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

Fig. 20G, I, shews the normal form of a right maxillary palp. l-'i.;. 206, II, represents the right palp of this specimen. Tin- 1st and 2nd joints arc much thickened and the latter has N hairs (instead of 4) and two apical articulations, the anterior bearing FIG. 20f>. Nebria gyllenhnlli, No. 8I')0. I. Normal right maxillary palp. II. Right palp of this specimen. III. Left palp of the same, m, terminal membrane. (The property of Dr Kraat/..) an apparently normal terminal joint, the posterior bearing a symmetrical piece ending in a sharp point with no membrane like that at the apex of the normal. The left palp of this specimen is shewn in Fig. 206, III. In it the 2nd joint has 8 hairs instead of 4, and the terminal joint though very much enlarged is not divided at all. For the loan of this specimen I am indebted to Dr G. KRAATZ who first described it in Berl. ent. Zt, 1S~:>. xvn. p. 433, yfy. 12.

861 . Carabus splendens : penult, jt. of 1. labial palp enlarged, and bearing two nearly similar jts. [broken before seen by me]. MOCQUKKYS, I.e., p. 29, .////. 862. C. auratus: 1st. jt. of 1. maxillary palp bears two similar branches at rt. angles to each other, each with two jts. [Specimen not scon.] MOCQUEHYS, /. <•.. p. 30. ji<i. 864. Lucanus. Three cases are recorded in which one of the mandibles bore an extra process of considerable size. Whether ;in\ of these are examples of duplicity, or whether the jaw, morphologically single, has in them varied towards a state of greater complexity, cannot well be said. The cases are L. cervus J, M <)(•(,» r K i ;vs, /. c., p. 10(5 [figure fairly true]; L. cervus </, KitAATZ, Dent. ent. Zt., 1881, xxv. p. Ill, fig.; L. capreolus J, id., 1. c., 1876, XX. p. -^8, fig.

THE evidence as to repetition of appendages in vertebrates is of great extent and has been studied by many, but in the morphology of these repetitions there is still much that is obscure. Speaking generally, the phenomena are similar to those seen in Arthropods, but there is no approach to the same regularity. Nevertheless when two extra limbs are present, it is usually possible to recognize that they are together a complementary pair; and if the extra part is apparently a single limb it is, I believe, never a normal limb and may very often be shewn to contain parts of a pair of limbs. The fact that the geometrical relations of the parts are less regular than they are in Arthropods may probably be ascribed in some measure to the circumstance that the surfaces of the vertebrate limbs do not maintain their original relations but are more or less rotated in the course of their development.

In Insects it appeared that repetition of the peripheral parts in Secondary Symmetry was not much more common than repetitions of whole limbs, but apparently this is not the case in vertebrates. Perhaps it would be more true to say that in vertebrates it is only in those extensive repetitions which include the greater part of the limbs beginning from the girdles, that the parts are clearly in Secondary Symmetry. From this circumstance doubt suggests itself whether some of the phenomena of polydactylism, at present regarded as repetitions of digits in Series, may not really be of the nature of Repetitions in Secondary Symmetry (see p. 378). But however this may be, there are, with the exception of some Artiodactyle cases, no examples of paired repetitious of digits or phalanges at all suggesting a comparison with the double extra tarsi &c. of Insects, or the double extra dactylopodites of Crustacea.

In the most usual forms of extra limbs in vertebrates a more or less amorphous pair of limbs, compounded together for a great part of their length, are attached to a supernumerary piece fitted into some part of the shoulder-girdle, or more often into the pelvic girdle. It is important to notice that though, as many (especially Kin OLAXI) have shewn, a complete series can be constructed, ranging for in-tanee from the ordinary pygomelian up to complete posterior duplicity, yet repetition of limbs may be and often is wholly independent of any axial duplicity, being truly a repetition of appendicular parts only.

The question naturally arises whether there is ever an extra limb placed as a single copy of a normal limb of the same side as that on which it is attached. As to this the evidence is not wholly clear, but I incline to think that no case known to me can properly be so expressed. Perhaps the condition which comes nearest to this is exemplified by a ease of a Frog fully described by KlNGSLEY1, where a single extra left hind leg is said to have been attached to the left side of the pelvis, it is difficult to question that this was actually the fact, for the figure clearly represents the extra limb as a left leg; but though the muscles are fully described, the bones are not, and it still seems possible that there was in reality some duplicity in the limb. The leg was admittedly abnormal in its anatomy and the naming of the muscles must in part have been approximate.

But though perhaps it should not be positively stated that no siwjle extra limb is ever formed in a vertebrate in Succession to the normal limb of the same side of the body, it is certainly true that in the enormous majority of polymelians the extra repetition consists of parts of a complementary pair. These phenomena are thus of interest as bearing upon the morphology of repetitions in Secondary Symmetry, but in all probability are not of the nature of variations in the constitution of the Primary Symmetry.

A just view of the details of these phenomena can only be gained from the specimens or from numerous drawings. The cases of extra limbs in Batrachia may be conveniently studied as exhibiting most of t In- different kinds of Secondary Symmetries both in the fore and hind limbs. In all, some fifty cases are recorded. These may be found from the following references. The evidence up to 1865 was put together by DUM£RIL, and an abstract of it is given also by LUNEL, and by KINUSMCV. A fuller bibliography is given by ERCOLAXI. The best papers on the subject are marked with an asterisk. T have added a few references of less importance not included in the other bibliographies.

* CAVANNA, G., Pubbl. del I\. 1st. di Stitdi super, in Firenze, 1879, p. 8, Tar. i. Four important cases; one, fig. "2, apparently resembling Kingsley's in some respects. From these Batrachian cases most of the chief features of the phenomena may be learnt. To those wishing to get a general view of the subject of repetition of Vertebrate linibs in a comparatively small compass the valuable memoir of ERCOLANI quoted above is especially recommended.

Before proceeding to a consideration of the significance of the phenomenon of Repetition in Secondary Symmetry it must be expressly stated that there are in vertebrates a certain number of cases, perhaps even classes of cases, which it is likely differ widely from the rest ; but as was said above, the chief difference between the Vertebrate and Arthropod cases lies in the comparative simplicity of the latter. It may be stated further that this greater simplicity of the Arthropod cases consists especially in the maintenance of the relation between the extra pair and some normal limb.

Remembering always the existence of unconformable cases we may, I think, safely gather up from the simple cases several points relating to the problems of Natural History at large. I only propose here to make allusion to those considerations which are not developed in the ordinary teratological treatises. Of the fact that any regularity can be discerned in these strange departures from normal structure, and of the bearings of this fact on current conceptions of the causes determining the forms of animals it is now hardly necessary to speak further. Other points not before noticed remain.

In the Arthropod cases that were spoken of as ' regular ' it was seen that the polarity of the Secondary Symmetries has a definite relation to that of the body which bears them. This is quite in harmony with the supposition that they are related to the normal body somewhat as buds are related to a colony, for in most colonial forms the morphological axes and planes of the buds are definitely related to those of the stock. But in the Vertebrate cases though there is generally a relation of images between the extra pair, a definite geometrical relation between them and a normal limb is seen more rarely.

That this is so may, I think, be in part at least attributed to the normal twisting of the vertebrate limb, especially of the hind limb, from its original position (see Note on p. 459). A question brought into prominence by facts of this kind is that of the nature of the control which determines how much of a body shall be repeated, or be capable of repetition, in a Secondary Symmetry. With repetition of a whole body we are familiar. Apart from the processes of sexual reproduction, we know this total repetition in the maiiv forms of asexual reproduction, whether occurring by budding, or by division either of adult bodies or of embryos ', and we thus commonly look on the whole body of any organism as in a sense a unit, capable of repetition or of differentiation — the latter especially in gregarious and colonial forms. Again, we familiarly use the conception of cells as units of repetition or of differentiation. Besides these we have come to recognize that members of series of segments are, in their degree, similar units. And generally, the same attribute of separateness may in undefined senses be properly attached to all organs that are repeated in Series, and to appendicular parts especially.

The attribution of some of the undefined properties of "unity2 to some at least of these various groups is very ancient, and thencan be no doubt that it is in the main a right and useful induction. The chief interest of repetitions in Secondary Symmetry lies in the fact that they give a glimpse of new light upon the nature of this unity, shewing a new form in which it may appear. For in Secondary Symmetry there is not a simple repetition of a part in Series, taking its place as a member of that series, but an addition of paired parts, whose intrinsic relation to each other is the same as that of any pair of parts occurring in the Primary Symmetry.

The addition is thus a unit, is in form complete in itself, and seems to have no place in the Primary Symmetry of the whole body any more than a late side-chapel — also a unit with its own focus and polarity— had a place in the design of the original architect of the Cathedral. From analogy, and from general knowledge of vital processes it would I think have been impossible to foresee the very curious indetiniteness of the (jmmfiti/ of the parts repeated in systems of Secondary Symmetry. It seems, especially in Arthropod cases,

1 As a normal occurrence notably in the case of Cyclostomatous Polyzoa of the genus Cm/*/ described by H.UOIKU, S. F., Q. J. M. .S'., 1891, p. 127, Plates. '• This somewhat incorrect term is used here to express some of the meanings commonly still more incorrectly rendered by the word •• individuality"— a word etymological ly most unhappy in this application to things endowed with divisibility as a conspicuous attribute. that the repetition may begin from any point in an appendage and include all the parts peripheral to the point, of origin. Seeing that the repeated parts are, in their degree, comparable with a whole organism, this indefiniteness is remarkable. We have thus to recognize that the property of morphological "unity" may attach not only to a pair of appendages beginning from the body, or from some definite surface of articular segmentation, but also to a pair of parts having no semblance of morphological distinctness.

Strangest of all is the repetition of the index of Crabs and Lobsters in Secondary Symmetry. The dactylopodite is of course a separate joint. Double extra dactylopodites in Secondary Symmetry present no feature different from double extra tarsi, &c. But the index we think of as merely a large spine or tubercle. It is in no sense a joint or segment. Yet a pair of indices may be added to a normal body. The interest of this fact is in its value as a comment on the principle given on p. 476 that extra parts in Secondary Symmetry contain the structures peripheral to their point of origin. The case of extra indices shews that the term peripheral, if it is to include the case of indices, must be interpreted as meaning not morphologically but geometrically peripheral '.

We have spoken of parts in Secondary Symmetry as having no place in the Primary Symmetry of the body. This is on the whole a true statement, but there are a few cases which make it uncertain whether it is absolutely true. These cases are those few where repetitions in Secondary Symmetry were present on appendages of both sides of the body. Cases of this class were Odontolabis stevensii, No. 799, and Melolontha hippocastani, No. 795, where such extra parts were present on both antennae, suggesting that the similarity of the repetition of the two sides is due to the relation of Symmetry between the right side and the left. But against this view may be mentioned the cases Prionus coriarius, No. 750, and Carabus irregularis, No. 760, where two legs of the same side each bore extra parts, and the Lobster, No. 821, having two pairs of extra points on one dactylopodite. These cases suggest that bilateral simultaneity in such repetition may perhaps represent merely a general capacity for this form of repetition. The case of Prionus calif ornic us, No. 843, would no doubt bear on this question, but unfortunately the facts in that case are scarcely well enough known to justify comment.

1 A case is given by FAXON (llarv. Bull., vin. PI. n. fig. 8) of Callinectes hastatus in which the left lateral horn of the carapace, instead of being simple as in normal specimens, had three spines. It is just possible that two of these may have been in Secondary Symmetry. All other cases known to rne are in appendicular parts. One further point remains to be spoken of. Wo have said that a system of parts in Secondary Symmetry is in a sense analogous with a Imd, but in one respect the condition of these parts differs remarkably from all phenomena of budding or reproduction that are seen elsewhere. In a bud the various organs always present the same surfaces to each other, or in other words, the planes of division always pass between similar surfaces. In Secondary Symmetries this is not the case. As illustrated by the diagram on p. 481, the extra parts may present to each other, or remain compounded by any of their surfaces, whether anterior, posterior, or otherwise. This seems to be altogether unlike anything ever met with in animals and plants. It is as if in a bud on a plant two leaves on opposite sides of the axis could in their origin indifferently present any of their surfaces to each other.

It wrill be remembered that the symmetry cannot be the result of subsequent shillings, but must represent the original manner of cleavage of the two extra limbs from each other. We must theretore conceive that in the developing rudiment of the two extra limbs either surface may indifferently be external, the polarity l>e ing ultimately determined by the relation of the bud or rudiment to the limb which bears it. OF the evidence as to double and triple "monstrosity" and of the classification of the various forms no account can be given here. This may be found in any work on general teratology. In this chapter are put together a few notes on points respecting these formations of interest to the naturalist, and having relation to what has gone before.

It is now a matter of common knowledge that in animals [and plants] division may occur in such a way that two or more bodies may be formed from what is ostensibly one fertilized ovum (cp. multipolar cells). But by a similar division, imperfectly effected, the resulting bodies instead of being complete twins or triplets may remain united together, frequently having a greater or less extent of body in common. In other words, speaking of simple cases in bilateral animals, the whole body, resulting from the development, may contain more than one bilaterally complete group of those parts which normally constitute the Primary Symmetry of an " individual."

If well developed, the component groups are most often united by homologous parts, so that there is a geometrical relation of images between the groups together, forming the compound structure, the whole being one system of Symmetry. Concerning the relations of the several parts of such a system to each other numerous questions of interest arise, but with these it is not nowproposed to deal. To those unacquainted with facts of this class it may be of use to point out in the fewest words the direction in which this importance lies. It arises, briefly, from the fact that in the resemblance between a pair of homologous twins, whether wholly or partially divided, there is once again an illustration of the phenomenon of Symmetry, and of the simultaneous Variation of structures related to each other as symmetrical counterparts.

The frequency of close resemblance between twins is a matter of common knowledge. If it be true that such twins may result from the development of one ovum — a fact that cannot be doubted in face of the complete series of stages intermediate between total and partial duplicity — the resemblance between these twins is then of the same nature as that subsisting between the two halves of any other bilaterally symmetrical system. A wide field of inquiry is thus opened up. For, as suggested in the Introduction (p. 36) if the very close resemblance of twins to each other is a phenomenon dependent on Symmetry of Division, the less close resemblance between members of families may be a phenomenon similar in kind.

It will In- remembered that the resemblance between twins is a true case of similar and simultaneous Variation of counterparts. This is clearly proved by the fact that when distinct Mrristie Variations are exhibited by one twin they are not rarely pr.--.-ni in the other alsi). Cases of this simultaneous Variation are familiar to all who have studied this subject. A useful list of examples in completely separate twins is given by WiMii.i;1. One of the best known cases in twins incompletely separated, is that of the Siamese Twins'-', who had each only eleven pairs of ribs (instead of twelve).

Inference must lastly be made to a particular corollary which may naturally be deduced from the fact that the bodies of incompletely separated twins are grouped as a single system of Symmetry. If the whole common body were bilaterally symmetrical, one twin must lie the optical image of the other. But if the organs of one twin be normally disposed, the organs of the' other must be tr<i>tx]i<n«-<l in completion of the Symmetry. This th < -lira! expectation is in part borne out by the facts.

With a view to this question Kirnwu.i>:; examined the evidence as to thoracopagous double monsters (including xiphopagi, Ac. |, and found that in ahm»t every case one of the bodies shew( d some transposition of viscera, though to a varying extent4. There are nevertheless a few cases even of thoracopagi where neither body exhibits any transposition5. Moreover, contrary to natural expectation, it does not appear that in ordinary cases of completely separate twins either twin has its viscera transposed; and conversely, of 152 cases of transposition collected by Kiicheiimeister only one could be shewn to have been a twin". It seems therefore that the frequency of transposition in double monstrosity depends in some way upon the iiKiiiiii-niince of the connexion between the twins; and that if the separation be completed early, as it must be supposed to be in cases of homologous twins born separate, then both bodies as a rule develop upon the normal plan, like the bodies of multiple births of other animals. But as the evidence now stands there is no tea son to suppose that individuals with transposition of viscera, born as single births, have ever had a counterpart any more than individuals whose viscera are normally placed, tempting as it is to imagine that both may have had some counterpart which in the ordinary course does not develop.

For the present we need not go beyond the fact that between complete duplicity resulting in " homologous twins," and the least forms of axial duplicity, consisting in a doubling of either extremity of the longitudinal axis almost all possible degrees have been seen7. By persons unfamiliar with abnormalities it 1 Kichwald supports the view that in these cases it is the right twin which shews the transposition. As KUCHENMKISTEU (/. c.) points out, this cannot by the nature of the case be a universal rule; for the relative position of xiphopagous twins may result simply from the way in which they happen to be laid by the mother or the midwife. Of the Siamese Twins, besides, it was Chang, the left twin, in whose body there were indications of transposition. The twins may also remain face to face. The expression " right twin " must always need further definition, and it should be qualified as the right when the livers are adjacent, or when the hearts are adjacent, as the case may be. Whether the rule is wholly or partially true for either of these positions seems to be very doubtful.

7 The fact that some of the degrees are much more common than others has an obvious bearing on the question of Discontinuity, which might with profit be pursued. A statistical examination as to the angles at which the bodies are most frequently inclined to each other would also probably lead to an interesting result. is sometimes supposed that axial duplicity is a phenomenon more or less peculiar to Man and to domesticated animals [and plants], and the occurrence is looked on as a part of that Meristic instability which is ascribed to absence of the control of a strict and Natural Selection. This view is far from sound. Such phenomena have on the contrary been found in many classes of animals, vertebrate and invertebrate, and the unquestionable frequency in domesticated animals may in great measure be fairly attributed to the comparative ease with which the births of these creatures can be observed. As considerations of this kind have weight with many it has seemed worth while to give references to examples taken from a variety of different groups, shewing not only that such compound bodies may be produced in wild animals, but also that they may sometimes be able to carry on the business of life without artificial help.

In Mammals and Birds I do not know an authentic case of a double monster that had groii-n -up in the wild state. *8G5. In Reptiles many such cases are known and are referred to by most of the older writers. Of Snakes having complete or partial duplicity, nearly always of the head, some twenty cases are recorded. Several of these were animals of good size, aucl must have had an independent existence for some considerable time. Some of the cases have special points of interest, but into these it is not now proposed to enter. As bearing on the question of the frequency of Meristic Variation in families and strains attention is called to the circumstance that MITCHILL'S three specimens were all found in one brood of 120 which were taken with the mother. The following is a list of records of snakes having the head wholly or partially double.

Pelamis bicolor. [Remarkable case1 : the duplicity appearing only in the fact that there were 4 nasal plates instead of 2, each with a nostril] BOETTGER, O., Ber. iib. d. 8enck, nat. Ges. in Frank/, a. J/., 1890, p. LXXIII. 1 Compare with MitchilPs two last cases, and also with a case in Alytes obstetricans. HEEOX-EOYEB, Bull. Soc, Zool. France, 1884, ix. p. 104. Fio. 2()7. Cliri/seniy* jiictn, 2 or 3 days old. I, II, normal. Ill and IV, twoheaded specimen. In the latter the nuchal and two pygal plates are normal. Be- tween them are 12 plates on each side, 11 being the most usual number. Among the costals an extra plate is wedged in on the rt. First vertebral divided by suture; fifth is made up of 4 irregular plates. In the plastron there is a doubling of the gular plate. The rt. femoral has a suture. (From Barbour.)