■t*! -S 1. ••■ KJiifiJii?" >!wr;.I^i, k -J>3 Cornell IHntverstt^ OF THE IRew ^ox\\ State College of agriculture ,h<^. ^'^..'i.T- , P^'l - 6 mTT Cornell UnNerslty Library QH 431.B71 Vigour and heredity. 3 1924 003 062 209 Cornell University Library The original of tiiis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924003062209 VIGOUR AND HEREDITY YIGOUH AND HEKEPITY BY J. LEWIS BONHOTE M.A., F.L.S., P.Z.S. WITH COLOURED AND UNCOLOUEED PLATES : : AND DLA.GBAMS IN TEXT : : LONDON: WEST, NEWMAN & CO. 54, HATTON GABDEN : : 1915 LONDOK : PBINXBD BY WEST, NBWMAN AND CO. HATTON GABDEN, B.C. PREFACE. The study of live animals has been my chief pleasure from my early schoolboy days to the present time, and during the whole of that period, as my notebooks can testify, I have always kept careful notes on all the numerous little happenings which were daily taking place among the different individuals of my stock. In 1900, when the re- discovery of Mendel's Law was given to the world, I com- menced, at the late Professor Newton's suggestion, a series of experiments in the crossing of various wild species of duck, and from then to the present day my whole energies have been devoted to the study of heredity. There can be little doubt that inheritance is the funda- mental bedrock on which the whole of Nature has evolved; and that were the laws which govern it thoroughly under- stood, results of enormous scientific and practical import- ance would be within our grasp. Scientific men have, until lately, largely neglected this study, and I venture to think that much recent work would have been considerably modified had those zoologists who have undertaken practi- cal experiments really understood live animals. A living animal cannot be treated like a museum speci- men ; it is continually changing from year to year and from b vi VIGOUR AND HBEEDITY month to month, according to its age, its food, and its surroundings. Further than this, animals are as much individualized in their own way as human beings, and in consequence the true study of heredity must really be a study of individuals. One of our leading Mendelians once said that when he went to a cattle, poultry, or pigeon show and saw the hundreds of various races that had been produced and bred to perfection by our stock-raisers and fanciers, he doubted whether science would be able to find any law by which their results could be surpassed. The keynote of the fancier's method is " individuality," and it is for this reason that what to the lay mind appears a fabulous price is often given for prize stock. Unfortunately, scientists as a body have not yet grasped this principle — which is, in our opinion, a fundamental one. To be able to draw any deductions or formulate any hypothesis from the study of individuals entails the making of numerous notes and observations over a considerable period of time ; and this is work which should undoubtedly be done by zoological gardens ; yet a perusal of the Eeports of the various institutions throughout the world gives us practically no notes of any scientific value on the animals. We are generally told (I am referring to no gardens in particular) the number of visitors, the amount of the money received, the number of keepers, the number of animals, and the total aggregate of food consumed in the gardens. We are not told the sexes of the animals bred, the period of gestation or incubation, any peculiar mating or nesting habits, the conditions and exact food of any particular animals and the effect of these conditions on their temper, PEEFACE vii actions, or breeding. We are not told of the moult, or the times of the moult, as compared with other years ; and, lastly, no attempt whatever is made to breed the animals on any scientific lines. Nowhere, except in the large zoological gardens, can these observations be made in sufficient bulk to be of any real use for scientific work, and yet the records on these points from gardens throughout the world, with scarcely an exception, show a dismal blank. In a small way I have tried to practise what I have just been preaching, and the results and thoughts that have been arrived at are embodied in this book. In my experi- ments I have worked with both domestic and wild animals, and I have at the same time carefully considered the forms and races found wild, and the conditions under which they live. This has led to the enunciation and working out of an hypothesis which, though it has been mentioned by many writers, has not hitherto been enlarged or considered of really fundamental importance. Whether my views are right or wrong, time alone will show ; but I would ask those who take up this book to read it with an entirely unbiased mind, for that is the spirit in which it has been written. As a free lance I hold no brief for or against any of the accepted theories. I have reviewed them openly and impartially, and, as will be seen, I have attempted to show how the Galtonian and Mendelian hypo- theses may be made mutually complementary. In this spirit, seeking only for the truth, and devoting many years of careful thought in trying to elucidate the many conflict- ing facts, and so add something to the sum-total of human viii VIGOUE AND HEREDITY knowledge, this book has been written ; and all I ask is, that readers will study it in the spirit in which it was written, asking themselves, with an honest and open mind : Does it help to elucidate any of the insistent problems in the zoology of to-day ? J. L. BONHOTE. P.S. — The whole of the MS. was completed and in the publishers' hands at the end of 1913 ; the publication has been delayed owing to the war. CONTENTS. PAGE PREFACE .......... V PAET I.— ENUNCIATION. CHAPTEB I. INADEQUACY OF PBESENT THEOEIES 1 Importance of the study of heredity — Effect of Darwinism on work of last century — The Mendelian Theory, its apparent failure in some cases, both among domestic and wild forms — Galton's Law and where it fails — Suggested hypothesis to reconcile conflicting theories. CHAPTER II. BOUGH OUTLINE OF VIGOUR 6 What is meant by Vigour — Initial Vigour — Fluctuations of Vigour — Vigour and Mendel — Prepotency — Inbreeding — Infertility — Vigour and Sex — Sterility — Coloration. CHAPTEE III. COLOEATION IN MAMMALS AND BIEDS 16 Origin of Colour — Vigour and Environment — Conditions in Polar Eegions — Dark Coloured Polar Animals — Migrant birds of Polar Begions — The Tropical Kegions — Light coloured Tropical forms — Seasonal change of pelage among tropical animals — The Temperate Begions. CHAPTER IV. COLOEATION AN INDEX OP VIGOUR . . . . . 27 Nature of animal pigments — Their relation to fat — Disease and Colour— Oxidising ferments— Variability of domestic animals — Uniformity of wild races — Suggested causes — Changes of colour due to captivity and artificial environment — Insular forms — Isolation — Environment of domestic races — Variable species in Nature — Purity of certain closely allied species in same localities — Possibly due to action of Mendel's Law. CHAPTEE V. COLOUE, SEX, AND MENDEL's LAW 41 Outward effects of Vigour limited by Mendelian characters — And by Sex — DiflEerent effects of Vigour in males and females — Animals in which the sexes are alike in colour — Bats — Seals — Birds — Sex of hybrids — Sex of wild animals in captivity — Causes determining sex — The Mendelian view and the effects of Vigour. VIGOUE AND HBEBDTTY PART II.— EXPEEIMBNTAL. CHAPTEE VI. DOaS, GOATS, CATS 50 Dogs : Table of matings — Black dominant to red — Colour of different litters compared in relation to time of year. Goats : Previous work — Mr. Davies's summary — Table of matings — a suggestion. Cats : Previous work — Author's "working hypotheses " — Table of matings — Discussion of results — Alteration of Mendelian proportions by Vigour — Pedigree of male Tortoiseshell — Eecessives carrying dominants — Abnormal result in Exp. 4 — Production of Tortoiseshell male. CHAPTER VII. EATS 61 Primary object of experiments— Notes on management and habits in captivity — Abnormalities and "sports" bred — Fawn Rats — Description of original wild stock — Characters investigated and results — Inheritance of fawn variety — Conditions giving rise to new characters — Nature of the change in environment and its effect on metabolism. CHAPTER VIII. MBEIONES 77 Management — Period of Gestation — Proportion of sexes — Number of young at birth — Fertility — Colour — Effect of damp heat — Differ- ences in successive generations — General remarks on inheritance of colour — Measurements — Skull characters — Consideration of results and their bearing on Vigour. CHAPTEE IX. PIGEONS 97 Webfoot character — Proportion of sexes — Proportion of sexes in Silvers and Mealies discussed — Sex limitation in Silvers and Duns — Production of Dun male on Vigour lines — Colour — -Deviation from Mendelian Theory — Increase of white in successive generations — Colour and Sex — Possible cause of sexual dimorphism — Period of development at which sex and colour is fixed — Part played by Vigour and Mendelism respectively. CHAPTEE X. DUCKS Original object of experiments — ^ Fertility — Sex — Coloration — Description of P. M. S. trigens and general remarks — Other crosses described and compared — Differences in crosses with the same blood proportions (Galton's Law) — Variations — Loss of colour — Possible reversion — White under parts — Grey Mutation — White Mutation — Summary — Discussion of results, 122 CONTENTS xi PAET III.— DISCUSSION. CHAPTBB XI. THE EVOLUTION OP SEX ....... 155 Messrs. Geddes and Thompson's hypothesis — Katabolism and Anabolism — Definition of Vigour — Apparent objections to sex and Vigour answered — Effect of nutrition and temperature — Effect of Vigour on sex — Initial Vigour ; of ova, of spermatozoa — Fluctuating or Enviromental Vigour — Inheritance of acquired Vigour — Cumu- lative effects — Nuclear and protoplasmic elements of the gametes — Inheritance of blends — Infertility of hybrids. CHAPTER Xn. PSYCHOLOGY OF EEPBODUCTION 186 Necessity of stimuli for actual reproduction among Mammals — Among Birds — Flamingoes — Conditions in Greenland — Eider Ducks — Sparrows — Ducks and Finches in confinement — Bird's song an expression of Vigour — Song periods critically examined and considered in relation to plumage and environment. CHAPTEE XIII. THE MENDELIAN HYPOTHESIS 196 Mendel's Law — DifSoulties for its complete acceptance — Sex and colour in Pigeons — Dominants from apparently pure recessives — Webfooted Pigeons — Formation of new characters with a Mendelian inheritance — Eeversion — Mendelian proportions affected by Vigour — Domestic breeds and wild races compared — Direct effect of Vigour on colour — Somatic expression of determinants modified by Vigour. CHAPTEB XIV. GALTON'S law CONTRASTED WITH MBNDBL's . . . 215 Both theories true in certain cases — Galton's Law in terms of Vigour — Dragoon Wattles — Tortoiseshell Cats. CHAPTEB XV. DISCUSSION OP OTHER HYPOTHESES , 224 Dr. Archdall Reid's hypothesis — Patency and latency main function of sexual inheritance — Single or paired characters — Mr. Cunningham on secondary sexual characters — Not necessarily connected with sexual organs — Messrs. Geddes and Thompson's hypothesis criticised — Sexual selection criticised — Period of development of sex — Mendelian theory of sex — Mr. Cunningham's hypothesis — Chemical changes —Hormones. CHAPTEE XVI. SUMMARY AND CONCLUSION 244 Chemical change the basis of life — Fundamental difference between the sexes — Effect of Vigour — Possible future results. LIST OF ILLUSTRATIONS. PLATE I. "Toby" to face page 50 II. Mus alexandrinus and Mus tectorum . . „ 64 III. Eats used in "Pawn" Expeeimbnts . „ 70 IV. Living Meriones crassus .... „ 77 V. Skulls of " Meriones " . . . . „ 93 VI. Pigeons' Webbed Feet .... „ 98 y-rj ( PiNTAIL-MalLAED-SpOTBILL, PiEST GbNESA- I jng I TION, FULL PLUMAGE . . . j " Ylll (PiNTAIL-MaLLAED-SpOTBILL, FieStGeNEEA-] -.no 1 TION, ECLIPSE PLUMAGE . . . j " j-^ |Pintail-Mallaed-Spotbill,ThiedGenbea-] ..„« 1 TION, FULL PLUMAGE . . . j " X. Heads op Teigens ,, 140 XI. Heads op Pigeons to show Wattles . „ 216 Eeathbes feom a Young Dun Pigeon .... 211 Goat Pedigrees 54 Table showing inheritance of Mus alexandrinus and Mus tectorum Table showing inheritance of Pawn character Genealogical Table of Ducks 66 70 124 Hypothetical Table to show changes of Vigour in Wild Duck 168 ,, „ ,, ,, Polar Bear 169 Eat . . 170 PAET I.— INTEODUCTOEY. CHAPTEE I. INADEQUACY OF EXISTING THEOEIES. One of the greatest fields for zoological investigation at the present day is the study of Heredity, and we may even be tempted to say that it is the only road along which a permanent advance is likely to be made. Gould we know the laws governing heredity, many of the most insistent problems which to-day confront the zoologist would be to a large extent settled. For instance, the inheritance of acquired characters, that long-debated question which, although not a burning argument at the present time, needs nevertheless much further elucidation, the evolution of geographic species or sub-species, the question of many very closely allied species inhabiting the same localities and under the same condi- tions, and yet differing slightly and constantly without intermingling of the characters, the ready adaptability of some species to changed environment, the plasticity of some species, the immutability of others, the occurrence of sports, the infertility of some hybrids compared with the fertility of others, the difference between the sexes in many species, these and many other problems await a satisfactory solution, and it cannot be doubted that a close study of the laws of heredity is likely to materially help to solve them. During the latter half of the last century, although the general interest in zoology became more widespread, it can- not be said that much real advance towards the solution of these questions was made. Darwin's theory took the world by storm, and gave an enormous impetus to the seeking out and describing of new forms of life — a necessary work, but it is to be feared so absorbing as to have caused the main laws which govern zoology to have been overlooked, and species have been merely described and figured without, in many B 2 VIGOUE AND HEEEDITY cases, more than a passing thought being given to their antecedents and the laws which brought them into being. With the advent of the twentieth century an advance was at once made in the rediscovery of Mendel's Law, and the " science of genetics," as it is now called, is receiving widespread attention from zoologists all over the world. Mendel's Law, as is now well known, is founded on the purity of the gametes; each individual is assumed to result from the development of a conglomeration of paired characters, and, as every individual springs from two parents, he will contain the properties of one pair of characters (one from each parent) as regards any particular part of his body. Thus, for instance, if a web-footed Pigeon be paired to a normal footed bird, each individual of the progeny will contain the character for web foot and the character for normal foot, and similarly pairs of characters for all other portions of his body. Such an individual containing a dissimilar pair of characters is known as heterozygous in respect to that character, and it follows that if two heterozygous individuals are bred together, the offspring will be of three kinds, heterozygous like the parents, or homozygous {i. e. containing a pair of similar characters), or either one or other of the characters. Homozygous individuals must breed true. This hypothesis of course refers solely to the gametes ; the external appear- ance may differ considerably — in a heterozygous individual one character will often swamp the other — or an entirely different appearance may cover externally the gametical properties of another pair, though the inheritance of the " covered up " pair will still continue as though it was displayed. Such very briefly is Mendel's Law, and on these lines many apparently complex forms of inheritance have been satisfactorily explained by scientists. To the practical breeder, however, the matter will not appear so simple. He will know of countless cases which crop up yearly with him, and are apparently in direct contradiction to the main facts of Mendelism ; he will know of cases of prepotency, in which certain individuals transmit their sex or their colour to an inordinately large percentage of their offspring; he will know the value of ancestry; he will know how in some cases an outcross may apparently be bred out in a few generations, and in others how the effect of an INTRODUCTOEY 3 outcross, although apparently bred out and absent for generations, may again suddenly appear ; he will know how certain colours tend to breed out to white in successive generations, and the almost impossibility of breeding his stock true to any particular shade. To the thoughtful zoologist, also. Nature herself offers many difficulties to the unconditional acceptance of Mendel's Law as the sole basis of heredity. For instance, the presence of geographic races breeding absolutely true (or nearly so) to a particular shade of colour (a matter, as pointed out above, yet unattained by breeders). If any race originally arose as a sport, one shade must have been dominant to the other. The recessive one might certainly breed true, and it is possible that the dominant colour would be exterminated by Natural Selection, but in some geographic races surely the dominant colour would be the suitable one, and it would then require untold ages before the heterozygous but externally dominant individuals were all exterminated by Natural Selection. Or, again, are we to believe that in animals inhabiting the desert the necessary "sand coloured" mutation was always present ; but even if this were so, and all animals that had tried to inhabit the desert and had no "sand coloured " mutation had been killed off, how are we to account for the difference of shades ? One species may be pale coloured and another nearly allied species reddish, yet each will live side by side in the same conditions and in approximately equal numbers ; their enemies are presumably the same, and yet their extermination does not evidently depend on their colour ; go some miles to the south, the paler one is now much more rufescent, the reddish one is unchanged.* Mendel's Law and Natural Selection are not sufficient to account for this. We must allow that environment, or, as I would prefer to call it, " climate," is the cause, but here we are met with another difficulty. Climate and environment are obviously external agencies, and consequently the effect they have on the animals exposed to them must come under the heading of "acquired characters," and as such cannot, according to present-day thought, be inherited. Consequently, each new * Gerbillus gerhillua and Qerhillus tarabuU from Egypt and the Sudan are examples of this. B 2 4 VIGOUE AND HBEBDITY generation must receive those characteristics de novo, and, therefore, if the conditions be altered, the animals should have assumed by at least the second generation a different colouring. We must, therefore, come to the conclusion that Mendel's Law is not of itself sufficient to account for many v?ell- known facts of inheritance ; Galton's Law, though holding true for some cases, has been absolutely disproved in certain simple cases of Mendelian inheritance, and we are thus driven to the fact that the effects of environment must be inherited. This last hypothesis, however, is hardly to be accepted as it stands, since up to a certain point the case against the inheritance of acquired characters is well substantiated. Can we, therefore, in any way reconcile these laws and bring any new factor into consideration which, without breaking down any of the more or less established laws, can put them in their proper perspective, and at the same time satisfactorily account for some of the difficulties both in Nature and in artificial breeding which I have just outlined ? I have bred animals of one kind or another most of my life, as well as studying Nature as a zoologist from a scientific standpoint, and I can speak from personal knowledge and experience of the difficulties raised. The solution I propose to offer is that all these matters are due to certain physiological causes which, during the course of this book, I shall bring forward and define under the name of "Vigour." This term has previously been used more or less vaguely by various scientists,* and especially by Dr. A. E. Wallace in his explanation of Darwin's theory of sexual selection. In fact, it will be seen that many scientists have vaguely felt the presence of this law, and I hope that the years of study that I have devoted to it will prove that it is of fundamental importance in governing heredity and evolution. * The following writers may be said to have written on the subject from this standpoint, viz. Bateson, Barrett-Hamilton, Beddard, Cunning- ham, Geddes and Thompson, Newbigin, Taylor and Wallace. CHAPTER II. OUTLINE OF VIGOUE. What is meant by Vigour? By vigour is meant "activity of nutrition and function," i.e. a state of high metaboHsm, and where the rate of metabolism in one individual is greater than in another the vigour is said to be higher. A not inapt simile of vyhat is meant by vigour may be taken from two clocks, in one of which the ticking is twice as fast as the other — ^viewed externally the hands of each do the same work in the same time, but internally the mechanism of the one works at double the rate of the other ; and so it is with animals during their lifetime, they each perform similar functions, eat, sleep, and reproduce their kind, but the metabolism of the one may be twice as active as that of the other. How, then, can we tell the amount of vigour present in any individual? At present we have no index, and no method of definitely ascertaining or measuring the rate of metabolism; but by actions, colour and condition we can get a very fair idea of the vigour of any individual. Before, however, dealing with that point, which is more cognate to experimental breeding, let us consider the causes which determine vigour. First and foremost, we have the vigour of the parents, and the combination of their respective vigours gives an animal what we may call its " initial vigour." This is the vigour with which the zygote starts life before environ- ment in any shape or form has begun to act. This " initial vigour " depends upon the rate of meta- bolism of the parents at the time of copulation, and this rate of metabolism or vigour will depend on their (the parents') initial vigour modified by the environment in which they have lived.* The initial vigour of any animal will bear a fixed relation to the mean vigour of its parents, and colour is, as * See Chapter XI. 6 VlGOtJE AND HEEEDITY I shall attempt to show later on (Chapter IV.), largely an index of vigour ; for the present, however, let us assume it. If two parents have the same vigour, their offspring will also be of that vigour ; if, however, the vigour of both these parents be raised to the same extent by favourable conditions, then their second offspring will have a higher initial vigour than the previous offspring, since the initial vigour of the offspring varies directly as the vigour of their parents at the time of copulation. Now the extent to which vigour can vary owing to the environment probably bears a fixed relation to the initial vigour, that is to say, it can only vary to a limited extent round the initial vigour as mean, so that if an offspring be born to parents at their highest limit of vigour, such off- spring will have a higher initial vigour, and be always capable, under suitable environment, of reaching a higher limit of vigour than was possible either to its parents or to any other offspring born when the parents' vigour was lower. Consequently it follows that a rise of vigour in the parents, brought about by favourable environment, is to a certain extent passed on to the offspring of the first genera- tion, and, if the conditions remain stable and favourable, inherited by the offspring of the second generation. If the vigour of any two parents be widely different, then the initial vigour of the offspring, which varies as the mean of the parents' vigour, will be intermediate between that of its two parents, and hence, if colour be an index of vigour, the offspring must be of a shade intermediate between that of his two parents, or, as we say, is a blend of their two colours. Suppose we carry this matter on a generation further, the offspring of the first generation, which are in colour inter- mediate to their parents, will all be of the same vigour, and their offspring will therefore also be of the same vigour, and will resemble their parents in colour, and be of a shade intermediate to that of their grandparents, or, in other words, the blend will apparently have been inherited. We must, of course, remember the Mendelian Law, and this hypothesis is not meant to deny or contradict that law, but rather to qualify it. In some apparent cases of continu- ous variation the followers of Mendel note certain points, which they fix as characters or points of discontinuous OUTLINE OF VIGOUR 7 variation, while, the intervening varieties they regard as " fluctuating variations " caused by soil, climate, food — in other words, environment. These "fluctuating variations" are not supposed to be inherited, the Mendelian character being the only inherited part; my argument is that the fluctuating variations brought about by environment are due to differences of vigour. I shall try to show how vigour affects the offspring, and how we thus get superimposed on the Mendelian characters an heredity brought about by environment. As long as we believe that the causes of environment have no effect on the offspring, it is at once obvious that we need not in our matings worry about the differences of "fluctuating variation"; we shall look upon a character and its "fluctuating variation" as gametically identical, and need not worry from which end of the fluctuating scale we choose our matings, and, in recording the results, we can ignore many of these slight variations. On this reasoning Mendel's Law holds good for many cases ; but if, on the other hand, we are always selecting, let us say, the tallest amongst fluctuating variations, then, provided the environment from which those tallest came is maintained we shall gradually raise up a taller and taller race until the initial vigour becomes the highest that can be produced by that environment — the parents of each generation will still be in the centre of the fluctuating varieties, but each generation will have advanced by the increase due to only choosing the parents from the top of the scale. This has been the method of breeders for past years — and who will venture to doubt its efficacy ? Unfortunately, Mendelians in proving and working at their theory ignore the slight differences of "fluctuating variations," so that a reference to their records would throw no light or very little on the slight differences in the heterozygote offspring: little differences which, being due to vigour, are, I maintain, inherited. Prepotency. — It may have occurred to the reader that the " complete dominance" of Mendelism is thus, as far as the F^ generation is concerned, only another term for "prepotency" ; the Mendelian Law, however, proves its truth in the next or F^ generation, and it will perhaps not be out of place here to enquire if the vigour theory can in any way explain this. 8 VIGOUE AND HBEEDITY It is obvious that if the offspring, F^ generation, started with the same amount of vigour, and, if they were to breed at that vigour, the initial vigour of the F^ generation would still be the same, and the "ids" of both the parents would be equally strong, and consequently the characters would be more evenly blended, and the prepotency would not neces- sarily be marked in any way. (I am only referring here to the Mendelian heterozygotes ; the pure dominants and re- cessives would of course appear as usual.) Suppose the initial vigour of the father very low and of the mother very high, the initial vigour of the offspring will then be intermediate; but it is more than possible that, where the difference in vigour between the parents is great, the characters of the parent with weaker vigour are unable to develop,* and have, therefore, to lie latent (though not by any means to necessarily become lost).t This is, I think, a possible explanation of some forms of prepotency, which, as we can see, may therefore be complete or partial. It is, however, more likely to be partial, since, as the vigour of the offspring is intermediate, it is unlikely that all the characters of a parent with a high initial vigour should develop : they- may all get a start ahead of those of the weaker parent, but only some will reach their full develop- ment. Let us, however, trace the case a little further; if these offspring be paired at their highest vigour, then their children will have the same initial vigour, and are therefore more likely to resemble their grandmother, since their initial vigour is nearer to hers than to that of their grandfather, but as both the parents' germs start equal in vigour the characters of the grandparents will be more blended, and the prepotency (which in this case does not really exist at all) much less marked. Inbreeding. — Another point to be considered is In- breeding, and how this affects vigour and inheritance. Inbreeding is popularly supposed to have two effects — * For further details, see Chapter XI. t The action may perhaps be roughly compared to a tree or plant in which the main stem or leader has been out ofif — several branches may try and grow, but as a rule one eventually gets the lead and the others remain undeveloped — they do not die, and are still potentially capable of development, but their development is retarded and crushed owing to the lead gained by the foremost. ■ OUTLINE OF VIGOUE 9 (a) weakening of the stock, {b) increased stability of the characters. One of the most marked cases is the fancier's so-called " light mules," i. e. Canary hybrids showing a preponder- ance of the yellow Canary colour; this is brought about by using " sib bred" (inbred) Canary hens, and hence a reason for the belief in the stability and prepotency of inbred characters. Without entering into details for the present, there is much evidence to show that inbreeding causes a general and gradual loss of vigour leading to infertility ; this, I think, we may assume, though it must not be forgotten that some animals may be inbred for several generations without any apparent harm. We may, therefore, assume that an inbred female has considerably less vigour than an ordinary one, hence when mated to a wild bird the vigour of the resulting hybrid is, compared with that of the male parent, low, and consequently we should expect a light coloured bird ; now since in this case the colours do not blend, we shall as a result of this low vigour get a bird which is chiefly yellow but shows patches of colour, and this is the "light" mule of fanciers, the coloration of which we may reasonably suppose to have been brought about by vigour, and not owing to an increased stability of inbred characters. Inbreeding does undoubtedly lead in many cases to infertility, and this infertility must necessarily take place if there is decrease of vigour. (See Chapter X. on Ducks.) The actual amount of fluctuation that is able to take place round the initial vigour as mean cannot in the present state of our knowledge be defined with exactitude, as we have no definite unit of vigour, or any means of measuring it ; the arguments, however, hold good none the less and give a clear and intelligent idea of the phenomenon. We shall enter into more detail giving concrete examples in the succeeding chapters, and formulate the whole hypothesis more minutely in Chapter XI., the object of the present chapter being merely to give the reader an idea of the lines on which we are working. As I point out in my remarks on vigour and zones (see Chapter III.), an animal is so constituted that if his vigour rise too high or fall too low, he dies, consequently there is a state of vigour either high or low, which means death to the individual, although the actual amount of vigour beyond 10 VIGOUR AND HEREDITY or below which life is impossible will depend upon the initial vigour of each individual and the amount of variation possible from that vigour. Before, however, the death limit is reached there is a stage when the sexual activities stop and the animal becomes temporarily infertile. All breeders know well that breeding is impossible with an impoverished and out of condition animal, hence, if we produce an animal whose vigour is below a certain limit, it will be unfertile. Can this be brought about by inbreeding on our vigour theory ? Yes, given conditions unfavourable for vigour in each succeeding generation, but inbreeding under favourable conditions will not necessarily lead to infertility. Suppose both parents have an equal vigour to start with, then their offspring will have a similar vigour ; now if the conditions are not very suitable, the offspring's vigour will fall slightly, and, breeding at that lower vigour, their offspring would start at a lower level, and under the same bad conditions fall still further and so on, so that in this case inbreeding would soon reach the limit of vigour results in infertility, but that infertility would have been brought about not so much by inbreeding but rather by using the same stock brought up on the same general conditions for low vigour. On the other hand, the conditions might be those of high vigour, and in this case we should similarly have a progressive change resulting in infertility, and even- tually in death. This, to my mind, gives a very reasonable explanation of inbreeding and its results, for unless the conditions are exactly normal for the vigour of the first parents, the inbred stock must inevitably go steadily up or down until infertility puts an end to it. Any fresh blood introduced will obviously tend to raise or lower the standard, except in the very improbable case of the fresh blood having the same initial vigour as the inbred stock, for it is probable that by the time the stock is inbred sufficiently to need fresh blood it has got below the normal initial vigour of the average. We are, of course, dealing with this matter very roughly, and in practice results will not be so definite ; no two individuals have as a rule exactly the same vigour, consequently the infertility will be reached in some strains sooner than in others, but the general trend is the same, and this, to my mind, gives OUTLINE OP VIGOUR 11 the best and simplest explanation of the different results attained by breeders when inbreeding their stock. Infertility. — There can, I think, be no doubt that in many cases infertility is due to loss of vigour. The com- monest and most widespread examples of this are wild animals in confinement, which are in most cases notoriously infertile ; as a rule, many of them show by their coat or plumage their lack of vigour, and their infertility may well be put down to that cause. This is also shown in the reverse way, for some species, e. g. Quails and possibly Pigeons, which adapt themselves well to conditions of captivity, far from being infertile show an increased fertility; and one may perhaps instance Lions, which not only breed freely but show their increased vigour in their finer appear- ance and more shaggy mane. The number of Lions reared is not very great, but this is due to other causes — difficulty of handling, and destruction of the young by the mother, &c. Hybrids are another group of notoriously infertile animals ; this infertility is probably due to other and more deep-seated causes, chiefly the non-development of the generative organs ; why this should be so it is not our province to enquire here — the fact remains that some hybrids are perfectly fertile, others are not. Among fertile hybrids the birds or animals of any particular cross usually closely resemble each other; in infertile hybrids we find much more variation, and they are often highly coloured. The sexual organs form a great safety-valve for using up and lowering the vigour, and when the sexual organs are active the whole animal has more stability and works more normally, so that animals of those crosses are generally more or less like each other ; among infertile hybrids, however, the vigour which should be used up by the sexual organs has to find some other means of working itself off, and this is generally shown by larger size or more brilliant colouring — sometimes favouring one parent and sometimes another — but it does not lead to an aggrandisement of a special secondary sexual character. What, then, is the cause of these fluctuations of vigour ? Briefly, environment, of which the main factors are : — (1) Temperature, (2) Humidity, and (3) Food. 12 VIGOUE AND HEREDITY It may be shown that the general coloration of mammals and birds throughout the world is governed by these factors. In the polar regions the change from summer to winter is marked and severe, and its effect on the organism is apparent in a marked summer and winter dress, the latter being usually white, while the former is fairly dark and rich, agreeing in its general coloration with that found in more temperate regions. Whereas in the tropics the whole tone of coloration is brighter and seasonal changes of pelage far less marked. Before, however, dealing more fully with this aspect of the question, let us briefly consider vigour in relation to sex. ViGouB AND Sex. To what extent is sex controlled by vigour? One thing must be borne in mind carefully, and that is that katabolism and anabolism* will exist with either a high or a low state of vigour. In this connection an animal may be compared to a steam engine, in which vigour represents the driving force or energy, e. g. the steam. In the case of katabolism the first safety-valve is much larger than in the case of ana- bolism, so that a much greater portion of a male's vigour is spent in energy and colour, while the female husbands her resources, or many of them, for sexual activity ; or, in other words, her generative organs need a greater pressure for their development, and so have a greater demand on the products of her vigour, leaving less for outward show. Can, therefore, the amount of the parents' vigour have any effect on the sex of the offspring ? The answer is undoubtedly Yes. If the vigour of both parents be low the offspring will probably be male, or, if a female, it will pro- bably not survive. There is at present not enough evidence to show whether the actual sex may be influenced by low vigour and only males born, or whether those young pre- destined to be females (if we are to accept the Mendelian belief) die. There is a certain amount of evidence on both sides. Yungs classic Tadpole experiments (which we may note were obviously influenced by artificially reducing the vigour) seem to point to the fact that vigour can influence * This question is more fully dealt with in Chapter XI. OUTLINE OP VIGOUE 13 sex ; on the other hand, I found that for some years I reared a very large proportion of male hybrid Ducks, but the broods were small, many young dying off. I then fed the young more generously, and found that I was able to rear full broods, and that the sexes were nearly equal in numbers. This, therefore, seems to point to the determina- tion of sex by something more than vigour. Possibly and probably the truth lies in the happy mean — if the parents' vigour is very low only males will be produced, but if their vigour be slightly higher both males and females may be produced; but the females will require so much more vigour to develop that, unless kept in specially favourable conditions, they will not be able to survive. Some broods of my hybrid Ducks showed an interesting relation between colour and sex, which is well accounted for by this idea of vigour. These particular birds (Pintail, Mallard, Spotbill trigens *) were of two types, a light and a dark, and this variation was common to both sexes. Light Drakes appeared in almost every brood, but no Duck would appear unless there were both light and dark Drakes, and no dark Duck unless there were both forms of Drakes and light Ducks, though sometimes when dark Ducks appeared light Drakes would be absent. No. of Brood. Males. Females. 4^ ^^ 3 light 1 dark . bs- 3? 2 light 3 dark 1 light 2 dark ^$ 4? 2 light 1 dark 2 light 2 dark 1$ 4? — 1 dark — 4 dark 3^ — 3 light — — — 1$ 12 — Idark 1 light — ^s- 22 2 light 1 dark — 2 dark A^ 22 1 light 3 dark 1 light 1 dark 23' 42 1 light 1 dark 1 light 3 dark 1$ 42 » 1 dark 4 dark These actual figures show the general trend, and in three out of the four cases where the females were in excess * See Proc. IV. Int. Orn. Oongrees, 1905, p. 254, and also Chapter X. 14 VIGOUE AND HEEEDITY dark birds were also in excess, and in five cases where the males were in excess only two showed an excess of light. We must, however, remember that these were not always full broods, and the colour and sex of those that did not survive might turn the balance. This point should not, therefore, be pressed too closely ; but the evidence, such as it is, tends in the direction indicated. On my first return from Egypt in the spring of 1909 a pair of wild Eats that I had just brought home had a litter of young. The weather was cold for the time of year, and they had just been removed from their travelling box, both conditions not being conducive to high vigour ; the brood was a remarkably backward one, and consisted of seven males, three of which were of a light fawn colour. That pair did not breed again — another significant fact, and pointing to low vigour. Broods from other pairs born later in the summer produced the sexes in fairly equal numbers. In a paper I had the honour to read before the Linnean Society some years ago, I gave "the rise and fall of sex activity" as one of the main causes affecting vigour, and, if that be true, the question at once arises as to what affects " sexual activity." In point of fact this state- ment requires correction, for we now know that it is vigour itself, together with a psychological stimulation, that in- creases the activity of the sexual organs, and that unless animals be in a moderately high state of vigour they cannot breed. This is further exemplified by the sterility of many, if not most, species in confinement, and even among domestic races breeders often know of cases where an individual may not have bred for several seasons, and yet on being sold and sent to a new owner has immediately proved fertile ; now, in such a case the only difference has been a change of food and climate, which, as we have just shown above, undoubtedly influence the vigour ; hence we may well assume that, by getting an otherwise healthy animal into a high state of vigour, we bring on an activity of the sexual organs. Now, it is also a well-known fact that breed- ing has a weakening effect, and thus reduces the vigour of an animal, and so it comes about that at the end of the breeding season an animal's vigour is considerably reduced. OUTLINE OF VIGODE 15 In this condition many animals moult, and their new coat or plumage is, with hardly an exception, duller than that assumed prior to the breeding season, when their vigour was high. Examples of this are most easily found in birds, though they are by no means unknown among mammals when the breeding season is well-defined. In all birds, however, the breeding season is of comparatively short duration, and in consequence we find their breeding plumages well marked. As a good and typical example of this, let us take the Knot {Tringa canutus), a very common and numerous species of cosmopolitan range. After the breeding season it assumes a dull grey dress in conformity with its low state of vigour. During the winter it inhabits the tropics in a warm and genial atmosphere' and with an abundance of food, and in spring, having reached a high state of vigour, it assumes a gorgeous livery of red and black. Its vigour, however, still remains high, and since high or surplus vigour must have some means of ebullition, it becomes expended in nesting cares. In confinement the full brilliancy of the red dress is never assumed, nor has it been known to breed, because the conditions were not suitable to a sufficiently high vigour. 16 VIGOUR AND HEREDITY CHAPTEE III. THE COLOEATION IN MAMMALS AND BIEDS. Although the literature on colour and coloration in the animal kingdom has reached huge dimensions, and the subject is one which has occupied zoologists from the earliest days, yet we are still a long way from a complete under- standing of the causes and use of colour. On the one hand, chemists and physiologists have re- stricted their investigations to the extraction and analysis of pigments, and, on the other hand, the majority of zoplogists have studied coloration from the standpoint of its utility to the organism in its environment. Few, however, have considered the fact that colour has probably its primary cause and utility in satisfying some physiological need of the animal, while Natural Selection has come in secondarily and eliminated the unsuitable, or perfected those colours and markings that were able to be adapted for purposes of protection, warning, &c. It is not my purpose here to enter into a long discussion, pointing out special cases where Natural Selection seems to have failed, or where its advocates seem to have overstepped the limits of probability. This has already been done by many writers, and although they have partially succeeded in showing that Natural Selection is not sufficient to account for the cause of coloration, yet their work has been, in the main, destructive rather than constructive, and it is pro- bably owing to this tendency that the physiological aspect of colour is not more widely accepted to-day. I do not, therefore, propose to deal in any way with Natural Selection, but rather to try and show that colour is primarily due to the vigour of an animal, so that where we fihd conditions suitable to a high state of vigour, we shall in those cases find a cbrresponding increase in the colour. Those who have kept and studied live animals know that the state of the coat or plumage is an unfailing criterion of OOLOBATION IN MAMMALS AND BIRDS 17 health or sickness, and that without any moult a bird, on recovering from an illness, becomes much brighter and more glossy, and I know of a case in which the black plumage of a bird became quite brown during sickness, and regained, to a limited extent, its black and glossy appearance with the return of health. We may, therefore, take it for granted that, when an animal's health becomes in any way affected, the • change will be visible externally in the animal's coat or plumage. Vigour, being the rate of metabolism, any circumstances bringing alDOut a physiological change in an animal (whether pathological or otherwise) will affect its vigour, and there- fore if alterations in an animal's health are reflected in its coat or plumage, as we know to be the case, alterations in its vigour will also be indicated in its coloration. As we mentioned in the previous chapter, an animal's " vigour " is affected by environment, of which the main factors are : — (1) Temperature, (2) Humidity, (3) Food supply. And considerable proof will be given to this statement if we can show that, where the environment presents marked seasonal changes, the difference in the coloration is also marked. Let us, therefore, consider the mammals and birds according to the regions (Polar, Tropical, or Temperate) in which they live. 1. The Polar Begions. — In the Polar regions the period of extreme cold and scarcity of food is so long as compared with the short, bright summer, and the transition from the one to the other so abrupt, that the inhabitants thereof will be subject to violent changes. The winter follows so quickly after the breeding season that there is no time, as in more temperate regions, for the animal to recuperate after the lowering effects of sexual activity. Such reserve material as is formed will, therefore, be stored up in the form of fat, a condition in which it is most accessible for any subsequent demands that may be made upon it by the body. This storing up of fat, as 18 VIGOUR AND HEREDITY Captain Barrett-Hamilton has shown,* is one of the concomitants of sluggish metabolism. Here, therefore, we have conditions of low vigour, and in correlation with them we find the animals for the most part white or grey. In the spring the higher temperature and more plentiful food, and the more genial conditions generally, will tend to a higher state of vigour, but it must be remembered that the privations of the winter have so lowered the animal's condition that it is unable to derive full advantage from the more favourable circumstances, and its vigour is, there- fore, unable to reach a high level. In correlation with this we find a predominance of duller colours amongst the majority of polar animals. How, then, are we to account for the darker animals of these regions, as, for instance, the Musk Ox, Penguins, and Eaven ? If our suggestion holds good, the fact of these animals being dark coloured in these regions proves that they are able to maintain a fairly high vigour throughout the year, in spite of the cold and scarcity of food.t Now, if this be so, we ought to find these animals throughout the world equally highly coloured or more so; or it is conceivable that if their " vigour " is so great in cold localities, it might become too great in warmer regions, and being unable to find any outlet for this excess (the animal being already highly coloured) death might ensue, + and such animals would, in consequence, be restricted to the colder portions of the world. The Eaven is a good example of the first of these conditions, for the Corvidas exist practically unchanged throughout the world, and the Musk Ox and Lemming fulfil the second. The Mouse Hares (OchotoTia), although Alpine rather than Arctic, give us a good idea that the general tone of colour in two allied species may be different * Proc. Boy. Irish Acad., vol. xxiv. sect. B., pt. 4, p. 307 (1903). t It might be argued against this that, in these cases, the colour having been placed in the hair, it could not be withdrawn, and not being harmful, but even, perhaps, useful (as has been suggested for the Musk Ox), the animal was enabled to survive. If these animals only moulted in spring, such an argument might hold good ; but in the case of the Eaven, the moult is in autumn, and I fancy the same holds good for the Peuguins. I The excessive energy of white men on first coming to the Tropics is a frequent cause of their over-taxing their powers, so frequently culminating in drink or fever. COLOEATION IN MAMMALS AND BIEDS 19 although they live under similar climatic conditions. This difference may be explained if we accept the differences of colour as being due to each species having a different vigour level. For example, two species, 0. Ladacensis and 0. Boylei, inhabit adjacent countries where the climatic conditions are very similar. 0. Ladacensis is white, or nearly so in winter, and light brown in summer; whereas 0. Boylei, which becomes bright red in summer, never becomes paler than iron grey in winter. So that here we have two nearly allied species inhabit- ing the same country, each equally influenced by a similar climate which causes them to undergo a seasonal change,v yet in winter one is white, and the other, in defiance of the laws of protection, dark. Captain Barrett-Hamilton has also brought to my notice another similar instance in the case of the Siberian Lemmings {Dichrostonyx and Lemnus) , where the light or protectively coloured species is much scarcer than its more conspicuous relative, which is, in fact, to be expected under our hypothesis, as the light-coloured one should be less robust. The Penguins, again, are birds whose nature is such as to enable their " vigour " to maintain a high level in Antarctic regions at all times of the year. It is not surpris- ing, therefore, to find several species, e.g. the Emperor Penguin, showing patches of bright colours, such as yellow, orange or red, which, if the colour be an index of vigour, we should expect to be the case. Thus we get further evidence pointing to the fact that the climate only influences " colour " through vigour, and that where the rate of metabolism is high the animal will be brightly coloured whatever may be the climate. If, therefore, our hypothesis be true, we should expect to find the vigorous, dark Arctic species breeding earlier than those that are paler coloured, and in confirmation of this we find the Eaven breeding very early in April, while the Ptarmigan and Grouse wait till the latter end of May. The Lemming is another instance, for it is a brightly coloured animal, and in certain seasons produces young more numerously and rapidly than any other Arctic animal, while as a result of this excessive " vigour " we find it most pugnacious and energetic. It is also an early breeder, c 2 20 VIGOUE AND HEEEDITY and is able to maintain a high vigour throughout the year, since in winter it lives under the snow, where it finds abundance of food and is well protected from the cold and winter storms. Let us now study for a short time the migrant birds that breed in Polar regions. These may be divided into two classes : — (1) Those from the Temperate regions, which gradually breed as far north as is consistent with their needs. (2) Those that winter in the Tropics and breed in the extreme north, but are only found in the intervening region on migration. We need not notice the first class here, as they belong properly to the north temperate region, from which they are stragglers. The second class consists, for the most part, of bright- coloured birds that assume their bright livery in the Tropics, and that wear in winter a duller dress, assumed in northern regions.* The Lapland Bunting (jOalcarius lapponicus) is a good example of these, for the cock has a bright black and red breeding plumage, which is much more intense than that of its near ally, the Snow Bunting (Plectrophenax nivalis), which has a much more northerly winter habitat. Another good instance of a bright-coloured northern region bird is the Eed-breasted Swallow {Hirundo erythro- gaster), which passes the summer in North-East Siberia, and the winter in Burma and countries to the south. This bird does not, as far as I am aware, moult in these northern regions, and consequently, as we should expect, retains its bright red colour, which was assumed in the Tropics (where the temperature and food are conducive to high vigour) throughout the year. It would be impossible to enumerate all the birds which may be considered in this class ; the Limicolse offer many examples, and the Knot (Tringa ccmutus) may be taken as very typical. * It has been poin1>ed out to me that some birds migrate in autumn before they moult, which may sometimes undoubtedly be the case, although I believe that in the majority of instances this is due to their being driven south by early storms before the moult has been completed. In any case, the moult would follow so soon after migration that their system would not have had time to respond to the more genial influences of a southern clime. COLORATION IN MAMMALS AND BIRDS 21 This bird assumes in early spring, when in the Tropics, a very deep chestnut and black plumage. It then migrates to the extreme north to breed. After breeding it moults and becomes a dull greyish white bird, so that both its plumages are in keeping with the colour characteristic of the regions in which they are assumed. Now it has long been a puzzle to me why * the young Knot, bred and reared in the north, should yet be able to assume a browner and pinker plumage than its parents, although in Arctic regions. The explanation may, however, possibly be due to the fact that when the old birds moult they have had a long journey immediately followed by the strain of the breeding season, compressed into the short Arctic summer, so that we would expect their vigour in autumn to be very low; whereas, on the other hand, the young bird has only had to grow, which is probably a far less strain on its system than that under- gone by the parents. The result is that the young are enabled, although in Arctic regions, t to assume to a slight extent a colour belonging rather to the Tropics. This case is, to my mind, of considerable importance as emphasizing again that colour is far more dependent on " vigour " than on the particular latitude in which it is produced, so that if an animal be sufficiently vigorous it will assume bright colours, although in northern latitudes. Knots when kept in captivity rarely, if ever, assume the full deep chestnut which they do when wild, the amount of change depending largely upon the individual. For two winters I endeavoured, by keeping these birds at a fairly high temperature, to induce them to assume their full red colour, but it produced no appreciable effect ; showing, therefore, that heat has', in this species at all events, but little power. The Golden Plover, at first sight, offers rather a puzzle ; for in this species the birds, which nest in the southern limit of its range, are duller and in less perfect plumage than those to the north, which is apparently exactly the reverse of what ought to take place. It has, however, * For further suggestions on this point see P.Z.S. 1901, p. 325. -j- Of course the tendency of the species to assume red must be taken into account, as it causes the bird to be pink instead of brown on the breast, though the brown colour appears on the back. As I previously pointed out {loo. cit. ante), the red plumage is in this case probably the older plumage. 22 VIGOUE AND HEEBDITY been proved for some species of birds, e.g. Geothlypis, that those individuals of a species that breed the furthest north, winter furthest south, and, consequently, the full plumaged Golden Plovers that breed in the far north may presumably have wintered in the Tropics, while those that breed with us may not have entered the torrid zone. Further evidence in support of this suggestion is given by the fact that in the nearly allied species the Grey Plover, whose breeding range is entirely in the north, and that only occurs in the temperate zone on migration, all individuals assume their full summer dress. To take a further example. The Ducks may be roughly divided into two groups : — (1) Those in which the male assumes a dull plumage after the breeding season. (2) Those in which the male and female are similarly coloured. (Chiefly found in the Tropics.) Of this latter group, the Long-tailed Duck (Harelda glacialis) is the only species which occurs in the Arctic regions. This bird, moreover, differs in its plumage from other members of the Anatidae by the fact that there is a summer and winter plumage common to both sexes. According to our present theory, this is quite easy of explanation. The light-coloured winter plumage is assumed in Arctic regions after the breeding season ; the winter is spent in temperate regions, and the summer plumage produced there (under con- ditions which must be very mild to an Arctic species) is brown. 2. The Tropical Begions. — Leaving the Polar regions, where the contrasts of the season are greatest, let us now turn to the Tropics, where the contrasts are least. In the first place, the conditions are favourable to a high rate of metabolism during all seasons of the year, and abundance of food in the shape of fruit and insects may be had in every month, the only seasonal differences being those of wet and dry. Although these naturally have a considerable effect in determining the breeding seasons of many species of mammals and birds, yet the periods of drought and wet are so local, and determined to such an extent by the geographical conditions of the country, that they need not be considered in detail at present. COLOEATION IN MAMMALS AND BIRDS 23 We are, therefore, not likely to find in the Tropics marked seasonal changes of colour, and it will be evident that the life of an animal will run much more evenly so far as its metabolism is concerned; while food and temperature being favourable to a high state of vigour, we should expect to find animals deeply coloured and remaining so throughout the year. In some animals, however, the rate of metabolism will be so low that they can only exist where the conditions of life are most favourable. Such animals, therefore, accord- ing to my argument should be white or pale-coloured and restricted to the Tropics, while the brighter-coloured Tropical animals should be found to have a much wider range, the majority decreasing in colour as they spread north or south. Of these brighter animals the Tiger and Leopard afford good examples. The Buffalo of India is black, but in Assam a light variety is found, and the more northern species of Bos are found to be lighter in colour. The squirrels, another brightly coloured group, are, like the oxen, cosmo- politan, shading through brown to grey in the northern regions. White or grey tropical mammals are very scarce, but we may notice the Bamboo Eat {Bhizomys sumatrensis) , which is dirty white in colour and does not range north of Burmah. B. pruinosus, a rather darker species, ranges as far north as Assam ; while B. hadius and B. sinensis, both dark brown species, range from the south as far north as Bhutan, Nepal, and even Thibet. As regards the birds, we may notice the Kingfishers and EoUers, typical Tropical families, which range to more northern climates. The Parrots, which by the same process of reason one would expect to find in more temperate zones, only extend northwards to a limited extent. This is probably due to two causes : — (1) Their sedentary habits. (2) The fact that many of them are not as brightly pigmented as they appear on the surface, the predominating pigment being yellow. On the other hand, many of them are perfectly hardy, and stand our climate out of doors well ; so that probably their sedentary habits have been no small factor in restrict- 24 VIGOUE AND HEREDITY ing their range. Humming-birds also range far to the north, but have to retreat before the winter, probably from lack of food, while the Sunbirds of India are probably restricted by the Himalayas. The Phasianidse are another example of bright-coloured Tropical groups extending north- wards. Dull-coloured Tropical birds (not counting those from temperate regions that have migrated to the Tropics) are not numerous. The Hoatzin (Opisthocomus) is, however, a good example of a dull-coloured Tropical bird, which, as we might expect, we find restricted to the Tropics, while the Bell birds (Ghasmorhynchus) form another good example. Let us now consider for a moment the question of those mammals whose pelage changes at certain seasons of the year. Eoughly speaking, in the Tropics the wet and dry seasons correspond to our summer and winter, so that, if an animal is to be at all affected by the climate, the change will show itself, roughly speaking, in spring and autumn, at the same time as when the cold or heat are affecting the animals of more northern climes. Mr. Lydekker, in an article in the ' Field,' * pointed out that a Squirrel {Sciurus comiceps) was the only Tropical mammal that, to his knowledge, had a seasonal change. This particular Squirrel, in company with another (S. atro- dor sails), assumes in mid-winter a very bright or intense coloration on the back, which, as I pointed out t some years ago, is coincident with the breeding season, when we may assume their vigour to be at a high level. On the other hand, we have certain species, such as Sciurus maclellandi, Fwnambulus berdmorei, &c., in which their season of most intense colouring coincides with the summer or wet period, but in this case the change is merely one of relative brightness and not so marked as in the other cases. Among birds, the changes seem to be almost always " breeding changes," for they take place just before the breeding season, and are probably due almost entirely to the vigour which is bound to be high at that season. We * ' Field,' 1903, p. 675. t P.Z.S. 1901, p. 52. COLORATION IN MAMMALS AND BIEDS 25 would suggest that any changes of pelage in the Tropics, which are primarily due to climatical variations, are to be found among the representatives of northern races where the necessity of a change arose, and has not yet been eradicated.* A good example is to be found in Mustela flavigula. This animal, which is of a light brown and yellowish colour in Siberia, has in Nepal a marked seasonal change, becoming much darker in summer. In the Malay Peninsula, although both pelages are darker than those assumed in the north, the seasonal change still per- sists, while in Java and Sumatra the dark brown form is permanent throughout the year. In the Peninsula of India is also found another species {M. gwatkinsi) which is uni- formly dark, and has, so far as we know, no seasonal change. To sum up, we find that in the Tropics the contrast between a summer and winter coat is far less marked than among animals found in higher latitudes ; and, further, that when there is a change the brighter pelage is not always worn in summer. The explanation seems to be that, although the differ- ences in coat, are due to and represent changes in the rate of metabolism, these changes are brought about rather by the fluctuations dependent on the breeding season than on the climate. In other words, the climate may be always favour- able to a high vigour, but the natural lowering of vitality following a breeding season will require a longer or shorter period for its recuperation, according to the species. When that period is very short the animal will have a practically continuous breeding season, and there will be no marked difference in the pelage throughout the year. When, on * For a further instance of the seasonal change persisting under altered conditions, see Captain Barrett-Hamilton, P.Z.S., 1899, p. 598; see also W. C. Beebe, ' Zoologica,' I. 14, p. 253 (1914). It is not advisable to press matters too minutely in a general paper, but it is perhaps worthy of note that in Cervus eldi, an inhabitant of Burma, Malay Peninsula and Cochin China, the brightest pelage is that assumed and worn in winter, and is, therefore, hardly analogous to the changes in C. dewanceli, whose brightest phase is in summer. I would, therefore, suggest that the change in C. eldi is purely a breeding change, and that it has progressed a stage further than C. axis and C, unicolor, that have a similar pelage throughout the year, in having first eliminated the seasonal change, and then adopted the breeding change ; this would merely depend on the balance between the sexual and climatic impulses. 26 VIGOUE AND HEREDITY the other hand, the quiescent period is prolonged, so that there is only one definite breeding season a year, we may often find a marked difference in the pelage. 3. The Temperate Begions. — After all that has been said on the other regions there is not much to note here, except to point out that the arguments for the other regions still hold good. In the first place, it is fairly self-evident that in most of the mammals and birds of this region the prevailing colora- tion is brown ; white mammals are as scarce as they are in the Tropics, and bright-coloured forms belong, without exception, to Tropical or cosmopolitan families. The Squirrel and the Fox are good instances of mammals that have partially retained their bright colour, the former during summer only, the latter throughout the year ; the Deer, also, are red in summer. The birds also call for little comment on these lines. The bright species, such as the Kingfisher and EoUer, being typical of the Tropics, and the Sylviidse, or Warblers, typically birds of the Temperate region, are brown. The chief interest of the Temperate region, however, lies in those families that, stretching from the north and even penetrating the Arctic Circle, may also be found on the northern edge of the Tropics, becoming brighter and brighter gradually and throughout the whole distance as they approach the southern limit of their range, where the conditions are less severe. More heat, better and more abundant food, more time between the periodic breeding season and the cold of winter, or the gentle advent of spring ; all these are causes which, as we have shown, may and do affect the " vigour " of the individual, and it is "vigour" which we are suggesting as the main factor by which colour is influenced, and not the direct action of the environment. CHAPTEE IV. COLOEATION AN INDEX OF VIGOUR. Thkoughout this work frequent reference is made to colour, so that it will be necessary to consider the colour question more fully, and to show why of all characters, many of which may be of far greater importance to the animal itself, it becomes of great importance from the " vigour " point of view. To do this, it will be necessary, in the first place, to consider shortly what colour actually is. The colours of animals have been carefully investigated by several chemical physiologists, and are found to consist of minute granules of pigment, often, though by no means invariably, of a fatty nature. In all cases, however, these originate in the blood, by which they are carried and deposited in various regions of the body. Among mammals the chief pigments are generally found in the " bile," and although they may serve a useful physiological purpose, they are in the nature of waste products. The three main sources for the elimination of waste matter are the urine, faeces, and the skin, and thus it comes about that the colour-pigments, in being eliminated from the body, find their way to a considerable extent into the hair, feathers, or other external integument. In some cases, though, this can in no way be considered as proven, the colour-pigments, being in the nature of fatty animal oils, may, far from being waste products, be rather reserve nutrition, and in such cases it will be found that the animals are pigmented in a very different manner from the majority (e. g. Badger, in which the under surface is darker thian the upper). Fat. — It may not be out of place here to place more clearly the relationship of fat and pigment, as physio- logically it plays a most important part in the economy of 28 VIGOUE AND HEREDITY. most animals. The subject was gone into by Mr. Barrett- Hamilton, some years ago.* He shows that pigment in most animals is waste material of a fatty nature, or, in other words, that the pigment is conveyed and excreted through the skin in those portions of the body where fat also accu- mulates. In a condition of sluggish metabolism we have (a) a diminished oxydisation (leading to less formation of pigment), and (6) an increased storage of fat, resulting, if a moult takes place at such a time, in a paler coloration. Consequently, therefore, an accumulation of fat is an index of sluggish metabolism (condition of low vigour). Fat, however, being a reserve material, may also arise primarily through an excess of nutriment, especially if accompanied by only moderate vigour. In the case of mammals in northern regions this accumulation of fat is brought about in autumn, partly through the greater ease with which food may be procured, and partly by its greater abundance. This causes a slackening of the metabolism, which is still further increased by the cold of the late autumn, till finally the extreme cold induces a metabolism so sluggish that the animal can no longer carry on the functions of life, but hybernates ; the enormous layer of fat protecting his vital organs from the cold, as well as supplying in that state of sluggish metabolism sufficient nourishment to carry him through the winter. In animals in captivity this hybernating instinct becomes largely lost, owing to these two predisposing causes being almost or entirely removed. Birds, on the other hand, have a much more active metabolism than mammals, as is indicated by their higher temperature. Like mammals, the abundance of autumnal food causes them to store up a vast amount of fat. A few species spend the winter in the north, conforming in colora- tion to the general rule and getting white, but their meta- bolism at no time becomes sufficiently low to induce torpor or hybernation. The rest, the majority, migrate, and thus expend their accumulated fat on their journey to distant lands. I have pointed out elsewhere t that certain portions of * B. Hamilton, Proo. Boy. Irish Acad. xxiv. sect. B, pt. iv. p. 303 (1903) ; Proc. Camb. Phil. Soc. x. pt. v. p. 279 (1900) ; Ann. Mag. Nat. Hist, ser, 7, vol. ix. p. 106 (1902), f 'Knowledge,' March, 1906, and Proc. Int. Orn. Congr. 1905, p. 2.58. COLOEATION AN INDEX OF VIGOUR 29 the body (both in birds and mammals) are frequently of a different colour to the surrounding parts, and to these spots I have given the name " poecilomeres." Now the majority of these are either (1) spots where subcutaneous fat tends to accumulate, such as the interclavicular region of birds, or (2) where the skeleton comes close to the skin or on the extremities, where the circulation is probably poor. In those spots where the fat accumulates, the colouring is generally paler or absent. This may possibly be explained by the fact that when pigments are waste products they would not be deposited with the stores of fat or reserve energy, but when, on the other hand, they are reserve products, they would most naturally be stored with the fat. It is possible, also, that in some cases when the pigment is of an oily nature, it would naturally be stored with the fat. All this, however, is in need of much further study; my main argument is that pigments, being the product of metabolism, will vary with the metabolism ; Vigour, being merely a term to indicate a high or low rate of metabolism, will therefore be indicated by the amount of pigment formed, and hence colour becomes all-important as an easily seen index of vigour. In the previous chapter we have shown how conditions for high vigour are correlated with bright or intense coloration, and some apparent exceptions to this rule have been carefully gone into and explained. If, however, further proof were necessary, we know how, in man, bad health is associated (except in certain diseases) with loss of colour, and when associated with an increase or change of colour, it is always due to the' presence of poisons (waste material) which the body is trying to eliminate through the skin. Again, advancing age, in which the vigour is lower, is marked by the whitening or loss of pigment in the hair, and to a less extent this is sometimes found in animals. The period of most active metabolism is the commence- ment of the breeding season, especially among those species in which that season is sharply defined, and here, again, without exception we find each individual in his finest and best livery, so that from these points of view we again see that colour is essentially an index of the metabolism. Some of my readers, whose knowledge of living animals 30 VIGOUE AND HEEBDITY has been limited, may perhaps doubt whether " colour " is in reality such a sensitive thermometer to vigour as I have tried to indicate, and to such one can only say that an animal in high vigour is, to the experienced eye, a very different animal to the same individual in low vigour. It may be, and probably is, true that the eye of the breeder takes into consideration unconsciously many more points than mere colour — an animal's carriage, the sleekness of his coat, the brightness of his eye, his alertness, &c., all of which are signs, but the most tangible and conspicuous sign is the " bloom of health," which shows itself in an animal's colour. Before leaving this aspect of the question, mention must be made of some recent work by Miss M. Wheldale and also by Chodat and Bach.* These workers have shown that the colouring of certain plants is due not merely to the presence of a pigment (chromogen), but also to an oxidising ferment, and that the resulting colour may vary in intensity owing to a more complete oxidisation — so that the purple sweet pea differs from a red one by the more complete oxidisation of its chromogen. Now, a state of active metabolism means a more rapid oxidisation, and hence we see that in some and perhaps the majority of cases deep or intense coloration is dependent on a fairly complete oxidisation, which is bound to be a con- comitant of active metabolism or high vigour. We may, therefore, accept as proven for the large majority of cases that the colour of an animal to a certain extent is an index of its vigour, and we may now proceed to consider further questions of coloration from this point of view. Breeders of domestic animals realize the great diffi- culty of producing all individuals to an exact standard or shade of colour. This, though varying in extent, is, I believe, generally true of all domestic breeds, whether they be cattle, dogs, poultry, or pigeons. For show purposes a certain standard has been set up, and it is for breeders to produce animals as near that shade as possible. It would not be far wrong to say that there is no breed which will produce 100 per cent, of an exact shade, in spite of years of * ' Progressus Eei Eotanicse,' iii. p. 457 (1913). COLOBATION AN INDEX OE VIGOUE 31 careful selection, although some breeds are much truer than others. On the other hand, in Nature animals breed, in most cases, very true to exact shades, so true in fact that large series from any one locality are exactly alike, while others from a neighbouring place also resemble each other, but differ from the first batch in a slight difference of shade. These differences are so well known and constant that many systematists give them definite names, and they are looked upon as geographical races or subspecies. Here, there- fore, we have a problem which does not seem as yet to have had as much attention as it deserves. Why should it be almost impossible, under the strictest artificial selection, to breed 100 per cent, true to a particular shade, whereas under Natural Selection it is the rule in almost every species ? We have pointed out elsewhere (p. 3) that it seems doubtful if Natural Selection can have much to do with the elimination of those individuals who do not in colour come up to the exact standard for their particular environment, and the failure of rigorous artificial selection to produce such a result confirms our supposition that, in the production of an exact shade, selection (artificial or otherwise) has had very little influence, and has at all events not been the main factor in determining the colour. As breeders have admittedly failed in this respect, it will be more profitable in the first instance to turn to Nature, and study the causes other than selection, which may have been at work, and then to see if these causes are absent in the case of the domestic forms. Putting aside Natural Selection, but allowing for its operation in removing those individuals that vary too widely from the type, what are the circumstances which differ in the case of the geographical races of the species ? Their habits will be the same, also their enemies (I am thinking of cases of nearly allied, continental races), and the factors that will be different will be temperature, rain- fall and (to a lesser extent) food. All of these, as we have already seen, largely influence the metabolism, hence the vigour becomes altered, and through it the shade of colour. Now, granting that vigour influences colour, and is in turn influenced by the environment, are we justified in assuming that the constancy of any particular shade is 32 VIGOUE AND HBEEDITY maintained by vigour, bearing in mind the innate tendency of all living things to vary ? To answer this question, it would be necessary to enter at some length into the cause for this innate tendency to vary, and to which no completely satisfactory answer has as yet been given. Many variations appear to arise spon- taneously, as sports and some of these cases we now know to be due to a rearrangement or recombination of characters under the Mendelian Law. In other cases, to which we shall refer in the course of this book, the variations appear to have been largely brought about through changes of vigour that have affected the gametes. The larger varia- tions will, of course, come under the ban of Natural Selection, and thus need not concern us here, but the smaller variations are probably entirely due to differences in the somatic development of the various characters, and this development will be directly controlled by vigour. So that identical conditions will produce a practically identical level of vigour, which will be reflected in a particular shade of colour, found among all individuals of the same race under those conditions. Now, if an average temperature of, say, 76 degrees Fahrenheit, with an average rainfall of 36 inches, produces in a normal individual of any species a certain degree of vigour, those same conditions are bound to produce the same amount of vigour in every normal individual of that species. I use the word normal advisedly, since in any abnormal individual the effects of the climate will obviously be different, and such an individual will come under the ban or otherwise of Natural Selection, and we are dealing now only with those individuals that remain after Natural Selection has eliminated the more striking departures from the type. Normal individuals, therefore, under identical conditions will have an identical metabolism and an iden- tical vigour, and hence will be of the same identical In a neighbouring country, however, where the tempe- rature and food will be slightly different, the individuals from this region will have a slightly different vigour, and therefore be of a slightly different shade, but all the individuals from any one region will be, as we have shown, of an exact shade. COLOEATION AN INDEX OP VIGOUE 33 If this argument be true, animals brought from one environment to another ought very soon to show a shght variation from their original shade of colour. It being obviously impossible to carry out such an experiment in Nature, we must have recourse to data from animals in con- finement. We have not been able to find any observations on these which deal with this point, but many species are remarkably constant, even in a state of Nature,* and a change is not necessarily to be expected in the actual individual, but rather in its progeny born in confinement^ and we must be careful not to compare any young with their parents, whom they might exactly resemble, but with wild individuals of their own race. To take a good example, in confinement a Linnet {A. cannahina) usually gets a yellow breast after his first cage moult. Young born in captivity resemble their parents in having the yellow breast, and these must obviously be compared, not with their parents but with genuine wild examples. This case is so obvious that an excuse is almost needed for introducing it, but in other cases, where the differences are merely those of slight and almost imperceptible shades, the neces- sity for reference not to the parents but to wild specimens is liable to be overlooked. Apart from the aforementioned case of the Linnet, where altered conditions have obviously caused alteration in colour, we may note Dr. Stanfuss's valuable experiments on moths and butterflies, where, by subjecting the pupae to heightened or lowered temperatures, the. colour of the resultant imago was affected in the direction of the Arctic or sub-tropical races of its kind. Finally, I have undertaken some experiments with two races of a small desert rodent (Meriones) , one from Lower Egypt, and another and paler race from the Soudan. My experiments with these animals are related elsewhere (Chapter VIII.), but both races kept under the same con- ditions in England lost their racial character so far as colour was concerned, and became of a shade slightly paler than either form in their wild state, t We see, however, * The constancy of many species throughout a wide area is dealt with elsewhere (Chapter XII.) \ A pair of Meriones placed in an incubator became markedly darker in the course of a few weeks (see Chapter VIII.). 34 VIGOUR AND HEEEDITY from these cases, that an alteration of environment can produce an alteration of colour, and that different races placed under a similar environment tend to become of the same shade of colour, and that the shade acquired is slightly different from the shades in v?hich they are found in Nature. It may, hov?ever, be well to mention in this connection another factor in causing geographical races, which is not directly due to vigour alone, i. e. isolation. In the cases we have been considering, I was careful to note that they were continental forms ; that is to say, species which may range uninterruptedly over many thousands of miles, and under different climates we find them of different shades of colour, and in the intermediate districts inter- mediate forms are found which can only with difficulty be referred to one or other of the extremes. The occurrence of these intermediate forms in the intermediate districts is, of course, another point which tends to prove my hypo- thesis. Now, however, we must consider those races which occur in practically identical climates, and in which no intermediates are found. I refer, of course, to insular forms. Naturally, the animal does not occur in the sea, so that intermediate forms are impossible, and I would not have mentioned them, were it not for the fact that certain systematists use " the absence of intermediate forms " as an argument for designating animals which differ by the slightest of characteristics as true species and not geo- graphical forms of the same species. That, however, is by the way ; the fact remains that in many parts of the world, e. g. various islands of the West Indies, or of the China Sea, where the conditions of tempe- rature, rainfall, &c., are practically identical, we find that the representatives of a given species will differ slightly on the various islands ; this is, of course, due in the main to isolation. The insular race has probably been derived from only a limited number of individuals, and these have trans- mitted their characters to their descendants, and as a result we are bound to get in every small community a " family likeness," which will differ in some slight degree from other insular families. This faetor may, therefore, acting inde- pendently produce these local races, or, of coui-se, it may COLOEATION AN ^tNDEX OP VIGOUE 35 also act in conjunction with the other factors (temperature, huraidity, &c.) that we have aheady noticed, and in the case of island forms as with the continental it will in all probability be the " vigour " which keeps them to an exact shade, though in the former case the factor determining the exact vigour will be the hereditary family one,* whereas in continental forms the " environment " will be the pre- dominant cause. By this means, therefore, I suggest that uniformity may be attained within a species or race, without the action of Natural Selection. Having now seen how uniformity has been brought about, namely, by temperature, rainfall, food, and isolation, it remains for us to show that these causes are variable in the case of domestic animals, and then not only will their " shade " variations be accounted for, but another proof will have been brought forward to show that "vigour" is the determining factor. Apart from climatic differences, and these are fairly large even within our own islands, domestic animals are more or less confined. In the case of stock, the farm may be on a sheltered hillside, or on a spot with a cold and bleak aspect ; it may be in a place where rainfall is heavy, or where rainfall is light, and the animals being kept to this limited spot have to assimilate their vigour to those condi- tions. Now, if no extraneous blood is introduced, and the animals on each farm were to be allowed to gravitate to their "normal" shade for that environment, it would doubtless be found easy to produce a breed of the exact shade suitable to that locality which would breed true, but in practice the shade required has not been fixed with a view to the locality where the exhibitor resides ; hence in striving for the desired shade he imports stock from else- where, and from the mixed strain he gets a few individuals of the right shade, and that is as near to " breeding true " as he usually gets. Sometimes a breeder, from the situation of his land, and perhaps from his methods of feeding, succeeds in producing a large percentage of right-coloured stock; then, realizing the difficulty of maintaining the characters pure if he cross with other stock, he tries to follow Nature (probably unconsciously) by isolation, i. e. he makes no outcrosses, * See Chapter XIV., on Galton's Law. D 2 36 VIGOUE AND HEREDITY but this factor, " isolation," having been introduced, the "family likeness" begins to assert itself, and his strain gradually shifts from the normal " standard " to which he wishes to keep ; to correct this he brings in an outcross, the vigour is altered, and the work must be begun again. This, then, in our opinion, is the cause of the variability of domestic races; each breeder's stock is under different local conditions, some warmer, some colder, some fed on one food, some on another, and consequently the different strains are continually trying to correct themselves to the conditions under which they live, and the continual cross matings from different strains prevent the normal level for those conditions ever being reached. In those cases where strains have been isolated and a rough level attained the stock almost invariably becomes inbred; it loses its vigour, and either dies out a prey to disease, or an outcross is brought in to restore its vigour, and by that means the rough level attained is again broken. We see, therefore, that among domestic animals the causes that tend to a stability of vigour are, and must be, essentially lacking, and, in consequence, the external index of vigour (the shade of colour) is as variable and impossible to fix as the vigour of the animal itself. There is yet one other cognate point to be reckoned with among domestic animals, and that is the shortness of the period during which any strain is kept under " identical " climatic conditions. In Nature aeons may have passed during which the race was gradually settling to its " normal " level; in domestic races forty or fifty years would be a maximum period for the strain to be kept pure, and the longest case of which we have probably any records are those of the wild cattle of England. Here we find the different herds, each with its family likeness due to isola- tion, while their colour has gravitated to two distinct types, white or black, of which the black is apparently a Mendelian recessive to the white. We thus claim to have shown that in Nature, under equable climatic conditions, and with a sufficiently large stock to prevent inbreeding, the metabolism (vigour) of the animal settles down to a normal level, and thus a definite shade of colour is reached and maintained; in another locality, under slightly different climatic conditions, a diffe- COLOEATION AN INDEX OF VIGOUR 37 rent level of metabolism is reached, and the animals there- fore differ by a constant difference in shade of colour ; in intermediate regions intermediate shades occur. In our opinion, the differences of shades are far too small to have been brought about by Natural Selection, and we would suggest that the alteration in their metabolism is probably the reason, for it must be borne in mind that the main differences between geographical races are those of colour, and that the habits and enemies of different subspecies are, as a rule, the same, the only differences in the environment being those of " climate and food " ; therefore it seems more than probable that differences of shade must be due to the effect of " climate and food." These factors, we know, have a great influence on the metabolism (vigour), and hence we are justified in claiming " vigour " as the controlling cause of these differences. Further, we have attempted to show how domestic races live in very variable environments, owing to the different methods of feeding, the various situations of the farms, &c., therefore, it is only to be expected, as in fact we know to be the case, that the vigour of the various stocks and strains will be very variable, and, as a result, breeders find it almost impossible to breed 100 per cent, true to any particular shade. Let us, however, carry the matter a stage further, and see if there may not be in Nature any very variable species com- parable to domestic races. Such cases are very few. If we eliminate dimorphic forms, which from their very dimorphism suggest " a pair of characters " and therefore a Mendelian inheritance, and also the many species which were formerly considered extremely variable and in which we now know that the varieties are in reality local races — if we eliminate all these, we shall find very few instances. Among the Amphi- bians and Eeptiles, certain cases mwy occur, e.g. in the Common Salamander, but here, again, further research, such as the work of Kamerer* in Vienna, seems to show that such variations are largely the effect of environment. The only really good case that occurs to us is that of the common Alexandrine Eat {Mus raitus), which is now found throughout the civilized world. This animal varies to almost any extent (within certain limits), and apparently * Eamerer, Arch. f. Entwickelungsmecli., xvii. p. 69 (1904), Heft. 2; Boulenger, E. G., P.Z.S., 1911, i. p. 323. 38 VIGOUR AND HEIREDITY largely irrespective of locality. Drs. Hossack* and Lloyd t have written at length on its varieties in India, and we have also dealt with them from Egypt and the Malay Peninsula,! and made an attempt to classify them into local forms. In the light, however, of our more recent knowledge, and without going into the question here, it must be allowed that (except for certain instances of "family likeness," and for the inheritance of certain definite characters — such as white under parts and grey under parts — which are probably inhei^ited on Mendelian lines) no true geographical races can be made out, but throughout its range the same different varieties are continually cropping up. This species, therefore, seems to behave very much as domestic races do ; we must therefore look more closely into its environment to see if we can find in it any reason for its great variability. Unlike the Brown Eat (Mus norvegicus) , which lives chiefly in fields, drains, stacks, &c., we find that Mus rattus is essentially a house rat, parasitic on man. We are told that in India it lives in the roofs or floors of the native houses, where it is tolerated and even shares the evening meal with the family, like any domestic animal. Further than this, it is a great traveller, infesting the shipping, and being thus carried from port to port throughout the world. Here, then, are the causes which lead to its extreme variability : its food, and the situations where it is found are extremely variable ; it may live solely on grain in a granary, on mixed scraps from a native's food, on vegetation and wild fruits in a garden, &c., so that within the limits of one town different colonies may be living in twenty different environments. Here, then, comes in its analogy to domestic races, an analogy made still more perfect by the continual introduction of fresh blood (out- crosses) brought from overseas by shipping, and hence, in obedience to these different environments, acting directly on its vigour, we get the numerous varieties found in this species. And, as would naturally be expected, the variation is greatest where outcrosses are most frequent, e. g. in the large seaports, and least in the remote country districts, * Hossack, Mem. Ind. Mus. 1, No. 1 (1907). \ Lloyd, Eeo. Ind. Mus. iii., parts i. and iii.. No. 18, p. 261 (1909) v., pt. ii., No. 9, p. 105 (1910). X J.L.B., Fasci. Malay. 1. Zool. pp. 28 et seq. (1903) ; P.Z.S,, (1910), p. 651. COLOEATION AN INDEX OF VIGOUE 39 wliere, to a certain extent, local true-breeding races tend to be found. It may be objected that, if this argument is sound, the other species {Mus norvegicus) ought also to show a similar variation, whereas in point of fact it is extremely constant. This may possibly be accounted for, since it is, as we have already noticed, much more constant and restricted in its habits. It is a typical sewer rat, feeding largely on garbage, and much less parasitic on man. All species do not tend to vary to the same extent, some species, as is well known, being remarkably constant throughout their range. A few pages back we referred to the extreme varia- bility of Mus rattus within certain limits, and also to the fact that dimorphism might possibly be a Mendelian character, and the present is perhaps a fitting place to discuss the part played by Mendel's Laws in some of Nature's species. To anyone who will take up a systematic study of the Oriental Squirrels or Oriental Eats (and the same is true of other species of animals and birds all over the world), it will be apparent that several species, differing merely by very slight differences in colour or size, are found inhabiting the same locality and living, so far as our knowledge goes, under the same conditions; were they to be found in slightly different localities, they would merely be considered as local or geographical races, but occurring together they have, for want of a better system, to be considered as different species. Furthermore, local races of all these so-called species are also found. Two problems confront us here : — (i.) Since their environment is identical, their diffe- rences cannot be due to vigour, and, therefore, what causes them ? (ii.) Do these species, which are practically iden- tical, never crossbreed, or, if they do, why are there no intermediates ? Our answer to this point must, I think, be that in these cases the differences are due to Mendelian characters. The only method of proof can be that of experimental breeding. This I have done in the case of certain characters in Mus rattus (see Chapter VII.), and have proved them to have a true Mendelian inheritance, and in default of further exact proof we must, I think, assume that a similar explanation would account for most, and perhaps all, of the other cases. In some instances, 40 VIGOUE AND HEEBDITY animals which differ only slightly from each other may be in reality two distinct species, and in such cases inter- breeding may not take place, and thus the characters remain pure and distinct. It has been suggested, and probably with much reason, that the males of one species are not excited by the females of another, and thus inter- breeding would be prevented. We have tried in this chapter to show — (1) That coloration is in many cases an index of vigour, since it is largely due to the products of metabolism. (2) That vigour is, therefore, the controlling factor in causing geographical races and maintaining them at an exact shade. (3) That where nearly allied species or races exist under the same conditions, and yet show slight but constant differences, these differences are probably due to their being Mendelian characters, or that the animals belong to two distinct species which do not interbreed. CHAPTER V. COLOUE, SEX, AND MENDEL'S LAW. We have now shown that the coloration of animals varies within limits according to certain causes, e. g. climate, temperature, humidity, and food supply, all of which un- doubtedly exercise an influence on the metabolism, *. e. vigour, of the living animal ; hence, therefore, coloration is to a certain extent an indication of the vigour. It is for this reason that coloration, considered by many to be such an extraneous and unstable character, is in reality of an extreme importance, because it forms our best and at the present time our chief index of vigour. Colour as an index must, however, be used with care, as in any particular individual it can only vary within certain limits. The first of these limits is the Mendelian character. If a pigeon, for instance, contains, from the Mendelian point, only a blue pigment,* the colour of that bird can only fluctuate within the limits of the " blue " series, and cannot, say, become yellow ; whereas if that bird contained a red pigment, yellow might have resulted if the vigour had been sufficiently decreased; yellow and red being the same character, whereas blue is due to another allelomorph. This is most important to bear in mind, as in my experience a Mendelian character is at the basis of most inheritance, but the normal results of the inheritance of the characters are profoundly modified by vigour A So far, therefore, we find the inheritance of colour ruled by two causes: first, the Mendelian character, which is modified by the vigour of the parents, this character remain- ing practically unaltered from generation to generation, * There are apparently only two pigment factors in Pigeons. The various colovirs and shades are due to combinations of these factors with each other or with other non-pigment factors. f For exceptions to this, see Chapters VII. and XI. 42 VIGOUE AND HEREDITY while the initial vigour will be a variable quantity — varying even in individuals of the same brood. . Secondly, there is another cause which must be taken into account before we proceed to consider concrete examples to prove my case. This is the question of sex. From the metabolic point of view, sexes are said to be fundamentally different, the male being kataholic and the female anabolic. Put in plainer language, this may be expressed by saying that the male is continually spending the greater part of his income (nutrition) as he receives it, whereas the female stores up a reserve, which is expended periodically and used as nutriment for the young.* Vigour, or rather the surplus physiological energy due to a high state of vigour, must be expended by some means or other. With regard to the females, we see that it is temporarily stored and then, under normal conditions, used up in the cares and strain of maternity. With the males the case is different; in some species it shows itself in bright pigments (Mandrills and many birds), in excess or length of hair (Man, Lions), in howling or song and fight- ing (many species of mammals and birds), in frequent copulations or travelling far in search of a female (many mammals), in displays and dances, nest-building, feeding the female and young (many species of birds) . Energy produced by high vigour viust express itself or the individual will probably perish, and this accounts, as we have tried to show elsewhere, for the restriction of certain dark- coloured (and therefore high-vigoured) animals to the more unfavourable regions of the globe (e. g. Musk Ox, Penguin). It may be asked, however, how we can account by this argument for the large majority of mammals in which the sexes are alike — species that have a distinct breeding season, and in which the males have no song, have no sexual display, and do not attend the female during her maternal cares, e. g. Eats, Bats, and many others. How do they show their expen- diture of that surplus energy which is correlated with a high vigom- ? With some, and possibly the majority, it may be accounted for by (i.) Increase of size and activity; (ii.) Exces- sive copulation. Among Eats, Eabbits, Cats, &c., life runs on a moderately level basis, the food supply is fairly constant throughout the year, in temperate and tropical climates * See " Evolution of Sex," Geddes & Thompson ; also Chapter XI. COLOUE, SEX, AND MENDEL'S LAW 43 thej' are not largely affected by extremes of temperature, and the males take no part in family cares ; but here we find the breeding season is almost continuous or extremely prolonged ; and the surplus energy thus easily finds a vent. With regard to Bats, the case in most climates is probably governed by hibernation or semi-hibernation, which results in the accumulation during the autumn of large quantities of fat ; the males thus become for a few months anabolic, and during the winter their vitality becomes extremely low. Gradually during the summer their vigour increases, till it becomes reduced by the breeding season of the early autumn, and then, with an abundance of food and low vigour, they again accumulate fat for their winter sleep. Of the Tropical and fruit Bats it is difficult to speak, since we have no exaet knowledge of their life history ; it is pro- bable, however, that their breeding season is prolonged, and that in the case of the fruit Bats a period of scarcity of their usual food keeps their vigour low at certain periods of the year. These foregoing remarks apply only to the males, for it is only the males that have to dispose of the products of their rapid metabolism ; the females, on the other hand, are continually storing up a reserve. In winter and during periods of scarcity this reserve is naturally accumulated more slowly than under more favourable conditions, but as soon as it reaches a certain point the surplus is expended in sexual endeavour. It follows from this that the breeding season and the number, colour, and — as we hope to show later — even the sex of the young will therefore depend to a considerable extent on the vigour of the parents, which is in turn dependent on the environment and climatic conditions in which they happen to live. It must not, however, be forgotten that the metabolism of a female runs on more level lines than that of a male, and is consequently not so visibly affected by the conditions. Under conditions of high vigour a female will accumulate energy till a certain stage is reachedj when if her vigour be high she will breed. Under moderate conditions of vigour this stage may never be reached, and she becomes tempo- rarily sterile, while under adverse conditions she may have to use up for her own needs some of the potential energy that was accumulated under more genial conditions. With males the case is different. Under high vigour 44 VIGOUR AND HEEEDITY conditions the limit of their powers for storing energy is soon reached, and any surplus has to be expended in one of the ways indicated above. Under moderate conditions, the small surplus, if there be one, becomes naturally expended in their normal restlessness and greater activity as com- pared with the females ; and, under low conditions, there will be no surplus. In both sexes, however, a state of low vigour and abundant nutrition may occur at the same time, more especially in the autumn after the breeding season ; and when these conditions are accompanied, as often happens, by colder and shorter days, the males become temporarily " anabolic," and accumulate vast stores of fat. The result with the females is also the same, but owing to their normal anabolic habit the effects are not so noticeable. If this fat accumulates beyond a certain point and the temperature remains low, a state of coma or hibernation ensues. A similar condition may be brought about (though actual instances are scarcer) by excessive heat (c/. aestivation of some of the lower Orders) . Apart from some species of Bats, to which we have already referred, some of the fur Seals offer perhaps the best example of an apparent "anabolic " condition among males. Early in spring the males, in a tremendously fat condition, resort to their breeding islands, where they remain without food, living on their own fat for two or three months.* After the breeding season, with their vigour low, the tem- perature cold, but with abundance of food, they would naturally, on our hypothesis, accumulate fat and become apparently anabolic ; till with a return of longer and warmer days their vigour gradually increases, and incited doubtless by the females being ready (see Chapter XII.) they repair to their island homes. They are now in a high state of vigour, but owing to their accumulated stores need no further food supply, and thus remain on land, living on their own fat till hunger or the absence of females causes them to return to the sea with reduced vigour. * It is a well-known fact that domestic animals and animals in cap- tivity are often temporarily sterile because they are too fat. It will be seen from the above that in these cases the accumulation of fat is due to a low vigour accompanied by an abundance of food, probably of too fat- forming a nature, the real cause of the sterility being due in the first instance to a reduced vigour. Fat, therefore, is only the effect and not the cause. COLOUE, SEX, AND MENDEL'S LAW 45 We have given this as an extreme instance, but in the majority of cases, and especially among the mammals of the Arctic or temperate regions, similar though far less exaggerated cases of apparent " anabolism " w^ill be found among the males. Turning to the birds, we may notice that the song and other displays of the male are at their maximum while the hen incubates and cease altogether on the hatching of the young ; or, in other words, the song, displays, dances, &c., reach their highest expression when all other outlets for vigour, such as nest-building, courting, and feeding the young are temporarily in abeyance. Females are then, as we have said, anabolic, and on that account accumulate fat more readily than the males ; * but since an anabolic condition is essential to them, and they cannot survive unless there is a small surplus nutrition I in excess after the normal needs of their bodies have been I satisfied, it follows that a higher vigour is necessary to lenable a young female to grow and survive than is necessary pr a male. In the females the sexual organs and their accessories demand a larger share of that nutrition, which in males is able to be almost entirely devoted to growing and carrying on the normal functions of life. In this last sentence lies the whole crux of the difference between the sexes. Hybrids are more often males ; why ? Because only very vigorous individuals survive, and conse- quently more females die from lack of vigour ; and this is again a further reason for the slight preponderance of males throughout nearly the whole of Nature. Mendelians have lately stated that maleness and f emale- ness are Mendelian characters, in which the female is hetero- zygous for sex and the male homozygous. This may be true. The case is only half-proven, but, if so, how do they ac- count for the fact that instead of getting equality, the expecta- tion under their hypothesis, there is almost always a slightly larger percentage of males. I have suggested the answer above, and would further point to this as an excellent example of the way in which Mendelian results may become obscured by the effects of vigour. * In the human race, for example, they have a much thicker layer of suhcutaneous fat, which gives their bodies the more rounded curves, and enables them to wear thinner clothing without feeling the cold. 46 VIGOUE AND HEREDITY To go back, therefore, we noted above that inheritance was due to MendeUan characters modified by vigour, and we may now note that the external results of that vigour may be still further shown by the sex of the offspring. Let us instance a few concrete examples before leaving the sex question. It is a well-known fact that a large pro- portion of animals and birds bred in confinement are males; this is easily accounted for, if my theory be correct. The vigour of wild animals is lowered under artificial conditions, the accumulated vigour of the parents is more often than not only sufficient for the production of males, and the females, if born, are weakly and do not survive.* In breeding and rearing somewhat complicated cases of hybrid Ducks, I found at first that a large proportion of the broods reared were males ; subsequently by feeding up the young on more abundant and nourishing food I found that, when full broods were reared, the proportion of the sexes was nearly equal. Breeders of Pigeons know well that if only one Pigeon be reared out of a nest, it is almost always a male. The two young birds in a pigeon's nest are usually a pair ; but if they are both females they will generally be found among the earlier nests of the season, and if both males, in the later nests. (See Chapter IX.) I have bred, as noticed elsewhere, many Meriones, a small desert Mouse allied to the Gerbilles — my original stock was caught wild during the winter. The first litter was produced at the end of March, and consisted of one male and four females, another pair gave three litters, con- taining, respectively, one male and four females, three males and one female, and three males, showing a marked increase * Although this is a well-known fact, I have tried in vain to find a sufficient number of definite statistics to prove it, but in all the records and reports of Zoological Gardens there has been great negligence in accurately recording the sexes of the young born and bred. Such notes as I have been able to gather are somewhat contradictory. In fome Gardens we get a large percentage of males in certain Orders and not in others; in other Orders a larger percentage of females. This might entirely bear out my hypothesis if the facts given were complete, and if the actual conditions under which the animals were kept were known. In default of these particulars, however, it becomes impossible to argue from the facts, and we urge on all Directors of Gardens to record to the fullest extent possible all details relating to the birth of animals under their charge. COLOUE, SEX, AND MENDEL'S LAW 47 in the proportion of males as the vigour decreased during the breeding season. This is by no means an isolated case, but is the normal course, both for Mice and Pigeons, and doubtless in other animals as well. (See Chapters VIII. and IX.) It will be noticed that in the aggregate (eight males to nine females) the Mendelian expectation of equality was practically attained. The same year I also bred several litters of the common Egyptian house Eat {Mus rattus), but in this case the food was similar to that to which they were accustomed in the wild state, and the sexes were approximately equal. There was, however, a litter of seven which were very weakly when young, and when a month old were hardly larger than they would normally be at a fortnight. It is therefore perhaps striking to note that they were all (seven in number) males, and that three of them were of a light fawn colour, previously unknown among these Eats. (See Chapter VII.) In birds the amount of reserve material required by the hen is much less than in mammals, for the special strain on the female practically ceases on the laying of the egg, incubation being often shared by the cock, and no reserve being required for lactation ; but any excess is expended by the hens as well as by the males in searching for food for young. The curious aberrant species in which incubation and care of the family are undertaken by the cock need only be noticed here, and are probably easily accounted for by a nearly anabolic condition of the male bird, and a corresponding increase towards katabolism on the part of the female.* As just noticed, the actual habits of a female bird do not call for a marked condition of anabolism as compared with the male ; and we may note that the difference in the sexes throughout the avian kingdom is most marked in those species where the male takes least interest in the female and young, e. g. game birds, Birds of Paradise, some Ducks ; and it is probable, therefore, that among birds generally the katabolism of the males and anabolism of the females is not very marked. Females, especially in the foetal or embryonic stage, cannot continue to live unless they are accumulating energy, and if the vigour be low sufficient reserve is * See Chapter XI. 48 VIGOUE AND HEEBDITY not accumulated, and consequently the females die ; hence with low vigour we get a higher proportion of males and small litters. There is a certain amount of circumstantial evidence that a low vigour, acting of course from the time of fertilisation, may result in the development of a male instead of a female. If the Mendelian hypothesis that the females be heterozygous for sex be true, the matter may be explained by the reversal of the dominant character, and this leads to the necessary corollary that such males must be heterozygous. We are inclined to favour this idea, but it might well follow in such cases that the determinant for femaleness may have been quite destroyed. This latter matter is still quite unproved ; but as even in adult life the destruction of the female organs leads to a growth of male characters, it is more than probable that the " death " or crippling of the female determinant at an extremely early age might well lead to the development of maleness and the birth of a male, and this fact is borne out by the statistics. Sex is apparently the primary and all-important factor in development. If an animal is female, it absorbs in its femaleness a larger proportion of vigour than a male. Con- sequently the amount of vigour left for the development of other characters is less than in a male. Therefore, the vigour on which characters (other than sex) have to draw is less, consequently it frequently happens that this amount of vigour in the female is one that favours the recessive character rather than the dominant ; and hence females are often differently coloured from males, their colour being almost always of the weaker (or recessive) character. Vice versd, in the case of infertile hybrids both sex gametes having been killed or remaining entirely undeveloped, the whole of the vigour is available for the other characters, and the animals are as a rule very strongly coloured, e. g. Pheasant hybrids. Thus we are able to realize that, although colour is an index of vigour, it may become modified by (i.) the Men- delian character and (ii.) the sex. These instances, together with the further evidence given in the succeeding chapters, will, I hope, be sufficient to show that the sex of any animal is partially determined by the vigour of its parents, and that when the environ- ment of the parents is of such a nature as to tend towards COLOUE, SEX, AND MENDEL'S LAW 49 a high vigour, females will tend to predominate among the offspring. Before closing this chapter, however, one doubtful problem remains. Does the amount of vigour under which a zygote develops actually determine the sex, or is the effect of a high vigour to enable a larger proportion of females to survive ? This problem we will refer to at greater length in Chapter XI., after discussing our experi- mental evidence; but we may call attention to it here, so that some of the reasoning in the intermediate chapters may be more easily followed. The Mendelians' present idea is that sex is a Mendelian character, and if this hypothesis be strictly true, it would then follow that the effects of high vigour would be to allow a larger proportion of the " potential " females to survive, and thus the percentage of females would be increased. On the other hand, by the same reasoning, since under the Mendelian expectation the number of each sex is about equal, no amount of vigour should raise the number of females to over fifty per cent., though a low vigour might reduce them to considerably below that percentage ; on the other hand, if the sex be not a Mendelian character, but rather the expression of a physiological state, then a zygote would be sexless, and the sex of the individual determined by its vigour. We need not go further into this very complex and difficult question at present, for the evidence on which a definite hypothesis might be founded is still much too scanty ; nevertheless, our experiments as detailed in the following chapters certainly tend to show that, when we have reason to suspect that the vigour of the parents is low, we find an increase in the proportion of females born. 50 VIGOUE AND HEEEDITY PAET II.— EXPERIMENTAL. CHAPTEE VI. DOGS, GOATS, AND CATS. Dogs. The experiments on Dogs were undertaken not for any special purpose except to retain a strain which had from a sporting point of view certain desirable mental characters, which may be summed up as high intelligence and a good " nose." Erom the heredity standpoint they are chiefly of interest as showing the relative inheritance of black and red. The strain was commenced by crossing a Eed (Gordon) Setter dog with curly coated Black Eetriever bitch, both of unknown parentage, out of which were produced some blacks and reds. One of these, a red dog (Toby), lived with me for fifteen years, and showed considerable mental development. The general appearance of this animal may be gathered from his photograph, and externally his mental qualities showed themselves in his eyes and general expres- sion. When mated back to his mother the resultant puppies were of various colours : pale red, brown and black ; the reds and browns being of various shades. The mental qualities could generally be foretold by the expression and shape of the head, and, although none of this progeny were kept very long, it may safely be stated that all those that resembled their father in this repect had always his character. On the other hand, it is quite possible that others without this " expression " might also have possessed his qualities. After his mother's death Toby was frequently mated to another pure curly coated Black Eetriever, and the resultant progeny of these matings were all black. Subsequently, being desirous of producing another dog similar in colour and mental qualities to " Toby," a pair of his later black offspring were kept and mated. One of these, the bitch, resembled her father in his mental qualities, shape Plate I. "TOBY." (Red Setter \ Black Retriever. Photo by J. L. Bonhote. DOGS, GOATS, AND CATS 51 of head, &c., while the dog, although having a good nose and sporting qualities, was not quite as good mentally. This latter pair were mated on two occasions. The first litter contained ten puppies, five males and five females, all black. The second litter contained eleven puppies, six red, (three males, three females) , and five black (three females, two males). A red dog from this litter proved to be an exact counterpart of his grandfather in every respect. This dog was subsequently mated to his mother, the resulting litter containing nine puppies, five red (two males, three females), and four black (one male, three females). The following genealogical table shows the matings at a glance : — Eed Setter ^ x Black Retriever (Gyp) J Several Beds Sev. Blacks Eed $ (Toby) Many Blacks Black $■ x ^iB,ck 2 (Eover) Blacks Beds Browns 15 Blacks 5 Eeds Eed ^ (Toby II.) 5 Beds 4 Blacks It is evident from the above that red in this strain acts as a normal recessive to black, and therefore the red dogs are pure recessives. The result of (1) Bed J X hetero (?) black, giving reds and blacks. (2) Eed ^ (Toby) x hetero (?) black, giving reds and blacks and browns. (3) Eed ,y (Toby) X pure black, giving blacks only. (4) Black J* (hetero) x black 9 (hetero) giving 6 reds to 15 blacks. (5) Eed ^ (Toby II.) x black 2 (hetero) giving 5 reds to 4 blacks. Although it is impossible to be equally definite about the inheritance of the mental quaUties, as the external characters are not so defined, and to have kept all the puppies for any length of time would not have been practicable, yet there can be no doubt that individuals 52 VIGOUR AND HEREDITY varied greatly in this respect, and that high mental quali- ties were associated with a definite (hut indescribable) expression and shape of head. In the ordinary way these experiments might be con- sidered as entirely confirmatory of the Mendelian theory: it is perhaps worthy of note that in the two matings of the heterozygous blacks, where the expectation was three blacks to one red, the first litter contained blacks only. These were born on July 21st, and therefore the period of gestation took place during the summer months, a period of high vigour, and we got a large preponderance of the dominant colour. The second litter was born on April 4th, and in this case the period of gestation was in the colder months of the year immediately following the winter, a period of lower vigour, and we got a preponderance of the recessive colour. The combined result of the two litters giving the exact Mendelian expectation. Goats. The inheritance of horns in Goats awaits as yet a satisfactory solution. In cattle, horns are recessive to their absence, in Sheep the inheritance is sex-limited; but in Goats, although horns are usually recessive to their absence, there seem to be certain complications. My friend, Miss Wilde, of Little Gaddesden, Herts, who has kept Goats for some years, has kindly put at my disposal the pedigrees which appear on p. 54, and as I have frequently seen and examined her Goats myself, these notes are put on record in the hope that they may prove of use to future workers. Before dealing, however, with Miss Wilde's results, we may briefly recapitulate the present state of our knowledge as regards this character. Mr. Bateson in his ' Mendel Principles of Heredity ' refers to this question in a footnote (p. 170), in which he states that the details are complex, and that he has not been able to extract a consistent scheme from them. Since then Mr. C. J. Davies has been critically through the Herd Books of the British Goat Society, and the results of his investigations, which, if we mistake not, were first given in the pages of the 'Bazaar,' have now been published DOGS, GOATS, AND CATS 53 in full in the ' Mendel Journal,' No. 3, p. 103 (September, 1912). Mr. Davies's conclusions, which we will quote in full, show that the horned character needs much further in- vestigation, and is not comparable either to sheep or cattle. He writes (p. 106) : " First of all, the twenty-four matings of a polled animal bred from polled parents with a horned animal bred from horned parents give equal numbers of polled and horned offsping (six horned males, six horned females, three polled males, three polled females) . "^If inheritance had followed the same lines as that in cattle, this mating would have given all polled offspring. If, on the other hand, it had been analogous to the phenomena observed in sheep, all the male kids would have been horned and all the female kids polled, which is equally not the case. So we are apparently confronted at the outset with the fact that there is no dominance (sexual or otherwise) of either character, and that the horned or polled condition is equally likely to be inherited by cross-bred Goats of either sex. " In cattle, the horned individuals of the F^ generation are pure (namely, breed true). In sheep only the F'^ polled males and F^ horned females breed true to their respective characters. In Goats, a pair of horned animals descended from four parents, all of which were also horned, breed true to the horned character, without exception, in the few cases of this precise mating which the writer has been able to collect; but practically all matings of two polled individuals throw a proportion (1 : 3) of horned progeny, whatever the precise germinal constitution of the parents may be." Mr. Davies then proceeds to give the detailed statistics of the breeding of some 1370 Goats. His general results show that in the main the inherit- ance is similar to that in cattle, namely, a dominance of the polled over the horned, but that the dominance is not invariable, and 3'50 per cent, of 882 polled Goats were bred from two horned parents. The following tables give the pedigree and breeding of the Goats kept by Miss Wilde, and it will be seen that, as 54 V V VIGOUR AND HEREDITY P- P- VulcftD p (7) Lao Witcb hb. Therefore Volc&o is beterozygouB Sonlighv p. Ill Grey Bock p Trixie p ETrito p (bj fi Bee p. (hj a Lad; Bird p Qu> E Beatrice p (h.) IV B. Beatiu« p. fh.} EMio p. Ou a Lad; Bird p. Oli ^ b. h. < p. < DOGS, GOATS, AND GATS 55 far as they go, they seem to show fairly clearly the dominance of the polled over the horned; or rather, we might say, that they show nothing contrary to that assumption. The cases, therefore, that chiefly need investigation are those (very few in number) in which horned individuals have thrown a certain percentage of polled. On this point we have one suggestion to make. Some of the Goats bred by Miss Wilde, although "horned," had much smaller horns than the normal, and it may be that these " horned " individuals were in reality heterozygous, and that the dominance in these cases was incomplete. Unfortunately, we were unable to test the matter, but if the suggestion be the right one, it would be a simple solution to a problem that seems otherwise somewhat complex. Our only reason, however, for bringing the subject forward is in the hope that others interested may take the matter up. Gats. For some years past Miss Wilde, to whom I am very greatly indebted for the carrying out of these matings, has also been experimenting on Cats, the main object being to study the reason of the great rarity of tortoiseshell males, and the possibility of producing one. The only details that have hitherto been published on this character are contained in a paper by Doncaster (Proc. Camb. Phil. Soc. xiii. pt. i. p. 35, 1904). The author had not at that time undertaken any experiments himself, but had collected and digested a fair number of pedigrees, and came to the conclusion that a female tortoiseshell was a heterozygote between orange and black, in which the dominance of the orange was incomplete, and that in the case of the heterozygote males the dominance was complete and the males were in consequence orange. A careful study of Doncaster's paper, together with other pedigrees, convinced us that his conclusions were correct, and might be accepted as the basis for future work. We, therefore, started with the assumption that orange is dominant to black, and that cream and blue are the dilute forms of these colours. What was required, therefore, in order to produce a tortoiseshell male was to produce a strain with a tendency 56 VIGOUE AND HEREDITY to black, in the hope that some of the heterozygous males would thus show some black. For this purpose two methods suggested themselves — (1) to procure or produce a strain of tortoiseshell which was bred from black rather than from orange sires; (2) to breed a strain whose vigour* would favour the production of black rather than orange, and, in the present case, black being recessive, it seemed that a diminution of vigour would be the most likely method to favour the production of blacks, and that this would be most easily brought about by inbreeding. Thus the appearance of a character recessive under normal conditions may be anticipated in greater numbers than under the strict Mendelian interpretation if we alter the conditions to favour a lower vigour than the normal. The following matings were carried out : — Pabents. Offspwno. Female Male Males Females 1. Tortoiseshell (Eainbow) x Lord Coke (black) 1 orange, 2 black 2 black 2. 11 ,. X Bed Spider (orange) 2 orange, 1 black 1 tortoiseshell 3. n i» ^ orange from 1 2 orange, 1 black 1 orange, 2 black and 3 blacks not sexed 4. Black from 1 x orange from 1, also used 3 4 black, 3 blue 1 black, S bine, 5 tortoiseshell 5. Black from 4 x orange from l.also used in 3 and 4 1 black 1 blue,2 tortoise- shell 6. Tortoiseshell from 4 x orange from 2 1 black, 1 orange 2 orange, 1 tor- toiseshell 7. Black from 1, also used x in 4 orange from 2, also used in 6 2 black, 1 blue 2 tortoiseshell 8. Tortoiseshell (Bainbow) x orange from 2, 2 orange 1 orange used in 1, 2 and 3 also used in 6 & 7 The blues in Experiment 4 come through Lord Coke, who had blue in his ancestry. The numbers bred in these experiments are too small to form any estimate of the Mendelian expectation being realised or not. Experiment 4 is, however, of interest and worthy of * We cannot at this present moment enter fully into the reasons for this hypothesis, but we have certain evidence (at present not sufficient to be definitely formulated) that under certain conditions the recessive character is favoured, and will appear in greater proportion tliaii expected under the Mendelian doctrine. DOGS, GOATS, AND CATS 57 further note. The black female being recessive should be pure, and the orange male, judging from his progeny in 2, is almost certainly heterozygous for black. So that we get BB X OB, which leads us to expect among the males an equal number of orange heterozygotes and black, and among the females an equal number of blacks and tortoiseshells. Working on the lines of our second hypothesis, we en- deavoured to reduce the vigour by inbreeding (in this case brother to sister), in the hope of producing a stronger tendency to the recessive colour (black) and thus to produce a tortoiseshell, instead of the dominant orange among the heterozygotes of that colour. In so far as we did not pro- duce a tortoiseshell male the experiment may be said to have been abortive, but in so far as we succeeded in entirely eliminating the dominant orange and in producing nothing but black males, the experiment may be said to have more than confirmed our hypothesis. It is only fair to add that Mr. Doncaster informs me {in litt.) that he has never known orange males to be produced from a black female by orange male ; but, on the other hand, mine are the only records he has of black females being produced from orange males ; females from orange males being, in his experience, always orange or tortoiseshell, according to the colour of the mother. There is now one other case on record in which a cream male to tortoiseshell female gave one blue male and two blue females. Unfortunately, lack of space and money has prevented the black males from being tested to see whether some of them were in reality not pure recessives, but hetero- zygous, and carrying the usually dominant character, orange. Our first hypothesis was tried in Experiments 1, 2 and 3. " Eainbow," the tortoiseshell female, was very dark, and bred from a black and a tortoiseshell, while Lord Coke had only blue and black in his pedigree. The expectation from (1) therefore was heterozygous orange and black males in equal numbers, and tortoiseshell and black females in equal numbers. Or, taking the sexes together, the total number of blacks should have equalled the orange males and tortoiseshell females taken together. Our hypothesis was, however, to increase the power of the black, and though no tortoiseshells were produced, the total blacks (4) outnumbered the single orange. Experiment (2) was a heterozygous orange to the same 58 VIGOUR AND HEREDITY female, undertaken as a control to Experiment (3). The expectation was three orange or tortoiseshells to one black, which was exactly the result achieved. Experiment (3) was a heterozygous orange paired to his mother. From the Mendelian standpoint, the expectation was identical with (2), but from our point of view the loss of vigour by inbreeding should give a larger propor- tion of the recessive character (black). Here, again, although no tortoiseshell males were produced, the results from the vigour standpoint coincided exactly with the expectation : instead of the proportion 8:1. Orange and tortoiseshells to black appearing, we got only three orange to six blacks. In Experiment (5) the numbers are too small for any deductions to be drawn. In Experiment (6) the male is possibly a pure orange, and in any case has much more orange blood in his ancestry than the male of Experiments 3, 4 and 5. If he is pure orange, no black male should appear. The presence of the black male, therefore, tends to prove that he is heterozygous, and among the females the orange and black together should equal the tortoiseshells; but from such small numbers no deductions can be drawn. In Experiment (7), presuming from the foregoing experiment that the male is heterozygous, our expecta- tion would be the same as in Experiment 4, with probably a slight reduction in the number of blacks owing to the stock being less inbred. Among the males the results are practically (though the number is very small) in accordance with Experiment 4, since no orange males were produced. In Experiment (8) the results are in a way curious, owing to the absence of any black, but this is possibly due to the disparity between the ages of the parents. If the experiments, therefore, have so far failed in the production of a tortoiseshell male, they show conclusively what is perhaps of greater general importance, namely that the Mendelian expectation of proportions may be altered by inbreeding, which is acknowledged to be a matter of reduc- ing vigour. If, however, the Mendelians do not like to accept that conclusion, the only other alternative is to admit that an animal will reproduce more strongly and in larger proportions a character which is strongly represented in its DOGS, GOATS, AND GATS 59 ancestry. To admit this is to disprove Mendelism ; but on the vigour hypothesis, as we shall explain later, pure Mendelism and the effect of ancestry may both be accepted in slightly modified forms. Since the above was written, Miss Wilde has kindly favoured me with the pedigree of two tortoiseshell males that have recently been bred by another fancier : — Colour $ X Cream 2 Blue J X 5 ? Colour (heterozygous ?) Blue. $ x Blue J Tortoiseshell $ Tortoiseshell $ Tortoiseshell $ Tortoiseshell $ From the above pedigree it will be seen that the Tortoiseshells were produced from two blues, one of which at all events may have been heterozygous, and the orange must have been almost certainly introduced through this male. This pedigree, which is somewhat incomplete, and therefore unsatisfactory, tends however to show that the tortoiseshell males were produced from heterozygote blues, which on our hypothesis are dilute blacks. According to the strict Mendelian interpretation, black being recessive to orange, cannot carry it, but several exceptions to this rule among other animals are known, and in our opinion the question as to which of a pair of characters is dominant and which recessive is merely dependent on vigour. Under certain conditions of vigour one character will be dominant (i. e. it will show itself in the development of the soma) , under other conditions it will lie dormant, and its allelo- morph will appear in the somatic development. The question of the heterozygous blue need not there- fore trouble us, and the fact that these tortoiseshell males have been produced from two blues certainly seems to favour our original hypothesis, namely that they may be produced from a strain with a tendency to black. Writing to me subsequently, Mr. Doncaster tells me that he has now several cases of blue females from cream 60 VIGOUE AND HEREDITY males, but with the exception of the single black female bred in Experiment (4), he has no records of black females from orange males. He suggests, therefore, that possibly blue does not behave like black in inheritance (i.e. that it is not black dilution) , but a separate character. In view of our solitary black female and the general behaviour of dilute colours, we are not inclined to agree with this suggestion, but rather to accept his alternative that the apparent blues are in reality heterozygous. Although Miss Wilde had got rid of the blues in Experiment 4, as we were only dealing with the concentrated colours, we have, in view of these subsequent facts, succeeded in getting one back, and hope shortly to be able to state definitely whether or no it is heterozygous. Since the above chapter was written, Miss Wilde has bred two more litters from Experiment (8) . The results being one tortoiseshell male, two orange males, one orange female, and two tortoiseshell females. Unfortunately the tortoise- shell male died young. We may note, however, that this mating was a combination of our two hypotheses (p. 56), being inbred (mother to son), the mother being aged and the ancestry of both parents containing a large amount of black or blue. The absence of any pure black and the prepotency of the orange is possibly to be accounted for by the disparity in the vigour of the parents (see p. 8). CHAPTEE VII. EATS (Mus rattus). When staying in Egypt some years ago I had oppor- tunities of studying this very variable species, not only when dead but also in its natural haunts. The results of my investigations, which have already been published elsewhere,* led me to the conclusion that some of these variations had arisen as mutations, and had in all probability a Mendelian inheritance. I therefore determined to try a few breeding experiments, with the object of endeavouring to elucidate the matter, and brought home three genuine wild-caught pairs of these Eats. The first pair was typical of the variety known as Mus tectorum, distinguished by having the hairs of the under parts white to their bases, the upper surface of the hind feet white, and the average length of the hind feet 35 mm. The next pair was what I have distinguished as the grey variety of M. tectorum (Group II., loc. cit.). It resembles the typical M. tectorum, except that the upper surface of the hind feet contains a certain admixture of brown hairs. The third pair were typical of what is known as M. alexandrinus, in which the hairs of the under parts are slate-coloured, usually with whitish tips, the upper surface of the hind feet is - brown, and the average length of the hind feet 33 mm. General Notes on the Management and Habits. Contrary to expectation, it proved to be somewhat difficult to keep this species in good health in confinement, and it breeds none too readily. For these reasons, there- fore, as well as from the fact that it has to be kept in zinc-lined cages, through which it cannot gnaw, and which * P.Z.S., 1910, p. 651. 62 VIGOUE AND HBEBDITY are rather expensive to make, I have not been able to keep a very large stock, and the numbers bred have been much fewer than one vsrould have wished ; nevertheless, the results, as will be seen later, are very definite, and there seems no reason to suppose that larger numbers would have appreci- ably altered them. All the experiments which have hitherto been conducted on Eats, such as those of Cuenot, and Castle and MacCurdy,* hav been carried out with the domestic Eats, which are varieties of the Brown Eat (M. iiorvegicus) , and as this species is apparently much more easily bred and kept than M. rattus, a few notes on the most satisfactory method of keeping the latter may not be out of place, and may save future workers from disappointment. As to food, mine have barley or other hard grain and soaked dog-biscuit, with occasionally a little green food during the hot weather. They will not thrive on grain and water alone, and it seems that a little meat in some form or other is almost essential. Although a minor point, the eating of their tails, a habit to which many rodents are prone, is reduced to a minimum if they be given some meat. Water by itself was discontinued owing to the difficulty of so fixing the vessel that it neither become spilt nor became contaminated. I am not yet sure that my present method may not be capable of further improvement, but at all events they live on it fairly well, and breed much more freely than before. Another difficulty with these animals is that they are quarrelsome, and that many cannot be kept together, nor can the young of different ages be put in the same cage. In practice I never keep more than four in one cage, and find that the best number is a pair of either the same or opposite sexes. If one be kept alone, it is apt to take cold during the winter. Never keep the sexes mixed, as if there are more than a pair it usually leads to fighting and the death of whichever sex is in the minority. Young of different broods of the same age and sex may be safely mixed together, and perhaps ten or a dozen will live safely in one cage, but they must be split up into smaller lots when about two months old. When putting them together at first, it is essential that * Castle and MacCurdy, Publ. Carnegie Trust. No. 70, 1907. EATS 63 they be placed in a cage new to both parties, otherwise those already at home will fight and probably kill the new- comers. To some these particulars may appear trivial, but it is only by attention to such details that one can hope to be successful in breeding wild animals in any quantity. Breeding. This species will breed during any month of the year, but in a wild state no pregnant females were obtained between December and the middle of February, and no very small Eats before March, so that there would appear to be a distinct " resting " period. In confinement (judging only from two individuals) litter succeeds litter at intervals of from six to eight weeks for about nine months, after which breeding stops for at least three or four months. Young Eats will breed when six months old, but when bred from at this age the litters are small and weakly. I fancy, however, that this may largely be due to the unnatural surroundings, and that in a wild state litters from Eats of that age would probably be quite normal and healthy. The period of gestation is about twenty-one days, and it is best, though not always essential, to remove the male before parturition. Young Eats may be weaned at about twenty-one days, and the male returned two days later, but on no account must the male be put back directly the young have been taken away. Some pairs will go together directly without any trouble, while at other times they will fight so furiously that they can only be separated by forcible means. I therefore usually introduce them first in a cage with a wire partition and sliding-door ; they should remain separated for at least twenty-four hours, and then the door may be opened in the evening ; if they fight seriously they must again be separated, but if it is only a slight scuffle they may safely be left together, and by the next day or the day after they will be found in the same bed-place, and may then be kept together for just under three weeks. The female very seldom destroys her brood, but having once ascertained that the brood has been born, it is advis- able as a precautionary measure not to disturb them again till they are weaned. All particulars of sex, colour, &c., may then be taken, and the sexes put into different cages and allowed to grow. They are practically full grown, so 64 VIGOUE AND HEREDITY far as their bones are concerned, at three months old, but for the next two or three months their bodies fill out considerably, and they become heavier and bigger. Eesults of the Experiments. The main experiment, namely, the testing of the inheri- tance of the wild varieties, has led to others in a rather unexpected way. The first brood, born about a month after reaching home, contained seven young, four of which were fawn-coloured — a mutation which, in spite of the innumerable varieties of these Bats found in a wild state, had never before been observed or obtained. Among domestic Eats, which belong to the Brown Eat {M. norvegicus), the fawn colour is also entirely unknown, but I have since seen specimens and received notes of wild fawn-coloured examples of this species.* In the F^ generation blind Eats, in which the eye was undeveloped and the optic nerve absent, were produced, and also some Eats showing the typical spinning or waltzing movements well-known in a certain breed of Mouse but hitherto unrecorded in Eats. Experiments between Two Wild Forms — M. tectorum and M. alexandrinus. In the paper quoted above I divided the wild Eats into four groups : — (1) Typical tectorum, with pure white under parts, white hind feet, 35 mm. in average length. (2) Tectorum (grey), with pure white under parts, white and brown hind feet, 35 mm. in average length. (3) Alexandrinus (grey), with slate-coloured bases to the hairs of the under parts. Hind feet mostly brown, and 33 mm. in average length. (4) Typical alexandrinus, with hairs of the under parts entirely slate-coloured. Brown hind feet, 33 mm. in average length. * The late Mr. E. Connop, of Wroxliam, told me that he had a fawn-coloured Eat, which had been shot in a neighbouring parish ; several have been procured at Westfield, near Hastings, one of which I saw in Mr. Bristow's shop ; Mr. Pocook (P.Z.S., 1912, p. 671) exhibited a live example caught on an island in Lough Corrib, Co. Galway, where he states that there are many others. Plate II. TWO ITITER FlfiURKS : MI'S nATTI'S ALEXAXDniXrs. Photo by W. S. -idge •two Lmi'ER FKU'RES : .Vl'S llATTUS TECTORUM. EATS 65 From this it will be seen that there are three chief characters to be investigated : — (i.) The colour of the under parts, (ii.) The colour of the hind feet, (iii.) The length of the hind feet. Of these the first is the most conspicuous and easily seen, and has been the one chiefly investigated, the matings having been made for that character alone. The colour of the hind feet proved so variable and difficult to define that only a few notes were made on this character. When brown appears on the feet in conjunction with white under parts, the evidence, which is very scanty, seems to show that such individuals are hetero- zygous for the " colour of the under parts " character. Other evidence seems to show that it may be a separate allelomorph to the main character (i.), though in these experi- ments characters (i.) and (ii.) seem to be inherited together. The third character (" length of the hind foot ") was the one chiefly dealt with in my former paper. From the experimental point of view with live Eats it offers many practical difficulties, as it cannot be accurately ascertained until the Eat is full-grown and dead. Notes, however, have been taken on all those that reached maturity and were killed, but, as the majority are still alive, the ascertained facts are too few to be of much value. So far as they go, however, they quite support the results given in my former paper, and show that the dark- bellied form has a smaller foot than the white-bellied one.* We may now investigate the results dealing with character, " the colour of the under parts " ; in these experiments. Groups (3) and (4) were counted as the same (alexandrirms) , and Groups (1) and (2) as tectorum. (See table, p. 66.) From an analysis of the results, we find : — (1) That the pure white-footed tectorum gave nothing but tectorum. Experiment a. The fawn may be of either variety, tectorum or alexandrinus : in this case they were all tectorum. (2) That seven matings of heterozygous tectorum gave 40 tectorum,, 15 alexandrinus, the expectation being 41 tectorum to 14 alexandrinus. * Tectorum, 7 ex. ; av. 35 mm. Alexandrimis, 4 ex. ; av. 34-2 mm. P 66 VIGOUE AND HEREDITY a I M EH » l-H > O & EH 13 H EH O 00 H I o •«■ ^1 P3 S ea ^ w a i-i *: a CO o u OQ o EH CO EH p 02 Q <) z n EH o o p m -;:- -(- ^-^ -i- CH- CH CH CH 0+ CH CH 0+ CH CH CH CH OQ O CO CO Cq 05 i-H CO oq r^ iH iH •* fb *o *o fo'^ '=o fb t^ fo *b fo ^ t- c^q CO tH i-H crq CD O CO tH CO 05 CO CO CO 00 00 Co rg § 53 ?3 53 53 ? s pA s s s . s g--£ 1 .-«. _•« _•« g •« l|l|ll|-s| "3 CO ■^ lo m T-H CD CM cq i-H cji 05 CO c^ cq CM ■* cq i-H -^ 1 ^ — ' ■ — , — ' ■ — . — ■ * — , — ' ■ — , — - 00 CO s* -t= _fl « 2 ..-1 tH ^ f -« s % .a 00 00 00 CO ,rj o g : : C3 c3 ^ § a o 2 5 e3 00 ,— ^ l-H , , '-' 00 fl '? 9- s ^ £ •g s a c3 1 g O -5 «*-! S '^*~' S y *. ^-' "* „=* £ ^1 =S Oh o ■2 § 1^ •3 03 1 s e § g o ^ 1 2 'S -S -2 •Ks CO pi] a 43 d ^ -a >*-< I-i Ti a 2 c3 = r-o g x> CO 00 00 CO 00 t3 o OQ a 2 o t ^ CO la ^ o .Q 00 ^-s h-i H 1 g 1 -i 1 c3 5g y £ .9h-+ O) .a > f*s bo a«2 tDrtJ o o p o a n CD o oo EATS 67 (3) Pure tectorum to pure alexandrinus (ex heterozygous tectorum) gave 27, all tectorum, in accordance with the expectation. (4) Pure tectorum to heterozygous tectorum gave 9, all tectorum, in accordance with the expectation. (5) Alexandrinus to alexandrinus, whether bred from alexandrinus or tectorum parents, gave 18, all alexandrinus. This shows very clearly that the tectorum and alexan- drinus character have a Mendelian inheritance, in which the tectorum or white under parts behave as a simple dominant. This result is important from several points of view. First, it can claim to be the first recorded instance of a Mendelian inheritance being proved to exist among natural races of wild animals, previous works on these lines having been confined to plants, insects, and other invertebrates. Such a result was to be expected, and is bound, when further facts are known, to have an important and far- reaching effect on the generally accepted views of evolution. Nevertheless, we must advance on these lines very cautiously, for just as the experiments of Doncaster, Bacot, Prout, Newman, MacCraken, and others show the existence of a Mendelian inheritance among certain forms of the same species, so with equal certainty is a blended form of inheri- tance proved for other cases (see Ducks, Meriones, &c.) ; while instances in which hybrids of the F-"^ generation show blended characters are innumerable (e. g. Mules, Finch hybrids, Pheasant hybrids. Duck hybrids, &c.) . The time is there- fore a long way from ripe for enabling us to draw any certain results from these apparently conflicting results. One fact, however, is certain : Mendelian inheritance exists in a state of Nature, and gradual evolution brought about by the blending of slight variations with the normal is accepted, hence it follows that heredity and evolution are due to the working of at least two main laws, neither of which by itself has been sufficient to account for even the comparatively few cases which have up to now been critically and carefully investigated. Secondly, this result is important as enabling us to p2 68 "VIGOUE AND HEREDITY understand how certain closely allied forms can exist under practically identical conditions without any blending (see also Chapter IV.), each form retaining unimpaired its characters, however slight they may be. Instances of such cases are numerous, and will at once strike anyone who has carefully attempted any systematic work on mammals or birds. I mention these two classes, as they have come under the range of my personal experience, but the same pheno- mena exist throughout the animal kingdom. In England we have a common instance in the Field Mouse {Mus sylvaticus and M. flavicollis) ; in the Oriental region (on the mammals of which most of my systematic work has been done) innu- merable instances occur, especially amongst the Squirrels, and some of the Eats (e. g. Mus jerdoni group, &c.). At present, when animals differ slightly in colour, size, or in some other respect, we rank them as separate species, if they occur in the same locality and as geographical races, if they have a different distribution and do not occur together. Such a method is probably the best that can be at the present devised, but from a strictly evolutionary point of view it has its weak points, since the difference between geographical races is often greater than that between species. From a practical, common-sense point of view it is almost impossible to believe that two obviously allied species living in the same environment, and differing only by characters which require critical examination to detect, should not be constantly interbreeding and thus produce one intermediate race. The knowledge of Mendel's Law and the proof that it governs the inheritance between two forms of one species renders it more than likely that such may be the case in other so-called species.* Experi- ments on living animals can alone prove this, but we have now at least grounds for a legitimate and probable assumption: that these species differ gametically in Men- delian characters, and thus remain externally pure to those characters. * It has been suggested, possibly with some truth, that many of these closely allied forms never interbreed, and that this is due to the fact that the males of one species are not excited by the females of the other, (See " Psychology," Chapter XII.) EATS 69 expeeiments to determine the heredity op the Fawn Colour. As mentioned above (p. 64), the first litter bred after my return from Egypt contained three fawn-coloured Eats, all males.* At the time, I was not aware that Eats of this colour were quite unknown, thinking that in domestic Eats the same colours existed that are usually found in domestic Mice. In addition to this I was at that time unskilled in the special precautions needful for the successful keeping of these animals, so that by the time I had discovered their interest, some six months later, I found only one fawn remaining out of the original three. On inspection, a fawn Eat differs from one of the normal colour in the absence of any black pigment. The long black hairs so characteristic of the normal Mus rattus are present but colourless, as are also the tips of the yellow hairs, which are normally dark-coloured, while the under fur which is normally slate-grey in the alexandrinus variety is, in the case of the fawn examples, of a dull drab. A distinction may in the case of Eats be drawn between a colourless hair and a white hair, although neither of them contains pigment. The former is partly translucent and has not got the solid look of the pure white hair. The difference is not easy to explain, and the apparent difference may only be due to the isolation of the " colourless " hairs, but it is perhaps worth noting. Experiments on the fawn character (in reality, absence of black) were able to be undertaken on the same individuals as those in which the tectorum and alexan- drinus character was being tested, and it may be noted that the two varieties tectorum and alexandrinus are equally distinguishable in fawn examples ; those belonging to the alexandrinus variety being much paler, owing to the greater number of colourless hairs. In young fawn Eats the eye is ruby red, becoming darker (apparently black) with age, so that in this respect it follows the eye colour of the Cinnamon Canary, although no sex limitation has been noted, but far too few have been bred to enable one to be positive on this point. * P.Z.S., 1910, ii. p. 688. 70 VIGOUE AND HEREDITY Table of Matings and Eesults to Show the Inheritance OF THE "Fawn" Colour. Exp. No. 8 9 10 12 13 14 17 Female Parent. tectorum alexandrinus tectorum Origin from Exp. Male Parent. wild caught tectorum b 8 fawn fawn 8 tectorum 8 8 8 »> (8 used in 14) i» Origin from Exp. wild caught 8 (used in 13) 8 (used in 10) Besult. 4 normal, 3 fawn 27 normal 7 normal, 2 fawn 8 normal, 4 fawn 4 normal 1 normal, 1 fawn 5 normal, 1 fawn 14 normal, 6 fawn On analyzing these results we find : — (1) Homozygous normal mated to fawn gave all normals. (2) Heterozygous normal to fawn gave seven normal : two fawn, the expectation being equality. (3) Heterozygous normal to heterozyous normal gave thirty-two normal to twelve fawn, the expectation being 33 : 11. There can therefore be little doubt that the fawn colour (absence of black) is a simple recessive to the normal coloration (presence of black) . The real interest of this set of experiments lies in the question as to how and why the absence of the black originated. We must remember that such a character was previously entirely unknown in this species, and when we consider that this animal is found throughout most of the world in close connection with man, by whom it is killed by thousands; and, further, that of late years, owing to its connection with the plague, many thousands have been critically examined by scientific men, yet in spite of its being, in some respects, one of the most variable species known to science, not a single fawn-coloured example has Plate III. Parents F^ ys PHOTOS OF SKINS ILLUSTRATING -FAWN" EXPERIMENTS. (T. Alexandrinus, female 6. Fawn (tectorum), male c. Pi generation, all tectorum (heterozygous) - Alexandrinus - r e. \ (fawn /. tectorum., fawn g. tectorum (heterozygous, same as F' generation) h. tectorum F'^ generation EATS 71 ever been obtained or recorded. Under these circumstances, I think we are justified in concluding that such examples are not found in a wild state, and that it is so unlikely as to be practically impossible that the two wild caught tectorum, which were brought over should have been heterozygous for that character, which is the only accepted possibility along strictly Mendelian lines. That possibility, however, we must throw aside, and we are thus driven to the conclusion that, from some cause or other, a change was brought about in the germ-cells that caused the loss of a character which, on investigation, shows a Mendelian inheritance. What could that cause have been ? The parents were typical examples of the wild Mus rattus tectorum; they were caught in the same garden, and were therefore probably not very distantly related ; there would have been a fair possibility of their having met and bred together in a wild state, yet the evidence given above is overwhelmingly against their producing anything but normal individuals, had they so met and bred. One conclusion, therefore, only remains, namely, that the fawn examples were produced owing to their changed conditions ; in other words, their environment — be it captivity (lack of exercise), change of food, or change of climate — has had an effect on their germ-cells which has been the means of removing a character whose absence may be fixed under the Mendelian Law. We need not press this point further for the moment, as we shall have to return to it later. Commencement op New Chaeactebs. Five of the six matings from which the F^ genera- tion were bred were those of brother and sister, and in each of these matings some of the young were weak and sickly. In Experiment 14, two of them died before being weaned. In Experiment 13, one was blind and exhibited " waltzing " movements, and the other was so weak that it had to be killed before it was six weeks old. Lastly, in Experiment 10, three individuals were waltzers. In one of the blind ones the eyes were formed but the cornea was thick and opaque ; in the other two, not only were the eyes undeveloped, but the optic nerve was entirely 72 VIGOUR AND HEREDITY absent. * It was interesting to note that these individuals seemed to have no difficulty in finding their food or in finding their way about their cage, and they only differed in their habits from the others by being apparently unaware of the approach of a human being, unless some noise was made, when they would bolt into their bed place as quickly as the rest. As, in addition to their blindness, these animals were sickly, none of them were kept alive for any length of time, though it would have been interesting to have found out whether this character was in any way inherited. Be that as it may, it is of great interest to have an instance of sudden congenital blindness, and we may be excused if we digress somewhat on this point. Are we sure that blind animals such as Moles, Mole- rats, cave-haunting fishes, &c., have become blind through the disuse of their eyes ? This has been a point which has always troubled me, for it seems to be generally accepted that, while many organs have been lost through disuse, it is equally denied that any have been gained or augmented (from the point of view of inheritance) by use. For instance, if Whales lost their hind limbs from disuse, why should they have become diminished to a degree far greater than would be brought about by Natural Selection, and yet still persist in some species as minute bones embedded in the flesh? Or if the loss of limbs be allowed by disuse, why should not a Hare's legs, for instance, gradually increase from use till, let us say, they become as long as those of a Jerboa? In other words, as disuse and use are agents acting continuously so long as a species exists in its living state, why should there be a limit to the extent of their powers ? A Whale must use its hind limbs less now that they have disappeared externally than when, under our assumption, the limbs first began to be partially unused — why therefore does not their diminution still go on ? This same argument, though with more force, holds good for the Hare, for if in this case long legs imply speed, which they undoubtedly do, there can be no question that the longer- footed ones would have a better chance of survival, and there can also be no doubt that Hares are not yet swift enough to escape from all their enemies, so that the only * I am indebted to Dr. Batten, who kindly critically examined one of these individaals. EATS 73 possible argument in this latter case is that their enemies are now so swift that the fastest Hares are captured equally with the slowest. The answer that would be given by the Natural Selec- tionists is that a longer foot would probably inconvenience the animal, and thus he would fall an easier prey to his enemies, and that, therefore, the longest possible foot com- patible with agility has been evolved. This reply, though difficult to answer, is hardly satisfying, for, even if we grant for the sake of argument the truth of their remarks, we should still surely expect to find some Hares with longer feet — they might be in a very small minority, they might even seldom live to propagate their species, but none the less some ought to be found in museums — yet on this point we are struck by the extremely small range of variation found in measuring these animals. In fact, throughout Nature one is struck by the constancy of the measurements of any particular species. Now those who have. followed me so far may be wonder- ing what connection this has with blind Eats. My point is this : if a variety of blind Kats arose in a cave or other dark place, they would not be eliminated by Natural Selection, since that particular infirmity would not place them at a disadvantage in the dark ; in fact, if anything, their other senses would probably be more acute, and they might be therefore at a positive advantage. Consequently they would tend to increase as fast as, if not faster than, those individuals that were normally endowed. But how did they originate? This is, of course, a different point. I have made it clear in the case of the fawn-coloured Eats (and I have little doubt that the argument for my blind Eats is the same) that they (the fawn ones) arose owing to a change in their environment. If, therefore, this holds good for the blind ones (and I think we may assume it), then in my hypothetical case the cave environment which gave rise in the first place to some blind ones would still be acting, and every year the normal ones living in that cave and under those conditions would give rise to a certain pro- portion of blind ones, till eventually, and in a comparatively short time, bearing in mind the slight advantage possessed by those born blind, a blind race would be produced. In other words, if, owing to " lack of vigour " brought about 74 VIGOUR AND HEREDITY by environment, inbreeding, or any other cause, blind Bats can be produced, it is quite possible that animals living in caves may have lost their eyes, not through disuse but owing to the unhealthy conditions of the environment. Blindness not being a handicap, but possibly even an advan- tage to those living in the dark, such individuals V70uld not be eliminated by Natural Selection. Further, the loss of sight, v?hich might happen suddenly, as in the case of my tame Eats, v?ould tend, judging by analogy, to be inherited in the same degree as when it first appeared, and, con- sequently, I suggest that in many cases of apparent loss of parts from disuse, the loss has not taken place gradually, but has arisen as a sudden mutation due to the environment acting on the vigour. Where these mutations have not been a disadvantage to the animal possessing them, they have not been elimininated by Natural Selection but, on the contrary, owing to the environment favouring such muta- tions, they have tended to increase in greater proportion than the normal form. Now blindness might (as in the case of waltzing Eats, to which we shall shortly refer) be a Mendelian character, and as such would continue to be inherited with a minimum of variation. It might be merely that the eye was defective ; it might be a normal eye with an absence of the optic nerve ; it might be, as in my case, an absence of eye and optic nerve ; but whatever state was originally produced, that state would remain constant and fixed till some other sadden change of environment should again cause a spontaneous variation. There would have been no gradual decline from disuse stopping at some arbitrary point, apparently beyond the reach of Natural Selection. The character would have had a spontaneous existence, complete or incomplete as the case might be, and would, if allowed by Natural Selection, continue in the state in which it was first formed. In short, change of environment may cause sudden and marked changes in an organism, which may he inherited, without further modification. "Waltzing" Eats. The case of the " waltzing" character, of which I have bred three or four individuals, is on a par with the blind ones, except that " waltzing " is a character known in EATS 75 a certain breed of Mice, though it has not hitherto been found in Eats. This character has been investigated by Mr. Darbishire,* who found that it had a MendeUan in- heritance, although the actual proportions were not exactly according to expectation. That point, however, need not trouble us here, as the fact that it is inherited is sufficient for our argument. According to our present knowledge, therefore, the fawn, waltzing, and blind Eats form three distinct examples of characters which have been acquired owing to the condi- tions under which they have been kept and bred. In the first case, they have been proved to have a Mendelian inheritance; in the second, a Mendelian inheritance has been shown to exist for the same character in other species; in the third, heredity has not been proved but, judging from analogy, we may in all probability expect that it will occur. Natueb op the Change op Environment. Can we in any way account for the kind of change that has brought about these results ? The fawn ones were born within a few months of capture ; the parents were fed on grain and water, a diet which in a wild state would probably have been supplemented by meat in some form or other ; there was the change of climate, from the dry heat of Egypt to the damp cold of the summer of 1909, and they were kept in a confined space. With regard to the other (F^ matings), they were brother and sister matings, and young animals (about six months old) nearly full-grown and apparently normal and healthy. The change of climate may be dismissed as a primary cause. It has produced no marked effect on desert rodents imported at the same time, and Mus rattus itself is found all over the world without any marked differences ; the food also need not concern us much, for the young Eats made normal growth and reached a normal size, which would have been impossible had the right and necessary constituents not been present ; lastly, we have lack of exercise : this, by process of elimination, must have been the chief factor, though by itself it could hardly, by any stretch of imagina- * Biometrika, iii. pt. 1 (1904). 76 VIGOUE AND HEREDITY tion, have affected the germ-cells. What, however, is the first and direct result of lack of exercise ? Chiefly a want of renewal of the body-tissues due to lack of active metabolism. We need not go into technical details, important as this point is, for every doctor and physiologist would prescribe good food, exercise, and bracing climate for any living being who was run down and who needed increased vitality. The case of these Eats is precisely similar ; lack of exercise and, in a lesser degree, change of food and climate had an ener- vating effect on them and lowered their vitality. One of the first results of lowered vitality is sterility (as witness the number of wild animals that become sterile in confine- ment) ; but before that stage is reached we have a stage in which young are born, but decrepit, deformed, and weak — a study of the human race in the poorer parts of any large town will emphasize the truth of this. So with the Eats. The lowered vitality has affected their generative organs, and their young are degenerate ; there are degrees of degeneracy in the fawn Eats— it was merely the absence of a colour factor which did not appa- rently affect their health — but in the second generation, with brother and sister matings, a further stage in degeneracy was reached. I have hinted above that the actual physiological cause was a slackness of metabolism, whch I have referred to elsewhere as a condition of low " vigour." I need not carry the argument further here, except to repeat and emphasize that a change of environment, by creating a condition of low vigour, has brought about marked changes in an organism, which are not necessarily degenerate but which are inherited. Plate IV. .■.■.-,:->^ Photos by W. S, Berridge. MERIONBS CRASSVS SELLYSII. Photos of Living Examples. CHAPTEE VIII. MEEIONES. The original stock of these animals consisted of two pairs from Giza, Egypt : of M. crassus sellysii, called in this article " North," and one male and two females of M. crassus pallidus,* from Atbara, in the Sudan, called in this article " South." The main object of the experiments was to determine whether the transference of these animals from their native dry climate to our damp one would have any appreciable effect on their colour. Management. Unlike the Bats referred to elsewhere, these animals proved exceedingly hardy, keeping in good health and breeding freely, with a minimum of trouble. Their food consisted of barley and lettuce or some other kind of green-meat. Their breeding qages are wire-fronted and about one foot square, with a small box for a bed place. Except in a few cases it was found unnecessary to remove the male when the young were born, pairs living happily together during the whole of the breeding season, or, if necessary, all the year. The young are born naked, and leave the nest at about three weeks old. While in the nest they may be examined with impunity, and no young were lost through interference with them, while on several occasions young from another pair were given to foster parents and successfully reared. At about a month old they are weaned and the sexes separated ; if left with their parents longer, most of them were, as a rule, killed. The young when put together must be about the same age, and a dozen or more will live comfortably together till the * M. crassus palUdm Bonh. F.Z.S. 1912, p. 226. 78 VIGOUE AND HEREDITY following spring. Other individuals, even when full-grown, should never be introduced into a cage where ten or a dozen are comfortably settled, as they are almost sure to be killed. If it is absolutely necessary to mix two batches, they must be put into a cage strange to both. In this country the breeding season commences in April and ends in July or August, according to the season. In the cold and wet summer of 1912 breeding had in most cases ceased by the end of June, and during this time each female will have three or four litters. In one case in which two litters were destroyed three more litters were reared, but otherwise I have not known any female to have more than four litters. In their native country the breeding season (in confinement) would appear to commence about a month earlier. Pbeiod op Gestation. No special pains were taken to determine the period of gestation, but as during the breeding season these animals were always kept in pairs, and the movements of the indi- vidual males always carefully noted whenever it was found necessary to separate them, a certain number of facts have been collected under this head, which, as they show remark- able differences in time, are probably worth noting. In 1910 (as noted elsewhere) * the period of gestation was proved in two cases to have been at least seventeen days. For 1911 I have the following notes : — Pair 21. Male removed and killed July 19th. Young born July 19th. Male introduced August 12th. Young born August 22nd. Gestation 10 days. Pair 28. Male introduced May 21st. Young born June 3rd. Gestation 14 days. Pair 30. First paired May 8th. Young born May 29th. Gestation 21 days. Male removed June 4th. Young born June 23rd. Gestation 19 doA/s. From these data it is absolutely certain that in the first case (Pair 21) the period cannot have been more than ten days, unless it was thirty-four days. There is no alterna- tive, and the latter suggestion of thirty-four days may be =" P.Z.S. 1911, p. 6. MEEIONBS 79 dismissed as most unlikely, since the male was killed within an hour or two of the birth of the previous young. In Pair 28 the female was separated before giving birth to her previous litter on April 26th, therefore the period cannot have been longer than fourteen days ; it may have been less. In Pair 30 they were seen to pair on May 8th, but not on the subsequent days ; but the period may possibly have been less. In the second case with this pair the young were born on the nineteenth day after separation, so that the period cannot have been less. Our observations, therefore, show beyond a shadow of doubt that the period may vary between ten and nineteen days. This shows such a considerable variation that we can hardly believe that a development that is able to be consummated in ten days can be retarded to twice that length of time under similar environment and conditions ; and we are not aware that any such extent of variation has been previously noted in any animal. One can only hazard the guess that the zygote may sometimes remain quiescent for a few days before undergoing further seg- mentation (a case which is, I believe, paralleled by the Eoe Deer and possibly by some Bats) ; or that in default of any ripe ova the sperm may lie in the uterus and maintain its vitality for some days. We incline, however, to the former suggestion, as the female would be unlikely to accept the male except during oestrus. Sex. As in most animals, the sexes are produced in approxi- mately equal numbers, and, as may be seen by the following table, these rodents agree with the usual rule, 127 males and 138 females having been born. I noticed, however, at an early stage of my experiments that the proportion of the sexes in the litters was by no means usually equal, but on the contrary, that whereas females tended to predominate in the early litters, the males predominated in the second litters. The number of litters is unfortunately rather small, and I should perhaps hardly have considered it worth while to draw attention to it except for the support it gives to the sex question which I have dealt with elsewhere. (See " Pigeons," Chapter IX.) 80 VIGOUR AND HEREDITY We may note, therefore, that in the first litters the males numbered fifty-two to the females sixty-two — 45'6 per cent, males ; in the second litters, the males numbered forty-four to the females thirty-five — 55'7 per cent, males, a difference, therefore, of 10 per cent, between the first and second litters. By looking more closely into the table, we see that in the first litters the males outnumbered the females in only seven cases (rather less than one in three), and that in the second litters the females outnumbered the males in only four cases (about one in four) ; so that in spite of the small numbers we are justified in saying that, on the average, there is a larger percentage of females born at the beginning than at the end of the breeding season. Table showing Peopoetion op Sexes in Dipfbeent Littees. Pair Exp. No. 1 Litter 2 Litter 3 Litter 4 Litter Bemarks 1909 3 % (? 2 = 66%^ * These numbers do not agree exactly with those previously given, owing to the elimination of fourth and subsequent clutches. PIGEONS 101 From a close study of these figures we note that in every case (except the MeaUes) the smallest proportion of males was produced in the first clutch, and a slight rise in the proportion of females in the third clutch. Were we only to take the aggregate result, significant as it might appear, it would be unwise to lay too much stress upon it ; but on looking at the figures piecemeal and critically, one cannot but be struck, in spite of the smallness of the numbers, by the close agreement there is between them. Thus taking the first four sets, and reducing to roughly similar figures, we find : — 3' 2 oooo oooo Cream Mealy. oooo CO P3 P O O a Q M 1 O O S-2 OrHOO It tH Cq O rH ^.2 OO OCM It iH O iH O S 3 i-( i-l CO O Mealy Expeeiments (Males). o ^ . . . 00 I-H o cq 00 X C3 o iH cq th 00 00 00 CO iH iH r-( iH ! = = = 60 P ! = = = o = = = J-< i o CQ OnSOOO OJ hi CO o U ^ fl O as '^ O I-H T-l ^1 I-H P i-^ ft as »o O O O > o r£3 r-H S .Efg tH OiHt-( o U P ►^a iSl < 1 O cq o tH H , o .M >. o . o §•3 ^ ooo ?j .3 o thoo n -» o 6D o " " " ^ CM 00 rH CO t-t O o I CO ■* 00 (m CO eo o 0+ fo'" I I I I 1^ IS I I I CO I CO 1—1 lO I iH l-l I I I I I I I kl I I I I I I I . I :o CD , ,— X : rH 00 lO ^^ ^ CD i« -- -■ ,■ iH oTio a 1 00 S °? 52 l!or-|SCDCO lO I 00 H CD -* S CD CO ScDrH a S""* g rH I .S t- a! .9 s S _ a fl (D Si s. ■ i-i S ? g ^1 ^ 'S ^ o o o o o*^^-*^-" .42-»S4^4^-^ O O © © H u S-1 H ij'^'^'N'N i>>>>i>gg.2.2 COb20C!CQD2PQWOO o "^ 1 (MrH 1 O CO B . m [^ CO 1 1 iH t- i(5 ^1- Ot tho) 1 cq >o <»5 [fo |°° 1° o d-^ ^ 1 CO iH rH CO GO bD [^ 1 ►^cS .^ r-IOOO iH ^-i 0+ cq O (M tH »o cs.H ' «^ fO 1 rHr-lrHrH T« a >> l'^^ 1 ^ 1 1 >o 2|- ,^o |o 1 1 o . ot 1 =^«=<^ 1 1-1 2 iH ^a fb 1 '«=^==' 1 Oi 00 V iH /■ m . o+(Mcoocq 1 c- EH ^i? 1 15 s S- H «^ f^O t- CO o 1 s « V Ph >< : -* : : CD .00^-00 : th o : "-1 • -03 E5 -Soo a! 3 O " T-l rH

^^r^ S.2 -3 x!H afl 'a1-2§:i> Mealy to Da Mealy to Gri Mealy to Gi Mealy to Bl Grizzle to L ^ j^SS-a § ^3P3>^ Qfih^^^qfi PIGEONS 119 we must, however, go a stage further. It would seem as if (where a weak colour is possible) we tend to get a pre- dominance of females. By the phrase " where a weak colour is possible," I mean where the Mendelian character for a weak colour, say Silver, is present; if it be not present, these females will not appear, and we must therefore assume that these pale females will be born as dark females, and it may be for this reason that males usually pre- ponderate slightly over the females. Let us go into this matter in a little more detail. It is certain that at the moment of fertilisation the sex of the individual is not determined, and from experiments of various people, and especially from Yung's famous work with Tadpoles, it has been proved that in some cases, at all events, the sex may be influenced during the young or larval stages by the environment. Now what holds good in the lower animals is probably equally true among the higher orders, though the period of plasticity as regards sex is much shorter, and, from the absence of a free living larval state, much more difficult to influence. Nevertheless, work on these lines has been attempted,* and Lalanie has shown that in an embryo chick the sex is determined during the hermaphrodite period between the seventh and tenth days, that after that period the sex is definitely determined, that before then the cells of the future generative organs have not begun to develop. We may, therefore, accept this evidence as showing that during a certain period of embryonic life there is a " struggle" for the determination of sex, and that its determination may be influenced by nutrition and temperature, i.e. vigour. Now a "struggle" similar to this between maleness and femaleness may, and probably does, take place between other characters, as has in fact been suggested by Weismann, but it does not follow that the " struggles " all take place at the same period of * Professor Sutton (P.Z.S. 1888, p. 432, "Evolution and Disease," 1890) has suggested that there is a transitory hermaphrodite period in many animals, and that unisexuality has been brought about by the hyper- trophy of the female cells (side) on the preponderance of the male, and Messrs. Gteddes and Thompson, arguing on these lines, point out ('Evolu- tion of Sex,' p. 54) that favourable conditions of nutrition, temperature, and the like, tend towards the production of females, the reverse increase the probability of male preponderance. Lalanie (Gomptes Bendus C.I., pp. 393-395, 1885) recognises in the chick three stages of development, which he calls : — (1) Germparity, (2) hermaphroditism, (3) unisexuality. 120 VIGOUE AND HEEEDITY embryonic life. We know that within certain limits temperature and nutrition will cause an adult to become brighter and darker ; hence we may well assimie that the " character " for darkness or strength will require, even in the embryonic state, more and a larger share of nutrition than a character for paleness. Now everything goes to prove that femaleness demands abundant nutrition, hence after the character for femaleness has been definitely de- veloped, it stands to reason that the amount of nutrition available for other (somatic) characters will be less, and in consequence there may not be sufficient to enable the "dark" and " strong " characters to develop, but the "light" ones needing less may be able to do so, and thus we should expect, so far as colour is concerned, to find the majority of females pale or sombre coloured. Supposing this argument to be sound, we have, as yet, no evidence as to whether the sex or " colour " character is the first to become definitely fixed, but the argument so far holds good either way. For if, contrary to the way we have expressed it above, the colour be determined before the sex, then if the " dark colour " is predominant, less nutrition is available for the sex determinant, and a male will result. It, however, by no means follows that these struggles dependent on " vigour " are the sole means by which the development of any particular character is determined. The Mendelian theory gives us quite another explanation, which, though in our opinion it breaks down in numerous cases, and in its strictest sense is not borne out by the facts, yet none the less its principles roughly accord with the results attained. Now to revert once more to our concrete case of Blues and Silvers. If an individual be homozygous for the Silver character, then the amount of nutrition available will be greater than if it was homozygous for the Blue character; therefore, presuming that the sex character is fixed before the colour character, the probability is that most hen birds will be Silver, and this is in accordance with the facts ; similarly it follows that among males the majority will be Blue, but it also follows on this reasoning that if there be sufficient nutrition we may get Blue females, or from a lack of nutrition Silver males. On this line of reasoning, therefore, we can see quite clearly not only why hens should as a rule be pale coloured, but also why, when PIGEONS 121 the character for pale colour is present in the germ, we may get a surplus of females. It tends to follow from this that Mendelian pairs of characters differ in degree rather than in kind, and that a higher vigour may cause a character to have one " out- ward " expression, and a lower vigour will give the same character a different outward expression. If this be so, Dr. Archdall Eeid's suggestion* of "patent" and "latent" reproduction comes near the mark, since but one character is inherited which develops one way or the other according to the "vigour," the difference between Dr. Eeid's explana- tion and ours being that he supposes the undeveloped character to be latent, whereas we do not believe it to exist. It is, according to our view, analogous to what obtains in certain plants, e.g. Berheris vulgaris, which bears leaves in a moist, and spines in a dry, atmosphere. The gametical character of these leaves, whether it be a single or a double one, obviously owes its somatic expression merely to their environment. This point of view is borne out also in Lalanie's works, for this observer claims for the chick to have observed the growth of male cells and degeneration of the female cells, or the persistence of these latter in the embryo. * ' Laws of Inheritance.' 122 VIGOUE AND HEEEDITY CHAPTEE X. DUCKS. My experiments on the hybridizing of Ducks date back to 1900, when the theory of Mendel had only just been brought forward, and when the study of genetics as we now know it was in its infancy. The first and primary object of the experiment was to test the fertility between different wild species ; secondly, when it was found that for a certain number of generations some species were perfectly fertile inter se, attention was directed to the different vari- ations, especially those which tended towards other species that were not included in the ancestry ; thirdly, with the spread and general acceptance of the Mendelian theory, matings were made to test, if possible, whether any Men- delian inheritance was to be found ; and, lastly, the whole work has been revised, and further matings made to test the results on our present hypothesis. Several papers* have been published on the interim results, and although the facts already published will be again mentioned in the following pages, the main portion of this chapter will be devoted to considering the results as a whole, and how they lend themselves to a plausible ex- planation on the lines of our hypothesis. Fektility. The species used in my experiments were as follows : — The Mallard - - - {Anas boschas). Mellers Duck - - {Anas melleri) (figured P.Z.S. 1864,p. 487). The Spotbill - - - {Anas poecilorhyncha) . The Australian Duck - {Anas superciliosa) . The Pintail - - - {Dafila acuta). * " On Hybrid Ducks," P.Z.S. 1902, p. 318 ; 1905, p. 147 ; 1909, p. 598. Proo. iv. Int. Orn. Oongr. 1905, p. 235 (1907). DUCKS 123 The Mallard is too well known to need any description ; the sexes are very different, and the male assumes an eclipse plumage after the breeding season. The Mellers Duck is a very dark form of the Wild Duck, in which both sexes are alike. Eestricted to Madagascar. The Spotbill is an Indian Duck. The sexes are alike, and there is no visible eclipse plumage, both sexes being very dark brown with lighter edgings to most of the feathers. The Australian Duck is a uniformly dark brown bird, with lighter markings over the eye and on the face and chin. The sexes are alike. The Pintail is a well-known species ; the sexes are dis- similar : the male has the under parts white, the flanks vermiculated like the Mallard, but the vermiculations coarser, the inner secondaries and central tail feathers elongated. There is an eclipse plumage. The whole build is longer and slimmer than the Mallard. The female is dull-coloured, of a paler and greyer brown than the Wild Duck, but of characteristic Pintail build. The accompanying genealogical table will show at a glance the nature of the matings made. In order to make it clear, the full number of matings and young are not given, but every different cross, either from species or relationship, is shown. It is evident from this that these species when crossed are perfectly fertile inter se, and that the addition or sub- traction of another species made no difference to their fer- tility, the " pentagens " * or combination of all the species being just as fertile as the first crosses of two species only. When, however, brother and sister matings were made, and carried on in successive generations, infertility, or a partial infertility, was soon reached, usually in the third or fourth generation from the last introduction of a pure species. The introduction of a pure species, even though different to one already in the cross, led to partial rejuvenation or further increase of fertility, and showed clearly that the infertility was not due (to any extent, at all events) to the crossing of the species, but must be ascribed to the usual * The word "hybrid " meaning merely a "cross," a new word had to be coined to imply crosses of more than one species. "With the assistance of the late Prof. Newton and Prof. Skeat it was decided to adopt the root yivos prefixed by Greek numerals. »°'0 S lO 0] ■§3° .^s 5 Sag 3 M n &. s - - " ° _ " 1 S o m"— ' ki m 3 o S cS € ^ t^ S II II Qj goPS o 'it 4^-" R SI. -Si 3i judy - _T_ _:;:__. i_. ... : :: ^^ __ -_ . . . _.:. _-^ : .v.._ . .__ . , . - _--tI V--- - I -I - _ _. ? : .\._ SjmaB[ - - _ _ i . .; .. . ^~ ::: . : .-_:. -.3_ ._-! ::_,, . . - -_ _ ..i: .1 _ - _- -i-i: :i. : _ _ JBqtneAOK - _ -__: t : :; I- . :: 5 I : -iiii-i-S t :_::■ ..: . V. loqtnoqdo^ - \ :::::::::: i:::: .::::: : %-- : : : _ 2::: : : . ;_ . _::: -^z _: : : - : *pr " _ :. : : ^it : : . __ __>,.,. __ _ - -f ~ -- :".:_: z^" ::: ::: " 5 • t-m ~ _ _ _ _ \. : : _: 5 : : fudw :._:, i^^-s, .: ., : .. IiKiy - __: - S--. _ _ -_ - .L .! tj:>iii^ __, \_ _ __ ._ ^ c -^ : :: _: :t 4 --■ -'-■■ - naq g_ _ ] Ooniref J ... _. _ — f .. jaqureAOi; IMOTOn --_i .___. joqaiBldeB . :._ it .. _ . : It ,_g_ -::::.;:: X .: : . _ £.jat . t -- -_ ::: u .- ::::::: it : :. .: : : :: ji : .- - :.: :: ji- . _ ^m E : _ - H:::t" :: TrtlV L 1 1 1 1 M, M NJ, = - -3:.,t-,i -?^ 00 -^ §.S S 00 ^ a O-ao 53 o O o • ^^M to toy. If THE EVOLUTION OP SEX 169 li c 1 1 1 c > c b < < s St c <: ^ c 3 r > ^'--r / aeqoqooL [ ? \ "7 5;: -' 5 ^S ■ qsnSnv • s" ' '- ' ^ "5 Gunf 5.5 . 1 V , V \ ludv V V v> \ \ 1 imuqaa / - f '• /' joqmBCQQ ,-■'''-- > ^ "1 ^' joqoioo > / N £ S "I ^^ ' ^BnSny ^/_ c^^ _; ^ \ T V, eoof 1 S \ « ^ ^ V 1 > judV V \ i _ / / iimuqaa V 1 • ^ - y ■ H ^ 1 3qra«»a ^A^ 4"^ / 1 1 1 170 VIGOUE AND HBEEDITY O 1 n 1 If f 1 r,r [ 1 "1 1 1 .i-i-^t-j f c 1 I'l j,i 1 lit V [1 ' ■^sl ^-}l 4 ?. ^ ^ f «> ■ iiUlIii- O cP~ ft, _ _^ _ , ! 3 I ^ BOrtJI ^ ■ _ S' _ ^_ -_ ^E.- _ v: _ ., . _ nuqBJ I - - :[^ : it innf 1 2 ^.^ . _ ._ t --i - i-t :: L. __ - _ _ 5 - ^ : __ . - r \ J I- MAON _ 10 ! 5 3 : : _ - - S' -,- - ■ .-^ ofdss ' :: c s :: _ _ ; _ - ■■ - - -17 ■■ - - -_ _ .. JL^ _.. iw _ U-.- _ ^ _ J - ^ 1% - - fliraj j . - im ^ : . I s . " ^ c/^ _. _ ._ ^ ^ -I tOCTjq 2 . _ :: r ' ■na«f : :: ^.v : ^ "-i i ' . _ : ':.± !l^ :: \^ »AOK : :: ? : ^ " «mn . .. d\.. btdog . .. ■■ \.. jZ ^ ^ - z " - - }-■ ; ^. : _ ' », \ ipi r^ ^ r » T einlf. 1 J 1 1 1 itlB Iflj i *A 1 i ' A J il^ it_ .. 3_ : 1 ill M 5° "" El --I o tao"' ^^^^ •** -*^ rrt . S13 » 2 -" +=-a o »-( Zi _ rj ^H +3 ■"" 03 ^3 fe 08 ts o o => _ cSja .9 J3 •*^ ^ *"-s a- s s s ° eS*! 3 a " ^ a <3 too o ''^ © •S "^ =s £ ffl.TS O 'm'-< a *= O O o 00 'O — "^ © C8 M THE EVOLUTION OP SEX 171 effects of vigour than pages of writing ; one point, however, still needs explanation: we have hypothetically considered each animal as commencing with an initial vigour of 50; in practice, as we have tried to illustrate, this initial vigour will vary slightly in each individual according to the breed- ing vigour of its parents, though among wild species this individual variation will only be very slight, and for the moment we may disregard it ; but as between wild species and families this normal initial vigour will show a consider- able range of variation. In the Polar Bear, for instance, it is probably considerably lower than, let us say, in the Musk Ox. Now these animals of normally low vigour will, provided they can obtain sufficient food, accumulate fat more rapidly than animals of higher vigour, and hence it follows that when under the more favourable conditions of spring their vigour rises, the "katabolic crisis," which is equivalent to the reproductive period, is reached at a much lower level of vigour than is the case with animals of higher initial vigour. By the same process of reasoning it follows that in animals of normally higher vigour the level of torpidity will be reached under more genial circumstances than in the case of the lower vigoured animals, and consequently death may be reached when the other animals are just becoming torpid, and in the same way animals of lower vigour will die under warm conditions suitable to those of higher vigour. In Table C we have illustrated the domestic hen, which we may take as typical of domestic animals, although it would apply equally well to many wild species (e. g. Eats) ; here we see that the -effect of the seasons, so marked in the case of the Polar Bear, is practically absent, and the main fluctuations in vigour are almost entirely due to exhaustion from breeding, the normal level being but very slightly below the reproductive line.* Now, although the fluctuations are comparatively slight, the vigour does fluctuate, but almost entirely above the line of reproduction; hence it follows that young will be born in almost all states of parental vigour — instead of at a fairly fixed * It must, however, be carefully borne in mind that though the reproductive and other limits are reached at different vigour levels according to the species, the colours are indicative of certain fixed degrees of vigour, so that if they are typical of a vigour of 60, they will never appear in an animal whose reproductive limit is 55. 172 VIGOUE AND HEREDITY level, as is the case in wild animals ; hence it follows that the initial vigour of young, born under these conditions, will tend to vary much more than in the case of other animals, the yearly fluctuations of whose vigour is much greater. In the case of pure bred domestic stock, the "characters" being the same in all individuals, this variation in breeding vigour will not show itself except in difference of shade ; but if the stock be crossed, this of itself will enlarge still further the limits of the breeding vigour, and the young being heterozygous we get a combination of factors that, to our mind, must play no inconsiderable part in producing the somewhat complicated and confusing results with which we are always being met. If 50 degrees of vigour represent the initial vigour of any animal, and if we suppose that this vigour may fluctuate, owing to its environment, 30 degrees on either side of this mean, then, as the vigour rises, we first reach a state of song, activity and brilliancy of coat, plumage, and other secondary sexual characters. Very slightly above this we reach the lowest limit of reproductive powers which may extend through five degrees. Below the mean torpidity results at 35, and death at zero ; while, in a similar manner, torpidity and death may result at the highest points, viz. 80 and 100. The degrees and the number covered by the fluctuations are of course arbitrary, but that there are limits of torpidity, death, activity, and reproduction are facts, and may be demonstrated on any animal with whose conditions and modes of life we are sufficiently acquainted (e.g. young birds on getting cold reach a state of torpidity which would end in death, but if their temperature be raised they immediately recover). For instance, in birds (Table A) we have therapid fall in vigour due to the exhaustion of the breeding season and dis- played in many by their dull winter plumage; this is followed by a recovery in early autumn, especially if it be a mild one, leading in many species to a resumption of song, and, in specially mild seasons, to reproduction. The advent of cold weather in December and January will cause the song to cease, then gradually the coming of spring leads towards the reproductive limit, and thus the cycle is complete. We therefore see that as the conditions affecting the vigour vary, so does the animal offer a ready response in its habits and actions. Some may be inclined to think that individuals THE EVOLUTION OP SEX 173 do not so readily respond to alterations of "food and temperature," and may point to the fact of birds led on by an early spring still endeavouring to attend to the duties of maternity in spite of adverse conditions which may super- vene ; but we must remember that, just as in the beginning, " sex" is fixed, and once fixed is immutable, so a nest built and eggs once laid, attempts will be made to carry the matter through. Sometimes it may succeed, but one could point to many instances of nests deserted through adverse conditions. Besides this, among the higher animals, psycho- logical forces come into play, with which we are not at present dealing. Furthermore, if wild animals, more especially birds where the song and mating impulses are more marked, be kept in a semi-domesticated condition, their behaviour during the early spring will be found to vary exactly according to the external weather conditions. In wild species, where the conditions of every individual are approximately the same, the vigour will tend to vary regularly with the seasons, and the initial vigour will always remain at a fixed point, whereas in domestic animals, as we have shown elsewhere, the vigour will be found to fluctuate widely. Now the actual limits of fluctuation, which is brought about by external environ- ment, will depend entirely on the " initial vigour " of any particular individual. Thus in our case, where we have presumed a fluctuation of 30 degrees on either side of the mean, an animal whose initial vigour is 50 may fluctuate between 20 and 80, whereas one whose initial vigour is 40 may fluctuate between 70 and 10, and the limits of breeding, song, &c., will also fluctuate. For instance, if conditions in which an animal is are suddenly changed, it is bound to cause a physiological shock, and a certain amount of physiological energy will be expended in adjust- ing itself to the new surroundings, the result of which may be that it is unable at first to reach the normal reproductive line of vigour, and in consequence becomes sterile ; this is primarily the cause of infertility among so many wild animals in confinement. The new conditions may, however, not be altogether unfavourable, and in the course of a year or more the animal slowly adapts itself to these conditions, and is thus able once more to reach a high rate of metabolism and reproduce its kind. 174 VIGOUR AND HEREDITY Now " adaptation " in this sense has probably involved the " getting up steam," so to speak, on a different fuel (nutrition), and consequently the semi-digested nutritive material composing the ova vyill be of a different value to that produced under the old and normal conditions. It may be better or it may be inferior, but its result will obviously affect the progeny. We have already shown that the vigour of the female is inherited through the nutritive material of the ova, and hence, conversely, it follows that the "initial vigour" of the progeny will be affected, and on this " initial vigour " depend the develop- ment or elimination of many characters which, once formed, remain immutable during the life of the offspring. Now what is true for one individual is true for many, and so if a whole country undergoes a change of climate all the individuals of any particular species will be simultane- ously and similarly affected, and in a very short time a new race adapted to its environment may result. As the genera- tions succeed one another, this new race will become more and more adapted to its surroundings, i. e. its physiological metabolism will become more complete, leading probably to a normal fertility, and gradually the nutritive plasm of the ova, which is due to the nutrition, will alter in quality, and thus the characters of a new race become formed, fixed and inherited. This, of course, is but one of the factors in the foundation of a new race ; the continuance and success, or otherwise, of that race must then depend on the innumerable factors of the evironment which we group together under the heading of Natural Selection. Thus it is that we suggest geographical races of various widely distributed species have arisen. The slight differences of climate are chiefly those of heat, light and moisture — factors which, as we have already seen, act most readily on the vigour, and thus we find that slight differences in shade and tone of colour and other characters have been brought about. This view receives added confirmation from the fact that these racial differences merge, as a rule, gradually the one into the other, varying directly with the climate. Those systematists who interest themselves chiefly with these matters, from the purely morphological point of view, frequently find that although these climatic differences, slight though they be, hold good THE EVOLUTION OE SEX 175 for practically 99 per cent, or 999 per 1000 individuals, yet examples continually occur which are not true to the type or normal form of the region in which they have been captured. Ornithologists, unwisely as we think, describe these abnormal varieties as wanderers from their true habitat, and in this manner the range of various forms is thereby extended; thus, for example, the Siberian form of the Yellow Wagtail , has been recorded from England ! The assumption that this individual has wandered from Siberia or India, of which actual proof is, at present, impossible, would be plausible enough if these aberrations were only found among birds — but they occur also among mammals, and more especially among the smaller mammals, where any idea of migration is out of the question. It is, therefore, obvious that some other explanation of this phenomenon must be sought. To our mind vigour supplies it, for if, in spite of any change in climatic conditions, an individual is able to maintain a rate of metabolism typical of another locality, that individual will automatically tend to resemble his foreign congeners, irrespective of where he was born or the actual colour of his parents. The natural corollary that follows from this is that "vigour" may frequently be the origin of variations, but as this has been fully discussed previously, when dealing with the Eat experiments, we need not pursue the matter further at this juncture. Now, conversely, if climatic conditions acting on "vigour" bring about variations, it follows that if an animal is able to maintain a constant rate of metabolism, in spite of changes in the climate, he will not vary throughout his range. It is doubtful, when we come to know more about any particular species by comparisons of large series from widely different localities, whether any species actually reaches a stable form throughout a wide range. But it is a fact that many wide- spread species are remarkably constant. Among mammals the Brown Eat {Mus norvegicus) is probably the best example, but, as we have already shown, he is omnivorous, and conse- quently is probably able to digest with equal facility any nourishment which may come his way, and is thus able to maintain a normal and equable rate of metabolism wherever he may be found. At the same time he is not absolutely 176 VIGOUE AND HBEEDITY constant — those living in sewers are, as is well known, much larger than those found in stacks and the open country, while in England, and to a still greater extent in Ireland, black varieties are by no means uncommon. Among birds, which are great travellers, we find more examples of those of wide range differing but little in whatever part of the world they may be found, but the bulk of such species are found among the Terns, Gulls, Waders and Petrels, birds whose food chiefly comes from the sea, and is therefore much the same throughout the world ; and, in addition to this, though they may range far and wide during many months of the year, their breeding quarters are, as a rule, comparatively restricted and often circumpolar; the conditions therefore, under which the young are born, are restricted and enjoy similar climatic conditions, so that the " initial vigour " of the young does not tend to vary, and the race is thus kept stable, e.g. Waders, G-annets, Terns, Gulls, Albatrosses, &c. Thus, then, the nourishment throughout the year being fairly constant, and the breeding-places restricted and nearly the same as regards climate, each individual of any par- ticular species starts with the same initial vigour, and the nutrition being constant, the rate of metabolism remains so also, in spite of the variations in temperature. There is, however, one case among land birds which is worthy of further consideration, namely, that of the Hooded or Eoyston Crow {Gorvus comix) . This species, which is now generally considered as a race of the common Black Crow, ranges throughout Siberia and North and Eastern Europe, including Egypt. Where the breeding-grounds of these species meet they interbreed, and the resultant offspring is fertile, so that in those localities individuals in all intermediate phases of plumage are found. The resident Egyptian race is indistinguishable from those whose home is in Northern Russia, and we may well wonder how the two races can remain identical under such dissimilar conditions. Two explanations are possible : (1) That though the conditions are so dis- similar, the net total of their results may be the same. Siberia, roughly speaking, is cold and damp ; Egypt hot and dry. Now the conditions favourable to a high vigour are heat and damp ; consequently, the gain to vigour from THE EVOLUTION OE SEX 177 the extra heat may be counteracted by the lack of moisture. (2) The other suggestion, and the one we are inchned to favour, is that the metabohsm of this species reaches its maximum in temperate localities, and that extremes of either heat or cold are detrimental and lead to a slower rate of metabolism ; and in support of this we may notice that several other tropical species, e. g. Corvus splendens, show a tendency to the markings of the Hooded Crow, but when we reach the temperate regions south of the Equator, the black type again predominates, e. g. Corvus capensis, of South Africa, and C. a/ustralis, of Australia. And this seems to be a plausible and logical explana- tion of a^fact, of which an explanation has hitherto not been attempted. Before finally quitting the cases of wide and discon- tinuous distribution, mention should perhaps be made of two other groups : (1) The Snowy Terns * (Gygis), (2) the Gannets {Sula sula, S. cyanops, S. piscator) . These species are found nesting in the Tropics in both the Eastern and Western hemispheres, and yet those from the East are practically indistinguishable from those in the "West. The explanation on our hypothesis is simple. Oceanic life throughout the Tropics does not vary to any extent ; both the temperature and the fauna are in the main similar, and thus it is only to be expected that the same species under similar conditions will show no definite or marked variations, whether its home be in the Eastern or Western hemisphere. Acquired Charaotees. We may now consider briefly the very large question of the inheritance of acquired characters. We have seen how the effects of climate acting through vigour may cause changes of form and colour, which, as long as the conditions remain the same, will be possessed by every, or nearly every, individual of that race. The first question, therefore, that we have to answer is. Do the individuals of each generation acquire their characters de novo, owing to their vigour being influenced by the climate from birth, or do they inherit from their parents or grand- * These have recently been differentiated. 178 VIGOUR AND HEREDITY parents the vigour which these latter have acquired from their environment ? We have already shown that the initial vigour, which means so much to the newly developing embryo, is almost entirely inherited from its parents, and consequently it fol- lows the initial vigour of an embryo will vary directly as the vigour of its parents ; hence, any alteration in the vigour of the parents brought about by environment will be inherited to some extent by their progeny. We may thus state definitely that acquired vigour is inherited, and therefore characters which are dependent on that vigour for their development will be inherited so far as their somatic development is concerned. This is, of course, not quite the same as saying that acquired characters are inherited. For instance, an adult animal (A) might find itself under conditions leading to high vigour, its progeny (B) develop- ing under these conditions, and possessing a high initial vigour might grow much larger than its parents (A) ; the children (C) of that progeny (B) under the same conditions would grow as large as B. Now it might be said that C inherited the character for size from B, but although the method and form of inherit- ance in each case are identical, we could not say that B inherited this character from A, since A did not possess it. What was inherited was the rate of metabolism, or vigour, which was acquired by A and inherited by B and C. In everyday life doctors and others have tried to beg the question of inheritance of disease by saying that the disease is not inherited, but that the tendency to the disease is. Whether or not those who argue thus are fully aware of all the great truths underlying that word " tendency," it would be hard to say ; but those who have grasped the full significance of " vigour " will see at once that " tendency " and " vigour " are synonymous terms. We need not here enter further into the pros and cons of the many arguments for and against the inheritance of acquired characters. Weismann has shown fairly con- clusively the continuity of the "germ-plasm" and its " segregation," so to speak, from the soma ; but while accepting that statement, it is equally certain that the germ plasm is dependent for its nourishment on the soma, while the soma in turn depends for its nutrition on its environ- THE EVOLUTION OF SEX 179 ment. Therefore, to a modified degree, the germ-plasm is dependent on the environment, but it only receives its nutri- tion after the raw material has been partially assimilated by the soma, and the chief factor regulating that assimilation is the rate of metabolism, or vigour ; hence, reaching the crux of our subject by another line of reasoning, we see that any alteration in the germ-plasm, by which alone inheritance can be affected, must be governed, to some extent at least, by vigour. These considerations lead us to the question of the differ- ence in the part played in fertilization and development between the nuclei of the ova and spermatozoa and the protoplasm which surrounds them. The matter is still undecided. Herburg and Strasburger agree in considering the nuclei as all-important, and the action of the surround- ing protoplasm as of but little account. On the other hand, Boveri, and more recently Delage, have succeeded in obtaining normal larvae of Sea-urchins by fertilizing non- nucleated fragments of the protoplasm of ova. Although we are unable to add any new suggestions on this subject, the question must to a certain degree affect us, since it will be the cell protoplasm rather than the nucleus which will be most affected by the "vigour " ; and we are inclined to the belief that all the germinal changes which we have shown to be due to vigour must have been brought about through the influence of the cell protoplasm. This is, how- ever, a further aspect of the question which must be left for the future to decide. CuMtTLATIVE EFFECTS. Our arguments on the inherited effects of the alteration of environment have been experimentally proved by Delage and Drinkwater, to whose work we have referred in detail elsewhere. These observers found that not only did a change of en- vironment lead to variations, but that when the conditions were again made normal these new characters tended to persist for a few generations. To our mind this points definitely to the conclusion that the alteration of environment has affected the germ-plasm, for otherwise we should have had an immediate return to the original form when the normal conditions were resumed. N 2 180 VIGOUE AND HBEBDITY Now it is a matter well known to practical breeders that a character tends to show itself more frequently in the off- spring the longer the time {i. e. the greater the number of generations) during which it has existed. Thus, for example, say a certain character X appears in an individual, if that individual be immediately crossed with an individual of the normal form, X will tend to disappear. If, on the other hand, this individual be carefully mated (preferably to its blood relations), in course of a few generations, by a careful selection, a strain will have been produced which will give a large percentage of young bearing that character. These young, when mated with any normal individual, will have progeny in which the particular character X will be far more predominant than at first ; and the longer the strain is kept pure for the particular character the more strongly inherited will that character become. This is a fact of inheritance that has not so far been taken scientifically into account. Mendel's theory does not help us much, as the characters with which he deals are well marked and often differ only in kind, whereas these other characters are usually quanti- tative, i. e. increased or lessened size, intensity of colour, and such like. Galton's Law does not touch this aspect of the question either. We have, however, on this point a hypothetical sugges- tion to make. To our mind the solution will probably be found to lie in the different parts played in inheritance by the nuclear and protoplasmic elements of the ova and spermatozoa. If we consider the true inherited characters to be transmitted by the nuclear elements alone, and the pro- toplasmic contents to serve merely as nourishment for those characters in their early stages of development, we may then realize that, owing to the quantity and quality of this protoplasmic nutrition, a certain character may develop and show itself in the soma. That development may, how- ever, have taken place after the segregation of the germ- plasm in the new embryo, and thus the " character " con- tained in the germ-plasm will not have been affected by any subsequent development of the soma and will, therefore, not be transmitted. If, however, this individual be mated back with one of his blood relations, the protoplasm in the germ- cell of each will tend to have the particular properties which THE EVOLUTION OF SEX 181 led to the somatic development of the character and, in the formation of the germ-plasm of their progeny, this particular protoplasm will be the nourishing agent of the nuclear ele- ments of that progeny, and thus the character itself will tend to become modified; this modification will thus increase with each succeeding generation, so that when eventually after several generations an individual showing this character is paired back to the normal form, the modified character will be inherited instead of merely a somatic growth being stimulated. In short, the first individual showing that character does so owing to the somatic development being nourished by a certain germinal protoplasm which has, however, not taken part in nourishing the nuclear elements ; by continued in- breeding the nuclear elements themselves owe their deve- lopment to this same protoplasm and, consequently, the characters themselves become altered and are thus for a few generations inherited in their altered state, although the external conditions are again normal. It thus follows that an external character of an animal will, in these cases, tend to be slightly more exaggerated than the character as actually transmitted, but that in each successive generation the character as transmitted will tend to approach more and more to its somatic development, till in course of time they become identical. With Mendelian characters we shall deal in a subsequent chapter, but it may be well to mention here that Mendelism deals with the total suppression of one of a pair of charac- ters rather than the greater or lesser development of a single character. In short, the environment affecting the vigour will influence the germinal protoplasm, but only to a lesser extent the nuclear elements of the germ. In the first generation, the characters of the soma being nourished in their early stages by the modified protoplasm, will them- selves be modified. In. the second and succeeding genera- tions, however, the nuclear elements themselves will owe their development to this modified protoplasm, and will thus in their turn become modified. The argument for vigour has now been so far explained that we may for a moment consider further some hypo- thetical explanations. Mendel's Law implies a belief in the 182 VIGOUE AND HEREDITY absolute segregation of gametes, yet it is a matter of common knowledge that in some cases {e.g. Duck hybrids) the crossing of individuals leads not to a segregation but to a blending of characters and that this " blend " is inherited. Studying the matter from the vigour point of view, can we offer any possible solution ? It must, of course, be clearly understood that any suggestion is from the nature of the case and from our present limited knowledge be purely hypothetical. On the other hand, we know the facts to exist, and if they are capable of an explanation by following up an hypothesis which we may claim to have established (partially, at all events) in many simpler cases, we are justified in trying to show how that hypothesis might explain other more complicated facts. Let us assume that X is any species living, say, in Europe, and that some of its members migrate to a climate of lower conditions for vigour and thus form a new geo- graphic form. Am; at the same time, other members migrate to a climate of higher conditions of vigour, and thus form another geographic race, Af. In course of time the actual gametes will have become affected and the distinguishing characters of Am and Af will be inherited for a generation or so, even though the vigour conditions be altered. Now let us suppose that these two races be brought together and crossed and, for the sake of argument, let us suppose that they have each deviated equally from the mean X, then the resulting initial vigour of the offspring will be approxi- mately that of X, and the resulting individual will be a blend between Af and Am resembling X, but gametically the two characters, having become altered, will not be blended but separate and distinct. Now when this offspring is mated together they will gametically be of three kinds : pure Af , mixed Af and Am, and pure Am ; while the vigour will be roughly that of X, though fluctuating doubtless to one side or the other. The characters of Am and Af are identical in kind and have only been modified by the rate of metabolism, consequently the resulting (F^) generation will show the external character according to its particular vigour, i.e. those of high vigour will tend to resemble the Af and those of low vigour the Am ; while those of inter- mediate vigour, which will be in the majority, will tend to THE EVOLUTION OF SEX 183 resemble X, the original form from which we presumed both Am and Af to have originated. Now a good example of this is the so-called reversion. If X be the wild type and Am and Af two widely dissimilar domestic varieties, we see at once why, if Am and Af be crossed, they tend to revert to X. We have assumed a case in which Af and Am had diverged equally from the mean ; now let us assume that Af has but slightly diverged, whereas Am has greatly diverged, the resultant initial vigour of the offspring would in consequence be low, and possibly too low to enable the assumed character of Af to influence the development, consequently Am would be what is known as completely dominant and the whole of the progeny (F^ gene- ration) would resemble Am ; in the next generation, how- ever, we should get the usual fluctuations, in some of which the vigour would be high enough to enable the character Af to develop. Or, again, if Am and Af, two domestic varieties, become feral, these conditions will tend to bring their vigour back to that of their original wild ancestor, and thus in a very few generations they will tend to resemble X ; that is to say, they will have " reverted " to their original ancestor. Now let B and C represent two distinct species, of approximately equal vigour. The fact that these are species, in contradistinction to geographic races, implies that most of their characters are essentially and gametically different. Consequently, when these species are crossed, the vigour being approximately the same, there will be a struggle between the respective determinants, and since they differ in kind and not merely in quality, as in the case of geo- graphic races, the resultant progeny will be a " blend " of its two parents, the determinant from each parent having had an influence on the developing embryo. Now it cannot be doubted that this unwonted development has caused an abnormal physiological strain, and as a result of this it follows that the vigour of the developing embryo becomes rapidly lowered. It also follows that the decrease in vigour will vary according to the energy expended in forming the blend. Embryos develop in certain definite sequence, hence as each fresh pair of determinants have to exercise their influence on the development of their particular character in the soma, that development will in turn be further modified by the actual vigour at that particular period. 184 VIGOUE AND HBEBDITY Let us try and follow this out to its logical conclusion and test it with known facts as regard sex. First, the vigour may be so reduced in the earliest stages of develop- ment that no offspring results. Next, development will go on for a time, but at the period of sex determination will be so reduced that neither sex determinant is able to develop, and the animal becomes a true sexless hybrid. It often happens in these cases, e. g. many Pheasant crosses, that the hybrids are dark, full and brilliant in coloration, which may probably be accounted for by the fact that, the sex organs being non-existent, the animal is able subsequently to develop a high vigour, all the products of which can go to the soma, and hence also we get the large size of many hybrids. Closely following on this line, we reach a stage when the vigour is only suflBiciently great to allow the development of males.* Next comes the stage when both sexes develop but both are infertile {e.g. Horse and Ass hybrids), or when the male alone is fertile (e. g. hybrids between Turtle and Barbary Doves), and, finally, we have the perfectly fertile hybrid, in which both sexes fulfil their normal functions (Cattle hybrids and some Duck hybrids) . This hypothesis, therefore, gives us a plausible explana- tion both of the cause of the blend in hybrids, instead of one character being dominant, and further accounts, fairly satisfactorily to our mind, for all the different gradations between sexlessness on the one hand and normal fertility on the other. This period of development with which we have been dealing is, of course, the early development of the embryos, long before the individual has been born or hatched. When discussing initial vigour (p. 166), we pointed out that under normal conditions there must be a loss of vigour due to development, and that owing to this loss the stability of the species is maintained, the initial vigour of the young being approximately the initial vigour of the parents, the periodical * This class is practically identical with the former and cannot with certainty be distinguished. A true sexless hybrid is from its nature essential katabolic, and therefore resembles a male with more intense coloration. Where the male organs tend to become functional, its appearance will be as before, but will show more of the male characters and less intense coloration. If the vigour is sufficient for a functional male some functionless females will, probably appear, and thus we reach the next class ; no hard-and-fast lines can be drawn. THE EVOLUTION OE SEX 185 yearly increase at the beginning of the breeding season tending to counterbalance the loss due to development or blending. Now, in the case of hybrids, as we have shown, this loss is greater, and consequently the initial vigour of young hybrids is less than that of their parents. It follows, therefore, that among fertile hybrids, if paired inter se, the initial vigour of each succeeding generation will become lower, till in the course of a very few generations we reach infertility. I have dealt at length with the details pointing to these conclusions from my Duck experiments, so need not enter into the matter here, except to note that in each succeeding generation the colours of the male became more and more dull, till in the third or fourth generation his full plumage resembles the eclipse plumage of the first genera- tion. We have noted elsewhere that the eclipse plumage is a sign of weak vigour, so that we have here a further fact pointing to the infertility of these former fertile hybrids as being due to loss of vigour. Careful distinction must be made between these degene- rate fertile hybrids and the true sexless one ; in the latter case, the sexual organs, not having developed, make no demand on the soma which, therefore, becomes entirely katabolic and deeply coloured ; in the former the sex organs would be functional, but the vigour is so low that it cannot reach the breeding level. We may thus sum up briefly the difference between crosses of di£ferent races and hybrids between different species. The characters of different races do not differ in kind, but they are the same characters altered by the rate of vigour at which they developed, whereas in many cases the characters of different species differ in kind, and if the species crossed be of approximately equal vigour, the de- terminant of each pair of characters will exert its influence on the development of their own character in the soma and a blend will result. 186 VIGOUR AND HEREDITY CHAPTER XII. PSYCHOLOGY OF EEPRODUCTION. We have dealt at some length (p. 158) with the causes that bring about a breeding condition, and we have tried to show how, when once an animal has been brought up to a certain physiological condition, it is bound to reduce its reserve material by natural outlets, of which reproduction is the chief. In the higher animals, however, another factor comes into play, which must not be disregarded, namely, the nervous or psychological stimulation. The actual act of copulation is brought about by vigorous nervous stimulation and expenditure of nervous energy and, per contra, the absence of such stimulation may tend to postpone or even put off the breeding season altogether. With each race, species or order, the actual nature of the stimulus is different, though in all cases a stimulus is required. Among the mammals, where the breeding arrangements are much less complicated than among birds, the " stimulus " is more definite and more easily traceable. The female, owing to superabundance of "reserve" material, comes "on heat," and this state forms the "stimulus" necessary to excite the male, while the act itself is the stimulus necessary to bring into use the accessory details, such as the preparation of the uterine wall for the attach- ment of the placenta, &c. Should, however, no male be met with, the anabolic surplus is disposed of and a fresh accumulation takes place, followed by a fresh oestrus period in due course. If, however, young are formed, they by their development create the necessary stimuli which divert the anabolic surplus to their own needs. In regard to the males, their anabolic surplus (sperma- tozoa) is much less and their power of storing the same much greater ; they have so many other outlets (e.g. activity, song, fighting, pelage) for their surplus that in the absence PSYCHOLOGY OF EBPEODUCTION 187 of stimulation this surplus increases but slowly, or may even cease altogether. We have hitherto for simplicity's sake vyritten as though the majority of anabolic surplus vpas stored directly in the generative organs; this of course is not so, the greater part remains in the soma (probably chiefly as fat, glycogen and other reserve products), and it is on stimulation that the sexual organs requiring more nourishment consume much, both of these products and of the fresh nutrition that is eaten. Physiologically, therefore, these stimuli may be regarded as locally increasing the vigour of more or less quiescent parts of the body — in this case, the generative organs. In the case of birds the matter is similar, but the stimuli are different. Birds do not come " on heat " in the same way as mammals, although from close observation we are inclined to think that "heat " in some species, at all events, is not entirely absent. As bearing this out, we may mention that, if several Ducks and Drakes be kept together in a confined space during the breeding season, certain individual Ducks will sometimes be persecuted by all the males, although they may nominally be specially paired to a particular Drake and though the other males may also have their own particular Duck. The result of this perse- cution, unless the Duck be removed, invariably ends in death, as she will never for a single moment be left alone.* The stimulus derived from " heat " in birds may, however, as compared with other stimuli be practically neglected. We are not taking as of much account, from this point of view, the possible stimuli derived from the courting displays of most birds, or their song, for these are indulged in when no females are present, and are probably merely indications of a high vigour, though it is not to be denied that they may also act as further stimulations at the right season, nor do we include the nervous twitchings and actions that immediately precede actual pairing, such, for instance, as the habit of Pigeons to give sharp pecks ^ It is on lines similar to this that certain naturalists advocate that the best and surest method for reducing the numbers of a "pest" is to destroy only the females «s, owing to the artificial surplus of males so created, the remaining females are not able to breed. This plan has, I believe, been practically demonstrated with moderate success on Babbits in some parts of Australia, and Sparrows on an estate in Germany. 188 VIGOUR AND HBEEDITY behind their wings ; these mcuy partially act as direct stimuli to the act itself, more probably they are meaningless actions due to the intense nervous energy of the moment. Birds are so much more dependent on the conditions of their environment for the feeding of their young and the successful building of their nests or concealment of the same than mammals, that it is only to be expected that the most powerful breeding " stimuli " should be those supplied by the environment. That the vigour may be high and the birds in breeding form without actually breeding, and that nesting is directly brought about by suitable environmental conditions is, I think, clearly manifest from the following facts. Flamingos have from their peculiar breeding habits long attracted the attention of naturalists. The Flamingo, as is well known, builds a large conical nest some twelve to eighteen inches high, in or near the edge of the water, and the locality chosen is usually a large brackish lagoon, where owing to the tide and rain, fluctua- tions in the depth of the water are considerable. Various colonies have been visited and described by different ornithologists, some of whom found the nests in the water ; others found them on hard, sun-baked mud, and most of them agree in noting that the date of laying varies con- siderably in different years. In 1902 I had the good fortune to visit a nesting colony in the Bahamas, and there found the birds just commencing to build on the soft damp mud at the edge of a lagoon. The mud used for the nests was collected from the imruediate vicinity of each nest, hence it is obvious that when the birds are actually building the mud must be soft ; on the other hand, they cannot be built in the water, since in that case the mud would never harden sufficiently to form a firm foundation, and consequently the requisite for nest building is soft mud at or near the edge of a lagoon. The site chosen must be above the usual fluctuations of tide, as otherwise each successive rise would wash away or soften the foundations ; consequently, I believe that in many localities, at all events, the first heavy rains at the right season, which are sufficient to soften the mud, form the desired stimulus to bring about the nest building and, consequently, breeding season. Mr. Frank M. Chapman, PSYCHOLOGY OF EBPEODUGTION 189 of New York, who has had considerable opportunities of studying these birds during the nesting, and with whom I made a joint expedition, formerly agreed with the view I have just stated. But on a subsequent visit he changed his mind, for he writes : " Since the rainy season is not inaugurated before May 16th, and these birds must evidently begin to breed late in April, there can be no close connection with these events." * These remarks, however, refer to one particular season, and as the "rainy." season in the Bahamas varies considerably in its time of appearance, it is more than probable that in that season there had been a heavy rainfall towards the end of April. Mr. Manniche,+ who accompanied the Danish Expedition to Greenland in 1906-8, gives a most interesting and instruc- tive account of the conditions of life at the different seasons of the year, and his remarks on the breeding season of 1907 are very much to the point in this connection. Mr. Manniche spent three summers (1906, 1907, 1908) in Greenland, and was thus able to compare the different conditions in each season. The summer of 1907 differed from those of 1906 and 1908, in the fact that ice round the shore did not break up, and in consequence there was no open water in the estuaries and bogs, which in other seasons had afforded abundant nourishment for the rearing of many species of birds such as Gulls, Terns, Eider Ducks, Long-tailed Ducks, &c. These species, as a rule, begin to nest immediately after their arrival from their southern winter quarters, but in this particular year, however, Mr. Manniche is of opinion that none of these species laid, and he is quite certain that no young were hatched. Species breeding on the inland pools reared their young normally, the others remained in the neighbourhood for some weeks, and then left for the south at an earlier date than usual. These facts seem to us conclusive evidence of the necessity of a stimulus of some kind, for, since the birds usually commence to nest immediately after their arrival, their vigour must be already sufficiently high and therefore * See Abel Chapman, ' Ibis,' 1884, pp. 86-88. Henke, ibid. 1882, p. 237. H. H. Johnstone, ' Ibis,' 1881, p. 174. F. M. Chapman, Bull. Am. Mns. N. H., pp. 56 et sqq. Bonhote, ' Avicnltural Magazine,' 1903, p. 87. f Meddelser om Greenland. 190 VIGOUE AND HEREDITY independent of the actual environment at their breeding place. Since also some of these species habitually feed out at sea, where the conditions were normal and there was no lack of food for themselves, and further those species not actually dependent on the open sea water for their young nested normally, therefore the failure of some species to breed cannot have been due to any general climatic con- ditions, and must therefore (all other possibilities being eliminated) have been caused by the absence of suitable open water near the shore, where they might rear their young. Now we have seen that the parents' food was in normal quantities, and since open water near the shore is not a necessity until the young are hatched, there was no reason why laying and incubation should not have pursued a normal course, which would have been subsequently followed by the death of the young when hatched ovnng to the lack of suitable environment. Yet Mr. Manniche is convinced that no eggs were laid : first, because he found no eggs ; secondly, because the birds were always to be seen in pairs, and there is little doubt that his conclusions are correct.* Therefore the evidence in this case is over- whelmingly strong that open water near the shore forms the necessary stimulus to bring about the laying of the eggs. To avoid being misunderstood, let us say at once that we are not bringing forward a teleological argument, we do not wish to imply that these birds consciously or sub- consciously realised that their young, if hatched, would perish, and therefore saved themselves the trouble of incubation. It is well known among the human race that many things, having in themselves no connection with reproduction, tend to lead to excesses, and from a similar though more legitimate reasoning, I would suggest that the presence of an environment suitable for their young gives the required stimulus which leads to the functional activity of the generative organs. A crucial case in Greenland is perhaps given by the two species of Eider Duck (Somateria inolUssima and S. spectabilis). They both rear their young * The pith of Herr Manniche's observations is summed up in his book (p. 25) as follows : " One will not be surprised to find that the birds on the here-mentioned territory appear at the normal time and in the normal number without regard to the weather and ice conditions, and when the circumstances are considered too unfavourable they fail to bred." PSYCHOLOGY OF REPEODUCTION 191 on the seashore, but the latter lays its eggs near inland pools, the other along the seashore. According to Mr. Manniche, neither species bred in that year, although the actual nesting localities of the former were entirely normal. Passing from this very striking, and, to our minds, con- clusive instance, let us take another one nearer home. The common House Sparrow is a species, now parasitic on man, resident in our own country, and never, from its omnivorous habits, short of food. We are therefore justified in concluding that it is of fairly high vigour; yet it is a late breeder, seldom hatching its young before the end of April. Nesting as it does on houses and feeding largely on man's produce, we have apparently all conditions necessary for early breed- ing, but we suggest the stimulus is lacking. What is likely to be the stimulus ? The young when first hatched are fed almost entirely on insects, whereas the adults feed largely on grain, and we would suggest that it is not until insects begin to appear that the necessary stimulus, which leads to the laying of eggs, is present. A close study of birds in captivity, whether belonging to wild or domestic species, still furthet confirms this idea. For years past I have been experimenting with hybrid Ducks. During the winter the birds are all kept together, and just previous to the nesting season are placed in adjacent pens along a stream. In some of these pens there is a certain amount of natural cover in the shape of bushes, but not until the grass has begun to grow is the cover on the ground in any way thick. In other pens there is no cover, and artificial cover is made with trusses of straw, which are much more efficacious in concealing a sitting Duck than the natural cover in other pens. In penning the birds no special selection is made as to which pairs have the straw shelters and which the natural bushes — yet every year, and almost without exception, the Ducks with the straw shelters lay from two to three weeks before the others. Bearing in mind that in all other respects the conditions are identical, we are again driven to the conclusion that a well-concealed place suitable for a " nest " forms a " stimulus " and forces on the laying. I can quote another similar case from the keeping of small Pinches; if several pairs of these birds be kept in a large aviary, no signs of pairing or nest-building or egg- 192 VIGOUE AND HEBEDITY laying would be apparent till late on in the spring, but if, at about the right season, nesting-boxes and suitable material be placed in the aviary, eggs would invariably be laid within a week or ten days. Another case comes to my mind, which is perhaps worth quoting. The male Common Sheld-Duck assumes in spring, as is well known, a large and swollen knob at the base of the upper mandible. I once had a single male that was kept by himself; in the spring in question he did not assume his " knob," appeared listless and in anything but breeding condition. Towards the end of April I got a hen and placed her in his pen. A wonderful change at once took place ; instead of being listless and silent he immediately became noisy and active : he commenced to "make up" to the hen, and in the course of a few days he had assumed a full and swollen "knob," and appeared in all respects like a bird full of health and vigour. We need not, I think, pursue this point further, for it is one that concerns practical breeding rather than the in- heritance of characters, and enough evidence has already been brought forward to prove its existence. It has, how- ever, a considerable bearing on the main theme. Eepro- duction is a natural result of high vigour, and cannot other- wise take place, but, as Messrs. Geddes and Thompson point out, reproduction itself is actually a katabolic crisis, and cannot therefore take place without a previous period of anabolism or the storing up of reserve material. This reserve is probably not stored up in the generative organs themselves; they are lying quiescent, but that reserve having been accumulated and stored in some other part of the soma, all that is required is the stimulus. The genera- tive organs then become vigorously active, absorb the greater part of the daily nutrition, as well as the reserve nutrition, and thus bring about the katabolic crisis which leads to reproduction and the formation of a new organism. We have previously referred to the song of birds as one of Nature's outlets for superabundant vigour, but as the song period in most species reaches its maximum during courtship, that is to say, during the period immediately preceding breeding, it may be as well to consider shortly the facts in some detail, with a view to showing that song is primarily an outlet for vigour, i. e. coincides with those PSYCHOLOGY OE EBPEODUOTION 193 times when the conditions are favourable to a high vigour, and has thus probably only become a courting measure as a secondary consideration. Although we have ourselves noted many of the facts, we would refer our readers who may wish to look up actual details to a paper by H. G. and C. Alexander, which appeared in 'British Birds,' vol. i. p. 367, vol. iv. p. 274, and in which the various months when some of the commoner birds may be heard singing is carefully given. Let us consider the cause of song. It is usually accepted as a form of mating and enticement to attract the female. To a certain extent this is undoubtedly true, but, on the other hand, it is often undertaken during the autumn and winter when there is no question of mating or of attracting the females, and as the nesting season is usually not longer than two or three months, and the song period of many species extends for at least twice as long, it is evident that in the first instance song was not brought about as a mating measure, but only secondarily evolved and perfected, to be used for that purpose. Without at present going into tedious detail on this subject, we may note that, almost without exception, the best songsters are dull plumaged birds, and in fact I know of no instance of a really bright plumaged bird having any song worthy of the name. This at once makes us suspect that " song" may, in the first instance, be an expression of high vigour. Looked at from this point of view, let us see if the causes which tend to produce a rise in or excess of vigour have any correlation with the song period. The period of highest vigour is, as we know, in spring ; the returning warmth, increasing food and longer days all combine to heighten the vigour. This may be said to be the beginning of the chief song season ; at such time most birds sing or show increasing vigour in other ways. The conditions for continual increase of vigour still persist, and we then have the breeding season, which forms a culminating point of high vigour. The song would probably still con- tinue, and, if our theory be correct, should still do so ; but other causes supervene, which, by affording other outlets for vigour, reduce the ardency of the song ; thus we find that, on the hatching of the young, the song ceases, the excess of vigour being expended in finding nourishment for 194 VIGOUE AND HEREDITY the family. As soon as the first family is fledged, however, the song is again partially renewed, till it dies down during the cares of the last family. By this time also the system has become weakened by the strain of breeding, and the vigour reduced below song level. Then in most cases follows the moult, owing to which the song level of vigour is not reached for some weeks. But we are still in early autumn; conditions were never more favourable for a high state of vigour, and in correlation with this we again find many, if not most, species again starting to sing. In exceptionally mild seasons some of the hardier species even start nesting cares, but usually the shortening days, increasing cold, and growing scarcity of food slowly reduce the vigour, and the song ceases, only to recommence again on the first fine and warih days of the succeeding spring. If we look at these facts, therefore, we see that the song is practically continuous throughout the year, as long as conditions are favourable to a high state of vigour, and that it only ceases during such conditions when the individuals are employed in pursuits which afford other outlets for their superabundant energy. So far, therefore, there are good reasons for believing that song birds always sing when the conditions for vigour are high, and, if further proof were required, we might instance cases of wild species in captivity, whose song period under such circumstances often becomes practically con- tinuous throughout the year, and this same reasoning will also explain why most species sing better in small cages (where exercise and other outlets for vigour are non-existent) than in large aviaries. We have stated that almost without exception all song birds (properly so called) were dull in plumage, and that birds with brilliant plumage had usually no power of song. Between these two classes, however, there are many grada- tions : the majority of birds are neither exceptionally brilliant nor remarkably dull, and among these we find many gradations of song. The commonest form of song consists of very frequent repetitions of the call-note, often uttered with slight differences of intonation. This is the style of " song" found among Owls and birds of prey. As might be expected, where the "song" is so primitive it is but seldom used, and in fact is only heard at the period of PSYCHOLOGY OP EEPEODUCTlON 195 highest vigour, namely, that period immediately preceding the breeding season, the first attempts at breeding sufficing to reduce the vigour below the " song level " for that particular species. In the large group of waders, song is nearly though not entirely absent. Many of these birds assume bright breeding plumages, which therefore take the place of song ; others, like Snipe, indulge in peculiar flight, by that means expending their surplus vigour, while others again, like the Dunlin and Stint, have a peculiar flight and indulge at the same time in a musical little trill. Gulls and Terns, though having no song proper, indulge in a series of cries and noises which are continually heard during the breeding season, and in these birds that live under fairly equable conditions throughout the year the " song," as we should therefore expect, may be heard during almost any month of the twelve. We have thus made a long step from the typical song birds to those whose "song" is only represented by a few raucous cries, heard practically throughout the year. If, however, we regard song as primarily an outlet for vigour, and only therefore to be heard at periods when circumstances are favourable to moderately high vigour, we shall thus see how, under these conditions and in default of other outlets (such as brilliancy of plumage or excessive^exercise) , most, if not all, species will give vent to a series of vocal utterances which will vary from the sharp piercing cry of the Hawk to the finely modulated and varied song of the Nightingale, but all proceeding and prompted by the need for an outlet for high vigour. It is perhaps superfluous to add that this idea in no way affects the question as to the high evolution of the Nightin- gale's song having been perfected owing to its use at mating time. To argue or express an opinion on this matter would be out of place here, as our present object is merely to show that song was not originated as a courting measure, and does not owe its existence to any special impulse, but owing to its being an outlet for, or expression of, superabundant vigour. o 2 196 VIGOUR AND HBEEDITY CHAPTEE XIII. THE MENDELIAN HYPOTHESIS. The theory of Mendel is now too well known and has become too complicated for us to deal with it in full within the limits of this chapter, but, as we have alreadj' shown, many practical experiments conducted by others and our- selves have offered difficulties in the way of a complete acceptance of the entire hypothesis as enunciated at the present day by its supporters. We will, therefore, attempt to show in a few words how these contradictions, which are continually occurring in the course of practical work, are susceptible of an explanation on the lines of our hypothesis. The Mendelian hypothesis in its simplest form supposes that all individuals are made up gametically of a number of paired characters known as " allelomorphs," one of which is derived from each parent. When the character derived from each parent is the same, the individual is known as " homozygous " for that character and can transmit that character only to his progeny; if the characters derived from each parent are different, the individual is said to be " heterozygous," and may transmit either of the pair {hut not both) to each individual descendant. So far we have been only dealing with the " gametical " aspect of Mendelism, which is its true foundation stone. The external appearance or somatic development of an individual does not necessarily show its gametic properties. If an indi- vidual be homozygous, its somatic development can naturally only exhibit the development brought about by that particular character in the gamete, but if the individual be heterozygous, its somatic development may show either the development of one of the characters only, to the complete exclusion of the other, or a blend between the two characters (which blend, however, is not inherited), or an appearance different from that produced by either of the two characters were either alone contained in the gametes {e.g. Andalucian Fowl). THE MBNDBLIAN HYPOTHESIS 197 To put it in other words: if A and B are a pair of allelomorphic characters, and if an individual homozygous for " A " be crossed with an individual homozygous for " B," all their progeny will be gametically heterozygous for "A" and "B." Somatically, let us assume that heterozygous individuals resemble "A" — then if homozygous individuals for "A" and " B " respectively be mated, the progeny will be heterozygous, but all will be externally indistinguishable from their homozygous parent " A." The character " A " is then said to be dominant to the character "B." This, then, is the explanation of the basis of the whole of the Mendelian hypothesis, but when we come to test it practi- cally we find that, although many cases bear it out, yet in a large number of experiments the results do not fit in ; these results are generally explained away, unsatisfactorily to our way of thinking, by those who believe this hypothesis to be the " be all and end all " of heredity, but in the works of others who, without being full-blooded Mendelians, are nevertheless carrying on experiments on Mendelian lines, one frequently meets with cases in which extracted recessives which are to all appearances pure, that is to say, which show externally none of the dominant characters, throw, nevertheless, a certain proportion of dominants. As in- stances of this which have come under my notice, I may mention Webfooted Pigeons, Silky Fowls, Horned Goats, &c. If a Mendelian be approached on the matter, his invari- able and obvious answer (if his theory of the purity of the gametes is to hold good) is that the apparent recessive is in reality a heterozygote, in which the preponderance of the dominant character externally has been entirely obliterated. This is, of course, a feasible suggestion, but one that is not by any means absolutely convincing, for at its best it is an assumption and lacks definite proof. Dr. Archdall Eeid, whose criticism of Mendelism is largely based on this point, suggests that, as characters tend to appear in alternative generations, it is a case of alternative reproduction and not alternative inheritance, and that every animal is in reality a heterozygote, with, as a rule, one factor latent and one patent. This hypothesis, which will partially account for the 198 VIGOUE AND HEEEDITY appearance of dominant factors from recessives, and to which we shall refer again in a subsequent chapter, we cannot accept as it stands, since to our mind the number of pure recessives which support the Mendelian hypothesis is overwhelming, though that hypothesis itself obviously needs modification to include the comparatively few exceptions. Let us then consider this matter from a " vigour " standpoint, and see if vigour will help us to explain the actual facts. To commence with, I propose to take a partially hypo- thetical case, but one which is only hypothetical, because I personally have not met with all the facts within the limits of a single set of experiments, though I have been told that they exist. The case is that of the silver colour in Pigeons, which I have elsewhere shown to be a recessive to blue, being in reality dilute blue. On page 117, when dealing with the sex of Pigeons, we found that in the crossing of colours all the Silvers heterozygous for colour were females, a fact which Mendelians would suggest tended to show an affinity of the Silver (dilution) character for femaleness ; when, however. Silvers are mated together (males, of course, sometimes appear) we find this excess of females reduced to the normal, i.e. equality. Now if this dilute or recessive character has, as the Mendelians would say, a repugnance for maleness and an af&nity for femaleness, we ought always to get a much higher proportion of females, and their assumption, therefore, breaks down " at the first fence," so to speak. Now our hypothesis is this : namely, that Silvers are usually females, because the recessive character shows itself in the soma under a certain state of vigour, and since femaleness absorbs more of the products of that vigour than a male, there is in consequence only a reduced amount of vigour among females to nourish the other "factors," of which colour is one, which go to form a zygote — hence, given heterozygous males and females of equal vigour, the males will be of the dominant or stronger colour, and the females of the weaker or recessive.* We must always remember, and this is the impor- * See note on Dun males, Chapter IX. pp. 110 et sqq. THE MENDBLIAN HYPOTHESIS 199 tance of the Mendelian theory, that within certain limits the characters of an animal are governed by the consti- tution of its gametes; thus, for example, a really pure blue Pigeon could not as a rule throw Silver females, owing to the absence of the dilution factor. (There may be exceptions to this.) The females of low vigour produced by such birds would either: (1) be poor coloured Blues or (2) would be born weak and die; or, and this is most likely, (3) they would be born males — that is to say, we should have a large proportion of males produced. Let us take the initial vigour of a Silver female at 70, and the minimum initial vigour of a Blue female at 75, and for a male 65 and 70 respectively ; then it follows that any female whose vigour lies between 76 and 70 will be Silver, and any male whose vigour lies between 65 and 70 will be Silver. Now we started with the fact that Silvers from heterozygotes are usually females. This becomes at once obvious. If the male be Silver and the female Blue, their minimum vigour respectively must be 65 and 75, which gives their offspring a minimum initial vigour of 70 ; and if the male be Blue and the female Silver, their minimum vigour by the same reasoning must be 70, so that whichever way the cross be made the minimum vigour of the offspring will be 70, and they would therefore be bound to produce young with an initial vigour between 70 and 75, which gives us what we get, namely. Blue males and Silver females, with possibly a Silver male as a rare exception when the vigour of parents is very low. This being a partially hypothetical case, we have presumed both males and females to be at their minimum vigour. In practice, of course, we have to take into account the fluctuating vigour caused by environ- ment, breeding trim, respective ages of the parents, &c., so that in actual breeding we very rarely get Silver males, and more frequently Blue females. But the large preponderance of Silver females always remains to emphasise the truth of our statement. Although this case deals only with a single colour character, it is obviously capable of a much wider interpretation, as the fluctuating levels of vigour would of course affect the other characters as well. Now let us carry the same argument a stage further, and presume that we have two Silvers of high vigour. The highest initial vigour of the males is, according to our 200 VIGOUE AND HEEBDITY reasoning, 70, and of the females 75 ; allowing, therefore, for a maximum fluctuating increase of 5 (of course the figures are purely arbitrary ; we have as yet no unit of vigour). We then might have males of 75 and females of 80, and this would give us young with an initial vigour between 75 and 80, which in theory would all be Blues. Therefore, on this perfectly logical reasoning, we have shovsTi how a dominant colour rmght be produced from two recessives, and we suggest that this hypothesis will meet the facts of the case better than any other which has hitherto been brought forward. Although possibly in some cases the actual character of the gamete might be altered during the development of the embryo, yet after that character is once fixed in the soma no subsequent rise of vigour can affect it, and it is thus quite possible to have adult Silvers of a high vigour — as high or higher than that of adult Blues. Our figures, of course, refer only to initial vigour, which is the only vigour that could possibly fix or alter the colour or sex as expected under the Mendelian hypothesis. We have chosen the question of Silvers ; first, because it helps to elucidate the so-called " sex limitation," and, secondly, because the first two stages on which the hypo- thesis for this particular aspect of the question is founded have come under our personal observation and are founded on facts, for which my friend Mr. Smalley and myself are jointly responsible. The second and all-important part of this case is hypothetical, but, as was pointed out earlier, is well authenticated and borne out in many other cases. In one experiment (No. 159) from a Silver Chequer and pure Blue, one Silver female was produced. Such an event is im- possible according to a Mendelian interpretation, unless the Blue was heterozygous, which we have no reason to believe. There was certainly no Silver in the blood for several generations previously, but, on the other hand, everything goes to prove a lack of vigour ; first, only two young were produced in the season, and these in different nests; secondly, the first nest produced a Silver male, the second nest a Blue, which died in infancy. This case, though not conclusive, is important as showing that in a pair that was obviously weak the heterozygote took on the recessive THE MBNDELIAN HYPOTHESIS 201 colour character. Had they been a strong healthy pair producing many Blues and but a few Silvers, we might have accepted the Mendelian explanation; but under the circumstances, and considering that the pedigree of the Blue was known for several generations back, we are, I think, justified in assuming that the appearance of the Silver was due to lack of vigour. In our own experiments it is paralleled by the case of the web-footed Pigeons (P.Z.S. 1910, p. 25) where two web- footed birds of different strains, in both of which the web had previously proved to be a pure recessive, when mated together gave a certain percentage of the dominant or normal footed offspring. According to our hypothesis this appearance of the normal feet — which is analogous to the production of Blues from Silvers in our hypothetical case — should have been caused by an increase of vigour, and that this is so receives a certain amoimt of support from the fact that it originated from matings in which the parents were of different strains, a fact which, as is well known and as we have also argued elsewhere, tends to an increase of vigour. Now, according to the main hypothesis (see p. 48), maleness by consuming less vigour allows other characters to develop at a higher vigour than in a female of the same initial vigour, hence it follows that males should tend to show the dominant character more frequently than the females ; hence if, owing to an increase of vigour, we get the re-appearance of a dominant factor from pure (?) reces- sives, we should find that the larger proportion of dominants thus bred would be males. In this case the dominant is the normal foot, and out of sixteen normals reared from two recessives twelve were males, which to our mind is very substantial confirmation of the truth of our hypothesis, and per contra, out of fifteen webs only six were males. It will be apparent, therefore, from this that some of the exceptions to the expected Mendelian results may be accounted for on our vigour hypothesis. We noted above that these differences which we allege to be modified by the individual -vigour were at the same time often limited by the gametic constitution of the individual, and we also only accepted in part Dr. Eeid's hypothesis, which would (sex limitation apart) well account for the appearance of dominants from recessives. In the 202 VIGOUR AND HEREDITY majority of cases so far examined this reasoning has been borne out by the facts, but as more and more cases are investigated, other facts come to hght which seem to show that under certain circumstances definite changes may take place in the gamete, and new characters (which have an apparently Mendelian inheritance) may be formed. As instances of this, one may notice the work of Mr. W. B. Agar on 'Hereditary Variations in a Cladocera ' (Simocephalus vetulus), and Messrs. Delcourt and Guyenot on Drosophila* In each of these cases a new variation that was inherited appeared when the environment was changed, and persisted for several generations when normal conditions were re- sumed. We may be perhaps excused if, in this instance as before, we take our own experiences for an example, since we are, of course, in a better position to argue out the details. The case we are about to refer to is that of our Rats (p. 69). From normal coloured wild individuals we have raised a race of fawn ones, hitherto unknown in that species among the many thousand individuals that have been critically examined. The details of this case have been gone into fully elsewhere, so we need only briefly recapitulate the facts here, namely, in one litter, from two pure wild individuals three fawn males were produced, only one survived, and in the course of breeding proved his colour to follow a Mendelian inheritance and to act as a pure recessive to the normal colour. Further, in the course of these experiments waltzing Eats were produced — another character never found among wild individuals (although known in a domestic variety of Mice, and proved in these latter to be a Mendelian reces- sive), and we have also produced Eats without eyes or optic nerves. Now how are these facts explicable under existing theories ? Take the fawn Eats ; what has really happened is that the " character for black " is absent. How are the Mendelians to account for this ? Is it sufficient, or in any way convincing, to presume that both the parents were heterozygous ? For surely it would be a marvellous coinci- dence that two wild Eats, taken haphazard from among many caught, should happen to be those extremely rare * Proc. IV. Int. Congr. Genetics, Paria, p. 478 (1918). THE MENDBLIAN HYPOTHESIS 203 " heterozygous " individuals, even supposing such indi- viduals to exist, of which we have no proof. Or if these individuals were really heterozygous, then surely by all laws of probability it follows that such individuals must be of fairly frequent occurrence among wild Eats. This sug- gestion, however, breaks down from the fact that, were it true, fawn individuals resulting from mating of two hetero- zygous individuals must have occurred not infrequently in a wild state. Yet among the thousands of Eats of this species that have been caught and critically examined, they have never been noticed or recorded. We must, I think, therefore, in all fairness, assume that when caught these parents were in all respects normal Eats. Now let us consider their subsequent history and the conditions under which they were kept after their capture. After being kept alive in their native country, Egypt, for some months, they were placed in a small cage and taken to England — an entirely fresh environ- ment both as regards climate and food. Within a few weeks of their arrival in England they gave birth to seven young, four normal and three fawn, the whole litter being males. Straws show the direction of the wind ; the fact of this litter being all of one sex, and males, points on my hypothesis to a weakness or loss of vigour. Before they left Egypt spring had come, and the normal rise of vigour had (in April) begun to make itself felt on the sexual organs. Then came the journey, the colder climate, and the change of food, all tending to check and reduce the vigour ; but the sexual organs had already begun to respond to the normal conditions of spring, and had so far developed that a litter resulted — a litter produced at a minimum of vigour, which under ordinary conditions would probably never have pro- duced young. The first result of this mating under these abnormal physiological conditions was an entire absence of females, the next an absence in several individuals of the "character for black." Are we to suppose that owing to this change of environment the gametical cha- racter itself has been changed ? Is that character dead and entirely lost, unless it be re-created under the reverse con- ditions, or has it, as Dr. Eeid would suggest, merely lost its power of reproduction and of affecting the soma and is laying "latent"? That question cannot be answered at present, nor is it one of fundamental importance, since the 204 VIGOUE AND HEEEDITY actual results will be the same, whichever point of view be taken. It has none the less a theoretical importance, since, if Dr. Eeid's suggestion be near the mark, every animal is heterozygous, there is no gametical segregation, and the whole of the Mendelian fabric falls to the ground. On the other hand, if my vigour hypothesis holds good, it explains equally well the results which called forth Dr. Eeid's sug- gestion; and if a character can be removed by violent changes of environment at critical periods, there is no reason why, by an analogous method, characters should not be created de novo. Were Dr. Eeid's suggestions correct, we should expect to find the " reproduction " of latent characters far more frequently than is the case, since it calls for no special changes to bring about the "patency" of the hidden character, and therefore it would follow, in our opinion, that the permanence of species and races would be far less stable than experience proves it to be. The other peculiarities that appeared in our strain, such as waltzing movements or blindness, involve the abnormal development or entire suppression of the somatic characters, and if these, as in the case of the fawn colour, were to show a complete Mendelian inheritance, we can come to but one of two conclusions — either the gametes may be affected by physiological changes brought about by changes of environ- ment, or the foundation of the Mendelian hypothesis is false. Of these we prefer the former. The appearance of these " sports," and in fact of the whole range of "mutations," whose inheritance is presumed nowadays to have always a Mendelian inheritance, leads us to the question of " reversion." It is well known that domestic animals, when they become feral, tend to lose many of the special features that have been brought about by domestication, and to take on to a greater or less extent the characters of the original wild ancestors from which they sprung.* These facts we need not dilate upon here, as it has been fully and completely expounded by Darwin (' Animals and Plants ') . Nevertheless, in the light of our new hypo- thesis the question must inevitably arise as to whether this recrudescence of ancestral characters is really (as Dr. Eeid * See also Chapter XI. p. 183. THE MENDELIAN HYPOTHESIS 205 would have us believe) a " patency " of characters that have for untold generations been latent, or a fresh development of new characters brought about by environment, and which resemble those of previous ancestors. In the chapter on our Duck experiments (p. 140) we have given some argu- ments that, to our mind, point to the latter hypothesis as being the true one — an idea which is still further confirmed in many of our Pigeon experiments. In these we have made a large number of matings, crossing both different breeds and various colours; yet we have not found any special tendency to " revert " to the original type or colour of the Wild Eock Pigeon, from which the domestic races have been derived. Without going further into the general question of " reversion " among Pigeons, let us, however, consider shortly one particular character, viz. chequering, which occasionally appears "as a mutation " among truly wild Pigeons, and which therefore presents features of some considerable interest. In a paper on Pigeon inheritance by Mr. Smalley and myself (P.Z.S. 1911. pp. 605-607), we dealt with the in- heritance of this character, and showed that when birds known to be heterozygous for this character were mated together, the chequering behaved as a Mendelian dominant. Mr. Staples Browne came apparently to the same con- clusion (Journ. Genetics, vol. ii. No. 2, p. 131, 1912). As against this we have records of some seventy-five chequer to chequer matings in which no self-coloured birds appeared, thus pointing to the supposition that all the parents were homozygous. On the other hand, since the Wild Eock is the ancestor of all domestic Pigeons, chequer- ing must at some period have arisen from unchequered birds; that is to say (if we accept the Mendelian inter- pretation), a recessive must have been carrying a dominant, and in the paper quoted above we mention a somewhat unsatisfactory instance in which a fully chequered bird was produced from a pair of undoubted " Wild Eocks." Since then we have, in the course of our experiments, bred quite a number of chequered birds from self-coloured (un- chequered) parents, so that the whole question of chequer- ing will need to be revised. In our former paper we noted, " although the indi- vidual matings show a very close approximation to the 206 VIGOUE AND HEEEDITY expected (Mendelian) results, yet at the same tirae the ten- dency to vary from the anticipated results is all in one direction, viz. to a greater number of chequered birds. . . . "It is possibly due as much to this tendency as to the unconscious selection by breeders that this character has become perfectly true and stable in some strains ; as pointed out in the earlier portion of the paper, we are restricting our remarks for the present to the consideration of the Mendelian inheritance of certain characters ; and that theory seems to fit in well with the main lines of inheritance as borne out by the facts. None the less, it is equally evident that there are other factors at work which are able to modify to some extent the results anticipated by the Mendelian hypothesis." There must undoubtedly be some factor, or factors, which override the strict Mendelian inheritance and cause the tendency towards an over-production of chequers or the production of chequers from self-coloured recessive birds. What these factors are, how caused, and to what extent they may influence the purity of the gametes, are questions for the future. If we suppose that some natural species have arisen from mutations on the lines of inheritance laid down by Mendel, and on those lines alone, we must also allow that, in some cases at all events, the dominant cha- racter has been the one that, proving the fittest, has sur- vived, and has now been established as a certain species. Natural selection can only act on the external {i.e. somatic) characters ; therefore gametically impure dominants, as well as pure dominants, will be selected for survival by Nature, and the recessives alone will be killed off. On the most modest calculation, therefore, at least half the individuals of the species should consist of impure dominants ; but if this be so, how is it that natural species breed so true ? Why, as the chances are very nearly equal, do we not get a proportion of the recessives continually appearing ? No one will deny that wild animals breed excessively true to type, so we may well assume as an axiom that the reces- sive character, whatever it may be, does not show itself. How then — assuming the Mendelian hypothesis to be the sole theory of inheritance — are we to account for the facts, except on the supposition that when a mutation which THE MENDELIAN HYPOTHESIS 207 was favourable occurred, it was always a recessive ? This supposition, however, cannot be held for a minute, and hence we are compelled to try and find some other factor dominating inheritance. A careful study of these chequered Pigeons seems to show that, without upsetting the basis of Mendelian inheritance, there may be a tendency on the part of some characters to assert themselves, and to be inherited in greater numbers and to breed more true than their opposite allelomorphs. This, then, seems to point to a factor in evolution in which natural selection has no part. Natural selection may eliminate the recessives, but between pure and impure dominants she cannot distinguish, yet gradually the impure are got rid of, and a race that breeds pure and true results. Our suggestion on this point, as we have already stated, is that although the gametic character is usually fixed and obeys the Mendelian law, yet its " somatic expression " is controlled by the particular vigour at which it 4evelops ; and, further than this, in the case of heterozygote indi- viduals the " somatic expression " will usually be that of the character which develops at the particular degree of vigour which is possessed by the individual. Thus, if a and 6 be a pair of characters of which a develops at a vigour of " x" and h at " ic+lO," then if the vigour of the heterozygote be "a;+2," the character "a" will influence the "somatic expression" if at " x+S" the character "b" will be dominant. Of course, if the individual be homozygous for b, and have a vigour of " x+2," we should probably get a "somatic expression " of h tending to resemble " a." It has been suggested that the so-called " pairs " of Mendelian characters are in reality not two "separate" characters, but only different degrees of the same character; and this suggestion is in part borne out by the work of the Mendelians themselves, who in a large number of cases con- •sider as a " pair " the " presence " or " absence " of a certain character. We are inclined to accept this suggestion, and would consider in such a case that the difference between a " pair " of characters is really a difference of " somatic expression " of one character developing at different vigour levels. This point, however, is not one of vital importance at the present moment, as our main argument is merely to 208 VIGOUE AND HBEEDITY suggest that the " somatic expression " of a Mendelian character may be largely influenced by vigour. This line of argument leads to a further point. It has been questioned how natural selection has been able to grade matters down to so fine a point that even the least shade of colour or marking has become fixed — i.e. all the individuals produced in any one place are of exactly the same shade (see Chapter IV.). If we once allow the tendency as shown above, namely, to an over-production of the dominant or recessive character, as expected under the Mendelian hypothesis — and there is much to be said in favour of it — then surely we can, at the same time, allow a tendency to a particular shade or mark- ing. Natural selection may rough-hew the species, but this tendency to a certain colour, pattern, or marking will fix the finer characters and keep the species from too much variation, which might in turn lead to the extermination of too many individuals, and thus lead to the destruction of the species. Domestic races prove most difficult to breed to type. In all the higher fancy breeds it is found almost impossible to get anything like a large proportion of the progeny to breed true in their finer points, and therefore fit for the show bench ; in other words, man just fails where Nature succeeds. This is sometimes accounted for owing to the absence of natural selection ; but artificial selection by breeders must be almost, if not quite, as strong a factor, and yet even in the oldest and most carefully selected strains the proportion of the progeny true to type is excessively small ; and, as Bateson has pointed out, only about four per cent, of the racehorses bred are really good. We must therefore come to the conclusion : — (1) That natural selection of itself is not sufficient to have brought about the minute differences which are in- herited absolutely true among wild species. (2) That artificial selection has similarly failed to fix any particular type so that minute differences are inherited pure. Now what does this show? First, it is intelligible that natural selection could not act with such exactitude as to control these small variations, which are none the less absolutely fixed in Nature. This, therefore, points to the need of some other factor, which, acting independently, has power to modify and fix variations. THE MBNDBLIAN HYPOTHESIS 209 Similarly, artificial selection, in which the most minute differences are carefully studied and selected, has failed to fix these differences, and so we are led to the conclusions: — (i) That even if natural selection did act on the minutest of variations (which we do not allow), it would still be in- capable ot fixing them. (ii) That whatever the factor is that fixes these varia- tions in Nature, it is either absent or has lost its power in domestic breeds. What is this factor, this tendency ? Does it exist ? Do we know what forces it obeys? If we want to find out about this factor, let us consider the main differences between wild and domestic animals. The chief difference that first strikes us is the almost entire absence of natural selection among the domesticated breeds ; this undoubtedly at first allowed varieties, which would otherwise have been destroyed, to exist ; or, in other words, under new conditions natural selection, by ceasing to act, allows potentially new species to arise, then by arti- ficial selection these rough varieties are perpetuated. But though artificial selection is able to carry on the work and take the place of natural selection, it is unable to fix these varieties as they are fixed in Nature. Artificial selection, consequently, has replaced natural selection, and it cannot, therefore, be the absence of natural selection that causes the fluctuating qualities of domestic breeds. In what other respects, then, do domestic breeds differ from their wild prototype ? In one word : " environment." Exercise, shelter, food, enemies are all markedly different for the domestic as com- pared vfith the wild animal. Now, as we have already seen, food, exercise, and shelter (which make up the differences in environments) are all agents which directly affect the meta- bolism (i. e. the vigour) of an individual, so that we may well ask ourselves whether vigour is not the factor which controls and fixes the finer points. Suppose; for argument, that vigour is the factor, then, as I have tried to show elsewhere, the vigour of the animal will be reflected to a greater or less extent in the vigour of the gamete, and therefore in the vigour of any determinant in that gamete. Now, if a favourable and dominant muta- 210 VIGOUE AND HEREDITY tion arises, under natural selection the race will gradually come to consist of dominants and impure dominants (hetero- zygotes). The vigour produced by the environment will modify and fix the exact type (shape, colour, &c.). That is to say, the fluctuating varieties of the dominant will all tend to a certain fixed type, in the same way the impure domi- nants will all tend to reach the same standard, and thus the race will outwardly tend towards a particular and exact shade or type. More than this, evidence given elsewhere tends to show that the somatic appearance of a heterozygote for any character may sometimes be determined by the vigour; and consequently, if the environment be one adapted to the vigour of the dominant mutation, fewer recessives will be produced, so that here also we have a further strong tendency towards the exact type as determined by vigour. Further, as the generations go on, this particular state of vigour will have begun to make its mark on the germ plasm, so that this exact variation will tend to be innately inherited. In course of time, therefore, we find not only nearly all the impure dominants eliminated by the Mendelian Law and the continued destruction of the recessives, but the fluctuating power of the dominant reduced to a fixed type by the ever-present vigour, and also that same type will tend to be inherited, owing to the prolonged action of a particular degree of vigour on the generative organs. Con- versely, if the environment favour the vigour of a recessive mutation, more recessives will tend to be produced or even the recessive may become the dominant — i. e. the somatic expression of the heterozygous individual may become that of the normally recessive character. The result of this is that not only are all the characters of any species in a particular environment fixed to an exact degree, but if the environment be changed they will still for a few generations continue to inherit that expct shade; for although the vigour may have altered owing to the change of environ- ment, it has not had time to impress its alteration on the germ plasm. This will account, therefore, both for the fixity of natural species as well as for the occurrence of fixed geographical races. Why, then, are domestic races not similarly fixed ? The answer is, probably, chiefly dependent on time. Compared with the most recent of species, the oldest domestic race is THE MENDELIAN HYPOTHESIS 211 a comparatively new product, apart from which, its environ- ment is by no means as stable as a natural one. Every breeder has his own ideas as to the best food and methods of keeping his stock, and the stock is always changing hands and thus passing into a slightly different environment quite sufficient, in my opinion, to prevent any definite standard of vigour becoming fixed ; the same might be said about breeding. Some breeders breed from their animals at an This photograph shows in a striking manner tlie effect of vigour on colour. The feathers are taken from a young Dun Pigeon ; when the feathers had just broken the quills this bird was deserted by its parents and found stone-cold in the nest. It was warmed back to life and hand-fed for two or three days, and then given to a foster mother, who also deserted it, and for about ten days it hovered between life and death, being often found stone-cold, until finally it was successfully reared by another pair. As will be seen in the photograph, each feather is crossed by an unpigmented bar, representing the starvation period in its growth, when its metabolism was often so low that it was stone-cold and unconscious. early age; others, in the case of birds (poultry, &c., for instance), prolong the breeding season; in short, the con- p 2 212 VIGOUE AND HEEEDITY ditions under which domestic stock are kept are as varied as the ideas of the men who keep them ; and even the hfe of one man is as nothing compared with the Hfe of a species, so that not even in the most favourable cases are the con- ditions both stable and long enough to fix the vigour definitely. We have, therefore, I submit, good grounds for believing that vigour is the factor influencing to a large extent the normal Mendelian inheritance, and having applied it to existing facts we see both how it may have led to the fixity of wild species on the one hand, and to the variability of domestic races on the other. Before closing this chapter, let us return once more to the chequered Pigeon. In this case we saw how in a certain strain almost all the birds which had been merely artificially selected had become pure, i. e. the impure dominants had been apparently eliminated, but by dealing with birds known to be impure dominants we found, although Mendel's Law was borne out in the main, there was yet a tendency to a larger proportion of Chequers, due, as we may now allow, to the particular state of vigour in which the birds were kept. The appearance of chequering from self-coloured birds may possibly in the same way be ascribed to a change of vigour. Let us return to the Wild Blue Bock. Chequering, probably, must have arisen as a mutation, but, be that as it may, our wild English stock is continually being added to by feral birds bearing the chequer character. This we have seen to be dominant, and since natural selection can hardly be acting so as to eliminate the chequered birds, why do not chequered birds predominate round our coasts? Here, again, vigour gives us a plausible reason. The wild bird is unchequered ; chequering is, therefore, among feral birds being largely eliminated, owing to the fact that the normal vigour of wild birds is such as to make the usually recessive blue colour more or less dominant. To what extent this dominance may influence all individuals we do not know ; but at all events, sufficiently so as to cause the greater pro- portion of the birds to become of the recessive type. (In other words, a character recessive in wild birds becomes dominant among domestic ones.) Now when chequered feral birds, having an apparently dominant character, meet THE MENDBLIAN HYPOTHESIS 213 with wild birds having the recessive character, and under conditions of vigour favouring the recessive type, it is more than probable, from the reasons given above, that the cha- racter of the feral birds at once acts as a recessive and then becomes gradually eliminated. Thus we may begin to realise, though very dimly, the cause of the reversion of feral domestic races to the wild type : namely, because the conditions of vigour in the wild environment are those that have originally fixed the wild race, and thus, independently of natural selection and independently of their enemies, feral races return to their former appearance, because the return to their wild life and environment has brought about a return to the state of vigour which their wild progenitors enjoy. To sum up briefly, the Mendelian doctrine is based on the purity of the gametes, which are composed of certain " determinants " which are paired, and in any individual each pair is made up of a determinant from either parent. Further, it is found that usually the somatic expression follows one determinant or the other, and is not usually a blend ; but even when a blend occurs, the blend is not inherited among all the offspring but a segregation takes place, so that a certain proportion of individuals of the second generation are as pure for any particular character as though they had never been crossed. This theory further implies that the characters will be inherited in certain definite pro- portions. To this rule all practical breeders will know that there are many exceptions, and we have attempted to show that the somatic expression of these determinants can be modified according to the state of vigour of the parents at the breeding period, and that possibly, in exceptional cases, a dominant character may become recessive and vice versd. The frequent and constant deviation from the expected Mendelian proportions, which has occurred in many experi- ments, points to the fact that vigour, by controlling and regulating the outward expression of the Mendelian determi- nants, may have played no small part in fixing the finer points of shape, colour, size, &c., which are found to be so constant among wild species, and which are, at the. same time, so slight and apparently so unimportant to the exist- ence of the individual, that natural selection of itself is not sufficient to account for the uniformity. 214 VIGOUR AND HEEBDITY The uniformity of races found in any particular region, and their constant variation from those of a neighbouring district, when the Mendelian characters must be the same, may also be accounted for on the same lines; as can also the tendency of feral races to revert to the external appearance of their vyild progenitors. In short, while accepting the basis of the Mendelian doctrine, we claim to have shown that vigour plays no small part in modifying and regulating the strict Mendelian expectations. CHAPTER XIV. GALTON'S LAW. Next to the Mendelian theory the most important and widely accepted hypothesis on the laws governing heredity is that of Sir Francis Galton, which is now known as Galton's Law, and was first enunciated in 1889.* The main point of this law is the acceptance of the principle of ancestral heredity ; that is to say, that every ancestor in the direct line of any individual has had some influence on the character and qualities of that individual, the influence diminishing in geometrical ratio according to the distance of any particular ancestor from the individual. In other words, any individual inherits half his qualities from his parents, one quarter from his grandparents, one eighth from his great-grandparents, one sixteenth from his great-great-grandparents, and so on to infinity. Therefore, we see that, according to this law, if any single one of an individual's progenitors had been different, the effect of this difference would make itself felt to a greater or lesser degree on that individual. Now Mendel's theory upholds gametic segregation and the purity of the gametes ; in other words, it does not believe in the inheritance of " blends," though it has to admit their occurrence as somatic characters. Consequently, with regard to any particular character, it is from the Mendelian standpoint only necessary to know the parentage for two generations back, and the rest may be ignored ; or, in the case of characters known as reces- sives, the mere presence of these as external somatic characters is amply sufficient to establish the purity of that individual as regards that character, whatever may have been his former ancestry. With individuals showing a dominant character, how- ever, the recessive might be carried for an indefinite number of generations, and would never show itself until crossed * ' Natural Inheritance,' by Francis Galton. London, 1889. 216 VIGOUE AND HBHEDITY with another dominant carrying a recessive or with a recessive itself. In no case, however, should a " blend " be inherited, although it might frequently appear as a somatic character (c/. Blue Andalucian fowls), but the gametes will always remain pure and stable for any particular character. It will thus be seen that, as they stand at present, these two theories are mutually antagonistic, and that if either of them be universally true the other must fall. Now practical experience shows both theories to be partially true. Among breeders it is quite a common practice, if they wish to introduce into any particular variety a character possessed by an entirely different breed, to make one such cross, and then by selecting for that character and breeding out the other imported characters to produce the desired result. Two such instances (at least) have come within the range of my own experience. In conducting my experiments on webfooted Pigeons (Experiment 196) I crossed a pure bred Dragoon with an ordinary Homer. Now the Dragoon (see illustration) is a Pigeon with a very heavy development of the wattle and cere, whereas a Homer has none round the eye and but a small amount round the bill. Although a consider- able number of these birds (F^ and F'^ generation) were crossed on Mendelian lines, the large wattle and cere of the Dragoon nevef reappeared. The first F^ generation showed a "blend" much nearer, however, to the Homer than the Dragoon, and the subsequent generations gave still smaller wattles. This entirely upholds Galton's theory, and is quite contrary to the Mendelian hypothesis. The second example that I have personally come across is in my hybrid Duck experiments. In the Pintail-Mallard- Spotbill cross a blend was produced which in the main breeds true, and at any rate there was no sign of any segregation of characters.* To take one character only, viz. the coloration of the bill. The wild forms show two distinct types, viz. the " Spotbill," in which the beak is blackish with a yellow tip, * The only case of any apparent segregation was an individual of the F^ generation of Mallard, Spotbill, Australian and Pintail cross which practically, but not absolutely, resembled a pure Mallard. See P.Z.S., 1909, p. 598. Plate XI. a. Head of Dragoon. 6. Head of Homer. c. Head of F^ Generation- Dragoon x Homer. GALTON'S LAW 217 and the Pintail bill, in which the bill is greenish with a longitudinal black line dow:n the centre. In the F^ genera- tion a certain portion showed the Pintail bill, but it never reappeared in any subsequent generation. This, again, confutes Mendel and supports Galton, since, as the pure Pintail was never introduced after the first cross, its characters would natural'ly become more and more swamped with each succeeding generation. If Mendelism held true for this particular case, that character should nevertheless have appeared unimpaired amongst a few individuals of the subsequent generations. It is thus certain that in some cases. Galton's Law holds good and Mendel's is entirely refuted. But, on the other hand, there is the other aspect, namely, those cases in which Mendel's Law is borne out and 'per contra Galton's Law refuted. During the last few years so much attention has been centred on Mendel, to the almost entire disregard of other theories, that the number of cases substantiating Mendel is considerable; but studying the matter as I have been for a number of years, with an entirely open and unbiassed mind, 1 have come to the conclusion that the evidence in support of both theories is about even. I need not quote instances of pure Mendelian cases of inheritance — they are now so numerous and may be found in any text-book on the subject — ^but without denying for one moment the truth of the Mendelian theory in some instances, e.g. among my Bats,* yet many of the accepted and quoted cases need further investigation, such, for instance, as the webfoot in Pigeons, the Andalucian Fowl (where many different shades ignored by Mendelians come into play), the black pigmentation of the Silky Fowl, the chequer character in Pigeons, and many others. The only inference that can be drawn from these apparently antagonistic facts is that both laws must con- tain a substantial substratum of truth, and that what is needed is some further hypothesis which, by suggesting a slight addition or modification of these existing laws, will make them mutually complementary instead of antagonistic, without, however, destroying the essential points of either. As supplying this need we would suggest our present hypothesis. As regards its bearing on the Mendelian theory, we need not discuss it here, for we have given many * P.Z.S. 1912, p. 6. 218 VIGOUE AND HEEBDITY instances throughout this book, where we believe that modifications of that theory have been brought about by alterations of vigour. To quote three examples found in this book, we would refer our readers to our remarks on Cats, Eats, and webfooted Pigeons. Galton's Law in Teems op Vigour. Let us, therefore, now consider the relationship of vigour to Galton's Law, and take some hypothetical cases in terms of vigour. Suppose each of a pair of individuals have a vigour of 80, and each of another pair a vigour of 60, then on the lines of our reasoning elsewhere we get — 80 X 80 60 X 60 I I 80 X 60 I 70 This last individual, therefore, has a vigour half the mean of his parents and a quarter the mean of his grandparents [i.e. ^ (80 + 60) = 70, i (80 + 80 + 60 + 60) == 70]. Let us take another hypothetical case, thus — 60 X 70 50 X 60 I I 65 X 55 I 60 and we find the same result [viz. J (65 x 55) = 60, i (60 + 70 + 50 + 60) = 60]. Now, in this latter case, let us presume that the vigour of the first grandparent was 70 instead of 60, the vigour of their progeny then becomes 70, and of the individual under consideration 62'5. In other words, if the vigour of any grandparent be changed, the vigour of the offspring is altered by half that amount ; hence it follows that the initial vigour of any individual is dependent on the vigour of all his ancestry in a regularly decreasing ratio according to their remoteness from himself* * We have discussed on p. 167 the reasons for taking the initial vigour as the mean of the sum of the parents' vigour ; further investiga- tion, more especially in regard to sex-limitation and reciprocal crosses, gives us reason to think that this view may subsequently have to be amended, but whether that be so or not, any modification will not afifect the general argument but merely the ratio of the decrease, with possibly a different ratio for each sex. GALTON'S LAW 219 Now what does this actually amount to? We have already shown on our hypothesis of vigour that the vigour of any individual is dependent not merely on the vigour of his parents but on that of all his ancestors, so that, subject always to the fluctuations in vigour due to environment, if the vigour of a single ancestor of any individual had been different the initial vigour of that individual will also have been different. In other words, therefore, our hypothesis is only a restatement of Galton's'Law in more detail, and, if we may say so, standing on a firmer basis. Elsewhere we have tried to show the effect alterations of vigour may have on the gametes themselves, or, again, how such alterations of vigour may bring about alterations in the expected Mende- lian proportions, and thus we begin to see how these two great laws, which at first seemed to stand in mutual antagonism to each other, and in support of which many ^acts were forthcoming on both sides, may, if our hypo- thesis of vigour be anywhere near the truth, have both mutually assisted to the inheritance of the characters possessed by every single living thing. This matter is one of such extreme importance that we may perhaps enlarge on it a little further. The cases which are apparently entirely dominated by Galton's Law are those in which the gametic character has remained unchanged. For instance, the wattle of the Dragoon. Every Pigeon inherits the character for a small wattle similar to that found in the Wild Eock Pigeon; in certain strains an in- crease of vigour, possibly combined with a tendency (" character ") to growth on the wattle, has brought about the large wattle. By continually selecting large wattles, breeders without knowing it have always been selecting Pigeons of a vigour especially suitable to the development of that character, and so, as long as the strain is bred together, large wattles are produced. Immediately, however, that vigour is altered by an outcross with an individual whose vigour is suitable to the development of a small wattle, the wattle immediately diminishes, and nothing short of raising its vigour will again cause it to reappear. Be it remembered, also, that since every ancestor has its effect on the vigour of the progeny, a single outcross which reduced the wattle would require several generations to obliterate its effect, which would not occur, in fact, until the outcross had 220 VIGOUR AND HEEEDITY become so remote an ancestor that his vigour had become an almost negligible quantity compared with that of nearer relatives. Such cases are probably comparatively rare, since it is doubtful if any permanent change of vigour could exist for many generations vyithout also altering the character of the gametes. Mendel's Law is only (if ever) absolutely true when the vigour remains absolutely constant, and thus affects neither the gametic inheritance nor the somatic development. Personally, we have grave doubts whether such a perfect case exists for an individual as a whole. Our experience shows us that alterations of vigour are invariably reflected in some way or another, but it by no means follows that any alteration in vigour will show itself with equal force in all the characters. In some races it will show itself in size, in others in colour, in others in fertility, &c., and thus it may be quite possible that if only one character be studied at a time, that character, if it is not one of those most affected by alterations in vigour, will follow the Mendelian expectation fairly accurately. For instance, in the case of my Eats, the colour characters followed the Mendelian expectation with wonderful exactness, yet the same in- dividuals showed their degeneracy in other ways — being blind, weak, waltzers, and so forth. Thus then, to put it briefly, Galton's Law represents mainly the apparent inheritance of changes in the somatic development where the gametic characters are the same and remain unchanged. Though it must not be forgotten that sooner or later alterations in vigour are bound to produce alterations in the gametes. Mendel's Law, on the other hand, deals only with the gametes, and is only apparently true when the somatic development of a pair of characters is not unduly altered by slight alterations of vigour. Pure Galton's Law is entirely a vigour development, the gametes being the same. Pure Mendel's Law is entirely a gamete develop- ment, the vigour being constant. In actual practice all inheritance is a combination of both vigour and gametical inheritance, further complicated by changes in the vigour of the same individual due to time of year, environment, nutrition, and period (beginning or end) of breeding season. GALTON'S LAW 221 Before closing this chapter, we will try to put the matter graphically. First, let us take the simple Galton's Law case, e.g. wattle of the Dragoon; where W indicates the wattle character and the figures the index of vigour. The wattle character being simple and unchanged in every individual is found in all examples. Dragoon = W 80. Homer = W 50. W 60 X W 50 I W 55 X W 55 W 55 ad lib. never reaching W 60, the Dragoon wattle. Now let us, equally shortly and graphically, describe a pure Mendelian case, e.g. the white under parts of Eats; where W = white under parts, w dark under parts; the vigour in this case is the same. Thus we get — W 60 X w 60 I Ww60 WW 60 Ww 60 wW 60 ww 60 Let us finally consider roughly a more complicated case, but one which is of far more general, if not universal, occurrence. I propose to take the case of tortoiseshell Cats, where = orange, B = black, and, for the sake of argument, let us presume that the orange and black are of equal vigour to commence with ; let us further assume that, since each character can only develop at a certain minimum of vigour, the orange minimum is 55, and the maximum for orange is 70, while the black minimum is 45, and the maximum 60. This, it must be remembered, is the initial minimum ; when the animal is adult and in breeding condition the vigour will of necessity be above the minimum. Let us suppose that both the orange and black parents are pure {i.e. homozygous) for their respective colours, and let us suppose that their vigour at the 222 VIGOUE AND HBEBDITY time of breeding is respectively 65 and 55. Then we get — 00 65 X BB 55 I OB 60 Now femaleness (see p. 48) requires more vigour for its development, so that after femaleness has been established in the embryo the vigour available for the development of the colour determinant may lie betT7een 60 and 55, or in general terms between the common limits which allow the develop- mentof both black and orange, and hence atortoiseshellresults. In the case of maleness the vigour may not fall low enough* to come within the limit of the development of the black character, i.e. it may be slightly above 60, and, therefore, beyond the limit of the development of the black, and hence an orange male (heterozygous for black) results. The figures, of course, are quite arbitrary, and we have at present no means of determining the ratio between orange and black. Our experiments, however, show that altera- tions in vigour are able to influonce the colour results, and therefore we claim that we are justified in speaking of the facts in general terms. Let us now take the E^ generation, and we may presume that the vigour of the P^ remained at 60. This may give us four permutations, but only three combinations, as follows : — ■ OB 60 X OB 60 00 60 OB 60 BB60 we can then follow the same line of argument ; as before, those combinations which contain both and B will give us orange males and tortoiseshell females. Those which contain only may be either males or females, and also those which contain only B. * Some readers may find a difficulty in allowing for the greater lowering of vigour due to femaleness ; we have tried to explain it more fully elsewhere, but we may again mention here that "femaleness" being anabolic will absorb more nutrition, and lack of nutrition for the other characters would in this case be tantamount to a lowering of vigour, even if it did not actually lower the rate of metabolism. The " fierceness " of a fire must partly depend on the nature and a/mount of the fuel available. GALTON'S LAW 223 An obvious question, of course, arises on this line of argument, namely, what happens to those individuals that, containing only 0, have not, after the needs of femaleness have been satisfied, sufficient vigour to lead to the develop- ment of that character ? This question is one to which at present no satisfactory answer can be given. Several attempts have been made to solve similar problems : — (1) That femaleness and orange have a natural repulsion ; this can, I think, be dismissed as hardly in accord- ance with the facts. (2) That complete segregation of gametes does not exist, but that if two heterozygotes be crossed, all the progeny will be to a greater or less extent heterozygous. This suggestion -would meet the case, but involves the sacrifice of one of the foundation-stones of Mendelism. (3) That the character may be in reality the same as the character B, the external difference being due to the vigour at which it develops. Or a lower vigour may be able to change into its other allelomorph B. The possibility of loss or assump- tion of a character owing to a change of vigour has had a fair amount of evidence to back it up, as mentioned elsewhere in this book (see p. 71 et sqq.). This may, therefore, be the possible explanation. (4) The failure of to develop may lead to the death of the embryo. This is an easy solution, but it can be neither proved nor disproved, for obvious reasons. We have thought it well to bring forward this last case in spite of its rather hypothetical character and the difficul- ties it raises. Other chapters will show, we hope, clearly enough the basis on which our hypothesis rests — the alteration in the proportions of orange and black Cats raised in our experiments shows that these colours can be affected by working on alterations of vigour, and that is all we claim for the moment. Subsequent work will, we have no doubt, clear away existing difficulties. In this chapter we have merely tried to show how the Laws of Galton and Mendel, instead of being antagonistic, are in reality both at work as integral parts of the Laws of Inheritance. 224 VIGOUE AND HEEBDITY CHAPTEK XV. DISCUSSION OP OTHEE HYPOTHESES. Dk. Abchdall Eeid's Hypothesis. We have already had occasion to refer to an hypothesis brought forward by Dr. Archdall Eeid in his book the 'Laws of Heredity,' and although personally we cannot altogether accept his conclusions, yet we must acknowledge that his views are more in accordance with many of the facts than the Mendelian doctrine, and if they be correct they account for several cases in which Mendel fails. Un- fortunately, to our mind, though he makes out a good destructive case against Mendel, his constructive hypothesis is hardly as convincing. Put briefly, he believes in alternate inheritance but not in alternate reproduction, and he instances the case of sex, in which he points out that either sex has latent the potentiality of reproducing the characters of the other, but that under normal conditions the characters are inherited alternately and pure {i. e. not as a blend) . Following up this line of argument, he is driven to the conclusion that most, if not all, of the Mendelian characters are of a secondary sexual character, and have thus an alternative inheritance, but like the question of sex the other character is lying " latent," and that there is no such individual as a pure " homozygote." The Mendelian proportions are explained as merely due to the universal laws of chance. This line of argument leads up to the crux of his hypothesis, which is that every individual is " gametically " a blend between his parents, but that one of the characters is "patent" (dominant, on Mendel ideas) and the other ' ' latent ' ' (recessive) . So far he does not differ greatly in facts from the Mendelians ; he then proceeds to argue that these characters are inherited "blended," and that the latent one may remain permanently latent for ever or until a change in DISCUSSION OP OTHEB HYPOTHESES 225 conditions forms the stimulus necessary to again awaken it into " patency." . He further argues (though I cannot quite follow his reasoning) that characters that are not used (whether the absence is visible externally, as in wingless birds, or in- visible owing to latency) tend to disappear gradually or retrogress, and consequently the longer any character has been latent the greater the stimulus necessary to awaken it, or it may have been lost past recall. This awakening of latent characters and consequent tendency to reversion is, he maintains, often brought about through the retrogres- sion caused by sexual propagation. Since new characters have only existed for a comparatively short time, a very small amount of retrogression, brought about by crossing, will be sufficient to eliminate them, the stimulus of the crossing being sufficient to reawaken the "latent" character. It follows from this that one of the functions of " sex" is to lead to a swamping or "watering down" of varieties, which are too marked or differ too widely from the normal, and thus to keep the race more level. He argues that, since the conditions likely to be met with by any particular individual of a race will be those under which the majority of the race has to exist, it becomes important that no individual should differ too much from the average or normal individual. Now if reproduction was unisexual, it follows that any abnormal individual would tend to re- produce other individuals as abnormal or even more so than himself ; this. Dr. Eeid argues, would be bad for the race, and that by sexual propagation the progeny of any abnormal individual is " watered down " and thus tends to " retrogress " towards the normal, i. e. towards that character that is "more or less" latent in the "sport" If, however, the variation be a favourable one, it will be more likely to survive than a normal individual even when " watered down," and thus the race will gradually adapt itself as a whole to altered conditions, and not merely through a few of its members. It must be allowed that his somewhat loose hypothesis does to a certain extent fit in with actual facts, whereas the Mendelian theory often breaks down in practice; but it seems to me distinctly weak in direct proof; and he has, Q 226 VIGOUE AND HBEEDITY to my mind, fallen into the very trap of which he warns others, viz. he has built up an hypothesis on facts but has not tested it. His argument that evolution has naturally proceeded from mutations seems to me unsound and unproven, and equally so is the statement that almost all artificial selection has been made from mutations. Further than this, it is obvious that both blended and alternative (to use his words) inheritance occurs in domestic races {e. g. Pigeons) ; this seems to have been a matter he did not take into account, which is strange, since Galton's Law is founded on blended inheritance. Our own impression of the facts brought forward by Dr. Eeid is that vigour will easily fit in and explain the matter. Whether all or any characters are paired or not seems to us at present an open question, but in the case of " blended inheritance " we should be inclined to answer it in the negative (e. g. Ducks) . When the inheritance is blended, the matter seems most easily explainable on the supposition that there is gametically only one character, and that the different somatic shapes or forms it takes on are merely due to its vigour; hence we can easily under- stand how a " prepotent " character will become to all intents permanently " patent," and that the so-called " latent " character, which may after a thousand genera- tions make its appearance, is merely the same character developed under a different amount of vigour. We are, therefore, convinced that difference of "patency" and " latency " is in reality a difference of vigour. It follows, therefore, that if the Mendelian thesis hold good and that in some cases characters are paired, and also that the dominant or patent character is dominant because of its vigour, then if the vigour of the next generation be altered by sexual intercourse to a level at which the character, which was latent in the parents, can develop — then that (recessive) character may become dominant {i. e. patent) , and vice versd the character previously patent will become latent. This will explain one of the difficulties to the complete acceptance of Mendelism, which is of fairly fre- quent occurrence in practical breeding. One point must, however, be carefully noted. In paired characters (if such exist) the dominance alone is due to DISCUSSION 0¥ OTHBE HYPOTHESES 227 vigour and not the character itself, which is transmitted unaltered whether it be dominant or recessive. In single characters, in which blending must always occur in the progeny, the somatic development is entirely dependent on its vigour, and the gametical property of such a race is the same in all individuals ; so that, although the race may externally show two somatic developments, there can for that character be no question of "homozygous " or " heterozygous " individuals. I need hardly further discuss at the present time the evidence as to whether characters are paired or not, or whether both kinds of characters paired and single exist, for the matter is still unsettled. If both kinds do exist, then the differences between paired characters must be gametical ; of a single character somatic. It could possibly be shown that by alteration of vigour certain pure dominants would become recessive ; this, however, would not necessarily prove the absence or non- existence of paired characters, but merely that in that particular case the character was a single one. Following on this line of reasoning, we might find the Mendelian cases reduced to very few in number, since all single characters would owe their apparent alternative inheritance to vigour. To our mind the evidence is against all characters being paired, and, in fact, the " presence and absence " method of considering the matter (as largely used by Mendelians) is a tacit acknowledgment that difference between many pairs of allelomorphs is merely one of degree, i. e. that the so- called pair is in reality a single character. In practice all we can be certain about is the difference between the two individuals to be paired, and this may be most easily denoted by x (= presence in one individual) and o (=its absence in the other). The character itself is of course present in both, but the actual difference between the individuals is the only thing that matters. Suppose, for instance, I owe a man £100 and 2d. and he owes me £100 and 4:d., the only thing that matters is the 2d. We may both be millionaires or bankrupt — as long as he has 2d. the matter can be settled. So, therefore, in this case the actual character may be anything or nothing, all that matters is x, which is the difference between the two individuals, that we can see with our eyes. Q 2 228 VIGOUE AND HEEEDITY Another point dealt with by Dr. Beid is the question of retrogression and loss of parts when selection ceases to act. " Eetrogression " occurs, as we have shown above, when a variety mates with a normal and is most likely to happen, if not in the first generation then in the second, and thus the progeny of the variety retrogresses towards normal. Lloyd, however, when dealing with Eats shows that varie- ties owing to inbreeding often tend to reproduce themselves, and thus form a small colony or family race of that variety. Eeid further points out that each animal goes through its ancestral life-history more or less accurately during its development, and thus the so-called "retrogression" may be an advance if the species has at some previous period retrogressed; e. g. suppose a hornless species develops horns, and these horns are again subsequently lost, the retrogres- sion from the second change brings back the horns which, in their time, were an advance, so that in this case we get an advance by "retrogression" which, even if true, sounds absurd. Now if we consider the matter in terms of vigour, we get a much simpler explanation. Suppose the carrying of horns to be a single character, and therefore potentially present in every individual. According to our hypothesis, therefore, the patency or latency of the horns will be entirely dependent on the vigour of the individual. If, therefore, the conditions of life improve, the vigour of almost every individual will be raised, and the horns instead of lying " latent " will become "patent." This "patency" will of course only take place among the newly born individuals, but it will quickly affect the whole race, since horns being "patent" or "dominant" under the new conditions of vigour, these young, even when mating back with older hornless individuals, will produce the majority of their young with horns. So that although the crossing produces a tendency to retrogress, this tendency may, owing to high vigour conditions, be practically negli- gible. Similarly, a return to low vigour conditions would result in the loss of horns. In the first case, therefore, the gaining of the horns was an advance and not a "retrogression." To our mind, however, the logical method of considering these matters is merely as an alteration of development, which need not necessarily be either an advance DISCUSSION OP OTHBE HYPOTHESES 229 or a retrogression. We are quite willing to agree with Dr. Eeid that the longer a character has been latent the greater the stimulus required to awaken it to " patency " ; this, however, is a matter which we have not yet critically investigated, and as Dr. Eeid does not deal with it fully we need not discuss it at present. Our only object has been to show that Dr. Keid's ideas on "retrogression," "rever- sion," " watering down" of varieties, &c., are capable of a much simpler and more direct explanation on the lines of our vigour hypothesis. At the same time, it must not be forgotten that we have described a case in its simplest form, and without regard to the effect of natural selection, and the benefit or the reverse of " latent " characters be- coming " patent," factors which would have, in practice, a considerable effect on the actual development of the race. In the 'Archiv Bntwickelungen,' xxvi. p. 372 (1908), Mr. J. T. Cunningham has published a suggestive paper on the origin and nature of secondary sexual characters, to which, as his hypothesis has many points in common with our ideas, we will refer at some length. He shows, first of all, that secondary sexual characters are somatic sexual characters, and may show themselves on any part of the soma, and are not confined especially to those parts concerned in sexual intercourse or in the nourishing of the young. They may also be absolute (as in the case of the antlers of Deer) and solely confined to one sex, or they may be merely questions of degree, such as the larger canine teeth of the males in some Monkeys, or larger size as in many species of mammals, or brighter colour as in many species of birds, or, again, in the presence of specialised glands (such as the lacteal glands) which are functional in the females and rudimentary in the males. Whatever be the nature of these characters, however, they are entirely somatic, and have not, in most cases, any special relation to reproduction (if we except the actual "accessory glands and reproductive organs), but at the same time their somatic development is dependent in some way or other on the sex of the individual. Mr. Cunningham's paper is in the main an attempt to show how these characters develop, and also how they may even have originated. 230 VIGOUE AND HEEEDITY Before propounding his theory, however, he reviews and rejects several of the principal hypotheses that have been brought forward to account for these sexual dififerences, and in the first place discusses Messrs. Geddes and Thompson's suggestions of katabolism and anabolism. This hypothesis we have accepted in this book and have referred to at length (Chapter XI.), so that we need not discuss Mr. Cunning- ham's objections in great detail here. He raises two main objections : — (1) That if this hypothesis be true, it should hold for Pigeons, Eats and Cats, as well as for Fowls and Birds of Paradise. At first sight this may seem a well- founded objection, but in reality it is not so. First, as we have already pointed out, sexual differences are not only somatic differences of shape, form, colour, &c., but are also temperamental, e.g. activity, voice, pugnacity, &c. ; so that although, to all appearances, the sexes may be alike yet their character and habits, which are largely dependent on their metabolism, may be widely different ; or, again, the breeding season in some species may be practically continu- ous throughout the year, and the katabolic excess of the males may thus be expended in frequent sexual intercourse. For instance, in Cats, Eats, and Eabbits, where the breeding season is practically continuous throughout the year, the sexes are externally alike though differing slightly tempera- mentally, and, on the other hand, we may cite the case of birds where the definite breeding season is most marked and the sexual differences are most striking, but in those species where the male is specially noted for his song the sexes are externally practically alike, e.g. Nightingale, Lark, and Canary. Mr. Cunningham seems to have taken it for granted that secondary sexual characters must be of somatic or " material " nature, and to have overlooked the fact that "temperamental" differences are equally "secondary sexual characters " and equally indicative of a state of metabolism. Then, again, as we have tried to point out elsewhere, the metabolism in some species is much more active than in others, and when the metabolism is high both the katabolism and anabolism are accentuated, and consequently the sexes differ more widely than if the metabolism be low. For example, among the indigenous animals of the Arctic regions, where the conditions are conducive to a low meta- bolism, we have no instances of marked sexual differences, DISCUSSION OP OTHEE HYPOTHESES 231 one of the most striking proofs of which is the Eeindeer, in which both sexes bear antlers. Seals must, of course, be excepted, since, although they live under rigorous conditions as regards temperature, nutrition is always abundant. To deny the existence of secondary sexual characters araong Pigeons is not, strictly speaking, true, although it appears so at first sight; our experiments show, for instance, that certain colours tend to be found more frequently among females than males ; in Dragoons there is a distinct difference in the size of the wattle, if birds of the same age and different sexes be compared; in almost all breeds there is a slight difference in the shape of the skull, and in every species there is a marked difference between the sexes in actions and voice. Therefore we must allow that in every dioecious species each sex has certain characters, whether somatic, vocal or temperamental, which must be regarded on Mr. Cunning- ham's broad basis as secondary sexual characters. This being so, Mr. Cunningham's first contention — that such characters cannot be due to katabolism and anabolism, since in some species the sexes are alike and in others they differ — falls to the ground. His second objection to Messrs. Geddes and Thompson's suggestion is that certain characters which in some species are typical of the males (such as size) are in other species typical of the females, and that katabolism if it produces a larger size in one species could not produce a smaller size in another. This objection may be answered in several ways : first, as Messrs. Geddes and Thompson point out, an animal cannot remain permanently in a state of very active katabolism, since in such a state it is expending more than it assimilates, consequently among males periods of very slight katabolism or even of anabolism must alternate with periods of violent katabolism, or, as Messrs. Geddes and Thompson put it, with a " katabolic crisis." This is well shown in many species of birds which at certain times of the year put on a plumage which resembles the female, and is, therefore, typical of anabolism ; or, again, we have the examples of the males in many species of insects which, in the imago or sexual state, take no nourishment, and as compared with the females are very short-lived. Now if the males have, even after maturity, anabolic periods, it is 232 VIGOUE AND HEREDITY still more obvious that during their youth, before their maleness becomes fully established, they will be anabolic ' (as is eYidenced in most species by the similarity of the/ sexes in the young and their resemblance to the adult female), hence males will grow and will continue to gro"W until their maleness becomes su£&ciently pronounced to check the growth, and, as is well known, early breeding tends to check growth. Consequently, the relative size of males and females will largely depend on the period at which their sexuality becomes active. Now by virtue of her anabolic character a female does not become sexually active until not only the gonads but her other accessory organs have assimilated a fair amount of reserve material, and this reserve will be greater in a female than in a male. Consequently in those species in which the sexes mature at the same age the preparation for that maturity must have begun at an earlier period in the female than in the male; hence her rate of growth must have been checked at an earlier period, and hence females are usually rather smaller than males. In many species, however, the female becomes sexually mature later than the male, and in these cases her growth will have been checked at the same time as in the male or possibly later, and she will therefore be equal in size to the male or larger. In this way we see that the relative sizes of the sexes are by no means incompatible with their katabolic and anabolic tendencies. We have dealt so far with the general size which is fixed throughout life, but we must also consider the other question of growth, such as the antlers of Deer, which are associated only with sexual maturity. Mr. Cunningham argues that these sexual characters, being cases of growth, show an anabolic tendency, and therefore contradict the assumption that males are kata- bolic. His error arises from the fact that he has not realized that males have periods in which they are " ana- bolic," and that, although the general tendency of males is towards katabolism, it need only be in the actual " breeding season " that the katabolism is really marked and becomes a " katabolic crisis." In temperate regions the breeding season of Deer is in autumn, whereas the antlers grow during the summer, that is to say, in the anabolic period preceding the katabolic period. DISCUSSION OF OTHEE HYPOTHESES 233 The female does not grow antlers, since all her anabolism is directed towards preparing herself for the nourishment of her young. The oestrus period and giving off of the ovum from the ovary is the sign of the "katabolic crisis," which is in both sexes the concomitant of reproduction. The question of the more brilliant coloration of males we need not discuss here, since Mr. Cunningham allows that pigments may be waste products and the sign of increased metabolism (katabolism) of males, and in dealing with the question on p. 160 we have tried to explain the rare cases in which the female is the more brilliantly coloured. Although denying that the fundamental difference between the sexes is one of katabolism and anabolism, Mr. Cunningham allows that the secondary sexual characters probably owe their development to physiological (metabolic) causes, and that they are connected with increased meta- bolism of the gonads. Our point of view is that the activity of the gonads is due to increased metabolism, which in turn is largely dependent on the environment (temperature, food, &c.), and that the secondary sexual characters owe their development to the essentially kata- bolic or anabolic properties of the respective sexes, and that any developments of the soma can only take place after the demands of " sex " as regards nutrition have been satisfied. Wallace's vigour theory* is, in our opinion, correct in its main points, and as merely dealing in the rough with the general brighter plumage of males. It has not, however, met with general recognition and acceptance because it does not go far enough. The brighter colours of males are only one feature of maleness, and a theory which only accounts for these without dealing with the many exceptions and the other sexual distinctions found throughout the animal world could not hope to be generally accepted. Daewin's Theoey. Mr. Cunningham next discusses the Darwinian theory of " sexual selection," and points out that it fails for two main reasons. First, that sexual dimorphism exists in the absence of sexual selection, and that, therefore, sexual selection cannot have been the cause of the dimorphism. * Wallace, ' Tropical Nature.' 234 VIGOUE AND HEREDITY Secondly, if the theory held good, the characters would be inherited equally by both sexes, though not necessarily by all individuals. Hence it follows, according to Cunningham, that the inheritance of the secondary sexual characters is " linked " with that of the gonads. He allows, however, that the characters are inherited equally by both sexes, but remain latent in one sex and patent in the other. As proof of this he notes : (1) that sterile females may often assume male plumage, and vice versd ; (2) that either males or females when crossed with another species may transmit to their progeny the special characteristics of the opposite sex of their own species ; (3) in some parthenogenetic insects {e.g. Bees) unfertilized eggs produce perfect males. With these arguments we entirely agree. He then goes on to show how castration causes a suppression of the secondary sexual characters, and further quotes more particularly the work of Geoffrey Smith on certain species of Crabs {Garcinus trachas), which when infected by a parasite (Sacculina) entirely lose their, secondary sexual characters, and he deduces from this an intimate connection throughout life between the gonads and these characters. He considers as the most absolute proof of this the fact that the first result of this parasitism is the atrophy of the gonads, subsequently followed not by a general atrophy of all parts of the body, but only of the secondary sexual characters, and if it be a male that is attacked an enlarge- ment of the abdomen follows, which is a female characteris- tic. He allows that the atrophy of the gonads is due to absorption of nutriment by the parasite, but argues that absence of nutriment alone could never cause increased growth, and that, therefore, this increased growth must have some connection with the "latent " female characters which tend to become active after the atrophy of the male gonads. At first sight this line of argument appears entirely sound, and we are far from denying any connection between the secondary sexual characters and the gonads, but we cannot accept this argument as proof of it. Every organism assimilates more nutriment than is necessary for its immediate wants {i.e. to keep it alive) ; in youth this nutriment is expended in somatic development, DISCUSSION OF OTHEB HYPOTHESES 235 then comes the development of the gonads, the amount of nutriment stored by the female being in excess of that stored by the male. Now if the organism be placed in unfavourable circumstances, the lack of nutriment if it be not too severe is first felt by the gonads, and temporary sterility ensues. (If the check be sudden and severe a false " katabolic crisis " is brought about and reproduction is attempted, as frequently occurs in plants.) If this lack of nutrition be sufficiently severe and prolonged the gonads (as in the Crabs) may atrophy and cease to absorb any nutriment from the soma. As soon as this stage is reached, there will be a larger amount of nutrition available for the soma, and since we may presume that the parasite still absorbs the same amount as before, more nutrition is available for the host, and consequently the abdomen may increase without that increase being in any way due to latent female characters. Now we quite agree with Mr. Cunningham that increase of size implies anabolism, and consequently the increase in the size of the abdomen of these Crabs implies an anabolic tendency, and, therefore, might lead to the development of the latent female characters, and even subsequently, as has been stated, to the formation of ova. This leads to the question as to the period at which the sex is definitely determined, and whether it is possible to alter it after fertilization. The evidence on this point is very contradictory, and we do not propose to enter on the matter here ; but in our opinion, bearing in mind Yung's Tadpole experiments, Laulaine's work on the Chicks, the case of these Crabs, and many other minor cases, it seems that even among the higher animals the sex may be controlled after development has begun, though naturally at a very early period of festal or embryonic life. The chief controlling factor is nutrition, which will produce a " kata- bolic " or " anabolic " habit and thus determine the sex. On this line of argument I am of opinion that in the Crabs the " anabolic tendency," as shown by increase of size, which follows on the atrophy of the male gonads, is responsible for the occasional development of ova in the gonads, rather than vice versa, as Mr. Cunningham suggests. The fact that worker Bees may become functional females or " queens " by a more favourable diet during their 236 VIGOUE AND HEEEDITY embryonic life is a strong argument in favour of my point of view, as is also the fact that in certain insects the males in winter may develop ova in their testes, although these ova do not mature. The Mbndelian View. The fact that the sexes are usually produced in equal numbers has caused many, who hold Mendelian views, to attempt to explain sex as a Mendelian character. Several hypotheses differing slightly in detail have been devised to account for sex on Mendelian lines, and Mr. Cunningham quotes the hypothesis of Castle. This gentleman assumes that both sperm and ova are of two kinds, male and female, and that the only zygotes that can develop are those formed by the junction of a male sperm and female ovum, or vice versd. This hypothesis, if true, would account for every individual containing latent the characters of the opposite sex to that which it obviously possessed, and would also account roughly for the equality of the sexes. On the other hand, we fail to see any real proof of this hypothesis, and, in default of such proof, it need not in our opinion, and in that of Mr. Cunningham, be seriously considered. Geoffrey Smith maintains that, whereas males are really hermaphro- dite and contain latent the female characters, females are purely female. This is another Mendelian hypothesis, and, if true, would give the equality of sexes. In some ways it may be said to be a better hypothesis than Castle's, since it only makes one assumption instead of the two or more required by Castle. On the other hand, it is more easily capable of disproof, since if the females are purely female, how are we to account for the development of male characters in aged females or among those whose ovaries have become diseased ? Following on these lines, it has been suggested that the secondary sexual characters have become, in Mendelian terms, linked up with the sex character, and are thus not only inherited together, but the dominance of the one leads to a dominance of the other. This hypothesis, however, breaks down in many instances. For example, Mr. Cunningham quotes some experiments made in crossing the normal female of a moth {Aglaia tau) with a melanistic variety only found in males, and known as A . lugens. Now DISCUSSION OF OTHEE HYPOTHESES 237 on this hypothesis the lugens character should have an inheritance linked with maleness, and consequently there should have resulted an approximate equal number of males and females, and all the males should have been lugens and all the females tau. In actual experiment, however, there appeared males and females of both varieties, though a majority of the males were lugens and a majority of the females tau. Apart from this experiment, however, if certain characters had a definite linked inheritance with sex, it would have been impossible for us to have produced our male Dun Pigeon as described on p. 111. Here, again, to our mind the conception of vigour, which we are attempt- ing to outline, meets the case better than any of these hypotheses and without violently upsetting the basis of existing theories. That certain characters usually (but not always) develop together cannot be denied, and we would explain that fact by saying that they develop under similar conditions of metabolism. In a developing embryo the characters develop in a certain sequence, the development of some commencing earlier than that of others ; at the same time the metabolism, depending on the condition of ^ the mother or external circumstances, will vary, and each character will develop according to the state of the metabolism at the time when it is due to develop. Now if we presume, for the sake of argument, that the sex character develops before the colour character, and if, as we have already shown, femaleness makes a more exhaustive demand on the nutri- tion than maleness, then it follows that in females the amount of nutrition * available for the development of characters, of subsequent determination than sex, is less than in males. + Hence these characters develop under a lower metabolism, and hence in coloration females are * Nutrition and metabolism are in this connection nearly synony- mous terms, since at this period the metabolism is almost entirely dependent on the nutrition. f In the 'Cairo Scientific Journal,' vol. viii. p. 89 (1914), we have shown with regard to certain of our experiments on fowls that there is a marked tendency for females to inherit a larger proportion of recessive characters, and also for white birds to inherit a larger proportion than coloured ones. The actual figures for eight characters are : — Males average 1'3 recessive ; 0'6 heterozygotes ; 6'1 dominants. Females „ 1-71 „ I'O „ 5-29 Coloured „ 1-16 „ 0-73 „ 5-11 Whites „ 2 „ 1-2 „ 3-8 238 VIGOUE AND HEEEDITY usually less intensely pigmented than males, and thus normally we shall find a certain lack of intensity of colour associated with femaleness as compared with the males of any species. The development on these lines is, however, by no means absolute ; for instance, the metabolism from the first may be very high, and consequently, even after the sex has begun to develop, the available nutrition may be sufficient to allow the development of characters not usually associated with femaleness ; or, again, although in the first instance the development of femaleness may have lowered the metabolism, more favourable external conditions may come into play, and thus affect the de- velopment of some of the later characters. This, to our mind, affords a comparatively simple explanation of the normal association of certain somatic characters with a certain sex, and at the same time^ enables us clearly to understand the numerous exceptions which are continually arising. Having discussed and rejected these hypotheses Mr. Cunningham then proceeds to give his own theory, and he commences by stating that, although he previously con- sidered all activity and development to be due to nervous stimulation, he now considers it chiefly due to chemical origin. As proof of this point of view he instances the fact : (1) that the development of secondary sexual characters has taken place after all the nervous connections had been cut, and (2) that functional activity and metabolism have been proved to be correlated with the presence of specific sub- stances " circulating in the blood," e.g. the increased respiratory movements following increased activity of the muscles is due to the stimulation of the respiratory centre by the increased CO2 in the blood. He also quotes Ancel and Bouin (C. B. Soc. de Biologic tom. Iv. p. 1682 (1903) ) as showing that the interstitial cells of the testes supplied a secretion without which the male characters did not develop, as well as the experiments of Shattock and Selig- man (Proc. Eoy. Soc. Lond. vol. 73 (1904) ) on Fowls, which showed that a minute portion of a testis left in the body of a castrated Eowl, even if only adhering to the peri- toneum or some other part, was sufficient to cause the develop- ment of the male secondary characters. On this and other similar evidence he suggested that the development of DISCUSSION OF OTHBE HYPOTHESES 239 these secondary sexual characters was due to the presence in the blood of certain specific substances which he calls " hormones." These hormones (he suggests) originate in the gonads; the hormones from the testes are different from those produced by the ovaries, and in both sexes these hormones are produced in the greatest quantity when the organs are most active. It will be seen, therefore, that Mr. Cunningham's hypo- thesis is quite compatible with ours, in that we both agree that these developments owe their existence to physiological causes and are brought about through the metabolism of the body. Mr. Cunningham, however, believes that the develop- ment and alteration in the metabolism of certain definite parts have been directly stimulated by a secretion arising from the gonads, a point of view^ for which a good deal may be said, whereas we believe it to be due to a general alteration in the metabolism, indirectly due to the environment. The argument against Mr. Cunningham's suggestion is, to our mind, to be found in the case of the assumption of male characters by females, and vice versa, for if the development of secondary sexual characters be due to the presence of special hormones, which are only secreted in sufficient quantities to affect the soma when the gonads are active, it is difficult to conceive how the atrophy or removal of the gonad of one sex or the other will cause a secretion of the hormones of the opposite sex, whereas the removal of the ovaries, the controlling cause of the ana- bolism of the female, from the mere absorption of nutriment, would naturally tend to make the individual katabolic, since the accumulating warehouse, so to speak, has been re- moved ; and similarly the removal of the testes, the con- trolling factor in the katabolism, would naturally result in a more anabolic condition, that is to say, in an approach to femaleness. These two hypotheses need not, however, be absolutely antagonistic, for it may be that the " hormones " of Mr. Cunningham have an appreciable effect in increasing the katabolism or anabolism of the individual. We must remember, however, that the sex, or, as we prefer to consider it, the predominant anabolism or katabolism of an individual, is determined long before the gonads are active, and that, therefore, hormones produced by the activity of 240 VIGOUE AND HEEEDITY the gonads could not possibly have caused the growth of many sexual characters, which take on a definite form very early in life. Further than this, many characters may be considered sexual in some species and not in others (c/. antlers in Deer and Eeindeer), which could not be the case if their development was primarily dependent on " hormones from the gonads." Mr. Cunningham then proceeds to discuss the difficult question of how these secondary sexual characters first arose, pointing out that hitherto all work on the subject has been devoted to showing, or trying to show, how they develop but not why they arose, or how certain somatic factors became the concomitants of one sex rather than of the other. Mr. Cunningham's views are Lamarckian, and on these lines he naturally attempts to account for the origin of these characters. He commences by stating that the secondary sexual characters correspond in their use only with those habits or instincts which are directly connected with sexual im- pulse. He argues that Fowls fight and Pigeons do not, hence Fowls are fitted with defensive weapons and Pigeons have none ; but if he really knew Pigeons he would realize that they fight, and with extreme pugnacity, using their wings as weapons in common with most birds, and that therefore, according to Mr. Cunningham's theory, they should develop spurs on their wings. Mr. Cunningham argues that exercise" or use produces a hardening of the skin, or, if sufficiently severe, a growth of bone, and therefore the continual fighting of Stags by "butting" has produced the antlers. In the case of the Bovidse, how- ever, he maintained that the fights are not so severe, and hence have caused a hardening of the skin, which, not becoming torn, does not cause the bone to fall off, and hence they are permanent. We are not intimately con- cerned here with the origin of these growths, but in our opinion there is an essential difference between the horns of the Bovidffi and the antlers of the Cervidae. The former are, to our mind, essential weapons of defence for the individual, and are in no sense sexual — they are therefore persistent throughout the year and in both sexes. In the Cervidse the antlers are essentially " sexual," and do not owe their existence as weapons of defence, but must be DISCUSSION OP OTHEE HYPOTHESES 241 considered rather as sexual adornments; hence they be- come a feature of the males, and are only worn during the period of sexual activity ; their use for fighting being entirely a secondary consideration. Coming to birds, Mr. Cunningham cannot be taken seriously, for in one sentence he says that the displays of birds are never a movement of the body, and in the next that the feathers grov7 long by the habitual movement of the muscles. These remarks may be easily answered in a few words : first, the displays, though sometimes motionless as far as the individual as a whole is concerned, always imply the raising or display of certain feathers brought about by muscular contraction. But in many, and perhaps in most, species the female also dances and displays, although she has nothing to exhibit, and if the dances have brought about growth in males why not in females ? Then many species of birds show specialised wing feathers, e.g. the Nightjar {Macrodvpteryx) , in which only one wing feather grows to an excessive length, although the muscles to which it is attached equally control the neigh- bouring feathers. The same may be said of the elongated feathers on the heads of many birds which are not more erected than their less conspicuous neighbours. We are unable, therefore, to agree with Mr. Cunningham on the special functions and uses of these essentially sexual characters ; but, leaving the specialised cases, he comes to more general grounds in discussing whether or not these characters are of blastogenic origin or not. Why, for instance, should the hormone from the gonad of a Stag cause the growth on the forehead and not on some other part of the body ? Whether we accept his hormone theory or not, the question of why a male Stag should show a growth on his forehead instead of elsewhere on his body needs an answer. In the present state of our knowledge we prefer to consider that they have a blastogenic origin,, which may, under certain circumstances, develop ; but Mr. Cunningham, believing implicitly in the connection between function and development — in regard to which we must allow that there are numerous very striking instances in Nature — seeks to connect their origin with their function. The heredity of acquired characters is not generally 242 VIGOUE AND HEEBDITY admitted, owing to the fact that the gonads are developed separately from the soma, and thus the " determinants " cannot be in any way affected by changes in the soma ; but Mr. Cunningham points out, and in this we are in entire agreement with him, that the gonads are nourished by the blood, and hence they cannot be entirely indifferent to changes in the soma. It seems to us obvious that, if the soma be under favourable conditions, the metabolism will increase, and the gonads receiving a richer supply of blood will also have their metabolism or potentiality of develop- ment increased. Mr. Cunningham then goes a step further, and argues that if " hormones " from the gonads are able to cause the development of definite parts of the soma, then why should not the converse be true, and external circumstances causing a development in the soma might thus increase in the blood definite hormones, which would affect the determinant of that particular part in the gonad, and consequently there would be a tendency in the next generation for that determinant to exercise a slightly increased development in the soma ? This point of view is certainly logical, but it is as yet merely a suggestion. On his main basis, however, Mr. Cunningham's ideas are quite compatible with our own, but he has gone into more detail, and attempted to define the physiological cause which brings about changes in the metabolism, and thus in the development of parts of the soma. Towards the end of the paper, however, he seems almost to have been driven round to our standpoint, for he allows that the "hormones" may after all be merely waste products of cell metabolism ; and on p. 423 he writes : — " It is better perhaps to conceive the modifica- tions in the gamete as an increase of vigour in the determinant, which shows itself in increase of growth in the corresponding part of the soma, which develops from that determinant." We have dealt at length with this paper, as it contains many suggestions which are very closely connected with our general thesis. We mainly differ in the rejection by Mr. Cunningham of Messrs. Geddes and Thompson's ana- bolic and katabolic hypothesis of sex, and we have tried to show in still further detail the reasons which have led us DISCUSSION OF OTHER HYPOTHESES 243 to accept that hypothesis as it stands. The attempt to account for developments of the soma by hormones is merely an attempt to particularise more clearly the exact method by which changes of metabolism are brought about in the soma. From the " sex" point of view, we cannot accept the hormone theory; from the point of view of characters not specially sexual there may be something in it, though such characters, if the katabolic and anabolic theory be accepted, are very few in number. On the main essential basis, however, there is nothing incompatible between Mr. Cunningham's " hormone " theory and the present vigour hypothesis, since we both agree that all developments are directly dependent on metabolic activity, and further that the metabolic activity of the body must to a greater or less extent be reflected in the metabolic activity of the gonads, and thus have an appreciable influence on subsequent generations. B 2 244 VIGOUR AND HEREDITY CHAPTEE XVI. SUMMAEY AND CONCLUSION. All living things live by reason of the physiological activity of their bodies. This physiological activity consists in the consuming of raw material (nutrition) and convert- ing it into various chemical compounds v^hich can be, sooner or later, easily assimilated, and thus become part of the living organism or enable it to carry on its normal functions. The conversion of the raw material is known as meta- bolism, the rate at which that conversion takes place we have defined as vigour, a high vigour implying a rapid rate, a low vigour a slow rate. Since metabolism is a purely chemical change, it follows, like all chemical changes, that its intensity (i. e. rate) will depend on the temperature and on the kind of material on which the change is being conducted. In animal life this depends largely on the environment, which embraces climate, temperature and food. The vigoiir of an animal will therefore largely depend on its environment. When nutrition is abundant and the temperature favourable, it follows that the rate vigour will be high and, in consequence, more raw material will be converted into a state in which it may be assimilated by the living body than the body itself can comfortably consume. If the animal be young and growing, this surplus nutrition goes to form more living tissue, in other words, the animal grows ; but when once an animal has reached its full growth, a further increase of size becomes im- possible, and then follow an enlargement and activity of the sexual organs, so that the surplus nutrition, instead of adding to the body of the animal itself, is utilised to form the nucleus of a fresh organism. SUMMAEY AND CONCLUSION 245 Most living things, all the higher ones, are of two kinds, male and female, and these, according to Messrs. Geddes and Thompson, differ fundamentally in their metabolism. In the case of the female, this surplus of nutrition to which we have just referred is stored up, after the sexual organs have been well nourished, as fat and other compounds ready for the production of nourishment for the young. The females are, therefore, said to be anabolic, i. e. they do not immediately use up all the nutrition they consume. The males, on the other hand, are known as hataholic, and have much less power of storing up this surplus nourishment, so that after the generative organs have been well nourished, the superabundant nutrition must be got rid of, since it cannot be stored to any extent; it is consequently ex- pended in energy, activity, restlessness, fighting, song or other noises, brilliancy and length of plumage or coat. From this it follows that the colour, size and external appearance of an animal will be, to a certain extent, depen- dent on its vigour, and that therefore this vigour, which, as we have seen, is largely dependent on its environment, will have played no small part in bringing about evolution, since it is only after the characters have been fixed by vigour that Natural Selection can come into play. We have shown that the sex organs only tend to become active when the vigour has reached a height which produces a surplus of readily assimilated compounds, after the needs of the soma have been satisfied, and hence it follows that the vigour of the resulting ova, spermatozoa and ensuing zygote must be directly dependent on the vigour of the parents. The development of the characters in the zygote, there- fore, must also be in their early stages dependent on the vigour of their parents, and therefore vigour must largely control inheritance. This fact, in its many and varied forms, we have tried to make clear in the foregoing pages ; the matter is one of extreme complexity, and one which, in the present state of our knowledge, it is impossible to prove by direct means. We have, however, tried to make it clear by showing that where, owing to environment, inbreeding or obvious weak- ness of the parents, a low vigour would be expected, then the characters resulting in the young were also of a, weak type. 246 VIGOUR AND HEREDITY We have dealt with various debatable points such as prepotency, reversion, &c., and shown how these things also might be susceptible of a logical explanation on our hypothesis. The incompleteness of Natural Selection as the sole factor in evolution has also been mentioned, as well as the two great theories of inheritance enunciated by Galton and Mendel. The partial truth of all these theories we allow, but in pointing out several definite and concrete cases in which they break down, we have at the same time tried to show that these cases are largely cleared up by following our hypothesis. Lastly, we come to the arising of apparently new characters — whether there is " anything new under the sun," whether these characters are really new creations, or the recrudescence of latent possibilities, or the alteration of existing characters is a question which cannot at present be answered, but in many cases the arising of these new forms seems to have occurred when the circumstances have been of a kind to lead us to expect a definite change in the vigour of the parents. Then we come to the practical side of the question — our hypothesis does not enable us to definitely prophesy the result of any particular mating, and it is on this account that the Mendelian theory is largely discredited by many practical breeders who know only too well that, as far as our present knowledge goes, certainty of inheritance is an impossibility ; but we claim to have shown that vigour has an extremely important effect on inheritance, and that the probability of certain characters appearing may be largely increased by careful attention to the environment. At present we claim no more for our hypothesis ; but what we have attempted is to have treated it carefully and scientifically, and to have shown that the vague remarks on vigour, which are to be found in the scattered \?ritings of those who treat of live animals, are worthy of more serious attention, and that the physiological activity by which all living beings exist has been no small factor in the evolution of the various forms of life. It may seem to those who have followed us so far that the arguments laid down, having been founded chiefly on mammals and birds, may probably not have the universal SUMMARY AND CONCLUSION 247 application which it is sought to give them. In these days it is almost impossible for a zoologist to know accurate details of more than one or two groups : and to attempt to argue out an hypothesis on general rather than detailed knowledge would be to court failure; in a general way, there can be little or no doubt that when carefully examined the other orders of animals as well as plants will prove to conform equally with the arguments here laid down. And what of the future result ? Once these facts are clearly recognised, what a far-reaching effect will be at once possible from the economic aspect ! The advantages that will accrue and the saving of waste will be enormous. What about the human race? When it is once realised that characters may be largely eliminated by environment if brought to bear early enough, we have open to our vision possibilities far greater than any offered by the actual work of Mendel ; when we realise that gentlemanly qualities are largely due to "mens sana in corpore sano," that the failings of the submerged tenth are due as much to their surroundings as to the " sins of the fathers," what a vista of possibilities rises up before us! We see a nation living in cleanliness of miad and body, fine strong men and healthy women. In some the brain will be developed, in others the muscles, but in all, honesty, courage, truthfulness and energy. Many diseases such as consumption, insanity, drunkenness, will ipso facto disappear. I am raising no Utopian picture ; such things are possible, and the change would take barely two generations, while if at the same time restrictions were placed on the marriage of the unfit, the completion of the change would advance still quicker. The nation that first realises this can and will dominate the world and be the finest race that has ever yet been known. But I digress — my part is that of the seeker after truth, he who in this utilitarian age is practically neglected. What is the use of knowing what colour a Silver Pigeon will breed or the sex of a Mouse's young ? is what most of my friends and others think when they see me studying the matter year after year. The answer has been attempted in this book. Did zoological societies and others realise the possibilities of accurate and careful classification and 248 VIGOUE AND HEEEDITY attention to details, the results here set forth would have been obvious years ago, and much money have been saved or diverted to less vpasteful causes. It has not been so, and I, as a poor man, have had to keep only those animals that could be fed on the crumbs from my breakfast table; but as a straw shows the direction of the wind, so may the colour of a Pigeon or sex of a Mouse give us a hint of the direction in which the laws of Nature work. INDEX. Abnormal varieties of animals, and vigour, 175 Absence of Mendelian characters in Meriones hybrids, 92 Acquired characters, 177 vigour is inherited, 178 Activity, physiological, 246 Estivation and vigour, 44 Agar, "W. E., 202 AgUa iugens, 236 tau, 236 Alexander, C, 198 H. G., 193 Albatrosses, 176 Alexandrian Rat {Mus rattus), i&Yrn coloured. 47 Alexandrian Eat {Mus roi^Ms), vari- ability of, 37-39 All inheritance a combination of vigour and gametical inheritance, 220 Allied forms in same locality, 68 Alternative reproduction, 197 Anabolism, 44, 45 and Ducks, 157 and katabolism, second- ary sexual characters, 230, 231 increased by nutrition, 162 Ancel, P., 238 Andalucian Fowls and Mendel's Law, 217 Animals, captive, and sterility, 14 dark-coloured, and vigour, 42 domestic, and environ- ment, 209-210 domestic, and standard of vigour, 211, 212 domestic, and variation, 210, 211, 216 • excess of ova produced by lower, 163 Animals in confinement and in- fertility, 173 wild and domestic, differ- ence between, 208, 209 wild, and Mendel's Law, 67 wild, fixity of characters in, 210 Antlers of Deer, 159, 229 of female Beindeer, 160 Arobdall Keid's hypothesis, 224, 225 Artificial selection, 208, 209 Australian Duck, 128 Autumn song of Birds, 172 Bach, Dr. E. A., 30 Bacot, A., 67 Barrett-Hamilton, Major G. E. H., 4, 18, 19, 25, 28 Bateson, Prof. W., 4, 52, 109, 208 Bats, hibernation of, 43 Batten, Dr. P. B., 72 Beddard, Dr. P. E., 4 Beebe, C. William, 25 Birds, autumn song of, 172 bright colours in female, 160 coloration of, 16 fat in, 28 muscles of, 241 of Polar regions, 20 psychological stimulation in, 187 song and colour, 198, 194 song and vigour, 45 Black Eats, 176 Blended and alternative inheritance, 226 Blended races in Ducks, 67 in hybrids, 67 in Meriones, 67 Blending of characters, 182 " Blends," inheritance of, 160 Blind animals, 72-75 250 VIGOUE AND HEREDITY Bonhote, J. L., 13, 28, 38, 61, 69, 77, 78, 88, 122, 125, 141, 152, 189, 216, 217 Bouin, P., 238 Boveri, Theodor, 179 Breeders and Mendel's Law, 246 Breeding a male Dun Pigeon, 112 and environment, 188, 189 and loss of vigour, 167 — — — — of House-Sparrow, 191 of Eats, 61, 63 of Sheld-Duoks, 192 season and expenditure of vigour, 162, 163 — season of Meriones, 78 season, time of, dependent on vigour, 163 — — ■ stimulated in Finches, 192 stimulated in hybrid Ducks, 191 time of, in Arctic species, 19 Bright colours in female Birds, 160 Brighter colours of tropical species due to vigour, 23 Brown Kat (Mus norvegicus), 38 constancy of, 39 Butterflies, change of colour under different temperatures, 33 Calcarius lapponicus, 20 Canaries„colour of eyes in, 69 Canary hybrids, 9 Captivity, animals in, excess of males in young of, 46 Castle, W. E., 62, 236 Cats, breeding of tortoiseshell male, 60 colour and tortoiseshell, 55 colour inheritance in, 60 colour of, affected by vigour, 223 j experiments on, 55 genealogical tables, 561 Miss Wilde's experiments on, 55 pedigree of tortoiseshell males, 59 results of matings for tortoise- shell males, 57, 58 vigour and tortoiseshell, 57 Cattle, horn inheritance, 62 Wild, 36 Cervus dewanceK, 25 eldi, 25 Change of climate and new species, 174 of colour and humidity, 89 in gamete, 202 Chapman, Abel, 189 Frank M., 188, 189 Characters, blending of, 182 Characters, dominant converted into recessive, 114, 115, 227 fixity of, in wild animals, 210 gametic, developed by vigour, 95 loss of, a sudden muta- tion, 74 — loss of, from disuse, 72, 73 loss of one sometimes due to increase in another, 94 • paired, and vigour, 226, 227 pairs of, 207 presence and absence of, 227 recessive, change from dominant, 227 recessive, effect of vi- gour on, 48 secondary sexual, 230, 231 single, and vigour, 227 skull, of Meriones, 92, 93 somatic, developed by environment, 95 somatic, development of, 96 somatic sexual, 229 suppression of, governed by Mendel's Law, 181 Chasmorhynchus, 24 Chequered Pigeons and Mendel's Law, 217 Chequering in Pigeons, 205, 206 Chickens, sex of, 164 Chodat, Bobert H., 30 Climate affects vigour, 18 and colour, 12 change of, and new species, 174 influencing colour through vigour, 19 Cole, Leon, J., 109, 115 INDEX 251 Colour, 267 a waste product, 27 an index of vigour, 6, 29, 30, 36,41 and climate, 12 and health, 17 and Meriones, 33, 86, 87, 88 and natural selection, 16 and sex, 13 and sex in Ducks, 125 and sex in Egyptian Bats, 14 and sex in Pigeons, 99, 100, 116 and song of birds, 193, 194 and temperature, 22, 23 and vigour, 16, 18 bright, in female birds, 160 brilliant in hybrids, 48, 184 change of, and humidity, 89 change of, and vigour, 26, 177 change of, in butterflies, 33 change of, in confinement, 33 change of, in Linnets, 33 change of, in Meriones under different temperatures, 86 change of, in moths under different temperatures, 33 dependent on Metabolipm, 41 effect of vigour on, 21, 22, 24 in Oats affected by vigour,' 223 in Ducks, 125 in females, and sex, 48 in Meriones, 86, 87 in Pigeons, 112 in Pigeons, deviation from Mendelian expectations, 112-121 in Pigeons, silver, 198, 199, 200 in Polar circles, 18, 19 in tortoiseshell C^its, 55, 56 influenced by climate, 19 inheritance of, in Cats, 60 inheritance of, in Ducks, 127 inheritance of Pigeons, 97 intermediate shades of, in wild animals, 37 ' new variations of, in Ducks, 125, 126 of Dogs, 50, 52 of eyes in cinnamon Cana- aries, 69 Colour of eyes in fawn Rats, 69 of geographical species, and vigour, 37, 40 of hybrid Duck, effect of vigour on, 125 of hybrid Ducks, 125 of Pigeons, 41, 112 — of Eats, 65 of temperate region species, 25 of tropical species due to vigour, 23 of young, effect of vigour on, 43 red flank feathers of Wigeon, 152 reserve nutrition, 27 shades of, and vigour, 37 silver, in Pigeons, 110, 111 silver, in Pigeons due to lack of vigour, 201 stability of, in nature, 31, 32 stabihty of, in tropics, 25 variability of, in domestic species, 30, 32 white, a sign of low vigour in Pigeons, 114, 115, 116 Colours, nature of animals, 27 Coloration in mammals and birds, 16 Colouring in Crows, 176, 177 Common Sheld-Duck, breeding of stimulated, 192 Comparative vigour of ova and spermatozoa, 166 Confinement and sterility, 14 Connop, E., 64 Constancy of colouring in Crows, 177 ; in Hooded Crow, 176 Continuity of germ-plasm, 178 Conversion of dominant into reces- sive characters by vigour, 114, 115 Corvus cmstralis, 177 capensis, 177 cornix, 176 splendens, 177 Cross between Dragoon and Homer Pigeons, 216 Crosses between races, 185 Crossings and increased fertility, 95 Crossings increase fertility, 83 Crows, constancy of colour in, 176, 177 Cu^not, L., 62 Cumulative effects, 179 Cunningham, J. T., 4, 229-243 252 VIGOUR AND HEREDITY Cunningham's objections to various theories, 230-243 Cunningham's theory, 229, 238 Cunningham's theory and vigour, 238-243 Cunningham's theory reconcilable with vigour, 289 Cunningham's views on Darwin's theory, 233-236 ; on Geddes and Thompson's theories, 230-233 ; on Mendel's Law, 236-238 Darbiaher, E. D., 75 Darwin, Charles, 117, 204 Darwin's theory and vigour, 233-236 Cunningham's views on, 223-236 of sexual selection, 233-236 Davies, C. J., 52 Death and high vigour, 165 Deer, antlers of, 159, 160, 229 Definition of vigour, 5 Delage, Prof. Yves, 179 Delcourt, A., 202 Determination of sex influenced by vigour, 119, 120 Determination of sex in the embryo, 119 Development of gametic characters by vigour, 95 Development of somatic characters by environment, 95 Deviation from expected Mendelian results in colour of Pigeons, 112-121 Deviations from expected Mendelian results in sex of Pigeons, 100, 101-112 Dimorphism, sexual, 233 Disappearance of Pintail characters in hybrids, 153, 154 Discontinuous distribution, 177 Discussion of other hypotheses, 224r-243 Disuse, loss of characters from, 72, 73 Dogs, colour of, 50 experiments on, 50-52 — — matings and results, 51 vigour and colour, 52 Domestic animals and environment, 209, 210 Domestic animals and variations, 210, 211 Domestic animals, variability of, 85, 36 races and reversion to wild type, 213 races, variability of, 172, 208 species, variability of colour in, 30, 32 Dominant characters converted into recessive, 114, 116 characters becoming re- cessive, 227 Doncaster, Leonard, 65, 67, 59, 67, 109, 110, 115 Down of Duck hybrids, 149 Drinkwater, Dr. H., 179 Duck, Australian, 123 Mallard, 123 Hellers, 123 Pintail, 123 Sheld, breeding of, 192 SpotbiU, 123 Trigens, 18 Ducks, 122 and anabolism, 157 and Galton's Law, 137, 139, 153 blended races in, 67 colour in, blending of, 125 cross between Pintail- Mallard- SpotbiU, 216 down of hybrids, 149 eclipse plumage of Wigeon, 152 eclipse plumage of Wigeon, and vigour, 152 effect of vigour on colour and sex of hybrid Ducks, 126 experiments on, 46 fertility of hybrids, 123 flank feathers of Wigeon, 152 genealogical table, 124 grey variety of, 146, 147 heredity of white under parts, 144 hybrid, P.M.S. eclipse plum- age, 127 inheritance of blends of colours, 127 — — — Mallard- SpotbUl-Australian- Mellor crossings, 132-139 new variations in colour of, 125, 126 Pintail - Mallard - SpotbiU - Australian crossings, 132 INDEX 253 Ducks, Pintail-Mallard-Spotbill hy- brids, description of, 126-131 P.M.S.A.. and M.S. A. Mell. crosses compared, 138 P.M.S. hybrids discussed, 131 P.M.S. first generation, des- cription of eclipse plum- age, 127, 128 P.M.S. first generation, des- cription of full plumage, 126 P.M.S. second generation, description of full plu- mage, 129 P.M.S. third generation, des- cription of full plumage, 130 reversion, 141 sex and colour, 125 sexes of hybrid, 124 variations and resemblances to other species, 140 variation in flank feathers, 142 -white under parts of, 143, 145 white variety of, 148 Wigeon, eclipse plumage of, 142 " Dun " colour inheritance in Pigeons, 110, 111 Dunlin's flight, 195 Eclipse plumage of Ducks, 22 and low vigour, 151, 152 Effect of realization of law of vigour, 247 Effect of vigour on birds moulting in Polar regions, 19-21 on colour, 21, 22 oncolourintropics, 24 on sex, 164 Effects of environment inherited, 178 lowered vigour on em- bryo inherited, 76 nutrition and tempera- ture, 163 Eider Duck (8. moUissima), 190 ; 8. speotabilis, 190 Embryo, determination of sex in, 119 influenced by vigour, 76 initial vigour of, 167 Environment and breeding, 188, 189 Environment and development of somatic characters, 95 and domestic animals, 209, 210 and fertility, 84 and heredity, 7 and inheritance, 179 and MendeUan char- acters, 75 and stimulation, 188, 189 — and vigour, 167, 244 effects of, inherited, 178 Evolution of sex, 155-185 Exceptions to Mendel's Law, 197 Excess of ova produced by lower animals, 163 Expenditure of vigour in breeding season, 162, 163 Experiments on Dogs, 50-52 on Ducks, 46 on Pigeons, 46 on Kats, 61-64 — ■ Pigeon, proportion of females in, 97-112 proving truth of Gal- ton's theory, 216 refuting Galton's Law, 217 Eyes, colour of, in Canaries, 69 in Fawn Kats, 69 Factors modifying Mendelian Law, 206 Fat, 27 an index of low vigour, 28 and fur Seals, 156 and low vigour, 171 in birds, 28 reserve material, 28 Fawn coloured Kats, 47 Rats, 47, 64, 75, 202 Feathers, flank, of Wigeon, 152 Feet of Rats, 64, 65 Femaleness, effect of vigour on re- cessive characters, 48 Females, adoption of male char- acters, 48 necessity of high vigour for, 47, 49 sterility of, 43 high vigour produces pre- dominance of, 165 254 VIGOUE AND HEEBDITY Females, lack of colour in, and sex, 48 larger size of, 161 polyandrous, 160 preponderance of, in Meri- ones litters, 81, 82 proportion of, in Pigeon experiments, 97-112 Fertility dependent on environ- ment, 84 dependent on vigour, 83, 84,85 ■ in Meriones, 85 increased by crossings, 83,95 of hybrid Ducks, 123 of hybrids, 184 Finches, breeding of, stimulated, 192 First recorded instance of Mendelian inheritance in wild animals, 67 Fixity of characters, 180, 181, 209 ; dependent on nourishment of nuclear of ova, 180, 181 Flamingos, 188 FUght of Snipe, 195 Flower, Captain S. S., 88 Fluctuation of vigour, 171 Fluctuations in vigour and Galton's Law, 153 Fluctuating variations, 7 Food and vigour, 11 controls uniformity, 35 supply affects vigour, 17 Fowls, Andalucian, and Mendel's Law, 217 Silky, 197 Fimambulus herdmorei, 24 Fur Seals and fat, 156 Galton, Su: Francis, 138, 215, 246 Galton's and Mendel's Laws both true, 223 ; irreconcilable, 216 ; partially true, 216 Galton's Law, 215, 221 a vigour develop- ment, 220 and Ducks, 137, 139, 153 and fluctuations in vigour, 153 and vigour theory, reconciliation of, 217-223 and wattle of Dra- goon Pigeon, 221 Galton's Law, in terms of vigour, 218, 219 proved, 216 refuted, 217 Gametical changes, 202 inheritance, 220 Gannets, 176, 177 Geddes, Prof. Patrick, 4, 117, 119, 155-159, 162-165, 192, 230, 231 Geddes and Thompson's theory and vigour, 230-283 Geddes and Thompson's theories, Cunningham's views on, 230-233 Genealogical table of Cats, 56 Genealogical table of Ducks, 124 Geographical races caused by isola- tion, 34 races due to vigour, 174 species, vigour and colour, 37, 40 Geothlypis, 21 Germ-plasm and vigour, 1 79 continuity in, 178 Gestation, period of, in Meriones, 78 Goats, experiments on, 52, 53, 55 genealogical tables, 54 horn inheritance, Bateson on, 52 horn inheritance, statistics of, 53 horn inheritance, suggested solution, 55 Miss Wilde's experiments on, 52 Golden Plover, 21 Gonads, 234, 239 Grey Plover, 22 GuUs, 176 GuUs' cries, 195 Guyneot, E., 202 Habits of animals affected by vigour, 172, 173 Hair, length of, in males, 159 Hares, Mouse, 18 Hawks' cry, 195 Health and colour, 17 Henke, Dr. F., 189 Herburg, 179 Heredity and horns of Goats, 52-55 and effects of lowered vi- gour on embryo, 76 and environment, 7, 178 INDEX 255 Heredity, blends of colour in Ducks, 127 in Bucks, 144 of the fawn colour in Eats, 69 Hibernation, 28 and vigour, 44 ^ : of Bats, 43 High vigour and death, 165 produces predominance of females, 165 Hirundo erythrogaster, 20 Hooded Crow, 176 Hormones, 239, 243 Horn inheritance in Sheep, 52 inheritance in Goats, 52, 53, 55 of Cattle, inheritance of, 52 Horned animals and vigour, 228 Goats, 197 Horns, latency and patency of, 228, 229 Hossaok, Dr. William C, 38 House- Sparrow, 191 Human race and vigour, 29, 247 Humidity affects vigoiu:, 17 and vigour, 11 and change of colour, 89 Hybrid Ducks, breeding stimulated, 191 ; coloration of, 125 ■ eclipse plumage of, 127 effect of vigour on colour of, 125 effect of vigour on sex of, 125 fertility of, 128 sexes of, 124 Hybrid, sexless, 185 Hybrids and infertility, 11 between species, 185 blended races in, 67 brilliant colour of, 184 Canary, 9 disappearance of Pintail characters in, 153, 154 down on Duck, 149 excess of males, 46, 46 fertility of, 184 infertility and loss of vigour, 185 infertile, vigour and strong colours of, 48 initial vigour of, 185 Hybrids, loss of fertility in, 185 Mallard - Spotbill - Austra- lian-Meller Ducks, 132- 139 Meriones, absence of Men- delian characters in, 92 Pintail - Mallard -Spotbill, 216 Pintail -Mallard - Spotbill- Duoks, 126-131 Pintail -Mallard- Spotbill- Australian Ducks, 132 P. M. S. A. Ducks and M. S. A. Mell. compared, 138 ■ P. M. S. hybrids discussed, 127-141 size of, 184 Hypothesis, Archdall Keid's, 224, 225 Hypotheses, other, 224-243 Inbreeding, 8 and infertility 9, 10 and infertility in Ducks, 123 and loss of size, 94 and loss of vigour, 94 and vigour, 151 Mendelian expectation altered by, 58 Infertility, 11 amongst animals in con- finement, 173 and inbreeding, 9, 10 and inbreeding in Ducks , 123 in Ducks due to loss of vigour, 123, 124 of animals in confine- ment, 173 of hybrids, 48, 11, 185 Inheritance a combination of vigour and gametical inheri- tance, 220 and environment, 179 and vigour, 246 blended and alterna- tive, 226 horn in Sheep, 52 of " Blends," 150 of colour in Pigeons, 97 of horns of Cattle, 52 of silver colour in Pigeons, 110, 111 256 VIGOUE AND HBEEDITY Inheritance of vigour, 176, 178, 220 Initial vigour, 5, 6, 165 ; of embryo, 167 Intermediate sbades of colour in wild animals, 37 Isolation, cause of geographical races, 34 • controls uniformity, 35 Johnston, Sir H. H.,189 Kamerer, Dr. Paul, 37 Katabolism and anabolism, second- ary sexual characters, 280, 281 and vigour, 157, 158 Kennel, Prof., 160 Knot, moult of the, 15 Larger size of some females, 161 Latency and patency, 224, 225 — of horns, 228, 229 Laulaiue, 235 Lelanie, 119, 121 Lemmings, Siberian, 19 Length of hair in males, 159 Lepidoptera, 166 Linnet (A. cannahina), change of colour in, 33 ' Litters, effect of vigour on number in, 43 Lloyd, Dr. E. E., 38, 228 Loss of characters a sudden muta- tion, 74 from disuse, 72, 73 Loss of size from inbreeding, 94 Loss of vigouj due to inbreeding, 94 in breeding season, 167 in one character sometimes due to increased vigour in another, 94 Lowered vigour, reason of sterility, 76 Low vigour and fat, 171 Low vigour produces predominance of males, 165 Lydekker, Bichard, 24 MacCraken, J., 67 MacCurdy, G. Q., 62 Males, excess of, in hybrids, 45, 46 excess of, in young of captive animals, 46 length of hair in, 159 length of life in, 157 predominance of, caused by low vigour, 84, 165 preponderance of, in Meri- ones litters, 81, 82 restlessness of, 159 > — short lived, of various species, 157 Mammals, coloration of, 16 psychological stimula- tion in, 186 Mallard Duck, 123 Manniche, A. L. V., 189, 190 Management of Meriones, 77, 78 Matings of Chequered Pigeons, 112, 113 Maupas, 94 Mealy Pigeons, 100, 101, 102, 103, 104, 105, 107 Mean vigour, 5, 6 Mellers Duck, 123 Mendel, Gregor, 246 Mendelian character? is sex a, 45, 236 Mendelian character, sex a, 49 Mendelian characters, absence of, in Meriones hybrids, 92; acquired owing to environment, 75 ; deve- lopment of, determined by vigour, 120, 121 ; affected by vigour, 49 Mendelian expectations, deviation from, in colour of Pigeons, 112- 121 ; deviation frona in sex of Pigeons, 100-112 Mendelian hypothesis of sex, 48, 49 Mendelian inheritance among natural races, 67 ; in alexan- drinus and tectorum, 67 ; in Pigeons, 97; of "sports," 146, 148, 153; influenced by vigour, 212 Mendelism, 196 Mendel's and Galton's Laws both true, 223 Mendel's and Galton's Laws parti- ally true, 216 Mendel's and Galton's Laws unre- concilable, 216 Mendel's Law, 2, 39, 41 Mendel's Law, a gamete develop- ment, 220 INDEX 257 Mendel's Law and Andalucian Fowls, 217 and breeders, 246 and chequered Pi- geons, 217 and inbreeding, 58 ■ and Pigeons, 217 and sand - coloured animals, 3 and Silky Fowls, 217 and vigour, 198,220, 236-238, 246 and vigour theory, reconciliation of, 217-223 and webfooted Pi- geons, 217 and wild animals, 67 Cunningham's views on, 286-238 exceptions to, 197 factors modifying, 206 governs suppression of characters not development, 181 inadequacy of in Nature, 3 inadequacy of in practice, 2 modified by vigour, 41, 213 Pigeon exceptions, to, 201 webfooted Pigeons exceptions to, 201 Mendel's theory, 196 Meriones, 46, 101 and colouj , 88 blended races in, 67 breeding season of, 78 change of colour during temperature experi- ments, 86 change of colour in con- finement, 33, 87 crassus, 88 crassus pallidus, 77, 87 crassus selVysii, 77, 87 fertility dependent on environment, 84 fertility dependent on vigour, 84, '85 fertility in, 85 hybrids, 87 Meriones hybrid, absence of Men- delian characters in, 72 increase of fertility by crossings, 83 increase of males where vigour is low, 84 management of, 77, 78 measurements of, 90, 91, 92 number of yoimg depend- ent on vigour, 83 number of young in litters, 83 period of gestation, 78 preponderance of females in litters, 81, 82 preponderance of males in litters, 81, 82 sex proportions, 79, 80, 81 skull characters, 92, 93 Metabolism, 230 and colour, 41 and vigour, 244 inherited, 178 rate of, 157 rate of, determines sex, 164 Moths, change of colour of, under different temperatures, 33 Moult and vigour, 15, 20, 21, 22, 23 Moult of the Knot, 15 Mouse Hares {Ochotona), 18 Muscles of birds, 241 Mus flamicollis, 68 jerdoni, 68 norvegicus, 175 rattus, 75 sylvaticus, 68 Mutation and loss of characters, 74 Musk Ox, 18 Mutations, 204 Mustela flwoigula, 25 gwatkinsi, 25 Natural Selection, 207 and colour, 16 ■ and vigour, 245 uniformity at- tained without, 85 Nature of animals colours, 27 Nature's safety valves, 157 Newman, H. H., 67 S 258 VIGOUE AND HEREDITY Newton, Prof. Alfred, 123 New species and climate, 174 Nightingale's song, 195 Nuclei and protoplasm of ova and spermatozoa, 179 Nutrition and vigour, 244 colour a reserve of, 27 effects of, 162 increases anabolism, 162 Objections to vigour theory, 159, 160, 161 Ochotona Ladacensis, 19 Boylei, 19 OpistJiocomus, 24 Oriental Bats, variability in same locality, 39 Squirrels, variability in same locality, 89 Ova and spermatozoa, comparative vigour of, 166 excess of, produced by lower animals, 163 nourishment of nuclear of, and fixity of characters, 180, 181, 209 Ox, Musk, 18 Oxidisation and vigour, 30 Paired characters and vigour, 226, 227 Pairs of characters, 207 Patency and latency, 224, 225 Penguins, 19 Photograph of Dun Pigeon feathers, 211 Physiological activity, 246 Pigeon, breeding of male Dun, 112 Dun, photograph of feathers of, 211 exceptions to Mendel's Law, 201 experiments, proportion of females in, 97-112 Pigeons, 97 and sex, 98, 99, 100, 101 and sex limitation, 108 Chequered, 212 Chequered and Mendel's Law, 217 Chequered, matings of, 112-113 chequering in, 205, 206 colour of, 41, 112 Pigeons, cross between Dragoon and Homer, 216 deviations from expected Mendelian results, 100- 121 experiments on, 46 inheritance of colour in, 97 matings of Chequered, 112, 113 Mealy, 100-107 Mendelian inheritance in, 97 sex and colour in, 99, 100, 116 sex and vigour, 102, 103, 104, 105 sexual dimorphism in, 117 ■ Silver Chequeres and Pure Blue, 200 silver colour in, 198, 199, 200 silver colour in, due to lack of vigour, 210 silver colour inheritance in, 110, 111 silver coloured, and vigour, 200 wattle of Dragoon, 219 wattle of Dragoon and Galton's Law, 221 webfoot character in, 97, 98 webfooted, 2, 197, 201 webfooted, and Mendel's Law, 217 webfooted, exceptions to Mendel's Law, 201 -white, a sign of low vigour in, 114, 115, 116 - Wild Eock, 205 Pintail, disappearance of, characters in hybrids, 153, 154 Duck, 13 Pintail - Mallard - Spotbill hybrids, 216 Plumage, eclipse in Wigeon, 152 eclipse, of Wigeon and vigour, 152 eclipse, of hybrid Ducks, 127 Pocock, B. I., 64 Polar Bears, 157 Polar circles, colour in, 18, 19 INDEX 259 Plectrophenax nivalis, 20 Polar region animals, 17 regions, birds of, 20 Polyandrous females, 160 Prepotency, 7 Presence and absence of characters, 227 Proportion of females in Pigeon experiments, 97-112 Protoplasm and nuclei, 179 influenced by vigour, 179 Prout, Louis B., 67 Psychology of reproduction, 186 Psychological stimulation in birds, 187 ; in mammals, 186 Punnett, E. C, 109 Eaees, crosses between, 185 BainfaU controls uniformity, 35 Kate of metabolism, 157 ; determines sex, 164 ; inherited, 178 Eats, Black, 176 blind, 64, 71, 75, 202 breeding of, 63 colour and sex in, 14 colour of, 65 colour of hind feet of, 64 colour of under parts of, 65 fawn, 47, 64, 75, 202 fawn-coloured Alexandrian, 47 fawn-coloured and heredity, 69 fawn, eyes of, 69 length of hind feet of, 65 Mendelian inheritance in, 67 Mus alexandrinus, description of, 61, 65, 67 Mus norvegicus, constancy of, 39 ; fawn-coloured, 64 Mus rattus, breeding of, 61 ; causes of variability, 88. Mus tectorum, description of, 61, 65, 67, 71 ; grey, des- cription of, 61 notes on management in cap- tivity, 61, 62, 63 Oriental, variability in same locality, 39 results of experiments on, 64 variation in, and vigour, 203, 204 waltzing, 64, 71, 74, 202 Eavens, 18 Eecessive characters, change from dominant, 227 Eed flank feathers of Wigeon, 152 Beid, Dr. Archdall, 121, 201, 203, 204, 224-229 Beid's, Dr., theory and vigour, 226 Beindeer, antlers of, 160 Beproduction, alternative, 197 — psychology of, 186 Bestlessness of males, 159 Eetrogression and vigour, 228 Beversion, 204 and vigour, 183, 213 • of domestic races to wild type, cause of, 213 BMzomys badius, 23 pruinosus, 23 sinensis, 23 sumatrensis, 23 Sciurus maclellcmdi, 24 Seals, 44 Secondary sexual characters, 230, 231 Selection, sexual, 233 Seligmann, C. G., 238 Seth-Smith, David, 160 Sex, a Mendelian character ?, 45, 236 and colour, 13 and colour in Ducks, 125 and colour in Pigeons, 99, 100, 116 and colour in Eats, 14 and Pigeons, 98-101 and vigour, 12, 45, 156, 244 and vigour in Pigeons, 102, 103, 104, 105 determined by rate of meta- bolism, 164 determination of, and vigour, 119, 120 determination of, in embryo, 119 effect of vigour on, 42, 43, 44 effect of vigour on Mendelian characters, 49 effect of vigour on sex of young, 46 evolution of, 155-185 excess of males in young of animals in captivity, 46 in Pigeons, deviation from Mendelian expectations, 100-112 260 VIGOUE AND HEEEDITY Sex of hybrid Ducks, effect of vigour ou, 125 of young, effect of vigour on, 43 proportions in Meriones litters, 79, 80, 81 limitation, 167, 200; in Pigeons, 108 Mendelian hypothesis of, 48, 49 of Chickens, 164 theory accounting for less colour in females, 48 Sexes of hybrid Ducks, 124 Sexless hybrids, 185 Sexual characters, secondary, 230, 231 characters, anabolism and katabolism, and second- ary, ,230, 231 dimorphism, 233; in Pigeons, 117 selection, 233; Darwin's, 233-236 somatic characters, 229 Shattuok, S. Q., 238 Sheep, horn inheritance, 52 Short lived males of various species, 157 Siberian Lemmings, 19 Silky Fowls, 197 ; and Mendel's Law, 217 Silver Chequer and pure Blue Pigeons, 200 " Silver " colour inheritances in Pigeons, 110, 111 Silver colour in Pigeons due to lack of vigour, 201 Single characters and vigour, 227 Size, loss of, from inbreeding, 94 Skeat, Prof. W. W., 123 Skull characters of Meriones, 92, 93 Smalley, F. W., 97, 112, 200, 205 Smith, Geoffrey, 234, 236 Snipe's flight, 195 Snowy Terns, 177 Somatic expression controlled by vigour, 207, 208 sexual characters, 229 Song, an expression of vigour, 198, 194 and colour in birds, 193, 194 and vigour, 159, 192 Nightingale's, 197 of birds in autumn, 172 Sparrow, House, breeding of, 191 Species and vigour, 151 Spermatozoa and ova, comparative vigour of, 166 " Sports," 204 and vigour, 246 " Sports," Mendelian inheritance of, 146, 148, 153 Spotbill Duck, 123 Squirrels, 68 Squirrels, Oriental, variability of, 89 Stability of colour in nature, 31, 82 ; in tropics, 25 Standard of vigour in domestic animals, 211, 212 Standfuss, Dr. Max, 88, 166 Staples-Brown, K. C, 109, 205 Sterility, 43 and vigour, 44 lowered vigour the cause of, 76 of species in confinement, 14 Stimulation, breeding of Finches, 192 from environment, 188, 189 — psychological, in birds, 187 psychological, in mam- mals, 187 Stint's flight, 195 Strasburger, Julius, 179 Summary, 244 ; of Chapter X., 149 Sutton, Prof., 119 Table, genealogical, of Cats, 59 Table, genealogical, of Ducks, 124 Table, genealogical, of Goats, 54 Table of vigour of domestic hen, 170 Table of vigour of Knot, 168 Table of vigour of Polar Bear, 169 Table of vigour of Wild Eat, 170 Table of vigour of Wild Duck, 168 Table showing heredity of fawn colour, 70 Tadpoles, 166 "Temperamental" differences, 230 Temperate regions, colour of species in, 25 Temperature affects vigour, 17 and change of colour in butterflies, 33 and change of colour in Meriones, 86 and change of colour in moths, 33 INDEX 261 Temperature and colour, 22, 23 and vigour, 11 controls uniformity, 35 effects of, 162 increases vigour, 162 influencing colour, 19 " Tendency " and vigour the same, 178 Terns, 176 ; cries, 195 Terns, Snowy, 177 Theory, Cunningham's, 229, 288 Theory, Darwin's, Cunningham's views on, 223-236 Theory, Geddes and Thompson's, Cunningham's views on, 230-233 Thompson, Prof. J. A, 4, 117, 119, 151, 155-159, 162-165, 192, 280. 231 Time of breeding of Arctic species, 19 Time of breeding season dependent on vigour, 163 Tortoiseshell Cats, 221-223; male, 55-60 Tringa canutus, 15, 20 Tropical species, colour due to vigour, 23 Uniformity attained without Na- tural Selection, 35 Uniformity dependent on tempera- ture, rainfall, food, and isolation, 35 Variability of domestic races, 208 ; of colour in domestic species, 80, 82 Variability of Eats, 38, 39 Variability of Squirrels, 39 Variation and domestic animals, 210, 211 Variations due to vigour, 32 Variations, fluctuating, 7 Variations in domestic species, 172 Varieties of animals and vigour, 175 Vigour, a development of Galton's Law, 220 affected by food supply, humidity, temperature, 17 affecting colour in Cats, 223 — and abnormal varieties in animals, 17a Vigour and sestivation, 44 and breeding seasons, 168 and climate, 18 and climate influencing colour, 19 and colour, 16 and colour and sex of hybrid Ducks, 125 and colour of Dogs, 52 and colour of Polar ani- mals, 18 and Cunningham's theory, 238-243 and Darwin's theory, 233- 286 and death, 165 and development of gametic characters, 95 and development of Men- delian characters, 120, 121 and Dr. Keid's hypothesis, 226 and eclipse plumage of Wigeon, 152 and eclipse plumages, 151, 152 and environment, 167, 244 and fat, 28 and fertility of Meriones, 84,85 and food, 11 and Galton's Law, 153, 218, 219 and gametical inheritance, 220 and Geddes and Thomp- son's theories, 230-233 and habits of animals, 172, 178 and hibernation, 44 and horned animals, 228 and humidity, 11 and hybrids, 185 and inbreeding, 151 and increase of males in Meriones, 84 and infertile Ducks, 128- 124 and infertile hybrids, 185 and inheritance, 246 and katabolism, 157, 158 and Mendel's Law, 198, 220, 286-238, 246 and metabolism, 244 262 VIGOUR AND HEREDITY Vigour and moult, 15, 20, 21, 22, 23 and number of brood in Merionea, 83 and nutrition, 244 and Natural Selection, 245 and oxidisation, 30 and paired characters, 226, 227 and predominance of female births, 165 and retrogression, 228 and reversion, 183, 213 and sex, 12, 45, 156, 244 — ■ and sex in Pigeons, 102, 105 and shades of colour in geographical species, 37 and silver coloured Pigeons, 200 and single characters, 227 — and somatic expression, 207, 208 and song, 159 and song and colour in Birds, 192, 193, 194 and species, 151 and sports, 246 and sterility, 44 and strong colours in in- fertile hybrids, 48 and temperature, 11, 162 and " tendency," 178 and the human race, 247 and tortoiseshell Cats, 57 and variation in Bats, 203, 204 and white colour in Pigeons, 114, 115, 116 and zones, 9 and zygotes, 245 cause of bright colours in tropical species, 23 cause of dark-coloured animals, 42 cause of geographical races, 40, 174 causes change of form, 177 causing change of colour, 177 compared to a dynamo, 159 compared to fire, 158 compared to steam, 157 controlling variations, 32 Vigour converting dominant into recessive characters, 114, 115 definition of, 5 direct cause of change of colour, 26 effect of lowered vigour on embryo inherited, 76 effect of, on colour, 21, 22, 24 effect of, on Mendelian characters, 49 effect of, on moulting birds in Polar regions, 19, 21 effect of, on number, colour, and sex of young, 43 effect of, on recessive char- acters, femaleness, 48 effect of, on sex, 42, 43, 44 — effect of, on sex of young,' 46 effect of reaKzation of law of, 247 expenditure of, in breeding season, 162, 163 expressed in song, 193, 194 fluctuation of, 171 governs germ-plasm, 179 hypothesis reconciling Gal- ton's and Mendel's Laws, 217-223 in man, 29 influence of, on embryo, 76 influences Mendelian in- heritance, 212 ; cell pro- toplasm, 179 influencing determination of sex, 119, 120 inherited, 178 initial, 5, 6, 165 initial, in embryo, 167 is inherited, 178 lack of, and silver colour in Pigeons, 201 loss of, due to inbreeding, 94 loss of, in breeding season, 167 loss of, in one character, 94 low, and fat, 171 low, produces predomin- ance of males, 165 lowered, reason of sterility, 74 — mean, 5, 6 INDEX 263 Vigour modifying Mendel's Law, 41, 213 necessity of high vigour for females, 47, 49 of domestic hen, table, 170 of Knot, table, 168 of ova and spermatozoa compared, 166 of Polar Bear, table, 169 of Wild Bat, table, 170 of "Wild Duck, table, 168 shown by colour, 19 size and appearance, 245 standard of, in domestic animals, 211, 212 theory, objections to, 159, 160, 161 theory reconcilable with Cunningham's, 239 "Waders, 176 "Wallace, Dr. A. R., 4, 117 "Wallace, Prof. Alfred, 233 "Waltzing Bats, 64, 71, 74, 202 "Waste product, colour a, 27 "Webfoot character in Pigeons, 97, 98 "Web-footed Pigeons, 197, 201 Pigeon exceptions to Mendel's Law, 201 Pigeons and Mendel's Law, 217 Weisman, August, 119 "Whales, 72 "Wheldale, Miss M., 30 "White a sign of low vigour in Pigeons, 114, 115, 116 "Wigeon, eclipse plumage and vigour, 152 "Wigeon, red flank feathers of, 152 "Wild animals and fixity of char- acters, 210 "Wild and domestic animals, differ- ences between, 208, 209 "Wild Cattle, 36 "Wilde, Miss Maude, 52, 54, 55, 60 Young, effect of vigour on number, colour, and sex of, 43 Yung, B., 119, 166, 235 Zones and vigour, 9 Zygotes and vigour, 245 INDEX TO TABLES AND CHARTS. Changes in vigour in domestic Fowl, 170 Changes in vigour in Polar Bear, 169 Changes in vigour in "Wild Duck, 168 Fertility in Meriones, 85 Galton's Law in terms of vigour, 218 Genealogical Table of Ducks, 124 (jonealogical Table of hybrid Ducks, 124 Inheritance of fawn colour in Eats, 70 Inheritance of Mus rattus teotorum and Mus rattus alexandrvnus, 66 Matings of Cats, 56 Matings of Bed Setter and Black Retriever Dogs, 51 Number of young in litters of Meriones, 83 Pedigree of grey variety of Pintail- Mallard Spotbill Australian Ducks, 146 Pe(iigree of white variety of Pin- tail-Mallard - Spotbill - Australian Drake, 148 Pedigrees of Goats, 54 Pedigrees of Tortoiseshell Cats, 59 Percentage of males in litters of Meriones, 85 Photo of feathers of Dun Pigeon showing effect of vigour on colour, 211 Possible carrying of the dilute recessive through four genera- tions of Pigeons, 108 Proportion of colours and sexes in mealy experiments, 106, 181 Proportion of males amongst Pigeons, 100, 101, 103 Proportion of sexes in different lit- ters of Meriones, 80, 81 Proportions of sexes and colours in Pigeons, 118 Sexes of Mealy Pigeons, 99 Showing relation between colour and sex in Ducks, 13 Skull measurements of Meriones, 93