The Law of Heredity - a Study of the Cause of Variation, and the Origin of Living Organisms, Second Edition, Revised

MALE AND FEMALE WOOD DUCK, TO SHOW SECONDARY SEXUAL CHARACTERS.

LFronL photographs of slufled specimens in the Collection of the Maryland Academy of Sciriices 1

THE

LAW O F HEREDITY. A STUDY O F TEE

CAUSE. O F VARIATION, AND THE

O R I G I N OF

LIVINGORGANISMS, BY

W.

I(. BROOKS,

ASSOCIATEIN BIOLOBY, Joms HOPKINS UXIVEBSITY.

BALTIMORE AND NEWY O R E : JOHN XURPHY & CO., PUBLISHERS.

COPYaIOaT,

1883,

BY

JOHN IklURPHY & CO.

TO

THEMEMORYOF

CHARLES DARWIN, FROM WHOSE STORE OF PUBLISHED FACTS I HAVE DRAIVN NOST OF THE MATERIAL FOR THIS

.

V 0LU M E

PREFACE. The subject which is tre:itcd in this book has occupied my thouglits for ten years or more, but I have refriiined from pubIishing my views, as I hope that I may some time be able to submit them to the test of experinlent. Many experiments have suggested theniselves t o me, but as most of them involve the cultivation ar,d hybridization, for many generations, of such animals arid plants as mill thrive aiitl multiply in confinement, they can only be carried out by some one who has the niems for experimental researches, arid who lias also a permanent home in thc country, where organisms of many kinds may be kept under observation for years, and where many specimens of hybrids between various wild and domesticated spec& can be reared to maturity. My own studies hare been in a different province of natural science, and it has therefore seemed best t o publish this volume in order to call renewed attention t o this most fascinating subject. I hare little hope that my views will be permanently accepted in the form in which they are here presented, but I do hope that they may Berve to bind together and

viii

Preface.

to vitalize the mass of facts which we already possess, and that they may thus incite and direct new experiments. If this book should serve to turn the attention of others into this channel, and should thus ultimately help us to a clearer insight into the nature of the forces which have acted, a d still act, to guide the evolution of life, this result will far outweigh the acceptance or rejection of the epculations which we here advanced.

CONTENTS. CHAPTER I. WHAT IS HEREDITY

?

The development of an animal, with the complex and beautiful structural adjustments, the instincts, habits, and individual traits of its parents is one of the most wonderful phenomena of the material universe-Heredity is not due to the external conditions which act upon the ovum, but to something within the ovum itself-The phenomena of reversion-Asexual and sexual heredity-Possibility of a n explanation of heredity-Characteristics which are now hereditary were at one time new variations-Heredity and variation are opposite aspects of the same problem-We may hope that a more perfect acquaintance with the laws of heredity will remove many objections to the theory of natural selection.. .. ... .. . . .

. .. ..... . . ...... ... . .....

6

CHAPTER 11. HISTORY OF TEE THEORY OF HEREDITY.

Requisites of a theory of heredity-Historical sketch of speculation on heredity-Evolution hypothesis of Bonnet and Haller-Ovists and spermists-Modern embryological research has sbown that it is impossible to accept the evolution hypothesis in its original form-Buffon’s speculations upon heredity fail to account for variation-Hypothesis of epigenesis-This hypothesis is logically incompleteThe analogy between phylogeny and ontogeny gives no real explanation of the properties of the ovum-Haeckel’e plastidule hypothesis-This hypothesis is not logically complete unless it involves the idea of evolution-Jager’s hypothesis-Ultimate analysis shows that this is at bottom an evolution hypothesis-No hypothesis of epigenesis is satigfactory-No escape from some form of the evolution 16 hypothesis-This conclusion is accepted by Huxley

......

x

CONTENTS.

CHAPTER 111. HISTORY O F THE THEORY OF HEREDITY-(&?LtZhUed).

Some form of tlre evolution hypothesis a logical necessityDarwin’s pangenesis hypothesis-This is an evolution hypothesis, since all the cliaracteristics of the adult are supposed to be latent in the germ-Miscellaneous objectious to it-These objections do not show that it contlicts with fact-Difiiculty in imagining detailed working is no reason for rejecting it-Galton’s experimental disproofThcre are m m y reasons for believing that the sexual elements have differetit functions-The evidence from partlietiogeuesis-Polar-cell hypothesis-The evidence from hybrids, from variation, and from structures confined to one sex-The pangenesis hypotliesis recognizes no such difference 10 tlie functions of’ the reproductive elementsWe must therefore distrust its absolute correctness-Sum47 mary of last two chapters

.............................

CHAPTER IV. A NEW TEEORY OF HEREDITY.

The objection to the liypotliesis of pangenesis would be almost entirely removed if it could be simplified-statenicnt of a new theory-Heredity is due to the properties of the egg-E:rcli new character lras been impressed upon tlie egg by the transmission of gemmnles-Tendency to form genimules is due to the direct action of external conditions-The ovum is the conservative element-The male cell is the progressive element-This theory has features of resemblance most of the liypotheses wliich have been noticed-It fills most of BIivart’s conditions also-It is not necessary to assume that the ovnm is as complicated as the adult-There are many race characters w u c h are not congenital-There are many congenital characters which are not hereditary-Direct action of external conditionsOur theory stands midway between Darwin’s theory of natural selection and Lamarckianism.. 80

.................

CHAPTER V. ON TEE OPINION THAT EACH SEX MAY TRANSMIT ANY CHARACTERISTIC WHATEVER.

The argument from hybrids-This argument is inconclusive -The argument from the homology between the ovum and the male cell-Homology does not involve functional similarity-The argument from the dual personality of each individual; from reversion; and from polymorpliism -These phenomena admit of a simpler explanationQ9 Summary of chapter..

.................................

CONTENT8.

xi

CHAPTER VI. THE EVIDENCE FROM HYBRIDS.

Importance of the subject-It

fUrRiSlJeS a means Of analyzing or isolating the influence of each sexual element-Hybrids very variable-Hybrids from domesticated races more v;~ri:tblethan those from wild races-The descendants of b y brids more variable than the hybrids themselves-The offspring of a male hybrid and the female of a pure ppecies are inucli more variable than those of a feniale hybrid and the i n + of a pure species-These facts inexplicable on any view. except the one liere presented-Reciprocal crosses-They differ in fertility and in structure-The difference is exactly what our theory requires-Difficulty in expliniug transmission of characters witiioiit fusiouIteversion caused by crossing-Two kinds of reversion.. . .... 118 Summary

.......... .... .. . ........ ....

.........

CHAPTER VII. TEE EVIDENCE FROX VARIATION.

Causes of variation-Changed conditions of Iife induce variability-No particular kind of change is necessary-Varinbility is almost exclusively confined to organisms produced from rertilized ova-Bud variation very rare-History of the Italian orange-The frequency of variation in organisms produred from sexual union, as compnred with its infrequency in those produced asexu:dly, receives a direct explanation by our theory of heredity-Bud variation more frequent in cultivated than in wild plants-our theory mould lead us to expect this-Changed conditions do not act directly, bnt they cause subsequent generations to vary-Tendency to vary is hereditary-These facts perfectly explicable by our theory-Specific cliaracters more variable than generic-Species of large genera more variable than those of small genera-A part developed in an unusual way liiglily variable-Lnw of equable variation -Secondary sexual characters variable-Natural selection cannot act to produce permanent modification unless many individuals vary together-Our theory is the only explanation of the simultaneous variation of many individuals-This theory also simplifies the evolution of complex structures-Saltatory evolution-This is explained by our theory of heredity-Correlated variation of homologous parts-Parts confined to males more vari. able than parts confined to females-Males more variable 140 than females-Summary of last two chapters..

... .. .. ..

CHAPTER VIII. THE EVIDENCE FROM SECONDARY SEXUAL CHARACTERS..

... 166

Xii

CONTEiVT8.

CHAPTER IX. TEE EVIDENCE FROM SECONDARY SEXUAL CHARACTERE coxT1NUED.-TEE CAUSE OF THE EXCESSIVE MODIFICATION O F MALE CIIARACTERS.

The explanntion of Daines Barrington and Wallace-Reasons for considering it inadequate-Darwin’s cxplanationHistory of domesticated races shows t l m this does not go to the root of the matter-The view that the male js more exposed than the female to the action of selection-A more fundamental explanation is needed-This is furnished by our theory of heredity-Special difficultiesSummary 207

............................................. CHAPTER X.

THE EVIDENCE FROM TEE INTELLECTUAL DIFFERENCES BETWEEN MEN AND WOMEH..

............................

2&

CHAPTER XI. TEE THEORY OF HEREDITY CONSIDERED AS SUPPLEMF,HTARY TO THE THEORY O F NATURAL SELECTION.

Darwin believes that variations are purely fortuitous-Natural selection cannot give rise to permanent race modifications unless many individuals vary in nearly the same n’ay, at about the .same time-The chances against this are very great if variations are fortuitous-Argument from North British Review-Dakwin acknowledges tlie great weight of this objection-It is removed by the theory of heredity -The co-ordinated modification of complicated organsThe time demanded by Darwin practically infiniteMurphy’s argument from the complexity of the eyeHerbert Spencer’s illustration-Our theory removes this difficulty-Mr. Conn’s objection-Saltatory evolutionEvidence that it occurs-Spike horn buck-Ancon and Mauchamp sheep-Black-shouldered peacock-The theory of heredity accounts for saltatory evolution-Parnllel variation-Evidence of its occurrence-Evolution of the medusse-General and special Homologies.. 275

.............

CHAPTER XII.

......................

RECAFITULATION AND CONCLUSION..

3ia

HEREDITY, CHAPTER I. WHAT IS HEREDITY?

The development of an animal, with the complex and beautiful structural adjustments, the instincts, habits, and individiial traits of its parents is one of the most wonderful plienomcna of the inaterial universe-Heredity is not due to the external conditinnq wliieli act upon tbe ovum, hut to something within the ovum itself-The phenomena of reversion-Asexual and sexual heredity-Possibility of a n explanation of heredityCharacteristics aliicli are now hcreditaiy were at one time new variations-Heredity and variation are opposite aspects of the same problem-We may hope that a more perfect acquaintance with the laws of lieretlity will remove many objections to the theory of natural selection.

TOthe ordinary nnscicntific reader the word heredity may perhaps suggest nothing more than s few curjoiis cases where an odd peculiarity of the parent has been transmitted to the children, or i t may recall the hereditary transmission of a tendency to certain diseases, or the mental or moral idiosyncrasies of the parents. To the breeder of domestic animals or plants it has a somewhat wider significance, and recalls the transmission by choice or fancy breeds of the features which give them their value. To him heredity is the law which enables him t o modify his animals and t o build up and perpetuate new varieties.

6

Heredity.

T o the naturalist, on the contrary, the word is filled with deep meaning, and instead of recalliIig to liis mind a few odd eases, tlie tricks and accidents of heredity, i t briugs before him the most i n a n ellous of all the plienomena of the material universe: the production from a simple egg of a living animal, with tlie intricate structure and complex bodily and mental functions of its proper species. Thoughtful men i n all ages have regarded the strncture and faculties of tlie higher animals as :I proper field for life-long study. Yet the acute intellects, thc powers of patient observation and profound reflection which generations of naturalists hare brought to this fascinating subject, have not yet given l i b iL complete knowledge of the life of a singlc animal. I n crery agc and country wlicre science Iins flourished men liave devoted tlieir lives to this subject, and h v e felt that tlieir lisrdly-enrned resnlts could scarcely be called :I beginning. So vast is tlic field, so many are the plienoniena, that tlic povincc of natural science is practically infinite, for each animal and each plant presents special problcms which open out in all directions before tlie student in an endless vista. Wonderful and various as tlic attributes of each :minial are, Iiowevcr, they ;ire not myste:.ious; for, at tlio same time t1mt we discoiw in an organism tlie power to do wonderfiil things, TVC also find in i t :Lmaterialorganization, a Inccliaiiism, adopted to do these very thiiigs. It is true tliet we cannot perfectly understand this meclianism, tlmt i n nmiy cases we fail completely i n our attempts to trim its working, i t n d that in most cases our insight is very crnde indeed. Still me are able to sliow that the niacliinery exists; and anatomy, or the study of structure, goes hand inhand with the study of the bodily

WXat is Heredity?

7

anil mental :tctivitics of animals. We do not underEtarld the ni:teliinery, but we find that i t is there, and we c:m intcrrulbt it5 work by obstructing or iiijnring it. Oar rvondcr is iiot :I feeling of mystery, a sense that tlic l)lienornc~i:itranscend kiio~ledgcj i t is due to a perceptioii of tlie :tmouiit of knowledge required. We rcgard an :tdult anirnal with feelings siniilsr to those with mliicli an intelligent savage might regnrd a telephone or a steamboat. A dog, with a11 the powers and faculties which enable him to fill his place as m;tii’s companion, is a wonder almost beyond o w powers of expression; but we fiiid in his body tlie machincry of muscles and reins, digestive, rcsi)iratory, and circulatory organs, eyes, ears, etc., mliicli adatpts liim t o his place; and study lias taught us ciiough about the action of this machinery to ‘‘1-Bsure us greatcr knowledge woiild show 113, in tlic structure of t h e dog, an esplmation of all that fits the dog for Iiis Iifc; an explanation :13 satisfactory as that which a s:ivage might rcacli, in the case of tlic steambod, by studying its anatomy. Lct our savage find, however, while studying an iron stcambo:it that small masses of iron, withont structure, so far as tlic mc:m at his command sllow liim to examine sad decide, are from time to time broken off arid thrown o v e r b o d , and that each of tIicse contains i i i itself the power to build up all the macliiiiery and applimces of a perfect stemiboat. The wonderful thing now is not the adaptation of woiiderf id m:icliiiiery to produce wonderful rcsults, but the prodiiction of wonderful results without any discoverable meclianism; and this is, in on~line,tlie problem which is brought before tlic mind of the nltturalist by t h e word heredity. Every o m knows that each dog exists at some time

S

flerediity.

as an egg, and the microscope shows in this egg no traces of the organs of the dog’s body or of anything a t all like them. So far as oiir means of examination go the egg is 110 more likc a dog tliaii the mass of iron is likc a stcamboat. I t may be said, though, that the dog’s egg is not left to itself, but is fertilized and is carried inside the body of the mother until the new animal is matured; that i t is there iiourished and built up from substlances supplied through the body of a full-grown dog; t h t it may be acted upon a t this time by agencies mliicli liare a direct tendcncy to build up out of it an organism like the parent; that the egg does not wtually contain a 110tential dog, but simply supplies the proper material to be acted upon by the surrounding conditions, and that the structure of tlic new animal is due to tlicsc conditions; that the embryo becomes a dog because it is bathed by a dog’s blood, nourished through a dog’s body, and is complctcly surrounded by influences which arc peculiar t o dog nature. Those persons who arc not nucumlists derive their knowledge of the animal world chiefly from our common domestic animals, and to such persons tliis explanation may seem probable; but naturalists, with wider experiencc, know that animals which carry tlicir young inside their bodies arc exceptions, and that tlie organization of thc future animal must cnist potentially in the egg, since the conditions to which it is exposed cannot possibly have any tcndcncy to prodrice from it :L being ml~ichdoes not already exist, in some form, within it. A bcc is almost as wonderful as a dog; its anatomical structure is cxqiiisitely dclicnte and complex, and every one is acquaintcd with the wolidcrful work mhich it accomplishes. A t the time it is laid tlie egg which is to become a worker-bee contains no ~iisibletrace of it6

What is Eieredity?

9

body, or of anything like it. It has becn carefully stndicd with all t h e resources of modern science, but examination shows nothing within it which is more like a bce than a mass of iron is like a n iron ship. This egg is not evcii fertilized, but i t dcvelops into a perfect worker, with all its wonderful structure and instincts, by virtue of something which it contained when i t left the ovary of its mother. It is true that it is not left quite to itself, but is carcfiilly attended and caredfor by other bees; but ercrytliing which they d o for i t might be done just as well by dc1ic:tte machinery, and t!ie attention has no tcndency whatever to miinufacture a bee. Proper heat a n d access t o air are :is necessary as attcntion, and attention hns no more power to prodnce it bee than air or heat. No onc who is familiar with marine anim4s can believe for a n instant t h t tlic conditions to wliich an egg is exposed liavc anything whatever to do with the character of tlic animal t o which i t gives rise. We map artificially remove eggs from the ovaries of severs1 dillercnt m i m d s , fertilize them :Lrtificially, a i d tlicn p h c c thcni together i n a tnmbler of fie%m:iter, and expose thcm to exiictly siniihr e s t e r i d conditions, yct c : d i one will follow its own d ~ ~ t ~ ~ i i conrse, i i ~ i e dand wc nxty rear in tlio Fame tumbler of wtrter from eggs which :&re h d l y diatinguishalle animnls which liare less in common than a dog :md a bird. Jf tlierc is no mystcry in the p~rform:tnceby the complicated organs of : ~ nadult aniiml of all its complicated IiinctionS, mliat slid1 me sny mlicn we find the ~ O W Wto perform thcsc functions cxisting i n a 1:ttent state in the egg, witliont tlio corresponding organs? Thi3 is tlic problem of heredity. I n thc mind of the natur:dist tlic word calls np t h e greatest of all tlic wonders of tlic niatcrial universe: the cxistcnce, iu a simple,

10

Heredity.

unorganized egg, of a power to produce a definite a d d t animal, with all its characteristics, even down to the slightest accidental peculiarity of its pauents; a power to reproduce in it all their l i d i t s and instincts, and even the slightest trick of speech or action. This is by no means the whole of the problem of heredity. One of the most interesting plicnorncnil coilnccted with our subject is what is known as rcversion, or the appearance in the child of peculiarities which were iiot present in either parent, but are diic to inlieritsnce from a grandparent or a more remote mcestor. An interesting illustration of this law is the occasional appearance in horses of stripes on tlie body and legs. Such stripes are not usually present in the horse, although Darwin has given reasom for believing that our liorscs :&redcseended from a striped zebra-like ancestor. The power to revert to this :tnccstrd form is handed down from generation to generation in the egg, acd i t may show itself a t any timo by the production of a striped colt. &version is, in a certain sense, exccptional, but it is not a t all rare, and me must add this power t o thc wonderful properties of the egg. Darwin gives the following case, which mill serve to illustrate the nature of reversion: A pointcr bitch produced some puppies; four were marked with blue :uid white, which is so unnsual :I color in pointers that she was thought to have played false with onc of the greyhounds, and tlie .whole litter '(VRS condemned, but the gamekeeper mas permitted to s aw one as a curiosity. Two years afterwards a friend of the owner saw the young dog, :md declared that lie mas tlic image of his old pointer bitch, S:tpph, tlie only blue :mtl white pointer of pure descent which he had eyer seen. This led to close inquiry, and it was proved that he was the great-

WiLat is Heredity? -

11 ~-

great-grandson of S:~pplio; SO that, according to the comnion expression, he had only one-sixteenth of her blood in liis reins. Another aspect of our subject must be kept constantly in mind. Among tlic higher aniin;rls heredity usually maiiifests itself oiily by what is known as sexual reproduction,-that is, the dcvelopment of iiew individnala from fertilized eggs; but in the lower forms of life another k k d of reproduction, the development of new inclividuds by bndding or by analogous processes, is even more common. Among the liydroids lierecljty may manifest itsclf by the formation of ncw :mimals, widi all tlic cliaracteristics of the parent, on almost any part of tlie body of the latter, and in certain plants the smallest fragmcnt of tissue may become a new and perfect p!mt, e:tpable of producing otlicrs in the same may or by seeds. The most surc and rapid way to get iicw sea-anemones is to tear an old one to pieces. As a rule this power i, confilled to the lower forms of life, but certain animals wliidi are by no means low or simple in strnctnrc multiply ;~scxu:tlly,and the offspring thus produced inlierit, like those developed from eggs, all the cliaracteristics of Clie parent. This then is the problem of hercdity, certainly one of the grandest secrets of nature. When we reflect upon its obscurity and complexity we may fairly ask what hope tlicre is of discovering its solution; of reaching its trno meaning, its hidden laws and causes. If i t is true lliat, i n each egg, all the functions and faculties of a dcfiliite mature allinla1 lic hidden, withont any corresponding organs, must we n o t regard heredity as a mystery too grc:it for solution; as something which must be acccptcd as it is without scientific explanation? Thirty years ago the adaptation of each organ of axi

12

IIeredity.

adult anim;il t o its proper purpose seenied to be a m:stery of the snnie hind. and ni:tny profouiid tliitiI\ers $:it isfied themselves and t:hiIglit otl1er3 that this adaptat ion iviis n o t broiiglit ahout by the laws of matter and 1)y secondary canses; t1i:it it must be accepted in itself, without explanation, and that the methods o f yhysical science are liere of no use. Darwin’s work has taught 11s that this is not trne; t h a t in the law of natnml Election we luivc a t least a p a r t i d cxplanation of the origin of the adaptiltioiis of nature; that while natural sclcction may n o t be the exclusivc means by ~vliichthey have been produced, it is, so far :is i t goes, :I true scientific explanation, for it cien pats i t iu our power t o produce, in domestic animals, similar adaptstions to special purposes, by the selectio~iof t h e fit test r a i h t ions. Darwin, in his first and in all his later bools on the snbjcet, pointcd out that liis discovcry did not complete the solntion of t h e problt3m; t h t ’ I natui*~il sclcctiori is ii great h i t not the exclusive means of motlification.” Tlic greatest valne of liis work lies in the yroof ivliicll he lias furnislicd, tlint tlic origin of the structure of animals is not be~oiidour rcacli, but that obscrwtion and reflection, tlie mcaiis ivliich 1ioi.e unlocked for 11s so many of the secrets of inorganic natnre, arc eqiuilly uscful in this fieltl; that tlic adaptations of nature nisy be studied and understood liltc a problcm in astronomy or physics. The aim of this work is to sliow that the eamc thing is trno of the problem of heredity. We may n o t be able. as yet. to pciietrdc its secrets to their inmost dcptlis, I ) a t I 11ol)et o d10w that observnt’ion and reflection do enable US to discoier sonit3 of the l a m upon which heredity depends, and do farnish us with at

What is Heredity?

13

least a partial solution of the problem; that we have every reason to hope that in time its hidden causes will all be made clear, and that its only mystery is that which it shares with all the phenomena of the universe. In this introductory statement we have presented one side of the problem of heredity: the transmission from parent to child of the established congenital hereditary characteristics of the race. We must not forget, though, that there is another aspect which is fully eqnal to this in importitnce. We know that each characteristic has been gradually acquired through a long series of modifications; that all thc wonclcrful adaptations which fit animals to their surroundings, and meet their particular needs, have been evolved step by step by the natnral selection of the fittest congenital variations. Each racecliaracteristic has at one timc been a new variation, and the process of modification is still going on and perfecting the harmony between the structure of each organism and its needs. No theory of heredity has any value unless it explains the may in which new features, which may become hereditary, continnally make their appearance as congenital variations, a t the same time that it accounts for the may in which established peculiarities are handed down from generation t o generation. The problem is two-sided; what is nom hereditary vas tit one time variation, and each new variation may soon be hereditary. Heredity and variation are opposite aspects of the same thing, and an explanation must be examined and tested on the one side, as well as on the utlier, before i t can be accepted. There is still another conoideration which remains to be noticed. Darrrin has never failed to perceive, and he has frequently pointed out,that the law of natural selection is not

14

Heredity.

complete explanation of the origin of species, and that it is exposed t o certain very serious difficulties.

zb

Still he concludes that the theory is supported by such a mass of evidence that we may fairly believe that our own knowledge, not natural selection, is a t fault, and that furtlier research will remove tlie difficulties by tlie discovery of otlier laws. Naturalists all over the world have acknowledged the justice of this claim, and some, less candid arid broadminded than Darwin, seem to have even lost sjglit of the difficulties. Kom natural selectiou can act only by the preserratiori of such variations as chalice to appear, and until we know the laws which govern the appearance of variations it miist be iinpossible to decide how far the course of OF ganic erolution has been determined by these uuliiiown laws, and how far by natural selection. We may tlicrefore entertain a rensonable hope, tliat when the true tlwory of heredity is disco\-ered, it mill, by revealing to 11s the laws aiid causes of variation, place the law of natural ,.election upon a firmer basis, and E ~ O W that its apparent diEEcnlties are simply due to the aarrow Iiniits of our knowlcdgc. K i t h this introduction I mill pass t o the discussion of our subject, tlie riatiire of heredity. The attempt t o generalize from the whole field of natural science is bcset with many difficulties, s ~ c the e field is so vast that an attempt to give in advance a statement of all tlie facts upon wliich rcasonii~gis based would simply confuse the mind of the render, arid burden liim with a mass of detail. I t seems best tlien to start with the generalizntions which are believed to bind the facts together, so that the reader may then approach the specific proofs with more

WLat is Heredity ?

16

interest. The latter method is open to objection, since the reader may bc called upon to listen to views which are opposed by accepted authorities, and to wait until the proofs are presentcd in due course. I must therefore request the reader to suspend judgment, and to lay aside established opinions, until he has examined the subject upon all sides. The examination of the history of the subject will furnish an introduction to its scientific discussion, and I have therefore adopted the following plan: 1 shall give, first, an outline of the chief hypotheses which have been proposed, from time to time, as an explanation of heredity, with reasons for rejecting them. I shall then present briefly, i n outline, a statement of what I believe to be the true explanation. I shall then try to show that this theory furnishes a basis or foundation for the theory of natural selection, and removes the most serious difficulties which have been urged against thc latter theory. I shall tlien show that there is no a priori reason for rejecting the theory of heredity; and that i t furnishes an explanation of many well-known facts which cannot without it be seen in their true relations. I shall then attempt to show that i t is supported by direct proof, and finally I shall give a statement of the theory in a more extended form.

CHAPTER 11. HISTORY OF THE THEORY OF HEREDITY.

Requisites of a theory of heredity.-Historical sketch of spcculation 011 heredity-Evolution hypothesis of Bonnet and 1131ler-Ovists and sperrnists-Modern embryological research 118s sliown that it is inipossible to accept the evolution liypothesis in its original form-Buff on’s speculations upon heredity fzil to account for variatien-Ilypotlicsis of epigenesis-This liypotliesis is logically incompletc-The analogy between pliylogeny and ontogeny gives uo real explanation of the properties of the ovum-IIaeckel’s plastidule hypothesis -This liypotliesis is not logically complete unless it involves tlie idea of evolution-Jager’s hSpotliesis--ultimate analysis sliows that this is at bottom an evolution hypothesis-No hypothesis of epigenesis is satisfactory-No escape from some form of the evolution hypothesis-This conclusion is accepted by Huxley.

9 1. Requisites of a theory of 7~eredity. The following list is a brief summary of what seem t o me the most important characteristics of the reproductive process in living things: 1. New organisms may be produced by tlie wrious forms of asexual generation and from ova. 2. Ova may develop, in certain cases, without fcrtilization. 3. As a rulc the ovum does not develop iiito a new organism until it has been fcrtilized by union with a malc cell. 4. The ovum and male cell will not unite unless t h y are derived from organisms with the same or nearly the same systematic affinities.

History of the Theory of Heredity.

17

5 . T h e new org;inisni, mliether produced sexually or asexnully, is essentially like its ancestors, although it m:Ly be quite different from its immediate ancestor, a s i n cases of alternation. 6. Orgiinisms prodnced from fertilized ova differ in t h e following points from those produced asexually: a. As a rule the development of the egg embryo is indirect, and a more or less complicated metamorphosis or alternation of generations must be passed tliroiigli before tlic adult form is reached, and t h e circuitous p a t h thus trarersed bears a resemb1:mce to the line of evo111tion of tlie species. An organism formed ascxmlly traverses o n l so ~ much of this path as remains to be traversed by the orgrnnism which gives birth to it. 6. Reversion, or the appearance of characteristics not exhibited by the parents, but inherited from remote ancestors, is not a t all unusnal in egg embryos, b a t it is niore rare in those prodnced asexually. c. New variations, or features which are not inherjtcd, appear con tinnally in organisms prodnccd from fez= tilized ova, and they may be transmitted either sesnnlly or asexually to future generations, thus bcconiing establislied as hereditary race-cliarncteristics. Hereditary variations are extremely rare in organisms produced asexually. 7 . Tlie ovnm and the male cell are homologous with each other, and are morphologically equivalent to the other cells of the organism. We must therefore believe tlint their distinctive properties ha1.e been grdiially acqnired, and that their specialization has been bronglit about by the action of the same laws as those in accordance with wliicli tlie other specializatjons of t h o organjsm have been produced. 8. Changed conditions do not act directly, b u t they cause subsequent generations to vary.

18

Heredity.

3. In the higher animals, where the sexes h a ~ clong been separated the male is more variable thaii the female. 10. The result of crossing is not the same wlien crosses are made reciprocally. 11. The sex of the parent-species affects the degree of variability of hybrids; arid when a hgbrid is used as the father, and either one of the pure parent-species, or a third species, as the mother, the offspring arc more wriable than when the same hybrid is used as the molhcr and either pure parent-species or the same tliird species as the fdher. There may pcrliaps be other requisites wliicli should be included in this list, but I think there can be no doubt that a theory of heredity must recognize and be in harmony mith all mliich are here given. $ 2. A sketch of the history of speculation on ihs timry of heredity. The laborions rcsearclies of the students of tlie science of embryology hare yielded a ricli harvest of valnable facts, and we now know that the process of cell division by rvliich an unspecialized unicellular egg becomes converted into a many-celled, highly-specialized organism bears the closest resemblance to the process of growth or of ordinary cell-mnltiplication. We know that all the various forms of reproduction, cell-multiplication, fission, gemmation, conjngation, sexual reproduction, and parthenogenesis, are inter-related in such a way that me must believe that they are different manifestations of the same power, and that they have been evolved one from the other. We know that direct development, metamorphosis, and alternation of generations are not separated from each other by any hard and fast line, and we know too that

Bistory of the Theory of Heredity.

19

~-

the changes through which the embryo passes on its road from the egg to maturity show a wonderful parallelism to the series of changes tlirough which tlie organism has passed during the liistory of its evolution from lower forms. These results are well worth the labor they have cost, and they illustrate, more clearly than any otlier facts in biology, the common naturc of 311 living things. They do not, however, contribute directly to a clearer insight into the lams of heredity. IIere we are still compelled to go beyond the visible plienomena, and to attempt by tlie scientific w e of the imagination to discover the as yet unseen relations d i i c h bind them together. As me.entcr upon this snbject it will be well to bear in mind tlie wide diffcrencc between the end we lime in vicnr-the discovery of the secondary laws of heredityand llie attcrnpt t o undcrstand its ultimate cause. The power to reproduce itself, to imprcss upon dead inorgmic niattcr its own distinctive properties, is oiie of the fund:imcntal cliaracteristics of living matter; and mliile im may hope that increase of knowlcdge may some day enable us to trace the origin of this power, such an attempt forms no part of our present undertaking. We shall accept without explnnation the fact that living matter does thus reproduce itself, and we shall coniinc oursclvcs to the attempt to discover why tlie egg of a star-fish for instance, reproduces a stnr-fish, and the egg of a bee a bee; to discover the origin of the differenccs betmecn the various forms of reproduction, rather than the cause of what they have in common. Tho plicnomena of heredity in the higher animals, as well as the niechpiism of ova and male cells through which these phenomena are manifested, hwe certainly

20

Neredity.

been produced by slow modification, throngh the inflnenc2 of conditioris which arc to a great extent open t o study. ‘I’hc attempt to trace tlieir origin and significiincc is not a pure speculation, but a legitiniate exercise for the scicntific intellect. As we should expect from the fascinating nature of the subject, there has been no lack of speculation in tlic past, and various hypotheses have been proposed from time to time t o account for the phenomemi of heredity. These hypo theses diEer greatly aniong themselves, but they may be rouglily classed as epigeiicsis ligpotlieses, and evolution hypothesis: the word evolution being hcre used, of course, in its old sense, as contrasted with epigeucsis. The liypotliesis of evolution, pnre and simple, as advocated by Bonuet and Ballcr, is that thcrc is contained in the egg or seed or in tlie male element a perfect but minutc orgmi3in, and tliat tlic subsequent development of the cgg is simply tlic “evolntion” or unfolding of this germ. Up to tlic end of the last century the prevailing opinion was that each egg contains, in alatcnt or dorinant state, a completcly formed o ~ p n i s n i . The fertilization of the egg was supposed to awaken this dormant germ, t o call its latent potential life into activity; and the process of devclopmcnt was regarded as the iinfolding and growth of the already fully formed and perfect embryo. The embryo m:is held to be not prodnced by, but simply anfolded from the egg, and tlie act of reproduction was therefore regarded as eduction riot pi*odiiction. According t o Hnxley (Eneyc. Brit., Art. Evolution) “Boiinet affirms that before fecnndstion the hen’s rgg contains an excessidy minutc but complete chick, and tliat fecundation and incubtition simply cause this germ t o absorb nutritions matters, which are deposited in the Snterstices of the elementary structure of which the min-

IZstory of the Theory of Heredity.

21

i.iturc chick or germ is made np. The consequence of tiiis intiissusceptive growth is the “ devclopnient” or “evoliitioii” of this germ into tlie visible bird. Tli~is nil orgnnizcd individual “is a composite body consisting of tlie original or elcmentary parts, and of tlie rnattcrs wliicli have been associated with tlietn by the aid of nutrition,” so t h a t if tliese matters could be extracted from the individual, it monld, so to spealc, become c b x n t r a ted in a point, and wonld thus be restored t o its prirnitivc condition of a germ “fjnst as b y extracting from a bone the c;ilcarcoud substance mliicli is tlie source of its hardness it is reduced to its primitive state of gristle and mcmbrane.” cLEuolzitio~zand develnpmeitt are, for Bonnet, synonymous terms; and since, by evolution hc means simply the expansion of t h a t which was invisible into visibilitv, lie was natnr,:lly led t o the conclusion, at wliicli Lcibnitz liad arrived by a different line of reasoning, that no such thing as generation exists in nature. Tlic growth of a n organism bcing simply a process of enlargement, as a particle of dry gelatine may swell iip by the intussnsccption of miter, its death is a shrinkage, such as the nielted jelly niiglit undergo on desiccation.” M,~nchmore anciently the erolution hypothesis found acccptance i n a somewhat different form. and the rniniature organism W:LS believed to exist in t h e male element, a i d t o receive from t h e cgy t h e nonrishmenb needed for its growth and perfect dcvelopmcnt. Leeuivenhoek’s discovery of the motile spcrnmtozoa of animals was reg:irdcd as a new basis for this view, anJ the “ sperrIi-aiiiniiilcule ” w:is held to be the perfect and living animal ready for unfolding or evolution, tlie term “sperniatozoon,” still retained in scientific nomenclature, being a remnant of this old hypothesis. Leeuwenhoek’s

22

€leered ity.

discovery inaugurated, in tlie first lialf of the last ccntn-

ry, the warm dispute between the Aninidculists and the Ovists, one side holding that the germ is contained in the egg, and the other that it exists as tlie seminal animalcule. It is obvious that, in either form, the evolution theory, as above stated, is logically incomplete, siiice it only accountsfor a single generation. Its advocates mere therefore compelled to enlarge it, and to assume that, as each organism thus exists, in a perfect form, in the prcccding generation, each germ must contain, on a still smaller scale, the perfect germs of a11 subsequent generations. Thus Bonnet held, in his hypothesis of eniboitement, "that a11 living things proceed from pre-existing germs, and that these contain, one inclosed within tlie other, the germs of all future Iiving things; that nothing really new is prodnced in the living world, but that the germs which develop have existed since the beginning of things." (Huxlcy, Evolution.) The advocates of thc evolntion hypothesis appealed to such facts n5 the presence Qf u minute plant inside the acorn, or to the butterfly inside the pupa-skin, in srrpport of their views; but the hypotliesis, i n its crude state, was quickly overthrown by the first discoveries of modern microscopic embryology. Harvey's studies on the development of the chick, followed by the researches of Wolff, Pander, Von Baer, and thehost of embryologists of the l a t fifty years, show concInPiveIy that the embryo is not nnfolded out of, but gradnally built np from the egg. We now know that the eggs of all animals, mhen t h y are not complicated by the presciicc of food, or of peculiar coverings for protection or attachment, are essentially alike in optical structure, and that they are not only like

History of the Theory of Heredity.

23

each other, but like tlie constituent cells of ail parts of the body of the organism. Fiir from being preformed in tlie egg, we know that the body is built up gradudly, step by step, by repeated cell-division, and that tlie earlier stages of development do not result in tlie formation of the parts of the perfect body at all, but that they simply give rise to germ-layers, or tracts of cells out of which organs’ are gradually formed, and that cells which were a t first quite widely separated in the embryo may come at last to enter into tlie formation of a single organ. For instance, wlien tlie nervous system of a vertebrate first makes its appearance in the embryo, there are no traces of the brain, of the spinal cord, of the nerves or of sense organs. It at first consists of a long group of cells running along themiddle line of the body, and presenting no difference from the otlier cells of its surface. I n most cases this elongated group of cells bccornes converted into a furrow, and afterwards into a closed tube, tlie nerve-tube, by tlie folding together of its edges. Tlie primitive nerve tube is at first simply a long tube of embryonic cells running along tlie middle line of tlie back, and it is a veIj different thing from tlie find nervous system of an adult mammal, nor is it in any sense a mammaliua nervous system in miniature, for tlie changes by wliicli i t becomes converted into the latter are great arid numerous, as well as gradual. Certain parts, such as the eye, are fornied only in part from this tract of ce!ls, for tlie vertebrate eye is the result of the combination of an outgrowth from tlie embryonic nervo:is system and an ingrowth from the surface of the head. The whole history of tlie nervous system and sense organs is thus seen to directly oppose tlie view that these organs are present in miniature in the germ.

24

Heredity.

Still more opposed t o the hypothesis of evolution is tlie remarliable fact tliat tlie clianges which take place i n tlie derelopingegg are not sucli :ts would lcad directly to the formation of the adult animal. I n most c a m a circuitous or indirect path is followed, a i d this indirect path leads at first towards tlie adnlt form of lower members of tlie group. This, the most suggestive fact of modern embryology, may perliaps be made clearer by an illnstr at’ion. Let us try to c o m p r c the growth of an egg into an adult animal with tlic growth of some manufactured prodiict in the 1iaiids of its m‘‘I 1-Ier. Tlie evolutionist I iew of the clewlopment of an organism may be illustrated by tlie tnannfactnre of a yarn base-ball. A boy, wishing to makc n y a m ball, procures, if possible, a small rnbber ball, and winds his yarn on to this until tlie desired size is reached, tlie only clianges during tlic growth of the ball being the change of size and of material. The observed facts of embryology show that tlie developmcnt of an. enibryo does not take p1:ice in any sucli vay as this. It may, liomrcr, be illnetrated by the growth of a ste:im-sliip in tlic hands of the builder, wlio first lays down an indefinite skclcton, arid outlines i n a vague my the niorc prominent featiires, before any of t h e details are finished. 111order to make tlie illustration perfect, howerer, n-c must imngine the builder to commetice work upon liis steam-ship by 1;ij ing oiit tlie skeleton of a big triareme; me must imaginehim tocnrry this some stiiges towards completion, and to p u t into it certain contrirauccs, snch as i~owers’-benches, ~vliicliare of 110 nse i n a steam-ship. We must imngiiie that lie then abanclons liis plan, tcars down his benclies, a11d nEes tlrc material to mukc a deck; that he cliaiiges 11ic shape and

IIistory of the 7’JLeory of Heredity.

25

proportions of his Iiull a little to fit i t for sailing instead of rowing, tliat he piits in masts and spars, and makes everything rcndy for a sliip’s rigging; that he then changes the shape of liis hull once more; tears out part of liis cabiil, puts i n bnlklieads, coal bunkers, and an engine and boiler ; shortens 111s masts, alters his rigging, and fiunlly converts liis unfiiiished ship into a finished stcamer. This is not by any means a forced illnstration, bnt a very fair outline of tlic dc~elopmcntof an animal. In nearly every case TVC Gild that the development of tlic embryo as :I,n~hole,or else the d e d o p m e n t of certain organs, takes place in this ronndabout, indirect way, and repeats, nsnally in :in imperfect manner, the structnrc of a related but lower animal. As an example, we may refer to tlic history of the bIood-vesscIs of a m:inimnl. The breathing organs of tlic lower vertebrates arc gills on the sides of the neck, and tlic venous blood is driven from the heart through a scrics of branchid srtcries to the gills, where it is aerated and conveyed into a scries of brancliial veins which carry it, not back to the heart, but to the various organs of the body. In a mammal thercare no traces of gills at any stage of development; the adult animal brentlies by lungs, and tlic blood which has been aerated in the lungs goes back to the heart; before it is distributed throughout the body. Now the early stages in the d e v c l o p c n t of the bloodvcsscls of a mammal would, if carried out to completion, Icad to thc formation of the system found in fishes. The manimalian embryo has no gills, but it does have bmncliial arteries and veins, and its blood at first follows tllc sainc conrsc tlmt i t follows in a fish. I t is plain h a t the fishlike circulation is not an outline or sketch Df that of a niammal; that i t is not R ncccssary stage in

26

Heredity.

the formation of the latter, for tbe branchial vesscls are soon, in part pulled down and destroyed, and in part profoundly modified, in order to conform to the mammalian type. Cases of this kind are almost univcr~nl,and the law of resemblance between the early stages of higher mimnls and the adult condition of lower animals is a fundamental law of embryology. I t is obvious that the hypothesis of evolution of a ~ m * fectly formed germ contained in thc egg, is utterly irreconcilable with this lam, and me may therefore state with confidence that this hg'pothesis is refuted by the observed facts of embryology. We must not forget, hon-erer, that there were other less superficial forms of the evolution hypothcsjs, and that these cannot be d i s p r o ~ dso easily. Buffon, for jnstance, held that the embryo is built up by the union of organic particles mhicli are given off from every part of the body of the piirent, and which, assembling in the scxnal ~ e r r c t i o i i assume ~, in tlie body of the offspring positions like those wlricl~they occupied in the parent. This is e~sentiallyan evolntion hg'potliesis, bnt it is logically complete, since it acconntsfor the prodiiction of siiccessive generations witliout the ncccssity for assuming that they were all contained in cmbryo i n thc body of a reniote ancestor. Microscopic exnmination cannot orerthroiv this hypothesk, for a f;tilnre to diseowr these organic p"rtic1es -with an? payiicular mngnifying power does not, of coiirse, dispro~etheir existcncc any more than a failrire to see them without a microscope. Although Bnffon's h ~ p o t h e s i sdoes n o t nccount for the fact that derelopment is indirect in most cases, t h a t the egg does not build up the adult aninial in the simplcst way, but takes a roundabout circuit, this fact is 112t

History of the Theory of Heredity.

27

directly opposed to liis liypothesis, for .we can easily conceive that after an indirect method of deve1opme:it Ii:is been establ;~l~ed i t might be perpetuated by E~iffm’s orgnuic niolectil~,provided tliese are given off by the p r c n t organism at all stages of its life, and not simply after it lias reaclied its final form. Tlicre is, however, another class of phenomena of even greater importance-the 1)lienomcna of variation. Buffon’s liypotlicsis accoimts for the rebemblaiice between tlic cliild and tlie parents, but we now know that tlic cliild is not exactly like its parents or eyen inidway between tliem, tliat nnimds aiid plants are born with tl tendency to vary, that this variation may affect any part of tlic body, and tliat by tlie selection of tliese congenital varistions the most profound changes of hereditary strusture may be produced. Tlie fact of congenital variation is as profound, as universal, and as cliaracteristic of living things as tlie fact of heredity, and tlie constant appearance of new variations is as fatal to B~iffoii~s liypotliesis of evolution as it is to tlint of Bonnet. With tlie growth of tlie modern science of morphology tliese liypotlieses have been abandoned arid tlie liypothesis of epigenesis almost universally accepted in their place. This hypothesis, first bronght into notice by the researches of Harvey and Wolff on tlie development of tlie chick, lias gradually assumed a more definite shape with the progress of embryology, and has been especially mod~ficdby the growth of the cell theory. In its modern form i t may, for convenience of discussion, be divided into two parts-a statement of the observed facts, and an explanation of the origin of the phenomena.

28

Her edit y.

So far as it is a statement of facts, it cannot be called an hypothesis, for it simply affirms that t h e rgg is optically an ordinary unspecialized cell; that i t gives rise, during tlie process of segmentation, in q miinircr vliicli is identical with ordinary groivtli by cell division, to a number of cells mliicli gradually become specialized for ccrtain functions, and are set apart as tlie foundations of tlic various organs of the body; that the repetition of this process gives riae, a t last, to the perfect body of the mature animal; that the reproductive elements which are to give rise to the nest generation, originate, like d.11the cells of the body, by cell division during the process of development, and that they are simply cells specialized for’ tlie reproductive function as other cellsare specialized for other functions. Every one who has the slightest acquaintance with modern biology will accept this statement, not as an hypothesis, but as an obserwd fact, and will agree that between this and tlie old evolution hypothcsis there can be but one choice. The old liypotliesis of evolution, lionww, c1:iimccl to be sometliing more than a statcmcnt of fact, for the presence of the germ within the cgg accounted for tlie wonderful properties of the egg itself. We are at once compelled to ask, then, horn, on the hypothesis of epigcnesis, has the egg acquired these properties ?‘ If it is simldy an nnspecialized ccll; if, as cfegenbaucr states, “ the egg is notliiiis more nor less thim a cell; tlie egg-cell does not diffcr from other cells in any essential points” (Conip. Anat., Bcll’q Trans., 13. 18), how can tlie cgg of a horse develop into a horsc, wliilc another cell, mliich “does n o t differ from i t in any cssclltial points,” develops into a bee or an alligator or an oyster ? Nothing in nature is more marvellous than the devel-

History of the Theory of Neredity.

29

opmcnt of each egg into its proper organism, and if i t is true tliat tlie egg which is to give rise to a man differs in no essential point from that wliicli is togive rise to :in insect, mc m iy conclude tliat the mystery is too great to be fnthomcd by oiir intelligence, and we may fairly ask what possi blc explanation can, on this hypothesis, bc given of tlic wondcrful properties of tlic egg. The answer wliicli has been given, and wliicli seetna t o linve been tlionglit satisfactory by many students, is tliis: We know, from a mass of evidence mliicli is constantly and rapidly increasing, and to wliicli each new observation adds cumnlntire weight, tIiat the various forms of life liave been slowly evolved, during long ages, from older and simpler forms; tliat as me trace back tlie history of any two animals or plants we find eridence tliat in tlic past tliey liad for a common ancestor a species mliicli liad iiot yet acquired tlic distinctive fenturcs of either of them; that a little fiirtlier back we trace tliis s p c i e s to :m :tiicestor with still midcr relationships. Every day tlie evidence grows strongcr to sliow that morc complete knowledgc will id timately provc that tlie same tliing is true of still larger groups ; tliat families, classes and ordcrs of organisms I i a ~ ebeen formed in the samc wiry by gradual modification and divergence ; tliat complctc liiiowledge of tlic ancestry of any organism would Icad us back tliroiigli simpler and simpler forms to a remote unspecialized nnicelliilar ancestral form. It is unnccessary to review in this place the evidence for this conclnsion, for t h e fact that i t is fully acccptcd by those best qiialified to judge of its truth, is perfectly familiar to all students. Now it is said, and tlie explanation is pretty generally accepted, t h a t since any particular organism, a horse for

30 Instance, lias been slowly erolred from an nncestral rliizopod, and siiicc tlie ovum of a liorsc is Iiomologous with a rliizopod, or is morpliologically eqiiirulent to it, me have iu tlic gradual phylogeiietic erolutioii of tlie horsc species from an unicellular ancestor, a satisfactory cxplanation of tlie ontogcnetic derelopnient of tlie individtial liorse from an unicellular ovum. As soon as atteiition is fairly fixed upon tlie subject, tlie meakriess of this cxplaiintion bccoines so evident that I take tlie liberty of m:iliing the follorviiig qnotation from a well-known antlioritp, in order to sliow that the explanation has been soberly advanced. I n making tlic extract from Haeclicl’s writings I am not actuated by a desire to attack liis views, for tlie same idea can be found, expressed pretty definitely, in the works of many other writers, and tliis particular selection is simply a matter of convenience. 1i:ieckel says : ‘‘ Uiitil reccutly the greatest students of embryology, Wolff, Baer, Reniack, Schleidcn and tlie whole school of embryology founded by tlieni, have regarded tlie science as cxclusively the study of individual development. Far otlierwise to-day, ~5 lien tlic mysteries of the wonderfnl history of tho development of indiriclual organisms no longer face 11s as an iiicompreliensil~le riddlc, hut Iiave clcarly revealed their deep sigiiificancc : for the clianges of form which tlie germ p s s e s tlirongh under oar eyes in a short time are, by tlie lam of inheritance, a condensed and shortened repetition of tlie corrcspondiiig changes of form mliicli the ancestors of tlic organism in question have lmsed through in tlie coiirse of many million years. To-day, wlien me lay a hen’s egg in an incubator, and in twenty-one days see t h e cliick break out of it, we no longer gaze in dumb wonder on the marvellous changes which lead from the simple egg

IZistory of the T h o r p of IIeredity.

31

to the two-layered gastrula, : from tliis to tlie worm-like and skulless gcrm, and from this to later stages wliicli rcpcat, csseiitially, tlie organization of fish, ampliibiun, reptile, until a t last we liave a pcrfcct bird. On the contrary, me unravel from this history the corresponding series of ancestral forms, which linve led u p through tlic a m a h , tlic gastraa, the worms, tlie acrania, t h e fiJics, the ampliibia and tlic reptiles to t!ic bird. ‘‘ Tlic scrjcs of c1i:~ngcs i n tlie llrii’s cgg gives u s a n ontlinc s!cetcli of tlie series of ancestors. This ancestra2 oi*pJi!jlogeiictics i ~ ~ ~ i ~ f i c?f n rthe ~ cplieiioineua c of ontogeny or iiitliridzud decelopnieizt i s up t o the present time the only explaiiatiou, of the latter.” (“ Gcsanirnelte Pupulsire V o ~ t r i i g e , I[., ~ ~ 1’. 103.) ‘‘ Any one who ‘accepts tlie law tlint individual dewlopmcnt is a rccapitnlation of the evolution of tile species toill find it simp1y siatwal tlint tlie microcosm of tlic ontogenetic cell-tree should be tlie dimiuntivc, and i n part distorted, reflection of the tnacrocosni of the pliylogcnetic gcnca!ogical tree of tlie species.’! (“ Gesammelte Populiirc Vortrige,” II., p. CS.) No one can set too liigli a d u e upon the scientific law liere expressed-that individual dcvclopment is a recapitulation of the history of the evolution of the species. It must be regurded a3 one of tlic greatest generalizations of modern science, bot I do not think i t is possible to agree with IIaeckeI that with its discovery tlie mystery of individnal dcvclopment lias clearly revealed its deep significance, and iio longer faces us as a riddle. I t may be trnc tliat i t is “ simply natiiriil” t h a t t h e egg of a horse should recapitidate tlie ancestral history of horses, and tlic cgg of a bird the ancestral history of birds, but the statement tliat this is tlie case is in no sense an explanation of heredity. For that matter it is

Her edity . “ simply natural ” that a bird’s egg should give rise to a bird, and a horse-ovum to a horse, but n o one mould accept the statement as an eqdanation. We have in the natural selection of variations a true ezpZnnatiol8 of the manner in which an unicellular rhizopod lias bceii slowly and gradually modified by an almost iiifini te number of slight clianges, extending through countless millions of generations, into a bird. The clinngc is one of the most monderful of tlie phenomena of nature, but i t is in no sense a mystery, for the skill of the breeder may even now, by the employment of tlie same means, produce similar results, only on a much smaller scale ; by tlie nietliodical selection of congenital variations an organisni may be, in a few gctierntions, sliglitlg modified in atiy desired direction, aiid we can fairly and truly a E r m that we understand tlic evolution of birds from their aniceIlular aiicestors ; biit the resemblance between the evolution of birds from tliesc remote ancestors by natural selection, and tlic derelopnierit of an individid bird from an uniccllulnr ovum, is simply an analogy. I t is true that i t is an aimlogy of the grentest significance, but we must not lose sight of the fact that the nicaiis by which the end is acconil~lisheci-the natural selection, through a long series of gcncrations, of congenital variations-is absent i n the second case. If the epigenesis hypothesis is true, if the egg is simply, like the rliizopod, an unslmializecl cell ; if the egg of a bird docs not differ from tlie egg of a star-fish in any essential points, we rnnst aclriioivletlgc that the mystery of individual derclolinicnt is iiot only as yet unsolved, biit absolutely insoluble. The student at the sea-shore may collect at the siirface, with his dip-nct, t h e e similar traiisparent splierical eggs. Each of these is, optically, simply a nucleated

Nistory of the Theory of Heredity.

33

cell, and each mlicn placed undcr tlie microscope will 60011 bc seen to pass through almost exnctly the same cliangcs, giving rise by division to a spherical h j e r of cells. Yet i1 these three eggs are placed together in a tnmblcr of water and exposed t o identical coiiditions, one may a t last become a star-fish, anotlier a crustaccau, and anotlicr a vertebrate. Similar tliings tinder similar conditions cannot give rise to widcly diffcrent results, and there seems no escape from tlie conclusion that tliesc threc eggs are not similar, or eTcn cssentially alike, but tliat one of tlicm is a potcntial star-fish, anotlicr a potential crnstaccan, and a third tl potential vertebrate. Tlitit tlicrc is in cach of them tl something which scparatcs it very widcly from the otlier two, and determines its future histoiy. ‘i’lie liypotlicsis of epigcncsis proves, then, on careful annlysis to be as unsatisfactory as the speculations of Bonnet and Bnffon, and mc must acknowledge tliat we are as yet unablc to picture to onrselves the hiddcu significance of tlie plieiioniena of individual dcvelopmcnt, witliout returning to sonic modification of the old evolution hypothesis. The attempt to escape this necessity, 2nd. to hold fast to the hypothesis of epigenesis, lias given rise witliin recent years to much* ingenions speculation, and an examination of some of tlie published papers will help, rather tlian retard, our argnment. Among tlicse, o m of the most jiigcnions and snggcstivc is IIacckel’s paper, “ Ueber die Wellcnzciignng der Lebenstlieilclien oder die Pwigcnesis dcr P1:tstidulc.” Tlie following extract ( I ‘ Gcsammclte PopulBre Vortrigc,” II., pi?. 66-72) mill, I hope, give a sufficiently clear statement of his views: “ I n order to penetrate still farther into t h e mechan-

34

Heredity.

ics of the bingenetic process, we must descend into the deep obscurity of pl~~stid-life, and search for its trne efficient cause in tlic motion of organic molecules (Plustidule-BeweSrutiS). “ I n fine, this question remzjns ta be anstrered, Are we in a position, by tbe aid of coniparison mitli annlogous plienoniena, of motion, to form a satisfactory provisional hypotliesis regarding the trne nature of the plastidiile motions wliicli are hidden from onr direct observation? Oar hypothesis of pcrigenesis is an attempt to answer this question in tlie affirmative. “As we reviem, from tlic IiigIiest and most cornprehensi-rc point of view, the sum of tbc phnomcna of organic der;elopmeut, the most general resnlt of our survey is the conclusion that t h e biogenetic process is a periodic motion, which we can best pictore to on;.selves as P wave motion. Adliering a t first to facts which are beyond dispute, and which admit of direct observation, we may commence witli our own ancestry: either confining ourselves to the so-called historic period, in wliicli we can pass from man to man by direct proof; or else following, by tlio methods of anthropogeny, our ancestry still further back, through the Tertebrates to amphioxus, and tlirougli the group of invertebrates to the gastrm, and nt last to the amceba and the moner. I n either case the course of development (entwickelungsbemegung) of onr series of ancestors can be most simply represented by a wave-line, in whicli tlie individanl life of each orgauisnz answers to a single wave. ‘‘ If now me enlarge our field of view t o embrace not simply our own direct ancestry but the whole of our blood-relations, we can make clear by a genealogicul tree their relatioiisliip to each ohher. As the history of the

IILstory of the Theory of Heredity.

35

cT-olution of enell person is represented by a wave-line, the entire tree will have the form of a branched mavemotion, Q ramified undulation. “ A nntural system of classification is notIiing but a gcncalogieal tree of allied species of organisms, a n d each branch a n d twig of tlie tree corresponds to a greater or smaller group of descendants from a common ancestral form. ‘f’his commuidty of descent unites all the forms of a class, an order, and so on. Since each class is divided into various orders, each order into several fitniilies, each family again into various genera, each genus into a number of species a n d l-arjeties, there is a similar brancliing i n tLe wave-motion wliicli is carried from the eoniiiion aceestral form to tlie entire group of its descendants; and each iindulating branch iniplanCs i n the same way its individual motion on its various descendants. “Now the fundamental lam of embryology jeaciies us that this liistory of t h e phylogenetic evolution of organisms is mirrored in miniature in tlie ontogenetic development of each individual. Here the single waves answer to tlic life of the constituent plastids (cytodes a n d cells). T h e cytiila, or tlie first segmentation cell mliicli originates from tlie fertilized egg, and out of wliicli the muny-celled organism is developed, bears the same relation to tlie various cell-generations which originate from i t by division, and which give rise later by specialization of function to the vnrions tissues, that the stcmform of a class or order bears to tlie various families, genera and species whieli diverge from it, and wliieli have been diffcreiitly evolved through adaptation to diversified conditions of existence. ‘“I!lie ontogenetic ‘cell-tree’ of the former bas exactly the same form as the phylogenetic ‘species-tree’ of t h c

...

36

EIeredity.

lattcr. The developing impulse which in tlie one case is trsnsfcrrcd from the ancestral species to t h e entirc group of species, and in tlic other case from the anccstral cell to tlic cntirc group of cells, assumes in both cases tlic same forin of a branching wave-motion. Any one wlio accepts the fundamental law of d e v c l o p c n t Trill find i t oiily natural that the microcosm of the ontogenetic ‘ cell-tree’ should be a diminution, and to some degree distorted reflection of the pliylogenetic ‘speciestrce. ’ “As we can only explain and render intelligible a complicated atid obscurc plienorncnon by dividing i t into its separate elements, and by tlic exact analysis of theso p u t s , so i t is necess:iry to licnetratc to t h e nltimate elemcntary f : t t s of our nicclianical tlicory of developlllcn t. “ Now tlic biogcncfc process as a whole is the liiglily compouncl resnltant of the developmental history of cll species of- organisms. These consist :(gain of tLe life histories of the individuals, just as tlic latter are again made up of the histories of thc constituent ldustids. “ T h e development of each plastid, howcvcr, is in its t u r n only the prodnct of tlie active rnovemcyts of its constitocnt plastidules. Now we 1iaT-e seen that the devclopmcntal impulse of the branches and classes, tlic orclcrs and families, the genera and epccics, the individuals and plastids, always and everywhere has for its fnnd i m m t a l characteristic the branched navc-motion. Accordingly tlic molecular plnstidulc-motion, which lies a t tlic bottom of rll tlie plienomena of life, can liave no other form. Wc must conclndc tliat tliis ultimatc cause of all tlic phrnomcna of lifc, tliat the invisible activity of t h c organic molccnlcs is a branclicd wavc-motion. This true and ultimate cazisa ejicieus of the biogenetic

History of the Theory of Heredity.

37

process I propose t o designate by a single mord-Perigeiiesis, tlic periodic wave-generation of the organic molccnles or plastidules. " This mcclianical hypothesis is a true cxplmation of tlic proces3 of orgmiic development. . " The designation of this branched mare-motion of the plastidule by the word perigenesis or wave generation serws to emphasize the distinctive characteristic which separates this 6 m n c h e d motion from all similar periodic phenomena. Tliis peculiarity depends upon tlic reproductive power of the plastidule, and this again is brought about by its peculiar atomic composition. This power of reproduction mliicli alone renders possible tlic multiplication of the plastids is, however, the eqnivalcnt of the memory (Gediichtness) of the plastidule. " This brings us tQ Ewalii Hering's ably established view that unconscious memory is the most importnnt clinracteristic of organized matter, or more properly of thc organizing plastidnles. Memory is the chicf factor in tlic process of dcvclopmcn t of organisms. Through tlic memory of tlie plastidnles the plasson has the power t o carry over from generntion to generation by inheritance, in continuous periodic motion, its characteristic peculiarities, and to add to these the new experiences which the plastidnles have acquired through adaptation in the course of their evolntion. "I have shown that each organic form is the necessary product of two mechanical factors-an inner factor, heredity, and an outer factor, variability, or a powcr of ada13t nt i on. " By tlie hypothesis of periqenesis me are able to more sharply define these two fnndamental lmvs of the modification of organisms, for heredity is the memory of the plastidules: viiriaBility tlreir power of perception (Die

. .

3s Erbliclikcit ist (Ins (>ciliiclitniss dcr Plnstitliilc, die TT3riabilitiit is die F’:tssltngslir;tft tler l’1;isticltilc). ! h e onc brings about thc const:tiicy xiid tlic other tlie diversity of orgnnic foriiis. In t h very sinil.de :md pcrsisknt forms of life tlic plastidnles li:tve, EO to y i e d i , 1c:iriicd nothing :11id forgotLen nothing. I n Iiiglily perfected and variable orgi~iiismsthe plastidulcs 1i:~veLoth 1e:iriied and forgot ten m ach ” This sonicnliiit long qnotation coiitnins n tliorongh and csliuiistivc statement of the Iiwigenesis Iiy~otlirsis, and it. is tliei-cforo interesting to irotioe that its only rral claim to rccngiiitioii : I S :I true c,rplmdioib of the ])lieno men:^ of lieredit? is I)iised upon or a t IC:ISL dcmaiitls the acccptiince of some forin of tlic evolution Ii~-potltesis. 1Iowever p e n t niny be tlic im1mrt:iiice of the anallogy between the g r a d i d erolntion r.)f tho species liy tlie specialization of t . 1 1 ~consiitr1c:it iiidivi~I~i:~ls, arid tlic develol)ment of the individu:tl l i tllc ~ spcci:tliz:ition of cclle, and plnstidiilee, we II:LVC itlreiitlp pointed oii t iliatr i t is in no sense nu es1)l:in:ition of tlic latter, Eiiice the r e d cause of tlic evolution of the species, tlie selection of congmitul ~ i w i : i t i o ~ i sis, nlmnt. !l’lie on?? p r t of 1T;icckcl’s liypotlicsis of perigenesis which 1i:u :my clilirn to be considered nn explanation of t,he rcprodnctivc power of :~iiinials,is tlic stnt.emcnt t l ~ t heredity is memory, n i i d wri:ibility the :tcquisitim of new experiences. Slated by itself, without expl:iti:~tion, this nia! seem to tltose who are nnf:umili:ir wj111 tlIe siihject F C ~ J : nincli like nonsense, for the pr(ilount1 truth upcm which it wats is not :it all obviorrs :It first sight. IIerbci-t Sprnccr Im, i n his m:istcrly tlisrnssim of the natnrc 2nd diatinctivc c l i : ~ r : \ c t ~ r i ~ of t i c life, given 118, as tlic sum and substaiicc of his analysis, the statcment

.

History of th.e Theory of IJeredity.

39

t l i ; ~ t“life is the continuous adjustment between intern;d reli~tionsand. esteriid rc1:Ltions.” ‘lliis, like 1I:ieckel’s st:ttenient, that 1ieredit.y is mcmoiy, is not very clcar willio\it e q ~ h i i : & m ,but its nie:uiing may perliays be broiiglit out by ail illnstixtioii. If 1 kick :istone I prodiiee in it ccrtnin oliaiiges, siicli as motion, hciit, ctc.; tliese cliariges being directly prodiiccci by tlic kick are siinplv manifestations of ilie eiicrgy traneferrcd from my foot to the s t m e . If, instead of :I stone, I kick a clog, I produce :Lsimilar set of cliunges, niid sonietliing more. Ylic cxpcrieiice of tlic dog*aiicl of his nnccstoibs lias tanglit hiin t1i:it siicli violent ntt:icks :ire : t l w ; i y nssociiitcd mitli :t rl isposition t o commit still fartlicr violencc, SO, when the dog feels tlic blow lie iniincdiately performs actions -wliicli have its tlicir object, esc;q)c from or nroi~liuiceof the danger mliicli lie 11as not p t cspcricncccl, b u t wliicl~lie knows to bc imminent. Tlicsc actions :we not t h e cffoct of tlic liiclc, for tlic cncrgy expencled m:ig bc linnclrcds of times gmiter. Their c1iar:ictcr is ilc:crmincd, riot IJJ any cli:iiigc in tlie dog, but by tlic character, tlic tlisposition, w\\ich he lins iiilicbritei\; uiid x\ieiher he retaliates by :in :tttnck on his o \ v n lxirt, piits Iiis tail b e t w e n his legs :ind riiiis, or cronclics at niy feet, his actions :ire tlic effect, riot of tlic kick, but of p s t experience as to tlie best mcans of cscn.ping further iiijury. !Tl~ercis n relation, cxtcninl to the dog, bctwc.cn t h e kick arid 5 disposition t o iiijiire tlic dog, and tlrcrc is within the clog a rc1:ttiou between the scnsntion of injury :mtl the actions vliicli e s l m % w x lins shown t o be tlie proper ones for cscnping fnrtlier iii,jnry. T h ; vliich cliatitigiiishcs tlic dog from tlie stone is t h e power t o adjust these internal relations t o the esternal relations, to conform his coiiduct to the lams of

40

Heredity.

the world around him. The dog, as a living thing, differs from all inorganic bodies, in his power to nia1;c tliis adjustment: so long as lie retains this power he lives; liis life is a (‘continuous adjustment between internal rela tioiis and extcrnnl relations.” I t is plain that this power depeiids npoii experience, but experience dcpends upon “ memory.” So we may state, witli truth, that in a certain sense, life is memory; and as tlic power to reproduce its like is c1i:iractcristic of all living tliings, we see that there is in €Iaeckel’s Etaternelit a profound

t 1’11t 11. We know memory, hornever, only in conncetion n i th organization, and i f i t is true that heredity, the power of an organism to reproduce its like, is simply the memory, by the ovum, of the expcrience of its ancestors, we must believe that therc csists in the o w m an organization of some kind to correspond to each of tlicsc past experiences. We are therefore driven by thc liypothesis of perigenesis back from tlic hypothesis of epigenesis to some form of tlic old evolution hypothesis, for TC caunot conceive that cornplicatcd experiences should exist without complicated structure. We are thus compelled to conclude that, while it 1111doubtcdly expresses a great truth, IIaeckel’s hypotlicsis of perigenesis is not a satisfactory aiid final exl)lanation of the phenomena of reproduction. A satisfactory theory of heredity must explain what i t is, in the strncture and organization of tlie ovum, which determines that each ovam should produce its proper organism. To statc that this organization can be expressed in ternis of memory, is. simply to statc thc familiar truth that matter and force are different aspects of the same thing; that all problems of matter may be put iiito the

Ifistory o f the Theory of Heredity.

41

terms of force. The statement does not help 11s at all to picture to ourselves the essential hidden structure of the egg, the organization upon wliich its wonderful prol'erties depend. JIiger has recently brougli t forward an hypothesis whicli seems at first sight to be a satisfactory epigenesis. hypdthesis, but exaniinatiod sliows that this too, like IIaeckel's perigenesis hypothesis, must be turned into an evolution hrpothesis before it can be accepted. Tlic follomiug extract from his paper (" Zur Parigenesis," von Prof. Dr. G. Jager. x o s r t i o s iv. 376. 1879) gives, I believe, a fair statement of his Triews. ' ''Each organ and tissae of a n animal or plant contains, in the molecules of its albumen at least, a specific ~flnvo~-n~ibodo~.-substniice (Daft-nnd-Wiirzestoff) which me can easily recognizc by our chemical sense, for each orgnu of an animal 1 ~itss distinctive fl:~vor. Wlienever a full-grown animal experiences hunger, decomposition of albumen takes place in all its o r p n s and tissues, so that their various ~avor-and-odor-sii~stalices, that is their soul-substance (Seelenstoffe), become free, and penetrate to all parts of the body. " Now, if there exists in any part of the body protoplasm with the power to attract this substance, this protop1:ism acquires in this way its sires formatiam. " I have already referred with empliasis to the embryological fact that the formation of the reproductive elements takes place a t a rery early stage i n the embryonic life of an animal, and I have designated this as the reservation of germinal protoplasm. As soon as the embryonal cells of the developing animal have become specialized into ontogeni tic and phylogenetic cells, khe following will occur. Whenever any decomposition of albumen occurs in the developing organism, from hunL

42

hk-edity.

ger or any otlicr cause, the ontogenetic cell-material whicli builds up t h e org:inism will set free sod-stuff. ‘‘ By t h e law of gascons diffusion this will not oiiiy escape from the body as an excretion, but j t wjll also p e n e trate to the germinal or pliglogenctic protolhsm. This process I sliall now term soul-reception (Seelenfiingcrci) in the following sense. The diemica1 substance wliich forms the greater part of the ova and male cells ha Iittely been callcd nuclein, since i t shows the closest YCsemblance to thc cell nucleus. T h e jolk-enbstance is now rcgnrded, not :is vite!lin, b u t egg-nuclein, rind the substancc of the male cell riot spermatin but sperm-nuclein. Wc also know that nuclein consists of albumin, and phosplioric lecithin. “ T h e qiiestion then is the origin of the nncIcin in the egg, and the male cell, and this may bc ansncrcd as f 0110 ws : ‘‘ The reproductive orgins do not receim albnmen from the body of the mother, since ;iccording t o the law of Trsnbe, the moiecnles of a substance which fornis B membrane cannot, on ncco~intof flicir size, pass through the porcs of t h a t membrane. Thc germ-ccll is an albuminous membrane. and hence i t .will not aIIow the passage of albumin molecules. “It simply contains the albumin-nuclens, which PCmains after the decomposition of the sod-substance, niid this is n peptonelike substance ~rhicli,having lost its soul-eubstance, has a smaller moIccule. It is therefore nnspecializcd, or deprived of its sod (mtepcsificirt, entseelt), and t h e process of assimilation in tllc germ m:iy be termed Eoul-restoration (~-iederbeseclung). Tlic necessary soul-substance is supplied by the decomposition of albumen in the ontogenetic cell-rnateri:~l. ‘‘Thus, for examplc (p. 380), it is kiioxn that tho re-

Bistory of the Theory of IIeredity. -

43

~~

lwocluctire organs of :t catcrpilhr are already formed bcfdrc i t lcadcs tlic cgg. During its lifc 111 t h e egg, and :i$ ;I ci~tciyillar,ci~t~rl)ill:~r-~in~leiil is formed iii its gcrmi11;11 cell-material. During tlic pupa stage pupa-iinclein is storcd 111) i n its iel)roCltictivc clcniciits, arid finally, wlicii it bcconies :I butterfly, butterfly-nucleia is stored up. ‘i’lic ripe cgg and tlic rilbe niule cell thcreforc contain nnclcin of three kind?, catcr~illar-nuclcjn,p u p i i ~ ~ c l c i iand i , butt crfiy-niiclein.” It will be sccn t l i i t t Jiiger’s hypothesis is, in n certain sense niidmay between evolution aiiti epigeuesia. lie liolds that at first both thc oviini :tiid tlic mile-ccll arc iinspccialiaed (cntscclt) ; tliat they exist i n th e w r y piing embryo :is cinbryonic ova or spermatozoa, :uid that, as the embryo gFows up, the reprodiictivc crIIs g~ttluallybecomc spccidizcd by tlie :mimilstion of soulstnff, whicli is thrown off by the decomposition of :tlbnmcii in varions parts of tlic body of tlic growing oYg:inism, and pcnctrating to tlic cnibryonic O V R a i d sl)crmatozos is asPiInilatcd by tliern, so that when the :mirn:tI becorncs se.uu;~llym:itiire, tlie cells of its reprodwAivc org:ins cont:iin :ill the “soul-stuff ” ncccssary to 1nwdiicc n nc\v organism like tlic parent. Tile statcmcnt wliicli I linve given is a frce translation of Jlgcr’s outiinc of his tlieory, and I think it may be rcg:irdcd ;IS a fair exposition of his tiews. d fatal objection to his hypothesis is found in the fact that wlicrc a parcut givcs birth to young beforc it hns rcaclicd f u l l maturity and before it lius acqnircd :dl the cliaractcristics of tlic spccics, tlie yonng n c ~ e r t l i c less inherit thesc chctructcristics. Tlic jouiig which w e h r i i c by a Cpcedomia lurm inlicrit :111 tile cliuractcristics of the full-grown ndtrlt insect, a d a brill may transmit t o fcmalo childrcii the good milking qualities of his

44

Heredity.

mother. I t is plain that the child of a be;irdlcss boy could not inherit tlie “ soul-stuff ” of a beard i n the way Jager imagines, arid this f w t alone js erioiigli to s l i o ~ tliat he h ~ i snot discovcrecl tlie true secret of Iicrcdily. We know, too, tlint ~ C V C I ~ S ~ O the I I , :ip~)e:~r:tnce in the child of tlie features inherited from a remote nnccstor but not shared by its parents, is not a t all unnsnal, and mast be regtardecl as otie of tlie lcading cliaracteristics of Ilercdity. I t is plain that if tlicenibr~-onicoviini is, ns Jager states, in specialized or ” de-souled,” rcvcrsion is inesplicable. Accordingly, when lie conies to discmss reversion lie makes ii fiindamcntal cliange i n liis ligpotliesis, nnd holds that wlien tlie oviim divides, a t a very early stage of its development, into two parts, zn ontogenetic portion, which givcs rise to the iicw orgaiijsni, and a pliylogenetic portion, mliicli ultimately forms tlic germinative cells of its rcprodactire organ, tlie second part is not unspccinlizcd or ‘‘ clc-soulecl”at all, but really retains all the clinractcristics of tlic ovum irliich gives risc to it, and is tlicrcfore capable, like tlie ovuni, of giving rise t o a new organism. As thus remodelled, I bclicw, and hope t o show in the scquel, that Jager’s hypotlicsis is a close approximation to tlie truth, but i t is only fair to point out tlint in its nltcred form it is not original with JHger. Tlie author piiblislied almost exactly the same view in 15’76 (“ On a Provisional IXypothesis of Pangcnesis,” Pmc. Amer. Assit., 1S76, and Aiiiericau A’ntzii*alist, March, 1877), and i t liacl been stated as long ago as IS49 by Prof. Omcn, i n his p:ipcr on Parthcnogcuesis, nlthough this author, in liis ‘‘ Anatomy of Vvrtcbrates,” afterwards states tlint lie now lsclicves it to be funclamentnlly erroneous. It. is plain, too, that in its sccontl form Jager’s hypothesis is one of evolution, pure and simple,

Bistory of the Theory of Heredity.

-~

46

for the egg is, at no stage of its growth, unspccialized, : ~ l di t does not require the assimilation of “ soul-stuff ” iii ordcr to devclop into a n organism. We must conclnde, then, that however satisfactory anll :tccordant with obscrved fact t h e hypothesis of epigcncsis seems to be a t first sight, more careful analysis sliows that it is i n no sense a true explanation of the p11en om en o 11 of d eve1opni cn t. Tlic analogy betwcen t h e evolntion of the species from an nnicellular mce&oy, and t h e development of the individual from ~ 1 ~~nicclInIar 1 egg, is simply an analogy, for tlic cause of Llie first plienomeiion, the selection of cong r n i t d mriation., is wanting in tlic second case, and tlicrc is nothing to take its place if i t is truc t h a t a n egg is rcally, like a rliizopod, an unspecializcd ccll. H;ieckcI’s statement tliat heredity is memory, howc ~ c truc r it may be, cannot Le accepted as an explanation, for me liavc no knowledge of the existence of memory apart from organiza,tion, and we cannot conceive t1i;it a n ovuni can rctain tlic mcniory of the past history of its species, iinless it possesses a corresponding organiznt ion. Jiigcr’s view that ‘the embryonic ovum is unspecializcil, and that its specialization is gr:idnally assirnilntcd during tlic development of t h e organism mhicli contains it, fails l o acconnt for the plienorncna of rerersion, and to acconnt for rcrcrsion lie is comIwlled to assume t h a t tlic c g is organized from the time of its origin i n tlie dcwloping egg of thc preceding generation. In each case we are drircn to tlic same conclusion, tli:it the cpigenesis Iiypothesis is inadequatc; and me are forccd t o accept some form of the evolution hypothesis. This neccssity Iias not escaped the notice of some of our most acute thinkers. Huxley, for example, says

46

Heredity.

(Encyc. Brit., Art. Evolution), " Harvey's definition of a germ as matter potentially alive, and having within itself the tendency to assume a definite living form,' appears t o meet all the reqnirements of modern science. For notwithstanding it might bc' justly qnedioned whether a germ is not merely potentially but rather actually alive, though its vital manifestations are reduced to a minimum, the term pot'cntial may fairly be used in a sense broad enough to escape the objection. And the qualification of potential has the advantuge of reminding ns that tlie great cliaracteristic of the germ is not so much wlint i t is, but what it may under suitable conditions become. '' Prom this point of view the process, which in its superficial aspects is epigenesis, appedrs in essence to be evolution, and deyelopment is merely the expansion of a potential organism or organic preformation according to fixed laws."

...

CHAPTER 111. HISTORY OF THE THEORY OP HEREDITY-(

Continued).

Some form of tlie evolution hypotliesis a logical necessityDarwin’s pangenesis hypotliesis-This is an evolution hypothesis, since all the cliaracteristics of the adult are supposed to be latent in the germ-Miscellaneous objections to it-These objections do not show that it conflicts with fact-Difficulty i n imagining detailed working is no reason for rejecting it -Gilton’s cxperimental disproof-There are many reasons for believing that tlie sexnal elements have different functions -The evidence from parthenugenesis-Polar-cell hypothesis -The evidence from hybrids, from variation, and from structures confined to one sex-The pangencsis hypotliesis recognizes no such difference io the functions of the reproductive elements-We must therefore distrust its absolute correctness --Summary of last two chapters.

8ome Form of the Bvolution Hypothesis a Logical Necessity. Most of the hypotheses which have been proposed, of late years, to account for the phenomena of heredity, are like the two we have quoted, epigenesis hypothesis, for they are attempts to show that the ovum is in reality, as well as in form, an unspccialized cell. Analysis shows, however, that they all rest ultimately upon the assumption that this is not true, but that the ovum really contains, in some form or other, actually or potentially, the €uture organism, with all its hereditary characteristics. We know that eggs which are to all appearances essentially alike, may, when artificially removed from the ova-

Neredit y.

48 ~

ries and artificially fertilized, and when kept niidcr csactly the sanie conditions, develop into widely differciit organisms, and as like things csnr ot, under. lilic conditions, give rise to different results, me are forced to conclude that these eggs are not essentially alike, but that each contains within itself in some form tlie organism to which it is to give rise-that iiidividual development is, in some sense, the unfolding of a germ which already exists in the egg. There is no escape from this conclusion, a t least there is none which can be accepted by the scientific student, and we see that logical tliinkers like Prof. H L X Xare ~ Cdriven ~ to conclude that the process which in its superticial aspects is epigenesis, appears in essence to be evolution.

Darwin’s Hypothesis of Pangenesis. In contrast to the views already quoted me hare the well-known pangenesis hypothesis of Darwin, an hypothesis which is thoroughly one of evolution, since Darwin believes that the whole organization of the syecics’ is present not only in the egg but in the male cell also; that each of these not only contafns the complete organization of the parcnt, but an indetinite series of similar organizations, inherited from a long line of ancestors. It is true that Darwin does not believe that each of these ancestors is represented in the ovum and in the male cell by a minute but perfect animal, like those imagined by Bonnet, but he imagines what is essentially the same thing, that each of the cells of each parent, and every cell of each ancestor for a long aiid practically an unlimited series of generations, is represented in each ovum and each male cell by a germ capable of producing that particular cell with all its distinctive characteristics. Darwin’s original statement ( Va?.iuiio?a,chaps. xxvii.

History of the Theory of Heredity.

49

and xxviii.) is rcadily accessible, but it mill not be out of place to quote it before entering upon its critical discnssion. IIc says: “In tlie previous chapters large classcs of facts, such as those benriiig oil bud-variation, tlie various forms of inlieritance, the caiiscs and laws of variation, have been discusscd, and it is obvious that tliese subjects, as well as tlie several modes of reproduction, stand in some sort of relation to each other. I hare been led, or rither forced, to form a view, which to a certiiin extent connects tlieso facts by a tangible method. Every on0 Tvould wish to explain to himself, even in an imperfect manner, horn it is possible for a character possessed by some remote ancestor suddenly to reappcar in tlie offspring; how the effects of increased nsc or disuse of a limb c n ~ be i transmitted to the child; how the male scxual clement can act not solely on the ovule, but occasionally on tlie niotlier form; horn a limb can be reprodaced on the exact line of amputation, vith ncither too inucli nor too little added; how tlie rarions forms of rcprodnction are connectcd, and so forth. I am amare tlint my view is nicrcly a pro~-isionalhypothesis or spcculntion, but nntil a better one be adranccd i t may be scrviccable by bringing togetlier a multitude of facts which are at prcsent left disconnected by any efficient came. As TVhewell, the historian of tlie inductive scicnces, rcmarks, hypotheses may often be of service to science, wlien they involre a certain portion of incomplctcucss or ercn of error. “ Under this point of vicm I ventnre to advance the liypotlicsis of pangenesis, m1iich.implies that the whole organization, in tlie sense of cvery separate atom o r unit, rcproduces itsclf. Hence ovules and pollen grains-the fcrtilized seed or egg, as well as buds-include and con-

80

Heredity.

sist of a multitude of germs thrown off from each separate atom of the organism." From the extract we see that the hypothesis is a n attempt to shorn that all the phenomena of generation and development, including those of variation as well as those of heredity, depend upon the fact that each strnct u r d unit of the body is the direct offspring of a similar unit i n the body of a parent or of ,a more remote ancestor. The cells of the body of one of the liighcr organisms are not only niorphologically but ;ictnally indcpendent inclivid i d s , reprodncing themselves directly. in the next generation: and the germ of such an organism is in reality an aggregate of these cell-germs. Stated more at length, the li~'p0t11esisis as follows : " 1 assume that cells, before their conversion into 'form matcrial,' throw off minute granules or atoms, which circnlate freely throughout the system, and when siipplied with proper nutriment, multiply by self-division, snbsequently becoming developed into cells like those from which they were derived. These granules, for the sake of distinctness, may be called gemtnules. They are supposed to be transmitted from the parent to the offspring, and are generally developed in the generation which immediately sncceeds, but are often transmitted in a dormant state during many gcnerntions and are then developed. Their development is supposed to depend on their union with other partidly developed cells or gcmmules, which precede them in the regular order of growth. Why I use the term union will be seen when we discnss the direct action of pollen 011 the tissues of the mother plant. " Gemmules are supposed to be thrown off by every cell or unit not only during the adult state but during all stages of development. Lastly I assume that gemmules

. . .

€-€istory of the Theory of Beredity.

61

in their dormant state have a mutual affinity for each otlier, leading to their aggregation either into buds or into the sexual elements. Hence, speaking strictly, it is not the reproductive elements nor the buds whicli generiite new organisms, but the cells themselves throughont the body. These assumptions canstitate the provisional hypothesis of pangenesis.” Darwin’s genimules are, of course, entirely imaginary, that is, a belief in their existence does not rest upon direct observation. We cannot deny that the liypotliesis furnislies an explanation of most of the phenomena wliicli Iic attempts to interpret by it, although it seems possible that there may be a simpler explanlttion. If the existcnce of tlie gemmules were proven we could nnderstand not only the rvonderfnl facts of ordinary inheritance by sexual reproduction, but the various forms of asexual reproduction as well. We should have a simple explanation of the manner in wliich the characteristics of a remote ancestor may suddcnly reappear after they have been dormant for many generations. We should understand how the embryological history of a species may become simplified by tlie omission of larval forms or appendages. I n a mortl, nearly all the phenomena of heredity admit of explanation by the liypothesis, and those who have criticiscd it have not usually attempted to show that it conflicts with fact, but have simply objected to it as apurely imaginary explanation. It is urged that the transmission of all tlie cliaracteristics which we know to be inlicrited from near and remote ancestors demands that the number of gemmules should be almost unlimited and practically infinite; that not only are the gemmules imaginary, but that the aggregation of such numbers in masses as small as the reproductive elements requires

62

Beredity.

that they sliall bc of inconceivable minuteness, and that naturc fiiriiislies no analogy for :tttribotiiig t o such s n i A prticlcs the vital properties which we know otily it1 bodies which itre compiwatively gigautic. It is also urged that the gemmnles ninst be endowed with eiitircly imaginary and wonderfully spccinlized elcctive affini Lies, in virtue of which e x I i dcveIops onIy at the proiler tiuic and place. In order to account for the ni:mtier i l l which tlie characteristics of each parent are mingled in 1110 child we must regud each individual as tlie product of a struggle for existence among the gcmmules, meult\og in the selection a d development of the fittest. The for. mation of several indiyiduals :isexiially by budding from a parent stock demands tliut the gcniniules thcnieelves must be c a p b l c of mnltiplication, and that they niust have the power to transmit tlieir ln-operties to tlieir offspring. To explain alternation of generations we must suppose that the enibrgo receives scvcnil complete scts of gemmules, which are not dnpkutcs, and it is almost impossible to follow ont, in thought, tlie complicated relstions mliicli must exist between the gcmniulcs of tlic egg-embrjo of such an organism as a Sipliouophorc. These aid similar objections may be fairly iu’gcd, and mliile tlicir great weight is oblions, we must iiot attwli undue importance to them, for they do not show tlmt the hypothesis conflicts with any known law or ubs c r ~ c ifact, l and the great drafts made upon tlic im:igiiintion should not, alone, r re vent its pro~isionalaccclltanco so long as me liave no simple esplaniition of tlic phenomena, for difficulty in imagining tlic dctidls of :in hypothesis is a purely snbjectirc matter, vliich vuics will1 the age and with the individual.

History of the Theory

of Heredity.

53

Gcr lt on’s Expe~ intents. Bcsidcs these theoretical objections, we hnve the experimental disproof furnislicd by Galton. In order to test the hypothesis this cxpirimcntcr selected the silvcrgray rabbit-a variety mliicli has, in itself, little tendency to ~ar.y,although i t readily crosscs with otlier varieties, and breeding freely with them gives birth to hybrid offspring. Into the bodies of eighteen of tliese silver-gray rabbits Iic trailsfL1scd tlic blood of otlicr varieties, in some cases replacing onc lialf of the blood. F r o m the ciglitecn rabbits tlina opcratd upon eighty-six soung wcrc produccd, ancl i n no case did tlic offspring cxliibit any of tlic charactcriatica of tlic variety from mliicli tlic blood was talicn, but all of tlie eighty-six mcrc pure silver gr‘iy. Froin tliesc cxpcrimeiit s Galton concludes tlint “ t h e doctrine of pangencsis, pure a n d simple, is incorrect;" a n c l I tliink mc ninst agree with him tliat this conclusion is justified by tlic results mliicli h e rcaclied, altliongli I hope to sliow tliat i t is possible t o restatc tlic hypothesis in a form which is so modified as to escapc this objection.

The Sexual Elenieiats Perfom Different Fuiictions i7a I h e d i t y. There is another objection which seems to me to be of almost eqnal weight, b u t which has never, so far as I a m amare, been pointed ont. T h e early writers upon heredity attributed certain fanctions to the male cell and others to the ovum; but me now know that their means of observation mere so inadequate, and their knowledge so limitcd, t h a t tlicir conclusions wore of littlc value, and that both orjsts and spermists wore wide of the mark. T h e fact that they crroneously attrib-

Her edity.

54 ~~

~~

~

-

~-

uted certain fnnctions to tlic ornm and certain otliers

to thc male cell does not, of course, prore that there is no difference in tlie fnnctions of these elements; but in modern times we actually find that thinkers hare gone t o this opposite cxtremity of the subject, arid liare either tacitly implied or directly accepted tlie vicw that the two sexn:tl elements play similar parts in heredity. Nei thcr IIacckel’s hgpotliesis nor Jiiger’s recognizes any difference in tlieir fn?ctions, w.hile JBger secrns to believe, and Darwin explicitly statcs, that tlieir sliarcs in hereditary transmission are alike. Many facts indicate t1i:it this view is, to eny tlie lcast, very improbable, and I mill give, bricflp, a statenicnt of some of the arguments against it, and will then devote a little space to a discussion of tlic reasons wliich 1i:ire been given by Darwin and others for accepting it. The structural difference between the ovum and the male cell is one of the most widespread and fundamcntnl characteristics of organic beings, and it is found in all except tlic very lowest animals and plants. I t is, to eny the least, very improbable that a strnctnrnl difference so fiindamental and SO nearly universal should 1i:lre 110 functional significance, and the fact that in many marine animals, when the ripe unfertilized ova arc thrown ont into the ocean, like the ninle fluid, to be swept away by the tide, the sexual elements differ in tlie same a a g that they do in aniinals whose eggs are fertilizcd inside the body of the female, forbids 11s to believe t1i:it the difference depends simply upon the fact that tlic nialc ccll must make its way to the ovum. Many of the secondary cliaracteristics of thc ornm, such as its p e n t size in birds and reptiles, and tlic 1msence in it of food-material in so many animals, are no

Bistory of the Theory of Heredity.

55

doubt traceable to the fact that, in most animals, the

cgg is stationary, while the male cell can be conveyed irom place to place; but me must bclieve that there is some more fundamental and primitive difference. Even if the phenomena of Parthenogenesis did not slioiv us that the part played by the ovum is more essential to the perpetuation of the race than the part played by tlic malc cell, we should still be justified in the belief that the difference in form corresponds to some profound difference in function, and the possibility of Parthenogenesis shows beyond question that this is the case. Pnrtlieiaogenesis. Siebold has proposed the term parthenogenesis for the yowx which is possessed by certain female animals, cspcci;~llythe arthropods, to produce descendants without scsnal union with a male. Tlie existence of this power was first pointcd out by Aristotlc (DE Geiisratioiie Animnlittnz, Lib. III., Cap. 10, 21, 22, 23). As this remarkable dmeruer had no menus for exact research at his command, he was, of course, unable to-make w e of rigid tests, or to furnisli the scvcrcly exact proofs wliicli have been given 11s by more modern naturalists ; but he gives many reasons for suspccting that the unfertilized eggs of the honey-bee may give rise to perfect animals without sexual union ; and although we now know that some of the reasons he urges do not really prove tlie case, yet modern science has given tlie most convincing proofs of the correctness of liis general conclusion. I sliall dcrote considerable space to this subject in ordcr to sliom the unscientific reader that the existence of fertile virgin female animals is proved by the obser-

56

Herediiy.

vations of a great number of compctcnt nntnralists; that the subjcct has been thoronglily and carefully studied, with every precsntion against error, and i l x t oiir belief in its existence docs not rest upon tlie unverified statenients of a f e v observers. I n this summary I sliall give many referenccs t o authorities, bnt as my p r p o s e is not to give a conlpletc bibliography, but simply to ~shomhow tlioronglily the subject hns been studied, many imrnes are omitted. Most of the following facts arc takcn from Gerstaecker's history of tlie subject in Volume V. of lh 0 1 1 11's El(isseu 0)d)m):gleu rEes TlLierreidis, altliough I Iiare rcfcrrcd to many of tlie origiiiul pipers and have :dded many facts which are not nientioiied by Gerst:iecGer. 'the snbject is perfectly fsrniliar to most naturalists, and the amount of space devoted to i t may seem oniicccssarily grcat to such 11wsons, bnt it is importnnt to j m p e s s u p o ~n n s e i c ~ ~ t ~ f i c readers :I scnsc of the exact and definite character of tlre cvidcncc for the cxistcnce of partlicirogencsis, and a short history of tlic subject seems the most effective means for accomplishing this prpose. Among the crnstnccn and iiisccts, lmrthenogenesis is by no mcans unusnal. It occurs in some groups mlicre impregnation by m ; h s is so nearly oniversol that nnturalists hare been slow to credit any exceptions. I n 0 t h gronps it is the general rule, and fcitilization by a male is the exccption. In some genera and species the powr is shown only by a few individuals, ~vliile in others it is shnrcd by d 1 the females. In some cases ihe unfertilized eggs give rise to feniales only, in other cases to males, and in still other cases to both Ecxes. In 1'775, Sciibffer, of Begensbnrg, discorcrcd its occurrence in fresh-water crustacea, although Dr. Albrecht

Nistory of the Theory of Heredity.

57

-

]lad made the same discovery in insects in 1’701. Scliiiffw found (‘ology does not iiivolve functional similarityThe argument from the dual personality of each individual; from reversion ; and from polymorphism-These phenomena admit of a simpler explanation-Summary of chapter.

The Arpmeizt from Bybrids. According to the view to be presented in this work, the functions of the two sexual clements, in inheritance, arc not alike. Tlic proof of this mill be presented rnrther on, when tlic snbjcct is reached in the logical course of the dcvelopni en t of o 11r argument. Some of thc very liigliest autliorities have been led to a view wliicli is directly opposite, and lime lield that citlicr parent may transmit to the offspring any characteristic whatever. Lest any reader should assume, a t the beginning of this book, that the work involves an absurdity, and that my conclusion is already d i s p r o d , it seems best to at oncc examine the reasons for the opposite view. If I can show that these reasons are inconclusive, and tliat tliere is and can be no proof for tlic statement that each sexual element transmits to the offspring every characteristic of the parent, we can then enter into the subject without prejudice, and can wait

100

Heredity.

for the proper time to present tlie proof of tlre opposite fview, tliat tlie two sexual elements play different parts i n heredity. If tlic authority of great names counted for anything Tvlmtever in sciciice, the case against me would be w r y strong, but where an appeal to natiirc is possible, authority counts for nothing. Darwin’s place among the students of heredity is certainly the highest, and lie takes very strong ground indeed upon this subject. Thus lie says (Variation of Animals and Plaids, Vol. ii. p. 8 8 ) : “I am aware that such cases (of prepotency) have been ascribed by various authors to snch rules as that the father influences the exteriial characters, and tlie motlicr the internal cliaracters. “ But tlie great diyersity of tlie rules given by various authors almost proves their falseness. Dr. Prosper Lucas lias fully discussed this point, and has shown tlii~tnone of tlic rnles (and I could :tdd others to those quotccl by him) apply to ;ill animals. Similar rules have been announced for plants mid have been proved by Giiirtucr t o be all erroneous.” In the Atiatoiiiy of Inverte6rnted Aniiiinls, 11 30, IIuxlcy states tliut “ n o structural modification is so slight, and n o fonctional pcca1i:trity is so insignificant i i eitherparent, ~ tliat i t may not malie its appearaim in the offspring.” D ~ r w i n ,in many parts of his mritings, is still more explicit. Thus he says (Vcci.iatioit of Aizin~uls nqzd Platrts, Vol. ii. 13. 431): “Ovules and tlic male clement, before tbcy become united, have, lihc bnds, :ti1 indcpcndent existence. Both have tlic p o ~ ufr t i x i i s m i ttiiig every sinyZc character possessed by the parent fol ni. We see this clearly when hybrids are paired inter se, for

Various Opinions o n Xeredity.

101

c

the characters of either grandparcnt often reappear, citlier perfectly or by segnicnts, in the progeny. It is I r i L error t o sappose that the iiiale traiisnzits certain charac. t p r s mail the jEinnZe other chni*acfers.” I tliiiik a little cxarnination will show clearly the impossibility of proving this statement from the phcnomeI n order to brecd together animals must i i i t of crossing. I)c closcly related; they must belong to the samc species or to two closcly allied species. Since the individuals which belong to two closcly related species are the dosccudants ‘of a common, and not very remote, anccstral spccies, it is clear that almost the wliole course of their evolntion has bccn shared by tficm in common; all their gcneric characteristics being inherited from tliis ancestor. Only the slight differcnccs i n minor points, ~vliich distinguish one species of a genus from another, liare been acquired since the two diverged, and not even all of these slight differences, for a diffcrence brtmcen two allied specics may be due to the fact that wliile one has been modified the otlier has retained, unmodified, certain resemblances to their commoIi ancestor. We know that thc dtiration of even the moat persistcnt species is only an infinitesinial part of the wholc history of their evolution, and it is clear that the common cliaracteristics of two allied species must outnumber, thousands of times, tlie differences between them. It follows that the parents of any possible hybrid must be alike in thousands of features for one in which they differ. I t is therefore out of the question to attempt to prove, from the phenomena of crossing, that each parent can transmit to tlie child all its characteristics. Crossing simply results in tlie formation of a germ by the union of a male and a female element derived from two essentially similar parents, with at most only a few secondary

102

HerecZity.

and comparatively slight differences, all of which have been recently acquired. If a perfect animal couId be developed from the spermatozoon of a male parent, as it can be, in cases of parthenogenesis, from the ovum of a female parent, we should have a means of proving that each sex transmits its entire organization to its offspring. The phenomena of parthenogenesis prove that the female does actually thus transmit its entire organization, but there is nothing to show that the male parent docs also, for it is clear that, from the nature of the case, the phenomena of crossing are incompetent to prove it.

The Argument from the Homology of the X a l e aitd Female Sexual Elenients. Many authors have gone much further than the statement that any charactcristic whatever may be transmitted by either parent, and have held that the offspring is actually a dual personality, made up of a complete organization or indjvidualitg inherited from the father, and another, equally complete, inherited from tlic mother. This view has found favor with a number of modern writers, and frequently makes its appearance in tlic literature of the subject. Thus Huxley says (Em~cZoop. Brit., Art. Evohitioii), “It is conceivable, and indeed probable, tIiat cvcry part of the adult contains molecules derived from the malc and from the female parent; and that, regarded as B mass of molecules, the entire organism may be compared to a web, of which the warp is derived from the female, and the woof from the male. And each of these may constitute an individuality in the same sense as the whole organism is one individual, although the matter of the organism has been continually changing.”

Qarious Opinions on Heredity.

103

+

I t will be found, on examination, that there is niuch to be said in support of this view, although I believe that there is a much simpler explanation of the facts which seem to favor it. The only reason given by Huxley, in the article above quoted, is the homology between the ovum and the spermatozoon; the fact that in all the higher animals and plants the germ is formed by the unioii of one nucleated cell, the ovum, with another more or less modified nucleated cell, tlie male cell, and that the structural components of the body of the embryo are all derived, by a process of division, from the coalesced male and female germs. I n answer to this we may point out that while the hypothesis requires that a wasp born from a fcrtilize& egg should differ essentially from one born from a parthenogenetic egg, the one being a dual person and the other a unit, we do not find any obvious difference corresponding to tlie supposed molecular difference. We should not expect a wasp with a dual personality to be, to all appearances, exactly like one with a single personali ty. A fatal objection to Huxley’s argument, above given, is that, at bottom, it is simply an assumption that the homology or morphological equivalence of the ovum and male cell proves their functional equivalence. The fallacy of this assumption hardly needs notice, since it is well linomn that hotnology is no evidence whatever of functional resemblance. The quill feathers which fit a bird’s wing for flight are homologous with the scales whicli cover and protect the arms. and fingers of a crocodile, but we conld hardly name two structures wliich scrve more different purposes. The homology between them simply indicates that, at some time in their his-

104

Berediiy.

tory, both scales and feathers have had a common origin i n an epidermic structure, which has gradually Become specialized into these organs. While tlie homology between the ovum and the male cell is no reason for assuming that their fnnctions are now dike, the constant differences between them, througliout almost all of the organic world, seem to afford a very convincing reason for believing that their functions have been specialized in two divergent dircctions. If we can sliow that good might have resulted to the organism from such specialization, and from the rostriction of certain parts of the rcproductive fnnction to one element, and tlie restriction of otliers to the other, we may feel confident that, provided rariations i n tliese directions have a t any time arisen, natural selection would ham seized upon and yerpetnated them. I h o p to show the great usefulness of a specialization of this sort, and if I can do so, it is clear that the known differences between tlic ovum nad the ~perrnatozoonare reasons for a belief in its existence, while the only conclusion mliicli can be drawn from tlie homology bctween them is, that at one time tlieir functions were alike.

The Arguments .frona the Transmission of Latent S'ezua2 C?baracteristics; from Beaersion, und from AItermt i o n of Generulions. I n addition to tlie reason given by Xuxley for a belief in the dual natiire of each orginism, lie might have adduced the fact that tlie cliuracteristics of each sex are potential and latcnt iu the organism of t h e opposite sex, as is proloved by the trmsmision by ,a falber to his d:llrghter of characteristics inherited from his grandmother. The fact that the characteristics of one sex are latent

Various Opinions o n Heredity.

105

in the organism of the other is proved by countless wellknown illustrations, and i t seems, at first sight, t o afford evidence of the dual persopzlity of each animal. The fact in itself is so interesting that, while I believe in tlie possibility of a much simpler and more satisfactory explanation, it will not be out of place to devote a little space to the subject. ‘‘In every fcmnle all the secondary male characters, and in every male all the secondary female characters, apparently exist in a latent state, ready to be evolved nnder certain conditions” (Darwin, Variation. VoZ. ii.

p. 68). A perfect beard often begins to grow upon the face of a woman after the power of reproduction is lost by age or disease. Such women are often alluded to by Roman authors nnder the name of “ viragines,” and Hippocrates (De iI401-b. PuZg., Lib. vi. 55-56) has left us the description of two well-marked instances. Aristotle (Hist. Aizimal, ix. cap. 36) gives an account of a hen which had ceased laying, and assumed the characteristics of the male bird, and similar change i n female birds has been recorded by many writers. It has been observed in tlie hen, common pheasant, golden phcasant, silver pheasant, turkey, pea-hen, partridge, bnstard, pelican, various ducks, cuckoo, cotinga, chaffinch, bunting, and other birds. The change may be produced by age, by disease of the ovaries, removal of the ovaries, and even (Yarrel, Phil. Tran,?, 182‘;: ii. p. 268) by r e m o d of part of the oviduct. Old liens which have stopped laying often acquire a comb, wattles, spurs, tho brightly-colored plumage and long tail-feathers of the cock, assume the habits of the malc, and w e n learn to crow. The bad character, as lajers, of crowing hens, has even given rise to a proverb.

106

Heredity.

According to Darwin, Waterton gives a curious case of it hen which had ceased laying, and had assumed the plumage, voice, spurs and warlike disposition of the cock: when opposed to an enemy she mould erect her hackels and show fight. Female deer often acquire the horns, peculiar hair, ears, odor, and sexual desire of the males. On the other hand, it is well known that the secondary sexual Characteristics of male animals are more or less completely lost when they are subjected to castration. Darwin states, on the authority of Yarrell, that if t h e operation be performed on a young cock, he never crows again; the comb, wattles and spurs do not grow to their full size, and the hackels assume an intcrmcdiate appearance between the true hackels and the feathers of the hen. Similar results are said to be produced by confinement. Buffon states (Hist. Nat., Tom. vi. p. 80) that the horns of a stag castrated during the rutting season become permanent, but that new horns do not usually appear if i t is castrated when out of heat. Simpsonsays (Cyc. qfAnat., 702. ii. p. 71’7), “From the frequency with which castration is performed, the effect of the testes in evolving the general sexual peculiarities of the male have been more accurately ascertained than that of the ovaries upon the female constitution. These effects vary according to the age a t which the removal takes place. When an animal is castrated some time before it reaches the term of p b e r t y , the distil;ctive characteristics of the male are in general nemr devcloped; and the total absence of these cl~aractcrs,togcthcr with the softness of their-tissues, the contonr of their form, the tone of their voice, and their want of encrgy and vigor, assimilate them more in appearancc and

Various Opinions o n Heredity.

107

habits to tlic female than to the male type. If the testicles are rcnioved nearer the period of puberty, or a t any time after that term lias occurred, and when tlie various male sexual l~cculiaritieshave been already developed, the effect is seldom so striking: tlie sexual instinct of tlie animals, and the energy of cliaracter wliicli these instincts impart, are certxirily more or less completely destroyed, and tlic tone of the voice is sometimes changed to tliilt of puberty, but the general male cliaracter of form, siicli as tlic beard in man, and the liorns of ruminants, gciicrnlly continue to grow.” D;1rivin, after reviewing these facts, concludes as follows: 66 We thus see that in many, probably in all cases, tlic secondary sexual characters of each sex lie dormant or latent in the opposite sex, ready to be evolved under peculiar circumstances. “ We can thus undcrstand horn, for instance, it is posz:lsle for a good milking cow to transmit her good niilking qualities through her male offspring to future gcncrations, for we may confidently believe that ‘these qualities are present, thougli latent, in the males of each generation. So it is with the game-cock, mlio can transmit his snpcriority in courage and vigor through his female to his male offspring; and with man it is known that diseases neccssarily confined to the male sex can be transmittcd through tlic female to the grandson. Such Eascs are iiitelligiblc on the belief that characters common to the grandparent and tlie grandchild of the same sex are present, though latent, in the intermediate paren t of tlic opposite sex.” Facts of this sort certainly seem, a t first sight, to show the existence in each individnal of two complete individualities, one from each parent; and the presence in each

...

108

Neredit y.

sex, in a latent condition, of the organization of the other sex ;but i t is not difficult to show that the plienome m in question adniit of a much simpler explanation. I n most cases when the sexes differ from each other i n what are known as secondary sexual characteristics, that is, features which are not directly concerned in the reproductive function, the mature male is more different than the maturefemale from the young. I shall discnss this subject more fully in another place, so I shall give only a few illustrations a t present. I t will be sufficient to call attention to the resernblance betmcen tIic smooth face of a woman and the face of either a boy or a girl, as contrasted with the bearded face of a miin. The voice of a wonian, the voice of a girl, and that of a boy, a11 resemble each othcr, and all differ from the voice of st man in tlie same, or nearly tlie same, respects. I n fowls tlic young of both sexes are much like the adult female i n form and color. These familiar instances are enough for our present purpose, and they show that, so far as the secondary sexuil characteristics are concerned, the female is, as a rule, distinguished from the male by her failure to acquirc tlic fully developed clinracteristies of the race. I n these respects the female j s an arrcstecl male, and this is me11 shown by that fact that while the females and young of two closely related species of mild animals may be so mncli alike tlint they can Iiardly be rlistingriished, thc adult males may bc very different from each other. All we need t o m u m e , then, i n ordcr to each a simplc explanation of the sccondayy sexual diff crenccs betwccii tlic sexes, is that each ovum has the pomer to derelop into an organism with all the characteristics of the species, but that the fernale function acts, in some

Various Opinions on Heredity.

109

J\':I?, to :wrest the general organization somcwhat short of f u l l lwfection. Jvc can also mdcrstand that the power to develop perfccily :ind to assnmc the charactcristics of the species niiglit rcmain Iatcnt in tlic female, and might come into action aftcr the loss of rcprodnctive power. Accorcling to this view, thc possession of a beard must be rcgmlcd as a general cliaracteristic of our race, inhcritod by all children, girls as well :is boys. The derclopmcnt, in the girl, of the fenide rcprodactive fnnction, or the 1;iclr of the stimulus which comes, in the male, from the dcvclopmeiit of tlie malc fnnction, arrests the dcrelopmciit of tlie beard, :iltho~igliits po\vcr for growth may rcninin htcnt, and may come into more or less perfect activity aftcr the period of reproduction is past. A careful cxamination of tlic examples given above mill bring out the interesting fact that when a female, from disease or mutilation or old age, assumes a resemblance to tlic male, the cliange is an advance, nnd consists in t he acqnjsition of structures not usually present i n the female. When, on the otlicr hand, the male, from castration or confinement, comes to resemble the fcmJC, the resemblance is due, in most cases, to arrest, or a failnre of the malc to acquire the adult male characteristics of the species. Sinipson (Hermaplwoditism, Cyc. of Annt. and Phys., Val. ii. 11. '719) gives the following summary of the subjcct: " The consideration of the rnrious facts that we have now stntcd inclines us to the belief that the natural history c1i:irnctcristics of any species of animal are certainly nct to bc songlit for solely either in the system of the mclc or in that of the female; bnt as Mr. Hunter pointed out, they are to be found in those properties that are

110

Heredity .

common to both sexes, and which w Iiaw occasionally seen combined togetlicr by m ~ t u r cupon the bodies of hermnplirodites, or evolved from tlic intcrfercnce of art upon a castrated male or a spaled female. “ I n assuming at tlic age of puberty tlic distinctire secondary peculiarities of his sex, tlie male, so f:ir as regards tliese secondary peculiarities, CT idently passes iiito a higher degree of dcvelopmcnt tlian tlic fcm:ilc, and lcavcs her more in poswssion of those cliamcters that arc common to thc young of both sexes, and which he himself never loses when his testicles are early removed. Tlicse and other facts connected with thc CTOlntion of both the primary and the secondary pccnliarities of the sexes farther appear t o us t o show that, ph~siologicallyat least, we ought to consider tlie male type of orgnnization to be tlie more perfect, as respects the individual, and the female as respects the species. Hence we find that, when the femalc is malformed in the sexnal parts so as to resemble the male, the m d formation is almost always one of excessive development, and, on the other hand, wlien tlie male organs are malformed in sncC a manner as to simulate the female, tlie abnormal appesrance is generally to be traced to a defect of development. I n tlie s:bme WRY, alien tlie female assnmes the secondary characters of the malc it is either, first, when by original malformation its own ovaries and sexnaI organs are so defective in structure as not to be capable of taking a part in the function of reproduction, and of exercising tliat inflnencc over the general orgnnization which this faculty imparts to them; or, secondly, when in tlie eoiiree of Rgc the ovaries hare ceased to be capable of pcrforming the action allotted to them in the reproductive process. I n both of tliese cases we obserre the powers of the female organization, now that its

Various Opinions o n Heredity.

111

capabilities for performing its particular office in the continuation of tlic species are wanting or lost, expend themselves in perfecting its own individual system, and herice the aninial gradually assumes more or fewer of the sceondary sexual characters tlirit belong to the male.” It is true that, in a few instances, the male has been knowii to acquire true feminine characteristics, foreign to normal males. Thus, according to Darwin, “characteristics properly corifined to the female are likewise acquired: the capon takes to sitting on eggs, and will briug u p chickens; and what is more curious, the utterly sterile male hybrids from the pheasant and fowl act in the same manner, their delight being to watch when the hen 1e:tves the nest, and to take on themselves the office of a sitter. Many male birds normally sit, and hatch the eggs, and there are reasons for believing that the incubating habit was originally shared by both sexes, and I am therefore inclined to attribute such cases as this to reversion to a remote male ancestor, rather than to the acquisition by tlic male of a female characteristic. We may conclude, then, that the transmission by one sex, in a latent condition, of the secondary characteristics of the opposite sex, does not compel us to believe in the dual sexual personality of each individual, since we have a much simpler explanation in the view that each embryo inherits the power to derclop all the characteristics of the species, but that this power does not fttlly manifest itself in the female. It may seem di6cnlt to explain in this way the transmission by a bull of the good milking qualities of his mother, or the capacity occasionally shown by male ma!nmnls of yielding milk, but it is surely simpler to a ~ n i i i c that each male inherits, like the females, the

112 power of devclopiag perfect functional mammze, and that this power is arrested in the male, than to assume that each male animal includes in itself a complete female duplicate. An illustrationmay make the subject more clear. Certain embryo bees, when exposed to certain conditions, develop into sterile workers, but when exposed to another set of conditions they become fertile females. The differences between the workers and the queens are not confined to the reproductive organs, but extend to the shape and size of the body, the general organization, and to the instincts of the animals. These differences are not due to the direct action of the conditions to wliicli the young are exposed, but are truly hereditary, as we see from the fact that the workers of different species are as distinct and as characteristic of their species as the maIe or the fertile females. Now which is simplest, to assume that each female embryo has a complete worker organization and a complete queen organization, or to hold that it has the power to develop a11 the characteristics common to both, and also the distinctive characteristics of each; that one set of conditions suppresses the distinctive characteristics of a perfect queen, while another set of conditions arrests those of a perfect worker ? The argument in favor of the multiple personality of individuals which is furnished by polymorphic comniunities is at least as strong as that furnished by the latent transmission of secondary sexual characteristics. I n the case of the polymorphic hjdroids an egg-embryo may give rise, by budding, t o certain descendants with fully developed digestive organs, but with no organs of locomotion or reprodiictive organs, to other descendants with organs of locomotion, but without diges-

Various Opinions o n Heredity.

113

tive organs or reprodnctive organs, and to still others with rcprodnctive orgalls, but with no organs of digestion or locomotion. All these forms are hereditary and arc characteristic of the species, so there is no escape from tlie conclusion that they a11 are present in some form i n the egg-embryo, and i t is certainly natural to snspect that the entire organization of each one of them is latent in this embryo, but the explanation which I 1iay-eproposcd to account for tlie transmission of secondary sexual characteristics, applies to such cases as this jnst as well. The hypothesis that the egg-embryo inherits and transmits to each of its descendants, those produced asexually as well as those produced sexually, all the characteristics of the species, and that it also inherits and transmits to each of them a tendency to suppress certain of these cliaracteristics under certain conditions, seems to furnish a simple and satisfactory esplanatioil of all tlie facts. According to this view the feeding zooids of a polym orpli i c Si 1’11on oph ore are in d i v i d n a1s mli i ch h ave i nheritcd in full all the cliaracteristics of the race, but which do not attain to perfcct development in all respects. The swimming zooids are similar individuals, with other characteristics suppressed, and so on. This explanation seems mnch more satisfactory than the snpposi tion that the egg-embryo contains one complete personality for feeding zooids, one for locomotor zooids and one for reproductive zooids, and J hope that this case will make clearer the lack of necessity for asassuming the dual personality of each malc or female animal, so long as we liave a mnch simpler explanation in tlic liypothesis that each embryo lias the power to develop all the clinracteristics of the species, together with a tendency to suppress certain ones in each sex.

114

Heredity.

A little thought will show that if there were no explanation of the transmission of latent sexual characteristics more simple than the hypothesis of a dual personality, this hypothesis would then be too simple, and would nced to be made much more complicated. The characteristics of the opposite scxare not the only ones which may be latent, aiid in cases of reversion it parent may transmit to children characteristics which were exhibited by neither parent nor grandparent, and whic!: may have remained latent for many generations. If we must assume the existence of a dual personality t o acconnt for the latent transmission of the charncterietics of the grandparent of the opposite sex, we must assume still other personalities to account for rel-ersion t o more remote ancestors, and Darwin has not hesitated to carry the hypothesis to this, its logical conclasion. He says (Variatior~,ii. 65), " Smeral a u t h o ~ sl m e maintained that hybrids and mongrels include all the characteristics of both parents, not fused togetlier but merely mingled in different proportions in different parts of the body; or, as Naudin has exprcsscd it, a Iiybrid is a ]iring mosaic work, in which the cye cannot distinguish the discordant elements, so completely are they intermingled. We can Iiardly doubt that, i n a certain sense, this is true, as when we behold in a hybrid the elements of both species segregating thcmselocs into segment in the same flower or fruit-bya process of selfattraction or self-affinity-this segregation taking place either by seminal or by bud propagation. Naudin further believes that the segregation of two specific elements or essences is eminently liable to occur in the male and fenmle reproductire matter, and he thus explains the almost universal tendency to reversion in successive hybrid generations. . But i t would, I suspect,

. .

Various Opinions

on

Heredity.

115

be more correct to say that tlie elements of both parent species cxist in every hybrid in a double state, namely, bIcnded together and completely separated.” In anotlier place (J’hrintion, ii. p. 80) he says: “ O n tllc doctrine of reversion, as given in this chapter, the germ becomes a far more marvellous object, for besides the visible changes to which it is subjected, we must believe that it is crowded with invisible characteristics, proper to both sexes, to both the riglit and left sides of the body, m d to a long line of ancestors, male and fcmale, separated by hundreds or cveii thousands of generations from the present time, aud these characters, like those written on paper with invisible ink, all lie ready t o be cvolrcd under certain known or unknown conditions. ’’ I shall discuss the phenomena of reversion somemhat at length i n another place, mid wish to simply call attention a t present to the fact that here, as in the case of secondary sexual characters, me have a much simpler explmntion in tlie liypotliesis of arrc-t, and tlierefore do not necd to call in an iiiiknown factor, such as the multiple 1~ersonalityof each individual. I tliitik that the plienomcna of alternation of generations favor this latter supposition even more than the facts of reversion. The egg embryo of a hydro-medusa may give rise by budding to an indefinite number of hydroids like itself, arid each of these may give rise to other hydroids, and so on indefinitely. Em11 one of thcse may also, under certain conditions, give rise to medusa quite different from the hydroids and like tlic original medusa. As the medusa which are thus produced inherit through a long series of hydra ancestors all the specific characteristics of the origi-

116

Heredity.

we are forced to conclude that e:ich 117 clrciicl ~oiitaiiis,i n a latent state, the 1)ower to rcprodnc~a dcGnite specific medosn. As the hydra and its rncdusn differ from cach otlrcr rcry milch more than :Lmalc and :t female ni:iiiim:il. xntl have little i n common except the general plan of tlicir org:inization, there seems :kt first to be 110 r e c a p ft 0111 tlic coticlusion that the mcdasa structnrc exists side L? side with the hjdra striictiire, in each lijdruicl, as :i sccond personality. I hope to show, in the chapter on asexual rcprotliiction that alternation of gcncriitions is a seco1iii:ii.y condition of things, and tliat i t lias been bronght abutit Iiy a modification of ordinary nietamorpliosis. I think there is every reasoil to Lelicve t h t a t one time the hydra-larva wliich liutched from a rnetltis:i c ~ bcg came rnetamorphoscd, by a gmdual cliauge during gron tii, into a medusa. If this mere the case now, there mould be 110 more reason for believing in D hgdra pcrhoti:ility nud n rncdus:~ personality than thcre is for belie\ing tliiit a 11li111i~t1 child contains a distinct adnlt personality. Now me can understand that if snch D 1nrr-s slioirl~l give rise by budding to otlicr hydroids like itself, tlioy also wotild have the power t o grow into matnre medusae. We can also understand that circumstances might wise to cause the later stages in the development of some of these hydra-larv2e to become latent. We should tlrrn have two generations-hydroids without a nicdtisn st .get and hydroids with a medusa stage. The suppression of tho ligdra features of tlic l#ittcr would tlien gire 11s a generation of niedosre w i t h no liydra stage, giving birth t o a generation of liydruids with no medusa stage, and these iu turn producirig a

11111 medusa,

Various Opinions o n Heredity.

137

generation of mednsz with no hydra stage. We should then linvc a case of alternation like that which is presented by ordinary hydro-niednsae.

Szwnmwy of Chapter. A careful review of the reasons mhicli have induced various authors to believe that either sexnal element may transmit any cliaracteristic whatever, leads to the coiiclusion that its truth is not proven. It is impossible to prove i t by the phenomena of crossing, since the only animals mliicli can be made to cross are essentially alike, and differ only in minor points. The homology bet&en the ovum and the male ccll is no reason for supposing that their functions are similar, and the differences between them should lead us to believe that their functions are not alike. There is no reason for assuming that each sex transmits its entire organization to the offspring, in order to acconnt for the latent transmission of secondary sexnal eh:ir:~eteristics, since this triznsmission can be more simply explained by assuming that each embryo inherits but docs not necessarily develop all the characteristics of its Fl’ccies. Reversion and alternation of generations admit of a similar explan at’ion. We may therefore conclude that there is and can be no proof that each sexual element transmits d l the characteristics of the parent, and that there is no a priori absurdity in the liypothesis that the male and female reproductive elements are unlike i n function, and are specialized in different directions. We can therefore enter without prejudice into an examination of the evidence for this latter view.

.

CHAPTER VI. THE EVIDENCE FROJZ HYBBIDS. Importance of the subject-It furnishes a means of analyzing or isolating the influence of each sexual element-Hybrids very variable-Hybrids from domesticated races more variable than those from wild races-The descendants of hylirids more variable than the Iiybritls themselves-The offapring of a m:dc lixbrid and the female of a pure species arc miicli more vari'ible than those of a frmnle hybrid and tlie male of a pure spccies-Tl,ese facts inexplicable on any view, except tlie one lirre prcseirtccl -Reciprocal crosses-They differ in fertility and in strncture --The difference is exactly v h a t our theory requires-Difficulty in cxplnininq transmission of c1i:ir:ictcrs tritliout fusionReversion cilused by crossiiig-Two kiuds of reversion-Sum-

mary.

THEstndy of hybrids and crosses is of especial interest t o us, sincc it affords 11s a means, somewhat imperfecl, it is true, for recognizing, i n the offspring, the structure which it owes to each parent. I n ordinary sexual rcproduction between animals or plants of the same race, the parents are almost exactly alike, except for their sexual differences; and as nearly erery structural feature of the young is a fcatnre of rcsemblance t o each parent, there can be nothing t o show that it is inherited from the one rather than from tlie other. T h e n distinct races or species arc crossed, the case is somewhat different. I t is true that thc two parents are still very much alike, for species cannot be m:ide to brccd together at all unless they arc very closely rehted. Still they are more different from each other than indiridnals

T h e Evidence from Hybrids. -

119 ~

of the same species, and the study of crosses and hybrids is therefore a means of separating, to some extent, the iiifliience of one parent from the influence of the other. Tliis is true, however, only with reference to characteristics which are of recent acquisition, for the greater part of the history of two allied species has been the same, and they show in comnion everything except wliat lias been acquired by each one since they diverged from their common ancestor. Crossing gives no way of showing whether these comni& characteristics arc or are not transmitted by one parent or S11e other or by both, but it does give us this iiiforination regarding characteristics which appear in one species but not in the other, and it is therefore the best means at oiir disposal for studying the influence of each parent upon the offspring.

Crossing as a Came of Variation. Accordiiig'to our theory of heredity, we can easily see how the crossing of two species or varieties shonld lead to variability, for when two species or varieties are crossed certain cells of the body will be hybrids between tho gemmules of the male parent and the ovarian particles inherited through the female from the egg of the preceding generation. Now the ovarian particle transmits the properties of a cell like that of the female parent, while the gemmule transmits those of a corresponding cell in the father. I t is plain that corresponding cells of a female of one species or variety and of a male of anotlier species or variety must be more different from each other than corresponding cells in a male and female of the same species or variety. The hybrid cell formed by their union would, therefore, be expected to differ more from each of them, that, is, to vary more than it

-~ ~

~

does in the offspring of parents of the same variety. It is weli known that this is the case; that, in domesticated animals and plants a t least, crossing is a great canseaccording to some older writers the only cause-of variution. Darwin says that it is probable that tlie crossing of two forms when one or both have long been domesticated or cnltivated, adds to the variability of the offspring, independently of tlie commingling of the characters derived from the two parent forms. He believes that new cliaracters arise in this way in hybrids betmeen domesticated forms, forms which Iiave been rendered mriable tlirougll cultiration, but he doubts whether we have, at yresent, sufficient evidence to prove that the crossing of species which have n e w been cultivated lcads to the appearance of new characters. The following illnstrations of this lam are quoted from his Vkriuflo?~ (Vol. ii. 13, 319): ''Gartner declares, and his experience is of the highest d u e on such a point, that when he crossed native plants which had not been cnltiv:ttcd, hc never once saw in the offspring any new character; bnt that from the odd manner in which the characters deriwd from the parents were combined, they sometimes appeared as if new. When, on the other hand, he crossed cultivated plants, he admits that new characters occasionnlly appeared. . According to Kijlreuter, hybrids in tlic genus Mirabilis vary d m o s t infinitely, and he describes new and singular cliaracters in the form of the E C C ~ S , in the colors of the anthers, in the colyledons being of immense size, in new and liiglily peculiar odors, in the flowers expanding early in the E ~ B S O I I ,and in their closing a t night. With respect to one lot of these hybrids he remark6 that they presented characters exactly the reverse

. .

T h e Evidence from Hybrids.

121

of what might have been expected from their parent-

age.

‘‘ Professor Lecoq speaks strongly to tlie same effect in rcgard t o this same gcnus, and asserts that many of the hybrids from Mirabilis j a l a p and niultiflora might easily bc mistaken for distinct spccies, and adds thal, they differed in a greater degree tlian the other species of the gcnus from M. j a l a p . Herbert has also described tlie offspring from a hybrid Rhododendron as being as unlike all otlicrs in foliagc as if thcy had been a separate species. The common experience of floriculturists proves that t h e crossing and recrossiiig of distinct but allied plants, such as the species of Petunia, Calceolari:i, F ~ i c l i ~ i a , Verbena, etc., indnces excessire variability: hence the appearance of quite IICW cliaracters is probable. M. Carriers has latcly discussed this subject; he states that Erytlirina cristagalli 11:d been multiplied by seed for niany yczirs, b u t has not gicltled any varieties; i t was tlien crossed with the allied E. hcrbacia, and tlie resistance was now overcome, and varieties were produccd with flowers of extremely different size, form, a n d color.’’ Darwin, tlicrefore, concludes t h a t crossing, like any otlicr change in t h e conditions of lifc, seems to be a n element, probably a potent one, in causing variability. The variability of hybrids is qnite as explicable by Dnrwin’s Pangenesis hypothesis as i t is by our theory of hercdity, nltliough I do not see why, on the hypothesis of pangenesis, the hybrid offspring of domesticated forms should be any more variable than those produced between wild species.

122

Heredity.

The Ofspring qf Hybrids inore variable than the Fimt Geizemtion There is another aspect of the variability of hybrids which is very remarkable, and wliiclt is in perfect ngreement with our theory of ltcrcdity, but, so far as I am aware, absolutely inexplicable withoiit it. This is the law that although the offspring of the first generation are generally uniform when two species or races are crossed, the subsequent generations of children produced by these hybrids display an almost infinite diversity of character. (Darwin, Variation, ii. p. 321.) Darwin also refers to this curious law in the Origin of Species, p. 260, and attempts an explanation of it. He says: “ The slight variability of hybrids in the first generation, in contrast wit11 that in the succeeding generations, is a curious fact, and deser-res attention. For it bears on the view which I have taken of one of the causes of ordinary variability, namely, that the reproductive system from being eminently sensitive to changed conditions of life, fails under these circumstances toperform its proper function of producing offspring closely similar in all respects t o the parent form. Now, hybrids in the first generation are deecended from qecies (excluding those long cultivated) which liave not ltad their reproductive systems in any may affected, and they are not variable; but hybrids themselves have their reproductive systems seriously affected, and their descendants are highly variable.” According to this view, the variability of the descendants of hybrids’is a sort of monstrosity, due to the failure of the reproductive organs to perform their proper functions; ordinary variability is not monstrosity, but is perfectly normal, and as the variability of hybrids

me

Evidence from Hybrids.

123

has precisely the same character, I think we cannot regard it as due to unnatural disturbance. According to our theory, nuithion is due to the action of cliatngcd or unnatural conditions upon certain cells of a preceding generation. Now, :is characteristics of both parents are mingled in a hybrid, i t must nearly always happen that certain cells with peculiarities of one parelit mill be in contact with, or will depend in some way upoti, cells with peculiarities inlierited from tlie otlier species. Tliere will therefore be a lack of the perfect adjustment between each cell and its neighbors, tvliicli has been brought about in each parent by natural selection, and this imperfect adjustment will cause the cell wliiuli is unfavorably placed to tlirom off gemmules. The cells of the body of a hybrid will therefore be unusually prolific of gemmules, and will transmit variability to later generations. According to our hypothesis, a hybrid is more likely to transmit Variability than a pure species, because more of its cells are placed under circumstances favorable to the production of gemmules. For tlie same reason a hybrid between two domcsticnted or cultivated forms must have more tendency to vary than one produced by crossing two wild species, for tlie domestic or cultivated parents live under unnatural conditions, and therefore have more tendency than wild species to transmit gemmules, and thus cause variability.

The Xex of tlbe Parent affects the Variability of Hybl.ids.

I hare shown that the body of a hybrid is peculiarly favorable for the production of gemmules, and that, for this reason, the descendants of hybrids are variable

124

Heredity.

to an unusual degree. Now, if our theory of heredity is true, if the seniirial flnid is especially adapted for the transmission of gemmules, ivlijle their tmnsmission by:)ii ovum is a niattcr of :tcidcnt, the tendency to Yary must be transmitted by the mule hybrid. Wlien children are born from two hjbrid parents i t is impossible to show that the variability which follows comes from the father rather than from the mother, but the subject can be put to a test by crossing tlie male hybrid with a female of one of the pure species, and the male of one of the pure species with the female hgbrid. Neither pure species has any especial tendency to transmit variation, mliile the niale hybrid has such a tendency. If, then, we cross tlie female hybrid with the male of one of tlie pnre forms, the offspring wonld not be expected to he ~~nus u a l T-nriable; ly but if the male hybrid is crossed with one of the pure females we should expect the offspring to be 1111 u sn:dly variable. Now it is very interesting to find that this actually is the case. T ~ L Gartner U states (Rnstni.derze.ligzi~zg,p. 452, 50‘7) that when the seeds of Dianthus barbatus were fertilized by the pollen of the hjbrid Dianthus chinensibarbatus, the seedlings were more wriable than those which were raised from the seeds of the hybrid fertilized with the pollen of Dianthus barbatus. Darwin states tliat Max Wichura obtained the same result with willows. Giirtner concludes from a number of experinleiits tliat when a hybrid is used as the father, and either one of the pure parent species or a third spccies as tlie mother, the offspring are more ~ a r i a b l etliaii when the same hybrid is used as tlie mother, and either piirc parent or tlie third species as the father. Darwin’s pangenesis hypothesis f urnishes n o explanation whatever of this curious fact. On the contrary, as

The Evidence from Hybrids.

125

i t requires t h a t each sexiial clement slionld contain gemmules from cvcry part of t l ~ ebody of tlic parent, it is directly opposed to any such result, and tlierc is 110 place for i t in any other hypothesis of hcrcdity. Oiir theory fits i t exactly, however, and a more crucial test could hardly be proposed than an experiment like those detailed by Gartner.

Rec roca 1 Bybl'ids. According to Darwin the two sexes play similar parts i n heredity, and any characteristic whatever may be transmitted by eitlier sexual element. This conclusion is based upon tlie phenomena of crossing, lint a little thought mill show t h a t it is impossible, from the nature of the case, to prove it from e d e n c e of this kind, althongh, as I hope to show, i t is capable of disproof. Only animals of t h e same species, or of closely related spceies, can breed together. Closely allied animals are alike i n aII respects, except as regards tile slight diffcrences wliicli distingiiish species, varieties and indiriclunls from each other. Sincc no animals or plants can cross except those mliich haye most of their past history in common, and whicli arc tlicrefore alike in nearly e n x y respect, it is plainly impossible t o prove, from tlie plicnornena of crossing, tllat encli p r e n t has power to trmsniit the features mliicli are shared by tlie other parent as well. The phenomena of parthenogenesis, or reproduction by virgin femnles, as in the case of bees and ~vasps,show t h a t the ovum alone may transmit all the established hereditary structure of tlie species, bnt there is and can be no evidence to sliom t h a t tlie malc element can accomplish the same thing. The facts of crossing, while they cannot prove that the

126 fnnctions of the t v o reprodnctive elements are alike, d o fornish convincing proof of the contrary, and show that they are not alikc. A reciprocal cross is a donblc cross bctweeii two spc. cies or varieties, one form being used in onc case as the father, and in the othcr case as the rnothcr. Thus A reciprocal cross between a horse and an ass is a double cross, between the male horse and the feniale ass on tho one hand, and the female horse and m& ass on the other. Now, if it is trne that the fuiietion of the oviim is like that of the male cell, the offspring of rccjpi.oca1 crosees shonld be d i k e in all respects, but this is by no means the case. I n the first place, the degree of steriIity often differs greatly in two species when reciprocally crossed ; for the male of the f i r s t will, in some cases, readily fertilize the ovum of the second, and thus give rise to dcscendants; whilc hundreds of attempts to fertilize the ovum of the first by the male of the second, result i n nniform failtire. It often h:qqxms also that even mlien both cross~sresult in the production of offspring, the lijbrid in the one case is sterilc, while in tlie other case it is perfectly fertile. Not only do the results of reciprocal crossing sl~orvt h i s difference, but they sliom what is still less reconcilable with the view that the fnnctions of the sexual elements arc alike, namcly, great differcnces of structure. In some cases where a rcciprocd cross is ~)crfectly fertile on both sides, the Iiybrids mliicli are thns produced are not a t all alike. When tlie male of species A a i d the femdc of B are crossed, the offspring is an entirely difftwnt being from thc one born from A as a mother w i t h B as a father. We know that allied species of animals are the deecend-

The Eoidence from Hybrids.

127

ants of a coninion ancestral form, from which they iiilicrit all that tlicy l i a ~ ei n coninion, while tiic distinctirc ~~cculiarities which distinguish them from each otlier arc niorc recently acquired. According to onr hypothesis tlie ovnin transmits estublislied characteristic?, while the cells which have rcccntly varied in the body of the male transtnit gemmules. If, then, me sclect two allied species or varieties and cross tlie male of one with the female of the other, and tlicn, reversing the process, crom the female of the first form with the male of the second, we should expect to find, in many cases, a difference in tlie offspring. Wliere tlie male of species or variety A is crosscd with the female of B, the offspring will inherit from its mother the common characteristics of both parents, a d i t will also receive from its father gemmules from those cells ~vliiclihave recently varied in the species A. The corresponding cells of its body will therefore be hybrids, and will bear a closer resemblance than the other parts of its body to the species A. That is, the hybrid will sliare, to some extent, the peculiarities which are distinctive of the species A as compared with B. The offspring of the opposite cross will, on the other hand, join, more or less perfectly, to the common race characteristics, some of the distinctive peculiarities of the species A r rod aced in it by tlie hybridization of the cells of its body by gemmules received from its father. Iteciprocal crosses between the horse and the ass have bccn reared for domestic purposes for ages, and Huxley gives the following iiiteresting account of the result: “ T h e offspring of the ass and the horse, or rather of tlie he-ass and the mare, is what is called 8 mule; .and, on the other hand, the offspring of the stallion and the

125

Heredity.

she-ass is what is called a hinney. It is a very rare thing i n this conntyy to see a hinney. I never saw one myself; but they have been very carefully studied. Now the curious thing is this, that altliongh you have the same elements in the experiment in each case, the offspring is entirely different in character, according as the male influence comes from the ass or the horse. When the ass is used as the male, as in the case of the mule, you find that the head is like that of the ass, that the eam are long, the tail is tufted at the end, the feet are small, and the voice is an unmistakable bray; these are all points of similarity to the ass; but, on the other hand, the barrel of the body and the cut of the neck are ninch more like those of the mare. Then if y o ~ look i at the hinney-the result of tlie nnion of the stallion and the slie-ass-then you find it is the horse mliieh has tlie predominance; that the head is more like that of the hdrse; the ears are shorter, the legs coarser, and the type is altogether altered, while the voice, instead of being a bray, is the ordinary neigh of the horse. &re, you see, is a most cnrious thing; yon take exactly the same elements, ass and liorse, bat yon combine the sexes i n a different niaiiner, and tlie result is niodified accordingly. " It would certainly be a wonderful thing if the conibination of the same elements should give such different results, and I think we must conclude that the elemcnts are not the same, but that the ovum and the male cell do not play the same parts in heredity. There are not many cases in which reciprocal crosses have been made so frequently, and single obscrvations are not of very great value. I will, however, cite a few, to show that the one given is not cxceptional. The Nnnx cat is a variety of the domestic cat 1ieculi:ir to the Isle of Man. I t differs from the ordinary cat i n Laving no tail,

The Evidence f r o m Bybrids.

129

a n d in some other slight pecn1i:trities; its hind legs are longer, and its habits peculiar. According to Mr. Orton (Physiology of Breeding, 1S55, p. 9; quoted by Darwin, T’cwintion, ii. SG), Dr. Wilson crossed :i male Miinx cat mitli common cats, and, out of twcnty three kittens, serentccn were destitute of tails; but when tlie female Manx mas crossed by conimon male cats all the kittens had t d s , though they mere gcnerally short and imperfect. Darwin gives tlic following in his Variation iuider Domestication (ii. 85): ‘‘ Godina lias giren a curious casc of a ram of a goat-like breed of slieep from tlie Cape of Good H o p , which produced offspring hardly to be distingnishcd from himself mlien crosscd with ewes of twelve other breeds. But two of these lialfbred ewes, wlicn p u t to a merino ram, produced lambs closely resembling the merino breed.” I quote tlic following from Darwin also (p. 8 7 ) : ‘‘ T h e silk fowl breeds true, and there is reason t o believe tliilt i t is a very ancient race; but when I reared a large numb ber of mongrels from a silk hen by a Spanish cock, not one cxliibited even a trace of the so-c:illect silkiness. Nr. Hewitt also asserts that in’ no instance are tlie silky featliers trarisniittcd by this breed when crossed with any otlicr variety. B u t three birds out of many raised by Mr. Orton from a cross between a silk cock and a bnnt:im hen liad silky feathers. Thei~eare some instances of reciprocal crosses which seem at first sight to give directly opposite results, arid therefore to contradict our theory. Thus Darwin says t h a t a hybrid which had for its mother a bay mare and for its fiither a hybrid between a male ass and a €emale zebra, had, when young, zebralike stripes npon its shoulders, flanks and legs. Here tlie only recent striped ancestor is t h e paternal grand-

130

Neredity.

mother. As the possession of stripes is a cliaracteristic which distinguishes the zebra from the horse and the ass, i t seems at first as if its transmission by a fcnxtle ancestor is opposed to our theory. We .know, home\er, that all the species of the horse genus are the descendants of a striped form, and the presence of stripes in the zebra is not due to recent variation, but to the fact that i t has not varied. The transmission of stripes by a female zebra is therefore nothing more than me might expect. We know, too, that both the horse and tlie ass show a tendency to revert to tlic striped ancestral form, and I s h l show in the next section that reversion is often excited by crossing. It is therefore quite probable that the stripes in this colt mere due to reversion. It is said that young animals born from a tigress by a male lion, as well as those born from a lioness by a male tiger, are striped, bnt many cat-like animals show a tendency to revert to a striped form, and in this case also we may explain the presence of stripes in the young by attributing it to reversion excited by crossing. Darwin says that a good authority assures hini that in South America, when hiata cattle are crossed with common cattle, though the niata is prepotent whether males or females are used, the prepoteiicy is strongest through the female line. The origin of the niata breed is not known, but there is no doubt that it originated in Paraguay from common cattle; and the fact that the niata pecnliarities arc not shared by any other living cattle, but are very much like those of the extiiict Sivatherium, seems to show that in this case also the peculiarity may be due to reversion to some remote ancestral form.

The Evidence from Nybrids.

131

Di$cdtjy of Zxpluiiziiag the Tyansinission of the CJLarn c f e m of Ti00 For.nzs ~ i t h o u tFusion. A mucli more serious difficulty is found in the fact that while a hybrid is usually somewhat intermediate between its parents, i t occasionally happens tliat the cliaracteristics of one or both parents refuse to blend and are transmitted in an unmodified state. Thus Darwin states that when gray and white mice are paired the young are not piebald nor of an intermediate tint, but are pure white or of tlie ordinary gray color. This particular case may perhaps be explained as follows : The bromn form is the ancestral form, and ivlien no hair gemmules are transmit ted tlie young are brown. All the hairs are homologous with each other, and are derived from the same part of tlie egg, and when gemmules are transmitted they may hybridize alike all the cells which are to form liairs, mid the hybrid animals will therefore be entirely white or entirely brown. I t is stated that when u black game fowl is crossed with a white, the young are either pure black or pure white. but this case is precisely like tliat of the mice. Darwin gives a number of interesting illustrations of this singular phenomenon, 'among which are the follow1ng:

TVlien tirrnspit dogs and aneon sheep, both of ivliich liarc dwarfed limbs, are crossed with common breeds, the offspring are not intermediate in structure, but resemble one parcn t only. W h e n tailless or Iiornless animals are crossed with perfect animals, it frequently but by no means invariably happens that the offspring are either perfectly furnished with these organs or are quite destitute of them. When Dorking fowls with five toes are crossed with

132

Heredity.

other brccds, the chickens often have five toes on one foot and four on the other. When the red flowered stock of Antirrliiniini is fertilized with tlie pollcn of tlic pwple Qneen stock, h u t half t h e seedlings resemble tlie inotlicr 11l:tnt, wliile t h e other lialf )Je:ir rich purple blossoms like tliose of the paternal plant. Darwin says t h a t lie fcrtilizcd the pnrplc sweet-pea, wliicli has a dark reddish-purple standard-petal and 1 iolet-colored wings and kcel, with pollen of the p i n t c d lady smcct-pca, mliicli lias a pale cherry-colored standnrd and almost wliite wings and kecl, and from the e:inie pod twice raiscd 1)lants resenibling both sorts, tlie greater n 11mbcr rcscni bl i ii g the fat 11cr. These cases are difficult to exl)lain, b n t tlie ldienonicna are so complicated that i t is l i a ~ d l ysafe to spccu1:ttc upon tlicrn until they are rc-csnn:incd by an observer who can devote liirnsclf to this subject cqwcidly. Some of them may be due to t h e c:iLiscs aboTe indicated, and sonie, possibly, to fertilization by two fatlicrs.

Crossing as n Cause of Recersion. According to Darwin’s view rcwrsion must in all cases be dne to the nianifcstation of a tendency which lias lain dormant in tlie cgg and lias bceu transniittcd for gcnerations in a latent condition, for tlic chances against tlic r c p t i t i o n , by an accidental variation, of a cliaractcristic of ii remote ancestor, arc inconccir:ibly great. According to our theor~l)tliisis not tlic case, for the conditions which causcd a cell in t h e anccstrd form to throw off genimnles and thns to produce a glvcn pecu1i:irity niity c;uise the corres1)onding cell of the parent to throw off gcinmulcs in the same may, and these, uniting with t h e .corresponding part of the cgg, will produce

The Evidence from Hybrids.

133

variation. As the gemmule and the ovarian element are both very similar to those wliicli produced the variation i n the ;~ncestor,the chmees arc not very great against t!lc rcl)rodiictioti of tlic same pecnliarity. In this ciisc IVC Jiould liave D new vmation with all the charactcristics of :I true reversion, but due to the transmission of a gcmmnle, rather tliaii to tlic sudden awakening of a tendency mhicli lias long Tain dormant in tlie egg. It is possible, therefore, tliat there may be two kinds of ixwxsion-truc hereditary reappear:tnce of features ~vliiclih a ~ lain c latent in the egg, and new variations ~ v l ~ i crepent li again certain old cliaracteristica of the race. Tliere arc, I tlilllli, certain reasons for believing tliat reversions of the latter kind arc the most common, the chief one beiiig tlie fact tliat most of the causes of variability are also c a p s of reversion. Tlins, crossing, wliicli is :t very efficient cause of vaiiiation, is also one of the chief causes of reversion. I);irwin gives a number of examides to show that, indcpcndently of the well-Bnown tendciicy of hybrids and mongrels to revert, after a nnmber of generations, to one of the parent forms, the act of crossing in itself gives a n inipnlse towards reversion, and often results in the reappcarance of long-lost characters. Tlic following interesting acconnt, from D;irwin’s Vnrintioir (Vol. ii. p. 57), will serve t o illustrate this law: “ In the chapter on tlic horse, reasons mere assigned for believiug that the primitive stock mas stripcd and d u n colored, and details were given showing tliat in all parts of the world stripes of a dark color freqnently allpear along the spine, across the legs and on tlie shoulders, mlierc they are occasionally double or treble, and even sometimes on the face and body of horses of all breeds and of all colors. But the stripes appear most

134

Neredity.

frequently on the various kinds of dun. They may sometimes plainly be seen on foals and subscqnently disappear. “The dun color and the stripes are strongly trnnsmitted when a horse thus c1i:iracterized is crossed with any other, but I was not able to prove that striped diins are generally prodnced from the crossing of two distinct breeds, neither of which are duns, although this does sometimes occur. “ T h e legs of the ass are often striped, and this may be considered as a reversion to the wild parent form, the Asinus f m i o p s of Abyssinia, wliicli is thus striped. In the domestic animal the stripes on the sl~oulderare occasionally double or forked at the estrenrity, as i n certain zebrine species. There is reason to believe tliat the foal is frequently more plainly strjped on the legs than tlie adult animal. As with the horse, I have not acqnired any distinct evidence that the crossing of differently colored varieties of the ass brings O U ~the stripes. “ B u t now let us turn to tlie result of crossing the horse and ass. Although mules are not nearly so nnmerous in England as asses, I have seen B much greater number with striped legs, and with the stripes far more conspicuous than in either parent form. Such mules are generally light-colored, and might be called fallow-duns. The shonlder stripe in one instance was deeply forked at the extremity, and i n another instance was double, thoLJgh united in the middle. Mi-. Martin gives a figure of a Spanish mule with strong zebra-like marks on its legs, and remarks that rnnles are particularly liable to be thus striped on the legs. I n South Amcrica, according t o Roulin, such stripes are more frequent and conspicnonsin the mule than in the ass. In the United States, Mr. Gosse, speaking of these animals, says that in a great

The Evidence from Hybrids.

135

number, perhaps in nine out of every ten, the legs are banded with transverse dark stripes.” Moles with striped legs can be seen in great numbers every d;iy in tlic strects of Baltimore, and the peculiarity is not in the least uncommon. Darwin gives a number of cases i n which the same reversion has been produced by the crossing of otherliorselike forms, and we must regard the tendency to revert to a striped form when crossed as characteristic of the horse family. Darwin says that when lie crossed diffcrent varieties of fowls he often got birds w i t h faint traces of the peculiar red p1um:ige of the wild GCLZZZLS bnizkivn, and that this pl timage mas almost perfectly reproduced in one magiiificcnt bird, the offspring of a black Spanish cock aiid a white silk hen, although either of these pure breeds may be reared by tens of thousands witliont the appearance of L: single red feather. Even long-lost instincts may be made t o reappear by crossing. The original mild ancestor of our domestic fomls must, like all wjld incubating birds, have had the incubating instinct. Now when two non-sitting breeds of fowls are crossed, the mongrcls frequentLy recover their incubating habit and sit with remarkable steadiness. It is said that hybrids between perfectly tame domestic :mirnals are often as wild as their wild ancestors. This has been noticed in cattle, pigs, fovls, ducks, and it is probnblc that the same thing frequently shows itself when widely separated human races are crossrd, as such good anthoritics as Liyingston and IIumboldt have remnrked uyon the savage character of half. caste liunian beings. Another intere,+ing resemblance between reversion and

136

Heredity.

ordinary variation is the fact that tlie descendants of hybrids are more apt to revert than the ligbrids thcmseives. Darwin says (Variation, 11. 65) that this is a general role. Now, whether rerersion be dne to the snddcn excitement of a tendency which has long been transmitted i n a dormant state by tlie ova, or whether it is due to t h e appearance of a new variation which rescmblcs a11 old one, we can readily understand how, according to our theory of heredity, crossing should call this poiver into action. During the evolution of t h e species eacli hereditary peculiarity has been cstablished i n tlie cgg by gemniules, and anything which prevents the egg from following its normal co~irseand developing the recently acquired c l i a r x teristics of the species, would sllow older characteristics to appear in their place. We know that animals which are w r y widely scparated are infertilc, and i r e can undcrstand tliat c w n when tlie difference betwcen two specics is not great enough to prcvent them from crossing, those cells of their bodies mliich have raried most may be so different from each other that gemmules from the one cannot fertilize tlie egg-particles which are to produce the other, or when tiley do fertilize them they may give rise to a variation which is so different from tlic 11orma1 cell that it cannot live. T h e cells which lmcede these in the order of growtli being less different in the two lmxwte, would be niuch more favorably situated, andwould thus give to tlie embryo a characteristic of longcr standing tlGn tlie peculiarities of eitlicr parent. On the other h:uid, if reversion is simply variation, me can sce that crossing might excite reversion just as i t cxcitcs Fariability.

The Boidence from Hybrids.

137

Sumrncmj of C l q t e r . The study of liybrids gives us a means of comparing, witliiii certain narrow limits, the parts which the t w o sexual elements play in,heredity. The influence of each sex can, in a certain sense, be studied by itself when a given species is used in the one case as the father of a hybrid, and in another case as the mother. The value of crossing as an experiment in heredity is greatly limited, hornever, by the fact that, although me can stndy the inflaeiice of one sexual element unobscured by the other element from thc Same species, it is obscured and complicated by the influence of this element from an allied species, and in all organisms which can breed together tlic rcproducti-re elements must be essentially alike. Hybrids do, however, present a number of peculiarities wliicli agree perfectly with what we should expect according to our hypothesis, and certain of these are hexpl i cabI e .ivi t 11o 11t it. I-Iybrids and mongrels are highly variable, as me should expect to bc the case, according to Darwin’s pangenesis liypotliesis. This hypothesis fails to account for tlic fact that hybrids from forms which have long been domesticated are more variable than those from wild species or varieties, or for the very remarkable fact that the children of hybrids are much more variable than the hybrids themselves. Onr theory not only explains the variability of hybrids, but it also accounts for the two latter peculiarities, for crossing will not give rise to a marked or conspicuous variation unless the hybrid inherits numbers of gemmules, and as domesticated animals and plants live under unnatural conditions they are more favorably placed than wild forms for the production of gemmules.

138

Her edity.

The body of a hybrid is in itsclf a new thing, and therefore in a certain sense nnnatural, and a male hybrid is, accordingly, more fitted for the production of gemniulcs than a male of a pure or unmixed race. When a male hybrid is crossed with the female of either pure species or with a third species, tlie children are much more variable than those born from a hybrid mother by a male of a pure species. It would be difficult to devise an experiment better fitted than this t o show that variation is caused by the influence of the male, and that the action of nnnatnrd or changed conditions upon tlie male parent rcsults in the variability of the child. The remarkable histor? of reciprocal hybrids is directly opposed to Darwin’s view that the functions of the two reproductire elements are esaentially similar, for in some cases it is impossible to breed from a female of one species by the male of a second species, wliile the male of the first species readily fertilizes lhe oviim of tlie second and gives rise to fertile offspring. Even wlicn both crosses are fertile the one is often much more so than the other. The hybrids of one cross often differ remarkably from those of the other cross in general structure, and in many cases they show, in addition to the common characteristics of both parents, a tendency, more or less pcrfectly pronounced, to develop the recently acqnircd characteristics of that species which is used as the fath er . This law is often obscnred by the appearance of rcwrsions, which are peculiarly apt to occur in Iijbrids, n n d by the presence, in certain cases, of a tendency f o r cncli pnrent to transmit its peciiliarities to the hgbrid, without fnqinn with t h n w nf t h o nthor nai-ent

Rnt a-Lon

ma

T72e Eoidence from Hybrids.

139

consiclcr tlic great obscurity and complexity of the case, and the great difficulty in conducting rigid cxperirnents, tlic balaiice of tlie evidence from hybrids seems to be greatly in favor of our view of the nature of heredity. It certainly presents features wliicli are ioexl)lie;ible in any other wtiy, :md perfectly simple and natural if our view is accepted.

CHAPTER VII. THE EVIDENCE PRO31 VARIATION.

Causes of variation-Clinnged

conditions of life induce vnriability-No particula~kind of cliange is necefsnry--vari:iLility i3 almost exclusively contiued to organisms produced from fertilized ova-Bud variation very rare-Ilktory of the Italian orange-The frequency of variation in organisms produced from scxual union, ns compared Tvitli its infrequency in those produced nscxunlly. rcreivcs n direct explannlion by our theory of Iicredily-Bud wriation more theory freqiient in cnllivated tlian in wild p1:rnts-Our would lead us to expect this-Clinnged conditions (lo not act directly, but they cause subsequeut generntions to vary -Tendency to vary is Iicreditnry-These facts perfectly expliaible by our tlieorj-Spccific clinractcrs more y:ui:ible t h n generic-Species of large genera innre v;iri;iblc tlinu those of sm;ill gcnera-A part tlcveloped in an utinsiuil way sexIiiglily variablc-Lnw of cqii:ible vnri:ition-Secondnry ual c1,:ir:icters v:iri~~)JIc--Nntur:i1 selection cannot act to produce pertii:incnt modificxtion unless Inany individuals theory i s tlic only expl;ination of tlie vary together-Our simultaneons variation of ~ n a n yindividoi~ls-This theory also simplifies the evolution of complex strnctures-Saltntory evolution-This is exphined by onr theory of lieredity --Correlated variation of liomoiogous parts-Parts confined t o males more variable tlian parts contined to feinales3Iales more variable than females-Summary of last two chapters.

TIM Cmses of Vnvintio?b. CERTAIN authors hare licld that rariitbility is a Iieccssary accompaiiiment of lqroduction; that it is determined by something mitliin rather than without the or-

The Evidence from Variation.

141

gmism, bnt Darwin, after long and careful study nf the snbjcct, rcachcs tlic conclusion that each variatiou is exci tcd by a chnngc of some kind in the environment. It is impossiblc to expose animals for any length of time to absolutely uniform conditions, and we therefore find that when careful attention is given to the subject, minute individual differences may be detected in animals which arc appwcntly most nniforni. A shepherd easily lcarns to recognize each sliccp in a large flock, and ants arc able to pcrccivc a difference between the members of their o ~ r ncomninnity ant1 those from another nest. I t is inipossiblc to show by direct proof tliat uniform conditions of life mould provent variation; but i t is qnitc possiblc to approach tlic subject from tlic other side, and to shorn t!iat slight external changes cause slight variability, and greatcr cliaiigcs greater variabil-

ity. Wild animals and plants mr y somewhat and hare indiridual peculiarities, for each one is under slightly different relations to tlic cxternd world from all the others, biit :is cotn~~~irect with domesticated species their contlitions of life arc very nniform. A wilJ iinimal has become habituated to the circumstances undcr which it lives, by exposure, for generations after gcuerations to the action of natural selection, and a host of competing animals tend to keep i t in its place, but domesticated animals arc protected from their enemics and competitors, they arc removed from their natnr:d conditions, and they are frequently carried from their native land and are exposed in other: countries t o unnatural food and climate. They are compelled to c1t:ingc their habits, and they arc never left long at rest, or exposed for any considerablc length of time to closely similar conditions, but they arc carried from dis-

142

Heredity.

trict to district, and their food and treatment varies considerably. We accordingly find that, with fcw exceptions, all our domcsticnted animals and plants rary more than their wild relations. Even tlie goose, one of the least variable of domesticated animals, varies more than almost any wild bird, and according to Darwin, hardly a single plant can be named, which lias long been prop:igated and cultivated by sced, tliat is not highly variable. These considerations force us to conclude tlia’t rariability is not a necessary contingent of reproduction, but that tlie production of tlie gemmnlcs which give risc to variation is excited by changes in external conditions, and we must agree with Darwin that “ it is probable that variability of every k i d is directly or indirectly caused by cliangcd eoiiditions of life; or to put the case under another point of view, if. it were possible to expose all the individuals of a species during many generations t o absolntely uniform conditions of life, there monltl be no variability.” TVlicn me come to examine tlie effect of different conditions of life we find thatwe cannot attribute the Tarinbility to one rather tliaii tlie other. The essential tliing is change, but not any particular kind of change. Variation is frequently caused by a change of climate, but this is by no means cssential, f o r most ciiltivnted plants yield more varieties when cultivated in their native country tlian when removed to other climates. (Darwin, Variation, ii. p. 310.) Change of food is often a cause of variation, but that this is not necessary is sliomn by the fact pointed out by Darwin, tliat fowls and pigeons arc the most variable of domesticated animals, although their foodis nearly the sitme as that of their wild allies, but is much less varied

The Evidence f

Tom

Variation.

143 -

than that which they would find for themselves in a state of nature. Excess of food often causes variation, yet the turkey and goose have been encouraged and tempted for generations to fecd to excess, and they have varied but little. These examples show that the chamcter of the change is uniinportant, and that variability cannot be attributed to the exclusil-e influence of any particular class of external conditions; that the exciting cause of variation is clinngc. but not any particular kind of change. Darwin quotes a number of cases to show how slight a chaugc may result in variability. Thus the mild horses of the pampasof South America arc of one of three colors, and the wild cattle are of one color; but when the same horses and cattle are domesticated, although they arc not confined, but aae allowed to run at large like the wild €orms, they entirely lose their similarity of color, and display the greatest diversity in tliis particular. I n India several species of fresh-water fishes arc rcarcd in great tanks as large as natural ponds, and they arc all very variable. Darwin quotesfrom Downing tlic statement that varieties of the plum and peach which breed truly by seed, lose this power, and like other worked trees give variable seedlings when grafted on another stock.

Variability almost Exclusively Conjined to Organisni Produced f r o m Fertilized Ova. The only method open to us besides the study of hyhrida for observing tlie influence of the sexes i n heredity, is by a comparison of sexual with asexual heredity. As I shall show in another place, all the various forms of asexual reproduction are so connected that we may pass from h i o n , or the formation of two new organisms

144

Heredity.

by the splitting of one old one, to parthenogenesis, or reproduction from unfertilized ova, without findiiig any important gap in the series, and wc may safely conclude that all these forms of reproduction are fundamentally alike. So far as regards the physical side of the problem of heredity, the only essential difference between asexual reproduction and sexual reproduction is the absence of fertilization or union with a male cell in the one case, and its occurrence in the other case. It is therefore extremely important to compare the two processes, in order to discover whether this physical difference is accompanied by any difference i n the result. I n the one case we have heredity with the male factor omitted, and in the other we have heredity with a male factor, and if there is any constant difference in the result, we may safely attribute i t to this factor. I n making this comparison we are almost compelled to restrict ourselves to plants, for although asexual reproduction is not a t all unusual in animals, it is restricted, with one exception, to animals which are not domcsticated or reared by man, and me thercfore know too little about the minute details of their life to make use of thcm for our purpose. The number of plants which hare been cultivated and carefully observed and studied by man is verygreat, and as most of them multiply asexnally by budding, as well as by fertilized seeds, we here have abundant material for comparative study, and it is well established by hundreds of thousands of observations that the prcsence or absence of the influence of the male element does have an influence upon the result of the reprodactire process, and that this result is exactly what our view of the nature of the process would lead us to expect. Plants produced from fertilized seeds differ from those

T 7 ~ eEoidence from Variation.

146

produced from bnds only in their greater tendency to riiry. Bud mriations do occur, but they are very uni i - i ~ i l ,while more or less variation in seedling plants is almost universal. As we suppose that any cell may, mlicii excited by unfavorable conditions, throw off genimules, the gemmnlcs may find their way, by a sort of accident, t o growing buds, and thus cause variation. We should thercfore expect bud variation to occur occasionally, but very much less freqwntly than variation in seedlings. This is so well known to be tlie case that many antliors have held tliat tliere can be no variation mithout sexusl union. Darwin has shown, however, by a long list of instances of bud variation in plants, that this is not absolntely true, and the weight of his authority has led to the almost universal acceptance of his conclusion that tliere is no essential difference between asexual and sexnal heredity. I shall discuss this conclusion a t length in another place, as I believe tliat the facts demand a n intcrpretntion which is somemhat different from the one mliicli Darwin furnishes. At present I simply wish to call attention to the fact that all authorities agree t h a t variation is almost infinitely more common i n sexual than i t is in asexual offspring. Asexual multiplication in animals is restricted to the lower forms which are of little use to man, and as tliese forms have not been domesticated and carefullyobscrvecl, our knojvledge of the rariability of organisms produced asexually is almost entirely derived from the study of pl;1nts. Thc only instance in domesticated animals of anything like asexual reproduction is the parthenogenetic reproduction of bees, and i t is therefore interesting to note

146

Heredity.

that the hive-bee is the least variable of all domesticated animals (Darwin, Variation, Vol. ii. p. 307). Darwin says (Variatioiz, Vol. i. p. 3 G O ) that lie procured a hive full of dead bees from Jamaica, mlierc they have long been naturalized, and on carefully comparing them under the niicroscope with his own bees, could not dctect a trace of difference. Wit11 plants it is well known to all cultivators that forms which are highly variable as seedlings can be kept perfectly true by asexual propagation, and we have Darwin's authority (Variatioiz, Vol. ii. I'. 307, and Vol. i. p. 429) for tlie statement that mliile hardly a single plant can be named which has long been cnltiratcd and p r o p agated by seed that is not liiglily variable, tlie total number of instances of bud variation is as notliiizg in comparison with seminal varieties. This contrast is the more remarkable when me recollect that in most of our cultivated plants the number of buds wliicli develop is thousands of times greater than the number of seeds which give rise to plants. It is clear that if the chance of variation were the same in both cases the number of bud variations would be thonsaiids of times greater than tlie number of seedling variations. If there were thousands of chances of seedling variation for one chance of bud variation, the number of bud varieties would still be equal to the number of seedling vane t ies. The fact that with all this probability in their fayor, bud rarieties are very rare as compared with seedling varieties, shows tliat the chance of bud variation is almost infinitely small as compared with the cliance of seedling variation. While we cannot deny that rariation may sometimes occur in organisms produced asexually, I think we are

The Eoidence from Variation.

147

jnstified in giving great emphasis to the lam that Tarisbility is a1most exclusively tlie cliaracteristic of organisms produced from fertilized ova. Di~rwinsays ( V w i u t i o i ~ Vol. , ii. pp. 361 and 377), “ When wc reflect on tlie millions of bods wliicli many trees have produced before sonic one bud lias varied, we arc lost in wonder m1i:tt the precise cause of each variation can be.” :‘ Habit, liowcver much prolonged, rarely prodaces any effect on a plant propigated by buds: it apparently acts only through succeEsive seminal generations.” ‘l‘lic curious history of the naturalization of the orange in Italy, quoted by Darwin on the authority of Gallesio (Theoria della RQivodwione Vegl, 1816, p. l%), is very interesting in this connection. Dnring inany centuries the sweet orange was propagated exclusively by grafts, and so often suffered from frost that it required protection. After tlie severe frost of 1709, arid more especially aftcr that of 1763, so many trees were destroyed that seedlings from tlic sweet orangc mere raised, and to tlic surprise of the inhabitants their fruit was fonnd to bc smeet. Tlie trees thus raised werc larger, more prod iictivc and hardier than the former kinds, and seedlings mcrc now constantly raised. Hence Gallesio conclitdes that much more was effected for the naturalization of tlie orange in Italy by the accidciital production of iiew kinds from seeds during a period of about sixty years than had been effected by grafting old varieties during many ages. It is liardly necessary to give other illustrations of this law, fur no one with any knowledge of the subject will be inclined to question it. It is strange that its significmce has been ovcrlooked, but this is probably due to the failure of students of the subject to perceive that it

148

Neredity.

is possiblc to bclicvc that tlic t m i s m i r s i o u uf I-ariaLili t y is the peculiar function of tlic ni:ilc ccll, mid iilso to :icknowledge that variation may occasioiiollg occur witlioat its influence. Oar theory that varintion is caused by tlic trausmission of gemmules, and tlint there is 110 espccitll xri’aiigenieiit for their transmission to buds or to unfertilized eggs, while there is a specia1 ndaptntioii which has bccii slowly evolved during the evolution of sex for trniisnritting them to fertilized eggs, gives us a simplc c.ul~l;~nation of tlie fact that while bud wriation is perfectly possible, i t is extremely rare as compared with the rariability of sexual offspring. Darwin has been lcd, through tlic study of nwi:ibility, to a conclusion ~vliicliis very pinch like the cslilalitition ivliicli is Iicrc presented. He sajs ( ITnrliltioa, Vol. ii. 13. 323) that ‘‘ we rnny infer from tlic occurmice of bud \-ariation tlint tlic affection of the fcm:ile clenient tliroiigli external conclitions m:iy iiidnce r ; n i d d i t y , for a bud seems 10 bc tlic ardogoe of 211 O Y l 7 k . But tlie itmle eleriaeiit i s nppai.eiitly nazicl~ qfteiier gffecfed by chanyed conditions, nt least iii n visible waiiiier, iliaii the ,female eleineiit 01’ ouule.” Brtd vnriatioti is mticli morc frcqiicnt jti cultirntcd plants than it is i n mild ones. Very fcw iiistanccs h i r e ever been observed in plants growing wild or iintlcr strictly natnral conditions, and Darwin states that ‘‘ bud variation is most comnioii in 17litnts nhich hnvc Lecu highly cultivated for a long time.” The adjustmcnt betmeeii a cultirntcd orginism and its artificial or unnatural enrironmcnt must, in most C ~ E C S , be less perfect than tliilt ivliicli lina LCCII sloivly cstablislied between a wild orgauism and its natural cnvironment. We should, therefore, expect domesticated and

The Evidence f r o n 17ar/aGoz.

149

cultivated forms to be more po?ifir of gcnmules than Tvilcl species. Tlic fact that Lad variation, like ordinary variation, is most common ii: cu!tivated fortiis, seems to shov t1i:tt the tcndency to vary is excited i n b~ids,as i t is in fcrtilizcd ova, by tlie inflnence of gemmules m1:ich are tlirown off by tlie cells of tlie body under new or unnutu r d conditions, and we can easily understand wliy it sIio:iId be niorc freqncnt wlicre gemmules are a’ouiiclant than in n forni with few gcmmulcs, for t h e chance in f:tvor of the accidental transmission of n geniriinlc to a growing or nascent bnd mill increase as tlie nnnilier of gcnimules increases.

Cliniiyed Conditions do mt act directly, b u t they cazis~

Subseqzierat r7eiieralion.s t o vary. This strange and, as I hope to sliow, higlily significant I:uv lias been noted by nimy obscr~.ers,and a long list of illustrations might be quoted. As Darwin points out, it is certainly a remark:tble fact that changed conditions should at first produce, so far as wc can see, absolutely no effect, bnt that they should snbsequently came tlie character of tlie species to cllallge. The late Dr. Jared P. lcirtland told me that for more than forty years he tried in vain to obtain varieties froxi the conimon red cherry, but that when a t last varieties began to appear the variability was yery great: tliat after it had once become established i t continued for many years with no diminution. It is well known that when new flowers are first introduced into gardens they do not vary, although all, with rarest exceptions, ultimately rary. Darwin, in his Variation, Vol. ii. p. 316, quotes the followiiig illustrations of this law: “Mr. Salter re-

150

Ner ed ity.

marks that every one knows that the chief dificnlty is in breaking tlirougli the original form and color of the species, and every one will be on the lookout for any natural sport, either from sccd or branch; that being once obtained, liowerer trifling the change may be, the result depends upon liimsclf. M. dc Joiiglie, with rcference to pears, s:iys tlie morc a type has entered into a state of variation, the greater is its tcnclciicy to continue doing so, and the more it IS disposed to vary still further. Vilmoriii says that when any particular variation is desired the first step is to get the plant to rary in any manner whaterer, and to go on selecting the most variable individuals, even though they w r y in the wrong dircction; for the fixcd cliaractcr of the species once broken, the desired variation will sooner or later appear. Darwin gives quite a list of authorities to ~ 1 1 that 0 ~ after English clogs have been bred for a fern generations in India they degenerate, not only in their mental filcnlties, but in form. According to Bachman, turkeys reared from the eggs of wild ones lose their metallic tints and become spotted with white in tlie third generation. - It will be seen from tlie instances which lime been given that the number of generations mhicli are csposed to the new conditions before variation is induced varies greatly. In the case given by Dr. Kirtland, fifty Tears elapsed before variations of the red cherry began to appear. I n the case last quoted, variation appearcd in the third generation, and Yarrell says that Australian diugos bred in the Zoological Gardens of EngI:ind, almost invariably produced in the first generation puppies marked with white and other colors. Sir Charles Lye11 mentions that some Englishmen en-

The Euidence from Vuriation.

151

p g c d in conducting the operations of the Real del 1\Ioiitc Company in Mexico, carried out with them some grcyliounils of the best breed to hunt t%e hares which abound in that country. It was found that the greyhonnda could not support -the fatigues of a long chase in this attenuated atmosphere, and before they could come np with their prey they lay down gasping for breath; bnt these same animals hare produced whelps, ~vhichhare grown up, and are not in the least degree incomnioded by the want of density of the air, but run down the hares with as much ease as do the fleetest of their race in this country. It is interesting to note in this connection that a tendency t o vary is strongly inherited independently of the inl%ritaiicc of any particular variation. Darwin believes that this tendency to vary may be transmitted by either parent, and he says (Vuriation, ii. 325) i t is certain that variability niay be transmitted through either sexual clement, whether or not originally excited in them, for Kijlreuter and GBrtner found that when two species mere crossed, if either one was variable the offspring were rendered variable. We have already pointed out that the crossing of species is in itself one of the most efficient causes of variation, and me can hardly base upon the observations above given the conclixsion that variability may be transmitted by either sex. The fact that changed conditions do not directly produce variation, but cause subsequent generations to vary, is precisely what we should expect, according to our theory: for a change in the environment of an animal or plant must disturb the harmonious adjustment which natural selection has brought about between the cells of its body and their conditions of life. Such a

162

Heredity.

change, if considerable, could hardly fail to affect certain cells unfavorably; and it woald tliereforc cause the production of gehimules, thus inducing variation i n later gener‘d t’1011s. We can also understand horn a tendency to vary may be lieredit:vy, for if certain cells of tlie body vary, tliey will exercise a disturbing effect upon adjacent or related cells, and these, transmitting gemmules, will hand on t h e tendency to vary to succeeding generations.

Secondary Laws of Varintioit. T h e law that variability is itself liercditary involves a number of secondary laws, all of wliicli find a ready explanation in our tlieory of heredity. Among these secondary laws is tlie law t h a t Y specllfic charaefers ore more vayinble than yeizeric clmmters.” Darwin 11asgiven the evidence of the existence of this law (“ Origin of Species,” p. 122), so it will not be necessary to discuss it, or to d o more than point out that t h e tlieory of heredity furnislics an explanation of it. Tlic cliaracters Tvliicli are common to all tlie spccies of n genus, and mliicli distinguish it from other genera, arc, as a rule, much older than those which dietingaish one species of tlic genus from t h e other spccies. T h e specific cliaracters or features which distinguisli each species of a genus from tlie others, are features mliich hare appeared as new variations since tlie time wlicn the various species diverged from tlie common ancestor from whom they inherit their conimon or generic characters. As spccific characters are of more recent acquisition than gencrio clinractcre, natural selection will hare had less time to act upon tlie former tlian upon the latter. T h e adjustment between a specific cliaracter and its environment will therefore be, as a rule, less complete and per-

The Evidence from Variation.

153

fect, and thc cells which are involved will therefore have a greater tcndciicy than those involved in generic characters to tlirom off gemniules. Tliese characters will tliercfore be more variable in the descendants than generic c1i:uxctcrs. Another lam,4tlie evidence for which is given by Darwin on page 4.1- of the “Origin of Species,” is that ‘(species qf fhe larger geiiera i 7 a each country vary more frepzrently than the species of the smaller genera.” When a country contains a great iiuniber of species of a genus i t is gciierally safe to conclude that they have recently varied and diverged from each other. As the tendciicy to vary is in itself hereditary, and as one variation is in itself a cause of other variations, our theory of heredity would lead us to expect species which have recently undergone considerable change to show a tendency to vary still further, and we should therefore expect tlie species of large geiiera to be, as a rule, more variable tlian tlie species of small genera, although there is no reason mliy this rule sliould be absolute. A still more interesting lam is that ‘ ( a part developed i n any species in a n extraordinary degree or inanner, in coinparison zuit7~the same part in allied species, tends to be highly vnrinble” ((‘ Origin of Species,” p. 119). When one species of a genus agrees with the other species in most particulars, but differs from them all in some one respect, we may conclude that the peculiar organ or feature has recently been modified. N a t u r d ‘selection has tlierefore had less time to perfect the ndjustnient bctwcen this part and the remainder of the body than it lins liad to perfect the relations between other parts, or between the same parts in the other species. This peculiar part will accordingly be in a favorable state for the production of gemmules, and it will there-

154

Heredity.

fore be more likely than a part which has not recently varied t o vary still farther. Walsh has called attention (“ Proc. Entomolog. SOC.,” Philadelpliia, October, 1863, p. 213) to what he calls the “Law of Equable Vnri:ition,” which is, “if any given chai.actcr is very variable iiz one speries of a yrozip, it will tend to 6e variable in allied q e c i e s , asid if any given character i s perfectly coilstant in one species of a group, it will tend to be constant in allied species.” This is by no means an absolute law, but simply a general rule. Darwinpoints out that something of tlic same kind occurs in domesticated races, and that in the forms which are now undergoing rapid improvement those parts or characters which are most valued vary the most. We can readily see that circumstances which cause a certain part t o throw oft getnmules, and thus induce variability, in one species, will be likely to produce the %me effect on allied species living under similar cii*cnmstances. We can also nnderstand that the divergent modification which has resulted in tile formation of several species or races from a parent form, will in itself be a cause of still further modification in the same general direct ion. Another well-known law, of which many examples will be given in Chapter IX. is that secondary sexual cliaracters are highly variable. I n the chapter on this subject I shall show that the distinctive sexiixl characters of a species are usually due to recent modificat’1011. Their g r p t variability is therefore due to tlic same cause as that which renders specific characters more variable than generic, and is exactly what our theory would lead us to expect.

The Evidence from Variation.

155

2CTrdurnl #election caniiof produce Race Motlijicaiion tinless the #ante Part tends t o vary i l l . a iVumber of Iiidiuiduals at the Sirvne Tints. Tliis argument, which seems to me to be the most importnnt one wliicli has ever beeii adduced against the theory of natural selection, was first adraiiced by a writer in the North B1itish Review in June, 1876. The anthor points out tliat since the cliance of surviral of m y lx~rticularindividual wliicli is borii is very slight indeed, the birth of an individual with any particular slight advant:ige, and its conseqnen t superiority over its fellows, would not be sufficient to over-balance the cliance of its destruction. The objection, wliicli is purely logical, and not experimental, will be stated at length in amtlier place. At present the fact that those who are best qualified to judge, Darwin among them, hare acknowledged its great weight, will suffice to shorn that it is a real and' valid objection, and that the foothold of the theory of natural selection wonlcl be greatly strengthened if we could show that the causes which produce variation act in such a way as to cause the mme part to w r y at tho same time in great numbers of individuals. According to our theory of heredity, this will generally be the case. We suppose that an unfavorable change in the environment of a particular cell causes this cell to throw 08 gemmnles. It is plain that a change i n the external world, which unfavorably affects any particular cell or group of cells in one individual, will usually affect the corresponding cells of other individuals of the species at the same time. JVlien any particular cell is prolific of gemmnles in one individual of a species, the same thing will usually be true of the same cell in other

156

Heredity.

individuals, and tlic corrcsponding cell will tliereforc bc a hybrid, and will tend to vary in many dcscendants. I n each of tliese descendants this hybrjcl will be composed of almost identical clements, aiid they will all tcncl to vary in the same or nearly the same manner; and as each variation causes other cells to throw otT gemmnles, the number of individuals which arc similarly modified will tend to increase from generation to generation, and natural selection will therefore act, not on a single exceptional individual, but upon a grcat number, all of wliicli are modified in essentially the samc way.

If Fariatioft is Purely Fortuitous, the Evolzction of a Corriplicated Organ composed of H a i q Pnrts by h-atural Selection demands a Period o j Time is almost Injinite. This obvious objection .to the law of natural selection lias been so frequently discussed that it is unnecessary to dwell upon it at present, cspecially as I shall examine it in detail in another place. At prcscnt I will onlycallattention to thefact tliatavariation inany part of a complicated organ will, in itself, disturb the harmonious adjustment of other parts, and mill thus cause them to throw off gemmules, and thus to induce variability in the next generation. The fact that change is needed in any part will be the cause of variation in this part, and the time which is needed to restore all parts of an organ to a position of equilibrium will thus be almost in6nitely redirced. The argument of those who hold that life has not existed upon the earth long enough for the evolution of :dl the adaptations of nature by the sclection of fortuitous variations will thus lose all its weight.

The Enidence f r o m Variation.

157

Saltatory Evolution. Darwin believes that the evolution of mild species is due, like the formation of many domesticated races, to very slow modifichon by the natural selection of great numbers of very slight and inconspicuous variations, b u t mAny other antliora have given reasons for believing that this is not the case. Many of our most peculiar domestic races have originated suddenly, and there are reasons for believing that the liistory of the evolution of each species is divided into periods of abrupt and extensive modification, alternatiirg with periods of comparative stahility. Tliis subject, likc those which have been biiefly noted in the last two sections, will be fully discussed in Chapter XI., :tnd I will only dwell upon it long enough at present to point out that our view of the cause of variation implies that any particular change should in itself be a fruitful source of still greater modification, so that as soon as a tcndency to rary becomes established it will continue to iiicrc;ise until an equilibrium is again established by the natural selection of tliose modifications which are adapted t o the environment.

Correlated Variation. Tliis snbjcct will be fully discussed in the chapter on homology, but a few words upon it will not be out of place 11cre. Damin, who freqnently nses the term, inclades under i t facts ivliicli belong t o two somew~11;ttdifferent c hsses. . Whcn any part varies, the oi.gnus with which it is most directly associated also tend to vary in such a w:~y:is to restore the harmonious adjustment bctwecn the various parts: and a variation in one part is often accompanied by mriation in homologous parts.

168

€&edit y.

These two cases shade into each other somewhat, but

i t mill be convenient to treat them separatcly. The first hits just hccn briefly cxamincd, 11. 156, and what follows rcldcs only to tlic second class of cases-the variation of honi 0 1 ogous p i r ts. The most fiimiliar illustration of this law is the fact t h z t in most bilxteral organisms homologous parts on both sidcs of tlic body tend to vary togctlier. The law holds in radially pymnictrical organisms also. All the petals of a regnlar flower generally vary in the same manner, but tlicre are many exceptions. Tlie front and hind linibs of vertebrates tend to ~ a r y iu tlic same mannci., as wc we in long and short legged or in thick and thin legged races of horses and dogs. It is stated tliat wlien the muscles of t h e arm depart i n number or arrangement from t h e proper type they almost almajs imitate those nf the leg, :tnd so converscly t h e varying mnsclcs of the leg imitate the normal muscles of the arm. There are many cases where a parent with extra fingers has produced a child with extra toes, or tlic reverse, and in otlier cases a parent with only one extra digit on one hand has liad children with supernumerary digits on both hands and both fcet. In certain pigeonsand for~ls,especiidly in the trumpeter pigeon, long feathers, like t h e primary wing feathcrs, grow on tlic outside of the leg and on the two outer tncs, and in pigeons with the feet thus feathcrcd the two outer toes arc partially connected by skin, thus slioming a marked :matomica1 resemblance to a wing. Tlie wrious appendages which are foymed from tlic skin, such as Iioofs, horns, hair, fcathers, teeth, etc., are homologous organs, and i t is interesting to notice liow frequently a pecnli:irity in one of these structures is associated with similar peculiarities in others.

T 7 ~ eE o ideizce f r o m Vuriation,.

150

Tropical sheep with long C O ~ I ' P hair ~ usually 11:~ve go;tt-lilcc horns. Iiilicritecl baldness irl nian is: often acconil)anied 1 y dcGcicnt teeth, iLritl tlic rctietval of the hair in old t i p by :t i~nev;nIof tlie tcctli. The famous liairg Burniehe 1i;tcl dcficient teeth, and Loth peculiarities mcrc liercdis,iry. A Spanisli dancer, Julia Pastrana, liad a f d l b ~ a r nand a double set of teeth, and the daily papers Iiavc rcce~tlycontained ail account of a man, Iivingnear Lebanon, Pennsjlvania, witli no hair, teeth, or sweat gla MI s. The homologons p u t s of plants often vary i n the same way, ns is well shown by certain comljound flowers, ill which tlic stamcns and pistils closely resemble petals. According to our view of the cause of viriation we can easily see how gemmules from a cell in one hand might lijlritlize, and thus c;tuse variation in tlie correspoiicling cclk of a11 four extremities, or perliaps iu the embryoiiic ccll from which all these cells arc derived, f m in tlie Snmc way that an animal can u i t e sexualiy either with mother of its own race or with one wliicli is somcwliat less closely related to it, so I assume that a gem111ule may unite with tlie particlc of the ovnm wliicli COP rcspoiids to it, or with sonie other closely related particle. For example, agemmule wliicli is thrown off from L: partku1;rr epit lielial cell may simply cause modification i n the corresponding cell of tlic offspring, or it may c m s c modification in a cell which is to produce this particular celI and a number of others. If each variation is purcly fortuitous the number of generations which mould be necessary in order to convert a species with bIacli hair into a spwies with every hair brown or with every hair red IS almost inconceivablc, bot this difficulty entirely disappears as soon as we recogtiizc tliat gemmules from oue part of the parent may

160

Heredity.

affect all the homologous parts of tile offspring in the same way and a t the same time.

Mules more Variable thaiz E’emnles. One of the most remarhble and suggestive of the laws of variation is that in all the higher anini;ils a part nhich is confined to males, or is more developed or of more functional importance in males than i t is in females, is Tery much more variable than a part vhich is confined to feniales or is more important in females 1h:m it is in males. The evidence for this remtirkahle law mill be presented at length in Chapters VIII. and IX. The cxlsteuce of such alaw is absolutely inexplicable willlout the theory of heredity, but it is exactly mliat lhis theory woulcl lead us t o expect, for an organ which is most iuiportant in one sex is most likely to be influenced in this sex by changed conditions, and is thercforo more likely to form gemmoles in the body of the sex wlierc it is most important than in the body of the opposite sex. An organ mlmh is most important in males will thercfore be most prolific of gemmulcs in males, while an orgnn mliicli is most important iii females will be most prolific of gemmules in females. Gemmules which are formed in the niale body are mstly more likely to be transmitted t o dcscciidauts than those which are formed in the female body. It fol1on.s that an organ which is most developed or most important in males must be vastly more liboly to transmit gemmules to descendants, and therefore t o \nrg in successive generations than an organ which is most developed or most important in females. Another law which follows from the one which has just been stated is that males are as a rule more rariable than females. This Inw has been noticed by Darwin and others, but no exidanation has ever becn aclranced. ,

The Eridence from Variation.

161

Summary of Last Two Chapters. The study of hybridx and of variation has led to the discovery of a great number of general laws, all of which are perfectly explicable by the theory of heredity, and are precisely what it would lead us to look for, although most of them are absolutely inexplicable without it, and have no place in any other hypvtliesis which has ever been proposed to account for the phenomena of heredity. The study of hybrids gives 11s a means of analyzing to a certain extent the influence of each sex in heredity, but our experiments in this direction are limited b j the fact that organisnis must be very closely related in order to breed together, and parents which are very closely related must be essentially alike in everything except the most recently acquired modifications. So far as tliey enable us to analyze the influences of the sexes, the results furnished by hybrids agree with the demands of our theory. This fnruislies an explanation of the great variability of hybrids, as compared with the p w e parents, and it also enables us to understand why hybrids from domestic races shoulc! be more variable than those from mild races. The remarkable fact that the descendants of hybrids arc more variable hnn the hybrids themselves receives a simple explanatio by our assumption that exposure of the various cells of the body to unnatural conditions is the prime cause of variability, and that it acts indirectly by causing the prod action of gemmule? Some of the recorded facts regarding hybrids are so very pecnliar that it mould be difficult to devise better tests tlian they furnish of the truth of our theory. What could be more curious or more opposed to the

d

162

Neredity.

view that tlie sexes play similar parts in heredity than the fact that the o3spring of a male hybrid and tlic female of a purc species is much more variable than the offspring of a female hybrid by a father of pure blood ? D:irwin’s pangcncsis hypothesis f iirnishes no explanation of this most remarkable fact, and none of the hgpotheses of heredity which haye been proposed from time to time are sufficiently dcEnite to have any bearing upon a concrete case like this, but our theory that changed conditions of life cause a production of gemmules, and that these are stored up in and transmitted by the male element, fits tliis case exactly. The curious plienomena of reciprocal crosses, a g i n , are just what our theory would lead us to expect, and it also farnishes 11s with an explanation of the fact that crossing so frequently causes reversion. A comparison of sexual with asexual reproduction also gives us a means of analyzing the inflnences of the two sexual elements, for asexual reproduction is essentially reproduction with the male element left out, and the result of this omission is, as we should expect, tlie reduction of the tendency to vary to a minimum. At the same time that OLU theory explains the great rarity of bud variations, it admits of their occasional appearance, and i t gives an explanation of the singular fact that bud variation is much less rare in plants which have long been cultivated than it is in mild forms. The most remarkable of the laws of variation is the well-known law that changed conditions do not directly produce variatibh, but cause subsequent generations to vary. As changed conditions do not in tlicnisel.F-escause hereditary modification, but simply lead to the production of gemmules, we see why their effect should be manifested in succeeding generations, and we also see

The Evidence from Variation.

163

why variation is itself hereditary, for the variation of any p:irticular cell will cause adjacent or related cells to tlirom off gemmules, and thus to produce variation in successive gcnerstions. We can also nnderstand why specific characters should be more variable than generic characters; why the species of large genera should vary more than the species of small genera; wliy a part developed in an unusual way or to an unnsual degree should show a marked tendency t o ~ a r y and , why secondary sexual characters should exhibit a similar tcndency. Uiiless our theory is true, what possible reason can tlicre be wliy a part whicli is excessively developed in males should vary more tlian a part whicli is similarly dereloped in females alone, or why the males of our Iiiglier domesticated animals should be more variable than the females? I t s power to deal with and interpret special cases of this kind scparates our theory from all other attempts to exldnin the phenomena, and seems to show that there can be but one choice between it and any otlier explanation wliicli has ever been proposed. If we accept Darwin’s view that variations are purely fortuitous, tlicre are certain grave difficulties whicli must prcvent 11s from giving the theory of natural selection unqualified acceptance as an adequate and complete explanation of the origin of species. Natural selection can rarely lead to permanent modification unless many individuals tend to vayy in nearly the same may at about the same time, and if variation is fortuitous the cliancc agxinst this is very great indeed. While tliere is no reasoii to doubt that natural selection might briiig about $1 the clianges which have led to the formation of a complicated organ, by the preservation of fortuitous variations, if time enough were given, there is

Heredity.

164 ~~~

reason to doubt wlietlier life has existed long enough to permit tlie evolution of existing forms in this way, and natural selection gives no account of the sudden appearante of considerable modifications, although the history of domestic animals shows us that sncli saltations do sometimes occur. On the one hand we find that Darwin’s assumption that variations are fortuitous involves us in grave dificnlties, but on the other hand we find scarcely any evidence to show that permanent hereditary race modifications are eves directly produced by the action of external conditions, while we do find evidence for the opinion t h a t race modifications are, as a rule, not due to this direct action, but to congenital variation. Our theory furnishes an explanation wliicli lies midway between Darwin’s view of the origin of variation and tlie Lamaskian view, and thus enables us to escape both these dificulties, for it shows us how the influence of changed conditions upon an organism may give rise to congenital variation in 1atei. generations, and it also shows‘us why variations tend to appear a t the time and place where they are needed. I t also shows how a considerable modification may appear suddenly and become hereditary. The correlated variation of homologous orgacs and the correlated modification of the various parts of a complicated organ are accepted by Darwin without explanation, but the theory of heredity shows us tliiit tlicse phenomena, the chance against the fortuitous occurrcnce of which is almost infinite, are due to the working of a very simple law. When we reriem the ground and sce how all the phenoiiiena of hybridization and variation fall into their proper places; how the same simple explanation fits the

The Evidence from Variation.

165

niost anomalous and exceptional phenomena as well as tlic more ordinary and simple cases, I tliiiik we must acknowledge that our theory is a t least an approximation to tllc t r d r . If it leads us to tho discovery of truth, and thue ultimately contributes to the establisliment of an explanation of tlie phenomena of heredity, its final acceptance in its present form is R matter of little moment. That it is R great advance beyond all the attempts which have been recorded s e e m obvious, and an examination of the ground whicli it covers ccrtainly seems to show that it is a step in the right direction.

CHAPTER VIII. T H E EVIDENCE FROIT SECONDARY SEXUAL CHARACTERS.

T h e N a t u r e of this Sort of Evidence.

~

I HAVE already given many reasons for believing that the male reproductive organ is especially adapted for gathering up the gemmules which are thrown off by the cells of the body; and for transmitting them to tlie next generation by impregnation, thus giving rise to variation; while the transmission of the gemmules which are formed in the body of the female is not thus provided for. If this supposition is correct, w e should expect to find that a variation wliicli first appears in a male should have more tendency to become hereditary than one which first appears in a female. Any slight change in either the male or the female body will, as we have already seen, cause all the cells which are either directly or indirectly influenced by the change to throw off gemmules. This will happen in a female body as well as in a male body, but the gemmules are, in tlie latter case, much more likely to be transmitted to descendants, and thus to give rise to more extended modification. We should also expect to find that an organ which is confined to males is much more likely than one which is confined to females to undergo hereditary changes, for even if the parts of the female body give rise to gemmules as frequently as the parts of the male body, tho chance of transmission is much less. We should also expect to 6nd that parts which are

The Evidence from SexuaZ Characters. 167 confined to males are more variable than parts confined to females; for variation in any part is due to inheritance of a gemniule from the corresponding part of one parent or the other, bat when the part is found in only one parent the gemmule must come from that parent. As transmission of gemmules by the mother is more rare than transmission by the father, i t is plain that parts which are confined to the male should be expected to vary more than parts found in the female alone. Finally we should expect thc male body as a whole to be more variable than the female body, for the same reason. In most cases it is impossible t o trace any particular variation back to its first appearance. This is almost out of the question with wild animals, and most doGesticated races have been formed so slowly that it is impossible to say whether the successive steps appeared in males or in females, nor can we be sure that a variation is new when it first attracts attention. Stilkit is interesting to note that the sudden variation which resnlted in the ancon breed of sheep was first noticed in a male, although it is, of course, impossible to say whether it was due to inheritance of gemmules from the father ratlier than from the mother. Certain hereditary diseases and montrosities, such as albinism or polydactylism, are fully as often traceable to a male origin as they are to a female origin, but as me know that peculiarities of this kind frequently skip a generation or two, we can xievcr be snre that we have traced them to their origin. In the secondary sexual cliaracters of animals we have a class of phenomena which are not rare and exceptional, for they are numbered by hundreds of thousands, and they can bc observed and studied by every one.

168

Heredity.

A secondary sexual character is a peculiarity which is not directly concerned in the reproductive process, although it is normally either confined to one sex, or else is much more developed in one sex than it is in the other. The presence of a beard is a secondary sexual character of man; the comb, wattles, spurs and brilliant plumage of the domestic cock, the horns of a stag, the tusks of an elephant, the mane of a lion, or the brilliant plumage of the peacock or of the drake, are all of them examples of this sort of organs, for they are either confined to one sex, or else tiley are macli more conspicuoukand important in one SCX than they are in thc other. They furnish, like hybrids, a means of disentangling or analyzing to sonic extent the influence of the two sexes in heredity, and I hope to show in this and the following chapters that they fizrnish evidence to prove1. That in most animals vith separate sexes the males of allied species differ more than the females from the ancektral type. 2. That organs which are confined to males or are of more importance or are more perfectly developed in them than in the females, are much more likely to give rise to hereditary modifications than parts which are confined t o or are most developed in females. 3. That a part which is confined to or is most developed in males is more likely than a similar female part to vary. 4. That males are, as a rule, more variable than females. 5. That the male leads and the female follows in the evolution of new races. There are two criteria which are of great use in the attempt to trace the path which a species has followed in its e-iolution. One of these is by comparison of a

The Evidence from Sexual Characters. 169 species with its nearest allies. The other is by comparison of the young with tfie adult. If most of the species of a genus resemble each other i n certain chamcters, while one species presents a marked deviation, we may in mosL cases safely conclude that the latter species has undergone recent modification in this respcct. Of course this rule does not hold good where the peculiarities of the exceptional species are featnrcs of resemblance to other genera of the family, for in this casc we must conclude that it has remained comparatively stationary, while all the otlier species of the genus have been modified. If in the second place we find that the adults of several related species differ greatly, while the young are much alike, we must attribute the difference in the adults to the fact that they have recently diverged from a common stock. Now I hope to shorn that throughout the animal kingdom, wherever the sexes differ from each other, the gencral law holds good that the males of allied species differ from each otlier more than the females do, and that the adult male differs more than tlie adult female from the young. There are many marked exceptions to this law, but the existence of the law has long been recognized by all naturalists. Every one who has worked at the systematic zoology of insects or vertebrates knows how difficult it often is to decide upon the specific identity of an immature or a female specimen, emn in cases where the mature males can be recognized and identified without difficulty. Darwin’s interesting essay on “ Sexual Selection” is well known. It is almost entirely devoted to the study of secondary sexual characters, and to a masterly discussion of tlie subject in all its aspects and relations.

170

.Beredity.

Darwin has gone over the whole field so thoroughly and exhaustively that little remains t o be said upon the subject, and the reader who is familiar with the essay will discover that almost all the facts in this chapter are borrowed from this source. Darwin’s aim, however, is simply to show the potency of sexual selection, while our present object ia to show the frequency of hereditary male modifico‘1t’ion as compared with female modifications, and I have therefore rearranged the facts, so as t o gire especial prominence t o this aspect of the subject. The critical reader will discover that in many cases I have borrowed the descriptive portion of one of Darwin’s paragraphs, but have said nothing about the theoretical portion. As Darwin’s conclusions are in many cases opposed to my own, this may convey to some the impression that I have made an unfair me of the weight of his anthority, and have quoted him in support of conclnsions which h e in reality opposes. I will refer such readers t o the chapter which follows this, where I have devoted a section to a statement of Darwin’s view of the origin of secondary sexual characters, and have given my reasons for beliering that it is only a partial explanation of the phenomena in question. Examples from Various Groups of the Animal hringdom to show that in ull Groups zultere the Sexes are Separate the Mule is, as a Rule, more Afodijed fhan the Female, and that the Adult Nales of Allied Species differ more, as a Rule, fhan B e Females or Eomg. R O T I F E R A. - h 1849, Dalrymple (Description of an Infusory Animnlcnle allied to the Genus Notommata, Pl~il.Trans. 1849) made the interesting and remarkable discovery that, in one species of the Itotifera, Notona-

The Evidence from Hexual Characters. 171 niatn A~zyZica,the animals are not hermaphrodites, as earlicr writers liad supposed, but that the mnles, which are rarely met wilh, arc very much smaller than tlie females. The latter sex is furnislied with a digestive tract which is quite complicnted in structure, and is armed at the mouth wjtli a lijglily specialized masticating apparatus. Tlie digestive organs of tlie male, 011 the other hand, are almost absent. Thc jaws, the msophagns and the mouth are wanting, and thc stomach and intestine are reduced to a functionless rudiment. The males reccive no'nonrishment after they lcavc the egg, and they live only a short time. Tile prcseiice of a digcstive tract is characteristic of all groups of animals above the protozoa, so we are compelled to believe that the ancestral form from which the Rotifera are descended had, likc the ordinary metazoa, a mouth, a stomach, and an intestine; and no one who is a t all familiar with comparative anatoniy can doubt that the male, in which it is absent, rather than the female, in which it is present, is the sex which has been modified. Tlie digestive tract is usually one of the first parts to be developed in tlie embryo, and its disappcarancc or absence in the adult male rotifer is therefore very different from the absencc of the wings in certain fcmale insects. Wings appear very late in life, and the failure of the female to acquirc them is simply at1 arrest short of perfect development, while the absence of digestivc organs shows active degeneration. I n 1855 Leyclig verified Dalrymplc's observation (Zeit. f. Wiss. 2001. vi. p. 96) in the samc s1xcies, and also in a second species of the same genus; and as Iic mas able to distingnisli tlie outline of tlic malc inside the egg, mliile this was still contained witliin the body of tlic fenialc, he remored all reason for doubting that the two sexes belong to one

172

Eeredity.

species. In these two species the feniales were much alike, while thc males mere not only very different from the females, but also froni each other.

FIQ. 1. Young female Rotifer. Hydutina senta. a. cloaca. b. contractile vesicle c. water tubes. d. stomach. e. ovary. f.ganglia. g. stomach. i . mouthparts.

FIQ.2. Male of the same species. a. orifice of penis. b. contractile vesicle. c. testes. d. ganglion. e. setigerous apparatus.

Sitice the year 1S55 the subject has been studied by many naturalists, and tlie ninles have been found in such a number of species that it is probable that the

The Evidence from Xexua.l C7~uructers.173 scxcs arc scptratc in a11 tlic Rotifern. In somc forms tlie niillcs itrc cven niorc siniplilicil than in Xotownzatn, wliilc i n otlicrs t h y arc lcss so, and in a few tlicy :ire like tlic fcni 11csi n size wid sirticturc, aiid liarc tlic digestive urgttiis perfectly dcvcloped. ANsmms.--Among tlic mnrinc p o l j c l ~ ~ t o uannelids s tlierc IS uf ten cwsidcr:tblc diffcrciicc between the sexes, and tlic points in wliicli tlie male diffcrs from tlie female are :dso points in ivliicli the males of various spccics differ from cacli other. ARTHROP0DA.-Among the Artlir0110d~,the Insects, Crnstacea, ete., the female is often very grcatly modified, and in somc case3 the females of allied spccics differ from each other inuch more than the males, and in oilier cases i t is linrdly iossible to say whether the males or tlie females of allied species differ most, but, taking the group as a whole, the Antliropods seem to follow the law which prevails in other groups of animals, and male modifications are more numerous than female modifications. I n the Branchiopod Crustaces the males are smaller than tlie females, and are mnch less abundant. The male differs from the female in the possession of a numbcr of sccond:wy sexual characters. Tlie second antennae of tlie malc are more richly supplied with sensory hairs than those of tlie female, and various appendages of tlie male may be so modified as to form clasping organs for holding the female. In Branchippus tlie second antcnnae of the male arc greatly modified for this 1mrl)osc. Fignre 3 shows the head of a female spccinicn of Brnizc?i@ps Grzibei, 'figure 4 tho liead of tlic m:dc of the same species, aiid fignres 5 and 6 tlie heads of thc males in two closely allied species. These figures show horn much the males of the various species

Heredity.

174

difler from each otlier iu this respect. Thc shape a d structure of tlie first antennae a i d of the abdomen may

FZQ.3. Eead of female specimen of Branchippus Crubei, greatly enlarged. a . Arst antennm. b. second antenna

FIQ.6. Head of male Branchippw stagnalia

Fro. 4. Head of the male of the Fame species.

FIG.6.

Head of male drtemia salina.

also show considerable modification in tho males of sarious species of Branchiopods.

The Evidence from Sexual Characters. 175 Among the Cladocera, of which the common waterpea of our fresh-water ponds and lakes is an example, the female is provided with a brood poiich, within which the eggs are carried and the young developed. I n the male these structures: are absent, and the second antenns are especially modified as organs for discovering and holding the female. They are richly supplied with sensory hairs, and they are often armed a t their tips with grappling hooks, which differ in the males of closely allied species.

FIQ.7. Antenna of male Cyclops aerrulatun.

FTQ.

8. Antenna of male Cyelopr canthocarpoiden.

The Ostracoda present sexual differences like those in the Cladocera, and in many of them it is certain that the male part deviates, more than the female part, from the typical form. I n the non-parasitic Copepods, of which the freshwater CycZops (Fig. 9) is an example, there is not very much difference between the sexes, although certain appendages, which are unmodified in the female and retain their typical form, sometimes differ greatly in the males of allied species, and may be specially mod-

176

Heredity.

ified for discovering or holding the female. The modification of the first antenna of the male for this purpose is quite general, and a comparison of this part in the males of various species of Cyclops (Figs. 7 and 8) with the same part in a female (Fig. 9), sliows how much the males of allied species differ in this respect. The second antenna, the maxillary feet, and the last pair of

FIQ.9. Female specimen of Cyclops canthoeawoides.

FIG 10. Female specimen of Notodelphys Allmani.

swimming feet, are sometimes modified in the same way i n the male. I n the male Saphirrina the wonderful display of brilliant colors is due t o the presence of peculiar color-producing organs, wliicli are absent in the female. Among the parasitic Copepods we find a departure from the ordinary typical structure, which is so remark-

The Evidence from Sexual C‘haructers. 17’7 able that no one, on first examining one of the more modified parasitic forms, such as tlic oiic sliown i i i Fig. 15, would detect any resenlblancc to tlie frce or nonparasitic members of tlie groul), or mould even suspect that the animals are crastaccans. The femalcs, which are known as “$sh-Zice,” are parasites upon fishes and other aquatic animals, while the males are parasites npon the bodies of the females, and are usually of minutc size as compared with the females. The adaptation to a parasitic life has not only produced the most profound changes in the general structiire, but it has also brought about an almost unparalleled difference between the sexes. It is true that, this is not due to the modification of males alone, for the females as well as tlie males exhibit the moPt extreme departures from the organization which is characteristic of typical or non-parasitic crustacea, and it is difficult to decide from structure alone whether the male or the female is most modified. The fact that the male has been adapted to a life as a parasite npon the body of the female, while the female has simply become adalitcd t o a parasitic life on other animals, seem to show that the male organism is somewhat more plastic than the female. Simple parasitism may be brought about by indefinite variability, but parasitism npon a parasite demands definite variation to meet the definite changes which hare taken place in the host. The liiglily specialized parasitic Copepods arc joined t o the non-parasitic forms by a long series of intermediate species, in which the parasitic habit is only slightly developed, and I give aL’femfigiires to illustrate some of the steps in this most interesting series. The female Notodelphys (Fig. lo), which lives in the body cavities

1'7s

Her edity.

of marine invertebrates, and has very limited powers of locomotion, liardly differs from the non-parasitic Cyclops (Fig. 9), except that two of the body segments are modified to form a chamber in which the eggs undergo their development. The male (Fig. 11) is somewhat

FIQ. 11. Male specimen of the same species.

Fro. 12.

Female specimen of Lernentoma corunta.

smaller than the female, but bears a close resembance to her, and to ordinary copepods. The female Lementornn (Fig. 12) is very different from the nialc (Fig. 13), and both dcpart very greatly from the typical copepod struoture, although a slight resemblance can be traced between the female and cyclops. The female is very much larger than the ordi-

The Evidence from SexuuZ C7mracters. 179 nary non-parasitic forms; the segmentation of the body is hardly visible, the power of locomotion is entirely lost, and the appendages are either rudimentary or are changed into hooks for clinging to the animal infested by the parasite. The male, like the female, has no power of locomotion, and is very much smaller than the female, the difference in size being much greater than the two figures would indicate. It is found nowhere except upon t,he body of the female, to which it clings by its rudimentary feet. The fernale of another form, Anchorella, is show1 in Fig. 15, and the male in Fig.

FIG.13. Male specimen of Lernent o m eorunta.

FIQ.14. Male specimen of dnchb rella uncinata.

14. I n tliis species the males are very small as comlmrcd with the females, to whose bodies they are firmly fastened by their rudimentary hooked limbs. We can liardly state with confidence that either sex is more modified than the other in these parasitic copepods, for both have undergone such great changes that they have lost all traces of their crustacean affinity, but in the very similar case of the barnacles, we have sufficient evidence that the males do depart further than the females from the ancestral type. The barnacles, or acorn-shells (Fig. IS), are crustacea

Heredity.

3 80

which arc prctty closcly related to tltc copcpods, n hich t h y rcecniblc somewhat, during tlic carly stilgcs of their devclolmcnt. Tlic j o n n g swim freely i n tlic water for a time, but fiii:tlly :itt:ich tliemeclvcs to foreign bodics, hc:d downrviirds, by tlicir aiitcnna, and arc scdcntiiry for tlic rest of their 1 3 . Ti1 the stitlkcd or p~di~iicuulated barnacles, the antcnns of the free 1arv:a bccortie replaced in the adult by a long peduncle, at the top of

Fro. 15. Female specimen of Anchorella unciiiata.

16. An hermaphrodite stalked barnacle. Pollicipes corn7wopia. c. csrina. t . tergum. 8 . scutum. r. rostrum. p peduncle.

FIG.

which there is an irregularly triangular box, the cnpitulum, made up of a number of calcareous plates. Inside this box tlic animal is pliiced, liead domnnmds, :ind altlioiigh it is grciitly niodifcd t o fit it for t h i s protccted scdcntary life, i t still prcsei:ts t1nniiStakiible cridcnccu of its crustnccaii affinity, such ns the mouth-pirts, tlie segmented body and limbs.

The Evidence from Sexual Characters. 181 Onc of the most rcmarIialsIc characteristics of the Bmiacles is that, with a few execptions, they arc herni:q)hroditc. Tlic Artliropoda include a vcry considcrable proportiou of a11 tlie animals mliicli arc known to US, and as all of them, except the Barnacles and a few closely rclated parasitic forms, have the sexes separated, the f x t that these few sedentary forms arc hermaphrodite is certainly very remarkable, and we must believe that they are tlie descendants of crustacea with separate sexes. The stalked barnacles resemble typical crustacea much more closely than do the sessile ones, and we must regard' the fornicr as more closely related than the latter to the ancestral form with separated sexes. It is, therefore, interesting to find that a few species of stalked barnacles are male and fcmale, and also that in a few others the ordinary hermaphrodite form is accompanied by a parasitic male, which has been called by its discoverer, Darwin, a complementary male. The study of tlie few species with separate sexes and of those with complemental males has brought to light some of the most remarkable phenomena of natural science, and the subject is well worthy of extkndcd notice. Figure 1 6 is an ordinary hcrrnaphrodite stalked barnaclc, Pollicipes. I t belongs to a genus in which no triie males or true females are ever found. Figure 17' is a species belonging to a closely related genus, Scalpellunz, and it will be seen a t once that it closely resembles Pollicipes, even in the arrangement of the plates o f d h e capitulum. I t is an hermaphroditelike Pollicipes, but with a difference, for it carries inside its shcll a small parasitic complemental male, which is shown in Fig. 18. This male is very much smaller than the hermaphrodite, and Fig. 18 is considerably magni-

fied, while Fig. 17 is of nearly the natural size; b a t with this exception the complementary male is essentially like the hermaphrodite, and i t has the structure of an ordinary stalked bai*nacle. There is a distinct pednncle, which carries a triangular capitulum, and although tile plates arc somewhat reduced in number they agree in form and position with the chief plates of such a species as PoZZicipes. The animal inside the capitulum is milch like an ordinary barnacle, the essential diffei~encebeing the total absence of female reproductive organs. I t is a male and nothing more.

FIG.17. An hermaphroditebarna-

Fro. 18. Complemental male of the

cle. Scalpellurn villosum.

same speciea

Figure 19 shows the female of another species, IbJa.

Cumnzingi, which does not differ essentially from the forms shown in Figs. 16 and 17, but the female of lbla Cunzmingi is a true female instead of an liermaphrodite, and tlierc are no traces of male reproductive organs, but inside her shell, and planted by a long rootlike process, there is a minute parasitic male, shown] in ~ i g 20, . magnified thirty-two times, wliiB t ~ i efigure of the female is magnified only five times. In Fig. 20, b is part of the -wall of the body of the female, and a is the long root by wliicli the parasitic male is planted.

The Evidence from Sexual Characters. 153 The male has a capitulum, but no calcareoiis plates, and its antenns, an., are not completcly merged in tlie peduncle. It also differs from tlie Eem:ile in the possession of an ocellus, or eye-spot It lias mouth-parts and limbs, and, except for the fact that all its parts arc somewhat rudimentary, it does not differ very greatly from other barnacles, except as regards its reproductive organs. I n other species of Scalpellum, however, as in 6cal-

Fig. 19. Female specimen of Ibla Cumingii.

F I ~20. . Parasitic male of the saw species.

pellum Regium, the male is still more rudimentary, and has no mouth or digestive organs. In two other genera, Alcippe (Fig. 21) and Cryptopliyalns, the femalee, which are true females, with no trace of male reproductive organs, differ very essentially from ordinary barnacles, and they have fastened to the outside of their bodies a number of very small males. I n the males of these two species, which are shown greatly niagnified a t Fig. 22, there are a few faint traces of muscular fibres, but the organs of digestion

Heredity.

1S4 -

are entirely gone, and tlie inside of the body is entirely filed witli a great testis, while t!ic posterior end is prolonged into a11 enormow penis; and the aninial hardly deserves to be called an animal at all, as it is scarcely more than an independent male reproductive organ attached to tlie body of tlic female. This is certainly on0 of the most remarkable cases of djfPereiiee bctmecn tlic sexes, and no one who compares Figs. 1s and 22 witli Figs. 16, 17 and 19, can doubt that

Fig.21. Female of AZcippe Zumpus.

Fig. 22. Male of the same species.

among these bariiacles the males differ from each other much more than the females. Amoiig the higher cmstacea we find great numbers of cases where tlie young male is like tlie adult female, or tlic young of both &exes, but at mntnrity acquires distinctive sexual characters. Any one who is familiar with the crnstacca will acknowledge tlie existcnce of tliis plienomcnon, and i t mill only be 1iecess:iry to give a few illustrations. The adult male Lucifer is distinguished from the adult female by tlie posses-

The Evidence from Sexual Characters. 185 sion of a very peculiar clasping organ, figure 23, upon the first swimming appendage of the abdomen. The corresponding appendage, fignre 24, of the adult female is like the other abdoniinal appendagcs of both sexes, and me must believe that die peculiar form, in the male, is due to recent n1odific:ition. I t is therefore interesting to note that, before the male reaches maturity, the limb in question is exactly like the other abdominal appcnddagcs of the adult male or adult female. The male Lucifer differs from the fernale in the sliape of the last segment of the abdomen, and the outline of the

BIG.23. First abdominal appendage of a mule Lucifw.

FIG.24. The corresponding appendage of the female.

exopedite of the tail-fin is peculiar. These differences are very slight, as mill be seen by comparing the terminal segments of the male, figure 25, with those of the female, figare 26, and it is hardly possible that they are of any direct service in reproduction. Tlie fact that, in the young of both sexes, these parts are like those of the adult female, and that their peculiarities in the adult male arc due to a final change which does not occur in the female, indicates that race-modification has gone n little further in the male Lucifer than it has in the female.

186

Heredity.

Darwin says that it seems to be a general rule among the crnstncea, that the remarkable differences of structnre Tvhich distingaish the male from the female, do not make tlieir appearance until the male is nearly mature. In proof of this he refers t o the fact that the male sandhopper, does not acquire his large claspers, which are very differently constructed from those of the female, until nearly full grown, while the claspers of the young male resemble those of the female. The history of the abdomen in crabs seeins to show

FIQ.25. Tip of abdomen of male Lucifer.

FIG.26.

Tip of abdomen of female.

clearly that this difference is due to the fact that the male has deviated further than the female from the ahcestral type. The long-tailed cmstacea, like the crayfish, have a long free movable abdomen, ending in an enlarged tail-fin, and composed of a number of segments, each of which carries ;Ipair of ;ippendages, In the female cray-fish the first of these appendages are like those behind, but in the male, the first ones are peculiarly modified and form copiilatory organs. We have ample evidence that the true crabs are the modi-

The Evidence f r o m XexuaZ Characters. 187 fied descendants of an ancestral form which had, like the cray-Esh, R long free tail, -which Tvas used in swyiniming. T h e fact tliet the young crab does have such an abdomen, is one of the proofs of the correctness of this view; but as tlie crab grows up, the abdomen becomes curled forwurds under the body; it ceases to be used as a swimming organ; its separate rings become fused together, and its appendages become rudimentary or disappear. This very instructive change goes further in the male than it does in the female, for in the latter, more of the riiigs remain distinct; a greater number of appendages persist in the adult, and these are much more like those of the young, or of the crayfish, than are those of the male. The great modification of the male as compared with the female is well shown, among the crustacea, by the fact that there may be in the same species two different m d e forms. This sexual dimorphisnz, as it is called, is well shown in a Brazilian amphipod, Ordesiia Darzoinii, in wliich species there are two male forms which differ -from each other in the structure of their large cla~vs. These claws are used for holding the female, but as both forms are now used for this purpose, either shape would certainly have sufficcd as well as the other, and this case therefore differs greatly from that of the social insects, where one form performs a certain duty in the community, while another form is adapted to fill a different place and perform a different duty. The two male forms in Orchestia seem to be due simply to the tendency of the male o i p n i s m to become modified more rapidly than the female, and not to any great advantage which has resulted from the divergent modification. In discussing this case Darwin says that the two male forms have originated by some having varied

188

Heredity.

in one manner, and some in another: both forms having derived certain specid but nearly equal advantages from their differently sliapcd organs. Dr. Hagcii lias called attcntion to tlie fact that in certain of our Anicricm species of cray-fislies, there are two slightly different male forms, and Fritz ?duller, who pointed out the existence of tlie two male forms of Orchestia, has also described a remarkable dimorphic species of Tanais, in which tlie male is represented by two distinct forms, n c v x gradnatiiig into each other. I n the one form the male is furnislicd with more nnmerous smelling threads, and in the other form with more powerfnl and more elongated claws to hold the female. Fritz bInllcr suggests that tliese differences betmeen the two male forms of the same species must have originated i n certain individunls having varied in tlie number of tlicir smelling threads, while other indi~-iclualsvaried in tlic shape and size of their claws, so that of tlic forber thosc mliicli mere best able to find tlie female, and of tlie latter those which mere best able to hold her when fonnd, have left the grcatest number of progeny t o inherit-their respcctim advantages. Wlicncvcr a number of species of a genus have any part more developed in tlie male tlian it is in the female, tliis part, as a rule, varies in the males of tlie different species, and is tIierefore of great systematic importance, since ib furnishes diagnostic cliaracters f o r distinguishing tlie species from e:di other. This rule is of gciieral application, in a11 groups of animals with separate sexes, and cvcry one who is a t all fumiliar with the systematic zoology of 0111' Iiiglier animals ILI~OIVS how difficult i t is to idciitify species without mature male specimens. Tlie crustacea furnish an abundant supply of illnstrations of tliis law, but we have space for only one.

The Evidence from SexuaZ Characters. 189 In t h e fiddler crabs, one of tlic clams of tlie malc is enomionsly dcvelopd, so t h a t i t compares with thc otlier about as a base-Tiol does w i t h its bow. In t l ~ female both clams are alike, and both small. Tlicrc arc a iinnibcr of species of fiddler crabs, forming together tlie genns Gclassimus, and the big claw of tlic male, in each species, lias certain points of differciicc from all tlie other species. Tlie fact that tlie fcatnres mliicli characterize males as distingnislied from females, arc alsn the featiircs mliicli distingnisli sl>ccies from encli other, certainly indicates t b a t the origin of specific differcncc is to be songht in some peculiarity of tlic niulc org:tnism. IKSECTS.-&ny insects haw stridiilxting organs, by ~vliicli,:is in tlic house-cricket, tlicy proclucc their sharp music. Iii many cases tliesc organs arc csclnsiwly coiifined to tlic m:iIcs; in otliers t h y are present h i t rutlimentary in the fenxrle, mliilc they are pcrfectly dewloped in both sexes of ccrt:iin others. I n :ill cases we find tliat tlie organs for this purpose differ greatly in closely rclated forms, and thus show that they arc of comparatively recent acquisition. I n the Cicadas the females are mute, and the sound is prodneed in the male, by tlic vibration of tlie lips of tlic spiracles, which arc set into motion by a ciirrcnt of air discliarged from the tracliez. It is increased by a wonderfully complex resonating app:ir>itus, consisting of two cavities covered by scales. Tliis appamtiis is present, very much less developed, in tlie female, but i t is never nscd for producing sound. The males of tlic crickets, grasshoppers, and Locnst i h e , are a11 i*emarkable for their niusical powers, wliicli are absent i n tlie females. Altliongh tliesc three g r o u p of insects are pretty closely related to each other,

190

Heredity.

and dthoiigh tlic general clinracter of tlic sonnd, and its mcc1i:micnl c:insc, arc cssentinlly nlilic in all of them, the posiilion and clinrnctcr of the sonnd-prodncing mechanism wiies greatly. In tlic nialc cricket the under surface' of each wingcover h:u a row of sharp transverse ridges or tecth, which is rapidly scraped across a projecting ridge on the outer snrface of the opposite wing, thus prodncing the music. First one wing is rnbbed over tlic other, and then the movement is reversed. Both wings are raised a little a t the same time, so as to increase tlie resonance. I n the Locustidze the oppoeitc wing-covers differ in structure, and their action cannot be reversed, as it is in the crickets. The left wing acts as the bow, and is scrapcd over the riglit, wliich serves as the fiddle. I n some forms tlie posterior part of the pro-thorax is elevated into a sort of resonating dome over the wing-covcrs. I n tlie grasshoppers tlie sound is produced in a very differcut mmner. There is usually a long row of nearly a hundred minute teeth on the inner surface of the femur, and t h i s IS scraped across the sharp projecting nermrcs on the wing-covers. I n one South African form the femur is rubbed, not against the wing-cover, .but against a notched ridge on the side of theabdomen, and the whole abdomen of tbe male is distended with air, like a great bladder, to increase the resonance. The female grasshopper has the stridnlating apparatus in a rudimentary condition, and it is interesting to note that the yonng male is like the adult female in this respect, for I m d o i s states that the teeth on the femora of the female remain throughout life in the condition in which they appear in both sexes dnring the larval state, but in the male they become fully developed and *

The Evidence from Sexual Characters. 19 L acquire their perfect structure at tlie last moult, when tlie insect is mature and rcady to breed. Many bcetles have rasp-like ridges with fine teeth on certain parts of their bodies, fur producing a stridolatiiig noise, by scraping against hard ridges or angles on the adjoining parts. Iu most stridulilting beetles they are equally developed in both sexes: in some they are rudimcntary or entirely absent in the female. These organs arc situatcd on widely different parts of the body in different beetles, even when they are very nearly related. I n the carrion beetles there are two parallel rasps with fine transverse ribs on the fifth abdominal segment, and they are rnbbed against the posterior edge of the wing-cover. I n otlier beetles the rasp is on the dorsal apex of the abdomen. I n others it is on the side of the first abdominal segment, and is scraped by ridges on the femur. I n others the rasps are on the lower surfaces of the wing-covers, and the edges of tlie abdominal segments serve as scrapers. In others the horny tip of t h e abdomen is scraped against a rasp on the wing-covers. 111a great number of species of long-horned beetles. tlie rasp is on the meso-thorax, and is rubbed against tlic pro-thorax. In still otlier beetles there is a ribbed rasp ruiiiiing obliquely across the coxa of each hind leg, and this is scraped across a specially projecting ridge on one of the abdomiiisl segments. In still others the rasp is on the pro-sternum, and tlie scraper on the meta-sternum. I n the cases where the stridulating organs are confined to the male, or wliere they are rudimentary and functionless in the female, we have every reason to believe that the successive variations which have led to their production have originated in males. I n the cases wliere each ;ex has inherited them in full perfcction, there is, of course, no direct cvidcnce to show

192

Beredity.

tliat they have originated in one sex rallier tliaii in the other. The organs are essentially alike in structure, whether they are confiued to the male or are present in both sexes; and as we liave good reason for believing, in the first case, tliat they have originated i n males, and no reason for doubting that they liave so originated in the second case, the conclusion that they all have liad a male origin certainly has a great probability in its favor. The great diversity of the males of allied species, as compared wit11 the females, is well slio~vnin those beetles where tlie males, and not tlie females, liave great horns rising from various parts of the body, as from the h a d , thorax, clypeus, or the under surface of t l ~ ebody. Darwin gives the following account of tlicse structures: '' These horns, in the great family of Lanielli.cor~is,rcsemble those of various qnadrupeds, siich as stags, rhinoceroses, etc., and are wonderful both from their size and diversified shapes. The females geac~allyerliibi t rudiments of tlic horns, iii tlic form of small knobs or ridges, but some arc destitute of even arudimcnt, wliiie i n a few others they are almost as well developed in tlie female as they are in the male. In abnost a2Z cases the horns are remarkable from tJJeir excessive cwiubiliiy; so that a graduated series can be formed, from the most highly developed males to otliers so degenerate that they can hardly be distingaished from tlie femalcs. The extraordinary size of the horns, and tlieir widely different structure in closely allied forms, indicate tliat t l ~ c yhave becn formed for some important pnrpose, bnt their cxcesaivc variability in the males of tlie same species leads t o tlie infercnce tliat this purpose cannot be of a definite nature. T h y do not sliow evidence of friction as tliey would if used for o r d i h r y work. They are not usually sharp, and do not seem well adapted for defence,

The Bnidexce f r o m EexuaZ Characters. 193 and they are not known to be used by the males in fighting %witheach other. The conclnsion mliicli best agrees with the fact of the horns having been so i n i m e n s e l y ~ c ~ t n o t fixedly dcvelopcd, as showii by tlicir txtrcme xiriability in tlic s;inic species, and 1)yllicir extreme diiersity i n closely allied spccies, is tbat t h y liave been acquired as orii a m e n t s.” One fact conncctccl with these horn-like projections gives as clear evidence as coula be desired, t h t tlie male is more liable to modification in this respect tlian tlie femde. I t sometimes happens tlixt t h e liorns are absent in the malcs of a species, h i t present i n a number of closely related spccics, and in such cases we must believe tliat t h e departure from tlie gener:il rule is due to tlic fact that tlic species in which they are absent lias becn recently modified. NOW,in such forms tlie fcmale shows her close relatioiisliip to the typical, nnnioctificd, or ancestral form by the possession of rtir1imciit:~ry horns. DtLrwin says tliat it is n higlily remnrltnble foints of resemblance betv,een tlie two forms of nieilnsae are, therefore, not due to Common inhcritancc, but liave been secondarily acquired. They are due to the f w t t h a t the two groups of mednsae have been evolved along parallel but distinct lines. Haccltel's familiarity with the medusae entitles him to speak with great authority; but still he may possibly be wrong, and the origin of the two groups may not have been as he supposes. There are four possible hypotheses as to the origin of the medusae, in favor of each of which something may be said. We may hold with Haeckel that the two larval p"1yps are the divergent descendants of a common anccstral polyp, which had no medusa stage, and that each has subsequently devcloped medusae, or me may bel i e w t l ~ tlie t common ancestor was a medusa without a p01.q~larva, and that the hydra larva and the scyphistoma larva have been independently acqnired, or we may believe that the ancestral form liad both a Iarv:il polyp-like stage, and an adult medusa stage, or finally we may assume what seems to 11s the most probable riew, that the ancestral form was neither a true swimming medusa nor a true sedentary polyp, but something half-way between, like the actinula of Tnbularia or the euibrgo of Cnnina. I do not see any fifth alternative, and one of these four suppositions must correspond with the actual evolution of tlie group. Now

806

Heredit?/.

whicherer one we accept, TVC are compelled to believe that there lias been parallel evolution, and that certain homologies between thc various forms are not due to inlie1 itance, but to independent modification. Haeckcl’s view ,compels us to believe that the yesemblaiices between the two groups of medusae hare been independently acquired. If we accept the second alternative, and derive the two forms from an ancestral medusae, the resemblances between the larvae must be due to parallel variation. Suppose, then, that we accept the third or the fourth view, and derive both groups from an ancestral form which had a polyp-like l a r d stage, and a medusa-like adult stage, or else from an ancestral form wliicli nnited in itself certain of the larval and certain of the adult characters. Among the veiled medusae there are some which, like Liriope, do not pass through a hydra stage, but lay eggs, which develop diyectly into m e d u s q and there aye also forms which, like the fresh water hjdra, hare no medusae stage. Among the veillcss forms there are also some which have no medusa stage, but which, like Luceriiaria and the Tesseridae, remain permanently as scypliistoma-_polyps, and i t is probable that in others, as the Charybdeadze, the eggs hatch into medusae, as thcy do in Liriope, without the intervention of alarval polyp stage. 1 t . k therefore impossible to frame any hypothesis as to the origin of the medusae, which will do away with the necessity for the belief that parallel modificatioii, along independent lines, has occurred in the different subdivisions of the group. If we accept Darwin’s view of the origin of variation, parallel modification is not absolutely impossible, although the chances against it are very great indeed.

H e r e d i t y and Natural fleeleetion.

307

The cases where i t can be proved to have occurred are not very numerous, it is trnc, but there are enough of them to present a serious difficulty. On our view that an external change which acts upon a certain part of tlie body may cause variation i n that psrticnlar part, the climices against the parallel modification of allied organisms are very grcatly diminislied, so much so that the occasional occurrence of such modifications might be expected. If such cases were the rule they mould be equally fatal to the theory of natural &election, wlicthcr our theory of heredity were accepted or not; bnt the cases are very far from frequent. Geiieral and Special Homology : and the Signijcancs of Serial Homology, Symmetry and Polymorphism. We have, eo far, been occupied in studying tlie evidcnce for tlie lam of heredity which is affordwl by the slight and recently acquired differences between the sexes of the same species, between the young and the adnlt, between domesticated and wild races, between the Iiyhrid offspring of allied species, be tween reciprocal hybrids, etc. Thc bearing of this law upon the more profonnd problems of morphology has hardly been referred to, for the field wliicli we have examined, although me h:Lve passed over it very rapidly, has furnished material for a treatise of considerable length. The discussion of the general problems of morphology mould require another voIunie of even greatcr length, for I believe, and hope to show in another place, that the acceptance of my view will lead to considerable change in our manner of liandling these problems ; and will so shift our view as to remodel some of the fundamenta1 principles of tlie science. I believe that it will throw light upon many obscure and perplexing qnestions, such as the significance of

308

Heredity.

symmetry and general homology, the origin of polxmorphism, the definition of an individual or person, etc.; but the end of a book is not the place to enter upon a a new field, and I am forced to reserve this subject for future discussion, although I mill now indicate very briefly the nature of this explanation. The basis of modern morphology i5 the doctrine that homology indicates genetic relntionsliip. Homology is fundamental similarity ofaplan, as distinguished from difference or similarity in physiological €unction. For example a man’s arm and hand are fitted for grasping, atid a bird’s wing for flight, and their different uscs rcnder them unlike ench otlicr in a superficial view, althoagh there is below :ind behind tliis obvious differcnce a more deep-seated resemblance. The feathers which cover the bird’s wing have the same histological structure and the same origin as the hairs upon the human arm; the skin which covers the limb has the same character in both cases; the wing, like the arm, has a siipporting skclcton, which consists of a shoulder and upper arm, a forearm, a wrist, and a hand; the muscles have the same structure and the same generd arrangement, and the way in wliich they are supplied with nerves and blood-vessels is the same. This fundamental itientityof strncture which is obscured, but not dcstroyed by the difference of use, is homology. I n a superficinlpiew the wing of a bird resembles the wing of a dragon-fly more closely than it resembles a man’s arm, but careful examination shows that the insect’s wing is not a limb n t all, but npeculiar outgrowth from thebody. The resemblance between a hird’s wing and an insect‘s wing is not an homology, and it has no morphological significance : it does not indicate that there is any close relationship between a bird and an insect.

Heredily and Natural Selection.

309

-

I t is sometimes dificnlt to determine whetliera resemblance is an homology or not, and in simple cases we decide by asking whether it can be due to similarity of use. We know that a bird’s wing is more closely rclated to a man’s arm than it is to an insect’s wing, bccause the resemblance between the two wings is 110 more than we should expect in organs adapted to the same purpose; but there is nothing in the use of the arm and wing to explain what they have in common. I n cases too complicated to be settled in this simple way we appeal to embryology, and ask mlietlicr the resemblance becomes more marked or lcss marlied when we study it in its younger stagcs. The arm and tlie wing arc more :dike in the embryo tlian they are in thc adults, and the features wliicli they share in comnion make tlicir appearance earlier than their distinctive characteristics. An homology then is a resemblance w11ic:h is not due to similarity of use, and which is more conspicuous in the embryo than in the adult. This is the doctrine of homology considered from its structural side; historically considered, an homology is a resemblance doe to community of descent, as distinguished from one due to recent modification. The modcrn morphologist believes that the resemblances between a bird’s wing and a man’s arm are due to inheritance from a remote ancestor in which the limb had all the characteristics which arc common to the wing and arm; that during the evolution of birds and mammals along two divergent lines from this ancestral form, the distinctive features which fit the wing for flight and the hand for grasping have been gradually acqnired. The doctrine that homology is an indication of ancestral relationship, and that the past history of organisms

310

Werediig.

can be traced by studying their anatomy and embryology, is tlie basis of the modern science of morphology. Now there are two kinds of ho:noIogy, special homology and generd liomology. Homologies between corresponding parts of different aiiiniuls are known ns special homologies, and those between different parts of the same animal as general homologies. As examples of general homology wc may instance tlic serial homology of a way-fish, the bilaterzll symmetry of mammals, and the radial symmetry of a star-fish. So far as structure goes tlie lion~ulogybetween a man’s arm and a man’s leg is precisely like the liomology between his arm and a bird’s wing. It is a resemblance wliicli is not due to similarity of use, but to fundamental resemblance, and it is more marked in the embryo than it is in the adult, and me seem, at first sight, to be jnstified in concluding that, if special homologies indicate genetic descent, general homologies must also; and that if general homologics cannot be explained iu this may, tlie explanation of special liomologies cannot be accepted. Mivart has pointed out that it is impossible to explain general homologies by attributing them to inlieritnnce from a common ancestor, and he therefore concludes that special liomologies do not prove genetic evolution. On tlie othcr hand many authors have licld that since special homologies indicate descent, general homologies must have the same meaning, and this belief has led to such speculations as the attempt to trace the vertebrates back to an annelid with a number of equivalent segments, to trace the echinoderms back to a commnnity of worms, and to trace tlie polymorphic siphonophores back to unspecialized commnnities of hydroids. I hope to show in another place that the acceptance

Heredity and Natural #election.

311

of my view of the natnre of heredity enables us to ayoid both of these results, since i t shows that special homologies may be due to lieredi ty of one sort, and general homologies to lieredity of another sort. Since corresponding cells in tlie homologous parts of the body of any individual are derived from closely related parts of the egg, t h y may be affected by similar gemmules and may thus give rise to what Darwin calls analogous variations. This form of inheritance I propose to call ontogenetic heredity, to distinguish i t from ordinary inheritance from an ancestor. I shall point out, in another place, that while special homologies are due to ordinary or phylogenetic heredity, that is, to descent from a common ancestor, general homologies are, in many cases, due to ontogenetic liercdity ; that special homologies are old, and that they jiidicate genetic relationship, and thus cnable us to trace the origin and history of animals, while general liomologits arc, in many capes, new, and recently acquired by secondary modification, and they do not indicate ancestry.

CHAPTER XII. RECAPITULhTION AND CONCLUSION.

THEobscurity and complexity of the phenomena af heredity afford no ground for tlie belief that the subject is outside tlie legitimate province of scientific inquiry. The existence, in a simple and unspecialized egg, of the potentiality of a highly organized and delicately adjnsted animal, with special functions, instincts and powers of adaptation, with tlie capacity for establishing and perpetuating harmonious relations to the clitlnging conditions of the world around it, is certainly one of the most profonnd problems of the material universe. The fertilized egg is one of the greatest wonders within oiir knowledge, but this is no reason for refraining from studying it. If we believe that living things lime become wliat thcy now are by a process of gradual evolution, and that tliey owe their cliaracteristics to the influences to which'they have been exposed in tlie past, we must believe that tlie properties of the egg arc capable of explanation, as far as these determining causes are open to study. If we accept tlie generalizations of modern science, and hold that an unicellular ovum is homologous with and is descended from a remote ancestral unicellular organism, and that its properties have been gradiially acquired by the natural selection of favorable variations, we must believe that the origin of its properties is 3s much within our reach as the origin of species. Tlic most proniiiient characteristic of heredity is that

Recapitulation a n d conclusion.

313

it may be brought about not only by tlie varions forms of asexual reproduction, but also by tlie sexual union of two reproductive elements, each of which is liomologous with the other cells of the body. In the lower aninials and plants the cells mliich thus unite with each other, or conjugate, are similar in form, and probably in function also; but i n a11 the higher organisms the male cell is very different from the ovum in form, size, and structure, as well as its mode of origin. The prcsent structure of each organism is the resnltant of two factors, wliicli we niay call adlierelice to type aiid adnl3titt ion to iiem conditions, or if the nse of t e r m withoiit teleological implicatioas is desired, we may speak of them :IS heredity and variibtion, or we may follow Haecliel and call tlieni memory of past experiences, and perception of new relations. The precise ternis to be used is a matter of little consequence. Tlie cssential thing is the recognition nf the fact that each organism is tlie resultant of two factors, and tliet evolution is two-sided. An animal is wliat it is because i t has the power to hold on to tile cxperiences and adaptations which fitted its parents for their place i n nature, and the parents acquired those peculiaritics in virtue of their pomcrs to gradually adjust their structure and habits to their environment. Tliis is the morphological side of erolution. Looking at it from its dyiiariiical or functional side, we notice that each step in tlic process of advancement has been remliccl by divergent specialization and by p11ysiologic:~l division of labor. Animals diverge from each other by acqniring the power to occupy different fields, to procure and use different kiuds of food, to exist in different media, etc., and the organs and tissues and cells of a highly specialized animal or plant are adapted to perform

314

Heredity..

definite, restricted functions exactly and efficiently, while each part of a low organism fills maiiy offices, but fills them all imperfectly. We find in all except the lowest organisms that heredity is brought about by two dissimilar reproductive elements, and we find that each organism is the resultant of two factors-heredity and variation. It is natuxal to inquire whether there may not be some connection between these two relations; wlietlier the natural selection of favorable variations has not acted upon the reproductive elements as it has upon the mature organisms; mlietlier it has not brought about a pliysiological division of labor between these elements; whether their originally similar functions have not gradiially become specialize$ until one has become tlie conserrati\ e medium, and tlie other the agent of progress in Iiercdity. According t o tlie view advocated in this book, such has actually been the history of the evolution of sex, and natural selection has evolved, in all the higher organisms, a secondary law of heredity, mliicli enables it to do its morlr rapidly and effectirely, with little waste. I n the metazon and in the higher plants, natural selcction is not a crude, rough ‘‘ hit or miss” method of evolution, for the law of heredity, itself a result of the law of natural selection, is that the ovum is tlie vehicle of heredity, while gemmnles or cell germs from cells in all parts of the body, are transmitted to the ovum by the male cell, thus causing variation when and where it is needed. This view is opposed to the conclusion of many high antliorities that there is no difference in the functions of the sexual elements, but exaniination shows that the reasons which they have given for this conclusion admit of another simple explanation.

Recapilulation and Conclusion. -

31.5

~

Darwin’s rcason for his Etatcmcnt that each sexual elcrnent has tlte power to ti-ansniit every singlc chwracteristic of the parent form, and that it is an error to suppose that the male transmits certain chnrnctrrs and tlic fcjrnde other chmictcrs, is that when lijbrids arc paired and bred inter so, the characters of either grandparent often reappear in the progeny. A little thought willsliow that it is impossible to prove any such conclnsion in this may. If two animals which differ from each other in cvcry rcspect could be made to cross, the m u l t mould fnrnish concltisire eyidcnce as to the correctness or incor~ectncssof Darwin’s statement, but in any possible cross the parents are essentially alike, and they diEer only in minor features of recent acqnisition. The possibility of parthenogenesis proves that the ovum does transmit the entire orgauizatim, bnt it isimpo3siblc to show, from the plienoniena of crossing, that the male elcrnent has the same power. Tlie reason given by HUXICY for his opinion that an animal inherits every characteristic of each parent, is that the ovum and the male cell are homologons with each other, and that all the cells of the body are descended, by a process of division, from the compound germ which is formea by their nnion. Homology, or similarity of origin, is no ground for assuming similarity of function, and the fact that the male cell and the egg are homologous with each other is no reason ~vhateverfor a belief that their parts in heredity are alike. The fact that either sex ay, under certain circumstances, acquire t hc secondary sexual characters of the otlier, seems at first sight to show tliat the mhole organization of the male exists in a potential and latent state in the body of every female, and that the whole organi-

...

316

Heredity. -

zation of the female is latcnt i n ewry male; that each individual is a complete double person. If we accept this conclusion i t is only logical to conclude that tlie power to revert or acqiiire tlie characteristics of remote ancestors provcs the existence, in a latent state, in encll individnal, of the coriiplete organization of each of a long series of ancestors of both sexes. This snttle metaphysical conception is so foreign to t h e methods and tendencies of modem thought, that when we conipnre i t with Hunter’s simple and definite statement that the natural history characteristics of any species of animal are to be found in those properties that are common to both sexes, there does not seem to be any room for choice. Tlie view that each individnal inlicrits all tlie characteristics of the species, and tliat tlie distinctive characteristics of tlie male are arrested in ccrtain ones, while tlie distinctive features of the female remain latent in others, furnislies a simple and adcqnate explanation of the facts, and remoyes all neccssity for tile subtle, complex and unthinkable, compouiid personality ligpothesis. In tliis connection the interesting and practical question, wlist determines the sex of the embryo, can hardly fail to suggest itself to the reader. I have rchaincd from a discussion of this important paint in the body of this work, as i t has no direct bearing upon our ni~gumcnt and I have no solntion to oeer. A s I liavc so far omi ttctl all reference to tlie subject, 1 will take occasion now to call attention. in tliis connection, t o the facts detailed on pp. 55-69. Tlie render will see tliat all female bces arc born from fertilized eggs, and all male bees from nnfertilized eggs; wliile tlie unfertilized eggs of t3aplinia gire rise to females only, and in many of the gall wasps both males and females are born from parthonogenetic eggs. There-

Recapitulation and Conclusion.

317

is no necessary or constant connection between the fertilization of the egg and the sex of the embryo, and the conclusion which I have reached from the study of these cases and of oiir scanty information upon the subject from other sources, is that sex is not determined by any constant law; that in certain animals and plttnts the sex of the embryo is determined by certain conditions, while in other groups it is determined by quite different conditions. However this may be, it is obvious that since perfect males and perfect females may arise from eggs which are fertilized, and also from eggs which are not fertilized, the necessity for fertilization does not come from the necessity for transmitting to the offspring the organization of each parent. A review of the opinions and reasoning of various authors shows that there is no good ground for believing that the two reproductive elements play simjlar parts in heredity and transmit every cliaracteristic of each parent. I t is impossible to prove i t by the phenomena of crossing, since the only animals which can be made to cross are essentially alike, and differ only in minor points. The homology between the ovum and the male cell is no reason. for supposing that their functions are similar. There is no reason for assuming that each sex transmits its entire organization to the offspring, since the latent transmission of secondary sexual characters can be wore simply explained by assuming that each embryo inherits, but does not necessarily develop, all the characteristics of its species. Reversion and alternation of generations admit of a similar explanation, and we may conclude that there is and can be no proof that each sexual element transmits all the characteristics of the parent. There is therefore

318

Heredity.

-

no a priori absurdity in the hypothesis that the ovnni and the male cell fill different offices. While there is 110 reason for believing that the functions of these elements are alike, there are ninny reasons for believing that this is not the case; for example, the almost universal occurrence of differences in f&m, size, and structure; the possibility of pdrthenogenesis; the differences between reciprocal hybrids; the fact that the offspring of a male hybrid and a female of a pure species is much more variable than the offspring of a female hybrid by the male of apure species; and the fact that a part Fhich is more developed or is of more functional importance in the male parent than it is in the female parent, is much more apt to vary in the offspring than n part which is more developed or more important in the mother than it is in the father. I n the absence of all evidence to the contrary I think we may safely conclude from this positive evidence that a division of physiological labor has arisen during the evolution of life, and that the functions of the reproductive elements have became specialized in divergent directions. The only way to discover the exact nature of this specialization is to study the influence of each element separately, and the comparison of sexual with asexual reproduction is the best available method of doing this, since asexual reproduction is essentially reproduction without a male element, while sexual reproduction is reproduction with a male element. Organisms produced from fertilized ova differ from those produced asexually only in their greater tendency to vary, and the hypothesis that the male element has become specialized for the transmission of n tendency to vary naturally suggests itself. Variation is not dependent upon fertilization, for plants produced from buds

Recapitulation and Conclusion.

319 -

vary as well as those born from fertilized seeds, although bud variatioiis are extremely rare as compared with seedling variations. I n any attempt to frame an hypothesis of heredity we must therefore recognize all the following facts : that the two reproductive elements are homologous, and that their functions were originally alike; that the possibility of parthenogenesis, together with many other well ascertained facts, shows that their functions are not alike, in the higher orgaiiisms, a t present; that their present functions are due to divergent specialization or physiological division of labor; that variation is possible without sexual union, but that the introduction of a male element in reproduction greatly increases the frequency of its occurrence. Among tho unicellular organisms variability is provided for by conjugation, or the fusion of two entire individuals so that the new generation is derived from 8 compound germ which contains particles to represent all the parts of the body of each parent. I n the metazoa and the many celled plants the reproductive bodies are localized and they are single cells, and there must therefore be some mechanism or organization in virtue of which they represent cells from all parts of the body, and thus provide for further variation. These various considerations have led us to believe that each cell of the organism inherits from its unicellular ancestow the power to throw off cell germs or gemmules; that these germs penetrate to all parts of the body, and that those which thus reach the developing reproductive elements insure variation, in the next generation, in the cells which they represent; that originally the two sexual elements were alike in function; that each inherited from the fertilized ovum o€ the p r e

320

Heredity.

ceding generation the power to give rise to a new organism with all tlie established hereditary cliaracteristics of the race; and that each element also had, by virtue of its contained gemmnles, the power to transmit variabili ty. The existence, in each element, of the power to transmit the hereditary clinracteristics of the species is obviously superfluous, since the object of sexual union, the transmission of a tendency to vary, wonld be equally well secured if only one element hsd the power to transmit the common characteristics of both parents. I therefore believe that, as organisms gradiially increased in size, as tlie number of cells in their bodics grew greater, and as the differentiationand specialization of these cells became more and more marked, one element, the male cell, became adapted for storing up gemmnles, and, at thesame time, gradually lost its unnecessary and useless power to transmit hereditary characteristics. This process was gradual, and eveu in the highest animals the power of the male cell to transmit heredicary characters does not seem to be completely lost, although few traces of it remain. I also suppose that natural selection has acted upon the various cells of the body to restrain them from throwing off unnecessary gemmnles, and that this power is exercised only when a change in the surrounding world renders variation necessary. After framing this hypothesis the next step is to tesC it by applying it to the various observed phenomena of heredity in order to see how far it explains and interprets them. I have attempted to do this in chapters VI. to X. of this book, and I think we are jtistified in conclnding, as the result of this review, that, while there are many facts which the hypothesis does not explain, they are not of such a character as to directly contradict it, while it

RecapituZation and Conclusion.

321

~~~

does group and illuminate many classes of facts which are quite inexplicable without it. The evidence from hybrids seems to be strongly in its favor, and it presents niany features wliich are perfectly simplc and natural, according to onr view of heredity, altliougli no oth'er explanation of them has ever been offered. Hybrids and mongrels are highly variable, as we should expect from the fact that many of the cells of their bodies must be placed under unnatural conditions, and must therefore have a tendency to throw off gemmules. Darwin's pangenesis hypothesis accounts for the variability of hybrids, but it does not account for the very remarkable fact that hybrids from forms which have loiig been cultivated or domesticated are more variiable than those from wild species or varieties, or for the fact that the children of hybrids are more variable than the hybrids themselves. Our view not only explains the variability of hybrids, but it also accounts for the excessive variability of hybrids from domesticated forms, and of the children of hybrids, for domesticated animals and plants live under unnatural conditions, and they are therefore more prolific of gemmnles than wild species, and as the body of a male hybrid is a new thing tlic cells will be much more likely than those of the pure parent t o throw off gem?ules. The fact that variation is due to the male influence, and that the action upon the male parent of unnatural or changed conditions results in the variability of the child, is well shown by crossing the hybrid with the pure species, for when the male hybrid is crossed with a pure female the children are much more variable than those bornfram a hybrid mother by a pure father.

322

Heredity.

The remarkable history of reciprocal crosses is, on the whole, exactly what we should expect, and although there are many difficulties, they are no greater than the complexity of the subject would lead us to anticipate. The study of variation brings out a number of secondary laws, all of which might have been derived from our view of the nature of heredity. The law that sexual offspring are more variable than those produced asexually has just been discussed, and i t is clearly in perfect accordance with our view. Another most interesting and remarkable law-that changed conditions do not act directly, but that they cause subsequent generations t o vary-receives a simple explanation as soon as me recognize that variation is due to the transmission of gemmules, not t o the direct modifying influence of external conditions. We can also understand why variation should itself be hereditary, why specific characters should be more variable than generic characters, why species of large genera should vary more than species of small genera, why a part developed to an unusual degree or in an unusual way should also be extremely variable, and why secondary sexual characters should show a marked tendency to vary. The study of secondary sexual characters aids us, like the study of hybrids and of variation, to analyze or disentangle the influences of the two sexes in heredity. These characters, therefore, possess especial interest in connection with our subject. They are found, upon examination, to present many striking peculiarities which might have been directly deduced from our view of the nature of heredity. As gemmules which w e formed in the male body am much more likely to be transmitted to descendants, and

Recapitulation and ConcZusion.

323

thus to give rise to variation, than gemmules which are formed in the female body, we should expect to find, in a variation which first appears in a male, much more tendency t o become hereditary than in a variation which first appears in a female. For thesame reason we should expect t o find an organ which is confined to males much more likely than one confined to females to give rise t o hereditary modifications. For a similar reason we should expect the males of unisexual animals to vary more than the females. We can form some conception of the amount of modification which a n animal has recently undergone by comparing the adults with the young, and by comparing allied species with each other. When we make comparisons of this kind we find that throughout the animal kingdom, with Fery few exceptions, wherever the sexes are separate and differ from each other, the males of allied species differ from each other more than the females do, and the adult male differs more than the adult female from the young. This law ia ao pronounced and conspicuous that its existence has long been recognized by all naturalists. We also find that organs which are confined to males, or which are of more importance or are more perfectly developed in the males than they are in the females, are very much more likely to give rise to hereditary modifications than parts which are confined to or are most developed in females; that a part which is tlius confined to males is much more likely to vary than a similar female part; that males are, as a rule, more variable than females; and that the male leads and the female follows in the evolution of new races. The scientific accuracy of most of these generalizations regarding secondary sexual characters has long been

384

Heredity.

recognized, although no one, so far as I know, has attempted to trace them back to a fundamental law of heredity. On the contrary, most of the authors who have discussed them have treated them rather as special cases than as the results of a general principle, and analysis shows that none of the explantitions which have been. advanced are sufficiently broad to cover the whole ground. Daines Barrington and Wallace have held that the explanation of the fact that male birdsand male insects are often so much more brilliantly colored and conspicuously ornamented than the females is to be found in the fact that the female, while laying her eggs or while incubating, is much more exposed to the attaclisof enemies than the male, and that inasmuch as the perpetuation of the race depends upon the safety of the females at this time, natural selection has gradually exterminated the conspicuous females, and has preserved those with the least striking colors. We know, however, that the male is usually mora brjlliantly colored than the female among reptiles which do not incubate, and evcn among certain fishes where the male attends to the eggs and young. It is therefore clear that Wallace’s explanation stops short of the whole trnth, and Darwin’s exhaustive review of the subject seems to prove tliat amon$ birds i t is the male and not the female which has been directly modified. Darwin believes that the greater modification of the males as compared with the females is due to sexual selection. The males have struggled with each other for the possession of the females, or have been selected by the females, and this process long continned is believed by him to have resulted in the perpetuation of the strongest, best armed, or most attractivc males.

RecupituZution and ConcZuusion.

325

It is plain that sexual selection must have the effect which Darwin attributes to it, but the fact that even in choice breeds of domesticated aninials which are mated according to the wislies of tlie breeder, and not according to their own selection, the males are more modified than the females, shows that behind the action of sexual selection some more profonnd 1.mmust exist. Darwin believes that tliia explanation is to be found in the fact that the male usually has stronger passions than the female, aud is consequently more exposed to the action of natural selection. IZe says that the perpetnation of the race depends upon the existence of the sexual passion, and that, since the male mnst in most cases seek the female, the most eager males have left the greatest number of offspring, and have thus become selected. Wlicn me bear in mind the fact that the parental instinct is fully as important to the race as the sexual instinct, and that this is usually most developed in the female, we see that the failure of tlie female to undergo modification for the good of the species as frequently as the maIe cannot be explained without the recognition of some more general law. The singular history of secondary scxual characters receives a ready explanation by the lam of heredity, for this law leads 11s to look to the cells of the male body for the origin of most of the variations through which the species has attained to its 'present organization. Since gemmules which originate in a male body are more likely to be transmitted than those formed in a female body, and since gemmules are most likely to be formed in the sex in whichan organis of most functional importance, and therefore most subject to disturbing influences, we can readily see why a part which is im-

326

Heredity.

portant to males should vary more than a part which is important to females. We are thus able to understand tlie great difference in the males of allied species, the difference between the adult male and tlic female or young, and the great diversity and variability of secondary male characters, and we should expect to find, what actually is the case, that among the higher animals, when the sexes have long been separated, the males are more variable than the females. I n the chapter on the intellectual differences betmeen men and women, I have shown that those philosophical writers who have devoted especial attention to the subject have reached conclusions which are exactly what our 4iypothesis would lead us to expect. The viev that the male mind is the progressive element in intellectual development, and the female mind the conservative element,. accords with the views which have been generally recognized and accepted by the common consent of mankind, and although our opinions upon this very complicated subject may possibly be very far from accurate, a certain conformation to the demands of our hypothesis cannot be denied. The facts relating to hybrids, to variation, to the secondary sexual characters of animals, and to the intellectual differences between men and women, which are stated at some length in chapters V. to IX., cover a very wide and diversified field, and any law mliich groups and explains them all is certainly worthy of careful examination. The most candid review which I am able to give to the evidence from all these sources, convinces me that tlie explanation which I have offered in this book is at least a step in the right direction, and that whether it be accepted in its present form or not, it does serve to en-

RecapituZation and ConcZusion.

827

large our insight into the hidden relations between the phenomena of nature. Chapter XI. is devoted to ail examination of the law of natnral sclcction, as modified by the law of heredity, s l i d I have liere atfeniptcd to sliow that the acceptaiice of this secondary law will remove the most serious objections to the view that our present f o r m of life have been brought into existence through the survival of the fittest variations, qnd I have also called attention to the fact that tlic law of heredity is itself a result of the b w of na t u r d select ion. No oiie can deny that tliere are grave objections to the law of natural selection in its origiiial form. Darlyin admits this in many places, and ttble but dissenting critics h a w stated most of tliese objections with great ability. The evidence for tlie 1:im of natural selection ia so many sided, so extensive, niid so satisfactory, that we may fairly conclude that the dificultics will disappear with greater knomlcdgc, and as none of its hostile critica have proposcd anything mliatever to take its place, the difficulties which they have pointed ont have liardly ieceivcd from naturalists the attention which they deserve. One of tlie most serious objections is that natnral selection cmnot effect any permanent modification of a race, unless g e n t numbers of individuals vary in essentially the same may at nearly the same time, and that tlic chances against this arc great beyond computation if variations are purely fortuitous in Darmin’s sense of the word. Darwin lins acknowledged the weight o € this objection, and tliere is no escape from the conclusion that natural sclection fails to account for thc origin of species, unless we can show that many ii1dividn:ils tend to vary at tlie same time. According to our view, the production of

328

Heredity.

gcnimulcs and the consequent variations are due to the dircct action of changed conditions upon certain cells of the body, and any cliangcvvliicli affects all tlie individuals of a specics will cause tlic same part to vary i n all of tlicm a t the same time. This objcction to tlie law of natural selection is thus entirely removed. The evolution of a complicated organism, or tlie modification of any part which includes a number of complicated structures, without destroying tlicir liarmonioue adjustment to each other, dcmands a w r y greet number of variations, and if t h e are fortuitous, we may well doubt mlietlier tliere has bccn time enough for tlic cvolution of life by nataral selection. Acearcling to our theory of hercdity, a change in one part of the body is i n itself a cause of variation in rclatcd parts; and as cliariges thus tcnd to occur wlierc and when they are needed, the time wliicli is dcmandcd for tlie cvolntion of a complicated organ by iiatural selcction is brought mitliin reasonable limits, :md one of tlie most fundamental objections is thus completely remorcd. There are many reasons for believing that vari at'ions under nature may not be so minute as Darwin snpposes, but that evolution may take place by jumps or saltations. According to our view a cliangc in one part will disturb tlie harmony of related parts, and will cause their cells to throw off gemmules. A slight change in one generation may thus become in following generations a very considerable modification, and tliere is no reason why natural selection should not be occasionally presented with great and important saltations. The lam of heredity also enables us t o understand the occasional occurrence of parallel or analogous variation, and tlie parallcl evolution of polyphylletic-groups.

INDEX. Adjustment between internal and external relations, 40 Adler on parthenogenesis in' gall-wasps, 6!! Adult organism, 6 Albrecht on parthenogenesis,

56, 58 Alcippe, 183 Alternation of generations, 115 American naturalist on spikehorn deer, 298 Anchorella, 179 Ancon sheep, 299 Adimdculist, 22 Annelids, 173 Anolis, 198 Antirlhinum, 139 Apple, 90 Apus, 58 Arbacia, 66 Argus pheasant, 201 Aristotle on parthenogenesis, 55 Aristotle on latent sexual characters, 85, 105 Artemia, 58, 91 Artliropoda, sexual diff ersnces in, 173 Asexual reproduction, 11, 17, 143, 249 Ass and horse, hybrids from, 197

Babyrusa, 205 Bachman on variation of turkey, 150 Balfour on polar globules, 71 Barnacle, sexual differences in, 179, 181 Barrington on colors of birds, 207 Basset on parthenogenesis, 62 Beclistein on spurred hens, 210 Bee, reproduction of, 9 " partlienogenesis in, 60 " variation of, 145 Beetle, sexual differences in, 191 Bell-bird, sexual differences in, 202 Birds, female modification i4

203 Birds, sexual differences in, 199 " weapons, etc., originally acquired by male, 199 Bischoff on parthenogenesis,Blumenbacli on Polishfow1,226 Bom byx, 58 Bonnet on evolution, 20, 85 Branchippus, 173 Buceros, 201 Budding in hybrids, 11 Bud-variation, 84, 144 Buffon on development] 26,85

330

Index.

Buffon on effect of castration, 106 Butterffies, sexual differences in, 196 Callionymus, 197 Cattle, hybrids from, 180 Cause of sex, 316 Ceratophara, 198 Chamelion, 399 Changed conditions cause snbsequent generations to vary, 149 Cherry, variation of, 149 Cladocera, sexual differences in, 175 Claus on parthenogenesis, 68 Cohn on parthenogenesis, 66 Color changed by a change of food, 90 Congenital characters not always hereditary, 93 Conn on cause of variation, 294 Copepoda, sexual differences in, 175 Correllated Variation, 84, 157 Crabs, sexual differences in, 1S6 Crayfish, dimorpliism in, 188 Crossingas a cause of reversion, 132 Crossing as a cause of variation, 119 Cryptophyalus, 183 Cultivated plants, variation of, 146 Cyclops, sexual differences in,

175 Dall on saltatory evolution, 88 Daphnia, parthenogenesis in, 67

Darwin, 85 'L on bud variation of of orange, 147 'I on causes of variation,

275 '6

.' '1

.I

.I

,' I 6

dd

., .#

I'

I,

81

dI ,d

Id

Id

'd

on compound personality, 114 on eomplexity of the germ, 114 on correllated variation, 84 on direct action of cxternal conditions,QO on evolution of eye, 282 japanned peacock, 299 on latent transmission of sexual characters,

105 on modification of male butterflies, 196 on pangenesis, 48 . on parallel or analogous variation, 302 on reciprocal hybrids, 129 on reversion, 10, 115, 132 on reversion in horse, 133 on selection, 12 on sexual charactera of birds, 199 on sexual selection, 169, 212 on transmission by each sex, 100 on transmission without fusion, 131

Indez

332

Darwin, on variation, 141,146

Fish-lice, sexual differences in,

149,153 on variation fron crossing, 120, 122

177 Fowls, hybrids from, 129, 131 " modification of female, 222 " reversion in, 135

"

on variation of specie of large genera, 15 Development often indirect, 2 Dianthus, variability of hybrid "

124

Gall-wasp, parthenogenesis in,

62

Direct influence of external con ditions, 89 Dog, reversion in, 10 " hybrids.from, 131 " variation of, 91, 150 Domesticated animals more va riable than wild ones, 142 Edrnnndstoii on sea-gull, 93 Education and culture, 272 Elephant, 205 Epigenesis, 20, 27 " and evolution, 24 Evadne, 58 Evolution and division of labor,

Gallus banlriva, reversion to,

135 Galton on pangenesis, 53 " on saltatory evolution,

296 Giirtner on variability of bybrids, 120, 124 Gegenbauer on nature of ovum, 28 Gelassimus, 189 Gemmules, 82 Gerstaecker on parthenogene. sis, 56 Godine on hybrid sheep, 129 Goose, inflexibility of, 142

313 " "

"

" "

"

"

Bonnet on, 20 of coniplicated or. gans, 156 definitions of, 20 Huxley on, 20 hypothesis logically imperfect, 82 morphological aspect of, 313 saltatory, 157,296

Falconer on variation of dog and goat, 91 Female modification, 235 Fishes, sexual differences in,

196

Haeclrel on perigenesis, 33, 45 '' on phylogeny of Me. dusze, 304 '' on significance of ontiw y , 30 Hagen on dimorphism of cray. fish, 188 Haller on evolution, 20 Huxley on evolution, 20 Harvey, 22 Hemp, variation of, 90 Hen, parthenogenesis in, 67 Xeredity, theory of, 16, 80 " ontogenetic aiid phylogenetic, 311 " and memory, 37

332

InBex.

Heredity, speculations on, 18 " meaning of word, 6 '' proper subject for study, 12 '' how caused, 81 Hinney, 127 Hippocrates on viragines, 105 Holmgrim on pigeon, 93 Homology of ovum and male cell, 102 '' significance of, 307 Hornbill. 201 IIorse and ass, hybrid from, 127 " in mines, 91 " reversion in, 10, 133 Humming bird, 201 Hunter on Intent transmission,

109 on sea-gull, 92 Huxley ou evolution and epigenesis, 46 " on reciprocal crosses, 127 '' on saltatory evolution, 83, 296 on transmission by each sex, 100, 102, 103 Hybrids, argument from, 99 " * evidence from, 118 " summary of chapter on, 137 " variation of, 18 Hydroids, budding in, 11 '' heredity in, 113 "

.I

Intellectual differences between men and women, 242 Jgger on heredity, 41, 45, 85 Jurine on parthenogenesis, 57 Kipp on parthenogenesis, 59 Kirtland on variation of cherry,

IS9 Kiilreuter on hybrids of mirabiIis, 120 Latent transmission of sexual characters, 104, 117 Leckey on male and female mind, 269 Leptodora, parthenogenesis in,

57 Leydig on rotifera, 171 Life, definition of, 39 Life and memory, 38 Lion, hybrid from with tiger,

130 'Lizards, sexual differences in, 198 Lyell on variation of dog, 150 Leeuwenhoek, discovery of spermatozoa, 21 Lubbock on parthenogenesis,

57 Lucifer, sexual differences in, 184 Male mind progressive, 257 Male more eager than female,

230 Ibla, 182 Inheritance of a tendency, 88 Insects, sexual differences in,

189

Male and female types of character, 259 Male cell, properties of, 81 " functions of, 84

Ingex, Male cell, discovery of, 21 Itlale more variable than female, 160 Male fish originally varied, 198 Male more modified than female, 170 Manx cat, hybrid from, 128 Many individuals must vary together, 155, 277 Mammals, sexual differences in, 204 &hn,sexual differences in, 204 BIaiicliamp merino sheep, 299 3IcCrady on polar cells, 70 Yednss, parallel variation in, 303 Mice, hybrid, 131 . Mill on subjcction of women, 2G4 Milabilis, variation in, 120, 121 Mivart on Iiomology, 310 " on requisites of a theory of heredity, 86 " on saltatory evolution, 83, 297 'I on sudden variation of plants, 301 " on time needed forevo. lotion, 284 Molliensia, 197 ItIoquin-Tandon on variation of plants, 94 Moths, pnrtlienogenesis in, 69 Mule, 127, 134 Milller on dimorphism i n crustacea, 188 Natural selection, 12, 14, 275 Narwhal, 205

333

Nervous system, development of, 23 Neuroterus, 63 Niata cattle, hybrids from, 130 North British Review, argument against natural selection, 277 Notedelphys, 177 Oelacher on parthenogenesis, 67 Onitis, 193 Orange, bud-variation in, 147 Orcliestia, 187 Origin of sex, 314, 316 " of complicated organs, 281 Ornitliorinchus, 205 Ostracodu, 175 Ovist, 22 Ovum, complexity 'of, 87 " development of, 23 " function of, 84 " homologous with other cells, 23 " properties of, 8, 81 " structure of, 22 Ovum and male cell, difference in fnnction of, 53, 74 Ovum and male cell, honiologous, 17 Pander, 22 Pangenesis hypothesis, Dar. win's, 50, 121, 124 Pangenesis hypothesis, Qalton's objection to, 53 Pangenesis hypothesis, objections to, 53 Pangenesis hypothesis, remodelled, 80

Index.

334

Paradise birds, 201 Par:illel or analogous variation, 302 Parthenogenesis, 55 Peacock, variation in, 299 Perigenesis, 33 Pheasant, 203 Pigeon, direct modification of, 94 " modification of male, 218 Polar cells, 70 Polisli fowl, 225 Pollicipes, 181 Psyche, 61 Pterotnalus, 65 Reciprocal hybrids, 125 Reproduction, 19 " by immature animals, 43 " of bees, 9 " of marine animals, 9 " sexual and asexURI, 11,17,249 Requisites o€ theory of lieredity, 16 Reversion, 17 " by crossing, 132 '' Darwin on, 10, 115 " in dogs, 10 '' i n horses, 10 " and .Jilger's hypotheEis, 44 " two kinds, 133 I' in offspring of hybrids, 136 " of lost instincts, 135 Rhodites, 65

-

Rotifera, 66, 171 Ruminants, 206 Saltatoryevolution. 83,157,296 Salter on variation of strawberry, 91 Sapliirrina, 176 Sassafras in Europe, 90 Scalpellum, 181 Sclisfferon parthenogenesis, 56 Sea-gull, 92 Sea urchin, 66 Secondaiy sexual characters,

iGG Semper on Lyninaeus, 92 Sex of parent, influence of, 18, 123 Sexnnl differences in Agrion, 239 Alcippe, 183 ancliorella, 179 argonauta, 237 artliropoda, 173 barn:iclw, 179, 181 beetles, 191 braiicliippus, 173 call duck, 236 cave beetles, 237 cicada, 188 clndocera, 175 copepoda. 175, 177 crabs, 186 crayfish, 188 cryptophyalus, 183 cyclops, 175 gelassinius, 189 ibln, 182 insects, 188, 236 lernentoma, 176 locusticlae, 188

Index. Sexual differences in lucifer,

184 notodelphys, 177 p i t i s , 193 orcliestia, 1S7 ostracoda, 175 papilio turnus, 237 paradise birds, 237 pliasmidae, 236 pollicipes, 181 rotifera, 171 snphirrina, 176 scalpellum, 181 shrike, 237 social insects, 237 tanais, 188 Sexual dimorphism, 187 " reproduction, 11,17,143,

289 selection, 169,212 Sheep, hybrid, 129,131 " modification of male, "

222 '' variation in, 297 Siebold on parthenogenesis, 55 Simpson on hermaphroditism,

109 "

on latent transmission, 106

Sitana, 198 Smerinthus, 59 Solenobia, 61 Sow, parthenogenesis in, 67 Spurred hen, 210 Spathogaster, Species of large genera, 153 Speculations on heredity, 18 Spermatozoa, discovery of, 21 Spencer, definition of life, 38 '' on Irish elk, 287

826

Spike-horn deer, 298 Staphylinid3c, 194 Strongylocentrotus, 66 Sweet pea, hybrid, 132 Tanais, 188 Theory of heredity, 319,16 Tiger, liybrid,.l30 Transmission without fusion, 131 Turkey, variation of, 150 Uhler on polymorphism, 238 Variability, how caused, 82 " of.offspring of hybrids, 122 " of sexual characters, 154 Variations, 13, 17 cau'jes of, 140,275,293 caused by climate, 142. " change of food,142 " crossing, 119 " excessof food, 143 correllated, 157 of exceptional parts, 153 of homologous parts, 158 of hybrids, 18 law of equable, 154 of male butterflies, 195 of organisms produced sexually, 143,249 parallel or analogous, 302 and sexual reproduction, 249 of species of large genera, 153 summary of chapter on, 161 Viragines, 105 Von Baer, 22

336

Index.

birds, 208 on polymorphism, 238 Walrus, 205 Walsh, law of equable variation, 154 Wart hog, 205 Waterton on hen with male characters, 106 Weismann on parthenogenesis, "

1

brid willow, 124 Wilson on parthenogenesis, 66 Wollff on heredity, 22 Xiphophorus, 198 Parrell on removal of oviduct,

''

105 on variation of dog, 150