PART I VITALISM AND MECHANISM SUMMARY (continued 2)

This I think is the main bone of contention between the two attitudes which are generically denoted by the terms mechanistic and vitalistic. The real issue has shifted from deciding whether the hypotheses of physics and chemistry suffice for the interpretation of vital phenomena to deciding whether there is an essential difference between the logical structure of those branches of science that deal with living matter and those which deal with inanimate objects. This is a welcome change, because it presents a much more genuine and concrete problem for solution. It is somewhat surprising that the controversy should undergo such a metamorphosis at the present moment. The recent development of evolutionary biology is especially calculated to reinforce the belief that biological theory only progresses, when the biologist adopts towards the subject matter of his investigations the same attitude as that which the chemist and physicist adopt towards the objects which they study. In our generation it is possible to find in those aspects of biology which are most recalcitrant to the application of physico-chemical hypotheses the most conspicuous examples of a fundamental similarity in the logical procedure which the biologist on the one hand and the physicist or chemist on the other employ in constructing their hypotheses. It would not be possible to select from the whole field of biological science a more striking illustration of the success of quantitative and experimental methods than the recent extension of Mendel’s hypothesis by Morgan’s school. This advance has entailed an extensive elimination of teleological concepts in the interpretation of the evolutionary process. Yet the phenomena of heredity and variation at present lie completely outside the scope of physico-chemical analysis in the ordinary sense of the term; and any attempt to formulate the problems of genetics in physico-chemical terms is still a matter of pure conjecture.

In this sense we may agree with one writer of the vitalistic school in saying that to speak of the “mechanism of heredity” is a meaningless collocation of words. But if our interest is primarily directed not to the end product itself but towards the way in which the scientist proceeds to elaborate his hypotheses, the study of heredity provides a particularly clear example of how a hypothesis developed without any departure from the principle of mechanism can yield verifiable conclusions about the behaviour of living systems. From this point of view it is both legitimate and intelligible to speak of the mechanisms of heredity and variation; and the expression is as permissible as the analogous phrase, the mechanism of chemical reaction. A comparison of the growth of the Mendelian principle with Dalton’s atomic theory of the structure of matter will help us to see whether the biologist does actually interpret his observations in a manner different from that adopted by the student of non-living matter, and whether the biologist has recourse to a kind of logic which is different from the logic which the physicist and chemist employ in framing their own generalizations.

When Mendel took up the problem of hybridization, the nature of fertilization in plants was known in a general way. Just a century before Mendel began his work Kolreuter by painting pollen from one individual on to the stigmas of another variety, and vice versa, had shown that hybrids inherit equally from the pollen and seed plant. At the end of the eighteenth century and the beginning of the nineteenth, Knight and Goss in England had made further progress in crossing pure bred varieties by calling attention to the “splitting” of hybrids, or reappearance of parental types when intercrossing hybrid offspring. Contemporaneously with Mendel, Naudin in France studied this phenomenon more closely, and came very near to formulating Mendel’s principle. His results were published in 1862. These pioneers in hybridization laid down the necessity of working with what to the geneticist is like pure chemicals to the chemist, pure breeding stock. They fell short of arriving at far-reaching results, because their attitude to heredity was dominated by the holistic standpoint. They could only think of the plant in terms of a preconceived notion of individuality. They refrained from focusing their attention on the separate parts, and following out the fate of discrete characteristics in their crosses.

We must not overlook the debt which Mendel owed to the pioneers of hybridization. There would have been no modern chemistry if the Arabs and alchemists had not devoted years of laborious study to the clarification of our idea of a pure substance; and there would have been no genetics, if the idea of pure breeding stock had not been laid down by Mendel’s predecessors. Chemistry failed to progress beyond the stage of describing new compounds so long as it remained entangled in the vitalistic “phlogiston” concept; and genetics, the study of heredity and variation, remained purely descriptive, until it was emancipated by Mendel from the holistic tendency to concentrate upon the organism as a whole. Naudin did in fact envisage less definitely than Mendel the atomistic concept of heredity, just as William Higgins had partly visualized the chemical possibilities of atoms before Dalton published his theory.

Mendel used in his researches pure breeding stocks differing only in well-defined particulars, employing single characteristics as units of study, and recording the progeny of every cross separately for comparative observation. In his original work Mendel chiefly dealt with the common pea, which possesses two advantages which recommend it for such experimentation, namely, that its flowers are capable of self fertilization (i.e., the pistil can be pollinated from the stamens of the same flower) and that it has a number of well-marked varieties distinguished by tangible characteristics such as the shape (round or wrinkled) and colour (green or yellow) of the seeds, or the stature (tall or dwarf) of the shoot, etc. In all his crosses involving a single difference of this kind he found that the first generation of the cross resembled one of the parents. When these crossbreeds were self fertilized, they produced offspring resembling the original parents in the constant ratio of three to one. One-quarter of the offspring of the crossbreeds resembled one parent and bred true; one-quarter resembled the other parent and bred true; and the remaining half being like the “dominant” parent, which the first generation of hybrids resembled, behaved exactly like the latter, when self fertilized.

An investigator who had not the attitude which makes a capable chemist might have been distracted by the peculiar circumstance of dominance, or the resemblance of the impure individuals to one of the parents exclusively. Mendel rightly judged this to be insignificant. A chemical analogy will perhaps assist to make this clear. Sodium and potassium yield colourless salts with most common acids, but the permanganates of both are purple in solution. The salts of copper are generally of a bluish or greenish tint in solution. In the one case the anion, in the other case the kation, is the dominant factor in determining the physical property of colour; but in both cases the other component behaves in any reaction with no less characteristic efficacy, because its presence is seemingly masked. So likewise Mendel looked beyond the bodily resemblance of the dominant parental and hybrid individuals to their hereditary makeup; and recognized in his experimental data two general conclusions which prompted special consideration. One was the fact that the original parental types can be recovered in all their purity. The other was the fact that the various hereditary types produced by hybridization regularly appear in the same numerical ratios. Both conclusions are of universal validity, though Mendel had the very good fortune to select materials which yield the simplest type of numerical results which occur in crosses between pure strains. When Dalton formulated the atomic hypothesis two fundamental empirical generalisations of chemistry were fully accredited. The law of the conservation of matter and the law of constant proportions had been established. Mendel found in his data the proof of what we might call the principle of the conservation of genetic materials and the law of constant genetic proportions. To the recognition of these empirical generalizations he added a conceptualization of the basis of their existence in terms of discrete factors. These factors were according to Mendel’s hypothesis (or Mendel’s “first law”) units of hereditary combination, just as Dalton’s atoms were units of chemical combination.

Each character involved in his crosses was regarded by Mendel as determined by a factor derived from the maternal and one derived from the paternal parent. A pure individual was thus represented by aa or bb, and an impure individual by ab. Mendel assumed that a and b are atoms of heredity in the sense that they retain their separate entities through the whole course of development. Having introduced this conception, he showed that all his numerical data followed from the laws of chance, if the maternal and paternal factors which determine a particular character separate in the formation of the gametes (pollen and ovules) so that one-half of the gametes contain only the maternal and one-half only the paternal factor for the character considered. The combinations which may occur as the result of fertilization are compatible with the assumption that any given male gamete (pollen or sperm) may fertilize any given female gamete (ovule or egg cell). Mendel’s first law may then be stated thus: characters distinguishing different hereditary strains depend upon factors which are inherited from both parents and segregate in the formation of the gametes, so that one-half contain the paternal and one-half the maternal factor. Mendel tested the implications of this hypothesis by crossing his hybrids to pure types with verifiable results. He then proceeded to make crosses involving two or three character differences. This led him to enunciate a second law which might be compared with the law of multiple proportions in chemistry, for its validity is of less general significance than the first law. It served eventually to direct attention to the much more complicated numerical results which arise in dealing with character differences attributable not to one but several pairs of factors. The analysis of such cases was left to Mendel’s successors.

There is internal evidence in Mendel’s writings to support the view that Mendel himself realized that the atomistic conception of inheritance would demand a drastic revision of the prevailing notion of variation. To Mendel’s generation, to Darwin and the pioneers of Natural Selection, variation and heredity were co-extensive terms. Offspring were always on the whole like their parents, but always on the other hand a little different. So the species in conformity with sound liberal principles broadened down from precedent to precedent. But on the atomistic view heredity is essentially conservative, and variation essentially revolutionary. For an indefinite number of generations the atoms of heredity remain unchanged. But times come, when the political barometer falls, and the change when it happens is a discontinuous one. Something new has been brought into being, as when lead is produced from the disintegration of radium or another allotropic modification of an element is formed. The full implications of this were not destined to be realized till forty years had elapsed. Meanwhile the evolutionary ship drifted upon an uncharted ocean of speculation without the compass of experiment to direct its course.

§2

Mendel’s work published in an obscure horticultural journal remained neglected for forty years, till in 1900 his principle was independently rediscovered by three continental workers—de Vries, Tschermak, and Correns. During that period the study of the reproductive process had progressed rapidly. The way was being paved for new and spectacular developments of the atomistic standpoint in heredity. To appreciate the subsequent elaboration of Mendel’s hypothesis in its historical perspective a brief digression into the anatomy of the cell is necessary.

Mendel’s researches were confined to plants. When he started his work the nature of fertilization in animals was still obscure. The bodies of animals like plants were known to be built up of microscopic bricks, or cells as Robert Hooke had called them. With the use of more powerful microscopes this had gained general recognition during the thirties and forties. Two centuries had elapsed, since Leeuwenhoek with the first microscope had seen seminal fluid teeming with minute vibratile bodies, the spermatozoa. At the end of the eighteenth century that inquisitive ecclesiastic Spallanzani had shown that the sperm is the essential constituent of the seminal fluid. By 1841 Kolliker had traced the development of the spermatozoa from single cells of the testis. It was not until 1875–9 that Hertwig and Fol working on sea urchins independently observed for the first time in history the penetration of the egg by the sperm, and established the universal rule that fertilization involves the union of a single sperm with a single egg cell. All modern discussion of genetic differences takes its starting point from the fact that anything which is implied by the word inheritance has its material basis in the microscopic sperm contributed by the father or in the egg cell with which it unites.

In all animals the sperm is a microscopic entity. In all animals from the jellyfish to Man with very few exceptions its appearance is extraordinarily similar. It consists of a thicker portion to which is attached a long vibratile process, or flagellum. The eggs of different animals are of very different dimensions. Sometimes they contain immense stores of food material (yolk). Sometimes as they pass to the exterior by the female generative tract they are invested with an additional slimy coat and a leathery or calcareous shell, secreted by special glands. The immature egg of all animals is essentially similar. In the living condition it is a spherical or ellipsoidal body in which a clear spherical vesicle is seen; this vesicle present in all cells is called the nucleus. The thicker part or body of the sperm consists mainly of the nucleus of the cell from which it is derived. At fertilization it swells up and unites with the nucleus of the egg. The fertilized egg then divides into two separate segments or cells, and the process of dividing is repeated an indefinite number of times. The cells or segments into which the fertilized egg divides each contain a nucleus, and the process of segmentation which involves the division of cells into two is accompanied by the division of the nucleus of each dividing cell. Like the testis or ovary the substance of all the organs of the animal body is built up of the microscopic bricks which we have called cells. In some tissues like bone and cartilage the bricks are separated by a good deal of mortar. Others, such as the lining membranes of the body, consist simply of cells packed tightly together. At the beginning of embryonic existence all the cells are very much alike. In the course of development the cells of different tissues are considerably differentiated. Throughout all stages the process of cell division always involves the partition of the nucleus in a highly characteristic manner.

The details of this peculiar process, first elucidated by Flemming and others during the seventies, has proved to be of astonishing significance for the further understanding of Mendel’s hypothesis. When a cell is about to divide, the nucleus looks like a tangle of fine threads; and this tangle of fine threads resolves itself into a number of readily distinguishable filaments which become progressively shorter, assuming the appearance of stout rods staining deeply with basic dyes. These rods, visible only with high powers of the microscope, are the chromosomes, whose behaviour has provided us with a tangible basis for Mendel’s conception of inheritance, and have thereby permitted an extensive clarification and amplification of the original hypothesis. From one point of view they might be said to have done as much for the Mendelian conception of heredity as the discovery of alpha particles has done for our belief in the atomic structure of matter. As the dividing cell begins to constrict, the chromosomes arrange themselves at its equator, and split longitudinally into halves, each half travelling to opposite poles, where they spin out again into fine threads from which the nuclei of the daughter cells are built up. Thus each of the chromosomes in the nucleus of any cell in the body is structurally equivalent to a corresponding chromosome in the preceding or succeeding cell generation. About the year 1875 it was recognized that this numerical constancy extends beyond the life of a single individual. In every species of animal or plant the number of chromosomes which can be counted in dividing nuclei is a constant for the species.

With the discovery of this fact a new problem arose so soon as the essential features of fertilization were appreciated. How is this constancy maintained from generation to generation of new individuals? Two investigators, Van Beneden and Boveri (1881–3), who worked on the horse threadworm, a form which has only four chromosomes in the dividing cells of the segmenting egg, showed that the egg and sperm each contain only half the number of chromosomes characteristic of the cells of the embryo. This conclusion turned out to be a perfectly general one. Attention was immediately directed to the nuclear changes which happen in the formation of the gametes. Innumerable cell divisions occur in the testis or ovary of an animal. These are at first similar in all respects to those which occur in the segmentation of the developing embryo; but cell division goes on in the testis or ovary throughout life. If we trace backwards the history of an individual sperm or egg in the testis or ovary in which it originates, we find a reduction of the number of chromosomes effected during the last division but one, leading up to the formation of a sperm or ripe egg. This penultimate division of the germ nuclei is preceded by the fusion of the chromosomes lengthwise in pairs. When the division actually takes place, each pair behaves like a single chromosome, splitting in such a way that one member of each pair goes to form each daughter nucleus. The succeeding division being normal, each gamete receives half the number of chromosomes present in ordinary cell division. At fertilization the normal number is restored. Thus each ordinary cell of the body has a chromosome set of which half the components are paternal and half maternal in origin.

In many animals and plants the chromosomes are very distinctly of different sizes and shapes, and can be sorted out into corresponding pairs. Such arrangements are constant for the species, and could only be maintained constant, if each gamete contains one representative of each pair. This means that the maternal and paternal constituents of a pair are distributed in the reduction division to different cells. The chromosomes therefore exist in pairs of which one element is of maternal origin and one of paternal origin. Each gamete receives one element of each pair, just as Mendel supposed that each gamete contained either the paternal or maternal element of his paired “factors.” By a curious coincidence this far-reaching conclusion was first established in the very year which witnessed the application of Mendel’s principles to animals by Bateson in England and Cuenot in France (1902). Its recognition accompanied the elucidation of another peculiarity of nuclear division, also destined to have important theoretical consequences. In many animals there is found to be an unequally mated pair of chromosomes, the XY pair. When this occurs, it occurs in one sex only. In the alternate sex there is a corresponding equal pair (XX). In birds and moths the female is the XY, the male the XX individual. In other animals the male is usually found with sufficiently careful measurement to have an unequal (XY) pair which is equally mated in the female (XX). During the nineties it was found that some animals had in one sex an odd number of chromosomes, a fact which at first sight seemed to conflict with the numerical constancy of the chromosomes. In the early years of the present century American zoologists provided the key to an understanding of the discrepancy. In all such cases the alternate sex has one more chromosome. The case of the large cockroach will serve as an illustration. The male of Periplaneta americana (its technical name) has 33, the female 34 chromosomes. The eggs will all have 17 chromosomes. One-half of the sperm will have 17, the other half 16 chromosomes. If a sperm of the former class fertilizes an egg, the individual produced will be a female (17 + 17 = 34); and if a sperm of the second type fertilizes an egg, the individual produced will be a male (17 + 16 = 33). In an animal with an unequally mated (XY) pair of chromosomes in the male reduction will result in one-half of the sperm carrying the X and one-half the Y chromosome. The eggs will all have the X, since this chromosome is equally paired in the female. Thus an egg fertilized by a Y-bearing sperm will become a male, while an egg fertilized by an X-bearing sperm will develop into a female.

§3

By statistical reasoning Mendel had deduced from his experimental data the existence of entities which behave just as the chromosomes do. He had no direct evidence that his factors had any material basis in the architecture of the germ cells. The new cell anatomy provided independent confirmation of his predictions from an unexpected quarter; but it was not immediately recognized that this was so. Antagonism to the belief that the chromosomes fulfilled the requirements of Mendel’s hypothesis is easily explicable. To Mendel’s first disciples his second and first laws were equally sacrosanct. Mendel’s second law implies that different pairs of hereditary factors behave quite independently of one another. On such an assumption one of two deductions is inevitable. Either the applicability of Mendel’s first law is extremely restricted; or the number of factors is too large to permit of their localization in the chromosomes. The sweet-pea, for instance, has only seven pairs of chromosomes. If Mendel’s second law were as general as the first, only seven pairs of factors could be accounted for by the behaviour of the chromosomes. From this dilemma further development of the atomistic view of heredity was rescued, when it was discovered that Mendel’s second law is only a particular case of the possibilities inherent in the first.

In 1910 Bateson and Punnet first discovered in the sweet-pea what they then called “coupling and repulsion,” or as we now say, linkage. Without going into the experimental data, we may define the phenomenon of linkage in the following way. Suppose that these are two varieties A and B which obey Mendel’s first law and two other varieties C and D which likewise conform to its requirements, when crossed with one another. Mendel’s second law stated that in a cross between AC and BD the second generation will consist of the types AC, AD, BC and BD in numerical proportions agreeable to the assumption that it is equally likely that the factor determining A will be present in the same gamete as the factor determining C or the factor determining D. Bateson and Punnet found that this does not always happen. There is another category of cases in which the factor which determines A sticks more or less completely to the factor for C in preference to the factor for D. The detailed analysis of these cases was at first made difficult by Mendel’s literal symbolism, and his way of thinking of factors in pairs. But the discovery of linkage at once led Lock to formulate the fruitful suggestion that factors located on the same chromosome pair would satisfy the requirements of linkage, while factors located on different pairs of chromosomes would fit in with Mendel’s second law.

From this point onwards the most spectacular development came from the study of inheritance in animals, and the significance of the chromosomes was immensely reinforced by newly gained knowledge of sex determination. Almost contemporaneously with the discovery of the sex chromosomes or XY mechanism, as we now say, Leonard Doncaster had elucidated in moths the phenomenon of sex-linked inheritance. This was soon found to be of common occurrence in animals. Till this discovery, which was made in 1905, the same results had always been obtained in crosses of pure-bred varieties, whether the male or the female parent displayed one or the other characteristic distinguishing them. Doncaster’s work on the currant moth showed that there is a category of cases which at first sight obey Mendel’s first law in its simplest form when the cross is made in one way, but yield a different type of result when the cross is carried out reciprocally with respect to the sex of the parents. In such cases one sex is only able to transmit certain characters to its offspring of the opposite sex. It was already known that the XY sex (male in Man and most animals) can only transmit its X chromosome to the XX type. The facts did not dovetail at first sight, because sex-linked inheritance was originally elucidated in birds and moths of which the female is the XY type. There was still an attitude of hesitancy towards accepting Lock’s hypothesis, strengthened by the persistence of an incorrect interpretation of the process of reduction which had been made the basis of Weismann’s metaphysical speculations concerning “germinal selection.”

When in 1914 Doncaster summed up the case for regarding the chromosomes as the material basis of Mendel’s first law, a new era had already dawned. Thomas Hunt Morgan, the central figure of a group of ardent investigators at Columbia, had initiated a body of enquiries which within half a decade eclipsed all other achievements that had succeeded Mendel’s pioneer labours. About the time when Bateson first encountered the phenomenon of linked inheritance Morgan began to rear the fruit-fly Drosophila for breeding experiments. Till then genetic experiment had been held in check by the slow rate at which most convenient animals and plants reproduce and the expense entailed in breeding them in sufficiently large numbers to permit statistical inference. The fruit-fly completes its life cycle, if kept in warm laboratory conditions, in a period of ten days. It is prolific. It feeds on rotten banana skins. It therefore costs little to breed. To these immense advantages it adds two others of supreme importance. It has only four pairs of chromosomes readily distinguishable from one another by size and shape; and it has produced in the laboratory a crop of several hundreds of sports or mutants. Each mutant type differs from the wild parent stock in some well-defined characteristic inherited in crosses with the wild type in accordance with Mendel’s first law. The mutant characters are extremely varied. One is distinguished from the red-eyed parent by having white eyes, another by having purple eyes, another by having no eyes at all. One is distinguished by having wings that are practically vestiges, another by wings that turn up at the tips, another by wings that are truncated at their extremities. From the wild type which has a greyish body, one mutant is distinguished by a deep black, another by yellow coloration. The mutant characters are thus in general clear-cut differences lending themselves to easy identification. With an animal that breeds so rapidly and prolifically information accumulated with astonishing rapidity. From data based on the study of a large assemblage of mutant characters there soon emerged the precise requirements of Lock’s hypothesis. All the mutant characters of Drosophila fall into four groups. Members of the same group always tend to stick together in hereditary transmission. Members of different groups like Mendel’s dihybrids behave independently of one another. Of several hundred mutant characters in Drosophila every one belongs to one of these four linkage groups; and the number of chromosome pairs in Drosophila is four.

This discovery was only the beginning of what might well be called one of the faery tales of modern scientific research. In the way of accepting Lock’s hypothesis there were still difficulties. It was in evading the principal difficulty that Morgan’s school extended the atomistic concept of heredity much further than his predecessors had done. Till then the main outcome of experiments on breeding had been to show that Mendel’s principle was of vastly wider applicability than was at first supposed, and to engender the suspicion that the patient unravelling of difficult and elusive cases would establish its universal validity. As yet the world of Mendel’s atoms was without form. Morgan and his colleagues gave it a map. Not content with showing that Mendel’s atoms of heredity have their material basis in the chromosomes, nor with actually identifying which chromosome is significantly associated with a particular mutant character, Morgan went further and localized the region of an individual chromosome in which a particular Mendelian factor resides. He thus gave to Mendel’s factors spatial co-ordinates in the living cell.

At the outset the study of linkage upon which the chromosome map is based was facilitated in the case of Drosophila, because the varieties dealt with were all known to be mutants from a fixed wild type. Thus it was possible to break away from Mendel’s conception of “pairs” of hybridizing characters. The Mendelian factor was replaced by the mutant gene, by saying which is implied that a mutant arises because at some point on a particular chromosome a physical change has taken place. The gene is the Mendelian factor for the mutant condition, but no assumptions are made about what determines the wild-type condition. The inter-relationship of different characters is greatly simplified by thinking only of the relation of one mutant gene to another. The discovery that all the genes fall into four groups corresponding to the four groups of chromosomes presented one stumbling-block. Members of the same group in general do not invariably stick together. When two mutants are crossed the numerical proportions of the various types of offspring give a definite value for the probability that the gene A and the gene B will stick together or separate apart. This is a constant for A and B. The constant used in practice is the tendency for A and B to separate. Expressed as a percentage, it is called the cross-over value. Some additional information was necessary to explain why A and B do not always stick together, if they are associated with the same chromosome. It was from the solution of this problem that the chromosome map took shape.

Here the sex chromosomes came to the rescue. One very interesting type of sport which has turned up in breeding the fruit-fly is not recognizable by any discrete bodily peculiarity but merely by an abnormality in the number of chromosomes. Of these the first to be discovered was a type of female which has in addition to its usual four pairs of chromosomes an additional Y chromosome. The XXY females yield very extraordinary numerical results both as regards the sex ratio and other characteristics, when used in making crosses involving mutant characters. There is a class of mutant characters in Drosophila, more than a hundred in all, distinguished by the fact that they are not inherited symmetrically with regard to sex. They display linkage inter se. They behave as “sex-linked” characters. The introduction of XXY females into crosses involving such mutant characters results in numerical ratios which are inexplicable on any assumption other than the view that the sex-linked gene is referable to the X chromosome alone. Yet, although the sex-linked genes are all borne on the same chromosome, they do not invariably stick together in crossing. The holistic chromosome clearly would not do. An atomistic chromosome had to be put in its place.

The clue to this was provided by studying more closely the extent to which the different genes stick together. Taking all the genes located on the X chromosome this remarkable generalization emerged from Morgan’s researches. If A, B, and C are three sex-linked genes; if the probability that A and B will not stick together is x, and the probability that B and C will not stick together is y, the probability that A and C will not stick together is either the sum or the difference of x and y. The correspondence here stated is, of course, subject to the margin of error permitted by the theory of probability. To interpret this new law of the linear alignment of the genes Morgan made use of a structural peculiarity of the reduction process. When the chromosomes pair in the reduction division, they appear to become twisted. The appearance suggests that in the ensuing split corresponding lengths of the original pair are interchanged. It is very natural to assume that the likelihood that two points will be separated from one another in such a manner is proportional to their distance apart. So if the sex-linked genes are arranged in a series along the length of the chromosome, the probability that A and C will not stick together must be the sum of the probabilities that A and B and B and C will be separated. This is just what experiment had shown to be true. Thus all the genes on the X chromosome can be arranged in a linear series. The intervals between consecutive genes in such a series represents a space dimension.

The law of the linear alignment of the genes was soon found to apply to the other groups of linked characters. Abnormalities in the number of chromosomes have made it possible to identify each of the remaining three linkage groups of the fruit-fly with its corresponding pair of chromosomes. The first chromosome map of Drosophila was constructed in 1916. It revealed the suggestive coincidence that the number of ascertained points on each pair of chromosomes is roughly proportional to its size. There is now very little doubt that the work of the Columbia school has revealed an aspect of inheritance which is of general significance. After years of patient work with the relatively slow breeding sweet-pea, Punnet has at length elucidated seven linkage groups corresponding to its seven pairs of chromosomes. He has constructed a chromosome map of a seed plant on the basis of the principle first established for the fruit-fly. A law which holds good for two organisms so far apart in the evolutionary scale can hardly be supposed to be lacking in universal validity. The chromosome hypothesis may now take its place as one of the major generalizations of biological science. The law of linear alignment has transformed Mendel’s original conception of inheritance in a way which might be compared with the elaboration of Dalton’s hypothesis after the discovery of the law of combination of gases by volume. Mendel’s atoms of heredity are now units spatially localized in larger units of microscopically visible dimensions. These supermolecules are the chromosomes.

Being a portion of living matter the chromosome is constantly undergoing chemical change. Some critics of the chromosome hypothesis have based objections upon this circumstance. The difficulty is more apparent than real. Like the individuality of the modern atom the individuality of the chromosome must be conceived in statistical terms. For the discussion of the more familiar chemical reactions the statical atom of traditional chemistry is adequate. For the interpretation of hybridization experiments the diagrammatic chromosome of the text-book suffices. In the field of radioactivity the statical atom makes way for a dynamical model. So also in the domain of cell physiology we conceive the chromosome as an ever-changing entity. The logical situation is analogous in the two cases. Those who hold with Dr. Haldane that the biologist must interpret his data in a manner different from that in which the chemist or physicist interpret theirs have now to fall back on the contention that the Mendelian view is only a partial picture of heredity transmission. Anything which might have been said in favour of this contention ten years ago has been weakened by recent work on the inheritance of size. The pioneers of Mendelism selected clear-cut hereditary differences which ordinarily manifest themselves in any environment in which the animal or plant can live. They succeeded in showing that a vast number of hereditary differences involving a great variety of anatomical and physiological features conform to the requirements of Mendel’s hypothesis. There was one category of phenomena which remained obscure till quite recently. Differences in size, height, body weight and the like vary greatly with environmental conditions. Two stocks may be distinguished from one another by the fact that the average member of one is measurably different from the average member of another; but any given individual of one stock may be indistinguishable from another individual of the other, because, even when the environment is standardized as much as is practicable, the range of variability of the two stocks overlaps. The analysis of such cases cannot be undertaken by the ordinary technique of Mendelian experiments; but certain statistical requirements of Mendel’s laws may nevertheless be verified. By elementary statistical reasoning we can deduce that the coefficient of variability of the progeny of a cross between two inbred stocks must be a minimum in the first generation and a maximum in the second. This has been shown to be true in a number of crosses in which it is impossible to distinguish individual genetic types by direct observation.

There is no longer any adequate reason to support the contention that Mendel’s atomistic concept leads us to an incomplete understanding of biparental inheritance in animals and plants. Those who assert that it is so are now forced to fall back upon the last resort of obscurantism by appealing to the magnitude of our ignorance. The modern theory of the gene is a statistical construction consistently developed by a logical interpretation similar to that adopted in elaborating the great generalizations of physical science. The mechanist is often accused of attributing vital processes to “chance” combinations of phenomena. If the word chance is used to imply that we do not know the precise conditions which determine such combinations, the statement is hardly exceptionable. It might also imply that the phenomena which biologists study can be successfully interpreted in terms of the mathematical laws of chance. The history of Mendelism shows that these laws provide a fruitful basis for predicting the behaviour of living systems, even where physico-chemical hypotheses at present fail to throw light on the phenomena which the biologist studies. The biologist is able to progress to greater certainty of prediction only when he interprets his data with the same logical method employed by the chemist and physicist to deduce physical “laws.” Whatever the future holds in store for further interpretation of heredity and variation on physico-chemical lines, the progress already achieved has at every stage involved elimination of holistic concepts by the ruthless application of mechanistic logic. To the application of physico-chemical hypotheses no branch of physiology has proved more recalcitrant than the physiology of inheritance. No branch of physiology might more suitably be chosen to cast doubt on Dr. Haldane’s recent statement that “anything which can properly be called scientific physiology is impossible apart from the assumption of holism.”

III. THE NATURE OF LIFE

“I am sorry then, I have pretended to be a philosopher: for I find your questions very perplexing; and am in danger, if my answer be too rigid and severe, of passing for a pedant and scholastic: if it be too easy and free, of being taken for a preacher of vice and immorality. However, to satisfy you, I shall deliver my opinion upon the matter, and shall only desire you to esteem it of as little consequence as I do myself. By that means you will neither think it worthy of your ridicule nor your anger.”—David Hume, The Sceptic

§1

Since a man must needs live before he can be a philosopher, no problem of philosophy is more fundamental than the nature of life. There is also no issue which provides more scope for vague, barren and undisciplined discussion. A Regius Professor of Moral Philosophy, whether he accepts the fact with resignation or refuses to do so, is a piece of living matter. Perhaps this is why physicists are more vocal than biologists in promoting a pacific solution of the territorial dispute between science and traditional philosophy. At the present moment it is the fashion among those who are writing on scientific philosophy either to neglect the contribution of the biologist to the world symposium, or to assume that the biologist in dealing with living matter operates with different methods and different concepts from those employed in physics. No phase in the history of biology is more fitted than the present to illustrate the fundamental unity of scientific method. In no branch of science is the limit of applicability of scientific method a more significant issue.

Since a scientific concept is only a way of describing a class of properties, the nature of life cannot refer to anything but the nature of the properties of living things. Having arrived at some general classification of the characteristic properties of living things, a discussion of the nature of life in the light of modern biological science presents two issues of pre-eminent interest. One is how far the methods employed in physical science have been successful and are likely to continue to prove successful in dealing with the properties of living matter. The other is how the increasing measure of success which attends the utilization of purely physical concepts to interpret the properties of living matter is calculated to influence our evaluation of the place of science in human thought. Whatever differences of interpretation may exist among biologists on matters of detail, it should at least be possible to infer from a survey of the progress of biology whether the study of living matter is progressing satisfactorily along the lines of quantitative analysis of experimental data towards greater certainty of prediction, and whether there is good reason to believe that the preservation of the teleological standpoint in dealing with living matter is likely to ensure conspicuous success in the same direction.

It may be admitted that there exists among biologists more unanimity with reference to the first than towards the second issue. Every infant science makes use of notions peculiar to its own province. Chemistry has but lately passed beyond the stage when the concept of affinity first became amenable to interpretation in thermodynamical quantities. There are still many biologists who would assert that the concept of adaptation demarcates the province of biology from that of physics and chemistry by an impassable gulf. There are others, fewer in number, who, surveying the teleological growing pains of the more exact sciences and bearing in mind that only 300 years have passed since the properties of familiar chemical compounds were literally personified as spirits of wood, spirits of salt and the like, do not feel compelled to regard the concept of adaptation as final. They are able to entertain the possibility that those properties which enable an organism to maintain its continued existence as an organism are not permanently more incapable of physical interpretation than the polarization of a voltaic battery, a phenomenon which the consistent teleologist would presumably regard as an attempt on the part of the latter to save its own life. Clearly the onus of defining what precisely is implied in the concept of adaptation lies on those who assert its uniqueness. Until the vitalist is more definite on this issue, the mechanist is under no obligation to refrain from classifying the properties of living matter in the light of his own experience. The mechanist denies that anything is to be gained by clinging to the teleological standpoint with its implication of some extra- or intra-mundane purpose which has been abandoned in all branches of science that lay claim to exactitude. He refuses to deal with living matter except in as far as it is considered as a series of “events” whose characteristics must be interpreted with rigid economy of hypothesis.

In approaching any lump of living matter, let us say the author of these essays, as an object of the external world, the maintenance of economy of hypothesis compels the enquirer to seek as far as possible a common basis for the characteristic properties of living and non-living systems. This necessitates a clear definition of the distinction between the two. Taking a comparatively complex organism, as, for instance, the common frog, a distinction might be attempted along the following lines. In the first place, its possibilities of behaviour are more varied than those of any machine which can be manufactured by man; yet, while possessing a greater range of reversible response than any non-living system, it would be difficult to specify in a living system any single activity which could not be reproduced by a mechanical system. Apart from this diversity, which we may refer to under the generic term reactivity, living matter is characterized in general by the wide range of external influences which are significant in determining its characteristic reversible responses. This peculiarity, in view of the subjective preconceptions implicit in the older terms irritability, sensation, etc., may be denoted by the term receptivity. Here again it is impossible to isolate any single agency (or “stimulus”) capable of evoking reversible change in any living system and incapable of evoking reversible change in any non-living system. Finally—and at first sight—a more diagnostic difference between living and non-living matter is seen in the property of reproduction (taken in the broader sense of the term, to include growth). A given piece of living matter comes into being in our experience only through the agency of other pieces of living matter closely resembling itself.

§2

Were the more obscure process of sexual reproduction universally characteristic of living matter, this distinction would appear especially fundamental. Experimental biology is far from the achievement of a complete physico-chemical analysis of asexual reproduction in any type of organism. On the other hand, in the life cycles of those multitudes of micro-organisms which multiply by simple fission after attaining a certain limit of growth, there is nothing which compels an unprejudiced investigator to regard the process as more intrinsically incapable of physical interpretation than the splitting into two of a liquid drop. Although our knowledge of the nature of sexual reproduction is fragmentary, in this very field some particularly spectacular advances have been registered in substituting physical agencies as effective instruments for initiating processes which at one time were only amenable to the influences of living matter itself. Thirty years have now passed since Loeb’s discovery that changes in the osmotic pressure of the external medium or alteration of the permeability of the egg itself, leading to changes in its own internal osmotic pressure, can initiate without any assistance from the sperm the development of the ovum into a new and complete organism. That discovery was the starting-point of a body of investigations whose influence has radiated into many other fields of biological enquiry. Especially noteworthy in this connexion is the work of Warburg during the last decade. Warburg was able to show that sea-urchin eggs, and later animal cells in general, if rapidly dehydrated and ground to a powder, will, like the intact cell, absorb oxygen for some time when moistened. He showed also that this property, like respiration in the intact cell, can be abolished by the action of cyanides and other classes of tissue poisons. By doing so, Warburg has taken a characteristic and highly complex property of living matter out of the realm of vitalism into that of physical chemistry. His analysis went further. Experiment showed that three classes of poisons which inhibit tissue respiration can be distinguished by their quantitative relations. Of these, the efficacy of one class, the cyanides, was shown by Warburg to be correlated with the iron content of the cell. On the hypothesis that iron catalysis is the main factor in the oxidation of organic material in the cell, Warburg manufactured suspensions of charcoal with a high iron content capable of catalysing the auto-oxidation of sugars, fats, etc. The catalytic activity of these suspensions was found to be related quantitatively to the three categories of respiratory poisons in a manner closely parallel to the action of the latter on tissue respiration.

Though reproduction is, in some respects, to the biologist at least, the most fundamental of all the three features which I have defined above, the ever-changing reactivity and manifold receptivity to external influences so characteristic of living matter pre-eminently engage our attention in connexion with the more intimate and subtle issues of a field of enquiry which biology may yet claim. I refer to the analysis of human behaviour. In this connexion I shall mention progress in three directions as illustrating the transition from teleological to quantitative treatment during the last half century; namely, the physical analysis of the events which constitute an isolated unit of response or reflex, the integration of reflexes in the normal behaviour of animals, and the determination of new behaviour patterns along the lines laid down by Pavlov’s school. With regard to the first, we will consider the effect of flashing a bright light upon an animal that has been previously kept in the dark. The characteristic response, let us say, blinking of the eyelids, and the intervening events involved are, first, a physical change in a receptive area, namely, the retina; secondly, the propagation of the disturbance there set up along a certain path, the nervous system; thirdly, the liberation of a considerable quantity of energy at the seat of response or effector organ, that is to say, the muscles of the eyelid.

Our knowledge of the nature of receptivity is least complete. That it is a measurable physical event is beyond dispute. When light impinges upon a given area of the retina there follows a characteristic series of changes in electrical potential of the excited area with reference to a non-excited area. Through the work of Jolly, Adrian and others the sequence, the time relations and the magnitude of these changes are being related to the intensity and duration of the stimulus within predictable limits for a given species. These events initiate the propagation of the disturbance known as the nervous impulse. The nervous impulse is a physical event whose space-time relations can be defined as concretely as the passage of an electric current through a wire. Three-quarters of a century ago Helmholtz showed that the time which elapses between the application of a stimulus to a nerve and contraction of its attached muscle is a linear function of the distance between the latter and the point of application of the stimulus. The conception of the nervous impulse as a physical event had been, till this discovery, entirely repugnant to scientific thought. We now know not only, as Helmholtz showed, that the nervous impulse has definite space-time co-ordinates, but that it has the dimensions of energy. Its passage corresponds to the rate of propagation of an electrical change of an analogous character to the electrical response of the excited retina. The total energy of its propagation has been recently measured by Gerrard and A. V. Hill from determinations of heat production during its passage. Its mass relations are attested by a measurable increase in the carbon dioxide production of stimulated nerve. The rate of propagation of the nervous impulse varies like all chemical reactions in a characteristic way with increase in temperature. The goal of the nervous impulse after it has traversed one or more synapses in the central nervous system is the effecter organ itself—in the case of blinking of the eyelids, a muscle fibre. During the past two decades a series of brilliant researches based on calorimetric methods have revolutionized our knowledge of the final component of the reflex. A. V. Hill and Meyerhof have correlated the chemical and energetic changes accompanying muscular contraction with a precision of the order expected in purely physico-chemical determinations. They have shown that the total energy of muscular contraction can be quantitatively related to the energy liberated in vitro by the breakdown into lactic acid of an amount of glycogen equivalent to that which is converted into lactic acid in the actual contractile process.

Passing from the analysis of the constituent events of the reflex to the integration of reflexes in normal behaviour, we are faced with a striking change in the attitude of enquiry adopted in the study of those aspects of behaviour determined by generalized stimuli such as light and gravity and denoted by the term tropisms. Three-quarters of a century ago, after Helmholtz had dispelled the belief that identified the nervous impulse with an imponderable psychical principle, biologists like Lubbock were content for the most part with the statement that the moth flies towards the candle because it likes the light. The work of Loeb and others has shown that the state of contraction of particular groups of muscles is reflexly determined by the stimulation of particular areas of the retina. It is a mechanical necessity that when different areas are unequally stimulated, differences in tension of different groups of muscles will bring the body into such a position that symmetrically opposite areas will be equally illuminated. The animal must move, as in fact it does, along the path of the incident beam, whether by so doing it brings itself into a brighter, or, as can easily be arranged experimentally, a darker situation. Whereas the older and purely teleological attitude permits us to predict nothing of consequence, the objective interpretation of tropisms by experimental methods permits us to make many verifiable predictions, as, for instance, the fact that the moth will move in circles, if one eye is blackened, owing to the fact that the muscles on that side will be more relaxed.

By the end of the nineteenth century, experimental biologists were generally disposed to the belief that the analysis of the reflex and the integration of reflex systems were problems not of apologetics but of energetics. Investigation had been confined to those aspects of behaviour which are for practical purposes invariable responses to a particular situation. From the human standpoint the most fascinating feature of the behaviour of an organism is, after all, the extent to which its behaviour is conditioned not by the immediate but by the antecedent situation. In the opening years of this century the researches of Sherrington were elucidating the integration of reflexes in normal behaviour. Restricted as they were to the decerebrate animal, the traditional distinction between reflex and voluntary activity remained as a defeatist formula in biological nomenclature. The distinction was not a gratuitous olive branch to introspective philosophy. It had its objective basis in the domain of behaviour which is not uniquely determined by the immediate stimulus, when all synchronous conditions have been standardized. That distinction has been superseded to-day by the work of Pavlov’s school, which has shown, first, that in the higher animals with the cerebrum intact, new reflex systems can be built up experimentally under perfectly definable and reproducible conditions; that the relations between such conditioned reflexes can be defined in the language of space and time; and that the concept of sensation can be externalized by reference to the ability of a given stimulus to become a specific agent in the building up of a new reflex system. In short, it is legitimate to anticipate the possibility of giving a complete specification of how such an animal as a dog will behave in a given situation without recourse to the traditional nomenclature of memory, consciousness, sensation, etc.

Pavlov’s work is now accessible to the English reader through two translations of the Russian physiologist’s own writings and several excellent résumés, such as the one given in Lovatt Evans’ Recent Advances in Physiology. How far-reaching are its consequences has not been widely recognized even by biologists themselves. Experimental biology, during its brief career, has attempted to accommodate itself to the introspective temper of traditional philosophy by a compromise explicitly formulated in the writings of Descartes, who bequeathed to physiology the dualism of mind and matter. In conformity with the Chaldæan mythos, many philosophers, Descartes among them, have endowed Man alone with soulfulness. The coming of the Evolutionary hypothesis has broken down so inflexible a distinction between Man and other forms of living matter. Evolutionists in the nineteenth century, like Haeckel, were prepared to equip the Amœba with a soul. In our time the Cartesian compromise has again shifted its boundaries. By the beginning of this century the moth once more had gone to join the candle. Still Man stood with a little family of mammals around him, each with one leg on either side of the frontier that separates the universe of space and time from the Platonic world of universals. Pavlov has taken those aspects of behaviour which would have been referred twenty years ago to exclusively introspective concepts, and has treated them successfully as predictable configurations in a space-time framework. The little family of mammals has been let through the tollgate of the Cartesian compromise. A new school of psychologists has come into being with the express object of making psychology a physical science, relieving Man, the celestial pilgrim, of his burden of soul.

Philosophers have always had a legitimate cause for complaint that biologists were unable to deal with those aspects of human life which interest people most. The distinction between reflex and voluntary activity provided the fullest absolution for that amiable libertarianism which we all entertain under the influence of alcohol and love. Because that distinction was implicit in the outlook of the most radical mechanists of the last generation, Loeb among them, Dr. Haldane finds it so easy to point out the inadequacy of the mechanistic outlook. In the light of Pavlov’s work we can now envisage the possibility that the methods of physical science will one day claim the whole field of what can be properly called knowledge. If I am right in cherishing such an opinion, it would thus appear that the investigation of the conditioned reflex initiates a new epoch in biology, pregnant with more far-reaching philosophical implications than the evolutionary speculations of the nineteenth century. The fact that no reference to the conditioned reflex is contained in Dr. Haldane’s Gifford Lectures may in part account for the fact that he can so easily dispose of the mechanistic position. The modern mechanist does not say that thought and love and heroism do not exist. He says, show me behaviour to which you apply the adjectives thoughtful or loving or heroic, and we will, one fine day, endeavour to arrive at predictable conclusions with reference to it by following the only method of enquiry which we have learned by experience to trust. When Dr. Haldane goes out of his way to dispose of the puerile formula that thought is a secretion of the brain, as bile is a secretion of the liver, and does so, I gather, under the impression that mechanists either believe it to mean something or alternatively shut their eyes to the major problems of existence, I can only respectfully suggest that he is flogging a dead horse, while the living ones are getting out of the vitalistic stables.

I have endeavoured so far to indicate the increasing measure of success that has crowned the application of physical methods and the use of physical concepts in modern biological investigation. I have attempted to illustrate the continuous retreat from teleological concepts that has accompanied this advance. In asking what progress may be anticipated from encouraging the teleological attitude to the nature of life, I wish now to urge that the important advances of biological science during the last hundred years have not only involved continual abandonment of teleological concepts, but have consistently been made in the teeth of opposition from the vitalists, organicists and holists of their time. A century ago, in the same year that witnessed Wöhler’s announcement of the successful synthesis of Urea, the great chemist Henry wrote (1827) concerning organic compounds: “It is not probable that we shall ever attain the power of imitating Nature in these operations. For in the functions of a living plant a directing principle appears to be concerned peculiar to animated bodies and superior to and differing from the cause which has been termed chemical affinity.” Only six years before Helmholtz’s determination of the velocity of the nervous impulse in 1851, Johannes Müller had declared that to measure the propagation of that imponderable psychical principle was a theoretical absurdity.

It is not unlikely that before another celebration of the centenary of Wöhler’s achievement, Fischer’s synthesis of an octadecapeptide will have been surpassed by the manufacture of complex proteins in the laboratory. Looking forward a little in the light of what success has crowned the construction of physical models of vital processes, it is, as Sir Edward Sharpey Schafer scandalously suggested at a meeting of the British Association some years ago, perfectly legitimate to entertain the possibility, even the likelihood, that scientists will one day construct from artificially synthesized organic materials, systems with so wide a range of reversible reactivity and receptivity to external influences that they would be called organisms, if met with in Nature. While taking a more hopeful view in this matter than some biologists, I would remark that the validity of the mechanistic outlook is quite independent of this possibility. The security of any dynamical system of treating the motions of the heavenly bodies is independent of the possibility that human effort could manufacture a new satellite for Jupiter.

§3

If we can assert that the present phase of biological enquiry is a peculiarly fruitful one, and that there is no reason to see any immediate cessation of progress in the use of physical concepts as the basis of our analysis of the properties of living matter, can we not go further and state that we have absolutely no encouragement for entertaining the hope that any deeper knowledge will accrue from apostrophizing under the sobriquets of entelechy, life force, élan vital that elusive entity to which, perhaps, the poet William Blake referred as Old Nobodaddy? It is doubtful whether we shall see a recrudescence of such frankly animistic devices as these. As biology becomes more technical and more exact, an aptitude for rehabilitating oriental mysticism in somewhat unusual verbiage will be regarded as an insufficient equipment for entering the field of biological controversy. The investigator who abandons physics for the pursuit of biology will contribute new ideas. Fruitful contributions need no longer be expected from those who combine the pursuit of literature with an amiable interest in natural history. The days of Butler and Bergson are passed.

Dr. Haldane, the most vigorous contemporary critic of the mechanistic standpoint, is very anxious to avoid any suspicion of being tainted with the cruder forms of vitalism. He disowns any allegiance to the life force, élan vital et hoc genus omne, except in so far as he, somewhat mysteriously, contrives to introduce an adventitious deity into the latter portion of his Gifford Lectures. This does not make its appearance until his major thesis is complete. Anxious as is Dr. Haldane to disclaim adherence to the tenets of vitalism, he is very definite in denying the possibility that atomistic concepts will ever successfully deal with the problem of what he calls “conscious behaviour.” In the light of Pavlov’s work we see that the problem of what is usually called conscious behaviour, or as we should rather say conditioned behaviour, can now be approached as a problem in the study of those conditions which determine whether a new reflex will, or will not, be brought into being. We may state this in other words by enquiring how the passage of impulses along particular tracts in the central nervous system influences the facility with which the nervous impulse will pass across a particular type of synapse. Since the problem of the conductivity of the synapse is, as we have seen, an essentially physical problem, it is not overstating the case to say that the work of Pavlov’s school has brought the study of what Dr. Haldane calls “conscious behaviour” within the realm of physical enquiry. Once this is fully grasped it no longer seems incredible that the interpretation of conditioned behaviour will eventually come within the scope of physico-chemical analysis. Contrary to the holistic standpoint, we are thus led to an atomistic concept of individuality. This I shall venture to formulate as the statistical probability that in an immensely elaborate system of reversible reactions a certain number of states characteristic of any given moment will be reproducible at another moment.

In his Lowell lectures Professor A. V. Hill lays down two general conclusions derived from the extension of modern biological enquiry. First, as I have endeavoured to show, there is no limit to the extent to which the mechanisms of life can be elucidated with the aid of physical methods and concepts. Second, that, however far we get, we shall still find function, adaptation, organization and purpose in the processes we explore. I would venture to suggest that, however alluring such a compromise between vitalism and mechanism may appear, these two conclusions, though formally in nowise inconsistent, are, nevertheless, in practice incompatible. As Henderson points out in his Fitness of the Environment, if we wish to indulge in teleological phantasies, we can find as much scope in physics and chemistry as in biology. We do not dismiss the hypothesis that thunderstorms occur when a blue unicorn sneezes on Uranus, because it is actually possible to disprove so engaging a fancy, but simply because other ways of treating thunderstorms lead to more useful conclusions.

Hence it seems to me that as we come to understand more and more about the mechanics of living systems by using methods of which Professor Hill is so brilliant an exponent, we shall inevitably find ourselves talking less and less about purpose and function. In consequence many of the problems which now engage the attention of philosophers will be relegated to the same status as the philosopher’s stone. No doubt such a change will come very gradually, so gradually that we shall hardly notice it. Nevertheless, one may venture to predict that philosophers, already forced by the developments of modern physics to divert their attention from the pretentious crossword puzzles of the Hegelian tradition, will sooner or later be driven to take account of the post-evolutionary developments in biology, and more especially those which have their starting-point in Pavlov’s researches.

By undertaking the analysis of the characteristics of conscious behaviour without departing from the methods of the traditional physiology of reflex action, biological science in our generation has shown that there is no nicely defined boundary at which physiology ends and moral philosophy begins. Hitherto physiology and academic philosophy have developed independently, because physiologists themselves have accepted the common sense dualism of mind and matter. Moral philosophy can no longer claim that there is any distinctive aspect of the Nature of Life, which lies beyond the province of physiological enquiry. If any fundamental distinction between mind and matter remains, that distinction henceforth defines the antinomy of a public world of common beliefs which all can share, the conceptual world of science in which ethical neutrality and economy of hypothesis reign supreme, and, in contradistinction to that public world, many private worlds which for the present remain impenetrable through the medium of discourse. Biological science is continually socializing our beliefs. What seems irrevocably part of the private worlds of one generation becomes irrevocably part of the public world of its grandchildren. Thus the new pluralism will not be, like the Cartesian system, static, but dynamic. It is ever tending towards a monistic outlook as a limiting case. Such a monism, unlike traditional materialism and traditional idealism will be regarded not as a formula but as an asymptote. It is evidently immaterial to public discourse whether we privately entertain the view that the public world is more or less real than our private worlds. It would thus seem that as biological science invades the province of human behaviour the concept of publicity, as I venture to call the communicability of beliefs, will come to occupy the status of importance which reality has held in the systems of egocentric philosophers.

The public world, as I have conceived it, is a construction based on the continuous extension of the principle of mechanism. The principle of mechanism, that a complex system is interpretable only by reference to the properties of its constituent parts, is not urged in the spirit of dogmatic assertion, but because it has served us well in the past. We still await any single verifiable conclusion that is uniquely developed from any alternative principle. Holism, the newest form of Vitalism, claims to have found an alternative or supplementary principle that is essentially teleological. The holist does not specify by reference to any single concrete situation how he proposes to use his principle. It is admitted even by the mechanist that we are not in a position to construct a symbolic relation which will completely describe the vagaries of a Ford car in terms of the field equations of the proton and electron. Does the holist wish us to believe that we can help anyone to drive a car by assuring him that at every level of complexity between the internal structure of the atom and the newly licensed automobile there emerges an ever-increasing urge to a wholeness or unity which is somehow indefinably different from the interaction of the parts? Verily the mechanist of all people knows that we know in part and we prophesy in part. For this very reason, because he is prepared to await with patience the slow advance of science, he refuses to subscribe to high-sounding pseudonyms for ignorance and principles that are never seriously intended to be put into practice.

In the recent symposium on The Nature of Life before the British Association both General Smuts in his exposition of the holistic standpoint and Dr. Haldane who supported him dwelt upon the supposed collapse of mechanistic principles in physics itself. The former cited in support of his point a somewhat rhetorical remark by Dr. Whitehead in this sense. It is of course evident that if our mechanical principles undergo modification our biological interpretations must share in the general change of outlook. It is, therefore, beside the point to criticize the mechanistic standpoint on the ground that our mechanical principles are undergoing revision. Let us examine this objection a little more closely. Experimental biology, we are told, has been directed towards the attempt to describe the properties of living matter in terms of the traditional physical concepts of mass, length, time, energy, etc. Since these concepts now appear to be less fundamental than we once believed, the hope that a complete mathematical description of the universe is realizable, has, as Mr. Sullivan asserts with triumphant naïveté, “no longer any plausibility.” Surely it is evident that a signal advance towards a more monistic interpretation of nature has been made, when the analysis of any biological phenomenon has been achieved with the aid of traditional physical concepts, and when concepts once peculiar to biology, as affinity was once peculiar to chemistry, have been translated into the traditional language of physics. Physics to-day is seeking a new synthesis to take into one system of equations all the old data, and many new ones which have lately accumulated. This is not a new situation. The old mechanics remains as valid as ever for the realm in which it was developed to operate. To effect a more comprehensive scheme it has been necessary to examine many of the old postulates. In the meantime we have to recognize that we are not so near to a single unifying hypothesis as the rise of energetics led the physicists of Kelvin’s generation to hope. What does this signify? Certainly not that mechanics has abandoned the principles of mechanism. Is it not rather a fact that the modern physicist is complaining that the inadequacy of Newtonian principles is in part attributable to teleological implications insufficiently recognized till now? The very hope of finality which Kelvin’s generation entertained seems from the new mechanistic standpoint, as I have stated it, to savour of scholasticism.

In taking this line General Smuts and Dr. Haldane seem to me to have laid bare the source of a misunderstanding that lies at the root of most of the criticism which vitalists old or new direct against the new or the old mechanistic standpoint. Those who have the scholastic predilection for finality and the scholastic predilection for the abstract noun, do not seem to be able to believe in the existence of people who are not like themselves. They cannot, it appears, understand that unless one starts off with the obsession that the universe can be summed up in a monosyllable, one is under no imperative necessity on the one hand to be resentful towards or disappointed with science because it lays no claim to the finality of religious dogma, nor on the other to make the assumption that such finality ought to be obtainable. The mechanist does not claim that his system is, or ever will be, complete in the sense that science will one day find an answer for all the conundrums which the scholastic temperament dictates. On the contrary, it is the essence of the mechanistic position that there is a technique of asking questions profitably as well as a way of answering them satisfactorily. All the mechanist claims is that as far as we can see at present his way of dealing with things leads to the most complete unanimity which it is possible to attain. Against the old vitalism, that of Dr. Haldane, who denies that the principle of mechanism can ever deal with conscious behaviour the older mechanistic outlook was secure in the assurance that, if the principle of mechanism failed at such a level, no other principle led to verifiable predictions in the same field. Against the new vitalism or holistic standpoint of General Smuts which no longer asserts dogmatically that the principle of mechanism is inapplicable at any specific level of existence, but contends that it does not anywhere give a complete account, the new mechanistic or publicist standpoint which I have outlined contends that if the principle of mechanism fails to give a complete account at any level no alternative or supplementary principle has been discovered. The reply of the mechanist old or new to the vitalist old or new is that of Mr. W. B. Yeats’ faeries:

It follows that, in any discussion between the two, the combatants are generally at cross-purposes. The mechanist is primarily concerned with an epistemological issue. His critic has always an ontological axe to grind. The mechanist is concerned with how to proceed to a construction which will represent as much about the universe as human beings with their limited range of receptor organs can agree to accept. The vitalist or holist has an incorrigible urge to get behind the limitations of our receptor organs and discover what the universe is really like. What we mean by really in this connexion evidently depends upon whether we view the question socially or individually. In our relation to other human beings the nearest approach to what the universe is really like is found in the schematization of our common experiences. If there is any other reality its sanction is non-social. Thus in contradistinction to the reality of traditional philosophy which is an individualistic concept, the concept of publicity, which it is proposed to substitute as the goal of synthetic philosophy, is an essentially social one.

IV. THE CONCEPT OF ADAPTATION

“No philosopher who is rational and modest has ever pretended to assign the ultimate cause of any natural operation, or to show distinctly the action of that power which produces single effect in the universe. It is confessed that the utmost effort of human reason is to reduce the principles productive of natural phenomena to a greater simplicity.... The most perfect philosophy of the natural kind only staves off our ignorance a little longer, as perhaps the most perfect philosophy of the moral or metaphysical kind serves only to discover larger portions of it.”—David Hume, Sceptical Doubts

§1

By those who hesitate to commit themselves to an explicit advocacy of either the vitalistic or mechanistic views about the Nature of Life it has often been urged that the concept of adaptation is fundamental to biological science. Professor A. V. Hill is perhaps the most brilliant physiologist now living. He adopts a hopeful attitude to the progress which awaits further analysis of the properties of living matter in physico-chemical terms. He also thinks that, however far mechanistic principles are extended, the biologist will always encounter “adaptation” in the phenomena which he studies. Another distinguished physiologist, Professor Lovatt Evans has expressed himself in rather more emphatic terms.

“Physiologists,” he states, “in attempting to know what life is, have in my opinion attempted too much, and I think that a new point of view is essential.... The idea of adaptation, urged by Claude Bernard, should be adopted by physiology as its basal principle, as the chemist accepts the conservation of matter or the physicist the conservation of energy. We need not seek to know why it is so, that is the province of the philosopher.... It is not a definition of what life is, but a brief statement of its way.... Life is conserved by adaptation.”

When I first read these words I was not sure that I agreed with them. I was not quite certain that I knew what they meant. I had already come to the conclusion that the word adaptation is frequently used by biologists without a very clear agreement as to its content. I cannot subscribe to the view that there is a sort of trade union of philosophers to which physiologists are ineligible, unless they can show their articles of apprenticeship. Nor can I conceive what is meant by a concept of life except such as is implied in a statement of its way. A scientific concept defines a class of properties. A scientific concept of life or adaptation must conform to this requirement. In this essay my object is not to criticize Professor Lovatt Evans for whose breadth of view I entertain a very sincere respect. I have quoted his words, because they focus attention on some significant and controversial issues. They serve to reveal how imperative it has become that biologists should agree about the sense in which they intend to use the word adaptation.

The quotation given above might be interpreted to mean two very different things. If the term adaptation is used to define certain very general characteristics of living systems, it becomes almost co-extensive with a scientific concept of life itself. If we use principle in a somewhat archaic sense to indicate a field for investigation, like the principle of affinity or the active principle of the thyroid gland, there can be no question that the idea of adaptation is the basic principle of physiology. The comparison of the biologist with the chemist or physicist seems to go beyond this, and imply that adaptation is not something to be explored and interpreted, but part of the logical procedure of biology, something by the aid of which we can predict conclusions of universal validity in the field of biological enquiry. I do not think that Professor Lovatt Evans really means this. I do urge that biologists continually confuse within the compass of the concept of adaptation the notion of a problem for solution and of a vera causa. This in everything but verbiage is precisely what the cruder type of vitalist does, when he invokes the vital principle. He first introduces a term to describe a large number of things about which we are ignorant and wish that we knew more. He then falls into the trap of imagining that the invention of a new term has solved the problem.

Quite apart from this difference which, if it is to define the scope of our scientific enquiries, cannot be dismissed as metaphysics, biologists differ a good deal concerning the extent of the phenomena and the kinds of phenomena they are dealing with, when they speak of adaptation. The physiologist—in the restricted sense of the term—is usually referring to something which might be called the self-regulating characteristic of the body. The evolutionary biologist—who to-day is a physiologist in the broader sense of the term—is usually thinking of “a change in the structure, and by implication also in the habits of an animal which render it better fitted” for life. I here quote Professor D. M. S. Watson’s suggestive address on adaptation from the evolutionary standpoint.⁠[3] Sometimes the word adaptation has a more comprehensive significance and includes both definitions which I have distinguished. It then amounts to saying that living systems are self-regulating and self-propagating, which is one way of defining the nature of life as a scientific concept. None of these technical uses of the word adaptation imply anything that the most dogmatic mechanist could decry. If we define adaptation as the self-regulating processes by which living matter retains its recognizable characteristics, it is a truism to say that life is preserved by adaptation. In that case, if adaptation is to be made the paramount issue for biological enquiry, we can hardly upbraid our predecessors for presumptuously seeking to know what life is. If we are to reach any agreement about the use of the word adaptation we must therefore retrace our steps, and examine more closely what are the characteristics of a living system. It is useless to define the goal of biological enquiry in terms of a concept which is as vague as life itself. I suggest that when, in its various uses, the term adaptation has any objective utility, it refers to these two more or less distinct categories of characteristics which living beings display, i.e. self-regulating and self-propagating. They are separable issues inasmuch as a worker bee and a Dominican friar are self-regulating but not self-propagating systems. There is no particular reason to object to the use of the prefix, so long as no personalistic implications of the word self are imposed upon it without further discussion.

§2

Of the two ways in which the word adaptation is used in biological discussion, that which implies the notion of self-regulation is most fundamental. A living organism is an extremely complex system in dynamic equilibrium with its environment. The idea of dynamical equilibrium is not peculiar to biology. The atom, which for traditional chemistry was a statical concept, is no longer regarded in that way by the modern physicist. What is more peculiar about living matter is its amazing complexity, and the idea of adaptation in the sense of self-regulation calls attention to the fact that a system of such extreme complexity, a system with so many characteristics, continues to maintain its individuality, i.e., its manifold characteristics, in spite of all the changes that are taking place within it and without. The recognition of this complexity is common ground. If the mechanist underrates the difficulty of the problem, he is certainly to be discouraged, except in so far as the scientist in attacking any problem must always focus his attention on a limited range of data and rule out certain things as insignificant for his present purpose. It may be true, as Professor Lovatt Evans opines, that mechanistic interpretations tend to become arrogant and superficial. Is he on surer ground in holding that “it is unthinkable that a chance encounter of physico-chemical phenomena can be the explanation”? Might we not reflect with David Hume that “our own mind being narrow and contracted, we cannot extend our conception to the variety and extent of nature, but imagine that she is as much bounded in her operations, as we are in our speculation”?

Scientific hypotheses are not always thinkable, if by that we mean pleasant, easy or conformable to common sense. Our grandfathers thought it “inconceivable” that her Gracious Majesty, Queen of Great Britain and Ireland, Empress of India and Defender of the Faith, could be descended from an ape. The atomic structure of matter was unthinkable to many people little more than a century ago. To-day the quantum atomic model is unthinkable; but we think it is the best way of interpreting the data. Given this amazingly complex system in dynamically stable equilibrium with its environment, we have to decide consistently with the fullest requirements of the problems what is the most economical way in which we can envisage its existence. Seeing that a mechanistic interpretation is evidently the most economical one, the real issue is to decide whether there are any characteristics of the complex which are inconsistent with such an attitude.

From the modern standpoint the individuality of the atom is a statistical concept. The atom is in dynamically stable equilibrium with its surroundings. It might, therefore, be argued that a Ford car is an example of a complex mechanism which is in dynamical equilibrium with its environment. This would be a superficial analogy for the order of complexity which we encounter in living matter. The molecular constitution of the parts of a Ford car is comparatively static. In the minutest parts of an organism new molecules are being built up and replacing others that have been broken down. Nevertheless, in all this astonishing panorama of microscopic revolutions which underlie the microscopic continuity of the organism we know of no events which are in conflict with the great generalizations of physical science.

On this point Professor Hill speaks with special authority, when he declares:

“Fortunately for physiology several of the generalizations of science appear to be fairly strictly true, even when applied to the living organism. Although such exact experiments are not possible on man, or animals, or plants, as may be made on non-living objects, there is little evidence—indeed, I would be bold and say there is no evidence—that such living creatures can, in any manner or degree, evade the ordinary laws of mechanics, chemistry and physics, the principles of the Conservation of Energy and Mass.... There really is no evidence that momentum and kinetic energy, that chemical transformations, that electrical and magnetic phenomena, occur in the living body in any manner, or to any extent, which differs from that obtaining in the more readily investigated non-living world.”

In the same lecture Professor Hill replies to a statement which has been frequently reiterated by vitalistic writers including General Smuts and Professor Julian Huxley. Referring to the Second Law of Thermodynamics, he says:

“Philosophically speaking, the Second Law of Thermodynamics, dealing with the limitations of the availability of Energy, is more liable to doubt. It is known to rest on a statistical basis, and when we are dealing with units, complete, self-producing, yet as invisible and intangible as the filter-passing or other micro-organisms, it is, theoretically speaking, possible that some means may be available of evading the statistical relations which govern the behaviour of larger systems. But here again we must ask for evidence—and there is none of a precise or definite character which suggests, in the least degree, that the living cell can escape the jurisdiction of the Second Law.”

We are thus forced to consider the order of complexity of the living system maintained in dynamical equilibrium with its surroundings as the essential feature which distinguishes it from non-living things. This complexity can be arbitrarily divided into many levels; but for convenience we may confine ourselves to two, the macroscopic and the microscopic. Let us be explicit about the meaning of this distinction. In the more familiar animals, we are accustomed to recognize a variety of responses to a variety of external agencies. Generally speaking in the more complex animals each kind of reactivity and each kind of receptivity is spatially localized. For instance, light impinging upon the retina evokes contraction of the pigment cells in the toes of a frog. From this macroscopic complexity of the gross architecture of the animal body arise two types of problems: first, the problems of co-ordination dealing with the way in which a disturbance recurring in some receptive area is propagated to an effector organ (gland, muscle, etc.) in some other region; and second, the problems of metabolic exchange, dealing with how the supply and distribution of sources of energy for all this display of activity is maintained. The first involves the study of the nervous impulse along the peripheral nerve fibres and through the central nervous system; it also involves the study of the internal secretions. The second involves the study of digestion and assimilation of foodstuffs, the intake of oxygen to burn up the waste products of chemical activity, and the removal of carbon dioxide, water and other products of oxidation. In contradistinction to the gross complexity of organs or populations of cells, we have to take into account the microscopic complexity of the cell itself. This presents a more general issue, because there exist many organisms whose complexity is of the same order as that of the separate cells which make up the bodies of familiar animals of visible dimensions. Two of the major problems of cell physiology concern the way in which the cell maintains its semi-permeability, and the way in which it maintains a constant renewal of chemical materials by utilizing the energy liberated in certain organic oxidations.

If we remove the magneto from a car, we can keep it intact for an indefinite period: it is fundamental to our idea of a mechanism that it can be taken to pieces and put together again. We are so accustomed to think of a leg or an arm as dependent for their activity on the rest of the body, that the conception of a living mechanism is repugnant to common sense. In the laboratory it is possible to study properties of nerve, muscle, the cell membrane, absorption of food in the gut, oxidation of nitrogenous materials in the liver, etc., as isolated events. A person who is not a biologist almost invariably expresses bewilderment when he sees the isolated heart of an animal beating regularly in a perfusion apparatus. There exists the idea that the living organism differs from a mechanical system in that the parts cannot persist without the whole.⁠[4] Behind this illusion of common sense the holistic concept of adaptation stands securely entrenched. The holistic conception implies that for living systems the part must be interpreted in relation to the whole, and not the whole from the interaction of parts. We have seen that the ultimate non-biological constituents of living matter, molecules, atoms, etc., do not behave differently when united to form a living system. In transcending this level of organization we are faced with an equally striking conclusion. The contraction of an isolated muscle preparation is essentially the same as the contraction of a muscle considered as an isolated aspect of the behaviour of the intact organism. The passage of the nervous impulse along an isolated nerve is not fundamentally different from the passage of the nervous impulse in the normal animal. The conversion of sugar into alcohol by the isolated enzyme zymase obtained from crushed yeast cells is a process like that of the conversion of sugar into alcohol by the living yeast fungus. The whole development of physiology, from the time when Haller first made an isolated muscle preparation and Spallanzani produced animal light by moistening a dessicated powder of luminescent jelly fishes, bears witness to the conclusion that the separated constituents of a living whole do not at any level of complexity behave differently from the way in which they behave as parts of a more complex order. Thus, when the fullest recognition is given to the extreme complexity of living systems, the problem of self-regulation submitted to experimental analysis does not bring forward any confirmation for the holistic view of adaptation as the creative interpolation of new irreducible properties at different levels of complexity. The holist may reply that it is one thing to take the living machine to pieces, and another thing to put it together again. Even here the analogy with the machine holds good. To graft the eye of one salamander tadpole on to the head of another individual is now a commonplace of experimental embryology. Five-legged and two-headed newts are now manufactured in the laboratory.

Self-regulation, the way in which an organism maintains a seeming continuity of arrangement in spite of the uninterrupted and ubiquitous flux of macroscopic and microscopic changes which its existence implies, defines the sense in which the term adaptation is ordinarily used by the physiologist. In contradistinction to this physiological and individual use, adaptation is employed in biological discussion in a morphological and specific sense, when we consider how one animal comes to be distinguished from another by some architectural arrangement appropriate to a particular kind of environment. In this sense the problem of adaptation has played a prominent part in the evolutionary speculations of the past century. Given the fact that organisms are not only self-regulating but self-propagating, the evolutionary theory sets out to explain how living systems come to exist in so many specific forms, and how it is that these specific forms are on the whole fitted or adapted to their respective surroundings. The qualification on the whole is highly significant. Organisms display many peculiarities of architecture which by no stretch of imagination can be regarded as necessarily fitting them better for their conditions of life. To assume that every peculiarity of structure in an animal is useful to it in the struggle for existence is a pure assumption unfounded on anything but teleological prejudice.

Adaptation in the morphological sense really includes two ideas which to some extent coalesce, and are therefore all the more readily confused. At times the word implies nothing more than viability. In this sense adaptation is the whole problem of evolution. Up to a certain point an organism must be “suited” to its environment in order to live at all. At other times adaptation is extended to mean an essential utility in every detail of the structure of an organism. This is a mischievous implication which, as will be seen later, has hindered the formation of a clear conception of the evolutionary process. Even if we could justify the belief that female peafowl are as much impressed by peacocks as are some male biologists, we have still failed to supply a criterion of survival value which has any satisfactory significance. The enthusiast who describes an adaptation is often like the advertizing manager who tells us how many customers we shall get, if we advertize with him, but is inclined to be reticent about whether the profit derived from more customers is commensurate with the fees he proposes to exact for his services. Bateson, who first applied Mendel’s principles to animals, wrote five years before the Mendelian Renaissance in terms which may still be commended to the thoughtful examination of every student of the evolutionary problems:

“Whereas the only possible test of the utility of a structure is a quantitative one, such a quantitative method of assessment is entirely beyond our powers. To know that the presence of a certain organ may lead to the preservation of the race is useless, if we cannot tell how much preservation it can effect... unless we know also the degree to which its presence is harmful, unless, in fact, we know how its presence affects the profit and loss account of the organism.” (Materials for the Study of Variation).

That animals do in fact display many structural characteristics which are in no sense useful to them is generally admitted to-day. It thus becomes as much the function of any theory of the evolutionary process to explain the origin of useless as to explain the origin of useful devices. There is a practical limit to the use of the concept of adaptation in morphology. There is a no less obvious limit to the use of the concept of adaptation in physiology. An animal is a self-regulating system up to a point; but it cannot in every contingency take arms against a sea of troubles and by opposing end them. If we could define in some general terms where this limit lies, we should be justified in speaking of a principle of adaptation in the sense that we speak of a principle of conservation of matter. The ideally self-regulating unit of living matter endowed with the secret of perpetual youth is as imaginary as the Economic Man. At present the fact that organisms cease to regulate themselves and die is as fundamental a problem of biology as the converse fact that they regulate themselves and thereby continue to live. The fact that the organism has a good deal of useless anatomical equipment seems to be as true as the fact that on the whole its anatomy is suited to the requirements of its surroundings. In whichever way we employ the term adaptation we are forced to the conclusion that it is only legitimate to speak of a principle of adaptation in the sense in which we speak of the active principle of the thyroid gland. Adaptation defines a field of problems which await solution. In that sense the concept of adaptation is as fundamental to mechanistic as to any other theories of the organism.