Weismann August
Essays Upon Heredity and Kindred Biological Problems
Hence the instance which Detmer looks upon as plainly conclusive, can hardly be accepted in support of such a far-reaching assumption as the transmission of acquired characters.
It is therefore clear that none of the facts brought forward by Detmer really afford the proofs which he believes that they offer. But another botanist, Professor Hoffman of Marburg, well known for his long-continued experiments on variation, has recently called attention to certain other botanical facts in support of the transmission of acquired characters. These facts are indeed conclusive, if we accept the author’s use of the term ‘acquired,’ but it will be found that they lead to hardly any modification in the state of existing opinion upon the subject.
In a short note, dated Jan. 1, 1888, the author communicated to this journal (‘Biologisches Centralblatt’) the statement that changes in the structure of flowers caused by poor nutrition can be proved to be hereditary to a greater or less extent288.
A more elaborate account of the experiments will be found in several numbers of the ‘Botanische Zeitung,’ and the author expresses his final results in the following words (see Bot. Zeit. 1887, p. 773):—‘These experiments prove with certainty (1) that insufficient nutrition may cause considerable morphological changes (viz. qualitative variations) which are in the first place acquired by the sexual apparatus of the flower, (2) that the “transient” (Weismann) characters acquired by the individual can be transmitted289.’
The data upon which Hoffman bases these opinions are certain experiments conducted upon various plants, in order to determine the conditions of life under which abnormal flowers or any other variations occur most frequently: to decide, in short, how far variations are caused by the change of conditions.
It is obvious that the attention of the author was not at first directed to the question of the transmission of acquired characters. His experiments are of a much older date than the present condition and significance of the question before us. Hoffmann has, in fact, re-examined his former results from the new point of view, and this explains why his proofs are not always sufficiently convincing when applied to the present issue. But this is of no great importance, inasmuch as there is no necessity for me to question the correctness of his assumptions.
The essential details of the experiments to which he directs attention are as follows.
Different plants with normal flowers were subjected to greatly changed conditions of life for a series of generations. They were, for example, crowded together in small pots. Under these circumstances the plants were of course poorly nourished, and in the course of generations, several species produced a variable proportion of abnormal—viz. double-flowers. This, however, was not always the case, for such flowers did not appear in Matthiola annua and Helianthemum polifolium. In other species, such as Nigella damascena, Papaver alpinum and Tagetes patula, they appeared and often increased in numbers in the course of generations, although this was not a constant result. For instance, four successive generations of Nigella damascena, when closely sown, produced the following results:—
But it was not always the case that the double flowers continued to appear after they had been once produced. In Papaver alpinum, which Hoffman has cultivated in successive generations since 1862, other changes in addition to the doubling of the flowers first appeared in 1882, viz. a slight variability in the form of the leaf, and a greater variability in the colours of the flowers. The production of double flowers appeared to be favoured by poor nutrition caused by crowding the plants. The results as regards the number of double flowers produced in this species by close sowing, from 1882-1886, have been as follows:—
Although in these and some other series of generations the double flowers again disappeared in the later generations, yet there can be hardly any doubt that their first appearance was due to the abnormal conditions of nutrition. This conclusion is also unaffected by the fact that double flowers appeared in nearly the same proportions in consequence of cultivation in ordinary garden soil. The plants which were crowded in pots produced 2879 normal flowers, and 256 (=8·8 per cent.) abnormal and mostly double ones, while 867 normal and 62 (=7·0 per cent.) abnormal ones were produced on garden beds. Hoffman will not indeed admit that such a comparison can be fairly made, for the plants in the garden beds were raised from seed which was in part taken from the double flowers, and was therefore, he believed, under a strong hereditary influence. But this latter assumption is not supported by the results of his own experiments.
Thus experiment XVIII., conducted upon Papaver alpinum, is described in these words,—‘Seeds yielded by double flowers from experiment XI. (1883), were sown in pots, and the resulting plants produced from 1884-1886, fifty-three single flowers and no double ones.’
In the converse experiment XIX. ‘The seeds of single flowers from different stocks were sown in pots, and the resulting plants produced in 1885 and 1886 forty-three flowers, of which all were typical except one;’ while plants produced in the garden by seed from the same sources, yielded 166 single and five double flowers. Hoffman also describes other experiments in which the seeds from double flowers produced plants which also yielded many double flowers. Thus, for example, in experiment XXI. seeds yielded by the double flowers of Papaver alpinum were sown in the garden and produced numerous plants, which in 1885 and 1886 bore 284 single and twenty-one double flowers, that is 7 per cent. of the latter.
It will therefore be seen that the transmission of the abnormality is by no means proved beyond the possibility of doubt, for who can decide between the effects due to heredity and changed conditions in the last experiment? I have no doubt however that the results are at any rate in part due to the operation of heredity, for I do not see how the phenomena can be otherwise understood. Nevertheless I cannot admit the transmission of acquired characters on this evidence, for the changes which have appeared are not ‘acquired’ in the sense in which I use the term and in the sense required by the general theory of evolution. It is true that they may be described by the use of this word: inasmuch as they are characters which the plant has come to possess; we are not however engaged in a mere dispute about terms, but in the discussion of a weighty scientific question. Our object is to decide whether changes in the soma (the body, as opposed to the germ-cells) which have been produced by the direct action of external influences, including use and disuse, can be transmitted; whether they can influence the germ-cells in such a manner that the latter will cause the spontaneous appearance of corresponding changes in the next generation. This is the question which demands an answer; and, as has been shown above, such an answer would decide whether the Lamarckian principle of transformation must be retained or abandoned.
I have never doubted about the transmission of changes which depend upon an alteration in the germ-plasm of the reproductive cells, for I have always asserted that these changes, and these alone, must be transmitted. If any one makes the contrary assertion, he merely proves that he does not understand what I have said upon the subject. In what other way could the transformation of species be produced, if changes in the germ-plasm cannot be transmitted? And how could the germ-plasm be changed except by the operation of external influences, using the words in their widest sense; unless indeed we assume with Nägeli, that changes occur from internal causes, and imagine that the phyletic development of the organic world was planned in the molecular structure of the first and simplest organism, so that all forms of life were compelled to arise from it, in the course of time, and would have arisen under any conditions of life. This is the outcome of Nägeli’s view, against which I have contended for years.
If we now use the term ‘acquired characters’ for changes in the soma which, like spontaneous abnormalities, depend upon previous changes in the germ-plasm—it is of course easy to prove that acquired characters are transmitted; but this is hardly the way to advance science, for nothing but confusion would be produced by such a use of terms290. I am not aware that any one has ever doubted that spontaneous characters, such as extra fingers or toes, patches of grey hair, moles, etc., can be transmitted. It is true that such characters are sometimes called ‘acquired’ in pathological works, but His has rightly insisted that such an obviously inaccurate use of the term ought to be avoided, in order to prevent misunderstanding. If every new character is said to be ‘acquired’ the term at once loses its scientific value, which lies in the restricted use. If generally used, it would mean no more than the word ‘new’; but new characters may arise in various ways,—by artificial or natural selection, by the spontaneous variations of the germ, or by the direct effect of external influences upon the body, including the use and disuse of parts. If we assume that these latter characters are transmitted, the further ‘assumption of complicated relations between the organs and the essential substance of the germ becomes necessary’ (His), while the transmission of the other kinds of characters do not involve any theoretical difficulties. There is therefore obviously a wide difference between these two groups of characters as far as heredity is concerned, quite apart from the question as to whether acquired characters are really transmitted. It is at all events necessary to have distinct terms which cannot be misunderstood. His291 has proposed to call those characters which are due to selection ‘changes produced by breeding’ (‘erzüchtete Abänderungen’), those which appear spontaneously—‘spontaneous changes’ (‘eingesprengte Abänderungen’), and these two groups of characters would then be opposed to those which he calls ‘acquired changes’ (‘erworbene Abänderungen’), of course using the term in the restricted sense. Science has always claimed the right of taking certain expressions and applying them in a special sense, and I see no reason why it should not exercise this right in the case of the term ‘acquired.’ It appears moreover that this word has not always been used in this vague sense by pathological anatomists, such as Virchow and Orth; for Weigert and Ernst Ziegler have employed it in precisely the same sense as that in which it has been used by Darwin, du Bois-Reymond, Pflüger, His and many others, including myself.
It is certainly necessary to have two terms which distinguish sharply between the two chief groups of characters—the primary characters which first appear in the body itself, and the secondary ones which owe their appearance to variations in the germ, however such variations may have arisen. We have hitherto been accustomed to call the former ‘acquired characters,’ but we might also call them ‘somatogenic,’ because they follow from the reaction of the soma under external influences; while all other characters might be contrasted as ‘blastogenic,’ because they include all those characters in the body which have arisen from changes in the germ. In this way we might perhaps prevent the possibility of misunderstanding. We maintain that the ‘somatogenic’ characters cannot be transmitted, or rather, that those who assert that they can be transmitted, must furnish the requisite proofs. The somatogenic characters not only include the effects of mutilation, but the changes which follow from increased or diminished performance of function, and those which are directly due to nutrition and any of the other external influences which act upon the body. Among the blastogenic characters, we include not only all the changes produced by natural selection operating upon variations in the germ, but all other characters which result from this latter cause.
If we now wish to place Hoffmann’s results in their right position, we must regard all of them as ‘blastogenic’ characters, for no one of them can be considered as belonging to the group which has been hitherto spoken of as ‘acquired,’ in the literature of evolution: they are not due to somatogenic but to blastogenic changes. The body of the plant—the soma—has not been directly affected by external influences, in Hoffman’s experiments, but changes have been wrought in the germ-plasm of the germ-cells and, only after this, in the soma of succeeding generations.
There is no difficulty in finding facts in support of this statement, among Hoffmann’s experiments. The proof chiefly lies in the fact that in no one of his numerous experiments did any change appear in the first generation. The seeds of different species of wild plants, with normal flowers, were cultivated in the garden and in pots (thickly sown in the latter case), but no one of the plants produced by these wild seeds possessed a single double flower. It was only after a greater or less number of generations had elapsed that a variable proportion of double flowers appeared, sometimes accompanied by changes in the leaves and in the colours of the flowers. This fact admits of only one interpretation;—the changed conditions at first produced slight and ineffectual changes in the idioplasm of the individual, which was transmitted to the following generation: in this again the same causes operated and increased the changes in the idioplasm which was again handed down. Thus the idioplasm was changed more and more, in the course of generations, until at last the change became great enough to produce a visible character in the soma developed from it, such as, for example, the appearance of a double flower. Now the idioplasm of the first ontogenetic stage (viz. germ-plasm) alone passes from one generation to another, and hence it is clear that the germ-plasm itself must have been gradually changed by the conditions of life until the alteration became sufficient to produce changes in the soma, which appeared as visible characters in either the flower or leaf292.
In addition to the above-mentioned cases Hoffmann also quotes some facts of a somewhat different kind. He succeeded in inducing considerable changes in the structure of the root of the wild carrot (Daucus carota) by means of the changes in nutrition implied by garden cultivation. These changes also proved to be hereditary.
Unfortunately, I have not the literature of the subject at hand, and hence I am unable to read the accounts of these older experiments in extenso; but it is sufficiently obvious that in this case we are also concerned with a change which did not become visible until after some generations had elapsed, and which was therefore a change in the germ-plasm.
Many instances of a precisely similar kind have been long known, and one of them is to be found in the history of the garden pansy, which Hoffmann has succeeded in producing from the wild form, Viola tricolor, in the course of eighteen years. Darwin some time ago pointed out in his work upon ‘The Variation of Animals and Plants under Domestication,’ that, in the case of the pansy and all other ‘improved’ garden flowers, the wild form remained unchanged for many generations after its transference to the garden, apparently uninfluenced by the new conditions of life. At length single varieties began to appear, and these were further developed by artificial selection and appropriate crossing, into well-marked races distinguished by peculiar colours, forms, etc.
In these cases also, changes in the germ-plasm are the first results of the new conditions, and there is no evidence for the occurrence of acquired characters, using the term in its restricted sense.
I now come to the last botanical fact brought forward by Hoffmann in support of the transmission of acquired characters. He states that specimens of Solidago virgaurea brought from the Alps of the Valais, commenced flowering in the botanical garden at Giessen, at a time which differed by several weeks from that at which specimens from the surrounding country, planted beside them, began to flower. In other words, the time of flowering must have been fixed by heredity in the alpine Solidago, for the external conditions would have favoured a time which was simultaneous with that of the Giessen plants.
What conclusions can be drawn from these facts? Hoffmann of course sees in them the proof of the transmission of acquired characters, but this presupposes that the time of flowering was originally an acquired character. Hoffmann indeed appears to entertain this opinion when he somewhat vaguely states that the time at which flowering begins has been acquired by accommodation—that is by the influence of climate—during a long series of generations, and has become hereditary. But what does Hoffmann mean by ‘accommodation’? He presumably means that which, since the appearance of Darwin’s writings, has been generally called adaptation:—that is a purposeful arrangement, suited to certain conditions. The majority of biologists have followed Darwin in believing that such adaptations have been produced by processes of natural selection. Hoffmann seems to imagine that they have arisen in some other way: perhaps he believes, with Nägeli, that they have been directly produced by external influences.
The fixation of the time at which flowering begins, is an adaptation which formerly could have been very well explained as the direct result of external conditions. The question we have to decide is whether such an explanation is the true one. We might imagine that the plant would be forced into quicker development by an earlier appearance of the warm season. Hence when transferred into a warmer climate the plant would at first flower rather earlier, the habit would then be transmitted, and would increase in successive generations from the continued influence of climate, until it advanced as far as the organization of the plant permitted. But in this explanation, as in so many others of the same kind, it has unfortunately been forgotten that the transmission of acquired characters which is presupposed in the explanation is a totally unproved hypothesis. It is sufficiently obvious that by interpreting a phenomenon in a manner which presupposes the transmission of acquired characters, we cannot furnish a proof of the existence of such transmission.
It always seemed to me that the fixation of the commencement of flowering, together with similar physiological phenomena in the animal kingdom (for example, the hatching of insects from winter eggs), could be explained very satisfactorily by the operation of natural selection: and even now this explanation appears to me to be the simplest and most natural. In Freiburg, where the vine is largely grown, the harvest is often injured by frosts in spring, which kill the young shoots, buds and flowers. Accordingly, different kinds of vine, which do not push their buds so early, have now been planted. Any one, who has seen all the shoots of the former destroyed by the frosts at the end of April, while the latter, not having opened their buds, were spared, would not doubt that the former must have been long ago exterminated, if they had been compelled to struggle for existence with the others, under natural conditions. Now the time of flowering fluctuates slightly in the individuals of every species of plant, and can therefore be modified by natural selection. It is therefore difficult to see why the time at which each plant flowers should not have been fixed in the most favourable manner for each habitat, by natural selection alone.
Hoffmann is obviously unaware of the fundamental distinction between the characters primarily acquired by the soma, and the secondary characters which follow from changes in the germ-plasm.
If the author had appreciated this distinction he would not have attempted to strengthen his opinions by following up the botanical facts which exclusively belong to the second class of characters, with the enumeration of certain instances selected from the animal kingdom (viz., the supposed transmission of mutilations), all of which belong to the first class. I will not discuss these latter instances, for most of them are old friends, and they are all far too uncertain and inaccurate to have any claim on scientific consideration.
I believe that I have shown that no botanical facts have been hitherto brought forward which prove the transmission of acquired characters (in the restricted sense), and that there are not even any facts which render such transmission probable.
A. W.
Naples, Zoological Station,
Jan. 11, 1888.
