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

The importance of the phenomena of Merism to the Study of Variation is thus, in the first instance, a direct one, for the Variations which have resulted in the production of Meristic Systems are a direct factor in Evolution. In addition to this direct relation to the Study of Variation, the phenomena of Merism have also an indirect relation, which is scarcely less important ; for they are a factor in the estimation of the magnitude of the integral steps by which Variation proceeds. This will be more evident after the second group of Variations has been spoken of.

We have thus far spoken only of the processes by which the living body is divided into parts, and we have thus constituted a group which is to include Variations in number, Division, and geometrical position. From these phenomena of Division may be distinguished Variations in the actual constitution or substance of the parts themselves. To these Variations the name Substantive will be given. Under this head several phenomena may be temporarily grouped together, which with further knowledge will doubtless be found to have no real connexion with each other. For the present, however, it will be convenient to constitute such a temporary group in order to bring out the relative distinctness of Variations which are Meristic.

These two classes of Variation, Meristic and Substantive, may be recognized at the outset of the study. There can be no doubt that they are essentially distinct from each other, and the proof that they are thus distinct will be found in the evidence of Variation, for it will be seen that either may occur independently of the other. An appreciation -of this distinction is a first step towards a comprehension of the processes by which the bodies of organisms are evolved.

A few simple illustrations may make the nature of these two classes of Variations more clear. For example, then, the fiower of a Narcissus is commonly divided into six parts, but through Meristic Variation it may be divided into seven parts or into only four. Nevertheless there is in such a case no perceptible change in the tissues or substance of which the parts are made up. All belong to and are recognizable us belonging to the same -<>rt of .Y'//r/.s-\//\. ( )n the other hand manv Narcissi, \. corbularia^ t'»r example, are known in two colours, one a dark yellow ami the other ;i sulphur-yellow, though the number of parts and pattern of the flowers are identical. This is, therefore, an example of a Substantive Variation.

Again, the foot of a Pig may, through Meristic Variation, be divided into five or six toes instead of into four ; or, on the other hand, the number may, by absence of the median division between the digits III and IV, be reduced to three, though the tissues composing the toes may not in structure differ from the normal. Again, the tarsus of a Cockroach (Blatta) may, through Me- ristic Variation, be divided into only four joints instead of into five, the normal number, but the joints are still in substance or quality those of a Cockroach.

I am aware that Meristic and Substantive Variations often occur together, and that there is a point at which it is not possible to separate satisfactorily the changes which have come about by the one process from those which have come about by the other. Instances of this kind occur especially in the case of series of parts such as Teeth or Vertebra?, in which individual members or groups of members of the series arc differentiated from the others. For example, we may see that it is through Meristic Variation that the vertebral column of a Dog may be divided into a number of Vertebra? greater or less than the normal ; and though in such cases all the Vertebra? have distinctively canine characters, yet there are nearly always Substantive Variations occurring in correlation with the Meristic Variations, manifesting themselves in a re-arrangement of the points of division between the several groups of Vertebrae, and causing individual Vertebrae to assume characters which are not proper to their ordinal positions.

Further inquiry into the questions thus raised cannot at this stage be profitably undertaken, though when the evidence has been considered it will perhaps be advisable to recur to them ; all that is now intended is to indicate broadly the general scope of Meristic and Substantive Variation respectively. As has already been stated, it is proposed to begin the Study of Variation by an examination of Variations which are Meristic, leaving the consideration of Substantive Variation to be undertaken hereafter. But nevertheless in the consideration of Meristic Variation it will be necessary to refer to phenomena of Substantive Variation in so far as their occurrence or distribution in the body arc affected by Meristic phenomena. For in the determination of the magnitude of the integral steps by which Variation proceeds, the existence of Merism plays a conspicuous part, and it is in con-

sequence of this that the subject of Symmetry and Repetition of Parts has a second and indirect bearing on the Study of Variation which is scarcely less important than the direct bearing of which mention has been made above. This indirect bearing on the manner of origin of Specific Differences arises from a circumstance which in treatises on Evolution is commonly overlooked. In comparing a species in which parts are repeated, with an allied species in which the same parts are repeated, it commonly occurs that each of the repeated parts of the one have some character by which they are distinguished from the like parts of the other. This differentiating character may be a qualitative one, or a numerical one, or both. In such cases it very frequently happens that this character occurs in each member of the series of Repetitions. For example, the tarsi of the Weevils have only four visible joints, while those of the majority of beetles have five ; but the characteristic division into four joints occurs in each of the legs. Before the four-jointed character as seen in the Weevils could be produced it was necessary that not one but all of the legs should vary from the five-jointed form, and in this particular way. The leaves on a beech tree are all beech leaves, and if the tree is a fern-leaved beech, they may, and generally speaking do, all shew the characters of that variety ; and so on with other particular species and varieties.

The limbs of a bilaterally symmetrical animal, in which the right side is the image of the left, are of course alike, and any specific character which is present in the limbs of the one side must in such an animal be normally present in those of the other side. The same is true of many forms in which appendages are repeated in series, as for example, the fore-legs and hind-legs of the Horse, the fore- and hind-wings of the Brimstone Butterfly (Gonepteryx rhamni); of the patterns on several segments of many caterpillars ; of the patterns of the segmental set® of many worms, and so forth. In series whose members are differentiated from each other, it of course frequently happens that the same specific characters are not present in all the members of the series, and in nearly all such cases these characters are not presented by all in equal degree ; nevertheless substantially the phenomenon remains that similar characters often are presented by the several members of a series of repeated organs.

To many this will seem little better than a truism, nevertheless I offer no apology for its introduction ; for though, as a common and obvious fact, it is a truism, it is besides a truth, the far-reaching significance of which is scarcely appreciated. For, in the consideration of the magnitude of the integral steps by which Variation proceeds, we shall have this to remember : that to produce any of the forms of which we have spoken, by Variation

from another form, it is not enough that the particular Variation should occur and become fixed in one member of the series, but it is necessary that the character should sooner or later be taken on by each member of the series which exhibits it. In such cases therefore, this question is raised. Did the Variation come in h'r>t in one member of the Series and then in another? Did it occur, for example, simultaneously on the two sides of the body? Did the right and left forelegs of the Horse cease to develop more than the present number of digits simultaneously or separately? \Vas the similar form of the hind-legs assumed before, after, or simultaneously with that of the fore-legs ? Were the orange markings which are present on both fore- and hindwings of the Brimstone, or the ocellar markings of the Peacock (V. Iu), and of tin- Knqteror (Saturnia carpini), assumed by both wings at once? Were the four wings of the Plume Moths split simultaneously into the characteristic " plumes " ? Did the brown spots on the three leaflets of Medicago, the fimbriation of the petals of Ragged Robin (Lychnis flos-cuculi}, the series of stripes on the Zebra, the pink slashes on the segments of Sphinx larvae, the eyes on the sent* - of Chitons, and the thousand other colour-marks, sense-organs, appendages and structural features, which throughout organized Nature occur in Series, vary to their present state of similarity by similar and simultaneous steps, or did each member of such Series take these characters by steps which were separate and occurring independently? To this question, which lies at the root of all progress in the knowledge of Evolution, the Study of Variation can alone reply.

That in the facts which follow, the answer to this question will be found, cannot of course be said ; but these facts, few though they are, do nevertheless answer it in part, and they suggest that more facts of the same kind would go far towards answering it completely. But beyond this, the facts are of value as an indication of the part which the phenomenon of Merism may play in determining the magnitude of Variations and the manner of their distribution among the several parts of the body. On examining the evidence it will be found that between parts related to each other in the way that has been described, there is a certain bond or kinship, by virtue of which they may and often do vary simultaneously and in similar ways, though the fact that they may also vary independently, and in different ways, will of course also appear.

The phenomenon of the Similar Variation of parts which are repeated Meristically in Series is a fact which will be found to have important bearings on several distinct departments of biological study. As was shewn above, it is by recognition of the existence of such similar and simultaneous Variation that the manner of origin of the similar complexity of several organs belonging to the same system or series becomes comparatively comprehensible ; for

it is not then necessary to conceive a separate origin for the complexity of each member of the series. For example, it is difficult to conceive the manner of evolution of an eye of a vertebrate ; nevertheless, for each vertebrate two eyes have been evolved. If it were necessary to suppose that each arose by separate selections of separate variations, the difficulty would be thus doubled. If, however, it is recognized that the complexity of both arose simultaneously, the phenomenon becomes the more intelligible as the number of integral variations and selections demanded is reduced.

The case chosen, of paired organs in bilateral symmetry, is a very simple one, but it will be found that similar relations hold between other parts repeated in series. For in the same way it is not necessary to suppose an independent evolution for each of the tail-feathers of the Peacock, for the legs of the Horse, and the like, since in the light of the facts of Variation it is as easy for all to take on the new characters as for one. If the manner of development of Repeated Parts is considered, this fact will not seem surprising. For all these parts arise from the undifferentiated tissues by a process of Division, and whatever characters were potentially present in the undifferentiated tissues may appear in the parts into which it subsequently divides. A somewhat loose illustration will perhaps make this more clear. Everyone knows the rows of figures which children cut out from folded paper. There are as many figures as folds, each figure being alike if the folds coincide. If the paper is pink, all the figures are pink ; if the paper is white, all the figures are white, and so on. If blotting-paper is used, and one blot is dropped on the folded edges, the blot appears symmetrically in all the figures. So also any deviation in the lines of cutting appear in all the figures ; a whole row of soldiers in bearskins may be put into helmets by one stroke of the scissors. Of course it is not meant to suggest that the process of division by which parts of the body are produced bears any resemblance to that by which the figures are cut out, but merely to illustrate the fact that since it is by a process of Division of an undifferentiated mass that the Repeated Parts are produced, so the characters of these Repeated Parts depend upon the characters which were present in the original mass and upon the modes by which the parts were divided out from it.

At this point it will be well briefly to recapitulate the preceding Sections. We are proposing to attack the problem of Species by studying the facts of Variation. Of the facts of Variation in general we have selected a particular group upon which to begin this study. The group of variations thus chosen are those which relate to Number of parts, Division, Repetition, and the other phenomena which are to be included under the term Meristic. With variations in quality and Si distance it is not at present proposed to deal, except in so far as it is necessary to refer to them in their relation to the phenomena of Merism, and in illustration of the structural possibilities or necessities which in the body follow as corollaries upon the existence of Meristic Repetition.

It has also been proposed to limit the consideration to Variations which are Discontinuous. As has been already stated, Discontinuous Variations may belong to the Meristic Group or to the Substantive, but it is to the former that attention will first be directed. In what has gone before, the two conceptions now introduced, namely the distinction of Variations into Meristic and Substantive, and iuto Continuous and Discontinuous, have been sketched in outline. The significance of the facts which follow will be made, more evident if these two conceptions are now more fully developed in some of their aspects.

Under the name Merism I have proposed to include all phenomena of Repetition and Division, whenever found and in whatever forms occurring, whether in the parts of a body or in the whole. The consequences of the admission of this proposition are considerable and should be fully realized; for on recognition of the unity of these phenomena it is possible to group together a number of facts whose association will lead to simplification of some morphological conceptions, and to other results of utility.

That the phenomena of Merism form a natural group is in some respects a familiar idea, but in its fullest expression it is as yet not generally received, still less have the consequences which it entails been properly appreciated. Every one who has gone even a little way into morphological inquiry has met some of the difficulties to which we shall now refer. It is with respect to the phenomena of Segmentation that these difficulties are most familiar, and it is in this connexion that. they may be best discussed. Segmentation is a condition which reaches its highest development iii Vertebrates, ih,. Annelids, and the Arthropods, and it is in these groups that it has been most studied. In them it appears as a more or less coincident Repetition of elements belonging to most of the chief systems of organs along an axis corresponding to the long axis of the body. To segmentation of this kind the name 'Metameric' has been given, and by many morphologists the attempt has been made, either tacitly or in words, to separate such Metameric Segmen- t at ion from other phenomena of Repetition elsewhere occurring.

It has thus been attempted to distinguish the Repetitions which occur along the long axis of the body from those occurring along the long axis of appendages, such as for example the joints of antennae or of digits, and some have even gone so far as to regard the Segmentation of the Vertebrate tail as a thing different in kind from that of the trunk itself. It would be apart from our present purpose to recur to these subjects, were it not that this suggestion of the existence of a difference in kind between Metameric Segmentation and other Repetitions has led to several notable errors in the interpretation of the facts of morphology and in the application of these facts to the solution of the problems of Descent. In order to lay a sound foundation for the study of Meristic Variation these errors must be cleared away, and to do this it is necessary to break down the artificial distinction between the phenomena of Metameric Segmentation and other cases of Repetition of Parts, so that the whole may be seen in their true relations to each other. When this is done, the mutual relations of the facts of Meristic Variation will also become more evident.

The first difficulty which has been brought into morphology by the suggestion that Metameric Segmentation is a phenomenon distinct in kind, is one which has coloured nearly all reasoning from the facts of Morphology to problems of phylogeny. For the existence of Metameric Segmentation in any given form is thus taken to be one of its chief characters, and, as such, is allowed predominant weight in considering the genetic relations of these forms. By the indiscriminate though logical extension of this principle the conclusion has been reached that Vertebrates are immediately connected with, or have arisen by Descent from Annelids, or from Crustacea and the like, for the Repetition in these forms is closely similar. Others again, being struck with the resemblance between the Repetition of Parts along the radial axes of Starfishes and those which occur along the long axes of Annelids have hazarded the conjecture that perhaps this resemblance may indicate the actual phylogenetic history of these Repetitions. Though such speculations as these are little better than travesties of legitimate theory, some of them still command interest if not belief.1 All alike are founded on the assumption that resemblances between the manner and degree in which Repetition occurs are unlikely to have arisen save by community of Descent. A broader view of Meristic phenomena will shew that

1 These modern "Instances" recall many that once were famous but are now forgotten. For example : Item non absurda cat shnilitudo et conformitas illn, ut homo sit tanquam planta inversa. Nam radix nerronim et facilitation inuiiialiiim est caput ; partes aitteni seminales xtntt ivftnue, non computatis cxtreinitatibiig tibiarum et brachiorum. At in plaiita, radix (qua .ins tar capitis est) rci/nluriter iiifuiiu loco collocatur ; semina aittem supremo. BACON, Nov. Org. Lib. n. 27. In non computatis extremitatibus, amateurs of INSTANTI.E CONFOEMES may still find matter for warning.

this assumption is unfounded; torso far are the expressions of it which are called Metamerism from standing alone, that it is almost impossible to look at any animal or vegetable form without meeting phenomena of Repetition which differ from Metamerism onlv in degree or in extent. Between these Repetitions and Metameric Repetitions it is impossible to draw any line, and the Variations of all will therefore be treated together. This error in the estimate of the value of Metamerism as a guide to phylogeny is one by which the evidence of Variation is only indirectly affected. The other errors now to be mentioned are of a much more serious nature, for they concern the general conception of the nature of Homology which is the basis of all morphological study.

In introducing the method of the Study of Variation I have said that it can alone supply a solid foundation for inquiry into the manner by which one species arises from another. The facts of Variation must therefore be the test of phylogenetic possibility. Looking at organs instead of species, we shall now see that the facts of Variation must also be the test of the way in which organ arises from organ, and that thus Variation is the test of Homology. For the statement that an organ of one form is homologous with an organ of another means that there is between the two some connexion of Descent, and that the one organ has been formed by modification of the other, or both by modification of a third. The precise way in which this connexion exists is not defined, and indeed has scarcely ever been considered, though such a consideration must sooner or later be attempted. We must for the present be content with the belief that in some undefined way there is a relationship between ' homologous ' parts, and that this is what we mean when we affirm that they are homologous.

We have however assumed that the transition from one form to another takes place by Variation. If therefore we can see the variations we shall see the precise mode by which the descent is effected, and this must be true of the parts or organs as it is true of the whole body. In like manner then as the Study of Variation may be hoped to shew the way by which one form passes into another, so also may it be hoped that it will shew how the organs of one form take on the shape of the homologous organs of another.

In the absence of the evidence of Variation reasoning as to Homology rests solely on conjecture, and assumptions have thus been made respecting the nature of Homology which have coloured the whole of morphological study. Of these, two demand attention now. I. As to Homology between the Members of one Series. We saw above (page 29) how the resemblance between Repetitions occurring in divers forms has led to the belief that those forms had a common descent : in a somewhat similar way it has happened that the resemblance between individual members of a series of Repeated Parts has led to the belief that they must originally have been alike, and that they have been formed by differentiation of members originally similar. Many who would hesitate thus to formulate such a belief nevertheless have taken part in inquiries which can succeed only on the hypothesis that this has been the history of such parts. Of this nature are the old attempts to divide the skull into vertebrae, recognizing the several parts of each ; the modern disquisitions on the segmentation of the cranial nerves ; the attempts to homologize the several phalanges of the vertebrate pollex and hallux with the several phalanges of the other digits ; similar attempts to trace the precise equivalence of the elements of the carpus and tarsus, and many other quests of a like nature. In all these it is assumed that there is a precise equivalence to be found with enough searching, and that all the members of such series of Repetitions were originally alike. If the series of ancestors were before us it is expected that this would be seen to have been the case. In the light of the facts of Variation this assumption will be seen to be a wrong one, and these simple views of the Repetition and Differentiation of members in Series must be given up as inadequate and misleading, even though there be no other to substitute.

II. As to the individuality of Members of Series. In seeking to homologize a series of parts in one form with a series of parts in another, cases often occur in which the whole series of the one is admittedly homologous with the whole series of the other. In such cases the question arises, can the principle of Homology be extended to the individual members of the two Series ? If the two Series each contain the same number of members this question is a comparatively simple one, for it may be assumed that each member of the Series is the equivalent or Homologue of the member which in the other Series occupies the same ordinal position. If however the number of members differs in the two Series, how is the equivalence to be apportioned ? This is a question again and again arising with regard to Meristic Series such as teeth, digits, phalanges, vertebra?, nerves, vessels, mammae, colour-markings, the parts of the flower, and indeed in almost every system whether of animals or plants. To decide this question there are still no general principles. But though we yet know nothing as to the steps by which Meristic Variation proceeds, there is nevertheless a received view by which the interpretation of the phenomena is attempted, and though in the case of each system of organs the application of the principle is different, yet the principle applied is essentially the same.

Thus to compare the members of Series containing different members it is first assumed that the series consisted ancestrally of some maximum number, from which the formula characteristic of each descendant has been derived by successive diminution. Here, again, I do not doubt that many who employ this assumption would hesitate to make it in set terms, but nevertheless it is the logical basis of all such calculations. Now this hypothesis involves a definite conception of the mode in which Variation works, and it is most important that tins should be realized clearly. For if it is true that each member of a Series has in every form an individual and proper history, it follows that if we had before us the whole line of ancestors from which the form has sprung, we should then be able to see the history of each member in the body of each of its progenitors. In such a Scries tinrise of an individual member and the decline of another should then be manifest. Each would have its individual history just as a Fellowship in a College or a Canonry in a Cathedral has an individual history, being handed on from one holder to his successor, some being suppressed and others founded, but none being merged into a common fund. In other words, according to the received view of the nature of these homologies, it is assumed that in Variation the individuality of each member of a Meristic Series is respected.

The difficulty in applying this principle is notorious, but when the evidence of Variation is before us the cause of the difficulty will become evident. For it will be found that though Variation may sometimes respect individual homologies, yet this is by no means a universal rule ; and as a matter of fact in all cases of Meristic Series, as to the Variation of which any considerable body of evidence has been collected, numerous instances of Variation occur, in which what may be called the stereotyped or traditional individuality of the members is superseded.

This error in the application of the principle of Homology to individual members of Meristic Series has arisen almost entirely through want of recognition of the unity of Meristic Repetition, wherever found. In the case of a series of parts among which there is no perceptible Differentiation, no one would propose to look for individual Homologies. For example, no one considers that the individual segments in the intestinal region of the Earthworm have any fixed relations of this kind ; no one has proposed to homologize single leaves of one tree with single leaves on another ; it is not expected that the separate teeth of a Roach have definite homologies with separate teeth of a Dace, for such expectations would be plainly absurd. But in series whose members are differentiated from each other the existence of such individuality is nevertheless assumed. To take only one case: a whole literature has been devoted to the attempt to determine some point in the vertebral column or in the spinal nerves from which the homologies of the segments may be reckoned. This is a problem which in its several forms has been widely studied. Some