Читать книгу Possible Worlds and Other Essays - J. B. S. Haldane - Страница 7

Table of Contents

‘Darwinism is dead.’

—Mr. H. Belloc.

THERE is a singularly universal agreement among biologists that evolution has occurred; that is to say, that the organisms now living are descended from ancestors from whom they differ very considerably. One or two, including a distinguished Jesuit entomologist, try to narrow down its scope, but so far as I know none deny it. To do so it would be necessary either to affirm that fossils were never alive, but created as such, presumably by the devil as stumbling blocks; or that species were wiped out, and their successors created, on a slightly fantastic scale. For example, the members of one single genus of sea urchins would have to have been destroyed and replaced by barely distinguishable successors some dozens of times during the course of the deposition of the English chalk. This is a reductio ad absurdum of a view which was tenable when only a few groups of extinct organisms belonging to very different epochs were known. But if evolution is admitted as a historical fact it can still be explained in many different ways.

The iguanodon has been replaced by the sheep and cow, the Austrian empire by the succession states. Some few people will attribute both these events to the direct intervention of the Almighty, a few others to the mere interaction of atoms according to the laws of physics and chemistry. Most will adopt some intermediate point of view. We have therefore to ask ourselves whether evolution shows signs of intelligent or even instinctive guidance; and if not, whether it can be explained as the outcome of causes which we can see at work around us, and whose action is fairly intelligible.

Popular ideas of evolution are greatly biased by the fact that so much stress is laid on the ancestry of such animals as men, horses, and birds, which are, according to human standards of value, superior to their ancestors. We are therefore inclined to regard progress as the rule in evolution. Actually it is the exception, and for every case of it there are ten of degeneration. It is impossible to define this latter word accurately, but I shall use it to cover cases where an organ or function has been lost without any obvious corresponding gain, and in particular the assumption of a parasitic or sessile mode of life.

To take an obvious example, the birds were almost certainly derived from a single ancestral species which achieved flight. This achievement was followed by a huge outbreak of variation which has given us the thousands of bird species alive to-day. The essential step was made once, and once only. But the power of flight has been lost on many different occasions, for example by the ostrich and its allies, the kiwi, the dodo, the great auk, the penguin, the weka, the pakapo (a flightless parrot), and so on. Only the auk and penguin converted their wings into flippers and may, perhaps, be absolved from the stigma of degeneracy. Similarly, hundreds of groups have independently taken to parasitism, and in many cases very successfully. On the average, every vertebrate harbours some dozens of parasitic worms, whose remote ancestors were free-living. Blake asked somewhat doubtfully of the tiger,

‘Did he who made the lamb make thee?’

The same question applies with equal force to the tapeworm, and an affirmative answer would clearly postulate a creator whose sense of values would not commend him to the admiration of humanity.

But in spite of this he might be an intelligent being. Now it is perhaps the most striking characteristic of an intelligent being that he learns from his mistakes. On the hypothesis of an intelligent guidance of evolution we should, therefore, expect that when a certain type of animal had proved itself a failure by becoming extinct the experiment of making it would not be tried repeatedly. Yet this has often happened. Both reptiles and mammals have on numerous occasions given rise to giant clumsy types with from one to six short horns on the head. One remembers Triceratops, Dinoceras, Titanotherium, and others. Not only did they all become extinct, but they did not even, like some other extinct animal types, flourish over very long periods. And the rhinoceros, which represents the same scheme among living animals, was rapidly becoming extinct even before the invention of the rifle. But all these animals were evolved independently. Among the titanotheres alone, eleven distinct lines increased in size, developed horns, and perished.

Two or three such attempts would have convinced an intelligent demiurge of the futility of the process. That particular type of mistake is almost the rule in vertebrate evolution. Again and again during Mesozoic times, great groups of reptiles blossomed out into an inordinate increase of bulk, a wild exuberance of scale and spine, which invariably ended in their extinction. They doubtless enjoyed the satisfaction of squashing a number of our own ancestors and those of the existing reptilian groups, who seem to have been relatively small and meek creatures.

It would appear, then, that there is no need to postulate a directive agency at all resembling our own minds, behind evolution. The question now remains whether it can be explained by the so far known laws of nature. In the discussion which follows we do not, of course, raise the questions as to how life originated, if it ever did; or how far the existence of an intelligible world implies the presence behind it of a mind.

Darwin recognized two causes for evolution, namely, the transmission to the descendants of characters acquired by their ancestors during the course of their lives, and selection. He laid more stress on the latter and was the first to point out its great importance as a cause of evolution; but—as might be noted by certain anti-Darwinian writers, were they to read Chapter i. of the Origin of Species—he was far from neglecting the former. Nevertheless, thanks in the main to Weissmann, the majority of biologists to-day doubt whether acquired characters are transmitted to the offspring. A vast amount of work has been done to demonstrate the possible effect on an organ of its use or disuse throughout many generations. To take a recent example, Payne bred Drosophila—a fly which tends to move towards light—in darkness for seventy-five generations. At the end of that time no visible change had occurred in the eyes; and when one thousand such flies were given the opportunity of moving toward a light, no change was found from the normal, either in the proportion which moved within a minute, or in the average rate at which they moved. The majority of the experiments on the inheritance of the effects of use and disuse lead to equally negative results.

Some of the apparently successful experiments can be explained by selection. For example, wheat taken from Scandinavia to Central Europe and brought back again after some years was found to germinate earlier than its ancestors, and the results were attributed to the effects of earlier germination in a warmer climate. But whereas in Scandinavia the earliest germinating shoots would tend to be nipped by frost, in a warmer climate they would get a start over the later and be represented in greater numbers in each successive generation. Hence, if there was any inheritable variation in time of sprouting, selection would occur, and the wheat as a whole would sprout earlier.

Nevertheless a certain number of cases remain which can hardly be explained away in this manner, nor by the transmission of micro-organisms. It must be remembered that, however many experiments fail, it is always possible that the effects of use and disuse may be impressed on a species at a rate not susceptible of experimental verification, yet rapid enough to be of importance in geological time. But the acceptance of this principle, and in particular of the corollary that instinct is in part inherited memory, raises difficulties at least as great as it solves. The most perfect and complex instincts are those of the workers of social insect species, such as bees and termites. Now a worker bee is descended almost if not quite exclusively from queens and drones. None, or extremely few, of her ancestors have been workers. If therefore memory were inherited, the instincts of workers should slowly alter in such a way that their behaviour came to resemble that of sexual forms, and insect societies should be inherently unstable—whereas in fact they appear to date back for at least twenty million years.

The case for natural selection is far stronger. Let us first be clear what is meant by this phrase. Among the offspring of the same parents variations occur. Some of these are due to accident or disease and are not transmitted to the next generation, others are inheritable. For example, a single litter of rabbits often contains both coloured and white members. If the whites are bred together, they produce only white young. The coloured will produce a majority like themselves and a proportion of whites. That is to say, both characters are more or less markedly inherited. If now the animals bearing one inheritable character produce on the whole more offspring which survive to maturity in the next generation, the proportion of the population bearing that character will tend to increase. The phrase ‘survival of the fittest’ is often rather misleading. It is types and not individuals that survive.

Of two female deer, the one which habitually abandons its young on the approach of a beast of prey is likely to outlive one which defends them; but as the latter will leave more offspring, her type survives even if she loses her life. Hence, in so far as courage and maternal instinct are inherited they will tend to survive, even if they often lead to the death of the individual. Of course, the fact that nature favours altruistic conduct in certain cases does not mean that biological and moral values are in general the same. As Huxley pointed out long ago, this is by no means the case, and an attempt to equate moral and biological values is a somewhat crude form of nature worship. But that is not to say that the moralist can neglect biological facts.

The assertion is still sometimes made that no one has ever seen natural selection at work. It is therefore perhaps worth giving in some detail a case recently described by Harrison. About 1800 a large wood in the Cleveland district of Yorkshire containing pine and birch was divided into two by a stretch of heath. In 1885 the pines in one division were replaced by birches, while in the other the birches were almost entirely ousted by pines. In consequence the moth Oporabia autumnata, which inhabits both woods, has been placed in two different environments. In both woods a light and a dark variety occur, but in the pine wood over ninety-six per cent. are now dark, in the birch wood only fifteen per cent. This is not due to the direct effect of the environment, for the dark pine wood race became no lighter after feeding the caterpillars on birch trees in captivity for three generations, nor can the light form be darkened by placing this variety on pines. The reason for the difference was discovered on collecting the wings of moths found lying about in the pine wood, whose owners had been eaten by owls, bats, and night-jars. Although there were more than twenty-five living dark moths to each light one, a majority of the wings found were light coloured. The whiter moths, which show up against the dark pines, are being exterminated, and in a few more years natural selection will have done its work and the pine wood will be inhabited entirely by dark coloured insects. Naturalists are at last beginning to realize the importance of observations of this kind, but they require a combination of field observations with experiment such as is too rarely made.

Now it is clear that natural selection can only act when it finds variations to act on. It cannot create them, and critics have therefore objected that it cannot really be said to create a new species. It would follow from this line of reasoning that a sculptor who hews a statue from a block of marble has not really made the statue. He has merely knocked away some chips of stone which happened to be round it! Natural selection is creative in the same sense as sculpture. It needs living organisms exhibiting inheritable variations as its raw material. It is not responsible for the existence of organisms, but it remains to be shown that without it organisms would display any tendency to evolve.

Of course, if variation is biased in some one direction, a new problem arises. Variation has been adequately studied only during the last twenty years, and it is necessary to digress on the results of this study. Most inheritable variations which have been investigated are transmitted according to Mendel’s laws, except that complete dominance is rather rare. That is to say, they are due to the handing on from parent to offspring of a unit which we call a gene, and which is a material structure, located at a definite point in the nucleus of the cell and dividing at each nuclear division. Characters which appear to vary continuously generally prove on analysis to be due to the interaction of a number of such genes. Now apart from non-inheritable ‘fluctuations’ due to the environment, there are two distinct types of variation. The first and commonest kind is caused by a mere reshuffling of genes. If we mate a black and white rabbit with a blue angora (long-haired) doe, the offspring, if the parents were pure bred, will be black short-haired rabbits; but among their children, if they are mated together, will appear an outburst of variation. Black, blue, black and white, blue and white rabbits will appear, some of each kind having short hair, some long, due to a reshuffling of the genes contributed by the parents. This sort of variation obeys the laws of chance, and selection will only be able to pick out one most favoured combination, say short-haired blue rabbits. Almost all variation in the human race is due to this cause.

But there is another and far rarer kind of variation, known as mutation, which consists in the origin of a new gene. I might breed a million rabbits without getting more than a dozen or so well-marked mutations. But the sort of mutations I should expect would be on more or less familiar lines. I should not be surprised if I got an outbreak of hereditary baldness, or came on a new race of rabbit with pink eyes and a yellow coat, for these types have arisen in mice; but I should be dumbfounded if one of my rabbits developed hereditary horns, and still more so if feathers were to appear! As a matter of fact, there is a marked parallelism between the new genes which have arisen in nearly related species; and this is intelligible because the structure of their nuclei is similar, and the changes likely to occur in them are therefore also similar. New genes appear to arise as the result of accidents—that is to say, causes which are no doubt determined by the laws of physics, but are no more the concern of the biologist than those governing the fall of a chimney-pot, which has been known to alter the shape of a human head, though not in an inheritable manner. Mutations have been provoked in mice and flies by mild injury of the germ plasm with X-rays. The vast majority of mutations are harmful, resulting in an impairment of some structure or function, and are eliminated by natural selection. Others are neutral. In a fly of which some tens of millions have been bred in laboratories, over four hundred mutations have occurred, some of them on many different occasions. Only two have yielded types as healthy as the normal. Advantageous mutations are still rarer—that is why evolution is so slow. But they do occur.

On a Sumatran tobacco plantation a new type of tobacco plant, due to a mutation inherited on Mendelian lines, arose suddenly. It was found that the new variety, though no better off than its ancestors in Sumatra, gave distinctly better crops in a cool climate. If it had arisen in the wild state it would have enabled the tobacco plant to extend its northerly range and form a new subspecies. It must be remembered that a mutation which in most circumstances would be disadvantageous, may be useful in a special environment. Wingless varieties of normally winged insects are common on small oceanic islands, though by no means universal. Mutations causing loss of wings are also common in the laboratory. It is clear that after an island has been colonized by a winged insect carried by the wind from an adjoining continent, hereditary loss of wings, if not accompanied by degeneration of other structures, will be of value in preventing its successors from being blown out to sea.

It is clear, then, that in mutations of this type we have a means by which subspecies may be formed in nature, and there is strong evidence that they have been so formed. For example, the three varieties of the black rat, which have different geographical distributions, differ from one another by single genes quite similar to those which arise by mutation in the laboratory. But there is no evidence at all that mutations are biased in a direction advantageous to the species. The possibilities of mutation do, however, limit the directions in which a species can evolve. Whether it will do so along any of the lines thus laid open to it depends on natural selection. In some cases, as among flowering plants, a good many species seem to be neither better nor worse off than their ancestors—and therefore to owe their origin primarily to variation. However, a slight change in leaf or flower form can hardly be called evolution.

In many cases a change in one character will only be of advantage to a species if some other varies simultaneously in the same direction. This has been used as an argument against natural selection. But in the first place, although one gene may affect one structure only or mainly, others will modify a whole group. Thus of the genes which alter the wing of the fly Drosophila some have little effect elsewhere, some also affect the balancers (rudiments of the second wing pair), others the legs, and so on. A mutation will, therefore, often be found to kill the two birds with one stone, so to speak. Should this be impossible, selection can still work.

Suppose it is to the advantage of an animal that two structures A and B—say bones—should increase together, but that variations in them are inherited independently. We can classify the animals according as the two are of less than the average size, greater than the average or about equal to it. So that we get nine classes in all. Those in which the two are unequally developed will be at a disadvantage, only where both are increased will there be any gain. Putting the number of the normal type surviving at 100, we should get survival rates somewhat as follows:—

	A+	A=	A–
B+	101	98	96
B=	98	100	98
B–	96	98	99

where the figure 101 represents the fact that animals with both A and B increased have a one per cent. better chance of survival than the normal. It will be seen that the A–and B–groups will tend to die out, so that both structures will increase in size.

To my mind the most serious argument against selection on these lines is that it does not explain the origin of interspecific sterility, except where it is due to external causes such as differences of size or breeding time. It is on these grounds that Bateson, a thorough believer in evolution, has criticized natural selection. But I have pointed out elsewhere,^[1] a difference of a single gene between two animals may cause the production of an excess of one sex on crossing, as occurs in fowl-pheasant and cow-bison crosses; and several such genes may well cause complete sterility.

Moreover, there is a second type of inheritable variation, leading to a change in the chromosome number, which causes inter-varietal sterility, often without a very marked change in external characteristics. This is quite common in plants, less so in animals. Although, therefore, the problem of interspecific sterility is serious, we are already well on the way to solving it.

We must now turn to the palaeontological evidence. In a few groups we can trace the course of evolution in some detail. Thus we know over five thousand species of ammonites, and nearly two hundred of extinct horses. In the horses, advance took place along several parallel lines, only one of which has left living descendants. In each line the toes were gradually reduced from three to one, while the molar teeth increased in length and complexity. When in the past we find two different species competing in the same area, one is usually further on the road towards a single toe, the other towards a long molar. We know that these two characters were of value, because we find fossils in which the thin lateral toes—reduced to mere vestiges in the modern horse—had been broken during the animal’s life, as shown by subsequent healing. We also find that in the more primitive types the teeth were often worn down to the roots, leading to death from starvation. Hence for two species to compete equally their advantages in these two respects must be balanced, since species combining both advantages—as does the modern horse—would oust those possessing one only. Evolution in the cases where the evidence is most complete is known to have been very gradual. Such large changes as those produced by most genes so far studied were rare in evolution. This is natural enough. Geneticists have concentrated their attention on genes which produce striking effects. Now, however, that they are beginning to look for those causing very small effects only, and often apparently continuous variation, they are finding them.

A more serious objection is that rudimentary characters sometimes appear which can be of no use to their owners, but only become so on further development some thousands of years later. This is almost certainly true and is at first sight fatal to the selection hypothesis. But it can be met along several lines. A change in one organ, as Darwin pointed out, generally carries with it a change in others. Hence an increase in the complexity of one molar brought about by natural selection may cause the beginning of a new cusp in its neighbour. This cusp will at first be useless, but as it increases selection will begin to act on it also, so that the process will gather momentum until we arrive at the extremely complex grinders of the elephant or horse. Moreover, we can trace just the same gradual beginnings of apparently quite useless organs, the excessive skeletal outgrowths which have been the harbingers of extinction in many animal groups, both vertebrate and invertebrate. If we knew more about these creatures’ soft parts we could perhaps elucidate these problems. Some light is thrown on them by recent work of J. S. Huxley and others. They have shown that, in certain animals, growth of the whole body leads to disproportionate growth of one part. Thus in a group of crabs, whenever the body doubles in weight, the large claw increases three times, until it finally becomes almost as large as the rest of the animal. Any cause promoting growth of the whole body, therefore, leads to a disproportionate growth of the claw. And such a cause is to be found in competition within the species, more especially the competition between males for females by fighting, as is common among mammals, rather than display, as seems to be the custom with many birds.

Still the possibility of some deeper underlying cause of evolution is often suggested by the study of a whole great group, such as the ammonites, which furnish the best available material, for the following reasons: They were sea-beasts, hence their shells were preserved far better than the skeletons of land animals. The number of known fossil species is nearly double that of living mammals. Their shells tell us of their development, for the whorls formed by the young animal are preserved in the middle of the complete structure. Finally, their history is over. The last of them died in Eocene times, about sixty million years ago.

The earliest forms were often not coiled at all and always had very simple patterns on the sutures between different shell chambers; and their descendants still made these simple patterns in the embryonic stages. In the great ages of ammonites during the first two-thirds of the Mesozoic era, the most complex ornamentation was generally made by the adult animal. But as time went on, it showed a tendency to slur its work. The most complex patterns were made by the half-grown creatures, and in Cretaceous times the adult shells were sometimes even uncoiled, as in the very earliest forms. Now this ‘second childhood’ occurred independently in some scores of different lines of descent, always as a prelude to extinction. In other groups the same phenomenon may be observed, though the stigmata of degeneration are different.

This degenerative process is often described as the old age of a race, but we must remember that this phrase is only a metaphor. Some very obvious explanations for it are as follows:—

A step in evolution in any animal group is followed by an evolutionary advance on the part of their parasites. When our fish ancestors came out of the water, they lost their louse-like crustacean parasites; and it was only after some time that insects can have taken their places; and later still that micro-organisms such as those of malaria and typhus were evolved, which pass part of their life-cycle in insects and part in vertebrates. So the apparent degeneration of a group may only mean that evolution of their enemies has caught up with their own. Again, specialization—while it leads to temporary prosperity—exposes a species to extinction or at least to very unfavourable conditions when its environment alters. A small change of climate will lead to a disappearance of forests over a wide area, and with them of most of the animals highly adapted to life in them, such as squirrels, woodpeckers, wood-eating beetles, and so forth. A few, like our own ancestors, adapted themselves to a new environment; but the majority, and all the more highly specialized, died out, the new population of the area being recruited from among the less well adapted forms. Also, as pointed out above, competition within the species, man included, may lead to results desirable for a few individuals, but most undesirable for the species as a whole.

To my mind the closest analogy to the evolution of a given group is the history of the art and literature of a civilization. The clumsy primitive forms are replaced by a great variety of types. Different schools arise and decline more or less rapidly. Finally a period of decline sets in, characterized by archaism like that of the last ammonites. And it is difficult not to compare some of the fantastic animals of the declining periods of a race with the work of Miss Sitwell, or the clumsy but impressive with that of Epstein. The history of an animal group shows no more evidence of planning than does that of a national literature. But both show orderly sequences which are already pretty capable of explanation.

To sum up, no satisfactory cause of evolution other than the action of natural selection on fortuitous variations has ever been put forward. It is by no means clear that natural selection will explain all the facts. But the other suggested causes are unverified hypotheses, while selection can be observed by those who take sufficient trouble. Some of the alleged causes, moreover, are difficult to reconcile with the facts of palaeontology and genetics. The evidence as to the earth’s age from radio-active minerals shows that about six hundred million years have elapsed since the first known fossils were laid down, and perhaps twice as long since life appeared on the earth. This is a larger time than the early supporters of Darwin demanded, and seems long enough to satisfy any quantitative objections as to the slowness of evolution. There are qualitative objections, such as those connected with the origin of consciousness. But consciousness arises anew in every human being. Its first origin on the earth presents no more and no less mystery than its last.

Finally, no facts definitely irreconcilable with Darwinism have been discovered in the sixty years and more that have elapsed since the formulation of Darwin’s views. Such a fact would be, for example, a convergence in the course of geological time of members of two or more groups to form a single species. Actually, we observe the convergence of forms as we go down and not up a geological series. And there have been quite enough anti-Darwinian palaeontologists to have seized on such a case had it existed.

As an explanation of evolution Darwin’s ideas still hold the field to-day, and subsequent work has necessitated less modification of them than of those of his contemporaries in physics and chemistry. Just as physiology has found no case of interference with the order of nature as revealed by physics and chemistry, the study of evolution has brought to light no principle which cannot be observed in the experience of ordinary life and successfully submitted to the analysis of reason.

[1]	Journal of Genetics, vol. 12.