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

Supernumerary parts having the structure of wings have been occasionally recorded in Lepidoptera, but their occurrence is exceedingly rare. In a subsequent chapter detailed evidence will be given respecting supernumerary legs and other of the jointed appendages of Insects and it will be shewn that in very many and perhaps all of these cases the supernumerary parts constitute a Secondary Symmetry within themselves (see p. 90). Extra wings however are of a different nature altogether, and there is so far as I am aware no indication that any of their parts are disposed as a Secondary Symmetry. In other words, an extra wing if on the left side is a left wing, and if on the right side a right wing.

In some cases the extra wing is a close copy of a normal structure, in others it seems to be more or less deformed. No genuine case of an extra wing present on both sides of the body is known to me. From the fact that no specimen of supernumerary wing has ever been properly dissected, it is not possible to make any confident statement as to the attachments or morphology of such parts. (See also No. 78.) The cases of X. carpini, No. 422, and of Bombyx quercus, No. 12!i, nevertheless sugge-t that Variation in number of wings is of the s;,nic nut un- a- that seen in teeth, digits, or other parts -landing in a Meristic S.-ries. In the specimen of S. carpini it is especi.-dly aoticeable that on the side having three wings, both the wiir^s formed as secondaries were smaller than the secondary of tinnormal side; but in other cases, G. rhamni (No. 4'27 ! tor instance, this was not the case, and the wing standing m-\t to the extra wing was normal. Both these conditions aiv freejiiently fnund in cases of the occurrence of supernumerary parts in Aeries: for two members of a varying series may clearly correspond jointly with a single member of the normal scries, or on the contrary a new member may stand adjaeent to members in all respects normal as in G. rhamni (No. 427.)

422. Saturnia carpini $ , having a supernumerary hind wing. The specimen is rather a small female. The right wings and the left anterior wing are normal, but in the place of the left posterior \\in-, there are two rather small but otherwise nearly normal posterior wings. Of these the anterior is rather the larger and to some extent overlaps the posterior. The costal border of the posterior wing is folded over a little so that its width cannot From the fact that the bases of these two wings are greatly overgrown with hair, it is difficult to distinguish their exact points of origin from the body, but so far as may be seen, the second arises immediately behind and on a level with the first. The neural ion of each of the two small wings is identical with that of a normal hind-wing. The scaling is perfect on both surfaces of both wings, but is perhaps a little more sparse on the anterior of the two abnormal ones. In colour the anterior abnormal wing is rather light, but the posterior one is identical with that of the other side. The markings on each of the wings are normal, butare on a reduced scale in proportion to the size of the wings. This is especially remarkable in the case of the ocelli, which are both of a size greatly less than that of the ocellus of the normal hind wing of the right side.

The two wings were in every respect true left hind-wings and were in no way complementary to each other. [Specimen in collection of and kindly lent by Dr MASON.] 423. Bombyx rubi $ : 5th wing on left side. The additional wing was placed behind the left posterior wing. It was of normal Structure as regards scaling and coloration. Its length was that of the hind-wing but in breadth it did not exceed 6 mm. The insertion of this wing into the body was immediately above that of the normal hind-wing. The extra wing bore 4 nervures, of which 3 reached to the margin but one was shorter. The proper hind-wing of the same side was rather narrower than that of the other side and was not so thickly covered with scales, but its neuration was complete and normal. SPEYER, A., Stettiner Ent. Ztg., 1888, XLIX. p. 206.

424. Samia cecropia ^, having a fifth aborted wing. Bred in captivity : ordinary size, expanding about 5| inches : a smoky variety in which red portion of transverse bands on wings is much narrowed. Right primary and both secondaries normal in shape and marking. Left primary in length from base to apex exactly the same as the right, but in width from inner angle across to the costa is —% of an inch less; the markings are the same, but condensed into the narrower space. Neuration normal in all wings. Left primary also somewhat narrower at base, where it joins the body. The inner margin is in exact line with its fellow; hence the costal line of the left primary is somewhat posterior to that of the right primary. The supernumerary wing emerges from the side of the collar and runs parallel to the normal left primary. It consists mainly of the costal and subcostal nervures, a small part of the median nervure and a strip of wing about i inch wide which was much curled in drying. The supernumerary wing is in no way connected with the normal one.

[The author regards this supernumerary wing as a repetition of the anterior part of the left primary wing.] STRECKER, H., Proc. Ac. Sci. Philad., 1885, p. 26. 425. Limenitis populi, having four normal wings and a fifth wing behind the left posterior one. This supernumerary wing was 20 mm. long and 9 mm. wide. It slightly overlapped the left secondary and was attached to it for a length of 12 mm., but its outer end was free. It is described as exactly resembling the part of the secondary which bears the three anterior nervures, and it is stated that both surfaces were normal as regards scales and colouration. ROBER, J., Correspondenzbl. d. ent. Ver. "Isis" z. Dresden, 1884, i. p. 31.

426. Vanessa urticae, having an additional hind-wing on the right side. This structure is inserted into the thorax dorsal to and between the two normal wings. It is shorter and of about i the width of the normal hind-wing. In colouring it is a close copy of the anterior third of the hind-wing. WESTWOOD, Trans. Ent. Soc., 1879, pp. 220 and 221, Plate. [Now in Brit. Mus.] 427. Gonepteryx rhamni with additional imperfectly developed hindwing on the right side. In this case the normal right hind-wing is only about two-thirds of its normal size. It overlies the additional hind-wing. The latter is coloured like the normal wing and bears an orange spot. From the neuration of the two wings Westwood considered that the supplemental wing contained missing parts of the normal wing.

Only two legs existed on the side of the abnormal wing, but for fear of injury the specimen was not sufficiently examined to shew whether the missing leg had been broken off or whether the extra wing was in its place. WI>T\V<>OD, ibid., p. 220. A specimen of G. rhuinni having live wings was caught at Brandon, Norfolk, in An-. 1*7.'* l.y Mr J. 'Woodgate, and exhibited to the Ent. Soc. by Prof. Mcldola, 1'rof. Knt. Soc., 1*77, p. xxvi. A similar specimen of this species was bought at Sti-vens'> auction-rooms and exhibited to Linn. Soc. by Prof. C. Stewart, in April, Ls'.il. This specimen is now in Mus. Coll. Surg. Whether it is the same as that taken by Mr \Yoodgate, or that described by Westwood, or not, I cannot say, but possibly the references are all to one individual.

42s. Lycaena icarus </. A coloured figure is given of a specimen of this f.irin with 5 wings from Taurus, Asia Minor. [No further il' •-'•! -i|>t !<>M is given. The figure is not very clear. It shews however that all the wings are normal except the right anterior. This wing is represented by two wings, which together are about a third wider than the normal wing. The costal portion of the foremost of these wings appears to be nearly normal in neuration, and the posterior part of the hindmost seems to be also normal. The two taken together shew several supernumerary nervures a- ei.nipared with the normal wing, but the details are not shewn with sufficient clearness to justify a more precise statement.] HONKATH, K. G, Jkrl. Ent. Ztschr, xxxn. 1888, p. 498, Taf. vn. fig. <>.

429. Bombyx quercus $ : specimen having 5 wings figured in colour by HONUATH, with statement that the left anterior wing shews a double structure. [No further description given. The figure shews the left anterior wing represented by two wings. Of these the posterior appears to represent a nearly complete anterior wing on a reduced scale. It bears the white ocellar mark of the anterior wing. The pale-yellow submarginal band is curved inwards over the ocellus upon the costal border as in a iKirmal wing and thus shews that the foremost wing is not merely the scp:ir;iicd costal part of this wing. The foremost wing is anomalous. Its central half is rather darker in colour than that of the normal wing and its peripheral half is pale in colour, decpi-ning towai'ds the margin. It bears no ocellus. The neuration s cannot be made out from the figure with precision but the two wings together contain many more nervures than the normal anterior wing. The legs are not described.] HONRATH, E. G., ibid., fig. 10.

430. Zygacna minos, having a fifth wing on the left side, inserted above and between the normal wings. The Dentation of this wing is peculiar. The colouring of the supernumerary \\ing was that of the anterior wing. [Dr liogenhofer kindly informs me that thr legs \\crc normal.] I\OUI:MIOI i.u, A., Sit:.-H<'r. d.zool.-bot. OYs. JJ'iY/i, In the same place the following instances of fivewinged Lepidoptera are given : Orthosia lacvis with an additional posterior wing on the left side, in the Mu-eutn of I'esth. TKKITM IIKE, Bd. vi. Abth. n. p. 407. 432 Pygaera anastomosis with a wing-like appendage to the left anterior wing in

433. Naenia typica with an additional posterior wing in the collection of NEUSTADT at Breslau. 434-. Crateronyx dumi with five wings in the collection of WISKOTT in Breslau. 435 Penthina salicella : left fore-wing about £ wider than the normal right fore- ' wing. The apical border was markedly ernarginated, giving it a bilobed appearance. The nervures were as in the normal wing, except that the cells between the branches of the subcostal nervure were enlarged. ROGENHOFKH, ibid. [I am indebted to Dr

[Palloptera UStulata (Diptera): specimen having a large upright scale on the thorax. This abnormal structure is like a third wing in appearance, and is fixed on the thorax, passing from the head, backwards between the wings. Its upper border is circular, and in all respects it resembles the upper wing-scale of one of the Calypterous Muscidas. GERCKE, G., Wiener Ent. Ztg., 1886, v. p. 168.] 436. Sheep. Repetition of the horns in sheep is well known. The best account is that of H. VON NATHUSius1 of which the following is chiefly an abstract.

Commonly there is a pair of extra horns placed externally to the usual pair, but there may be three pairs in all, and even higher numbers are recorded, though Nathusius had seen no such case. The numbers on the two sides may be different, two on one side and one on the other, and three on one side and two on the other being sometimes met with. It is noticeable that in all cases the horns stand in a transverse series, and not in a longitudinal series as they do in the Fourhorned Antelope (Tetraceros quadricornis). The bases of the horn-cores are generally in contact, standing one outside the other at the same transverse level on the skull. Nathusius observed that in development the outgrowth for the horns of one side is at first single, but afterwards divides into two or more points, but he surmises that the division may appear earlier in other cases.

The external horns are generally smaller than the internal ones, but this is not universal. In some cases of two pairs of horns a small fifth horn is placed between the external and internal horns of one side. In another form of double horn the horn-core of one side or other may be a double structure, both cores being enclosed in a single horn, which on being separated has a double-barrelled appearance. Several examples of permanently four-horned breeds occur in various localities, being described as common in Cyprus and notably in Iceland and other northern islands. YOUATT (p. 169) stated that there were two breeds of sheep in Iceland, the one small and the other large, and that the greater part of both breeds

had more than two horns, some having eight. I am informed however by Mr E. H. Acton, who has spent some time in the country, that many-horned sheep are by no means common in Iceland at the present day. In Kishtwar (district of'S.E. Kashmir) a breed of 4-horned sheep is carefully preserved, in which the horns anas a rule very symmetrical, somewhat resembling No. 4381. Natlmsius states that a four-horned ram does not always beget four-horned offspring even when the ewe has the same character, and the valuation between father and son in respect of horns is frequently considerable.

'I'll.' bi-st figures of many-horned sheep are those given by lln-Tox, Hist, nut., Vol. xi. 1'ls. :-Jl and 32 (3-horned and 4-horned); YOUATT, The Xh<cp. \>\>. 141 and 171, cojiinl t'rnni r.riruN. Numerous other figures are referred to by Xathusius, but fuw of them are satisfactory. 437 Goat. A family of goats on an isolated farm near Bozen had -t horns, which had been inherited for many generations. In most cases thf two ordinary horns were typical in shape and direction; and in addition to these there were two lateral ones, which were laterally curved, being sickle-shaped and bent into a semicircle. GREDLER, V., Kurrespondenzbl. d. zool. min. Ver. Regensburg, 1869, xxin. p. 35.

*4,3S. Rupicapra tragus (Chamois) : skull bearing two well-formed and symmetrical extra horns. The cores of these horns were a little outside and posterior to the normal pair. ALSTON, E. R., P. Z. S., 4,30, Capreolus caprea ( Roebuck) : specimens having a supernumerary beam are probably not very rare, and a number of such antlers were shewn among the hunting-trophies exhibited by H. H. the Duke of Saxe-Coburg-Gotha, and H. S. H. the Prince of Waldeck-Pynnont at the German Exhibition held in London in 1891. The normal antler of the roebuck has a single beam rising vertically, then bifurcating, the posterior branch again dividing. In the abnormal specimens from the single burr of one side arose a supernumerary beam in addition to the normal one. In one specimen, in which the supernumerary beam was nearly as long as the normal one, the latter bifurcated as usual but was rather more slender than that of the other side (Fig. 75 I.). In another case (Fig. 75 II.), from the left burr, which was much enlarged, arose (1) an innermost beam, in thickness and texture resembling that of the normal right horn, though it was much shorter and bore no tine ; ('2) an external beam at once dividing into two almost equivalent branches having about the same length as the innermost beam. In such a case I know no criterion by which one of the three beams can be certified to be the normal to the exclusion of the others. As in the sheep and goats, the several horns resulting from subdivision seem to be generally in or nearly in the same transverse plane.

FIG. 75. Abnormal horns of Roebuck (Capreolm caprea), No. 438. by rne the horns were fixed upon heads modelled in plaster.) Haliotis gigantea (Japan) having two rows of perforations in the shell. In addition to the ordinary row of perforations, of which 12 were present in this specimen, there was a series of 8 additional perforations which began within an inch of the apex. Of the normal series the last four remained open, but all the perforations in the abnormal row were closed with nacre. Specimen in Brit. Mus. SMITH, E. A., Ann. and May. of N. H., 1888 (1), p. 419.

Haliotis : two specimens, of different species, in which the perforations were entirely absent, their place being taken by a continued convex, spiral rib, like the second rib of Padollus. "Probably in this individual the mantle was without any slit, and hence the malformation, the water being admitted to the gills by the slight notch in front of the ribs, as in some Emarginulce, or Scuta." (JKAY, J. E., Proc. Zool. Soc., 1856, p. 149. **• albicans : several specimens in which the perforations were united to form a continuous slit. The appearances were so uniform that Gray was disposed to think that these specimens might represent a new genus, but on comparison with types they seemed to belong to the species named. In some fossil genera (Scissurella) the perforations are replaced by a more or less continuous slit over the mantle. The specimens in question were greatly eroded and had a diseased appearance, ibid. Plate.

UP<'\ the bodies of animals belonging to many classes are markings which consist of a central patch of colour surrounded by a variable number uf concentric rings of different colours. Such markings arc known as ocelli or eye-spots from their resemblance to the pupil and iris of vertebrates. Eye-spots are perhaps best known in Lepidoptera, but similar markings are not unfrequent in other groups and especially on the feathers of Birds and in Fishes.

In one of the best known chapters in the Descent of Man' the nature and mode of evolution of these markings is the subject of a full discus-ion, the case of eye-spots on feathers being chiefly r.-ikeii in illustration. As is well known, Darwin by the comparative method, comparing the eye-spots found in different species, on the different feathers of the same bird, or on different parts of the same feather, tiiiind that it was possible to construct a complete progression from a plain spot to a fully-formed ocellus. Though no one examining such a series can possibly doubt that the simple -pot and the fully-formed ocellus are really of the same nature and that the one represents a modification of the other, there remains nevertheless the difficulty that members of a series of parts cannot be assumed to represent conditions through which the other mem- 1" rs of the same series have passed, and it is of course clear that the conditions found in some forms do not necessarily correspond with pliylogeiietic. phases of other forms. In the present instance however Darwin is not specially urging this view, but brings for wan I the comparative evidence chiefly in illustration of the possibility that such structures may exist in an imperfect state and so may be conceived of as having had a gradual origin.

1 The evidence concerning eyespots of Lepidoptera is taken here because eyespots when rrpi'iitr.l in MTH s, though borne on appendicular parts, are nevertheless arranged chirlly with roft-n-nce to the chief axis of symmetry of the body. In some ft\v forms, !•. ;i. 'I'lii/iii-tif, there is a conspicuous Minor Symmetry within the limits uf ;i .-ingle wing (tinposterior), but this is not often the case. Though doubtless the eye-spots of Birds are in their nature not different from those of Lepidoptera yet their manifestations in the latter are usually in some respects simpler than they are in Birds. From the abundance of material also the Variation of eyespots is most easily studied in Lepidoptera and it is to them that the present evidence chiefly relates.

In preface to the evidence a few remarks are needed to direct attention to certain features in the mode of normal occurrence of eye-spots and in the manner of their Variation. On a survey of the facts it is at once seen that eye-spots are extraordinarily variable both in number and size, some of the best formed being occasionally absent, and large and perfect ocelli being sometimes added in situations having normally no trace of such marks. With this fact Darwin was well acquainted and he refers to observations in illustration of it. In speaking of Gyllo leda he concludes that from the great variability of the eye-spots " in cases like these, the development of a perfect ocellus does not require a long course of variation and selection ; " and again, that bearing in mind " the extraordinary variability of the ocelli in many Lepidoptera, the formation of these beautiful ornaments can hardly be a highly complex process, and probably depends on some slight and graduated change in the nature of the tissues." The facts to be given and the circumstances attendant on the variation of ocelli tend to support this conclusion.

Considered from the point of view of Meristic Variation the chief feature in the manner of occurrence of eye-spots in Lepidoptera is the frequency with which they are repeated. A single spot may be repeated in homologous places in both pairs of wings ; in other cases there is a series along the margins of one or both wings. Besides the repetitions thus occurring it is especially worthy of notice that ocelli are very commonly repeated on both surfaces of the wing (Satyridas, &c.), the centres of the upper and lower ocelli coinciding. It need scarcely be remarked that this effect is not produced by transparency of the wing-membranes and scales, but is an actual repetition, the scales of both surfaces being so coloured as to form an eye-spot on each side, the two having their centres coincident. In some cases, e.g. Saturnia carpini (the Emperor Moth), the rings and centres of the upper and lower ocelli have nearly the same colouring, but in the majority e.g. Pararge megcera (The Wall), Erebia blandina, &c., the upper and lower spots, though coincident, have quite different colours. In considering the Variation of the spots these facts as to the repetition of the spots should be remembered, for, as has been often insisted on in other cases of repetitions, we are concerned with the evolution of the series and not of one member only. Here therefore regard must be had to the degree of correspondence between the variations of the eye-spots in the fore and hind wings, on the

upper anil lower -uriace> of the same wing, in the several eye-spot.- aloiiM- the margin of th'- -aimwing, or in all of these, as the case niav 1)'-. 'I'll'' evidence \\ill >hew that there is sometimes a close correspondence between the variations of eye-spots in these several positions. Hi it tin nigh these are the matters with which we have now tin' in.. iv din-ct concern it will be convenient to speak at the same time more generally of eye-spots. It should be remembered first that there are eye-spots of various complexity. In the simplesl all the band*, are circular, having one centre; the ocellus is then ml.. complete in one cell of the wing, though sometimes tin outer /ours of colour overspread parts of the adjacent cells. In some cases th.- spot is double, having two centres, the bands being di>pos,-d round them in an hour-glass shape. As to the visible structure of e\v— | i..ts it can be seen with the microscope that the colour of the eye-spot lies in the colours of the scales. The scales an arranged in parallel rows running (with little crossing or anastomosing) as 1 1< 'arly as possible at right angles to the nearest nervures, being disposed in regard to them much as the circular threads of a cobweb are in regard to the radial threads. Across these rows of scales run the colour-zones, in no way limited or guided by them. On the other hand it can be seen that the patterns are almost wholly made up by the colours of single scales, each having its n\\n colour, particoloured scales being exceptional. The effect thus seen is very like that of a mosaic picture made of similar pieces, or of a design worked in cross-stitch on canvas, all the stitches being in rows and each stitch having its own colour.

As regards the position of eye-spots it should be noticed that the simpler sort, e.g. those of Morpho or of Satyridoe, are usual/;/ jilni',',1 /// xin-li <i iinxitinn flint eni'li if their centres is on the line of one of the creases or fnlfl-murlcn of the winy, and it sometimes happens th: it these creases seem to begin from the centre of an ocellus. 1-Yoin the fact that the creases for the most part run e\. nlv between two nervures, bisecting a cell, it commonly results that the centre of the eye-spot is exactly halfway between two n.Tvures. The large spots on the hind wings of some Pieridas, e.g. Parnassius <i/i<iflo, are an exception to this rule.

In that cell of the hind wing which lies between the submedian and first median nervures in many ocellated forms (Satyrida?, Mnrftlio, &c.) there are two creases, and it is especially interesting to notice that in this cell there are commonly two ocelli, one on each civase; but if there is only one ocellus its centre does not correspond with the middle of the cell but is nearer to the first median nervuiv, being placed exactly on the anterior of the two creases. In spite of the excessive variability of ocelli, in for instance Satyrid;e, it appears that they are not formed in situations other than these, being so far as I have seen always on one of the creases '. 1 These remarks refer to simple ocelli with one or more definite centres.

On looking at such a series of repeated ocelli as those on the hind wing of Pararge megcera, from this fact that the ocelli are on these creases or folds the question naturally arises whether the wing may not have been, in its development, folded along these creases so as to bring the ocelli into contact with each other like the fold-edges of a fan. If this were the case it might be supposed that the repetition of the ocelli was due to the action of some one cause on all the folded edges together. As a matter of fact, however, so far at least as can be judged from the condition of the wings in the pupal state before scales or pigments are excreted, there is no such folding, but each wing is laid smoothly out, and the increase in extent of the wings of the imago is attained, not by a process of unfolding, but by a stretching of the elastic wing-membranes on inflation from the trachea?. On the whole it does not seem likely that the repetition of similar eye-spots on the Lepidopteran wing arises in any way more immediately mechanical than that by which other repeated patterns are elsewhere formed on animals.

The Variation of eye-spots as already stated may be very great, and examples are to be given both of the total absence of large eye-spots present in the normal, and of the presence of perfect eye-spots in abnormal places. Besides these extreme cases there is immense Variation in the degree to which eye-spots are developed, and such variability is nearly always to be seen in any species possessing simple ringed ocelli. In the manner of Variation of ocelli the following things are noteworthy.