Weismann August
Essays Upon Heredity and Kindred Biological Problems
Thus the very careful and extensive investigations of Poulton284 upon the colours of certain caterpillars have distinctly shown that some species possess the possibility of development in two directions, and that the actual direction taken by the individual is decided by the influence of external conditions. Poulton surrounded certain larvae of Geometrae with an abundance of dark branches, in addition to the leaves upon which they fed. When such conditions prevailed from the beginning of larval life, the caterpillars as they developed, gradually assumed the dark colour of the twigs and branches upon which they rested. When other larvae of the same species (and in many experiments hatched from the same batch of eggs) were similarly exposed to the green leaves of the same food-plant, they did not indeed become bright green like the leaves, but were invariably of a much lighter colour than the other larvae, while many of them gained a brownish-green tint. The larvae of Smerinthus ocellatus285 also possess the power of assuming different shades of green and of thus approaching, to some extent, the green of the plant upon which they happen to live. It is quite impossible to explain the phyletic development of the green colour of these and other caterpillars as due to the direct action upon the skin of the green light reflected from the leaves upon which they sit. The impossibility of such an effect was pointed out long ago by Darwin, and also followed from my own investigations. Here, as in the other cases, the only possible solution is afforded by natural selection. The colour of the caterpillars has become gradually more and more perfectly adapted to the colour of the leaves,—and often to the particular side of the leaves upon which these animals rest,—not by the direct effect of reflected light, but by the selection of those individuals which were best protected. Poulton’s experiments quoted above prove that certain species which occur upon different plants with different colours (or even in some cases upon the differently coloured parts of the same plant), present us with a further complication in the process of adaptation, inasmuch as each individual has acquired the power of assuming a lighter or darker colour286. The light which falls upon a single individual caterpillar during the course of its growth determines whether the lighter or darker colour shall be developed. Here therefore we have a case exactly parallel to that of the Thuja-shoot in which the palisade or spongy parenchyma is developed according to the position in which the shoot is fixed.
As far as it is possible in the present condition of our knowledge to offer any opinion upon the origin of sex in bisexual animals, it may be suggested that this problem is also capable of an essentially similar solution. Each germ-cell may possess the possibility of developing in either of two directions, the one resulting in a male individual, and the other resulting in a female, while the decision as to which of the two possible alternatives is actually taken may rest with the external conditions. We must, however, include among the external circumstances everything which is not germ-plasm. Moreover, this explanation is by no means certain, and I only mention it as an instance which, if we assume it to be correct, further illustrates my views upon the phenomena presented by the Thuja-shoot.
The two other facts brought forward by Detmer as proofs of the transforming power of external influences can be explained in precisely the same manner. These instances are—the fact that Tropaeolum when grown in moist air produces leaves with anatomical characters different from those produced when the plant is grown in dry air; and the differences in the structure of the leaves of many plants, according as they have been grown in the sun or shade respectively. Such differences do not by any means afford proof of the direct production of structural changes by means of external influences. How would such an explanation be consistent with the fact that the leaves are, in all these cases, changed in a highly purposeful manner? Or is it assumed that these organs were so constituted from the beginning, that they are compelled to respond to external conditions by the production of useful changes? Any one who made such an assertion nowadays, or who even thought of such a thing as a possibility, would prove that he is entirely ignorant of the facts of organic nature, and that he has no claim to be heard upon the question of the transformation of species. The very first necessity in any scientific question is to gain acquaintance with that which has been thought and said upon the subject. And it has been frequently shown that whole groups of useful characters cannot by any possibility have been produced by the direct action of external influences. If a caterpillar, which hides itself by day in the crevices of the bark, possesses the same colour as the latter, while other caterpillars which rest on leaves are of a green colour, these facts cannot be explained as the results of the direct influence of the bark and leaves. And it would be even less possible to explain upon the same principle all the details of marking and colour by which these animals gain still further protection. If the upper side of the upper wings of certain moths is grey like the stone on which they rest by day, while in butterflies the under side of both wings which are exposed during rest, exhibits analogous protective colours, these facts cannot be due to the direct influence of the surroundings which are resembled, but, if they have arisen in any natural manner, they must have been indirectly produced by the surroundings. One may reasonably complain when compelled to repeat again and again these elements of knowledge and of thought upon the causes of transformation!
Any one who remembers these things, and is aware of the countless number of purposeful characters which cannot possibly depend upon such direct influences, will be very cautious in yielding to any single instance which at first sight appears to be the direct consequence of external conditions. If Detmer had been thus cautious he would hardly have written the following sentence as a résumé of the physiological experiments on plants which have been already alluded to: ‘In certain cases it is possible, as we have seen, to artificially modify the anatomical structure of certain parts of plants. In such cases the relation between the structure and the external influences is undoubtedly clear: the latter act as the cause; the anatomical structure of the members of the plant is the consequence of this cause.’ A little more logic would have prevented the author from expressing such an opinion, for, as has been already shown, it is founded on a confusion between the true cause of a phenomenon and one of the conditions which are necessary for its production. We might as well consider the phenomena of geotropism, hydrotropism, and heliotropism—which have been established, and investigated in such a brilliant way by modern vegetable physiologists—as the direct results of the attraction of the earth, of water, and of light; and it is not improbable that some botanists are even inclined to make this assumption. And yet it is perfectly easy to show that this cannot be the case. By geotropism we mean the power possessed by the parts of a plant of growing along lines which make certain angles with the direction of the earth’s attraction. For example, the chief root grows parallel with the earth’s attraction, viz. towards the centre of the earth, and it is described as positively geotropic: conversely the main shoot grows along the same line but in an opposite direction, and it is negatively geotropic. But geotropism is not a primitive attribute of the plant, and it is even now absent from those plants which, like many Algae, have no definite position. Geotropism cannot have arisen before plants first became fixed in the earth. If any one were to assume that the direct influence of gravity, continuous through countless generations, had at length conferred upon the root the power of growing in a geotropic direction, how would it be possible to explain the fact that the shoot which has been under precisely the same influence has acquired the power of growing in an exactly opposite direction? The characteristic differences between root and shoot cannot have appeared until the plant became fixed in the ground, and how can we imagine that the same influence of gravity has since that time directly produced the two antagonistic results of positive and negative geotropism, in two structures, which were originally and essentially similar? It should also be remembered that it is only the main root which exhibits true positive geotropism. The lateral roots form angles with the main root, and do not therefore grow towards the earth’s centre; and the same is true of the lateral shoots which grow obliquely, and not perpendicularly upwards, like the main shoot. Moreover the angles which the lateral roots make with the main root, and the lateral shoots with the main shoot, are quite different in different species. How is it possible that all these different modes of reaction witnessed in the different parts of plants can be the direct results of one and the same external force? It is quite obvious that these are all cases of adaptation. The main root has not acquired the power of growing perpendicularly downwards under the stimulus of gravity, because this force has acted upon it for numberless generations, but because such a direction for such a part was the most useful to the plant. Hence natural selection has conferred upon the root the power of reacting under the stimulus of gravity by growing in a direction parallel to this force. For the main shoot, the opposite reaction was the most useful and has been established by natural selection, while still another reaction has been similarly established for the lateral roots and another for the lateral shoots.
Each part of a plant has received its special mode of reacting under the stimulus of gravity because it was useful for the whole plant, inasmuch as the position of its different parts relatively to one another and to the soil became thus fixed and regulated. These modes of reaction have become different in different species, because the conditions of life peculiar to each require special arrangements.
The same argument also holds with regard to heliotropism. The power of growing towards the light possessed by green shoots cannot be a primitive character of the plant: it must have arisen secondarily. If it were an essential and original character it could not be reversed in certain parts of the plant; but, the roots are negatively heliotropic, for they grow away from the light. There are also shoots, such as the climbing shoots of Ivy, which are similarly negatively heliotropic. Whenever the heliotropic power is thus reversed in shoots, the change is of a useful kind. Thus the shoots of the Ivy gain the power of clinging closely to a perpendicular wall or to some horizontal plane287. In this case, however, it is only the shoot which is negatively heliotropic, its leaves turn towards the light; and the same is true of the flower-bearing shoots which do not climb. All these are clearly adaptations and not the results of direct influence. The light only provides the stimulus which calls forth the characteristic reaction from each part of the plant, but the cause of each peculiar reaction lies in the specific nature of the part itself which has not been produced by light, but as we believe by processes of natural selection. If this explanation does not account for the facts we may as well abandon all attempts at understanding the useful arrangements in organisms.
Sachs has used the term anisotropism to express the fact that the various organs of a plant assume the most diverse directions of growth under the influence of the same forces. He also states that anisotropism is one of the most general characteristics of vegetable organization, and that it is quite impossible to form any idea as to how plants would appear or how they could live if their different organs were not anisotropic. Since anisotropism is nothing more than the expression of different kinds of susceptibility to the action of gravity, light, &c., it is obvious that the configuration of the plant is to be traced to such specific susceptibilities.
Now these specific susceptibilities cannot have been produced by the direct effect of the various external influences (as was shown above), and the only other possible explanation is to recognise them as adaptations, and to admit that they have arisen by the operation of natural selection upon the general variability of plant organization.
Simple as these conclusions are, I have failed to meet with them in any of the writings of botanists, and they may perhaps be of use in helping to shake the vaguely-felt opinion that the characters of plants are to be chiefly referred to the direct action of external influences.
At all events it cannot be maintained that the phenomena of anisotropism support the opinion mentioned above; and the mere assertion that it is highly probable that hereditary characters arise as the result of external influences, is no more than the expression of an unfounded individual opinion. It is remarkable that Detmer should make such an assertion as the outcome of his discussion of the reversed Thuja-shoot, &c., for even if we admit that the dorso-ventral structure of the shoot is—as Detmer believes—the direct and primary effect of the action of light, the experiment with the reversed shoot would prove that no part of this effect has become hereditary. Although the upper side of the shoot has produced the palisade parenchyma under the influence of light for thousands of generations, there is nevertheless no tendency towards the establishment of any hereditary effect, for as soon as the upper side of the growing shoot is artificially transformed into the under side, its normal structure is at once abandoned. Hence so far from lending any support to the assumption that acquired characters can be transmitted, Detmer’s experiment rather tends to disprove this opinion.
I think I have sufficiently shown that Detmer’s reproach—that I have under-estimated the effects of external influences upon an organism—may be fairly directed against its author. If we can believe that every structural arrangement in plants, which depends upon certain external conditions, has been produced in a phyletic sense by these latter, it becomes very easy to explain the transformation of species; but in accepting such an explanation we are building without any foundation, for the proof that acquired characters can be transmitted has yet to be given.
As a further disproof of my views Detmer quotes the so-called phenomena of correlation in plants, and he believes that these instances help us to conceive how the acquired changes of the body (soma) of the plant may also influence the sexual cells. If the apical shoot of a young spruce fir be cut off, one of the lateral shoots of the whorl next below the section rises and becomes an apical shoot: it not only assumes the orthotropic growth of such a shoot, but also its mode of branching. The phenomenon itself is well known, and I have often observed it myself in my garden without making any botanical experiments; for this experiment is not uncommonly made by Nature herself, when the apical shoot is destroyed by insects (for example the gall-making Chermes). The change of the lateral into an apical shoot occurs here in consequence of the loss of the true apical shoot, and is therefore really dependent upon it. The only difficulty is to understand how these and many other kindred phenomena can be considered to prove the transmission of acquired characters. That correlation exists between the parts of an organism, that correlated changes are not only common but nearly always accompany some primary change, has been perfectly well known since Darwin’s time, and I am not aware that it has been disputed by any one. I further believe that hardly any one would maintain that it is impossible for the reproductive organs to be influenced by correlation. But this is very far from the admission that such changes would occur in the germ-cells as would be necessary for the transmission of acquired characters. For such transmission to occur it would be necessary for the germ-plasm (the bearer of hereditary tendencies) to undergo a transformation corresponding to that produced by the external influences;—such a transformation as would cause the future organism to spontaneously develope changes similar to those which its parent had acquired. But since the germ-plasm is not an organism in the sense of being a microscopic facsimile which only has to increase in size in order to become a mature organism, it is obvious that the developmental tendencies must exist in the specific molecular structure, and perhaps also in the chemical constitution of the germ-plasm itself. It therefore follows that the changes in the germ-plasm which would be required for the transmission of an acquired character must be of an entirely different nature from the change itself acquired by the body of the parent plant: and yet it is supposed that the former is produced by the latter as a result of correlation. I will illustrate this by an example. Let us suppose that the influence of climate had caused a plant to change the form of its leaves from an ovate into a lobate shape: now such a change could not be transferred to the germ-plasm in the pollen and the ovules, as anything similar to leaves or the form of leaves; for such specialized morphological features have no existence in the germ-plasm. The only thing which could happen would be changes in its molecular structure which bear no resemblance to those changes which are implied by the direct alteration of the form of the leaf in the parent plant. Any one who clearly appreciates this difficulty will hesitate in admitting the possibility of the transmission of acquired characters, because it is possible that the sexual cells may be affected by correlated influences. If the change in the form of a leaf exercises any influence at all upon the germ-plasm, why should it produce a corresponding (in the above-mentioned sense) change in its molecular structure? Why should it not produce some other out of the immense number of possible changes? There must be as many possible changes in the structure of germ-plasm as there are possible variations in each part of a plant that arises from it. Why then should the corresponding change always occur,—a change which had never previously existed in the whole phyletic development of the organic world; for the plant with the latest modification can have never existed before? The occurrence of a particular change out of the countless possible changes would be about as likely as if one out of a hundred thousand pins thrown out of a window were to balance on its point when it reached the ground. The assumption scarcely deserves to be called a scientific hypothesis, and yet it must be made by all who accept the transmission of acquired characters,—that is unless they adopt the hypothesis of pangenesis, which is quite as improbable, and which even Darwin did not look upon as a real, but only as a formal explanation.
Detmer is also greatly mistaken when he says that I refuse to admit the transmission of acquired characters, because I am prejudiced in favour of my doctrine of the continuity of the germ-plasm. This doctrine is either right or wrong, and there is no middle course: to this extent I quite admit that I am prejudiced. But the question as to whether acquired characters can be impressed upon the germ and thus transmitted would not be by any means settled in this way; for even if we admit that the germ-plasm is not continuous from one generation to another, but that it must be produced afresh in each individual, this would by no means necessarily imply that it would potentially receive and retain every change produced in every part of the individual, and at any time in its life. It seems to me that the problem of the transmission or non-transmission of acquired characters remains, whether the theory of the continuity of the germ-plasm be accepted or rejected.
I will now proceed to examine the last group of phenomena which Detmer brings forward in favour of the transmission of acquired characters. He charges me with not having taken into account, in discussing the problem of heredity, the very important facts which are known about the strange phenomena of ‘after-effect’ in plants. Among these ‘after-effects’ are the following.
If vigorous plants of the sun-flower, grown in the open air, be cut off close to the ground and transferred to complete darkness, the examination of a tube fixed to the cut surface of the stem will show that the escape of sap does not take place uniformly, but undergoes periodical fluctuation, being strongest in the afternoon and weakest in the early morning. Now the cause of this daily periodicity in the flow of sap depends upon the periodical changes due to the light to which the plant was exposed when it was growing under normal conditions. When plants which have been grown in darkness from the first are similarly treated, the flow of sap does not exhibit any such periodicity.
Another instance is as follows:—it is well known that darkness accelerates, while light retards the growth of plants, and therefore plants usually grow more strongly by night than by day. If now plants are transferred from the open air into constant darkness, the periodicity in their growth does not immediately disappear, and often persists for a long time as a phenomenon of after-effect.
The opening and closing of the leaves of Mimosa pudica also takes place periodically under natural conditions, the leaves closing at dusk as a result of changes in the stimulus provided by the light. In this case also, when the plants are transferred to constant darkness, the periodicity in the movements of the leaves continues for several days.
All this is certainly very interesting, and it proves that periodical stimuli produce periodical processes in the plant, which are not immediately arrested when the stimulus is withdrawn, and only become uniform gradually and after the lapse of a considerable time. But I certainly claim the right to ask what connexion there is between these facts and the transmission of acquired characters. All these peculiarities produced by external influences remain restricted to the individual in which they arose; most of them disappear comparatively soon, and long before the death of the individual. No example of the transmission of such a peculiarity is known. Although successive generations of sunflowers have been exposed for thousands of years to the daily alternation of light and darkness, the periodicity in the flow of sap has not become hereditary, and does not take place at all in plants which have always been kept in darkness. Detmer specially tells us that we can even reverse the periods of opening and closing the leaves in Mimosa pudica by keeping them in darkness during the day, but exposed to light at night; an experiment which was performed by Pfeffer. Here again we see the proof that influences which have acted upon countless generations have left no impression whatever upon the germ-plasm.
Detmer himself admits this when he says that the after-effects are only witnessed during the life of the individual, but he nevertheless adds that he has been for many years convinced that the phenomena of heredity and after-effect differ in degree and not in kind. He even goes so far as to assert that, in spite of the obvious non-transmission of after-effect, the similarity between the natures of these two classes of phenomena cannot escape the intelligent observer.
It seems to me that this question does not demand the attention of the observer (for the observations have already been made) so much as that of the thinker. It is not a correct train of reasoning to conclude that after-effect and heredity are identical in nature, from the fact that certain periodical influences, acting upon a single individual, set up periodical physiological processes which continue for a time after the influences have ceased to act. We might almost as well argue that the oscillations of a pendulum, which continue as after-effects when the pendulum has been set going, are of an identical nature with the process of heredity. All these phenomena have indeed this much in common:—a cause which acted at some time in the past, but which is no longer visible at the time when the phenomenon appears. But the likeness ends here, and the supposed identity in nature merely depends upon wild speculation. One difference is very obvious, for the phenomena of after-effect gradually cease after the withdrawal of the stimulus, just like the oscillations of the pendulum, while the phenomena of heredity continue without any interruption. As far as heredity is concerned the physiological processes of after-effect are not distinguishable from any of the other well-known acquired characters which are recognizable as morphological changes. After-effects are not transmitted, and compared with this fact but little importance can be attached to the use of vague analogies by Detmer, who would wish to conclude that heredity is only the after-effect of processes which had been set going in the parent organism.
At the end of his paper Detmer applies the ideas which he has gained from the consideration of after-effect to certain phenomena in the normal life of plants. He suggests that the periodical change of leaf in trees and shrubs may have been produced by the direct effect of climate. If branches bearing winter buds are cut off in the autumn and are placed in a hot-house, with their cut ends in water, the buds do not at once develope, and months may often elapse before they begin to break. He argues that this experiment proves that the annual periodicity of the plant no longer depends directly upon external influences; these latter produced the periodicity at some earlier time, but it has been gradually fixed in the organism by after-effect and heredity(!), so that its disappearance does not now take place when the stimulus is withdrawn, and changes would only happen very gradually under the influence of changed climatic conditions. He considers that this is proved from the fact that our cherry has become an evergreen in Ceylon.
Such are Detmer’s opinions, and every one will agree with him in believing that the periodical change of leaf in temperate climates has been produced in relation to the recurring alternation of summer and winter. This is certainly the case, and it cannot be doubted that the character has become fixed by heredity. Where, however, is the proof that this hereditary character has been produced by the direct influence of climate? What right have we to look upon the hereditary appearance of the character as an after-effect of the direct influence exerted by changes of temperature upon previous generations? Such an opinion derives but little support from the previously described experiments upon after-effect, which showed that these phenomena were never hereditary.
It appears to me that there are certain points in this change of leaf and its accompanying phenomena, which distinctly indicate that natural selection has been at work. Can Detmer imagine that the brown scales which form the characteristic protective covering of winter buds have been produced by the direct action of the cold? If, however, the peculiar structure of these buds is to be referred to the specific constitution of the individual rather than to the direct effects of climate, would it be so very improbable for their physiological peculiarity of lying dormant for several months to have been developed simultaneously with the structure, by the operation of natural selection? And if this explanation be correct, we can at once see why the character has become hereditary, for natural selection works upon variations of the germ-plasm, and these are transferred from one generation to another with the germ-plasm itself.
But Detmer attempts to establish the converse conclusion, and he argues that the hereditary change of leaf has been abandoned under the long-continued effect of changed climatic conditions; but this opinion is based upon the single instance of the alteration in the habit of the European cherry in Ceylon. If it were proved that our cherry, grown from seed in Ceylon and propagated by seed for several generations, became evergreen gradually and not suddenly in the first generation: if, under such circumstances, it came to retain its leaves in the autumn and ceased to produce the dormant winter buds:—then indeed the transmission of acquired characters could hardly be doubted. I am not a botanist, but I believe I am right in supposing that the wild cherry reproduces itself by seeds, while the edible domesticated cherry is propagated by grafting. Grafts are, however, parts of the soma of a previously existent tree, and we are not therefore concerned, in this method of propagation, with a succession of generations, but with the successive distribution of one and the same individual over many wild stocks. But no one will doubt that one and the same individual can be gradually changed during the course of its life, by the direct action of external influences. The really doubtful point is whether such changes can be transmitted by means of the germ-cells. If, as I presume, the English in Ceylon do not care to eat wild cherries but prefer the cultivated kinds, it follows that the branches which bear fruit in that island have not been developed from germ-cells, at any time since their introduction, and there is nothing to prevent them from gradually changing their anatomical and physiological characters in consequence of the direct influence of climate.
