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

areas and external poriferous zones are like those of a normal ambulacrum ; but the poriferous zones which touch one another are fused together, with the pores irregularly arranged. The combined poriferous zones are not quite equal to the sum of two normal ones. The whole of this area, formed of the union of two ambulacra, projects as a ridge which is continued down the whole of the side of the shell. STKNVAKT, C., Jour. Linn. Soc., xv. p. 130, PI.

689. Hemiaster latigrunda : right posterior ambulacrum double, the two resulting ambulacra ;uv closely adjacent peripherally and a small inter. iinliiilacral area is formed between them in their more central parts. There are G oculars but no extra genital. GAUTHIER, /. c., tigs. 5 and -") ///.s-. 690. Hemiaster batnensis : right anterior ambulacrum double, the two ambulacra are in contact through all their length. COTTEAU, Pal. [For interesting evidence as to variation in the number of genital pores on the costals in several genera of Echini, see LAMBERT, Bull. Soc. Yonne, 1890, XLIV. Sci.

nat., p. 34; also GAUTHIER, Comptes rftidus Ass. fr. pour I'av. Sci., Toulouse, 1887, and other references given by these authors.] Individuals with various numbers of arms are often seen, especially in the genera Ophiothela, Ophiocoma, Ophiacantha and Ophiactis, and in many of the species there are most usually six arms. In these forms the evidence as to Meristic Variation is complicated by the circumstance that in several of them change in the number of arms may take place in the ontogeny, by division and subsequent regeneration (see note on p. 433).

OF the organs repeated in Linear Series whose variations have been illu>t rated, many are bilaterally repeated also ; but thus far we have considered them only in their relations as members of Linear Scries. It now remains to examine the variations which they exhibit in virtue of their relation to each other as members of a Bilat'-ral Series. Meristic Variation in this respect is manifested in two ways. A normally unpaired organ standing in the middle line of a bilateral symmetry may divide into two so as to form a pair of organs ; and conversely, a pair of organs normally placed apart from each other on either side of a middle line may be compounded together so as to form a single organ in the middle line.

In animals and plants nothing is more common than for different forms to be distinguished from each other by the fact that an organ standing in the middle line of one is in another represented by two organs, one on either side. The facility therefore with which each of these two conditions may arise from the other by discontinuous Variation is of considerable importance. Admiral ile instances of the bearing of this class of evidence upon the question of the origin of Species are to be seen in zygomorphic flowers. Veronica for example differs from the other Scrophulariacese especially in the fact that it has only one posterior petal, instead of two posterior petals one on each side of a middle line. But there is evidence not only that forms having normally two posterior petals may as a discontinuous variation have only one such petal, placed in the middle line, hut also that the single posterior petal of Veronica may as a variation be completely divided into two. Similarly the single anterior petal of Veronica may also as a variation be divided into two, thus giving three posterior and two anterior petals as in for example SalpiffloBsis1, In these cases, which might be indefinitely multiplied,

1 An account of several discontinuous variations in the structure of zygomorphic corollas was ^ivi-n by Miss A. BATESOX and myself. Jour. Linn. Soc., 1892, xxvin., there is thus a clear proof that so far as the variations in number and symmetry are concerned, the transition from the one form to the other may be discontinuous. Analogous phenomena in animals are so familiar that general description of them is for the most part not needed, and an account will only be given of a few less known examples both of union and of division of such parts. Besides these strictly Meristic Variations in the amount of separation between the two halves a few examples are introduced in further illustration of the relationship that subsists between the two halves of a bilateral animal.

In considering the evidence both of median union and of division it must be remembered that the germs of most of the organs in question are at some time of their developmental history visibly double, and that when organs that should normally unite to form single median structures are found double in older stages, this duplicity is strictly speaking only a persistence of the earlier condition. But to appreciate this comment it should be extended. For, in every animal in which at some period of the segmentation of the ovum, the plane of one of the cleavages corresponds with the future middle line, all median organs must in a sense be paired in origin, and the distinction between paired and median organs is thus seen to be only one of the degree or amount of separation between the symmetrical halves. Nevertheless the evidence of Variation bears out the expectation that would be formed on examination of normal diversities between species or larger groups both in animals and plants, namely that whenever structures are geometrically related to each other as optical images, instability may shew itself as Variation in the degree to which such parts unite with or separate from each other. It is remarkable that this instability appears as much in the case of organs bilaterally symmetrical about an axis of Minor Symmetry as it does in the parts paired about the chief axis of Symmetry of the whole body.

Examples of such Variation in bilaterally symmetrical parts of a Minor Symmetry have been already given in the case of the feet of the Horse and of the converse phenomenon in the feet of Artiodactyles (q.v.). A good illustration of the way in which duplicity about an axis of Minor Symmetry may pass into the unpaired condition is seen in the case of ocellar markings on bilaterally symmetrical feathers. By comparing different feathers on several species of Polyplectron, DARWIN found that it was possible to find most of the gradations between the complete duplicity shewn in Fig. 140, I. where each half of the feather bears an almost symmetrical

ocellus, and the partially confluent condition shewn in Fig. 140, II, which is not far removed from the state of the ocellus in the Peacock's tail-coverts, where the whole ocellus has no peripheial indentation and is very nearly symmetrical about the rachis of ill.- t'.-atli.-r, though eacb -if it> 'li' lives has no axis of s\ mim-try. Fie;. 140. I. Part of tail-covert of Pnlyplectron cliitiifiiis, with the two ocelli of nat. size. II. Part of tail-covert of Polyplectron malaccense, with the two ocelli )i;irtiiilly confluent, of nat. size.

Attention should be called to the fact that abnormal division along a middle line may in many cases represent one of two different phenomena which are not readily distinguishable. For when a normally >ingle or^aii is represented l»v two, standing on either side of a middle line it is often possible that there may be not only a division of the organ but a partial duplicity of the axis. These two conditions are of course morphologically distinct ; for in the case of division of the organ onlv, the two parts are still in symmetry about the original axis of Major Sviiimet rv of the body, but in the case of duplicity of the axis there are two equivalent axes of symmetry, about which each half is separately symmetrical But though this distinction is in a sense a real one it cannot be applied to cases of duplicity occurring in any organ whose halves assume a bilaterally symmetrical form when separate. For example in the case of the foot of the Horse, or of the ha'inal spines Arc. of Gold-fishes (r. infra), when division occurs, each of the two halves is only hemi-symmetrical, and this duplicity is no more evidence that the axis is double than is the ordinary double condition of the vertebrate kidney; Imt in the case of duplicity of the central neural canal in 31. -in for instance, or in the case of the tail-spine of LliiiH/Hfi described below, it is not clear that there is not a partial duplicity of the axK

of the organs which in a vertebrate stand in a median have been seen inoiv or less often in a divided condition. Examples of such division in the middle line were, I believe, first put together by GEOFFROY ST HILAIRE, and a very full collection of the evidence seen in Man is given by AHLFELD \ The organs most often divided are the sternum, neural arches, uterus, penis, &c., and of these, specimens may be seen in any pathological collection. Organs more rarely divided are the tongue2, epiglottis3, uvula4, and central neural canal5. The following are special cases of variation consisting in a median division.

*691. Cyprinus auratus (Gold-fish). The following account of the multiple fins of Gold-fishes in China and Japan is taken chiefly from Pouchet6 and Watase7. There is evidence to shew that these animals were first imported to Japan from China. Three distinct breeds of Gold-fishes are kept in Japan. The first, called "Wakin " has a slender body closely resembling that of the common carp. The second "Maruko or Ranchiu" has a very short body, being in some cases almost globular in shape and in it the dorsal fin is generally entirely absent. The head is usually disfigured by rough-looking protuberances of the skin which often attain a considerable size.

The third or "Riukin" has a short body with a rounded abdomen. Of all the breeds, this has the most beautiful tail which is very large and often longer than the rest of the body. Gold-fish breeders of the present day can freely produce the " Riukin" or "Maruko " from the " Wakin." Various intermediate forms between the above-mentioned breeds exist. In all gold-fishes, irrespective of the breed to which they belong, the tail-fin is, above all other parts, subject to the greatest variation. It is to be found in one of the following three states ;

(2) It may consist of two separate halves ; each of these halves is to all appearance a complete tail and the two tails pass backwards side by side, but are united dorsally at the point where they join the body. (3) The two tails thus formed are united by their dorsal edges to a variable degree and their lower edges may be bent outwards, so that the two combined tails come to be spread out into a three-lobed, nearly horizontal fin. Fio. 141. Caudal and aiial fins of Gold-fish (Cyprinus auratu*).

I. Normal tail, seen from side, r, dorsal lobe, d, ventral lobe. II. Abnormal form divided a- far as the notochord. c' r', two ventral lobes, d' d', two dorsal lobes. III. Abnormal form, tht> two ventral lobes, v' v', separate. IV. Penultimate vertebra of normal Carp (''. cm-pin). «..-•, neural spine, h.s, haemal spine. Y. Penultimate \<-it.'Kra of a Gold-fish with trilobed caudal fin. h'.s', double li,-i uial .spine. VI. I Marram of transverse section through region of anal fin of normal < luld-ti-h. VII. Similar section through a specimen having the anal fin il. milled. /..•.-, interha-mul spine. /. r, tin ray. n, bony nodule, i.s', f'.r', n', correspniidm" partdoubled. (After WATABB.)

variation. It is either median and normal ; or it may be distinctly are all stages of caudal and anal fins, intermediate between tinnormal and the completely paired states. Thus the tail-fin with its IO\\<T portion alone in a double state, or the anal h'n with cither its anterior or posterior portion double and the remainder -in-le. is of quite common occurrence. These different conditions of tintwo (ins coniliine in various ways in different indi\ iilnaN t hns ^i\ ing rise to manifold varieties of form.

This doubling of the tail-tin consists essentially in a longitudinal splitting of the morphologically lower lobe of the tail. The tir>i step in the process of doubling is seen in the case of s in which there is a slight longitudinal groove in the ventral margin of the tail-fin. This groove may be extended up through all the rays of the lower lobe of the tail, which then consists of two tails side by side. The small dorsal lobe, which lies above the notochord, is never involved in the process, but always remains single. There is therefore in this case no doubling of the axis of the body. Examination of the skeleton shews that in those fishes which have two tails the hsemal spines of the last three vertebra are longitudinally split1 and diverge to carry the two tail-fins (Fig. 141, V).

POUCHET lays stress on the fact that the size of each of the paired tails is greater than that of the normal tail of a Gold-fish ; but as Watase states that in the variety "Riukin" the tail may be as long as the body, it is clear that this hypertrophy may exist without any repetition. In cases where the anal fin is doubled the process is exactly the same, resulting from a longitudinal splitting of the rays of which it is composed. This may only affect the outermost parts of the fin or may be carried up further so as to divide the interha?mal spines, in which case the two anal fins arise from the body wall at separate points and diverge from each other.

POUCHET, who has extensively studied the history of Gold-fishes in Europe, believes that it is almost certain that those which were brought to Europe in the eighteenth century were all more or less of the double-tailed order. He refers especially to the figure given by LlNN^EUS2 representing the double-tailed form as a normal. POUCHET states that the evidence goes to shew that this anomalous race is not maintained in China by any rigid selection. He quotes a Chinese encyclopaedia to the effect that the double-tailed Gold-fish is found in running streams, and gives the evidence of KLEYN3, a missionary in China during the eighteenth century, who states that "In fluvio Sleyn Cyprini sunt qui caudam habent trifarcam et a piscatoribus Leid-brassen vocantur, quasi diceres aliorum Cyprinorum conductor es."

Though the duplicity of the haemal spines may be unaccompanied by other variations it should be noticed that the extraordinary "Telescope " Gold-fish not mifrequently has also the double tail-fin. In the Telescope Gold-fish the eyes project from the orbit to a greater or less extent, in the extreme form being entirely outside the head and attached by a small peduncle only. The various forms of abnormal Gold-fishes are generally to be seen in large quantities in the shops of the dealers in aquariums &c. which abound near the Pont Neuf in Paris. One of these dealers told me that he bred considerable numbers of them every year, and that in fish from the same parents there was little uniformity, many normals being produced for one that shewed any of the extreme variations. It is recorded that of the Gold-fish hatched in Sir Eobert Heron's menagerie about two in five were deficient in the dorsal fiu and two in a hundred or rather more had a " triple" [so. three-lobed as described above] tail-tin, and as many have the anal

1 It should be observed that there is no want of original union between the haemal spines, for these close in the haemal canal as usual. The phenomenon is thus altogether different from that of spina bifida in the neural spines. lin doubled. The deformed fishes were separated from the others but did not produce a greater proportion of varying offspring than the normals (Ann. -!/</<;. For a iMHj-'iiiucent series of plati/s illustrating the various forms of Goldfishes see liii.i.AKi.ux M SATU.INY, Hist. nat. den Dorades de la Chin<-, Paris, 1780. [In Brit. Mus. cnj.y text wanting ; I do not know if it ever appeared.]

Division of i/H-'li'iit structures in Coleoptera. The following list includes every case known to me. C92. Anisoplia floricola (Lain.): Algerian specimen having the epistoiiie (chaperon) completely divided into two parts in the middle line. Attention is called to the fact that this is a normal character in certain genera of Lamellicorns, for example, Diphucephalu and Tnca, FAIKMAIKE, L., Ann. Soc. ent. France, 1849, Ser. 2, vu. Bull., p. LX. In Coleoptera the pro-thoracic shield or pronotum is normally a single plate continuous from side to side. The following is a list of cases in which this structure was composed of two lateral parts. In Nos. 695 and 706 the division was not completed through the whole length of the shield. The two halves were in most cases symmetrical, but in Nos. 700 and 703 they were unequal.

As is shewn by No. 704 £c., there is in these variations more than a mere fault of union between two chitinous plates, for in t his case the adjacent or inner edges of the plates were beset with yellow hairs such as occur on the anterior and posterior margins of the normal pronotum. In No. 703 again the adjacent edges of the two plates are everted and form definite margins. 693. Melolontha vulgaris (Lam.), prothoracic shield consists of two symmetrical pieces which do not meet in the dorsal middle line. The prothorax is greatly reduced in length and the head consequently is almost in contact with the scutellum (Fig. 142, I). KUAATZ, G., Dent. ent. Ztscltr., 1880, p. 341, PL u. fig. 8.

FIG. 14'2. Mi'lnlonthti riil'inrix, the Cockchafer, two cases of division of pronotum. (After KRAATZ. ) 1 With these cases compare the following : Hydrobius fuscipes, specimen having ]>ronotum fonued into three lobes, one being central, and two lateral. The lateral lobes projected from each side as considerable expansions. KRAATZ, G.,Dcut. cut. Zttchr., 1889, p. 22'2, fig. 21. p. 57,Taf. i. Jirj. 2. 605. A male in which the pronotum was similarly divided, but

the division was not quite complete. DE LA CHAVIGXERLE, Ann. 698. Oryctes nasicornis <£ (Lam.): anterior part of pronotum divided into two parts by a longitudinal suture : posterior part of pronotum undivided. Head normal, ibid., PL V. fig. 7. 699. Onitis bison (Lam.): pronotum divided in the middle by a longitudinal suture, the lateral pieces being raised up. ibid. 700. Heterorhina nigritarsis (Lam.) : specimen in the Hope Collection at Oxford having the pronotum completely divided into two somewhat unequal halves, of which the left is the largest. The posterior angle of each of the pieces does not occupy its normal position, but lies internal to the outer border of the elytron. Owing to this disposition the mesothorax is exposed for a short distance on each side and for a considerable extent in the centre.

701. Attelabus curculionides (Rhyn.) : specimen of moderate size ; head, elytra and legs normal. Structure of prothorax peculiar in that the two lateral halves do not meet in the middle line, leaving betwixt them a membranous space. The prothorax is shortened and the head is pushed back into the thorax as far as the level of the eyes. The edges of the plates of the prothorax are well formed and properly finished. Scutellum present, but is not at all concealed by the prothorax. DRECHSEL, C., Stettiner ent. Ztg., 1871, xxxii. p. 205.

702. Chrysomela fucata (Phyt.) : Pronotum divided centrally into two parts, each of which is triangular. The parts of the head and scutellum which should be covered by the thoracic shield are thus exposed. KRAUSE, Stettiner ent. Ztg., 1871, xxxii. p. 137. 703. Telephorus nigricans (Mai.) : the pronotum is divided into two unequal halves. The left half is nearly twice as large as the right, and projects beyond the middle line, covering a part of the right side of the prothorax. The right portion is small and very concave. Both of these two parts of the pronotum are everted at their edges to form a definite margin. The margins are continued all round each piece, and thus two margins are adjacent in the contiguous parts of the plates. This specimen was kindly lent to me by M. H. GADEAU DE KERVILLE.

704. Carabus scheidleri: thorax dorsal ly covered by two completely separate and symmetrical plates, whose inner edges are beset with yellow hairs [as the anterior and posterior margins normally are]. The rest of the animal was normal. KRAATZ, G., form t\v-i triangular pieces which only unite at a single point. Th«- head is drawn back into the thorax. DUPONCHEL, Ann. Soc. pn-s, ni a deep final filiation both before and behind [description in a symmetrical manner. Looked at from the ventral surl'ace the met asternal plates are seen to be divided in the middle line by a deep depression so that the abdomen consists superficially of two luhe^ : these two lobes are united together in the last segment in which the metasternal plate is undivided. The two lobes are of equal size and the longitudinal depression which divides them is shewn in the figure to be regularly and symmetrically formed. The animal is otherwise normal. [No dissection was made.] BAUDI, L. V., Bull. Soc. Ent. Ital., 1877, ix., p. -I*}, fig.

709. ^ case °^ "double proboscis" is recorded in Sphinx ligustri. The specimen was a pupa, and through the juijial skin it could be seen that the two mandibles had uot unitril to form the single proboscis, but were divaricated. KHAATZ, l>eut. int. /.t.^-lir., issQ, xxiv., p. MI."., li-. 7K). Ascidians. Pn.f. W. A. llcrdman tells me that he has several times met with Asridiaiis having a supplementary lateral atriopore. lie retail Is this as a retention of a larval character, since in the young thenare t u o at riopores which in normal individuals afterwards unite dorsally.

711 Limulus polyphemus : lai ^e .specimen found at Fort Macon, N. havin<r a forked caudal spine ( Fi^. 14.'5). This variation is 712. Palamnaeus borneensis (Scorpion): specimen in which the terminal poison-spine was double, as shewn in Fig. 144. The two halves were not quite equal and there was no opening of a poison-gland on the shorter spine. This specimen, which is in the Brit. Mus. was kindly shewn to me by Mr R. I. POCOCK. 713. Chirocephalus $ : specimen having the generative sac with two horns instead of one. [Normally there is only one such horn which forms a median