Essays Upon Heredity and Kindred Biological Problems

The theory of natural selection explains the rise of new species by supposing that changes occur, from time to time, in those conditions of life to which an organism must adapt itself if it is to continue in existence. Thus a selective process is set up which ensures that only those out of the existing variations are preserved, which correspond in the highest degree to the changed conditions of life. By continued selection in the same direction the deviations from the type, although at first very insignificant, are accumulated and increased until they become specific differences.

I should wish to assert more definitely than Darwin has done, that alterations in the conditions of life, together with changes in the organism itself, must have advanced very gradually and by the smallest steps, in such a way that, at each period in the whole process of transformation, the species has remained sufficiently adapted to the surrounding conditions. An abrupt transformation of a species is inconceivable, because it would render the species incapable of existence. If the whole organization of an animal depends upon adaptation, if the animal body is, as it were, an extremely complex combination of new and old adaptations, it would be a highly remarkable coincidence if, after any sudden alteration occurring simultaneously in many parts of the body, all these parts were changed in such a manner that they again formed a whole which exactly corresponded to the altered external conditions. Those who assume the existence of such a sudden transformation overlook the fact that everything in the animal body is exactly calculated to maintain the existence of the species, and that it is just sufficient for this purpose; and they forget that the minutest change in the least important organ may be enough to render the species incapable of existence.

It may perhaps be objected that the case is different in plants, as is proved by the American weeds which have spread all over Europe, or the European plants which have become naturalized in Australia. Reference might also be made to the plants which inhabited the plains during the glacial epoch, and which at its close migrated to the Alpine mountains and to the far north, and which have remained unaltered under the apparently diverse conditions of life to which they have been subjected for so long a time. Similar instances may also be found among animals. The rabbit, which was brought by sailors to the Atlantic island of Porto Santo, has bred abundantly and remains unchanged in this locality; the European frogs, which were introduced into Madeira, have increased immensely and have become almost a plague; and the European sparrow now thrives in Australia quite as well as with us. But these instances do not prove that adaptation to external conditions of life is not of primary importance; they do not prove that an organism which is adapted to a certain environment will, when unmodified, remain capable of existence amid new surroundings. They only prove that the above-mentioned species found in those countries the same conditions of life as at home, or at least that they met with conditions to which their organization could be subjected without the necessity for modification. Not every new environment includes such changed conditions as will be effective in modifying every species of plant or animal. The rabbit of Porto Santo certainly feeds on herbs different from those which form the food of its relations in Europe, but such a change does not mean an effective alteration in the conditions under which this species lives, for the herbs in both localities are equally well suited to the needs of the animal.

But if we suppose that the wild rabbit, occurring in Europe, were to suddenly lose but a trifle of its wariness, its keen sight, its fine sense of hearing or of smell, or were to suddenly acquire a colour different from that which it now possesses, it would become incapable of existence as a species, and would soon die out. The same result would probably occur if any of its internal organs, such as the lungs or the liver, were suddenly modified. Perhaps single individuals would still remain capable of existence under these circumstances, but the whole species would suffer a certain decline from the maximum development of its powers of resistance, and would thus become extinct. The sudden transformation of a species appears to me to be inconceivable from a physiological point of view, at any rate in animals.

Hence the transformation of a species can only take place by the smallest steps, and must depend upon the accumulation of those differences which characterise individuals, or, as we call them, ‘individual differences.’ There is no doubt that these differences are always present, and thus, at first sight, it appears to be simply a matter of course that they will afford the material by means of which natural selection produces new forms of life. But the case is not so simple as it appeared to be until recently; that is if I am right in believing that in all animals and plants which are reproduced by true germs, only those characters which were potentially present in the germ of the parent can be transmitted to the succeeding generation.

I believe that heredity depends upon the fact that a small portion of the effective substance of the germ, the germ-plasm, remains unchanged during the development of the ovum into an organism, and that this part of the germ-plasm serves as a foundation from which the germ-cells of the new organism are produced[181 - Compare the second and fourth of the preceding Essays, ‘On Heredity’ and ‘The Continuity of the Germ-plasm as the Foundation of a Theory of Heredity.’]. There is therefore continuity of the germ-plasm from one generation to another. One might represent the germ-plasm by the metaphor of a long creeping root-stock from which plants arise at intervals, these latter representing the individuals of successive generations.

Hence it follows that the transmission of acquired characters is an impossibility, for if the germ-plasm is not formed anew in each individual but is derived from that which preceded it, its structure, and above all its molecular constitution, cannot depend upon the individual in which it happens to occur, but such an individual only forms, as it were, the nutritive soil at the expense of which the germ-plasm grows, while the latter possessed its characteristic structure from the beginning, viz. before the commencement of growth.

But the tendencies of heredity, of which the germ-plasm is the bearer, depend upon this very molecular structure, and hence only those characters can be transmitted through successive generations which have been previously inherited, viz. those characters which were potentially contained in the structure of the germ-plasm. It also follows that those other characters which have been acquired by the influence of special external conditions, during the life-time of the parent, cannot be transmitted at all.

The opposite view has, up to the present time, been maintained, and it has been assumed, as a matter of course, that acquired characters can be transmitted; furthermore, extremely complicated and artificial theories have been constructed in order to explain how it may be possible for changes produced by the action of external influences, in the course of a life-time, to be communicated to the germ and thus to become hereditary. But no single fact is known which really proves that acquired characters can be transmitted, for the ascertained facts which seem to point to the transmission of artificially produced diseases cannot be considered as a proof; and as long as such proof is wanting we have no right to make this supposition, unless compelled to do so by the impossibility of suggesting a mode in which the transformation of species can take place without its aid. (See Appendix IV, p. 310 (#x27_x_27_i17).)

It is obvious that the unconscious conviction that we need the aid of acquired characters has hitherto securely maintained the assumed axiom of the transmission of such features. It was believed that we could not do without such an axiom in order to explain the transformation of species; and this was believed not only by those who hold that the direct action of external influences plays an important part in the process, but also by those who hold that the operation of natural selection is the main factor.

Individual variability forms the most important foundation of the theory of natural selection: without it the latter could not exist, for this alone can furnish the minute differences by the accumulation of which new forms are said to arise in the course of generations. But how can such hereditary individual characters exist if the changes wrought by the action of external influences, during the life of an individual, cannot be transmitted? We are clearly compelled to find some other source of hereditary individual differences, or the theory of natural selection would collapse, as it certainly would if hereditary individual variations did not exist. If, on the other hand, acquired differences are transmitted, this would prove that there must be something wrong in the theory of the continuity of the germ-plasm, as above described, and in the non-transmission of acquired characters which results from this theory. But I believe that it is possible to suggest that the origin of hereditary individual characters takes place in a manner quite different from any which has been as yet brought forward. To explain this origin is the task which I am about to undertake in the following pages.

The origin of individual variability has been hitherto represented somewhat as follows. The phenomena of heredity lead to the conclusion that each organism is capable of producing germs, from which, theoretically at least, exact copies of the parent may arise. In reality this is never the case, because each organism possesses the power of reacting on the different external influences with which it is brought into contact, a power without which it could neither develope nor exist. Each organism reacting in a different way must be to some extent changed. Favourable nutrition makes such an organism strong and large; unfavourable nutrition renders it small and weak, and what is true of the whole organism may also be said of its parts. Now it is obvious that even the children of the same mother meet with influences different in kind and degree, from the very beginning of their existence, so that they must necessarily become unlike, even if we suppose them to have been derived from absolutely identical germs, with precisely the same hereditary tendencies.

In this manner individual differences are believed to have been introduced. But if acquired characters are not transmitted the whole chain of argument collapses, for none of those changes which are caused by the conditions of nutrition acting upon single parts of the whole organism, including the results of training and of the use or disuse of single organs,—none of these changes can furnish hereditary differences, nor can they be transmitted to succeeding generations. They are, as it were, only transient characters as far as the species is concerned.

The children of accomplished pianists do not inherit the art of playing the piano; they have to learn it in the same laborious manner as that by which their parents acquired it; they do not inherit anything except that which their parents also possessed when children, viz. manual dexterity and a good ear. Furthermore, language is not transmitted to our children, although it has been practised not only by ourselves but by an almost endless line of ancestors. Only recently, facts have again been worked up and brought together, which show that children of highly civilized nations have no trace of a language when they have grown up in a wild condition and in complete isolation[182 - Compare Rauber, ‘Homo sapiens ferus oder die Zustände der Verwilderten.’ Leipzig, 1885.]. The power of speech is an acquired or transient character: it is not inherited, and cannot be transmitted: it disappears with the organism which manifests it. Not only do similar phenomena occur in the vegetable kingdom, but they present themselves in an especially striking manner.

When Nägeli[183 - ‘Sitzungsberichte der baierischen Akademie der Wissenschaften,’ vom 18 Nov. 1865. Compare also his ‘Mechanisch-physiologische Theorie der Abstammungslehre,’ p. 102, etc.] introduced Alpine plants, taken from their natural habitat, into the botanical garden at Munich, many of the species were so greatly altered that they could hardly be recognized: for instance, the small Alpine hawk-weeds became large and thickly branching, and they blossomed freely. But if such plants, or even their descendants, were removed to a poor gravelly soil the new characters entirely disappeared, and the plants were re-transformed into the original Alpine form. The re-transformation was always complete, even when the species had been cultivated in rich garden soil for several generations.

Similar experiments with identical results were made twenty years ago by Alexis Jordan[184 - Jordan, ‘Remarques sur le fait de l’existence en société des espèces végétales affines.’ Lyon, 1873.], who chiefly made use of Draba verna in his researches. These experiments furnish very strong proofs, because they were originally undertaken without the bias which may be given by a theory. Jordan only intended to decide experimentally whether the numerous forms of the plant, as it occurs wild in different habitats, are mere varieties or true species. He found that the different forms do not pass into one another, and are in all cases re-transformed after they have been altered by cultivation in a soil different from that in which they usually grow, and he therefore assumed that they were true species. All these experiments therefore confirm the conclusion that external influences may alter the individual, but that the changes produced are not transmitted to the germs, and are never hereditary.

Nägeli indeed asserts that innate individual differences do not exist in plants. The differences which we find, for instance, between two beeches or oaks, are always, according to him, modifications produced by the influence of varying local conditions. But it is obvious that Nägeli goes too far in this respect, although it may be conceded that innate individual differences in plants are much more difficult to distinguish from those which are acquired, than in animals.

There is no doubt about the occurrence of innate and hereditary individual characters in animals, and we may find an especially interesting illustration in the case of man. The human eye can with practice appreciate the most minute differences between individual men, and especially differences of feature. Every one knows that peculiarities of feature persist in certain families through a long series of generations. I need hardly remind the reader of the broad forehead of the Julii, the projecting chin of the Hapsburgs, or the curved nose of the Bourbons. Hence every one can see that hereditary individual characters do unquestionably exist in man. The same conclusion may be affirmed with equal certainty for all our domestic animals, and I do not see any reason why there should be any doubt about its application to other animals and to plants.

But now the question arises,—How can we explain the presence of such characters consistently with a belief in the continuity of the germ-plasm, a theory which implies the rejection of the supposition that acquired characters can become hereditary? How can the individuals of any species come to possess various characters which are undoubtedly hereditary, if all changes which are due to the influence of external conditions are transient and disappear with the individual in which they arose? Why is it that individuals are distinguished by innate characters, as well as by those which I have previously called transient, and how can deep-seated hereditary characters arise at all, if they are not produced by the external influences to which the individual is exposed?

In the first place it may be argued that external influences may not only act on the mature individual, or during its development, but that they may also act at a still earlier period upon the germ-cell from which it arises. It may be imagined that such influences of different kinds might produce corresponding minute alterations in the molecular structure of the germ-plasm, and as the latter is, according to our supposition, transmitted from one generation to another, it follows that such changes would be hereditary.

Without altogether denying that such influences may directly modify the germ-cells, I nevertheless believe that they have no share in the production of hereditary individual characters.

The germ-plasm or idioplasm of the germ-cell (if this latter term be preferred) certainly possesses an exceedingly complex minute structure, but it is nevertheless a substance of extreme stability, for it absorbs nourishment and grows enormously without the least change in its complex molecular structure. With Nägeli we may indeed safely affirm so much, although we are unable to acquire any direct knowledge as to the constitution of germ-plasm. When we know that many species have persisted unchanged for thousands of years, we have before us the proof that their germ-plasm has preserved exactly the same molecular structure during the whole period. I may remind the reader that many of the embalmed bodies of the sacred Egyptian animals must be four thousand years old, and that the species are identical with those now existing in the same locality. Now, since the quantity of germ-plasm contained in a single germ-cell must be very minute, and since only a very small fraction can remain unchanged when the germ-cell developes into an organism, it follows that an enormous growth of this small fraction must take place in every individual, for it must be remembered that each individual produces thousands of germ-cells. It is therefore not too much to say that, during a period of four thousand years, the growth of the germ-plasm in the Egyptian ibis or crocodile must have been quite stupendous. But in the animals and plants which inhabit the Alps and the far north, we have instances of species which have remained unchanged for a much longer period, viz. for the time which has elapsed between the close of the glacial epoch and the present day. In such organisms the growth of the germ-plasm must therefore have been still greater.

If nevertheless the molecular structure of the germ-plasm has remained precisely the same, this substance cannot be readily modifiable, and there is very little chance of the smallest changes being produced in its molecular structure, by the operation of those minute transient variations in nutrition to which the germ-cells, together with every other part of the organism, are exposed. The rate of growth of the germ-plasm will certainly vary, but its structure is unlikely to be affected for the above-mentioned reasons, and also because the influences are mostly changeable, and occur sometimes in one and sometimes in another direction.

Hereditary individual differences must therefore be derived from some other source.

I believe that such a source is to be looked for in the form of reproduction by which the great majority of existing organisms are propagated: viz. in sexual, or, as Häckel calls it, amphigonic reproduction.

It is well known that this process consists in the coalescence of two distinct germ-cells, or perhaps only of their nuclei. These germ-cells contain the germ-substance, the germ-plasm, and this again, owing to its specific molecular structure, is the bearer of the hereditary tendencies of the organism from which the germ-cell has been derived. Thus in amphigonic reproduction two groups of hereditary tendencies are as it were combined. I regard this combination as the cause of hereditary individual characters, and I believe that the production of such characters is the true significance of amphigonic reproduction. The object of this process is to create those individual differences which form the material out of which natural selection produces new species.

At first sight this conclusion appears to be very startling and almost incredible, because we are on the contrary inclined to believe that the continued combination of existing differences, which is implied by the very existence of amphigonic reproduction, cannot lead to their intensification, but rather to their diminution and gradual obliteration. Indeed the opinion has already been expressed that deviations from the specific type are rapidly destroyed by the operation of sexual reproduction. Such an opinion may be true with regard to specific characters, because the deviations from a specific type occur in such rare cases that they cannot hold their ground against the large number of normal individuals. But the case is different with those minute differences which are characteristic of individuals, because every individual possesses them, although of a different kind and degree. The extinction of such differences could only take place if a few individuals constituted a whole species; but the number of individuals which together represent a species is not only very large but generally incalculable. Cross-breeding between all individuals is impossible, and hence the obliteration of individual differences is also impossible.

In order to explain the effects of sexual reproduction, we will first of all consider what happens in monogonic or unisexual reproduction, which actually occurs in parthenogenetic organisms. Let us imagine an individual producing germ-cells, each of which may by itself develope into a new individual. If we then suppose a species to be made up of individuals which are absolutely identical, it follows that their descendants must also remain identical through any number of generations, if we neglect the transient non-transmissible peculiarities caused by differences of food and other external conditions.

Although the individuals of such a species might be actually different, they would be potentially identical: in the mature state they might differ, but they must have been identical in origin. The germs of all of them must contain exactly the same hereditary tendencies, and if it were possible for their development to take place under exactly the same conditions, identical individuals would be produced.

Let us now assume that the individuals of such a species, reproducing itself by the monogonic process and therefore without cross-breeding, differ, not only in transient but also in hereditary characters. If this were the case, each individual would produce descendants possessing the same hereditary differences which were characteristic of itself; and thus from each individual a series of generations would emanate, the single individuals of which would be potentially identical with each other and with their first ancestor. Hence the same individual differences would be repeated again and again, in each succeeding generation, and even if all the descendants lived to reproduce themselves, there would be at last just as many groups of potentially identical individuals as there were single individuals at the beginning.

Similar cases actually occur in many species in which sexual reproduction has been entirely replaced by the parthenogenetic method, as in many species of Cynips and in certain lower Crustacea. But all these differ from our hypothetical case in one important respect; it is always impossible for all the descendants to reach maturity and reproduce themselves. The vast majority of the descendants generally perish at an early stage, and only about as many remain to continue the species as reached maturity in the preceding generation.

We have now to consider whether such a species can be subject to the operation of natural selection. Let us take the case of an insect living among green leaves, and possessing a green colour as a protection against discovery by its enemies. We will assume that the hereditary individual differences consist of various shades of green. Let us further suppose that the sudden extinction of its food-plant compelled this species to seek another plant with a somewhat different shade of green. It is clear that such an insect would not be completely adapted to the new environment. It would therefore be compelled, metaphorically speaking, to endeavour to bring its colour into closer harmony with that of the new food-plant, or else the increased chances of detection given to its enemies would lead to its slow but certain extinction.

It is obvious that such a species would be altogether unable to produce the required adaptation, for ex hypothesi, its hereditary variations remain the same, one generation after another. If therefore the required shade of green was not previously present, as one of the original individual differences, it could not be produced at any time. If, however, we suppose that such a colour existed previously in certain individuals, it follows that those with other shades of green would be gradually exterminated, while the former would alone survive. But this process would not be an adaptation in the sense used in the theory of natural selection. It would indeed be a process of selection, but it could form no more than the beginning of that process which we call natural selection. If the latter could only bring existing characters into prominence, it would not be worth much consideration, for it could never produce a new species. A species never includes, from the beginning, individuals which deviate from the specific type as widely as the individuals of the most nearly allied species deviate from it. And it would be still less possible to explain, on such a principle, the origin of the whole organic world; for, if so, all existing species would have been included as variations of the first species. Natural selection must be able to do infinitely more than this, if it is to be of any importance as a principle of development. It must be able to accumulate minute existing differences in the required direction, and thus to create new characters. In our example it ought to be able, after preserving those individuals with a colour nearest to the required shade, to lead their descendants onward through successive stages towards a complete harmony of colour.

But such a result is quite unattainable with the asexual method of reproduction: in other words, natural selection, in the true meaning of the term, viz. a process which could produce new characters in the manner above described, is an impossibility in a species propagated by asexual reproduction.

If it could be shown that a purely parthenogenetic species had become transformed into a new one, such an observation would prove the existence of some force of transformation other than selective processes, for the new species could not have been produced by these latter. As already explained, the only selection which would be possible for such a species, would lead to the survival of one group of individuals and to the extinction of all others. Thus in our example that group of individuals would alone survive, the ancestors of which originally possessed the appropriate colour. But if one group alone survived, it follows that all hereditary individual differences would have disappeared from the species, for the members of such a single group are identical with one another and with their original ancestors. We thus reach the conclusion that monogonic reproduction can never cause hereditary individual variability, but that, on the other hand, it is very likely to lead to its entire suppression.

But the case is very different with sexual reproduction. When once individual differences have begun to appear in a species propagated by this process, uniformity among its individuals can never again be reached. So far from this being the case, the differences must even be increased in the course of generations, not indeed in intensity, but in number, for new combinations of the individual characters will continually arise.

Again, assuming the existence of a number of individuals which differ from one another by a few hereditary individual characters, it follows that no individual of the second generation can be identical with any other. They must all differ, not only actually but also potentially, for their differences exist at the very beginning of development, and do not solely depend upon the accidental conditions under which they live. Moreover, no one of the descendants can be identical with any of the ancestors, for each of the former unites within itself the hereditary tendencies of two parents, and its organism is therefore, as it were, a compromise between two developmental tendencies. Similarly in the third generation, the hereditary tendencies of two individuals of the second generation enter into combination. But since the germ-plasm of the latter is not simple, but composed of two individually distinct kinds of germ-plasm, it follows that an individual of the third generation is a compromise between four different hereditary tendencies. In the fourth generation, eight; in the fifth, sixteen; in the sixth, thirty-two different hereditary tendencies must come together, and each of them will make itself more or less felt in some part of the future organism. Thus by the sixth generation a large number of varied combinations of ancestral individual characters will appear, combinations which have never existed before and which can never exist again.

We do not know the number of generations over which the specific hereditary tendencies of the first generation can make themselves felt. Many facts seem to indicate however that the number is large, and it is at all events greater than six. When we remember that, in the tenth generation, a single germ contains 1024 different germ-plasms, with their inherent hereditary tendencies, it is quite clear that continued sexual reproduction can never lead to the re-appearance of exactly the same combination, but that new ones must always arise.

New combinations are all the more probable because the different idioplasms composing the germ-plasm in the germ-cells of any individual are present in different degrees of intensity at different times of its life; in other words, the intensity of the component idioplasms is a function of time. This conclusion follows from the fact that children of the same parents are never exactly identical. In one child the characters of the father may predominate, in another those of the mother, in another again those of either grand-parent or great-grand-parent.

We are thus led to the conclusion that even in a few sexually produced generations a large number of well-marked individuals must arise: and this would even be true of generations springing from our hypothetical species, assumed to be without ancestors, and characterised by few individual differences. But of course organisms which reproduce themselves sexually are never without ancestors, and if these latter were also propagated by the sexual method, it follows that each generation of every sexual species is in the stage which we have previously assumed for the tenth or some much later generation of the hypothetical species. In other words, each individual contains a maximum of hereditary tendencies and an infinite variety of possible individual characters (see Appendix VI, p. 326 (#x28_x_28_i6)).

In this manner we can explain the origin of hereditary individual variability as it is known in man and the higher animals, and as it is required for the theory which explains the transformation of species by means of natural selection.

Before proceeding further, I must attempt to answer a question which obviously suggests itself. For the sake of argument, I have assumed the existence of a first generation, of which the individuals were already characterised by individual differences. Can we find any explanation of these latter, or are we compelled to take them for granted, without any attempt to enquire into their origin? If we abandon this enquiry, we can never achieve a complete solution of the problems of heredity and variability. We have, it is true, shown that hereditary differences, when they have once appeared, would, through sexual reproduction, undergo development into the diverse forms which actually exist; but this conclusion affords us no explanation of the source whence such differences have been derived. If the external conditions acting directly upon an organism can only produce transient (viz. non-hereditary) differences in the latter, and if, on the other hand, the external influences which act upon the germ-cell can only produce a change in its molecular structure after operating over very long periods, it seems that we have exhausted all the possible sources of hereditary differences without reaching any satisfactory explanation.

I believe, however, that an explanation can be given. The origin of hereditary individual variability cannot indeed be found in the higher organisms—the Metazoa and Metaphyta; but it is to be sought for in the lowest—the unicellular organisms. In these latter the distinction between body-cell and germ-cell does not exist. Such organisms are reproduced by division, and if therefore any one of them becomes changed in the course of its life by some external influence, and thus receives an individual character, the method of reproduction ensures that the acquired peculiarity will be transmitted to its descendants. If, for instance, a Protozoon, by constantly struggling against the mechanical influence of currents in water, were to gain a somewhat denser and more resistent protoplasm, or were to acquire the power of adhering more strongly than the other individuals of its species, the peculiarity in question would be directly continued on into its two descendants, for the latter are at first nothing more than the two halves of the former. It therefore follows that every modification which appears in the course of its life, every individual character, however it may have arisen, must necessarily be directly transmitted to the two offspring of a unicellular organism.

The pianist, whom I have already used as an illustration, may by practice develope the muscles of his fingers so as to ensure the highest dexterity and power; but such an effect would be entirely transient, for it depends upon a modification in local nutrition which would be unable to cause any change in the molecular structure of the germ-cells, and could not therefore produce any effect upon the offspring. And even if we admit that some change might be caused in the germ-cells, the chances would be infinity to nothing against the production of the appropriate effect, viz. such a change as would lead to the development in the child of the acquired characters of the parent.

In the lowest unicellular organisms, however, the case is entirely different. Here parent and offspring are still, in a certain sense, one and the same thing: the child is a part, and usually half, of the parent. If therefore the individuals of a unicellular species are acted upon by any of the various external influences, it is inevitable that hereditary individual differences will arise in them; and as a matter of fact it is indisputable that changes are thus produced in these organisms, and that the resulting characters are transmitted. It has been directly observed that individual differences do occur in unicellular organisms,—differences in size, colour, form, and the number or arrangement of cilia. It must be admitted that we have not hitherto paid sufficient attention to this point, and moreover our best microscopes are only very rough means of observation when we come to deal with such minute organisms. Nevertheless we cannot doubt that the individuals of the same species are not absolutely identical.

We are thus driven to the conclusion that the ultimate origin of hereditary individual differences lies in the direct action of external influences upon the organism. Hereditary variability cannot however arise in this way at every stage of organic development, as biologists have hitherto been inclined to believe. It can only arise in the lowest unicellular organisms; and when once individual difference had been attained by these, it necessarily passed over into the higher organisms when they first appeared. Sexual reproduction coming into existence at the same time, the hereditary differences were increased and multiplied, and arranged in ever-changing combinations.

Sexual reproduction can also increase the differences between individuals, because constant cross-breeding must necessarily and repeatedly lead to a combination of forces which tend in the same direction, and which may determine the constitution of any part of the body. If, for instance, the same part of the body is strongly developed in both parents, the experience of breeders tells us that the part in question is likely to be even more strongly developed in the offspring; and that weakly developed parts will in the same manner tend to become still weaker. Amphigonic reproduction therefore ensures that every character which is subject to individual fluctuation must appear in many individuals with a strengthened degree of development, in many others with a development which is less than normal, while in a still larger number of individuals the average development will be reached. Such differences afford the material by means of which natural selection is able to increase or weaken each character according to the needs of the species. By the removal of the less well-adapted individuals, natural selection increases the chance of beneficial cross-breeding in the subsequent generations.

Every one must admit that, if a species came into existence having only a small number of individual differences which appeared in the different parts of different individuals, the number of differences would increase with each sexually produced generation, until all the parts in which the variations occurred had received a peculiar character in all individuals.

Moreover sexual reproduction not only adds to the number of existing differences, but it also brings them into new combinations, and this latter consequence is as important as the former.

The former consequence can hardly make itself felt in any existing species, because in them every part already possesses its peculiar character in all individuals. The second consequence is, however, more important, viz. the production of new combinations of individual characters by sexual reproduction; for, as Darwin has already pointed out, we must imagine that not only are single characters changed in the process of breeding, but that probably several, and perhaps very many characters, are simultaneously modified. No two species, however nearly allied, differ from each other in but a single character. Even our eyesight, which has by no means reached the highest pitch of development, can always detect several, and often very many points of difference; and if we possessed the powers necessary for making an absolutely accurate comparison, we should probably find that everything is different in two nearly allied species.