Russell1930 14

From Embryology
Revision as of 15:18, 3 December 2019 by Z8600021 (talk | contribs)
Embryology - 28 Mar 2024    Facebook link Pinterest link Twitter link  Expand to Translate  
Google Translate - select your language from the list shown below (this will open a new external page)

العربية | català | 中文 | 中國傳統的 | français | Deutsche | עִברִית | हिंदी | bahasa Indonesia | italiano | 日本語 | 한국어 | မြန်မာ | Pilipino | Polskie | português | ਪੰਜਾਬੀ ਦੇ | Română | русский | Español | Swahili | Svensk | ไทย | Türkçe | اردو | ייִדיש | Tiếng Việt    These external translations are automated and may not be accurate. (More? About Translations)

Russell ES. The interpretation of development and heredity. (1930) Oxford. Univ. Press.

   The interpretation of development and heredity (1930): 1 Introductory | 2 Aristotle’s ‘De Generatione Animalium’ | 3 Preformation and Epigenesis | 4 The Germ-Plasm Theory | 5 The Theory of the Gene | 6 Some Modern Epigenetic Theories | 7 Wilhelm Roux and the Mechanics of Development | 8 The Mnemic Theories | 9 Retrospect. The Use and Misuse of Abstraction | 10 The Organismal Point of View | 11 The Physiological Interpretation of the Cell Theory | 12 The Cell and the Organism | 13 The Cell in Relation to Development and Differentiation | 14 The Organism as a Whole in Development and Reproduction
Historic Disclaimer - information about historic embryology pages 
Mark Hill.jpg
Pages where the terms "Historic" (textbooks, papers, people, recommendations) appear on this site, and sections within pages where this disclaimer appears, indicate that the content and scientific understanding are specific to the time of publication. This means that while some scientific descriptions are still accurate, the terminology and interpretation of the developmental mechanisms reflect the understanding at the time of original publication and those of the preceding periods, these terms, interpretations and recommendations may not reflect our current scientific understanding.     (More? Embryology History | Historic Embryology Papers)

XIV The Organism as a Whole in Development and Reproduction

The point of departure for every approach to the prob J- lems of inheritance and development is given by the fact that the germ is originally a single cell, equivalent in all ot its essential features to any one of the tissue-cells of which the body is composed.’ This dictum of Wilson’s 1 well represents the current state of opinion as to the proper method of interpreting heredity and development. It is generally held that these are essentially cellular phenomena, and that the clue to their understanding is to be found in the study of the cell and of the constituents of the cell. Every organism begins life as a single cell; development is brought about by the multiplication of this original egg-cell, and by the subsequent interaction and differentiation of the multitude of cells so formed. From this point of view the proximate task of developmental physiology is to analyse the processes of development in terms of cell-activities, the cell being regarded as the fundamental morphological and physiological unit.

We have seen in the last three chapters that the cellular view of development requires correction and modification. It expresses only part of the truth. Its method is abstract and analytical, and it leaves out of account the fundamental truth that the developing egg is from the very beginning a unitary organism — at first monoenergid, later polyenergid. This primary unity of the developing organism the cellular theory ignores. We have seen (Chap. IX) that the method of analytical abstraction resolves the organism into a coordinated assemblage of parts, and invests the parts so distinguished with a certain independence and separateness which they do not in reality possess. From these conceptualized ‘parts’ it is, we have seen, impossible fully to reconstitute the original unity from which they have been derived by abstraction. The whole cannot be completely explained from the parts into which conceptual analysis resolves it; the unity of the developing organism is therefore not explained on the cellular view, which has the inherent faults of the uncorrected analytic method.

1915, p. 980.


Quite apart from these theoretical considerations, the facts of development clearly show that the germ is primarily an organism, and only incidentally a cell. The evidence is conclusive that segmentation and differentiation are two distinct processes ; cell-formation often facilitates differentiation, but differentiation may go on quite independently of it. The egg as a whole, whether in its monoenergid or its polyenergid form, shows a primary differentiation as organism, and this is independent of the degree to which it is subdivided by the process of segmentation. In respect of this primary organismal differentiation the polyenergid blastula is approximately the equivalent of the fertilized monoenergid ovum.

We have further seen that the formation of discrete cells is not a universal phenomenon in the organic realm. Early development may be syncytial, and adult organisms that are polyenergid but unicellular are known. Hence multiplication of nuclei is a more general phenomenon than multiplication of cells, and a ‘nuclear’ theory is of wider application than a cell-theory. We have been able, following Sachs, to suggest a reason for the multiplication of nuclei and their dispersal throughout the body — it is because they are necessary for all processes of constructive metabolism, particularly of growth and repair. They can apparently act effectively only if their surface is large in proportion to their volume, that is, given their spherical form, if their absolute size is small; they cannot function save in continuity with the cytoplasm, and as a rule their sphere of influence is strictly limited spatially. On the energid theory the developing and adult organism must be regarded as essentially a protoplasmic whole, just as it is in the egg-stage. The number of energids necessary to maintain the vital metabolic activities is in the main a function of the size of the organism — large organisms are necessarily polyenergid, small organisms may be monoenergid. It is irrelevant to the energid conception whether the cells of the organism are spatially distinct from one another or not. Cell-formation is something superadded to multiplication of nuclei; it has an important function to perform in facilitating differentiation — no very high degree of organization is reached by non-cellular organisms, even when their nuclei are numbered in thousands — but it still remains something secondary.

For all these reasons it seems clear that for the cellular theory of development, which contains no more than a part of the truth and that expressed abstractly, there must be substituted an organismal conception, which shall take account of those broader characteristics of development with which the cellular theory is powerless to deal. We have already gathered the first fruits of the organismal method in that it has shown us the inadequacy of the cellular theory of development.

2. If the cellular conception of development fails by reason of ignoring the essential unity of the organism from the eggcell onwards, how much more does the nuclear or chromosomal theory fall short of completeness and adequacy. If development is not primarily a matter of cells and their activities, it is still less a matter of the chromosomes and their presumed organizatory powers.

We have in previous chapters outlined our general criticism of the germ-plasm theory, the gene theory, and the underlying conception of nuclear dominance and control. Into detailed criticism of the special theories we need not enter again, but it is desirable to consider once more the general chromosome theory, to discuss what its shortcomings are, and also to utilize what amount of positive truth it contains in the building up of a more adequate, less abstract, conception of development and heredity.

According to the chromosome theory there is contained in the nucleus an organized structure, built up of discrete selfperpetuating units, which constitutes the material basis of heredity and is the main determining cause of development. Suitable general environmental conditions are admittedly necessary for development, and particular conditions may be essential for the appearance of particular characters — duplicity of legs, for example, in Drosophila is dependent not only upon a ‘mutant gene’ but upon low temperature. But in general the environment does not account in any way for the specificity of development — this is entirely due to the nuclear or genic organization. Even the cytoplasm, according to the theory of nuclear dominance, plays a very minor role in the essential determination of development; Morgan roundly asserts that ‘. . . except for the rare cases of plastid inheritance, the inheritance of all known characters can be sufficiently accounted for by the presence of genes in the chromosomes. In a word, the cytoplasm may be ignored genetically’. 1 We may recall in this connexion the dictum of McClung that the chromosomes represent the sum total of the elements of control over vital processes (see above, p. 1 54).

If inheritance is thus determined by the chromosomes aloqe, it follows that the course of development is also so determined, for inheritance is merely an expression for similarity or quasi-identity of developmental process.

As a moderate and judicial statement of the chromosome theory let us take the following passage from Wilson: 2 ‘All the available evidence indicates that the nucleus is indeed a kind of “original preformation” in which are contained great numbers of self-perpetuating, definite entities grouped in a definite though shifting pattern. Their nature is unknown. They are conceivably single molecules of nucleoproteins, in which the protein complex may perhaps be the determining element to which the different genes owe their several specific characters. Heredity may thus be determined fundamentally by the chemical constitution of the various proteins.’ It is, however, more probable, Wilson considers, that the heredity units are molecular aggregates which multiply by division. Even when characters appear to be

1 T. H. Morgan, ‘Genetics and the Physiology of Development’, American Naturalist , lx, 1926, p. 491. 2 1925, p. mi.


preformed in the cytoplasm, as in germinal pre-localization, the most likely explanation is that they have been formed under the influence of the chromosomes. It would in Wilson’s opinion be ‘highly misleading to state that the “embryo in the rough” is determined solely by the cytoplasm. The cytoplasmic characters of the ovum are themselves the product of biparental heredity, even though they may be determined before the sperm enters’ (p. 1108). Wilson’s matured and reasoned view is ‘that in respect to a great number of characters heredity is effected by the transmission of a nuclear -preformation which in the course of development finds expression in a process of cytoplasmic epigenesis’ (p. 1 1 12).

These quotations from Wilson leave no doubt as to the essentials of the chromosome theory. The basis of heredity and the determining factors of development are internal to the organism and internal to its cells. There is a mechanism inside the nucleus to which are due the visible transformation of the organism during development and that stereotyped repetition of the course of development which we call heredity. This theory of an internal and invisible mechanism that dominates and controls the outer manifestations of vital activity is stated with peculiar clearness in a recent paper by Conklin. ‘We know’, he writes, ‘that the organism consists of machines within machines. The inner machine in every cell is the nucleus, usually containing two sets of chromosomes and genes, any one set of which is capable of giving rise to an entire organism if it is not prevented by the outer machine consisting of cytoplasm and the products of differentiation.’ 1

We have earlier on discussed in general terms this theory of an internal mechanism, and noted the curious analogy between the formative germ-plasm and the entelechy of Driesch. In both theories there is evident a dualism of agent and material; the germ-plasm, even in its modern genic form, is something which itself remains unaltered while acting as the cause of visible change in the organism. Aristotle would have recognized in this almost mystical conception something strangely like his ‘soul’! We cannot however enter farther into the curious methodological and psychological problems here involved. We are concerned primarily with the validity and adequacy of the chromosome theory of development and heredity as an explanation of these phenomena, and we must pass on to consider the evidence on which it is based, and the logical processes by means of which it has been built up.

1 E. G. Conklin, ‘Problems of Development’, American Naturalist , lxiii, 1929, p. 30.



3. The strongest point in favour of the chromosome theory has been the fact that in fertilization there is a union of exactly equivalent male and female pronuclei. This is the essential thing in normal fertilization. The egg and the sperm may be utterly different in size and in structure, yet they are exactly equivalent in respect of their nuclei. It appears on the face of it obvious that the paternal and the maternal contributions to the hereditary equipment of the offspring are approximately equal. Hence, it is argued, each contribution must be carried by the only exactly equivalent structures in the gametes, namely the chromosomes. Hence the chromosomes represent the ‘material basis of heredity’.

In illustration of this usual train of reasoning let us take a passage from each of two recent and well-known text-books of cytology. The first from Doncaster : 1

‘we know that in general the father and mother contribute equally to the hereditary features of their children, and yet in Man the whole of this almost infinitely complex web of family likeness is transmitted, on the father’s side, by a spermatozoon about one five-hundredth of an inch in length’ (p. 207).

The second, more elaborate, from Agar : 2

‘The fact that, on the average, offspring inherit with approximately equal intensity from both parents, whereas the macrogamete as a whole is nearly always enormously larger (often a million or even more than a billion times larger) than the microgamete, immediately suggests that the hereditary substratum is not the substance of the

1 L. Doncaster, Cytology , Cambridge, 1920,

W. E. Agar, Cytology with Special Reference to the Mammalian Nucleus , London, 1920.


gametes as a whole, but some special portion of them which is of more approximately equal mass in the two cells. This consideration led Nageli in 1884 to postulate two substances in the gametes, one of which is present in equal amount in the micro- and macro-gamete. This is the bearer of hereditary qualities — the idioplasm. The other has mainly a nutritive function and is present in far greater amount in the egg and is indeed responsible for its larger size. Knowledge of the processes of fertilization naturally led to the idioplasm being identified with the nucleus (independently by O. Hertwig and Strasburger in 1884), since the nuclear substance appears to be the only one that is contributed in approximately equal amounts by the two gametes. Mere study of the anatomy of the gametes therefore at once leads us to suspect the all-importance of the nucleus and the essential passivity of the cytoplasm in the transmission of hereditary qualities’ ( P . 154). 1

We have already exposed the weak spot of this commonly accepted argument (see above, p. 68), which lies in the fact that it is quite impossible to prove that the maternal and paternal contributions are in fact equal, that the offspring inherits from both sides a full complement of hereditary potentialities. When we say that a child shows a hereditary likeness to its father, we mean that it resembles its father more closely than it does the average of the population. The likeness is observable in respect of those individual characteristics that distinguish the father from the rest of his race. If the child shows no particular resemblance to its father, but takes after its mother, we say that it has inherited specific hereditary traits from the mother — those traits or some of them that distinguish the mother as an individual from other individuals of the same race. If by exception the child showed no particular resemblance to either of its parents we should then say that there was no specific paternal or maternal inheritance at all. But yet in all three cases the child would show the characteristics of its species and its race — it would be a human child, distinguishable as belonging to the same racial type as its parents.


1 This passage is interesting also as showing how essentially morphological and abstract is the separation of nucleus and cytoplasm.


This general resemblance in type may then be distinguished from the special resemblance to one or both of the parents, or to more remote ancestors of the direct line. It is these special resemblances and differences that have been the subject of the modern study of heredity, whether by biometrical or by genetical methods. The broad general resemblances of type give no hold for experimental or statistical treatment, and have accordingly on the whole been ignored. But it is this general hereditary resemblance which constitutes the main problem.

We saw in discussing the gene theory that it deals only with differences between closely allied forms, and with the modes of inheritance of these differences ; it leaves the main problem quite untouched as to why, for example, from a pair of Drosophila only Drosophila arise. It takes for granted the inheritance of Johannsen’s ‘great central something’ — the general hereditary equipment of the species.

Now there is no means of telling, even by breeding experiments, whether the general hereditary equipment comes from one parent only or from both. In intra-specific crosses all the main characters, those of the phylum, the class, the order, the genus, and the species, are common to both parents; what characters in the offspring can be definitely ascribed to one or other parent can only be those that distinguish one parent from the other, characters therefore of less than specific rank. In inter-specific crosses the differences may be larger, and the respective contributions may thus be more readily distinguishable, but even here the overwhelming majority of important characters are common to both parents. The fact that true crosses between quite diverse types are impossible to realize imposes of course a strict limit on this mode of analysis.

To put the matter in another way — if in a normal cross the maternal and paternal contributions were equivalent, they would be almost identical, that is to say, they would have almost everything of importance in common. It would be possible then to relate a character in the offspring to the maternal or the paternal side only if this character were different in the male and the female parent. Such differences could by the nature of things be only minor differences — either differences in minor characters, or minor differences in major characters. Exactly the same result in crossing would be obtained if the main hereditary 'potentialities came from one parent only , and the other parent merely modified in its own direction the fundamental hereditary constitution . It is therefore impossible to be certain that the main hereditary equipment comes from both parents — it might equally well come from one only, and that of course, from other evidence, the female.

The ‘axiom’ that the maternal and paternal contributions are equal is therefore no axiom at all, but a begging of the question, and the argument from this supposed axiom falls to the ground. The correctness of the chromosome theory is therefore not logically established by this line of argument, and the question as to the material basis of heredity (if there be such a thing) remains completely open.

Though most authors have accepted uncritically this faulty chain of reasoning, one or two have realized its lack of cogency, notably Winkler . 1 Regarding the supposed equivalence of the paternal and maternal contributions in heredity, Winkler writes :

‘The union of nuclei is only part of the process of fertilization, and its closer fathoming, in all those cases where pure isogamy is not found, has led to the recognition of the fact that the two parents contribute quite unequally, both as regards amount and as regards quality, to the formation of the cell which is the origin of the new individual. It would certainly be natural to conclude from this that the two parents were unequally concerned in the transmission of the genotype. If the opposite conclusion is commonly drawn, this is by reason of the petitio principii that both parents are concerned in heredity in approximately equal measure. As proof of this there is advanced the intermediate character of many hybrids. But obviously this merely shows that both parents take part in general to an equal degree in the transmission of those characters in which they differ from one another. The behaviour of hybrids however can shed no light upon the question as to how and by what means there is brought

1 H. Winkler, ‘Ueber die Rolle von Kern und Protoplasma bei der Vererbung’, Zts . indukt. Abstammungslebre , xxxiii, 1924, pp. 238-53.


about the reappearance in the offspring of all those characters in respect of which the parents are alike . It is possible that both parents are equally concerned in the appearance of these characters also; but it is equally possible that the primordia for the development of these characters are transmitted by one parent only’ (p. 239).

Winkler concludes from this that the common ‘proof* that the hereditary characters are borne exclusively by the chromosomes lacks solid foundation. He himself considers that the cytoplasm must contain the genes responsible for general as distinct from special heredity. These cytoplasmic genes form a homozygous ‘Grundstock’, which is answerable for the appearance of the more important and fundamental characters of the organism. The nuclear genes have to do only with the transmission of relatively unimportant and superficial peculiarities, such as obey the Mendelian laws.

It follows from this way of looking at things that ‘the two kinds of gametes are not of equal significance for heredity, but to a high degree unequal, inasmuch as an important part of the genotype is transmitted only by the female germ-cell. . . . And if we ascribe to the female germ-cell so much greater importance for the transmission of heritable characters than to the male germ-cell, this does not mean that a greater effect is exercised by the mother than by the father on the genotypical particularity of the offspring. For what the mother alone transmits to the child consists only of those things in which she is identical with the father; it is just the continuity of cytoplasm solely in the female line that brings about complete identity of the cytoplasms in all descendants, whether they are of the male sex or the female. Even if the father contributed on his side cytoplasm to the constitution of the zygote, he could only transmit approximately the same characters as the mother. And that would be superfluous* (p. 252).

4. We can take this fundamental question of the equivalence or non-equivalence of the gametes a stage farther by considering the point of view developed by Brachet. 1 He draws the same distinction as we have done above between general resemblance and special resemblance.

‘In the development of every individual*, he writes, ‘two distinct groups of hereditary tendencies are concerned; the first constitute 1 A. Brachet, VCEuf et les Facteurs de l* Ontogenise, Paris, 1917.


together general heredity , the second may be united under the name special , or more properly, personal , heredity. . . . General heredity is the totality of the causes, factors and laws, thanks to which a fertilised egg gives rise to an individual of the species to which it belongs. Its composition, the wheels of its mechanism, limit the egg to two possibilities only — regular development or death. 1 The egg possesses this essential part of the hereditary patrimony, this capability of building up a new organism according to definite laws, but all the facts so far known unite to prove that the spermatozoon is devoid of this power. ... As for special heredity, it comprises that which, in each individual egg, is added on to general heredity; it is that which gives the “personal turn” to heredity, if one may use the expression. This it is which gives to the development of each individual, or rather of all the individuals born of the same parents, a particular impress, sometimes easy to discover, sometimes on the contrary indistinct and obscure. General heredity and special heredity are never superimposed; there is never any clash between them, and the latter is in effect only a special case of the former. Accordingly it is incorrect to limit, as some have done, the action of general heredity to the first phases only of development, conceiving that it yields place progressively to special heredity; the latter impresses its personal seal from the beginning, provoking differences, minimal it is true, but none the less real. . . . These premisses being agreed, it is easy to conceive that in the development of a fertilised egg, the heredity contribution of the spermatozoon affects only the special side of heredity ’ (pp. 176-8).

The egg is, as we have seen, demonstrably the organismto-be, the new organism in its simplest, monoenergid state ; it is so, not in respect of its chromosomal constitution, but as a whole , as a functioning nucleo-cytoplasmic system. We know further, from the numerous examples of natural and artificial parthenogenesis which observation and experiment have revealed, that the egg contains within itself the full potentialities of development, and can in fact develop to a fully adult stage without the help of the spermatozoon. The case is c^uite different with the spermatozoon. It cannot develop into a new organism, in spite of the fact that on the chromosomal theory it possesses all the necessary genes. All

1 Cf. Delage, p. 80 above. There is much similarity of thought between Delage and Brachet.


the experiments made to induce spermatozoa to develop their own potentialities in a suitable cytoplasmic medium have failed. We have no right then to assume that the spermatozoon does in fact possess the complete hereditary equipment of the species; the plain facts teach otherwise, and it is only theoretical and poorly based considerations that would persuade us otherwise.

It is doubtful even whether the spermatozoon can properly be called a cell, for it is probably incapable of constructive metabolism, growth or division ; it can certainly not rank as an organism, for it cannot maintain itself for any length of time as an active unity; it is merely a mechanism for transporting a nucleus. The egg and the sperm are therefore, as plain observation shows us, completely unlike — as unlike as two ‘cells’ can possibly be. The only equivalence is between the nucleus of the sperm and the nucleus of the egg. In fertilization the nucleus of the sperm is added to and becomes an integral part of the nucleus of the ovum. The ovum is not by this process essentially changed ; it remains the organismto-be, with perhaps some re-arrangement of its cytoplasm as the first step in development (and this start in development can be brought about by other means than the entrance of the sperm) ; it may be modified as regards its special heredity by the introduction of the male chromatin, but there is no reason to assume, having regard to the similarity between parthenogenetic and sexual development, that its general hereditary potentialities are seriously altered, or added to.

This is the general view of the relative importance of egg and sperm which we are led to adopt if we regard only the facts and eschew the theoretical interpretation offered by the germ-plasm theory. It is in general outline the view adopted by Brachet, with the sole difference that he prefers to think of the egg as a cell — the ‘type-cell’ of the species — rather than as a monoenergid organism. Brachet lays much stress on the fact, which we have already indicated, that in respect of their developmental potentialities the sperm and the ovum are totally dissimilar. His argument is as follows. If we place the sperm in a suitable nutritive medium, for example an extract of crushed eggs of the same species, as has been done by De Meyer, we do not find that it can assimilate and organize this material into an egg nor initiate any developmental process. At most the sperm head swells up and forms a sort of pronucleus. Again, experiments on intra-specific merogony merely show that the male pronucleus is the exact equivalent of the female pronucleus, and can replace it in development. Giard has attempted to regard merogonic development as a case of male parthenogenesis, but this possibility is negatived by the results of merogonic crosses between animals belonging to different families, and by the facts of polyspermic development in Anura. In the wellknown experiments of Godlewski, in which an enucleate fragment of an Echinus egg was fertilized by Antedon sperm, the early development is purely maternal; the sperm remains inactive and exercises no influence on development. It is clear from such cases that the cytoplasm of the egg is much more than a nutritive medium for the nucleus, for it determines by its own powers 1 the early course of development.

The case of polyspermy in Anura is even more demonstrative. Under certain conditions, as when sperm are supplied in high concentration, the egg of Rana can be penetrated by a considerable number of spermatozoa at once. The sperm heads turn into typical pronuclei; one of them unites with the pronucleus of the egg, while the others, scattered throughout the substance of the egg, develop asters and divide up the egg between them into so many zones of influence, forming as many ‘spermatic energids’ as there are male pronuclei present. Now when development starts, all the nuclei of the polyspermic egg enter simultaneously into typical bipolar mitosis. Each male pronucleus acts exactly like the nucleus of a segmentation cell, and the egg segments and develops as a whole. If the spermatozoon were really totipotent, if it imposed on the surrounding cytoplasm a development determined by its own powers, each energid would act on its own, and the development of the egg as a whole would be quite chaotic. On the contrary, in successful

1 Provided some sort of nucleus is present, or at least nuclear matter.


cases, perfectly normal larvae are obtained, capable of hatching and living for several days (p. 125). If each sperm nucleus really contained the potentialities of complete development, we should expect each to initiate the formation of an embryo; instead of which each fits in with the others and co-operates in the developmental processes of the egg as a whole.

‘One essential conclusion emerges from these experiments; it is in contradiction, not only with the idea of Giard, but also with all the theories that maintain the potential equivalence of the male and female gametes. It may be formulated as follows: the destiny of a spermatozoon, or more generally of a pronucleus or even of any nucleus whatsoever, is absolutely dependent on the quality of the cytoplasm in which it finds itself situated. It is an agent of division, it plays undeniably a role in the carrying out of cellular metabolism, and it can even, eventually, impress on the cytoplasm certain characters of minor importance, but that is all. In sum ... we may say that in the present state of our knowledge the spermatozoon of the Metazoa possesses neither virtually, nor even in a latent state, the potentialities necessary for the formation of an organism similar to that from which it is itself derived. Experiment has so far proved the correctness of this notion, whose biological significance is considerable; the objections that may be raised to it, if they have no other basis than a priori reasoning, cannot upset it’ (pp. 126-7).

While the spermatozoon by itself is quite incapable of development, the case is quite different with the egg. Experiment clearly proves that the egg has no need of any contribution from the sperm in order to develop normally. We can safely deduce that the potentialities of the egg differ fundamentally from those of the spermatozoon, that the egg is really totipotent, and that the union of the two pronuclei, with the consequent re-establishment of the normal number of chromosomes, is not a condition sine qua non of the development of the egg (p. 13 1).

5. It may be well here to refer to the famous experiments of Boveri (1889) on the fertilization of enucleate fragments of Spbaerechinus eggs by the sperm of Parechinus or Paracentrotus. The original experiments appeared to show that the resulting pluteus larva might exhibit purely paternal characters, and this result was hailed as affording conclusive proof of the primacy of the nucleus in the determination of hereditary characters. Boveri however at a later date submitted his original conclusions to very close and critical scrutiny; he went to immense pains to repeat his experiments on an adequate scale, tracking down every possible source of error. In a remarkable paper 1 published after his death, he came to the conclusion that his original experiments were faulty, and that no definite proof existed of the determining influence on larval characters of the paternal chromatin introduced into the enucleate egg. He showed that in the original experiments the egg fragments that gave rise to plutei were almost certainly nucleate, or at least contained a portion of the egg-nucleus. Real merogony he found to be possible between Parechinus ? and Paracentrotus <J, but the larval characters in the two forms are so similar that no deduction could be drawn as to the influence of the male pronucleus. Elaborate tests of the original merogonic cross Sphaer echinus $ X Parechinus or Paracentrotus <J gave the following simple and definite result :

‘Judged by the criterion of nuclear size, the enucleate fragments of Sphaerechinus eggs after fertilisation with Paracentrotus or Parechinus sperm develope at first equally well with the nucleate fragments. After the completion of the blastula stage, however, they do not keep pace with these, and soon cease to develope. The furthest developed that we found in the three cultures had stopped short during gastrulation. Two tiny three-pointed stars were the most they produced in the way of a skeleton’ (p. 441).

Thus instead of the well-developed plutei, so often figured as the result of the merogonic cross, the real products of the fertilization of an enucleate fragment by foreign sperm were little aborted larvae which did not get beyond the gastrula stage. In his general conclusions Boveri draws attention to the fact that ‘certain very general form-relations of the developing individual are undoubtedly predelineated in the

1 Th. Boveri, ‘Zwei Fehlerquellen bei Merogonieversuchen und die Entwicklungsfahigkeit merogonischer und partiellmerogonischer Seeigelbaatarde’, Arcb. /. Ent.-Mccb. xliv, 1918, pp. 417-71.


arrangement of the egg-cytoplasm. We recognize in eggs of various groups more or less detailed plasmatic differentiations, by which the axis of the embryo, the point of gastrulation, and of the appearance of other primitive organs is already marked out’ (p. 465). It would appear, Boveri considers, from the results of his experiments, that any chromosomal equipment is good enough for the earliest development of the embryo, up to the beginning of gastrulation. But thereafter chromosomes of the same species, or of a closely allied species (as in the Parechinus x Paracentrotus merogonic crosses) are required for further development. Thus ‘with Sphaerechinus protoplasm the Paracentrotus nucleus cannot carry development beyond that first period in which we may consider formative relations between nucleus and cytoplasm to be absent. If then we mean by heredity the totality of the internal conditions which relate to the unfolding of the characters of the new individual, a much more specialized importance attaches to the protoplasm than most people have hitherto been willing to suppose, and the opinion that it should be possible to induce a sperm to develop in an artificial culture medium appears more absurd than ever’

(p. 466)*

Though Boveri still adhered to the view that the nuclear substance constitutes the material basis of heredity, he admitted that the cytoplasm might play a considerable part, and he also indicated the possibility that the nucleus might be regarded as being merely a ‘factor of development’ and not the sole bearer of the hereditary equipment (p. 467).

It is of methodological interest to note that Boveri considered the ‘Grundfrage’ or fundamental question at issue to be whether the egg-cytoplasm without its own nucleus could develop specific maternal characters. This, however, as we saw in Chapter IX, is not the correct formulation of the problem. Nucleus and cytoplasm considered in isolation from each other are abstract notions; the only real self-existent thing is the cell or cell-organism. There is clearly a danger in attempting to separate and contrast too absolutely the respective parts played by nucleus and cytoplasm in heredity and development. A false antithesis of nucleus and cytoplasm is thus set up, and their natural unity as an organic system destroyed. As we have seen, the true formulation consists in regarding the egg as the organism in its earliest form, possessing as a whole the full potentialities of development. If for the pronucleus of the egg there is substituted a male pronucleus of the same species, development goes on normally, because the original nucleo-plasmatic system or individual is reconstituted approximately as it was. If a pronucleus which is too different is substituted, development proceeds only a little way, presumably because the foreign chromatin cannot settle down with the egg-cytoplasm, and co-operate with it in metabolism and development. Actually the complete egg is the organism-to-be, and the sperm nucleus is merely a possible modifier of development, a factor conditioning its course. The sperm is then in no sense the equivalent of the egg; it is the equivalent merely of the egg-nucleus. There is no proof whatsoever that it contributes to the egg a complete set of hereditary potentialities; the evidence in fact all points the other way.

6. We can now formulate a working hypothesis of the part played by the chromosomes in heredity and development. We have seen that they are not to be regarded as the sole bearers of the hereditary potentialities. The evidence indicates clearly that they are to be considered as conditioning factors of heredity and development, not as fully determinative effectors of these processes.

It will be remembered that in Chapter X we discussed in general terms the relations between the organism, the cell, and the subordinate parts of the cell. We saw that the action of the whole could not be fully accounted for by the actions of the parts — that for example the activity of the organism as a whole was not a mere summation of the activities of its cells. The cell-activities condition and implement the action of the whole, but they are not independent elements from which the action of the whole can be derived by addition. The unity of the whole has been destroyed by the very fact of distinguishing separate, semi-independent parts or centres of activity, which have no real existence apart from their relations with the whole, and it cannot be reconstituted by summation of these abstract parts. The method of analytical abstraction introduces in fact discontinuities and separatenesses which do not exist in nature; the tissue cell, for instance, considered as an independent centre of activity, apart from its relations to the whole, is to a very large extent a fiction or conventional abstraction.

We distinguished in general the modes of action of higher and lower unities — from the mode of action of the organism as a whole down to the modes of action of those parts of the cell which, like the chromosomes, show a certain measure of independence and individuality. We came to the conclusion that the modes of action of the subordinate unities condition, both in a positive and a negative sense, the modes of action of the higher unities. Being integrated into the activity of the whole they render possible the vital manifestations of the whole, and at the same time they limit the mode of manifestation of these activities by imposing on them a particular form.

It is along these lines that we must interpret the nature and activities of the chromosomes. They are semi-independent units, possibly genetically continuous from one cellgeneration to another, if the theory of the individuality of the chromosomes is to be believed. They have obviously an important role to play in cell-metabolism ; they appear to be absolutely indispensable for all processes affecting the growth and repair of the cell; and it is highly probable that they exercise a profound influence on the general metabolism of the organism. Their exact physiological functions are not known with any certainty, but the classical experiments with enucleated Protozoan cells appear to indicate that the nucleus is the main source of the enzymes concerned in constructive metabolism. Some of the Mendelian results, especially as regards colour inheritance, also appear to point towards the conclusion that the chromosomes are the centre of production of oxidases and other enzymes. Many authors have in fact looked upon the nucleus as the main producer of enzymes and hormones. ‘How genes or chromosomes operate’, writes Wilson, ‘is unknown; but we may suspect that they, like plastids and other cytoplasmic bodies, are centres of specific chemical action, and possibly may serve for the production of soluble enzymes or hormones’ (1925, p. 1113). However this may be, there can be no doubt that their physiological influence is profound.

The fact that, owing to the exactitude of mitotic division, the chromosome complex is apparently identical in all the cells of the body seems to show that the individual nuclei exercise no direct or specific influence upon the differentiation of the cells severally containing them (see above, pp. 70-3). They are necessary for the life and maintenance of the cell, and in this sense they are the most important of the cellorgans, but that they exercise a direct or active modifying or morphogenetic action upon the cytoplasm is both unproved and improbable. Being universally distributed throughout all the living parts of the body, it is more probable that their effect is a massed one, exercised through the milieu interne upon the general metabolism of the body.

If then we may assume that the chromosomes as a whole profoundly affect general metabolism, any slight modification of one or more of them may be expected to alter in a definite and specific way most or many of the characters of the organism. This is in fact what we find in Mendelian inheritance. The evidence is fairly clear — quite apart from its interpretation in terms of the gene theory — that the individual chromosomes of a set differ from one another in their effects upon the organism; it is probable even that different parts of the same chromosome are qualitatively different from one another. It is nearly certain that the transmission of Mendelian differences is bound up with, and explicable in terms of, the distribution of particular chromosomes in reduction and fertilization. It is generally agreed by geneticists that the effect of a modification of a chromosome, the effect of a mutant gene, is widespread throughout the organism, affecting to some degree many or most of the characters of the organism, but some much more definitely and visibly than others.


This is all perfectly understandable on the view that the chromosomes normally play an important role in general metabolism, for any slight modification of the chromosomal complex, e.g. the substitution of a mutated chromosome for a normal one in fertilization, is perpetuated throughout all the cells of the body; it is therefore to be expected that the metabolic processes of the organism will be definitely modified as a whole. But because rather definite and specific differences between individuals can be tracked down to differences between their chromosomes, it does not in the least follow that the chromosomes are the sole determiners of all the characters of the organism. It is clear that the physiological functions of the chromosomes are important for the life, and hence for the development, of the organism. It follows that the characters of the organism are in part determined by the particular modes of metabolism manifested by the particular chromosomal set which is present in all its cells. This comes to light especially clearly when differences between homologous chromosomes arise, and the visible effects of such mutations can be traced by genetic experiment. The chromosomes therefore, being essential elements in general metabolism, clearly take part in the determination of the characters of the organism, but they are not the exclusive determiners; they condition, but they do not fully condition (i.e. determine), the appearance of these characters.

In heredity and development the whole nucleo-plasmatic system of the egg is involved ; the general hereditary potentialities are the possession of the egg-cell — the monoenergid organism — as a whole. The chromosomes brought in by the sperm merely modify slightly the general potentialities of the egg-cell or egg-organism; they do not duplicate these potentialities nor even transform them profoundly.

If the male chromosomes merely modify development, then the chromosomes of the egg, which experiment shows to be their exact equivalent, must also be regarded as partial determiners, as conditioning factors, not as sufficient causes or full determiners of heredity and development.

The truth of this conclusion, that the chromosomes cannot by themselves be responsible for development and heredity, may be verified by considering a further implication of the organismal point of view. If the activity of the organism as a whole is not completely reducible to the modes of action of its parts, then it follows that the modes of action of the whole, whether actual or potential, can be transmitted only by a whole, i.e. by the egg in its entirety, which at the very beginning of development is the new individual. Subordinate parts of the egg-organism can transmit only their own particular modes of action, and not the modes of action of the whole ; they cannot transmit even their own modes save as integral parts of the whole. A plastid, for example, which is transmitted en bloc in the plant ovum, carries over only its own capabilities, and these cannot be manifested except in connexion with the whole. In the same way, the chromosomes, considered in isolation, can transmit only their own modes of activity, presumably particular metabolic rhythms, and that only while they remain parts of the whole. Finally, any specific chemical substances that are handed on as such from one generation to another, transmit merely their own chemical characteristics and modes of reaction. There are thus diverse grades or levels at which transmission is possible, but only the whole, which includes all these grades, can transmit the potentialities of the whole.

From the organismal point of view, then, it is impossible that heredity and development should be accounted for by the sole action of the subordinate parts of the egg-cell, for these cannot transmit the potentialities of the whole.

7. The conception which we have reached of the role played by the chromosomes in development and heredity is, as we have seen, consonant with the facts of Mendelian inheritance, though not with the theory of the gene, which is an abstract and morphological rendering of these facts, restating them in terms of hypothetical particles. Instead of trying to separate and distinguish the action of the nucleus in development and heredity from the action of the cyto { >lasm, according to the analytic method, we have been ed, by adopting the synthetic view, to regard nuclei and cytoplasm as integral parts of one unitary system or organic individual, as being equally concerned in the life and activities of that individual. We do not consider for example, like Conklin and Loeb, that the ‘embryo in the rough’ is determined by the cytoplasm only, any more than we agree that the chromosomes are solely responsible for the finer characteristics which appear later in development. Nor do we agree with Jenkinson 1 that the cytoplasm determines the broad characters of the organism, those of its phylum, its class, order, and family, while generic, specific, and individual characters are transmitted by the nucleus. For us, nucleus and cytoplasm are indissolubly wedded in their action upon development ; the influence of the chromosomes is exerted from the very beginning, so that a mutated chromosome may modify development from the very start.

We have further drawn no hard and fast distinction between the individual in its monoenergid state, as an egg ‘cell’, and its polyenergid state as an embryo; for us it remains always one and the same organic individual. The polyenergid system which develops from the monoenergid egg we regard as a unitary individual, not as the cell-state or colony, into which the cell-theory would have us resolve it. Nor on our view is the developing organism formed and actuated by a mysterious internal and invisible mechanism concealed in the chromosomes, as the germ-plasm theory would have us believe.

The germ-plasm is indeed a totally superfluous and otiose conception. The facts do not demand it; it arises purely from theoretical considerations, which do not impose themselves, and its main effect is to introduce a maze of complexities and artificial problems into the theoretical interpretation of development and heredity.

The trend of opinion is fortunately away from belief in a fully determinative germ-plasm, and it may be hoped that when the purely Active nature of the gene is more generally realized the last traces of this scholastic notion will disappear.

1 J. W. Jenkinson, Three Lectures on Experimental Embryology , Oxford, 1917, p


Describing the historical evolution of the germ-plasm theory, Wilson writes :

Tn the light of these discoveries [by Roux and Boveri] and their later development it became evident that the nucleus cannot be thought of as composed of a single, homogeneous idioplasm or germplasm. It is a biological system , built up from a specifically organized group of different chromosomes which are themselves highly complex bodies; and it is only one part of a larger protoplasmic system, represented by the germ-cell as a whole. The term idioplasm or germplasm thus lost much of its original meaning; nevertheless it is often convenient as a collective name for all those components of the cellsystem, whether nuclear or cytoplasmic, that are transmitted from generation to generation and which embody the primary and essential factors of determination.* 1

Wilson’s matured view with regard to the influence of the chromosomes on development is in fact not so very different from that which is advocated here.

‘We here again emphasize*, he writes, ‘the conception that the cell is a reaction-system and that the whole cell-system may be concerned in the production of every hereditary trait. In practice all the purposes of experimental analysis are sufficiently met if the hereditary “units’*, “genes** or “pangens” be thought of merely as modifiers which call forth responses, this way or that, according to their specific nature. To speak of them as “determiners’* is to make use of a convenient figure of speech; but this need imply no more than that they are differentials by the use of which we are enabled accurately to analyze the observed results.* 2

That chromosome differences act as modifiers of the general course of development is the conclusion at which we have ourselves arrived in the foregoing discussion.

A word or two more about the functions of the chromosomes. It would seem that they are best regarded as embodying particular modes of metabolic activity, as being the stable producers of particular chemical substances, such as enzymes or hormones, which are all-essential for the life, growth, and differentiation of the organism . 3 The fact that

1 1925, p. 1039. a ! 9 2 5 > PP- “U-H*

Cf. D. Noel Paton, ‘The chromatin must be considered as an anabolite of protoplasmic activity, an anabolite transmitted from one generation to another, but even

'they are walled off from the cytoplasm in the interkinetic phase When they are in a sol state, and are divided with such meticulous accuracy in their gel phase at mitosis, suggests

that it '.is highly important they should be preserved essentially unchanged from one cell to another, and from one generation to the next. They may represent therefore the conservative element in the cell — the part which does not materially alter in spite of cytoplasmic differentiation. If this be^ so, they have some claim to be regarded in a general

way as important for the maintenance of specific type, of 1 specific modes of metabolism. Nuclear continuity or identity becomes from this point of view one of the most important conditions of that repetition of type which we call heredity.

As Gates put it : ‘The chromosome, as the most conservative

body m the cell, thus makes possible the phenomena of heredity.’ 1 The chromosomes may do this while not being in any Sense the full determiners of hereditary characters, nor . consisting of genes or determinants, or any other hereditary ' units. ‘The autogenetic particles which it [the chromosome] i; is assumed to contain tend with increasing knowledge to lose their ‘^representative” character and to become more purely chemical aggregations’ (ibid., p. 500). A biochemical or 1! physiological interpretation of the facts of Mendelian in, heritance is probably not far off. It will come when the

physiological functions of the chromosomes are better understood When more is known as to the function of the nucleus in the cell, and as to the effect of the nuclei as a whole on the ! general metabolism of the body. It may well be that a study I of the, kinds of effects produced by modifications of single

chrojrnosomes, or of special loci of single chromosomes, as revealed by genetic experiment, will throw much light upon the bi^in physiological functions of the chromosomes. For what we see in mutation are merely the incidental effects of

' in the germ-plasm augmented in each generation. In its increase and in its distribui tion it follows the course of hereditary inertia, and its action in each generation may ! be (1) to stimulate the chemical changes that lead to division and multiplication, and (2) to stabilize the direction of chemical change which determines development on ! ancestral lines’, The Physiology of the Continuity of Life , London, 1926, pp. 63-4.

1 R. Ruggles Gates, in Nature , vol. cxv, 1925, p. 500.


slight but definite changes in chromosomal metabolism, but from these minor effects it may be possible to gain some inkling of the major functions of the chromosomes, both as a set and as individuals. An interesting line of study is here indicated.

Our conception of heredity and development as functions or activities of the whole unitary organism, which cannot be fully accounted for by the effects produced by its separate cells, or ascribed to the separate action of its nuclei, makes, as we have seen, the concept of a determinative germ-plasm, or internal formative mechanism, entirely superfluous. The germ-plasm theory was already on its last legs when it was given a new lease of life by the theory of the gene. Now this also is becoming physiological, and we shall soon have heard the last of both these abstract and morphological concepts.

We have seen in our historical survey of theories that the unity school has always maintained, in opposition to the particularist doctrine, that the germ-plasm is a superfluous fiction, and we have quoted several writers to this effect. We may here refer to two more, one from the botanical, the other from the zoological side.

F. Noll, in a remarkable series of papers published in 1903, 1 points out how the chief theorists have tried to solve the problem of development by assuming a material and particulate basis, without however attempting to explain how the mere presence of material elements could exert a controlling influence on development. They have been forced indeed to ascribe to such abstract material units properties and powers with which they would hesitate to credit the cell as a whole. They have thought of the characters of the adult organism as being represented in the nucleus by material particles, and have conceived it necessary to assume a sort of ‘infection’ by these particles of the generically indeterminate egg-cytoplasm in order to explain the visible phenomena of development (p. 323). But the concept of material representative particles is quite unnecessary. ‘If the egg-cell of a

1 F. Noll, ‘Bcobachtungen und Betrachtungen liber embryonalc Substanz’, Biol . Centralblatt , xxiii, 1903, pp. 281, 321, and 401.

lime tree is already a young lime tree, there is no need of any idioplasm, germ-plasm, pangens, or heredity-substance to render possible its development into a lime tree; the eggcell as a whole is the heredity-substance 5 (p. 325).

Our second illustration of the unity view is taken from Brachet, who writes :

‘We have not considered it necessary, in the course of our discussion, to use the expressions “germ-plasm 55 , “idioplasm 55 or the like, borrowed from the phraseology of Weismann, Nageli, O. Hertwig and other theoricians of heredity. So too it has seemed to us unnecessary to discuss the question as to whether these substances, which are supposedly the material basis of hereditary tendencies, are represented by the chromatin of the chromosomes or by the mitochondria of the cytoplasm. As we have already said before, we consider with Delage, Conklin, Godlewski, Prenant and many others, that there is no one morphological substratum of heredity, any more than there exists, in the sexual cells or others, a substance of which it is the specific attribute. Heredity finds its total expression in the physical and chemical composition of the cells. It manifests itself in all the activities of life, from fertilisation right up to death, by the special turn it gives to these activities in each species and even in each individual. All the parts of the egg are necessary in order that the potentialities it contains may be harmoniously realised; all the substances play their part in this process, whether they exist in the form of organs or “organites 55 (nucleus, mitochondria, etc.) or in a state of ordinary or of colloidal solution.’ 1

Or again, explaining in the last pages of his book why he has paid no attention to the particulate theories of development and heredity:

‘Without exception the authors of these theories: Darwin, Nageli, de Vries, Weismann, O. Hertwig, to mention only those in the front rank, have thought it necessary, in order to explain heredity, i.e. the properties of the egg, to imagine the existence of special particles or substances to which they have given various names and whose structure even they have in some cases guessed at. For our part we do not see the need for these complications. We believe that the gemmules, pangens and other formed idioplasms, invented by these authors, give only the misleading appearance of an explanation, and prejudge as to structures which have not been observed and, in some cases at least, cannot and never will be observed.’ 1

The view which we have worked out in this book — that the egg is from the very beginning the organism-to-be — is very similar to that advocated by Noll and Brachet, and like their theories renders a germ-plasm superfluous. We ourselves prefer to state this view in its organismal form, maintaining as we do that the concept ‘organism’ is more concrete and more adequate than the concept ‘mechanism’ or ‘physicochemical system’.

The integrative view may however be stated also in terms of mechanism, if the developing organism be regarded as being from the beginning a unitary system, which increases in size and degree of differentiation. The important thing is that the developing organism shall be considered primarily as a unitary whole, and not as a mere composite of parts.

One objection may be raised to this jettisoning of the germ-plasm theory which is worth some consideration. If there is no specific substance of a complex and definite organization which is responsible for heredity and development, how does it come about that development is so extraordinarily specific, and the repetition of type so extremely exact in every detail ? Does not this point to a high degree of original preformation in the egg ? The answer to this objection is given in principle by a remark of Delage, to which we called attention when discussing his views (see above, pp. 80-3). He points out that it is only necessary to assume that the egg has a very precise organization, a very exact and specific total structure, in order to account for the specificity of development. We need not assume any extremely elaborate preformation in the egg — to do so is merely to beg the whole question, and to put into the egg what is not actually there, namely the characters of the embryo and adult which develop from it. Insistence on the need for elaborate pre-organization in the egg is indeed merely a confession of weakness on the part of the intellect, which finds it difficult to conceive how from the relative simplicity of the egg there develops the enormous complexity of the adult.

The exact repetition of type by the minute living egg is indeed an amazing thing, and it may never be fully explicable, but there is no need to imitate the old preformationists and adopt in despair the view that because we cannot understand development there is in reality no development, but merely an unfolding of complexities already present in the egg invisibly.

That the nucleus plays an important part in maintaining stability of type we have seen to be highly probable. If the nucleus represents essentially the conservative element in the cell, as the bearer of certain fundamental metabolic rhythms and the producer of certain essential enzymes and hormones, the fact that its continuing identity is accurately conserved by its exact division in mitosis indicates one means by which precise repetition of type is favoured. Another way in which the exact specificity of development is furthered may be the transmission of specific proteins. There is much evidence that allied species differ from one another in respect of their proteins, and these differences, though slight, are perfectly definite and stable. It is quite possible then that the differences between the eggs of two allied species may depend in part, but only in part, on the differences between the proteins of which they are respectively built up, and that in general the specificity of development may, to some extent and in some respects, be due to the specific nature of the proteins involved. 1

8. One of the most potent ideas which the genius of Weismann imposed upon his contemporaries was the conception of the separateness and distinctness of somatic and germinal cells. The latter, containing the full hereditary equipment of the germ-plasm, lived an isolated and independent life in the midst of the somatic cells, which, having become specialized through the qualitative disintegration of the Id, were no longer capable of reproducing the species. The contrast between the immortal germ-plasm, ensnrined

1 See the views expressed by Loeb and R. S. Lillie, pp. 88-9 above.


in the germ-cells, and the fugitive somatic plasm, was regarded as absolute. The individual organism was looked upon as passing and evanescent — as it were a mere temporary shoot from the continuous stolon of the germ-plasm. On such a conception, the value and importance of the individual was hardly apparent, save as a means of protecting and nourishing the germ-plasm. Nor was it clear on the pure germ-plasm theory why the germ-cells must undergo a long process of maturation before they become capable of reproducing the species.

Under the influence of the Weismannian conception, much attention was devoted to tracing back the germ-line (Keimbahn) to the earliest stages of segmentation, and to demonstrating wherever possible an early segregation of the germinal cells. On the modern view of the germinal substance, conceived as having its seat in the chromosomes and as being exactly divided quantitatively and qualitatively in every mitotic division, Weismann’s conception of the separateness of germ-plasm and somatic plasm is seen to have no foundation in fact. The ‘germ-plasm’ is present unaltered in every cell of the body, and, theoretically at least, every nucleus contains the complete potentialities for the development of a new individual. It follows then, to put the matter somewhat crudely, that the only difference between somatic cells and germinal cells must be a difference in their cytoplasmic organization.

It is an interesting fact in this connexion that in certain cases, where the germ-track can be followed right back to the egg, the determining factors for the primitive germ-cell are cytoplasmic. Thus in As carts the experiments of Boveri have shown that the difference established at the 4-cell stage between somatic cells with diminished chromatin and the primitive germ-cell with the full complement of chromatin depends on a cytoplasmic differentiation. 1 So too in some insects where the germ-cells are early differentiated the conditioning factor appears to be a cytoplasmic body, the s o-called germ-cell determinant.

1 See Wilson, 1925, pp. 1091-2,


It seems quite certain that there is no such thing as an unequal division of the nucleus, leading to differences being established between nuclei in respect of their composition and potentialities, and hence that there can be no distinction between germ-cells and somatic cells, so far as their nuclei are concerned. It is of course a definitely established fact that cells recognizable as primitive germ-cells are in many forms early set apart and cut off from the differentiations characteristic of the cells of the body. They pursue an evolution of their own, which culminates in their becoming mature ova or mature spermatozoa. This evolution is largely a cytoplasmic one, and results in the formation of two extreme and highly specialized types of cell. There is therefore solid ground for drawing a general distinction between germ-cells and somatic cells, quite apart from any theoretical considerations.

But this distinction also is no absolute one. The germtrack is often well defined, and may show direct continuity from the primitive germ-cells to the mature gametes; but the track can be interrupted, and germ-cells may arise de novo. Cases are certainly known where, after the complete elimination of the original line of germ-cells, new germ-cells have been formed from undifferentiated or dedifferentiated tissue cells. Two instances may suffice in illustration.

In his studies of regeneration in Planar ia, Morgan 1 found that head-ends cut off well in front of the reproductive organs could develop into complete individuals which formed new reproductive organs. One of these examined by means of sections was found to be normal in every respect and to have well developed ovaries ; others actually laid eggs. This result shows ‘that new germ-cells can develop from the somatic tissues, or at least from cells not included in the old reproductive system’ (p. 186).

Our second instance relates to the mouse. Davenport 2 has

1 T. H. Morgan, 'Growth and Regeneration in Planaria lugubris', Arcb. Ent,» Mecb. xiii, 1902, pp. 179-21 1.

2 C. B. Davenport, 'Regeneration of Ovariei in b/[\ct\Joum. Exper . Zoo/, xiii, 1925, pp. 1-12.


recently shown that when the ovary in this form is completely removed a new ovary may regenerate from the remains of the stalk or even from the peritoneum.

G.T.Hargitt, 1 who has given much attention to the origin of the germ-cells in Coelenterates and other forms, well expresses the modern view of the relation of germ-cells to body-cells as follows:

‘The dedifferentiations of cells, the probable origin of vertebrate germ cells from a new production following degeneration of primordial cells, the formation of germ cells whether primordial or definitive from such tissues as the peritoneum, the absence of demonstrable evidence of a specific germ track in Weismann’s sense — all these point to the view that germ cells are as much differentiated from the embryo as are other tissues and cells. ... It would seem proper and logical, therefore, to discard our previous views of the difference between body cells and germ cells, and the complete isolation and insulation of the latter from the influences of the body, even to discard the concept of the germ plasm and its continuity, since the germ plasm (nucleus) is equally present in every cell of the body, and the germ cells are no more continuous with the egg than all other cells’ (pp. 92-3).

We see then that the distinction between germ-cells and somatic cells, though real, must not be pressed too far; both are the result of differentiation, along two different and divergent lines.

This question of the origin and differentiation of the germcells leads us naturally to the consideration of reproduction, which must be regarded as one of the master-functions of the organism, and as in a sense the crown and aim of development.

9. Like development, reproduction has been looked at too much from the cellular point of view, and too little from the point of view of the organism as a whole. It is one of the main activities of the living thing, and its influence on the whole being of the organism is profound and far-reaching.


1 ‘Germ-Cell Origin in the adult Salamander, Diemyctylus viridescens’, Joum . Morph. Physiol . xxxix, 1924, pp. 63-111.


How then shall we regard reproduction from our synthetic or organismal standpoint ? We have got rid of the germplasm hypothesis; and we have seen that development is not merely a cellular process, but requires for its understanding the concept of the organism as a whole. Unhampered by the preconceptions of the cell-theory and the germ-plasm theory, we are in a position to look on reproduction in a different light, and in particular to bring under one general concept both sexual and asexual reproduction.

It is to K. E. von Baer that we owe the essential definition of reproduction, which will serve as a guide to our discussion, namely the taking on by a part of the potentialities of the whole (see above, p. 36). In reproduction, a part, be it egg or bud or fragment, escapes as it were from the domination or control of the whole, and develops a new individuality. This definition is comprehensive enough to cover all the modalities of reproduction.

Let us consider first asexual reproduction. Confining our attention to the animal kingdom, we may distinguish the following main types of asexual reproduction, namely, (1) division into two, as in many Protozoa and in some Metazoa, (2) the splitting off of fragments, as exemplified by some Actinians, (3) budding and strobilization as in some Coelenterates and worms, (4) sporulation, or the formation of unicellular or multicellular spores.

Of these, division into two is the simplest, and will afford us an insight into the essential nature of asexual reproduction. Let us follow what happens in the reproduction by fission of a highly organized Ciliate, such as Stylonichia. The process is described by Faure-Fremiet 1 as follows. In the very first stage of division, even before the coming of fission is indicated by a constriction of the body, new ciliary elements appear. These are four supplementary rows of marginal cirri, two anterior and two posterior, one on the right side of the peristome, below the lateral cirri, the other below the peristome, between the abdominal cirri; a new peristome finally is formed to the left, below the old one, in a sort of intracytoplasmic cavity. Later the new peristome opens to the

1 £. Faure-Fremiet, La Cinetique du Dheloppement , Paris, 1925.


exterior, with its fringe of adoral membranellae and its two undulatory membranes, endoral and paroral; then a constriction of the body indicates the coming separation of two individuals. ‘It is remarkable to observe that at this stage, if we regard the extremely exact disposition of the vibratory apparatus, three forms are superimposed on the same protoplasmic body; the form of the original individual, preserved almost entirely, and those of the two new individuals, interwoven with it. At the later stages, the cirri of the original individual, shared between the two daughter-cells, disappear, and new cirri take their place’ (pp. 20-1). The important thing to note is that there is not merely or principally a division of the existing organization into two, but a reorganization of the whole, including the reconstitution of two complete new individuals. The parts become wholes, even before separating from each other.

Such a process of reconstitution is common among Protozoa of a certain degree of complexity. There is generally some amount of dedifferentiation as a prelude to the redifferentiation of the two daughter-cells. In Flagellates, writes Calkins, 1

‘Reorganisation is indicated to some extent by those cases in which the old flagellum is absorbed. It is also evident in those forms of Chrysoflagellida, Cryptoflagellida and Euglenida which reproduce in the palmella or quiescent phases after the exudation of a gelatinous matrix, and after loss of the characteristic swimming organs. It is still better indicated by a number of flagellates in which the cytoplasmic kinetic elements, as well as the flagella, are all absorbed and replaced by new combinations in each of the daughter cells’ (p. 210).

Reconstitution of wholes from parts is most marked in Ciliates — ‘The processes through which the ciliate cell passes during division indicate that the organism is restored to a generalized condition practically equivalent to an encysted cell. Except for the cytostome the entire array of complex cortical organs is withdrawn and a new set is formed from the cortical protoplasm’ (p. 223).


1 G. N. Calkins, The Biology of the Protozoa , London, 1926.


Now exactly the same process of the reconstitution of a new whole from a part of the original organism takes place in the asexual reproduction of Metazoa. When Hydra forms a bud, or the scyphistoma of Aurelia cuts off successively little medusae, or a detached fragment of Sagartia elegans grows into a new individual , 1 there is the same reconstitution of a new whole from a part. The fact that in the Protozoa the process is carried out within a single cell, and that in the Metazoa many cells are concerned in forming the new individual, may be regarded as purely secondary, and as being in the main a dimensional difference.

The point is well put by Whitman : 2

‘In the infusoria we see most complex organizations worked out within the limits of a single cell. We often see the formative forces at work and structural features established before fission is accomplished. Cell-division is here plainly the result, not the cause, of structural duplication. The multicellular Microstoma behaves essentially in the same way as the unicellular Stentor, or the multinucleate Opalinopsis of Sepia. The Microstoma organization duplicates itself, and fission follows. The chain of buds thus formed bears a most striking resemblance to that of Opalinopsis, and the resemblance must lie deeper in the organization than cell-boundaries’ (p. 115).

In the same way, there is a close analogy between the unicellular spores produced by many Protozoa and Protophyta and the multicellular ‘spores’ of some Metazoa and Metaphyta, e.g. the gemmules of Sponges, and the gemmae of Liverworts; these are all parts separated from the whole, which have acquired the potentiality of reproducing the whole.

Brachet, who discusses in an interesting manner the question of asexual reproduction, sums up the general characteristics of this type of reproduction as follows :

‘At a given moment, and under the influence of conditions which must be determined in each particular case, one or more parts of the body of an organism — Protozoon or Metazoon — cease to function in

1 T. A. Stephenson, ‘On the Methods of Reproduction as Specific Characters in Sea Anemones’, Journ. Mar. Biol. Assoc, xvi, 1929, pp. 131-72.

C. O. Whitman, The Inadequacy of the Cell-Theory of Development', Woofs HoU Biol. Lectures for 1893, Boston, 1894, pp. 105-24.

their usual way and manifest properties, if not new at least unsuspected hitherto; they abandon the role assigned to them by their position in the organism to take up functions of an autogenetic origin which are infinitely more complex — the building up, by themselves, of a new individual, similar to the original one.’ 1

This manifestation by a part of a hidden power to produce the whole, whether it be a part of a monoenergid or of a polyenergid organism, Brachet, following Child , 2 connects up with its physiological isolation, however caused, from the general vital processes of the parent organism. We see the effects of isolation most clearly in cases where a part is separated mechanically from the whole and regenerates the whole.

Particularly suggestive are those numerous cases of regeneration where the separated part, instead of merely adding to and completing its structure, undergoes dedifferentiation and reorganization, remodelling its structure so that from the part there is formed, without any growth or cellmultiplication, a new individual of reduced size, by the process named by Morgan morphallaxis. As instances may be adduced the formation of tiny but complete individuals from small sections of Planaria , the reorganization of pieces of Pubularia , and the reconstitution of whole Clavellina from cut up pieces of the original whole. The same process of reconstitution of parts occurs in the regeneration of unicellular organisms. Apropos of such experiments, Morgan’s 3 remarks are worth quoting:

‘Numerous experiments on Protozoa and Protophyta have shown that a nucleated part is capable of forming a new individual. So far as we can see there is not in most cases, perhaps in none, the formation of new indifferent protoplasm in which the new parts are developed, but the entire piece is changed over into a complete animal or plant of smaller size. At first sight there seems to be here a marked difference between the regeneration of unicellular and multicellular forms, for


a C. M. Child, Die pbysiologiscbe Isolation von Teilen des Organismus ah Auslosungsfaktor der Bildung neuer Lebetvesen und der Restitution . Roux's Vortrage, xi, Leipzig, 1911.

T. H. Morgan, ‘Some Problems of Regeneration', Wood's Holl Biol . Lectures fir 1898 , Boston, 1899, pp. 195-207.


in the latter it is usual for a knob of new tissue to appear, and out of this the new part develops. However, Trembley saw that when a Hydra is cut longitudinally the cut edges bend in and fuse, forming a new tube of smaller diameter. Nussbaum has also observed in Hydra the rolling in and fusion of the cut edges. In both cases the new form develops without the previous formation of new tissue. In a tubularian hydroid Bickford has found that when a piece is cut from the stem the new tentacles appear in the old tissue, and I shall describe more fully below the results of some experiments on planarians which show that the old part plays an important role in the formation of the new individual. We see then that the difference between unicellular and multicellular forms is not so great as appears at first sight. . . . Do not these cases of regeneration in the multicellular form indicate that the individual is a whole in the same sense that the unicellular form is a whole ?’ (p. 196).

10. Sexual reproduction differs from asexual only in the fact that the origin of the new individual is a single cell, and that in this egg-cell, after the process of fertilization, half the chromosomes are of paternal, half of maternal, origin.

As a form of reproduction it is widespread throughout the organic realm, and it is undoubtedly of greater general significance than asexual reproduction. It results as a rule in the production of a much greater number of new individuals, and it is clearly a process economical of substance and energy in proportion to the number of new individuals started in life. Natural parthenogenesis is clearly derivative from ordinary sexual reproduction.

Apart from the two differences mentioned — unicellular origin and fertilization — there is however, as we shall see, the closest analogy between sexual and asexual reproduction, for in both the essential process is the redifferentiation of a part so that it becomes both potentially and, later, actually a whole.

All experimental work goes to show that sooner or later during the process of segmentation, according to the degree of original cytoplasmic prelocalization, all the cells lose their power to reproduce the whole when separated from their neighbours. Sometimes this power is lost at the very first division, or it may persist in certain cases till the 16- or 32 cell stage, but in no case is it retained for very long. It is clear then that if this power is lost by the segmentation cells it must be re-acquired by the cells which give rise to the ova, for the ova are essentially characterized by their power to reproduce the whole. There must therefore occur, just as in asexual development, a reconstitution or remodelling of a part or parts in such a way that these parts re-acquire the potentiality of forming a new whole. The difference between the two processes is mainly one of scale or dimensions; the bud or segment or part in asexual reproduction is usually polyenergid and of a certain size, while the ovum is as a rule relatively small (apart from its stores of yolk) and is invariably monoenergid. Even this difference disappears if we consider asexual reproduction by fission in Ciliates, where the part which is reconstituted a whole is small and monoenergid. As compared with the state of affairs in Ciliates, the process of egg-formation in Metazoa differs merely in the fact that the parts separated off are usually very numerous, and are almost invariably minute in comparison with the large polyenergid body of the parent; further, redifferentiation of structure and the re-acquirement of the potentiality to form a whole occur in Metazoa after the separation of the parts as primitive germ-cells, but to some extent before division or separation in Ciliates. But essentially, from the organismal point of view, the two processes are the same, and the differences relate to unimportant details. The egg may therefore be considered as equivalent to a uni-cellular bud.

It will be remembered that in Chapter I we indicated that the main characteristic of differentiation was the progressive limitation of the potentialities of the parts formed during the course of development. Functional differentiation entails physiological division of labour and a loss by the differentiated parts (cells or organs) of the more general potentialities possessed by the egg. The germ-cells constitute in some measure an exception to this law, for they have reacquired and preserve this original total potentiality, which was possessed by the egg-cell from which they are derived.

Brachet expresses this general difference, between the differentiated tissue cells and those cells and groups of cells that have re-acquired the potentiality of the whole, in terms of the ‘specific composition of the egg-protoplasm’, as follows:

‘Modified at the very first segmentation, but often capable of reestablishing itself in the two daughter-cells if their connection is broken, able to repeat the process, under the same conditions, in the four blastomeres formed from the first two divisions of the egg, capable of reconstitution sometimes, in certain privileged cells, at more advanced stages, this composition is lost to an increasing degree in proportion as differentiation progresses. But it is evident that it is not altered at the same time and to the same extent in all cells or in all groups of cells. There are some cells which, by reason of their origin, will very early diverge from the fundamental composition of the species-protoplasm; the connections constantly renewed which become established between them and their neighbours, the special character of their metabolism, determined by the position they occupy, and the effect of their correlations, will create in them functional structures which rapidly become fixed and indelible. This is the case for example with the nervous system in all Metazoa.

‘On the other hand, in other more favoured groups of cells, the imprint of correlations, being less accentuated, will leave open the possibility of a return to the initial stage, 1 which, being the normal state, typical of the species, must have a natural tendency to reestablish itself as soon as favourable conditions allow. There enter into this category without a doubt the mother-cells of the sexual products, the elements that form buds, gemmules, statoblasts, etc.* 2

It is interesting to note that Brachet states the matter also in terms of potentialities, j ust as we do — actually of course ‘specific composition’ and ‘total potentiality’ are equivalent expressions for the same thing. Thus he writes:

‘If the real and total potentialities of blastomeres and of the groups of cells which are derived from them were known at all stages in the development of an organism, we could predict with a high degree of probability the “virtual” or “latent” properties of the different tissues and organs of the body. One could enumerate in advance those which have produced all they are capable of, those which are capable of regeneration and to what degree, those which can form

1 Cf. Delage, p. 79 above. * 1917, pp. 308-9.


buds or gemmules, and those finally which are capable of becoming sexual cells.’ 1

Along such lines, by the use of such concepts, a rational analysis of development seems possible.

Accepting as he does, to a very considerable degree, the organismal conception of development and reproduction, Brachet naturally lays emphasis on the essential similarity between sexual and asexual reproduction ; for him, as for us, reproduction is essentially the taking on by a part, whether monoenergid or polyenergid, of the characters of the whole.

The same view is expressed by Child, who as we have seen (p. 92, above) is no believer in the classical theory of nuclear determination, which is so completely opposed to the organismal conception. In his remarkable book, Senescence and Rejuvenescence , 2 he writes :

‘We can dispense entirely with that remarkable conception, the germ plasm of the Weismannian theory, and say that germ plasm is any protoplasm capable under the proper conditions of undergoing dedifferentiation and reconstitution into anew individual of th^species. Reproduction, whether it is the process of reconstitution in a piece experimentally isolated from an animal or plant body, or the process of development from the fertilised egg, is fundamentally the same physiological process and involves both regressive and progressive changes, both rejuvenescence and senescence’ (p. 427).

Summing up our discussion at this stage, we may say that both the ovum and the bud (using the word in its widest sense) are parts of the organism which re-acquire that potentiality of forming a whole which in the first stages of development is lost by the cells and parts of the parent organism. From the organismal point of view it is comparatively unimportant that the one is monoenergid, the other polyenergid, for both are new organisms in their simplest, most undeveloped state.

In the fly Miastor the larva produces by parthenogenesis a number of viviparous young. It is hardly fanciful to see in this an actual case of budding, differing from ordinary internal budding only in the fact that the bud or embryo originates from a single cell. Actual budding of parts from group of undifferentiated cells is exemplified in the development of the imaginal disks in holometabolic insects.

1 1917, p. 310. Cf. Child, p. 92 above. * Chicago, 1915.


It is possible that sexual reproduction by means of single cells (which differentiate as organisms at first within the limits of the single cell) is historically a specialization of development from a primitive mode of asexual reproduction by fission or budding. It has many advantages over the latter, and it permits of that mingling of paternal chromosomes in the nucleus of the egg which clearly has some deep significance, judging from the predominance of sexual development, though what that significance may be is not yet fully clear.

II. Reproduction, or the formation of a new individual, begins when a cell, or a number of cells, or in unicellulars a portion of a cell, ceases to behave as a part and takes on progressively the characters of a whole. The development of ova therefore, regarded as the future individuals, begins long before they are mature, and long before they are fertilized and commence embryogeny proper. In a sense therefore, all organisms producing eggs are viviparous, even though fertilization and further development may take place outside the body of the parent, for the egg is already the beginning of the new organism.

At the same time, the production and ripening of germcells may be regarded as the final term and completion of the development of the parent organism. Hence the development of the parental generation overlaps the development of the filial generation. We have to picture then not a continuous lineage of germinal cells, giving off at intervals shoots representing successive generations, successive flowerings of the germ-plasm, but an actual new production in each generation of the generation to come. As Samuel Butler said, the hen on this view does actually make the egg and the chick, and the chick in the fullness of time makes another egg, just as its mother did before it. 1 It is in this way that we

1 The same view is held by A. Cohen-Kysper, The germ-cell also reconstitutes itself. It returns again to the same orderly determined state in which it began its must conceive the continuity of life — it is a continuity passing through the individuals of each generation, not an abstract continuity separate and apart from the lives of individuals. Eggs are parts of organisms which become reconstituted as wholes, just as buds do; they are not mere links in an endless chain of germ-cells transmitted unchanged from generation to generation.

Erom the organismal point of view, the long maturation which eggs must undergo before they can be fertilized and start embryogeny becomes understandable. It is not understandable on the germ-plasm hypothesis. If the nuclei of the primitive germ-cells — or for that matter the nuclei of any other type of cell — contain within them all the determining factors of development, why is it that they do not develop straight away ? And if it is averred that the cytoplasmic maturation of germ-cells is the work of their nuclei, of the internal and invisible mechanism of the genes, pangens or what you will, why is it that tissue cells, which contain the same complement of heredity units, do not normally manifest this reconstitution as wholes ? The period of maturation is essentially taken up by the long process of reconstitution whereby from a cell that commences as a part of the organism there is developed a new if embryonic whole.

It may be noted in passing that this life and development of the ova in the heart of the maternal organism enables them in a measure to share in the life and experience of that organism. It is true that the trend of their evolution is different from that of the tissue cells, and that they live to some degree a life apart, but the possibility cannot be excluded that they may be altered insensibly by the vicissitudes through which the mother passes, and there is here at least a slight opportunity for the transmission to the germ-cells of peculiarities acquired by the parent during the time they form part of her body. It may be difficult to conceive how the experience of the organism as a whole may be shared in by the germ in such a way as to become manifest in the course

development* (p. 18), Kontinuitat des Keimplasmas oder Wiederherstellung der Keim zelle. (Vortrag) Leipzig, 1923, 24 pp.


of its later development, but transmission at a lower level at least is easily comprehensible. One might expect that any radical alteration of the metabolism of the parent would be shared in by the germ-cells, which are bathed by the same milieu interne , and are subjected like the ‘somatic’ cells to the influence of the various substances dissolved therein. Direct effect at the chemical level on ‘soma’ and germ-cells alike, or on the germ-cells through the intermediary of the body-cells, has been often demonstrated. One need only mention in illustration the results obtained by Heslop Harrison on the artificial production and transmission of melanism in moths.

Coming back after this digression to consider in more detail the maturation of the primitive ova — we have already noted that germ-cells start from an undifferentiated or dedifferentiated state and undergo a long process of redifferentiation. The mature germ-cell is a highly differentiated cell.

So far as the female gamete is concerned, this differentiation shows a very special character. Much of it is to be interpreted as a preparation for the future embryogeny. In some cases there is, as we have seen, an actual building up of a cytoplasmic organization which foreshadows in its main lines, though in the broadest way, the organization of the future embryo; this preformed organismal structure, however, is not as a rule completed until the spermatozoon enters. But there is during the evolution of the ovarian egg a preparation for the assumption of the ‘organism-structure’ which characterizes the mature and fertilized egg. It is noteworthy by the way that the prelocalized areas of the fertilized egg are usually so arranged that on gastrulation they fall into their proper positions and relations in the developing embryo — this also is anticipatory development.

Another, process which obviously takes place in the cytoplasmic maturation of the egg, and is clearly preparatory of the future, is the accumulation in the egg of the yolk and other materials necessary for the maintenance and growth of the early embryo. Tnis is a process comparable, as Aristotle saw, with the formation of milk in the mammary glands of the Mammalia, in advance of, and in preparation for, the needs of the coming offspring. Both are typical cases of anticipatory action or anticipatory response.

Child, who has emphasized the point that the mature ovum is highly differentiated, though along different lines from the tissue cells, considers that it is essentially a senescent cell, which becomes rejuvenated through natural or artificial fertilization. Certain it is that towards the end of its cytoplasmic maturation the egg is a singularly inert and quiescent cell, with an abnormally large nucleus or germinal vesicle, and a cytoplasm often loaded and swollen with deutoplasmic material. It may perhaps be regarded as a cell which has reached the term of its development as a cell, and as being in this sense senescent. It has however vast potentialities dormant in it, which can be released or awakened by many means, natural or artificial. Activation of the egg must be regarded as the removal of the bar, whatever it is, that prevents the egg from developing farther — whether by overcoming its impermeability, or by remedying the state of semi-asphyxiation in which it is supposed by some to be, or by some other means. Segmentation, as we have seen, may be regarded as a process whereby the large and inactive nucleus of the egg is restored to activity by a division up into smaller units, which expose collectively a larger surface of action than the original whole nucleus. There is accordingly much to be said for Child’s view that fertilization, or more generally the stimulus to development, leads to a rejuvenescence of the egg. But it must be remembered that the potentialities of the egg are there already, only inhibited from expressing themselves ; the egg is not first endowed with them by the mere stimulus to development, a stimulus which may be purely mechanical, and is anything but specific.

12. In conclusion, let us consider very briefly the part played by the chromosomes in asexual reproduction. In discussing heredity and development generally, we saw that there might be distinguished various levels or grades of hereditary transmission. We reached the conclusion that the potentialities of the whole organism could be carried over or transmitted only by that which is itself a potential whole, e.g. the ovum, and that they must be carried by the ovum as a whole. Parts of the whole could transmit only their characters and potentialities as parts, and there was, we saw, something artificial and abstract in considering the potentialities of the parts in isolation from the activity of the whole. The activities of the parts can never fully explain the activity of the whole; they are to be regarded rather as conditions, both limiting and implementing that activity. The chromosomes therefore in sexual reproduction cannot be the full and sole determiners of the characters of the egg, or of the embryo, or of the adult; they can only carry over their own modes of activity, which we found reason to believe are essentially particular modes of metabolic activity.

Exactly the same considerations apply to asexual reproduction. The potentialities of the bud are not represented by the organization of the chromosome complex in each of its cells ; the developed characters of the bud are not determined exclusively or mainly by the chromosomes of its single cells, or even by the cells themselves. The bud as a whole has reacquired the potentiality of producing a new organism; the cells and their nuclei are conditions, not causes, of the realization of this potentiality. The chromosomes carry over into the new individual certain stable metabolic rhythms, certain stereotyped methods of chemical synthesis and chemical production, and represent thus a conservative element in hereditary transmission; they have nothing directly to do with the differentiation of the bud, beyond supplying the necessary means for growth. The unity of the new developing individual, the bud, has come into being somehow; we can probably discover the conditions which enable this unity to arise; but once it has arisen we must accept is as something given , and interpret the cell activities and chromosomal activities of the bud in this light, rather than attempt the hopeless task of explaining the unity of the developing bud in terms of the activities of its subordinate and abstract parts.

The general conception of reproduction to which we have been led by a simple application of the organismal method is therefore quite different from that which arises from theories of nuclear determination. It is hardly necessary to enlarge on this point, which has already been dealt with repeatedly in the course of this book. From the organismal point of view, the existence of any form of germ-plasm, of any specific substance or organization determining heredity and development, appears not only an unnecessary hypothesis but a misleading and erroneous one. The germ-plasm theory has in fact outlived its usefulness, and is now become a hindrance to the proper understanding of development and heredity. These processes must be regarded primarily as activities of the whole organism, affecting the whole organism, and carried out by the whole. Nothing but confusion will result if we attempt to ascribe these processes exclusively to the activities of certain parts of the organism — cells or nuclei, chromosomes or genes. If analysis of the organism is undertaken — and its value in increasing knowledge is fully admitted — there must follow upon this analysis, the abstract character of which must be recognized, a vigorous attempt at re-integration, a determined effort to put back all the part-processes artificially isolated by analysis into their proper relations with the whole. A concrete and comprehensive understanding of development and heredity can be attained only be keeping steadfastly in view the essential unity of the organism which is manifested in these activities.

We have in a previous chapter discussed the differences between the integrative and the analytic view, and we have steadily maintained the indispensability of an integrative or organismal conception. In these last few chapters, dealing with the cell-theory, with the problems of early development, and with the general interpretation of reproduction, we have tried to show the value of this conception. We have done no more than illustrate its use; and there remains yet much to be done in applying it to the problems of later development, and indeed to all the other general problems of biology. Mav it prove fruitful, and lead to a better understanding of the abiding mystery of life.