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

Thus far the connexion between the geometrical relations of the digits and the modes of their Variation is clear and simple, and does not differ from that maintained in the Major Symmetry. But in proceeding further there is difficulty. If, for instance, the manus or pes of a Horse possesses within itself the properties of a bilateral Symmetry, then the splint-bone II may be supposed to be in symmetry with the similar bone IV. It would then-fore be expected that on the occasion of the development of II to be a full digit, the splint-bone IV would at least not unfrequeiitlv develop, thus exhibiting that similarity and simultaneity of Variation which we have learnt to expect from parts in -vmmetry with each other. Nevertheless such an occurrence seems to be extremely rare. Then arises a further question: if the digit II develop simultaneously, say in the two fore feet, would the mechanical conditions of which Symmetry is the outward expression be satisfied without a corresponding change in the digit IV of the fore feet? Is the frequent absence of symmetry in the variation <>f the halves of the Minor Symmetry in any way connected with the possibility that th<- two Minor Symmetries together may be maintaining their relations to each other as pails of a Major Symmetry ' ( )f course as to this we know nothing, but the existence of this double relation should be remembered.

In several other phenomena of digital Variation the influence "t Symmetry is to be suspected. Reference may first be made to the -..-ries of changes seen in the Cat's hind foot in correlation with numerical change. The bones of this pes do not normally exhibit any very clear bilateral symmetry l. Yet on the appearance of new digits the foot is reconstituted and its parts are, to use a metaphor, ' deposited ' in a system of bilateral symmetry '2 whose completeness is proportional to the degree of development of the new digits. What may be the meaning of this extraordinary fact one cannot yet guess. The fancy is constantly presented to the mind that there is in the normal foot a condition of strain, that the balance between the right foot and the left is a condition of imperfect stability, and that upon the introduction of some unknown disturbance this balance is upset and each foot settles down as a separate system. But I see no way of testing this fancy and no way of following it further.

Still more complex are the facts seen in the human hand. There is here first the fairly complete series of conditions ranging from the normal, through the three-phalanged thumb up to the several Conditions in which extra digits upon the internal side of the limb seem to have sprung up to balance the four normal digits ; but on the contrary there is the exceptional case of the Macacque's foot (No. 504) where the extra parts are, as I believe, external. (Besides these there are the wholly distinct series of "doublehands," which will be spoken of below.) The former cases taken alone would certainly suggest that there is an imperfect balance or system of symmetry subsisting between the thumb and the four fingers of the normal manus, but to this suggestion there are numerous difficulties which need hardly be detailed in this preliminary glance at the phenomena.

With more confidence it can be maintained that the pollex and perhaps the hallux of Man is in itself a Minor bilateral Symmetry, apart from the four fingers, for it may divide into equal parts related as images. The same is true of the hallux of the Dorking (p. 390), and probably of the extra digit or digits sometimes arising from the tibio-tarsus of the Turkey for example (see No. 603). Besides this the facts of the frequent syndactylism between the digits III and IV of the human manus, taken in connexion with the phenomena of the Pig and Ox, suggest that the four fingers may have among themselves again a relation of the nature of Symmetry.

1 In the normal pes, though all the claws are retracted to the outside of the second phalanges, yet the claws of digits III and IV rest close together, that of III being external to its pad, while that of IV is internal to it* pail, forming, so far, a relation of images between these two digits. In the polydactyle foot it is a remarkable feature that, though the bones are in symmetry about an axis passing between II and III, the relation of the claws of III and IV to their pads remains almost normal, still giving a superficial appearance of symmetry between these two digits. (In the polydactyle pes the pads are mostly rather narrower.)

- It will be remembered that this symmetry appears not merely in the lengths of the several digits but in the manner of retraction of the claws and in the corresponding form of the second phalanges, three digits being fashioned (in the case of six perfect digits) as right digits and three as lefts. It has been mentioned that there is some evidence to shew that in the human ]»- it is tindigits II and III which are most frequently syndactvlf, .-ven up to the point of being (in No. ">±l) a|»]iaivntly ivpiv-eutrd by a single digit, and in this connexion it will be remembered that in the polydactyle pes of the Cat it is also between these digits that the new axis of Symmetry falls.

Th.'Sf >r;mty allusions to the possible influences which Symiiiftrv niav exercise over Meristic Variation of digits will suffice to indicate the nature of the problem to those who may care to f\aniiiM' it. It is with hesitation that so indefinite a matter is >|iok,-n "t at all. Nevertheless it is likely that if any one can find a way of interpreting these indications the result will be considerable. In the evidence as to the digits of Man facts were given iv>pecting the state known as Double-hand, and some similar '•asfS were referred to in Artiodactyles. In these instances the digital series, and to some extent the limb, is in its new shape made up of the external parts of a pair of limbs compounded together in such a way that there is a partial duplicity of the limb, the two halves being more or less exactly complementary to each other and related as images1.

This phenomenon in its perfect form must be essentially distinct from the other cases of increase in number of digits; for in the double-hands the limb developes an altogether new bilateral symmetry (see especially No. 4-92). Between cases of duplicity in limbs and the other forms of polydactylism confusion can only arise when the nature of the parts is ambiguous. As has been stated, in all certain cases of double-limbs the two are compounded by their internal or pra-axial borders, but the case of Macacque No. 50-i was peculiar in the fact that there \\as in it a presumption that the two limbs were not a pair but in Succession.

In Arthropoda there are a very few cases of true duplicity in appendages comparable with the double-hands. These cases will be dealt with hereafter. The fact that a structure naturally hemi-symmetrical, needing the limb of the other side to balance it, may on occasion develop as a complete symmetry is most ]>;u,i lexical, but no other interpretation <>f the facts seems possible. The phenomenon is of course comparable with that observed by DRIESCH in the eggs of J-lchinus, where each half-ovum developed into a whole larva on being separated from the other half-ovum (see p. 3"», \'nt<'). It will be shewn that in almost every case in which such un ;i|p|ir:ii.-in<v i^ found iii tlie extra appendages of Insects this appearance is misleading, and that the extra parts have a Secondary Symmetry of their own : but no such way through the dith'culty is here open.

(6) Homceotic Variation in terminal digits when a new member This is a principle that has been several times seen in Meristic Variation, and in Chapter x. Section 7, it was treated of at length in the case of teeth. Some few illustrations of the same principle occur among the evidence as to digits. It has been seen for instance how that, upon the appearance of an extra digit on the radial side, the digit which stands in the position of pollex may have three phalanges and resemble an index (No. 485, &c.). Similarly it was found that upon the formation of a large digit externally to the minimus the digit standing in the ordinal position of the minimus may have an increased proportional length (No. 509). Still more important is MORAND'S case (No. 510), in which the most external digit had muscles proper to a minimus, while the digit standing in the ordinal position of the minimus was without them.

The cases of extra digit in the Horse (No. 536, &c.) still more clearly illustrate the principle, if the view of the nature of those cases taken in the text be received. It should be expressly stated that in digits, as in teeth, it is not always that the terminal member is promoted on becoming penultimate. Such promotion is indeed rather exceptional in digits, but the fact that it may occur is none the less a phenomenon of great significance. (7) The absence of a strict distinction between duplicity of a given digit and other forms of addition to the Series.

This subject has been so often spoken of in connexion with special cases that it is unnecessary here to make more than brief allusion to it. The same principle was shewn to be true of teeth (p. 270) and of mammae (p. 193), and there is little doubt that it is true of Meristic Series generally. Facts illustrating the matter in relation to digits will be found in the evidence as to duplication of pollex and hallux in Man (p. 351), as to duplication of the hallux in the Fowl (p. 391), in the evidence of cases in the Horse of variation intermediate between division of III and development of II (p. 371), and in the cases of three-toed Cows (p. 377).

In almost all the animals in which any considerable range of digital Variation is to be seen it is possible to find a series of cases making an insensible transition from true duplicity, or division into two equivalent parts whose positions and forms are such that they maybe reasonably looked upon as both representing a normally single member, up to the condition in which while the series contains a greater number of members, each member nevertheless stands in a regular Succession to its neighbour.

Upon the proper understanding of this proposition and upon the recognition of its truth hang those corollaries before enuntiated touching the false attribution of the character of individuality to members of Meristic Series. The evidence that the Meristic Variation of digits may be discontinuous is often rather circumstantial than direct. If for example in the case of the Horse any one chooses to suppose that every polydactyle horse had in its pedigree an indefinitely long scrips of ancestors in which the size of the extra digit progressively increased, it would not be easy to produce direct evidence that this \\as not the fact. But as regards the human examples such evidence is abundant, many of the most marked cases being the offspring of normal parents and there can be no reasonable doubt that the same would lie found true of other animals.

But it may fairly be replied that until it shall have been shewn that formations like those described as variations may be established in a natural race or species the contention that the Variation of dibits may be discontinuous is so far weakened. To this I would reply by referring to the case of Cistudo, Chalcides, and the other similar examples ; for though in respect of these forms the evidence is sadly imperfect yet it plainly indicates that very distinct and palpable variation may be found between different individuals. And since it is actually known that there may in these points be considerable differences between the two sides of the body it may safely be assumed that at least the same differences may occur between parent and offspring.

\Ve may therefore take it that there is in these cases some Discontinuity of Variation, though until some one shall have examined statistically such cases as that of the Box-turtles or of the Kittiwakes, as to the magnitude of the Discontinuity it is not possible to speak. If hereafter Discontinuity shall be shewn to occur in many such cases it will be difficult to resist the suggestion that similar numerical diversity elsewhere characterizing the digital series of various forms may have come about by similarly discontinuous Variation.

(9) Relation of tJie facts of diyital Variation to the problem of This relation is both direct and indirect: direct, inasmuch as some of the conditions seen to occur as variations are not far removed from those known as normals in other forms ; and indirect, since.1 those strange and paradoxically regular dispositions of digits which are found among the variations bear witness to the influence of the principles of Symmetry, and prove that there are modes in which Variation may be controlled and may produce a result which has the quality of regularity and order of form independently of the guidance of Natural Selection.

Of actual variations from the arrangement of digits characteristic of one form to that characteristic of another there are as yet scarcely any examples. The cases given on pp. 395 to 398 being the most evident. For the rest, that is to say examples of arrangements happHi- ing as variations matching no normal, some may say in haste that with their like Zoology has no concern. It would be convenient if those who make this careless answer (as many do) would mark the point at which it is proposed to begin this rejection of the evidence of Variation. Few perhaps realize how impossible it is to give a real meaning to these distinctions. As regards digits, for instance, I suppose that no one who holds the doctrine of Common Descent would refuse to admit the evidence of Variation as to the hallux of Hedgehogs (No. 612) as exemplifying the way in which species may be built up — if indeed species are built up of variations at all. And if this case is admitted, by what criterion shall we exclude cases of the formation of a hallux in the Dog? But if these are not excluded it is difficult to shew good reason for not admitting the case of the three-phalanged digit placed as a hallux in the Cat (No. 472) with all the curious series of which that is only the first term. Are we quite sure that because there is no Carnivore with a three-phalanged hallux therefore such a creature could not exist in nature ? Still more difficult is it to shew cause why duplicity of the hallux should be set apart as a variation not capable of being perpetuated or of becoming part of the specificcharacters of an animal, seeing that there is actual evidence both in the case of the Dorking fowl and in the St Bernard dog that itmay become at least an imperfectly constant character.

In connexion with the subject of this section many suggestions with special bearing on particular cases, both positive and negative, will strike every reader. In the present imperfect state of the evidence it would be premature to pursue these. It may however be well to mention that several writers, especially JOLY and LAVOCAT (No. 55-t). have seen in the cases of divided digit III in the Horse an indication that the digit III of the Horse corresponds with the digits III and IV of the Artiodactyles. The evidence as to syndactylisrn between these two digits in Ox and Pig would probably be considered to give support to the same view. But while we may note that the relations of the digits with the carpus and tarsus of these forms,, were comparative evidence absent, should absolutely prevent any one from seriously maintaining such an opinion, nevertheless the fact that such closely similar systems of Symmetry may thus arise independently of each other is of interest.

.MKIMSTIC Repetition along the axes of appendages is very like that along the axis of the body. Just as particular numbers of -e^meiits or repetitions along the axis of Major Symmetry characterize particular forms, so particular numbers of joints characterize particular appendages. Such numbers frequently differentiate species, genera, or other classificatory divisions from each other. In the evolution of these forms therefore there must have been change in these numbers.

Those who are inclined to the view that Variation is always <-"iit iniious do not perhaps fully realize the difficulty that besets the application of this belief to the observed facts of normal .structure. For in those many groups whose genera or species may be distinguished from each other by reason (amongst other things) of difference in the number of joints in some particular appendage or appendages, will any one really maintain that in all these the process by which each new number has been introduced was a gradual one? To take a case: even were evidence as to tlie manner of such Variation wanting, would it be expected that the Longicorn Prionidae, most of which have the unusual number of \'l antunary joints, did, as they separated from the other Longicorns which have 11 joints, gradually first acquire a ne\v joint as a rudiment which in successive generations increased ? Or, conversely, did the other Longicorns separate from a 12-jointeil form by the gradual "suppression" of a division or of a joint ' If any one will try to apply such a view to hundreds of like examples in Arthropods, of difference in number of joints in appendages of near allies — forms that by the postulate of Common Descent we must believe to have sprung from a common ancestor — he will rind that by this supposition of Continuity in Variation he is led into endless absurdity. Surely it must be clear that in many such cases to suppose that the limb came through a phase in which one of its divisions was half-made or

one of its joints half-grown, is to suppose that in the comparatively near past it was an instrument of totally different character from that which it has in either of the two perfect forms. But no such supposition is called for. With evidence that transitions of this nature may be discontinuously effected the difficulty is removed. The frequency of Meristic Variation in appendages is much as it is in the case of body-segments. On the one hand there are series containing high total numbers of repetitions little differentiated from each other (e.g. the antenna? of the Lobster), and in these Meristic Variation is common ; on the other hand in series containing few segments much differentiated from each other, such Variation, though not unknown, is rare. Of the latter a few instances are here offered. That the}r are so few may perhaps be in part attributed to the little heed that is paid to observations of this class. Records of this kind might indeed be hoped for in the works of those naturalists to whom the title " systematic " has been given ; but unfortunately the attention of these persons has from the nature of the case been drawn rather to features whereby species may be kept apart than to facts by which they might be brought together.

From the lack therefore of records of such variations their absence in Nature must not lightly be assumed. To quote but one case : in the common Earwig the numbers and forms of the antennary joints are exceedingly variable, but in many special treatises on Orthoptera, I cannot find that this variability is spoken of, and if alluded to at all the only notice is given in the form "antennae 13- or 14-jointed." I am indebted to Dr D. Sharp for the information that the number of antennary joints in certain Prionidse varies. In Longicorns generally the number of joints is constantly 11. Dr G. H. HORN of Philadelphia who is specially acquainted with this group, has kindly written to me that of six species of N. American Prioni four species have 12 antennary joints constantly in both sexes. Besides these he gives the following cases of Variation. It will be seen that in both of these the normal number is much greater than it is in the other species1. *616. Prionus imbricornis: females have very constantly 18 joints; males have 18 to 20. A male in Dr Sharp's collection has only 17 joints in each antenna.

1 In Prionus imbricornis and P. fissicornis doubt may be felt whether the trifid apex should be reckoned as one joint or as two, but this applies equally to each individual. I have counted it as one. *G17. Prionus fissicornis: the female has 25, and the male 27- 30, the note on the preceding species applying here. 618. Polyarthron. A Prionid beetle, in which the male has curious many jointed feather-like antennae, according to SERVILLE has always 17 joints, but THOMSON (>'_y.s/r. Ceramb., 1866, p. 284) says the number varies with the species and individually. A male in Dr Sharp's collection has 45 joints in each antenna and a female has 31 in each.

G19. Lysiphlebus is a Braconid (Hymenoptera) parasite on Aphides. From a colony of Aphides on a bush of Baccharis viminalis 121 specimen^ of Lysipklebus were reared : of these 57 were males and 64 were females. In those having a different number of joints in the right and left antenna?, the last joint of the antenna which contained the fewest joints was longer than the last joint of the antenna with the larger number of joints. Nevertheless this relation did not hold throughout : for example in the case of the male with 16 joints, the last joint was of the same length proportionally as that of the males with only 14 joints. As a rule the specimens with fewer antennary joints are smaller than the others.

Variations were also seen in coloration, in the proportional length of the tarsi, and in the presence or absence of the transverse cubital nervure, but none of the characters divided the sample consistently, it was therefore inferred that the individuals belonged to one species of Lysiphlebus, (L.citraphis, Ashm.) From another colony of Aphides living on a rose-bush 58 specimens of Lysiphlebus were bred, and no characters were found by which these could be separated from those bred from the Aphis of Baccharis. In the case of the second sample the joints of antennas were as follows :

The number of antennary joints is employed as a specific character in the classification of Lysiphlebus by ASHMEAD, Proc. U. S. Nat. Mus., 1888, p. 664). COQUILLET, D. W., Insect Life, 1891, Vol. in. p. 313. *620. Donacia bidens. (Phyt.) A female found by Dr 1). Sharp at Quy Fen in company with many normal specimens had in each antenna eight joints instead of eleven as in the normal. As shewn in the figure (Fig. 123) the antennae of the two sides were exactly

FIG. 123. Donacia bidens ? . I. Normal antennas, eleven joints in each. II. Abnormal specimen, having eight joints in each antenna. No. 620. alike, and the insect was normal in all other respects. I am much obliged to Dr Sharp for shewing me this specimen. Forficula auricularia, the common Earwig. In the various species of Forficula the number of joints in the antennas differs, the numbers 11, 12, 13 and 14 being all found as normals in different species1. As regards F. auricularia most authors give 14 as the number of antennary joints. SERVILLE2 gives 13 or 14. A number of adult earwigs examined by myself with a view to this question shewed that there is great diversity in regard to the number of antennary joints. The whole matter needs much fuller investigation but the preliminary results are interesting.

The commonest number is 14, which occurs in perhaps 70— 80 per cent. The next commonest is 13, which was seen in a considerable number, while 12, and even 11 occur in exceptional cases. Different numbers were frequently found on the two sides. 1 BRUNNER VON WATTENWTL, Prodr. eur. Orth., 1882. The number in F. auricularia is given by Brunner as 15, but I have never seen this number. It is no doubt an accidental error. The same mistake is repeated by SHAW, E., Ent. Mo. Mini., 1888—89, xxv. p. 358.

A- is ii-ual with appendages the whole length of the antenrue diffnvd a good <1< -al imlfpnidmtly of the number of joints. 621. On comparing antmn:L' with different numbers it seemed that tli.- pi-oporiional length of the first two joints was nearly the same in all, but in the third joint there was great difference, as shu\vn in Fig. 124. The left antenna in Fig. 124, I may be taken to be the normal form with 14 joints. In it both 3rd and 1-th joints ansmall. The right antenna of the same specimen has 'l 3 joints and in most of the 13-jointed antenna1 the arrange-