Materials for the Study of Variation Treated with Especial Regard to Discontinuity in the Origin of Species

saj that among these it now and then gives off one which approaches the first normal, that shooting at the new mark it n..\\ and th.-n hits the old one. But all that we know is that now and then it shoots wide and hits another mark, and we assume from this that it would n<»t have hit it if it had not aimed at it in a bvgone age. To apply this to any other matter \\onld be ab-urd. '\\V might as well say that a bubble would 1..- r-niiid if the air in it had not learned the trick of roundnesfi bv having been in a bubble before: that if in a bag after pulling '"it a lot of white balls I find a totally red one, this pn. M-.S that the bag must have once been full of red balls, or that tin- \\hite ones must all have been red in the past.

lie-id.--, tli.- logical absurdity on which this use of the theory ot' Reversion rests, the application of it to the facts of Variation bp-ak- d'.un again and again. I have already mentioned some cases of this, but there are many others of a different class. For in-tan.-.-, it will be shewn that the percentage of extra molars in the Anthropoid A p.-- is almost the highest reached among mammals. On the usual interpretation, such teeth are due to Ke\. -i-sioii to an ancestral condition with 4 molars, and on less evidence it habeen argued that a form frequently shewing such |;< \ersion " is older than those which do not. From this reasonin:; it ,-hould follow that the Anthropoids are the most primitive form, at lea-t ..f monkeys. It is surely time that these brilliant and facile deductionwere no more made in the name of science.

[nquiry into the cau-es ,,f Variation is as yet, in my judgment, premature. The often-repeated -latemeiit that " useless " parts are .-pecially \ariable. finds little support in the tacts of Variation, • pt in a- far a- it i- a misrepresent at ion of another principle. Th. examples taken to supportthis statement are commonly m- -landing at the end of a Meristic Series of parts, in which there is a progression or increase of size and degree of development, starting from a small terminal member. In such cases, as that of the last rib in Man, and several other animals, the wisdom-teeth of Man, etc., it is quite true that in the terminal member Variation is more noticeable than it is in the other members. This is, I believe, a consequence of the mechanics of Division, and has no connexion with the fact that the functions of such terminal parts are often trifling. Upon this subject something will be said later on, but perhaps a rough illustration may make the meaning more clear at this stage. If a spindle-shaped loaf of bread, such as a "twist," be divided with three cuts taken at equal distances, in such a way that the two end pieces are much shorter than the middle ones, to a child who gets one of the two large middle pieces the contour-curves of the loaf will not matter so much ; but to a child who gets one of the small end bits, a very slight alteration in the curves of the loaf will make the difference between a fair-sized bit and almost nothing, a difference which the child will perceive much more readily than the complementary difference in the large pieces will be seen by the others. An error in some measure comparable with this is probably at the bottom of the statement that useless parts are variable, but of course there are many examples, as the pinna of the human ear, which are of a different nature. It is unnecessary to say that for any such case in which a part, apparently useless, is variable, another can be produced in which some capital organ is also variable ; and conversely, that for any case of a capital organ which is little subject to Variation can be produced a case of an organ, which though trifling and seemingly " useless," is equally constant.

With a knowledge of the facts of Variation, all these trite generalities will be forgotten. In examining cases of Variation, I have not thought it necessary to speculate on the usefulness or harmfulness of the variations described. For reasons given in Section II, such speculation, whether applied to normal structures, or to Variation, is barren and profitless. If any one is curious on these questions of Adaptation, he may easily thus exercise his imagination. In any case of Variation there are a hundred ways in which it may be beneficial, or detrimental. For instance, if the " hairy " variety of the moorhen became established on an island, as many strange varieties have been, I do not doubt that ingenious persons would invite us to see how the hairiness fitted the bird in some special way for life in that island in particular. Their contention would be hard to deny, for on this class of speculation the only limitations are those of the ingenuity of the author. While the only test of utility is the success of the organism, even this does not indicate the utility

In the view of the phenomena of Variation here outlined, thenis nothing which is in any way opposed to the theory of the on-m .it' Species " l>v means of Natural Selection, or the preservaii' MI <if favoured races in the struggle for life." But by a full and unwavering belief in the doctrine as originally expressed, we shall in no way be committed to representations < if that doctrine made li\ those who have come after. A very biief study of the facts will Mith'cu to gainsay such statements as, for example, that of Glaus, that "it is only natural selection which accumulates those alteration.*, .*•> that they become appreciable to us and constitute a variatioii \vhi--h is evident to our senses1." For the crude belief that living In ings are plastic conglomerates of miscellaneous attributes, and that order of form or Symmetry have been impressed upon this medley by Selection alone ; and that by Variation any of these at tributes may be subtracted or any other attribute added in indefinite proportion, is a fancy which the Study of Variation does not support.

II re t hi- [introduction must end. As a sketch of a part of the phenomena of Variation, it has no value except in so far as it may I. 'ail some to study those phenomena. That the study of Variation i-< tinpi-op'-r field for the development of biology there can be no doubt. It i- -i-an-ely too much to say that the study of Variation I" ars to the M'lence of Evolution a relation somewhat comparable with thai \\hidi the study of affinities and reactions bears to the science of <-henii-tr\ : tm- \\e mi;_dil almost as well seek for the "ii-in of chemical bodies by the comparative study of crystallo-

I'liy. a- tor i he origin of living bodies by a comparative study of normal f'.irm-. THE cases of Meristic Variation, here given, illustrate only a small part of the subject. The principles upon which these have been chosen may be briefly explained. It was originally intended to give samples of the evidence relating to as many different parts of the subject as possible, so that the ground to be eventually covered might be mapped out, leaving the separate sections of evidence to be amplified as observations accumulate. This plan would be the most logical and perhaps in the end the most useful, but for several reasons it has been abandoned. I have chosen a different course, first, because during the progress of the work opportunities occurred for developing special parts of the evidence; second^, since isolated observations have no interest for most persons, it is more likely that the importance of the subject will be appreciated in a fuller treatment of special sections, than in a general view of the whole ; and lastly, because as yet the attempt to make an orderly or logical classification of the phenomena of Merism, however attractive, must be so imperfect as to be almost worthless. For these reasons I have decided to treat more fully a few sections of the facts, hoping that in the course of time similar treatment may be applied to other sections also. The sections have been chosen either because there is a fairly large body of evidence relating to them, or on account of the importance or novelty of the principles illustrated.

As far as possible I have described each case separately, in terms applicable specially to it, deductions or criticism being kept apart. The descriptions are written as if for an imaginary catalogue of a Museum in which the objects might be displayed1. This system, though it entails repetition, has, I believe, advantages which cannot be attained when the descriptions are given in a comprehensive and continuous form. In speaking of subjects, such as supernumerary mammae, or cervical fistula?, where the evidence has been exhaustively treated by others, and upon which I can add nothing, it has not seemed necessary to follow this system, and in such cases connected abstracts are given.

As the evidence here presented consists, as yet, only of specimen chapters in the Natural History of Meristic Variation, and does not offer any comprehensive view of the whole subject, no --itiration of the facts is attempted. The evidence of :c Variation relates essentially to the manner in which changes occur in the number of members in Meristic series. Such numerical changes may come about in two ways, which are in some respects distinct from each other. For instance, the number of legs .m. I body-segments in Peripatns edwardsii varies from 29 to 341: In !•• the variation in number must be a manifestation of an original difference in the manner of division or segmentation in tinprogress of development. The change is strictly Meristic or di\ isional. On the other hand, change in number may arise by tli.- Substantive Variation of members of a Meristic series already c. -n-tituted. For example, the evidence will shew that the number of oviducal openings in Astacus may be increased from one I ..ii i to two or even three pairs. Here the numerical variation has come :il..iiit through the assumption by the penultimate and last th.-racic appendages, of a character typically proper to the append-

- .if the antepenultimate segment of the thorax alone. \ow then- i-, h.-re no change in the number of segments composing the Meristic series, but by Substantive Variation the number of op oings lias been increased. The case of the modification of the antenna of an insect into a to..!. - if the eye of a Crustacean into an antenna, of a petal into a stamen, and the like, are examples of the same kind. It is desirable and indeed necessary that such Variations, \vhi.-h con-i-i in the assumption by one member of a Meristic aeries, of th,- form or characters proper to other members of the - ries, should be recognized as constituting a distinct group of phenomena. In the case of plants such Variation is very common and is one "f the ino-t familiar forms of abnormality. MASTERS, in his t r.-at is.- mi Ve^-.-iahie Teratology-, recognizes this phenomenon and gives to ii th<- name •• .M etam< >r| ih v," adopting the word from i '"•• ' D< A- Masters says, so long as it is only proposed to use the \\oid m Teratology, no great coiifu.sioii need arise from the fact' that the s.-,ine term and its deri\afi\es are used in a different Jense in several branches of Natural History. Hut if, as I hope, 'he time has come \\lien the facts of \\ hat has been called " Tera- \\ill be admitted to their proper place in the Study of Variation, this contusion is ine\itable. In this study, besides, this I'.-iitn-iilar kind of \ariation will be found to be especially impor- l;"it -'"id 1 believe that in the future its significance and the mode of its occurrence \\i|] become an object of high interest. For this reason it is desirable that the term which denotes it should not

lead to misunderstanding, and I think a new term is demanded. For the word 'Metamorphy' I therefore propose to substitute tinterm Homoeosis, which is also more correct ; for the essential phenomenon is not that there has merely been a change, but that something has been changed into the likeness of something else. In the cases given above, the distinction between Homoeotic Variation and strictly Meristic Variation is sufficiently obvious, but many numerical changes occur which cannot be referred with certainty to the one class rather than to the other. Such cases are for the most part seen in Vertebrates : for in them what may be called the fundamental numbers of the segments are not constituted with the definiteness found in Arthropods or in the Annelids, and several Meristic series of organs are disposed in numbers and positions independent of, or at least having no obvious relation to those of the other Meristic series. The number and positions of mammae, or stripes, for instance, need not bear any visible relation to the segmentation of the vertebrae &c. The repetition of members of such a series may thus not coincide with, or occur in multiples of the segmentation of other parts in the same region. When such is the case, when the segmentation of one series of organs bears no simple or constant geometrical relation to the segmentation of other systems, it is not always possible to declare whether a numerical change in one of the systems of organs belongs properly to the first or the second of the classes described above. It is likely enough that in such a case as that of mamma?, there may sometimes be an actual Meristic division and subsequent separation of the tissues already destined to form the mammae, occurring in such a way that each comes to take up its final position, and indeed the numerous cases in which such division has been imperfectly effected go far to prove that this is the case.

But, on the other hand, it is not possible to know that the division did not occur before any tissue was specially differentiated off to form mammas, and that the separation may be as old even as the division of the mammae of the right side from those of the left, a process which almost beyond question occurs in the segmentation of the ovum. The distinction between these two alternatives is thus one rather of degree than of kind, and it is only in such forms as the Arthropods, the floral organs of some Phanerogams and the like, where the members of the several Meristic series have definite numbers, or coincide with each other, that this distinction is easily recognized. For this reason I do not think it well to attempt to carry out any classification of the evidence based on this distinction.

In the foregoing remarks I am aware that a very large question, which lies at the root of all accurate study of Meristic Variation, has been passed over somewhat superficially, but I scarcely think a fuller treatment possible in the present state of knowledge of the physics of Division, and in the absence of thorough observation of the developmental history of those tissues which ultimately become differentiated to form members of such non-coincident or iii-1. -p. -inli-iit Meristic series.

S. ,m.- \.-irsago1, in the course of an argument that Balanoglossus -In. ul. I b.-'eon-id.-red as representing some of tlie ancestral characters • •I ' • ..rd.ita, I had occasion to refer to some of these difficulties, and •••iallv ti, tindim-rent characters of the two kinds of segmentation : tli.it i-f t'he Annelids, in which the repetitions of the organs belonging t.. tli.- -.\eral systems are coincident, and, on the other hand, that the Chordata, for example, in whicli this coincidence may be

irregular or partial. At tliat time I was of opinion that these t\s<. plr. ic history, and have resulted from processes essentially Amu-lids on the one hand, segmentation of the various systems of organs had been coincident from the beginning, while in the Chordata tin- -egmentation had been progressive and had arisen by segmentation <>r repetition of the organs of the several systems independently. The il.lt- to arrange the lower Chordata in order of progressive segmentation ,,f the several systems. In particular such treatment was shewn to }>•• applicable to the central nervous system, the vertebral column and tiniiii->ol)lastic somites, and in these cases it was maintained that tin- e\ i.|. -nee of the lower forms of Chordata goes to shew that segmentati.m had occurred in these systems one after another, and that their

•Mentation was not derived from a form having a complete repetition H-li part in each segment: that these forms, in fact, shewed us the history . .f this progress from a less segmented form to one more fully segmented. Tin- \i.-us tlien s.-t forth have met with little acceptance. Those \\lni anoccupied with the search for the pedigree of Vertebrates still direct their in<niines on the hypothesis, expressed or implied, that in the am-e-tral fnnn there was a series of complete segments, each containing a r.-pre-entat i\ .- of each system of those organs which in tinpresent descendants appear in series. It is thus supposed that each

'"•nt ..f tinpriiuiti\e form must have liecn a kind of least common denominator of the segments of its posterity. The possibility that the seg nt.ition .,f V. -i-t. -lii-atc -s may have arisen progressively is, indeed, Though in tinlight of the study of Variation, it now seems to ""' ''"at the discussion of these questions must be indefinitely postponed, and that there are radical objections to any attempt to interpret the ta.-ts of anatomy and development in our present ignorance of Variation, I ha\.- seen no reason to depart from the view expressed '" ''"• paper rcferr.-d to: that interpreted by the current methods of morphological critieism, the facts go to shew that the segmentation of ''"• Choi-data dill'ers essentially from that of the Annelids tvc., and that it has arisen by progressive segmentation of the several systems of :m originally unsegm.-nted form. To those who hold as Dohrn, Gaskell, N1 irshall and oth.-rs ha\.- done, that the evolution of Vertebrates has

been a progress from a more fully segmented form to forms less segmented, I would again point out that this view is in direct opposition to the indications afforded by the lower Chordata, which are less and not more segmented than the higher forms. The hypothesis of an ancestor made up of complete segments is resorted to because it is felt to be difficult to conceive the progressive building up of a segmented form, but on appeal to the facts of Variation the evidence will clearly shew that Repetition of parts previously existing is a quite common phenomenon ; that such repetition may occur in almost any system of organs; and lastly that such new repetitions may be coincident in the several systems. To argue moreover that these repetitions, for instance that of oviducal apertures in Astacus, of niammje or cervical ribs in mammals are "reversions," leads to absurdity, for on the same reasoning, the occurrence, in the Crab, of a third maxillipede formed as a chela, would shew that these appendages had been originally chela?, that the occurrence of petaloid sepals shews that the sepals had originally been petals, and so forth.

These considerations will suffice to illustrate the great difference of degree, if not of kind, which probably exists between these two kinds of segmentation, that which arises by the repetition of budlike segments, each containing parts of many systems on the one hand, and the progressive and separate segmentation of the several systems on the other. For reasons already given, however, I shall not attempt in this first collection of evidence to separate the facts on these lines. Though some cases can at once be seen to be strictly Meristic while others are plainly Homceotic, many cannot be affirmed to belong to the one group rather than to the other. There is, besides, a serious doubt whether perhaps after all, Homoeotic Variation even in its most marked forms, may not ultimately rest on and be an expression of a change in the processes of Division, and be thus, at bottom, strictly Meristic also. In our present ignorance of the physics of Division, this doubt cannot be satisfied, and therefore it will be best to make no definite separation between the two classes of variations, though whenever the nature of a given variation is such that it may at once be recognised as Homceotic, it will be well to specify this.

In the absence of a more natural classification, the material has been roughly arranged with reference to the geometrical disposition and relations of the structures concerned. In the Introduction, Section IV. p. 21, reference was made to the fact that the Symmetry of an organism may be such as to include all the parts into one system of Symmetry, and for such a system the term Major Symmetry was proposed. Systems of this kind are seen in the Vertebrates and Echinoderms, for example. On the other hand systems of Symmetry occur in limbs and other separate parts of organisms, in such a way that each such system is either altogether or partially geometrically complete and symmetrical in itself. For example, the toe of a Horse, the arm of a Starfish, the

eye-spots of some Satyrid butterflies, &c., are each in themselves nearly symmetrical. To these separate systems of Symmetry the i, TIM 'Minor Symmetry will be applied. Minor Symmetries may or may ii"t be compounded into a Major Symmetry. Between tli. — - there is of course no hard and fast line. In each .-lass of Symmetry, Meristic Repetition may occur, and tiprepeated parts then stand in either ts meristically repeated may thus stand in one or more ,1 ivlatkms to each other, and the first part of the

lence »t M.-ristir Variation will be arranged in groups according • i- in on,- or other of these relations that the parts are affected. In each group cases affecting Major Symmetry will be given first, an. I those atV.-eting Minor Symmetries will be taken after. A- i; i- proposed to arrange the facts of Meristic Variation in ipe corresponding with these three forms of Meristic Repetition, ill he ii.-.-t'ul to consider briefly the nature of the relation in

which the members of such series stand to each other, and the .•liaia.-t.-r- distinguishing the several kinds of series. Reduced to th.- -impl- -t terms, the distinction may be thus expressed. In tin Liii'-ar or Successive series the adjacent parts of tint/ (,!>.. • nn-ii tbers of the series are not homologous, but the /•-///// liniimlniiniis parts of each member or segment form a .iltfi-iKitinff with each other. For example, the anti-ri'T and posterior -ur faces of such a series of segments may he n-pr. -. -nt.-d 1'V th.- .-.TICS

ujttii-nl mi'i'K "f t/ir "flier. Thtrxtt-rnal and internal surfaces of >in-h a pair may tht.-ivti>iv In- i-rjirrM-nti-d thus: If tinmaniit-r «•!' origin of tlu-s<- two kinds of Repetition be considered, il \\ill !»<• seen that though both result from a process • •I' I >i\ ision, yet the niann.-r of l>i\ision in the two cases is verv ditVi-n-nt. F..I- in tincase of division to form a paired st nu-t lire, ot uhidi similar and homologous parts ]j,. ,,n raeh side of the

plane of division ; \\hile.in the toi-mation ..('a chain of successive ments, each |>lane «.f divi-ion pa^-es li.-tween parts which are dis-iiiiilar, and \\h«>>e !ioinol«.gy is alternate. The distinction 11 the-e tw.i kinds of l)i\ivion is of course an expression of the ta.-t that the at t raet ioi,s and repnUion> from which Division results are differently disposed in the two cases. It is further to be observed that the distinction, though striking, is nevertheless one of degree, for the two kinds of Division pass gradually into each other. By one or other of these two modes, or by a combination of both, all Meristic Series of Repetitions are formed.

In Radial series, the Major Symmetry is built up by radial divisions of the first kind, producing segments whose adjacent parts are homologous, and related to each other as images. Each of these segments is therefore bilaterally symmetrical about a radial plane. There is no succession between the segments, and in a perfectly symmetrical series, Successive or Linear repetitions can only occur in Minor Systems of Symmetry. The considerations here set forth, though well known, have an importance in the interpretation of the evidence, for the connexion between the geometrical relations of organs and their Meristic Variations is intimate.

An arrangement of the facts with reference to these geometrical relations cannot, of course, be absolute, for it is clear that a Bilateral Symmetry, containing Linear Repetitions may be derived from a Radial Symmetry, and that these figures cannot be precisely delimited from each other ; nevertheless this plan of arrangement has still several advantages. Chief among these is this : that it brings out and emphasizes the fact that the possible, or at least the probable Meristic Variations of such parts depend closely on the geometrical relation in which they stand. This is, perhaps, in a word, the first great deduction from the facts of Meristic Variation. The capacity for, and manner of Meristic Variation appear to depend not on the physiological nature of the part, on the system to which it belongs, on the habits of the organism, on the needs or exigencies of its life, but on this fact of the geometrical position of the parts concerned. Linear series are liable to certain sorts of Variation, Bilateral Series are liable to other sorts of Variation, and Radial Series to others again. As I have ventured to hint before, the importance of all this lies in the glimpse which is thus afforded us of the essential nature of Meristic Division and Repetition. Such interdependence between the geometrical relations, or pattern, in which a part stands, and the kinds of Variation of which it is capable, is, I think, a strong indication that in Meristic Division we are dealing with a phenomenon which in its essential nature is mechanical. Since this is a thing of the highest importance, it will be useful to employ a system which shall give it full expression.

Evidence as to Meristic Variation in cell-division and in the segmentation of ova will be spoken of in connexion with the Variation of Radial and Bilateral series. The second section of evidence is less immediately relevant to the problem of Species; nevertheless it bears so closely on the nature of Merism and on the mechanics of Physiological Division, th.it in any study of this subject reference to it cannot be omitted. 'I'll-- t-vidt-nce in question relates first to abnormal repetition of liinl>- MI- other peripheral structures, (which in the normal form are

ipi-d into and form part of a system of Symmetry,) such abnormal n-pftitions occurring in such a way as to lie outside this /"./•//oil xi/sti-in of Symmetry and unbalanced by any parts within 1 phenomenon occurs in many forms, especially in bilateral animals, and may be exceptionally well studied in the case of -npernunirrary limbs in Insects and in supernumerary chel* in < 'iabs and Lobsters. It will be shewn that such extra parts ••rally, it' not always, make up a Secondary system of Symmetry in themselves ; and the way in which such a

• Hilary system is related to the normal or Primary system of Symmetry of the body from which they spring, constitutes an iiiMrnrthv chapter in the study of Meristic Variation. M"H- .-.\ tensive repetitions of this class, when affecting the axial parts "t the body, give rise to the well-known Double and Tri|>k Mi.n-ti-rs. which, as has often been said, reproduce in the higher animals phenomena which, under the name of fission, are commonly seen in th<- lower forms. The general evidence as to tln-M- abnormalities is so accessible and familiar that it need