Proceedings of the Washington Academy of Sciences
Vol. VIII, pp. 197‑403. February 13, 1907.

ASPECTS OF KINETIC EVOLUTION
By O. F. COOK

5. THE HEREDITY CONCEPT MODIFIED BY HETERISM.

Questions are debated with the most persistence and the least profit when diverse opinions are being expressed by means of the same words. The term heredity has figured largely in evolutionary discussions ever since the time of Darwin, and yet the ideas which it represents are by no means the same in the minds of the many investigators who use it. The meanings do not vary merely in the extent of their application to related ideas. They differ fundamentally in their standpoints, and in their conceptions of the nature of the causes of evolution.

The traditional concept of heredity, the supposed production of like by like, also enters largely into the composition of the various philosophical systems of evolution, so largely, in fact, that evolution, descent and heredity are often treated as synonymous terms. Indeed, the whole subject of evolution is often summarized and crystallized into heredity, so that no further thinking is possible which does not definitely adopt or as definitely reject the heredity conceptions of the various schools of evolutionary study. The extreme views are very widely divergent, and perhaps equally remote from the truth.

On the one side is the hypothesis of environmental causation, or a direct impression or moulding of characters by external conditions; on the other side is the hypothesis of prefiguration or definite predetermination of characters by internal character [328] unit mechanisms of descent. Some regard heredity as a summary of environmental influences, and some as the result of an intracellular mechanism of predetermination, having no relation to the environment.

The environment does not form organisms, but neither can organisms be thought of correctly without bearing in mind their normal diversities and powers of individual accommodation to different external conditions, powers which are as incompatible with ideas of complete predetermination from within as they are with ideas of direct causation from without. Heredity, as signifying the succession of organisms in continuous lines of descent, is an actual fact, though as yet quite unexplained. Heredity, in the sense of a normal uniformity of organisms in species, does not exist. Instead of like producing like, the rule of heredity is that unlike produces unlike. To assist in an understanding of evolution and of the processes of descent the conception of heredity must be modified, and for some purposes entirely replaced, by a recognition of the facts of heterism, the normal inherent diversity shown by the individuals, castes and sexes of the same species. It is only when the members of a species are compared with the members of other species that they can be said to be alike. Compared with the members of their own species, all organisms are different.

Heredity and variation are not uncommonly personified as two opposing agents or "forces," the one striving to make organisms alike, the other to make them different. The late Professor Hyatt and others have even gone so far as to definitely locate all the heredity inside the organism and all the variation outside, holding that the organisms would be identical in form and structure were it not for variable external influences. The conception of heredity as an ideal uniformity is more applicable to some species than to others, but is not completely true of any. Experiment has everywhere shown that the members of the species and varieties are alike‑as far as they are alikebecause they breed together, not because they live in the same environments or because their form is definitely predetermined by an internal mechanism. The network of descent is a part of the mechanism of heredity, quite as truly as any characterunit particles can be. [329]

The character‑unit hypothesis of heredity is one of the corollaries of the environmental causation hypothesis of evolution. It seemed necessary to predicate something in addition to the observed methods and sequences of organic existence, in order to explain the evolutionary progress of species. How could the environment change the characters of organisms, and how could the changes of the characters be inherited and bring about the transformation of the characters of the species? These are the questions which Darwin sought to answer by his hypothesis of pangenesis, a migration of determinant particles from all parts of the body of the parent to the reproductive cells, so as to repeat in the offspring the modifications which the parent organism had experienced. The doctrine of pangenesis never found any support or justification in fact, since it could not he ascertained that characters caused by the environment are inherited by pangenesis or otherwise. Nevertheless, the doctrine of determinant character‑unit particles has been kept alive by the speculations of Nägeli, Weismann, and many other mathematically inclined students of evolutionary problems.

The kinetic theory does not approach the problem from this standpoint, for it finds causes of evolution in the facts of symbasic interbreeding and normal intraspecific diversity. The first significant fact in the direction of an explanation of evolution is the method of interweaving of the network of descent in which evolutionary progress is carried forward. In place of the assumption by static theories of a hypothetical mechanism of character‑determination, with an equally hypothetical result of ideal uniformity, the kinetic theory presents for our study conjunctions of lines of diverse descent and results of continued diversity of offspring.

HEREDITY IN CELL SPECIALIZATION.

The fact that the germ‑cells of the higher plants and animals are so different from those of which the various tissues and organs of the adult body are composed, has been taken to mean that they have some special function of heredity. A long series of exceedingly difficult and detailed investigations have been made in the hope of discovering these causes of development [330] which were supposed to lie hidden inside the nuclei of the reproductive cells.

If we trace back the organic series to their more simple representatives we not only find that the body cells become more like each other, but that the distinction between somatic or body cells and reproductive cells quite fades out. When the unicellular stage is reached, the problem of heredity seems largely eliminated, for here reproduction consists merely in the repeated division of cells into two equal parts, the close similarity of which appears in no way mysterious. The difference between the higher plants and animals and the lower lies in the fact that in the former the cells do not repeat indefinitely the same size, shape and structure, but are greatly diversified, though remaining joined together in colonies or compound individual organisms. Viewed in this manner it becomes apparent that there is no particular point at which this mechanical idea of heredity becomes necessary, no definite stage where the similarity of parts of a divided cell ceases to explain the facts of organic structure.

Reproduction and growth frequently figure merely as two names for the same process. Division of cells, which is reproduction among the lowest organisms, means growth in the higher. The process of conjugation of cells commonly termed sexual reproduction, need not be allowed to complicate the question of heredity, since the same stages of gradual differentiation can be traced among double‑ or conjugate‑celled organisms as among simple‑celled. Organisms which have conjugated recently do not divide differently from those which have not, though they may not be able to continue to divide indefinitely without conjugation. Among the higher compound organisms, conjugation takes place only at the unicellular stage. All the cell divisions necessary to the building up of the plant or animal body must be carried on without any readjustments of conjugate relations. To this limitation is doubtless due the fact that as organisms increase in complexity and in specialization of tissues, conjugation becomes a more and more indispensable preliminary to the reproduction of each new cell colony, or compound individual. If, for example, there could be one hundred divisions between each conjugation, this would [331] suffice for one hundred generations of unicellular organisms but might provide only one compound individual. Plants and lower animals can be grown from cuttings or will regenerate lost parts, but among the higher animals these powers of asexual reproduction gradually disappear.

Divergence from the normal may occur at any stage in the development of the individual, which also varies continuously, and not merely in the germ‑cell. If the life‑history of a very simple animal or plant be considered, the concentration of interest on one point tends to disappear. The processes of growth and the preparation for spore‑formation in such an organism as Spirogyra do not appear less interesting or less fundamental from the biological standpoint than conjugation and reproduction. Moreover, we now know that adaptations arise inside of cells as well as outside. The chromosomes and centrosomes, no less than the larval stages of insects, may prove to be resultant phenomena of evolution, rather than causal or truly primitive.

It is easy to understand how those who have approached evolution through the study of complex and specialized higher groups should be led to think of heredity as a mechanism, but if we take our standpoint at the other end of the organic creation it becomes apparent that heredity is merely a name for the fact that cell divisions by which organisms are built up follow closely similar lines in each successive generation. Organisms are not different merely because they are built of different kinds of cells, nor merely by reason of different arrangements of the same kinds of cells. Both causes of difference are present together in all the higher groups. Both kinds of differentiation have gone forward simultaneously and it need not be thought more wonderful that the cells of the same compound individual are different than that different species should be found among unicellular organisms. Indeed, heredity is most perfect when the cells formed by successive divisions are all alike. It may be deemed a departure from strict heredity when they become diversified, as in higher organisms. But whether the individual consists of a single cell or of a colony formed by many cell divisions, we are still dealing with the same [332] fact of organic repetition, and have no more reason in the one case than in the other to view heredity as the function of any special organ. We may define heredity as the property of organisms with as much propriety as the chemist treats crystallization as a property of sugar. The cells know, as it were, how to arrange themselves repeatedly into similar colonies or compound individuals, just as the molecules of a chemical compound take repeatedly the same crystal form.

The causes of crystallization and of heredity are equally unknown; we can merely expect for the future that to which the past has accustomed us. We have no better reasons for expecting to find that the adult is definitely prefigured in the germ‑cell that we have for supposing that the crystallographic forms or other properties of inorganic materials can be determined by microscopical examinations of the substances in solutions or in amorphous states. The germ‑cells with their chromosomes and other internal organs do indeed carry the organic sequence from one generation to another, but this fact gives us no warrant that they contain any parts or particles which will afford a general explanation of evolution. And even if the germ‑cells do contain some feature of special bearing upon heredity, it does not alter the probability that the results of the agencies operating in the germ‑cells are shown to best advantage in the completed organisms. Sperms and egg‑cells are themselves organisms, quite as truly as the elephants and whales, but their infinitesimal size, which kept them unknown and mysterious so long, does not warrant us in ascribing to them any gratuitous mysteries, nor in failing to appreciate that evolution is a motion of the specific network of descent.

Whatever the nature and functions of nuclear organs may be in different groups of animals and plants, we may expect that these organs and functions will find their primary explanation and relations in the evolutionary network of descent, rather than as affording an independent basis for theories of heredity. Neither the relations of individual organisms to environment, nor the possibility that germ‑cells have predetermining relations to adults, will justify us in leaving out of account the network of descent in which the evolution of species goes forward. [333]

HEREDITY AS A RESULT OF ENVIRONMENT.

The strength of the predisposition toward theories of environmental causes of evolution finds many illustrations in the controversies which have raged about the Lamarckian doctrine of direct environmental influences. Thus Professor Lankester, even when opposing Lamarck, assumes environmental influences of a character which the facts may not justify. It is shown that Lamarck was illogical in supposing that new environmental characters could be preserved by heredity and thus replace at once the effects of the "long‑continued response to the earlier normal specific conditions," but it becomes evident, even while this excellent chronological distinction is being drawn, that it rests on a conception of heredity only slightly less objectionable than that of Lamarck himself. Though making no direct reference to mechanical theories of heredity, these assumptions are such as to suggest and to justify such interpretations.

"Normal conditions of environment have for many thousands of generations moulded the individuals of a given species of organism, and determined as each individual developed and grew 'responsive' quantities in its parts (characters); yet, as Lamarck tells us, and as we know, there is in every individual born a potentiality which has not been extinguished. Change the normal conditions of the species in the case of a young individual taken to‑day from the site where for thousands of generations its ancestors have responded in a perfectly defined way to the normal and defined conditions of environment, reduce the daily or seasonal amount of solar radiation to which the individual is exposed; or remove the aqueous vapor from the atmosphere; or alter the chemical composition of the pabulum accessible; or force the individual to previously unaccustomed muscular effort or to new pressures and strains; and (as Lamarck bids us observe), in spite of all the long‑continued response to the earlier normal specific conditions, the innate congenital potentiality shows itself. The individual under the new quantities of environing agencies shows new responsive quantities in those parts of its structure concerned, new or acquired characters."¹ [334]

If the environments controlled the character‑units and thus moulded the characters of organisms we should expect to find that each environment would have its own organisms, or that all the individuals of the same species in the same environment would be alike, or at least more alike than individuals from different environments, but these results have not been attained. Sexual and other analogous differences which have been developed among the members of the same species in the same environments are vastly greater than any of the diversities which differences of environments can cause or induce. Moreover, there are nowhere in nature any constant environments which suppress or tend to extinguish the potential of adjustment. Vicissitudes are ever at hand, ready to make selections in directions of adjustability. The highest types of organic life, those which have been able to travel farthest on the evolutionary road, are those which have responded most effectively to their opportunities for learning the arts of adjustment. Neither are these responses mere passive mouldings; the powers of individual adjustment, no less than the general adaptive characters of the species have been attained by the putting forth of variations, the steps by which species travel.

Heredity, the name we have given to the mysterious power of plants and animals to follow accurately the developmental pathway of the species, and even to repeat the individual peculiarities of the parents, is more similar to memory than to any other biological phenomenon. Professor Lankester's conception of the facts implies that the hereditary memory is imposed from without, that it is stamped or moulded upon the species by the environment, and that its strength is, or should be, proportional to the time during which the environmental impression is continued. It is true that new or recent environmental reactions, or direct adaptations, are not inherited, and do not replace the older responsive characters of the species, but this fact lends no support to the doctrine of environmentally moulded heredity, for other character‑modifications do appear suddenly, and do immediately and definitely replace the earlier type of the species, as shown in numerous and well established instances of genetic variation and mutation. These modifications of [335] heredity have no doubt adequate physiological causes resident in the species, but as far as the environment is concerned they seem to be thoroughly spontaneous and fortuitous. They appear without notice and bring their own new and complete heredity with them; their very appearance signifies and consists in an abrupt modification of heredity. The environment may reject the new character and extinguish all the individuals with the modified system of heredity; it may limit heredity through selection, but it does not mould or modify heredity.

Heredity has been defined, in accordance with Professor Lankester's view, as the sum of past environments, but this statement, as usually understood, is only partial and misleading. It is true only to the extent that it means that the heredity of a species is a summary of the variations which the environments have permitted it to retain. The idea, for example, that improved environments will change the inherent characters of , backward races of mankind or of the deficient and criminal classes of our populations, as often stated by philanthropists, is founded on teleological inferences, and not on concrete observations. New environments may permit new and desirable characters to be put forth which the selection of adverse conditions has forbidden hitherto, but humanitarians seldom have patience with such time‑consuming methods of improvement. Moreover, if they were to view the subject from a biological standpoint they would soon appreciate the desirability of selecting the good stocks for further amelioration instead of wasting their efforts, relatively, at least, upon unworthy materials, in the vain hope of realizing an unnatural ideal of equality. Ethical considerations which concern only the relations of individuals and organized social bodies are often applied to racial and other questions as purely biological as those of the relations of species and subspecies in any other department of nature. Our chief duty with reference to the really backward and deficient races is to keep them from bringing about the deterioration of our own, as almost inevitably occurs when a higher race comes in contact with a lower. The qualities and standards which conduce to fitness in a higher civilization are of little or no significance in a lower, and rapidly deteriorate. This does not [336] prove that the higher qualities are caused by the environment, but only that they require certain conditions in which to develop and maintain themselves.

Environment is of the first importance to individual organisms, but the inference so widely drawn in scientific and general literature, that the environment causes and controls evolution, is essentially fallacious. It controls, in a measure, by limiting some of the avenues of advance, or by setting higher and higher requirements for continued progress, but life finds millions of different ways to solve its environmental problems. Given a particular environment and a particular selection of individuals with their hereditary qualities and habits known, and we may with confidence expect a fairly definite reaction in line with previous experiments of the same kind. But this does not mean that evolution is an environmental cul de sac. Changes are not passive merely, but kinetic. The environmental possibilities are persistently tested by many variations. Species have retained in this way the power of ameboid motion, and have thus crept over the whole face of nature, and into all the crevices.

The progress possible in a single life‑time or generation may be small, but the lesson is plain. The largest, most practical, and most precious factors of amelioration for plants, animals and men, lie in the discovery and preservation of those individuals which are in the line of evolutionary advancement for the breed — those possessing the qualities required by the environment, and which at the same time strengthen the species and help to maintain the necessary vital motion in courses of beneficial change.

THE PURITY OF GERM‑CELLS AND CHROMOSOMES.

In the search for causes of natural phenomena an important step appears to have been taken when definite quantitative relations have been established. It is not strange, therefore, that the discovery of Mendelian or "disjunctive" hybrids should have aroused much interest, and even a certain amount of excitement, among biologists. Mathematical considerations have been allowed to obscure biological facts, and Mendel's "principles of inheritance" have been declared to be as fundamental [337] and significant for biology as Dalton's law of definite proportions for chemistry. Deductions from Mendelism followed in rapid succession, such as the purity of germ‑cells, inheritance by character‑units, and the localization of these in chromosomes.

Mendelism as a phenomenon is both interesting and suggestive, but it lacks warrant as a generalization, because the conditions imposed by the experiments are as likely to be the cause of the results as the general principles of heredity alleged to have been revealed. There are, in fact, many reasons for believing that the inbreeding which is deemed an essential preliminary to experiments in Mendelism, induces the "disjunction" of the hybrids, instead of the purity of the germ‑cells or the antagonism of "dominant" and "recessive" characterunits. It is, perhaps, to be expected that Mendelism can be found whenever the conditions of the experiment can be met, but this does not prove that the phenomenon is a normal one. Still less has it been shown that Mendelism has been a contributing factor in evolution, since in Mendelian hybrids the more recently derived characters are held not to be dominant, but recessive, and would thus have the less chance of being preserved under natural conditions of unrestricted crossing.

Some writers have claimed for Mendelism a practical utility as determining the methods of procedure in breeding, and many plants and animals are being bred to learn which characters are dominant and which recessive, it being taken for granted that such facts have a fixed and definite value for each species or variety, thus enabling the results of breeding combinations to be known in advance. The utility of such knowledge is, nevertheless, negative rather than positive; it may keep the breeder from attempting the impossible, but it seldom gives him new leverage in attacking practical problems. The danger is rather that the acceptance of erroneous theories of heredity may delay his perception of facts and discourage his efforts.

It seems to be agreed by several experimental evolutionists that white fur or feathers is a recessive character; but no attempt has been made to test the general basis of this assumption by comparing interbred white mice with inbred gray mice. Albinism is one of many mutations induced by [338] inbreeding, and this debilitating process has been continued with white mice ever since the original specimens were caged, while gray mice have mostly remained at liberty until needed for breeding experiments. To overlook these historical differences is to neglect factors of known significance for those of purely hypothetical meaning.

A second series of pertinent facts commonly ignored is the frequent and perhaps general dominance or prepotency of mutations when bred upon their own immediate blood‑relations. Commercial white mice are a long standing breed, with no close and equally inbred gray relatives. To test prepotency fairly a new mutation would be required. There are numerous instances in literature, but experimenters naturally attach special importance to what happens in their own cages.

For a third experiment which might afford conclusive evidence on the pure germ‑cell theory, some of the more recently developed varieties of mice might serve. If two varieties of independent origin which had been crossed separately with mice of the ancestral type and found to mendelize, were then crossed with each other and found to revert to the parental type, experimentalists might admit that the doctrine of pure germ‑cells had been definitely disproven. The mice which in the Mendel experiments had produced pure white, yellow or black germcells would later have produced gray germ‑cells. And yet this possibility in crosses of selected domesticated varieties has been known since the time of Darwin's experiments with pigeons.

The arrangement of the chromatin granules into chromosomes, to which so much importance is ascribed, is a very temporary phenomenon. The chromosomes do not appear to retain their separate identity either during sexual fusion (mitapsis) or during vegetative growth, when the activities of the cells are bringing to expression the qualities which have been transmitted through the gametes. The diversity in number of chromosomes in closely allied species, or even in the same species, also tends to weaken our faith in the idea that chromosomes as such, or as character groups, play a very definite or determining part as governors of the form of the organic structure of the individual plant or animal. The chromosomes may prove, after [339] all, to be merely crowds of chromatin granules which are being assembled from the vegetative nucleus for mitapsis, and redistributed after mitapsis to resume the functions of control over vegetative growth.

Adult organisms, with their various characters, do develop out of germ‑cells, but until we know something more of the nature of protoplasm, there can be no certainty that the individual characters of the adult are in the germ‑cell in any such form that we can look in and find them. As well might we undertake to find in human embryos or infants the mental and moral characters of adult persons. All that we can be sure of is that the potentialities are there, but the nature, form and residence of these potentialities can be discussed only by means of abstract inferences, and are not yet accessible to the concrete imagination. This explains why the theories of hereditary mechanisms are merely philosophical or mathematical, not biological. Even if the conception were correct and it were possible to ascertain by some extension of microscopic vision that chromosomes or granules are prefigurations of adult organisms, the fact would still have little use as an explanation of heredity, or even as a working hypothesis, until we could learn, or at least imagine, how the models could build the structures. It is as though some barbarous tribe, on being visited for the first time by a modern man‑of‑war, should think to explain the structure by finding a small model of the ship in a glass case in the saloon. There would simply be two ships to explain, instead of one. Indeed, the discovery of the character‑unit mechanism has been so long and so vividly anticipated that it is not altogether unjust to mention the fact that no very definite uses for such a contrivance have been suggested.

The studies of Boveri tend to show that in one group, at least, there is a definite necessity for the presence of one full series of chromosomes to make normal development possible, but this is still very far from showing that individual chromosomes or granules correspond to different parts of the animal. A mutilation or disarrangement of the organs of the germ‑cells might well interfere with their development into normal individuals, even if the adult organism were not prefigured, preformed, [340] or prefixed, inside the reproductive cell. It is highly important, of course, that the nature and extent of all determinative relations be known, but until the nexus, the modus operandi of the process has been learned, predetermination by material particles has no special standing as a theory, especially where the resulting concept of heredity fails to accord with concrete facts, such as the need of normal heterism and free interbreeding.

To those who view the matter from the mathematical side only, it is still impossible to trove that essential changes occur in mitapsis which make the chromomeres and chromosome aggregates different from what they were before the fusion took place. Nevertheless, there are three facts of nature, universal and much accentuated among all the higher plants and animals, which these theories of construction of organisms by character‑unit mechanisms leave entirely out of account, without physiological meaning or explanation, (1) the diversity of the individual members of species, (2) the elaborate adaptations for interbreeding, and (3) the conjugation of the granules in mitapsis. The different assortments of chromosomes or granules might explain the diversity, but they show no use or reason in it. They may cause, too, the adaptive characters of interbreeding, but still for no purpose. Finally, they perform the elaborate evolutions of mitapsis, but all without result, according to these hypotheses of purity of germ‑cells or of chromosomes.

For numerical purposes it may be that all these complexities of symbasis are useless and unnecessary. The diversity of genera and species, and of the individuals inside the species, could all be worked out arithmetically if we could be provided beforehand with the determinant mechanisms and a system of permutations for combining them. But from the biological standpoint it seems equally clear that this is not the way the organisms were developed in nature. The character‑unit plan might have avoided all these unexplained and apparently unnecessary complications of heterism and symbasis. The difficulty is that, like its progenitor, the static theory of evolution by environmental causes, it seems not to be followed in the organic creation. Organisms are not naturally uniform and [341] they do not tend to stay uniform. Organisms are not naturally pure‑bred, and their tendencies are ever to be mixed more and more. This is the overwhelming testimony of the facts of nature, which the inventors of character‑unit mechanisms would do well to canvass before entering upon their labors.

Chromosomes and granules as parts of cells are morphological entities, in the sense that they exist and can be made visible by microscopical technique. It does not follow, however, that they are biological or evolutionary entities, or that they can properly be thought of as having any general evolutionary significance, except as parts or organs of cells or of organisms, which are the units of life. Moreover, as already indicated from other considerations, not even organisms can be considered units of evolution, which requires the coherent network of descent of a normally diverse, interbreeding species.

CONTACTS BETWEEN LINES OF DESCENT.

The fact that the lines of descent are joined only in reproductive cells should not be taken to mean that there is merely a single or casual contact between them, nor prevent our recognizing the possibility that the functions of the chromatin granules may be physiological rather than morphological. It is through them, evidently, that the reorganization of the protoplasm of the cells is accomplished. They represent the citadels of life, the most vital points of the cell substance. The final stage and apparent purpose of the process of conjugation is to bring them into contact with other granules from other lines of descent. The nature of this contact, whether the granules exchange particles, or renew their vital energy by molecular or other adjustments, is still unknown.

The most recent results of cytological investigation are in accord with the supposition that the ability of the higher plants and animals to lessen the number of conjugations and prolong the intervals of vegetative growth, has been attained by the development of more and more efficient methods of conjugation. A few years ago the opinion was held that the process of synapsis involved only a fusion and reduction of the number of the chromosomes; it now appears that the [342] chromosomes are not the ultimate units of the nuclear structure, but are merely aggregates of granules of chromatin. In the final stage of conjugation (mitapsis) the chromosome aggregates no longer appear distinct, but are subdivided into small clusters of granules called chromomeres. The chromomeres are strung out like beads in single file along two slender, protoplasmic threads which finally lie parallel and close together, so that the individual chromomeres can be paired off and fused with each other. Instead, therefore, of thinking of conjugation a simple bulk fusion of protoplasm or of nuclei, we must view it as involving a long line of many scores, hundreds, or even thousands, of contacts or combinations between the much smaller granule‑groups or chromomeres. Chromomeres appear, therefore, to have important physiological functions as specialized contact points in the fusion and reorganization of the protoplasm, and do not need to be thought of as bearers of hereditary character‑units.

There remains one other stage of elaboration of mathematical hypotheses of heredity, to treat the chromomeres as permanent entities of descent and deduce the infinitely multifarious diversities of individuals in nature from the infinity of combinations and rearrangements of which the chromomeres may be capable. This theory is complete and unimpeachable mathematically, but is as indefensible biologically as its predecessors; for like them it rests on the assumption that the bringing of the chromatin granules into contact in mitapsis has no significance in descent. It takes for granted that nothing of importance occurs when the granules appear to fuse, and that they separate again without mixture, interpenetration, or combination, of the granular or fluid constituents of the protoplasm.

The character‑unit assumption requires us to imagine some way in which the particular granules could create or bring about the existence or the accentuation of the particular character, whereas the other interpretation, by lines of descent, does not needlessly destroy the unity of the problem of heredity. It avoids the necessity of elaborate and gratuitous hypotheses in a field which science is scarcely prepared to enter. As in the adjoining regions of instinct and memory, it is easy to ascribe the phenomena [343] to positional or other relations of molecules or atoms of the cerebral tissues, but impossible to imagine an adequate nexus of association with the concrete facts, actions or functions. The opinion has already been recorded in another place that truly mechanical solutions of this series of problems are likely to await the recognition of additional properties of matter, which physical researches are now revealing with such startling rapidity.¹ As clearly perceived and definitely stated by Lord Kelvin, the current conceptions of physics are not adequate for the treatment of the problems of biological evolution.

The wonderful and altogether unexpected results of studies of the internal structures of cells are but poorly appreciated by those whose hopes have dwelt on the discovery of mechanisms of heredity. From the morphological standpoint it may appear that little has been obtained except to open another chapter in the vast complexity of nature. The internal organs and processes of cells have their multifarious similarities and diversities, like all other phases of organic existence. Reproduction is carried on by as many different methods as assimilation, respiration or locomotion. The great and surprising result of cytological investigation is not in learning that such diversity exists, which might have been anticipated, but in ascertaining that the evolution of the large and complex bodies of the higher plants and animals has been made possible by the evolution of superior methods of reproduction. Mechanical theorists have been so intent on finding a mechanism of heredity that they have failed to recognize the physiological significance of an improved process of conjugation.

The older idea was that reproduction, that is, the production of a new individual plant or animal, followed the conjugation or complete union of the parental germ‑cells, but it has been found that this is not true of any of the higher types of life. What has been considered conjugation among the higher groups, that is, the process in which the characters of the new organism are determined — as far as they are determined in the germs — is not a complete conjugation of the germ‑cells, but only the beginning of a conjugation which continues throughout the life of the new individual. [344]

This fact has bearing upon the conception of heredity, for it takes us another step away from the older idea of a mechanism in the cell, and shows us that the intracellular organs, which some look upon as the mechanisms of heredity, are capable of change and adaptation like other parts of organisms, and that the problem of evolution is not to be solved by the supposition that evolution is determined in advance by mechanisms of heredity.

In the lower groups the union of the gametes is completed before vegetative growth is resumed, or before the new generation begins. But in the remote ancestors of the higher groups this procedure was abandoned, and the completion of conjugation was deferred. Vegetative growth began to be carried on while the cells were still in the double, conjugating condition. If the form of the adult were strictly predetermined by the internal organs of the cell, the double‑celled organisms could have existed only as monstrous doubles of the simple‑celled organisms which are built up after conjugation is completed. But, as a matter of fact, the structures which were built up from these double, conjugating cells proved to be entirely different from those which had been built previously from simple cells. New evolutions began on entirely independent lines, without reference to the character‑units or other equipment of heredity resident in the cells of which the new structures were built. Moreover, the old form of heredity continued to be transmitted, even after new and higher types of organic structures had been intercalated into the life‑history of the primitive organism.

All the liverworts, mosses and ferns continue to build up the two different kinds of cellular structures, one during conjugation and the other after or between conjugations. The two kinds of heredity, the conjugate and the post‑conjugate, continue to run peaceably along the same lines of descent, like multiple telegraphic messages on the same wire.

Such complications do not, of course, dismay the inventors of hereditary mechanisms. Difficulty only adds zest to their ingenuity. Having invented one set of determinants, it is easy to invent another and have them working by turns, as Weismann gravely proposed in explaining the alternative heredity of [345] sexes. For the bees and ants three kinds of mechanisms were provided, and for the termites four kinds, though in reality upwards of a dozen sorts would be needed to account for the strange diversity of types found in some of the African species. And the most curious thing about the ants and termites is that the animals which exhibit the supposed results of these diverse kinds of mechanisms do not transmit them at all, but are descended independently in each generation from sexual insects. Here again it is apparent that new methods of development have been entered upon without requiring any change or displacement of the old. With the bees, at least, the heredity is not determined when the egg is laid, or even when it hatches. It is still possible for two or three days to induce the young larva to develop either into a queen or into a worker, by varying the nature and amount of food. The environment determines, evidently, which of the mechanisms shall continue in play and which retire into desuetude.

There is no need, of course, to continue the discussion in this direction; doubtless it is too long already. There are those who think only in relations of numbers and spaces; and for these mechanical forms are a necessity. But for those who approach from the biological side, who are curious to understand nature, and yet not so impatient as to accept even scientific fiction at the expense of ascertainable fact, these character‑unit mechanisms of heredity do not appear to help, but rather to hinder, clear perception and exposition.

ALTERNATIVE OR POLARIZED HEREDITY.

From the standpoint of the kinetic theory it appears possible to reconcile the proposed character‑unit phenomena of Mendelism with other facts of alternative descent, without invoking the hypothesis of character‑units and pure germ‑cells. The phenomena of heterism and symbasis, that is, normal diversity and broad‑breeding in specific groups, do not necessitate the character‑block assumption. They only require us to suppose that diversity of descent affords a certain amount of molecular tension or attraction, a polarity, as it were, between protoplasmic elements derived from the different lines of descent. [346]

There also appears to be a complete series of stages of accentuation of this polarity of descent. The most primitive condition is that of indiscriminate or unspecialized heterism, in which a character shows all degrees of expression from the lowest minimum to the highest maximum, with a preponderance at some intermediate or optimum point.

The physiological advantages of diversity of descent not only prevent the species from concentrating or stagnating on a central average or optimum point, but they often favor the development of two optima. The connecting series of character‑stages may weaken, or it may entirely disappear, except for rare abnormalities, the normal form of the species being represented by the two separated extremes. The typical and most familiar instances of specialized heterism is to be found, of course, in the phenomena of sex. The primary sexual characters are now so intricately involved with the functions of reproduction that their significance as specializations of heterism is much obscured, but large numbers of secondary sexual characters are quite functionless for any purpose thus far detected, except this of increasing the diversity of descent inside the species.

When once a species has reached the stage of sex‑differentiation, and has thus established a polarity of descent, the tendency seems to be for other specializations of heterism to group themselves with sex. The result is to give each generation the benefit of full diversity of descent, instead of losing this advantage in cases where similar individuals might breed together. No doubt it is easier, too, for a new character to join with and accentuate an already established polarity than to establish a new one for itself. Even among the plants which have not attained differentiation into separate sexes there are definitely alternative characters, and sometimes there are not merely two alternatives, or two groups, but several, and in a variety of, combinations, as in the genus Lythrum. In insects the phenomena of alternative descent reach their highest accentuation and complexity, for there they are superposed upon the sex‑differentiation. There may be two distinct forms of one of the sexes, as among the bees. In some species of termites both sexes are capable of specialization in several different [347] directions, so that more than a dozen different and distinct types of individuals may be found in the same colony, and no intermediate forms.

The equal sharing of the two sexes in these wonderful specializations of the termites is a reminder of the general fact of numerical equality between the sexes. Among the bees where the male sex is completely useless in the social economy and environmental relations of the colony, the reduction of the number of males has been accomplished only by the very remarkable specialization of the reproductive process. The sex is no longer determined by a polarity or other simple relation which would give equality of sexes, but by the queen herself, who has the power of laying at will either fertilized or unfertilized eggs, the former developing into females, the latter into males. This arrangement appears peculiar because it constitutes so radical an exception to the general rule of equality in the choice by individuals of one or the other of the two routes of development possible in all sexually differentiated species. If these relations depended upon merely mechanical arrangements or upon the relative numbers of different kinds of pure germ‑cells, we should expect the frequent occurrence of many definite deviations from equality of sexes.

Experiments have shown that in some groups of animals and even in plants the sex‑determination may be influenced by the conditions of existence, and particularly by nutrition and temperature. The changes are supposed, however, to occur in continuous series of gradations, as though brought about by general influences upon the constitution of the organism, rather than by the abrupt changes of adjustment which might be expected to result from the action of character‑unit devices.

The phenomena of Mendelism constitute an extension of the facts of alternative descent; for they show that this is not limited merely to secondary sexual characters and to the form differences of polymorphic species, but that closely similar effects can be obtained in a somewhat artificial manner, by combining domesticated varieties with properly opposed characters. Instead of producing merely averages or miscellaneous gradations of intermediates, well established and contrasted differences [348] are preserved separately, like alternative sexual differences. Instead, therefore, of considering that Mendel's Laws explain sexuality, it seems more reasonable to assimilate the Mendelian phenomena with those of normal alternative descent as shown generally in sex‑inheritance.

If the principle of alternative or polar heredity applies to Mendelism, the earlier explanations by the special characterunits, segregated in different germ‑cells, will be superfluous. The phenomena would still be abnormal, as are the conditions under which they appear, but they would no longer need to be associated with the phenomena of incompatability of chromatin, described by Guyer in sterile hybrids between diverse species.

"When germ‑cells are to be matured, before the real reduction, there is in most forms a so‑called false reduction, in which the chromosomes fuse in pairs so that there appears to be only half the normal number present, though in reality each is double (bivalent) and equivalent to two of the simple (univalent) type. The doubling of chromosomes which normally occurs at such times is frequently incomplete, or lacking, in hybrids. This is especially true if the hybrids are from widely separated species. Instead of a normal spindle bearing the usual number of bivalent chromosomes, multipolar spindles, or two separate spindles may appear, thus apparently permitting the two kinds of parental chromatin to remain apart. In the most extreme cases a complete separation may occur subsequently, the entire chromatin of one parent occupying one cell, that of the other a different cell. Such visible separations, however, only occur extensively in sterile hybrids from markedly different parent species. Fertile hybrids from closely related forms, for the most part, display spindles normal in appearance.

"In the case of these milder fertile crosses, then, where reversions follow the Mendelian law, the germinal incompatibilities must be narrowed down to the qualities themselves rather than confined to the respective germ‑plasms as a whole. These qualities must separate and each take up its abode in a different germ‑cell irrespective of whether the other qualities of that particular germ‑cell are of a different parentage or not. The cases in which the entire plasmas are segregated are then probably [349] but magnified images of what occurs among the specific qualities of the milder crosses. The interesting possibility arises that if fertile hybrids can be secured from widely different species the plasmas of which must be more incompatible than those of nearly related forms, such hybrids will give rise to offspring in which there is reversion, not only of one character, but of many or all characters in the same individual, due to a more thorough segregation of the parental germ‑plasm as a whole. In other words, the farther apart the parent species are, the more complete will be the return in any given offspring which shows reversion."¹

Instead of representing germinal incompatibility, the Mendelian phenomena may prove to be merely examples of the preservation of welcome and desirable contrasts. Nor is it unreasonable to suppose that the polarity or other form of alternative reaction is rendered more definite and intense by the process of inbreeding which is considered a necessary preliminary for the exhibition of the Mendelian phenomena. Contrary to Dr. Guyer's supposition, the "disjunction " of characters does not appear to depend upon the extent of diversity, but upon conditions of inbreeding. Experiments with Mendelism seem to succeed only with closely inbred domesticated varieties, not with wild species. Indeed, it is only among narrow‑bred domesticated varieties that materials for such experiments can be found, that is, definitely contrasted pairs or small groups of uniform characters.

SEXUALITY OF CONJUGATE ORGANISMS.

The sexual differentiation of the higher plants and animals affords another fairly definite indication that sexual and other alternative characters are determined by some such general principle as polarity, rather than by specialized character‑unit mechanisms of the reproductive cells. It is now known that the bodies of higher plants and animals are not the result of a completed conjugation of the parental sex‑cells, but are formed before [350] the conjugation is completed, and are thus a joint or conjugate product of the two germ‑cells.

The sexuality of the higher plants, known to the ancients, and to the aborigines of tropical America, reasserted by Bacon, rediscovered by Sprengel and substantiated by Muller and Darwin, has been denied on technical grounds by recent botanical writers, as a result of the prevalence of certain morphological theories of alternation of generations. This doctrine has led to the inference that the bodies of our higher flowering plants represent an "asexual generation," and it is held to be absurd to ascribe to such organisms the qualities and specializations of sexuality.

Some botanists accordingly refuse to call the stamens and pistils sexual structures, or the staminate and pistillate plants male and female, because they do not represent the same kind or stage of sexual differentiation as that shown in male and female moss‑plants or male and female fern‑prothallia. The fact remains, however, that the sexuality of such a plant as the date palm is completely analogous to the sexuality of the higher animals and of man himself. In other words, it has been proposed to deny sexuality to exactly that form of sexdifferentiation to which the word was originally applied.

The significant fact is that the sexual differentiation of organisms should have taken place on the two different planes of structural organization, both in the simple‑celled lower types and in the conjugate‑celled higher types. Indeed, there are three grades or stages of development where sexual diversification has taken place.

1. Sexual differences of the single gametic cells, as of the sperms and ova, or the pollen‑grains and the egg‑cells.

2. Sexual differences of simple‑celled gamete‑bearing structures, as of the male and female thalli of liverworts, the male and female plants of mosses, and the male and female prothallia of ferns, Isoetes, Selaginella and Equisetum.

3. Sexual differences of double‑celled or conjugate structures, as of the male and female individuals of the higher plants and animals.

Nor does the reckoning end here, for the separation and [351] diversification of the sexes has not taken place twice only among the plants, but probably hundreds of times, independently, and in different and unrelated natural groups, the ancestors of which were bisexual. Separate sexes, though well‑nigh universal among the higher animals, both arthropods and vertebrates, show, nevertheless, numberless independent specializations. In short, no tendency of evolution has been so definite and so general as that leading toward the accentuation of sexual differences. This can hardly mean anything less than that diversity of descent, to which sexuality ministers, has a general physiological importance and is not merely incidental to fortuitous collocations of character‑units. No doubt it will be found that the details of sex‑determination differ much in the different groups of animals and plants, but this will not diminish the general significance of the phenomenon.

Sex‑determination by purely mechanical means might still serve the purposes of symbasic interbreeding, but the heredity which might be due to the existence and operations of such mechanisms would not afford the basis of a complete theory of evolution. It would still be in need of an evolutionary explanation.

VEGETATIVE MODIFICATIONS OF HEREDITY.

Further reasons for preferring this idea of polar or positional relations of the ancestral hereditary elements to that of character units or determinants, is to be found in the fact that the hereditary attributes of form and structure are apparently capable of change at any time in the life‑history of the organism, and not merely at the time of conjugation when under the more mechanical theory the nature of the individual should be determined, once for all.

As a matter of fact, plants do make extensive and permanent alterations of their characters during the vegetative period. Such cases, though relatively rare, are numerous in the aggregate. The best known instances are those of bud variations or "sports," as the gardeners call them, where a single bud produces a branch as different from the others as seedgrown individuals, or more so. A bud mutation of coffee found [352] in Guatemala in 1904 showed characters often approached by seedling mutations, but somewhat more accentuated than any of the similar mutations which have been raised from seedlings.

Fasciation is, perhaps, to be looked upon as a form of bud variation, but it must rise in some instances, at least, through a derangement of the apical cells, rather than as a mutating adventitious bud. This has been observed very frequently in fasciations of asexually propagated plants like Dioscorea and Ipomoea. A normally round stem broadens gradually to several times its normal width, but retains its original thickness or even becomes thinner than before.

Another instance in which heredity, in the usual sense of the word, is suspended or set aside during vegetative growth, may be found in the familiar phenomenon of galls, where the presence of the insect parasite or the substances secreted by it, is able to cause the formation of complicated and highly specialized structures, as though new ingredients of heredity had been added.

The mutations which often occur in the first generation of plants when grown in new regions are also to be reckoned as post‑reproductive changes of the hereditary type, for while we could not be certain in any individual case, that the mutation could not have occurred if the seed had not been transferred, the very great difference in the percentage and the range of mutations which can be secured from the same stock of seed will prove that the new conditions have been an inducing cause, able to act after the planting of the seed and long after the nuclear elements have been arranged on a basis which would normally have persisted throughout the life of the individual.

The fourth type of interference with heredity during the vegetative period is that of graft hybridism. The extent to which this takes place with normal plants has not been ascertained, but the power of communicating diseased conditions has been well established in a variety of instances ranging from peach‑yellows, peach‑rosette, and the mosaic disease of tobacco, to the only slightly abnormal variegations. Mr. Luther Burbank relates also an instance in which a graft of a red‑foliaged variety of Prunus influenced the foliage and the progeny of the stock. [353]

RELATION OF HEREDITY TO HETERISM.

The recognition of normal diversity inside the species necessitates a modification of the older view of heredity which predicated an exact likeness among the members of a species. The uniformity which the older authors had chiefly in mind was that of the members of one species compared with those of another species. This is indeed a wonderful phenomenon, and it is not surprising that mechanical explanations were suggested. It was also quite to be expected that when the idea of internal "mechanisms of heredity" had arisen it should have seemed necessary to predicate a complete uniformity of individuals as the normal result of the workings of such a device. The mechanical inference was carried even to the extent of suggesting that the diagnostic characters like those enumerated in systematic manuals are each represented by one of the chromosomes or minute masses of infinitesimal granules found in the nuclei of reproductive cells.

As a matter of fact, natural species do not differ merely by six or seven formally expressed characters. They are different throughout, and the diversity does not end with the distinctions between the species, but extends to the individuals of each of the groups. Appreciating the necessity of greater flexibility for the mechanisms of descent, Mr. Walter T. Swingle suggested several years ago that the expression of characters might not depend directly or entirely upon the chromosomes or granules themselves, but upon their positional relations. This suggestion avoids all occasion of resorting to the character‑unit hypothesis, and may afford a clue to a cytological explanation of the phenomena of heterism.¹

It is not necessary to think that the granules determine the characters as such; they need be considered only as representing the characteristics of the ancestral lines of descent. It is then [354] possible to suppose that if the granules derived from a given ancestor secure a favorable position the characters of that ancestor will predominate in the new individual. In this way the characters of different ancestors might assert themselves in endlessly varied degrees, even in the offspring of the same parents, as they often do. This theory has the advantage of affording a thinkable connection between facts which otherwise appear completely mysterious. Two collateral circumstances increase the warrant for applying the suggestion to the phenomena of heterism.

It has been indicated by several observers, but most directly by Prowazek¹ that the granules of chromatin, which compose the chromosomes at the period of the conjugation, migrate, during vegetative growth, to positions at the knots of the nuclear network, as though to direct the processes of assimilation and growth. It was found by Maupas in his experiments with infusoria that continual inbreeding causes the gradual deterioration and diminution of the nucleus, as though diversity of descent were necessary to maintain the nuclear network, either by keeping up the number of granules or by enabling them to stay at the right distance apart. Such a relation would explain the known facts, to the extent of indicating a reason for heterism and a means for bringing it about.²

It is also easier to conceive of the possibility of bud‑variations under the supposition that the influences exerted by the chromatin depend upon position, rather than upon the origination of new units or upon the making of different combinations. Modifications of hereditary forms and methods of growth do occur during the vegetative period, as already stated, and may be quite as pronounced as the mutations obtained from seed. Changes capable of accounting for bud‑variations would also be adequate for the explanation of mutative variations.

Those who begin with the assumption that evolutionary progress is actuated by external causes are compelled to argue that the diversities of individual organisms arise through varied [355] environmental experiences, but the inadequacy of this conjecture is made plain by the fact that the greatest of these intraspecific divergencies, those of sexes, castes and alternating generations are obviously not subject to such an explanation. Protoplasmic arrangement, and the specializations of the organs and processes of reproductive cells, were not, of themselves, effective for the problems of advancing organization. There had to be differences, vital tensions, as it were, between the protoplasms, if organic progress were to be maintained, and conjugation were to become adequate for the building up of large, complex and long‑lived organisms.

As fission suffices for the reproduction of only the simplest types, and haplogamy, apaulogamy and finally paragamy, have proved necessary to continue the propagation of organisms of successively higher degrees of complexity, so, for the very highest, sexual diversity and continuously maintained symbasis are requisite. The effect of prolonging the process of conjugation is to double in each organism the threads of the vital network. The separation of a species into sexes is a still more advanced category of specialized descent, since it doubles the whole specific network, permits accumulation of two sets of variations, and insures that each individual be descended from two diverse parents.

But even this provision of interbreeding does not suffice to maintain the perfection of organic excellence found in man himself, where the requirement of diverse descent is so acute as to forbid, on pain of degenerate offspring, the union of individuals separated by less than four or five generations, or by two or three strains of alien blood. Human descent is so difficult and precarious a fabric that the double network cannot be held in place merely by the joining of adjacent knots. The structure is likely to totter or fall if the lines of descent which join in the building of each new individual are not well braced by meeting each other at broad angles. Neighboring parallel or only slightly divergent lines do not afford the necessary stability of contrast, the vital tension which enables the conjugate cells to build a well‑knit body. The intricacies of relationships which fascinate the genealogist are not gratuitous or [356] accidental, but are a biological necessity in the elaboration of the framework of symbasic descent which sustains the organic vigor of the species.

In cytology, no less than in the more general fields of study, it is the physiological values which need first to be ascertained, before the morphological considerations can be correctly appreciated. Germ‑cells can indeed be viewed as mechanisms of descent, but speculations regarding them should not be made the basis of evolutionary thought nor the test of orthodoxy, to the exclusion of more definite and concrete indications of the nature of evolutionary processes.

The kinetic theory finds significance and confirmation in the now rapidly accumulating indications of an extensive series of fusions between the individual granules of chromatin, which previous cytological interpretations, based on static views of evolution, have denied. From the kinetic point of view the fusions of the chromatin are an important and altogether accordant part of the whole system of evolution; they are the actual knots and junctions of the fabric of descent. Static theories of cellular determinants, on the other hand, can see in these evidences of fusion only an elaborate deception, an unnecessary complexity of the process of reproduction, just as it was formerly held that sexual reproduction itself stood in the way of evolution, because it interfered with the subdivision of species and the isolation of new variations.

The traditional concept of heredity, the ideal of uniformity in descent, has furnished the basis of all preceding doctrines of evolution. Conditions of isolation or of restricted descent have accordingly been considered typical for evolution, because it was only in narrow bred groups that the ideal of uniformity could be approximated in nature. The kinetic theory breaks with all these traditions, and seeks to substitute for the abstract conception of a uniform, definite or mechanical heredity, a recognition of the concrete fact of normal diversity, inside the species.