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

1 specimen lias m- -- = ; in upper jaws m5 very small in crypts, but :!!»i. 1 specimen has m — m* being minute and lying in crypts. Lagorchestes. In tlii- ^enns ///•• is present and is a large tooth, not materially smaller than //* :;. Nc\»-rtheless it commonly falls short "f 'lie nther teeth and remains partly within the jaw. This was the case in 1<> skulls of /,. 1,'1,,,,-oiilfx and L. i-t>itxj>!r!//<ifntt. In one skull <-f L. /••jini-niil' x, ///' stood at the same height as the other teeth. I see no reason to suppose that all the other skulls were voting, and it seems more likely that this imperfect eruption of m4 is characteristic.

Some features characteristic of Meristic Variation are well seen in the case of the teeth of Sharks and Rays. Of these fishes there are many having little differentiation between the separate rows of teeth. In these a distinct identity cannot be attributed to the several rows, and numerical Variation is quite common. But besides these there are a few forms whose teeth are differentiated sufficiently to permit a recognition of particular rows of teeth in different specimens, and to justify the application of the term "homologous" to such rows. Nevertheless with such differentiation Meristic Variation does not cease.

In the following examples it will be seen further that in such Variation there may be not merely a simple division of single teeth but rather a recasting of the whole series, or at least of that part of it which presents the Variation, for the lines of division in the type may correspond with the centres of teeth in the variety. These cases also exemplify the fact that variations of some kinds are often only to be detected when in some degree imperfect ; for if the divisions in No. 396 for instance had taken place similarly on both sides, it would have been difficult to recognize that this was a case of Variation.

*396. Rhinoptera jussieui (= javcmica) : specimen in which the number and arrangement of the rows of teeth is different on the two sides, as shewn in Fig. G9, upper diagram. The disposition on the right side of the figure is normal, that on the left being unlike that of any known form. Specimen in B. M. described by SMITH WOODWARD, Ann. and Mag. N.H., Ser. 6, vol. i. 1888, p. 281, fig. As Woodward points out, the rows of plates on the left side may be conceived as having arisen by division partly of the plates of the central row and partly from the lateral row, marked I. But if this be accepted as a representation of the relation of the normal to the abnormal, in the way indicated by the lettering, the plates of the row marked 0 b, for instance, must be supposed each to belong half to one rank and half to a lower rank. The same applies to the plates in the row I b. By whatever cause therefore the points of development of the teeth are determined, it is clear that the centres from which each of the teeth in the rows I b and 0 b was developed were not merely divided out from centres in the normal places but have undergone a rearrangement also. With change of number there is also change of pattern.

The tessellation on the abnormal side is so regular and definite that had it existed in the same form on both sides the specimen might readily have become the type of a new species. There is indeed in the British Museum a unique pair of jaws in both of which (upper and lower) a very similar tessellation occurs in a nearly symmetrical way. This specimen is described as lil,;n,,i>t,,;i polyod&n, but it is by nu means unlikely that it is aetualK a Variation derived from the usual formula of Rhinoptera. It is figured by (Ji'-NTiiKi:. tftmli/ <>f Fishes, 1SSO, p. 34(3. Fig. 133.

*-'!!'7. Rhinoptera sp. ineerl.: teeth as in middle diagram, Fig. 69. On the left side ////•"' i-oNvs of small lateral teeth, while on the right siile t \\o of the^e ro\\ s are represented liy one row, \vhieh in one part of t Illseries s!ie\\ - an indieation of division. (.'. S. M. (llunt'-f'uni S]»'cime)i). Rhinoptera javanica : the row of teeth marked I is one side single, but on the other side is represented by two rows. Fig. 69, lower diagram. OWEN, Odontography, PI. 25, Fig. 2. C. S. M. (Hunterian specimen}.

399. Cestracion philippi : an upper jaw having the teeth disposed as in the figure (Fig. 70). C. S. M. On comparing the teeth of the two sides it will be seen first that the rows do not correspond individually, and secondly that they do not at all readily correspond collectively. Assuming that the rows marked 4 on each side are in correspondence (which is not by any means certain) several difficulties remain : for right 5th is larger than left 5th, but left 6th and 7th together are larger than right 6th; right 7th is about the same size as left 8th, but right 8th is larger than left 9th. The proportions in the figure were carefully copied from the specimen. 400. " Cestracion sp." [so labelled, but probably not this genus] : lower jaw as in Fig. 71. On the right side the second row of large plates is represented by two rows, properly fitting into each other, but on the left side the plates of the inner side are completely

divided, but tindi\i-i..n is gradually lost towards the middle of the ja\v ami tli.- .'Menial plates are without trace of division. C. 8. .M. l'i<.. 71. The lower jaw of a Selachian, No. 400. The proximal ends shewn (enlarged). The right is reversed for comparison with the left. The following e.N.-nnple of Meristic Variation in the teeth of a Mnlluscnn odmituphore may be taken in connexion with the subject of teeth, though the structures are of course wholly different in nature. For information on this subject I am indebted to the Rev. A. H. COOKE.

( leu. -rallv speaking the number and shapes of the radular teeth ait- \er\ characteristic of the different classificatory divi- There arc however certain forms in which a wide range of Variation is met with; of these the case of Buccinum inidatum is '401. Buccinum undatum. In most specimens the number of denticles "n ih<' central plate is ."> — 7 and on the laterals 3 — 4. In -1 -ecinicns from Haniiiierfest and Vardo the teeth were *402. The range of Variation may be still greater than this, the number of centrals being sometimes as low as 3. Fig. 72 shews the different conditions found. In it eight varieties are shewn,

I. Three centrals (Labrador). II. Four centrals. III. Five centrals, approximately symmetrical bilaterally. IV. Five centrals, not symmetrical ; the two external centrals on one side almost separate, correspond with a bifid denticle on the other side (Labrador). V. Six complete centrals (Labrador). VI. Seven centrals (Lynn). VII. Nine almost distinct centrals. VIII. Eight centrals; laterals asymmetrical (4 and 5). I. II. IV. — VI. from photographs made and kindly lent by Mr A. H. COOKE. III. VII. VIII. after FRIELE.

I. II. IV. — VI. being taken from Mr Cooke's specimens, III. VII. and VIII. from Friele's figures. As thus seen, in these variations considerable symmetry may be maintained. This symmetry and definiteness of the varieties in the cases with 3 and 4 centrals is especially noteworthy, inasmuch as these are abnormal forms and have presumably arisen discontinuously. As also seen in the figure, e.g. IV. and VI. this symmetry is not universal, and may be imperfect. The specimen shewn in VIII. is remarkable for the asymmetry of the lateral plates, which have 4 and 5 denticles respectively.

In connexion with the subject of symmetrical division interest attaches to cases like that shewn in Fig. 72, IV. in which on the outside of tincentral plate a pair of almost wholly separate denude-, nn one -ide r-f.nvspond with a large, imperfectly divided denude fit' tin- i.thi-r -ide. A very similar specimen is figured by KIM 1. 1. 1 Norake .Y"/v///-//W-,'./y,., vm. PI. v. fig. 16. The number found in one part of the radii la is usually maintained thruiighout the whole series, but this is not always so. A case in \\liidi the number of centrals at the anterior end of tin radula was G. and at the posterior end 8, is given by FKIELE, Norske Sordini c.^-K'j>., 1882, vm. p. 27, Taf. v. fig. 17.

IN this chapter I propose to speak of those matters which seem to have most consequence in the foregoing evidence as to the Variation of Teeth. Each of the following sections treats of some one such subject, specifying the cases which chiefly illustrate it. It will be understood that the sections do not stand in any logical collocation but are simply arranged consecutively. The treatment given is of course only provisional and suggestive, being intended to emphasize those points which may repay investigation.

The subjects which especially call for remark are as follows : (1) The comparative frequency of dental Variation in differ- (1) The comparative frequency of dental Variation in different The total number of skulls examined for the purpose of this inquiry was about 3000. From so small a number it is clearly impossible to make any definite statement as to the relative frequency of Variation in the different orders, but some indications of a general character may be legitimately drawn.

First, the statistics very clearly shew that while dental Variation is rare in some forms, it is comparatively frequent in others, but there is no indication that this frequency depends on any condition or quality common to these forms. Setting aside examples of the coming and going of certain small and variable teeth, the animals shewing the greatest frequency of extra teeth were the domestic Dogs, the Anthropoid Apes and the Phocidae. Attention is especially called to the fact that the variability of domestic animals is not markedly in excess of that seen in wild forms. From tluhypothesis that Variation is uncontrolled save bv Selection, there has sprung an expectation, now fast growing into an axiom, that wild animals are, as such, less variable than domesticated animals. This expectation is hardly borne out by t In- tads. It is true that, so far as the statistics go, supernumerary teeth were more common in domestic Dogs than in wild Canidae, and though the number of Cats seen was small, the same is true in their case also as compared with wild Felida1. But though it is true that the domestic Dog is more variable in its dentition than wild Dogs, it is not true that it is much more variable than some other wild animals, as for instance, the Anthropoid Apes or the genus I'liin-'i. Tindoctrine that domestication induces or causes Variation is one which will not, I think, be maintained in the light of fuller evidence as to the Variation of wild animals. It has arisen as the outcome of certain theoretical views and has received support from the circumstance that so many of our domesticated animals are variable forms, and that so little heed has been paid to Variation in wild forms. To obtain any just view of the matter the case of variable domestic species should be compared \\ith t hat of a species which is variable though wild. The greal variability of the teeth of the large Anthropoids, appearing n»t merely in strictly Meristic and numerical Variation, but also in frequent abnormalities of position and arrangement, is striking both when it is compared with the rarity of variations in the teeth of other Old World Monkevs and the com i>« rut ire rarity of great variations even in Man. If the Seals or Anthropoids had been domesticated animals it is possible that some persons would have

n in their variability a consequence of domestication. When the evidence is looked at as a whole it appears that no genera I i/at ion of this kind can lie made. It suggests rather that the variability of a form is, so far as can be seen, as much a part of its specific characters as ;iny other feature of its organization. Of such frequent Variation in single genera or species some curious instances are to be found among tile facts given. Of Canis cancrivorus, a S. American Fox, the majority shewed some abnormality. Of Felis fontanieri, an aberrant Leopard, two skulls only are known, both showing dental abnormalities. In Seals only four cases of reduplication of the first premolar were seen, and of these two were, in Ci/*t»j>!i<>r<i rrixt<tt<i. The number of cases of abnormality in the genus Ateles is verv large. Of six specimens of CriMsiirrliiix zcltrn, two shew abnormalities. Of the very few skulls of Myrmecobius seen, two shew an abnormal number of incisors. Three cases of Variation were given in Canis mesomelas, not a very common skull in museums. On the other

hand the rarity of Variation in the dentition of the Common Fox (C. vulpes} is noteworthy, especially when compared with the extraordinary frequency of Variation in the molars of S. American Foxes. The constant presence of the small anterior premolar in the upper jaw of Otters (Lutra) of most species, as compared with the great variability of the similar tooth in the Badgers (Meles} and in other species of Otters, may also be mentioned. The evidence given in the last chapter should not, I think, be taken as indicating the frequency of dental Variation in Mammals generally. The orders chosen for examination were selected as being those most likely to supply examples of the different forms of dental Variation, and it is unlikely that the frequency met with in them is maintained in many other orders.

With respect to bilateral Symmetry an examination of the evidence shews that dental Variation may be symmetrical on the two sides, but that much more frequently it is not so. The instances both of bilaterally symmetrical Variation, and of Variation confined to one side are so many that examples can be easily found in any part of the evidence. Besides these there are a few cases in which there is a variation which is complete on one side, while on the other side the parts are in a condition which may be regarded as a less complete representation of the same variation. Such cases are Ommatophoca rossii No. 320, Phoca grcenlandica No. 324, Dasyurus maculatus No. 385, Canis lupus No. 24-6, G. vetalus No. 248, &c.

In the remarks preliminary to the evidence of dental Variation, reference was made to a peculiarity characteristic of the teeth considered as a Meristic Series of parts. As there indicated, the teeth are commonly repeated, so as to form a symmetry of images existing not only between the two halves of one jaw, but also to a greater or less extent between the upper and lower jaws. It was then mentioned that cases occur in which there is a similar Variation occurring simultaneously in the upper and lower jaws of the same individual. Such similar Variation may consist either in the presence of supernumerary teeth, or in the division of teeth, or in the absence of teeth. It should, however, be noticed that examples of Variation thus complete and perfect in both jaws are comparatively rare. Speaking generally, it certainly appears from the evidence that similar Variation, (1) on one side of both jaws, or (2) on both sides of one jaw and on one side of the other, or (3) on both sides of both jaws are all rare. Of these three the following examples may be given :—

( )t these, further examples may be seen in the evidence given Among the cases of increase in number of teeth are many in which l.y ili.' appearances presented it maybe judged that two teeth in the varying skull represent one tooth in the normal, and have arisen l.\ the division of a single tooth-germ. < H such division in an incomplete form several examples have been given. The plane of division in these cases is usually at right angles to the line of the jaw, so that if the division were complete', tintwo resulting teeth would stand in the line of the arcade. Incomplete division of this kind is seen in the first pivmolar of UniiiKifnjiIinca rossii No. 320, in the fourth premolar of /'//or,/ iim-iil.inillca No. 324, in the incisors of Dogs No. 219, in the canine of Dog No. 221, in the lower fourth premolar of Das i/ a r us f/eo/ro>/i No. 383. The plane of division is not however always at ii-lit angles to the jaw, but may be oblique or perhaps even parallel to it, though of the latter there is no certain case. < lases of division in a plane other than that at right angles to the jaw are seen in ('. n/ljtes No. 230, Phalanger oriental-is No. 368, Pit <>ca yru'iil<iii<lic<i No. 326 and doubtfully in a few more cases. The existence of the possibility of division in these other planes is of some consequence in considering the phenomenon of duplicate teeth standing together at the same level in relation to that of the presence of duplicate teeth in series. Beyond this also it may be anticipated that if ever it shall become possible to distinguish tinforces which bring about the division of the tooth-germ, the ivlati..n of the planes of division to the axis of the Series of Repetitions \\ill be found to 1).- a chief element.

Teeth standing at or almost at the same level as other teeth which they nearly resemble may conveniently be spoken of as duplicate teeth, though it is unlikely that there is a real distinction of kind between >\\c\\ teeth and those extra teeth which stand in series. Duplicate teeth were seen in Felis domestica Nos. 286 and 2s7, t'anis mcsniHc/nx No. 22s. HcrjH'stt'ti ichneumon No. 300, [Pntoi-iiis] I'isuii Inirstii'lilH. No. 311, Helictis orientalis No. 312, ( 'i/sfii/il/nr/i i'1-istiit't No. 322, and perhaps in some other cases. That these cases are not separable on the one hand from examples of e\tra teeth in series may lie seen from Herpestes gradlis No. 2!»!>, rd crisfiifii No. 321 [compare with No. 322], Braclnjteles

hemidactylus No. 199 [compare with Ateles marginatus No. 200], Phoca vitidina No. 336 ; and that on the other hand they mergf into cases of supernumerary teeth standing outside or inside the series, and whose forms do not correspond closely to those of any tooth in the series, may be seen by comparison with Otaria ursina No. 325, Phoca vitidina No. 329, Phalanger oriental™ No. 372. Though in some cases the shapes of duplicate teeth make a near approach to the shapes of normal teeth, yet they are never exactly the same in both, and teeth whose forms approach so nearly to those of other teeth in the series as to suggest that they are duplicates of them and that they may have arisen by multiplication of the same germ, cannot be accurately distinguished from extra teeth whose forms agree with none in the normal series.

(5) Presence and Absence of Teeth standing at the ends of Series (first premolars, last molars): the least size of particular TeetJi. Of the cases of numerical Variation in teeth the larger number concern the presence or absence of teeth standing at the ends of Series. As was mentioned in introducing the subject of dental Variation, in many heterodont forms the teeth at the anterior end of the series of premolars and molars are small teeth, standing to the teeth behind them as the first terms of a series more or less regularly progressing in size. Not only in teeth but in the case of members standing in such a position in other series of organs, e.g. digits, considerable frequency of Variation is usual.

Variability at the ends of Series is manifested not only in the frequency of cases of absence of terminal members, but also in the frequency of cases of presence of an extra member in their neighbourhood. An additional tooth in this region may appear in several forms. It may be a clear duplicate, standing at the same level as the first premolar (e.g. Cat, No. 270). On the other hand, as seen in the Dogs (Nos. 232 and 233) there may be two teeth standing between the canine and (in the Dog) the second premolar. The various possibilities as to the homologies of the teeth may then be thus expressed. The posterior of the two small teeth may correspond with the normal first premolar, and the anterior may be an extra tooth representing the first premolar of some possible ancestor having five premolars; or, the first of the two premolars may be the normal, and the second be intercalated (see No. 224) ; or, both the two teeth may be the equivalent of the normal first premolar ; lastly, neither of the two may be the precise equivalent of any tooth in the form with four premolars, Of these possibilities the first is that commonly supposed (HENSEL and others) to most nearly represent the truth. But the condition seen in cases where there is an extra tooth on one side only, as in the Dogs figured (Fig. 42), strongly suggests that neither of the two teeth strictly corresponds with the one of the other side. Seeing that in such cases the single tooth of the one side stands often at the level

of the diastema on the other, it seems more likely that the one tooth balances or corresponds to the two of the other side, which may be supposed to have arisen by division of a single germ. On the other hand since the two anterior premolars found in such cases are not always identical in form and size, either the anterior "i- the posterior being commonly larger than the other, there is no strict criterion of duplicity, and it is clearly impossible to draw anv sharp distinction between cases of duplicity of the first premolar and cases in which the two small premolars are related to each other as first and second. These two conditions must surely pass insensibly into each other. If the case of the teeth is compared with that of any other Linear series in which the number of members is indefinite, as for example that of buds on a stem, the impossibility of such a distinction will appear. A good illustration of this fact may often be seen in the arrangement of the thorns on the stems of briars. For large periods of the stem both the angular and linear succession of the thorns of several sizes may be exceedingly regular; but it also frequently happens that a thorn occurs with two points, and on searching, every condition may sometimes be found bet \\eeii such a double thorn and two thorns occurring in series, having between them the normal distinctions of form or si/e. Very similar phenomena may be seen in the case of the strong dermal spines of such an animal as the Spiny Shark (Erliinorhinus .--/*///o.s-//.s-). These structures are of course from an anatomical standpoint closely comparable with teeth. In them, spines obviously double, triple or quadruple, are generally to be seen scattered among the normal single spines, but between the double condition and the single condition, it is impossible to make a real distinction.

The remarks made. as to the first premolars apply almost equally to the last molar. See Phoca vitulina No. 336, Mycetes niger N<». -<>i;. Man. MACJITOT, Annm. syst. dent., PI. v. figs. 4, 5 and (i. Canis cancrivorus Nos. 251 and 252, Crossarckus zebra No. 302. What is the least size in which a given tooth can be present in a species \\hidi sometimes lias it and sometimes is without it ? In other \\onis, what is the least possible condition, the lower limit of the existence of a given tooth ? This is a question which must suggest it-e|f in an attempt to measure the magnitude or Discontinuity of numerical Variation in teeth.

The evidence collected does not actually answer this question completely for any tooth, but it shews some of the elements upon which t lie answer depends. In the lirst place it is seen at once that the least size of a tooth is different for different teeth and for different animals. Considered in the absence of evidence it might be supposed that any tooth could be reduced to the smallest limits which are histologically conceivable ; that a few cells might take on the characters of dental tissue, and that the number of cells thus constituting a tooth might be indefinitely diminished. Indeed on the hypothesis that Variation is continuous this would be expected. Now of course there is no categorical proof that this is not true, and that teeth may not thus occur in the least conceivable size, but there is a good deal of evidence against such a view. The facts on the whole go to shew that teeth arising by Variation in particular places, at all events when standing in series in the arcade, have a more or less constant size on thus appearing. Within limits it seems also to be true that the size in which such a tooth appears has in many cases a relation to the size of the adjacent teeth and to the general curves of the series. For example in the Orang, the scries of molars does not diminish in size from before backwards, and extra molars when present are, so far as I know, commonly of good size, not wholly disproportionate to the last normal molar. The same is I believe true in the case of the Ungulates. In the Dogs however the series of lower molars diminishes rapidly at the back, and the extra molars added at the posterior end of the series are of a correspondingly reduced size. As presenting some exception to this rule may be mentioned two cases in the Chimpanzee, Nos. 178 and 181 and the case of Cebus robustus No. 19-i, in each of which the extra molar is disproportionately small.

The principle here indicated is of loose application, but speaking generally it is usual for an extra tooth arising at the ends of series to be of such a size as to continue the curves of the series in a fairly regular way. It would at all events be quite unparalleled for an extra tooth arising at the end of a successively diminishing series, as the Dog's lower molars, to be larger than the tooth next to it, and with the exception of cases of duplicate anterior premolars (see Dogs Nos. 232 and Cat No. 268) I know no such case. In these besides, the anterior tooth is very slightly larger than its neighbour, and it should be remembered that the first premolar, though the terminal member of the series of premolars, is not actually a terminal tooth.

Examples have been given of animals which seem to be oscillating between the possession and loss of particular teeth, the first premolar of the Badgers, p^ of some species of Otter, &c. In these cases we are not yet entitled to assume because in a given skull the tooth is absent, that it has never been formed in it, though this is by no means unlikely, but as already pointed out (p. 228), the fact of its presence or absence may still indicate a definite variation. Attention should be called to the case of Trichosurus vulpecula, var. fnliginosa No. 378, in which the first premolar is generally of good size if present, and there can be no doubt that it has never been present in those skulls from which it is absent.

Variation of unusual amplitude may be seen also in the molars df lii-tttni<ini No-.. .'>vi, &c., for while on the one hand the last or fourth mi ilar may In- absent, it may on the contrary be large and mav even l)i- succeeded by a fifth molar as an extra tooth. All i lies,- conditions were seen in looking over quite a small number <»f -pel-miens. ( 7 ) Il"/noeotic Variation in terminal Teeth ivlien a new member (']><>n tinremarks made in the last Section the fact here M. .tired naturally follow-. We have seen that there is a fairly constant relation between the size of extra teeth and that of the teeth next t«> which they stand, so that the new teeth are as it w.re, from the first, of a size and development suitable to their position. We have now to notice also that the teeth next to which the\- stand ma\ al-o undergo a variation in correlation with the presence of a new tooth behind them.