Chapter VIII: Part II: Darwinism and the Atomistic Interpretation of Inheritance (2)

A great deal of confusion can be dispelled if we recognize that Darwin never clearly distinguished between two distinct issues. His herculean labours in the field of geographical distribution urged him to seek a reason for the circumstance that different species of animals exist in different parts of the world. The struggle for existence does explain why some species have died out in one place while others have died out in other places. From that point Darwin went on to generalize about the Origin of Species, i.e. how species come into being. It is unfortunate that, though most of his earlier and enduring contributions to science are concerned with how certain species have ceased to exist, the title of his work laid emphasis on the process by which species are brought into being. It was naturally this part of his theory which made the greatest appeal to his contemporaries. Possibly it was not the one which was most significant to Darwin himself. Darwin used the term Natural Selection in connexion with both problems. With regard to the former his theory is as acceptable as ever. With regard to the latter it has now been superseded by exact experimental enquiry into the mechanism involved in the production and preservation of new hereditable types. The work of Gregor Mendel is the proper starting-point of such enquiry.

A third aspect of the Principle of Genetic Variation concerns the adequacy of geological time. It will only be touched on briefly. When the evolutionary theory was introduced to the biological world, it had to encounter a difficulty that no longer presents itself as a formidable objection. Kelvin had calculated the possible period of time during which life can have existed from considerations derived from the rate of cooling of the earth. The allowance which Kelvin conceded was subject to the qualification that no factors at that time undiscovered enter into the question significantly. Since that time the discovery of radio activity has removed the necessity to place any such restriction on the period of geological time, as Kelvin was led to deduce. To-day we have no reason for believing that geological time is too short to permit us to ascribe the faunistic changes of successive generations to the operations of the natural process of genetical variation. At the same time the Evolution Theory will not stand side by side with the retrospective hypotheses of astronomy in the hierarchy of scientific generalizations, until the frequency of genetical variation and the conditions which determine it have been correlated with more exact knowledge of the duration and climatic features of the intervals corresponding to geological strata.

§3

There remains another aspect of the Principle of Genetic Variation. This is of paramount importance in connexion with the evolutionary hypothesis. It is the Origin of Species _sensu stricto_. A good deal of confusion has arisen in the discussion of the species problem on account of the equivocal usage of the word _species_. It is therefore best to begin with a clear definition of the species problem. It is a universal experience that any dog resembles its father and mother in more respects than it resembles any cat or any fish; any cat resembles its father and mother in more respects than it resembles any dog or any fish; any fish resembles its father and mother more closely than it resembles any cat or any dog. We may express this by saying that cats, dogs and fish have certain specific hereditable properties. If we examine these hereditable properties we find that a cat has more hereditable properties in common with any dog than those which it shares with any fish. Thus organisms can be arranged or classified in groups expressing the extent of resemblance in their hereditable properties. The work of Ray and Linnæus in the early half of the eighteenth century led to the general belief that “like begets like,” and the publication of the _Systema Naturæ_ (1757) by the latter author marks the beginning of a century and a half of detailed anatomical studies directed to classification of this kind. According to the degree of resemblance of organisms with respect to their hereditary properties they are customarily grouped in phyla, classes, orders, families, genera and species. To illustrate the meaning of these terms let us consider the reader of this essay. He or she is said to belong to the species _sapiens_ of the genus _Homo_, which includes all living races of man. The genus _Homo_ includes in addition _H. Neanderthalensis_, the early stone-age heavy-browed first men, and is grouped with the genera _Pithecanthropus_ and _Eoanthropus_ (the fossil ape man of Java and Pilt Down man) in a family _Hominidæ_, within the order _Primates_, that comprises apes, monkeys and marmosets. The order _Primates_ is one of many orders of forms within the class _Mammalia_ that includes hairy animals that suckle their young. The _Mammalia_, along with birds, reptiles, amphibia (frogs, toads salamanders) and fishes, is placed in the phylum _Vertebrata_, which includes all forms with a backbone.

The degree of similarity implied by placing two species in the same genus, order, class, etc., is an arbitrary one defined by convenience and general consent. The degree of similarity implied in placing two individuals in the same species in the sense in which the term was defined by Linnæus implies something more than convenience. Linnæus placed within the same species all individuals which breed readily with one another. The structural difference between two Linnæan species of animals and plants may be negligible compared with the immense structural differences that distinguish varieties within a single Linnæan species, as for instance the difference between White Leghorns, Yokohamas, Silkies, Partridge Cochins, etc., which are all members of the species _Gallus domesticus_.

Though this definition of the species as a unit is the one sanctioned by priority, it is insufficiently emphasized by those who discuss evolution that the creation of new species in the daily routine of a large museum has very little to do with the Linnæan test. Preserved animals are sent by collectors to the taxonomist, who proceeds to classify them in new species, varieties or genera in the vast majority of cases without any experimental knowledge as to their breeding habits. Hence the terms species and variety are in practice used to a large extent interchangeably, though not deliberately. The historic problem of the origin of species is not that of the origin of museum species but of Linnæan species. If we can show that discrete hereditable properties arise, as we know that they do, discontinuously in the normal course of generation, we have all the materials we need to interpret the origin of varieties, genera, orders, families, classes, phyla. Whereas all these are arbitrary groups defined in terms of similarity and difference of the hereditable anatomical properties of animals and plants, the species, as defined by Linnæus, is a group limited not merely by the anatomical resemblance of its individual members but also by _their inability to breed successfully with other forms_.

What has been said so far about the origin of new hereditable properties bears directly upon the way in which new varieties arise. New varieties will only retain their characteristics if some external agency is employed to prevent them from hybridizing and thereby giving rise to an indefinite number of new combinations of characters. The Yokohama can be made to retain those characteristic differences which distinguish it from a White Leghorn by the mechanical device of separating the two strains with a partition of wire netting. No wire netting is required to prevent a White Leghorn and a turkey from losing their genetic individualities, when they are placed in propinquity to one another as are closely allied species in Nature. There is therefore in addition to the problem of the origin of new varieties a problem of the origin of species incompatibility. This cannot be dismissed as of no importance, so long as our experimental knowledge of the origin of varieties fails to suggest in what way this incompatibility may arise. If we can solve this new problem the evolutionary hypothesis presents no ulterior difficulties in the way of explaining the origin of differences which separate the larger systematic groups. The differences employed in distinguishing genera from varieties, and orders from genera or classes from orders are differences of degree. Between species, superficially at least, there seems to be a difference in kind. On this account the origin of species has always been taken to signify the core of the evolutionary problem.

In a somewhat panegyric vein Mr. H. G. Wells replying to Hilaire Belloc makes the following remark: “Darwin’s book upon the subject was called _The Origin of Species_. It was a very modest and sufficient title. He did not even go to the length of calling it the origin of genera or orders or classes.” Surely Darwin might much more appropriately have employed the latter. How types which are structurally different arise may or may not be accounted for by the selection hypothesis. How types which will not breed with one another arise within the same stock is not relevant to it. It is true that Darwin and Wallace vaguely referred in their writings to a natural tendency to infertility as forms become more sharply differentiated. This does not meet the difficulties of the case, even if it is a sound experimental doctrine. Bateson has used the following illustration to emphasize the irrelevance of Natural Selection to the species problem in the strict sense of the term:

“Sometimes specific difference (anatomical differences between
species) is to be seen in a character which we can believe to
be important in the struggle, but at least as often it is some
little detail that we cannot but regard as trivial which suffices
to differentiate the two species. Even when the diagnostic point
is of such a nature that we can imagine it to make a serious
difference in the economy, we are absolutely at a loss to explain
why this feature should be necessary to species A, and unnecessary
to species B, its nearest ally. The house sparrow (_Passer
domesticus_) is in general structure very like the tree sparrow
(_P. Montanus_)... They differ in small point of colour... The two
species therefore, apart from any difference that we can suppose
to be related to respective habits, are characterized by small
fixed distinctions in colour-markings, by a striking difference in
secondary sexual characters and by a difference in variability. In
all these respects we can form no surmise as to any economic reason
why the one species should be differentiated in one way and the
other in another way, and I believe it is mere self-deception which
suggests the hope that with fuller knowledge reasons of this nature
would be discovered.”

It is permissible to argue that the final justification of the evolutionary argument will be achieved when intersterile mutants have been shown to appear under experimental conditions. We shall then be able to state that new types which display not only anatomical but specific discontinuity have arisen in the ordinary course of generation. At present it is only possible to say that we have very good reason to believe they can do so. Bateson overemphasized the difficulty of the species problem when he said “the production of an indubitably sterile hybrid from completely fertile parents, which have arisen under critical observation from a common origin... is the event for which we wait.” Although the origin of the species barrier does introduce a novel issue into the discussion of the evolutionary problem, its novelty is not so fundamental as it appears to be at first sight. Morgan remarks with justice:

“The necessity of putting the mutation theory to the test that
Bateson calls for seems to me very doubtful, for while this is one
of the possible ways in which a mutant might split off at once from
the parent type, it is by no means the only way or even, I think,
the most probable way in which species have become separated....
There is no one problem of infertility of species and no one
problem of the sterility of hybrids, but many problems, each due to
differences that have arisen in the germinal material. One or more
of these differences may affect the mechanism of fertilization or
the process of development, producing some incompatibility.”

Bateson performed a most important task in emphasizing that the problem of species discontinuity exists. He made its solution assume more formidable proportions than the facts merit. There is no mysterious wholeness about the concept of the species barrier. Like other scientific concepts it defines a class of properties. When we examine the characteristics of species barriers, we at once see that they constitute a very heterogeneous assemblage of hereditable properties, many of which are recognizably similar to hereditable properties which we know to arise as mutants in genetic experiments. An individual may be placed in a different species from another individual because of some merely anatomical difference in the structures associated with the copulative act. Owing to the respective absence of neck hackles and tail feathers in two strains known as the Barbadoes and Rumpies, the male of the latter cannot successfully tread the female of the former, though each is interfertile with other breeds of domestic fowls. The origin of such differences does not constitute a problem of a different class from the origin of other varieties. High and low fertility are hereditable properties that can be studied as varieties within the species group. They have arisen as mutant characters in experiment. If there arose within a stock mutants with complementary genes for infertility either type would be infertile with respect to the other. They would constitute separate species in the Linnæan sense, when the parent stock died out. In the case of the donkey and the horse, we can go further and identify the complementary sterility factors in the structure of the chromosomes. Difference of size and shape in the chromosomes of the donkey and horse prevent them from pairing in the reduction division, so that no ripe sperm is formed in the testis of the mule. Mutants differing with respect to chromosome numbers and sizes arising by fragmentation or fusion are known both in plants and animals to have arisen under experimental conditions. In many plants they have been perpetuated by self-fertilization. Plough has raised a mutant strain of Drosophila which is more fertile _inter se_ than with the wild stock. The genetic basis of interspecific sterility, while worthy of much more extensive research, is now reaching a precision which places the experimental data of evolutionary theory beyond the plane of Malthusian speculation.

The foregoing illustrations do not exhaust the variety of biological characteristics which separate one individual from another as a member of a different Linnæan species. Nor do they exhaust the types which can be brought within the realm of experimental treatment. Other cases are discussed at length in Crew’s _Animal Genetics_. The species barrier is not one thing but many things. In the light of modern research there is no reason to regard the origin of species barriers as an essentially different problem from the origin of varieties. Nevertheless the two issues are superficially distinct. No discussion of the present status of the evolutionary hypothesis is complete unless the distinction is submitted to critical examination in the light of experiment.

In the opening years of the twentieth century it had become the fashion among biologists to treat evolution as a dogma. The growth of experimental study of inheritance and variation tends rather to make us value it as a hypothesis suggestive of further enquiry. The difference between the two attitudes is akin to a difference of method which mankind has adopted throughout the ages in the pursuit of knowledge. One method rationalized in its most rigid form in the philosophy of Hegel is to seek for some proposition to which every one is agreed and proceed by deduction to whatever conclusions may be reached from the starting-point. This method has proved invaluable to politicians and members of the legal profession in the discharge of their vocational activities. It is essentially like that of the schoolmen who would exhaust themselves in untiring search into the writings of the ancients for some authoritative statement regarding the number of teeth which the horse possesses, a statement that no one would dare to question. The scientific method is irreconcilably opposed to the Hegelian method. With no aspirations to good breeding it prefers to look the gift horse in the mouth. It is just those propositions which every one accepts that the scientist is most anxious to examine in the hard light of experience. In attempting to envisage a natural mechanism by which the graded differentiation of animal structure could have been brought about, Lamarck was content to employ the generally accepted belief in the inheritance of acquired characters without bringing it to experimental test. Darwin, fortified with newer knowledge of the historical succession of animals and plants as recorded in the rocks, sought to show that evolution was a necessary consequence of competition and the “strong principle of inheritance.” Darwin did not undertake the task of enquiring into the nature of the “strong principle of inheritance.” It was to him like one of Euclid’s axioms. Mendel alone at this time saw the necessity for an experimental study of inheritance, and pointed the way to a non-dialectical treatment of the problem.

VII. NATURAL SELECTION AND EXPERIMENTAL RESEARCH

“Heredity as something quite incomprehensible cannot be used as an
explanation, but only as a designation for the identification of a
problem. And the same holds good of adaptability.”--Nietzsche, _The
Will to Power_

§1

To large numbers of people evolution is Darwinism, just as to our fathers geometry was Euclid. In one of his writings Morgan has remarked that “it is not so important to find out whether Darwin’s ideas were as clear as our own, as to make sure that our own ideas are clear.” This is true; but an interest in the history of scientific thought is a blameless pursuit for its own sake; and there are ulterior reasons which justify an historical discussion of the criticisms which experimental discovery has brought to bear on the Selection doctrine in its original form. During the latter half of the nineteenth century the evolutionary hypothesis became entangled with the idea of a moral progress of mankind. On this account some philosophers, who are not biologists themselves, fail to recognize the ethical neutrality of biological enquiry. It is doubtful whether the promulgation of any scientific hypothesis has ever had so profound and, at the same time, so immediate an effect on the attitude of educated people towards personal responsibility and social obligations. The fate of Darwinism is as much the concern of the layman as of academic biologists.

Nor is it easy for those who are not biologists to gain definite enlightenment concerning the extent of the change that has taken place. With the rise of experimental method the discussion of evolution has become more technical owing to the accumulation of new data and on account of the introduction of a more intricate form of logic. It is a quantitative branch of science. There was a time when the biologist thought it worth his while to read and to reply to Samuel Butler. To-day there are biologists who read--and like the present writer enjoy--the works of Mr. Bernard Shaw. They do not feel it necessary to defend their philosophy against the arguments advanced in the preface to _Back to Methuselah_. Popular expositions of evolution are still written. More often than not one suspects that they are rather too popular to answer the questions which an intelligent reader who is not a biologist is most anxious to hear discussed.

As an exact science biology is still very young. Evolution is in its infancy. Only in our generation has it become the nucleus of a growing body of experimental research. It may be that when the history of the evolutionary hypothesis is written two centuries hence, Bateson’s _Materials for the Study of Variation_ will assume a more prominent place than _The Origin of Species_. It may be that the name of Thomas Hunt Morgan will be mentioned in its pages more often than that of Charles Darwin. We are too near the footlights to view the matter in its correct historical perspective. It is at least permissible to entertain such a possibility. Ancestor worship has no place in the ritual of science. If any display of sentiment is appropriate in scientific discussion, it might be said that the only fit way in which to honour the memory of a Darwin and a Newton is to press forward in exploring the fields which their labours have fertilized.

Without entering into technicalities I shall make the attempt in this essay to contrast the use of the term Natural Selection in Morgan’s writings with the Darwinian doctrine in its original form. My aim will be neither to justify in the one case nor exculpate in the other, but to discover whether a difference exists, wherein the difference lies, and how the difference has arisen. In contrasting the views held by two men of science it is of the utmost importance to lay emphasis on the type of data which they have respectively studied most. Morgan is an experimental geneticist. Darwin was pre-eminently a geographical naturalist. Morgan’s most brilliant contributions to the advance of science have been focused on the study of those conditions which are significant to the origin and transmission of new hereditable properties in animals. Before the publication of _The Origin of Species_ Darwin’s scientific labours had concentrated more especially on amassing a wealth of information about the way in which species are distributed in different parts of the world. In his long itineraries, it is not difficult to surmise what aspect of the species problem was constantly uppermost in Darwin’s thought. I think it is necessary to appreciate this bias in any attempt to understand the way in which the Selection hypothesis developed. Though Darwin spoke of the Origin of Species, he was interested primarily in why some species happen to be found in one place and other species in different places. Darwin had two distinct problems in view when he set out to write _The Origin of Species_. In the course of writing it he sometimes lost sight of the distinction between them. One was how different types of animals have come to persist in different parts of the world. The other was how an evolutionary process could take place at all. That the struggle for existence is the key to the former is highly plausible. No facts are known which contradict such a view. It is not really an issue with which the modern experimentalist concerns himself. Up to this point there is no divergence between the Darwinian and the Mendelian standpoint. But Darwin in very unequivocal language committed himself to the view that in building up new specific forms the struggle for existence makes use of all differences between parent and offspring of “whatsoever” origin. He thus implicitly encouraged the view that natural selection is a creative agency. Herein lies a fundamental difference between the standpoint adopted by Morgan and the Darwinian doctrine. Darwin really believed in the Origin of Species by natural selection. Morgan believes in the Origin of Gaps by natural selection.

It is perfectly true that Darwin did not formulate this deduction so explicitly or so prominently as did some of his followers. But it was logically implicit in his earlier writings and very definitely set forth in his later. It was in virtue of this aspect of the Natural Selection hypothesis that evolution captured the support of Darwin’s contemporaries. Till Darwin’s book appeared, biologists did not for the most part believe that evolution could take place. Darwin’s hypothesis demonstrated that evolution must take place in a world in which organisms had to struggle for their existence. The experimental data which Morgan employs as the basis for his conception of the evolutionary process imply that the reasons which led the pre-Darwinian biologist to think that evolution could not take place are unfounded. They also imply that the reasons which Darwin advanced to show that evolution _must_ take place are wrong.

It is easier to make this distinction clear at a later stage with the aid of a concrete example than by stating general propositions. This is because one result of experimental progress has been a change in our use of the concept of “variation.” Darwin used the term variation for any difference between parent and offspring. In affirming that the struggle for existence makes use of all variations for building up species differences, he logically implied one of two things. Either all differences between parents and offspring are genetic in origin, that is to say, referable to differences in the egg or sperm; or alternatively bodily modifications which occur during the lifetime of an individual influence the genetic structure of the offspring so as to produce an analogous result. This principle, usually associated with the name of Lamarck, was accepted by every one in Darwin’s time. Darwin himself, while ridiculing Lamarck’s idea of the _modus operandi_ of evolution, accepted the inheritance of acquired characters. There was therefore no need for him to make a distinction between the two alternatives. Neither the one nor the other is in harmony with the standpoint of a modern geneticist of Morgan’s school; but the difference between the Darwinian standpoint and that of Morgan concerns not only the question of fact but the deductions drawn from it.

The difference between either of these alternatives on the one hand and the Mendelian standpoint on the other can be illustrated by reference to one of Mendel’s original experiments on the hybridization of peas. In crossing pure-bred peas of the variety characterized by a dwarf shoot with the normal tall variety, Mendel obtained only tall types on the first generation, and in the second generation derived from crossing the latter _inter se_ one-quarter were dwarf and the remaining three-quarters tall. Now the individuals of either the tall or the dwarf class are not all alike. Any dwarf shoot grown under ordinary conditions is distinctly smaller than a tall shoot, so that the two classes are discontinuous and quite easily distinguishable; but when the conditions are standardized as much as possible small differences of light, moisture, soil-content, temperature or proximity exert their influence, so that no two dwarf plants are of exactly the same size. What is transmitted through the gametes is something which determines the extent to which an individual is capable of growing under appropriate conditions. This distinction greatly clarifies our thought about the so-called inheritance of acquired characters.

A criticism of the Lamarckian doctrine is irrelevant at this juncture. It is referred to in this connexion because it was only in the eighties, after the Lamarckian view was challenged by Weismann, that the full force of the logical implications of Darwin’s teaching was felt. It is true that his followers were far more definite than the author of _The Origin of Species_ in emphasizing the creative rôle of selection. It is true that the discredit into which the Lamarckian principle fell after the discovery of the nature of fertilization led the Selectionist writers to exaggerate this aspect of Darwin’s hypothesis. Nevertheless Darwin did express himself in unmistakable language with regard to this issue. His followers, forced to be more specific concerning the nature of differences between parents and offspring, made the bold, and, it transpired, unwarranted assumption that all those small differences between parent and offspring now referred to as fluctuating variability are in the main genetic in origin. The Selectionist doctrine thus assumed that hard outline which produced its first vigorous reaction in Bateson’s _Materials for the Study of Variation_ (1894), a work which laid down the main lines of investigation which have been elucidated by the Mendelian renaissance.

To avoid vagueness concerning what Darwin actually did say I shall quote once more from _The Origin of Species_:

“Any being, if it vary in any manner profitable to itself, under
the complex and sometimes varying conditions of life, will have a
better chance of surviving, and thus be naturally selected. From
the strong principle of inheritance, any selected variety will tend
to propagate its new and modified form.” (Introduction.)

“Each of the endless variations which we see in the plumage of
fowls must have had some efficient cause; and if the same cause
were to act uniformly during a long series of generations on many
individuals, all probably would be modified in the same manner.”
(Chap. 1.)

“A high degree of variability is obviously favourable as giving
the materials for selection to work upon, not that mere individual
differences are not amply sufficient, with extreme care, to allow
of the accumulation of a large amount of modification in almost any
desired direction.” (Chap. 1.)

“Over all these _causes_ of change, the accumulative action
of selection, whether applied methodically and quickly, or
unconsciously and slowly but more efficiently, seems to have been
the predominant power.” (Chap. 1.)

“Variations, _however slight, and from whatever cause_ proceeding,
if they be in any degree profitable to the individuals of a
species, in their infinitely complex relations to the individuals
of a species... will tend to the preservation of such individuals
and will generally be inherited by the offspring. The offspring
also will have a better chance of surviving, for of the many
individuals of a species which are periodically born, but a small
number can survive. _I have called this principle, by which each
slight variation if useful is preserved, by the term Natural
Selection._” (Chap. 3.) (Italics inserted.)

If, as Darwin believed, it were true, that variation occurs in every generation, the evolutionary process would be a continuous one. To Morgan the production of mutants is a discontinuous break in a normal routine of stability. To Darwin variation and heredity were co-extensive terms. The offspring are always on the whole like their parents. That resemblance constitutes inheritance. On the other hand they are never quite the same. The difference was what Darwin called variation. To Morgan heredity and variation are not co-extensive terms. The structure of the chromosomes is fundamentally stable. From time to time there occur disturbances of this normally stable equilibrium. New hereditable properties emerge into being in a quite discontinuous fashion. There is no self-evident reason why a particular stock should not remain indefinitely in a phase of stability. To the experimental geneticist there thus exists no difficulty in interpreting the fact that some animals have remained unchanged since the earliest rocks.

To the generation in which Darwin lived there seemed to be only one logical outcome of the view that variation is a continuous process involving all the individuals of every generation. This deduction was never stated very explicitly by Darwin himself, though it was definitely asserted by Wallace. There can be no doubt that this deduction gave the Selection hypothesis such a strong appeal to Darwin’s contemporaries, and contributed largely to the success of the hypothesis of Natural Selection. Before Mendel, investigators in hybridization had treated the individual as the unit for study. From this arose the belief that hybrids are intermediate between the parents. This belief in its turn gave rise to the notion that on crossing a new type back to the parent stock there would be a dilution of the new character, culminating after a number of generations in swamping it out of existence altogether. Evolutionists of the Darwinian period therefore introduced a variety of devices, such as geographical isolation and, above all, the survival of the fittest, to counteract the effect of this swamping and account for the persistence of new types. To Darwin’s generation it seemed that without selection there could be no evolution. The new type would always be swamped out in the long run. In the struggle for existence the less viable variations would tend to be eliminated, and since there would always be less of them on that account, the swamping process would favour the gradual moulding of the species in the direction of more favourable variation. On this view the struggle for existence is the agency which makes species change. Evolution becomes a necessity.

From Morgan’s standpoint evolution is only a necessity in so far as it happens that mutants do from time to time appear. The struggle for existence though eliminating the less viable types has no creative rôle in the Darwinian sense. Mendelian analysis shows that though the first generation of a cross between pure-bred parents may be intermediate between the parental types, both parental types appear in their original purity in the next generation, and will continue to breed true to type, whenever they mate with other individuals similarly constituted. The modern geneticist feels no necessity for an _argumentum ad hominem_ to explain how evolution can occur in spite of a supposed swamping process. To him the swamping process is an illusion based on imperfect knowledge of the facts of hereditary transmission. The importance of this difference in standpoint lies in the fact that the idea of natural selection would never have assumed so powerful an influence over biological thought, unless it had provided the evolutionist with train of reasoning which seemed to prove that evolution must be going on all the time.

This interpretation of the Darwinian standpoint is not a caricature drawn by the pen of an adverse critic. An enthusiastic contemporary exponent of Natural Selection, Mr. H. G. Wells, thus defines the selection theory in his _Outline of History_:

“the young which a living thing produces... are like the parent
living thing. But they are _never exactly like it_ or like each
other.... Suppose, for example, there is some little furry
whitey-brown animal living in a bitterly cold land which is usually
under snow. Such individuals as have the thickest, whitest fur will
be least hurt by the cold, less seen by their enemies and less
conspicuous as they seek their prey. The fur of this species will
thicken and its whiteness increase _with every generation_, until
there is no advantage in carrying any more fur.” (Italics inserted.)

Having cited the above, it is somewhat surprising to note that in replying to Mr. Belloc’s strictures, Mr. Wells makes the following statement with reference to the Natural Selection theory:

“Among questions bearing upon it but not directly attacking it
has been the discussion of the individual difference.... What
rôle is played by what one might call normal relatively slight
differences and what by the sports. Can differences establish
themselves while outer necessity remains natural? Can variations
amounting to specific differences... be tolerated rather than
selected by Nature?... What happens to differences in cases of
hybridization?... None of these subsidiary questions affect the
stability of this main generalization of biology.”

In explaining the Natural Selection theory, as quoted above, Mr. Wells himself states or implies every one of these “subsidiary” questions, and answers them in his own way.

Let us now see how a modern geneticist would interpret the evolutionary process by taking an analogous concrete example. He would argue somewhat as follows. Supposing a single white mutant hare arises in a grey parent stock, the behaviour of the chromosomes leads us to infer that eventually other white hares, pure for the white gene or genes, will reappear. These mated _inter se_ will breed true to type. On the assumption (not conclusively proved) that it is advantageous for a hare in temperate climates to be grey and in arctic regions to be white, there will be more white hares in the long run in northern countries and more grey ones in temperate countries. If there were no competitive struggle at all, there would in the long run be grey and white hares in northern and grey and white hares in temperate countries. There would have been the same amount of evolution. The only difference that the struggle for existence introduces is that the final picture presents a more discontinuous aspect. This was not at all what Darwin meant by Natural Selection. He would have said that a single mutant would be swamped out of existence by intercrossing. He would have formulated the problem in the following terms. Of all hares born to grey parents some are lighter and others darker. In a region where it is advantageous, the half that are lighter than the mean will have more chance of surviving to maturity. In any given generation there will therefore be more lighter than darker parents. The result of this will be that in every generation the swamping process will always be on the side of the lighter individuals. Darwin postulated that, if this process went on long enough, a white hare would eventually be produced. Such a race would only be produced in the region where natural selection favoured its survival. On this view natural selection is the creative agency, or at least a paramount creative agency in the evolutionary process. Without the struggle for existence hares everywhere would remain grey. In every generation the half that are lighter than their parents would always be swamped by the half that are darker.

To Darwin and more especially to Darwin’s followers selection was the agency which preserved not merely new individuals but new characters, since characters would otherwise be diluted out of existence. For Morgan the preservation of new characters ultimately resides in Mendel’s law of segregation. It has its material basis in the behaviour of the chromosomes. The contrast between the alternatives is at once made clear when we consider what would happen in a universe so large and so abundantly supplied with the necessities of life that no struggle for existence intervenes. Given unlimited time in a Mendelian universe in which natural selection did not operate, all the species we know to-day would be present, and many more besides. Evolution would have occurred; but the pageant of life would present to the taxonomist a more continuous appearance, and the striking gaps which we now see would be filled not by fossil relics but by living forms. Except in so far as he was prepared to invoke the Lamarckian principle to circumvent difficulties inherent in his own hypothesis, natural selection was to Darwin the necessary condition not merely for gaps but for any evolution to take place at all. In a Darwinian universe without natural selection there would be no progressive differentiation of new characters.

§2

When, out of deference to Darwin’s contribution to biological thought, the experimentalist of Morgan’s school asserts his belief in Natural Selection, he is in fact referring to something very different from Darwin’s Natural Selection, indeed to a view of the process which Darwin would have rejected emphatically. Of course it is admitted that all scientific hypotheses become modified as new data accumulate; and phrases imperceptibly change their meaning in the course of time. But the natural selection of Morgan’s school is not a continuous development from the original concept. Within two decades of the publication of _The Origin of Species_ the selection hypothesis had assumed a clarity of outline which had an influence on subsequent developments in biological thought, persisting till the present day, and not likely to disappear for some time. In 1881 Weismann challenged the prevailing belief in the inheritance of acquired characters. Thenceforth in the hands of the Selectionists environment became merely an agency by which the hereditary materials are preserved or rejected. As an aspect of the problem of development it faded into the background of the picture. To question the almightiness of heredity became equivalent to defending the Lamarckian principle, though the two issues are logically independent.

Educated people frequently use the words environment and heredity in a very different sense from that in which they are employed by the biologist. Unless we are accustomed to the study of embryonic and larval life, we are apt to think of an organism as a finished product. The rôle of environment and of heredity as seen through the eyes of a contemporary biologist can be made explicit by reference to recent work on the metamorphosis of tadpoles. We know to-day that the thyroid gland of all vertebrates contains a high percentage of iodine. Barger and Harrington have now prepared in pure crystalline form an iodine compound which has the same therapeutic properties as extracts of the thyroid gland. A few years ago the discovery that frog tadpoles will change very rapidly into adults if fed with thyroid gland, was followed up by the development of a successful technique for removing the rudiment of the thyroid gland in frog embryos. Thyroidless tadpoles never undergo metamorphosis. They continue to grow as tadpoles when the normal tadpole would change into a frog. The change into the adult in the normal tadpole is initiated by the liberation of the thyroid secretion into the circulation. It has also been shown that tadpoles reared on an iodine-free diet in water containing no trace of iodine remain permanently in the larval state. This clarifies what is meant by an _environmental_ factor in development. In contradistinction to the influence of environment the influence of inheritance in development may be illustrated by reference to an American salamander, _Amblystoma tigrinum_, which has a characteristic larval form. In the lakes around Mexico city there is a local race of this species which never undergoes metamorphosis in nature, reproducing in the larval form. It can be made to develop into the land-dwelling adult in a few weeks, if fed with thyroid gland in the laboratory. Addition of iodine salts to the water in which it lives or to its food will not induce metamorphosis. Its permanent fixation in the larval stage is due to the fact that it _inherits_ from one generation to another a deficient thyroid gland, which cannot make use of the iodine in its surroundings. Absence of iodine in minute quantities from the water, a purely environmental agency, or on the other hand a hereditary difference between two races with respect to the efficiency of thyroid secretion, may either of them be _independently_ instrumental in deciding whether a particular individual shall attain sexual maturity in the form of an air-breathing land-dwelling salamander, or an aquatic half-way house between a salamander and a fish. A geological epoch, if you like to put it in that way, is thus summed up in a mutant gene or in a trace of iodine.

In the attempt to understand the tenacity with which belief in the Lamarckian view persisted in biological thought, it must be borne in mind that embryology is the most recently developed branch of anatomical science. Until the classical researches of von Baer and Meckel were published in the first half of the nineteenth century, the prevailing idea about development was the teleological doctrine that an animal is from the very first complete in all its parts and only needs growth to make its minute structure manifest to the eye. Caspar Wolff in 1759 made observations on the hen’s egg, and was led to state the “epigenetic” as opposed to the prevailing “evolutionary” view. He sought to show that the hen’s egg is at the beginning without any gross anatomical organization and that structural organization within the egg is a gradual development. His work failed to attract attention. Von Baer’s researches on the same subject were published synchronously with the formulation of the Cell doctrine (1832). One might say that until the middle of the nineteenth century, the current conception of inheritance in biology was closely analogous to the legal notion. The parent was supposed to hand on its anatomy to its offspring in the same sense as the well-to-do hand on their belongings. With so erroneous a conception of the nature of development prevailing, it is little wonder that the idea of the inheritance of acquired characters seemed a perfectly reasonable one. It is not surprising that the doctrine of Lamarck should have been first challenged during the decade in which the nature of fertilization and the process of maturation of the germ cells were elucidated.

As stated by its author the Lamarckian principle implied that any reaction of the organism to its environment is carried over to subsequent generations. It was especially _adaptive_ reactions such as the effect of use and disuse which Lamarck emphasized in his evolutionary speculations. When the Lamarckian principle was first challenged, prominent scientists like Cope were willing to assert such fables as the story that a cock deprived of one eye transmitted eye defects to all his offspring. When it was conclusively proved that mutilations effected through several generations left no impress on the hereditable characters of the stock, the Lamarckians fell back on the gratuitous postulate that only “adaptive” changes could be transmitted. The precise meaning of this adjective was never defined, nor was any reason forthcoming to suggest the existence of a mechanism that could discriminate between mutilations and bodily changes that are “adaptive.” This is yet another example of the perils of introducing teleological preoccupations into the construction of biological hypotheses. If recent experimental research conserves any element of truth in the Lamarckian idea, it has robbed it of any special significance to the way in which adaptive structures originate.

Structural changes may arise in the course of development from two conceivable sources. The chromosomes which represent the hereditary materials may find themselves reacting to a different type of “internal environment.” The majority of modifications in the normal course of development undoubtedly come within this category. Modifications of this type, including in all probability relative sizes of organs, all mutilations and habits are clearly not hereditable. Belief in their hereditability was only possible so long as biology was dominated by teleology and the essential features of the reproductive cycle were undiscovered. There is another possibility which was entirely disregarded by Weismann in his Theory of the Germ Plasm. It is a possibility that has no bearing on the problem of adaptation. If environmental agencies can produce mutations by a structural change in the chromosome itself, there is no reason why such structural changes should be confined to the chromosomes of the germ cells. We must therefore preserve an open mind with regard to the possibility of encountering phenomena having a superficial similarity to what is implied in Lamarck’s doctrine. The exposure of young larvæ of the fruit-fly to X-rays has led to the production of individuals which show bodily resemblances to forms which have arisen in the ordinary course of events as mutants. The effect of X-rays may be to change the environment in which the chromosomes operate. But the recent investigations of Patterson indicate the likelihood that the modification is due to the action of the X-rays on the chromosome itself. We know that X-rays will produce mutant changes in the chromosomes of the germ cells. If Patterson’s interpretation is correct, it may well be found that X-rays can simultaneously effect mutant changes in all the chromosomes of the body. If applied sufficiently early in the course of development, radiation with X-rays would then produce bodily changes of a transmissible nature. This possibility resides in the fact that the agent is capable of acting on all the cells of the body in the same way at the same time. There is no inherent unlikelihood that temperature and the chemical constituents of an animal’s food may simultaneously produce bodily and germinal mutations. Strictly speaking this is not the same as the traditional belief in the “inheritance of acquired characters.” The Lamarckian principle completely disregards the distinction between modifications which arise from a change in the internal environment of the chromosomes and a physical change in the chromosomes themselves. It takes no account of the possibility that the environmental agent can act in the same way simultaneously on all the cells of the body.

There are still students of fossil forms who claim that the traditional Lamarckian view is necessary to explain the historic succession of animals by continuous generation. There seems to be no satisfactory reason to justify the statement that evolution can only be satisfactorily explained by assuming the inheritance of acquired characters. If there were, it would not be an argument in favour of the Lamarckian principle. It would be as an argument against the evolution theory. It would imply that the truth of evolution depends on assuming a mechanism whose existence is most unlikely. What is often called the neo-Lamarckian standpoint, the view that acquired characters only gradually become impressed on the hereditary constitution after countless generations, transfers the issue from the plane of verifiable experience to one of pure surmise, rendering further discussion profitless. In such a matter as this when experiment is silent, the student of fossils must also be silent.

The objection rests in fact on a misapprehension. The earlier phase of experimental enquiry along the lines laid down by Mendel was confined to the analysis of simple clear-cut hereditary differences which present themselves in almost any environment in which the animal can live. They were also largely concerned with differences that could be resolved into the simplest arithmetical ratios, or as Morgan would say with mutants that have arisen through a change at a single point on one pair of chromosomes. It is only as technique has progressed that it has been possible to analyse the more complex cases in which single characteristics depend on numerous Mendelian factors, or where the character differences are so variable that they can only be defined in statistical terms. The palæontologist being occupied very largely with size differences is sometimes disappointed, because such phenomena lie outside the scope of the simpler problems, which were once thought to define the scope of the Mendelian hypothesis. Recent progress which has led to the recognition that Mendel’s principle of segregation underlies the inheritance of size is therefore of no little significance to evolutionary theory. As we come to recognize the dependence of hereditary transmission on discrete particles which maintain their entities uncontaminated through all the cell divisions of the body, segregating in their entirety in the formation of the gametes, the unlikelihood of the Lamarckian principle in its traditional form becomes more and more evident.

If the Lamarckian principle in its traditional form was undoubtedly based on a confusion of ideas and an ignorance of fact, the Theory of the Germ Plasm put forward by Weismann shows how facts may be distorted to fit in with preconceived ideas which are in themselves logically flawless. The discredit into which the Lamarckian principle fell, almost as soon as the elementary facts about the nature of fertilization became known, led Darwin’s successors to assume that all those differences between parent and offspring which Darwin had referred to under the term variations are genetic in origin. The assumption was gratuitous, as later experimental analysis has shown. Without that assumption the Selection doctrine would have been robbed of the immense importance it had already begun to assume. From a complete misapprehension of the true rôle of the environment in relation to inheritance, the biological pendulum swung in the opposite direction to a complete disregard of the influence of the environment in relation to development. It is from Weismann’s writings that we can best appreciate the fundamental dissimilarity of Darwin’s Natural Selection and Morgan’s views. For Weismann’s “germinal selection” is the logical outcome of Darwin’s selectionism, once it had been purged of the Lamarckian principle. It is a triumph of Hegelian reasoning applied to biology. There is nothing wrong with it but its premises. Weismann’s theory embodied an atomistic conception of heredity. Unlike Mendel’s it had no connexion with experimental data. Weismann identified his hereditary determinants with the substance of the chromosomes. Unlike Morgan’s hypothesis, Weismann’s speculations were based on incorrect observations about the way in which the chromosomes behave. In the long run the influence of Weismann’s teaching has probably been more sterilizing than the Lamarckian doctrine which he challenged.

Weismann imagined that his atoms of heredity or “determinants” multiply in the cell and in some rather abstract way compete with one another for survival. Hence the hereditary constitution of the individual is never quite the same in two successive generations. Heredity and variation are thus co-extensive, as Darwin’s Natural Selection postulates. Weismann also thought wrongly, it transpired, that the reduction division of the germ cells takes place in such a way that each cell receives half a maternal and half a paternal chromosome of each pair and not, as we now know, a whole paternal or a whole maternal element. Hence he argued that the formation of the germ cells involves not, as Mendel proved by experiment, a segregation but a closer intermingling of the germinal materials. From this the swamping of new characters on crossing became an absolute necessity. To Weismann selection alone could prevent this swamping. Selection must act in every generation, because the mingling of the hereditary materials becomes more intimate with every generation. Only under the influence of continuous selection could any change be brought about. Without it universal stagnation would exist. In short Selection was the creator and the preserver of the benefits of variation. In all this Weismann, with the support of Wallace, went much further than Darwin himself. But the Selectionist doctrine in its main features was implicit in the Origin of Species. The sociological exploits of biologists belong especially to the period in which the Selection doctrine assumed this doctrinaire aspect. Doctrinaire Selectionism has persisted in our own generation in the writings of many eugenists.

§3