proceeded to develop that particular part of the embryo to which it would have given rise if the rest of the egg had not been killed. There was no regeneration of the part killed, no formation of a complete embryo. It may be pointed out that segmentation in the insect egg is peculiar. The nuclei multiplied by segmentation migrate into the superficial cytoplasm surrounding the yolk, and then this cytoplasm segments, and each part of the cytoplasm develops into a particular region of the embryo. This, of course, does not prove that the nuclei or their chromosomes do not determine the _characters_ of the parts of the embryo developed, but they show that the parts of the non-nucleated cytoplasm correspond to particular parts of the embryo. The most important object of investigation at the present time is to find the origin of these properties of the chromosomes. We may say, using the word 'determinant' as a convenient term for that which determines the adult characters, that in order to explain the origin of species or the origin of adaptations we must discover the origin of determinants. Mendelism does not throw any direct light on this question, but it certainly has shown how characters may be inherited as separate and independent units. When one difference between two breeds is considered, _e.g._ rose comb and single in fowls, and individuals are crossed, we have the determinant for rose and the determinant for single in the same zygote. The result is that rose develops and single is not apparent. In the next generation rose and single appear, as at the beginning, in separate individuals. When two or three or more differences are studied we find that they are usually inherited separately without connexion with each other, although in some cases they are connected or coupled. The facts of Mendelism are of great interest and importance, but we have to consider the general theory based on them. This theory is that characters are generally separate units which can exist side by side, but do not mingle, and cannot be divided into parts. When an apparently single character shows itself double or treble, it is concluded that it has not been really divided, but consists of two or three units (Castle). Further, although Mendelism in itself shows no evidence of the origin of the characters, it assumes that they arose as complete units, and one suggestion is that a dominant factor might at some of the divisions in gametegenesis pass entirely into one daughter cell, and therefore be absent from the other, and thus individuals might be developed in which a dominant character was absent. Bateson in his well-known books, _Mendel's _Principles of Heredity_, 1909, and _Problems of Genetics_, 1913, discusses this question of the origin of the factors which are inherited independently. The difficulty that troubles him is the origin of a dominant character. Naturally, if he persists in regarding the determinant factor as a unit which does not grow nor itself evolve in any way, it is difficult to conceive where it came from. The dominant, according to Bateson, must be due to the presence of something which is absent in the recessive. He gives as an instance the black pigment in the Silky fowl, which is present in the skin and connective tissues. In his own experiments he found this was recessive to the white-skin character of the Brown Leghorn, and he assumes that the genetic properties of _Gallus bankiva_ with regard to skin pigment are similar to those of the Brown Leghorn. Therefore in order that this character could have arisen in the Silky, the pigment-producing factor _P_ must be added and the inhibiting factor _D_ must drop out or be lost. He says we have no conception of the process by which these events took place. [Footnote: _Problems of Genetics_, p. 85.] Now my experiment in crossing Silky with _bankiva_ shows that no inhibiting factor is present in the latter, so that only one change, not two, was necessary to produce the Silky. Mendelians find it so difficult to conceive of the origin of a new dominant that they even suggest that no such thing ever occurs: what appears as a new character was present from the beginning, but its development was prevented by an inhibiting factor: when this goes into one cell of a division and leaves the other free, the suppressed character appears. This is the principle proposed to get over the difficulty of the origin of a new dominant. All characters are due to factors, and all factors were present in the original ancestor—say Amoeba. Evolution has been merely 'the rejection of various factors from an original complex, and a reshuffling of those that were left.' Professor Lotsy goes so far as to say that difference in species arose solely from crossing, that all domestic animals are of mixed stocks, and that it is easier to believe that a given race was derived from some ancestor of which all trace has been lost than that all races of fowls, for example, arose by variation from a single species, but the evidence that our varieties of pigeons have been derived from _C. livia_, and of fowls from _G. bankiva_, is too strong to be disregarded because it does not agree with theoretical conceptions.

My own experiments in crossing Silky fowls with Gallus bankiva (P.Z.S., 1919) show that the recessive is not always pure, that segregation is not in all cases complete. The colour of the bankiva is what is called black-red, these being probably the actual pigments present, mixed in some parts of the plumage, in separate areas in other parts: the Silky is white. There are seven pairs of characters altogether in which the Silky differs from the bankiva. Both the pigmented skin of the Silky and the colour in the plumage of the bankiva are dominant, so that all the offspring in F1 or the first generation are coloured fowls with pigmented skins. But in later generations I found that with regard to skin pigment there were no pure recessives. Since the heterozygote in F1 was deeply pigmented, it is certain that a bird with only a small amount of pigment in its skin was a recessive resulting from incomplete segregation of the pigmented character. The pigment occurred chiefly in the skin of the abdomen and round the eyes, and also in the peritoneum and in the connective tissue of the abdominal wall. It varied in different individuals, but in some, at any rate, was greater in later generations than in the earlier. The condition bred true, as pure recessives do; and when such an impure recessive was mated with a heterozygote with black skin, the offspring were half pigmented and half recessive, with some pigment on the abdomen of the latter.

Still more striking was the incomplete segregation in the plumage colour. The white of the Silky was recessive, all the birds of the F1 generation being fully coloured. In the F2 generation there were two recessive white cocks which when mature showed slight yellow colour across the loins. These two were mated with coloured hens, and in later generations all the recessives instead of being pure white, like the Silky, had reddish-brown pigment distributed as in pile fowls.



In the hens (Plate I., fig. 1) it was chiefly confined to the breast and abdomen, and was well developed, not a mere tinge or trace, but a deep coloration, extending on to the dorsal coverts at the lower edge of the folded wings. The back and tail were white. In the cocks the colour was much paler, and extended over the dorsal surface of the wings, where it was darker than on the back and loins (Plate I., fig. 2). These pile-coloured fowls when mated together bred true, with individual differences in the offspring.

The pile fowl as recognised and described by fanciers is dominant in colour, not recessive as in the case above described. In fact, a recessive pile does not appear ever to have been mentioned before the publication of the results of my experiment. From the statements of John Douglas in Wright's Book of Poultry (London, 1885), it appears that fanciers knew long ago that the pile could be produced from a female of the black-red Game mated with a white Game-cock. It would seem, therefore, that the pile is the heterozygote of black-red and 'dominant' white. Bateson, however (Principles of Heredity, 1909, p. 120), writes that the whole problem of the pile is very obscure, and treats it as a case of peculiarity in the genetics of yellow pigments. On p. 102 of the same volume he describes the results of crossing White Leghorn with Indian Game or Brown Leghorn, the F1 being substantially white birds with specks of black and brown, though cocks have sometimes enough red in the wings to bring them into the category known an pile. To test the matter I have crossed White Leghorns with a pure-bred black-red Game-cock, and in the offspring out of eight six were fairly good piles, but with not quite so much red on the back as in typical birds: one was a pile with yellow on the back instead of red, and one was white with irregular specks. Of the hens, four were of pile coloration with breast and abdomen of uniform reddish-brown colour, back, neck, and saddle hackles laced with pale brown, tail white. The other four were white with black and brown specks. Whether these pile heterozygotes will breed true I do not yet know.

These results tend to show that factors are not indivisible units, and segregation is rather the difficulty of chromatin or germ plasm from different race uniting together. It must be remembered that the fertilised ovum which forms one individual gives rise also to dozens or hundreds or thousands or millions of gametes. If a given character is represented by a portion of the chromatin in the original ovum, this has to be divided so many times, and each time to grow to the same condition as before. How can we suppose that the divisions shall be exactly equal or the growth always the same? It is inevitable that irregularities will occur, and if the original chromatin produced a certain character, who shall say what more or less of that chromatin will produce?

In the case of my recessive pile, my interpretation is that when the chromosomes corresponding to two distinct characters such as colour and absence of colour are formed they do not separate from each other completely. Whether the mixture of the chromosomes occurs in every resting stage of the nucleus in the successive generations of the gametocytes, or whether it occurs only in the synapsis stage preceding reduction division, it is not surprising that the colloid substance of the chromosomes should form a more or less complete intermixture, and that the two original chromosomes should not be again separated in the pure condition in which they came into contact. A part, greater or less, of each may be left mixed with the other. This is the probable explanation of the fact that the recessive white plumage has some of the pigment from the dominant form. Segregation, the repulsion between chromosomes, or chromatin, from gametes of different races may occur in different degrees from complete segregation to complete mixture. When the latter occurs there would be no segregation and the heterozygote would breed true. The most interesting fact is that a given factor in the cases I have described, namely, colour of plumage and pigmentation, of skin in the Jungle fowl and the Silky, is not a permanent and indivisible unit, but is capable of subdivision in any proportion. Bateson has already (in his Address to the Australian meeting of the British Association) expressed the same conclusion. He states that although some Mendelians have spoken of genetic factors as permanent and indestructible, he is satisfied that they may occasionally undergo a quantitative disintegration, the results of which he calls subtraction or reduction stages. For example, the Picotee Sweet Pea with its purple edges can be nothing but a condition produced by the factor which ordinarily makes the fully purple flower, quantitatively diminished. He remarks also that these fractional degradations are, it may be inferred, the consequences of irregularities in segregation.

Bateson, however, proceeds to urge that the history of the Sweet Pea belies those ideas of a continuous evolution with which we had formerly to contend. The big varieties came first, the little ones arose later by fractionation, although now the devotees of continuity could arrange them in a graduated series from white to deep purple. Now this may be historically true of the Sweet Pea, but I would point out that once the dogma of the permanent indivisible unit or factor is abandoned, there is nothing in Mendelism inconsistent with the possibility of the gradual increase or decrease of a character in evolution. I do not suggest that the colour and markings of a species or variety were, in all cases, due to external conditions, but if the effect of external stimuli can be inherited, can affect the chromosomes, then the evidence concerning unit factors no longer contradicts the possibility of a character gradually increasing, under the influence of external stimuli acting on the soma from zero to any degree whatever.

SEX AND SECONDARY SEXUAL CHARACTERS

The mystery of sex is hidden ultimately in the phenomenon of conjugation, that union of two cells which in general seems necessary to the maintenance of life, to be a process of rejuvenation. We know nothing of the nature of this process, or why in general it should produce a reinvigoration of the cell resulting from it. We know little if anything of the relation between the two conjugating cells or gametes, of the real nature of the attraction that causes them to approach each other and ultimately unite together. We have, it is true, some evidence that one cell affects the other by some chemical action, as for instance in the fact that the mobile male gametes of a fern are attracted to a tube containing malic acid, but this may be merely an influence on the direction of movement of the male gamete, while there are cases in which neither cell is actively mobile. What we know in higher animals and plants is that each gamete contains in its nucleus half the number of chromosomes found in the other cells of the parent, and that in the fertilised ovum the chromosomes of both gametes form the new nucleus, in which therefore the original number of chromosomes is restored.

The remarkable fact is that from this fertilised ovum or zygote is developed usually an individual of one sex or the other, male or female, other cases being comparatively exceptional, although each act of fertilisation is the union of the two sexes together. Various attempts have been made to prove that the sex of the organism is determined by conditions affecting it during development subsequent to fertilisation, but now there is good reason to believe that generally the sex of the individual is determined at fertilisation, though as we shall see there is evidence that it may in certain cases be changed at a later stage.

In Mendelian experiments, a heterozygote individual is one arising from gametes containing opposite members of a pair of characters, in other words, from the union of a gamete carrying a dominant with another carrying a recessive. A pure recessive individual is one arising from the union of two gametes both carrying recessives. If a heterozygote is bred with a pure recessive the offspring are half heterozygote and half recessive. The heterozygote individual in typical cases shows the dominant character. In the formation of its gametes when the reduction division of the chromosomes takes place, half of them receive the dominant character, half the recessive. When the division in the gametes of the recessive individual takes place its gametes all contain the recessive character. Thus, if we indicate the dominant character by D and the recessive by d, the constitution of the two individuals is

Dd and dd.

The gametes they produce are

D+d and d+d,

and the fertilisations are therefore

Dd, Dd, dd, dd,

or heterozygote dominants and pure recessives in equal numbers.

It is evident that the reproduction of the sexes is very similar to this. One of the remarkable facts about sex is that, although the uniting gametes are male and female yet they give rise to males and females in equal numbers. If one sex were a dominant this would be in accordance with Mendelian theory. In accordance with the view that the dominant is something present which is absent in the recessive, the Mendelian theory of sex assumes that femaleness is dominant, and that maleness is the absence of femaleness, the absence of something which makes the individual female. If we represent the character of femaleness by F and maleness or the recessive by f, we have the ordinary sexual union represented by

Ffxff;

the gametes will then be

Ff and ff

and the fertilisations

Ff and ff,

or males and females in equal numbers, as they are, at least approximately, in fact.

The close agreement of this theory with what actually happens is certainly important and suggests that it contains some truth. But it cannot be said to be a satisfactory explanation. It ignores the question of the nature of sex. According to the theory the female character is entirely wanting in the male. But what is sex but the difference between ovum and spermatozoon, between megagamete and microgamete? The theory then asserts that an individual developed from a cell formed by the union of male and female gametes is entirely incapable of producing female gametes again. Every zygote after conjugation or fertilisation may be said to be bisexual or hermaphrodite. How comes it then that the female quality entirely disappears? Whether the gametocytes are distinguishable at an early stage in the segmentation of the ovum, or only at a later stage of development, we know that the gametes ultimately formed have descended by a series of cell-divisions from the fertilised ovum or zygote cell from which development commenced. If segregation takes place at the reduction divisions we might suppose that half the gametes formed are sperms and half are ova, and that in the male the latter do not survive but perish and disappear. But in this case it would be the whole of the chromosomes coming from the original female gamete which would disappear, and the spermatozoon would be incapable of transmitting characters derived from the female parent of the individual in which the spermatozoa were formed. An individual could never inherit character from its paternal grandmother. This, of course, is contrary to the results of ordinary Mendelian experiments, for characters are inherited equally from individuals of either sex, except secondary sexual characters and sex-linked characters which we shall consider later.

Similarly, if we suppose that segregation of ovum and sperm occurs in the female, the sperms must disappear and the ovum would contain no factors derived from the male parent. But the theory supposes that the segregation of male and female does occur in the female, that half the ova are female and half are male. What meaning are we to attach to the words 'male ovum' or even 'male producing ovum'? It is a fundamental principle of Mendelism that the soma does not influence the gametocytes or gametes; we have therefore only to consider the sex of the gametes themselves, derived from a zygote which is formed by the union of two sexes. The quality of maleness consists only in the size, form, and mobility of the sperm in the higher animals and of the microgamete in other cases. In what sense then, can an ovum be male? It may perhaps be said that though it is itself female, it has some property or factor which when united with a sperm causes the zygote to be capable of producing only sperms, and conversely the female ovum has a quality which causes the zygote to produce only ova. But since these qualities segregate in the reduction divisions, how is it that the male quality in the f ovum does not make it a sperm? We are asked to conceive a quality, or the absence of a factor, in an ovum which is incapable of causing that ovum to be a sperm, but which, when segregated in the gametes descended from that ovum, causes them all to be sperms. It is impossible to conceive a single quality or factor which at different times produces directly opposite effects. The Mendelian theory is merely a theory in words, which have an apparent relation to the facts, but which when examined do not correspond to any real conceptions.

However, we have to consider a number of remarkable facts concerning the relation of chromosomes to sex. In the ants, bees, and wasps the unfertilised ovum always develops into a male, the fertilised into a female. The chromosomes of the ovum undergo reduction in the usual way, and are only half the number of those present in the nucleus before reduction. We may call this reduced number N and the full number 2N. The ova developing by parthenogenesis and giving rise to males segment in the usual way, and all the cells both of soma and gametocytes contain only N chromosomes. In the maturation divisions reduction does not occur, N chromosomes passing to one gamete, none to the other, and the latter perishes so that the sperms all contain N chromosomes. When fertilisation occurs the zygote therefore contains 2N chromosomes and becomes female. Here then we have no segregation of Fxf in the ova. The difference of sex merely corresponds to duplex and simplex conditions of nucleus, but it is curious that the simplex condition in the gametes occurs in both ova and sperms.

In Daphnia and Rotifers the facts are different. Parthenogenesis occurs when food supply is plentiful and temperature high. In this case reduction of the chromosomes does not occur at all, the eggs develop with 2N chromosomes and all develop into females. Under unfavourable conditions reduction or meiosis occurs, and two kinds of eggs larger and smaller are formed, both with N chromosomes. The larger only develops when fertilised and give rise to females with 2N chromosomes. The smaller eggs develop without fertilisation, by parthenogenesis, and become males. Here then we have three kinds of gametes, large eggs, small eggs, and sperms, each with the same number of chromosomes. It is not the mere number then which makes the difference, but we find a segregation in the ova into what may for convenience be called female ova and male ova.

In Aphidae or plant lice a third condition is found. Here again parthenogenesis continues for generation after generation so long as conditions are favourable, i.e. in summer, and the eggs are in the same condition as in Daphnia, etc., that is to say, reduction does not occur, and the number of chromosomes is 2N. Under unfavourable conditions males are developed as well as females by parthenogenesis, but the males arise from eggs which undergo partial reduction of chromosomes, only one or two being separated instead of half the whole number. The number then in an egg which develops into a male is 2N-1, while other eggs undergo complete reduction and then have N chromosomes. The latter, however, do not develop until they have been fertilised. In the males, when mature, reduction takes place in the gametes, so that two kinds of sperms are formed, those with N chromosomes and those with N-l chromosomes. The latter degenerate and die, the former fertilise the ova, and the fertilised ova develop only into females. The chief difference in this case then is that the reduction in the male to the N or simplex condition takes place in two stages, one in the parthenogenetic ovum, one in the gametes of the mature male. In Hymenoptera and in Daphnia, etc., the whole reduction takes place in the parthenogenetic ovum, and in the mature male, though reduction divisions occur, no separation of chromosomes takes place: at the first division one cell is formed with N chromosomes and one with none, and the latter perishes.

In many insects and other Arthropods which are not parthenogenetic the male has been found to possess fewer chromosomes than the female. The female forms, as in the above cases of parthenogenesis, only gametes of one kind each with N chromosomes, but the male forms gametes of two sorts, one with N chromosomes, the other with N-l or N-2 chromosomes. On fertilisation two kinds of zygotes are formed, female-producing eggs with 2N chromosomes, and male-producing eggs with 2N-1 or 2N-2 chromosomes. There is also evidence that in some cases, e.g. the sea-urchin, the female is heterozygous, forming gametes, some with N and some with N+ chromosomes, while the male gametes are all N. Fertilisation then produces male-producing eggs with 2N chromosomes, female-producing with 2N+.

Such is the summary given by Castle in 1912. [Footnote: Heredity and Eugenics, by Castle and Others. University of Chicago Press, 1912.] It will be seen that he treats the differences as purely quantitative, mere differences in the number of the chromosomes. Professor E. B. Wilson, however, who had contributed largely by his own researches to our knowledge of sex from the cytological point of view, had already published, in 1910, [Footnote: 'The Determination of Sex,' Science Progress, April 1910.] a very instructive resume of the facts observed up to that time. The important fact which is generally true for insects, according to Wilson, is that there is a special chromosome or chromosomes which can be distinguished from the others, and which is or are related to sex differentiation. This chromosome, to speak of it for convenience in the singular, has been variously named by different investigators. Wilson called it the 'X chromosome,' McCluny the 'accessory chromosome,' Montgomery the 'hetero-chromosome,' while the names 'heterotropic chromosome' and idiochromosome have also been used. For the purpose of the present discussion we may conveniently name it the sex-chromosome. It is often distinguished by its larger size and different shape. Wilson describes the following different cases:—

(1) The sex-chromosome in the male gametocytes is single and fails to divide with the others, but passes undivided to one pole. This may occur in the first reduction division (Orthoptera, Coleoptera, Diptera) or in the second (many Hemiptera). But it is difficult to understand what is meant by 'fails to divide.' In one of the reduction divisions all the chromosomes divide as in ordinary or homotypic nucleus division, but in the other the chromosomes simply separate into two equal groups without division. If there are an odd number of chromosomes, 2N-1, in all the gametocytes of the male, as stated in most accounts of the subject, then if one chromosome fails to divide in the homotypic division, we shall have 2N-2 in one spermatocyte and 2N-1 in the other. Then when the heterotypic division takes place and the number of chromosomes is halved, we shall have two spermatocytes with N-1 chromosomes from one of the first spermatocytes and one with N and one with N-1 from the other. Thus there will be three spermatozoa with N-1 chromosomes and one with N chromosomes, whereas we are supposed to find equal numbers with N and N-1 chromosomes. It is evident that what Dr. Wilson means is that the sex-chromosome is unpaired, and that although it divides like the others in the homotypic division, in the heterotypic division it has no mate and so passes with half the number of chromosomes to one pole of the division spindle, while the other group of chromosomes has no sex-chromosome. Examples of this are the genera Pyrrhocoris and Protenor (Hemiptera) Brachystola and many other Acrididae, Anasa, Euthoetha, Narnia, Anax. In a second class of cases the sex-chromosome is double, consisting of two components which pass together to one pole. Examples of this are Syromaster, Phylloxera, Agalena. In a third class the sex-chromosome is accompanied by a fellow which is usually smaller, and the two separate at the differential division. The sizes of the two differ in different degrees, from cases as in many Coleoptera and Diptera in which the smaller chromosome is very minute, to those (Benacus, Mineus) in which it is almost as large as its fellow, and others (Nezara, Oncopeltus) in which the two are equal in size. Again, there are cases in which one sex-chromosome, say X, is double, triple, or even quadruple, while the other, say Y, is single. In all these cases there are two X chromosomes in the oocytes (and somatic cells) of the female, and after reduction the female gametes or unfertilised ova are all alike, having a single X chromosome or group. On fertilisation half the zygotes have XX and half XY, whether Y is absence of a sex-chromosome, or one of the other Y forms above mentioned. The sex is thus determined by the male gamete, the X chromosome united with that of the female gamete producing female individuals, while the Y united with X produces male individuals.

Professor T. H. Morgan has made numerous observations and experiments on a single culture of the fruit-fly, Drosophila ampelophila, bred in bottles in the laboratory for five or six years. He has not only studied the chromosomes in the gametes of this fly, and made Mendelian crosses with it, but has obtained numerous mutations, so that his work is a very important contribution to the mutation doctrine. Drosophila in the hands of Professor Morgan and his students and colleagues has thus become as classical a type as Oenothera in those of the botanical mutationists. Different branches of Morgan's work are discussed elsewhere in this volume, but here we are concerned only with its bearing on the question of the determination of sex. He describes [Footnote: A Critique of the Theory of Evolution. Princeton University Press and Oxford University Press, 1916.] the chromosomes of Drosophila as consisting in the diploid condition of four pairs, that is to say, pairs which separate in the reduction division so that the gamete contains four single chromosomes, one of each pair. In two of these pairs the chromosomes are elongated and shaped like boomerangs, in the third they are small, round granules, and the fourth pair are the sex-chromosomes: in the female these last are straight rods, in the male one is straight as in the female, the other is bent. The straight ones are called the X chromosomes, the bent one the Y chromosome. The fertilisations are thus XX which develops into a female fly, and XY which develops into a male. Drosophila therefore is an example of one of the cases described by Wilson.

Dr. Wilson (loc. cit.) discusses the question of how we are to interpret these facts, in particular, the fact that the X chromosome in fertilisation gives rise to females. He remarks that the X chromosome must be a male-determining factor since in many cases it is the only sex-chromosome in the males, yet its introduction into the egg establishes the female condition. This is the same difficulty which I pointed out above in connection with the Mendelian theory that the female was heterozygous and the male homozygous for sex. Dr. Wilson points out that in the bee, where fertilised eggs develop into females and unfertilised into males, we should have to assume that the X chromosome in the female gamete is a female determiner which meets a recessive male determiner in the X chromosomes of the sperm. When reduction occurs, the X[female] must be eliminated since the reduced egg develops always into a male. But on fertilisation, since the fertilised egg develops into a female, a dominant X[female] must come from the sperm, so that our first assumption contradicts itself.

Dr. Wilson, T. H. Morgan, and Richard Hartwig have therefore suggested that the sex-difference as regards gametes is not a qualitative but a quantitative one. In certain cases there is no evident quantitative difference of chromatin as a whole, but there may in all cases be a difference in the quantity of special sex-chromatin contained in the X element. The theory put forward by Wilson then is that a single X element means per se the male condition, while the addition of a second element of the same kind produces the female condition. Such a theory might apply even to cases where no sex-chromosomes can be distinguished by the eye: the ova, in such cases (probably the majority), might also have a double dose of sex-chromatin, the males a single dose. This theory, however, is still open to the objection that the female gametes before fertilisation, and half the male gametes, have the half quantity of sex-chromatin which by hypothesis determines the male condition, so that here again we have the male condition as something which is distinct from the characteristics of the spermatozoon. But if this is the case, what is the male condition? The parthenogenetic ovum of the bee is male, and yet it is an ovum capable only of producing spermatozoa. If the single X chromosomes is the cause of the development of spermatozoa in the male bee, why does it not produce spermatozoa in the gametes of the female bee, since when reduction takes place all these gametes have a single X chromosome?

In biology, as in every other science, we must admit facts even when we cannot explain them. The facts of what we call gravitation are obvious, and any attempt to disregard them would result in disaster, yet no satisfactory explanation of gravitation has yet been discovered: many theories have been suggested, but no theory has yet been proved to be true. In the same way it may be necessary to admit that two X chromosomes result in the development of a female, and one X, or XY chromosomes result in the development of a male. But Mendelians have omitted to consider what is meant by male and female. The soma with its male and female somatic characters has nothing to do with the question, since somatic sex-differences may be altogether wanting, and moreover, the essential male character, the formation of spermatozoa, is by the Mendelian hypothesis due to descent of the male gametes from the original fertilised or unfertilised ovum. The Mendelian theory therefore is that when an ovum has two X sex-chromosomes it can only after a number of cell-divisions, at the following reduction division, give rise to ova, while an ovum containing one X sex-chromosome, or two different, XY, chromosomes, at the next reduction division gives rise to spermatozoa. The X sex-chromosome is not in itself either female or male, since, as we have seen, either ovum or spermatozoon may contain a single X chromosome. The ovum then with one X chromosome or one X and one Y changes its sex at the next reduction division and becomes male. In parthenogenetic ova this happens without conjugation with a spermatozoon at all: in other cases, since the zygote is compounded of spermatozoon and ovum, we can only say that in the XX zygote, the ovum developing only ova, the female is dominant, in the X or XY zygote developing only spermatozoa the male is dominant. Hermaphrodite animals, as has been pointed out by Correns and Wilson, cannot be brought under this scheme at all. In the earthworms, for instance, we have, in every individual developed from a zygote, ova and spermatozoa developing in different gonads in different parts of the body. The differentiation here, therefore, must occur in some cell-division preceding the reduction divisions. Every zygote must have the same composition, and yet give rise to two sexes in the same individual.

Further light on the sex problem, as in many other problems in biology, can only be obtained by more knowledge of the physical and chemical processes which take place in the chromosomes and in the relations of these structures to the rest of the cell. The recent advances in cytology, remarkable as they are, consist almost entirely of observations of microscopic structure. They may be said to reveal the statics of the cell rather than its dynamics. Cytology is in fact a branch of anatomy, and in the anatomy of the cell we have made some progress, but our knowledge of the physiology of the cell is still infinitesimal. The nucleus, and especially the chromosomes, are supposed in some unknown way to influence or govern the metabolism of the cytoplasm. From this point of view the hypothesis mentioned above that the sex-difference in the gametes is not qualitative but quantitative is probably nearer to the truth. Geddes and Thomson and others have maintained that the sex-difference is one of metabolism, the ovum being more anabolic, the sperm more katabolic. A double quantity of special chromatin may be the cause of the greater anabolism of the ovum. In that case the difficulty indicated in a previous part of this chapter, that the ovum after reduction resembles the sperm in having only one X chromosome, may be explained by the fact that the growth of the ovum and its accumulation of yolk substances has been already accomplished under the influence of the two chromosomes before reduction. Other difficulties previously discussed also appear to be diminished if we adopt this point of view. We need not regard maleness and femaleness as unit characters in heredity of the same kind as Mendelian characters of the soma. Instead of saying that the zygote composed of ovum and spermatozoon is incapable of giving rise in the male to ova, or in the female to sperms, we should hold that the gametocytes ultimately give rise to ova or to sperms according to the metabolic processes set up and maintained in them through their successive cell-divisions under the influence of the double or single X chromosome. There still remains the difficulty of explaining why the male gametocytes after reduction develop into similar sperms, with their heads and long flagella, although half of them possess one X chromosome each and the other half none. We can only suppose that the final development of the sperms is the result of the presence of the single X chromosome in the successive generations of male gametocytes before the reduction divisions.

The Mendelian theory of sex-heredity assumed that in the reduction divisions the two sex-characters, maleness and femaleness, were segregated in the same way as a pair of somatic allelomorphs, but the words maleness and femaleness expressed no real conceptions. The view above suggested merely attempts to bring our real knowledge of the difference between ovum and sperm into relation with our real knowledge of the sex-chromosomes and their behaviour in reduction and fertilisation.



CHAPTER III

Influence Of Hormones On Development Of Somatic Sex-Characters

We have next to consider what are commonly called secondary sexual characters. These are characters or organs more or less completely limited to one sex. When we distinguish in the higher animals the generative organs or gonads on the one hand from the body or soma on the other, we see that all differences between the sexes, except the gonads, are somatic, and we may call them somatic sexual characters. The question at once arises whether the soma itself is sexual, that is to say, whether on the assumption that the sex of the zygote is already determined before it begins to develop, the somatic cells as well as the gametocytes are individually and collectively either male or female. In previous discussions of the subject I have urged that the only meaning of sex was the difference between the megagamete or ovum, and the microgamete or sperm. But if the zygote, although compounded of ovum and sperm, is predestined to give rise in the gametes descended from it, either to sperms only or to ova only, it may be suggested that all the somatic cells descended from the zygote are likewise either male or female, although they do not give rise to gametes. It is evident, however, that the somatic cells, organs, and characters do not differ necessarily or universally in the two sexes. On the one hand, we have extraordinary and very conspicuous peculiarities in the male, entirely absent in the female, such as the antlers of stags, and the vivid plumage of the gold pheasant; on the other we have the sexes externally alike and only distinguished by their sexual organs, as in mouse, rabbit, hare, and many other Rodents, most Equidae, kingfisher, crows and rooks, many parrots, many Reptiles, Amphibia, Fishes, and invertebrate animals. In the majority of fishes, in which fertilisation is external and no care is taken of the eggs or young, there are no somatic sexual differences. Moreover, somatic sexual characters where they do occur have no common characteristics either in structure or position in the body. It may be said that any part of the soma may in different cases present a sex-limited development. In the stag the male peculiarity is an enormous development of bone on the head, in the peacock it is the enlargement of the feathers of the tail. In some birds there are spurs on the legs, in others spurs on the wings. It is no explanation, therefore, to say that these various organs and characters are the expression of sex in the somatic cells.

As I pointed out in my Sexual Dimorphism (1900), the common characteristic of somatic sexual characters is their adaptive relation to some function in the sexual habits of the species in which they occur. There is no universal characteristic of sex except the difference between the gametes and the reproductive organs (gonads) in which they are produced. All other differences, therefore, including genital ducts and copulatory or intromittent organs, are somatic. When we examine these somatic differences we find that they can be classified according to their relation to fertilisation and reproduction, including the care or protection of the offspring. The precise classification is of no great importance, but we may distinguish the following kinds to show the chief functions to which the characters or organs are adapted.

1. GENITAL DUCTS AND INTROMITTENT ORGANS.—According to the theory of the coelom which we owe to Goodrich, in all the coelomata the coelom is primarily the generative cavity, on the walls of which the gametocytes are situated, and the coelomic ducts are the original genital ducts. In Vertebrates we find two such ducts in both sexes in the embryo, originally formed apparently by the splitting of a single duct. In the male one of these ducts becomes connected with the testis while the other degenerates: the one which degenerates in the male forms the oviduct in the female, while the one which is functional in the male degenerates in the female.

Intromittent organs are formed in all sorts of different ways in different animals. In Elasmobranchs (sharks and skates) they are enlarged portions of the pelvic fins, and therefore paired. In Lizards they are pouches of the skin at the sides of the cloacal opening. In Mammals the single penis is developed from the ventral wall of the cloaca. In Crustacea certain appendages are used for this function. There are a great many animals, from jelly-fishes to fishes and frogs, in which fertilisation is external, and there are no intromittent organs at all.

2. ORGANS FOR, CAPTURING OR HOLDING THE FEMALE: for example, the thumb-pads of the frog, and a modification of the foot in a water-beetle. Certain organs on the head and pelvic fins of the Chimaeroid fishes are believed to be used for this purpose.

3. WEAPONS.—Organs which are employed in combats between males for the exclusive possession of the females. For example, antlers of stags, horns of other Ruminants, tusks of elephants, spurs of cocks and Phasiamidae generally, horns and outgrowths in males of Reptiles and many Beetles, probably used for this purpose.

4. ALLUREMENTS.—Organs or characters used to attract or excite the female. These might be called the organs of courtship, such as the peacock's tail, the plumes of the birds-of-paradise, and the brilliant plumage of humming birds and many others. The song of birds is another example, and sound is produced in many Fishes for a similar purpose.

5. ORGANS FOR THE BENEFIT OF THE OFFSPRING: for example, the extraordinary pouches in which the eggs are developed in certain Frogs. In the South American species, Rhinoderma darwinii, the enlarged vocal sacs are used for this purpose. Pouches with the same function are developed in many animals, for instance in Pipe-fishes and Marsupials. Abdominal appendages are enlarged in female Crustacea for the attachment of the eggs, the abdomen also being larger and broader.

The argument in favour of the Lamarckian explanation of the evolution of these adaptive characters is the same as in the case of adaptations common to both sexes, namely that in every case the function of the organs and characters involves special irritations or stimulations by external physical agents. Mechanical irritation, especially of the interrupted kind, repeated blows or friction causes hypertrophy of the epidermis and of superficial bone. I have stated this argument and the evidence for it in some detail in my volume on Sexual Dimorphism. It is one of the most striking facts in support of this argument that the hypertrophied plumage which constitutes the somatic sexual character of the male in so many birds is habitually erected by muscular action for the purpose of display in the sexual excitement of courtship. I doubt if there is a single instance in which the male bird takes up a position to present his ornamental plumage to the sight of the female without a special erection and movement of the feathers themselves. Such a stimulation must affect the living epidermic cells of the feather papilla. Even supposing that the feather is not growing at the time, it is probable, if not certain, that the stimulation will affect the papilla at the base of the feather follicle, so as to cause increased growth of the succeeding feather. But we have no reason to believe that erection in display occurs only when the growth of the feathers is completed, still less that it did so always at the beginning of the evolution.

The antlers of stags are the best case in favour of the Lamarckian view of the evolution of somatic sexual characters. The shedding of the skin ('velvet') followed by the death of the bone, and its ultimate separation from the skull, are so closely similar to the pathological processes occurring in the injury of superficial bones, that it is impossible to believe that the resemblance is only apparent and deceptive. In an individual man or mammal, if the periosteum of a bone is destroyed or removed the bone dies, and is then either absorbed, or separated from the living bone adjoining, by absorption of the connecting part. In the stag both skin and periosteum are removed from the antler: probably they would die and shrivel of their own accord by hereditary development, but as a matter of fact the stag voluntarily removes them by rubbing the antler against tree trunks, etc. When the bone is dead the living cells at its base dissolve and absorb it, and when the base is dissolved the antler must fall off.

The adaptive relation is not the only common characteristic of these somatic sexual characters. Another most important fact is not only that they are fully developed in one sex, absent or rudimentary in the other, but that their development is connected with the functional maturity and activity of the gonads. There is usually an early immature period of life in which the male and female are similar, and then at the time of puberty the somatic sexual characters of either sex, generally most marked in the male, develop. In some cases, where the activity of the gonads is limited to a particular season of the year, the sexual characters or organs are developed at this season, and then disappear again, so that there is a periodic development corresponding to the periodic activity of the testes or ovaries. Stags have a limited breeding or 'rutting' season in autumn (in north temperate regions), and the antlers also are shed and developed annually. In this case we cannot assert that the development of the antler takes place during the active state of the testes. The antlers are fully developed and the velvet is shed at the commencement of the rutting season, and development of the antlers takes place between the beginning of the year and the month of August or September. In ducks and other birds there is a brilliant male-breeding plumage in the breeding season which disappears when breeding is over, so that the male becomes very similar to the female. In the North American fresh-water crayfishes of the genus Cambarus there are two forms of males, one of which has testes in functional activity, while in the other these organs are small and quiescent: the one form changes into the other when the testes pass from the one condition to the other.

It has long been known that the development of male sex-characters is profoundly affected by the operation of castration. The removal of the testes is most easily carried out in Mammals, in consequence of the external position of the organs in these animals, and the operation has been practised on domesticated animals as well as on man himself from very ancient times. The effect is the more or less complete suppression of the male insignia, in man, for example, the beard fails to develop, the voice does not undergo the usual change to lower pitch which takes place at puberty, and the eunuch therefore has much resemblance to the boy or woman. Many careful experimental researches have been made on the subject in recent years. The consideration of the subject involves two questions: (1) What are the exact effects of the removal of the gonads in male and female? (2) By what means are these effects brought about, what is the physiological explanation of the influence of the gonads on the soma?

I have quoted the evidence concerning the effects of castration on stags in my Sexual Dimorphism and in my paper on the 'Heredity of Secondary Sexual Characters.' [Footnote: Archiv fuer Entwicklungesmechanik, 1908.] When castration is performed soon after birth a minute, simple spike antler is developed, only two to four inches in length: it remains covered with skin, is never shed, and develops no branches. When the operation is performed on a mature stag with antlers, the latter are shed soon after the operation, whether they have lost their velvet or not. In the following season new antlers develop, but these never lose their velvet or skin and are never shed.

CASTRATION IN FOWLS

The removal of the testes from young cocks has been commonly practised in many countries, e.g. France, capons, as such birds are called, being fatter and more tender for the table than entire birds. The actual effect, however, on the secondary sexual characters has not in former times been very definitely described. The usual descriptions represent the castrated birds as having rather fuller plumage than the entire birds; but the comb and wattles are much smaller than in the latter, more similar to those of a hen. It is stated that the capon will rear chickens, though he does not incubate, and that they are used in this way in France.

The most precise of the statements on the subject by the earlier naturalists is that of William Yarrell [Footnote: _Proc. Linn. Soc., 1857.] (1857), who writes as follows:—

'The capon ceases to crow, the comb and gills do not attain the size of those parts in the perfect male, the spurs appear but remain short and blunt, and the hackle feathers of the neck and saddle instead of being long and narrow are short and broadly webbed. The capon will take to a clutch of chickens, attend them in their search for food, and brood them under his wings when they are tired.'

It would naturally be expected, on the analogy of the case of stags, that when a young cock was completely castrated all the male secondary characters would be suppressed, namely, the greater size of the comb and wattles in comparison with the hen, the long neck hackles, and saddle hackles, long tail feathers, especially the sickle-feathers, and the spurs. As a matter of fact, the castrated specimen usually shows only the first of these effects to any conspicuous degree. The comb and wattles of the capon are similar to those of the hen, but he still has the plumage and the spurs of the entire cock. Many investigators have made experiments in relation to this subject, and most of them have found that complete castration is difficult, and that portions of the testes left in the bird during the operation become grafted in some other position either on the parietal peritoneum, or on that covering the intestines, and produce spermatozoa, which, of course hare no outlet. In such cases the secondary male characters may fee more or less completely developed. Thus Shattock and Seligmann (1904) state that ligature of the vas deferens made no difference to the male characters, and that after castration detached fragments were often left in different positions as grafts, when the secondary characters developed. In one particular case only a minute nodule of testicular tissue showing normal spermatogenesis was found on post mortem examination attached to the intestine. In this bird there was no male development of comb or wattles, a full development of neck hackles, a certain development of saddle hackles, a few straggling badly curved feathers in the tail and short blunt spurs on the legs. Lode [Footnote: Wiener klin. Wochenschr., 1895.] (1895) found that testes could easily be transplanted into subcutaneous tissue and elsewhere, and that the male characters then developed normally. Hanau [Footnote: Arch. f. ges. Physiologie, 1896.] (1896) obtained the same result.

The question, however, to what degree the male characters of the cock are suppressed after complete castration is not so definitely answered in the literature of the subject. Shattock and Seligmann in their 1904 paper make no definite statement on the subject. Rieger (1900), Selheim (1901), and Foges [Footnote: Pfuegers Archiv, 1902.] (1902) state that the true capon is characterised by shrivelling of the comb, wattles, and spurs; poor development of the neck and tail feathers; hoarse voice and excessive deposit of fat. Shattock and Seligmann, on the other hand, have placed in the College of Surgeons Museum the head of a Plymouth Rock which was castrated in 1901. It was hatched in the spring of that year. In December 1901 the comb and wattles were very small, the spurs fairly well developed, and the tail had a somewhat masculine appearance. In September 1902, when the bird was killed, the comb and wattles were still poorly developed, the neck hackles fairly well so; saddle hackles rather well developed; the tail contained rather loosely-grouped long sickle feathers; the spurs stout. The description states that dissection showed no trace of either testicle, and I am informed by Mr. Shattock that there were no grafts. The description ends with the conclusion that the growth of the spurs, and to a certain extent that of the long, curved sickle feathers, is not prevented by castration. With regard to the spurs this result does not agree with that of the German investigators, but it must be remembered that the latter speak only of the reduction of the spurs, not entire absence. It is important in discussing the effects of castration in cocks to bear in mind the actual course of development of the secondary sexual characters. When the chicks are first hatched they are in the down: rudimentary combs are present, wattles can scarcely be distinguished, and there is no external difference between the sexes. The ordinary plumage begins to develop immediately after hatching, the primaries of the wings being the first to appear. The feathers are completely developed in about five weeks, and still there is no difference between the sexes. The first sexual difference is the greater size of the combs in the males, and this is quite distinct at the age of six weeks. At nine to ten weeks in black-red fowls, in which the cocks have black breasts and red backs with yellow hackles, the black feathers on the breast and red on the back are gradually developing, both sexes previously having been a dull speckled brown, closely similar to the adult hens. The spurs are the last of the male characters to develop, these at the age of four months being still mere nodules, scarcely, if at all, larger than the rudiments visible in adult hens. This is the age at which castration is usually performed, as at an earlier age the birds are too small to operate on successfully. It follows, therefore, that the spurs develop after castration, and it would seem that their development does not depend upon the presence of the sexual organs. It is a question, however, whether castration in the cock is ever quite complete. In the original wild species and in the majority of domesticated breeds the spurs are confined to the male sex, and are typical secondary sex-characters, as much so as the antlers of stags or the beard of man, yet the above discussion shows that there is some doubt whether their development is prevented as much as in other cases by the absence of the sexual organs. Even if it should be proved that in supposed cases of complete castration, such as that of Shattock and Seligmann, some testicular tissue remained at the site of the testes, it would still be true that the development of the comb and wattles is more affected by the removal of the sexual organs than that of the spurs or tail feathers.

My own experiments in castrating cocks were as follows: On August 20, 1910, I operated on a White Leghorn cock about five months old. One testis was removed, with a small part of the end broken off, but the other, after it was detached, was lost among the intestines. On the same day I operated on another about thirteen weeks old, a speckled mongrel. In this case both testes were extracted but one was slightly broken at one end, although I was not sure that any of it was left in the body. An entire White Leghorn of the same age as the first was kept as a control. On August 27 the two castrated birds had recovered and were active. Their combs had diminished in size and lost colour considerably, that of the White Leghorn was scarcely more than half as large as that of the control. Such a rapid diminution can scarcely he due to absorption of tissue, but shows that the size of the normal cock's comb is largely due to distension with blood, which ceases when the sexual organs are removed. In the following January, the second cock, supposed to be completely castrated, was seen to make a sexual gesture like a cock, though not a complete action like an entire animal: this showed that the sexual instinct was not completely suppressed. In February this same bird was seen to attempt to tread a hen, while the white one, supposed to be less perfectly emasculated, had never shown such male instinct.

The White Leghorn cock was killed and dissected on May 13, 1911, nine months after castration. I found an oval body of dark, dull brown colour loose among the intestines: this was evidently the left testis which was separated from its natural attachment and lost in the abdomen at the time of the operation. I examined the natural sites of the testes: on the right side there was a small testis of considerable size, about half an inch in diameter. When a portion of this was teased up and examined under the microscope moving spermatozoa were seen, but they were not in swarms as in a normal testis, but scattered among numerous cells. On the left side was a much smaller testis, in the tissue of which I with difficulty detected a few slowly moving spermatozoa. The vasa deferentia were seen as white convoluted threads on the peritoneum, but contained no spermatozoa.

On July 29, 1911, a little more than eleven months after the operation, I examined and killed the second of these castrated cocks, the speckled mongrel-bred bird. I measured the comb and wattles while it was alive, in case there might be reduction in the size of these appendages when the bird was killed. The comb was 1-1/3 inches high by 2-3/8 inches in length. The spurs were 1 inch long, curved and pointed. Saddle hackles short, hanging only a little below the end of the wing. Neck hackles well developed, similar to those of an entire cock. Longest tail feather 15-5/8 inches, blue-black in colour.

I had no entire cock of same breed, but measured the entire White Leghorn for comparison. Comb 1-3/4 inches high by 3-3/4 inches in length. (It is to be remembered that the comb and wattles are especially large in Leghorns.) Wattle 1-1/4 inches in vertical length. Spur 1 inch long, stouter and less pointed than in the capon. Longest tail feather 12 inches long.

When killed the capon was found to be very fat: there were masses of fat around the intestines and under the peritoneum, which made it impossible to make out details such as ureter and vas deferens properly. I found a white nodule about half an inch in diameter attached to mesentery. The liquid pressed from this was swarming with spermatozoa in active motion. Two other masses about the same size or a little larger were found on the sites of the original testes. These also were full of mobile spermatozoa, and must have grown from portions of the testes left behind at castration.

In ducks the sexual characters of the male differ from those in the fowl, especially in the fact that they almost completely disappear after the breeding season and reappear in the following season. In the interval the drake passes into a condition of plumage in which he resembles the female; and this condition is known as 'eclipse.' The male plumage, therefore, in the drake has a history somewhat similar to that of the antlers in deer. Two investigations of the effects of castration on ducks and drakes have been recorded. H. D. Goodale [Footnote: 'Castration of Drakes.' Biol. Bulletin, Wood's Hole, Mass., vol. xx., 1910] removed the generative organs from both drakes and ducks of the Rouen breed, which is strongly dimorphic in plumage. One drake was castrated in the early spring of 1909 when a little less than a year old. This bird did not assume the summer plumage in 1909, that is, did not pass into eclipse. It was in the nuptial plumage when castrated. This breeding or nuptial plumage is well known: it includes a white neck-ring, brilliant green feathers on the head, much claret on the breast, brilliant metallic blue on the wing, and two or more upward curled feathers on the tail. The drake mentioned above was accidentally killed in the spring of 1910. Another drake was castrated on August 8, 1909: only the left testis was removed, the other being ligatured. At this time the bird would be in eclipse plumage. It appears from the description that it assumed the nuptial plumage in the winter of 1909, and did not pass into eclipse again in the summer of 1910. Thus in drakes the effect of castration is that the secondary sexual character remains permanently instead of being lost and renewed annually. Goodale, however, does not describe the moults in detail. In the natural condition the drake must moult twice in the year, once when he sheds the nuptial plumage, and again when he drops the summer dress. Goodale insists, from some idea about secondary sexual characters which is not very obvious, that the eclipse or summer plumage is not the same as that of the female. He states that the male in summer plumage merely mimics the female but does not become entirely like her. In certain parts of the body there are no modifications toward the female type. In others, i.e. head, breast, and keel region, the feathers of the male become quite like those of the female. 'It can hardly be maintained that this is an example of assumption by the male of the female's plumage, especially as the presence of the testis is necessary for its appearance.' The idea here seems to be that since the eclipse plumage is only assumed when the testis is present, therefore it must be a male character.

Out of five females on which the operation was performed only two lived more than a few days afterwards. One of these (a) was castrated in the spring of 1909 when a little less than a year old, the other (b) on August 13 when twelve weeks old. In October 1909 they showed no marked modifications. In July 1910 it was noticed that they had the male curled feathers in the tail, and (a) had breast feathers similar to those of the male in summer plumage, (b) was rather more strongly modified: she had a very narrow white neck-ring, and breast feathers distinctly of male type. The next moult began in September, and in November was well advanced. On the whole (a) had made little advance towards the male type, but (b) closely resembled the male in nuptial plumage. It had brilliant green feathers on the head, a white neck-ring, much claret colour on the breast, and some feathers indistinguishable from those of the male, and also the male sex feathers on the tail. Goodale concludes that the female owes her normal colour to the ovaries or something associated with them which suppresses the male characters and ensures the development of her own type. He considers it is quite as conceivable that selection should operate to pick out inconspicuously coloured females as that selection of brilliantly coloured males should bring about an addition to the female type. But as pointed out above, selection cannot explain the dimorphism in either case.

It may be mentioned here that owing to the fact that the single (left) ovary in birds is very closely attached to the peritoneum immediately covering the great post-caval vein, it is generally impossible to remove the whole of the ovary without cutting or tearing the wall of the vein and so causing fatal hemorrhage. The above results observed by Goodale are therefore all the more remarkable, and it may be assumed that he removed at any rate nearly all the ovary.

The research of Seligmann and Shattock [Footnote: Relation between Seasonal Assumption of the Eclipse Plumage in the Mallard (Anas boscas) and the Functions of the Testicle.' Proc. Zool. Soc. 1914.] begins with a comparison between the stages of the development of the nuptial plumage and the stages of spermatogenesis. In the young pheasant the male plumage is fully developed in the autumn of its first year, but no pairing occurs and no sexual instinct is exhibited till the following spring. The wild duck pairs in autumn or early winter, after the assumption of the nuptial plumage, but copulation does not occur till spring is advanced. The investigation here considered was made upon specimens of semi-domesticated Anas boscas, such as are kept in London parks and supplied from game farms. The testes attain their maximum size during the breeding season— end of March or beginning of April. At this time each organ is almost as large as a pigeon's egg, is very soft, and the liquid exuding from it when cut is swarming with spermatozoa. The bird is of course in full nuptial plumage. By the end of May, although the plumage is unchanged, the testes have diminished to the size of a haricot bean, and spermatogenesis has ceased. They diminish still further during June, July, and August, and acquire a yellow or brownish colour, while microscopically there is no sign of activity in the spermatic cells. The change from nuptial plumage to eclipse takes place between the beginning of June and the middle of July. The reappearance of the nuptial plumage takes place in the month of September, and while this process takes place there is no sign of change or renewed activity in the testes. During October and November, when the brilliant plumage is fully developed, the testes increase slowly in size but remain yellow and firm and exude no liquid on incision. Spermatogenesis does not commence until the end of November or beginning of December. The testes increase greatly in size in January and February, and again reach their maximum size by the end of March. It is shown, therefore, that the loss of the nuptial plumage takes place in June when spermatogenesis has ceased and the testes are diminishing in size, but the redevelopment of this plumage takes place in September without any renewed activity of the testis and long before the beginning of spermatogenesis. The case of the antlers in the stag is probably very similar.

The important statement is made with regard to castration (under anaesthetics, of course) that it was found impossible to extirpate the testes completely. When the bird was killed some months after the operation, a greater or lesser amount of regenerated testicular tissue was found either on the original site of the organs or engrafted upon neighbouring organs. This experience, it will be noted, agrees with my own in the case of fowls. There were, however, reasons for believing that the results observed within the first six or eight months after the operation are not much different from those which would follow complete castration.

Castration carried out when the drake was in nuptial plumage produced the same effect which was observed by Goodale, namely, delay, and imperfection in the assumption of the eclipse condition, but the observations of Seligmann and Shattock are more precise and detailed. One example described was castrated in full winter plumage in December 1906. On July 11, when normally it would have been in eclipse, the nuptial plumage was unmodified except for a diffuse light-brown coloration on the abdomen, which is stated to be due not to any growth of new feathers but to pigmentary modification in the old. By September 1 this bird was almost in eclipse but not quite; curl feathers in the tail had disappeared, the breast was almost in full eclipse, the white ring was slightly indicated at the sides of the neck, the top of the head and the nape had still a good deal of gloss. After this the nuptial plumage developed again, and on November 12 the bird was in full nuptial plumage, with good curl feathers in the tail. The only trace of the eclipse was the presence of a few brown feathers on the flanks. This bird was killed July 30, 1908, when the bird was in eclipse, but not perfectly so, as there were vermiculated feathers mixed with eclipse feathers on the breast, abdomen, and flanks. Dissection showed on the right side a series of loosely attached nodular grafts of testicular tissue, in total volume about the size of a haricot bean: on the left side two small nodules, together about the size of a pea, and two other grafts at the root of the liver and on the mesentery. Several other cases are described, and the general result was that the eclipse was delayed and never quite complete, while although the nuptial plumage was almost fully developed in the following winter, it retained some eclipse feathers, and was also delayed and developed slowly.

Several drakes were castrated in July when in the eclipse condition, and although the authors state, in their general conclusions, that this does not produce any constant appreciable effect upon the next passage of the bird into winter plumage, they describe one bird so treated which on November 18 retained many eclipse feathers: the general appearance of the chestnut area of the breast was eclipse.

It must be remembered that not only was the castration in these cases incomplete, but also that it was performed on mature birds. Birds differ from Mammals, firstly, in the difficulty of carrying out complete castration, and secondly, in the fact that the occurrence of puberty is not so definite, and that immature birds are so small and delicate that it is almost impossible to operate upon them successfully.

ASSUMPTION OF MALE CHARACTERS BY THE FEMALE

That male somatic sexual characters are latent in the female is shown by the frequent appearance of such characters in old age, or in individual cases. The development of hair on the face of women in old age, or after the child-bearing period, is a well-known fact. Rorig, [Footnote: 'Ueber Geweihbildung und Geweihentwicklung.' Arch. Ent.-Mech. x. and xi.] who carefully studied the antlers of stags, states that old sterile females, and those with diseased ovaries, develop antlers to some degree. Cases of crowing hens, and female birds assuming male plumage have long been known, but the exact relation of the somatic changes to the condition of the ovaries in these cases is worthy of consideration in view of the results obtained by Goodale after removal of the ovaries from ducks. Shattock and Seligmann [Footnote: 'True Hermaphroditism in Domestic Fowl, etc.' Trans. Path. Soc., Lond., 57. 1, 1906.] record the case of a gold pheasant hen which assumed the full male plumage after the first moult: it had never laid eggs or shown any sexual instincts. The only male character which was wanting was that of the spurs. The ovary was represented by a smooth, slightly elevated deep black eminence 1 cm. in length and 1-5 mm. in breadth at its upper end. These authors also mention three ducks in male plumage in which the ovary was similarly atrophied but not pigmented. They regard the condition of the ovary as insufficient to explain the development of the male characters, and suggest that such birds are really hermaphrodite, a male element being possibly concealed in a neighbouring organ such as the adrenal or kidney. This hypothesis is not supported by observation of testicular tissue in any such case, but by the condition found in a hermaphrodite specimen of the common fowl described in the paper. This bird presented the fully developed comb and wattles and the spurs of the cock, but the tail was quite devoid of curved or sickle feathers, and resembled that of the hen. Internally there were two oviducts, that of the left side normally developed, that of the right diminutive and less than half the full length. The gonad of the left side had the tubular structure of a testis, but showed no signs of active spermatogenesis, but in its lower part contained two ova. The organ of the right side was somewhat smaller, it had the same tubular structure, and in one small part the tubules were larger, showed division of nuclei (mitotic figures), and one of them showed active spermatogenesis.

In discussing Heredity and Sex in 1909, [Footnote: Mendel's Principles of Heredity. Camb. Univ. Press, 1909.] Bateson referred to the effects of castration as evidence that in different types sex may be differently constituted. Castration, he urged, in the male vertebrate on the whole leads merely to the non-appearance of male features, not to the assumption of female characters, while injury or disease of the ovaries may lead to the assumption of male characters by the female. This was supposed to support the view that the male is homozygous in sex, the female heterozygous in Vertebrates: that is to say, the female sex-character and the female secondary sex-characters are entirely wanting in the male. This argument assumes that the secondary characters are essentially of sexual nature without inquiring how they came to be connected with sex, and it ignores the fact that the influence of castration on such characters is a phenomenon entirely beyond the scope of Mendelian principles altogether. The fact that castration does affect, in many cases very profoundly, somatic characters confined to one sex, proves that Mendelian conceptions, however true up to a certain point, are by no means the whole truth about heredity and development. For it is the essence of Mendelism as of Weismannism that not only sex but all other congenital characters are determined in the fertilised ovum or zygote. The meaning of a recessive character in Mendelian terminology is one that is hidden by a dominant character, and both of them are due to factors in the gametes, particularly in the chromosomes of the gametes which come together in fertilisation. For example, in fowls rose comb is dominant over single. A dominant is something present which is absent in the recessive: the rose comb is due to a factor which is absent from the single. The two segregate in the gametes of the hybrid or heterozygote, and if a recessive gamete is fertilised by another recessive gamete the single comb reappears. But castration shows that the antlers of stags and other such characters are not determined in the zygote when the sex is determined, but owe their development, partly at least, to the influence of another part of the body, namely, the testes during the subsequent life of the individual. According to Mendelism the structure and development of each part of the soma is due to the constitution of the chromosomes of the nuclei in that part. The effects of castration show that the development of certain characters is greatly influenced in some way by the presence of the testes in a distant part of the body. The Mendelians used to say it was impossible to believe in the heredity of somatic modifications due to external conditions, because it was impossible to conceive of any means by which such modifications could affect the constitution of the chromosomes in the gametes within the modified body. It would have been just as logical to deny the proved effects of castration, because it was impossible to conceive of any means by which the testes could affect the development of a distant part of the body.

But this is not all. The supposed fact that female secondary characters in Vertebrates are absent in the male is completely disproved for Mammals by the presence of rudimentary mammary glands in the male. It is true that secondary sex-characters are usually positive in the male, while those of the female are apparently negative, but in the case of the mammary glands the opposite is the case. There is no room for doubt that the mammary glands are an essentially female somatic sex-character, not only in their function but in the relation between the periodicity of that function and those of the ovaries and uterus, and it is equally certain from their presence in rudimentary condition in the male that they are not absent from the male constitution.

INFLUENCE OF GONADS DUE TO HORMONES

The existence and the influence of hormones or internal secretions may be said to have been first proved in the case of the testes, for Professor A. A. Berthold [Footnote: 'Transplantation der Hoden,' Archiv. f Anat. u. Phys., 1849.] of Goettingen in 1849 was the first to make the experiment of removing the testicles from cocks and grafting them in another part of the body, and finding that the animals remained male in regard to voice, reproductive instinct, fighting spirit, and growth of comb and wattles. He also drew the conclusion that the results were due to the effect of the testicle upon the blood, and through the blood upon the organism. Little attention was paid to Berthold's experiment at the time. The credit of having been the first to formulate the doctrine of internal secretion is generally given to Claude Bernard. He discovered the glycogenic function of the liver, and proved that in addition to secreting bile, that organ stores up glycogen from the sugar absorbed in the stomach and intestines, and gives it out again as sugar to the blood. In 1855 he maintained that every organ of the body by a process of internal secretion gives up products to the blood. He did not, however, discover the action of such products on other parts or functions of the body. Brown-Sequard, in his address before the Medical Faculty of Paris in 1869, was the first to suggest that glands, with or without ducts, supplied special substances to the blood which were useful or necessary to the normal health, and in 1889 at a meeting of the Societe de Biologie he described the experiment he had made upon himself by the injection of testicular extract. This was the commencement of organotherapy. Since that time investigation of the more important organs of internal secretion—namely, the gonads, thyroid, thymus, suprarenals, pituitary, and pineal bodies—has been carried on both by clinical observation and experiment by a great number of physiologists with very striking results, and new hormones have been discovered in the walls of the intestine and other organs.

Here, however, we are more especially concerned with the gonads and other reproductive organs. A great deal of evidence has now been obtained that the influence of the testes and ovaries on secondary sexual characters is due to a hormone formed in the gonads and passing in the blood in the course of the circulation to the organs and tissues which constitute those characters. The fact that transplanted portions of testes in birds (cocks and drakes) are sufficient to maintain the secondary characters in the same condition as in normal individuals shows that the nexus between the primary and somatic organs is of a liquid chemical nature and not anatomical, through the nervous system for example. Many physiologists in recent years have maintained that the testicular hormone is not derived from the male germ-cells or spermatocytes, but from certain cells between the spermatic tubuli which are known as interstitial cells, or collectively as the interstitial gland.

The views of Ancel and Bouin, [Footnote: C. R. Soc. Biol., iv.] published in 1903, may be described in large part as theory. They state that the interstitial cells appear in the male embryo before the gametocytes present distinctive sex-characters. They conclude that the interstitial cells supply a nutritive material (hormone?), which has an effect on the sexual orientation of the primitive generative cells. In addition to this function, the interstitial cells by their hormone also give the sexual character to the soma. When castration is carried out at birth the male somatic characters do not entirely disappear, because the hormone of the interstitial cells has acted during intrauterine life. The functional independence between the interstitial cells and the seminal tubules is shown by the fact that if the vasa deferentia are closed the seminal gland (i.e. tubules) degenerates while the interstitial cells do not. In the embryo the interstitial gland is large, in the adult proportionately small.

There is complete disagreement between the results of Ancel and Bouin on the one hand, and those of Shattock and Seligmann on the other, with regard to the effects of ligature of the vasa deferentia. The latter authors, as mentioned above, found that after ligature not only the somatic characters but the testis itself developed normally. The experiments were performed on Herdwick sheep and domestic fowls. They state that on examination the testes were found to be normally developed, and spermatogenesis was in progress. The experiments of Ancel and Bouin were carried out on rabbits seven to eight weeks old, and consisted in removing one testis, and ligaturing the vas deferens of the other. About six months after the operation the testis left in situ was smaller, the seminal tubules contained few spermatogonia, though Sertoli's cells (cells on the walls of the tubules to which the true spermatic cells are attached) were unchanged; while the interstitial cells were enormously developed, by compensatory hypertrophy in consequence of the removal of the other testis. At the same time the male instincts and the other generative organs were unchanged. In a few cases, however, Ancel and Bouin observed atrophy of the interstitial cells as well as the spermatic cells. They believe this is due to the nerves supplying the testis being included in the ligature. This is rather a surprising conclusion in view of the fact that testicular grafts show active spermatogenesis. It is difficult to understand why nerve connection should be necessary for the interstitial cells and not for the spermatic, and, moreover, if the interstitial cells are really the source of the hormone on which the somatic characters depend, they must be acting in the grafts in which the nerve connections have been all severed.

The facts concerning cryptorchidism, that is to say, failure of the descent of the testes in Mammals, seem to show that the hormone of the testis is not derived from semen or spermatogenesis, for in the testes which have remained in the abdomen there is no spermatogenesis, while the interstitial cells are present, and the animals in some cases exhibit normal or even excessive sexual instinct, and all the male characteristics are well marked. It may be remarked, however, in criticism of this conclusion that the descent of the testes being itself a somatic sexual character of the male, its failure when the interstitial cells are normal and the spermatic cells defective, would rather tend to prove that the defect of the latter is itself the cause of cryptorchidism.

Many investigators have found that the Roentgen rays destroy the spermatic cells of the testis in Mammals, leaving the cells of Sertoli, the interstitial tissue, nerves, and vessels uninjured. Tandler and Gross [Footnote: Wiener klinische Wochenschrift, 1907.] found that the antlers of roebuck were not affected after the testes had been submitted to the action of the rays, showing that the interstitial cells were sufficient to maintain the normal condition of the antlers. Simmonds, [Footnote: Fortschr. a. d. G. d. Roentgenstr., xiv., 1909-10.] however, found that isolated seminal tubules remained, and regeneration took place, and concludes that both spermatic cells and interstitial cells take part in producing the testis hormone. The conclusions of two other investigators have an important bearing on this question—namely, that of Miss Boring [Footnote: Biol. Bull., xxiii. 1912.] that there is no interstitial tissue in the bird's testis, and that of Miss Lane-Claypon, [Footnote: Proc. Roy. Soc., 1905] that the interstitial cells of the ovary arise from the germinal epithelium, and are perfectly equipotential with those which form the ova and Graafian follicles. It seems possible, although no such suggestion has been made, that the interstitial cells might either normally or exceptionally give rise to ova and spermatocytes. The observations of Seligmann and Shattock on the relation of spermatogenesis to the development of nuptial plumage in drakes probably receive their explanation from the above facts. Spermatogenesis is not the only source of the testicular hormone: changes in the secretory activity of the interstitial cells or spermatocytes are sufficient to account for periodic development of somatic sex-characters, and the same reasoning applies to the antlers of stags.