r.O- ' r ■ <^ '^r- S. v^ •-^^ ^^- V^ > .# ^^" ,00 " I." J -7*, ^ \0 <^^ ' , 4 -^ > . \ <- ''/ COUNTRY LIFE EDUCATION SERIES Edited by Charles William Burkett Kansas State Agricultural College Types and Breeds of Farm Animals By Charles S. Plumb, Ohio State University Principles of Breeding By Eugene Davenport, University of Illinois Other vohtines in preparation PRINCIPLES OF BREEDING A TREATISE ON THREMMATOLOGY OR THE PRINCIPLES AND PRACTICES INVOLVED IN THE ECONOMIC IMPROVEMENT OF DOMESTI- CATED ANIMALS AND PLANTS BY E. DAVENPORT, M.Agr., LL.D. Professor of Thremmatology in the University of Illinois Dean of the College of Agricultiire Director of the Agricultural Experiment Station WITH APPENDIX BY H. L. RIETZ, Ph.D. Assistant Professor of Mathematics in the University of Illinois GINN & COMPANY BOSTON . NEW YORK • CHICAGO • LONDON [UBHARY of CONGRESS j Two CoDles Received OCT 4 »907 , Copyncttt Entry io^s A XXc, n/ eopY c.-'^ ^ ^ o1 Entered at Stationers' Hall Copyright, 1QO7 ^ By EUGENE DAVENPORT all rights reserved y^ 1-V^^^^ aCfyt atfjenxum j^retfcf GINN & COMPANY • PRO- PRIETORS ■ BOSTON • U.S.A. PREFACE Two classes of people have been in mind in the preparation of this text, viz. the student of agriculture in the college and experi- ment station and the practical breeder upon the farm . Both need to know all that evolution has to teach of methods that may be employed in still further adapting to our needs such animals and plants as hav^e been domesticated because of their valuable natural qualities. The general purpose has been first of all to define the problems involved in animal and plant improvement ; to free the subject from the prejudice and tradition that have always befogged it ; to bring to the study whatever facts are fully l^nown to biological science ; to recognize and define somewhat clearly the present limitations of knowledge, and to indicate as Veil as may be the directions from which further and much-needed light is most likely to come. Last of all and more than all, it has been the »«5^ ^" purpose to encourage, and if possible induce, more exact methods of study and of practice than have hitherto characterized this branch of agricultural science. It is yet too early to prepare an ideal treatise upon this most intricate subject, and no one is more conscious than the author of the many deficiencies and shortcomings of this attempt. Some effort, however, is surely needed at this time to clear the atmos- phere, to give the student of agriculture at least a rational point of view, and to bring him into comradeship with those who are earnestly studying biological problems and through whose efforts these vexed questions are sure sooner or later to find a solution. This, together with the pressing need of a text in his own class room, is the author's only warrant for the present volume. No new theories of evolution are proposed. The chief object has been to distinguish what is known from what is merely tradi- tional ; to give as much as possible, within the limits of available space, of the best established facts bearing upon this subject ; to vi PREFACE call attention to approved methods of study, and to indicate lines of research most likely to furnish valuable information in the not distant future. It is necessary to introduce a considerable amount of mathe- matical work in the later chapters. No excuse is offered for this introduction, and it is earnestly desired that the reader give special attention to this portion of the text, whether easy or diffi- cult of following, because it is by this road that many new princi- ples will arrive and that many of our future operations must be ordered ; for nothing is clearer than that the successful breeder of the future ivill be a bookkeeper and a statistician. For the convenience of the non-mathematical reader general formulae are placed in footnotes, and some of the more abstract matter is placed in the form of an appendix for the benefit of the more mathematically inclined. The writer has taught this subject for fifteen years and is fully aware of the pedagogic difficulties involved as well as of the limitations of knowledge. He has tried many different outlines and many different methods of presentation, and has chosen the one here employed because in experience it seems the most favorable for the presentation of the subject-matter involved and at the same time for putting the student in a frame of mind favorable for the undertaking of economic breeding operations and for the reception of new truths as they shall be discovered. Variation rather than heredity has been chosen for the initial and leading thought because better calculated, as experience has shown, to afford a favorable outlook and to develop such con- ceptions of evolution as are most useful later on. The evolutionist who might chance to scan these pages would be struck by the absence of some of the cardinal features of evolution, as he would also note the exceeding prominence given to certain other questions of seeming minor importance. Herein exists the difference between thremmatology and evolution, and this very matter has given the author more difficulty than all others, viz. to rearrange values and to determine proper relations of old questions in a new field. We must discuss the causes of variation even though we are told by the best students that such attempts are premature. A PREFACE vii minor matter in evolution, curious rather than otherwise, it is a vital one in thremmatology, and we must discuss the subject the best we are able, if only to learn how little we really know about it and to point attention in the right direction. No attempt has been made to include exhaustive references. On the other hand, they are confined for the most part to a few standard books easy of access, and to save time the references are mostly to definite pages. A general and more extended list follows the summary of nearly every chapter, enabling the student to pursue that particular subject further if desired ; but there is no attempt at a complete bibliography. It was hoped that if the list of references could be kept small the student and the breeder would be the more likely to provide themselves with standard literature bearing on the subject. I have made the freest use of standard authors, giving full credit in all cases, generally in the form of reference to text and page. This course has been dictated by the desire to furnish the student with reliable facts rather than a series of academic discussions upon disputed subjects. I desire to acknowledge the very great services of Dr. Rietz, to whom I am indebted for much assistance in the more statistical portions, and for the preparation of the appendix especially directed to the mathematical student, not as a text but as an introduction to further study in this special phase of science. I am also indebted to many of my colaborers in the Univer- sity of Illinois and elsewhere, as well as to numerous breeders in this and other states, who by their assistance have contributed much to any success which this volume may meet. Its possible merits, therefore, I must share with others ; its defects and shortcomings are my own. E. DAVENPORT University of Illinois Urbana CONTENTS INTRODUCTION Page PART I — VARIATION I. Variation in General .... I. Variation Universal among Living Beings II. Variability the Basis for Improvement III. Nature of Variability .... IV. Meaning of the Term "Character" V. Dominant and Latent Characters VI. The Unit of Variability VII. Distinctions as to Kinds of Variations References . II. MORPHOH)GICAL VARIATION References . III. Substantive Variation IV. Meristic Variation I. Symmetry . II. Meristic Variation in Linear Series III. Meristic Variation and Bilateral Symmetry IV. Symmetry in Variable Parts V. Meristic Variation in Radial Series VI. Importance of Meristic Variation References ...... V. Functional Variation .... I. Variation in the Degree of Activity of Normal Functions be- tween Different Individuals of the Same Species . II. Variation in the Degree of Activity of Normal Functions within the Same Individual ........ III. Modification of Normal Functions by External or Other Influ- ences ........... IV. Normal Functions exercised under Abnormal Conditions . References .......... 13 IS 17 24 25 29 30 33 34 39 65 68 70 73 74 75 77 9' 107 1C9 VI. Mutations no I. Distinction between Mutation and Ordinary Variation . .110 II. Examples of Mutation . . in ix CONTENTS Chapter III. Experiments of De Vries . IV. American Experiences V. Economic Significance of Mutations VI. Biological Significance of Mutations References . . . . . PART II — CAUSES OF VARIATION VII. iNTROnUCTION The Mechanism of Development and Differentiation I. Protoplasm the Physical Basis of Life . . . . II. The Cell the Unit of Structure . . . . . III. Mechanism of Cell Division (Mitosis) . . . . IV. Cell Division with and without Differentiation V. Physiological Units References ......... VIII. Internal Causes of Variation /. INTERNAL INFLUENCES AFFECTING PRIMARILY THE IN- DIVIDUAL I. Cell Division ......... II. Bisexual Reproduction a Fundamental Cause of Variation III. Maturation and the Reduction of the Chromosomes a Cause of Variation ........ IV. Bud Variation ......... V. Influence of the Condition of the Germ upon Development VI. Xenia, or Fertilization of the Endosperm VII. Telegony .......... VIII. Intrauterine Influences ....... IX. Reversion and Atavism ....... X. Individual Characters dependent upon Sex //. INTERNAL INFLUENCES AFFECTING THE RACE AS A WHOLE XI. Relative Fertility, or Genetic Selection .... XII. Physiological Selection ....... XIII. Selective Death Rate ; Longevity ..... XIV. Bathmic Influences ........ XV. Physiological Units ........ XVI. Germinal Selection ........ References ......... IX. External Influences as Causes of Variation I. General Effect of Locality upon Plant and Animal Develop ment ......... II. Influence of Food upon Variability III. Effect of Moisture upon Development IV. Effect of Contact upon Protoplasmic Activity . V. Effect of Gravity upon Living Matter ; Geotropism VI. Effect of Light upon Living Matter CONTENTS XI Chapter VII. Influence of Temperature upon Living Matter VIII. Effect of Chemical Agents upon Protoplasmic Activity IX. Effect of Saline Solution upon Development in Aquatic Animals ........ X. Influence of Use and Disuse upon Development XL External Influences as Causes of Variation in Type References ......... X. Relative Stability and Instability of Living Matter I. Evidence from .Stability of Type .... II. Evidence from Mutability of Species III. Evidence from Reversion and Atavism IV. Evidence from Disappearance of Parts V. Evidence from the Direct Action of the Environment VI. Evidence from Acclimatization .... VII. Evidence from Regeneration ..... VIII. Internal Factors in Regeneration .... IX. Evidence from Grafting ...... X. Evidence from the Origin of New Cells and Tissues XL Evidence from Development and Differentiation References ........ 254 264 282 285 290 294 295 296 298 305 306 307 308 316 332 335 336 338 345 XI XII. PART III — TRANSMISSION Transmission of Modifications due to External Influences I. Introductory ..... II. Evidence from the Nature of Variation III. Evidence from Mutilations IV. Evidence from Food Supply V. Evidence from Acclimatization . VI. Evidence from Habit and Instinct VII. Evidence from Use and Disuse VIII. Evidence from Disappearing Organs IX. Variations due to Causes not affecting the Germ are not Transmitted .... References .... Type and Variability I. Type ........... II. Variability, or Deviation from Type ..... III. Practical Hints on the Taking and Grouping of Measurement IV. Probable Error ......... V. Comparative Type and Variability for Different Characters in the Same Population ....... VI. Effect of Selection upon Type and Variability VII. Indirect Effects of Selection upon Type and Variability . VIII. Studies in Type and Variability of the Same Variety of Co raised under Different Conditions as to Fertility . References 348 348 356 364 370 374 386 404 409 416 418 419 420 425 435 437 444 445 447 449 452 Xll CONTENTS Chapter XIII. CORRKLATION Pace XIV. I. Meaning of Correlation II. Calculation of Coetiftcients of Correlation III. The Correlation Table .... IV. The Correlation Coefficient . V. The Regression Coefficient . VI. Studies in Speed Records of Trotters . References ...... Heredity I. How Characters behave in Transmission II. Statistical Methods of Study of Heredity III. The Regression Table .... IV. Like Parents beget Unlike Offspring and, conversely, Like Offspring may be begotten by Unlike Parents . V. Regression. In general, the Offspring is More Mediocre than the Parents ; that is, Whatever the Parentage, the Offspring exhibits a Strong Tendency to regress toward the Mean of the Race .... VI. The Measure of Heredity VII. The Mean of the Offspring not necessarily the Same as the Mean of the Parentage VIH. Extremes of a Race relatively Less Productive than the Means ...... IX. Progression. Parents in general produce a Few Individuals More Extreme than the Race X. The Exceptional Individual arises either from Mediocrity or from the Exceptional Parent XL Fraternal Variability, — Offspring of Same Parents nol Identical ... ..... XII. Characters tend to combine in Definite Mathematica Proportions ........ XIII. Mendel's Law of Hybrids XIV. The Law of Ancestral Heredity ..... XV. Limit to the Reduction of Variability .... XVI. Power of Selection to permanently modify Types by the Establishment of Breeds XVII. Breeding True, or Stability of a Character established by Selection ........ XVIII. Duration of Varieties, Breeds, and Family Strains References ........ XV. Prepotency I. Data from the Trotting Records illustrating Prepotency II. Prepotency in Sex ........ III. Influence of Age on Prepotency ..... IV. Influence of Constitutional Vigor upon Prepotency References ......... CONTENTS Xlll PART IV— PRACTICAL PROBLEMS Chapter XVI. Selection .... I. Ideals in Selection II. Historical Knowledge of the III. General Principles involved i IV. Rational Selection Breed Necessary n Selection XVII. Systems of Breeding . I. Purposes in Breeding . II. Grading .... III. Crossing or Hybridizing IV. Line Breeding V. Inbreeding VI. Breeding from the Best References XVIII. The Determination of Sex I. Theories II. Influence of Nutrition . III. Influence of Fertilization IV. Sex in Mammals . V. The Accessory Chromosome References XIX. Plant Breeding . I. Advantages and Limitations 11. Soil and Culture Conditions III. Systems of Planting References XX. Animal Breeding I. Advantages and Disadvantages II. Fewer Characters for Selection III. Fashion ..... IV. Show-Ring Consequences V. Testing of Sires and Dams . VI. Weathering a Period of Depressio VII. Records VIII. Disposal of Surplus Females IX. A Market for Sires X. Community Breeding XL The Young Breeder References .... XXL Development .... APPENDIX INDEX Sex Determinat the Herd 577 578 579 581 592 599 599 602 6c8 610 613 626 628 629 629 631 632 634 634 637 639 639 641 643 651 654 654 656 658 660 660 665 666 672 673 674 675 676 677 681 THE PRINCIPLES OF BREEDING THREMMATOLOGY INTRODUCTION The main title of the present volume was chosen because self- explanatory, though it less accurately expresses the scope of the subject than does the sub-title, which is only beginning to come into general use. Thremmatology, from the Greek tJircninia, a thing bred, is a term proposed by Ray Lancaster ^ to cover the principles and practices concerned in the improvement of domesticated animals and plants. The term is broader than the " principles of breeding " because it includes development as well as reproduction. It is distinct from evolution in general, which attempts to explain the princi- ples and forces connected with the origin and development of all forms of life but without the slightest reference to economic considerations. In evolution a protozoon is as important as a pig, a hydra of as much significance as a horse, and the njost pestiferous weed as much an object of interest as either corn or wheat. Thremmatology limits itself to those species and varieties whose natural qualities made them useful to man in the beginning, and it asks and seeks to answer this one question. How can they be made still more useful and better adapted to the purposes of an advancing civilization } In this study forms of life that are of no economic value are of no special concern except as their consider- ation may throw light upon domesticated forms. This latter is often the case, for it is doubtless true that the same principles apply to all species, economic or otherwise, because none were ^ See Encyclopedia Britannica, ninth edition, XXIV, 841. 2 INTRODUCriON si)ccially created lor man any more than lOr any other animal. It is a general biol<)L;ical truth that everythini;" lives unto itself, — not where it choses but where it can ; not upon what it likes best but ui)on what it can get. It may seem to the student that an undue amount of attention is given to variation and that a disjjroportionate amount of space is devoted to that subject. In that event I have to say that variation is not the antithesis of heredity but rather its constant and insep- arable attendant, and that the facts of variation constitute the best portion of that stock of information with which the student must become possessed before he is ready to study the principles involved in those generalizations ui)on which i)ractical operations may be safely based. One is painfully aware, too, of the necessity of ranging far and wide for facts, and the student cannot fail to feel ofttimes that the subject-matter in hand is far removed from agriculture. When this is the case it is because we arc forced to take what is avail- able and make the most of it. Unfortunately the workers in strictly agricultural fields are all too few and the reliable data deplorably meager, though some original and I trust valuable matter has been recently added to our stock of knowledge. While the same principles doubtless apply in thremmatology as in evolution, yet important distinctions are to be observed. First, there is every reason to suppose that even fundamental laws apply to different species in different ways. For examj^le, Indian corn seems particularly sensitive to close breeding, whereas wheat is almost exclusively inbred and has been so inbred for unknown generations. Again, the circumstances of the case often introduce into the problem certain economic considerations not resting upon general evolution. Iu)r examjile, man cannot afford the "countless ages " and "untold generations " which are accorded nature for accomplishing results. In practical breeding operations substantial results must follow at once and exhibit a high degree of success within the period of a lifetime or they will be discarded as valueless. Experience shows that the purposes, standards, and methods of a successful breeder are seldom handed down from one man to another, even to his own son. Even if that could be done, it INTRODUCTION 3 would constitute no exception to the rule that a man must real- ize the fruit of his own labors and in his own generation, for breeding is a business and must be made to pay. The breeder must therefore work faster than nature, and thremmatology can- not make use of a leisurely operating evolutionary principle unless its action can be accelerated or its cumulative effects exaggerated. Yet again, man cannot afford the immense numbers and the whole- sale destruction that characterize nature's methods of working changes. Animals and even plants cost money, and a relatively large proportion must meet the conditions, or the enterprise must be abandoned. Business considerations therefore set arduous limitations upon thremmatology in respect to both time and numbers from which evolution in general is entirely free. The breeding business has its own particular problems, some of the most important of which unfortunately the known facts of evolution are least able to answer. The profitable study of this subject will, however, be assisted by a clear statement of these problems. The problems of the breeder. Certain questions stand clearly out in the minds of practical breeders, and though an attempt to answer them seriatim would not be the best method of study, and though some of them cannot be answered with certainty in the present state of knowledge, yet nothing is of more con- sequence at the outset than that the student get a clear idea of the problems needing solution and towards whose solution the study of thremmatology is directed. They are substantially as follows : To what extent are the characteristics of an individual at matur- ity due to its ancestry (heredit)), and to what extent are they due to the conditions of life (environment), such as food, climate, exercise, and general care during development ? Are the influences of the conditions of life limited to the indi- vidual or are they in certain instances and to some extent carried over upon the offspring ? that is, are the effects of envirfmment inherited ? Can variations be directly controlled to any extent whatever, or only indirectly through selection and by special care during development ? 4 IN I'KOnill'lON I low rltiHtiM" is scK'it ion in conl i ollmi; \.iii,ition? ill, it is. an." i-onj;cnit.il x.ni.itions ilur rnliri'K to pau-nta^c or an- llu-n" hack ol the i\uontai;r ivitain inhcivMil and (.-onstitutional Iciulcncios that lai\!L;x"l\' iix the i;cnoial iHiciiion ol \aiiations, indcpiMulcnt ol solortion ? DiH's iniproxonuMit consist in raisins; the slandard ahsoUitoly or onl)' in raisin^; the _i;vnoral a\oia,i;c h\- ohminatini;' tho loss dosir- ahlo ? that is. s hioodinj;' improve upon ihe host or doos it only hrinj; tho _i;iMUMal mass noaror to iho nppor lowl ? I'o wliat oxtont is oxolution a <;ra(hial pnnoss and to wliat o\tonl ma\' prolound aiKani'os appear siuldonh . as in sports ? and is the one class ol improx oniont an\ more poimanont or roli- ahlo than the othoi' ? 1 )o all possihlo \ahios ol a \ariahlc oharaitcM' appear or arc certain values seKlom or nexer presented ? that is to say. is variation al\va\s eonlimious or is it sometimes diseontinuous, making' certain things impossihle hecause [\\c propiM' filia- tions or eomhinations i\i) not appear o\\ which selection may he hased ? \\ hat \ariatiiMis are most likeh to appear in successive gener- ations ol an\' _i;i\en hreed. \ariet\. or l\pe? .Are variations ci)rrelaleil ? that is, do they teml at ail to move together, sui^'-^estini;" lelations ol cause and effect ? What art." the jMoper standards h>i' selcctiiMi ? I low nuuh shall ho i;i\en \o utilit\- and hmv much to appearance? To what extent is individual excelleiu~i" a sale i;uide to breet.!- inj;" powers ? To what extent is the offspring;' like the inunediate parent and to what extent does it resemble more remote ancestors ? What is the relatixe inlhience of sire and dam with respect to transmission of chaiaclers ? To what extent k\o the condition iA the male at the lime of service and the care oi the female durini;' pre_i;nanc\- inlhience the olfsprins;" ? What aie the real dani;ers from (."lose breeding, it an\. and are the\- certain or only jirobable ? llow can the ad\anta.<;es of close or line breeding be realized without encounterinL! its danuers ? Ii\"l PC)!)! f I ir)N 5 Will a ^ivcn brccfl, variety, or family strain cnfliirc inrlcfi- nilcly iiriflcr pro|><;r cnifiitions or will it inevitably "run out," necessitating a constant return to founrJation stock for new cf>mbinations as the fjasis of improverl strains ? What arc the laws that determine the sex of offspring ? Do the same laws of breeding aj^jjly equally to animals and to plants anfl to all species and varieties alike, or rlo rlifferent species operate vmrler somewhat flifferent laws ? Is a given species, variety, or breed always subject to the same laws ? that is, are identical variations always due to the same causes and do given causes always jjrfxluce the same effects ? How can results be secured with the least wastage either in time or numbers ? Upon the answers to these questions will depenrl the policies of all breeding enter[jrises and the permanent value of particular family strains. Upon some of these points there exists much specific and reliable information ; upon others, unfortunately, the evidence is yet scanty and uncertain. At the present rate of progress, however, we will not have long to wait for much addi- tional knowledge. In the meantime we must make the best use f possible of the information and experience at hand. These problems can best be answered not by directing atten- tion tr> each separately, because they overlap, but rather by follow- ing out what are known to be the characteristic lines of study in the subject as a whole. The order pursued in this brK>k is the one believed to be most favorable both for this purpose and for the most successful answering of these definite questions.' ' A fair knowledge of general evolution i.s presumed on the part of the student and reader. If this is not in his possession, he will do well to read at least J Irwin's <^Jrigin of Species for a broad though now somewhat old and incomplete outlook upon the general field. If he desires to go further and enter the field of controversy, he can do so most directly by readmg Weismann's Kssays on Heredity and his Germ I'lasm, together with Romanes' Examination of Weis- mannism and his two volumes of I>arwin and y\fter Darwin. If this is done, it would be well to fini.sh with Habit and Instinct by Morgan. Part I — Variation CHAPTER I VARIATION IN GENERAL SECTION I— VARIATION UNIVERSAL AMONG LIVING BEINGS The most obvious fact about living beings is their variability. Not only do species differ from each other by many and widely different characters, but individuals within the species are distin- guished by differences readily discernible, at least by the trained observer. The general differences between horses and cattle, for example, are specific and distinct and therefore striking even to the casual observer ; but to .the trained eye all horses are not alike, and so it is that differences are detected within the species. Two individuals may be recognized as possessing the same char- acters and therefore related by descent, but invariably these characters differ in degree or in their proportions one to another. Two animals may be of the same or of different colors, but in either event the parts are differently proportioned. The leg of one is longer, larger, or more crooked than that of the other. The bones composing the two are not of equal, or even of pro- portional, lengths. Two cows of the same breed differ marvel- ously in the amount of milk they can yield in a year, and some are known to produce three times as much butter fat as others from the same amount of the same kind of feed.^ Again, some milk is rich in fat (6 or even 8 per cent) while other is poor (2 per cent or even less). Some horses, because of their con- formation, travel more easily or more rapidly than others, and some are more intelligent or more enduring or more docile. 1 See data from Agricultural Experiment Station, University of Illinois. 7 8 VARIATION Some dogs (bloodhounds) trail marvelously well ; others (grey- hounds) scarcely at all. Some hens lay more eggs than others, and of different color and size. Some animals are hard feeders, while others lay on flesh readily. Some beef is coarse in its grain ; other is fine and ten- der. Some is well marbled with fat ; other is not. Sometimes the flavor is dehcate ; again it is rank, and often it is insipid. No two trees are alike in their growth or branching habit, though similar within the same variety, and the widest difference is often found in leaves from the same tree. Differences extend to minute particulars and include all charac- ters. The student should early form a clear conception of the fact that differences extend to all characters however insignifi- cant or minute. Besides, he should understand that they include function as well as structure, and that not only external anatomy and conformation are involved but internal organs and their activity as well, and no greater mistake can be made than to define evolution as " a study in morphology." If we so define the word " variation " as to cover any change in detail of structure or function which our faculties enable us to detect, then we may say that variation extends to all charac- ters, internal or external, structural or functional, and if the study lay in the realm of ethics, economics, philosophy, or reli- gion, we should add, material or immaterial. The individual is therefore so distinctly a unit that its iden- tity is at once recognized and the principle is conceded that " no two are alike " and that variation is universal. Limitation of variability. The exception to the universality of variability is in the realm of non-living matter. The specific gravity and other properties of iron, gold, sodium, or chlorin are constant, and their relations and combining powers with other chemical substances are, under identical conditions, invari- able and therefore well known. Oxygen, hydrogen, and even nitrogen and carbon, combine always in definite proportions, and though their combinations are exceedingly numerous yet when the conditions are known the exact combination can be foretold ; moreover the properties of this combination will not only be definite but they will be identical with those of all other similar compounds. In this VARIATION IN GENERAL 9 way sodium chlorid, for example, has always definite and well- known properties not subject to variability. It is to be noted of course that the properties of the com- pound NaCl are totally different from the properties of the ele- ments that compose it, Na and CI. They are nevertheless distinct and invariable as well as new. This distinction between the varia- bility of living matter and the constancy of non-living matter should be borne in mind later on when discussing some of the causes of variation. SECTION II — VARIABILITY THE BASIS FOR IMPROVEMENT Improvement is possible only where variability exists. The compound NaCl being constant, it would be impossible to pro- duce an improved variety of sodium chlorid, because the com- pound is always the same and cannot be had with other than its standard and invariable properties. Improvement in this com- modity is limited, therefore, to its mechanical form and cannot extend to its constitution. Living matter, upon the other hand, while possessed of defi-, nite properties, does not exhibit these properties always in the same degree, and observation and experience have both shown that profound changes may be made in either the form or the constitution of both plants and animals by the simple method of judicious combinations of desirable deviations. If there were no variability, and if living matter were as con- stant in its properties as is non-living matter, then we should be certain of what we already have, but no improvement would be possible. As it is, with variability everywhere, living organisms are both capable of improvement and liable to degeneration, for both are the logical consequence of variability. Man must there- fore work for what he possesses in the way of animals and plants, and they will serve him well or ill according to his knowl- edge and skill in dealing with their variations. Accordingly he cannot know too much about the variations that are likely to occur, — their nature, their extent, and the causes that control their appearance and determine their permanency. lO VARIATION He cannot know too much about life and its vicissitudes, about living things and what they do. An animal is born into the world. Its energies are first devoted to nutrition and growth. It builds its own machine and builds it quickly out of materials lying close at hand. In good time it is finished and all its energies are at a maximum. It seems like a stable thing that must live forever. But repro- duction occurs, securing a succession of its kind. One after another of its faculties fail, and its condition is again reduced to that of bare existence, with youthful recuperative powers gone forever. By and by some vital function fails. Then life goes out ; the organism breaks down and returns its elements to the inorganic world. Such is the brief history of a bit of matter temporarily endowed with life, — fleeting as a breath; any service it may render us must be caught in the passing. SECTION III — NATURE OF VARIABILITY The exact nature of variability is a most obscure subject, and one that cannot be fully comprehended in the present state of knowledge. Whether the distinctions between living and non-living matter will always remain as marked as they now seem to be, only future discoveries will determine. We have as yet only touched the fringe of this great subject, but enough is known to enable us to begin to penetrate some of its mysteries. At least two general principles may be laid down in the present state of knowledge without much chance of error : 1. That all characters of plant and animal life, whether struc- tural or functional, are exceedingly variable. 2. That ordinary variation is the result of a change in the relations between a number of associated characters through the deviation of one or more of the members, and not the introduc- tion of an absolutely new character. We speak loosely of " introducing new characters," but in truth improvement consists, not in the introduction of absolutely new characters, but in the intensifying of desirable old ones and the subordination of those that are undesirable. For example, VARIATION IN GENERAL I i when we undertake to improve the quahty of wool we hmit our attempts to the sheep, with which wool bearing is a natural character. Whether the horse or the hog could be made to grow wool is a question. If the hen could be made to produce milk or the cow to grow feathers, that would be the introduction of a new character in the strictest sense of the term. But nothing similar to this has ever been accomplished by man. The particular group of characters that constitutes a given species appears to be strangely fi.xed, and improvement seems to con- sist in changing the relations of these characters among them- selves rather than in the introduction of new members. How this particular grouping arose originally and how a new member (character) might be introduced are questions for the student of general evolution. They are questions, moreover, upon which the present state of knowledge sheds little light, and, so far as is known, the study of the practical breeder is limited to methods of dealing with groups of characters already associated and con- stituting well-marked types and forms. SECTION IV — MEANING OF THE TERM "CHARACTER" This is a much-abused term, loosely used in a variety of mean- ings. For example, when an individual differs slightly from another we say he has different characteristics. What we really mean is simply that his characters differ in their development, not that he has different characters. His bone is not so round or his hock so crooked ; the crops are not so full or the milk so rich ; the eye of the potato is not so sunken or the color of the fruit so high in one specimen as compared with another, and we say loosely that the characters are different. Now the truth is the characters are not different in kind but only in degree and proportion. We say of one horse that he has speed and of another that he has not speed. The fact is that they both have some speed, but only one has enough to attract attention and be worthy of remark. This use of terms, unfortu- nate as it may be, is probably too common to be changed ; indeed, the mere use of terms is of less importance than a clear compre- hension of the facts. 1 2 VARIATION The term ** character " is employed in this text to designate one of those details of form or function which, taken together, consti- tute a well-marked group of animals or plants more or less closely related by descent, and this is the only sense in which the term oueht to be used. Thus the color characters of the horse are black, bay, brown, gray, etc., but not red, green, or blue, although these characters are not unknown to the animal world, being common with birds. Used in this sense, a " character " belongs primarily to the race or group of which the individual is a member. It is there- fore not peculiar to any particular individual and is in no sense personal property. Thus not only the color of the coat but the form of the body, the peculiar function of any of its organs, as in milk production or the secretion of poisons, any special mental attitude or intellectual function, or even a particular crook of limb or special body marking of any kind that runs commonly through the groups, is properly spoken of as a racial character. Those characters do not come and go, but on the contrary they remain with the race indefinitely. The individual horse, for example, will be marked by one or possibly more of the color characters of his kind, — black, white, bay, etc., — but he will not be marked with characters not of his kind, as red, green, etc. From this we see that the individual is not so good a unit for study as is the group to which he belongs and the racial characters that compose it. Now the personality of the individual so strongly impresses us that we instinctively regard him as an actual unit, and we speak loosely of his characters as if they were personal property peculiar to this individual alone, whereas he possesses nothing that is not common to his race. His differences are in degree, not in kind. What we mean to designate in the individual is the particular combination of racial characters that make up his personality, knowing perfectly well that the characters of all individuals within the race are racial characters and no other, and that every indi- vidual that may ever arise by descent will be limited as to his details to some combination of the characters of his race. Now the characters of any race are so many, their deviations are so VARIATION IN GENERAL 13 wide, and their power to move independently of one another is so great that, according to the doctrine of probabiUties, an ahnost infinite variety of combinations is possible. Hence no two indi- viduals are ever likely to be identical. We thus arrive at the conclusion that the proper study of the breeder is not so much the individual as it is the normal characters of the race to which he belongs. SECTION V — DOMINANT AND LATENT CHARACTERS The race as a whole clearly possesses more characters than can ever be utilized in the visible make-up of any single individual. Among" all the colors of horses, but one, or at most two, can be found in any special instance. The race is therefore a kind of composite of all the individuals that compose it, or, more properly speaking for purposes of study, it affords a wide assortment of elements out of which individuals are composed. The individual transmits the characters of the race. If the group of characters constituting a species is larger than that constituting an individual, as with color among horses ; and if an individual may transmit a character which (apparently) he does not possess, and experience shows that he does, then it follows that the individual is in actual possession of more characters than those directly involved in his visible make-up. For example, the offspring of two black horses will likely be black, but it may be bay, brown, or any other color characteristic of the horse kind. It is safe to say that it will not be red, green, or blue, because these colors are known not to belong to the horse kind, though all are freely found in nature. Milk secretion is confined to the female sex, yet a bull whose dam is a heavy milker will transmit milking cjuality almost as successfully as will a cow. In this instance the male transmits a quality that he does not apparently possess and that could not become functional in his case. From this we infer that the individual, whatever his particu- lar make-up, transmits all the characters of the race, and none other ; and that he is therefore possessed of all the racial char- acters of his kind in some degree visible or potential. From 14 . VARIATION this we conclude that the apparent make-up of the individual depends upon the particular characters that happen to be strongest, that is, most highly developed in his case, but that he is in actual possession and may transmit any and all the characters of the race to which he belongs, but no other. We now arrive at the distinction between dominant and latent characters, which is as follows: Those characters that are prom- inent in any individual are said to be dominant with him because well developed and plainly evident, and all other racial charac- ters are said to be latent because not evident, although they are known to be present from the fact that they are transmitted to the offspring, often becoming the dominant characters in future generations. The term "latent" should not convey the impression of hidden or lurking characters, but rather undeveloped possibil- ities of the race within the individual in question. With this conception the student will be saved much mental confusion when dealing with heredity and reversion. Elementary characters. In a biological sense the ultimate unit of variability, therefore, must be something less than the racial characters which we have been discussing, because they themselves are complex rather than simple. We speak of the leg of a horse or the quality of milk as a whole. Even if we narrow the point to the conformation of the hock or the propor- tion of fat, we yet have characters clearly made up of parts. The hock is an exceedingly complex structure, and seven, pos- sibly nine or ten fats and oils are found in the fat of milk. As almost unlimited color effects are made up by few pri- maries in different proportions, and as all ordinary materials are made up of a few chemical elements in different combinations, so in all probability if we could make the ultimate analysis we should find that all these characters are made up of definite liv- ing units, that we may call, for want of a better name, elementary characters. These elementary characters have received many and various names. They are the stirp of Galton, the biophors of Weismann, and the physiological units of biologists generally. In general they are the smallest conceivable living units, comparable with VARIATION IN GENERAL 15 the molecule of chemical compounds. Such elementary charac- ters are supposed not to be variable except as they effect dif- ferent combinations among themselves. SECTION VI— THE UNIT OF VARIABILITY The unit of variation is therefore not the individual but the racial characters that constitute the particular, group, and that run down the line of descent like the strands of a rope and out of which individuals are made up, — some with one combination, others with another, after the fashion of threads in a fabric, forming patterns here and there, now of one design now of another, as they wander apparently aimlessly here and there. ^ It is evident, however, that the actual basis of character devia- tion is sometimes exceedingly complex. Milk secretion, for example, while limited to certain animals and confined to the female sex, is properly recognized as a distinct character ; yet its successful functioning depends upon a variety of considera- tions, — the general health of the body, the nervous tempera- ture of the individual, the power of the stomach to provide large quantities of prepared food, the ability of the kidneys to do their work, and the power of every organ in the body to dis- charge its function successfully and fully under heavy strains. All these are as important to successful milk production as are large and active milk glands, and an accident at any point will cause deviation in milk yield either as to quantity or quality or both. Deviation in color, on the other hand, may be due to presence or absence of pigment, which may be regarded as a chemical substance secreted at a single point. In this and in similar cases the actual basis of deviation is simple and readily detected. From this it will appear that the ultimate seat of variation, whose fluctuations are responsible for character deviations, may be exceedingly difficult if not impossible to identify. 1 The unit of variability must not be confused with the unit of selection : the latter, of course, is the individual. We cannot separate his characters, but must take him as he is, for better or for worse ; but we must do so fully realizing that each of his separate characters has an identity of its own, so that the unit of vari- ability is far within the necessary unit of selection. 1 6 VARIATION It has been assumed that the ultimate unit of organized beings is the cell. This is true in a structural sense only, for there is positive evidence that the cell is itself made up of various and distinct elements capable of somewhat independent action in both cell division and growth. The content of a cell is not to be regarded as a mass of amorphous protoplasm to be halved or quartered by chance, but on the contrary it is an organized body with a distinct difference between the nucleus with its definite number of chromosomes (the supposed seat of the physiological units that give character to its activities) and its surrounding cytoplasm or food material. Again, a whole group of similar cells may constitute a special organ (liver, kidney, or heart), discharging a highly specialized function quite different from that of any other portion of the body. This colony of many cells discharging the same function appears to move together, thus constituting a kind of functional unit larger than and quite distinct from the ultimate physio- logical units that must reside within the cell. Correlated variation. Still again, it is found in practice that occasionally whole groups of characters seem to be so correlated as to move together, so that having found one we may reason- ably expect to discover the other. Familiar examples of this are found in nearly all cases of reversion. For instance, a white calf among Devon cattle will almost certainly show black or brown points (ears, nose, and legs), while a white shorthorn will not. The one is a case of reversion to the ancestral color of the breed, — the wild cattle of Britain; the other is simply the appearance of one of the normal color characters of the race. In increase of numbers of parts there is some tendency to repeat a whole group, as in cases of " double hand." ^ The same tendency for many distinct characters to move together in groups is found in cases of so-called " sports," in which we instinctively recognize something more than ordinary variation. These instances of grouping of characters normally independ- ent in such a way as afterward to move together is to be dis- tinguished from such variation as is involved in extreme milk 1 See under Meristic Variation. VARIATION IN GENERAL 17 production, in which results arc to be ascribed rather to fortui- tous variations among independent units than to anything like a linking together of the separate characters involved. The ultimate unit. The student must maintain a clear vision as to distinctions of this sort. From the biological standpoint we accept as the unit of variability the ultimate physiological units that must reside within the cell ; for purposes of everyday use, however, we assume as a unit that detail of form or function that is important to the breeder, understanding perfectly well that, considered physiologically, it doubtless consists of a num- ber of ultimate units which, like the elements in a chemical com- pound, are entirely capable of other and distinct combinations. To repeat, we may assume either one of these conceptions as the unit for study according to the purpose in hand. In the text the word " character " will be used to denote that detail of (racial) structure or function which is important to the farmer and to this study. The character will be considered as the practical unit of variation, knowing perfectly well that the ulti- mate unit of variability lies much farther back in the constitution of the protoplasm itself. When this is in mind the terms " ele- mentary character," "physiological unit," or some equivalent term will be used. Whatever the situation, the student must not consider the individual as the unit of variability or he will come to grief both in study and in practice, nor should he become confused by those occasional and remarkable correlations of characters that suggest a unit of variability unduly large. SECTION VII — DISTINCTIONS AS TO KINDS OF VARIATIONS In the critical study of variation it is necessary to observe certain distinctions that are often overlooked, resulting in more or less confusion as to what is really involved in the term "variation." Variation quantitative or qualitative. The first c[uestion that should arise in the mind of the student touching any deviation is this : Is the difference one in degree merely (quantitative) or 1 8 VARIATION is it a difference in kind (qualitative) ? For example, one horse is exactly like another, only larger : ' the difference is quanti- tative. Another is no larger, but he can draw more and has greater endurance : the difference is qualitative. One cow gives more milk than another : the difference is quantitative ; but a third gives better milk, and the difference is qualitative. One apple is larger than another of the same variety (quantitative variation), but another is different in texture and flavor (quali- tative variation). When, therefore, in the study of a racial character as repre- sented in the same or in different individuals it is found to have varied, the first question to ask and answer is this : Is the devi- ation one in kind or merely in amount ? Is it qualitative or merely quantitative .? Is the change to be regarded as one in nature or only in degree .'' If the student will carry these distinctions always in mind, he will avoid much needless confusion. Variation continuous and discontinuous. It is not to be assumed that variations differ from one another by infinitesimal increments. The differences may be infinitesimal (continuous variation) or they may be "discrete " (discontinuous variation). Darwin supposed, and it is commonly assumed, that variation is by nature continuous, and that new forms originate by the gradual accumulation of insensible differences through the agency of long-continued selection. This means that if all the individuals that ever lived could be assembled and so assorted as to bring nearest together those that are nearest alike, it would then be found that they would grade into one another by imperceptible differences, and that any gaps that might occur would be due to the effect of selection in blotting out intermediate forms. Now this is a hasty assumption, indicating in these days but a superficial acquaintance with the manner of variation. We cannot assume that all possible values in variable characters are presented ; indeed, we know very well that in many cases all possible values are not presented, and that some intermediate forms never arise. For example, peaches often give rise to nectarines, but there is a gap between the two that apparently is never filled. Darwin called an occurrence of this kind a "sport," as if it were an instance in which all ordinary laws were set aside, VARIATION IN GENERAL I9 whereas it only shows that the variations of the peach are often discontinuous, with wide gaps representing spaces not filled by- variation. To get a full understanding of this ground the student must form a clear conception of the distinction between continuity and discontinuity as used in this connection. A man grows from childhood to maturity. In doing so he passes through all possible weights and heights between those of infancy and maturity. We cannot represent all these values by any of our units of weight or measurements because all numbers are by nature discontinuous. The only measure of continuity is a line, because a line, curved or straight, represents all values sensible and insensible between its two extremes. We can thus plot continuity, but we cannot measure it except by cutting it into sections and measuring it at stated points as if it were dis- continuous, ignoring the intervals. Changes of temperature are continuous, as are those of humidity, illumination, and all growth in the sense of extension in size, whether plant or animal. All motion, whether regular or irregular, is continuous because all intervening spaces are included. Discontinuity, on the other hand, implies vacant spaces not represented by values. The good singer goes abruptly from one note to the next, giving a discontinuous series of tones, while the unskilled vocalist slides up or down the scale, giving rise to a continuous series of tones in his effort to find the proper note. The latter is not music because the ear is not pleasantly affected by this confused jumble of sound waves arising from the inter- mediate tones. Good music consists of a series of tones not flowing into one another but cut sharply off and cast into a discontinuous series, striking the ear at intervals with sound waves that fit with mathematical precision. All number is by nature discontinuous. By fractions we attempt to bridge the space between contiguous units, as between i and 2 ; but however small the fraction, there is yet a space, and a sensible and measurable one, between the fraction and the next unit. lyVfToTj- is not 2, nor will it ever become 2 this side of infinity by the addition of any number of nines to the numerator and ciphers to the denominator. It will constantly approach 2, 20 • VARIATION but will never reach it, because definite number is discontinuous. This is why we can never accurately measure continuity except by a line, and this is why we cannot express in numbers the growth of an animal or plant, except approximately in terms of discontinuity. All chemical compounds are made up of elements in definite proportion. They are therefore discontinuous. We have H2O and H.2O.2, but no intermediate is possible, — this again for numer- ical reasons. Plants and animals generally are dimorphic, the one form being male, the other female : this is discontinuity. Some species are trimorphic or even polymorphic. The ant is either male, female, worker, or soldier, and though they all belong to the same species there are no connecting links in this discontinuous chain. Dimorphism without respect to sex oc- curs in many beetles, and is exceedingly marked in the common earwig,^ as shown in the accompanying diagram. The shape of this curve shows clearly Fig. I. Dimorphism iiius- ^hat here are two distinct forms of male, tiated: two types of the a large and a small one, living together common earwig. B is and arising naturally out of the common more common than A, . , • , ^ • , t , , ,, . . mass, yet showmg almost no mtermediates. and the type is more pro- ' ^ => nounced, but there are Students of breeding familiar with the almost no intermediates, older types of Hereford will recall that the — After Bateson i i ■, . ■,• i--^ii,j_ ^• breed was almost dmiorphic m that two dis- tinct types tended to appear with singular perverseness, refusing either to blend or to undergo modification. There was the old, large, solid-bodied, thick-meated, deep-ribbed type, ideal except as to lateness of maturity ; then there was also the pony-built type, — short in the barrel and lacking in depth behind, though well proportioned in front. The shorthorns are almost polymorphic in possessing not one but a variety of types, each standing out with extreme distinct- ness and not readily merged. The more the matter is examined the more it will be seen that strange and unaccountable gaps are found everywhere. Many a ^ Bateson, Materials for the Study of Variation, pp. 36-42. VARIATION IN GENERAL 2 1 breeder has spent his life and his substance in the vain attempt to produce a desired intermediate between two forms, either one of which are easily secured. The question arises, therefore, Are some intermediates impossible ? Can we bridge the space between the nectarine and the peach ? between the apricot and the plum ? With all these examples before us, we see at once that to pro- ceed upon the theory of continuity is a gratuitous assumption not borne out by facts. Force and physical agents generally seem to be continuous in their different manifestations, shading one into another with imperceptible gradations ; but organized matter, living or non-living, seems to be constructed upon the plan of dis- continuity, in which case we may expect to find differences that are easily perceptible, and should not be surprised at the appear- ance of wide spaces between nearly related forms or at these remarkably distinct gaps that often occur between a standard form and its offset, which Darwin called a " sport " and which we in these days call a " mutant." With this view of the case, we should not expect to find all nature united by imperceptible gradations, even providing all living beings past and present could be assembled and assorted according to nearest resemblances. Realizing the discontinuous nature of all chemical combinations, living or non-living, we should expect to find notable gaps representing spaces not taken by any possible form, and appearing quite independent of any selective process. This distinction is exceedingly important at the outset of this study. If all variations are continuous, then all shades of differ- ence, however minute, may be expected to occur naturally, and we may hope to secure them by breeding. If, however, some variations are discontinuous, then for these characters minute gradation is impossible, and we may expect descent to follow along certain lines only. Most of the conditions of life are without doubt naturally con- tinuous in their variations. This is certainly true of temperature, moisture, light, and food. Discontinuity must therefore arise from within, and is evidently connected with the nature of organisms. This is not difficult to appreciate when we recall the essential discontinuity of all chemical compounds or other organizations built upon the basis of distinct units. 2 2 VARIATION The student must not therefore assume the possibihty of inter- mediate gradations and insensible differences when deahng with biological phenomena. Many of these differences are essentially and necessarily discontinuous. It remains to discover which these are and to discover the bearing of discontinuity upon the results to be accomplished by selection. If all variations were continuous we might hope to be able theoretically to accomplish any desired result and secure any desired shade of difference by selection ; but if not, then there will remain notable gaps that cannot be filled. The natural corol- lary of all this is that we can accomplish by selection almost any desired shade of result with those variations which are by nature continuous, but that with those variations which are by nature discontinuous, our efforts in this respect will be limited. Distinctions arising from the nature of the characters involved. Having determined whether the deviation is quantitative or quali- tative, continuous or discontinuous, we next inquire into the real nature -of the variation as it affects the organism. Manifestly this depends upon the character or characters involved. Those concerned with form will, in their deviations, give rise to morphological differences. On the other hand, deviation in char- acters distinctly functional will give rise to differences in organic activity without regard to form. Accordingly four distinctly different kinds of variation are recognized : 1. Morphological, relating to differences in form or size. By nature they are always quantitative, but may be either continuous or discontinuous. 2. Substantive, relating to differences in quality of the struc- ture as distinct from mere form or size. By definition they are always qualitative and generally, if not always, continuous. 3. Meristic, relating to deviations in pattern, especially as to repeated parts, as in extra fingers and toes, doubling of petals, stooling of grain, etc. Variations of this kind are either quanti- tative or qualitative, generally the former, but are of necessity discontinuous. 4. Functional, relating to deviations in the normal activity of the various organs and parts of the body or the plant, such as VARIATION IN GENERAL 23 muscular activity, glandular secretions, etc. They are either quantitative or cjualitative, continuous or discontinuous, though rarely the latter. A clear understanding of these distinctions is necessary to an intelligent study of the nature and causes of variation. Accord- ingly enough attention will be given to each to acquaint the student with the way in which variation behaves, partly for its own sake and partly as preparation for careful inquiry into methods of dealing with deviations in those plants and animals that we have domesticated and appropriated to our use, and which we would see still better adapted to the purposes of man. Summary. Variability is the universal rule among living beings. Literally no two are alike. The differences extend to all characters and to the most minute particulars. Non-living compounds exist in definite proportions, and their qualities are constant, not variable. Variability is the only basis for improv- ment. No improvment is possible, in the strictest sense of the term, with respect to inorganic compounds, but living matter being variable is capable of change and therefore of improvement. Variation consists not in the introduction of new characters but in different proportions or relations among the old ones. All characters are racial, and all individuals actually possess all the characters of the race and none other. This is shown by the characters that are transmitted to the offspring. The unit of variability is in no sense the individual, though he must be accepted as the unit for selection. The real unit of devia- tion is the racial character, but back of that, in a biological sense, lie the elementary characters or physiological units, whose vari- ous combinations constitute racial characters. Variation is both quantitative and qualitative, both continuous and discontinuous, and these distinctions should be clearly in mind at all times. Special Exercises Prepare a list of variations that are (i) quantitative, (2) qualitative, (3) morphological, (4) substantive, (5) meristic, (6) functional, (7) con- tinuous, (8) discontinuous. 24 VARIATION ADDITIONAL REFERENCES Animals and Plants under Domestication. By Charles Darwin. 2 vols. Discontinuous Variation. (An example.) By E. R. Saunders. Pro- ceedings of the Royal Society, LXII, 11-25. Origin of Species by Means of Natural Selection. By Charles Darwin, i vol. Theory of Organic Variation. By H. S. Williams. Science, VI, 73-84. Type, How Fixed. (On genetic energy of organisms. Is variability and not permanency the normal law of organic life ?) By H. S. Williams. Science, VII, 721-729. Variation and Some Phenomena connected with Reproduction AND Sex. By Adam Sedgwick. Science, XI, 881-894, 923-930. Variation Discontinuous. (Study of a recent variety of flatfish.) By H. C. Bumpus. Science, VII, 197. Variation in Plants. (A study of the portions of a leaf on which chlo- rophyll is found.) Experiment Station Record, XIII, 423. Variation in Trillium grandiflorum. (A record of the variations observed in 185 cases.) By H. W. Britcher. Maine Experiment Station Bulletin No. 86, pp. 169-196 ; also Experiment Station Record, XIV, 634. CHAPTER II MORPHOLOGICAL VARIATION Morphological variation has reference to differences in form. If two or more individuals possess the same structural characters and if they have all attained the same relative development, then the different individuals will differ only in size ; but if the characters have not attained proportional development in the different indi- viduals, then we shall note differences in form independent of mere size. This is the simplest of all forms of variation and is the one chiefly in the mind of older biologists, even leading to the mistake of supposing that evolution is essentially a study in morphology. The cause of morphological differences may lie in extremely favorable or unfavorable conditions of life, especially as regards food and climate, affecting different characters differently, or they may arise from internal and constitutional causes, as in giants and dwarfs, or in such extreme differences as in the mul- berry leaves shown in Fig. 2. Instances of morphological variation are so common and so easily noted as to scarcely require mention. Two apples are exactly alike except that one is larger than the other. It is a clear case of morphological variation. In this instance there is no difference in the characters of the two individuals except that cell division and growth have proceeded farther in one case than in the other. Aside from this they are identical. If two stalks of corn or if a number of pigs, sheep, cows, or horses are exactly alike except as to size, then their differences are cjuantitative only, and the effect is morphological merely. Again, two horses are of the same breed, — that is, possess the same characters, — but their characters are not equally, that is, proportionately, developed. In one the leg is longer, the hock shorter, or the face wider between the eyes. These differences 25 Fig. 2. Morphological variation illustrated : different forms of mulberry leaves picked from the same tree on the same day 26 MORPHOLOGICAL VARLATION 27 are all quantitative and morphological, but they influence form rather than size as a whole, because their development is not proportional one part with another. Variation seldom simple. Instances of the above kind are, however, extremely rare. Variation is so common that other differences generally accompany those of form. The two apples may differ in color, flavor, or texture as well as in size, in which case substantive variation has also occurred. One of the horses may have an extra rib, one of the pigs a solid hoof, or one of the sheep more fibers of wool to the square inch, in which case meristic variation is present. The pulse of one of the horses may be faster than that of the other, or the milk of one cow may be richer in fat, showing functional deviation. And so it is in practice that two or more forms of variation may be and likely will be found present in the same individual. But however that may be, all differences in form or size are regarded as morphological, no matter what other differences may be found, and it is important that the student early form the habit of distinguishing clearly between the different kinds of variation present, even in the same individual. The limits of size. Every species has a general average of size to which most individuals closely approximate. A few, how- ever (giants), greatly exceed this size, and others (dwarfs) fall far short of it. All investigators agree upon the conclusion that this difference in size is due to the number and not the size of individual cells ; in other words, size is dependent upon the energy of cell division.^ This energy is exceedingly active in youth, gradually decreasing to zero at maturity, except as to certain parts (reproductive organs, skin, and sometimes the teeth and horns). In most species accident to a part will stim- ulate cell division, leading to a more or less successful regenera- tion. For the most part, however, cell division does not proceed rapidly after maturity, and the limit of its activity is in general the limit of growth. Giants therefore represent excessive cell division above the normal, and dwarfs, arrested development, — an abnormally early cessation of cell division. 1 Wilson, The Cell in Development, pp. 388-394. 28 VARIATION The causes involved in this abnormal behavior of the cell in division are exceedingly obscure. They are certainly sometimes connected with food and care in early life, and no doubt they are often constitutional. Every stockman knows about the stunted pig, calf, or colt, and that it sometimes, but rarely, recovers; that is to say, cell division once checked does not easily return to the normal rate. The trait, however, easily becomes constitutional and hereditary, for whole families (strains) become undersized and others as much above the medium. Energy of growth not to be confused with bodily or func- tional activity. While the larger animal of his kind, whether it be the individual or the strain, represents the greater energy of cell division in body building, it by no means follows that the body when built will possess a greater degree of activity than will its normal or smaller neighbor ; indeed, the opposite is likely to be true, because the larger body works at greater dis- advantage, having greater inertia to overcome and more dead weight to carry about. This is eminently true of all animals whose service involves transporting the body from place to place, as among horses. It is manifestly impossible for a heavy- horse to equal a light one in speed without the expenditure of far more power in doing it. This is not only because of the extra weight but because of mechanical disadvantages as well. In activities not involving motion this difference in size, within reasonable limits, does not exist ; small men, for example, are doubtless no more and no less intellectual than are large ones. Importance of morphological variation. Next to those of color, differences in size are the most noticeable of all variations; but they are by no means the most significant, and their importance is likely to be greatly overestimated. Except in a few instances, as with draft horses, mere size is of far less consequence than is commonly supposed. Generally speaking, it is some quality other than bulk that determines value, and it will be fortunate for breeding when the popular notion that " the biggest is the best " shall have passed away. The largest apple is not the best for eating, nor the largest bull the best for breeding purposes. However, this does not free the student and breeder from considerations of size. MORPHOLOGICAL VARIATION 29 because extremes both ways are in most cases to be avoided, and the highest excellence and the most reliable breeders will commonly be found among the individuals of medium size. Differences in form, arising from relative inequality in devel- opment of structural parts, are of more consequence than are differences in mere bulk, in which development has been pro- portional. This is especially true in horses, in which differences in relative development of structural parts may seriously affect the appearance or interfere with the action of the individual. And so it is that, while morphological differences are of far more significance to the student of general evolution than they are to the farmer, they yet constitute a phase of variation not to be overlooked by the student who is interested in the improve- ment of domesticated forms. ADDITIONAL REFERENCES Darwiniana. By Asa Gray, i vol. Darwinism. By A. R. Wallace, i vol. Expression of the Emotions in Man and Animals. By Charles Darwin, i vol. From Greeks to Darwin. By H. F. Osborn. i vol. Lamarck: His Life and Work. By A. S. Packard, i vol. CHAPTER III SUBSTANTIVE VARIATION Substantive variation has reference to differences in quality as distinct from form or size. It regards the composition or make-up of the body or its members, and refers to the constitu- tion, or inherent nature, of the organism. Everybody recognizes differences in muscles, whether firm and strong, or soft, flabby, and weak. We distinguish the bone of a horse as dense or as soft, porous, and spongy. The horn of the hoof is hard and tough or soft and " shelly." Meat is fine in grain and high in flavor or coarse in grain and lacking in quality.^ It is either juicy and rich or dry and taste- less. The gamy flavor of wild meat, both of mammals and birds, is especially tasteful to the huntsman, and whether due to breed- ing or to feed, it is certainly characteristic of wild life everywhere. No two cuts of meat are alike, whether wild or tame, and these differences are so pronounced as to be commonly recognized ; indeed, language abounds in adjectives descriptive of differences in quality of food stuffs. At one time milk was sold on the quantitative basis only, but now the per cent of fat is the basis of value. The intelligence of an animal or of a man depends less upon the size and weight of the brain than upon the quality of the brain matter and the depth of the convolutions. One apple is sweet, another is sour, and still another is insipid. One fruit is highly flavored, another is tasteless. The sugar of beets, of cane, and of maple is the same ; but the two former are simply sweet, while the latter is accompanied by a highly volatile ether that adds a peculiarly delicious aroma.^ 1 Pigs fed heavily on cotton-seed meal make a pork strongly flavored with cotton-seed oil. See Grindley, Journal of Atnerican Chemical Society-, Vol. XXVII. 2 Isolated by Kedzie, of the Michigan Agricultural College, from samples fur- nished by the author. 30 SUBSTANTIVE VARIATION 31 The sugar content of beets varies from 4 or 5 per cent to over 20 per cent. It is also exceedingly variable in cane. Wheat is richer in protein than is corn, but both are variable, and corn has been bred with a protein content higher than that of wheat .^ Plants differ in their ability to withstand frost as do animals to resist disease. A single stalk of corn may remain fresh and green when all its neighbors have been killed. Certain individuals seem immune to particular diseases, and appear to be able to resist infection indefinitely. Color in general is based upon definite chemical constituents or upon the character of the surfaces presented for refraction of light. In either case it is a matter of inherent quality and is substantive. Importance of substantive variation. The significance of sub- stantive differences depends upon the instance. Speaking gen- erally, these differences are of high value. They are usually, though not invariably, correlated with efficiency, and in such cases they possess a utilitarian interest. A dense bone is better than a soft and fibrous one. Every- body prefers a good apple to a poor one ; we have a decided pref- erence for certain aromas ; the juicy, highly flavored steak is better than the dry, tough, and tasteless one. Color is a utility character among flowers. We buy them for their color, their form, and their odor. They have no other value, and of these characters their coloring is of the most importance. Color, however, is in general the most deceptive of all charac- ters, — deceptive because it is striking and because we greatly prefer certain color effects over others, even though not correlated with utility. We carry this preference beyond reason. A red apple will sell for more than one of any other color ; yet we buy an apple not to look at but to eat ; and no one has shown a cor- relation between color and quality in fruit. Horses with white skin are proverbially subject to certain diseases. For this and other reasons color has no little signifi- cance in horses, but among cattle it has practically no meaning whatever ; and yet how decidedly do color markings figure in ^ The lowest protein content discovered in the breeding experiments at the University of Illinois up to date (1907) is 6.13 per cent, and the highest is 17.79. 32 VARIATION many score cards for pure-bred animals ! Now the facts are that a cow is kept for what she can do, and there is nothing inherent in mere color that is indicative of her ability to convert feed into either milk or meat. She is therefore neither better nor worse for her color except as it is an index of blood lines when she is to be used as a breeder. So striking are color differences, however, and so distinct are our preferences, that we instinctively follow our prejudices in this respect, quite regardless of more important considerations, until the whole fabric of breeding is interwoven with "fancy points" in the shape of color markings that greatly confuse the breeder in his attempts to select individuals for breeding purposes. To-day the majority of prize-winning shorthorns are either roan or pure white. Twenty-five years ago no breeder would have dared to show a roan animal, and if a white calf had been dropped in his herd he would have destroyed it at once and kept the matter as secret as possible, so strong was the red color craze following the American worship of Bates cattle. This craze, which was always groundless, cost the breed and their admirers dearly and checked by a decade or more its progress upward. Probably of all substantive variations color is, excepting among flowers and ornamental plants, of the least consequence to man ; yet the prejudice is with us, and the breeder who expects to sell his product must reckon with it. He should do it intelligently, however, realizing fully that individuals vary greatly in inherent quality quite independently of color. In general it may be said that substantive differences, though not so easily detected, are yet of far more significance to the farmer than are those of either form or size. CHAPTER IV MERISTIC VARIATION Meristic variation has reference to a deviation in the number or arrangement of repeated parts involved in the plan or pattern upon which any particular organism is built. A plant or an animal is not an amorphous lump of living matter. On the contrary, it is made up of parts, each of which has a kind of identity of its own, many of which arc similar, and all of which are definitely related and placed in some sort of orderly arrangement. Reflection discloses the fact that each organism is developed upon a specific plan, essentially different from that of any other, and that with most organisms the pattern is composed of a defi- nite number of similar parts more or less repeated. Thus the chicken has two legs, and the horse has four legs, that are more or less alike. The flower has many petals, the corn plant many leaves ; the spinal column is composed of vertebra very much alike, and organisms generally possess many parts that more or less closely resemble one another. From this it appears that the individual animal or plant is not a unit in respect to form, but rather that it is made up of many units, some of which are practical duplicates. Thus the idea of multiple parts in orderly arrangement {inerism) comes at once into the study, and variation in the number or character of these repeated parts (meristic variation) is a broad and compli- cated subject affording considerable insight into the nature of variation. Accordingly it is profitable to pursue it at considerable length, not so much for the material involved, which consists largely of abnormalities of no practical interest or value, but because no other phase of variation affords so much information upon the real nature of living matter and its adherence to or deviation from a definite plan. 3J 34 VARIATION SECTION I — SYMMETRY The central thought in all mcristic studies is symmetry, by which is meant that opposite sides of an organism possess parts that are either identical or at least similar. Thus the petals upon one side of a flower are in most cases like those upon the oppo- site side, and the blossom is made up of a number of similar parts very much alike and several times repeated.^ Among higher animals, however, opposite sides are similar but not identical, and here arises the distinction between radial sym- metry and bilateral symmetry. Radial symmetry and radial series. By this is meant that kind of pattern in which the separate parts are identical and each part is capable of replacing any other in the series. Common examples are the petals of most flowers (leguminous and the like excepted), the leaves and lateral shoots of plants, the capsules of many seeds, such as the apple, orange, etc., the rows of corn upon the cob, the parts of the sea urchin, the starfish, and the Jladiolaria generally. In all cases of this sort the individual parts could each replace any other part of the series. The pattern is therefore spoken of as one of radial symmetry, and the parts as members of a radial series. Bilateral symmetry. Among higher animals a different sym- metry is observable. While each side has its counterpart upon the other, yet there is a distinction between the right and the left sides and between the dorsal and the ventral surfaces. In such cases the parts while similar are not alike, nor could they replace each other. 1 Symmetry is 7vcll-)iigh universal. All organisms arise through cell division in one or more planes, and some degree of symmetry is to be expected from the manner in which growth takes place. But symmetry is not confined to multicellular structures. Appendages consist- ing of single cells are frequently symmetrically placed, and many organisms, which are single celled and therefore microscopic, as diatoms, secrete a skeleton with regular markings as symmetrical as hoarfrost and quite as beautiful. All this is curious rather than valuable to the student of thremmatology who is interested in multicellular beings ; yet it all throws light on the method of life, and we are able to lay down the principle that symmeti^ is not only the natural corollary of development by cell division, but that it is also a general principle in living matter. MERISTIC VARIATION 35 The right hand and arm are made upon the same plan as the left, but could not replace them because they would not fit ; the one is the reverse of the other. Reflected in the mirror the right hand seems to be the left, but it is an illusion, for the right side is the negative or optical image of the left with all its elements reversed. Hence a part upon the one side could not replace the corresponding part upon the other. It is its counterpart, not its duplicate as among leaves and petals. Thus we arrive at the distinction between the com- plexity of bilateral symmetry and the simplicity of radial symmetry. It is also significant that bilateral symmetry is characteristic of higher animal life, and radial symmetry of lower animals and plants. Dorsal and ventral surfaces. The fundamental fact at the bottom of bilateral symmetry is the distinction between dorsal and ventral surfaces, necessitating differences in the quadrants that are forced to work in opposition to each other.^ Indian corn and the grass family generally are as distinctly bilateral as is the horse or man, yet there is no thought of dorsal and ventral differences, and hence no distinction between right and left. For example, let the leaves of a corn plant and the arms of a man extend east and west. Then the north and south sides of the corn plant will be alike. Not so with the man. In the one direction (we will say the south) will be his spinal column and the general framework of the body ; in the other (to the north) will be his face, his nose, his eyes, and all the active external parts. Moreover his hands are made to oppose each other and to work together with this (the ventral) side of the body. There is therefore bilateral symmetry in one direction but not in the other. With the corn plant the case is different. It does not move from place to place, and it presents its plain sides indifferently to the world. Accordingly no distinctions similar to dorsal and ventral are possible.^ 1 The word "opposition " is here used not in the sense of " antagonism " but as "placed opposite and working with"; as, The thumb is "opposed" to the other members of the hand, thereby making a working unit. 2 None of these distinctions should be confused with homologous parts or with analogous parts, nor should the ideas be confounded. Homologous parts belong to two individuals, not one, and they are such as bear corresponding structural relations to their respective organisms, suggesting 36 VARIATION Bilateral symmetry not complete. Curiously enough not all the parts follow the same plan as to bilateralism and symmetry. What has been said refers to paired organs standing on opposite sides of the body, as hands, arms, legs, eyes, ears, etc. Many organs not paired present curious facts to the evolu- tionist. The nose, for example, has no counterpart, but it stands on the median line and has a bilateral symmetry of its own, being made up of right and left halves. The liver and the heart, however, while consisting of right and left halves, are unpaired organs, placed not on the median line, but the one upon the right, the other upon the left. Each has a bilateral symmetry of its own, with distinct right and left sides, yet both are unsymmetric- ally placed. The stomach, on the other hand, is an unpaired organ lying unsymmetrically across the body, and its own bilateralism is not between right and left hut from front to back. The kidneys pre- sent the anomalous phenomena of paired organs with a bilateral- ism of their own but at right angles to that of the body, being also from front to back. Longitudinal symmetry and linear series. Inasmuch as all growth is by cell division, we might expect longitudinal symmetry as well as lateral symmetry. Owing, however, to the definite relations of both animals and plants to the external world, it is not much developed, and there is but a suggestion of longitudinal symmetry to be found in either plant or animal forms. Most plants are both geotropic and heliotropic ; that is, one part goes down into the earth in response to gravity and the other upward toward the light and against gravity. This makes common descent. Thus the leg of a man is homologous with that of a horse or a bird, because of structural resemblances. In the same sense his arm is homolo- gous with the fore leg of the horse or the wing of a bird. Analogous parts are such as serve the same purpose in different organisms though structurally distinct. Thus the flipper of the whale, which is a modified hand, is analogous to the fin of a fish, and the gill of a fish is analogous to the lung of a mammal, because it serves the same purpose, though there is no struc- tural relation between the two. The homologue or the analogue of a part is therefore to be found in another individual and of a different species. Symmetry, on the contrary, with its corollary of multiple parts, refers to individuals taken singly and to the interrelations of their parts. MERISTIC VARIATION 37 the two extremities at once very different, and forestalls the development of any very pronounced symmetry longitudinally. Animals in their locomotion establish different relations at their opposite extremities, thus preventing exact symmetry in this direc- tion, and yet reminders of inherent tendencies toward universal symmetry are constantly encountered. Long worms, for example, though distinctly different at the extremities, are yet compo.sed of rings very much alike throughout most of their length, even permitting locomotion backward with considerable facility. Longitudinal division, however, with or without corresponding symmetry, is everywhere found both in plant and animal life, espe- cially in the latter, and linear series of similar parts present as many opportunities for variation as are afforded by radial series either with or without bilateral symmetry. Thus the rings of worms, the vertebras and the ribs of the body, the joints of the fingers and of insect parts, — all these are fertile sources of meristic variation. Homoeosis in meristic variation. This is a form of variation in which one part assumes the characters or appearance of another, usually quite distinct. It is a frequent accompaniment of meristic variations. For example, an extra vertebra may be found in the dorsal series, increasing the number by one, — all normal. This is meristic variation of the simplest kind, with no homoeosis. On the other hand, it may be situated at the front of the dor- sal series and partake somewhat of the character of a cervical (forward homoeosis), or it may be located at the rear of the dor- sal series and in many respects resemble a lumbar (backward homoeosis). This posterior dorsal vertebra may bear a rib that is bifurcated at the extremity, one branch effecting a union with the sacrum, the other floating, in which case there is doubt as to the real character of the additional vertebra, whether dorsal or lumbar. In much the same way misplaced organs are often found. A leaf may be seen growing from a fruit, an antenna may spring from an injured eye, or foot appendages may develop instead of those proper to the extremity of the antenna. All cases of this order, in which one organ through some dis- turbance assumes the character of another organ, are known as 38 VARIATION homoeotic variations, and homoeosis of some sort is a frequent accompaniment of meristic variation in longitudinal series. The best instance of this is found in the petals of flowers which are recognized by botanists as modified leaves, and instances are not rare in which the specimen plainly shows the various transition states from leaf to sepal, sepal to petal, and petal to stamen. This tendency of one part to assume the character and dis- charge the functions of a neighboring part is an important phase of variation, throwing much light upon the general subject of development. With this introduction the student is prepared for the study of meristic variation somewhat in detail, in which he will find that while each species is built upon its own somewhat peculiar and definite pattern, yet this pattern is subject to many and profound alterations, and the organism is frequently able to exist upon an- other and much-distorted plan, all of which goes far toward enlight- ening the student as to the variations that may be expected in the organic world. Studies in meristic variation are useful to the stu- dent of thremmatology, not so much for their own sake as for the light they shed upon the nature and manner of variation. Examples of meristic variation. Examples of meristic varia- tion are to be found at every hand. In the doubling of flowers and the stooling of grain, in increased or reduced numbers of fingers and toes, in the four-leaved clover and the branching habit of many plants, — everywhere are seen alterations in the customary plan on which nature does its work. Fortunately an extended and valuable collection of meristic variations, mostly among animals, has been made by Bateson.^ He lists his data under 886 headings, each recording from one to several authentic cases. Equally complete data covering plants have not been collected, though it is among plants that meristic variation is most common. Indeed, it is so common and so evident that formal collection is hardly necessary. The student is therefore referred to plant life out of doors and to Bateson's collection for a fuller study of this important subject, a bare outline of which, as a guide, being all that is attempted here. 1 Bateson, Materials for the Study of Variation [Macmillan & Co., 1S94]. MERISTIC VARIATION 39 SECTION II — MERISTIC VARIATION IN LINEAR SERIES Vertebrae. Among fishes and snakes variation in the number of vertebrae may be very great. In mammals it is smaller but yet distinct, as in the following examples, instances of which, according to Bateson, could be multiplied indefinitely. Erinaceous Europ.'eus (The Hedgehog) ^ No. Cervical Dorsal Lumbar Sacral Coccygeal Total 1 7 14 6 4 II 42 2 7 '5 6 3 10 + 3 7 i6 6 3 9 + 4 7 15 6 4 12 44 5 7 15 6 4 II 43 6 7 14 6 3 9 + 7 7 15 6 3 II 0112 8 7 15 6 3 13 44 9 7 15 6 3 12 or 13 It will be noticed that i and 5 differ only in the dorsal region, which fact, however, affects the total, but that 4 and 8 differ both in the sacral and the coccygeal without affecting the total. Commenting on the phenomena of an additional lumbar or sacral vertebra, Bateson says : . . . There is a strong suggestion that (in cases of this kind) the num- ber of vertebras has been increa.sed by simple addition of a new segment behind, after the fashion of a growing worm ; the variation of vertebrae thus .seems a simple thing. But there is evidence of other kinds, which plainly shows this view of the matter to be quite inadequate. What this evidence is he proceeds to show by succeeding examples, a few of which are reproduced here : In a skeleton of Python tigris'^ (No. 602, Museum of the College of Surgeons) the vertebrae are normal up to the 147th inclusive. The 148th and 149th are, however, abnormally short from front to back, suggesting arrested development with imperfect separa- tion, although each vertebra bears a normal rib on either side. 1 Bateson, Materials, etc., p. 103. 2 ibid. pp. 103-105. 40 VARIATION Passing backward in the same specimen, the i66th vertebra is seen to be normal on the left but double and bearing two ribs on the right, thus greatly crowding the ribs on that side. The 185th vertebra is reported in the same condition, both being doubled on the right side and single on the left (see Fig. 3). Following, Bateson ^ gives two examples of the reverse condi- tion, namely with duplicity on the left side, and another with duplicity on the right, showing clearly that meristic variation in one side of a bilateral symmetry may or may not involve the other side. Ribs. Variation in the dorsal region necessarily involves the ribs. Aside from this, all evi- dence goes to show that j^artial division of the ribs is much more common than is variation in the number of vertebrae. In man, for example, the cartilage is fre- quently divided for a considerable distance (1.5 in.) back from the sternum, often involving a real bifurcation of the rib itself. Homoeotic variation in vertebrae and ribs.^ These may be out- lined as follows (all in man, except as noted) : I. Cervical resembling dorsal : backxvard homceosis. The dis- tinguishing character of dorsal vertebrae is the bearing of ribs, but this character is often assumed by neighboring cervical, being common on the seventh and not unknown on the sixth. Of fifty-seven cases examined by Struthers, forty-two showed ribs on both sides and fifteen on one side only, showing a tend- ency to preservation of symmetry. The completeness of develop- ment ranges all the way from the merest rudiments (rare) to a perfect rib connected by cartilage to the sternum (also rare), the commonest form ending free or being joined by cartilage to the first true rib. Fig. 3. Meristic variation in vertebrae : double on right side. — After Bateson 1 Bateson, Materials, etc., p. 105. 2 Ibid. pp. 106-128. MERISTIC VARIATION 4 1 2. Dorsal rcscnibling cervical : forzvard Jiovio^osis. Not so common as above, but Struthers^ describes a specimen in which the first pair of ribs is defective. On the left side the rib extends but two fifths of the way around, where it articulates with a process on the second rib. On the right side it joins the second rib about one inch beyond the tubercle. As seven nor- mal cervical vertebras are present in this specimen, it is to be regarded as a modified dorsal rather than an extra cervical assum- ing the characters of the dorsal, as in the preceding cases. 3. Dorsal resembling lumbar. Frequently the twelfth rib in man is rudimentary, in which case the last dorsal vertebra assumes the form and general appearance of a lumbar. 4. Lumbar resembling dorsal. Cases of a thirteenth rib are not unknown but are more rare than the reduction of the twelfth. 5. Homceosis betiueen lumbar, sacral, and coccygeal. The last lumbar may unite on one or both sides with the sacral, in which case the lumbar develops processes to assist in the support of the ilium. On the contrary, the first sacral may remain detached, thus becoming practically a lumbar. Similar relations obtain be- tween the sacral and the coccygeal. A careful study of this whole subject develops the following facts : 1. That an increase in the number of parts in one region may or may not affect the total number in the series. 2. That consequently a change in number in one region may or may not be accompanied by changes in other regions of the same series ; that is, changes in the dorsal do not imply changes in either the cervical or the lumbar. 3. That homceosis in vertebrae and ribs is confined to members contiguous ; that is, if a cervical resemble a dorsal, it will be that cervical lying next to the dorsal series. 4. That the tendency is for an extra member to resemble somewhat the members of the next region ; that is, an extra dorsal is likely to resemble a lumbar or a cervical, if not to entirely replace it, suggesting that it arose at the end, and not in the middle, of the dorsal series. 1 Bate.son, Materials, etc., p. 109. 42 VARIATION 5. That forward homcEosis in one region is not necessarily- attended by forward homoeosis in other regions of the same series. 6. That in general (especially in man, where this has been most studied) forward homoeosis is attended by a total increase of the series, and backward homoeosis by a decrease. 7. That an abnormality on one side may or may not be attended by a like abnormality on the other, though the tend- ency is strongly to the preservation of bilateral symmetry. 8. That when one part resembles another it is the member lying contiguous ; that is, a dorsal vertebra will resemble a cer- vical or a lumbar, not a sacral, and lying between the stamen and the leaf are the petal and sepal and all intermediate grada- tions, either present or obliterated. Meristic variation in spinal nerves. Branches from the spinal cord emerge between the vertebrae, so that in general the sys- tem of spinal nerves is determined by the vertebras. Aside from this, however, the emergence of the branches varies greatly both in number and in conformation, even when the vertebrae are normal. Fiirbringer's ^ studies in birds show that the minimum num- ber of spinal nerves forming the brachial plexus (supplying the wings) is three, but in some species it rises as high as six. Moreover, in some instances the number varied from four to five zvithin a single species^ and in one (the pigeon) the varia- tion was from four to six. As might be suspected, the two sides are often differently supplied. For example, in one specimen (goose) " the plexus was formed on the right side by nerves xvi, XVII, xviii, and xix, while on the left side it received a strand from the xxth nerve in addition to these." Fiirbringer's tables show that in some specimens of the goose the wings were supplied by the nerves xv to xix, while in others they were supplied by the xviith to the xxth. In the dove the brachial plexus was formed by the xth branches of the spinal nerve in some specimens, by the xiith to the xvth in others, and in one case by the xith to the xivth as an intermediary. 1 Bateson, Materials, etc., pp. 129-135. MERISTIC VARIATION 43 Herringham ^ dissected to their origin the nerves forming the brachial plexus in fifty-five human subjects (thirty-two fetal and twenty-three adult). Quoting from his work, Bateson says : The origin of the ulnar nerve was traced in thirty-two cases, fourteen being adults. It (the ulnar nerve) was found to arise in four different ways. Most commonly it arises from the vnith and ixth ; this occurred in twenty- three cases. With the vnith and ixth is sometimes combined a strand from the viith, as shown in five cases (four fetal, one adult). In three fetal cases it arose from the vnith only, and in one fetal and one adult case from the \iith and vnith. ... In several cases the branch from the vnith was much larger than that from the ixth, but the reverse was never met with. Similar conditions were found elsewhere with man, the gorilla, baboon, and chimpanzee, and the following principle was set forth : " Auj/ givett fiber may alter its position relative to the vertebral eolnmn, but xvill maintaiji its position relative to otJier fibers'' Homoeosis in insects and other small animals. The replacement of one part by another, while common among plants (modified leaves and stems), is compar- atively rare in animal life. It is, however, by no means unknown, and some striking examples are quoted from Bateson to show the remarkable manner in which a perfect part may arise in a most unusual Fig. 4. Homceotic variation 1 1-1 .1 r 11 • 2 in sawfly: right antenna place, among which are the following -.^ , , r . ^ ' ° ° normal; left antenna 1. Specimens of sawfly (Cimbex axillaris^ in ^^^""S ^ ^°°*- ^ ^"^^ ^' ,.,.,, r, . J J • ,, 11 r 1 enlarged. — After Bateson which the left antenna ended m "a well-formed * foot, having a pair of nomial claws and the plaiiiiila between them " (Fig. 4). Right antenna normal.'' 2. A male bumblebee {Bonibiis vaiiabilis) taken in Munich showed the left antenna " partially developed as a foot," bearing " a pair of regularly formed claws like the claws of the foot." 3. A male specimen oi Zygceiia Jilipejidulce "pos.sessing a supernumerary wing arising in such a position as to suggest that it replaced a leg" (Fig. 5). The extra wing was on the left side and projected from the underside of the body after the exact fashion of the leg, which was absent. The specimen ^ Bateson, Materials, etc., pp. 135-13S. - Ibid. p. 147. ^ Ibid. pp. 146-155. Professor Bateson vouches for the genuineness of this specimen, which he himself carefully examined, although it belonged to Dr. Kraatz. 44 VARIATION belongs to Mr. Richardson, and was examined by Professor Bateson as closely as was possible without removing- the hairs, to which the owner objected. It is well known that supernumerary wings may arise with the normal number of legs. In this case the closest examination failed to reveal even a rudimentary leg, and there was cer- tainly " no empty socket or other suggestion that the rest of the leg had been lost."' 4. Specimen of Palinurus penicillatus with an -^ antenna-like flagellum growing up from the surface of the (left) eye." 5. The female crayfish has normally a pair of oviducal openings on the bases of the ante- penultimate pair of walking legs. This .speci- men possessed in addition a pair also on the Fig. ;. Supernumerary wine on u- ^ 1 • ■ i ^ ^• , ^ ..^ ,^ ,/ *' ijenultmiate, showmg irregular segmentation, leftside. — After Bateson ^ . , . , . , 6. Another specimen, also of the crayfish, possesses an extra pair of oviducal openings, as in the last, except that they were placed on the last pair of legs, skipping the penultimate. It is note- worthy that this is the normal position for the sexual organs of the male, except that the openings were placed in their own proper position on the leg and not " at the posterior surface of the joint as the male openings are." 7. Bateson himself examined 5S6 female crayfish for abnormalities of oviducal openings. Of this number he found 563 were normal and 23 abnormal, as follows : 1. Extra oviducal opening on left penuUimate leg 7 2. Extra oviducal opening on right penultimate leg .... 10 3. Extra oviducal opening on both penultimate legs .... i 4. Exti^a oviducal opening on both penultimate and last legs . i 5. Single oviducal opening on left side only 3 6. Single oviducal opening on right side only i Total abnormal specimens 23 Bateson reports but one abnormal .specimen out of 714 males examined by him, and this abnormality consisted in the ab.sence of a generative open- ing on the right side. 8. Among earthworms will be found many cases of imperfect segmenta- tion, showing more rings on one side than upon the other, often suggesting a spiral rather than a series of rings. Great irregularity is also found in the position of generative openings, as to whether paired or single, ^ although the male parts are always posterior to the female, whatever the number of the ring on which either is borne. Cervical fistulae and auricular appendages in mammals.^ Cervical fistulae are openings in the neck, occurring singly or in pairs and located anywhere from the median line backward as I Bateson, Materials, etc., pp. 156-166. - Ibid. p]). 1 74-1 So. MERISTIC VARIATION 45 far as the angle of the jaw. The opening is sometimes sHght, but often it extends completely to the pharynx. In the latter case it is possible to pass an instrument the size of a small quill, provided the opening is comparatively straight, otherwise its completeness or incompleteness may be ascertained by the injection of a licjuid. Bateson cjuotes Fisher^ as describing sixty-five persons with seventy-nine fistulae. Fourteen of these were bilateral (occurring Fig. 6. Child with supernumerary auricle on each side of the neck. Bateson, from Birkett After on both sides), and fifty-one were unilateral, of which thirty- three were on the right side. He adds, " There was evidence of heredity in twenty-one cases." Auricular appendages, often called supernumerary auricles, are not at all uncommon. They are non-functional growths occurring in the neighborhood of the ear but below it, and are generally accompanied by some deformity of that organ. They consist of little flaps of skin or, more commonly, of cartilaginous growths identical in texture with that of the normal external ear. ^ Bateson, Materials, etc., p. 175. Obvious errors in figures prevent further quotations that would otherwise be of interest. 46 VARIATION One of the most remarkable cases ever described is that of an infant brought to Guy's hospital in 185 1.^ Another was of a child having a well-developed supernumerary auricle on each side of the neck (see Fig. 6). These appendages were easily removed and proved to be entirely cutaneous, though each was served by a small artery. Whether cervical fistulse are to be regarded as remains of unclosed gill slits, or whether they are to be regarded as repeti- tions of the external ear, in any event their presence shows a pronounced tendency to repeat certain characteristic structures in this particular region of the body. Growths of this character are by no means confined to man. Cervical auricles (the so-called " wattles ") are common in sheep, especially merinos. They are well known in goats and are ex- ceedingly common in many strains of unimproved swine. Strange as it may seem, these repetitions of the ear appendages are un- known in either the horse or the ox. Meristic repetition in mammae.- One of the chief distinguish- ing features of mammals is milk secretion. Speaking generally, this occurs at some point or points on either side of the ventral surface of the body on lines running from the armpit to the groin. In swine and in dogs it is distributed throughout the entire extent of these mammary lines. In cattle, horses, goats, sheep, etc., it is confined to the rear extremity of the line, and in the elephant it is as decidedly forward, the udder being located at the armpit. In the human being the point of normal activity is relatively further back (down) than in the elephant, but yet above the middle. This latter point is established by the fact that supernumerary nipples are found both above and below the normal. The fact that no less than three supernumeraries have been found above indicates that the normal mamm?e are perhaps fourth in a full series. It is to be noted in this connection, however, that in most cases supernumeraries are situated below rather than above the normal. These structures vary all the way from mere nipples resembling warts and entirely unaccompanied by mammary tissue 1 Bateson, Materials, etc., p. 178. ^ Ibid. pp. 1S1-195. MERISTIC VARIATION 47 up to well-formed organs fully functional. Curiously enough super- numerary mammae are more common in men than in women. Bateson ^ quotes Bruce as having found in 2 3 1 1 females fourteen cases (0.605 percent), and in 1645 males forty-seven cases (2.857 per cent). In another series 315 subjects were examined, show- ing twenty-four cases {^ .6 per cent), nineteen being male and five female. Bardeleben is also quoted as having examined 2736 recruits (all males, of course). In this series "637 cases (23.3 per cent) were seen, 219 being on the right side, 248 on the left, and 170 on both sides." The largest number of supernumerary mammas ever recorded was in a subject described by Neugebauer.^ This patient had five pairs of nipples, of which the fourth, numbered from above, was the normal. When the child was being suckled milk oozed from each of the uppermost pair, but all other supernumeraries yielded milk only with pressure. Extra teats in cows are too common to need mention except to call attention to their excessive number. The cow Rose, famous for her record at the Illinois Station,^ had in all eight mammae, six of which were fairly well developed, though only four were functional."^ It is noticeable that supernumeraries are nearly always posterior to the normal or else constitute a doubling of one of the normals. Every milker knows by sad experience that these supernumeraries are not only common but frequently functional. A close study of this subject shows that repetition of these parts may be by pairs or singly ; that the repeated parts may be on the same or on different levels ; that they may be out of line, being in some cases very near the median, and that the normal nipple may be doubled. From the latter fact we further establish the point that meristic variation may occur in two ways, — either by addition to the series or by division of a normal number. We shall find the same in teeth. 1 Bateson, Materials, etc., pp. 182-183. 2 Ibid. p. 183. ^ See Bulletin A^o. 66. * These supernumeraries were not symmetrically placed. On the right side the two extra teats were placed behind the two functional, as is commonly the case ; but on the left side only one supernumerary was so placed, while the other was between the two functional teats. 48 VARIATION Meristic variation in teeth. As Bateson remarks, "Teeth arise by special diilcrcntiation at points alon^ the jaw, as mammas arise by special differentiation at points along the mammary line," and we shall see that with teeth as with mamm?e these points of special differentiation may frequently lie outside the normal region, that they are subject to increase or decrease in number, and that the increase may be due either to the addition of a mem- ber to the series, to the interpolation of a member, or to the division of a normal member. Before considering special cases it is well to note that the similarity between the right and left jaws is that of ordinary bilateral symmetry, but that there is also a kind of symmetry, not very close but still marked, between the dentition of the upper and that of the lower jaw. It should be further noted that in many animals, as in the shark, alligator, etc., the denti- tion constitutes a series in which the separate teeth differ from one another mainly in size. But mammals for the most part are heterodont ; that is, the series is broken up into groups which differ among themselves, though the members of the separate groups resemble one another closely. Thus the incisors are quite different from the canines, which in turn differ from the premolars and the molars. The different incisors, however, are very much alike, and the same is true of the canines and the various molars and premolars. Meristic variation in a heteroge- nous series like this is manifestly much more complicated than in a simple series like the mammae or the ribs. With this intro- duction attention will be called to a few special examples quoted from the 237 cases that have been collected by Bateson. ^ 1. One hundred and fifty-two adult skulls of anthropoid apes showed twelve cases of extra teeth. One was an incisor, one was anomalous, and the others were molars. This is nearly 8 per cent abnormal, as against 425 Old World monkeys that showed but two cases of extra teeth, — less than one half of i per cent. 2. Adult orang, with an additional posterior molar on both sides above and on the left side below. No trace of extra molar on right side was dis- covered, " though there is almost as much room for it as on the left side." Extra molars perfect but slightly smaller than the normal. ^ Bateson, Materials, etc., pp. 105-273. MERISTIC VARIATION 49 3. Skull (orang) No. 2043^?, Oxford Museum, is normal except as to the second premolar in tlie upjjer jaw {/i'-). Both these teeth are missing from their proper place. There is plenty of space on the left side but somewhat less than the normal on the right side. The missing tooth of the right side is present in the skull, but instead of being in its proper place it stands up from the roof of the mouth within the arcade immediately in front of the right canine and almost exactly on the level of the second incisor, being in the premaxilla at some distance in front of the maxillary suture. Discussing this case, Bateson observes : That this tooth is actually the second premolar which has by some means been shifted into this position there can be no doubt whatever. It has the exact form of the second premolar and is of full size. It stands nearly verti- cally, but is a little inclined towards the outside. The canine is, by the growth of this tooth, slightly separated from the second incisor, and the first premolar is consequently pushed also somewhat further back. Hence it happens that the diastema space for the second premolar on the right side is not of full size. This should be understood, as it might otherwise be imagined that the contraction was due to a complementary increase in the size of the other teeth, of which there is no evidence. The missing premolar on the left side was not visible, but " on the left side of the palate there was a very slight elevation at a point homologous and symmetrical with that at which the second premolar on the right side was placed. ... A small piece of bone was here cut, away, with the result that a tooth of about the same size and formation as /^ was found imbedded in the bone." In this case, therefore, the upper premolars on both sides had " traveled away from their proper positions and taken up new and symmetrical positions in the palate, anterior to the canines." As Bateson pertinently remarks (italics and parenthesis mine), "The facts of this case go to show that the gc7'in of a tootJi con- tains ivithin itself all the clcuicnts necessary to its dei'clopnient in its oivn true form [even in an abnormal position], provided of course that nutrition is unrestricted." This is a significant point of peculiar interest to students of thremmatology, not because of its bearing upon dentition but because of the light it affords upon the basis of variability and the ultimate units of variation. 4. Gorilla from the Congo, with a fifth incisor standing almost in the middle of the lower jaw. It has the characteristic chisel shape of the incisor, but it is " turned half round so that the plane of its chisel stands obliquely." 50 VARIATION 5. Dog with lower jaw and teeth normal, but with upper canines imper- fectly divided. The division was more complete on the right side, forming practically two canines standing in line with the regular teeth. 6. Dog with first premolar in right side of upper jaw doubled, both teeth being normal in shape, the anterior somewhat the larger. Fig. 7. Merism in teeth : canines partially divided. — After Bateson 7. Dog with an extra premolar on both sides above and below, the denti- tion formula beinir t> ■ 8. Sledge dog : " All teeth normal, e.xcept left upper / -. This tooth nor- mally has two roots. Here it is represented by two teeth, each having one root." 9. Absence of first premolar frecjuently quite common in Eskimo dogs, suggesting a breed peculiarity. 10. Among domestic dogs supernumerary molars were found in twenty- eight cases out of 345 skulls examined, as follows,^ the normal dentition of the dog as to molars being two above and three below (/// 2). m^ on both sides and «■* on one side i case n^ on both sides 2 cases m^ on one side 9 cases n^ and w7* on one side only 2 cases m'^ on both sides 6 cases ;«* on one side only 8 cases This Strongly suggests the formula m |, which is that of Otocyon {Lalandes clog) and of the fossil Amphicyon, the sup- posed doglike progenitor of the bears. It calls to mind the further remarkable facts that Otocyon itself varies from m | to m I and that it is the only mammal outside the marsupials ^ Bateson, Materials, etc., p. 220. MERISTIC VARIATION 5 1 that ever has four molars on both jaws/ which goes far to indi- cate a marsupial ancestry. Remembering that the teeth are considered as one of the few most reliable bases for classification, the remarkable variation in their number, character, and position throws no little light on the manner in which variation behaves, which is the chief reason for their extended notice here. Supernumerary eyes. The development of extra eyes seems to be confined to insects, which afford a number of excellent examples of the development of normal tissue in abnormal situations. Bateson's Nos. 419 to 421 are all cases of the development of a third eye in Coleoptera. In every case these extra eyes are c}uite distinct from the normal. In No. 419 the supernumerary was small and lay abutting against but distinct from the right eye. Its color was brownish yellow, while the normal eye was black. In Nt). 420 the extra eye was on the left side but quite independent of the normal eyes, which were exactly alike. In No. 42 1 the extra eye was on the left side of the head, which was rather less developed than the right. This eye is borne upon an irregular chitinous loop, having a diameter of about 2.5 mm. This loop is attached to the substance of the head before and behind, and these two attachments are distant from each other about i mm. The diameter of the eye is about 2.5 mm., thus occupying the full surface of the loop, and its faceting is said to be "not cjuite regular, and finer and slighter than that of the normal eye." It is thus a very good attempt at a functional third eye. Supernumerary wings in insects. Bateson ^ reports and de- scribes fifteen cases of extra wings among insects, — sometimes on one side, sometimes on the other, but generally if not always smaller than the normal ; sometimes plainly identified with the fore wing, more frequently with the posterior ; occasionally nor- mal in coloring and scaling, but as a rule abnormal. In one in- stance it took the form of a large upright scale and in another of a winglike appendage to the left anterior wing. 1 Lydekker, Library of Natural History, p. 580. 2 Bateson, Materials, etc., pp. 281-285. 52 VARIATION Meristic variation in horns. ^ These appendages afford good material for studies in variation. They sometimes consist of horny matter (cattle, sheep, rhinoceros, etc.) and sometimes of true bone, as in the deer. They sometimes persist through life (cattle, sheep, goats, etc.) and sometimes are periodically shed, as with the antlers of the stag. They sometimes, as in cattle, have a bony case, which is a true outgrowth of the skull, but often, as in the rhinoceros, they have no connection whatever with the bone beneath. Again, the antlers, which are bony, sep- arate with a clean scar from the bone of the skull, as a leaf stem parts from its twig. The meristic variations of horns are no less remarkable than their substantive variations just mentioned. They are for the most part symmetrically placed in pairs on either side of the skull just above the eyes, though the horn of the rhinoceros is borne upon the nose and therefore upon the median line. Variation in number occurs cither symmetrically or asymmet- rically. If the rhinoceros has an e.xtra horn it will be just above and on the median line with the normal. Sheep may have an extra pair just external to (behind) the normal,^ or there may be three on one side and two on the other. In the latter case the third horn will be a little one lying between the normal and the more ordinary extra horn. In still other cases, according to Bateson, a double core will be found incased in a kind of "double-barreled " single horn. Among cattle no increase in the normal number of horns is known to the writer, but their entire absence is common. Indeed, the readiness with which the polled character appears is astonishing,^ particularly as it is associated with a peculiar prom- inence (the poll) lying between and often slightly below the normal base of the horns. In cattle, meristic variation in horns seems to be associated neither with divided horns or extra prongs. 1 Bateson, Materials, etc., pp. 285-2S7. 2 Four-horned breeds are not unknown. Bateson, Materials, etc., p. 2S5. 3 It is a well-known fact that if either parent be polled the horns are almost certain to be absent in the offspring, and Storer, in his Wild White Cattle of Great Britain, says there is evidence that these park cattle have been several times alternately polled and horned since their inclosure in the parks. MERISTIC VARIATION 53 Besides sheep Bateson gives three specific cases of increase in the number of horns, as follows: (i) a family of goats in which the four-horned character was hereditary for " many genera- tions " ; (2) chamois with two " well-formed and sym- metrical extra horns " ; (3) roebuck, two specimens of which are figured. Of these one has two horns on one Fig. 8. Abnormal horns in roebuck : all but one have undergone meristic variation. — After Bateson side and three on the opposite side, while the other has three on one side, the other being normal, consisting of a single horn with one prong near the summit (see Fig. 8). Meristic variation in digits.^ Variations in these parts are peculiarly complex. There may be an increase or a decrease not only in the total number but also in the parts or joints that compose the several members. The best example covering both these points in the same individual is Bateson's No. 48 5. ^ In this case the right hand 1 Bateson, Materials, etc., pp. 311-410. ^ Ibid. p. 327. 54 VARIATION has the usual number of digits, but the thumb has three instead of two phalanges, though its general shape is normal. In the left hand there is much confusion in the region of the thumb. There is an extra digit, but its true character is not so evident. It is sharp, like a finger, but functions as a thumb. Internal to this is a thumblike supernumerary with a true nail, but from its position it is functionless (see Fig. 9). Fig. 9. Meristic variation in the hand : right and left hands of the same indi- vidual, showing on the left hand a duplication of the forefinger at the expense of the thumb, and on the right hand an extra joint in the thumb. — After Bateson A fifth real finger, making six digits in all, is not uncom- mon. Its true position, however, is by no means always easy to determine. Speaking generally, extra digits may arise in three ways, — cither by addition to the series of an outside member next the thumb or the little finger, by the insertion of a member at some point within the series, or by the doubling of a member. Just which has taken place in any given case is not always easy to determine. Reduction in the number of digits is common and takes place in three ways, — by the loss of an outside member (generally MERISTIC VARIATION 55 the thumb, when the radius is absent), by the suppression of a member within the series (ectrodactyhsm), or by the union of two or more members (syndactylism). Syndactyhsm may occur in all degrees, from mere webbing to a real bony union, as in the case of solid-hoofed hogs. Fig. lo A B Fig. io. Degrees of syndactylism in digits: the general shape of the member is preserved even when one digit is suppressed. — After Bateson exhibits a case, Z>, in which the normal shape is preserved in the absence of a member (ectrodactylism), with nothing to sug- gest a union ; that is to say, digits iii and iv seem to be fully represented by a single digit, normal in character but replacing two members of the usual series. 56 VARIATION To fully appreciate the significance of this subject we need to remind ourselves of evolutionary history with respect to digits. Man, for example, has normally five digits in all extremities. The same is true of the bat. The ox has only two toes and the horse but one, yet there are rudiments of others in both cases, strongly suggesting that at some remote period the number might have been greater. All things considered, it looks as if, for some unexplained and at present unexplainable reason, animal life in most of its higher forms had been originally constructed upon a plan of five as regards the extremities. True, many, if not most species, have long since departed more or less widely from the original plan, and yet the numeral five is as distinctly characteristic of the digits in animals as it is of petals in the rose family among plants. How this number has been gradually reduced to a final form, sometimes of two, as in cattle, sometimes of one, as in horses, is a chapter in development that belongs to the ancient history of evolution. Moreover it is a chapter that, for obvious reasons, must be read backwards and reconstructed from its fragments. It will assist in this reconstruction, and, what is of more consequence to the breeder, it will throw much light upon the manner of development and the unit of variability, if the stu- dent will consider the present condition and evident ancestry of a few characteristic species with respect to digits. Bats have five digits in both wing and leg, though the thumb is modified into a strong claw. Birds have three digits in the wing, namely, i, the thumb, which makes the so-called bastard wing; ii and iii, which make the true wing ; iv and v, missing. Radius and ulna are both present. In the leg the fibula is a mere splint, lying by the tibia. There is but one metatarsus, but it is large and heavy, ending in three pulley-like surfaces, over which play the tendons that are attached to the three toes directed forward (ii, iii, and iv). This plan suggests that the three middle metatarsals of the normal foot have here become united along the shank into one, but with three surfaces preserved for attachment of digits. Most birds have also a toe behind. This is regarded as digit i, but MERISTIC VARIATION 57 no bird has shown a trace of v. Added together, this all means that birds have lost digit v from the leg, if they ever possessed it, and IV and v from the wing, with i -in a fair way to ultimately disappear from both wing and leg, except when functional in the latter. The cat has normally four digits (ii to v) on each foot, all with three phalanges, and all furnished with claws. Besides this, I is represented on the fore foot by a pollex (thumb) of two phalanges, and a non-retractile claw, while on the hind foot the hallux (great toe) is rudimentary, consisting of a small bone articulating with the cuneiform but bearing no claw. Of all animals, aside from man, the cat is the most subject to supernu- merary digits, especially on the fore foot. In the great majority of cases the doubling is in the region of digit i. Often the extra mem- ber is shaped, not like its neighbors, but rather as if belonging to the opposite foot, though sometimes it is indifferent. For exhaustive material on this subject, see Bateson, Materials for the Study of Variation, pp. 313-324. Speaking generally, the dog tribe has five toes in front ^ (digit I not touching the ground) and four behind (i absent). The seal has five digits on all extremities, though the hand is modified into the flipper, and the foot is but slightly functional and evidently well on the road to extinction. The whale generally has five digits in front incased in skin to form a flipper, though this number is often reduced to four, and in all cases 11 and iii have more than the usual number of joints. The only traces of a hind limb are a few small bones beneath the sacral region and occasionally a part of a limb.^ In the manatee and the dugong the process has gone farther. Though these aquatic mammals have exceedingly serviceable flippers with five digits, yet the hind leg has been entirely lost. The vertebrae in the sacral region are not united, and even the pelvis is represented only by a pair of splint bones, though some fossil forms show a rudimentary femur or thigh bone.^ 1 Excepting the African hunting dog, which has four (Lydekker, Library of Natural History, p. 496). 2 This is similar to the loss of wings in the case of the New Zealand apteryx. ^ Lydekker, Library of Natural History, p. 1156. 58 VARIATION The bear has five toes, all round, with an additional claw in digit II behind, which he uses for combing. In mice and rats digit I'in front is rudimentary. This case is unique because in most instances, where a difference is notice- able, the reduction in digits has proceeded farther behind than before. Snakes, especially the large ones, occasionally show external vestiges of hind legs, and internally are frequently found traces not only of the pelvis but likewise of the thigh bone or femur. ^ This shows clearly that the snake is a somewhat recent form developed from lizard-like ancestors with limbs, the hind pair of which must have been placed not far from the middle point of the much-elongated body. This view is strengthened by the fact that as a rule but one lung is developed, showing that the body is more slender than formerly. Of all studies in digits the most interesting is that of the ungulates or hoofed animals. It is also the most profitable, because the majority of our valuable domesticated animals are included in this classification. The interest arises from the fact that out of this stock have developed two very different forms of feet, viz. the two-toed (as cattle) and the one-toed (as horses), both evidently having descended from five -toed ancestors, each by a process of its own. For example, cattle, sheep, deer, pigs, etc., have two. toes (iii and iv) well developed into a serviceable foot, with two others (ii and v) standing behind, not touching the ground (pigs), often rudimentary (deer), and frequently represented merely by splints (cattle). Occasionally all trace of these digits is lost (giraffe). On the other hand, the horse and his kind have but a single toe (hi) ; but on either side is a well-developed splint, the re- mains of the second and fourth metacarpals in front and of the corresponding metatarsals behind. They are, however, without functional significance, being attached only above and extend- ing downward with slender shafts and free ends not supplied with digits. In this connection it is to be noted that the extinct protohippus of the United States and the hipparion of Europe, both decidedly 1 Lydekker, Library of Natural History, p. 2535. MERISTIC VARIATION 59 horselike animals and regarded as ancestors of the modern horse, had each three toes that probably reached very near the ground. Passing still further back (down) in geologic time and looking for a still more remote ancestor, we get beyond what can be called a true horse, as can the protohippus and the hipparion. But yet there is among these long-extinct forms sufficient horse- like character to suggest ancestry, as with the forest horse and desert horse of the Whitney find in Wyoming, forty inches high and three toes down.^ As we progress in this direction, however, the toes increase in number to four and even five, clearly indi- cating that the modern horse has developed from a five-toed ances- tor like the Eohippus, twelve to sixteen inches high and all toes down, also discovered in Wyoming where it flourished, according to Osborn, some three million years ago or thereabouts.^ If we begin with the modern two-toed species and attempt to read their story backwards, we soon land among the same four- or five-toed primitive forms just mentioned, forcing the conclu- sion that the one-toed and the two-toed species of recent times have each descended from five-toed progenitors, — indeed, we may even believe from the same five-toed progenitors. The manner of this descent is not difficult to trace by the comparison of modern species with similar extinct forms in suc- cessive downward (backward) geologic times. In almost the lowest tertiary rocks of both North America and Europe occur fossil remains of large ungulates. These '* Coryphodons " were supplied with five-toed feet much like the elephant of to-day, that has survived by virtue of his teeth and in spite of his feet. Ascending to the Miocene Tertiary, we find large ungulates still remaining, but digit i is gone, while the metacarpal (or meta- tarsal) has become much lengthened and the third and fourth members greatly strengthened, not only in their own development 1 Henry F. Osborn, Origin and History of the Horse. 2 The development of the horse from an ancestor only twelve to sixteen inches high and with five toes, all down, is the best instance of progressive evolution of which we have any knowledge. Doubtless the evolution of other species has been no less extended and fascinating, but of no other case do we possess so complete a history, thanks for which are in large measure due to the generosity of the late William C. Whitney and to the labors of Professor Henry F. Osborn. See also chap, x, sect. ii. 6o VARIATION but also as regards their articulation with the small bones above. This foot is now on the road to becoming indifferently either a two-toed or a one-toed form, depending upon whether ii and v reduce together or whether iii takes the lead. In this connection certain intermediate or stranded forms are of no little interest. For example, the elephant has five toes in front, with four and sometimes three behind. The rhinoceros has three both before and behind, but the extinct form often had four. The tapir, which is also regarded as a remnant of ancient life preserved until the present, has generally three toes, though sometimes four and occasionally two. In any case, however, digit III is largest and symmetrical in itself, showing affinity with the line of descent that has developed the single-toed forms. The camel has two toes, while the nearly related chevrotain has four, two being reduced. The hippopotamus has four short toes, all down, all hoofed and partly webbed, showing affinity with two-toed forms in that the symmetry is about a line drawn between digits iii and iv. This is the same plan as that of the pig, except that in the latter the foot is more contracted, the toes being flattened on the inside and the second pair not touch- ing the ground. The kangaroo is anomalous, having five toes in front and in general four behind, of which iv is much the largest ; v is small, and II and in are much reduced and incased in a common integument. With this brief survey of specific differences in respect to digits, certain individual deviations will have an added meaning. Bateson gives us the following : ^ 1. Horse having supernumerary toe on inside of right fore foot, presum- ably digit II. It articulated with an extra bone in the lower row of the carpus and was provided with a hoof, " convex both sides, resembling the hoof of an ass" (see Fig. i i). 2. Foal having two toes on each fore foot, otherwise normal. The car- pus was in this case normal, but the extra toes were again borne on the inside and were provided with a small hoof. 3. Horse having a rudimentary digit on inside of left hind foot. This again results from a slight development of digit 11, which is the most com- mon cause of polydactylism among horses. 1 For variation in the feet of the horse, see Bateson, Materials, etc., pp. 360-372. MERISTIC VARIATION 6l More rare than this are: (i) the development of iv, making an extra toe on the outside ; (2) development of 11 and iv, with iii normal, making three toes in all, after the fashion of the protohippus ; and (3) development of II and IV vf'ith iii aborted, resulting in an abnormal two-toed foot. All these forms are well known among horses. 4. Horse with supernumerary on outside of each fore foot, illustrating condition mentioned above (development of digit iv) (see Fig. 12). 5. Horse with both splint bones bearing digits on each foot, illustrat- ing condition 2 (11 and iv developed normal, making a three-toed horse). Fig. II. Right fore foot of horse (front view) : as the hoof of the horse is re- garded as digit III, this extra member is to be considered as digit 11. — After Bateson Fig. 12. Right fore foot of horse (rear view) : this extra toe is to be regarded not as digit 11 but as digit IV. — After Bateson 62 VARIATION 6. Foal with right fore foot bearing two complete digits symmetrically developed, each bearing well-formed hoofs'that are flattened on the inner sides and curve toward each other like those of the artiodactyles (cattle, etc.). This illustrates condition (3) just mentioned (see Fig. 13). Cattle, sheep, and pigs af- ford deviations no less inter- esting : ^ Fig. 13. Foot of horse: digit iii sup- pressed, digits 11 and IV developed. — After Bateson I. Calf having three digits on right fore foot, borne on a single common bone after the fashion of the birds and fully symmetrical (see Fig. 14). 2. Heifer having three fully developed toes on each hind limb. In this case the supernumerary was clearly digit il. 3. Calf with "supernumerary toe placed between the digits of the right manus (fore foot). This toe had a hoof and seemed ex- ternally to be perfect, but on dissection it was found to contain no os.sification, but was entirely composed of tibrous tissue and fat." 2 4. Cow, full-grown, right fore foot with four digits arranged in two groups of two each (see Fig. 15). This is clearly a case in which the increase is due, not to the reappearance of an ancient lost toe like il or V, but rather to the doubling of the nortnal digits III and IV through ordinary ineristic variation?^ 5. Calf, left hind foot with five toes, " an inner Fig. 14. Right fore foot of calf: three digits group of two toes curving present, each supplied with both flexor and toward each other and an outer group of three, of extensor tendons. — After Bateson 1 Bateson, Materials, etc., pp. 373-390. - Ibid. p. 377. 3 No case is better than this to suggest caution to the student of evolution. When an extra toe appears among those forms whose ancestors were known to MERISTIC VARIATION 63 which the middle one was ahiiost bihiterally symmetrica], while the hoofs of the other two turned toward it." ^ Bateson says of the pig that he knows of no case of polydac- tylism in the hind feet. All cases described are of the fore feet, and the extra toes are on the internal side of the digital series. Syndactylism in cattle, sheep, and pigs. By this term is meant a real union of digits II and iii into a single bone incased in a single hoof, as in the solid-hoofed hogs. According to Rosenberg, as quoted by Bateson,^ in the normal sheep " the meta- carpals II and V are distinct in the embryonic state, afterwards completely uniting (fusing) with iii and iv." "^ This throws some light upon the whole ques- tion, as tending to explain not only certain cases of polydactylism but all cases of syndactylism.* Again quoting Bateson : Fig. 1 5. Said to be the right fore foot of cow : digits in two groups of two each. — After Bateson (from Delplanque) 1. A young ox having the two digits of the right fore foot completely united. 2. Calf: each foot having only one hoof, though all the bones were normal. 3. Same as above, except that in the fore foot the normal digits (iii and IV) were completely united, bearing a single hoof. The same condition was found behind, except that the hoof was more pointed. 4. A fore foot and a hind foot of the same individual (pig), in which the two chief digits were completely united, viz. represented by a single series of bones. possess a greater number of digits, it is habitual with many to regard it as a case of atavism, — the reappearance of a long-lost character. But how is it in the case of man when a si.xth or even seventh digit appears ? This must be meristic varia- tion and not atavism, because no six-toed species of any sort has ever been described or its existence suspected. Meristic variation, therefore, is not limited to lost characters or to numbers once normal, but may go far in excess of either. Here, then, is need for discrimination, for even the appearance of a character that has been once lost is not absolute evidence of atavism. 1 Bateson, Materials, etc. p. 381. - Ibid. p. 3S3. 3 The former from failing ever to unite, the latter from a continuation of the fusing process to include in and iv. * That is, II unites with iii, and v unites with iv during development. 64 VARIATION Other and similar cases are given, though the latter is the only one described in which the syndactylism is complete in all four feet. It is, however, generally simultaneous in fore and hind feet.^ Absence of parts in a linear series. Men with hands but no arms, with feet but no legs, are not unknown. Whether the missing parts are really dropped out of the series, or whether they were originally present but suffered abortion during embry- onic development, being repre- sented at maturity by rudimentary parts, is uncertain, though zoolo- gists would incline strongly to the latter view. The exact fact would have an important bearing upon the unit of variability, the nature of heredity, and the manner of differentiation. Whatever the fact in this re- gard, variations of this order are manifestly rare as compared with the increase or decrease in strictly multiple parts. In other words, while considerable deviation in the number of similar members (as fingers) is common, it is exceed- ingly uncommon for an entire group (as the hand) to be omitted, — rarely from the end of the series (as the foot or hand) and still more rarely, if ever, from the middle of the series, as would be the case in a truly missing arm (humerus, radius, and ulna), but with the hand present, coming directly from the body. Extra legs. The repetition of a member as complicated as a leg is extremely unusual but by no means unknown. The writer 1 Bateson gives many similar cases, each with some peculiarity of its own. Solid-hoofed pigs are seen so frequently and at points so widely removed both in time and space (mentioned by Aristotle and reported from many regions of the earth) (see Bateson, Materials, etc., p. 3S7) that this would seem to be a variation that has often arisen afresh. Fig. 16. Meristic repetition in leg: right leg of beetle repeated in triplicate. — After Bateson MERISTIC VARIATION 65 saw one specimen of a leglike appendage growing from the left side of the neck of a calf near the point of the shoulder. The leg was not more than two thirds the usual length, and was twisted and functionless, though it terminated with a hooflike growth.^ Extra legs are common in insects, sometimes throughout their entire length, sometimes doubling at the femur (see Fig. 16). SECTION III — MERISTIC VARIATION AND BILATERAL SYMMETRY Meristic variation among paired organs and those standing singly on the median line throws no little light upon the nature of bilateral symmetry and also incidentally upon the manner of variation. Speaking generally, paired organs may double on either side separately or on both sides (digits, legs, wings, etc.), or they may unite into a single organ with its axis on the median line (horse- shoe kidney). Most of the examples of meristic variation already given are of repetition in paired organs in bilateral symmetry. It remains to call attention to the opposite condition, — the fusion of a pair into a single organ standing on the median line : 1. A good example of this is that of a roebuck having the horns com- pounded for full}' half their length into a single " beam " standing on the middle line "^ (see Fig. 17). 2. A honeybee '"^ having the two compound eyes united into one at the top of the head with no groove or line of division between them.^ 3. Posterior ends of kidney united (in man), forming a horseshoe kidney with three renal arteries on each side. This case is in sharp contrast to Bateson's No. 407, with a single large kidney on the left and two smaller, one below the other, on the right. 1 Thi.s .specimen is de.scribed from memory, as it was seen before these phe- nomena were matters of personal interest. 2 Bateson, Materials, etc., p. 460. ^ Ibid. p. 461. * The compounding of eyes has already been mentioned. It apparently occurs only in insects, but is a good example of the development of highly differentiated tissue in abnormal situations, illustrating not only meristic variation but functional variation as well. 66 VARIATION Conversely, unpaired organs standing on the median line may be divided so as to form a pair of organs symmetrically placed. It should be noted that in general a single organ standing on the median line, as the nose, is symmetrical both with reference to itself and to the median line, but that for the most part paired organs, though symmetrical with reference to the median line, are 7iot tJicinselves necessarily synivietrical bodies (ears, arms, hands). In other cases of paired organs, however (eyes, kidneys), the members do have a kind of symmetry of their own. Again, nothing is more common, especially among plants, than to find a single organ on the median line appearing as a paired organ in cer- tain individuals or in nearly related species or varieties. Bateson gives as examples of the last the posterior petal in Veronica^ which in most related species appears Fig. 17. Compounding of paired ^g ^ pair of petals lying on either side organs: the two horns of this ^ , • i n i- roebuck are united into a single ^f the middle Ime. beam for a considerable dis- After giving numerous instances of tance, but afterwards they sap- division of median Organs in fishes arate. — After Bateson ... , . , . ^ and m msects, he cites authority tor saying that " The organs most often divided in man are the sternum, neural arches, uterus, penis, etc., and of these, speci- mens may be seen in any pathological museum.^ Organs more rarely divided are the tongue, epiglottis, uvula, and central neural canal." ^ These latter are in reality cases of a.xial duplicity.^ 1 Bateson, Materials, etc., pp. 450-458. 2 Teratology is that branch of biology which treats of abnormalities, and it affords many cases of e.xtreme variations. This study has been considered as curious rather than profitable, and yet, as such abnormalities are coming to be regarded as frequently due to a defective germ, it may yet prove that attention to cases of this order may furnish the key to the solution of questions involving the unit of variability. ^ Bateson, Materials, etc., pp. 559-566. Fig. i8. Double-headed turtle compared with the usual spechnens two to three days old. Note effect on shell plates. In this specimen the movements of the legs on opposite sides were not well coordinated. — After Bateson 67 68 VARIATION Fig. 1 8 shows a case of double head in the turtle. Many similar instances have been described, but this is especially inter- esting because " the two heads seemed to act independently, and it is said there was no concerted action between the feet of the two sides." The same phenomena of double monsters are said to be frequently noted in fish-hatching establishments. Among snakes " some twenty cases are recorded of complete or partial duplicity, nearly always of the head. Several of these were ani- mals of good size, and must have led an independent existence for some considerable time." ^ Similar cases of doubling are known in birds and even in mam- mals, but among these higher animals the practical difficulties in sustaining existence with extreme abnormality are very great, and they commonly do not long survive. Between this division of a single organ lying on the median line and the doubling of so important a part as the head, there seems to be no clear line of demarcation. This doubling may even go further, as in the case of the Siamese twins, until the specimen is regarded as essentially two individuals united by some sort of attachment. SECTION IV — SYMMETRY IN VARIABLE PARTS Without a doubt meristic variation in one organ of the body is likely, but not certain, to be accompanied by abnormality in another. For example, a variation among the digits of the fore foot is likely to be associated with a similar variation behind, still more likely on the opposite side, but not positively with either. Again, there is some suggestion of symmetry within the part itself in which the variation occurs. A good example of this is Bateson's No. 495 (Fig. 19). This is a left hand, and the four extra fingers seem to repre- sent not the thumb of that hand but the fingers of the opposite (right) hand, thus seeming to aim at a kind of secondary sym- metry within the member.^ In the description of this case we are told that this double hand and arm were very muscular, so that it was not possible ^ Bateson, Materials, etc., p. 561. " Ibid. p. 335. MERISTIC VARIATION 69 to decide in the living subject whether or not there was a doubhng- of the bones of the forearm. The eight fingers were in two groups of four each, with a wide space between. The two " hands " were thus opposed to each other and could be folded upon each other. The power of independent action of these digits was limited, showing an insufihcient supply of muscles. If the two index fingers, iv and v (really ii and 11), were extended, the other six could be flexed ; either group of four could be flexed independently of the other, or the three fingers of either Fig. 19. Symmetry within the variable part. Here it would seem that an attempt has been made to repeat the hand, or rather that an attempt at repetition of the thumb has resulted in a doubling of the hand. — After Bateson hand could be flexed alone. The index fingers alone could not be flexed while the other six were extended. Bateson gives several other cases of "double hand" (Nos. 496—500), all giving the impression that the doubling is not simply of digits but of a JuDid as a wJiolc. His No. 513 is the case of a double thumb, in which the two are symmetrically opposed to each other. It is unfortunate for our purpose that so large a proportion of cases cited as examples of mcristic variation should be among human subjects. This is only because it is here that the matter has been most studied. The idea has been advanced that domes- ticated species are more variable than wild ones, and man more variable than his simian congeners. The point is not well taken, because careful study shows the ape, the chimpanzee, the baboon, and the gorilla to present the same meristic deviations in respect to diijits and the same abnormalities in dentition as are found 70 VARIATION in man. Again, while digital variation is exceedingly common in chickens, it is rare in birds generally, and is almost unknown in ducks and geese which have long been domesticated. The fallacy above alluded to seems to have arisen from the fact that domesticated species are better known than wild ones, and that certain variations at least are more likely to be preserved. In any event they are more strongly impressed upon our atten- tion. The truth seems to be that variability depends upon the nature of the part and the relative stability of the species in ques- tion, not upon its domestication or its place in the scale of life. We can therefore avail ourselves of any material bearing upon the general question wherever it may be found, hoping, however, for the early coming of the time when the variations within the particular field of thremmatology shall be better known and more accurately described. Asymmetrical development in symmetrical parts. The case of the narwhal illustrates a fact in xariation which, though seldom so apparent, is doubtless often potentially present, and if so, is certainly to be reckoned with by the breeder. In the narwhal the canine tooth (in the male only) develops as a tusk, often attaining a length of seven or eight feet, or half the length of the body. The peculiarity is that normally only the left tusk develops, and in the few cases seen in which both are developed the right tusk is spirally twisted from left to right, exactly like the left tusk, and not in the opposite direction, as we should expect. What is still more astonishing is that no case has ever been described in which the right tusk was developed alone instead of the left. Either both are developed or the left one only, and in the former case they are essentially both left tusks. SECTION V — MERISTIC VARIATION IN RADIAL SERIES Except in the lower forms, radial series are characteristic of plant rather than of animal life. In the branching of stems and the parts of flowers, members of radial series are everywhere about us. Their variations are always interesting (doubling of flowers) and often exceedingly valuable (stooling of grain). MERISTIC VARIATION 7 1 Botanists would say that what seem to us as radial series, with the members standing on the same horizontal level, are in most cases really shortened stems, bringing these parts into a relation which is apparent rather than actual, as would happen if we could telescope any long stem until the leaves, regularly dis- posed along its length, should come to occupy the same plane. ^ In this view of the case the petals of flowers and the branch- ing of stems, as in the stooling of grain, would be examples of linear series very much shortened rather than of radial series, according to the strictest definition of the term. For our purposes, however, this structural point may be waived, and all apparent cases of radial symmetry treated as actual. Observations indicate and experiments show that members of such series may be increased in number almost indefinitely. All the members may be doubled simultaneously (as five petals increased to ten), or any one member (original segment) may double or even triple, or it may be entirely suppressed without reference to other members of the series. The natural method of doubling seems to be for cell division to proceed one step beyond the normal, gi^'ing rise to two instead of one. If this occurs in all the members (petals), then the members will all be doubled, as ten instead of five ; if only in part, then only that portion will be affected, making six, seven, or even eight instead of five. Thus we have clover running all the way from the normal three up to as high as seven leaflets. Manifestly if cell division proceeds two stages beyond the normal, each of the twin pair again dividing, it will result in 1 Leaves are arranged in regular order upon the stems of plants according to a system constituting the mathematical series, J, |, |, |, etc., in which the numerator indicates the number of circuits around the stem to reach a leaf directly over the one with which the count was started, and the denominator the number of leaves that would be passed in such a circuit. It therefore repre- sents the number of members in a whorl of a shortened stem of this character. Corn, for example, belongs, with all other members of the grass family, to the fraction ^ , — built upon the plan of two. This number runs throughout the plant, and while the number of rows of corn on the cob may vary freely from eight to twenty-four, 710 case of an odd mimber of roivs //as ever been reported. This fact tends to set some limits to even so wayward a thing as meristic varia- tion, which seems never to have produced an ear of corn with an odd number of rows. This seems marvelous when we consider the havoc it works with digits and with even so complicated a structure as a head. 72 VARIATION quadrupling that member of the series. If, however, only one of this pair should divide again, we should then have one plus two, or three, new parts in place of the one that was normal. Again, if all four should start and one abort, it would likewise result in three developed members instead of four that should have appeared. All these various processes may take place, but whatever the final result, and whatever number ultimately develops, the method is that of doubling through cell division, giving rise naturally to even numbers. Odd numbers are explainable, however, by sup- posing that one of a pair continues the process one step farther than its twin, or else that one of the members fails to develop. In these ways an original member of a radial series may at any time develop into two, three, four, or more ; and if all the mem- bers take part, a true doubling results. Meristic variation and cell division. In the last analysis, there- fore, variation in the number of members in a radial series is reducible to questions of cell division. Indeed, we may go further and note that all cases of meristic deviation arise in this manner ; that the preservation of the normal number of multiple parts depends upon successful cell division up to a certain (normal) point and its abrupt cessation at that point ; and that all sorts of abnormalities may arise through excessive multiplication, through abortion, or through some other disturbance of the process of cell division. This view of the case helps to explain why it is that meristic variations in radial series are among the easiest to explain of all variations which may present themselves to the breeder. Considerations of this character make clear the futility and shortsightedness of appealing to reversion or atavism to explain what may be a mere incident in cell division, — an incident, more- over, that may never have occurred in phylogeny, may not be even common in ontogeny, and is therefore not to unduly im- press the observer.! 1 These terms will be frequently used in the text. Phylogeny refers to the development of the species, ontogeny to the development of the individual. The latter is supposed in a general way to repeat the steps of the former, though with this view of the matter important gaps are of frequent occurrence. MERISTIC VARIATION 73 SECTION VI — IMPORTANCE OF MERISTIC VARIATION Nothing is of more direct benefit to man than the stooHng of grain, and the doubhng of flowers is of prime importance to students of the beautiful. Digital variation, and indeed most of the examples among animals, are not only of no practical use but they constitute deformities that would at once be eliminated from the fields of any intelHgent stockman. Their study is, however, useful to the student in two ways : first, as showing him that freaks are by no means uncommon and therefore not to be specially prized ; and second, to show the manner in which variation operates and the size of the unit involved, together with something of its relations to other and similar units in the same body. The careful student will not, therefore, waste his time in trying to establish a race of solid- hoofed hogs the first time a specimen of the kind turns up in his yard, but he will utihze the information afforded in meristic variation generally to advance his understanding of the manner in which variation behaves and of the relations that obtain between the several parts of a highly differentiated body. The purpose at this time is to secure a mass of characteristic facts on which future studies may be based. Most of the dangers of erroneous procedure in this field arise from a paucity of well- authenticated instances and from restricted views of their real significance. Summary. Meristic variation refers to deviations in the plan or pattern on which the organism is built. Its central thought is symmetry. Symmetry may be radial with the members identical, or it may be bilateral with opposite members, as optical images the one of the other. Distinctions of right and left arise from those of dorsal and ventral, and have reference to the relation of the individual to the outside world. Organs symmetrically placed may or may not have a symmetry of their own, but the ten- dency is for a part to establish some kind of symmetry within its own members. When parts are multiplied they may be like the other mem- bers of the series in which they arise, or they may imitate those of neighboring series (homoeosis). Repeated parts are especially 74 VARIATION subject to meristic variation. The general plan is preserved, but wide variation in the details is common, as in the nerve branches from the spinal column. The part repeated may be simple, like a digit ; or it may be an entire group, as a whole hand or an entire leg. Meristic variation has its seat in cell division. It is of little utility in animals though highly useful in plants, but its phe- nomena are valuable for the insight they afford into the man- ner of variation, the general persistence of plan, and the unit of variability. Exercises. Let the student give ten separate examples of meristic variation not mentioned in the text and describe each fully, stating all that is involved of symmetry, homoeosis, etc. ADDITIONAL REFERENCES Variation. A cock with no spurs on the leg, but with well-developed ones on either side of the comb. By E. S. Dexter. Science, VII, 136. Vegetable Teratology. By Maxwell T. Masters. CHAPTER V FUNCTIONAL VARIATION By functional variation is meant a deviation, not in form or in the number of parts but in the functions that they perform. The Hving animal (or plant) not only is something, but it does some- thing, and plants and animals differ among themselves not only in what they at-c but in what they do. Each portion of a highly differentiated organism has its own peculiar activity, which is essentially different from that of any other part of the same organism. These activities are not con- stant but variable ; and inasmuch as many animals and not a few plants are kept not for their appearance but for what they can do, any deviations in their performance ability are of prime importance to the breeder, who is bent upon their increased efficiency and their permanent improvement for the service of man. Now plants and animals are considered as high or low in their development according to the degree of differentiation or division of labor between their different parts. In the protozoon the func- tions of life are few, and its relations to the environment are simple. Accordingly its activities are exerted and its obligations to life discharged by the common mass of undifferentiated pro- toplasm, perhaps without so much as a stomach, reproduction being effected by a direct division of the whole mass. In higher organisms (metazoans), however, life is more complex and the responsibilities of existence are heavier. These are met by specialized structures, such as the mouth to take food, the stomach and intestines to dissolve and prepare it for use, the liver to convert certain portions into specially usable form, the lungs to absorb air, the blood vessels to carry it and the digested food to all parts of the body, where each extracts what it needs and can use. 75 76 VARIATION Then there are organs, as the kidneys, whose function is to remove waste products that would otherwise accumulate and destroy the body. There are others, as the udder and various glands, whose function is to manufacture some particular product to be used either within or without the body. There is a system (the muscular) for moving the body as a whole or for the exercise of any of its parts, and a network of nerves forming a ready and rapid means of communication. There is a heart to drive the blood, and perhaps a bony skeleton to hold the complicated mass together. Now the activities or functions of these various parts are by no means constant and invariable from day to day. In other words, there is probably as much deviation in function as in form, and for the purpose of the farmer it is even more important. Evolution not a study in morphology only. The mistake is often made of defining evolution as exclusively a study in mor- phology.^ It means more than that. Living beings are some- thing besides form, and their evolution something more than the development of their form ; indeed, in their service to man both animals and plants are valued less for their structure than for their function, which is what they can do. And so it is that 1 " The problem of development is an acknowledged morphological problem." — C. B. Davenport, E.xperimental Morphology, Part I, Preface. The conception here alluded to is not difiicult to account for. The idea of evolution or development as opposed to the older assumption of special crea- tion was first announced at the very close of the eighteenth century, but was not generally before the public until the appearance of the Origin of Species, after the middle of the nineteenth century (1S59). At that time biologists were chiefly con- cerned in classification as based upon external structure or form. It is not strange, therefore, that the discussion should have first arisen, and the battles incident to a new, startling, and, in the popular mind, sacrilegious theory have been first fought out, in the field of morphology. Gradually, how^ever, biologists began to concern themselves more and more with internal structure (histology), and, quite to their surprise, they found them- selves still within the field of evolution. Then came studies in function (physi- ology), showing conclusively that this, too, is a matter of development and subject to variation and heredity. It is therefore not only erroneous but for the breeder dangerously misleading to consider the study of evolution as confined to the field of morphology, which is not the exclusive nor to him even the primary manifesta- tion of the great principle of evolution. There is evoluion of function as well as evolution of form. FUNCTIONAL VARIATION 77 the breeder, intent upon enhancing their service to man, sees in the variation of the functions natural to domesticated animals and plants the greatest opportunity for improvement. So true is all this that the successful breeder may be distin- guished from the novice at this point. The latter is likely to be attracted, first of all, by form or color, because differences of this sort are striking and easily noticed ; while the former will always keep foremost in mind the question. Why is this animal (or plant) valuable to man, and what is it to do ? The student cannot, therefore, know too much about the natural functions of domesticated animals and plants and the deviations to which they are subject. He should know this, not only as a guide to his selection but also as constituting valuable information upon the nature of evolution and the possibility of influencing the causes that control the development of living beings, functionally as well as structurally. Instances of variation in functional activity are easily divisible into four classes. 1. Variation in the degree of activity of normal functions between different individuals of the same species. 2. Variations in the degree of activity of normal functions within the same individual. 3. Modification of normal functions by external or other influences. 4. Normal functions exercised under abnormal conditions. SECTION I — VARIATION IN THE DEGREE OF ACTIVITY OF NORMAL FUNCTIONS BETWEEN DIFFERENT INDIVIDUALS OF THE SAME SPECIES Variation in milk secretion. This is a function peculiar to one class of animals (mammals). It is the product of a highly spe- cialized structure and is practically confined to the female sex. Moreover, it is of peculiar economic as well as physiological im- portance, and there is no better example to bring out much that is involved in functional variation. The structure of mammary-gland tissue is characteristic wherever found, but the quality and flavor of its product (milk) ^^ VARIATION are not the same for any two species (functional variation between species). Again, no two individuals of the same species can be depended upon to give exactly the same quality of milk, for herd records show that the milk of different cows varies naturally from less than 3 per cent to more than 6 per cent fat ^ (functional varia- tion between individuals). Nor is this dependent upon the food supply, for all authorities agree that the proportion of fat to other solids is dependent upon the individual and not upon her feed. Moreover, differences nearly as wide as these quoted may be found within the limits of a single herd and therefore under identical conditions as to feed. Still again, two individuals of the same breed will produce radically different amounts of milk or fat, whichever is measured, from identical amounts of the same kind of feed. This has been repeatedly and conclusively shown by Professor Fraser of the University of Illinois.^ Probably no fact in animal physiology is of more far-reaching importance than is this marked instance of functional difference between individuals. Three experiments were conducted in the attempt to deter- mine the limits of this difference between cows considered good enough for a place in a commercial herd. In the first ^ Eva produced 48 per cent more milk and 1 1 per cent more butter in ninety-one days than did Janet, and in doing so consumed no more grain and but y.G per cent more roughness. These cows were both mature, were fresh on the same day, and neither suf- fered accident during the experiment, yet Eva produced 1057 pounds of milk and 12 pounds of fat out of an extra feed of 112 pounds of hay and corn stover, — a difference greater than any margin of profit the dairyman may hope to realize. The second experiment was a comparison between Rose, a native cow nine years old, and Nora, a native cow six years old.* ^ The actual range in milk is far greater than these figures. Single milkings have been known to run as low as 1.8 per cent fat, and Jersey cows near the close of lactation often give milk with 9.0 per cent fat. 2 See Bulletin No. ji and Bitlletin A'o. 66, Agricultural Experiment Station, University of Illinois, May, 1898. 3 Ibid. (51) p. 103. * Bulletin A^o. 66, University of Illinois, November, 1901. FUNCTIONAL VARIATION 79 Rose commenced April 13 and Nora, May 22, 1899, and both were milked for a full twelve months. Both were in good health, and both continued in good flow until the last, Rose averaging over 18 pounds of milk per day and Nora nearly 14 pounds for the last seven days of the test. Each consumed all the feed she cared to take, the only restriction being that its composition was the same for both. Neither was in any sense beefy, but Rose gained 181 pounds and Nora 165 pounds from August i to April I, showing that they were evidently working at or near their limit of milk production. The grain fed was corn meal and oil meal, and the roughage consisted of clover hay, rape, green corn, and corn silage, always in combination with one or more of the following, — gluten meal, wheat bran, and ground oats. They were never on pasture during the experiment, and, as has been stated, the feed was identical in quality for both. Rose consumed slightly the heavier ration and yielded decid- edly the larger product both in milk and fat. The following table exhibits the total feed consumed and the product yielded for the entire period of twelve months : ^ Comparative Milk Production on Basis of Food Consumed Cow Feed ' Milk Fat Butter ' Rose Nora 6477.92 6189.06 11,329.00 7,759.00 564.82 298.64 658.95 348.41 Difference .... Per cent .... 288.86 4.67 3,570.00 46.01 266.18 89.13 310-54 89.13 Cast in verbal form this means that Rose was able to produce 47 per cent more milk and 89 per cent more butter than Nora, with the consumption of but 4.67 per cent more feed. Reduc- ing both to the same basis of food consumed, it appears that with a given amount of feed fo7- every WO pounds of milk giveti by Nora, Rose gave IJQ.§ founds ; a7id for every 100 pounds of 1 Feed in pounds of digestible nutrients. Butter reckoned at 16.66 per cent water, adding one sixth to the butter fat. Per cent of difference calculated on Nora as a base. 8o VARIATION butter fat {or butter) produced by Nora, Rose produced iSo.y pounds. For purposes of milk production, therefore, feed was worth 39.5 per cent more when fed to Rose than when fed to Nora, and for butter production it was worth 80 per cent more. This, then, is the true measure of the functional difference between these two cows, and it is good and sufficient ground on which to base breeding operations. Further, it is to be noted that this is not the difference between .a good cow and a poor one but between two good cows; for Nora produced 348.4 pounds of butter, which, as Professor Fraser remarks, is nearly three times the average yield (130 pounds) of cows in the United States, and almost one half more than the average yield (250 pounds) of what are considered profitable cows in Illinois. It may be added at this writing (1906) that Rose, though used in many experiments and exhibited at various state fairs and at the St. Louis Exposition, is still living, hale and hearty at sixteen years of age, and is still an economical producer of milk. She has an average yearly record of J84 pounds of butter fat for ten years, ^ and though she has been in many tests since the one just reported she has never been beaten but once. That was in the following case, which bears further on the present point. Three cows were in this test with Rose, — Tina Clay's Queen, known to be a poor cow, and two natives, known as No. I and No. 3, supposed to be two of the four best cows bought for experimental purposes out of a herd of one hundred. Reduced to the same feed basis, and taking the yield of Queen as 100, that of No. 3 would be represented by 121, of Rose by 304, and of No. I by 312. This is a rate of more than three to one against the poor cow, or over tzuo and one-Jialf to one betiveen good coivs on tJie same feed basis. This difference in the efficiency of individual cows is depend- ent not so much upon daily differences as upon the ability for long-time performance. Some cows will give a heavy yield for three or four months, and go dry in six or seven months ; others will give a profitable yield almost continuously. Both extremes are deceptive. The herdsman will almost certainly overrate the 1 Since the above was written Rose has completed a twelve-year record of 7258 pounds of milk and 360 pounds of butter fat as an average. FUNCTIONAL VARIATION 8l former and underrate the latter, so prone are we to remember striking and maximum data. These are not isolated and peculiar cases. Professor Fraser of the University of Illinois tested 554 cows in 36 commercial dairy herds of the state for a full period of twelve months each. He found that the best 25 per cent of the whole number tested were able to produce an average of 301 pounds of butter fat per year, while the 25 per cent of lowest efficiency were able to pro- duce an average of but 133.5 pounds, — a range of consid- erably more than two to one. The practical significance of this difference is pointed out by Professor Fraser as follows : If it costs thirty dollars a year to feed the poorer cows and thirty- eight dollars a year to feed the better ones, then at present prices a herd of twenty-five of the latter will produce as much net profit as would a thousand of the former. A little calculation will show the immense saving in labor in keeping the smaller herd, and, what is equally significant, the relatively smaller investment in animals, feed, and barns, and the smaller volume of business generally. The faculty of producing a high yield of milk manifestly depends not only upon the activity of the mammary glands but also upon the capacity of the stomach to handle a large amount of feed, and the ability of every organ of the body to discharge its normal functions regularly and to endure the wear and tear of sustained exertion under heavy pressure. This particular function of milk production is, therefore, a kind of resultant or algebraic sum of many body functions, and we should not expect to find its maximum except rarely and in few individuals. A simpler function practically independent of others would there- fore be unhampered by their weaknesses, and it would reach its maximum in a higher proportion of individuals. Variation in meat production. That the same principle is operative in meat production is abundantly shown by experi- ments. Steers were fed separately from calfhood to full maturity at the Michigan Experiment Station. ^ The experiment was com- menced as a breed test by Professor Samuel Johnson, and com- pleted by the writer as a test of individual differences in ability to put on gain in proportion to feed consumed. ^ Bulldtiii A'o. 6g, Agricultural E.xperiment Station, Michigan. VARIATION Gain in Proportion to Feed Consumed Steer Weight at Beginning Pounds of Gkain to One of Gain Jumbo Colby Walton Nick Milton Boy Barrington Disco 650 840 S70 740 920 485 605 440 6.16 6.08 7.00 6.30 6.48 4.56 4-93 4.78 Differences of this character are further shown by Professor Mumford's experiments with the various market grades of steers.^ Feeding cattle are divided in the markets into six grades, from fancy selected down to inferior. A car load (sixteen) of each of these six grades (ninety-six animals in all) were fed on the same ration for a period of 179 days. The animals were all natives, though the better grades showed a much higher percentage of good blood than did the lower. The following table shows the relative ability of these six grades of steers to handle feed and convert it into gain : Rel.ative Efficiency of Different Grades of Steers Grades Gain per Steer Total Gain 16 Steers Total Dry Mat- ter Consumed Dry Matter per Pound of Gain Fancy Choice Good Medium .... Common .... Inferior .... 460 455 419 381 395 350 7362 7284 6705 6095 6322 5607 73.267 88,093 81,017 79.535 75.S75 72,494 9-95 12.09 12.08 13-05 12.00 12.93 Here is a variation of over 31 per cent (460-350) in the total gain of sixteen steers under equal opportunities, and, what is 1 Bulletin xVo. go. Agricultural Experiment Station, University of Illinois. FUNCTIONAL VARIATION 83 more significant, a difference of over 30 per cent in the feed required for a pound of gain. This shows the inferior feeding quahty of the lower grades, due partly to age and partly to lack of breeding. Composition of corn as influenced by functional deviation. One hundred and sixty-three good seed ears were selected from a strain of corn known as Burr's White. ^ They presented to the eye no more differences than are usual with seed corn. Three rows of kernels from each were analyzed for protein and also for fat. As a result the protein in the various ears was found to range from 8.25 per cent to 13.87 per cent and the fat from 3.84 per cent to 6.02 per cent. Composition of Corn from 163 Different Ears Corn Carbo- Corn Carbo- Ash Protein Fat A SH Protein Fat No. hydrates No. hydrates 76 1.70 10.05 4-77 83.48 104 I 54 11.82 4-43 82.21 77 1-45 10.42 5.24 82.89 105 I 2,7 12.36 4.84 81.43 78 1-55 11.00 4.90 82.55 106 I 33 II. 15 5.21 82.31 79 1.62 10.89 4.88 82.61 107 I 33 9-47 4.97 84.23 80 1.63 I 1.50 4-58 82.29 108 I 30 11.04 4.67 82.99 81 1.47 11.49 4.26 82.78 109 I 45 10.82 5.65 S2.08 82 1-39 I 1. 78 4-83 82.00 IIO I 60 12.81 5.21 80.38 83 1. 17 9.08 4.05 85.70 III I 31 10.76 4.13 83.80 84 1-51 12.79 4.25 81.45 112 I 26 10.48 4.54 83.72 85 1.46 11.76 4.94 81.84 113 I 10 9.30 4-38 85.22 86 1.50 12.07 4.61 81.82 114 I 32, 9.12 4.10 85-45 87 1-59 12.40 4-74 81.27 "5 I 29 10.41 4.17 84.13 88 1-35 9-34 4.84 84-47 116 I 10 8.38 4-88 85.64 89 1.61 10.71 4.70 82.98 117 I 42 9-95 4-23 84.40 90 1-55 9.90 4-97 83-58 118 I 44 11.40 5.02 82.14 91 1.56 10.68 4.91 82.85 119 I 55 12.38 4.62 81.45 92 1.46 12.96 3-97 81.61 120 I 39 9-97 4.42 84.22 93 1.48 11.80 4.80 81.92 121 I 36 10.09 4.82 83-73 94 1.74 11.89 4-55 81.82 122 I 36 10.31 5.25 83.08 95 1-55 10.49 5-51 82.45 123 I 34 9.68 4.01 84.97 96 1.60 II. 10 4.38 82.92 124 I 44 11.87 4.61 82.08 97 1-59 11.84 4.96 81.61 125 I 34 10.73 4-53 83.40 98 1-39 10.23 5-51 82.87 126 I 49 13.87 5-72 78.92 99 1.42 8.40 4.91 85.27 127 I 43 "•53 4.31 82.73 100 1.65 12.28 4.76 81.31 128 I 33 11.64 4.57 82.46 lOI 1.30 10.08 4.86 83.76 129 I 36 11.25 4.16 83.23 102 1.49 11.83 4.51 82.17 130 I 35 11.86 5.01 81.78 103 1.44 11.25 4.78 82.53 131 I 47 10.49 4.86 83.18 1 This was the foundation of Dr. Hopkins' work in corn breeding at the Uni- versity of Illinois, as reported in Bulletin Xo. 55 and Bitlletin Xo. 100. It will be further discussed later on. 84 VARIATION CoMPOsn ION OF Corn FROM 163 DH'FERENT EaRS - — Continued Corn Ash Protein Fat Carbo- Corn Ash Protein Fat Carbo- No. hydrates No. hydrates 132 1-55 II. 13 4-55 82.77 186 1.48 10.78 4-74 83.00 ^33 1-39 II. 13 4.10 83-38 187 1.28 10.49 ■ 4-44 83-79 134 1.30 10.85 4-45 83.40 188 '•53 13.10 5-51 79-86 135 1-37 11.29 4-53 82.S1 189 1.32 9-58 5-63 83-47 136 1.59 "•43 5.10 8 1. 88 190 1-25 11.50 4-95 82.30 '^l 1.47 II. 61 4.41 82.51 191 1.29 II. 19 4-31 83.21 138 ,.36 1 1.36 4-53 82.75 192 1-51 11.49 4.07 82.93 139 1-57 9.81 5-23 83-39 193 1.36 9-47 4.51 84.66 140 1-34 IO-53 4.18 83-95 194 1.50 11.47 4.65 82.38 141 1-45 12.42 4.51 81.62 195 1.54 11.09 4-37 83.00 142 1-37 931 4.82 84-50 196 1.30 9-44 3-95 85-31 143 1.29 11-33 4-49 82.89 197 1.26 11.20 4.46 83.08 144 1.42 11-39 4.99 82.20 198 1.44 10.23 4-53 83.80 145 1-45 8.25 4.81 85.49 199 1.29 10.64 4.67 83.40 146 1.47 11.29 4-83 82.41 200 1-39 10.13 4-84 83.64 147 1.26 12.21 4-49 82.04 201 1.38 9.64 5.22 83.76 148 1-54 11.94 4-74 81.78 202 1-39 11.26 4.96 82.39 149 1.36 11.29 4.08 83.27 203 1.26 10.48 4-59 83.67 150 1.44 11.71 4-03 82.82 204 1.66 12.57 4.82 80.95 151 1.40 9-31 496 84-33 205 1.46 10.71 5-36 82.47 152 1.41 11.90 4.09 82.60 206 1-34 10.27 4-65 83.74 153 1-35 12.51 5-I9 80.95 207 1-25 11.09 4.27 83-39 154 1.42 II. 13 5.02 82.43 208 1. 48 12.05 4.78 81.69 155 1.44 1 1.05 4-53 82.98 209 1. 48 10.22 4-30 84.00 156 1-39 11.74 4.14 82.73 210 1-45 II. 16 4-75 82.64 157 1.46 10.02 4.88 83.64 211 1.48 10.44 4.21 83.87 158 1.45 10.66 4-51 83-38 212 1.27 9-75 4.12 84.86 159 1.48 11-53 465 82.34 213 1-53 12.40 4-75 81.32 160 1-43 11.50 4-83 82.24 214 1.58 10.22 4-43 83-77 161 1.47 II. II 4-93 82.49 215 1-45 9.22 4.60 84-73 162 1.48 12.09 5-6i 80.82 216 1.42 10.27 4-35 83.96 163 1.29 10.78 5.09 82.84 217 1.32 9-39 4-83 84.46 164 1.30 9-36 4-34 85.00 218 1.40 9-74 4.71 84.15 165 1.47 10.50 4-75 83.28 219 1-37 9-92 4-32 84-39 166 1.65 11.29 3-84 83.22 220 1-43 9-63 5-23 83-71 167 1-37 9-58 4.72 84-33 221 1.32 10-33 5.01 83-34 168 1.49 10.94 4-34 83-23 222 1.41 12.34 4-57 81.68 169 1.60 11.79 4.22 82.39 223 1-49 10.58 4.64 83.29 170 1.36 11.06 4-39 83.19 224 1-52 11.36 4-63 82.49 171 1.44 II. 18 5-75 81.63 225 ^■33 9-15 4-55 84.97 172 1-45 12.28 3-99 82.28 226 1.36 10.31 5.08 83-25 173 1-39 10.14 4-35 84.12 227 1.46 12.63 5-15 80.76 174 1.30 10.19 5.22 83-29 228 1.41 12.16 4.12 82.31 175 1.40 12.68 5-29 80.63 229 1.36 11.04 4-52 83.08 176 1-37 9.86 4-73 84.04 230 1-43 12.10 4.29 82.18 177 1.48 13.06 4-93 80.53 231 ^•33 10.95 4.60 83.12 178 1-37 10.93 4.76 82.94 232 1.52 12.76 4.10 81.62 179 132 11.87 5-03 81.78 233 1.40 9-75 4.14 84.71 180 1-39 11.27 4-55 82.79 234 1-39 10.78 4.76 83.07 181 1.47 9.66 4.21 S4.66 235 1.58 9-97 5.27 83.18 182 1-37 10.97 3-94 83.72 236 1.40 10.18 6.02 82.40 183 1-54 10.32 5-46 82.6S 237 1.47 II. 16 5-13 82.24 184 1.44 10.68 4.89 82.99 238 1.60 11.42 5.20 81.78 185 1.42 9-33 4.49 84.76 FUNCTIONAL VARIATION 85 The wide variation in all constituents, particularly in protein and fat, indicates that profound differences existed in the func- tional activities of the plants that produced these ears. In order to determine whether these differences are constitutional and therefore hereditary, the twenty-four ears highest in protein were planted (separately) for the "high protein plot" and the twelve lowest for the "low protein plot." In the same way the twenty-four highest and the twelve lowest in fat were planted for "high fat" and "low fat," respectively. This has been continued from its beginning in 1 896-1 897 until the present (1907), and is still in progress, the practice being each year to plant separately the ears that show the highest and the lowest values in respect to these particular constituents. The follow- ing table shows the average composition of the seed corn and the crop for the first year of the experiment: Protein Highest protein ear out of 163 analyzed Lowest protein ear out of 163 analyzed Difference Average of 24 high protein ears planted Average of 12 lovv' protein ears planted Difference Average of crop from high protein seed Average of crop from low protein seed Difference 13-87 8.25 5.62 12.54 9-°3 3-51 1 1. 10 IO-55 •55 Fat Highest fat ear out of 163 analyzed 6.02 Lowest fat ear out of 163 analyzed 3.84 Difference 2.18 Average of 24 high fat ears planted 5.33 Average of 1 2 low fat ears planted 4.04 Difference 1.29 Average of crop from high fat seed 4.73 Average of crop from low fat seed 4.06 Difference 67 86 VARIATION It is sufficient for the present purpose to note that there was a difference of 5.62 per cent (13.87-8.25) in the protein content of the highest and the lowest ears of the 163 analyzed ; of 3.51 per cent (12.54-9.03) in the seed planted; and of 0.55 per cent ( 1 1. 18-10.55) in the crop harvested the first year. It is not the intent to pursue this subject further at this point. It will be fully discussed under heredity and under plant breeding, but enough has been quoted to show that these func- .tional differences are both distinctive and hereditary, and that in it all the ear is the unit to such an extent that it is entirely prac- ticable to permanently influence functional differences by selec- tion. Indeed this has been done already to such an extent that corn has been produced with a higher protein content than wheat. Variation in sugar production. Sugars of various kinds are produced by many plant and animal activities. Certain plants excel in this particular function, and among these wide differ- ences have been found, leading to marked and permanent in- crease in the amount of sugar produced. The beet, for example, though originally producing but from 4 to 6 per cent of sugar, has been so improved and its sugar-producing activities have been so increased as to yield specimens containing as high as 25 per cent of sugar and whole crops averaging 14 per cent. Cane is also variable, and every one familiar with the maple knows that certain trees will yield a large amount of exceed- ingly sweet sap, while others yield but little, which little may be either sweet or tasteless, — indeed, even bitter. Variation in speed, scent, and organic activities generally. One horse is faster or more enduring than another, not so much from conformation as from inherent activity and power of endurance. Some dogs are especially keen in scent, others are defective, and the hearing instinct is much better developed in some individuals (dogs, cats, horses, cattle, birds, etc.) than in others.^ Mental qualities, personal tastes, and intellectual ability in general are conditioned not upon conformation but upon the 1 It is more than likely that some of these differences are connected with the degree of development of certain portions of the nervous system. They are none the less functional, however. FUNCTIONAL VARIATION 87 peculiar action of certain parts possessed by the race in com- mon, but whose special functioning in each particular case determines the place of the individual in the scale of life. Variation in vital functions. For present purposes the animal body may be regarded as a colony of organs, each endowed with its own peculiar function, the life of the whole and of every member being dependent upon the degree of success with which each portion does its work. The whole is, therefore, as strong as its weakest member, and when the whole is put to work in service for man, that service will depend not only upon the functional activity of the special organ involved, as the udder or the muscular system, but also upon the successful discharge of all vital functions when subjected to the unnatural strain involved in working under pressure. The point at which the machine will break down or fail to do successful work is, therefore, a matter of relative strength of parts, and in the last analysis the limiting element in performance is not infrequently one or more vital functions, which experience shows are as variable as are other and, from the biological standpoint, less important characters. The beat of the heart, in man for example, though steadily decreasing in rapidity from infancy to old age, yet varies be- tween different individuals at maturity all the way from less than fifty to more than eighty beats per minute. Athletes tell us that the slow beat is characteristic of long-distance running and sustained effort generally, but that individuals of this order are ill adapted to short-distance running or other work requiring quick response to stimulus. There is a marked difference in the digestive powers of different animals, and some individuals starve because the stomach and intestines are unable to dissolve sufficient food to meet the demands of the body, — and there are all degrees of starvation.^ Others with excellent digestion but with limited powers of assimilation fail to make use of the full supply that ^ Wide study of men, animals, and plants will reveal many cases in which the individual has accustomed itself to an abnormally small food supply. The effect is not necessarily fatal, but it is shown in reduced output either of labor, body product, or other form of organic activity. 88 VARIATION is put into the circulation. In the- case of Rose and Nora, pre- viously cited, what did Nora do with the excess of food con- sumed ? Digestion experiments with individuals indicate no such differences in digestion among healthy animals as the differences in milk yield that are known to exist between cows. The only conclusion is that in a case like this the surplus food passed on out of the body, laying excessive labor upon the excreting organs as well as incurring loss upon the one who provided the feed.^ From this it will be seen that the excreting organs them- selves act as a kind of safety valve, and that much depends upon their relative ability to discharge their functions well. This they are better adapted to do in some individuals than in others, but that every effort is made to keep up with demands is evidenced by the fact that if one kidney is lost the other acts for both, usually increasing in size. Speaking generally, a cow will give as much milk from three teats as from four. Whether this is from compensation, as with the lost kidney, or whether it is true only because cows are seldom worked up to their limits, the data at hand do not determine. The tremendous increase in the activity of the salivary glands on the part of tobacco chewers, the increase in size of muscles through use, and the marvelous development of skill in eye and hand by long-continued practice, as in the playing of musical instruments, reading on the part of the blind, etc., — these are all familiar examples of functional development through prac- tice. That this development is or may be greater in some indi- viduals than in others is too well known to need more than passing mention in this connection. Resistance to disease and invasions of the animal economy generally differ greatly in different individuals. Some are abso- lutely immune to certain diseases, others peculiarly susceptible. It is a matter of common observation that in fields of corn killed by frost an occasional stalk remains green and unaffected, showing unusual powers of resistance, due to some constitu- tional difference. Without a doubt these are the differences 1 This is conclusive proof of the fact that appetite is not a safe guide to the amount of food that can be profitably consumed. The most that can be said is that it is a good indication of body use among animals whose efficiency is known. FUNCTIONAL VARIATION 89 that go far toward constituting the essential distinction between annuals and biennials in temperate regions. And so it is that the value of an animal or a plant depends not only upon what it can do but also upon its powers to endure sustained exertion ; and this is indirectly dependent upon the vital functions, which are therefore of prime consequence to the farmer. The horse that died in a Michigan coal mine at fifty- four years of age after having worn out more than five genera- tions of harness mates ; Old Granny, the Galloway cow that died at nearly thirty-six years of age, having raised no fewer than twenty-five calves ; men who live to be a hundred or thereabouts, — these are examples not of individuals that have been shielded from hardships, but rather of those splendid pieces of animal machinery whose every part easily performs its proper function to any limit laid upon it by the exigencies of life.^ That these functional differences exist and that they are in a measure hereditary are facts that challenge the most careful attention of the thoughtful breeder. 1 Benjamin Franklin Harris died of pneumonia in Champaign, Illinois, May 7, 1905, at the age of ninety-three years, four months, twenty-two days. He was personally known to the writer as remarkable not so much for his advanced age as for the fact that he was in full possession of all his powers, and actively engaged in business until within a week of his death. He organized the First National Bank, which at his death was operated by three generations of the Harris family ; but he was president to the last in fact as well as in name, and the management deferred to his judgment even in matters of detail. He was the owner and operator of over five hundred acres of prairie land, and was one of the earliest and largest cattle men in the Middle West. He was a noted feeder before a market was established in Chicago, and was both a buyer and a seller on that market every year of his life afterward. He was always a believer in heavy cattle, and he finished and sold the one hundred heaviest cattle ever marketed in this country and, so far as is known, in the world, — an achieve- ment of which he was always especially proud. This instance of longevity is given not as an extreme in respect to years but in respect to retention for so long a time of all the powers of body and mind. Mr. Harris never had a second childhood, and was a good example of what a splendid machine is the human organism, which ordinarily breaks at the •weakest point but does not wear out. How full of weak points animal organ- isms really are and how weak these points must be are considerations forced upon the mind by the fact that Mr. Harris' span of life was more than three times that of the average man. These three instances of extreme longevity in animals and man, showing what is possible in animal life, afford to the breeder ample food for reflection. 90 VARIATION Variation in fertility. Certain birds regularly produce two eggs, others three, and still others four, before incubation. The average hen, following her natural habit, lays a "setting" (ten to fourteen) and then suspends for incubation. The " crop " of ova has been laid, and time is required for another to come forward. The Maine station, however, has succeeded in greatly increasing the production of eggs, and has produced one hen that has laid two hundred and fifty-one eggs during a single year. Most cows produce only five or six calves, many only one or two, and some not any, yet Old Granny (No. i in the Galloway Herd Book) produced twenty-five, the last one in her twenty- ninth year. The difference between regular and shy breeders is the difference in the functional activity of the reproductive organs, and next to performance ability it is the most impor- tant character in the eyes of the breeder. Even longevity itself is not its equal from the standpoint of the improver, because quality cannot be said to exist in the race unless the individuals that possess it are sufficiently fertile to insure its easy and certain perpetuation. Accumulation of functional variations. Having shown the marvelous differences in functional activity between different in- dividuals, and having shown that these differences are hereditary, as in corn, it follows necessarily that functional variations may be accumulated into true breed distinctions, and that strains of animals and plants may be permanently established with exceed- ingly high efficiency in desired lines ; indeed, this has been already accomplished, though we are still far short of what is possible. For example, the beef breeds are more economical producers of meat than are the dairy breeds, and the converse is true as to milk production. These two functions have, therefore, been largely separated along breed lines. But it cannot be said that one beef breed is more efficient than another in meat produc- tion, or that any one dairy breed stands out preeminently as the most economical producer of milk. This is partly because breed differences have been largely built up along lines other than those of efficiency, and partly because all breeds contain many individuals of low efficiency in their own breed characters and high efficiency in those of other breeds. Some Jerseys, FUNCTIONAL VARIATION 9 1 therefore, are better feeders than certain shorthorns, and there are to be found among the latter breed individuals that are more economical producers of milk than are many of the Jer- sey breed. It has been said that no race was ever taken by a part-bred horse over one that was racing bred. Whether literally true or not, it is substantially correct, so intensely have the racing ability and instinct been developed in certain breeds. These facts, together with what is known as to corn and beets, show clearly that much yet remains to be done in the way of developing functional activity, thereby increasing efficiency. SECTION II — VARIATION IN THE DEGREE OF ACTIVITY OF NORMAL FUNCTIONS WITHIN THE SAME INDIVIDUAL Of no less scientific or economic interest than the data given in the last section is the fact that functional variation is by no means confined to differences between individuals. The prin- ciple applies, though to a less extent, to the individual itself, whose activities are not constant but variable from day to day and throughout its life. Daily fluctuation in normal functions. It will be found upon investigation that the ordinary functions of the body are unex- pectedly irregular. Even the heart action is not absolutely constant ; it is slower when the body is at rest than when it is in action and is subject to great acceleration in certain diseases. All organs of the body work better some days than others ; indeed, there is a distinct periodicity with each, — a period of increase, followed by one of maximum activity, and that again by one of subsidence. This is the way organs rest. These periods are evidently of different lengths, and it is therefore only occa- sionally that the body as a ivJiolc is at its maxiimini. A good example of this periodicity with functional deviation from day to day is, again, that of milk secretion, which, as has been remarked, is a kind of resultant of all the activities of the body. The fol- lowing table gives the variations in the quantity and character of milk from a single cow for a period of one month. ^ 1 Bulletin No. ^i^ Agricultural E.xperiment Station, University of Illinois. 92 VARIATION O H ^2 a h < OOOO^OOOOOOOOOOOOOO-OOOOOOOO < Q fa 2 per. •< 2 Pi ^ < fa H 2 m Tf ro -t <^ -t ro ^ 'I- ro ro -+ ro ^ ro -h ro ro ro -:f <"0 -<*■ '"'^ ^ ro ro CO Tj- X a < £^ H in>- rOLOOr^M ro "IvjD or-^i-covO iom l/^m rovo "^ M ci Csoo M -0 5 o ^ CO vO M O "J^ CO ro ri 'f ri ro r^ CO d i-> l-Hl-CI-ll-IH-lH-lCIN M On r^ 1-1 ■* ro "^ On Cn M ro pi w w w ro M N N N M PI CN lO On O — i^ 1^ I^VO ro O "^ O 00 rO O t^O --I rororO'^ri pi nvO ro-l-i-Oi^O vd vd CO \d vd \d r^vd d^od d 06 i-< d\ pi d^ ■rt-CO PI 1-^ r^ t^ ^CO NO -> poonm onloq t^— m tONO 10 pi d POCO pi d^ pi dN pi On ro On < fa O Z 9 o CO fe^ 1* cu rou^roupr^- O c-i i-. p; ►- r; w moo "^T'T'" "^^''P'*"'^ T^^"^ '^ d^•G^Q^O^d^d^d^O^Q^O^O^d^d^ dNOO 0^0^0\O^O^O^C><^0^0^0^0\d^ T3 B 3 O p t^t^rocicOCOvO -:1- in rp rj- CO ?" '^. ''! ►HWMi-Ii-iwNN CO NO PI w-1 NO PI ■^ CO PO PO PO -^ pi pi pi pi fi pj s s ro r^ t^ M \0 r^NO VOVO 0:0C0 OCO-/D r^vO 00 t-^ t-^^O t^sO Oco 1- 00 " PI " t-~ r^ On -^co po pi pi -t- m w rO" poOnpOOnpoO poo tJ-o < B, »-4 H-l»-ll-t MHtl-H HI HHhHHHKHHHHH < fa J^ vO VO n CO '■0 t--.vq CO rp On fi CO -t O ro tJ- ro ^ rorj- ro ro ro ro ro r^ ro ro ro -i- PI NO -t p) -to •& n <-n\0 M NO ro ro PO -^i- PO -i- ro -t PO PO PO PO ■a c o ior^Loor-^"^»--i i0\0 u^w-lVO t^CO On w> « Tl- CO On Os On On (^ 00 1 5 g 00 CO Tf w CO OCO r^CO " r-^ M M ri <-i c-i ro ro n n <"' M ro ro -^ ro -t -t -f 'J- u-)W->„ -t-OPI ^0 Geo 1 c 3 O 111 COCOVD r-.q t>0 CT\ Tj-\d '^''id »-i ^-^ i-H M i-i w ri ri M M r^ ^ CO PI PI "^ NO LO ^ lA 10 Nd PI p) PI M PI PI r^ r-^ (> r^ t-^ r^od r^ d On n d* ""O >-< 't d "_ NO q -t ^ Tj-NO NO PI q 1^ q Tfpiuod^d'^d^'-<'^>-< " 1 u H < J M M ro-^'-">-<'-■0^■-|PIO"O^C^Pl►-PlOPOPO"P^PlPOPI■-'>-| 00000000-hOQ-"-0--"-00-"-""-"-^"">--- r^r^r^O pi -co O^O Oco •& O picO On r-.0O "^00 roi^i-oi-^Tj-u-)« -t irico r^co pi •+ POPOPOrfPOTfPO-^j-POPOrO^'^TtPOPOPOPOPOPOPOrOPOPOPOPOrOPOrOPl rOPI •^■'t r^co On PI POu^i-^>-H GO PI r^pi Ov^O m t^O POO ONONPOt^co ^ o — "-)vq rt-cooq pi pi poiopoonOvOmD t-^i-^t-^co r^r^vCJ^O PI PI M PI P< M PI PI PI P4 PI PI PI PI PI PI PI " vO ►- 1-^ vo -:f PO On\0 r^ m O i^ pi pood pi CO ONr^"ir^t-^0 pOON"CO pi pi O O pi pO "\Ooopi"-pi'-'OOOcoo ONoo "~> t^ onoo GNONTt-POLTi'-i o "^i-; •-; '^'"p'-; ^^ "? '^ poONpi ONPod PO" Pod LT)" pop! popi -^i-I POPOLOpi ^pn^^POTJ-PO^PO'^pi tj-p! PI ij~i^u^m^ir,^tj~>u~, -f\D NO o Lo t^ t^NO r~~ Lovo i~^ t^ t-^vq 00 00 t^oo 00 00 00 vq vq On Qn cS 0\ O^ ON O^ 0\ On On On On On On On C> On On On On On On ON On On On On ON ON On On On Qs On POO -' PI •- i^POLOOO POO -^CO O On PI CO Tj--t-"-ir^vO OnOnOnOnQ — O " "_ O O On pi pi pi pi pi pi pi pi pi po po ro pO po po po pi 00 inoo PI O >-> PO On pooO po pi -^ On O ionO pi\0 I-^OnmOONO OnOnOni-i n t^" ►- ►- r^ POO POO ■^•- -"^-M -^i-NO PO lo PO "1 Tf\o PO "^ -*no ■* "^ -^no -^ "^ i^\q -^vq -^vo po ■.^ LT) u-ioo o CO NO 00 Ti-oo NO pioo wNoooNONo -^j-no pi no ^no "^ ■* ■-. PO ^ "^^ ■> 9 ^! POPOPOPOPOPOPOPOPOPI PO-->pi'--''-"d'-< PI PI PI PI PI PO PO PO PO PO PO PO PO PO PO PO PO O'-'t-^r^OOO^r^O^ rocO O •-oco O >- CO i- lonO "-ipoOO pi PiNq ^ O -^ "^00 f^ LO^ "to lon- lOPOLOP) r^ ponO -t" lo lo r^ pOnO ^ r^ ^nO "-^ i^ ltjnO lovO >^nO "^nO ■•^ "^NO r-.CO CnO - PI poM-"-)nO r^OO OnO - — ■-'►- — « PI PI M PI PI PI PI PI PI PI PO PO 94 VARIATION Comment is hardly necessary to show the irregular working of the animal machine, which in this case was a mature, strong, and healthy cow. The per cent of fat varied all the way from 2.7 to 4.2 within the space of twenty-four hours. Influence of age upon functional activity. At birth the vital functions and those of growth are at their maximum. At this time growth seems to be proceeding with the " energy of embryonic development." It continues at a maximum for a time,^ gradually declining in rate until maturity, when growth of the general body ceases, although for certain parts (hair, teeth, horn, hoof, reproductive cells, etc.) it continues nearly or quite through life. In case of accident many parts not com- monly indulging in growth after maturity are able to regenerate with more or less success (bone, skin, blood vessels, nerves), but among the higher organisms generally growth practically ceases at maturity.^ It is for this reason that feeding enterprises are most profitable with young animals that are still growing. At this age functional activity is greater and general bodily efficiency higher. One reason why the lower grades of feeders are less econom- ical producers than the better grades is that the animals are older, the difference in age between fancy selected and inferior steers being one to two years. Strength and endurance are evidently at their maximum somewhat after maturity, although with the passing of that age is also lost something of the power of rapid recuperation and repair. The reproductive functions, undeveloped until a considerable period after birth, attain a high degree of energy, if not their maximum, somewhat before full maturity is reached. They then decline, and fail entirely in old age. Their duration is therefore considerably less than the life of the individual, often dropping below 50 per cent of the full life period. 1 The curve showing rate of growth at different ages has not been sufficiently worked out, but enough has been done to indicate that the maximum rate of growth is attained a few days after birth and that this maximum is never again reached. - Trees continue to increase in size, to some extent at least, during life, illustrat- ing a marked difference between the higher plants and animals. FUNCTIONAL VARIATION 95 At what period the reproductive functions are most success- ful in producing young of a high order of excellence, whether in youth, middle age, or old age, is not determined, though it is a point on which light is badly needed. It is certain that the practice of using young and immature sires is almost universal, especially among cattle. That there is danger in continued reproduction from immature animals, even though they are sex- ually vigorous, there is grave reason to fear, and yet, in general, reproduction antedates maturity. The duration of profitable service depends, of course, upon the nature of the function involved. The life service of a racing horse is manifestly less than that of a work horse, and the life of a meat animal is less than that of one kept for milk. Influence of exercise: use and disuse. The beneficial effect of use in developing and perfecting the functions of the body has been recognized from the most ancient times. Athletes train for this purpose. Musicians practice for many hours every day ; indeed their chief labors arise from the need of constant and severe exercise of the musical faculties in order to achieve any considerable degree of perfection and to hold it after it has been acquired. Horses designed for racing are worked almost from the first in order to make the most of any natural ability to trot or run. Cows are believed to be more efficient producers of milk if they begin at two years of age and are kept, so to speak, in constant practice, and barrenness is believed to be less likely if heifers are bred early than if left to attain maturity without produc- ing young. Darwin discovered that the wing bones of wild ducks as com- pared with their leg bones were relatively heavier than those of tame ducks, corresponding to their respective habits of life. The arm of the blacksmith and the wing of the ostrich ; the remarkable leg of the kangaroo and the remains of that of the whale ; the brain power of the busy man and that of the slug- gard, — these and scores of examples that might be cited show not only that exercise develops, quickens, and perfects the body functions, but they show, too, that their very retention or loss depends in the long run upon their constant and rational use. 96 VARIATION Blind people acquire a quickened and an educated sense of hearing and a touch that amounts almost to a separate sense. In the same way the developing of the bodily functions and activities generally depends upon judicious use (exercise) ; and the skill of the trainer and the results he is able to achieve depend not only upon his knowledge of methods that will most certainly insure the exercise of the desired parts but also upon his judgment as to how severe and protracted that exercise should be in order to secure the maximum effects of use and not incur the destructive consequences of overuse. This is as true in the feeding yard as upon the race track, and applies as well to the raising of good and profitable feeders as to the devel- oping of racing horses, the education of drivers and saddlers, the training of hunting dogs, the " trying out " of homing pigeons, or the teaching of canaries to sing by never allowing the young birds to hear a false note. Influence of feed upon functional activity. The relation of the amount of feed to its economical consumption is a subject needing careful investigation. Enough is known to warrant the assertion that animals can and do learn to take amounts far larger than can be really used. When a steer consumes over a bushel of corn a day he has simply formed the eating habit as the result of a morbid appetite, nor is this appetite an indi- cation of body needs or a guaranty of its powers to economically convert the feed into meat, milk, or labor. It is significant that steers very gradually brought into full feed will never take these enormous amounts. Professor H. W. Mumford of the University of Illinois finds that under such cir- cumstances twenty pounds of grain per day is all the animal will take. Consumption of extreme amounts is, therefore, evidence only of the quantities of feed the digestive tract can carry and dis- charge without calamity, of its power to secrete gastric and other digestive juices, and of the ability of the excreting organs to eliminate unused and unusable surplus from the body In the case of Rose and Nora the latter consumed the same feed as the former but returned but little more than half as much. She was undoubtedly, from the standpoint of economy, FUNCTIONAL VARIATION 97 overfed, but whether the same individual would make cheaper returns on less feed is not so well known as it should be. In the mechanical world the highest return of energy per unit of consumption is realized when the machine is working full but somewhat below its maximum capacity. Doubtless the same principle holds with living machines, but on this point we are sadly in need of accurate information. On one point we are certain. The animal (or the plant) is able to adjust itself to a wide range of food supply, providing the change be gradually made. Not only individuals but whole fam- ilies for generations live in a condition of semi-starvation, often quite ignorant of their real condition, if indeed they are not so indifferent as to prefer to continue in the old way rather than to disturb their tranquillity by increased exertion. Such a state may easily become chronic in man or animal, but it is unprofit- able because all other functions are suspended or reduced to a minimum in order that the vital functions may be discharged at all and the animal not die outright. There is no nicer problem for the stockman and the feeder than this : How much shall I put into this animal machine in order to realize the highest net efficiency, after first providing for those activities which are necessary to the life of the machine, — the vital functions ? Influence of hard conditions. Under hard conditions the func- tions of life may be disturbed but not destroyed. Under these conditions valuable activities are carried forward upon a reduced scale, and they often give rise to losses that are no less serious because invisible. The most common example of this is in the case of ill-fed or much-abused animals and of badly nourished crops or trees : some milk is secreted, but it is insufficient and its quality is poor ; the plant is weak, with little resisting power against insects or disease, and with little ability to mature its crop ; the apples are there, but they suffer for the means of development. Every one who has had experience with unthrifty animals or plants knows how difficult and how slow is the process of resto- ration of normal activity after it has once been seriously checked by neglect or disease. This is because the condition readily becomes constitutional, tending to continue through life. 98 VARIATION SECTION III — MODIFICATION OF NORMAL FUNCTIONS BY EXTERNAL OR OTHER INFLUENCES To what extent are normal functions dependent upon favorable environment and how do they respond to changed conditions ? A few notable facts will throw some light upon this all-important cjuestion. Galls. An insect stings a plant. Under the influence of the poison a gall is formed. If this gall be shown to a biologist he will at once state with certainty the plant which produced the gall and the insect that made the injury, so definite in form and appearance is the resulting growth and so distinct is it from any normal growth of the uninjured plant. In other words, specialized plant tissues subjected to a certain injury produce a new kind of growth almost as specific as when operating under the laws of heredity. Here the functional activi- ties of the plant at tJie affected point \\?mq been not dcstroyedXiwX. permanently altered m such a manner as to give rise to new struc- tures of definite form and often with specific chemical properties, as in the case of nutgalls ^ containing 30 to 80 per cent of tannin. In cases of this kind a new function has been developed suddenly out of old materials, and it at once gives rise to new and distinct products, both substantive and morphological. Abnormal overgrowth of disordered animal tissues. We have just noted that vegetable tissues subjected to specific injuries often suffer such a derangement of their functions as to cause the production of an abnormal but characteristic overgrowth of the injured part. Quite similar is the result of the invasion of the animal economy by specific germs from without, as in the case of Bacillus tuberculosis. Here the growths (tubercles) resulting from the apparent attempt to encyst the foreign material are sufficiently characteristic to serve as a name for the disease (tuberculosis). Tumors generally, whether malignant or benign, are perverse overgrowths of normal tissues of the body, whose ordinary 1 All formulas for good black writing ink include gallnuts (or nutgalls) as the characteristic ingredient. Those commonly used are produced on the oak by the sting of the gallfly i^Cynips galla-tinctoriie). FUNCTIONAL VARIATION 99 functions have been deranged by some cause, external or inter- nal, and whose activities have been diverted to the production of abnormal but characteristic tissue with " no typical termination to its growth." Nearly all tissues of the body ^ are subjected to this derange- ment of their ordinary functions, resulting in the suspension of all activities except that of growth, which proceeds with the "energy of embryonic development" and continues indefinitely. Though often supplied with special blood vessels and a system of nerves, these growths are entirely functionless and therefore useless to the body. Being parasitic they are always a drain upon its resources, and often from their nature or position they constitute a real menace to its existence. A tumor represents a bit of differentiated body tissue that for some reason or other has abandoned its characteristic func- tions, cut loose from all restraints of heredity, set up an inde- pendent existence of its own at the expense of the colony of which it has been a respectable and dependable member, and has now devoted all its resources to growth, which, as has been said, proceeds with the energy of embryonic development, result- ing in nothing but functionless masses of living matter, strongly suggesting a reversion to primitive undifferentiated tissue. Under conditions not well understood all sorts of abnormal growths may appear. In this way an antenna may appear where an eye ought to be,^ or it may end in a foot instead of a feeler.^ The writer knew a young lady of culture and of no little natural beauty except for the fact that, growing from one cheek, was a tuft of coarse black hair three or four inches long. Her normal hair was brown and her complexion clear. What functional dis- turbance could have given rise to such a growth is as mysterious as it was unfortunate. Ossification is a natural process, but under the influence of excessive strain it may proceed to an abnormal extent, as in spavin, where the entire hock joint becomes solid through the ossification of the fluid thrown out as the result of injury. ^ Muscles, fatty tissue, connective tissue, bone, cartilage, nerves, glands, blood vessels, the covering of the brain, etc. 2 Bateson, Materials, etc., p. 151. i ftp p ^ Ibid. pp. 146-147. lOO VARIATION Derangements of a more fundamental nature often arise dur- ing embryonic development, resulting in monsters of all degrees of abnormality. Teratology has little interest to the biologist generally, because these abnormal caricatures of life constitute nothing but sporadic offshoots of the species. Developing from defective germs and having no connection with the line of descent, they are of little interest to the evolutionist. Their interest to the thremmatologist lies in their bearing upon fiDic- tional activity and the degree of certainty with zvJiich specialized tissues may be depended upon to discharge their hereditary and proper functions. ■ Variation due to the suppression or failure of the reproductive functions. The abdomen of the crab Carcinus ni(S)ias normally has seven segments. In the female these are distinct. In the male the abdomen is much narrower, and the divisions between the third, fourth, and fifth segments are obliterated. Males, however, inhabited by the parasite Sacculina do not develop sexual characters, and in them the segmentation is complete, as in the female.^ A young male is castrated. The parts removed are in no sense vital, and they seemingly have no connection with other organs of the body. All the bodily functions except those of reproduction proceed, bnt not as before. In general the develop- ment of the shoulders and neck will be arrested, and they will remain lighter and finer. The voice, the nervous temperament, the disposition, and the general activity of the body are all affected. The mane of horses will be thinner, finer, and shorter. The hair of face and neck in cattle will be finer and less curly. In hogs the tusks and shoulder plates do not develop. The growth of the horns is stopped in sheep, but in cattle the only effect is to make them slightly longer and a little more slender, ap- proaching the female type. The hinder parts of the body as a whole develop rather more in castrated than in entire animals, and there is a general approach to the form of the female. It is noteworthy in this connection that the same general effect follows the failure of the sexual powers with advancing age, except that the body development has already taken place. 1 Bateson, Materials, etc., p. 95. FUNCTIONAL VARIATION lOi Females deprived of their ovaries develop to some extent the characters of the male. Spayed heifers are not at all like bulls, but they do resemble steers. Unsexing animals seems, therefore, to induce a kind of mediocre development, although it gives rise to four distinct types instead of two for each species. Many females in later life assume certain characters of the male. Cows bellow and paw dirt like bulls ; hens grow spurs and try to crow ; women sometimes grow a straggling beard and acquire a heavy voice. These changes do not by any means appear in all cases, but when they do appear they may be regarded as symptoms of loss of the sexual function and of cessation of breeding powers.^ This influence over the functions of the body by organs apparently having no connection with the parts affected is akin only to that of certain glands like the thyroid, whose function is entirely unknown, but in whose absence children grow up defect- ive both physically and mentally.^ We are at this point very near to the forces that determine the activities of living matter, but the mysteries involved are in no sense cleared up ; they rather deepen instead as they are studied. It is as if our vision were obstructed, not by a curtain that can be drawn aside afford- ing a view beyond, but rather by a solid wall fixing the limits not only of vision but of progress as well. Functional variation due to the modifying influence of the conditions of life. The conditions of life are most active in stim- ulating or depressing normal activities, but they are not without effect upon their character as well. Plants having a fixed abode are more dependent upon their environment and therefore less resistant than animals, though species living in confined waters are little better off than plants in this regard. 1 Though bearing but indirectly upon the present que.stion, it is worthy of remark at this juncture that many individuals of each sex seem to be naturally endowed with more than the usual proportion of the characters of the opposite sex and to be correspondingly short in those of their own. Thus we have our " mannish " women and our " effeminate " men, distinguished not only for their tastes and their mental characteristics generally but for their body conformation as well. These abnormal unions of male and female traits are often strange mixtures indeed, and may well be avoided in the breeding yard. 2 Loeb, Physiology of the Brain, p. 207. I02 VARIATION Plants, and animals too for that matter, growing in cold climates or under hard conditions suffer profound changes, to which they become accustomed (acclimated) and which are ever afterward constitutional. We become accustomed to cold or to heat and are thereafter less affected by extremes. Recent calorimeter tests show that the temperature of the human body is lowest from three to five o'clock in the morning and highest from one to three in the afternoon, thus following fairly close the minimum and maximum of outside temperatures. These conditions continue even if the subject works at night and sleeps in the daytime. Two conditions tend to produce hard and spiny growth in vegetation. These are intense light and extreme dryness. Both are found in tropical regions, and when they occur together their maximum results follow as to harshness and spines. These condi- tions can be verified in the laboratory, showing conclusively that the character of growth is, in a measure at least, dependent upon surroundings. Speaking generally, plant lice reproduce parthenogenetically during the growing season of the summer, and during this time only wingless females are produced. With the approach of cold weather, however, a winged bisexual brood is produced that lives over winter. These conditions can be produced artificially in the greenhouse at any time by lowering the temperature and allowing the plants on which the lice feed to dry up. Thus we may say that wings and sex may be developed at will by the manipulation of the condi- tions of life. The so-called conversion of one species into another by influ- encing its environment has been largely overstated, and yet the facts are that when Schmankewitsch ^ grew Artcmia salina in water whose saline content was gradually increased, the caudal fins and their bristles " progressively degenerated " until, in many cases, these appendages had disappeared, the animal thus assum- ing the character of ^. milJiausenii, which normally lives in waters of extreme density. These experiments were undertaken because he seemed to have observed this transformation taking place naturally in a lake crossed by a dam, and which was inhabited 1 Bateson, Materials, etc., p. 96. FUNCTIONAL VARIATION 103 by both species, the one above the other. This dam broke, mixing the two species, but in three years original conditions were practically restored. The experiments were undertaken to learn whether real transformation had taken place or whether the result had been brought about by selection. The experiment seems convincing, and the point is further strengthened by the facts that Schmankewitsch restored the caudal fins by reducing the solution, and that when the reduction was carried below the normal of A. salina, within three genera- tions the last segment of the body divided after the fashion of A. branchipus, another related species. The facts are the more remarkable as A. salina reproduces parthenogenetically, while A. branchipus is not known to do so. Biologists are extremely careful not to assume the absolute con- version of one species into another by any such direct methods as have here been noted, the question being, rather, whether they are all good species ; but that single characters are profoundly influenced by changed conditions, and come to resemble the same character in another and related species, is too well established to be longer questioned. What light this may finally throw upon the origin of species is problematical, but it serves the present purpose in showing the power of environment to profoundly modify the functional activities of living beings. If a cutting of willow, currant, or other suitable growth be planted in the earth, roots will start from the part below the ground, and leaves and branches from the part above. If, now, it be cut off at the surface of the ground and the top portion be planted again, it will again take root at the new point of sever- ance which had before borne leaves. The process may be con- tinued indefinitely, or until the piece is used up, showing that roots or leaves may be developed at will at any point along the cutting, according as it is placed below or above ground. If a cutting be planted, roots will develop only at the lower end. If, however, before planting it be cut into two pieces, each will develop roots on the part below ground, and in many species this will occur even if the pieces be inverted and planted top down.^ 1 This is most likely to take place in young wood, less likely in old wood. See Morgan, Regeneration, pp. 71-91. I04 VARIATION A maple tree growing in Urbana had forked into two nearly equal parts about six feet from the ground. One part was split down and torn off in a heavy storm, when it was seen that roots had developed in the crotch and were evidently at work upon the soil that had blown from the street and the moisture that had accumulated from rains. The writer was excavating for a basement. A black cherry tree stood some ten feet from the line of the wall. In taking: this out many of the roots were severed, their cut ends being left in the bank of undisturbed earth. In a few days these cut ends were clothed with a growth of green leaves. Here was tissue that, under normal conditions, functions only as roots, yet upon occasion readily gives rise to both leaf and stem. All plants and most animals maintain definite relations to light, and if free to move, orient ^ themselves with reference to the direction of the light rays striking the surface of their bodies. A plant bends toward the window because of the contracting effect of the light upon the protoplasm of one side of the stem. Many larvae are negatively heliotropic ; that is, the lighted side of the body is more irritable and they move away from the light, coming to rest only in dark places, where they feed and mature. Others are positively heliotropic when hungry and negatively heliotropic when fed. Such larvae will climb trees and feed upon the leaves or buds until filled, when, becoming sensitive to light, they descend and hide in the ground, under rubbish, or in any other place shielded from the sun. This action has been errone- ously attributed to a semi-intelligent instinct. It is nothing but functional dependence upon external stimuli. This is not the place to pursue the subject in detail, — which will be done when discussing "Instinct and Reflex Action," — but it is the place to note the wonderful dependence of certain normal functions upon external influences. An animal is invaded by a foreign germ and suffers from disease. It is in most cases ever after immune to that disease. What change has been worked in the animal economy ? We 1 By orientation is meant the direction in which the long a.xis of the body is brought to rest with reference to surrounding bodies or influences, such as grav- ity or light. FUNCTIONAL VARIATION 105 know that as long as the white corpuscles are able to discharge their proper function the resistance is complete. Why do they weary of their work, and what condition is left behind which assures absolute resistance to future invasions ? The phenomenon of acclimatization in general represents a condition in which an organism has undergone a permanent change in its vital functions, forced upon it by the exigencies of life. In the future studies, however, it will be seen that the disturbing effect of adverse conditions, if not too severe, may be gradually overcome, and the animal or plant resume its functions, either modified or unmodified ; and it will be seen further that if the changes be gradual, the immunization will extend to a point that would have been fatal at the outset. Thus organisms may be reared in a gradually intensified poisonous solution, or in a liquid whose temperature is slowly raised, and in this way a point may be reached many degrees above the power of normal organ- isms to withstand. The subject cannot be pursued further in this connection, for it is a large one, with many other bearings ; but the student should bear it in mind throughout the study. Irregular functioning. An interesting phase of irregular func- tioning is found in the so-called "instinctive acts," more properly reflex actions, which by popular conception are supposed to pro- ceed with unerring accuracy. This assumption is natural in view of the complex nature of many of these acts, all of which have the appearance of being under the control of reason. For exam- ple, note the complicated nature of the process necessary to the successful deposition of the egg of the yucca moth {Promiba yiic- casclla). We are tokP that these moths emerge simultaneously with the flowers of the yucca, which open but for a single night and are practically dependent upon this particular moth for ferti- lization. When ready to oviposit, the female gathers a bundle of pollen from one flower, flies with it to another, pierces the tissues of the pistil of the latter, and lays her egg ; after which she ascends to the stigma of the same pistil and " stuffs the fertilizing pollen pellet into its funnel-shaped opening." Now this process is necessary not only to the fertilization of the yucca, but also to the grub that hatches from the egg^ which 1 Morgan, Habit and Instinct, pp. 13-15. I06 VARIATION otherwise would be left without food. There is, therefore, a particular sequence in this complicated performance that must be observed or failure results ; and failure is fatal to the exist- ence of both species. Moreover, this act is performed but once in the lifetime of the moth, who has no knowledge of the acts of her predecessors, and is therefore not proceeding from simulation ; nor has she opportunity to learn the fate of her offspring and profit by the experience. Now the truth is that, unerring as is this performance, a good many ovules escape, from failure of the Qgg to hatch or from other causes, and thus the yucca is able to mature some seed. That complicated processes of this kind are not always carried out in proper sequence and full detail is shown by careful study of different individuals, as pointed out especially by Professor C. S. Crandall in his studies of the apple and plum curculio.^ Careful study of these complicated acts and of the body functions in general must convince the student not only of their nice adjustment but (what is of equal consequence) of their exceeding variability and irregularity within limits that certainly are by no means narrow. Cases of "double personality," in which the individual behaves for a time as another and distinctly different person, are too well known to require more than a passing notice. These are instances in which an entirely new set of functions is brought into play, — distinct from the normal, yet working together to the accomplishment of definite ends. But a few of the many modifications of normal functions have been mentioned. This is not the place to exhaust the subject. Only enough has been given to show the student that even the highly specialized functions are subject to the laws of variation. The matter will be more completely covered under " Causes of Variation." 1 Bulletin AV. g8. Agricultural Experiment Station, University of Illinois, pp. 500-502. The account of the different ways in which three different females performed their work is given in full under " Transmission of Modifications," section on " Habit and Instinct." FUNCTIONAL VARIATION I07 SECTION IV — NORMAL FUNCTIONS EXERCISED UNDER ABNORMAL CONDITIONS The mammary gland, normally confined to females, is com- monly functionless until after pregnancy ; but by manipulation of the udder, heifers and other females may be made to yield milk without bearing young. Again, rudimentary mammae, present generally in males, are occasionally accompanied by considerable development of mammary tissue, nearly always but not necessarily functionless.^ Most remarkable of all, mammary-gland tissue has been known to develop in extremely unusual places upon the body. Mammary tumors in the axilla (armpits) are described as of " common occurrence in lying-in women." ^ These tumors have no duct, but in squeezing they yield " both colostrum and milk," following in the same order as from normal mammae, and oozing through the skin " at the situations of the sebaceous follicles." Besides these there is " indisputable evidence of the presence of a mammary gland on the thigh . . ., on the cheek . . ., on the acromion (shoulder point) . . ., and in the labium majus. ... In the two last cases the mammary nature of the gland was proved by microscopic examination." ^ Similar conditions may be produced artificially by grafting, and all sorts of abnormalities may testify to the persistence with which highly specialized tissue continues to discharge its functions, often under the most discouraging circumstances."* For example. Hunter and Duhamel grafted the spur of a young cock into his comb, " where it continued to grow to its normal size." ^ " Bert transplanted the tail of a white rat to the body of M?(s deciimanus (the common brown rat), where it continued alive." ^ The same experimenter bent over the tail 1 Dr. Hottes, a personal friend of the writer, knew a young man in Germany who was suckHng an infant. 2 Bateson, Materials, etc., p. 185. 3 Ibid. p. 1S7. * As when a piece of mammary gland was grafted into the ear of a guinea pig ; when the pig became pregnant the gland commenced to secrete. ^ Morgan, Regeneration, pp. 17S-179. I08 VARIATION of a rat and grafted it back into its own body. After it had united he severed it at the normal base and thus provided the animal with a " reversed tail." He found, however, that the tail of the mouse did not grow as well in the body of the rat and would not unite at all with the body of either the dog or the cat.^ Born succeeded in uniting the anterior and posterior parts of the tadpoles of two different genera of frogs {Rana csculcnta for anterior and Bombinator igneits for posterior). The combi- nation lived for ten days, when it was killed because of patho- logical changes. 2 In the same way Harrison made up an individual of two species (Rana vircscens and Rana palnstris). This he kept alive until after its transformation into a frog, " each half retaining the characteristic features of the species to which it belongs." "^ This being true, it is not surprising that many varieties of apple can be grafted into the same tree top. Examples of this sort might be multiplied indefinitely, as in the making up of worms by grafting together pieces of two different species, in which each piece preserves its specific characters ; but enough has been given to show the persistence with which specialized tissue continues to discharge its natural function even under the hardest of conditions.^ The circumstances under which living matter can discharge its normal functions unaltered, either in character or in degree, and the limits beyond which these functions must cease or undergo alteration, — all this is not only a question of deep biological interest but it is one of special significance in breed- ing, because it throws no little light upon the real nature and causes of variability, a subject upon which we sorely need in- formation if variations are ever to be controlled either in their development or in their transmission. Summary. We are to regard variations in function as well as in form, of activities as well as of structure, of what an animal or plant does, as well as what it is. The body functions are not constant, but variable. They are variable as between different individuals and also with the same 1 Morgan, Regeneration, pp. 178-179. ^ Ibid. pp. 183-185. ^ Ibid. chap, ix, pp. 159-189, " Grafting and Regeneration." FUNCTIONAL VARIATION 109 individuals at different times. They are variable not only in degree but also in kind, and normal functions may be disturbed, even altered, by external influences. Conversely, usual func- tions may be discharged under most unusual conditions. All variation is either continuous or discontinuous, and contin- uity must not be assumed. Many of the fundamental qualities of living matter, such as definite composition, argue for discon- tinuity. ADDITIONAL REFERENCES Immunity and Adaptation. Biological Bulletin, IX, 141-151. Geldings more Susceptible to Disease than Mares. Experiment Station Record, XI, 896. Variation in Immunity to Anthrax (among sheep). By Martinet. Experiment Station Record, XIII, 186. Insectivorous Plants. By Charles Darwin. For good evidence on functional variation consult the speed records of trotting and running horses and the Advanced Registry of Jersey and Holstein-Friesian Cattle. CHAPTER VI MUTATIONS SECTION I — DISTINCTION BETWEEN MUTATION AND ORDINARY VARIATION The deviations from type heretofore considered are those of individuals rather than of groups. Whether quantitative, sub- stantive, meristic, or functional, they represent the fluctuations of individual members of a species or a variety about the nor- mal type of the race, not necessarily exhibiting any tendency to depart permanently from that type. The study of these deviations shows that, while no two individuals are alike, yet the departures of certain individuals in one direction are compensated by departures of other indi- viduals in the opposite direction. In other words, the members of a race cluster closely about what may be called a center of fluctuation, which is, in most cases, comparatively stationary. Because of this fact we may have a relatively fixed type, indicating a practically stationary race, even in the midst of considerable individual deviation.^ Mutations, on the other hand, mark sudden and distinct departures from type. The pendulum swings, but does not re- turn. A new center of fluctuation is established, from which individuals deviate in all directions as before. It is not that the old center is abandoned, — for the mass of individuals still cluster about it as before, — but it is that a new center is estab- lished, about which a new group clusters, and all close observers recognize at once that a new type has been born into the world 1 This fact is extremely confusing, especially to animal breeders. In the midst of wide variations and with but few individuals living at any one time, the breeder is often unable to tell whether his general average (or type) is improving, retro- grading, or standing still. This matter will be alluded to again under " Type and Variability." MUTATIONS 1 1 1 and a new race is established on the earth. This new group and its center constitute a mutation, and the individuals are spoken of as mutants. Variation has become discontinuous as well as continuous. These sudden offshoots from established species were noted by Darwin, who called them sports. He considered, however, that new species are formed only by the slow but continuous action of selection working with ordinary fluctuations (continuous variations), building up new types a little at a time through the gradual accumulation of slight but favorable deviations. His so-called "sports " were therefore mysterious, and from the fact that under natural conditions they generally disappear rapidly by crossing, he was led to attach little importance to these sudden departures from the established type.^ Later researches, however, have given them unexpected significance. The distinguishing feature of a mutation is that there are no intermediates between the old type and the new, which was therefore attained not by slow degrees but by a sudden leap ; that there is but a slight tendency to revert to the old form, but that if reversion takes place at all it is complete at once and the return is to the old type and not to an intermediate form. The mutant is distinctly a case of discontinuity. SECTION II — EXAMPLES OF MUTATION The classic examples of mutation are the weeping willow and the nectarine. They are to be regarded, however, as familiar illustrations of general principles widely operative and giving rise not to few but to many distinct types. When a seed germinates it puts out two sprouts. One is positively geotropic ; that is, it responds to the force of gravity and grows downward into the soil, developing the root system. The other is negatively geotropic ; that is, it grows upward 1 The student of evolution should gain the conception that the type of a race is not a fixed point from which deviations radiate ; it is rather the center of gravity of all the individuals of the race, its exact location depending upon the extent and direction of individual deviations, shifting slightly from time to time with the causes that influence variability. I I 2 VARIATION against gravity, and with an energy sufficient to maintain it in a fairly upright position, developing stems, branches, and leaves. Occasionally this latter geotropism fails, and the branches hang downward, forming a "weeping" variety. This is espe- cially common in the willow and the birch, though by no means unknown in other trees, notably the elm, maple, and beech. "Cut-leaved" varieties and "fan tops " ^ occasionally arise suddenly, all of which may be preserved by grafting or by bud- ding, so that with proper attention we may have weeping, cut- leaved, and fan-top varieties, not of a few but of many species of trees and shrubs, although the readiness with which a partic- ular mutation may appear in one species is no guaranty of its appearance in another. A tree which has always before borne peaches may suddenly bear nectarines, or more likely a single branch may make the departure, the remainder of the tree continuing to bear peaches as before. In any event the mutation may be propagated by bud or possibly by seed, — in which latter case a nectarine-bearing tree results. This tree may bear nectarines all its life, or it may occasionally bear peaches on the whole or a portion of its top. The significant fact is that there is no intermediate between the peach and the nectarine, and yet the one may arise at any time from the other.^ The mutation from peach to nectarine is clear- cut and distinct, and the reversion from nectarine to peach when it occurs is equally complete. The apricot appears to be related to the plum much as the nectarine is to the peach. In both cases the main stocks (peaches and plums) exist in many varieties, and the mutations (nectarines and apricots) in but few. In the case of the former (the peach) the main stock is downy, while the mutant is gla- brous, or destitute of downy covering. In the latter, however, the conditions are reversed, for the main stock (the plum) is glabrous, while it is the mutant that is downy. 1 A fan-top tree is one in wiiich the branches are borne on opposite sides, after the fashion of corn. In a small forest plantation belonging to the writer is a fan- top linden, now grown to considerable proportions. ■^ Inasmuch as the peach is considered as the main race, the nectarine is said to arise from the peach by mutation. Therefore when peaches are borne upon nectarine trees the case is considered to be one of reversion. MUTATIONS 1 1 3 Because the apricot has never been observed to arise direct from the phim as the nectarine has repeatedly been known to arise from the peach, and because the apricot trees have never been known to bear plums as the nectarine trees occasionally bear peaches, — because of these facts botanists have quite generally ceased to regard the apricot as a sport from the plum, and are agreed, I believe, in considering it as a distinct species. However, it behaves precisely like a mutant, and in consider- ing the means by which new types originate the presumptive evidence is strong that the apricot originally sprang from the plum stock. Though it is true that some mutations are fre- quently repeated, it is also true that others arise but rarely. The nectarine is unusual in the frequency with which it reap- pears, and the readiness with which the peach and its mutant exchange places has perhaps no parallel. In many respects the apricot appears like an intermediate between the peach and the plum. The external appearance of the fruit is that of the peach. The pit is smooth, resembling that of the plum. The bark of the tree is like that of the peach, but the leaf is like that of the plum. There is nothing to sug- gest a hybrid origin, though everything to suggest that this strange plant and its fruit are in some way composed of the elements of both the peach and the plum. Nor would a hybrid origin be at all necessary to this fact. Certain characters are general, running through many species quite independent of consanguinity. Thus the weeping or the cut-leaved habit is common to a great variety of species only remotely related. The downy character is common with both fruit and leaf, and almost every downy or pubescent species has its glabrous or smooth variety, — its mutant in all probability, and one that easily and frequently arises. So also, without doubt, the reverse is true by which smooth species occasionally throw off downy or pubescent varieties. Now this particular character of pubescence, while simple enough in itself, is yet exceedingly noticeable, and serves to insure a specific name, unless indeed the direct origin happens to be extremely well known, in which case the mutant is likely to get off with a varietal distinction. 114 VARIATION In the same general manner, color is likely to fail, and nothing is more common in nature than albino varieties. Thus we have our white currant, strawberry, raspberry, and even the black- berry, — almost every thicket affording its examples and speak- ing eloquently of the freedom with which nature creates new forms, and if we will only open our eyes to see what is going on about us, we shall learn much of how it is done. Albinism among animals is even more common than among plants. Men, dogs, cats, horses, cattle, sheep, bears, rabbits, rats, mice, and many other species are distinguished by albino varieties. These distinctions, marked though they are, arise doubtless from the simplest causes. For example, if an animal for any reason fails to secrete pigment in the normal manner it is from necessity an albino, and if the failure is hereditary an albino race is likely to be established, although unrestricted breeding greatly reduces its probability through crossing with other forms. SECTION III — EXPERIMENTS OF DE VRIES i Hugo De Vries, professor of botany in the University of Amsterdam, long ago became convinced that Darwin's theory of the origin of species through the gradual accumulation of fortuitous variations is not the only means of creating new types. Darwin taught not only that existing types had been preserved by selection because they in some way fitted the conditions of life, but that the intervening spaces between species and varieties represent extinctions through the agency of natural selection. De Vries came to believe that, in many cases at least, the new type springs suddenly from the old, without gradation and without intervening forms, and that while selection may shape up the new type and perhaps the better fit it for existence, yet the selective process is in no way responsible for its origin. Indeed, one of the earliest evidences, to his mind, that new types often arise without the agency of selection, was the notable fact that new forms arising spontaneously in nature are for the most 1 Hugo De Vries, Species and Varieties, their Origin by Mutation [Open Court Publisliing Company, Chicago]. MUTATIONS 115 part promptly exterminated by the rigors of natural selection, which therefore could not have been the chief agency in their creation. Accordingly he conceived the idea of cultivating a few unstable forms under conditions such as would protect and preserve any mutations that might arise, hoping in this way to throw some light on the origin of new types and to determine whether in the origin of species natural selection works principally upon indi- viduals or upon types. Experiments with toadflax ^ (Linaria vulgaris) . These experi- ments were designed to test the origin of the peloric form.^ The toadflax was chosen, first because the peloric form is known to have arisen repeatedly, and second because the change involved is slight, structurally speaking. These two considerations gave reason for the hope that if the species were put under careful observation and control, he (De Vries) "might be present at the time when nature produces another of these rare changes." The experiments commenced in 1886 with normal plants bear- ing "one or two peloric flowers," as is common with most indi- viduals of this genus. The roots were planted in the garden, and flowered and seeded in 1887. This second generation was grown for three years, producing in 1889^ one, and in 1890 two, peloric structures. The seeds of these were saved and produced the third generation in 1890— 189 1. Among some thousands of blossoms in this generation there was one five-spurred flower. This was pollinated by hand and luckily produced "abundant fruit, with enough seeds for the entire culture of 1892, and they only were sown." * 1 De Vries, Species and Varieties, etc., pp. 464-487. 2 The normal flowers of the toadflax are exceedingly unsymmetrical. Aside from bearing a short spur, they are described as consisting of a " two-lipped corolla, the lower lips spreading and three-lobed, with a base so enlarged as to nearly close the throat." Plants bearing such unsymmetrical flowers as do toadflax, snap- dragon, etc., are known occasionally to produce peloric, that is, symmetrical, flowers. Not only that, but peloric varieties are not unknow^n, and these experi- ments were designed to solve the manner of their origin. 3 The toadflax is a biennial. * Peloric flowers of this species are commonly sterile, but in any case are dependent upon artificial fertilization. They are by nature ill adapted to preserve themselves. Il6 VARIATION Up to this point in the experiment each generation required two years, as the toadflax is a biennial, not blooming until the second year. After this, however, the seedlings were started under glass and transplanted to the garden in June. By this means the new plants were made to produce flowers and seeds the first year. About twenty plants of this (fourth) generation were secured, and under this treatment most of them produced seed the first year. Only one peloric flower was observed, however, in the entire lot. The plant bearing this flower and one other were preserved, and all others were destroyed. These two fertilized each other freely and produced lo cc. of seed, but no more peloric flowers appeared. It is from this pair of plants, how- ever, that a peloric race finally sprung.^ In 1894 about fifty plants were in flower. There was no rea- son for considering these plants any more promising than pre- vious sowings, except that " stray peloric flowers were observed in somewhat larger numbers than in previous generations, — eleven plants bearing one or two, or even three, such abnor- malities." De Vries wisely remarks that this "could not be considered as a real advance, since such plants may occur in varying though ordinarily small numbers in every generation." However, besides these eleven individuals, each bearing one or two abnormal flowers, tJicre ivas a single plant bearing only peloric flozuers. The mutation had arisen and De Vries "was present at the time." This plant was carefully kept, all others being destroyed, and the next year it bloomed again, bearing only peloric flowers. It was true to its type. In this connection De Vries says : Here we have the first experimental mutation of a nonnal into a peloric race. Two facts were clear and simple : [first] the ancestry was known 1 It has been said that the flowers of one plant are sterile to pollen from the same plant. De Vries ascertained by careful experiment that this is true in about 50 per cent of the cases, so that, though a much higher degree of fertility exists between individuals than within the individual, absolute barrenness between all flowers of the same plant cannot be asserted. The point is not significant in the present connection, but it is important as demonstrating that fertility and sterility are not always in direct proportion to consanguinity, and that, though close breed- ing may be commonly infertile in certain strains, it by no means follows that it is always infertile even in the same strain. MUTATIONS 117 for over a period of four generations. . . . This ancestry was quite constant as to the peloric peculiarit}', remaining true to the wild type as it occurs everywhere in any country, and showing in no respect any tendency to the production of a new variety. [Second] the mutation took place at once. It was a sudden leap from the normal plants with very rare peloric flowers to a type exclusively peloric. The parents themselves had borne thousands of flowers during two summers, and these were inspected nearly every day in the hope of finding some pelorics and of saving their seed separately. Only one such flower was seen. . . . There was simply no visible preparation for this sudden leap. This leap on the other hand was full and complete. No reminiscence of the former condition remained. Not a single flower on the mutated plant reverted to the previous type. . . . The whole plant departed absolutely from the old type of its progenitors. The next object was to seek for other mutants from the same lot of seed ^ and to compare their proportion with the proportion coming true from the seed of the first mutant. Accordingly De Vries planted his entire remaining stock of seed, which, it will be remembered, was grown from the pair of plants one of which bore a single peloric flower, but both of which were immediately descended from the single five-spurred flower of the third generation. From this seed he grew about two thousand plants in well- manured soil. About 1750 of these bore flowers, and among these sixteen, or about i per cent, were wholly peloric. As these seeds were of the same generation that produced the first mutant, he concludes that the chance of a peloric mutant is not over one in a hundred. De Vries next undertook to determine whether the mutation would be repeated in another generation, for up to this point all the mutants had arisen from the same lot of seed. For this purpose he saved seeds from normal plants so isolated as to pre- vent crossing with peloric strains. In one instance he " obtained two and in another one peloric plant with exclusively many- spurred flowers," showing conclusively that mutations are itera- tive, and that the same conditions that produce one mutant will from time to time produce others altogether similar. 1 It will be remembered that the original stock of seed of this generation was ID cc, but that only enough had been grown to produce fifty plants, leaving a quantity still on hand. Il8 VARIATION New type persistent. Next he undertook to determine to what extent these mutant pelorics would " breed true," in order to compare the proportion with the previously ascertained i per cent. In this he encountered difficulty because of the high degree of sterility of peloric flowers. He had in all some twenty plants, and pollinated artificially more than a thousand flowers. Of these he says : Not a single one gave a normal fruit, but some small and rudimentary capsules were produced bearing a few seeds. From these I had 1 19 flower- ing plants, out of which 106 were peloric and the remainder (13) one-.spurred. The great majority (some 90 per cent) were thus shown to be true to their new type. Whether the 10 per cent reverting ones were truly atavists caused by stray pollen grains from another culture cannot of course be detemiined with sufficient certitude. This experiment determines not only the distinctness of the new type and the suddenness of its formation, but its essential purity as well ; for it bred true in 90 per cent of the cases, while the probability of original mutation was slight, certainly not over I per cent. The total lack of intermediate steps in the control experi- ments is significant. Their absence in nature is not less so (for if they were present as transition steps toward the formation of peloric races, they would certainly be discovered, particularly when we remember that the species is a perennial), and the con- clusion seems inevitable that the transition is abrupt, and the new type, repeatedly re-formed, is without doubt to be regarded as a true mutation. The common snapdragon, whose flowers are exceedingly un- symmetrical, also has a peloric race. " But the snapdragon is self -fertile, and so is its peloric variety," observes De Vries. These mutations are therefore much more easily preserved, and are, as we should expect, more common than in the toadflax, — so common and so distinct as, without a doubt, to give rise to real hybrids with the old form. What is true of toadflax and snapdragon is held to be true of unsymmetrical flowers generally ; namely, a strong tendency to give rise from time to time to peloric varieties, not by gradual change of the parent stock but by sudden offset, or mutation. MUTATIONS I 1 9 Experiments in the production of double flowers. ^ After remarking that mutations occur as often among cultivated as among wild plants, De Vries drops the caution that in all experi- mentation of this order hybridism must be carefully guarded against.'^ He observes, too, that white varieties seem compara- tively old, as they are common in the wild state, while double flowers are rare in the wild state and correspondingly recent, indicating their origin under cultivation, and thus making the matter of doubling a favorable character with which to conduct investigations upon mutation. In the experiments upon peloric toadflax nothing new was attempted. The object was to repeat what nature was known often to have done, but so to control conditions as to " be there " when it happened next time. In this experiment, however, De Vries determined to attempt a new mutation, — that is, to try to secure double flowers where they had never been observed in nature. He accordingly chose the corn marigold (ClirysantJievmvi scget?im), common in the grain fields of central Europe, and its cultivated variety, grandi- Jioniin. The number of ray florets is variable in both, but is, on an average, thirteen in the wild and twenty-one in the culti- vated. This indicated the latter as the more favorable for the experiment, and it was therefore chosen; but it is far from pure, for many of its heads have as few as thirteen rays. Only six out of the first lot of three hundred plants reached an average of twenty-one, and these were selected as the foundation. The seeds of each of these were sown separately. Five gave proof of being still mixtures with the wild form and were re- jected. The offspring of the sixth plant averaged twenty-one ray florets, and after counting some fifteen hundred heads the two plants were selected whose secondary heads made the best showing. The progeny of these plants also averaged twenty-one, 1 De Vries, Species and Varieties, etc., pp. 489-515. 2 If a new form is a mutant it will " breed true " to itself in the great majority of cases and perforce hybridize with the original stock. If, on the other hand, it is an ordinary hybrid, it will not breed true, but will observe the principle of Mendel's law (to be discussed later), by which a certain definite percentage is of the original types. Thus it is comparatively easy to ascertain whether a new type is the product of a single race by mutation or of two races by hybridization. 1 20 VARIATION and De Vries considered that the strahi was now pure and that " no further selection could be of any avail." One of these two plants was distinguished by producing two secondary heads with tiveiity-tivo rays, whereas generally only the terminal head reached so many as tzventy-one, the other retrograding often as low as to thirteen. This exceptional plant was distinguished only by these two secondary heads. Its ter- minal head had but twenty-one rays, and the average of all its heads was not exceptionally high ; but no other plant out of hundreds had ever produced secondary heads with more than twenty-one rays, and it was from this plant that the double- flowering line developed three years later. This plant appeared in 1896. Its seed was sown in 1897. The largest number of rays in the terminal head suddenly increased from twenty-one to thirty-four; next year (1898), to forty-eight ; next (1899), to sixty -six ; and during this time the general average for all the heads increased remarkably. No indication of doubling had, however, yet appeared. The im- provement was such as follows selective breeding with fluctuat- ing variability, — improvement by gradual change and without mutations. Late in the season (September) of this year (1899), how- ever, three secondary heads appeared on one plant with a few ray florets scattered over the disk. The mutation anxiously awaited for seven years had suddenly appeared in this small, belated way toward the close of the growing season, and in a manner that would have escaped the attention of any but the most painstaking investigator,^ and that would have invited extermination in nature. This was in 1899. The heads were of course pollinated with other and inferior flowers, but in 1900 the highest number of rays rose to one hundred, and in 190 1 reached two hundred. He remarks, " Such heads are as completely double as are 1 The student will note that every flower of thousands of plants was carefully examined, and that in every case the foundation of the mutation was in an incon- spicuous plant, certain to be overlooked by casual observers. The obvious lesson is that only the most careful and systematic examination will detect the founda- tion stock, so easily does it escape notice in the general mass and so readily is it lost unless isolated and protected. MUTATIONS 121 the brightest heads of the most beautiful double commercial varieties of composites." He adds : The race has at once become permanent and constant. Real atavists or real reversionists were seen no more after the first purification of the race. It has of course a wide range of fluctuating variability (considering all the heads), but the lower limit has been worked up to about thirty-four rays, a figure never reached by the grandijlonnn parent, from which my new variety is sharply separated. Unfortunately, the best heads now produced are sterile, so that seeds must be secured from inferior stock and the variety must be propagated from slightly inferior parentage. Selection has, therefore, reached its limit, unless a fertile strain arises, which is entirely possible. This mutation is decidedly new. It had never been known, nor had anything approaching it ever been discovered in this species. The only hope that it might appear was belief in the principle, and the fact that doubling had taken place in other compositae. Right royally was De Vries's prophecy fulfilled, and again was he " present " when it happened ; not only that, but in this case nature evidently would not have produced this mutation without assistance. Here nature has accomplished with help a work which she was powerless to accomplish alone, but abundantly able to achieve with a little assistance. Experiments in the production of new species.^ De Vries was not content with the simple production of varieties. He desired to show that the principle of mutation produces species as well.^ He cultivated many species of wild plants in his garden, choos- ing wild in preference to cultivated, because he regarded the latter as evidence of what had recently taken place and, there- fore, not the best stock for further mutation in the near future. In other words, he desired to be present before the mutation 1 De Vries, Species and Varieties, etc., ])p. 516-546. 2 He distinguishes sharply between varieties and species. The variety differs from the main stock in but a single character, progressive or retrogressive, while the species differs in all characters, some of which are perhaps progressive and others retrogressive. He likes to distinguish elementary species from all other types, as these are in his estimation the most stable forms in nature ; and when any race assumes the " mutative state " it is likely to throw off, if conditions are favorable, a large number of new elementary species, each with its new center of variability. 122 VARIATION happened, rather than to enter just afterward and only in time to note results with no evidence as to methods. Of all the hundreds of plants cultivated he foimd the evening primrose most fertile in distinct strains, both in the wild and in the cultivated state. Other species gave rise to varieties freely, but no other appeared to be sufficiently mutable to give rise freely to what he regarded as elementary species. Of all the primroses, GinotJicra Laviarckiana, commonly called Lamarck's evening primrose, was the most prolific in distinct forms, and accordingly this was chosen by De Vries for special attention in his experiments. It is described as "a stately plant with a stout stem, attaining often a height of 1.6 m. or more. When not crowded, the main stem is surrounded by a large circle of smaller branches growing upward from its base so as to form a dense bush. These branches in their turn have numerous lateral branches. . . . Contrary to their congeners, they are dependent on visiting insects for pollination." Ordinarily this primrose is a biennial, producing rosettes in the first year and stems in the second year. Both the rosettes and the stems are highly variable in nature, producing a num- ber of distinct races, some of which show a marked ability to hold their own under natural surroundings, while others are too weak to endure. Many of these De Vries regarded as new species. Experi- ments to determine this point were commenced with stock dis- covered near Hilversum, and three plans were followed : first, to transplant the apparent new species into the garden whenever the new race was sufficiently strong; second, to reproduce weak races by sowing seeds from "indifferent" plants growing wild ; third, to sow the seeds from the introduced plants. " These various methods," he adds, "have led to the discovery of over a dozen new types never previously observed or described." These new plants are divisible, according to De Vries, into five different heads or " groups " : (i) those that " are evidently to be considered as varieties in the narroivcr sense of the word," representing retrograde development ; (2) " progressive elementary species" which are "as strong as the parent species " ; (3) " progressive elementary species," but weaker than MUTATIONS 123 the parent, and "apparently not destined to be successful"; (4) certain forms that are "organically incomplete"; (5) "in- constant forms." Group (i), retrograde varieties. Of this class the following three forms were discovered, all produced in nature as well as in the garden : O. Icevifolia, the smooth-leaved variety, constant from seed and never reverting except from crossing. As strong and fertile as the parent. O. brcvistylis, the short-styled form. In this the ovary is so placed that it is reached by very few pollen tubes. Thus while the plant is vigorous it is but indifferently productive of seeds, and as De Vries says "many [capsules] contained no seeds at all ; from others I have succeeded in saving only a hundred seeds from thousands of capsules." These seeds, however, reproduce the variety without reversions to Lamatxkiana. O. nanella, the dwarf, "a most attractive little plant, . . . very short of stature, reaching often a height of only 20-30 cm., or less than one fourth of that of the parent." The flowers are as large as those of the parent ; the leaves are much smaller and with no reversion in seedlings, even in repeated and successive generations. Group (2), progressive elementary species, and vigorous ; two forms discovered : O. gigas, the giant, deserving its name not from being higher than its parent, but because it is " so much stouter in all respects." The stems are often twice as thick as in the parent {Lamarckiana), and the " internodes are shorter and the leaves more numerous, covering the stems with a denser foliage." The flowers are larger, and the seed capsules are smaller and filled with fewer but larger seeds than in the parent plant. It has a strong tendency to remain a biennial. O. rnbrinei'vis, the red-veined form. In this the veins of the leaves are distinctly tinged with red and the fruits are streaked with red. The plants are in many ways a counterpart of the giant, except for the red tinge and distinctly lighter foliage. This latter probably accounts for the marked tendency on the part of this form to become an annual. Like the giant, this form 124 VARIATION is true to type when grown from the seed, and its recurrence is far more common than is that of gigas, which is extremely rare. Group (3), progressive elementary species, but weakly. Two forms : O. albida, the albino, with whitish, narrow leaves, "appar- ently incapable of producing sufficient quantities of organic food." The seedlings are exceedingly delicate, and if left to themselves will be speedily overgrown by their more vigorous neighbors ; but if transplanted and given the best of care, they make fairly vigorous plants the second year, comparing fairly well with the parent stock but bearing fewer seeds. They come true even to the third generation and the type remains distinct. O. oblonga, the narrow-leaved form. It " may be grown either as an annual or a biennial. In the first case it is very slender and weak, bearing only small fruits and few seeds. In the alter- native case, however (biennial), it becomes densely branched, bearing flowers on quite a number of racemes and yielding a full harvest of seeds." The investigator says : We have now given the description of seven new forms which diverge in different ways from the parent type. All were absolutely constant from seed. Hundreds or thousands of seedlings may have arisen, but they always come true and never revert to the original O. Lainarckiaiia. He adds the remark that they have inherited the condition of mutability to some extent and are evidently themselves able to produce new forms, but that they do so but rarely. Two other forms belong to this group, — O. semilata and O. leptocarpa, — but their characters do not merit special description. Group (4), forms organically incomplete : O. lata is a pistillate variety, wholly dependent for fertiliza- tion upon other forms, and it had therefore no opportunity to establish its type, which, however, freely appeared. It is a "low plant," but with "dense foliage and luxuriant growth." Its presence can be detected in the seedling by the " broad, sinuate leaves with rounded ^ tips." Being pistillate, it produces seed only when cross pollinated, in which case its characters are MUTATIONS 125 transmitted to a portion only of its offspring, thus behaving hke hybrids. Indeed, he specifies that " on the average one fourth of the offspring are lata," the others assuming the character of the pollen parent," — a strict example of hybridism between a weaker and a stronger form, according to Mendel's law. Group (5), inconstant forms : O. scintillans is a perfectly fertile form, bearing smooth, dark- green leaves with glistening surfaces. It is a natural dwarf, easily cultivated as an annual. When fertilized with its own pollen to produce a " pure " strain, it is found that the seedlings all resemble the parent, but that soon afterward they diverge into various types. Some of these resemble the original parent stock {Laniai'ckiana) and others remain pure, but the proportion is very variable. These might be regarded as simple reversions, except that occasionally other types appear, especially oblonga, lata, and nanella, the first often constituting 10 per cent of the sowings. It thus shows a disposition to give rise to the same distinct forms as does its own parent, and is thus regarded by De Vries as being itself in a " highly mutable state." O. clliptica is a narrow-leaved, inconstant type, exceedingly "difficult of cultivation." Though fertile to its own pollen, it " repeats its type only in a very small proportion of its seeds." There are thus " a dozen new types springing from an original form in one restricted locality and seen to grow there, or arising in the garden from seeds collected from the original locality." Most of these types behave with a constancy that ranks them, for breeding purposes at least, as distinct forms, good elementary species, — new things in the earth that may be held constant or that may be slightly modified by the exercise of selection among the fluctuations to which all types both old and new are subject. The experimenter observes : It is most striking that the various mutations of the evening primrose display a great degree of regularit}-. There is no chaos of forms, no indefi- nite varying in all degrees and in all directions. On the contrary', it is at once evident that very simple rules govern the whole phenomena. History of the experiment. In all De Vries made four differ- ent series of pedigree cultures of the evening primrose, extend- ing from five to nine generations and including thousands of 126 VARIATION plants. The types that arose at different times have already been described, but considerable interest and no Httle profit attaches to the details of the experiment, especially in re- gard to the order and manner of the appearance of the new types. The following is abstracted from the experimenter's account of one of these four experiments, running through eight generations.^ Beginning in the fall of 1886 he took nine large rosettes of O. Lamarckiana from the field and planted them in the garden. The second generation was sown in 1888 and flowered in 1889. The seed produced fifteen thousand seedlings, of which ten were divergent at once, — five lata and five nanella. No inter- mediates appeared. "They came into existence at once," says De Vries, "fully equipped, without preparation or intermediate steps. No series of generations, no selection, no struggle for selection was needed. It was a sudden leap into another type, — a sport in the best acceptation of the word." ^ The third generation of ten thousand plants showed three lata and three nanella, besides one riibrinervis. Growing expert in detecting mutants at an early stage, he discovered 334 young plants out of 14,000 of the fourth gener- ation (1895). This is about 2.5 per cent. Of these 176 were oblonga, 73 lata, 60 nanella, 15 albida, 8 riibrinervis, i scintil- lans, and i gigas. The larger number and wider range of mutants discovered this year are to be ascribed to growing skill in detecting them at an early age. Manifestly such immense numbers must be greatly reduced at the earliest possible date, and without doubt some good forms were overlooked in the earlier generations. After this (fourth) generation the number of seedlings was greatly reduced, with the effect of reducing the number of mutants and also the chances of the rarer forms appearing at all ; indeed, ^z>^?j- never appeared again, and scintillans not after 1 De Vries, Species and Varieties, etc., pp. 549-556. 2 It may occur to the student to object to the conclusion on the ground that the parent stock taken from the field may not itself have been pure. If, however, the stock had been in any sense hybrid, the departures should have been, accord- ing to Mendel's law, more than ten ; but not in this or in later generations did either parent stock or mutant behave like a hybrid in this respect. MUTATIONS 127 the sixth generation. The entire results of the eight generations of breeding are given in the following table. Eight Generations of a Mutating Strain of Evening Primrose (^O. Lamarckiana) Genera- tions 0. gigas Albida OMonga Rjibri- nervis Lamarck- iana Natiella Lata Scintil- latis I 9 II 15,000 5 5 III I 10,000 3 3 IV I 15 176 8 14,000 60 73 1 V 25 135 20 8,000 49 142 6 VI I I 29 3 1,800 9 5 I VII 9 3,000 II VIII 5 I 1,700 21 I In the opinion of the experimenter here are numbers enough and types suflficiently distinct to warrant the enumeration of certain laws or principles that appear to govern the appearance of mutants, especially in the species under observation. This De Vries attempts to do, but without presuming to say how closely they may apply to other strains of plants or animals. Laws of mutability for evening primroses. De Vries' experi- ments. On the basis of his experiments with the evening prim- rose the investigator announces the following laws of mutability as applying to that species :^ 1. "That new elementary species appear suddenly, without intermediate steps." As proof he points out that no interme- diate forms appeared to fill the gaps, and that no selection was necessary to establish the type. 2. "New forms spring laterally from the main stem." " The current conception concerning the origin of species (or new forms generally) assumes that species are slowly converted into others. The conversion is assumed to affect all the individuals in the same direction and in the same degree. . . . The birth 1 De Vries, Species and Varieties, etc., pp. 558-575. These laws, while announced for the evening primrose, are without doubt of wide if not general application. 128 VARIATION of a new species necessarily seemed to involve the death of the old one," at least the old merged into the new. The experimenter points out, however, that through all the process of originating a dozen or more distinct forms, the parent stock continued unchanged, and still constituted the principal strain of all the primroses,^ and from this he deduces the law that mutants are laterals. 3. " New elementary species attain their full constancy at once." " Constancy is not the result of selection or of improve- ment. It is a quality of its own. It can neither be constrained by selection if it is absent from the beginning, nor does it need any natural or artificial aid if it is present." De Vries remarks that scintillans repeats its characters in but part of its offspring, and that he has " tried to deliver it from this incompleteness of heredity but in vain. . . . The insta- bility seems to be here as permanent a quality as the stability in other instances. Even here no selection has been adequate to change the original form." He regards it as itself in a state of instability. 4. " Some of the new strains are evidently elementary species, while others are to be considered as varieties." Elementary species are regarded as possessed of progressive characters, but varieties as differing from their parent stock in but a single character, and that in the way either of an assump- tion or of a loss. The elementary species is, therefore, a new aggregation of characters, while the variety is simply the old form minus a single character. Whether this distinction holds, remains to be determined. Much of the argument turns upon what is to be considered as a character and when it is lost. For ^ A natural corollary to this observation is to remark upon the erroneous popu- lar assumption that of similar and contemporaneous forms the more primitive are necessarily the progenitors of the more nearly perfect. For example, it is hastily assumed that if evolution is true then man must be the direct descendant of the ape. But the ape, though very old, is still an ape, and he is not descending into anything but apes. Though evidently developed from the same original stock at some time and in some way, whether by one or by many mutations nobody knows, yet the gap between us is evidently fixed and not growing less or being bridged at any point. Good evolution regards related forms as connected by ties of con- sanguinity, but whether direct, or, what is more likely, indirect, running to some extinct common ancestor, only a novice will attempt to say. MUTATIONS 129 example, is the smooth leaf or stem considered as having lost a character as compared with its downy relative ? 5. '* The same new species are repeatedly produced," that is to say, the same new forms arise again and again, showing that the tendency to their production is inherent and persistent. " This is a very curious fact," remarks De Vries. " It embraces two minor points, — the multitude of similar mutants in the same year, and the repetition thereof in succeeding generations. Obviously there must be some common cause. This cause must be assumed to lie dormant in the Lamarckians of my strain, etc. . . . The germs of the oblonga, lata, and nanella are very irritable and are ready to spring into existence at the least summons, while those of gigas, rubriiietins, and scintillans are far more difficult to arouse." May not the same be true in nature generally, and may not the same strain arise again and again, commonly fail- ing to persist because as a rule all conditions are against it } 6. Mutability is distinct from fluctuating variability. Darwin regarded the new type as built up by the operation of selection upon fluctuating variability, establishing a new type by the gradual accumulation of favorable variation, all others (inter- mediates) being exterminated. De Vries regards mutability as distinct from fluctuating variability, and considers that he has presented experimental evidence to show that it is entirely com- petent to give rise to new forms suddenly, without intermediates and without the aid of selection. He of course believes that all types, both old and new, are subject to fluctuating variability, and that through selection some improvement is possible, but that this is not the sole or principal method of securing new types. 7. " The mutations take place in nearly all directions." Some are larger, some are smaller than the parent ; some stronger, others weaker; some plainer, others more brilliant. The species is not, therefore, drifting ; it is sending out new types from all sides. SECTION IV — AMERICAN EXPERIENCES The experiments of De Vries are strongly confirmed by the experience of breeders, especially in the production of new varie- ties of fruits and vegetables. Many of these have been so long I30 VARIATION under cultivation that nothing is known of their origin. Of others, on the contrary, the hfe history is well known. When Europeans peopled America they naturally brought with them their fruits, their vegetables, their grains, their grasses, and their domestic animals. The new country was rich in native species, both plant and animal, but the European species had the advantage of being better known and better adapted to the special needs of man. Accordingly, wherever the introduced varieties succeeded, the corresponding native types were neg- lected ; but when the European varieties failed, then the natives were developed. It is from this latter class that some important observations may be made.^ The gooseberry .2 The large English gooseberry was too tender for the American climate, and withal was exceedingly liable to mildew. Native varieties flourished widely in the forests. Unfor- tunately the varieties bearing the largest berries were exceedingly thorny, both on bush and fruit. Side by side, however, with these prickly sorts were smooth varieties, free from "prickers" both on fruit and bush. These were freely transplanted to the gardens of the pioneers and furnished an acceptable fruit. ^ In good time they developed improved sorts, — first the Houghton, a seed- ling originated by Abel Houghton of Lynn, Massachusetts, some- time in the early forties. Then came the Downing, a seedling of the Houghton, first described in 1853, the fruit of which is said to be " the largest yet known, being about twice the size of the Houghton's seedling, its parent ; it is pale or light green, without any blush, and smooth. The skin is very thin and the fruit as delicate and tender as any European gooseberry on its native soil. The flavor and aroma are perfect." Bailey observes in this connection, " This berry, now known as the Downing, is the standard of excellence in American gooseberries, and is probably grown more extensively than all other varieties combined ; and yet it is only two removes from the wild species." 1 L. H. Bailey, Evolution of our Native Fruits [The Macmillan Company, New York]. 2 Ibid. pp. 389-399- ^ The writer remembers very well as a boy searching the woods, and espe- cially the swamps, of Michigan for these smooth varieties for transplanting. MUTATIONS 131 Is not this a natural mutation in the truest sense of the term ? If not, then it is merely a question of terminology and definition. The fact remains that it arose suddenly, a distinct type, and remains true with no characteristics of a hybrid. We need a term for this sort of thing which is occasionally occurring everywhere in nature, in our gardens and in our herds, and I know of none better then the one already in use — mutation. The strawberry.^ The wild strawberry grew everywhere in northern North America. There were not only many distinct types of the red, but, like the native raspberry and the blackberry, it had everywhere its albino race. Good progress had been made in the cultivation of the native strawberries, and without doubt good varieties would in time have developed ; but the introduction of the Chilean berry (the parent of most present varieties) seems to have put a stop to this. The most promising of all native strains was the Fragaria Chilocnsis, a native to Oregon and the Pacific coast ; but, as Bailey observes, " the garden progeny of its South American branch is already so good that there is little reason for returning to the wild for a new start." Here is a curi- ous instance of the successive supplanting of varieties. European sorts were vanquished by developments of New England natives. Then the wild type of Oregon came into the struggle and threatened to supplant them both, for it was full of promise. But its prosperity was its own defeat, for its own Chilean brother has now supplanted everything in that it is the stock which is furnishing our improved varieties. Any student of this subject will recognize the comparative readiness with which these new types spring up. The blackberry.- The blackberry grows wild l^oth in America and in Europe, but is .said to be cultivated only in North America. It is not more than fifty years since improved varieties were introduced, and its real cultivation dates only from about 1875. There are two principal types of the wild blackberry growing in the northern United States: (i) the "high bush," long and luscious, loving the shade, represented in its cultivated types, according to Bailey, by the Taylor and the Ancient Briton ; 1 L. H. Bailey, Evolution of our Native Fruits, pp. 424-432. 2 Ibid. pp. 29S-330. 132 VARIATION (2) a smaller variety growing in sunny, open places and bearing small, round, loose-grained fruits, ripening late and exceedingly sour. This type is represented in cultivation by Lawton, Kittatinny, Snyder, Agawam, Erie, and others. Neither of these yielded readily to cultivation and restraint, and this fact served in an early day to earn an evil reputation for what Professor Card calls this " gypsy of the fruits." Nevertheless, they yielded to persistent efforts, and have given rise, as Bailey puts it, " to a host of varieties . . . very many of them wildings, or chance bushes found in fence rows." The first-named variety was the Dorchester, introduced about 1 84 1. Its exact origin is unknown, though its originator (prob- ably Captain Lovett) is known to have transplanted wild plants for many successive years. Whether this first civilized gypsy was a sport or simply a strain improved by selection is not now capable of proof, and yet its constancy is good presumptive evidence. Wilson's Early was known in 1854, the Holcomb in 1855, and in 1857 the Lawton (first called New Rochelle) was introduced, being at once declared superior to the Dorchester. Of these the Wilson was "discovered in the wild about 1854 by John Wilson of Burlington, New Jersey"; and the Lawton, formerly New Rochelle, " was found in the town of New Rochelle, New York, by Lewis A. Secor." These two strains hav^e given rise to numerous distinct modern varieties. The "loose-cluster" strains are regarded by horticulturists as the descendants of the Wilson. The origin of certain other varieties seems to be as follows : In 1870 Mr. William Parry, of New Jersey, "selected a healthy young Dorchester and planted it in the same hill with a strong, healthy Wilson's Early (for breeders), located far away from any other blackberries." ^ In 1875 the seed from some of the largest berries growing on the Wilson were planted. One plant only was regarded as valuable, and all others were destroyed. This new strain was named Wilson Junior. The fruit was " large, early, and very fine," and so prolific that in 1884 "one acre yielded i loi- bushels of fruit, by the side of five acres of Wilson's Early in the same field, with the same culture, which averaged 1 Bailey, Evolution of our Cultivated Fruits, p. 316. MUTATIONS 133 but 53 bushels." The Eureka was produced in exactly the same way in 1877. In 1879 Rioter and Farmer's Glory were also produced from berries growing on the Wilson, and Gold Dust and Primordian from berries growing on the Dorchester. The Gold Dust was remarkable for the fact that its entire crop ripened within a period of four days. It was th-us distinct from all other blackberries in at least one important character. The Sterling Thornless arose as a chance seedling of the Wilson on the farm of John Sterling at Benton Harbor, Michigan. It is, as its name indicates, destitute of thorns, and is a distinct mutation, to be carefully distinguished from other strains of thornless blackberries, which, according to Bailey, are " specific- ally distinct from the common bush blackberry." Plums.i According to Bailey not a single commercial variety of plum has ever originated from the native stock of New Eng- land, New York, Pennsylvania, or Michigan. This is partly because the European sorts thrive well and partly because the natives of this region " are less prolific of large-fruited forms than those farther west." Some excellent varieties have arisen, however, from native stock elsewhere. The Miner was produced from seed of native stock planted in 18 14 by William Dodd in Knox County, Tennessee. The Robinson was a seedling from North Carolina stock. Wayland " came up in a plum thicket in the garden of Professor H. B. Wayland of Cadiz, Kentucky," and was introduced about 1876. The Missouri apricot was found wild in Missouri. The Golden Beauty was found in the same way in Texas, the Pottawattamie in Tennessee, and the Newman in Kentucky. The Wolf originated from seed gathered from wild trees in Iowa. The Pottingstone was found on the bank of Potting- stone Creek, Minnesota, and the Quaker was found wild in Iowa. Literally scores of well-defined varieties have arisen from native stock. It would be too much to say that none of these are hybrids. Undoubtedly many of them are the product of crossing, but this origin cannot be consistently claimed from chance seedlings found in a thicket of ordinary wild stock. Mutation, whatever 1 Bailey, Evolution of our Native Fruits, pp. 170-226. 134 VARIATION it is, must be credited with having produced many new forms spontaneously. Grapes.^ " North America is a natural vineyard," says Bailey, and yet with the most skillful and persistent attempts to culti- vate the European varieties for wine making, they have not suc- ceeded. Under these circumstances nothing is more natural than that valuable native varieties should arise, providing the capacity was inherent in the species. John Adlum wrote, about 1823, "The way is to drop most kinds of foreign vines at once and seek for the best kinds of our largest native grapes." He is to be remembered for the intro- duction of the famous Catawba, which was " found wild in the woods of Buncombe County in extreme western North Carolina in 1802." The Catawba is, therefore, almost certainly a sport of the wild grape growing in profusion in that region. In 1843 came the Diana, a seedling of the Catawba. In 1840 Mr. E. W. Bull bought a house in Concord. Some seedlings of wild grapes sprang up about it, one of which fruited in 1843. It was so excellent in quality that all others were destroyed and the new variety was named the Concord. This seedling has since given us the Worden, Moore Early, Pockling- ton, Eaton, and Rockland, of which the first has long been famous. The Concord, itself a mutant, seems to have been peculiarly rich in possibilities for still other races. " In the year 182 1 Honorable Hugh White, then in the junior class in Hamilton College, New York, planted a seedling vine in the grounds of Professor Noyes, on College Hill, which still remains, and is the original Clinton." These are only a few of the many varieties of grape of Ameri- can origin, tracing directly to wild native stock. Lost possibilities. Had other domesticated plants and animals brought from Europe succeeded less admirably, what enrichment might have come through the native flora and fauna of America ! The prairie chicken would have been improved if the domestic hen had not succeeded. The turkey was a new thing and was therefore seized upon. The buffalo would not now be extinct 1 Bailey, Evolution of our Native Fruits, pp. 1-126. MUTATIONS 135 if cattle had acclimated less successfully. Native grains other than maize would have been developed had it not been for this competition, and native grasses have not lived up to their possi- bilities. This is through no fault of theirs, though we still lack " the best American grass." SECTION V — ECONOMIC SIGNIFICANCE OF MUTATIONS Because of the waywardness of sports, — the impossibility of predicting their appearance, the readiness with which they disappear when interbred with the parent stock, and their very frequent inability to reproduce at all, — because of all these con- siderations it has become fashionable to declare sports in general to be of slight economic importance and unworthy the breeder's serious attention. The only course left open for improvement is, therefore, the slow one of gradual accumulation through selec- tion of minute but favorable variations, according to the theory of Darwin. The best of evidence exists, however, for believing that this is a hasty and unwarranted conclusion, and that many, if not indeed most, of our really valuable new types have arisen sud- denly as mutations and not gradually through infinitesimal differ- ences, as is commonly supposed. The experiments of De Vries and the American varieties of fruits both come near enough to the origin of types to more than warrant this view of the situa- tion and to afford ground for the greatest hope that unsuspected possibilities still exist in many if not most domesticated species, — possibilities of spontaneously giving off varieties representing essentially new combinations of the characters of the species and consequently possessed of different and perhaps enhanced eco- nomic value. The work of Luther Burbank^ and of our commer- cial seedsmen add confirmation to this hope, which, if well founded, promises new methods in breeding and vastly increased possibilities for improvement. The small numbers involved in animal breeding reduce enor- mously the chances of mutations appearing ; and yet nearly every 1 W. S. Harwood, in The Century Magazine, March and April, 1905 ; also New Creations in Plant Life [The Macmillan Company, 1906]. 136 VARIATION species has thrown off its albino variety, which in most cases is easily propagated. Hornless cattle occasionally appear in nearly all breeds, and the type is comparatively easy of preservation. It is more than likely that the different types in the larger breeds, which breeders find so difficult to break up, are in reality quite distinct. In future chapters dealing with the measurements of variation and the statistics of heredity in general, it will appear that even fixed types afford sufficient deviation to keep a breeder busy with selection ; in other words, that the animal breeder dealing, as he is, with small numbers will always find sufficient variation to lead him to suppose that he is getting results of his selection even when he has not shifted the center of variation the slight- est. Much that passes for breeding is nothing more than this ineffectual multiplication, and it is not too much to say that hundreds of breeders and thousands of animals have lived and died without affecting the breed in the slightest. The writer is strongly of the opinion that while selection is a powerful agent for "shaping up" and "finishing off" a fairly acceptable type, and while it is the only means of deciding what shall live and what shall disappear, yet much of the real advance in both animal and plant breeding is likely to come through distinct offsets which are now called mutations, and which in Darwin's time and until recently were erroneously, if not reproachfully, denominated " sports." SECTION VI — BIOLOGICAL SIGNIFICANCE OF MUTATIONS Too much mystery has surrounded this matter of sports, and there has been a too ready tendency to evoke the aid of latent characters to explain this and almost every other unusual phase of evolution. In truth, there is no more mystery about mutations than about heredity in general, which is a complication of mysteries. It is not a question of latency but of relative prominence of characters, of the possible loss of a racial peculiarity, or, what MUTATIONS 137 is more likely, a new combination of the elements out of which characters are made up. Every new being is the result of a new combination of racial faculties transmitted from two family lines, possibly differing in essential particulars. This new combination is certain to throw some characters into prominence and others into the background, and results occasionally in strikingly new effects. This is usually the case in hybridization, but it follows in less degree in ordinary reproduction, which differs from hybridiza- tion more in degree than in kind. Again, many characters, though exceedingly noticeable, rest after all upon a comparatively simple basis. Such, for example, is pubescence in plants, which depends upon the activity or non-activity of a few cells in developing a hairy growth. Nearly all species present both forms, — the one in which the character is present, and its opposite in which it fails to develop. Simi- larly, almost any character may fail, giving rise to a distinctly new creation. If the failure is not at a vital point it may be transmitted, in which case a new type has arisen. The origin of a new type by the addition of a character is, biologically considered, much more complicated and dif^cult of understanding; yet even this is not beyond some degree of comprehension. The probability is that what we call racial characters are less complicated than we may at first suppose. The unlearned savage could scarcely believe that the almost infinite variety of colors of natural objects are due to different combinations of very few primaries. The effects produced by three-color printing are almost beyond belief, yet we are fully advised as to the real basis for all these variations ; while the effects are striking, the means are simple. So it is, we may imagine, in the ultimate make-up of what we call racial characters : their elements are doubtless fewer than we have supposed, and the possibilities of their combinations and recombinations are greater than we have hitherto imagined. Whether all possible combinations of these elements actually take place we do not know, but all facts go to show that they occur in great variety, the most striking and permanent of which we call mutants. 138 VARIATION If a new race is produced by hybridization, then a new com- bination of characters has been effected, and it is fair to assume that the combination is richer in possibihties and possesses a larger number of characters than did either parent. Mutation teaches that new assortments of characters may take place, in some cases at least, without hybridizing. If a racial character, as color or hairiness, is lost, we recognize the new type and name it as a new creation. It may be more valuable to us than its parent, but it must be recognized biolog- ically as having lost something to which it was racially entitled. Again, if all normal characters acquire an unusual develop- ment, relatively or absolutely, as in giants, or if their develop- ment is abnormally arrested, as in dwarfs, we again recognize the new departure, and it is a good mutation. Still again, if certain characters only undergo change in devel- opment, while others remain normal, then relative values are changed, the effect is altered, and we recognize a different type. This, too, is a good mutation, provided the new relation persists. All these changes can be worked with the normal characters of the race, without the introduction of new characters or even the supposititious aid of latent characters. Soberly considered, these changes are none other than the student of biology would expect, unless indeed racial characters are bound together rnuch more rigidly than present evidence would lead us to suspect. Summary. Not all variations are continuous and connected with the type by insensible differences. Some deviations are dis- continuous, with a tendency for future variations not reverting to the main type but clustering about a new center of variability, thus setting up a new type. Such a deviation is called a mutant, and new strains may arise in this manner, as well as by the slower Darwinian method of heterogeneous variation, out of which new types are established by the slow process of selection. Both the experience of breeders — especially with new varie- ties in America — and numerous instances of experimental evi- dence show conclusively that new strains not only may, but in actual practice do, originate in this manner, suddenly and com- pletely, without any apparent preparation and with little tendency to revert to the original or main type, which continues as before MUTATIONS 1 39 Mutants, like their parent types, are subject to fluctuating variability, which is a necessary law of reproduction, and they may be improved — that is, shaped up — by judicious selec- tion, but their principal characters and main trend were fixed when the type arose. ADDITIONAL REFERENCES Alijixism. (A critical study of its causes.) By E. Pantanelli. Experiment Station Record, XV, 55. Atavic Mutation of the Tomato. By C. A. White. Science, XVII, 76-78, 234-235. Atavism in the Potato. By S. Rhodin. Experiment Station Record, XI, 710. Determinate Mutations. (De Vries and others quoted.) By M. M. Metcalf. Science, XXI, 355-356. Evolution and Adaptation. By T. H. Morgan. Science, XIX, 221-225. Evolution without Mutation. By C. B. Davenport. Science, XIX, 215. Inheritance of Monstrosities. (Experiments of -twelve years.) By Hugo De Vries. Experiment Station Record, XI, 546. Mutation and Selection. (What causes mutations.? Are they all in one direction?) By M. M. Metcalf. Science, XIX, 75-76. Mutation in the Tomato. By C. A. White. Science, XIV, 841-S44. Mutation Theory. (A review of Species and Varieties.) By C. B. Daven- port. Science, XXII, 369-372. Mutation Theory of De Vries. (Twenty-eight lectures by the author at the University of California, 1904.) Experiment Station Record, XVI, 745- Mutation Theory of De Vries. By D. T. McDougal. Experiment Station Record, XIII, 324-619; XIV, 226, 526. Mutation Theory of Organic Evolution. (A brief but pointed sur- vey of the subject.) By W. E. Castle of Harvard University. Science, XXI, 521-543; from standpoint of animal breeding, 521-524; from standpoint of cytology, 525-528. Mutations in Plants. By D. T. McDougal. American Naturalist, XXXVII, 737-770; also in Experiment Station Record, XVI, 23. Origin of Species. By Hugo De Vries. Science, XV, 721-729. Origin of Species through Selection contrasted with their Origin through Appearance of Definite Varieties. By T. H. Morgan. Popular Science Monthly, LXVII, 54-66. Prepotency of Individuals with Abnormal Variation or Muta- tion. (A study of cats with extra toes.) By H. B. Torrev. Science, XVI, 554-555- I40 VARIATION. Some Causes of Saltatory Variation. By C. H. Eigenmann. Pro- ceedings of the American Association for the Advancement of Science, 1900, XLIX, 207. Sports. (Author concludes there are other laws than Mendel's and Galton's.) By C. B. Davenport. Science, XIX, 151 ; also in Experiment Station Record, XV, 753. Sports on Grapevines. By J. C, Talback. Experiment Station Record XV, 478. Sports; the Peach-Nectarine. Journal of the Royal Horticultural So- ciety, XXVI, 596-598; also in Experiment Station Record, XIV, 45. The Mutation of Lycopersicum. By C. A. White. Popular Science Monthly, LXVII, 1 51-162. The Mutation Theory'. (A defense.) By Thomas L. Casey. Science, XXII, 307-309. The Orkjin of a White Blackberry. By Luther Burbank. Exper- iment Station Record, XIV, 1071. Theory of Mutations. By A. A. W. Hubrecht. Popular Science Monthly, LXV, 205-223. Part II — Causes of Variation INTRODUCTION Variation is at once the most promising agent for improve- ment and the most powerful and subtle force for undermining and destroying what has already been attained. Because of this and with a view to their possible control, the breeder is especially interested in the causes that lead to deviation in plant or animal characters. It is said that it is yet too early to inquire into the causes of variation, because our stock of accurate knowledge is too limited to permit a settlement of this most complicated cjuestion. That the matter cannot be fully settled in the present state of knowl- edge is beyond question, but the writer does not share the opin- ion that discussion at this stage of proceedings is unprofitable. The student of general evolution may well assume the role of a curious but disinterested observer, note what passes before his eyes, and take his choice as to the questions that shall engage his attention. Not so with the farmer and breeder. His funds are tied up in his animals and his plants. He is breeding them not for amusement but for profit, and he is interested in results not thousands of years hence but in those that may be confidently expected within the limits of a lifetime. He above all men, therefore, is interested in variation and the causes that induce it ; and we are bound in his interest to study the question assiduously, to determine what is known and what is not known on this most important point, and to indicate as well as we are able the direction from which further light may be expected. To this end everything is important that is connected with variability in a causative way, whether its effect is upon either the form or the function of living matter. 141 CHAPTER VII THE MECHANISM OF DEVELOPMENT AND DIFFERENTIATION Before specific inquiries can be profitably made into the causes of variation it is necessary to become fairly familiar with what is known of the essential constitution of living matter and of its manner of growth and differentiation. After attention has been bestowed for a time upon these con- siderations, it will be evident to the student that here, in the inner workings of living matter, are fundamental causes of pro- found variations, even in protoplasm seemingly the most stable. SECTION I — PROTOPLASM THE PHYSICAL BASIS OF LIFE Protoplasm is a general name for all matter that is endowed with life, but the student must never forget that the biologist, like the chemist, is dealing with matter composed of well-known chemical elements united in definite proportions. The known differences between living and non-living matter are, for our purposes, the following : 1. Living matter is endowed with a mysterious force called life. 2. Matter so endowed has a much more complicated chemical composition than has non-living matter, or than can be main- tained after the life principle has departed. Living matter at death, therefore, breaks up (or down) into ordinary chemical compounds. The true constitution of living matter cannot, therefore, be determined by any known methods of analysis, which reveal only the elements involved but not their exact relations during life. 3. Matter endowed with life is able to appropriate to itself other outlying matter and to increase its bulk through growth. 142 THE MECHANISM OF DEVELOPMENT 143 4. This growth is not of bulk merely, but it is attended by differentiation, so that one part is distinctly different from another. 5. As growth proceeds bits of this bulk are thrown off, each of which constitutes a new individual capable of independent existence, — reproduction. 6. The new individual is substantially, but never exactly, like the one from which it arose, and here lie the chief mys- teries of breeding. In reproduction there are no duplicates. Nothing approaches this in the inorganic world save crystal- lization. Crystals add matter to their bulk and thus may be said to grow. Moreover, the matter is added in an orderly manner, resulting in a kind of definite structure with exact angles always the same, but nothing like differentiation exists. One part of the crystal is like another ; it has no power of reproduction and is possessed of no force comparable with life. The student should early learn that the field of biology is distinct, but he should also fully realize that it lies within and not outside the range of chemistry, and that living matter is not freed from its ordinary affinities by reason of its association with life, but on the contrary it continues as before to be subject to the ordinary physical and chemical relations of matter generally. If he can do this, he will simplify many of his difficulties. SECTION II— THE CELL THE UNIT OF STRUCTURE If a bit of liver, bone, wood, or any other form of plant or animal tissue be examined under the microscope, it will be found to possess a definite structure, and to consist of a large number of separate divisions, each filled with a gelatinous mass called protoplasm. These separate divisions or cells are apparently alike throughout the substance of any particular tissue, - — as the liver, — but they differ greatly in different tissues of the same body (bone, muscle, brain). Biologists have been unwilling to consider the individual as the unit, because he is too large and his structure and activities are too complicated. They have, therefore, chosen to regard the individual as a colony of many and variously differentiated cells. '44 CAUSES OF VAklA'LION In assumiiif^ the rc-ll as a unit many structural (liffu ulties were s(jlvc(l to the entire satisfaction of the anatomist, but the l>hysi(jlo};ical and evolutionary problems were ( on)|)li( ated rather than simplified, brcansc this entire colony of many different cells and activities — heart, liai^t^s, liver, muscles, nerves, etc., ivith their many and diverse functions — sprani^ originally from a single cell ; moreover, this colony will throw off a succession (jf single c;ells, each of which will undergo s]jec:ific and orderly development and '(\x\\\Wj produce a colony like the pare^it. This bein^ true, the cell cannot be regarded as the ultimate unit oi living matter, unless we assume sf>mc kind of unity between all the cells ; some kind of intercellular force to insure that differ- entiation shall take place at the jiroper points and stages, — otherwise the original cell wfjuld develop into a lump of proto- plasm or into a ( olony of cells all alilce. The single cell frrtm wh'ch a new individual is to develop (the "germ cell" or mother cell) is gifted with ])otentialities for Ihe entire being, with all its ( om])lications of structure and with all its variety of function, liiologists at one time were inclined to rc;gard this germ < ell as " totipotent," that is, able to develop into almost any kind of stru( ture depending upon the surroundings. This view could not hold because different germ cells under identical conditions of life develop each into its own species. The cause of differentiation, therefore, lies ])rimarily within, and the germ cell is to be regarded as gifted with unlimited powers of (l(:velo|)n)ent only within the characters that belong to the species. Specific protoplasm is, therefore, possessed of s|)ecific prf)per- ties as truly as is any cluMiiical substance, and all the characters of structine or function belonging to the mature individual are to be regarded as in some way " inlierent in the germ." The cell is, therefore, like the individual, too laige and too complicated to be considered as the ultimate unit of living matter. This view is upiield not only upon theoretical grounds but also by the known facts of its com|)licated struct luc ;uid its re- markable behavior during cell division and growth, a subject 'i'lllv Ml'lCHANISM ()!'• l)KVI<:iX)l'MKNT 145 which it were well to consider before proceechn^^ furllier with the search after the " iiiliniate unit of nvin<^ matter," and therefore of f^nnvth, oi differentiati(jn, and of variability. SECTION 111— THK MKCIIANISM OF CKLL DIVISKJN (MITOSIS) Growth in the sense of increase of size is the direct result of cell division. Lar^e bodies do not have larger cells than small ones, but they have more of them. Growth is, therefore, in proportion to cell division, — the mechanism of which is exceed- ingly suggestive of the methods by which lines of descent are preserved and the proi)er development assured. When the pnjtoiilasm of an ordinary growing cell, |)lant or animal, has absorbed material until it has reached a certain maxi- mum size, it then prepares for division. This is not a lump division in which the new cells each get one half of the bulk of the jjarent cell, but it is (jualitative as well as cjuantitative, and is based on an exceedingly orderly procedure, which insures not only that each daughter cell shall receive its share of the mass but also that this share shall be identical in quality with that inherited by the sister cell of the same divisicjn. Those portions of the cell contents most intimately concerned in the process of divisicm, and therefore of chief interest here, may be briefly described as follows : Floating in the general protoplasmic mass (the cytoplasm) is a small body (the nucleus) of greater density than its surround- ing matter and the evident seat and initial jjoint of all construc- tive processes. Scattered through the mai^s of the nucleus and generally, but not always, in the form of minute granules is the so-called "chromatin matter," named from its intense reaction to staining agents. These granules of which the chromatin matter is (apparently)' composed arc the "chromatin granules" of some authors, the ^ The word "apparently" is inserted because the granular character of chro- matin matter has not in every case been made out, and because its granular character is less pronounced at some times than at others. 146 CAUSES OF VARIATION " chromomeres " of others, the "microsomes" of still others, and the " ids " of Weismami. Lying generally in the cytoplasm just outside the nucleus will commonly, but not always, be found an extremely minute highly staining body, the centrosome, about which, when division is about to occur, the near-by matter is thrown into radiating lines like iron filings about the poles of a magnet, giving the whole a kind of starlike appearance. These are the portions of the cell most concerned in cell divisions, and their special characters are most pronounced and the differences most distinct just previous to the act of division, and least well marked in the cell during its "resting stage" between divisions. The actual process of cell division whereby one cell gives rise to two and by which growth is attained is essentially as follows : When division is about to take place the chromatin matter (granules) assumes the appearance of a fine network running through the mass of the nucleus, the granules looking like beads strung upon a thread. This network commonly, though not always, condenses into a ribbon or thread (the spireme), which, however, speedily breaks up transversely into a definite num- ber of segments, generally in the form of short rods, straight or curved (the chromosomes). Whether or not the reticular or net- work form passes through the spireme stage, the result is always the same ; namely, the chromatin matter becomes divided into a definite number of chromosomes. Here are two remarkable and significant facts ; first, the miniber of cJironiosomes is con- stant in all individuals of the same species ; and second, " /;/ all species arising by sexual reproduction the number is even.'' ^ While the chromatin matter has been engaged in breaking up (or down) to form the chromosomes, another significant process 1 Wilson, The Cell, p. 67. The author here gives the number of chromosomes characteristic of certain species as follows: some of the sharks, 36; mouse, salamander, trout, and lily, 24; ox, guinea pig, and onion, 16; grasshopper, 12; Ascaris, 4 or 2 ; the crustacean Arleinia, 168; man, 16 or possibly 32. In this connection it is worthy of note that varieties of the same species often differ in the number of their chromosomes, the significance of which variation has not yet been determined. THE MECHANISM OF DEVELOPMENT 147 has been going on. The centrosome has divided and the two new bodies derived from it have separated and migrated to opposite sides of the nucleus, each surrounded by its radiating Unes, in which condition they are known as "asters" (stars). During this migration the asters are generally (not always) visibly connected by lines, but in either case by the time they have reached opposite sides of the nucleus they will be seen to lie at opposite ends of a spindle-shaped body (the amphiaster) consisting of lines, among which lie the chromosomes. Matters are now ready for the final and significant acts of cell division. The chromosomes arrange themselves end to end along the equator of the spindle, and at right angles to its axis ; whereupon each chromosome splits lengthwise, one group of halves migrating to one aster (centrosome), the other to the other, where each clusters about its own center, forming a new nucleus with its centrosome. The cell wall now becomes con- stricted, dividing the cytoplasm approximately equally (some- times very unequally) between the two new cells, and the division is complete. The resting stage ensues, during which preparation is made for another division ; indeed, the centrosome occasionally divides, in anticipation of the next division, even before all the details of the first division are complete. For a graphic outline of the complete process of mitosis see Figs. 20 and 21. This, in general, is the process of cell division which, with more or less variation, attends all growth. The significant facts brought to light in this complicated process are: (i) that the number of chromosomes is constant for all individuals within the species ; (2) that for all forms arising by sexual reproduc- tion the number is even ; (3) that however its details may vary, cell division consists essentially in a splitting of the chromo- somes, by which each daughter cell secures (apparently) an exact equivalent of what is received by the other daughter cell of the same division. Cell division is therefore not a lump division of the cell mass, but it is meristic, insuring a strictly qualitative division in which one half of each chromosome descends to either daughter cell.^ 1 These same facts have added significance when considered in connection with the germ cells, reproduction, and the problems of heredity. VARIATION / ''\ Fig. 20. Diagrams showing the prophases of mitosis A, resting cell with reticular nucleus and true nucleolus : at c the attraction sphere containing two centromoses. B, early prophase : the chromatin forming a continuous spireme, nucleolus still present; above, the amphiaster («). C, D, two different types of later prophases : C, disappearance of the primary spindle, divergence of the centrosomes to opposite poles of the nucleus (examples, some plant cells, cleavage stages of many eggs) ; D, persistence of the primary spindle (to form in some cases the "central spindle"), fading of the nuclear membrane, ingrowth of the astral rays, segmentation of the spireme thread to form the chromosomes (examples, epidermal cells of salamander, formation of the polar bodies). E, later prophase of type C: fading of the nuclear membrane at the poles, formation of a new spindle inside the nucleus ; precocious splitting of the chromo- somes (the latter not characteristic of this type alone). F, the mitotic figure established ; ep, the equatorial plate of chromosomes. — After Wilson THE MECHANISM OF DEVELOPMENT 149 SECTION IV— CELL DIVISION WITH AND WITHOUT DIFFERENTIATION The accomplishment of this minutely accurate division of cer- tain portions of the mother cell between the daughter cells at division suggests two points : (i) that the matter thus carefully / J Fig. 21. Diagrams of the later phases of mitosis G, metaphase: splitting, of tlie chromosomes {cp)\ n, the cast-off nucleolus. H, anaphase: the daughter chromosomes diverging, and between them the interzonal fibers {if), or central spindle; centrosomes already doubled in anticipation of the ensuing division. /, late anaphase or telophase, showing division of the cell body, midbody at the equator of the spindle, and beginning of reconstruction of the daughter nuclei. /, division com- pleted. — After Wilson divided is of special importance in shaping the activities of future cells ; (2) that daughter cells so provided should be identical, and their after growth should not only be alike but should 150 CAUSES OF VARIATION also be similar to that of the mother cell from which they are descended. For all cell division within the same tissue this latter is true ; that is to say, in the growth of liver tissue, bone tissue, or any other specific structure the cells appear to be identical and their resulting growths alike. But it must not be forgotten that these different tissues all arose originally from a single cell ; in other words, that some cell divisions are attended by differenti- ation. When .'' How } Here lies the chief mystery of variation. The mechanism of cell division would seem to be specially designed to prevent deviation and to insure absolute transmission ^ B Fig. 22. Pathological mitosis in epidermal cells of salamander caused by poisons A, asymmetrical mitosis after treatment with 0.05% antipyrin solution ; B, tripolar mitosis after treatment with o.-~/'/q potassic iodid solution. — After Wilson, from Galeotti from mother cell to daughter cell. It does not account for or indeed appear to admit of differentiation of tissues or variation in growth. But differentiation does take place and variation is a fact to be in some way explained. Irregularities in cell division. Not all cell division, it is true, proceeds with the regularity and perfection of plan indicated in the description, which is the common method in higher animals and plants and may therefore be regarded as fairly typical. The known abnormal cases are of several distinct kinds : I. Asymmetrical mitosis, in which the chromosomes are not equally distributed to the daughter cells, most of them massing THE MECHANISM OF DEVELOPMENT 151 at one pole, some of them perhaps being lost altogether in the mass of the cytoplasm. 2. Multipolar mitosis, in which the number of centrosomes is more than two and the resulting daughter cells three or more. Both these abnormal processes, however, are characteristic of abnormal growths, such as cancers and tumors, and are therefore considered as pathological. It is a suggestive fact that such irregularities may be artificially produced by poisons and other Fis D E F Pathological mitoses in human cancer cells A, asymmetrical mitosis with unequal centrosomes ; B, later stage, showing unequal distri- bution of the chromosomes ; C, quadripolar mitosis ; D, tripolar mitosis ; £, later stage ; F, trinucleate cell resulting. — After Wilson, from Galeotti chemical substances such as chloral, quinin, nicotin, antipyrin, cocain, etc.^ (See Figs. 22 and 23.) 3. Amitotic division,'^ — that is, division without the forma- tion of the amphiaster or the splitting of the chromosomes. This form of cell division is effected by constriction, resulting simply in a lump division of the mass of the nucleus, without reference to qualitative considerations. In this case the daughter cells would not, presumably, be alike. This form of cell division 1 Wilson, The Cell, pp. 97, 98. Ibid. p. 114- 152 CAUSES OF VARIATION is of rare occurrence, is never known in embryonic tissues, and is characteristic of tissues "on the way towards degeneration." ^ 4. Quite generally imicellular organisms display extreme irregularities in mitosis, some species omitting one and others another of the processes typical in higher species. Prominent among these deviations is the failure of chromatin granules to unite to form definite chromosomes. In place of this the indi- vidual granules themselves divide, suggesting that fission of the grannies is the elcnioitary and essential feature of nuclear division? Other minor deviations are known, though much of the field is yet unworkcd. These may account to some extent for differentiation during cell multiplication, and yet so far as is at present known all processes that do not accomplish an ecjuitable division of the chromatin granules through the splitting process are looked upon as distinctly pathological. Normal cell division, therefore, scents to he in the interest of constancy, not differentia- tion, and what poivcr it is that produces one sort of tissue from another, as must Jiappen in the developing embryo, is still a mystery. SECTION V — PHYSIOLOGICAL UNITS This difficulty has led to the assumption of some sort of phys- iological units, some of which are active at certain stages of development, others at other stages ; and the chromatin granules whose qualitative division is in most cases carefully insured are quite generally regarded as the repository of these units and the common vehicle of hereditary transmission. Such were the gemmules of Darwin,^ the stirp of Galton, the idioplasm of Nageli, and the determinants of Weismann."* 1 Wilson, The Cell, pp. 1 16-12 1. - Ibid. p. 90. ^ See Darwin, Animals and Plants, chap, xxvii. 4 Weismann's elaborate theory of heredity regarded the germ plasm as the original substance, of which the body is the natural expansion. This " ancestral germ plasm " is unchanging, unchangeable, and, so long as the species endures, is immortal. He regards this germ plasm as comprised ultimately of "biophors" (life bearers), which may be spoken of as living molecules. These biophors, or ultimate units, are combined in an orderly manner into " determinants," whose activity at development determines what the particular part shall be. These THE MECHANISM OF DEVELOPMENT 153 Whether or not any of these theories finally hold, this signifi- cant point remains, — that an adequate theory of heredity must account for the following facts : 1. A single cell thrown off from the sexual parts of a mature individual will, under proper conditions, produce an entirely new individual in all essential respects like the parent, but in minor respects different. 2. Commonly a cell or a number of cells taken from any other part of the body will wither and die, or, if growth follows, only one kind of tissue develops ; but in some instances (the begonia and others) the smallest bit of leaf, under favorable conditions, is able to grow and produce a new plant capable of bearing blossoms and seeds. ^ 3. The mechanism of cell division seems admirably adapted to insuring growth without differentiation. 4. But differentiation does take place, and in process of development a great variety of different cells arise from the single original germ cell. 5. These "differentiations" take place at different but proper stages, insuring orderly arrangement and, for the most part, uniform results. 6. There is always more or less variation between individuals, showing that the problem of development and inheritance is something else than absolute descent without change. We still seek, therefore, physical units with sufficiently exact properties to insure the general character of development, with such mutual relations as shall provide for orderly, not simultan- eous development, sufficiently elastic in their constitution (or combining powers) to admit of certain deviation, and each withal gifted with the power of nutrition, growth, and multiplication by division. Such in general are the properties of the physiological determinants are united into " ids," wliich are held to be identical with chromatin granules, and these in turn are assembled into " idants," which correspond with chromosomes. For a full explanation of Weismann's theory, see his Essays on Heredity, chap, iv, and his Germ Plasm, chap. i. ^ It i.s a significant fact that if two begonia leaves be placed on sand simultane- ously, one taken from a plant just about to blossom, the other from one just past the blossoming period, the plant from the former will flower first. For Weismann's views, see his Essays on Heredity, chap. iv. 154 CAUSES OF VARIATION units required to explain the function and achievements of the germ plasm, — the bit of vitalized matter that holds within its substance all the potentialities of its particular kind of life. Summary. The causes of variation are closely connected with the mechanism of growth and differentiation. The cell is the unit of structure and all growth is by cell division ; but it is not the unit of differentiation of different parts of the body, because all parts arise from one original cell, the germ cell. Cell division seems admirably adapted to insure absolute trans- mission without variability of any kind. But both differentiation and variability zxq facts. We seek, therefore, a "physiological unit" more minute than the cell, whose activities and possibly whose combinations with other physiological units of different properties are able to bring forth first differentiation within the body and later differences between different individuals. ADDITIONAL REFERENCES Chromosome Vesicles in Maturation. By \V. M. Smallwood. Science, XXI, 386. Cytological Features of Fertilization. By W. H. Blackman. Proceedings of the Royal Society, London, LXIII, 400-401. Fertility of Eggs after Removal of Cock. By L. G. Jarvis. Experiment Station Record, XI, 671. Laws of Embryonic Development : the Law of Von Baer. By Otto Glaser. Science, XV, 976-982. Mechanism of Development. By William Turner, F.R.S. Popular Science Monthly, LVII, 561-575. Ontogenetic and Phylogenetic Variation. By H. F. Osborn. Science, IV, 786-789. Problem of Development. By E. B. Wilson. Science, XXI, 281-293. Protoplasmic Structure. By E. B. Wilson. Science, II, 893-899; X, 33-45- Some Observations and Considerations upon Maturation Phe- nomena of Germ Cells. By T. H. Montgomery. Biological Bulletin, VI, 1904. Structure and Formation of Pus Cells. Experiment Station Record, XIV, 1016. Vitality of Pollen. (Roses, twenty-two days ; clivias, three months.) Experiment Station Record, XIII, 620; (Bear, thirty days) XV, 872. CHAPTER VIII INTERNAL CAUSES OF VARIATION While the causes of variation are both internal and external to the organism, the facts of the last chapter must satisfy the student of breeding problems that many of the processes attendant upon growth and reproduction are fruitful sources of variability. It is the purpose of the present chapter to discuss these internal influences somewhat at length. They are of two kinds, — (i) those affecting the individual only, and (2) those affecting the race as a whole. It is expedient to distinguish between these two classes, and the chapter will be divided into two parts, corresponding to these distinctions, as follows: (i) internal influences affecting primarily the individual ; (2) internal influences affecting: the race as a whole. I— INTERNAL INFLUENCES AFFECTING PRIMARILY THE INDIVIDUAL SECTION I — CELL DIVISION Growth is the result of cell division. Manifestly, therefore, all differences in size or in pattern are intimately dependent upon the extent and regularity of this process. Morphological variation due to cell division. Whatever influ- ences underlie the phenomena of mitosis, all questions of form or size are absolutely dependent upon the extent to which cell division and its attendant growth proceed. The individual cells in giants are not larger than those of normal specimens, but they are more numerous ; and in dwarfs they are not smaller, but fewer in number.^ What energies decide how far cell division shall proceed and when it shall stop in the case of each separate 1 Wilson, The Cell, pp. 38S-389. 155 156 CAUSES OF VARIATION organ we do not know. Food and climate undoubtedly exert a general influence, as we shall see, but altogether aside from this there must be profound internal forces or interrelationships, upon the normal exercise of which all typical results depend. Consider the development of a normal individual from the fertilized ovum to maturity. The circumstances require not only that arm, leg, and bone, heart, liver, and brain, arise at the proper time and place, but also that the attendant cell divisions tji each proceed to the requisite number and then stop. If the number be too few, a dwarf is the result; if too large, a giant ; and if too few in some parts (arrested development) or too large in others (hypertrophy), the individual is thrown out of proportion and is recognized as more or less of a monstrosity according to the degree of disproportion. To be sure, all these things occasion- ally happen, and yet, in the majority of cases, the process of cell division is adjusted with a nicety that is nothing short of marvelous ; in any event, the results secured, though varying somewhat in total development, are yet almost absolutely proportional (^.)} Whatever may be the controlling force to decide at what point cell division in each case shall stop and when the individual as a whole shall cease to grow, the plain physiological fact is that all considerations of size (quantitative variation) are fundamen- tally those of cell division. The cessation of growth at maturity does not imply the loss of power of cell division, because most forms of life, plant or ani- mal, have more or less powers of regeneration if a part is lost or injured. If a leg of a salamander be cut away, it will speedily be restored, bones and all, as good as new. A tail of a lizard is readily broken off, separating not between two vertebrae but at the middle of a vertebra (in some species generally the seventh caudal). 2 When the tail regenerates, however, the vertebrae do 1 At this point the author questions his own statement. As a matter of fact, the data involved have not been submitted to absokite mathematical determina- tion. We do not know whether the normal deviation in size due to variation in cell division is the same for all species ; nor do we know whether in giants and dwarfs all parts bear the same relative proportions as in normal specimens ; indeed, there is ground for believing that they do not. In the most general sense, however, the statement is true. 2 Morgan, Regeneration, p. 198. INTERNAL CAUSES OF VARIATION 157 not regenerate, and in their place there is only a "cartilaginous tube attached to the broken vertebra." ^ In the first case (that of the salamander) cell division, which would normally remain suspended through life, was able upon occasion not only to resume activity but also to begin back at the proper point in ontogeny ^ and repeat its normal processes from that point onward. Moreover, in this particular instance it can do this not once but many times. ^ In the lizard, on the other hand, regeneration is not complete, as no true vertebrae are formed. Higher animals generally have but slight powers of regeneration, but all have enough to repair ordinary injuries to the skin, bone, nerves, etc., showing that the power of cell division is not entirely lost at maturity; in other words, that cessation of growth when the normal size is reached is due to some cause other than the failure of the power of cell division. There are many cases of abnormal size of certain parts due to a failure of this process to arrest itself at or near the proper point. "Big heads," "giant kidneys," and similar pathological cases are instances in point, but whether the division is mitotic or amitotic has not, so far as the writer is informed, been determined. While the limitation of cell division can certainly be influenced, especially by the food supply and by exercise, it is manifest that its absolute control is, and doubtless always will be, largely beyond our power. All animals get feed enough to more than build their bodies, and the point at which growth ceases seems to be mainly constitutional. If we could regulate size directly, it would vastly simplify the process of breeding, but as it is now, we are obliged to "breed for size " and feed accordingly. 1 Morgan, Regeneration, p. 198. 2 Ontogeny refers to the development of the mdividual, as phylogeny refers to that of the race. 3 The absolute limits of regeneration are not known. Speaking generally, they are high in plants and low in animals. The salamander has been known, however, to restore tail and all four legs six successive times (Morgan, Regeneration, p. 5). The deer grows a new set of antlers every year. This is hardly a case of regener- ation, however, because successive growths are each more complicated than the former, each adding its characteristic prong; but it is a good instance to show the persistence of the power of cell division. 158 CAUSES OF VARIATION Meristic variation in general due to cell division. All differen- tiation involving numbers of duplicate parts manifestly has its seat in cell division. An additional division at the point of origin of the series doubles the mmber, but an extra division at the point of origin of a member adds a pair, if both daughter cells develop, or a single member if but one develops. When the number in a meristic series is even the series is easily conceivable as having arisen from a corresponding number of cell divisions. For example : 2 in the series, i division 4 in the series, 2 divisions 6 in the series, 2 divisions, with one pair dividing again 8 in the series, 3 divisions 10 in the series, 3 divisions, with one pair dividing again If the number of members is odd, it is only necessary to assume that one of the even numbers failed to develop, or, what is more likely, that one of a pair indulges in additional division, — its sister member remaining single ; thus : 3 members, i division, one member dividing again 5 members, 2 divisions, one member dividing again 7 members, 2 divisions, two members dividing again 9 members, 3 divisions, one member dividing again The frequent recurrence of five as a digital number is one of the mysteries in creation, and its singular persistence is another. It is, however, subject to many deviations, as was seen in the chapter on "Meristic Variation"; even in the rose family there is an occasional loss of one of the members. The frequent presence of six digits is not to be explained by reversion, as nobody supposes that number ever to have been characteristic in any species, — a fact that should be noted by some of our friends who are always ready to invoke the aid of atavism to explain every abnormality. Meristic variation, like other deviations arising from external causes, is to some extent hereditary, and capable of being in- fluenced, if not absolutely controlled, through selection. No other method is known, aside from the fact that external injury INTERNAL CAUSES OF VARIATION 159 to many plants and certain animals results in budding and mul- tiplication of parts. We cut the main stem of a small tree or shrub in order to increase the number of side branches. Some- what similarly, injured parts are often doubled in regeneration.^ In this way lizards may be made to produce an increased number of toes and even double feet, legs, and tail. It is supposed that double feet, sometimes seen even in mammals, may be produced by a "fold of the amnion constricting the middle of the begin- ning of the young leg " ^ in the embryo. This, however, is curi- ous rather than valuable to us, as it tends to explain abnormalities rather than to point a way to practical improvement. Irregularities in cell division a cause of variation.-^ The char- acteristic act in cell division seems to be the splitting of the chromosomes (or chromatin granules) and the migration of exact equivalents to each new daughter cell, strongly suggesting that the assortment of " physiological units " (whatever they may be) received by one daughter cell is an exact duplicate of that received by the other, thus insuring an orderly and systematic develop- ment through a strictly qualitative division of hereditary sub- stance at each and every stage of growth. The whole mechanism of mitosis seems adjusted to this end, and if the assumption is true its significance can hardly be over- estimated. If this careful adjustment of the mechanism of cell division is necessary to orderly development, it is manifest that any substantial deviation is likely, if not certain, to result in variation more or less profound. Such deviation is characteristic of amitotic division generally, and it is more than conceivable that the ordinary process is subject to occasional " slips." Some chromatin granule may fail to divide at the proper moment and may pass over to one daughter cell entire,* or, conversely, it may indulge in an extra division. Substantial deviations in the process are known to occur not rarely but frequently. For ex- ample, the splitting sometimes takes place in the spireme stage, sometimes after the formation of the chromosomes ; sometimes 1 Morgan, Regeneration, pp. 137-139. 2 Ibid. p. 139. ^ See previous cliapter. 4 This is known to occur in certain instances in maturation. l6o CAUSES OF VARIATION the centrosomc divides before the resting stage, more commonly afterward. Taking it ah in all, here is an exceedingly compli- cated procedure, only semi-mechanical and therefore subject to deviations. Absolute constancy demands no failure in the final object of exact qualitative division, but the student sees many possibilities for unequal division and therefore for deviation in growth. Is this the fundamental cause of mutations ? One thing is certain, — living forms are made up of elements, and these elements are subject to strange combinations throughout the entire range of plant and animal life, and the facts seem to teach that from time to time combinations may arise that are entirely new. Moreover, whole units seem occasionally to be " lost out," as when horned cattle suddenly give rise to polled strains, hairy species to smooth varieties, colored to albino, etc. Conversely, do vital elements like chemical radicles assume new combinations from time to time, giving rise to new char- acters and new types which we call " mutants " .'' We do not know, and yet we feel the conviction that at this point we are very close to the "origin of characters," the cause of mutations and of variation in general. Manifestly, in so far as irregularities in cell division may be a cause of variation, the matter lies absolutely beyond our control except that lines in which it is believed to occur may be avoided in selection. Here is a field, however, too far beyond our pres- ent knowledge to admit of anything more than the merest mention. We confidently believe that the future will shed more light on this obscure subject. SECTION II— BISEXUAL REPRODUCTION A FUNDAMENTAL CAUSE OF VARIATION Among higher animals and plants the new individual is the direct product of two others, — the male and the female parent, — and is of necessity different from either, being a product of both. In bisexual reproduction, therefore, biologists recognize a fundamental cause of variation, — slight if the parents are of like blood lines, extreme if of radically different, as in hybridism. INTERNAL CAUSES OF VARIATION i6l This view of the case is borne out by the facts of fecundation or fertilization of the ovum, which may be briefly described as follows : ^ The ovum. This is the finished product of the sexual cells of the mother parent, and consists of a nucleus with its characteristic chromatin granules surrounded by a comparatively large mass of cytoplasm. The sperm cell. This is the finished product of the sexual cells of the male parent. It is called a spermatozoon in animals, sper- matozoid in lower plants, and pollen grain in higher plants. It is in all cases vastly smaller than the corresponding ovum, being almost destitute of cytoplasm. The characteristic elements of the ovum are its nucleus and the cytoplasm, while the character- istic elements of the spermatozoon are its nucleus, borne in the "head," and a centrosome, generally carried in the "middle piece." The tail, formed from the small amount of cytoplasm, seems to have no function beyond providing motile power, and is absent in the pollen of higher plants. Fertilization. Both the ovum and the sperm cell have arisen in their respective organs by the method of cell division, display- ing in the process the ordinary phenomena of mitosis.'^ But both have reached the end of their powers of self-division, and if left alone they will be thrown off from their respective points of origin to wither and die. If, however, they are brought near together, mutual attraction ensues, the spermatozoon (or other sperm cell) enters the ovum, the nuclei approach each other and fuse, the centrosome divides, an amphiaster is formed, and cell division ensues. The ovum is now fertilized, segmentation proceeds, and a new individual is established in an independent existence. The new individual is thus the possessor of actual living mat- ter (physiological units) derived from both parents, and thus inherits literally the substance of both, having come into direct possession of material identical with the living matter of both parents. ^ For a fuller discussion of this subject, see Wilson, The Cell, pp. 178-231. ' For a brief statement of what is involved in maturation, see the next section. 1 62 CAUSES OF VARIATION All that is involvctl in fertilization is not well understood, but its essential feature is the jtnio)i or fusion of the jutclcar matter {clironiosovies) of the germ cells from tivo parents to form the cleavage or segmentation nucleus whose subseqiient groivtJi and divisions '^ give rise to all the nuclei of the body.'' This fertilized ovum becomes, therefore, the first cell of the new being, which inherits directly and equally a portion of the nuclear matter from both parents, so that " every nucleus of the child may contain nuclear substance derived from both parents." ^ Here, then, is the avenue of all inheritance, and, as the new individual is a kind of blend of both parents, we see in fertilization an initial and primary cause of variation. This is the only form of variation recognized by Weismann in his earlier writings as in any sense hereditary. All deviations in development due to external causes were conceived to affect the body (soma plasm ^) only, exerting no influence upon the ancestral germ plasm .'^ True, he later announced the theory of germinal selection, in which a kind of struggle for existence is conceived as taking place between the "biophors " (physiological units), by which some prosper and multiply exceedingly while others are crowded out entirely.* This would give another cause of variation within the germ plasm of each individual. Biologists generally recognize internal causes of variation other than these, and yet this union of the chromosomes from different individuals taking place at each new generation must be regarded as a very effective means of introducing variability. Even if the offspring of a single parent, as in parthenogenesis, should be an exact duplicate of the parent, — which it is not, — every one would recognize the fact that the blending of heredi- tary substance from two parents must of necessity produce an individual with a new combination of faculties. It is a variation, however, confined not only to the characters of the race but also to the family possessions of the particular parents. Bisexual reproduction cannot be looked upon as a means 1 Wilson, The Cell, p. 182. 2 " Soma plasm " is a term used to represent the protoplasms of the body in general as distinct from the output of the sexual cells (germ plasm). 2 Weismann, The Germ Plasm, chap. ix. * Weismann, Germinal Selection (pamphlet). INTERNAL CAUSES OF VARIATION 163 of introducing new characters into the race, and while it is mani- festly a fruitful source of never-ending combinations of racial characters in new individuals, yet variations so introduced are comparatively slight except when the two parents belong to sepa- rate lines. Fertilization of the ovum is something more than a stimulus to growth. It is a real union of material bodies, physiological units, or whatever they may be called, representing the hereditary substance of both parents. Bisexual reproduction is therefore not only a guaranty of transmission of racial characters but also an assurance of inheritance with some variation. Control. Here is a fundamental cause of variation practically under the control of the breeder through selection. True, his knowledge and his judgment are insufficient to insure him against mistakes in mating, and it is also true that there are many other influences at work to produce variations, but this is the field in which the breeder can exert the largest influence, and it is by selection that the greatest results in improvement have been attained up to date. Sexual selection, 1 preferential mating,- and assortative mating.'^ Powerful as are these influences in directing the trend of varia- bility, they yet belong to general evolution because they are ele- ments in natural selection, and they have no place in the present discussion. SECTION III — MATURATION AND THE REDUCTION OF THE CHROMOSOMES A CAUSE OF VARIATION Fertilization is a process whose inevitable consequence would seem to be the ^^ piling tip " of nuclear matter indefitiitely ; for if, with each new generation, the chromosomes (or physiological units) of the one parent are added to those of the other, it would seem that in time the resulting nuclear matter would speedily become " unmanageably large " and inconceivably complex, — an event certain to follow except for a series of very remarkable 1 Darwin, Origin of Species, see Index. 2 Pearson, Grammar of Science, pp. 425-428. 3 Ibid. pp. 4^9-437- 1 64 CAUSES OF VARIATION facts occurring just previous to fertilization and by which tJie mimber of cJiroinosomes in both the male and female germ cells is reduced to one half the nsnal or sojnatic number, so that their union at fertiHzation restores the true number of chromosomes typical of the race. Thus, if the somatic number of chromosomes is sixteen, the number in the germ cells at fertilization will be eight each, or sixteen after fusion of the nuclei. This process by which the number of chromosomes is halved in the germ cell is known as reduction, and is supposed to be the significant feature of the maturation process by which the male and female germ cells are prepared for union. Parallelism in the sexes. Maturation and its attendant phe- nomena of reduction in the number of chromosomes is a subject that must be considered separately in the male and the female, and yet there exists a strange parallelism worthy of notice. To quote Wilson ^ : Recent research has shown that maturation conforms to the same type in both sexes. . . . Stated in the most general terms this parallel is as fol- lows: In both sexes the final reduction in the number of chromosomes is effected in the course of the last two cell divisions, or jiiaturatioii dii'isions [as they are called], by which the definitive germ cells arise, each of the four cells thus formed having but half the usual number of chromosomes. In the female but one of the four cells [resulting from the two maturation divisions] forms the ovum proper, while the other three, known as the polar bodies^ are minute, rudimentary, and incapable of development. In the male, on the other hand, all four of the cells become functional sper- matozoa. This difference between the two sexes is probal:)ly due to the physiological division of labor between the germ cells, the spermatozoa being motile and very small, while the egg contains a large amount of protoplasm and yolk, out of which the main mass of the embryonic body is formed. In the male, therefore, all of the four cells may '^ become func- tional ; in the female the functions of development have become restricted to but one of the four, while the others have become rudimentary. 1 Wilson, The Cell, p. 234. 2 The author is here speaking specifically of reproduction in animals, as plants do not form polar bodies. The difference in plant and animal reproduc- tion is, however, more in form than in significance. ^ The author says " may " become functional. He means by this that each of the four cells (spermatozoa) arising from the last two divisions is capable of fertilizing an ovum, while of the four cells arising from the last two divisions in the female only one is capable of being fertilized. INTERNAL CAUSES OF VARIATION 165 Maturation and reduction in animals and in plants radically- different. This process is far better understood in animals than in plants, and in many respects it is radically different in the two. It is simpler in animals and more direct. In them the last two cell divisions always (apparently) give rise in the male to four functional spermatozoa, but in the female to one functional cell, retaining nearly all the cytoplasm, and to three polar bodies incapable of fertilization and destined to wither away and disap- pear. The same general facts seem to hold for animals of all species, and Wilson remarks ^ : The evidence is steadily accumulating that reduction is accomplished b)' two maturation divisions throughout the animal kingdom, even in the unicellular forms ; though in certain Infusoria an additional division occurs, while in some other Protozoa only one maturation division has thus far been made out. Among plants, also, two maturation divisions occur in all the higher forms, and in some at least of the lower ones. Here, however, the phe- nomena are complicated by the fact that the two divisions do not, as a rule, give rise directly to the four sexual germ cells, but to asexual .spores which undergo additional divisions before the definitive germ cells are produced.- [The end product, however, shows the same reduction in the number of chromosomes.] A brief description of reduction in animals is worth consider- ing somewhat in detail, as it is fairly well known and cannot fail to impress the student with its fundamental significance and the nicety of adjustment of the mechanism of living processes. Reduction in the female.-^ Among animals the production of the female germ cell (the ovum) is the special function of the ovaries. In the tissues of these organs cell division proceeds under the usual mitotic plan, giving rise to a series of cells known as oogonia. At a certain point mitotic division halts, and each cell prepares for the final (maturation) changes. Food material is absorbed, the cytoplasm increases in bulk, the nucleus greatly enlarges, and the cell, now known as an oocyte^ is ready 1 Wilson, The Cell, p. 235. 2 Ibid. pp. 235-236. Note that, in general, polar bodies are not formed in plants. 3 Ibid. pp. 236-240. This description applies to the animal. The details are distinctly different in plants, to be discussed later. 1 66 CAUSES OF VARIATION for the last two, or maturation, divisions. In this condition the egg cell remains until near the time of fertilization, when the process of maturation proper takes place. The significant details of this interesting series of changes are concerned with the nucleus and are substantially as follows : During the long resting stage preparatory to these final divisions the nucleus increases in bulk and the chromatin matter assumes the reticular form characteristic of the resting stage of dividing cells in general. In this condition the nucleus is known as the "germinal vesicle." Up to this point the number of chromo- somes is the same as that of the body cells in general. Their identity is, of course, now lost, but as the time for the first maturation division arrives, instead of the spireme of ordinary mitosis breaking up into the usual number of chromosomes, there appear more or less spontaneously a number of " primary chromatin masses " in the form of rods, rings, or V-shaped bodies, each of which ultimately breaks up into four smaller bodies. These groups of four are always one Jialf the Jisnal or somatic number of chromosomes. Whether the chromatin masses appear in the form of rods, rings, or otherwise, the final result seems to be always the same ; namely, the breaking up of each into four smaller bodies, either by two longitudinal divisiojts or by one (the first) longitudinal and one transverse. The details differ in different species and have been worked out in but few cases. It is not important here to trace the bewildering differences, but rather to describe typical behavior.^ Having assumed this condition the nucleus now migrates to the margin of the cell, each of the groups of four (tetrads in rod- shaped cases) splitting into two smaller groups of two each (dyads). ^ The mass now divides, one pair from each group 1 For a full discussion of the different forms of reduction, see Wilson, The Cell, V, 233-287. 2 The terms " tetrad " and " dyad " of course apply only in the case of rod- shaped masses. In the case of rings (common in animals) and V-shaped masses (common in plants) the parting into four takes place gradually as the work pro- ceeds, while in the case of rods the division into four takes place early and the parts are distinct from the first. In this formation the two divisions take place much more rapidly than in the case of rings, which split and divide slowly. The INTERNAL CAUSES OF VARIATION 167 remaining in the cell, the other passing outside, forming the first polar body, which may or may not undergo further division. The portion now remaining within the cell consists of groups of two each, instead of four, and their number is of course the same as before, namely, one half the somatic number of chromo- somes. Immediately now, without assuming the resting stage, the dyads, or groups of twos, turn one fourth around, taking a position at right angles to the margin of the cell, and (7^ oiice divide again, one member of each pair remaining behind iti the egg eel I, the other passing ont, forming the second pohxr body. The first polar body carried away one half the nuclear matter, and the remaining half has now been divided equally between the second polar body and the main cell, which is now ready for fertilization and is from this time on spoken of as the ovum. Neither polar body carries any appreciable quantity of cyto- plasm, and both are destined to degenerate and disappear. The first one, however, containing half of the total nuclear matter, commonly divides once, so that the first polar body represents not one, but two cells, — the first and the third polar bodies. The total result, then, of this complicated process seems to be the equal division of the chromatin matter between the ovum, capable of fertilization, and three polar bodies, destined to extinction. A group of four cells thus arises, - — namely, the mature ^gg (ovum), which after fertilization gives rise to the embryo, and three small cells or polar bodies (incapable of fertilization),^ which take no part in the further development, are discarded, and soon die without further change. The Q.gg nucleus (of the ovum proper) is now ready for union with the sperm nucleus,^ which process is known as fertilization. *' In some cases — for example in the sea urchin — the polar bodies are formed before fertilization, while the o^gg is still in tetrad form is always chosen for description because the details are capable of more definite statement. Whatever the form of the masses, however, the final result seems always the same ; namely, a reduction to one half the usual number of chromosomes, and this by the method of division and extrusion. 1 In rare instances the polar bodies have commenced to segment, but they never proceed far in development. 2 Wilson, The Cell, pp. 236-237. Fig. 24. Diagrams showing (lie essential-facts in the iiiaturatinn of the egg. The somatic number of chromosomes is supposed to be four A, initial pliase: two tetrads have been formed in the germinal vescicle. B. the two tetrads have been drawn up about the spindle to form the equatorial plate of the (irst polar mitotic figure. C, the mitotic figure has rotated into position, leaving the remains of the germinal vesicle at g.v. D, formation of the first polar body : each tetrad divides into two dyads. E, first polar body formed: two dyads in it and also in the egg (/•/'.'). F. preparation for the second division. (7, second polar body forming and the first dividing: each dyad divides into two single chromosomes. //, final result: tliree polar biidies and the mature ovum, eacli containing two single chromosomes, or half the somatic number ; c; the egg centrosome, which now degenerates and is lost. — After Wilson 168 INTERNAL CAUSES OF VARIATION 169 the ovary. More commonly, as in annelids, gasteropods, and nematodes, they are not formed until after the spermatozoon has made its entrance ; while in a few cases one polar body may be formed before fertilization and one afterward, as in the lamprey eel, the frog, and in Ainphioxus. In all these cases the essential j^henomena are the same. Two minvite cells are formed, one after the other, in rajiid succession and near the upper or animal pole of the ovum ; and in many cases the first of these divides into two as the second is formed." To what extent this division is qualitative is unknown. Of one thing we are certain : soincwJicrc in the process the nuviber of eJiroDiosouies Jias been reduced to exactly one half the number characteristic of the species. It was formerly supposed by Van Beneden, Weismann, and Boveri that reduction consists in the casting out and degenera- tion of half of the chromosomes. " Later researches conclusively showed, however, that this view cannot be sustained, and that reduction is effected by a rearrani^emoit and rcdistribjition of the nuclear substance, without loss of any of its essential constitu- ents." ^ This is said because the groups — tetrads, rods, rings, etc. — arise spontaneously in the nucleus in the reduced number. The loss occurs later in the extrusion of the polar bodies, but no corresponding loss takes place on the male side because all four cells are functional, though not all alike. Reduction in the male.- The maturation j^rocesses in the male and female are i)ractically identical in their results, with two exceptions; namely, first, in the male the four cells resulting from the maturation divisions are all alike and all functional, and second, they are exceedingly small in size as compared with the ovum, being almost destitute of cytoplasm. The spermatogonia, corresponding to the oogonia of the female, arise in the testes by mitotic division, with the full somatic number of chromosomes. As in the female, they reach a stage where division ceases for a time and enlargement ensues, in which condition the cells are known as spermatocytes (corre- sponding to oocytes in the female). 1 Wilson, The Cell, p. 233. - Ibid. pp. 241-242, 170 CAUSES OF VARIATION At the proper stage each spermatocyte undergoes two divi- sions (maturation divisions) into four cells, called spermatids, each of which develops a tail and becomes functional, in which finished condition it is known as a spermatozoon, when it is ready to enter and fertilize the ripened ovum. The history and distribution of the chromatin matter in the male is identical with that in the female, so that each sperma- tozoon inherits one foiirtJi the chromatin matter and one half the chromosomes of the original cell. In plants the process differs but the general results are the same. Significance of reduction. On the female side three fourths of the chromatin matter has been extruded in the polar bodies, and therefore lost to the line of descent. Whether reduction takes place by extrusion or by rearrangement, one thing is certain : when the second division is transverse, and possibly when it is longitudinal, it results in an unequal division of physiological nnits, if the identity of the chromosomes and the chromatin granules has any meaning. If this division be anything else than strictly qualitative, then the extrusion of the polar bodies means a loss of something qualitative on the female side. On the male side the loss is not absolute, because all four cells are functional, but if reduction has the meaning we attach to it, these four spermatozoa are not identical but different in the hereditary substance with which they are provided. Significance of fertilization. Here, then, are two sexual cells ready for union. Each has lost large portions of its chromatin matter, the evident vehicle of transmission, and each brings to the union but one half the number of chromosomes characteris- tic of its species, strongly suggesting a loss of certain chromatin granules and the hereditary qualities they represented. When fusion of the nuclei of these two germ cells takes place at fertilization, however, the act of union again restores the full and proper number of chromosomes, which will remain charac- teristic of the new individual throughout its life, and which it will hand down to posterity, always through the same compli- cated method we have attempted to describe. The number of chromosomes is evidently kept constant by the complicated process of reduction during maturation and by fertilization INTERNAL CAUSES OF VARIATION 171 afterward ; but what about their character ? In what condition have they emerged from this seemingly incomprehensible tangle ? Is nature as careful to preserve their quality as it is their number ? What opportunities for profound variation ! Certainly if chromatin matter has any fundamental meaning, and if chromo- somes are in any way representative of physiological units, and if they in their turn are in any way representative of racial characters, these processes must have some meaning in varia- tion. Certainly we have been very near in all this to the material basis of transmission of racial characters, and to fundamental and initial causes of variation. Something has been lost in the two peculiar divisions attend- ing maturation. Some definite groupings of hereditary substance have disappeared from the line of descent. They could not have represented, ordinarily, definite portions of a body, but they must represent something. What chances for accident ! And, in the light of these marvelous phenomena, do we wonder that individ- uals are sometimes born minus a leg, an arm, or some other part .? Do we wonder that vital parts are so often affected, and that one third of our children die in infancy ? How many die before birth, and how many more die at some stage in embryo ! Evidently all that is required to make a living being is a fairly perfect development of the vital parts ^ quite regardless of the presence or absence of the many other racial characters that should be present in the perfect individual. Is it surprising that perfect individuals are so few, and that defectives are frequently so far from the type ? Here is material for study on the part of criminologists and courts of justice, as well as students of methods of economic improvement. Reduction and fertilization in plants. It may be said in general that in animals the evidence tends to the assumption that reduction takes place at the extrusion of the second polar body ; that each group (rod, ring, or V-shaped body) is in reality a doubled (bivalent) chromosome, and that the first polar body removes one half of each (split) chromosome, while the next removes every alternate chromosome. While the facts of reduction in plants are yet in a hopeless tangle, it is safe to say that the evidence tends to show that no 172 CAUSES OF VARIATION true polar bodies are formed,^ but that the chromosomes sud- denly appear in reduced number at the first division, as if it were effected directly by the segmentation of the spireme thread (of the maturing germ cell) into half the somatic number of chromosomes. While details vary greatly, botanists recognize two stages in the development of the female germ cell of the plant, neither of which is identical with maturation in animals, though the first is fairly comparable thereto. The first stage, or sporogenesis, follows after that active massing of food material which marks in both plant and animal the preparation for reduction. At this time the nucleus divides quickly into four daughter nuclei, each of which is supplied with half the number of chromosomes that characterizes the species. The precise methods followed out are matters of much dispute among botanists, but the significant fact is that reduction is accomplished at this stage. Of these four daughter nuclei none are extruded, but three of them degenerate in the cytoplasm, while the fourth increases in size to form the embryo sac, which, without waiting for ferti- lization as among animals, continues to divide, — commonly twice, — giving rise to eight sub-nuclei, which arrange them- selves in definite positions. Two of these sub-nuclei remain near the center of the embryo sac and give rise to the endo- sperm ; three migrate to the extremity nearest the point of attachment with the pistil, and one of these (and one only) — the so-called egg nucleus — unites with the nucleus of the pollen grain to form the fertilized germ ; the three remaining migrate to the other extremity of the embryo sac and concern themselves with establishing a food supply with the parent plant. On the male side the process is simpler. The pollen nucleus divides, one half forming the pollen tube, along which the other half travels, dividing again at some point before uniting with the egg nucleus of the embryo sac. These divisions are evidently reducing divisions, as the pollen-grain nucleus brings to the union a reduced number of chromosomes. 1 Disputed by Chamberlain, who believes that "the egg with its three polar bodies constitutes a generation directly comparable with the gametophytic genera- tion in plants." See Botanical Gazette^ XXXIX, 139; see also under " Xenia," in this text. INTERNAL CAUSES OF VARIATION 173 Thus, while the plan is different in plants and in animals, the first stage, sporogenesis, in plants seems fully comparable with maturation in animals, and the same general end is accomplished. ^ After all, the manner of division is primarily of interest to the physiologist and does not concern us. Our interest is in the fact that maturation in general involves an actual loss of chromatin matter (hereditary substance) and a reduction in the number of chromosomes, and consequently of physiological units. In the present state of knowledge it seems safe to assume that both these results follow, whatever the mechanism of maturation in each particular instance. If this be true, here is a fertile and initial cause of profound variation, an excellent opportunity for losing important elements of the physical make-up, but, so far as we can see, no chance for positive gain, unless it be by new combinations, because nothing is introduced. Phenomena such as these are remarkable for what they sug- gest rather than for conclusions that can be positively drawn. The suggestion is that of substantial deviation in the very fun- damental process of transmission of the hereditary substance, — a deviation that cannot but be fruitful of variation in resulting individuals. Weismann's prediction. It is noteworthy that reduction was predicted by Weismann on purely theoretical grounds some years before it was known as a fact.^ He argued for its recur- rence as a physiological necessity to prevent the piling up of "ancestral idioplasm," — the physiological units to which he afterward gave the name of "ancestral units," '^ — and later developed the intricate system of biophors,"* determinants,^ ids,^ and idants,'^ by virtue of which he explained the constitution of the germ plasm," and which he used as the basis for his famous theories of heredity.^ 1 For full discussion, see articles by B. M. Davis, Amcricait A'atiiraiist, XXXIX, Nos. 460 and 463. ■^ Weismann, Essays on Heredity, I, 357, 363-396; II, 1 14-150; also Weis- mann, The Germ Plasm, chap. viii. ^ Weismann, Essays on Heredity, II, 116. * Weismann, The Germ Plasm, pp. 40-53. * Ibid. pp. 53-60. 6 Weismann, Essays on Heredity, II, 136-138; Weismann, The Germ Plasm, pp. 60-75. " Weismann, The Germ Plasm, chap, i, pp. 37-85. ^ Ibid. chap, ix, pp. 253-293. 1 74 CAUSES OF VARIATION The later discovery of the mechanism of maturation and of the extrusion of the polar bodies was a startUng confirmation of Weismann's prediction, and went far to fix his theories of hered- ity in the minds of many biologists. In this connection Wilson very pertinently remarks : ^ The fulfillment of Weismann's prediction is one of the most interesting ) esults of recent cytological research. It has been demonstrated in a manner which seems to be incontrovertible that the reducing divisions postulated by Weismann actually occur, though not precisely in the manner conceived by him, . . . but it remains quite an open question whether they have the significance attributed to them by Weismann. Just when the reduction occurs is not known. It was at first assumed that it occurs in connection with the extrusion of the second polar body, — an assumption based upon the development of parthenogenetic eggs. But plants do not form polar bodies, and again there is great uncertainty as to whether the rods (tetrads), rings, or V-shaped bodies (whose number is half the usual number of chromosomes) are to be regarded as represent- ing the usual number of chromosomes split and arranged in pairs, — in which case the second polar body would accomplish the reduction ; or whether the chromosomes as formerly known never emerge from the nucleus of the oocyte, so that the iden- tity of the chromosomes is in some way lost and the reduction is effected at this early stage by some sort of internal fusing, or perhaps by an entire rearrangement of chromatin granules. On this point the evidence is confusing, but on two significant points there is no doubt, — the loss of chromatin matter out of the line of descent, and a reduction of the chromosomes in the germ cells to one half the somatic number.^ Composition of the chromosomes.'^ In view of the important office of the chromosomes and the many theories of heredity based upon their nature and constitution, in view also of their evident importance in all studies on inheritance and variation, 1 Wilson, The Cell, p. 246. 2 By "somatic" number is meant the number characteristic of the soma,—' the body in general as distinct from the germinal matter whose function is not growth but reproduction. 3 Wilson, The Cell, pp. 294-304. INTERNAL CAUSES OF VARIATION 175 it may be well to note the substance of what is really known about their actual constitution. That they are not masses of homogeneous matter is certain, and that they consist of numbers of small granules capable of multiplication and division seems equally certain. To quote Wilson : The facts are now well established (i) that in a large number of cases the chromatin thread consists of a series of granules (chromosomes) imbedded in and held together by the linin substance ; (2) that the split- ting of the chromosomes is caused by the division of these more elementary bodies ; (3) that the chromatin grains may divide at the time when the spireme is only just beginning to emerge from the reticulum or resting stage. Because of these facts there arises the strongest tendency to attach individuality to the chromatin granules and to conceive them as built up of definite, though often diverse, physiological units, thus constituting a semi-mechanical basis for heredity, and incidentally for variation as well. This assumption Weismann and others have made. Whether the facts should be pushed to this extreme interpretation is, in the opinion of the author, as yet uncertain. The facts are extremely suggestive, to say the least, and it is certainly not too much to believe that at this point we have touched the physical basis of life and in some fashion the very root of inheritance and variation ; indeed we may proceed upon the conviction that transmission is a function of the chromatin granules. Reduction as a cause of variation. The most remarkable and suggestive fact about living beings is the numerical constancy of chromatin units (chromosomes) for each species, and the most remarkable and suggestive of all the vital processes is their reduction before fertilization. If, as we suppose, the chromosomes are the physical basis of inheritance, then in the loss of c1iroinati7i matter at maturation lies a fundamental cause of variatiofi, and one quite independent of the effects of fertili- zation afterward. Reduction would seem to be a process calculated to insure that no two germ cells, even from the same individual, should ever be alike, and this is the most evident reason for the 176 CAUSKS OK VARIATION essential differences in children of the same parents, even in the case of twins.' It is true, of course, that no two individuals, even twins, can be developed under conditions of life exactly identical ; and yet the differences of condition cannot account for the fact that, while one brother resembles his father, another may resemble 1 Twins are considered as arising from separate ova, as in the case of multiple births (pigs, dogs, etc.), and, of course, as exhibiting the deviations to be expected from different germs and distinct fertilization, as in litters generally. Some twins, however, are so nearly alike (identical twins) as to suggest the possibility of their having arisen from a single ovum in some way sejxirated into halves at its first cleavage, each half developing an individual. This view is evidently favored by Geddes and Thom.son (see The Evolution of Sex, p. 41). If twins should be developed in this manner, they would evidently be of the nearest possible similarity, for they represent but one ovum and but a single fertilization. This possibility is supported by the experiments of Roux, Endres, and Walter, in which each blastomere of the two-cell stage of the frog sometimes (not always) is capable of developing into a perfect individual. Driesch, working with echino- derms, established the same facts, which are also well known in the case of Amphioxus (see Wilson, The Cell, p. 419). Conversely, when two fertilized ova of sea urchin, or Ascari.s, adhere acci- dentally, they may develop into an embryo of unusual dimensions (see Loeb, Studies in General Physiology, Part II, p. 676). When, however, the blastomeres are not separated, but one of them is killed by a heated needle (Roux), the uninjured half alone develops, but it produces at the best a kind of half larva (right or left half), "containing one medullary fold, one auditory pit " (Wilson, The Cell, p. 399). Chun, Driesch, Morgan, and Fischel, working with ctenophore eggs, however, found that isolated blastomeres of the two-, four-, or eight-cell stages developed " defective larvae, having only four, two, or one row of swimming plates." Also " Crampton found that in the case of the marine gasteropod Jlynnnssa isolated blastomeres of two-cell or four- cell stages segmented exactly as if forming part of an entire embryo, and gave rise \.o fragments of a lawa, not to complete dwarfs as in the echinoderm" (Wilson, The Cell, p. 419). This attempt to form entire individuals from a portion only of a fertilized egg, resulting as it often does in dwarfs, seems to the writer a process closely akin to regeneration (which see in chapter on " Relative Stability and Instability of Living Matter"), and would seem to raise doubts as to its successful occurrence in the higher animals. Though not bearing especially upon the point in question, the matter of twins in cattle is unique and worthy of mention. Three kinds of twins are known in cattle: " (i) the twins may be both female and both normal; or (2) the sexes may be different and normal ; or (3) both may be males, in which case one always exhibits the peculiar abnormality known as a ' free-martin,' — the internal organs are male, but the external accessory organs are female, and there are also rudi- mentary female ducts" (Geddes and Thomson, The Evolution of Sex, p. 41). This is a kind of hermaphroditism, and not, as is commonly supposed, "a heifer twin with a bull." INTERNAL CAUSES OF VARIATION 177 his mother or one of her male ancestors. Differences such as these must arise from strictly internal causes, which seem to set a natural and inevitable limit to what may be accomplished through selection. Here would seem to be an irreducible mini- mum in variation, arising directly through reduction.^ Control. In so far as variations arise through changes in hereditary matter during the pnjcesses of maturation and reduc- tion, they are, and must doubtless always remain, entirely beyond the influence of the breeder. It is quite evident that here is a degree of deviation and an element in breeding that must be left to nature antl subject to the laws of chance within the range of characters natural to the race. That this will always be a bar to absolute success is evident, but that it constitutes the strongest known ai'gument for purity of blood is, in the opinion of the writer, beyond question, because the chances of tmfortujmte deviations are reduced in proportion to the purity of blood and the absence of undesirable characters. Variation in parthenogenetic reproduction.'^ Had Weismann's original assumption been correct to the effect that se.xual imion is the only constitutional cause of variation, then individuals arising from parthenogenetic reproduction should, barring the influence of surrounding conditions, be alike, because only the 1 Endeavoring to determine the function of the cytoplasm and nucleus, Boveri removed the nuclei from the eggs of sea urchins and afterward admitted sperma- tozoa to these enucleated ova. Development followed in a few cases, but the nuclei were smaller than in larvas normally fertilized, contained /'/// lialf tlie iiiiiii- ber of cJiromosoines, and the resulting larvai possessed the "pure parental characters." It is not supposable that anything like this occurs in nature, and yet it raises the cjuestion whether, after reduction, every reiiiaiiiing elenient of tlie inicleiis of botli parents always plays its part in development. Should it fail to do so for any reason, herein would lie a sufficient cause for the occasional remarkable resem- blance of offspring to one and not the other parent. 2 While in all higher animals and plants a union of a male with the female cells is necessary to each fertilization and to the production of young, it is by no means true among other organisms, especially in rotifers, crustaceans, and insects with which "parthenogenesis lias become a fi.xed physiological habit," through which the unfertilized female cell develops a perfect individual. It is now well known that the queen of the honeybee, if prevented from mat- ing, will yet lay eggs capable of development, but they will all be drones (males). After mating she can lay either fertilized or unfertilized eggs, the fertilized devel- oping into workers (undeveloped females), — or, if jiroperly fed, into Cjueens, — the unfertilized into drones as before. After the male element is exhausted (she I 78 CAUSES OF VARIATION female and her germ cell are involved. But individuals thus arising through unisexual reproduction vary zvide/y, a fact easily credited when the phenomena of reduction are remembered. Parthenogenesis being limited to lower animals, the range and character of variations are for the most part difficult of detection and measurement. It is known, however, that great differences in size occur among individuals parthenogenetically produced, and characters generally are so variable in such individuals as to lead to the statement that the variability of offspring never mates but once) she is, of course, capable of laying only unfertilized or " drone eggs." In this way, in crossing, the drones and the workers may actually be of different breeds. Plant lice reproduce parthenogenetically during the summer season, producing only females ; but as the temperature lowers with approaching autumn a mixed brood of both males and females appears, which, upon mating, produces the long- lived, winter-enduring eggs. It is noteworthy that the parthenogenetic eggs of bees develop males only, while those of plant lice develop females during the summer, both sexes appearing as autumn approaches. It is also noteworthy that under the artificial heat of greenhouses, approximat- ing perpetual summer conditions, parthenogenesis continues indefinitely, and males are not produced unless the plants become badly dried up. Parthenogenesis differs greatly in degree. It is supposed to be complete in certain minute crustaceans and in many rotifers among which " no males have ever been found." It is "seasonal" in the aphis (plant lice), and "partial" in the honeybee and " in some of the lower animals which are not themselves normally parthenogenetic, but have relatives which are." Occasional parthenogenesis has been frequently observed. An example is the silk moth, in which Nussbaum found that out of 1102 unfertilized eggs ... 22 developed . . . up to a certain point. It is supposed that in all cases of parthenogenesis many eggs fail to develop. In this connection it is noteworthy and extremely suggestive that among higher animals — frogs, hens, and even mammals — the unfertilized ovum occasionally begins segmentation, never proceeding far, however, on its parthenogenetic course. The student should understand that in all probability a large number of eggs fail to develop into complete individuals even in the most successful parthen- ogenesis, just as do many fertilized ova fail along the way (see Weismann, Essays on Heredity, I, 175). There are all degrees of parthenogenesis, from the perfectly successful down to zero. Bearing upon the general subject are the interesting e.xperiments of Loeb in artificial parthenogenesis, especially of the sea urchin, normally bisexual, but which, after immersion in a saline solution of high density and subsequent return to nor- mal sea water, commenced segmentation and afterward developed living larvas. Magnesium and potassium salts proved most effective, though in general any treatment avails that serves to withdraw a portion of the water from the unfertilized egg (see Loeb in American Journal of Physiology, III, 434; also Loeb, Studies in General Physiology, Part II, pp. 576-626, 638-692 ; Methods, pp. 766-772 ; Geddes and Thomson, Evolution of Sex, pp. 1S3-198). INTERNAL CAUSES OF VARIATION 179 asexually reproduced is not "immensely reduced below the vari- ability of the race." ^ In the honeybee only the male sex is produced parthenoge- netically. In plant lice it is commonly the female alone under high temperature, and both sexes under lower. Weismann bred separately two varieties of Cypris rcptans for some seven years, covering more than forty generations and " many thousand indi- viduals." One variety. A, was light in color ; the other, B, was dark. No males were ever discovered in either, and it is sup- posed that this species produces only parthenogenetically. While, for the most part, the descendants of each were extremely alike, yet " minute differences invariably existed." Not only was this true, but in 1887, three years after the experiment commenced, " some individuals of the dark green variety, B, appeared in the aquarium with the light variety." The same variation appeared a second and a third time, and in the last instance intermediate forms could be made out. In 1 89 1 another case occurred, and in the same year its converse appeared, — a few typical light individuals among the dark colony that had "bred true many years." ^ Does this experi- ment also throw light on the origin of varieties, and were these mutations ? However this may be, it clearly shows that vari- ability is not entirely dependent upon sexual union, and that even distinct varieties may arise without the intervention of sex. The first significant fact in maturation of parthenogenetic eggs is that they produce but one polar body.^ From this point on two alternatives seem possible. In the first place, a second polar body appears to be forming in the usual manner and the separation of the nuclear matter takes place, but instead of passing out of the o.^^ it remains beJiind^ fusing again zvitJi tJie nucleus of the egg proper, which straight- way undergoes development, with its chromosomes increased to 1 Pearson, Grammar of Science, pp. 472-473. 2 Weismann, Essays on Heredity, I, 1 61-164. ^ Often this polar body divides, giving the appearance of two, but a second one is not formed. It is quite remarkable, though entirely consistent, in the case of aphis, honeybees, and certain other forms that produce both se.xually and asexu- ally, that ihe fertilized eggs produce (too polar bodies but the purt/teiiogeiiefic eggs only one. l8o CAUSES OF VARIATION the proper number. In this case the second polar body appears in the role of a male element, so we may speak of this as a kind of '■'■ fertili:::ation by the second polar body." In the other form of parthenogenesis, however, there is little suggestion of a second polar body ; certainly no formal separa- tion and later fusing of nuclear matter takes place. On the con- trary, development takes place directly upon the extrusion of the first polar body, and it is significant that individuals a^'ising in this zvay possess but half the number of chromosomes as com- pared with those arising by sexual reproduction or by the method just described. Some species (as in Artemia) ^ reproduce par- thenogenetically by botJi methods, giving rise to two distinct varieties, one with half the number of chromosomes character- istic of the other. ^ Mutation as related to reduction and fertilization. Mutants seem to be departures characterized by a sudden loss of some racial character or by its possession in some unusual degree. They do not appear to be endowed with characters new to the race, except when artificially produced by hybridization. If the process of reduction means the loss of hereditary material, and if fertilization means its restoration, and if either means in any sense new combinations, then we can see in the two phenomena taken together, or even singly, abundant oppor- tunity for the most profound variations ; indeed, admitting their possibility through these causes, the wonder is that they are not yet more common and infinitely more remarkable. If there is in any sense, however slight, a qualitative loss through reduction, then by the law of chance the time is certain to come when something unusual will appear. Is it not more than likely that here lies a fruitful source of sweeping changes, as well as of the more obscure differences that are everywhere about us ? And is it not likely that still greater and more fre- quent changes would present themselves were it not that fertili- zation is for the most part restricted to comparatively narrow lines .'' 1 W^ilson, The Cell, pp. 2S2-283. 2 Artemia thus varies from 84 to 168, according to the particular method observed. INTERNAL CAUSES OF VARIATION i8l SECTION IV — BUD VARIATION ^ Variation is not necessarily connected with reproduction in the ordinary sense of the term. One Hmb of a peach may pro- duce nectarines. A single branch of a tree may assume the weeping habit or the cut-leaved form. Not only are these wide deviations between buds of the same tree well established, but also all shades of differences exist, showing that one part of a plant may vary independently of another, quite after the manner of meristic variation among animals. Bailey 2 calls attention to the fact that the plant is not an individual with a simple anatomy like an animal, but that "its parts are virtually independent in respect to (i) propaga- tion, ... (2) struggle for existence among themselves, (3) varia- tion, (4) transmission of their characters by means of either seeds or buds." Each bud, therefore, has a kind of individuality of its own. All but the first are developed asexually, yet all shades of differ- ences will be found among these different members of what we call a plant or tree ; hence each branch or phyton is a bud variety, and one which can be propagated by cuttings or by seeds or by both, and in either case can doubtless be improved by selection.^ Bailey makes the statement "* that "the seeds of bud varieties are quite as likely to reproduce the variety as the seeds of seed varieties are to reproduce their parents." ^ He quotes Darwin in saying that " moss roses (which are bud varieties) generally reproduce themselves by seed, and the mossy character has been transferred by crossing from one species to another." If this be true, — if bud variations are transmitted by the seed, even to the slightest degree, — then the changes wrought in bud vari- ation must be profound, extending as they do to the constitution of the germ, a fact which argues much for the ever-present 1 Bailey, Survival of the Unlike, pp. 80-106. - Ibid. p. 105. •^ Ibid. pp. 90-92. * Ibid. p. 94. ^ Professor Bailey does not intend to say that seeds of bud varieties are certain to come true, but rather that no seed exactly reproduces the parent plant. 182 CAUSES OF VARIATION liability to internal change and not at all, as is erroneously sup- posed, for the inheritance of acquired characters, because the characters in question were not "acquired" in the ordinary acceptance of the term, — they were the result of internal, not external, impulses. SECTION V — INFLUENCE OF THE CONDITION OF THE GERM UPON DEVELOPMENT Staleness. Both the male and the female germ cells are capable of living for a considerable time after maturation, so that fertilization may be somewhat delayed ; how long is not known, and what the effect of delay may be is not fully understood. Experiments by Vernon upon the ova and spermatozoa of the sea urchin of different ages, from nine to forty-five hours, indicate that the size of the larva is in some degree dependent upon the freshness of the germ at fertilization.^ The results of a number of trials were as follows : 1. With stale ova and stale sperm the resulting larvae differed but slightly from the normal (in which both were fresh). 2. With fresh ova and stale sperm the larvoe were distinctly larger (5.8 per cent). 3. With stale ova and fresh sperm the larvae were distinctly smaller than when both were fresh (4.9 per cent). It is certain that the above combinations as to staleness are possible in the fertilization of mammals by mating and of plants by pollination. Whether the results are the same and whether the differences persist through life are, of course, unknown. The facts recorded are suggestive, but whether they will ever be useful remains to be determined. Individuality of the germ. That successive germ cells from the same individual may be substantially different, even aside from considerations of maturation, is a fact beyond question. The ear of corn, like its tassel, matures from the base upward. The tip kernels are not only younger but decidedly smaller than their half-sisters at the base. The different peas in a pod are ^ Vernon, Variation in Animals and Plants, pp. 105-108. INTERNAL CAUSES OF VARIATION 183 not equally developed. One of the twin pair of oats is more or less undeveloped. Is this difference in size due to season, food supply, room, or to some peculiarity in the germ .? It may be lack of room in pod-bearing plants, but it cannot be that in the case of corn. The strong presumption is, in the opinion of the writer, that these differences in size are partly due to differences in food supply but more largely to inherent differences in the germs. SECTION VI— XENIA, OR FERTILIZATION OF THE ENDOSPERM, — DOUBLE FERTILIZATION If one kind of corn be fertilized by another, the mixture will show the first year. For example, if white and yellow corn be planted side by side, the white ears will have many yellow grains, showing at once the effects of cross fertilization. These " off" kernels are the mixed seeds, but, reasoning from analogy, we should not expect the mixture to appear until the grains are planted and the generation of mixed breeding is at hand. The visible part of the kernel is not the germ ; it is the " endo- sperm," or surrounding portion, which serves as food for the sprout until the young plant has established itself. It is related to the germ much as the white of ^gg and its shell are related to the yolk. Fertilization is of the germ. How, then, do these outside parts become affected } It will be remembered that in the animal the female germ gives rise to one mature functional cell, the ovum, and three non-functional, the polar bodies ; that the male cell gives rise also to four mature cells, the spermatozoa, all functional, and that the nucleus of the one unites directly with that of the other witJwiit intervening nuclear divisions. In plants, however, it is found to be substantially different. The mature female cell, corresponding to the ovum, instead of awaiting fertilization, continues its activity, undergoing gener- ally two (sometimes more) divisions of the nucleus, giving rise to eight, or some other corresponding number of " sub-nuclei," which remain floating within the cytoplasm. It will be remem- bered that of these eight sub-nuclei only one is capable of func- tioning as an Q.gg nucleus ; also that two others remain near 1 84 CAUSES OF VARIATION the center of the embryo sac to form the endosperm. It will be remembered, also, that the pollen nucleus undergoes a second division during its progress down the pollen tube and before uniting with the Qgg nucleus. Of this divided nucleus one portion unites with the single functional member of the female group, making the germ, and in cases such as are now under consideration the other joins with the minor members concerned with the development of the endosperm. In this way, by means of this kind of double fertilization, the endosperm is itself affected and the crossing is evident the first year. Of course this visible effect upon the endosperm is of itself purely transitory, having no influence upon the line of descent. The real effect of pollination is manifestly, as in all other ferti- lization, confined to the germ. Whether the two fertilizations are similar as to comparative influence of the two parents no one knows, nor does it greatly matter. The effect upon the endosperm enables us to detect the cross, if it is capable of detection, and to remove the con- taminated seeds if we desire to retain purity. If the object be to secure crossing, we shall of course subsequently deal with the products of the cross-bred germ, which only are significant from the breeder's standpoint. Just what species indulge in this double fertilization is not well known. It is, however, well established in a large number, and the process is supposed to be common rather than unusual. Effect of crossing upon fruit in general. What the layman calls fruit is commonly not the endosperm that has been under dis- cussion but the thickened and much developed fleshy receptacle on which the seeds are borne. It has been claimed that these parts are directly influenced the first year by crossing, so that the character of strawberries, apples, pears, melons, squashes, etc., depends much upon the source of the pollen used in fertilization. This claim has never been well substantiated by direct experi- ment. Dr. Burrill, of the University of Illinois, tells me that he crossed Crescent strawberries both with the Sharpless and with a wild berry especially selected for its insignificant, worthless fruit. Nobody was able to detect the difference in the resulting INTERNAL CAUSES OF VARIATION 185 crops. So far as is known to the writer, the same principle holds in other fruits. It is the endosperm and not the receptacle that is directly affected by fertilization, and any influence upon the latter must be indirect and comparatively slight. Possible indirect effect of pollination upon the development of fruit. Though the receptacle is not itself fertilized, its develop- ment is conditioned upon that of its superincumbent seeds, which are themselves directly dependent upon fertilization for their development. This fleshy growth of the receptacle is, therefore, the result of a kind of stimulus from the growing germ, and it is con- ceivable that this stimulus may differ somewhat in degree, depending upon the source of the pollen. In this way the size of the fruit might be indirectly influenced by the pollen ; and in fruits like the pear, which are not concentric about the seeds, even the shape might be influenced in the manner noted. All this is quite independent of certain markings of fruit which may arise by those dispositions of color which are every- where responsible for stripes and spots, and whose causes are not as yet understood. SECTION VII— TELEGONY The term " telegony " is synonymous with "infection of the germ" and the "influence of previous impregnation." By this is meant the supposed influence of the male upon the female in such a way as to affect future offspring by other sires. Breeders of animals quite generally believe that the influence of one impregnation, especiafly the first, is permanent and wifl affect all future offspring ; indeed, some go so far as to say that a female once mated to a male of a different breed is ever afterwards, for breeding purposes, herself a cross-bred animal} This supposedly permanent effect of the male upon the female has been especially claimed for horses, dogs, and men. Telegony in horses. The classic example among horses, and the one that is everywhere cited as proof of the theory, is the 1 This theory seems to be limited to animals. The writer is not aware that it has ever been claimed for plants. 1 86 CAUSES OF VARIATION instance of Lord Morton's mare mentioned by Darwin.^ This mare bore a colt by a quagga, which was of course striped after the manner of his sire. She afterwards bore two colts by a stallion, both of which were said to have been marked with bars on shoulders and legs supposedly showing the effects of the quagga upon the offspring of the stallion. Professor Ewart, of Edinburgh, has recently repeated this experiment on an extended scale, with results showing no trace of the quagga beyond his own offspring.^ Recent investigations in contemporary literature throw grave doubt upon the essential accuracy of the data at Darwin's hand. They seem to show that the supposed resemblance of the stallion colts to the quagga was exceedingly fanciful, probably being nothing beyond what appears frequently in young horses of the purest parentage, dun-colored horses as a rule showing more or less tendency to stripes and bars. It is one of the best evidences of the power of tradition that this single instance, happening more than a hundred years ago, has done duty ever since to prove (.?) an exceedingly doubtful theory and an almost unaccountable belief. It is remarkable that so uncertain a circumstance, and one so easy of repetition, with universal experience tending constantly to throw light upon the subject, should have been so excessively overworked. It shows, as no other instance has ever shown, the persistence of tradition, the extent of credulity in the presence of the phenom- enal, and the willingness of men to repeat an assertion, or even an opinion, until by mere repetition it comes to have all the force of authority. The thanks of the world are due to Professor Ewart for his excellent work in disposing, by direct experiment, of a citation that has done damage long enough. It is to be hoped that the question may at least be held open until some sort of positive evidence is brought forward that is worthy the credence of careful students. Telegony in dogs. Dog fanciers are pretty generally credited with believing in telegony, especially in case of first matings. ^ See Darwin, Animals and Plants under Domestication, chap, xiii, p. 17, of second edition by Appleton. (Quoted from PJiilosophical Transactions, 1821, p. 25). " Breeders^ Gazette, XLI, 1009. INTERNAL CAUSES OF VARIATION 187 The best students, however, insist that very little real evidence has been produced on the subject, and none at all tending to prove the existence of this influence.^ With a view to testing somewhat the real extent of this be- lief, the author addressed letters to the best-known dog fan- ciers of the United States. Of thirty-seven answers received, one writer is a believer in telegony ; six somewhat mildly express uncertainty ; two are non-committal ; and twenty-eight are outspoken against the theory. The most outspoken of them all is a well-known fancier of long experience. Judging from this small number, it would seem that this belief among dog fanciers has been overrated. Proof by the method of instance. Without a reasonable doubt belief in telegonic influence rests upon stray instances, difficult of understanding by those who happened to be the observers, and hastily accepted as evidence. Now nobody should be more careful than the breeder to judge accurately the nature and value of evidence. A single instance may be good negative testimony, but it is seldom worth much as positive evidence. The products of breeding are so many and so various, and the causes of variation are so numerous and so complicated, that a particular result can seldom be assigned to the operation of any single cause. It is more likely the mixed or composite result of many influences, both internal and external ; and in order to know the effect of a single cause it is necessary to isolate the case if possible, or, if not, to resort to the examination of large numbers of cases, subject to varying degrees of influence, and thus indirectly to estimate the effect of any special cause of vari- ation. For example, stripes and bars were once common color markings of horses, as they are now of asses, especially zebras and quaggas. Consequently a certain proportion of colts, what- ever the parentage, will be born with traces of shoulder and leg markings. Now, under the laws of chance, a certain portion of these will be the direct offspring of striped or barred sires, and will attract no attention, the markings being considered heredi- tary. By the same law of chance a certain (smaller) portion will be the offspring of parents not barred, and a still smaller number 1 Proceedings of the Royal Society, LX, 273. l88 CAUSES OF VARIATION will be the progeny of unbarred sires and out of dams once mated with baiird sires for other offspring. This smallest portion, get- ting its bars not by direct descent but by reversion, will most likely be erroneously considered to have derived them from the barred male not their sire. The same is true of other markings, and on such evidence as this the theory of telegony has been built up, and, so far as proof goes, it rests on no better founda- tion as yet. In order to secure evidence amounting to proof, it is necessary to examine large nnmbers involving both positive and negative evidence, in order to secure trustworthy averages. When- ever this has been done the theory of telegony fails of support. Telegony in man. The statistical method has been applied by Pearson^ in the case of man. He, together with Galton, pos- sesses data covering hundreds of individuals in English families. He reasoned that if the sire exerts a permanent influence upon the dam, tending to assert itself in all future offspring, then this influence must be in a sense cumulative, so that the younger sons in the family will tend to resemble the father slightly more than will the older sons, conceived before such influences have become established. His study covered 385 brother brothers and 450 sister sisters, taken two and two. In some instances there was considerable difference in ages, and in others they were successive children. His data covered both height and arm length, and after making the usual allowances for sex and age Pearson concludes that, so far as these characters are concerned, " n(j steady telegonic influence exists." Again, the many successive marriages of both colored and white women to men of opposite color should afford numerous examples of telegony were it a consequential force in heredity. Scientific objections to the theory of telegony. If telegony exists, its influence over hereditary characters must be explained, so far as present knowledge goes, in one of three ways : (i) some effect upon the tissues of the female such as will influence future ova in their maturation or the embyro in its development ; (2) something like a partial fertilization of immature and unde- veloped ova, in such a way as to influence their character at ^ Proceedings of the Royal Society, LX, 273. INTERNAL CAUSES OF VARIATION 189 maturation ; (3) the retention of the spermatozoa from the first mating, and their action in successive fertiHzations. As to the first, there is no scientific ground for assuming the sHghtest effect of the spermatozoa upon the tissues of the female. It is the ovum that is fertihzed, not the female, as was at one time supposed when fertilization was regarded solely as a stimulus. As to the second, there is no ground for believing that the nuclei of growing immature oogonia are in condition to unite, or that they are capable of uniting, with the nuclei of other cells in any capacity whatever. As to the third, there is every ground for believing that the spermatozoa are not retained for any considerable time, else successive births would occur from a single mating. Moreover, as but one spermatozoon takes part in fertilization, the blended effect of two sires is impossible. It is even impossible in multiple births when two services are close together. If a litter of pigs is the result of two matings by different sires, some may resemble one sire and some the other, but none will resemble both. SECTION VIII — INTRA-UTERINE INFLUENCES It is a widespread tradition that distinct characters, especially abnormalities, may be impressed upon the individual while /// 7itcro through the imagination or other strong mental impression of the mother. It has even been held in the case of hens, which would necessitate the exertion of the influence upon the ovum itself. The usual argument is that thg intimate contact between the mother and the fetus renders the latter peculiarly susceptible to influences affecting the former. Thus marks and deformities of all sorts are popularly attributed to unfortunate sights and experiences of the mother before the birth of the young. Pecul- iarly marked calves are said to owe their markings to the strong mental impressions created by a steer or by other cows, and colts are believed by many to owe their color not so much to the sire as to the gelding mate that worked beside the dam while she was carrying her young. Persons with whom the tradition is strong often display a blanket of a pleasing color before the eyes of the I90 CAUSES OF VARIATION mare at the time of service, and of course are extremely care- ful to protect her from unpleasant colors of any sort.^ The hold of this theory upon the popular mind is the best example afforded by breeding of the strength of tradition. The supposed reason on which it rests has slight basis in fact. The contact between the mother and the fetus is not so intimate as is popu- larly supposed. The fetus is absolutely dependent upon the mother for nourishment, it is true, and it lies floating in its fleshy incasement, which is in intimate contact with the tissues of the uterus ; but there is no organic connection, no nervous interrelation whatever. Anything which would curtail or shut off nourishment would of course injure or destroy the fetus. It is also subject to other accidents, as becoming entangled in its own cord, which may thus divide a limb or cause strangulation, — all of which, how- ever, is quite aside from the matter in point. The real question is whether, and to what extent, the fetus is influenced by peculiarities of nourishment during its develop- ment. It would of course be injured by poisons, and the danger from administering anaesthetics is great, but this discussion is limited to the direct effect of mental impressions. The indifference of the fetus to its source of nourishment is shown by an experiment of Heape,^ performed for another pur- pose, but throwing light upon these questions. In this experiment " two segmenting ova were obtained from an Angora doe rabbit which had been fertilized by an Angora buck thirty-two hours previously, and were immediately transferred to the upper end of the Fallopian tube of a .Belgian hare rabbit which had been fertilized three hours before by a buck of the same breed as herself. In due course this Belgian hare doe gave birth to six young. Four of these resembled herself and her mate, but the other two were undoubted Angoras.'*^ . . . Both of the Angoras were born bigger and stronger than any of the other young, and 1 For a good collection of alleged instances, see Miles, Stock Breeding, pp. 281-295, or consult any neighborhood oracle. 2 Vernon, Variation in Animals and Plants, pp. 1 19-120; also Procei'dim^s of the Royal Society, XLVIII, 457. 3 The Angoras were characterized and easily distinguished by their long, silky hair and their habit of swaying the head from side to side. INTERNAL CAUSES OF VARIATION 191 they all along maintained their supremacy in this direction." Whatever this experiment proves or does not prove, it shows conclusively that a fertilized Angora germ preserves and develops its inherent characters perfectly in an exceedingly foreign envi- ronment, on which it evidently depends only for nourishment. Mental impressions and nervous conditions are commonly in- voked to explain birth marks and other natural abnormalities, such as the loss of a finger. In this connection two facts are to be carefully considered : first, there is certain to occur a large num- ber of marks (" strawberry," " cucumber," and others) and many malformations of one kind or another. Scarcely an individual is absolutely free from something of the kind. Again, mothers are subjected to all sorts of sights, sounds, and experiences during the many weeks of embryonic development, and it would be strange indeed if out of the thousands of cases some correspondence between marks and experience could not be figured out, espe- cially by one whose belief is fixed and who, having the case at hand, needs only to find the proper " corresponding experience." The law of chance alone will insure an occasional correspondence between the two, — entirely enough to start the tradition and to maintain it afterward. As in theories concerning the control of sex, any theory stated will be verified half the time because there is but one alternative, so here, while the alternatives are more, the correspondence is certain sometimes to appear under the law of chance alone. Another fact to be reckoned with is that if the fetus were so sensitive to mental impressions as to require the display of properly colored blankets, — if females were so susceptible as this to surrounding sights, — what a jumble of colors our domes- tic animals would speedily display. In the opinion of the writer this tradition has neither a scientific basis nor well-established instances, and it is time it no longer occupied the minds of breeders to the exclusion of far more important matters. In this connection it is worthy of remark that if the average breeder were half as familiar with important facts, and half as attentive to their bearing upon his operations, as he is familiar with and attentive to floating traditions, we should have a far smaller proportion of worthless animals. 192 CAUSES OF VARIATION SECTION IX — REVERSION AND ATAVISM These two terms are used to designate characters appearing in the offspring but not visible in the parents. " Reversion " is used to indicate resemblance to a comparatively near-by ancestor, not the parent, while "atavism" refers to exceedingly remote ancestors, sometimes of other and foundation species. Thus, if a dash of impurity of blood enters a herd, it will appear occasionally for many generations. This would be spoken of as a " reversion to the strange blood." If a sire or dam has some peculiar character, like white stockings in horses, a peculiar horn in cattle, or a habit in man, it is likely to appear from time to time in future generations, even after its real origin is forgotten. This is a reversion. English breeds of cattle are developed from the ancient wild white cattle of Britain, and the occasional appearance in all these breeds of a white calf with red or brown ears, lower legs, and tail brush is to be expected. It is a reversion, not a proof of mixed blood. Of course the animal so marked is useless for breeding purposes, but no reproach to the herd, and none necessarily to the dam that produced it, for reversions for the most part seem to come singly. Atavism, on the other hand, goes farther back. For example, mammals during their early embryonic development still show traces of the gill slits, thus betraying their undoubted one-time connection with the same stock which gave rise to the aquatic animals. These gill slits occasionally persist, failing to close, and give rise to the abnormality known as "cervical fistula." It is an undoubted atavistic abnormality, — an extreme case of course, because of its antiquity. Cases of this kind are to be carefully distinguished from mere meristic variations. For example, the sudden appearance of a three-toed horse would be regarded as atavistic, for all horses once had three toes ; but a sixth digit in man is certainly not atavistic, for we have no evidence that man ever possessed normally more than five digits. The criminal instinct in certain men is undoubtedly atavistic, showing not so much a delight in evil doing as an entire absence of the higher sense of right doinof. INTERNAL CAUSES OF VARIATION 193 There are certain intermediate cases difficult to name. An occasional cow gives no more milk than her wild progenitor ; a hog resembles not his immediate kind, but his striped and long- nosed ancestor, the wild boar of the bush ; the horse or the sheep paws snow the first time he sees it, though cattle do not ; the dog turns many times around in lying down, as if making his nest ; the occasional horse has bars on his legs and stripes on his shoulders. Which term shall be applied .? These are undoubtedly line cases, and all would not agree as to whether they should be regarded as reversions or instances of atavism. Because of our very frequent need for a term to cover experiences met almost every day by the breeder, and due to more near-by causes, the writer is of the opinion that it is better for our purposes to extend the meaning of the word " atavism " well forward, making it cover cases of remote characters, even within the species (like those just given), leaving the term " reversion," which will be much more frequently needed, to cover the more near-by cases that occur every day in our herds, and that would be traceable, could we know all the facts, to an old but not remote ancestor. Inheritance is from the race. It is evident that inheritance is not limited to the visible characters of the immediate parent. We constantly forget that every individual is possessed of and capable of transmitting all the characters of the race to which he belongs. We forget that his visible characters are not his total possession, but only those which are relatively most prom- inent in his case. Other combinations are easily possible out of the same elements in slightly different proportions, and it is not so strange as we think that a character once in possession of a race tends to persist indefinitely and, perforce, occasionally to become visibly apparent. It is as bound to appear, under the law of chance, as is the one black ball in the box of a thousand or a million, if only throws enough are made. The work of Galton, while mostly confined to man, yet shows clearly that inheritance is not in strict line with the visible parental characters, but is in large measure independent of the immediate parents.^ ^ See the Regression Table, sect, iii, chap. xiv. 194 CAUSES OF VARIATION Galton has endeavored to assess mathematically the fraction of direct inheritance, or, more accurately, the similarity between the child and its various ancestors. From his studies he con- cludes that the visible or dominant characters of the child are, on the average, inherited (that is, correspond with those of the various ancestors), roughly, as follows ^ : From the immediate parents 50 per cent From the grandparents 25 per cent From the great-grandparents 12.5 per cent From the great, great-grandparents . . . 6.25 per cent Earlier ancestors in proportion Pearson, working with larger numbers and diverse characters, concludes that Galton's fraction of direct inheritance (0.50) is too high, and is inclined to believe it not above 0.40 for blended characters. The subject will be pursued farther under " The Law of Ancestral Heredity," but this glimpse into the nature of inheritance is the best method known to the author to dissolve the almost supernatural mystery that has been thrown around reversion and its natural corollary, latent characters. The study can be pursued no farther at this point, but what has been said will serve to show that reversion (regression) is a fertile cause of variation as calculated from the type of the parent. It will serve also as an introduction, preparing the student for the more serious study of inheritance later on, when we shall learn that the real type, from which all departures should be reckoned, is the type of tJie race, and not the special type of the parent, or even of the mid-parent. SECTION X — INDIVIDUAL CHARACTERS DEPENDENT UPON SEX That both sexes possess and transmit all the characters of the race is a well-established fact in evolution. It is also true that the particular characters to undergo development, and the extent 1 Vernon, Variation in Animals and Plants, p. 123; Proceedings of the Royal Society, LXI, 401, 1897; Galton, Natural Inheritance, p. 191. This does not mean that every individual will inherit in this proportion, but that the fractions express averages. INTERNAL CAUSES OF VARIATION 195 to which they will develop depends very much upon the sex of the individual. How much allowance to make on account of sex in comparing one individual with another of a different sex we are in most cases unable to say, — not from the impossibility of knowing, but from the fact that in respect to most characters the matter has not yet been worked out. It is easily possible, however. For example, in respect to stature, men are 8 per cent taller than women, so that when the heights of the latter are multi- plied by 1.08^ the two are strictly comparable, and not before. When this is done the difference due to sex has been eliminated, and the statures of men and women may be directly compared. In general, males and females exhibit the same characters, but in varying degrees. For example, the woman as well as the man has hair on the face, but in less amount ; the male as well as the female has nipples, but they are rudimentary. Among mammals and the domestic animals generally the male is heavier in front, generally of a more robust build, and considerably larger than the female, — a distinction that by no means holds in animal life generally. In the present state of knowledge we simply know that the general appearance of the individual and its character devel- opment are largely dependent upon its sex ; but to what exact extent remains in most cases to be determined, and the deter- mination must be made before we can compare individuals of different sexes with any degree of accuracy. Without a doubt distinctions in sex have been greatly overworked, the differ- ences being mostly of degree rather than of kind ,2 and far less consequential than has been supposed. Individuals deprived of their sexual organs by castration or by spaying do not develop their primary sexual characters. The castrated male is not a female, as is sometimes erroneously believed, but a male arrested in his development ; and the spayed female is an undeveloped female. As would be expected, both ^ These data are the result of Gallon's study of the stature of English people. See Galton, Natural Inheritance. 2 It is idle to attempt to prove that certain characters, aside from those of reproduction, are especially identified with one se.v. 196 CAUSES OF VARIATION take on the secondary sexual characters (those dominant in the other sex) prematurely young.^ Many entire individuals never develop strongly the primary characters of their own sex. There are effeminate males and mascuHne females, — those in which the characters of the oppo- site sex are unusually developed. It is needless to say that such individuals are not the best parents. , • 11— INTERNAL INFLUENCES AFFECTING THE RACE AS A WHOLE Over against those causes that may operate in the case of each individual to warp its development are to be considered those that influence the race as a whole, turning the line of descent more or less permanently aside from former channels. Some of these influences are clearly defined and easily recog- nized ; others are problematical, the discussion not having yet passed beyond the stage of a plausible theory. The student of thremmatology and the breeder should be always mindful tJiat the purpose of all good breeding is not simply to hold what zvc already have bnt to produce nezv types better adapted than the old to the purposes of man. Accordingly any and all lines that promise any hope of success should be assidu- ously investigated. SECTION XI— RELATIVE FERTILITY, OR GENETIC SELECTION ■' The assumption that all members of a race are equally fertile /;/ se and inter se (of themselves and between each other in all directions) is not only hasty but dangerously incorrect. To quote Pearson, " Fertility is not equally distributed among all individuals." 1 The entire animal with increasing age, its own sexual characters abating, begins to take on those of the other sex. Thus the hen grows spurs, the cow bellows and paws the dirt, women grow scanty beards, and old men's voices grow light. ■^ Pearson, Grammar of Science, pp. 376, 414, 437-449, 462. INTERNAL CAUSES OF VARIATION 197 If this be true, and practical breeders know that it is true, then an interesting and important question at once arises ; namely, What characters are correlated zvitJi the highest fertility ? This is important, because these are the ones that will become the dominant characters of the race, certainly unless opposed by the most rigid selection or by other powerful influences. This is genetic selection, — an ever-present influence over the line of descent, tending to establish what might be called a natural type. Unfortunately we possess no accurate data on this point among domestic animals, but Pearson's work ^ among men and plants is sufficient to settle the principle that such a definite influence exists. He finds, for example, that daughters are not taller than their mothers, but that they are taller than wives in general. Now not all wives are mothers, and these data mean simply that taller women are on the average more fertile. There is thus some correlation between fertility and stature. This is genetic selection, and under it the stature of women (English) may be expected to gradually increase until such correlation is satisfied, unless held back by other influences. Mothers are less variable, but daughters more so, than wives in general ; that is, progressive selection exists, for not all daughters marry, and not all who marry produce young. If the standard deviation from the race zvere the same for each, then no selection wonld be involved, but it is progressively less from daughter to wife and on to mother. TJie difference betzveen daughter and rvife is due to preferential mating, the especially ugly ijidividuals being less likely to find a mate ; but the differ- ence betzveen wife and fnother is due to relati'-oe fertility. The fact that in general the mother is nearer the average than is the wife shows that the race is fairly stable ; but the fact that wives are shorter than mothers has but one meaning, — that in respect to stature the race is yet unstable. Extensive studies in eye color indicate that dark-eyed indi- viduals, both men and women, are slightly more fertile than are the lighter-eyed. This means that the dark -eyed will progress (increase) upon the light-eyed and the race will grow darker-eyed, 1 Pearson, Grammar of Science, pp. 441-445. 198 CAUSES OF VARIATION unless the tendency shall be held in check by the greater attract- iveness of lighter eyes, — preferential mating. This would be a long and slow process, but it would avail much to reduce, though it could never overcome, the effects of the higher fertility of the darker-eyed individuals. Pearson collected 4443 capsules of wild poppy. ^ They showed the following distribution arranged according to the number of stiermatic bands : Bands .... 5 6 7 8 9 10 1 1 12 13 14 15 16 17 iS 19 Frequency. . I II 3- 56 148 363 628 925 954 709 397 155 51 12 I The largest number of capsules (954) had 13 bands and the next largest number had 12. Very few had so many as 18 or 19, or so few as 5,6, or 7. The type number of bands is then 13. He provided receptacles and kept the seeds of each group separate. He says : To my great surprise, however, my receptacles for 12 and 13 were filled up with the contents of very few capsules, those for i i and 14 more tardily, those for 10 and 15 only with emptying a great number of capsules, while I could hardly get any seed at all from those capsules with very many or very few bands ; they were practically sterile. The type capsules were enormously fertile, [while] those with even a moderate deviation from it [were] relatively or even absolutely infertile." This being true, the poppy has become about as stable as is possible, for its highest fertility is with its most numerous popu- lation. This plant was growing wild in nature. Obviously the great bulk of seeds distributed would be of the type number, 13 or near it, and the mass of descendants would arise from seeds close to the type. What chance now would there be in nature for a large colony of six-or seven-banded strains to arise } Very little, unless they happened to possess some decided advantage in the struggle for existence, in which case the type would speedily shift in that direction ; but as long as the highest fertility remained with the higher number of bands, the race would be unstable. 1 Pearson, Grammar of Science, pp. 443-444. 2 Ibid. p. 444. INTERNAL CAUSES OF VARIATION 199 Suppose it were the purpose of man to develop a poppy with fewer, or with more, than the natural number of bands, — say seven or seventeen. Under what disadvantage he would work as long as the fertility remained relatively low ! and he would never succeed unless he separated the plantings from the more prolific type. This is genetic selection. Breeders are constantly operating against the drag of infertility without knowing it, and are as often wondering why better results do not follow, especially when only approved mating is practiced. Consider the mathematics involved in, say, three lines of descent of different degrees of prolificacy. For the sake of simplicity in illustration let us suppose three cows were living in a herd together. One of these cows raises two calves and becomes barren ; another raises four before she ceases to breed, and another six. For the sake of further simplicity let us suppose that one half the calves are females, and that each daughter descendant exactly repeats the performance of her dam and then becomes barren. How will the account stand in a few generations ? ^ Cumulative Effects of Fertility as shown by the Rel.ative Number of Female Descendants of Cows of Various Degrees of Fertility Cows Calves Generations I 2 3 4 5 No. I No. 2 No. 3 2 4 6 I 2 3 I 4 9 I 8 27 I 16 81 I 243 This tabular presentation shows that after five generations of this kind of breeding there would be but on^ fertile cow of the first order in the herd,^ while if all had been kept there would be 32 of the second order and 243 of the third. What an 1 There is no longer any doubt that fertility is an inheritable character. 2 There might be any number of living barren ones if the strain happens to be a favorite and is long-lived. 200 CAUSES OF VARIATION opportunity for selection in the latter instance, with none what- ever in the former ! The first untimely death would render the line extinct, which is perhaps the best fate that could overtake a race which at best is able only to hold its initial number good. Of course artificial conditions have been assumed in order to bring out the principle. It does not work out in this regular and evident manner in our herds, but the principle of genetic selection is at work, nevertheless. It would be fortunate if it were more evident, for the herds are few that do not contain a large proportion of females that contribute nothing to the real line of descent, though they now and then give birth to excep- tional individuals. The quality is good, but the rate of repro- duction is too low. How many a breeder has spent fruitless years in ineffectual attempts to build up a strain excellent in itself but essentially infertile ! Witness the fate of that remarkable family of short- horns, the Duchess. This famous family, in its glory, was never surpassed, yet it was genetic selection that exterminated the line. Fortunate indeed is the breeder who knows this principle and realizes its full power whenever he finds himself opposed by it. The student must not get the impression that genetic selec- tion is an enemy only. A prolific line tends as strongly to establish and maintain itself as does a barren one to rush head- long to extinction. Genetic selection is therefore a friend pow- erful for good, as well as an enemy powerful for evil ; but it is as quiet and unobtrusive in the one relation as it is insidious in the other. The breeder has only to be eternally conscious of the fact that if he is to succeed he must have numbers, not occasional births, but regular and generous. Then he may be sure that he is not trying to do a thing on which nature has set the seal of her disapproval through non-production. However worthy and however valuable intrinsically the strain may be, it is worthless unless he can produce it with certainty and in any desired numbers. " Beware of the shy breeder, and treasure the old female that breeds regularly and true." This doctrine estab- lishes a cooperation with nature that insures results, and with- out it genetic selection will work against us, not for us. INTERNAL CAUSES OF VARIATION 20 1 SECTION XII — PHYSIOLOGICAL SELECTION The term "■ physiological selection " refers to the fact that cer- tain individuals, fertile enough of themselves, will yet absolutely fail to breed with 2^ particular individual of the opposite sex.^ This principle is now well established and is recognized as a large cause of fruitless marriages. Individuals are frequently barren in one marriage and perfectly fertile in another. Physio- logical selection is a phase of genetic selection, and while of extreme importance in the marriage relation it constitutes no special menace to our herds. In general it has little bearing upon the development of a breed, but is often exceedingly troublesome when it is desired to effect a particular combination of blood lines. SECTION XIII — SELECTIVE DEATH RATE; LONGEVITY The total population depends as much upon longevity as upon fertility and the prevailing type at any moment depends as much upon the individuals that die out of the world as it does upon those that are brought into it. If the draft by death is equal, or rather proportional, from all types of the race or breed, then the existing type will be the same as that born into the world ; if not, it will be different. As there is little use in attempting to breed a strain, however desirable, that is not at least fairly prolific, so there is little use in spending time and expense upon short-lived strains, especially of milch cows and horses, which are valuable largely in propor- tion to age. For reproductive purposes the " age " of an animal is the age at which he stops breeding, but for other purposes it is the age at which he can no longer render valuable service in the desired direction, such as labor. 1 This principle was first announced by Romanes (" Physiological Selection," Journal of the Liniuraii Society, XIX, 337-41 1), though what he had in mind evi- dently included what is now known as "genetic selection." It was proposed as showing that other principles are at work to fix types, aside from the struggle for existence. 202 CAUSI'lS Ol' \ AKI AtlON VVi'isniimn ' believes lluil in naline llie deatli point has been lixcil at an a^e niosl piolilable to tlie lace as a whole. Tliat is to say, it is best lor the raie (i) ihal onl\ llie slionj^i-sl sliould sui\i\e to the hii-edinj;' aj;e ; (.') thai these should live as lon^ as theN aie able to leproduei-; and thai (•;) lhe\' should then die and (H'ase lo oe( up\' looni and t onsuine lood whith would other- wisi- be axailabU- lt>i the sustenante ol nioie robust iiulivi(hials cn^aj^ed ii\ lepiodui lion. This li.\es the death jtoint theoretie- ally at llu' ei'ssation ol n-pioduetion, e.xeept in sueh spc-iies as those in uhieh the \ount; neeil the eare or the eduiatixe assist- ance ol the niother. I n thest" the theoietieal death point would bo at the nialuiily ol the last xouni;. 'This ol eouise is in lelerenec to wild species, and Weisniann believes that natuie has estab- lished the death |)oinl in elose eorrespondenee to this piineiple. llowi'xer that ina\' be, then- is a piobleni here lor the breeder, it is loi him to li\ the death limit well be\'ond the pc-riod ol tin- particular service rcqiiircit. In natuie there is but one object in lite.— sell preservation and repuxhulion. (^n our farms there aie olhei objeils. The horse is lor labor, and his sei\iceable ai;e as well as his dei;ree ol intellij;eiue needs lo be lengthened as nuuh as possibK-. In nature early and ra|)id leprochution is a lull e(|Ui\aleiU lor loni;e\it\'. it is not so on oui' laiins, wheie the u\di\idual counts lor more, and e\en lapid leproduction I annot lake the plaie ol lonj; lile and laithlul service. Si'.CriON \1\' r.Al'ilMU" INI'Il'llNClCS I )o species possess inheriMit UMulencies to \ai\ .■' II a race could be surrounileil by positiveh unehani;ini; conditimis, if it louKl pioduie asexuallv. and il all types were eiiually vii^orous and ei|ually lertile, K'oiild it remain constant? Some variation would aiise lhiouj;h reduction, but this wouKl be heleroi;eneous, — that is, now in one tlirection, now in another. The real c|ues- lions the balhmic evolutionist asks aie these : Is there a tend- enc\ lor the t\ pe to ilrill, indepeiulenl ol selection or surroundinj;" inlluences ^ Are its tiexiations iharacteri/.eil by a continual bias ' Wi'isinaiiM, l'",ss.iys on 1 K-iiHlity, 1, i i i i(>j; ,^00 also I'tMisoii, Cluiiuos of l>iMlh, PI). \-.\2. INTI'lkNAI, ('AlISi;S Ol' VARIy\ri()N 203 in favorite dircclions ? Docs it vary progressively because impelled in these directions by "^Movvtli force" or other inherent energy? Are s])ecies iu-hi to tlu-ir |)resent standards by oiitsitle influencx-s ? or, if ncU " held," arc they driftinj^ in spite of us? Is the life principle constant or |)eriodic in its activities; and are those internal energies that vitalize matter and that determine development and differentiation, are they indifferent as to the trend of the type, or do they run more easily in some channels than in others? Is variation in some sense subje( t to and directed by a natural bias? This is the field of bathniic evolution,' and these are the (|uestions involved. No one is more interested in their discussion thini is the bree(K-r ol domesticated forms. Two principal theories covering the field of bathmic evolution have been ])roposed, both incaj)able of absolute proof, as all such theories nnist be, but both of interest to the- breeder. Acceleration or retardation of growth force. This print i])le is annouiK ed by ("o|)e- as an intcM'iial and e\er-|)resent cause ol progressive evolution, rumiing thiough all loinis ol lile and beneath all ordinary influences, impelling unnoticeably but irre- sistibly in certain directions. It is, after all, according to this author, the most subtle and most potent cause; ol departure; I rom type. The horse has undergone steady progressive development from an animal of the si/.e of a jack rabbit up to his present |)ro|)ort ions and perfection. This is due, a((oi(ling to ("ope, not so much to selection as to a continuous, peilia|)s almost unprecedented, araliration of ^i^iv7vth force. This theory attem|)ts to e.\])lain much of evolution through the energy of growth, thus throwing into the discussion a dynamic element commonly neglected by evolutionists. In this connection Pearson pertinently remarks: 'I'licrc is iiotiiiiij; more (of less) mis( icMitific in iisin^ an inherent growth force to explain (lie .se< niar c lian^e.s in living forni.s tiiau in nsing tlie force oi f^ravitalion inlierent in tnatter to explain the (levclopnicnt ol jjlanetary ' i'e.irson, ( iiaiiMiiai of Si icn(c, ])p. t,-j <:, 577. 'i'lic (cini "l)atliini< " as liere used does not include j;enelii: selcclion 01 any olliei sclei live aj^enl, internal or external, l)etaiise tlie effects of all su( li influences lend to come to a rest and not to constitute a "continual bias." '■^ (Jope, Origin of the Fittest, pp. i}i-.30, ic;o-iy2, y)(>-y)^ \ I'riniary Factors of Organic JCvolution, pp. .17.3 -.1'>1. 204 CAUSES OF VARIATION systems from nebulie. The ultimate action of vital units in each other's pres- ence would be no more, nor less, of a mystery than the ultimate action of material units. . . . The real objection to bathmic evolution lies not in any a priori reason against an inherent growth force, but to the obvious histor- ical fact that such a force has been used to cover all sorts of obscure reason- ing and even sheer foolishness. Science would welcome above all things a description of the action between vital units as simple as the law of gravi- tation, provided it gave a causal account of variation ; and the welcome would be none the less sincere if the action showed that variation was biased and that evolution would be irreversible even with a reversed sequence of physical environments.^ Cope's theory of acceleration or retardation of growth force is of course merely quantitative, and would explain any differences that might arise through either size or proportions of parts, or faculties dependent upon such proportions. It does not attempt to explain the introduction of characters, and if it can be in any way controlled no method has yet been pointed out. We all allude to the same general thought when we use the words " vitality " and " constitution " to denote not so much ten- acity of life as vigor of growth, and we all recognize that some individuals and some lines possess this faculty in much higher degree than others. Some individuals never survive the embry- onic stage ; others die in infancy ; still others reach full maturity, and a few persist to an advanced age. As death comes only with the failure of some vital function, the individual may persist long after he is stripped of everything that makes life enjoyable. It is so with races. Some seem endowed with phenomenal vigor, while others are preserved from extinction only with the greatest difficulty and by the narrowest margin, not only because of low fertility but also by reason of inherent lack of vigor. The existence of these internal forces is not a matter of doubt, and their office in directing variation is an interesting and valuable problem which the present state of knowledge is insufficient to solve. Orthogenesis. Closely akin to Cope's conception is Elmer's theory of orthogenesis ^ (straight creation), or, as he calls it, "definitely directed evolution." 1 Pear.son, Grammar of Science, pp. 375-376. 2 Eimer, On Orthogenesis and the Importance of Natural Selection in Species- Formation (pamphlets, 56 pages) [Open Court Publishing Company]. INTERNAL CAUSES OF VARIATION 205 This theory of Eimer's is put forth as a protest and a counter proposition to the theory of Darwin, — afterward very much elaborated and extended by Weismann and others, — which was to the effect that all evolution is the result of heterogeneous growth trimmed down and shaped up by the extinction of indi- viduals possessing unfavorable characters. The natural assump- tion of the extreme selectionists is that utility is the basis of all selection, and that only useful characters will be preserved, the inevitable corollary of which is that all existing cJiaractcrs are iisefiil. Now the necessary consequence of selection is that after a time all existing forms and characters will come to " fit " or agree with the conditions of life, which are the natural agents of selection. This "fit " is so accurate as to deceive many observers and lead them to declare selection to be a fundamental cause of variation. Eimer points out two facts ^ : first, that there can be no selec- tion until a choice is presented, — therefore that the selective process follows and does not precede the origiji of a deviation ; that selection may and does cause the race to vary, but that it has nothing to do with the presentation of the variation in the first individual, — a position in which he is certainly correct. He argues, second, that it is not true that all characters are useful, but that many species endure those that are inconvenient and unfortunate, yet not sufficiently detrimental to be fatal, else the line would become extinct and no such instances would ever be seen. His position is that, first of all, "organisms develop in definite directions wdthout the least regard for utility, through purely physiological causes, and as the result of organic groivtliT ^ Then, after all the characters have developed together, they are passed upon by natural selection in the struggle for existence, this process blotting out only those sufficiently detrimental to unfit the individuals so afiflicted for continuing the struggle in competition with more favored forms. Selection does not remove a handicap, or relieve a race from all undesirable characters. It eliminates only the worst, and down to a level sufficient to establish a kind of "equilibrium of life." 1 Eimer, Organic Evolution, sects, ii and iii. - Eimer, On Orthogenesis, p. 2. 2o6 CAUSES OF VARIATION In this view of the case characters bad and good develop together. The luorst ones are ehminated, but many undesirable or indifferent ones are left behind as not being sufficient to turn the scale against the individual or the race. Thus many undesir- able characters linger in all races, and, what is more to the point, utility is not the cause of either the origin or the persistence of characters, but only of their obliteration tvJien sufficiently detri- mental to destroy the individual and therefore cut off descent in that particular line. The writer shares the opinion that this is the true limit of the selection process under nature, and that the presence of unfavor- able characters argues for their having arisen from causes quite independent of selection. In our yards and fields we control selection according to what- ever standards we may please to establish, but if unfavorable char- acters develop in nature, where selection aims directly at life, will they not be likely to develop also, unnoticed, under our own selec- tion, espscially when we do all within our power to preserve life ?^ The presence or absence of a principle aside from selection, yet responsible for the presence of characters, turns very largely upon the cjuestion as to whether, after all, there are well-established instances of characters independent of utility, and therefore of selection. The presence of such characters is easily shown. For example, what is the utility of the scrotum among mammals } Would it not have been better with sheep, for instance, if the testicles had remained within the abdominal cavity, where they develop, and where they would be safe, instead of descending into an external sack exposed to frequent injury } Undoubtedly it would have been better for individuals, for many have not only lost these organs, but their lives as well, from this unfortunate position ; but the number lost is not sufficient to seriously affect the racer' In other words, selection has aimed at this vulnerable 1 It is noticeable that nature allows reproduction to go on unrestricted, and then slays by the millions. Man cannot afford this wholesale destruction of num- bers. He seeks to prevent undesirable births, — a kind of advance selection that has both its advantages and its disadvantages. 2 This shows that what is bad for the individual is not necessarily bad for the race ; conversely, what is best for the race is often hard upon or even fatal to the individual. This is the very essence of selection. INTERNAL CAUSES OF VARIATION 207 point many times, and in numerous cases with effect, but mam- mals as a race have been able to endure the handicap, else they would long since have become extinct. This shows that some causes other than utility are responsible for the appearance and continuance of racial characters ; that teleology 1 is not a universal principle, and that the function of selection is restrictive, not creative. Other characters not teleological may easily be mentioned : 1. The peculiar minute markings on diatoms and on other inconspicuous organisms. 2. The green color of leaves, due simply to the fact that chlorophyll is green. This is no more of an inherent necessity than that coal should be black or gold yellow. 3. The digital number five which runs generally through vertebrates, which has often been modified and often left intact. Certainly the original number five could not have been teleolog- ical. It is not enough to assume that changed conditions might have rendered an organ detrimental which was once useful. There are too many obviously useless characters. 4. The phosphorescence of pelagic animals,^ and the pearl of the oyster, which is due to injury. Is this beauty useful or is it accidental ? ^ 5. The bright color of deep-sea fishes. Is it any more signifi- cant than the (accidental) color of chlorophyll-bearing leaves ? * 6. The horns of stags, — useful (.'') in battle, but weapons as dangerous to the possessor as to his enemy. The list might be extended indefinitely. Elmer's argument is that characters such as these have been produced not through selection but in spite of it, and through the agency of organic growth in definite directions, which is ortJiogenesis. It would be difficult to be always certain that no basis of utility exists or ever has existed simply because it is not now evident, yet no 1 The doctrine that development is in line with utility and that everything is useful is known as " teleology." 2 Eimer, Organic Evolution, p. xiii. 3 Eimer, Orthogenesis, p. 10. * In certain leaves of bright color the chlorophyll is unable to dominate the stronger shades of other chemical substances, and the leaf is not green but some other color. 2o8 CAUSES OF VARIATION careful student of evolution doubts any longer that there are many misfits in nature. Whether they have arisen, as Eimer asserts, by reason of organic groivth, and whether they are evi- dences of definitely directed deviation^ is quite another matter. There is no doubt of the persistence of a character once started, even in the face of selection, but whether it be necessary to invoke the aid of an internal directive force to explain it is a question upon which more evidence is sorely needed. It is exceedingly important for the breeder to know and recognize all the inherent tendencies with which he must finally reckon, and it may be necessary to go beyond physiological units, correspond- ing to chemical atoms or molecules, and invoke some form of "growth force " corresponding to chemical energy to explain the mysteries of development. Any theory, however, that will even reasonably account for these mysteries must be, in the present state of knowledge, largely an assumption, and let the assump- tion be as simple as possible until we can defend its complexity by a mass of well-established facts. In the opinion of the writer the existence of such a principle as orthogenesis is more than problematical, except as it is an expression of the relations that naturally obtain between physiological units, whatever they may be. SECTION XV — PHYSIOLOGICAL UNITS The "gemmules" of Darwin, the " stirp " of Galton, the "idioplasm" of Nageli, the " biophors," "determinants," and "ids" of Weismann, and "the "physiological units" of other writers are all attempts to explain inheritance of definite quali- ties by assuming that the germ cell which passes over from parent to offspring at the time of procreation is composed of definite units of living matter, each with its specific properties, among which are nutrition and multiplication, which together constitute growth, and — considering the separate properties of the different units of which a given individual is composed — growth in definite directions. In support of this general theory it may be urged that the indi- vidual is zvhat he is very largely because of internal qualities. Corn and wheat grow side by side, drawing their nourishment INTERNAL CAUSES OF VARIATION 209 from the same soil and the same atmosphere. The most nour- ishing food and the most deadly poison are produced side by side under identical external conditions. A man divides his dinner with his dog : one portion simply nourishes the dog and provides energy to watch sheep or perchance to kill them ; the other results in strength to bless the ages, or perhaps in crime to shock the world, — each according not to the nature of the food but to that of the animal that consumes it, and to the support of whose peculiar energies it contributes. This kind of difference in living organisms is traceable to the endowments of a single cell, — the only material that passes over from parent to offspring, — and, regard it as we may, we must see in this single bit of living matter all the potential qualities of tJie race, all the differences between the corn and the wheat, between the man and his dog. They are all there, represented in some material way in the constitution of the germ cell. There is thus a material basis to heredity. We must accept one horn or the other of the dilemma : either conceive this single cell as directly endowed with all the qualities of the race, defining its development, or else endow it with the capacity to develop in this fashion or that according to stimuli. But whence come the stimuli } Certainly not alto- gether from without, or the man and his dog would become alike, if consuming the same kind of food ; and to assume that the influences are internal is only to push the puzzle one step farther away and to assume possibly an immaterial in place of a material basis. The most simple and direct explanation of the phenomena of inheritance and definite development is to consider the germinal matter as consisting of units of some sort endowed with life and the power of growth. This assumption of the physiological unit is not so violent nor so different from other accepted scien- tific assumptions as it at first may seem. In the non-living world we assume the existence of the atom, whatever its ulti- mate constitution, as a minute, indivisible, and indestructible unit of matter. The association of some millions of like atoms makes a measurable quantity of an element like gold, silver, iron, chlorin, or sodium. 2IO CAUSES OF VARIATION Something over eighty distinct kinds of matter are known ; therefore some eighty kinds of atoms are assumed, and this ex- hausts the possibilities so far as unlike like atoms are concerned (unless other atoms are subsequently discovered or created). But this does not exhaust the possibilities of matter, for these atoms combine together, forming new units (called molecules) with distinct properties. Thus NaCl (sodium chlorid) is differ- ent in every way from either the sodium or the chlorin atoms that have united to produce it. In this way these (eighty) various atoms effect all sorts of combinations, many of them exceedingly complex,^ each constituting a new material unit, a sufficiently large number of which constitutes a measurable quantity of a substance whose real composition could rarely be predicted by any of its visible properties. The following table of chemical formulae is presented for two purposes : (i) to show the exceeding complexity of ordinary materials ; (2) to show how certain groups of atoms (as CH2 or C02H)^ behave as units, effecting profound changes in the properties of their compounds. It is evident that the possible combinations ^ even with the few most common atoms, — as C, H, O, N, Fe, Na, K, P, S, — are practically infinite when they are able to organize themselves into larger units, giving rise to complex series like the table on the following page.''^ In this table the radical CO2H runs through the entire series, giving a kind of genetic cjuality to the compounds, while specific differences accompany the varying numbers of C and H atoms present with the radical. It is to be noted, however, that these C and H atoms are in definite proportion to each other, namely, CnHg,,^! ; that is, for every atom of C there will be one more tJian tivice as many atoms of H.-^-CO^^, — all of which is extremely suggestive as early steps in the world of organized matter. 1 The composition of strychnine, CiiHonNoOs, and that of grape sugar, C12H22O11, are both exceedingly simple as compared with many known sub- stances. 2 Such groups of atoms that move together are known as "radicals." They are in every sense units and are capable of replacing or of displacing other atoms in their constructions. 3 We are told by the chemists that more than one hundred thousand separate compounds are now known. INTERNAL CAUSES OF VARIATION 21 1 Monobasic Acids of the Acetic Series, CnHon+iCOaH Acid Formula Formic . . . Acetic . . . Propylic Butyric . . . Valeric . . . Caproic . . Qinanthic . . Caprylic . . Pelargonic Rutic or capric Euodic . . . Laurie . . . Cocinic . Myristic . . Pentadecylic . Palmitic . . Margaric . . Stearic . . . Balenic . . . Butic . . . Nardic . . Behenic Lignoceric Hyeenic Cerotic . Melissic . . red ants, nettles .... vinegar oxidation of oils .... rancid butter valerian root rancid butter oxidation of castor oil . . rancid butter geranium leaves .... rancid butter oil of rue bayberries cocoanut oil nutmeg butter .... Agaricjis integer (a fungus) palm oil tallow- butter beech-wood tar beeswax H CHs C2H5 C3H7 C4H9 CsHu CeHis C7H15 CsHiY C9H19 C10H21 C11H23 C13H07 C14H29 C15H31 C16H33 ClvHsr; C18H37 C19H39 C20H4I C21H43 C23H47 C24H49 C26H53 CogHsg •CO2H •CO2H •co2n • CO2H •COoH ■COoH •C02H •C02H •C02H •C02H •C02H •C02H • C02H •C02H ■C02H •C02H ■C02H ■C02H •C02H •C02H ■C02H •C02H ■C02H •C02H •COoH ■ CO.,H Observe the exceeding complexity of these compounds, and notice that they stand in definite series with uniform differences ; there are no breaks in the series and no missing members until we reach C.^jH^g. Note too that these compounds may them- selves combine with others ; thus stearin, the characteristic fat of tallow, is CgHg (€^8113502)3, in which case 3 H from the acid radical and 3 (HO) from the glycerin radical have united to form 3 (HgO), or water. Above everything else in this series note the wide difference in physical properties arising from a slight difference in chemical 212 CAUSES OF VARIATION constitiition. For example, it is significant that in this series both the sokibihty in water and the acid strength diminish as the proportion of carbon increases. In addition to the properties of non-living units the theory of physiological units as the basis for specific characters in living matter requires one distinctive and additional quality, namely life, with its attendant phenomena, — the power of nutrition and growth. But what are nutrition and growth } Considered in general terms, nutrition is simply the power of one chemical compound (the living) to enter into the composition of another (the food) and break it up and readjust its elements on a basis like its own, leaving the residues to take care of themselves. This readjustment of the non-living food to the composition of the living plant or animal we call nutrition, and it means essentially an increase of the living at the expense of the non- living ; in other words, growth through the numerical increase of living units. There is often readjustment in the non-living world when two compounds are brought together, — the weaker giving way to the stronger affinity, — but there is no such wholesale "carry- ing over" of matter from one to another as in the phenomena we call nutrition and growth. This is a true invasion of the non-living by the living world, transferring matter almost indef- initely, unto itself, not only preserving its own identity in the meantime but impressing it upon the appropriated materials as well. These physiological or vital units are therefore conceived to be the smallest living units, like molecules in non-living matter, except that they are far more complex in constitution and are endowed with the power of self-multiplication through nutrition. This requires growth and division after the manner which has been noted in chromatin granules, except on a scale infinitely more minute. This conception of the action of living units has its similitude in the non-living. Crystallization is a growth, in the sense of increase of size, but it is not attended by transformations equal to those in living matter. Furthermore, crystallization is growth without differentiation, except as to geometric form. Either INTERNAL CAUSES OF VARIATION 213 the physiological units arc capable of a cycle of differentiated energies, or else the race is possessed of many kinds of units, each inactive until its turn, then playing its role in suitable order. What then establishes the order of activity and calls out each unit at the proper time ? Herein lies the mystery, and while the physiological unit seems a biological fact, it after all does not solve the mystery of differentiation ; it only pushes the problem one step farther away. Unsatisfactory though it is to attempt to solve the mystery of inheritance and differentiation by means even of vital units, still nothing else that has yet been proposed comes nearer satisfying the needs of the case, and we cannot fight off the following convictions, namely : 1. That there is a material basis not only of life but of racial characters as well, and this material passes to the individual by means of the germ cell. 2. That the processes of life are essentially chemical. 3. That if the whole truth could be known, the physiological units of vital activity may not be fundamentally different from atoms, molecules, and radicals actuated by chemical afifinity. Broadly speaking, there are suggestive similarities between the chemical behavior of living matter and that of laboratory mate- rial generally, and these similarities are constantly turning up, even where least expected. Whatever the truth may be as to the unit of vital activity, of two things we are sure : first, there is a unit of some kind, — a center of activity ; and second, it is a chemical material pos- sessed of life. Finally, to be useful, these units must be conceived as capable of absorbing nourishment, and of self- multiplication indefinitely. SECTION XVI — GERMINAL SELECTION The difficulty in seeking causes for inheritance and variation is that we are likely to prove too much. For example, if sufifi- cient plasticity is assumed to fully account for the high degree 'of variation that often occurs, then variation is sufificiently pro- vided for, but this view makes a thing like inheritance a matter 214 CAUSES OF VARIATION of extreme improbability.^ On the other hand, if influences are discovered which are really efficient in setting bounds to varia- bility, then they make the transfer of characters from parent to offspring so absolutely certain, regular, and fixed, as to seem to leave little or no possibility of variation. This latter is the case with the hypothesis of physiological units. Weismann recognized its limitations and proposed the theory of germinal selection ^ to account for variation as well as inheritance. This theory assumes that the " biophors," or the physiological units by whatever name they may be called, are engaged in a kind of struggle among themselves within the germ, much as are plants and animals in the larger world outside. Any theory of physiological units must include their absorp- tion of food and their power of self-multiplication. If these activities proceed at a uniform rate for each unit involved, then no variation would result from this multiplication ; but if propor- tions change, or if the vitality varies, then variation would neces- sarily result from these causes alone. Now these activities must be either constant or variable. Weismann assumes that they are variable ; that these units of various relative numbers and strengths are competitors among themselves, one with another, for food ; and that those most energetic in food absorption and capable of the most rapid multiplication will not only be the most vigorous but they will also exist in relatively the largest numbers. They will therefore tend the more to impress their characters upon the later development of the individual. Under this view of the case the "balance of power" is con- stantly shifting, always in favor of the most vigorous and rapidly multiplying units. Believers in physiological units must either follow Weismann in this conception or else assume on the part of the units absolutely equal powers of nutrition and multiplication, for multiplication there must be if such units avail anything in the role of inheritance. 1 All things considered, inheritance and not variation is the mystery. The wonder is, not that individuals vary, but that they follow as closely as they do the type of the race to which they belong. 2 Weismann, On Germinal Selection as a Source of Definite Variation (pam- phlet, second edition) [Open Court Publishing Company]. INTERNAL CAUSES OF VARIATION 215 Weismann reminds us that " By far the largest part of qualitative modifications ... rest on quantitative changes. A determinant," says he, " must be composed of hetero- geneous biophors, and on their numerical proportion reposes, according to our hypothesis, their specific nature. If this pro- portion is altered, so also is the character of the determinant ; " and further: "for fluctuations of nutriment and the struggle for nutriment, with its sequent preference of the strongest, must take place between the various species of biophors as well as between the species of determinants. But changes in the quantitative ratios of the biophors appear to us qualitative changes in the corresponding determinants." ^ And again : " By a selection of the kind referred to tJie germ is progressively modi- fied in a manner corresponding with the prod2iction of a definitely directed progressive variation of the part y^ In this way Weis- mann would "explain" Elmer's orthogenesis; but it is note- worthy that none of the theories yet pivposed ivill account for the original introduction of a new character in the race, whether represented and transmitted by a physiological unit or not. Germinal selection would provide for changes in relative pro- portions of characters, and even for their utter extinction, but not for their introduction, unless, indeed, characters may origi- nate by new combinations of old elements. One is almost forced to the conclusion that in nature loss or modification of characters is far more common than their origin and introduction. It looks as though most of the changes arise in this way, yet it is conceivable that an entirely new quality might arise through a relatively slight modification of the chemical or physiological make-up of the vital units. It is seen and recognized that in the non-living world a slight change in the radical is followed by a profound alteration in physical and chemical properties, and that this sweeping change may be induced by comparatively shght and even external causes. May not the same be true of vital radicals or units, and may not new characters arise more readily than we suppose, all per- haps out of elements fewer, and transformations simpler, than we have hitherto imagined .'' ^ Weissman, On Germinal Selection (second edition), pp. 46-47. ^ Ibid. p. 35, 2i6 CAUSES OF VARIATION Control of internal causes affecting the race as a whole. Whatever causes of this nature may be at work in our fields and yards, — and they are to be reckoned with, — our control over them is secondary and indirect. Their effects, if present, are at once insidious and sweeping. We can be mindful of the effects of genetic selection and the selective death rate, and provide against them, at least to a large degree. If growth force and orthogenesis are also forces, they can be assisted or held in check by selection, but they can never be absolutely controlled ; and if germinal selection is a fact, it is going on entirely independent of any control which present knowledge enables us to exercise except through selection. Summary. All that is involved in heredity is contained in a minute bit of living matter passed from parent to offspring, and whose development will constitute the new individual. The im- pulse to development, therefore, and its fundamental possibilities are forces internal to the germ and to the living organism. It is not difficult to see many causes of variation in the internal processes known to be involved in the activities of living protoplasm. Growth is the result of cell division, which seems to proceed upon plans calculated to insure qualitative as well as quantitative equality as between the daughter cells. Any deviation from the plan, however, — and deviations are known to occur, — must result in variation. This is especially true in the reduction process which is characteristic of matura- tion in both sexes, and which probably lies at the basis of bud variation and of many mutations. Fertilization and sexual union are processes calculated to effect new combinations out of the elements involved, though the possibilities in this direction would be rapidly reduced by close breeding or by any other circumstance which simplifies the ancestry. Doubtless the condition of the germ has some influence, but it is not well understood. The phenomenon of xenia, or double fertilization in certain plants, causes the seed coats to vary the first year in the same direction as the germ. Telegony is a myth, and intra-uterine influences are doubtless limited to those of nutrition, except in cases of disaster. INTERNAL CAUSES OF VARIATION 217 Reversion and atavism are but special instances under the law of ancestral heredity to be discussed later, serving to show that inheritance is partly from ancestors back of the parent. Certain internal factors are of such nature as to affect the race as a whole. Genetic selection is based upon the fact that all individuals are not equally fertile, and that the type tends strongly to assume that of the most prolific. Bathmic influences, such as "growth force" and orthogenetic bias, have been advanced as explanations of those inherent tendencies that appear to characterize most species and that give an underlying trend to their direction in descent. The basis of all vital activities is conceived as being some kind of living unit, comparable with the atom and the molecule in the non-living world, whose activities constitute growth and differentiation, and whose reactions with outside matter and with each other are fruitful causes of variation. ADDITIONAL REFERENCES Bud Variation and its Bearing upon Weismannism. By L. H. Bailey. Science, I, 281-291. Bud Variation, Causes of, and Illustration. By R. M. Kellog. Proceedings of the Michigan Horticultural Society, 1897, pp. 121- 134; also in Experiment Station Record, XI, 424-425. Bud Variation OF Concord Grape. By W. Paddock. Garden and Forest, No. 456, pp. 464-466 ; also in Experiment Station Record, VIII, 290. Color Effects in Crossing Sweet and Flint Corns. Bulletin Illi- nois Experiment Station, No. 21 ; also in Experiment Station Record, XIII, 740. Determinate Variation and Organic Selection. By J. M. Baldwin. Science, VI, 770-773. The Development of the Hybrids between Fundulus Hetero- CLITUS and MeNIDIA NOTATA, WITH ESPECIAL REFERENCE TO THE Behavior of the Maternal and Paternal Chromatin. By W. J. Moenkhaus. American Journal of Anatomy, III, 29. Effect of Fertilization upon Fragrance. Experiment Station Record, VIII, 55. Effect of Spaying upon the Quality of Milk. Experiment Station Record, XIV, 182. Essays on Heredity. By A. Weismann. 2 vols. Evolution in a Determinate Line. By Bashford Dean. Biological Bulletin, VII, 105-112. 2i8 CAUSES OF VARIATION Evolution on Predetermined Lines. By T. D. A. Cockerell. Science, XIII, 311-312. Experimental Study of Variation. By J. C. Ewart. Report of the British Association for the Advancement of Science, LXXI, 666- 680. Experiments in Crossing Corn and Watermelons. By F. C. Card and G. E. Adams. Experiment Station Record, XIII, 740. Experiments upon the Influence of the Sexual Cells upon the Somatic. By G. W. Field. Biological Bulletin, II, 346-347. Heterotypical Division in Maturation. By T. H. Bryce. Report of the British Association for the Advancement of Science, 1901, pp. 685-687. Hybridization of Corn and Watermelons. By F. C. Card. Experi- ment Station Record, XI, 928. Hybridizing Melons: Effect in Sugar Content. Experiment Station Record, XVI, 229. Hybridizing Zebra and Horse. (Experiments of Baron de Parana in Brazil.) Experiment Station Record, XI, 972. Individuality of Chromosomes. By H. Metcalf. Proceedings of the Nebraska Academy of Science, 1901. Inheritance in Parthenogenesis. By Ernest Warren. Proceedings of the Royal Society, LXV, 154-158. Organic Selection. By J. M. Baldwin (Science, IV, 724-725; V, 634-636), E. D. Cope (Science, II, 124), and H. F. Osborn (Science, VI, 583-587). Orthogenetic Variation. By H. F. Gadow, Cambridge. Science, XXII, 637-640. Orthogenetic Variation in Certain Mexican Species of Lizards. By H.F. Gadow. Proceedings of the Royal Society, LXXI I, 109-126. Physiological Selection. By G. J. Romanes. Science, VII, 606-608. Premature Fertilization Injurious to Fruit. Experiment Station Record, XIV, 634. Problem of Development. By E. B. Wilson. Science, XXI, 281-293. Secondary Sexual Characters. (A study of the effects of castration.) By S. G. Shattuck and C. G. Seligmann. Proceedings of the Royal Society, LXXI 1 1, 49-58. Sexual Reproduction ; Distinction from Asexual. By Richard Hertwig (translated by W. C. Curtis). Science, XII, 940-946. Song of Birds Kept from their ov^n Species. By William E. D. Scott. Science, XIX, 154. Sterile Fruit Blossoms. By S. A. Beach. New York Station Bulletin, CLXIX, 331-371- Sterility of Cattle, Causes of. Experiment Station Record, XI, 289. Telegony. By J. C. Ewart. Proceedings of the Royal Society, LXV, 243-251. INTERNAL CAUSES OF VARIATION 219 Telegony (disproved). By J. C. Ewart. Popular Science Monthly, LVII, 126-133. Telegony (disproved): Account of Ewart's Experiment with Zebroids. By J. C. Ewart. Breeders' Gazette, XLI, 1009; also in Transactions of the Highland Agricultural Society, 1901, pp. 81-134 ; and in Experiment Station Record, XI, 1077; XIII, 275 ; and XIV, 76. Telegony (disproved): E.xperiment with Guinea Pigs. By C. S. Minot. Report of the Briti.sh Association for the Advancement of Science, 1906, p. 606. The Chromosomes in Heredity. By W. S. Sutton (1903). Biological Bulletin, IV. The Germ Plasm. By A. Weismann. i vol. The Physical Basis of Heredity. By K. H. Eigenmann. Popular Science Monthly, LXI, 32-44. Theory of Heredity and Telegony. By J. C. Ewart. Nature, LX, 330-333- Xenia, Example in Apple. By L. H. Bailey. Science, IV, 498-499. Xenia, or Double Fertilization. (Review of work of De Vries, Correns, and Webber.) Experiment Station Record, XII, 421, 717; also XIII, 620. Xenia in Maize. By Hugo De Vries. Experiment Station Record, XI, 1016. CHAPTER IX EXTERNAL INFLUENCES AS CAUSES OF VARIATION It was long ago noted, and the most casual observer cannot fail to discover, that individuals of the same species vary greatly according to their environment, — meaning by that term all the external conditions of life, such as climate, food, friends, enemies, and all those outside influences, favorable or unfavorable, among which the individual finds itself born, and with which it must live upon the best terms possible if it would live at all. That these external agencies exert a direct effect upon living matter is beyond question, and it remains to give attention to the nature and extent of this influence as a partial answer to the cjucstion we would solve, — the dependence of organized living matter u{X)n the external world for the nature and range of its activities. Anything we may learn upon this point will be a contribution to the stock of knowledge out of which we shall one day determine all the causes of variation. Without a doubt the great bulk of variability is due to causes internal to the organism, mainly in the form of inherited tendencies. Pearson, after exhaustive statistical investigations, remarks, " The individual contains within itself, owing to a bath- mic law of growth, a variability which is itself quite sensible, being 80 or 90 per cent of the variability of the race." ^ Even then, however, these internal influences are dependent upon outside conditions for their opportunity. A born giant must have food in abundance, but no amount of food would make a giant out of a dwarf. Nor will it avail to awaken, late in life, forces that once might have been active. Some dwarfs are therefore born and others are produced by insufficient food. The external conditions of life affect variability in four dis- tinctly different ways : (i) through natural selection, influencing 1 Pearson, Grammar of Science, p. 473. 220 EXTERNAL INFLUENCES AS CAUSES OF VARLVriON 22 1 the type ; (2) by affording or withholding the opportunity for the proper development of the characters born into the individual and therefore representing internal forces ; (3) by exerting, directly upon the organism, a stimulating or a depressing effect upon its normal activities ; (4) in extreme cases by temporarily or per- manently modifying the character of normal functions. The first manifestly affects the type and the race as a whole, while the second, third, and fourth primarily affect the individual. It is with these latter that we are now concerned. The hasty student credits to external conditions all that tvould happen if these conditions zvere ivithdraxvn. This is erroneous. A very large part of all that happens is due primarily to internal causes, because different races are differently affected by the same conditions. The fundamental cause of variability is therefore to be sought in the form of inherited characters, even though these are dependent upon external conditions for their development, W'hich may themselves seem to be direct causes of variation. It is therefore proper enough to speak of external conditions of life as causes of variability, providing we know what we mean thereby and are careful to distinguish between their indirect effect, on the one hand, in affording or withholding the con- ditions of development in which their influence is secondary, and their direct influence, on the other hand, in stimulating, depressing, or altering the activities of the organism. With these distinctions in mind we may study the effects of outside conditions upon variability without danger of attributing to them what properly belongs to inherited faculties. SECTION I — GENERAL EFFECT OF LOCALITY UPON PLANT AND ANIMAL DEVELOPMENT It is a matter of common knowledge that the texture and quality of garden vegetables depend veiy much upon the con- ditions under which they are grown, and that the highest flavor of the orange, peach, pineapple, and edible fruits generally is found only in specimens from certain favored localities. Thus cantaloupes of extremely high quality developed first at Rockyford, Colorado, and afterward in a few other sections. 22 2 CAUSES OF VARIATION American seed growers generally seem to have settled upon Kansas as the spot most favorable to the development of the highest quality in the watermelon, and it is accordingly the favorite seed-producing locality. Darwin tells us that " the seed of the Persian melon yields near Paris a fruit inferior to the poorest market kinds, but at Bordeaux yields delicious fruit." ^ European varieties of grapes failed so utterly in eastern North America as to necessitate the developing of varieties from the native vine. Indian corn develops local varieties with extreme readiness, but they seldom succeed when transferred even short distances, at least until time enough for acclimatization has elapsed. The writer sent a standard white Illinois corn, ripening in about a hun- dred and twenty days and capable of maximum yields (seventy-five bushels per acre), to be grown in Michigan, Wisconsin, Maine, and Mississippi.^ In Maine it failed to ripen, but at all other points it ripened in about a hundred days, producing small, inferior ears, altogether worthless as a commercial crop. That it should hurry through its period of growth at the north was not surpris- ing, but that it should do the same at the south, where it had even more time at its disposal than at home, is unaccountable. Wheat, on the other hand, is a cosmopolitan crop, and while varieties succeed better in some localities than in others, yet a new variety seldom fails, and sometimes succeeds even better than in the locality whence it came. However, it is altogether likely that no known wheat-growing region equals England in natural advantages for maximum yields. This is supposed to be due to the humid atmosphere and cloudy skies during the later stages of growth, in sharp contrast to the bright skies and hot dry air of America at this season. If this be the true partial explanation of the occasional phenomenal and always high yields in Great Britain,^ we may hope for equal results some day in the similar climate of Oregon. 1 Dan\dn, Animals and Plants, II, 264. 2 Wisconsin, 200 miles north; Michigan, 200 miles north and 100 miles east; Maine, 300 miles north and 900 miles east ; Mississippi, 450 miles south. 3 The average wheat yield of the United States is between 12 and 13 bushels per acre, while that of Great Britain is almost exactly 30, and a maximum of 90 bushels has been reported. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 223 The experiments of Bonnier upon this general subject are classic.^ He divided dandelion, helianthus, and many other plants growing in the valley, and planted one half of each indi- vidual plant on the mountain at an elevation of 2300 to 2400 meters, leaving the other half in the valley (how much below he does not say). As the division was made by a vertical cut through the fleshy root, the two halves must have been practi- cally alike. He found that the portions on the mountain developed plants much smaller than those in the valleys. The difference was mainly in the length of the internodes, not in their number. The leaves of the dandelion were less than one fifth as long when drawn to scale, and the flower stalks were not one tenth as long. The mountain plants in general developed higher colors. Darwin tells us that the medicinal qualities of digitalis are '* easily affected by culture," and that " the wood of the Amer- ican locust tree {Robinia) when grown in England is nearly worthless, as is that of the oak tree when grown at the Cape of Good Hope." ^ The same author quotes Sir J. E. Tennent as saying that "in the Botanic Gardens of Ceylon the apple tree ' sends out numerous runners underground, which continually rise into small stems, and form a growth around the parent tree.' " ^ If this be true, its naturally slight tendency to sprout has in this locality developed into a pronounced habit. Sheep, especially the merinos, are cosmopolitan, and yet they succeed nowhere else as in New Zealand. On the other hand, according to Darwin they seem not to succeed in the West Indies or on the west coast of Africa, where " the wool disap- pears from the whole body except over the loins."'* The writer has seen the same thing in Brazil, except that the best-clothed portion of the body was the back of the neck, the same spot on which the vicuna bears its fur. The statement is frequently made that fat-tailed sheep rapidly lose this character when removed from their native saline pas-- tures, but the assertion needs confirmation, for the writer has ^ C. Bonnier, Recherches expeiimentales sur I'adaptation des plantes au climat alpin. Annales des sciences naturelles, 7*= Serie, Tome XX, 1895. 2 Darwin, Animals and Plants, II, 264. ^ Ibid. p. 266. * Ibid. I, 102. 224 CAUSES OF VARIATION seen a respectable development of fat when the sheep were kept under ordinary conditions. One of the most remarkable and seemingly best authenticated instances of the evil influence of locality upon character devel- opment is the almost uniform failure to maintain the quality of certain English breeds of dogs when bred in India. We are indebted again to Darwin ^ for the remarkable statement that in that country the bulldog rapidly loses his ferocity, and of all dogs the hounds decline most rapidly. Instances might be multiplied indefinitely, but two things must be borne in mind by the student when dealing with this class of facts : first, there is the greatest opportunity for error or exaggeration from inexact observation and report ; second, the plant or animal is exposed to a multitude of new conditions when transplanted to a new locality, only a portion of which are inherent in the conditions of life. Commonly the breeders or attendants are not familiar with the new form and do not afford proper conditions, as in not giving suitable food to animals or in failing to afford sufficient room to large-growing varieties of plants.^ After making due allowance, however, for all these considera- tions, the fact remains that the conditions of life evidently do exert a strongly modifying influence upon development. Locality a comprehensive term. There is little use in attempt- ing to determine the exact influence of each separate locality. The term is an exceedingly comprehensive one, including many things, — climate, by which we mean not only temperature, moisture, and light, but their comparative proportions in that particular spot ; season, by which we mean the succession of climatic conditions ; food (both as to quantity and quality), on which the creature absolutely depends, not only for life but also for growth ; habits of life, radical changes in which may be forced upon the animal by its habitat. ■ This is all too complex for profitable study and discussion. We must separate locality into its elements and determine, if we 1 Darwin, Animals and Plants, I, 39. 2 It will be noted in this connection that most of the instances cited are those of deterioration. EXTERNAL INFLUENCES AS CAUSES OF VARLATION 225 can, the particular modifying influence of each of the conditions of life with which animals and plants are surrounded. In this way we may get important information upon this most difificult and unsatisfactory subject. Accordingly we undertake to ascertain the effects due specifically to food, temperature, light, etc., — the elements that, taken together, constitute the conditions of life. SECTION II— THE INFLUENCE OF FOOD UPON VARIABILITY The best evidence goes to show that food affects develop- ment both quantitatively and qualitatively. It is expedient to consider the two separately. Quantitative effects of food. In general, as every stockman knows, full feed means increased size, provided always there has been no check during development. This is not only the experience in the yards everywhere, but the world over the largest animals are found on the best feeding grounds. Doubt- less other external influences affect size, but certainly no other equals the food supply, and if maximum development is expected food must not be withheld, especially during the early stages of growth. No amount of later feeding, after the individual has accustomed itself to a reduced supply, can make amends for early shortage. This is itself a deviation which easily becomes permanent and follows the individual through life. Development, however, bears no direct ratio to food con- sumed ; that is to say, the greater portion of all food is consumed in supporting the vital processes, altogether without reference to increase of weight or to labor performed. Under the best of feeding we rarely recover 10 per cent of the food consumed in the form of growth or increase of weight, and seldom realize as much as one sixth in the form of labor or other output of the body. The great mass of the food is either not digested at all or goes to support the internal activities of the body, or else is digested and passed out of the body without serving any useful purpose whatever. It is a significant fact that stunted animals (and plants as well) seldom recover from the evil effects of arrested development. 2 26 CAUSES OF VARIATION and that under-development due to insufficient food is quite distinct from dwarfing, especially among animals, in that the body does not develop proportionately. In underfed calves, for example, the head outgrows the rest of the body, the legs are long, and the joints are large. In general, full feed means not only increased size but early maturity as well, which is of even greater consequence. Because of the large proportion of food never recovered in gain, it is manifest that any shortening in the period of development re- sults not only in improved quality but also in the saving of feed ; in other words, early gains are economical gains and they tend to higher quality. Effect upon fertility. The amount and character of food often exert profound physiological influences. For example, the fer- tility of the female honeybee is mainly due to food, the sterile workers and the fertile queens developing from the same eggs. If, even after the worker eggs are hatched and the larvae well developed, they be taken from the worker cells, put into queen cells, and fed "queen's food," they will develop into queens, — a fact often taken advantage of by the bees themselves when by accident all the prospective queens have been lost. Here fertility is largely a matter of food, although an occasional worker is known to produce eggs. This general difference between worker and queen must therefore be regarded as one of development dependent mainly on the food supply. Speaking generally, excessive food supply leads to infertility among both plants and animals. The former vegetate luxuri- ously, but they do not blossom and fruit so abundantly as under a full but moderate supply of plant food. Whether the effect in question is due to overfeeding or to some one or two elements in particular is not well established. Enough is known, however, to justify the assertion that extreme proportions of nitrogen produce luxuriance in stem and leaf at the expense of flower and fruit, but there is exceeding doubt whether this effect would follow a well-balanced food supply with plenty of phosphorus. Excessive feeding of animals, especially females, tends to fatty degeneration of the essential sexual organs, and consequently to EXTERNAL INFLUENCES AS CAUSES OF VARIATION 227 sterility, and this result is hastened if the food contains unusual proportions of carbohydrates, especially sugar. Effect of feed upon variability. It is a common belief that plants and animals are more variable when well fed than other- wise. Doubtless this is true, so far at least as appearances go, for only under such favorable conditions are all the faculties that are born into the exceptional individual able to develop and become visible. When a race is living under mediocre conditions there is a dead level in development. The mediocre individuals have relatively the best chance, and few will rise above the con- dition of mediocrity. Whether full feed is a direct stimulant to variability or only brings potential differences to the surface is therefore an open question. The procedure indicated is, however, in either case the same ; namely, to provide maximum conditions if the breeder expects to realize the utmost from his best individuals or hopes ever to find variations worth preserving. Herein lies the fact that well-bred animals often require more feed than their scrub relatives. It was upon that point that they departed from their kind — not that they contracted to exist on less feed, but that they were able to handle more feed and put it to good use. If the purpose of the breeder were to develop races with a minimum maintenance ration, it could be done; but we keep domestic animals not for their society but for what they can do, — for what they can manufacture out of corn, oats, and hay. We improve crops, not to see upon how poor land they may live, but rather to increase their ability to construct valuable food materials from the mineral elements of the soil and the inorganic constituents of the atmosphere. Not minimum of consumption but economic consumption is therefore the virtue sought. Evil effects of overfeeding. The plant will seldom suffer from abundance of food, although it is not impossible. Excess of nitrogen causes rank growth of stem, but phosphorus is needed for seed. Something akin to a balanced ration is doubtless best for plant as well as animal, yet the former has more of selective ability than the latter. The animal is easily overfed, and if so the injury is likely to be permanent. It results (i) in disorders of the digestive tract; 2 28 CAUSES OF VARIATION (2) in disorders of the excretory organs ; (3) in excessive fat ; (4) in sterility, especially in females, through fatty degeneration of the essential sexual organs. Qualitative effects from the nature of the food. The color of plants and flowers, and even of animals, is said often to be directly influenced by their food, but it is exceedingly difficult always to separate fact from tradition in this particular field, because the average person has an exaggerated conception of the influence of food upon the constitution of the organism, often believing that a meat diet, for example, inclines to ferocity and uncontrollable temper generally. Darwin ^ mentions the practice of feeding hemp seed to bullfinches to darken their color ; the fat of a certain fish to parrots, causing them to " become beautifully variegated with red and yellow feathers." He also states that the shells of mollusks are largely influenced by the amount of lime in the water in which they grow, and Beal succeeded in growing blue hydrangeas from pink stock by occasional aj^plication of alum water to the roots. That the flavor of milk and eggs is largely dependent upon food has long been known, as has the specific effect of certain foods upon the texture and flavor of meat. Mumford found at Illinois that the fat of pork fed upon large amounts of cotton seed proved, on chemical examination, to contain a proportion of true cotton-seed oil, showing that a portion at least of the food had been carried over and stored unchanged in the tissues. That this is the exception, or perhaps more correctly the lesser result, is proved by the fact that in general the identity of the food is lost in the nature of the organism, showing that the organism and not its food dominates results. The same meat becomes dog or man indifferently, indicating that food com- pounds are not as a rule carried over but rather " suffer profound disruption," being reduced, if not to their elements, at least to com- paratively simple compounds, the energy set free in the disrup- tion being " utilized in the subsequent work of construction." ^ The higher organisms are comparatively independent of their food so far as qualitative changes are concerned. They seem ^ Darwin, Animals and Plants, II (second edition), 269-270. 2 Encyclopasdia Britannica, XIX (1885), article " Physiology," 21. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 229 able to extract the materials for their activities often from the most forbidding sources and under the most discouraging cir- cumstances. Indeed, many forms of animal activity are depend- ent for support not so much upon the materials of the food as upon its energy. This applies more especially to mature animals and their functional activity, but we also remember that in the business of body building, materials are required in large excess of the actual amounts retained in the body. Lower forms of life, however, seem often greatly dependent upon their culture medium. Thus many bacteria grow quite dif- ferently upon the potato and in agar or in beef broth, and a germ disease often presents in one species symptoms substantially different from those it presents in another. The animal and the plant (among the higher forms) are nourished upon fundamentally distinct plans. Both require that food be in the soluble form before it can be useful, and each makes use only of such available materials as it may need. But the animal is provided with an elaborate excretory system by which it frees itself of all residues, both of undigested food and of broken-down tissues, and also of digested food in excess of requirements. By this means the animal is promptly freed from all redundant food material. It is not so with the plant. It takes in food by absorption at its roots, and the water carries with it anything and everything that may happen to be dissolved. If nothing poisonous enters, the plant will live, but it will be loaded with residues, because it has no excretory system. The water passes off by evaporation at the leaf surface, leaving at that point large quantities of what- ever was dissolved in the juices of the plant. We are not sur- prised, therefore, that vegetation of the same species differs very widely in composition in different localities, especially with respect to mineral content, depending upon the character of the soil in which it was grown. This difference may be, as in the case just cited, quite independent of vital processes, and due to nothing more than accident. Plants, and the simpler organisms generally, are of necessity far more dependent upon their environment, and especially their food, than are the higher animals, that have to a large extent 230 CAUSES OF VARIATION freed themselves from the bondage due to the accident of birth- place, being able to move about and therefore to establish in the widest sense an independent existence. Notwithstanding all this, and after making allowance for the grosser influences over lower organisms, the fact yet remains that, to a slight extent, and to a slight extent only, the animal is influenced by the character of his food. That this influence is larger upon the products of the body than upon the body itself is certain ; and upon the subtler qualities, like flavor, rather than the more essentially biological characters such as structure. SECTION III — THE EFFECT OF MOISTURE UPON DEVELOPMENT Animals as a rule are quite independent of moisture, providing their direct needs for water are satisfied in the way of body consumption in addition to that needed to reduce the effects of internal heat by evaporation.^ Plants on the other hand do not have the circulatory system of animals, and they depend upon water, taken in by the roots and evaporated by the leaves, to actually carry food to all parts of the structure. Their need for water is therefore far above the amounts necessary for actual composition. For example, it requires the evaporation of something like the equivalent of eight inches of standing water over the entire field to mature an average corn crop. 1 It is often erroneously taught that animals consume carbonaceous foods to sustain the body temperature, while the truth is that all food actually utilized is broken down and its energy set free. This energy is disposed of in three ways : first, to a slight extent in effecting the recombination of the elements of the food into the exceedingly complex protoplasm of the body or its products; second, by the body or some of its parts in the form of internal or of external work ; or, third, it is radiated in the form of heat of low intensity. In this way the body is a factory that is constantly producing heat, which must be disposed of or the structure will be destroyed. There are but two ways of doing this, — by radiation and by evapora- tion. The body is thus constantly producing and is as constantly losing heat. Its actual temperature is certain to be something above that of the surrounding medium, — how much will depend upon the relation between the rapidity of production and the facilities for radiation and evaporation. The body temperature is thus a kind of algebraic sum, and it depends upon a great variety of conditions. Small animals radiate rapidly, large ones less rapidly, for radiation is a surface action, and their surface is less in proportion to the bulk. Hogs radiate slowly, because of their blanket of fat. They do not sweat, and thus are easily overheated. EXTERNAL INFLUENCES AS CAUSES OF VARLVl'ION 23 I Water, or the lack of it, is therefore in many countries and in many seasons the limituig element ; that is to say, the yield is limited not by the available fertility or the ability of the crop but by the moisture present. In excessively wet seasons crops are notoriously " soft," that is, lacking in substance. Just what the difference is has not been well established, but size has been attained at the expense of quality. There is good reason to assume that it is the result of abundance of water, leading to full cellular development, but deficiency of evaporation and transfusion of food due to cloudy skies, resulting in a lack of actual dry matter. Effect of moisture in the atmosphere. It is said that moist atmospheres produce fineness of hair or fur in animals and deli- cate foliage in plants, and that a dry atmosphere inclines to a harsh, dry coat and to spiny growth in plants. Under natural conditions, however, moisture is often associated with coolness and shade, and dryness with great heat and intense light. Certain it is that fur-bearing animals are found in cool climates and that vegetation is delicate in the temperate region but harsh, dry, and spiny in the arid sections. These facts are well known and universally recognized, but how much is clue to moisture alone cannot well be determined in nature. Resorting to direct experiment, however, we find that the same plant may be grown with or without spines according to the degree of moisture in the surrounding atmosphere. Spines are undeveloped leaves, as thorns are abortive stems, and anything that checks growth tends to their production. That this is mainly the result of a dry atmosphere, however, is easily shown in the laboratory. ** Lothelier has made numerous observations in which individ- uals of the same species were placed side by side, some exposed freely to the air and others kept moist under a glass shade." Under conditions such as this " Berberis vulgaris bore non- spinescent leaves in a moist atmosphere, but spines alone in a perfectly dry one. Again, the shoots which in Lychim barbarum, Ulex Eii7-opcBus, etc., would normally have formed thorns, by arrested development and sclerosis (hardening), in a very damp atmosphere continued to grow, and elongated into leafy branches." 232 CAUSES OF VARIATION Microscopical examination showed that in the dry-air specimens " the palisade cells were well developed and there was a special consolidation of fibrous tissues." ^ The same author continues : Again, the common water reed PJiraginites coz/n/nt/iis, when growing in the unirrigated areas of the Nile valley, forms a stunted growth with very short and sharp-pointed leaves. Close to the Nile, however, ... it grows nine or ten feet high, with long leaves almost exactly like the plants in English rivers. All observations go to show that the number of vessels in the fibro-vascular system is greater in the aerial than in the aquatic forms of the same species,^ and the evidence in general seems conclusive that the notorious abundance of spines in tropical vegetation is due primarily to a dry atmosphere, assisted to some extent, no doubt, by the retarding effect of intense light upon growth. A significant fact in this connection, possibly attributable to the scarcity of water, possibly to the lack of heat, is the well-known phenomenon that plant lice, producing females only during the summer, begin with approaching autumn to produce males, and that under the perpetual heat of the greenhouse the insects ob- serve summer habits indefinitely unless the plants on -cuhich they are feeding are allozved to become dry. As is well known, seeds may be kept for long periods if thoroughly dried. In this case the vital activities are reduced to a minimum, but probably not entirely suspended, because seeds will not last indefinitely. Certain lower forms of plant and ani- mal life have a marked power of apparently suspending life through desiccation and resuming its activities again with suffi- cient moisture.-^ The actual influence of water upon development is not yet well understood, except that it is one of the absolute conditions of life, and, being a fluctuating element, often limits development. The student should fully appreciate the bearing of all this upon the matter in hand : The degree of development of an individual at maturity is not a complete index to his inherited characters. 1 Vernon, Variation in Animals and Plants, p. 264; see also Henslow, Origin of Plant Structures, p. 40. ^ Vemon, Variation in Animals and Plants, p. 265. 3 C. B. Davenport, Experimental Morphology, Part I, pp. 59-65. EXTERNAL INFLUENCES AS CAUSES OF VARIATION ^33 It is both something /ess and something more. It is /ess by so much as the individual has failed to develop because of unfavor- able external conditions ; it is more by whatever development is due to the c/ireet influence of external conditions. This is the principal reason why breeders have difficulty in knowing how much to credit to inheritance, and it is on this account that more knowledge is needed of the nature and extent of the development due directly to external causes, and therefore independent of inheritance. After all, however, it will be found that the total development from all causes is well within hereditary limits except, perhaps, in rare cases where the normal functions may have been altered by unusual conditions. SECTION IV — EFFECT OF CONTACT UPON PROTOPLASMIC ACTIVITY 1 Unstable chemical compounds are exceedingly sensitive to mechanical contact either from solid bodies or from liquids or gases in rapid motion. Davenport says:^ Mechanical disturbance can induce in certain lifeless compounds violent chemical changes. Compounds which are so affected are preeminently unstable. This instability, however, varies greatly in degree. In some cases the blow of a hammer is required to upset the molecules, the result being often a violent explosion. In other cases (e.g. chloride or iodide of nitrogen^) the slightest touch of a feather suffices to produce an explosion. Now, most of the substances which explode upon impact [altering their chemical arrangement and properties] . . . are organic compounds, — fulminate, nitroglycerin, gun cotton, and picric-acid derivatives, — and therefore it is not surprising that we find the notoriously unstable proto- plasm violently affected by contact. " Of special interest in this connection " is the fact that '^ periodie disturbances produce very important molecular changes in [certain] chemical compounds. Certain substances have a 1 C. B. Davenport, Experimental Morphology, Part I, pp. 97-110, and Part II, pp. 370-388, from which most of the data on this subject are taken. N. B. Con- tact agents are technically known as molar agents. 2 Ibid. Part I, pp. 97-9S. ^ When chemical terms of this kind are used outside of quotations the new form will be used, as nitrogen chlorid, omitting the c. 2 34 CAUSES OF VARIATION specific rate of vibration, so that when this is reproduced by a vibrating cord or plate, explosion of the substance may occur, lodid of nitrogen is one of these substances which is exploded by a high note." ^ Living protoplasm is no exception to the general rule that specially unstable compounds are sensitive to contact. Effect of contact upon the metabolism of protoplasm.^ It is a well-known fact that phosphorescence is increased by mechanical irritation. So true is this that the water thrown from the pro- peller wheels of a steamer in tropical regions looks like liquid fire, and a brisk breeze moving over the .surface suggests at a distance the white foam of the surf. Contact also exerts a strongly stimulating influence upon secre- tions, not only with lower organisms which seek attachments, and the glands of insectivorous plants, but with higher animals as well.^ Effect of contact upon movement. The first effect of mechanical disturbance in protoplasm is to check all movement. Minute or- ganisms, tradescantia hairs, etc., cease their protoplasmic motion by the irritation of mounting under the cover glass. In higher plants a sudden jar causes cessation of movement and often a retreat of the protoplasm to one side of the cell so characteristic as to be spoken of as "fright." If an amoeba with pseudopodia out is touched or irritated, it immediately assumes the spherical form, and in general the effect of contact is to cause protoplasm to cease motion and assume approximately the spherical form, or, in other words, to occupy the least space possible. But this is to all intents and purposes contraction, and the general principle may be laid down that ex- ternal contact causes contraction, especially noticeable in muscle, which is par excellence the cojitractile tissue. This contraction most commonly, and of necessity, operates at first to draw the organism azvay from the irritating body (negative thigmotaxis), but if the body be long or large, so that locomotion continues, then the side next the foreign body will 1 This suggests, as the author observes, the phenomena of hearing. 2 C. B. Davenport, Experimental Morphology, Part I, pp. 98-99. 3 As is well known, the heifer that has never produced a calf may be made to give milk merely by persistent manipulation of the udder. EXTERNAL INFLUENCES AS CAUSES OF VARLA.TION 235 be shortened and the organism will move in a curve that will speedily bring it into actual contact. It is noticeable, too, that real contact, being once established, is broken with difficulty. Many lower animals, as in aquaria, coming in contact with a plain surface, move along that surface until they reach a point where side and bottom or where two sides join, and where they can place their bodies in contact with two surfaces. They are likely now to move along the groove formed by the two surfaces until a corner is reached where contact on three sides is possible. Here, if anywhere, the organism will come to rest. It is only that it is " more comfortable " ; that it moved under the molar impulse until it reached a point where further movement and more complete contact were alike impossible. Even higher animals come to the highest state of rest when in contact with foreign bodies on as many sides as possible. Effect of contact upon direction of movement, — thigmotaxis, or stereotropism.^ It is a well-known fact that roots growing in running water grow upstream, not downstream, and that many fish at the breeding season are possessed of an irresistible im- pulse to move against the current (rheotaxis).^ They therefore ascend the strongest currents, leap waterfalls, and surmount every possible obstacle in upstream movements, — a passion which ultimately carries them to their breeding grounds in shal- low water. It is at these times that salmon pile themselves up even above the water level and that they will follow any decoy that leads against the current, even into hopeless traps. Thus may external agents exert a strongly modifying influ- ence upon such essential activities of living matter as the contractility of protoplasm, resulting in definitely directed move- ments through their control of muscular contraction. As we shall see, contact is not the only influence capable of stimulat- ing contractility of protoplasm and controlling the direction of movement ; on the other hand, muscular tissue is exposed to the exciting influence of a great variety of circumstances both internal and external to the organism, any one of which will induce the characteristic reaction of this sort of tissue, which is contraction. ^ C. B. Davenport, Experimental Morphology, Part I, p. 105. ^ Ibid. p. 109. 236 CAUSES OF VARIATION SECTION V — EFFECT OF GRAVITY UPON LIVING MAlTERi; GEOTROPISM^ A germinating seed sends out two sprouts. From whatever position they emerge, one grows downward in response to grav- ity, the other upward in opposition. In other words, the root is positively and the stem is negatively geotropic ; that is to say, each contains within itself some quality that puts it into definite relation with the center of the earth, but in opposite directions. That this tendency is something consider- able is shown by the fact that it is capable of e.xertion against pronounced resistance, as in burrowing through the soil or persisting against mechanical obstructions. This definite relation to gravity seems to ex- ert itself in the manner of inward forces respond- ing to outside conditions ; for if a piece of stem be altered in its position, future growth readjusts itself as promptly as possible in 1 C. B. Davenport, Experimental Morphology, Part I, pp. 1 12-124 ; 'ilso P^i't II) PP- 391-402, from which mostof the facts here cited are taken. 2 Two series of terms are in use, of substantially the same meaning: one (see "geotropism" and "geotropic"), with endings derived from the Greek meaning to turn ; the other (see "geotaxis" and "geotactic"),with Greek endings signifying to arrange. We thus have also "chemotropism," with its adjective " chemotropic," over against " chemo- taxis," "chemotactic," " heliotropism," and " heliotactic," etc. The latter endings were supplied when the effect of chemicals, gravity, and light upon free-tti.ovmg organisms, as bacteria, swarm spores, etc., was first noted, which effects led to definite " arrangements " ; but when similar effects were noted upon larger organisms not free to move, like the fixed plants, but which manifested the effect by turning, the term " tropism " came into use, signifying a turning. This term and its derivatives seem to express more accu- rately what really happens in most cases, and they are coming to be preferred in all generalized considerations. Hence in the text the term " geotropism " is preferred to " geotaxis " and similarly in most places for all correspond- ing terms. Fig. 25. Influence of geotropism : be- havior of a grow- ing plumule in righting itself after being placed in a horizontal posi- tion, as at I. — After C.B. Daven- port, from Stras- burger EXTERNAL INFLUENCES AS CAUSES OF VARIATION 237 accordance with this inward polarity and the new conditions (see Figs. 25 and 26). Vochting's experiments as here figured, and as described by Morgan,^ indicate that the formation of stem or of root is a E Fig. 26. Influence of polarity and of gravity upon the character and direction of growth. — After Morgan, from Vochting A, piece of willow (cut off in July) suspended in moist atmospher?, with apex upward ; B, older piece of willow (cut off in March) suspended in moist atmosphere, with apex downward ; C, piece of willow with a ring removed from the middle, apex upward ; D, piece of root of Popiilus dilafafa, with basal end upward ; shoots from basal callus ; E, piece of root of same with two rings removed ; new shoots develop from basal callus and from basal end of each ring question of both internal causes (polarity) and external influ- ences (gravity), that is to say, the character of growth is a func- tion both of internal and external conditions. 1 Morgan, Regeneration, pp. 72-83. 238 CAUSES OF VARIATION A stem of willow severed from its parent plant and suspended, apex upward, in a moist atmosphere, will of course send out shoots from its apex and roots from its base. If a ring of bark be removed from the middle of the stem, then sprouts will issue from the apical extremities of the sections and roots from the basal end. Neither of these experiments determines whether gravity or polarity is chiefly instrumental in the production of stem and root, but if the piece be inverted and suspended apex downward (in a moist atmosphere), we shall get some light on the two forces, external and internal. Under these conditions the apical end, now downward, will yet produce stems, but they will change their direction with ref- erence to the axis and point tipivard, while the basal end will produce roots, but they will extend downward. In this case each end has produced its characteristic growth, and each has responded to gravity in the usual way (see Fig. 26), except that, if the piece be of the older wood, roots will appear throughout the entire length. The force that fixes the character of growth appears to be internal ; that which fixes its direction appears to be mainly external. If a begonia leaf be planted in the ground or suspended in moist air, whatever its position roots will start from the basal end of the stem at its point of severance, and afterward shoots will arise just above the point of origin of the roots, the body of the leaf withering away.^ (By "above" is meant between the origin of the root and the apex of the leaf, whatever its position.) By this we see that the stem has a distinct polarity, producing sprout and root, each at the proper end ; that the leaf has no true polarity, producing primarily only roots ; but that wherever and however produced, a distinct geotropism characterizes both stem and root. There is little geotropism in leaves, or in the horizontal stems of many plants running along or just below the surface of the ground. However, the stems and roots produced at the nodes of these underground stems are both geotropic. Geotropism in animals. Geotropism is much less marked with animals than with plants. We may say that only fixed organisms 1 Morgan, Regeneration, pp. 74-76. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 239 would be likely to develop decided geotropism ; or, conversely, we may say that organisms with marked geotropism would be likely to become fixed. In either event less geotropism would be expected among animals, and either assumption would square with the facts. Many lower organisms truly animal are distinctly geotropic, however, and most animals show a decided preference as to position with reference to gravity. Both with land animals, high or low, and with fish as well, the ventral side is carried downward, and the anterior portion in general upward. Effect of geotropism upon protoplasm. The protoplasm of cells, plant or animal, is not homogeneous. The nucleus is heavier than the cytoplasm, and together with chlorophyll granules and starch grains tends to settle to the lower side of the cell, giving it a kind of polarity due to gravity. " In many ova the yolk sinks to the lower pole and the cytoplasm floats on top, in whatever position the egg may be held," — a fact which " undoubtedly has an important effect upon development." ^ General effect of gravity upon development. There is no room for doubt as to the profound effect of gravity upon development. However, this influence of gravity has been continual and con- stant on all existing species for untold generations, and it may be looked upon as having already exerted the maximum of its influence upon all forms of life. The effect of gravity upon development has, therefore, long ago reached the position of a constant force to be reckoned with, and is now to be regarded as a fixed factor in development rather than as a present cause of individual deviation, — to be studied more for the sake of learning the degree of dependence of liv- ing matter upon outside forces rather than as a direct means of further change. SECTION VI — EFFECT OF LIGHT UPON LIVING MATTER In all chlorophyllaceous plants the amount of carbon fixed, and therefore the total of growth in the sense of increase in dry matter, is in almost direct proportion to the expanse of leaf sur- face and the amount of light that falls upon it. 1 C. B. Davenport, Experimental Morphology, Part I, p. 114. 240 CAUSES OF VARIATION Light has other influences, however, than those exerted through the fixation of carbon. For example, strong sunUght tends to check growth in the sense of increase in bulk, and when these two effects of light are combined, as they are in the tropics, they give us naturally the slow-growing, generally small, and extremely dense wood of the lower latitudes. Briefly, light, like gravity, exerts specific effect upon matter. Many of the effects of gravity (positive geotropism) may be regarded as arising from the elementary properties of matter, for naturally all matter is attracted by, and approaches as nearly as possible to, the surface of the earth ; that is, matter in general may be said to be positively geotropic. Sensitiveness to light, however, should be regarded as due to the special compounds that constitute living matter, rather than as a property of mat- ter in general, for matter in general is indifferent to light. Light exerts influence upon living matter, especially plants, in three distinctly different ways : (i) through its heat rays, affect- ing temperatures ; (2) through the so-called chemical (actinic) rays, causing definite chemical reactions in the protoplasm ; (3) through the luminous rays, influencing especially the direction of growth in those parts that are so fixed as to be incapable of free movement. Certain of these influences are worthy of somewhat extended consideration. Chemical effects of light. ^ While matter in general in its simpler compounds is quite indifferent to light, yet certain com- pounds are notoriously dependent upon its influence ; that is to say, many combinations are effected more readily and others only in the presence of light (photosynthesis). Oxidation of vegetable oils is much more rapid in daylight than in darkness. Hydrogen and chlorin unite explosively in tJic presence of light. Chlorin passed through alcohol in strong sunlight unites with it, forming chloral hydrate, and chlorin compounds generally are sensitive to light. The whole field of photography is dependent upon the action of light upon the halogen salts of silver, gold, platinum, and other metals, due to the so-called "chemical rays" extending ^ C. B. Davenport, E.xperimental Morphology, Part I, pp. i6i- 165, from which most of the instances under this heading are taken. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 24 1 from the blue upward and most pronounced in the invisible "ultraviolet" portions of the spectrum. These particular wave lengths seem closely akin to chemical energy, and their effect, invisible and subtle as it is, should not be overlooked. Certain organic compounds are readily formed only by the aid of light ; thus the reaction Cj4Hg02+CgH5CHO=Ci4H8 (O.H)(O.CO.CgHj;) takes place in sunlight, but in darkness the substances are indifferent to each other. ^ It is under this same principle that the vegetable substance chlorophyll is able to break up the CO.2 of the atmosphere and fix the carbon in the form of starch, setting free the oxygen. This is the most dis- tinctive act of plant life, and yet it takes place only in the pres- ence of light. The student is therefore prepared to realize that light is one of the controlling forces, not only in effecting chemical compounds in the non-living world but in the activi- ties of living matter as well ; that in many respects its action is fundamental (as in the fixing of carbon), in others incidental, and in still others even accidental (as in the color of chloro- phyll or of gold or silver). In any event it is an influence to be taken account of when one is engaged in the study of the circumstances that control the activities of living matter. Effect of light upon functional activity.'^ The effects of sun- light upon growth are of three kinds, — one due to the heat rays of the lower spectrum, the others to the luminous and the so-called chemical or "actinic" rays of the upper spectrum, from the blue to a considerable distance beyond the violet. Strange as it may seem, the influence of light upon the fixation of car- bon is greatest in the thermic rather than in the actinic region of the spectrum. Timiriazeff^ kept a plant in the darkness until the starch in the leaves had been absorbed. " Then in a dark room a prismatic spectrum was thrown upon the leaf and the position of Fraunhofer's lines indicated on the leaf. After three to six hours starch had formed under the influence of the light, only in the region of the absorption baruis of cJiloropJiyll lying between B and D^' as determined by treating first with 1 C. B. Davenport, Experimental Morphology, Part I, p. 163. 2 Ibid. Part I, pp. 166-180 ; Part II, pp. 416-436. 3 Ibid. Part I, pp. 169-170. The italics are mine. 242 CAUSKS C)K \ARIAT10N alcohol to decolorize, and then with iodin, which forms its char- acteristic blue with starch. The sjxxnrum between /> and D^ includes the upper part of the red, the orange, and the lower parts of the yellow. — the thermic rather than the actinic portion of the spectrum. Among both plants and animals light has an important influ- ence upon color. The chlorophyll of plants is formed only in its presence, and it is intimately concerned in the production of pigments in the skin. Not only that, but the arrangement and position of pigmentary matter, whether lying next the surface and well diffused — thus giving color to the animal — or lying collected in masses deeper in the skin and having little effect upon the color, are due largely to the direct effect of light falling upon the skin of the animal. In this way certain animals, as the chameleon, are capable of e.xhibitinga considerable range of colors, giving rise to the fiction that they are able to imitate any color near which they may be situated."^ It has been customary to cite the fact that cave animals are frequently less highly colored than their congeners of the land, as evidence that color is fundamentally dependent upon light. This cannot be tRie except in a very general sense. All material substances have some relation to light and therefore have some color. What the color of a body may be is therefore dependent primarily upon its composition, and in this sense its color may be said to be accidental, — a remark that is as true of chlorophyll as it is of gold or silver, or of red, white, or yellow brick. But when the particular compound happens to be one like chlorophyll, or a pigment that can be fanned only in the presence of light, then and then only can color be said to depend upon the presence of light. Deep-sea fishes are often highly colored ; rocks hidden in the earth have their characteristic tint ; the blood of vertebrates is reil. not from the presence of light but from the presence of compounds of iron. In all these cases the color arises from a substance in no sense dependent upon light for its formation and existence, and the case is distinct ^ The vibrations at this point are approximately 5-5 x 10^* per second (525»- 000,000.000,000). - C. B. Davenport, Experimental Morphology, Part I, pp. 192-194. ICXTIoKNAL 1NFLU1<:NC"I;S as CAUSl'lS Ol- VARIATION 243 fi'oni one ill which the color is due to ;i subsliince lornied only by the aid of li,i;hl. It is only in this latter case that li^ht- 'ii'iy truly be said to be a direct external cause of variation. Examples of this are found in the coloring of fruit, either under normal conditions or in " fruit photography," a process by which pictures may be made to api)ear on iiighly colored fruit by shading with a screen derived like a negative from the picture to be transferred.^ The sun is supi)osed to exert a direct effect upon the skin, ranging from the tan of the white man to the dark color of the tropical races. This seems an ill adaptation, and so it is as regards the heat, for black objects are warmer than white ones ; but the adaptation is not to the heat rays but to the chemical, for black pigment is almost totally non-actinic. Hence we may say that dark-skinned people have lost something in heat resist- ance, but they have gained what is of more conse(|uence, — a screen against the actinic rays.^ Light exhibits its most characteristic effec^t uiv)n the eye of higher animals. It here gives rise to two remarkable actions, — muscular contraction of the iris, by which the amount of light admitted is regulated, and a nerve stimulus, which forms a defi- nite image on the retina, as upon a mirror, and which is perfectly compreiiended by the mind. Whether the colors and images seen by all eyes are absolutely identical is obviously a matter that can never be determined. It is of course safe to assume that the images, in so far as fonn is concerned, are identical, because the outlines are due to the mechanical laws of refrac- tion, but the colors as comprehouied may be due in part, if not entirely, to physiological peculiarities. That is, the color which to one is red may look to him as yellow does to another, — a supi)osition entirely ])lausible when we remember that with some individuals sound always suggests color as well, so that the name Jones immediately suggests black, or red, or some other color, differing with different individuals. What relation or coordination between the auditory and the optic nerves can be responsible for this sort of mixed impression we do not know. 1 Lilcrary Digest, Septeml)er 16, 1905, \). 381. - Ibid. October 7, 1905, p. 485. 244 CAUSES OF VARIATION Vital limits. Chlorophyllaceous plants are absolutely depend- ent upon light for their very existence, but parasitic plants, and animals in general, are not dependent upon light in any vital sense, because they, like animals, subsist upon highly organized materials in which the fixation of carbon has been already accomplished by other organisms.^ All known facts indicate that animal life in general is essen- tially successful in total darkness. Mules have been kept in mines for twenty years, and beyond temporary sensitiveness of the eyes no effect was perceptible. Prisoners have spent their lives in dungeons. All embryonic development in mammals takes place in the total darkness of the mother's body. It is doubtless not too much to say that light has no effect whatever upon the vital functions of the higher animals ; that it is as unessential in this respect to animals as it is indispensable to plants. Bacteria, as a rule, not only do not need the light, and flourish best in darkness, but strong sunlight is almost uniformly and quickly fatal ; indeed, direct sunlight is recognized as one of the most successful germicides (which is of itself the principal reason why plenty of light should be provided wherever domestic animals are kept). This fact is illustrated by inoculating a gela- tin plate uniformly with bacteria, as Bacillus antliracis, covering the plate with a piece of black paper out of which some pattern is cut, as the letter E, and exposing it all to strong sunlight for a few hours. If the plate is then put into an incubator the bac- teria will grow, except over the area exposed to the light, in which area they have been killed.^ From the well-known fact that bacteria i?i a vacuum are not affected by light we conclude that death is due to oxidation in the presence of light, a phe- nomenon common in organized compounds. Light rigor. The movements of protoplasm in general are retarded, or even stopped, in the presence of intense light, so that rigor precedes death. There is therefore (for plants) an optimum, minimum, and maximum intensity, and between 1 This is written in general terjus, and regardless of the fact that certain lower organisms, whose nearest relatives are distinctly animal, themselves bear chlorophyll. 2 C. B. Davenport, Experimental Morphology, Part I, pp. 171-172. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 245 these limits protoplasm is stimulated by sudden changes, rapidly becoming accustomed, however, to alterations within a narrow range (phototonus), and soon resuming its normal activity, except that as the intensity approaches the point of rigor activ- ity appears to be permanently checked. Retarding effect of light upon the rate of growth. To the higher organisms generally, sunlight, however intense, is not fatal, but it not infrequently retards the rate of growth, espe- cially among plants. This accounts for the relatively flower growth of tropical vegetation as compared with that of higher latitudes, and for the fact that growth in the sense of increase in bulk is more rapid at night than in the daytime. Sachs found 1 that the curve of growth reached its greatest height at daylight, then commenced to decline, reaching its minimum a little before sunset. Davenport points out that this fluctuation is opposed to the effects of temperature, which is more favor- able in the day than at night, so that the final results are some- what less than the total influence due to light. This fact is well illustrated in experiments upon seedlings, grown both in darkness and in light, — investigations that are entirely feasible, because at this stage the young plant depends upon the old seed for its nourishment. It is invariably found that seedlings grown in the darkness have grown at the faster rate. That different rays have different effects upon the growth of plants is easily shown. Flammarion cultivated sensitive plants for three and a half months (July 4 to October 22) in red, green, white, and blue light. At the close of the experiment the plants had attained heights as follows : in the red light, 420 mm. ; in the green, 152 mm. ; in the white, lOO mm. ; and in the blue, 27 mm., with general appearances shown in Fig. 27. In this experiment the greatest heat rays were of course trans- mitted with the red light, but the general temperature was regu- lated by currents of air passing through the various chambers.^ It has been roughly stated that light has no effect upon germination. In general this is true, though careful experiments indicate that most seeds germinate slightly earlier in darkness ^ C. B. Davenport, Experimental Morphology, Part II, p. 421. 2 Ibid. pp. 427-429. 246 CAUSES OF VARIATION than in light, and a few, prominent among which are the poas, germinate more readily in the light, as do also the spores of ferns and the seeds of the mistletoe.^ Evidence as to whether sunlight influences the growth of ani- mals is inconclusive. Experiments have been conducted with tadpoles, snails, the eggs of certain fish, and with the young of higher animals, but while slightly better growth is reported Red Cireen White Blue Fig. 27. Effect of light upon rate of growth : sensitive plants grown for three and a half months in red, green, white, and blue light. — After C. B. Daven- port, from Flammarion during daylight, it is not certain that other conditions did not contribute to the results.^ Experiments in feeding fattening ani- mals in darkness and in light fail to establish special differences. All considerations point to the conclusion that light exerts a strongly modifying influence upon living matter in general, but that the higher animals are substantially free from its direct effect except to some extent in the matter of color. 1 C. B. Davenport, Experimental Morphology, Part II, pp. 419 and 424. 2 Ibid. pp. 425-426. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 247 Influence of light upon the direction of locomotion or of growth ; heliotropism. Irritability to light is one of the properties of protoplasm. This reaction is generally in the form of contrac- tion, as with muscle fiber ^ resulting in a shortening of the side next the source of light. With free-moving plants and animals this gives direction to locomotion, and they gradually swing about until both sides are equally lighted, when, if motion continues, the creatures will of necessity progress toward the light. This is positive heliotropism. Quick-moving forms are often carried into the source of light by their very impetus, before the repellent effect of great heat has time to act. In this way moths fly into the candle and are killed, while slower- moving forms are checked by the heat in time to save them- selves. In this latter case the future movements will be a resultant of the positive heliotropism and negative thermotrop- ism, by which the insects are held at a certain radius circling about the source of both light and heat, as if not able to leave it, as indeed they are not. Negatively heliotropic forms, like earthworms, are those whose protoplasm does not contract in the presence of light, but, on the contrary, expands. These are carried away from the light, and if, in their wanderings, a lighted area is approached, they are unable to enter it. When the organism is not free to move, as in the case of stems of higher plants, the effect will be manifested in the direc- tion of groivtJi^ which is all the response to heliotropism pos- sible under the circumstances. In cases of this kind the stem will bend toward, or away from, the source of light, according as the plant is negatively or positively heliotropic, until all sides are equally lighted, in which position it will remain during growth, as do other forms during locomotion. This placing of the body (or stem) with reference to the various " tropisms " is technically known as "orientation," and 1 As most of the examples to follow are confined to the lower animals, and to plants in which protoplasm is comparatively exposed, it is well to remind the student that the higher animals are not destitute of the same properties, and that they have one exposed region peculiarly sensitive to light, namely the iris of the eye, whose muscles contract promptly under its influence. 248 CAUSES OF VARIATION this particular orientation with reference to the rays of light, whether parallel to their direction or transverse, has received the special name of " phototaxis." ^ Effective rays. All experiments indicate that the blue rays are the effective ones in producing heliotropic effects.^ Some organisms seem sensitive to the yellow, but both red and ultra- violet are alike inoperative. Heliotropism is therefore an effect decidedly due to the luminous rays. Contributary conditions.'^ Heliotropism is dependent upon a variety of conditions both external and internal to the organism. 1 There is great uncertainty as to terms. The one just quoted ("phototaxis") is in its root a protest against the old term "heHotropism " in that it recognizes //i,'/// as the active agent rather than the siin {/lelios), which is a source not only of light but of heat and chemical energy as well; and it recognizes, too, that these influences are exerted as light, quite independent of the sun or any other particu- lar source. Of late there has been a strong disposition to substitute this root for the older term, giving us phototropisni in place of " heliotropism." Those disposed to this view would go farther and discriminate between those movements that appear to occur with reference to the direction of the rays, which is " phototaxis," and those which are made with reference to the intensity of illttnii nation, which is " photop- athy." These students also recognize the fact that irritability and the consequent movements, whether positive or negative, depend very much upon the intensity, so that organisms that are negatively heliotropic at high intensity are positively heliotropic at low intensity, suggesting a middle point at which the organism will be held, as with deep-sea animals that come near the surface at night but sink to considerable depths in the bright light of day, the water acting as a screen. This neutral point or satisfied condition is described as "phototonus." It may be necessary to recognize all these distinctions when the subject is better understood, but it is more than likely that these different behaviors are only different manifestations of the same natural irritability to light on the part of protoplasm in general, and that the so-called " phototaxis " or even "phototonus" is only the condition of the organism after it has brought itself into such position that one irritability is balanced by another (which is always easy with bilateral symmetry), or that it is a kind of acclimatization acquired by the protoplasm to the particular intensity at hand. The theoretical objection to direct reference to the sun is certainly sound. The attraction (or repulsion) is with reference to light as light, and not to the sun as its source. But the sun is preeminently the source of light, and for this reason, and with a proper understanding of the facts, the author prefers for our purposes to adhere to the single old term, at least under the present state of knowledge, for if properly understood it seems substantially correct and serves all practical purposes fairly well. 2 C. B. Davenport, Experimental Morphology, Part I, pp. 201-203; Loeb, Studies in General Physiology, pp. 29-31. 3 Ibid. pp. 196-201. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 249 It is first of all dependent upon intensity of light, feeble illumi- nation being ineffective, even if the characteristic rays be pres- ent. Not only that, but some organisms are positively helio- tropic in moderate light and negatively heliotropic in strong light. In this way many fishes are held in a nearly constant illumination, rising or falling in the water according to the intensity of sunlight, and coming completely to the surface at nightfall. Another element in heliotropism is temperature, its influence being most active at the maximum or normal, and lessening or disappearing as it falls. Still another controlling influence is the condition of the animal. Many caterpillars are positively heliotropic only when unfed. They are thus led to ascend trees when hungry and to descend when filled.^ Still again, certain animals are heliotropic only under peculiar circumstances ; for example, Loeb found that winged ants exhibited no reaction to light except at the time of their nuptial flight, when they were decidedly heliotropic.- Influence of light upon the direction of growth.'^ Instances of this influence upon the stems of plants are almost too common to need mention. The leaning of plants toward the window, or of trees over a stream, can be seen almost any day. It is noticeable that most leaves appear to be destitute of heliotropism, and yet it is not impossible to detect traces of its influence. On careful observation it will be noted that some leaves tend to present their upper surface at right angles to light rays, while others tend to present the edge. It is a general principle of orientation that organisms with radial symmetry, like most plants, present as nearly as possible the end of the long axis to the source of light, with the lateral parts equally lighted ; while organisms of bilateral symmetry, like most animals, tend to present the dorsal surface at right angles to the light, with the oral (head) end nearest its source, the right and the left halves equally lighted, and the ventral surface shaded. 1 Loeb, Physiology of the Brain, p. 189. 2 Loeb, Studies in General Physiology, pp. 52-53. 3 C. B. Davenport, Experimental Morphology, Part II, pp. 437-445. 250 CAUSES OP^ VARIATION In the very highest animals little but the eye is sensitive to light, most parts being protected by a heavy epidermis or other covering, so that heliotropism in its strictest sense is in them limited to the visual parts. In lower animals, however, with bodies less protected, light exerts a controlling influence upon movements. Influence of light upon the direction of locomotion.^ It has already been explained that some animals exhibit heliotropism only at certain periods of their lives, or only in certain condi- tions, as when hungry. Others, however, are constantly and uniformly sensitive. The common house fly is positively helio- tropic, while the larvae of the same, hatched in the dark, soon become strongly negative, and so continue while in the larval condition.^ This difference between the larval and adult stage is common, and led Loeb at first to suppose it to be a general principle, — a conclusion invalidated by the fact that caterpillars and their imagoes behave alike. ^ Both moths and butterflies are positively heliotropic, but moths are "attuned" to a lower intensity. This, with their more rapid flight, is responsible for their wholesale destruction by the naked flame, which the slower-flying butterflies avoid as a source of heat. The tendency of many small animals to creep into crevices as if to hide must not be understood as evidence of negative helio- tropism, much less as evidence of timidity. It is often due simply to contact irritability (stereotropism), for it is a well-established fact that living matter is sensitive to contact with other solid substances. This is the principle discussed in Section IV and the one that generally lies at the basis of the huddling together of individuals, or of their crowding into corners or crevices. Loeb brought out this principle very nicely with some negatively heliotropic butterflies which wedged themselves closely between two plates of glass in the presence of light, showing how one tropic influence — in this case contact irritability — is competent 1 C. B. Davenport, Experimental Morphology, Part I, pp. 180-210; Loeb, Studies in General Physiology (1905), pp. i-i 14, from which most of the examples are cited. ^ Loeb, Studies in General Physiology, p. 20. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 251 to overcome an opposite but weaker one, — in this case nega- tive heliotropism. It is a noteworthy fact that irritabihty to Hght, while a prop- erty of protoplasm in general, is more pronounced in some cases than in others, even within the same organisms. This is true not only in the ej/cs of animals, but, in general, the oral (anterior) end of eyeless animals is much more sensitive to light than is the aboral ; ^ as also is the dorsal surface more sensitive than the ventral. Light therefore operates strongly to influence not only the position but the locomotion of animals as well. The following conclusions from Loeb upon the influence of light are valuable.^ In substance they are : I. " The dependence of animal movements on light is in every point the same as the dependence of plant movements on the same source of stimulation." 1. "The direction of the median plane or the direction of the progressive movements of an animal coincides with the direction of the rays of light." 2. " The more refrangible rays of the visible spectrum are more effective than the less refrangible rays." 3. " Light of a constant intensity acts as a constant stimulant." 4. Heliotropism is in a large measure dependent upon the intensity of light, differing for different animals. 5. " Heliotropic movements occur only between certain limits of temperature." II. " The orientation of an animal toward a source of light depends on the form of the body, just as the orientation of a plant to light depends on the form of the plant." 1. " Symmetrical points on the surface of dorsi ventral animals possess equal irritabilities." 2. The " irritability of the oral pole of an animal is different from the irritability of the dorsal pole," and is generally greater. 3. " The irritability of the ventral surface is different from the irritability of the dorsal surface." " These three conditions taken together cause dorsi ventral animals to place their median planes in the direction of the ^ Loeb, Studies in General Physiology, pp. 76-78. 2 Ibid. pp. 81-84. 252 CAUSES OF VARIATION rays and to move toward or away from the source of light in this direction." 4. " Eyeless animals behave in this respect like animals hav- ing eyes." III. " Heliotropic irritability of an animal manifests itself frequently only at certain epochs of its existence." 1-4. In winged ants this epoch is at the nuptial flight; in plant lice it is when the wings are present ; in the larvae of Musca vomitoria negative heliotropism is most prominent when the larva is fully grown ; and in a large number of animals the irritability is opposite in the larval and in the adult stages. 5. " Both night and day Lepidoptera are positively heliotropic, and their heliotropism is similar to that of every other positively heliotropic animal. The period of sleep of the night Lepidop- tera, however, falls in the daytime, and only for this reason is their heliotropism manifested exclusively at night." ^ IV. "In many animals heliotropic irritability is connected with sexuality." Ants are sensitive only at the time of the nuptial flight, and in both ants and Lepidoptera the males are more sensitive than the females. V. " The behavior of an animal depends on the sum total of its different forms of irritability." VI. Many animals are "compelled to orient their bodies against the surfaces of other solid bodies," or to bring their bodies "in contact with other solid bodies on as many sides as possible (stereotropism)." VII. Animals " may be forced by light to move from diffused light into sunlight and to remain exposed to the high tempera- ture of the sunlight, even though it may cause their death." Considering all the "irritabilities" and " tropisms " to which animals and plants are exposed and to which they react, it is not necessary to appeal to instinct^ or even to the nervous system^ to explain all movements of animals, nor is it well to ascribe to them such anthropomorphic qualities as love of light, distaste for darkness, preference, avoidance, curiosity, reason, or other such bases of higher intelligent action. 1 The question naturally arises, Why is the daytime the period for sleep in a positively heliotropic animal ? The answer has not yet been given. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 253 To these conclusions two observations may well be added for present purposes : I. Heliotropism, like many other reactions of protoplasm, arises from the nature of the organism, and not necessarily from 14:31 V'" ft" f.c 1 10:55 10:44 I! 10:34 28 (2). 14:0O to 14:20 —> ^■^4 Interval not observed 13:06 13:00 '^ I 12:54 y 14:21 (3). 14:20 to 14:34 Fig. 28. Effect of light upon locomotion : movements of an amoeba in response to changing directions of light. Ar- rows show the direction of the light rays, andfigures showthehour(i3= i o'clock, etc.). — After Davenport f(1). 12:48 to 14:00 a basis of utility. For example, roots are in general positively heliotropic, — a quality that is not of the slightest usefulness, and which must be regarded as entirely accidental. Again, this 254 CAUSES OF VARIATION quality is often fatal, as in the case of moths. In general, how- ever, matters have long since become adjusted to these reactions as to other necessities governing the behavior of living matter. 2. All experiments indicate a high degree of variability among individuals, not only as regards the degree of response to hcliot- ropism, but also as regards the effect of all other outside influ- ences, even that of poisons. Among examples furnished by the extended investigations into this subject, we have space to note but two. The amoeba, which represents about the simplest form of animal life, is an excellent medium for illustrating sensitiveness to light. Fig. 28 exhibits the movements of one of these bits of living matter under the influence of light, whose direction, as shown by the arrows, was occasionally changed, the figures in- dicating the hour, — all of which is strongly suggestive of the process of driving sheep. The other example to be noted is the effect of light upon the position of the chlorophyll granules in the leaf cells, under dif- ferent degrees of illumination, whether on the exterior walls, the partition, or the interior walls. Conditions that determine heliotropism. Some organisms are positively heliotropic in one intensity and negatively so in another ; some are positive at one temperature and negative at another ; some are positive in a certain concentration (of sea water) and negative in another, and the general principle may be stated that decreasing the concentration has the same effect as increasing the temperature.^ It thus appears that the same organism can often be made positively or negatively heliotropic at will by altering the surrounding conditions of life. SECTION VII — INFLUENCE OF TEMPERATURE UPON LIVING MATTER The relation of heat to living matter is mainly, but not exclu- sively, quantitative ; that is to say, the effect of heat is princi- pally upon the rate of growth and activity. In general, each species has its maximum, above which protoplasm becomes 1 Loeb, Studies in General Physiology, pp. 265-294. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 255 inactive (heat rigor); its minimum, below which all activity ceases ; and its optimum, — that point at which growth is most rapid. Certain facts in this connection are noteworthy : 1. The maxima, minima, and optima are not the same for different species. 2. Protoplasm is killed if carried much above the maximum, — the organism decomposes and is destroyed. 3. Temperatures below the minimum are not fatal except in the presence of moisture, which, on conversion into ice, destroys the structure of protoplasm by the act of expansion. 4. The optimum at which growth is most rapid is nearer the upper than the lower limit. 5. Both the optimum and the maximum may be raised by careful methods involving gradual acclimatization. Specific effect of heat upon protoplasm. ^ Beginning at the optimum and decreasing both ways to the limits, it may be said in general that protoplasmic activity is in proportion to the tem- perature. This is true of the amount of oxygen absorbed, of carbon dioxid evolved, of chlorophyll formed, and of carbon fixed, — in other words, of metabolism. The same is true as to movements of protoplasm and its irritability to light, contact, or other stimuli. The following table exhibits the number of electric shocks per second recjuired at different temperatures to produce tetanus in the neck muscles of a tortoise.^ Effect of Temperatures upon Animal Activities Temperatures Shocks per Second re- OlIIRED TO produce Tetanus Tempekatukes Shocks per Second re- quired TO PRODUCE Tetanus 4°C. 9°C. I 5 21°C. 28° C. 25 34 Effect of heat upon the rate of growth in plants.-^ The relation of temperature to plant growth is well shown in the tables on the following page.* 1 C. B. Davenport, Experimental Morphology, Part I, pp. 222-231. 2 Ibid. p. 230. 3 Ibid. Part II, pp. 450-460. * Ibid. p. 451. 256 CAUSES OF VARIATION Growth in Millimeters of the Plumules and the Radicles of Seed- lings GROWN UNDER DIFFERENT TEMPERATURES FOR 48 HoURS (Sachs) Plumules Radicles Temperature C. Maize Bean Pea Temperature C. Maize Bean Pea 14-16° 17.0° 2.S1 4.0 1 16-18 4.6I 7-4^ 3,01 257 39 I 8- 20 26.3 24.5 47 20-22 28.5 34 41.0 22-24 1,?>-^ 390 30 17.0 24-26 34-0 55-0 28 26-28 S.6 II. 1 0.0 38.2 25.2 22 12.2 28-30 42.5 5-9 7 30-32 32-34 I I.O 10.5 5-7 34-36 13.0 15.0 5.0 36-3S 3S-40 9i 10.2 5-5 42.5 4.6 7-5 Maxima, Minima, and Optima for Various Species arranged ACCORDING TO THE OfITMA ^ Sfecies Bacillus phosphorescens Fenicillium Phaseolus multiflorus^ (bean) Fhaseolus multiflorus* (bean) Pisum sativum •* (pea) Sinapis alba* (white mustard) Lepidium sativum* . . . . Linum usitatissimum* (fla.x) . Lupinus albus* Hordeum vulgaie* (barley) Triticum vulgare* (wheat) . . Yeast Bacillus subtilis Bacterium termo Zea mais ■• (Indian corn) Zea mais"* (Indian corn) . . Cucurbita pepo * (gourd) . . Bacillus ramosus Bacillus anthracis Bacillus tuberculosis . . . . Bacillus thermophilus . . . Optimum 20.0" 22.0 26.3 33-7 26.6 27.4 27.4 27.4 28.0 28.7 28.7 28-34 30.0 30-35 33-7 34-0 33-7 37-0 37-0 38.0 63-70 Minimum 00.0" 1-5 9-5 6.5 0.0 1.8 1.8 7-5 5.0 5.0 0.0 -f- 6.0 5-0 9-5 137 13.0 14.0 30.0 42.0 2 + 2 + 7 5 o o + 1 Growth during 96 hours. 2 C. B. Davenport, E.xperimental Morphology, Part II, p. 454. 3 Radicle. < Plumule. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 257 Commenting on this table, its author observes in substance : 1. That in general the optima, the minima, and the maxima rise and fall together ; that is, a species with a high optimum will also have a relatively high maximum and minimum. 2. That species vary greatly ; so much so that the maximum of one (/>. phospJiorcscens) may be below the minimum of another {B. thcnnopltilits). 3. That the optimum for the radicle and the plumule may be widely apart, as in the bean. 4. That, in general, the optimum is in close relation to the natural habitat of the species, as in B. phosphorcsccns that lives in the moderate temperature of the sea, and in B. therviopJiiliis that lives in the high temperature of decaying manure. From collateral evidence this must be ascribed to acclimatization. 5. That of all the species noted, the bacteria have the greatest range in optimum, showing that they are, as yet at least, less fixed in their organization. 6. That the minimum never falls below 0° C, the freezing point of water, which is the minimum for vital activities. 7. That the maximum temperature tends to be rather constant with related species, and among flowering plants the range is but 9° (37°~46°)- But 46° is a fatal temperature for most proto- plasm, and 50° is the limit, showing how near the limit some species have been pushed. The extraordinarily high temper- atures of B. tJiermopJiilus must be regarded as an instance of acclimatization, of which other striking examples are found in hot springs.^ 8. That the range from minimum to maximum varies with the species. In this table the range is least for B. tuberculosis (12°) and greatest for B. pJiospJiorescens (37°). 9. That the "wonderful adjustment" of critical temperatures to the environment of the species is not to be regarded as evidence of selection, but, as is elsewhere shown under "Accli- matization," it is due to the modification ivrougJit in the pjvto- plasni by the temperatw'e itself. 1 To the above may be added the observation that the optimum lies nearer the upper limit ; that is, the difference between the optimum and the maximum is less than the difference between the optimum and the minimum. 258 CAUSES OF VARIATION Effect of heat upon growth in animals.^ All larger land animals have acquired facilities for maintaining practically a constant tem- perature. This is not true for all animals, many of which, like marine species, are notably dependent for their temperature upon the accident of environment, in which respect they differ but little from plants. It remains to note, therefore, what influence heat may exert upon the growth of animal life so conditioned as to be dependent upon the surroundings for its temperatures. Increase in Length (Millimeters) of Young Tadpoles of Frog and OF Toad, under Different Temperatures, from the 24TH to THE 48TH Hour after Hatching ^ Temperatures Average Growth Temperatures c. Average Growth C. Frog Toad Frog Toad 9-11° 4-5 3-0 23-^5° 41-3 11-13 5-3 5-3 25-27 31-5 39-0 13-15 4-3 (?) 155 27-29 40.0 15-17 .6.3 29-31 47-5 56.8 17-19 9-5 31-33 40.2 55-3 19-21 19.8 21.2 33-35 43-5 21-23 Twenty-nine to thirty-one degrees seems to be about the optimum temperature for both, from which the table shows that the land-living toad prospered rather better under the higher temperatures than did the frog, as he certainly suffered more under the lower, but that in botJi the rate of growth was sub- stantially /;/ proportion to the temperature. The author of the table reports that the interval between fertilization and hatching in cod varies from thirty days at a temperature of — 2°, to thirteen days at a temperature of 6°-8° ; that herring vary from forty days at 2°-4°, to eleven days at iO°-i2°; and that the time required for the frog to attain a development at which the head and tail are sharply defined is, at 15°, six days ; at 33°, one day. 1 C. B. Davenport, Experimental Morphology, Part II, pp. 457-460. 2 Ibid. p. 457. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 259 Birds develop only at high temperatures. The normal tem- perature for the chick is 38°. Fere^ incubated at temperatures varying from 34° to 41°. The individuals were all examined at the same absolute time, and the following figures express the percentages of development attained, taking that of 38° as the standard : ^ Temperature . . . . Index of development 34" 0.65 35^ 0.80 36° 0.72 38° 1. 00 39^ 1.06 40- 41" 1.51 The author remarks that some doubt attaches to the figures under 35°, 36°, and 37°, and calls our attention to the fact that somewhere not far above 41° the series would become zero. But for the range given, the development is in proportion to the temperature, although the highest given (41°) is considerably above that attained under natural conditions. The groiving cJiick therefore does not, in nature, achieve its optiminn. Effect of heat upon the direction of growth, — thermotropism.'-^ Without going into the methods of investigation, it appears that, independently of the influence of light or other " tropisms," plants are often positively or negatively thermotropic largely according to temperatures. The plumule of seedling maize, for example, is known to be positively thermotropic at ordinary temperatures, while the radicle is positive between 15° and 35°, indifferent at 37.5°, and negative above. The indifferent point with the bean (radicle) is given at 22.5°. The subject is little understood, and though the impulse of thermotropism is weak as compared with that of heliotropism or geotropism, it is supposed to be one which inclines the organism to align itself in accord with the direction of heat rays, although it is true that thermotropic plants are sensitive to conducted as well as to radiant heat. Temperature limits of life.'^ All experiments indicate that as the temperature rises above the maximum the first effect is heat ^ C. B. Davenport, Experimental Morphology, Part II, p. 459. 2 Ibid. pp. 463-467. 3 For e.xtended discussion, and for tables of temperature limits, see C. B. Daven- port, E.xperimental Morphology, Part I, pp. 231-249. 2 6o CAUSES OF VARIATION rigor, which soon passes into death. The more rapid the rise the lower the death point, and the more gradual the rise the greater the resistance. Again, if the temperature does not rise too high, the heat rigor may gradually pass off, and activity may be resumed, even at temperatures which at first were followed by entire suspension of activity, and even by rigor. This is the first stage in the- process of acclimatization. Sachs found that a sensitive plant kept at " 40° C. for one hour suffered loss of sensibility during twenty minutes after removal. Raised slowly to 50°, sensibility was only temporarily lost, but 52° proved fatal. Immersed in water, heat rigor occurred at a temperature 5° to 10° lower." ^ Hofmeister found that hairs from the stem and leaf of Ecballiiim agreste, showing lively movement, when gradually raised from i6°-i7° C. to 40° C. "became motionless," but that "after one or two hours move- ment returned and was very violent. Cooled and raised again to 45° C, the protoplasm was motionless at first, but after seven- teen minutes movements recurred but were not rapid." ^ The vital limit varies greatly with the species. Thus, roughly speaking, for bacteria it is 45° C; for cryptogams, generally 45°-50°, with an occasional one at 60° ; flowering plants, 45°-50°; protozoa, 40^-45°, with a few as high as 60°; mol- lusks, 30^-40°; worms, 45°-50° (tardigrades, dried, 98°); crus- taceans, 26°-43°; insects, 27°-4'3.7°; fish, 27°-40° ; salamander, 44°; frog, 40°-42°; dog, rabbit, and man, 44°-45°; vertebrate muscle, 40°— 50°.^ This series exhibits a wide range of resistance to excessive heat, yet few organisms can endure much above 50°. All experiments and observations indicate that death from high temperatures is caused by coagulation of the albumen of the protoplasm, a circumstance showing that albumen carries into its vital relations its ordinary property of coagulation by heat. That living matter contains many easily coagulable pro- teids no longer admits of doubt, and that their coagulation causes death is evidenced by the fact that once in this condition they do not return to their normal state. 1 C. B. Davenport, Experimental Morphology, Part I, p. 232. 2 For full table.s from which these abstracts are made, see C. B. Davenport, Experimental Morphology, Part I, pp. 234-237. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 26 1 Death from low temperatures appears to result from entirely different causes. Protoplasm seems to contain no substance but water that undergoes either chemical or serious physical change by low temperatures. Many yeast cells endure — 1 13.7° C. (Schumacher). De Candolle subjected " various dry seeds and spores of bac- teria to a temperature of nearly — 200°, at which temperature the atmosphere becomes liquefied, but without fatal effects." " Cilia from the mouth of the frog were cooled to — 90°, and recovered their movement upon raising the temperature." " Eggs of the frog, lowered slowly to —60°, can revive." From facts such as these Davenport concludes that ''there is no fatal teviperattire for dry protoplasm} The first effect of lowering temperature is a slowing of activity, followed, finally, by complete cessation. As is pertinently re- marked by the author just quoted, " The fact that cold rigor usually occurs close to the zero point (C.) indicates that the activities of protoplasms are closely determined by the fluid state of water," and " the critical point for vital activity has been adjusted to this critical point of water." ^ There is much lack of information upon the exact cause of death from excessive cold. Among the higher animals the immediate cause is without doubt asphyxia from the cessation of the blood flow ; but among the simpler organisms the matter is not so clear. What evidence w^e have seems to indicate that the primary cause of death is in all probability the mechanical rupture of protoplasm and cell wall by freezing water expanding as it solidifies. In any event experience and experiment agree in indicating that protoplasm is resistant to excessive cold in the absence of moisture, and in all study of this matter we are to remember two facts : first, that the freezing point of all protoplasm is lower than that of water only ; and, second, that as long as the slightest activity is present heat is being produced. From these two facts the protoplasm is able to resist actual solidifying much 1 C. B. Davenport, Experimental Morphology, Part I, pp. 240-242. Though not so stated in the text quoted these temperatures are C. 2 Ibid. p. 242. 262 CAUSES OF VARIATION longer and under much lower temperatures than we should at first s«ppose. Substantive variation due to temperature ; color markings. Early in the section it was remarked that the effects of tem- perature are qualitative as well as quantitative. Without doubt temperature exerts a controlling influence upon the color of butterflies, as has been determined by a number of direct ex- periments. For example, Vanessa levana and V.prorsa were long regarded as distinct species. Levana is " characterized by a yellow-and- black pattern on the upper side of the wings," while proi^sa " has black wings with a broad white transverse band and deli- cate yellow lines running parallel to the margins." ^ Later this was recognized as a case of " seasonal dimoj-phism," the yellow- and-black levana being the spring brood and the darker prorsa being the summer brood ; that is to say, levana, ernerging in the spring, breeds immediately, producing a summer brood {pT'orsa), and this brood in the same way gives rise to a genera- tion which passes the winter in the chrysalis form, emerging in the spring as levana. Thus these two " species " are produced from the same stock, the difference being that one passes the chrysalis stage in the summer, the other in the winter. That this difference is one of temperature was proved by direct experiment. Dorfmeister ^ succeeded in producing prorsa directly from prorsa by the application of warmth to the pupae, and " by the application of cold he obtained from levana not the pure levana form, but one intermediate between it and prorsa^' — an intermediate occasionally observed in nature and known as V. porinia? Weismann, repeating the experiment, found that by using lower temperatures levana could be produced directly from levana, and he adds, " The converse experiment was also occa- sionally successful, the pupae of the winter generation being forced to assume the summer form by the influence of a higher temperature during, or shortly after, pupation." ^ 1 Weismann, Germ Plasm, p. 379. 2 Vernon, Variation in Animals and Plants, p. 233. ^ Ibid. ; also Weismann, Germ Plasm, p. 379. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 263 Here temperature seems to exert a controlling influence upon pigment formation, although Weismann is careful to inform us ^ that the two patterns " do not correspond " ; that if we were to "superpose" one upon the other, "it would seem that the black parts in prorsa do not correspond to the yellow ones in levana, and that the white band in the former does not corre- spond to [either] a yellow or [a] black part in the latter. This band is, on the contrary, entirely zvanting in levana, and is represented by both black and yellow regions y ^ Again, Weismann experimented with Polyommatus pJiloeas^ a species " distributed over the whole of the temperate and colder parts of Europe and Asia." Toward the north (in Germany) the upper surface of the wings is of a "beautiful reddish-gold color," hence its popular name, " fire butterfly." But he says, " Farther south the reddish-gold color is more or less thickly dusted with black, and specimens from Sicily, Greece, or Japan often display only a few reddish-gold scales, the general appearance being almost black." " In Germany this butterfly is double-brooded, and the two generations are similar, but in certain districts of southern Europe . . . the first generation is reddish-gold, — the second, which flies in midsummer and is known as the variety eleiis, having the wings well dusted with black." " As in Germany during exceptionally hot summers individuals with a blackish tint have repeatedly been caught together with the ordinary form . . . ," and Weismann observes that it would seem " the butterfly becomes red when exposed to a moderate temperature and black when the heat is greater." Attempts to produce these forms at will, however, by regula- tion of temperature only partially succeeded. But the conditions were severe. There was no common ancestor. Weismann under- took to produce the southern form from the fiorthern stock and vice versa. Insects reared from German butterflies but kept in high temperatures were in many instances " dusted with black, but none of them resembled the darkest forms of the southern elens.'' Conversely, butterflies raised in cool temperatures from 1 Weismann, Genu Plasm, p. 379, 2 Ibid. pp. 399-4CO, 264 CAUSES OF VARIATION Neapolitan stock were lighter in color than in their native habi- tat, but " none were so light-colored as the ordinary German form." ^ This difference he ascribes to the cumulative influence of the natural seasonal temperatures, and is quick to protest against its interpretation as indicating an inheritance of acquired characters. He calls it a case of internal selection as between "winter and summer determinants." However, that is of no consequence in the present connection. The facts here given show beyond a doubt that outside tempera- tures exert a direct effect upon so important a character as color. Whether this occurs by chemical disturbance in pigment formation, by internal selection, or by other means does not greatly matter here. There is some evidence tending to show that the light color of polar animals is due to the direct action of cold.^ This, if true, argues for chemical action upon pigment as the cause of color changes due to temperature. Temperature an all-pervading influence. Temperature differs from most other external forces in being, for many species, at least, an all-pervading influence. Higher animals and plants are themselves centers of heat production, and in general their temperatures are the algebraic sum of their own heat production and the heat of their surroundings. Lower organisms, however, are very largely dependent upon their environment for their temperatures, and in cases of this kind the entire protoplasm of the body is affected. SECTION VIII — EFFECT OF CHEMICAL AGENTS UPON PROTOPLASMIC ACTIVITY All development, all differentiation, and all functional activity of living organisms are the result of protoplasmic activity ; but protoplasmic activity is, in the last analysis, chemical activity, and it is certainly subject to many of the laws controlling ordi- nary chemical reactions. It is noticeable in the study of vital 1 Weisniann, Germ Plasm, pp. 399-400. 2 The writer has somewhere read that animals on shipboard become rapidly lighter in color as the coat becomes exposed to intense cold, but he is unable to verify the report. EXTERNAL INFLUENCES AS CAUSES OF VARLATION 265 processes from the chemical standpoint that some substances exert no influence upon protoplasm, while others kill it out- right; that some accelerate and others retard its normal action; and that some suspend activities more or less completely, while others divert them into entirely 7iezu cJiannels. Here is varia- tion due to chemical disturbance of the material basis of life, and it is well to study somewhat in detail this " modification of vital actions " from chemical causes.^ In studying this class of phenomena it is necessary, of course, to make use of simple organisms of one or of few cells, and while we cannot reason directly from these to the higher animals and plants, still all evidence goes to show that the differences are not so much in kind as in complexity. Oxygen. All experiments indicate that no protoplasm can long survive in the absence of oxygen. In most cases it is taken directly from the air, but in others, as in anaerobic bacteria, it is probably extracted from surrounding compounds containing oxygen. Diminished oxygen retards and increased oxygen and ozone greatly accelerate the vital processes, ^ all zuitJiout cJiang- ing their character. A number of oxygen-containing substances greatly retard or even destroy vital activities, probably through " oxidation of the protoplasm." ^ If this be the case, and the material basis of life is subject to the ordinary chemical process of oxidation, it shows that vital processes in the last analysis rest upon a strong chemical basis. Hydrogen peroxide (H^Og), only slightly different from water (H2O), is a powerful oxidizing agent. One part in ten thousand (0.0 1 per cent) in hay infusion killed all ciliata in from fifteen to thirty minutes. Algae survived a o. i per cent solution but ten or twelve hours, and died in a 10 per cent solution in a few minutes. *' Salts of chromic, manganic, permanganic, and hypo- chlorous acids act as intense poisons, apparently by directly yield- ing oxygen atoms to the plasma proteins." Chlorin, iodin, and ^ C. B. Davenport, Experimental Morphology, Part I, chap, i, from which the data in this section are largely taken. 2 Small animals confined in an atmosphere of pure oxygen exhibit greatly increased activity and " live themselves to death " in a few hours. ^ C. B. Davenport, Experimental Morphology, Part I, p. 3. 266 CAUSES OF VARIATION bromin in the presence of water act " fatally upon all organisms by splitting [the] water, forming hydro-halogen compounds, and leaving the oxygen to unite with the living protoplasm." ^ Hydrogen. Amoebae subjected to an atmosphere of hydrogen for twenty-four minutes became motionless, some having assumed the spherical form. The same general result followed in trades- cantia hairs, but from the fact that normal activity was restored by admitting air it was assumed that the results arose not from any injurious effect of hydrogen but from the exclusion of oxygen.^ Oxids of carbon, — CO2 and CO. These two oxids of carbon have very different effects upon protoplasm. The former, like hydrogen, seems to act only by excluding oxygen, death, when it results, being due mainly to asphyxia, while the latter kills by attacking the protoplasm directly.'^ Catalytic poisons.* A large number of unstable carbon com- pounds, neither acid nor basic and therefore not characterized by intense chemical action, are yet violent poisons. Here belong the anaesthetics, as chloroform, chloral, ethyl, ether, alcohols, etc. These unstable compounds are characterized by a " lively condition of molecular movement " (Nageli), which is considered to disturb the normal movements of the protoplasm, or " to lead to chemical transformations in the unstable albumen of the protoplasm" (Loew). Catalytic substances are supposed to exert their action not by entering into and effecting new combinations but by disturbing, through their mere presence, the usual behavior of bodies. It is in this way that protoplasm suffers in their presence. Thus hydrochloric and prussic acids unite only at high temperatures, except in the presence of various ethers, when they will unite even at — 15°. The vigor of this catalytic action is in proportion to the molecular composition.^ Thus in the methan series with CHg as the base we have CH^, C2Hg, CgHg, etc., in which 1 C. B. Davenport, Experimental Morphology, Part I, p. 4. 2 Ibid. p. 5. 3 Ibid. p. 6. ^ A free extract from C. B. Davenport, Experimental Morphology, Part I, pp. 7' 8. ^ The facts here stated are taken almost literally from C. B. Davenport, Experimental Morphology, Part I, pp. 7-8. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 267 " the poisonous action increases up to a certain limit in proportion to the number of C atoms," while "above this hmit the com- pounds are more stable and are more indifferent, as for example paraffin {C^^H^^ to €271^53)." Again, in such a series, if the H atoms become replaced by one of the halogens, the poisonous properties correspondingly increase ; thus : ^ CH4, marsh gas, innocuous. CH3CI, slightly anaesthetic. CHCI3, chloroform, powerful anaesthetic. CCI4, very dangerous, stupefying involuntary muscles. Chloroform and ether affect all protoplasm, both plant and animal, higher as well as lower. They seem to produce at first (two to five minutes in a 25 per cent water solution) a "very intense excitement in the movement of the protoplasm," fol- lowed by "strong vacuolization, and then the cytoplasm grad- ually becomes immobile " and dies, if the influence is continued. In a similar way the various alcohols exert stupefying effects in proportion to the number of CH2 radicals present, and carbon disulphid (CSg) is one of the most powerful catalytic poisons. Protoplasm is therefore subject to catalytic disturbances, in which it is not different from other and more ordinary chemical materials, — a fact in itself exceedingly significant to the stu- dent looking for fundamental causes of variation. Poisons which form salts."-^ These are acids and bases which Loew believes, as stated by Davenport, " unite [directly] with the protein substances of the protoplasm, producing salts," — disturbances that of course soon lead to death. Thus " formic acid, even in small per cents, — 0.05 per cent to 0.006 per cent, — prevents the development of bacteria. On the other hand, some protoplasm has acquired a resistance to organic acids, the vinegar eel Uving in 4 per cent acetic acid," and the gland cells of some marine Gastropoda secrete from 2 per cent to 3 per cent of H2SO4, a strength which is fatal to most protoplasm. 1 The halogens — fluorin, chlorin, bromin, and iodin — form a group of sub- stances of very similar chemical properties, but form, in the order named, a decreasing series as to chemical energy. ■•^ C. B. Davenport, E.xperimental Morphology, Part I, pp. 12-14. 268 CAUSES OF VARIATION Nageli's experiments ^ indicate that water distilled in cop- per vessels, or standing four days in other vessels with twelve clean copper coins per each liter of water, was fatal to Spiro- gyra, though the proportion of copper to water was but i to 77,000,000. These reactions, resulting in death rather than in modified action, are important, not as showing primary causes of varia- bility but as proving again that living protoplasm is still subject to many of the chemical affinities that controlled its elements before they became organized into living matter. It must be remembered in this connection that many of these substances attack only living protoplasm, having no action upon dead pro- toplasm, showing that at death the highly complex materials have, to some extent at least, broken down. The action of some poisons, like nicotin, is proportional to the "differentiation of nervous substance" ; others, like cocain and atropin, first excite and then paralyze the central nervous system of vertebrates, but act as violent poisons upon undiffer- entiated protoplasm (Protozoa).^ Toxic poisons. It is not the germ that kills, but rather its specific toxin that deranges some of the vital functions beyond endurance. The dire effects of germ diseases are therefore due not so much to the organisms themselves as to their constant manufacture within the body of a chemical poison which the protoplasm cannot endure and preserve its normal functions. It may die in the attempt, or it may succeed and become accli- mated, but while the struggle is on, the body functions will be considerably disturbed. It is significant that compounds similar to those of disease-producing bacteria have been " extracted from the seeds of some phanerogams ; for example, ricin from the seeds of the castor-oil bean, etc." In this class may come the poisons secreted by certain animals, as the rattlesnake and cobra, fatal to vertebrates but innocuous to Infusoria and Flagellata. It is also noteworthy that the blood serum of one species rapidly dissolves the corpuscles of another (red and white), and is therefore injurious or fatal according to the amounts present.^ 1 C. B. Davenport, Experimental Morphology, Part I, p. 14. 2 Ibid. pp. 23 and 24. ^ Ibid. p. 22. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 269 Specific secretions and glandular activity. The fact just men- tioned introduces a subject full of interest. It appears that each species, and perhaps each individual, is engaged in the produc- tion of chemical substances (whether the result of metabolic or katabolic activity is uncertain) which exert specific action upon living matter. It is significant that some of the lower organisms producing definite substances die from the injurious effects of their own product,^ unless this is removed as formed, and that higher ani- mals are supplied with elaborate excreting apparatus, strongly suggesting that certain of their products are deleterious to the organisms that produced them, while in other cases they are clearly beneficial. In this connection Loeb remarks : ^ It is perhaps not impossible that those mental diseases that are heredi- tary are, in reality, chemical diseases caused by poisons that are formed in the body, just as special substances — for instance, alcohol, hashish, and other intoxicating substances — produce temporary mental diseases. The delirium of fever, as well as certain other mental diseases, may owe their origin to poisons which are formed in the body. It is quite possible that these poisons are also formed in the normal body. It is only necessary that they be formed in somewhat larger quantities or destroyed in some- what smaller quantities in the body of the insane than in the normal man.'^ It is further not at all necessary that the hypothetical poisons which cause mental diseases be formed in the central nervous system. They may be formed in any organ of the body. It is only necessary that they affect the central nervous system ; in other words, that they be nerve poisons. Nothing is better qualified to make this view clear than the result which the destruction of the thyroid gland has on the mental and physical devel- opment of children. We know that in case of degeneration of the thyroid gland the growth and mental development of the child are retarded. Idiocy may result from the destruction of the thyroid gland. It has been found that an improvement, or even a cure, can be attained by feeding patients afflicted with this trouble with the thyroid substance of animals.'' Baumann found that the thyroid gland contains an element which is contained in no other organ of the body, — namely, iodin. 1 The yeast plant that forms alcohol dies when the solution has reached a strength of about 20 per cent. 2 Loeb, Physiology of the Brain, p. 207. ^ It is said by Lombroso, and others agree, that the criminal is characterized by excessive amounts of urea. * Medicinal preparations from various glands are now regularly supplied by the large slaughterhouses. 2 70 CAUSES OF VARIATION Insect poisons. The poison of bees — formic acid — is fatal to insects and small animals, and in sufficient quantity to the larger vertebrates, including man, though frequent stings of a moderate number lead rapidly to acclimatization. It is note- worthy, too, in this connection that the sting of the insect (mud wasp, for example) does not always kill but often merely paralyzes, so that the creature stored with the egg will remain alive to furnish food to the larvae some weeks later. Galls. Galls are the direct result of the sting of an insect, leaving a specific poison to act upon a particular form of proto- plasm. The result is not death but a diverting of the activities into entirely new channels. Darwin states that " no less than fifty-eight kinds of gall are produced on the several species of oaks by Cynips with its sub-genera, and Mr. B. D. Walsh states that he can add many more." ^ Darwin further remarks that many gall insects are exceedingly small, and that consequently the drop of poison they inject must be exceedingly minute ; moreover, it is never injected but once. The growth that follows, however, is specific and continuous. He quotes Walsh as saying, " Galls afford good, constant, and definite characters, each kind keeping as true to form as does any independent organic being," ^ and he calls our attention to the fact that seven of the ten distinctly different galls produced on the willow are by insects which, " though essentially distinct species, yet resemble one another so closely that in almost all cases it is difificult and in most cases impossible to distinguish the full-grown insects one from another." The difference in the quality of the poison secreted by insects so nearly alike cannot be great, yet it is sufificient to give rise to galls widely different. Last, and not least, he mentions that " Cynips fccjmdatrix has been known to produce in the Turkish oak, to which it is not properly attached, exactly the same kind of gall as on the European oak. These latter facts apparently prove that the nature of the poison is a more poxvcrful agent in determining the form of the gall than [z>] the specific character of the tree ivhich is acted on." ^ 1 Darwin, Animals and Plants, II, 272. ^ ibid. p. 273. ^ Ibid, (second edition), 273. Tlie italics are mine. EXTERNAL INFLUENCES AS CAUSES OF VARLVFION 271 Tumors. Those overgrowths of various parts of the body, called tumors, arising from causes not well understood, have their specific characters as truly as if derived from inherit- ance. Whether the character of the tumor is derived primarily from the tissues affected or from some outside specific cause is not known, but it is a significant fact that the protoplasm, ivhich derived its cJiaracters originally by inheritance , has undergone permanent and definite alteration tJirongJi the operation of causes absolutely distinct from inheritance , whether internal or external to the organism, thus sJiowing the possibility of diverting the ener- gies of inJierited 7fiaterial into absolutely new channels through apparently slight causes. At this point it is well to call our attention to the profound changes (permanent variations) wrought upon the constitution of the individual by such internal-external circumstances as vacci- nation or the injection of antitoxin, as well as to the immunity acquired through a single attack of an infectious disease. Germination of seeds. It is a well-known fact that certain chemicals accelerate and others retard the process of germina- tion. Just why this is true is not clear on any other ground than that of the ordinary susceptibility of growing protoplasm to stimulants and sedatives. The process of germ development requires, in addition to what is contained within the seed, only oxygen, water, and a favorable temperature ; the influence of other chemicals must be indirect. Chemotaxis and chemotropism.^ The influence of chemical substances upon the locomotion of free-moving organisms is tech- nically known as chemotaxis, and their influence upon the direc- tion of growth (in plants) is known as chemotropism.'^ The distinction is hardly worth observing for our purposes, because both phenomena arise from a direct influence of chemical sub- stances upon living protoplasm, either attractive or repellent. If . attractive, it (the protoplasm) will move toward the material in 1 C. B. Davenport, Expeiimental Morphology, Part I, 32-45; Part II, pp. 335-342; Loeb, Physiology of the Brain, pp. 50, S8-90, iiS, 1S6-188. 2 As has already been noted, the same distinction is often observed between geotaxis, the influence of gravity upon locomotion, and geotropism, its influence upon the direction of growth ; also between heliotaxis and heliotropism as cover- ing corresponding influences of the sun (light). 2 72 CAUSES OF VARIATION question, providing it is free to do so (as in the case of the lower plants and all animals), or its growth will be directed toward it if (as in the case of higher plants) it is fixed and un- able to move. The former is, strictly speaking, chemotaxis ; the latter is chemotropism. For our purposes it is a distinction without a difference, because in the latter case the plant is unable to indulge in locomotion, and performs the hearest pos- sible act in changing the direction of growth. Both are loco- motion under the circumstances ; both are due to the same cause, — a chemical affinity or attraction, — and they differ because of differences in the organisms affected in respect to the power of locomotion, not because of differences in the nature of the forces in action. The two terms are, therefore, for our purposes, synonymous as denoting a power of attraction between protoplasm and certain ordinary chemical compounds such as to cause the protoplasm to approach if it be free to move, or, if not, to grow in that direction, — which is all that can be done under the circumstances to satisfy the affinity. Chemotaxis, or chemotropism as the writer prefers to call it,^ has but a slender hold upon higher animals. It appears to be localized in the nostrils and to manifest itself only in the sense of smell, agreeable or otherwise ; but in lower organisms, even in many insects, it is apparently not confined to a minute fraction of the surface, but pervades the whole organism with an influence that is all but overpowering. It may of course be aided or opposed by other tropisms, as gravity or light, in which case the total result is the algebraic sum of all the energies operative, but that chemotropism is a force to be reckoned with in development is a fact not to be doubted. Examples of chemotropism. ^ Englemann noticed that bac- teria under a cover glass will gather along the margin, or if green algae be introduced they will cluster about them so long as they are producing oxygen, but in the darkness they will not 1 The ending " taxis " is from the Greek, meaning assortment or arrangement ; the ending " tropism" is from " tropic," to turn. " Chemotropism " is pronounced ke-mot'rd-pTzm. ■^ Until otherwise noted what follows is a free though not exact transcript from data given in C. B. Davenport's Experimental Morphology, Part I, pp. 32-39. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 273 be affected. In the same way nearly all kinds of motile organ- isms are now known to be influenced by a variety of chemical substances. Lubbock has shown that ants retreat from essence of clove, lavender water, etc.,^ placed within one fourth inch of their path, and Loeb found that the larvae of flies creep towards a piece of fiesh brought nearer than 1.5 cm. Not only flesh and decaying meat, but meat juice in a glass, will allure, while fat has no effect. Every farmer knows how quickly flies are attracted by a dressed animal, and carrion birds by a carcass. According to Pfeffer's experiments the inorganic salts of potassium, sodium, calcium, ammonium, magnesium, and many other metals in 0.5 per cent solution act attractively upon Bac- terium termo. " Inorganic acids ... in general act repulsively," but phosphoric acid and the phosphates are strongly attractive. Dewitz states that mammalian spermatozoa are attracted by KOH. " Alcohol in grades between 10 per cent and i per cent acts repulsively towards bacteria," but "glycerin is neutral." Malic acid, which is of wide distribution among plants, is strongly attractive to spermatozoids, even in a 0.00 1 per cent solution, — a fact which is highly significant. This principle of chemotropism acting on higher organisms gives rise to characteristic movements. In Loeb's experiments on actinians ^ a piece of meat laid upon the tentacles so affected them as to cause a bending which carried the meat into the mouth, while a wad of water-soaked paper had no effect, but lay there until removed. If, however, the paper was soaked in meat juice it was received the same as a piece of real meat. (See Fig. 29.) Now the actinian, consisting simply of a sac with a row of tentacles around the edge, without brain or nerve centers of 1 Experimenting upon means of preventing the ravages of tlie corn-root aphis, Forbes of Illinois found that a small amount of oil of lemon on the seed corn, before planting (costing but ten cents per acre), is able by its strong odor to repel ants from the neighborhood of the corn hill for no less than six weeks after planting. As the young aphis is absolutely dependent upon the attentions of the ant, this treatment is effective. 2 Loeb, Physiology of the Brain, pp. 49-50. 274 CAUSES OF VARIATION any kind, is certainly incapable of exercising anything like in- telligent choice. The characteristic movement must have been due to the specific chemical action of the juices of the meat upon the protoplasm of the muscle fibers of the tentacle, while the paper had no such action and hence no movement followed. This movement and non-movement look like intelligence or instinct and have often passed for one or the other to the infinite con- fusion of the subject.^ ^' Lnmbrici foetidi live in the decaying compost of old stables, and prob- ably the chemical nature of certain sub- stances con- tained in the compost holds them there," for "when one half of the bottom of the box is covered Fig. 29. Stimulating effect of certain chemicals upon mus- cular action : a piece of meat laid upon the tentacles of the actinian stimulates their action and it is drawn into the mouth. A piece of paper similarly placed is inoperative with moist blotting paper and the other half with a thin layer of compost, all the worms will gather on the compost side." Decapitated worms behave in the same way,^ so that the effect is due to a general influence, not to " nerve centers." Loeb says,'^ " I have often placed pieces of lean meat and pieces of fat from the same animal side by side on the window sill, but the fly never failed to lay its eggs on the meat and not on the fat." (He tried, of course without success, to raise larvae in the fat.) " It can easily be shown that larvae of the fly are posi- tively chemotropic towards certain chemical substances which are formed, for instance, in decaying meat or cheese, but which are not formed in fat. The substances in question are probably volatile nitrogenous compounds," and "the chemical effects of 1 This subject will be pursued further under " Instinct and Reflex Action." 2 Loeb, Physiology of the Brain, p. 90. ^ Ibid. pp. 186-187. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 275 the diffusing molecules on certain elements of the skin influence the tension of the muscles," causing motion. The female fly is attracted by meat, the same as are larvce, and " as soon as the fly is seated on the meat chemical stimuli seem to throw into activity the muscles of the sexual organs, and eggs are deposited on the meat." This chemical stimulus is about all there is of the wonderful "instinct" by which insects are led " always to deposit their eggs in exactly the right places." These and similar examples show the effect of certain chem- icals upon free-moving organisms. It remains to illustrate their effect upon the direction of grozutJi among plants, which are not free to move. First of all, we are to note the effect of certain chemicals upon the tentacles of insectivorous plants. Darwin noticed that "when drops of water or solutions of non-nitrogenous compounds are placed upon the leaves of the sundew, Drosera, the tentacles remain uninflected ; but when a drop of a nitrogenous fluid, such as milk, wine, albumen, infusion of raw meat, saliva, or isinglass is placed on the leaf, the tentacles quickly bend inwards over the drop."^ Darwin found that of " nine salts of ammonia tried, all caused inflection, and of these the phosphate was the most powerful," and that " sodium salts in general caused inflection with extreme quickness." This action of nitrogenous, phos- phatic, and other chemicals common to animal life, was the same upon the tentacles of insectivorous plants as upon' the tentacles of lower animals subsisting upon the same kind of food.^ Thus plant and animal tissues appear to be subject to the same general laws in this regard. From this point of departure, common to both plant and ani- mal, we note that the animal, free to move, does so in response to this class of stimuli. What does the plant do that cannot move, even as tentacles move .-* In other words, how does chemot- ropism affect the direction of grozvth among higher plants ? Roots in general are supposed to grow toward oxygen, and pollen tubes will certainly turn toward the stigma of the flower, 1 C. B. Davenport, Experimental Morphology, pp. 335-336. - See the e.xperiment on actinians previously quoted from Loeb, Physiology of the Brain, pp. 49-50. 276 CAUSES OF VARIATION the supposition being that sugar is the special attracting sub- stance in the latter case. It is noteworthy that the pollen tube is attracted not simply to its own stigma but also to the pistil and the ovule of other species, even of a different genus with which it is unable to unite. ^ Davenport says :^ The results of experimentation upon chemotropism show that various substances may direct the growth of such elongated organs as tendrils, roots, and hyphae of plants. ... In many instances it can be shown that the direction of growth is on the whole advantageous to the organi.sm. ... In other cases, however, the response seems to have no relation to adaptation. The immediate cause of change of direction is "excessive growth on one side, due to excessive imbibition or to excessive or restricted assimilative activity." The evidence seems conclusive that the chemical elements constituting living matter have not entirely lost their ordinary affinities and properties ; that protoplasm is in many respects subject to the laws of other chemical compounds ; that its activ- ities may be accelerated or retarded ; that they may be tempo- rarily or even permanently modified in character by chemical alteration, or even entirely destroyed by entering into new and strange combinations with surrounding substances. Here is a real cause, at least of occasional variation, — possibly of those sudden changes we call mutation. Rhythmical contraction of muscle instituted by certain chem- ical substances. The irritability and consequent contraction of muscles due to influences of a chemical nature have already been noticed, as in the case of the movement of the tentacles of the actinian, the attractive or repulsive effect of certain odors, and, to some extent, in the placing of insect eggs in "exactly the right spot." It is now well known that certain salts exert a specific action upon muscle, exciting even rhythmic contraction. As long ago as 1 88 1 Biedermann discovered that if the muscle of a frog be carefully excised at a low temperature {o°-io° C), then weighted and dipped into a 0.6 per cent solution of sodium chlorid con- taining also small amounts of sodium phosphate and sodium 1 C. B. Davenport, Experimental Morphology, p. 339. ^ Ibid. p. 343. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 277 carbonate "one observes as a rule, after a longer or shorter period of rest, that the immersed muscle begins to beat rliytJimically.'" On this point Loeb remarks that sometimes only a mere tremor is noticeable, at others violent contractions ; that some- times only individual fibers are active, at others the whole muscle is involved ; and that " at low temperatures these phenomena may continue for days." ^ These facts, together with Ringer's and Howell's statements that calcium and potassium salts exert a direct action upon the heart, led Loeb to extend the Biedermann investigations. He subjected the gastrocnemius muscle of a frog (unweighted) to a series of solutions of chemically pure materials in twice-distilled water.2 These experiments not only confirmed Biedermann's findings that the salts of sodium were able to excite rhythmic muscular contraction but they also added lithium, caesium, and rubidium to the list of bases, and the salts of bromin, iodin, and iron to those of carbon, chlorin, and phosphorus. These movements are periodic and continue into the second day, even at room temperatures. Loeb determined that if a muscle be immersed in a 0.7 per cent solution of sodium chlorid, contractions will begin in from sixty to ninety minutes, but that " if a trace of alkali is added, con- tractions begin much sooner." This acceleration he attributes to the hydroxyl (OH) in the alkali added. Not only that, but he ascribes the effect to the H involved in the hydroxyl, because the same action follows the addition even of inorganic acids, as HNO3, " if the same number of hydrogen ions are contained in the unit volume";^ but Loeb hastens to assure us that neither the hydrogen ions nor the hydroxyl ions " belong to those which are capable of Hberating rhythmic contractions." They only accelerate the action of those which of themselves possess this power. The action of sodium and other chemicals in exciting contrac- tion is, in the opinion of Loeb, to be ascribed to their entering 1 Loeb, Studies in General Physiology, Part II, p. 518. 2 Ibid. pp. 519-538. 3 Ibid. p. 527. 278 CAUSES OF VARIATION the muscle and there forming with its substance definite com- pounds, and he beheves the accelerating effect of H or OH is due to their catalytic action in facilitating the formation of these compounds. In this connection it is to be remembered that the serum of the body which bathes the muscles is always, in health, strongly saline and slightly alkaline. Further experiments clearly showed that the salts of potas- sium and those of calcium, magnesium, strontium, manganese, and cobalt tend strongly \o prevent contraction, this being espe- cially true in the case of potassium and calcium, forcing the conclusion that certain definite substances are necessary to con- traction ; that certain others tend to accelerate and still others to retard this characteristic activity of in?iscular tissue. Artificial parthenogenesis through changes in the surrounding solution.^ The indefatigable labors of Jacques Loeb upon this subject have not only thrown much light upon the essential features of fecundation, but incidentally they have afforded results of high value in determining the nature and range of external influences upon the characteristic activities of living matter.- It had long been known that many of the eggs of sea urchins, arthropods, and marine worms, even zvhen unfertilized, would, if left for a comparatively long time in sea water, begin to segment, reaching the two- and sometimes the four-celled stage. Loeb, and later Morgan, found "that if the concentra- tion of the sea water be raised sufficiently by the addition of certain salts, a segmentation of the nucleus takes place with- out any segmentation of protoplasm [cytoplasm]. Such eggs, however, when brought back into normal sea water divide into as many cells as there were preformed nuclei." '^ In none of these experiments did the cell division "lead to the formation of a blastula. A heap of cells, at the best about sixty, were formed, and then everything stopped." As in the case of tumors 1 Loeb, Studies in General Physiology, Part II, pp. 539-691. 2 These investigations have been published from time to time, especially in the American Journal of Physiology, and later (1905) in book form under the title, Studies in General Physiology. ^ Loeb, Studies in General Physiology, Part II, pp. 540-541. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 279 and galls, here was growth without systematic differentiation ; cell division without the formation of an embryo. Encouraged by this degree of segmentation and by his experi- ments upon irritability of muscle, Loeb tried a great variety of solutions, in various degrees of concentration, in the hope of carrying the segmentation far enough to produce real embryos and live, free-moving larvae. He was greatly hampered by the fact that unfertilized eggs do not form membranes as do fertilized, so that growth tended to be formless, and even when assuming definite form in the blastula^ stage there were often formless masses of dividing matter lying to one side. Briefly stated, the following facts developed during the prog- ress of this systematic experiment : 1. In a solution of sodium chlorid the eggs were unable to reach even the blastula stage. 2. With the addition of MgClg, however, blastulas were formed, but they did not move. When afterward placed in normal sea water movement soon appeared. 3. With three chlorids (Na, K, and Ca) "the eggs not only reached the blastula stage and swam around in the most lively way, but they reached the gastrula and even the pluteus stage, with the exception, however, that practically no skeleton was formed." - Such larvae lived about ten days. 4. The addition of a trace of Na2C03 resulted in the formation of a skeleton, but it was not quite normal. It was made normal by adding a trace of MgClg. 1 Three early stages are characteristic of the early development of all embryos : (i) the morula, or " mulberry " stage, in which cell division gives rise to a globular mass of rounded cells, each more or less distinct, like the grapes on a bunch or the seeds of a mulberry; (2) the blastula stage, in which the outer cells become condensed, showing a distinct outer layer, — the blastoderm ; and (3) the gastrula stage, in which one side becomes pushed in (invaginated), as one would push in a hollow rubber ball with his thumb, forming a kind of mouth and stomach. A few forms never get beyond this stage, but most pass quickly through it, differentia- tion proceeding rapidly. In higher animals the outer layer (ectoderm) gives rise to the skin and its appendages, the inner (endoderm) to the internal organs. Among sea urchins, which were here under experiment, the next stage is known as Xhe pluteus, — the stage of free-moving larvae. It was this stage the experi- menter desired to produce. 2 Loeb, Studies in General Physiology, Part II, pp. 585-5S6. 28o CAUSES OF VARIATION 5. All experiments indicated that it is impossible to secure more than the beginning of segmentation from an unfertilized egg without raising the co7icentration of the sea water. 6. But for this purpose MgCl2 was peculiarly effective and normal plutei (free-swimming larvae) developed from unfertilized eggs lying in normal sea water after having lain for two hours in a solution of MgCl2 of proper strength. ^ 7. These effects seemed to be due to the increased concen- tration in the sea water, bringing about increased osmotic pres- sure and resulting in a loss of water on the part of the egg. This loss of water seems to be the active cause of rapid seg- mentation, and a variety of substances were discovered which were able to bring it about. 8. The principal difference noticeable in the plutei was that those developed from fertilized eggs swam freely at the top of the water, while those developed from unfertilized eggs " were all at the bottom of the dish and unable to rise." 9. Experiments upon the marine annelid Chaetopterus^ indi- cated that artificial development is easier than with the sea urchin, but that it is achieved by a different solution. In the words of the experimenter, " We may say that Chsetopterus possesses a higher degree of parthenogenetic tendency than the Arbacia [sea urchin] eggs," ^ and "if the sea water contained only a slightly greater proportion of K, we should find that Chaetopterus was normally parthenogenetic." * 10. If certain forms are prevented from becoming partheno- genetic by the constitution of the sea water, we may infer that those which are naturally parthenogenetic are so by the consti- tution of the blood or the sea water enabling the egg to develop.^ 11. "The bridge between the phenomena of natural and artificial parthenogenesis is formed by those animals in which 1 Loeb, Studies in General Physiology, Part II, p. 624. 2 " Mead had already found that if 0.5 per cent KCl is added to sea water the unfertilized eggs of Chaetopterus throw out their polar bodies, while the addition of 0.5 per cent NaCl produced no such effect." — Loeb, Studies in General Physiology, Part II, pp. 656-657. ^ Loeb, Studies in General Physiology, Part II, pp. 654-655. 4 Ibid. p. 665. 5 Ibid. p. 683. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 28 1 physical factors decide whether or not their eggs develop par- thenogenetically." ^ The consideration seems to be largely one of change in osmotic pressure, some organisms requiring increase and some decrease. Plant lice are parthenogenetic only at high temperatures and when the host plant has plenty of water. " If we lower the temperature or let the plant dry out, sexual repro- duction occurs." ^ It seems to be decided that Artcinia salina living in brackish waters is parthenogenetic, while its nearest fresh-water relative, Branchipus, is not.^ 12. From the fact that the beginning of segmentation is com- mon in unfertilized and untreated eggs of many forms, it seems to follow that the effect of fertilization or of treatment is largely to accelerate a process which is able to begin alone but which proceeds so slowly as to be overtaken by destructive processes and the death of the Q.^g before an embryo can form. 13. The introduction of a small amount of a catalytic sub- stance at the critically proper time (at maturity) seems in most cases necessary to a cell division sufficiently rapid to insure the continuation of life. 14. The function of the spermatozoon would seem therefore to be twofold, — first, to introduce such a catalytic substance, and second, to convey hereditary material. No student can consider these fundamental matters and fail to realize . the profound effect of external influences upon in- ternal activities, nor can he avoid the conclusion that we must revise our ideas as to the relation even of inorganic chemistry and physical forces to the processes of life. Much that we have considered as morphological and peculiarly vital is, after all, evidently due to the operations of ordinary chemical and physical laws. This does not make the facts of variability less significant, but it does show the extent to which living organisms have become accustomed to their ordinary sur- roundings. ^ Loeb, Studies in General Physiology, Part II, p. 683. 2 " Janosik has found segmentation in the unfertilized eggs of mammalians." — Loeb, Studies in General Physiology, Part II, p. 543. Loeb expresses the con- viction that possibly " only the ions of the blood prevent the parthenogenetic origin of embryos in mammalians," and that a change in their blood might be followed by parthenogenetic development. 282 CAUSES OF VARIATION SECTION IX — EFFECT OF SALINE SOLUTION UPON DEVELOPMENT IN AQUATIC ANIMALS Sea water differs from fresh water in two particulars, salinity and density, both of which exert marked influence upon animal life and between which it is often difficult to discriminate. A goldfish plunged into sea water at first shows " violent incoor- dinated movements," but shortly " becomes immobile and rises to the surface by virtue of its lower specific gravity." ^ " The effect of fresh water upon marine organisms is equally striking. They go immediately to the bottom and move with difficulty. Swimming animals swim badly if at all, and small fishes have to make much exertion to rise to the surface." ^ On many marine animals, as mollusks and fish, fresh water acts as an anaesthetic, the mollusks soon yielding to paralysis, the fish appearing to suffer from lack of air. "The respiratory movements become deep and rapid. . . . The tissues become swollen so that soft-bodied animals are visibly deformed, — in fishes the eyes are forced out, the foot of gastropods swells, the blood corpuscles swell up and burst, and muscular tissue may increase as much as six times in volume." ^ Many of these effects are clearly due to differences in pres- sure which may amount to many atmospheres, but it remains to separate, so far as may be, the effects of salinity from those of specific gravity. Rather startling claims have been made from time to time as to the conversion of one species into another by altering the degree of salinity. Further investigation seems to show that in all such cases intermediate forms are known to occur, which argues that the two forms which had been recognized as dis- tinct were not both good species ; that is to say, it is a case of one species with wide variability as to certain characters, not that of two distinct and well-defined species. If, however, differences in salinity are effective in bringing about alterations in even a single character, the fact is of interest here, no matter what specific lines should be drawn by the biologist. 1 C. B. Davenport, Experimental Morphology, Part I, p. 79. 2 Ibid. pp. 79, 80. EXTERNAL INFLUENCES AS CAUSES OF VARLVITON 283 The small crustaceans Artcmia salina and A. viilhausaiii have been recognized as distinct, the former living in brackish and the latter in still more concentrated waters, the two differ- ing mainly in the number and length of bristles borne at the extremity of the caudal fins. Early in the seventies Schmankewitsch published an account of the mutual conversion of each form into the other, but the facts as given by him have been greatly overstated, as frequently happens in repetition. They are sufficiently significant as first reported, and it seems well to give the original statement as recorded by Bateson,^ which is as follows : The salt lagoon, Kuyalnik, was divided by a dam into an upper and a lower part ; the waters in the latter being saturated with salt, while the waters of the upper part were less salt. By a spring flood in the year 1871 the waters of the upper part of the lake swept over the dam and reduced the density of the lower waters to 8° Baume (= about sp. g. 1.051), and in this water great numbers of A. salina then appeared, presumably having been washed in from the upper part of the lake or from the neighboring salt pools. After this the dam was made good and the waters of the lower lake, by evaporation, became more and more concentrated, being, in the summer of 1872, 14° B (about sp. g. 1.103); in 1873, 18° B (about sp. g. 1.135) ; in August, 1874, 23.5° B (about sp. g. 1.177), and later in that year the salt began to crystallize out. In 187 1 the Artemia [as first carried over] had caudal fins of good size, bearing eight to twelve, rarely fifteen bristles, but with the progressive concentration of the water the genera- tions of Artemia progressively degenerated, until at the end of the summer of 1874 a large part of them had no caudal fins, thus presenting the character of A. milharisenii. — Fischer and Milxe-Edwards. Bateson adds : A similar series was produced experimentally by gradual concentration of water, leading to the extreme form resembling A. iiiilhauseiiii. It was found also that if the animals without caudal fins were kept in water which was gradually diluted, after some weeks a pair of conical prominences, each bearing a single bristle, appeared at the end of the abdomen. The experimenter also relates that by breeding salina in still more diluted water he attained a form resembling Schaffer's genus Branchipus. But the principal difference between the genera is that in Artemia the last segment is about twice as 1 Bateson, Materials, etc., p. 96. 284 CAUSES OF VARIATION long as each of the others, while in Branchipus it is divided. It is extremely significant that this division should be produced in Artemia by culture in comparatively fresh water, but the fact is no warrant for the assertion that one genus can be converted into another by altering the environment. It rather casts doubt upon the wisdom of a classification which establishes generic dis- tinctions upon differences so slight and so easily brought about. The same experimenter studied species of the genus Daphnia, and found " in their case also considerable structural and physi- ological changes, the fresh- and salt-water forms differing, in his opinion, by characters usually held to be specific." ^ Bateson studied the common cockle, a mollusk, everywhere present in the Aral Sea and its outlying waters of different degrees of salinity. One of the lakes (Shumish Kul) on its western shore exhibited no less than seven distinct terraces, held to represent successive stages of the water levels during its long period of drying up, with corresponding increase in salinity. The most noticeable differences in the shells taken from these successive terraces, and presumably due to increas- ing salinity, are outlined as follows : ^ 1. A diminution in the tJiickness of the shells, first apparent in the third terrace. In the seventh terrace this change was so marked that the shells were almost horny, and their weight was not a third of that of the shells from the first two terraces. 2. Diminution of the size of the beak [with the lowering of the level]. 3. High coloration. [The author does not state which way the changes ran, whether up or down the terraces, but he re- marks that all the shells of a given terrace were " very nearly alike in texture, thickness, and degree of coloration." ] 4. Grooves between the ribs appearing on the inside of the shell as ridges with rectangular faces. 5. A great diminution in absolute size of the shells on the lowest terrace. 6. Alteration in proportion of length to breadth, ranging from I to 0.80 in the shells of the first terrace to i to 0.725 in those of the seventh and i to 0.66 on the shores of neighboring 1 Vernon, Variation in Animals and Plants, p. 275. - Ibid. pp. 275, 276. EXTERNAL INFLUENCES AS CAUSES OF VARIATION 285 lakes. In view of these facts Bateson remarks, as quoted by Vernon/ " It seems almost certain that these conditions are in some way the cause of the variations." Biological literature is full of similar examples of the char- acteristic effects of varying degrees of salinity. The limits of space forbid the further pursuit of a subject which might be extended almost indefinitely. It may be sufficient to say that the specific influence of salinity upon certain characters is, beyond a doubt, well established. SECTION X — INFLUENCE OF USE AND DISUSE UPON DEVELOPMENT No fact is better or more generally known than that use stim- ulates and disuse dwarfs the development of many organs. To say that development is in proportion to use would doubtless be true, roughly speaking, of certain parts, as the muscular system, secreting glands, etc. It certainly would not be true of many others, as hair, feathers, bony skeleton, etc., which develop independently of use, and some of which, as hair and feathers, involve no activity in the sense in which the term is here understood. This discussion should be limited to the distinctively active parts, and the influence of exercise or the lack of exercise upon their development. Of these parts it may fairly be said that perfect development is dependent upon, if not proportional to, the degree of their use, especially during the earlier stages of development. The classic illustration from Darwin, showing the leg bones of the tame duck and the wing bones of the wild duck to be relatively heavier ; the arm of the artisan and the body of the athlete ; the training of the track horse ; the marvelous coordi- nation of complicated nervous impulse and muscular response in the violinist and the pianist, — all these and a multitude of similar facts teach clearly that individual development of usable parts depends very much upon their early and continuous exercise. ^ Vernon, Variation in Animals and Plants, p. 277. 286 CAUSES OF VARIATION Upper limits of development. In what sense is development conditioned upon use ? Docs use simply enable the part to attain its normal z.Vi^ proper development, to which it is entitled under the laws of heredity, or does it stimulate development beyond the normal f Some biologists at once assume the latter to be impossible and that any unusual appearance is a case of atavism. It is true that in times long past there may have existed ducks that walked and others that flew more and better than those which Darwin examined ; but when did nature produce a running or a trotting horse as good as the one of to-day ? To what remote ancestor do our violinists and our pianists owe their skill, and what was the instrument on which they acquired it ? The accompanying cut is a facsimile of a properly attested let- ter written with \\\q feet by a young woman twenty-three years old who lost both arms at the age of ten. Among her other ac- plishments she numbers cutting, sewing (threading her own needle), drawing, sweeping, and a great variety of housework. ^ Here is a case of putting parts to an entirely new ?ise, demanding a nicety of adjustment that was never acquired even in one out of a million of the ancestors. Could there be better evidence of the fact that few individuals ever use, and there- fore few ever develop, more than a fraction of the capacities born in them ; that the possibilities of life are seldom realized, and that never are all faculties developed to their utmost in any single individual ? All this is clear but it is not so easy to determine where to draw a line and say, " All development below this is due to inheritance and all above to use." The truth would seem to be that development depends upon both inherent tendencies and external conditions affording opportunities for their exercise, and that the maximum of development is reached only ivJien both are at their optimum. There is much force in the word "optimum." Too much exercise, too much food, too much temperature, or too much of any of the conditions of life is as unfortunate as too little. ^ The author saw one man who wrote with his feet, but they were attached directly to the body, with no legs whatever. EXTERNAL liNFLUENCES AS CAUSES OF VARIATION 287 /i/>~ \^''/u 1^:^-!^ Mo-Y ^/fe/ i\./' .,v^.' -f-f ,i>.-i,-sro ;5t<,ii-t-,', s;. •;f/i5/ <«r ) t;^! n "^^feaj-n on her i, ■\ "Jotarv =i. K ; - ^^- ^fe-^S^-^Sf";,"':^"^ * \ - •. fr^....^. '