purely internal forces of development.

The total disappearance of a part which has become useless takes place with exceeding slowness; the whales, which have existed as such since the beginning of the tertiary period, have even now not completely lost their hind-limbs, but carry them about with them as rudiments in the muscular mass of the trunk, and the birds, which are even older, still show in their embryonic primordia the five fingers of their reptilian forefathers, although even their bird-ancestors of the Jurassic period, if we may argue from _Archæopteryx_, had only three fingers like our modern birds. A long series of similar examples might be given, and modern embryology in particular has contributed much that, like this example of birds' fingers, points to a certain orderliness in the disappearance of the individual parts of an organ which has become superfluous. Parts which, in the complete animal, have disappeared without leaving a trace, appear again in each embryonic primordium, and disappear in the course of the ontogeny. Speaking metaphorically, we might express this on the basis of the determinant theory, by saying that the determinants, as they become weaker, can only control an increasingly short period of the whole ontogeny of the organ, so that ultimately nothing more than its first beginning comes into existence. But this is only a metaphor; we cannot tell what really happens as long as we are ignorant of the physiological rôle of the determinants, and even of the laws governing the degeneration of a useless organ. In respect of the latter, much might still be achieved if comparative anatomy and embryology were studied with this definite end in view, and perhaps we should even be able to draw more definite conclusions in regard to the composition and activity of the determinants in the germ.

In the meantime we must be content with the knowledge that, on the determinant hypothesis, the disappearance of organs which have become useless may be regarded as a process of intra-selection going on between the 'primary constituents' (_Anlagen_) of the germ, and depending on the same principle of the 'struggle of parts' which William Roux introduced into science with such brilliant results. If a struggle for food and space actually takes place, then every passive weakening must lead to a permanent condition of weakness and a lasting and irretrievable diminution in the size and strength of the primary constituent concerned, unless personal selection intervenes, and choosing out the strongest among these weakened primary constituents, raises them again to their former level. But this never happens when the organ has become useless.

This explains why not only parts with active function, like limbs, muscles, tendons, nerves, and glands, disappear when they cease to function, but also passive parts like the colouring of the external surfaces of animals, the lifeless skeletal parts of Arthropods and the exact adaptation of their thickness to the dwindling function, the disappearance of superfluous wing-veins, and of the hard chitinous covering of the abdomen when it is concealed in a protecting house, as in the case of hermit-crabs, Phryganidæ, and Psychidæ. Here too we find a sufficient explanation of the fact that parts which have become functionless, such as the wings of ants, can disappear even in the case of sterile workers.

* * * * *

The principle of germinal selection, however, can only be understood in its full significance if we take the positive aspect also into consideration. We had reached the conclusion that because of the fluctuations of the food-supply one set of the homologous determinants represented in the various ids may vary in a minus direction, and another set in a plus direction, and that this direction will be adhered to as long as no intra-germinal obstacles come in the way. As long as this does not happen the determinant concerned will pursue the path of variation it has once struck out, and indeed the tendency will be strengthened, because every passive variation, upwards or downwards, results in a strengthening or weakening of the determinant's power of assimilation.