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

These two antennas curve in opposite directions and are in all respects complementary to each other, forming a trne pair. The most anterior of them, r, is disposed as a rir/lit antenna, while the posterior, I, is disposed as a left. This specimen was taken by M. Albert Mocquerys, and was kindly lent to me by M. Henri Gadeau de Kerville. 789. Melolontha vulgaris $ (Lamellicorn): left antenna bearing a pair of supernumerary clubs. The extra pair arises from the second joint of the normal antenna, and they have their third joints united at the base. The relative positions of the extra clubs and the normal one are those marked VP in the Scheme. All these three clubs are perfect and of the same size, but each is a little smaller than a normal club. At the thoughtful suggestion of Prof. Howes this specimen was very kindly lent to me by Mr E. E. Green, and has been placed in the Museum of the Royal College of Surgeons.

790. Melolontha vulgaris : Lrinrht antenna bearing a supernumerary pair of clubs in Position P. For details see original, where a different and I think untenable view is taken] LEREBOULLET, Rev. et May. de ZooL, S. -2, in., l8ol,X'A 791. Melolontha vulgaris £, with a pair of supernumerary antenna? arising from the left antenna. [The figure shews that the proximal joint or scape was of abnormal thickness and had two peripheral articulations in the same horizontal plane. The anterior articulation bore a normal antenna. The posterior articulation bore a single large first funicular (2nd) joint which in its turn bore a pair of clubs in the same horizontal plane, the anterior being a right club and the posterior a left, having their anterior surfaces adjacent : they are therefore a complementary pair in Position P.] KRAATZ, G., Dent. ent. Zt., 1880, xxiv. p. 3-il, figs. 7 and 7 a.

792. Amphimallus solstitialis (Lamellicorn): left antenna bearing a supernumerary pair of imperfect antenna? articulating by a common stalk 011 the anterior surface of the second joint. The two extra clubs are an imperfect pair, complementary to each other, being set on back to back, in Position A. The most anterior of the clubs has only two lamellar joints, one small and one large. The posterior has three lamellse. The normal club has three lamelhe as usual. Originally described by MOCQUERYS, /. c., p. 15,y?Vy.

793. Anomala junii (Lamellicorn) : left antenna bears 3 clubs, each having 3-jointed stem articulating with elongated 2nd joint of antenna. [Symmetry nut clear: possibly Position DPI'.] KUAATX, Dent. cut. Zt., 1881, xxv. p. Ill, PI. m. rig. 4. *794. Geotrupes typhaeus £ (Lamellicoru) : left antenna bearing a pair of supernumerary clubs compounded together. The antenna is normal up to the 7th joint which is dilated. The 8th is still more dilated and bears posteriorly the normal club composed of three lamella?; and anteriorly by a separate articulation a supernumerary structure (Fig. 174, inr,ml) consisting of three joints, each of which has the form of a complementary pair of lamellae joined by their morphologically posterior (sc. external) edges. The whole supernumerary structure is thus morphologic-

ally a j>"ir of dubs, a right and a left, compounded together. Tinhi.-tol.i-y i.f tin' -upenimnerary lamelke is just the same as Fin. 171. Geotrupet ti/i>)m-tn;. No. 794. Left antenna bearing a compounded «f i-lnli-i. ml, mr, morphological left and right of the extra parts. (The property of Di Ki.mt/.i that nf tinimnnal lamella-, all being covered with pubescence. Tinform of the compound eleventh joint is somewhat irregular. The extra parts are in the Position A of the Scheme. Specimen kindlv lent l.\ hi KIJAATX, and first described and figured by him in Deut. ent. Zt., I.SSD, xxxin. p. 221, fig. 13.

7!i.">. Melolontha hippocastani ^ having supernumerary parts of double structure upon both the right and the left antenna. ///;//// Antenna. Third joint elongated, thickened and pre- -'•ntin- t\\i» articular surfaces; of these one is terminal and hear- a normal antenna IT club, while the other is dorsal and bears a >u|ierniinieraiT double club. This structure has the form shewn in tindrawings. Fig. 17-"), A, shews its appearance when looked at t'rnni abnve, B shews the structure when seen from below and externally. It consists of seven pieces shaped like half-funnels, tit ted into each other.

Km. 17-''. Mflnloiitliii lii]>]><><-ii*t<nii. No. 795. D, view of tlie whole head and ant. IIII.T :ifii r vos Hi M'i N. C. view of right antenna. B, detail of right antenna from below. A, detail of the same from above. The morphological nature of this .supernumerary organ may be determined thus. The upturned edges of the folds bear hairs as shewn in the figure A ; since in the normal antenna the dorsal edges of the lamellae alone bear hairs, these edges are in this case dorsal morphologically as well as by position.

Since the outermost lamella (marked 4) is articulated into the third joint of the fimiculus, it is therefore the 4th joint, or proximal lamella, and the remaining lamellae are therefore oth, 6th, 7th, 8th, 9th and 10th respectively. Next, the surface mr is structurally like that of the internal (sc. anterior) surface of the proximal lamella of a normal club, and the surface ml is a similar surface : but ml, being an internal surface, faces towards the right and is therefore morphologically a left; while mr, being an internal surface and facing towards the left, is a right', hence this club consists of two clubs compounded together by their external or posterior borders, and the two are a right and a left, the left being next the normal right club.

Lastly, since the upper free edges of the lamellae are structurally dorsal, it follows that their lower edges are structurally ventral : but these lower edges do not exist as free edges, for the lamella? are continuous upon their ventral aspect ; therefore the surfaces which are adjacent in the extra right and left clubs, and by which they are compounded together, are partly ventral surfaces. This is approximately Position DP of the Scheme. Left Antenna. Second joint thickened and presenting three articulations as follows. 1. a peripheral articulation bearing the normal club ; 2. a ventral articulation bearing a 4th joint and club composed of 3 formless lamella? ; 3. a dorsal articulation bearing a small cylindrical joint only. The shape and formation of these extra parts is so indefinite that their morphology could not be determined.

For the loan of this specimen I am indebted to Dr L. vox HEYDEN, who first described it in Dent. ent. Zt., 1881, xxv. p. 105, fig.l. 795'. RhizotrogUS aequinoctialis (Lain.) : 4th joint of ri.ufht antenna bears a supernumerary structure projecting forwards and lying in the same horizontal plane as the normal club. This structure is lanceolate in form and its outer surface is in texture similar to the external surfa«-e> of a normal club. On the ventral aspect it presents a simple rid «,'»•. but on the dorsal side its outer coating is divided by a spindle-shaped sin through which part of the internal structure protruded. The ed^es of this opening and the protruding portion of the interior bear a te\\- hairs. There can be little doubt that this supernumerary body represents an imperfectly formed pair of chilis, and that it is in taet a more rudimentary condition of the parts found in No. 795. Specimen originally described and figured by MOCQUKKYS, <',>/. annnn., p. \^,fig.

797. Llchnanthe vulpina (Lam.) : right anU'iuia bears in addition to normal ehil> a ' small spherical club made up of three joints, arising from p.»t.-rior boni.-r of a long joint apparently representing the normal 4th, 5th, and Oth joints not segmented from other. [A* tlii- -ii|u i ilium T;U y part is in itself symmetrical it probably containwithin it-elf ]>aits of a pair of clubs compounded in Position P. Cp. No. 7!i.".. J -l.vYM . II. 1 ., Trim*. .Inn-r. Knt. Soc., 1880, vm. p. 158, PI. IV. fig. 8. Tl'V Polyphylla decemlineata 1 1. am. •lln-orn). A specimen in which the right antenna beai a partially dooble supernumerary branch in addition to the normal null nna. I hiadditional structure articulates with the second joint of the antenna by mi annt' a -ingle lain joint. This joint carries a double club consisting of two of lamellje, Si ren being iii each set. The two sets of lamella are united at thfir baseat an angle of fi.rty-fiv.- decrees. The plane of the normal club is perliciiiliculai tn that of tin- a I iin 'i ma) one-. The normal club itself is ith shorter than that nt tin uther side. [I hi' details of the structure of this specimen are difficult tn follow ami tin leader is n •('. mil for further particulars to the description and figures given in thr m iginal.] JAYXE, H. F., Trams. Amer. l-.nt. Soc., LSHO, vm. p.

Odontolabis stevensii$ (Lucanida-). As tin.- repetition in this specimen is almost complete and the relations of the parts fairly clear though in some respects peculiar, a detailed account Tinliody, legs, iVrc. arc' normal, save that the back of the head and thorax have been crushed by some accident. The antenna' are both abnormal in the way shewn in Fig. 170. The cuiiditioii \\ill be better understood if the normal antenna is first < I escribed.

FKI. 17'i. odantnliiltia .«ti'rt'n*ii, No. I'M. 'i'he head seen from below, and enlarged views of the two antenna •. H. right. L, left. There is some doubt as to which of the branches is the normal and which the supernumeraries. See The normal antenna of Odontolabix is much like that of its ally l.in-nniiK trrntx, the Stag-beetle. It is made up of 10 joints composing three parts differentiated from each other. The first., or "scape," is a single joint as long as the rest of the antenna. It widens a little from its central end or base tou.-mls I),,, apex, and is slightly flattened from above downwards. The second part, or " tuniculus,1' has six simple joints. The last three joints form the rlub. They are flattened from above downwards and lie in a hori/ontal plane. The anterior ("inner") border of each of these three joints is produced into flat expansions, covered with sensory pores, which together form a series of serrations along the anterior border. When in its natural

position the serrated border of the right antenna faces towards the left side, and that of the left is turned towards the right. The structure of the abnormal specimen is as follows. Left Antenna. Scape normal. Its plane however is not quite horizontal as usual, but is a little oblique, the anterior border being slightly higher than the posterior. In the funiculus the 1st and -nd joints (2nd and 3rd of the whole antenna) are a little thicker than usual but otherwise normal. The 3rd joint of the scape is enlarged and presents at its apex two sockets, each bearing a continuation as shewn in the figure. The two sockets are not in a horizontal plane, but their plane is oblique and nearly at right angles to the plane of the scape, the socket bearing the branch I1 being the higher. It is important that the pre<-i>«' relations of these parts should be clearly understood.

This outer socket of the 3rd funicular joint bears the branch I1, made up of three more funicular and three club-joints, turning their serrated border in the direction of the right antenna: I1 is therefore structurally a left antenna. Its surface is of the same nature as that of a normal antenna, but its size is a little smaller. It is in an oblique plane inclined to the horizontal at about 45°, the posterior (outer) border being the higher. The inner socket of the 3rd funicular joint bears a cylindrical joint not quite fully segmented off from the next joint peripheral to it. These two are 4th and 5th funiculars. The 5th again presents two sockets, bearing respectively the branches I2 and I3. The branch P has one small joint (6th funicular) and three club-joints, turning their serrated border towards I1. This branch is therefore structurally a right antenna. It stands in the same oblique plane as I1, the serrated border being the higher. In size it also agrees with I1, being rather smaller than the normal. The branch /3 is a normal left in size and shape, and it lies in a horizontal plane.

Here therefore there is a left antenna and a pair, one a right and the other a left. Which then is the normal, /' or I3 ? Inasmuch as Is and I- arise by a common stalk it may seem that they are the extra pair and that I1 is the normal. \\'e have now seen in many cases that extra parts in Secondary Symmetry are compounded together as /•* and I- are here. But considering the fact that P is of normal size and in the normal horizontal plane, whereas I1 and I- are both smaller and are in an oblique plane complementary to each other, I incline to the view that if one branch is the normal, it is I3, and that I1 and I- are the extra pair in Secondary Symmetry, though they do >/<>t a rise together. They are then nearly in Position DPP, but depart I'mm that position in the fact that /' is not horizontal (cp. No. 757).

If /l and I- are really the extra parts, in the fact that they do not arise together, but spring separately t'nun different points on the normal, we meet with a condition ran-ly seen, but that /.'////// Anti'ii/iii. Scape precisely as in left antenna. The I-: funicular (2nd ant. •unary) has two sockets at its apex, placed like those "ii tin- :!id funicular of the left side, the ant. Timsocket I., intinlower and the posterior socket being the higher. The anterior sockel bears a normal right antenna,?'3. The posterior bears the struciiire shewn in the figure. This appendage has unfortunately been broken, but enough remains to suggest the original structure. It con.-ists of five funicular and a 1st clubjoin!. The .1th joint of the whole funiculus bears a large socket lo. .king downuanls and forwards, its other socket looking backwards and upwards. rYom the former the original continuation habeen lo>t. The latter bears the 6th funicular and its 1st club-joint, this ,-IL;HIII ha-\ ing an empty socket.

The plane of the two sockets of the 5th funicular is oblique to the hnri/on. like that of /' and /'-'. Though it is clearly im- p"-- ible to shew how this antenna was in its unbroken state, \\e may note that it' it were continued in the way suggested by the dotted lines it would have borne a complementary pair of clubs, ?•' and /•'-'. like /' and /- of the other side, placed like them in an oblique plane nearly corresponding with DPP of the This specimen was kindly entrusted to me by M. Henri Cad.-au de Kerville. He tells me that he believes a description nf it has already appeared, but this I have failed to find. I have therefore ventured to describe it again, with apologies to the original describer. The specimen bears a label in the handwriting of the late Major Parry and was no doubt in his celebrated collection of Lucanidse.

800. Melolontha vulgaris : right antenna bearing a pair of incomplete supernumerary antennas (Fig. 177). The first joint is normal ; it bears a second .joint of singular form, consisting of a long anterior branch, and a short posterior branch ! the length of the anterior. The anterior bears two clubs in the manner shewn in the figure (Fig. 177). Of these one (R) is inwardly directed and is as wide as, but only £ the length of ;i normal club. The posterior of the two clubs (L) is directed b;i'-l. wards and has only four lamella; which are apparently united together. The other small club (A)() is also composed of only four lamella; which are similarly united together. In both L and /.'' t In- middle lamellae shew traces of further subdivision. The figure represents the three clubs as being all in one plane, but the club It' is really below //, which stands up from the normal antenna. It is mentioned that some of the tarsi u ere mutilated or defective. [Here L and K are clearly a complementary pair, though separately arising from the normal. It will In-, observed that as in Lereboullet's case (No. 790) the second joint, which is common to two clubs, is greatly elongated.] WKSMAEL, Hull. Ac. /!>•/!/., 1850, xvi. 2, p. 3S2,fig.

*801. Navosoma sp. (Longic.) Left antenna abnormal. The joints of the normal are a little flattened from above downwards and are nearly elliptical in section. But the anterior border is differentiated from the posterior by the presence of two elongated patches of tissue covered with sensory pores. The two patches are both on the anterior border, one being on the dorsal surface and one on the ventral, separated from each other by a chitinous ridge. Upon the general surface of the peripheral joints of the antennae are several other such patches, but none are so distinct as those of the anterior border. The abnormal left antenna has the form shewn in Fig. 178. So far as the 8th joint it does

FIG. 178. Navosoma, No. 801. Left antenna seen from below. Lettered on the view that E and L' are the extra parts. ,S, sensory patch. (In Hope Collection.) not differ from the normal. The 9th and 10th joints have besides their chief patches of sensory pores (&) on the anterior border, an additional patch (L' + R'} posterior to the chief patch. But up to the 10th joint there is no vertical division. The 10th joint however has two articular surfaces, anterior and posterior, in the same horizontal plane. Theposterior bears an apical (llth) joint of normal form, having anteriorly a sensory patch. But the apical joint borne by the anterior articular surface has two such sensory patches, an anterior and a posterior. This joint therefore contains in itself parts of a pair of joints. It is not quite fully segmented off from the 10th joint.

Nevertheless it is difficult to suppose that the anterior joint is the extra pair in Secondary Symmetry, for its anterior patch, L*. seems to continue the normal series of patches, S, S, etc. Therefore the ].at.-he>. R and L' seem to be the patches of the extra pair, though one of Ui-m is on a separate joint and the other is applied to the normal. Taken with the case of Odontolabis No. 799 and .!/< 'lolonthu, No. 800, thia must, I tliink. In- judged to In- a possible account, and in this case It and // are, as i >••_'••! i-.U symmetry, in Position P. It is of course possible that /,* ami A' are really the extra pair in Position A, hut the presumption is rather the other way1. Specimen in Hope Collection at Oxford.

$02. 1" none of tinfollowing can any confident statement be made as bo tinsymmetrical relations of the parts. Several of the cases 1 have mvself seen, Imt I noticed no clear indications as bo their symmetry. A good many of them however were examined b, .for.- 1 was fully alive to the importance of these matters in the case of filamentous antennae, and perhaps if they were studied with proper regard to the question of symmetry more mi-ht he made of them. Many cases that follow are mutilated or partly amorphous, ami of almost all the descriptions are very imperfect. Finour purpose some value attaches to these records as evidence of the distribution of such abnormalities, and to any person who may hereafter pursue the subject a fairly complete list of the references may be of use. To this therefore I >hall routine m\self: tor on reviewing the abstracts that I have mad.- of these examples it is clear that they only give the results of siipertirial examination.

Speaking generally, in these cases, from some one joint of an antenna thenarises either a pair of extra antennas compounded for a greater or less extent of their proximal parts, or two extra antennadistinct from their point of origin. The letters l\ ami L indicate the side affected, and the number following is approximately that of the joint from which the extra parts spring. In the greater number of sound cases the three branches lie in .»• nearly in a horizontal plane and are, I anticipate, in I'o-it ions A or P.

Cases which seem from the indications to conform to the Scheme are marked ||. Mutilated or partially amorphous cases are marked . - I >. -niptioii and ti^uic iiu-i.inrt. Apical joint of extra branch is litid. (Encn.) * + Carabus arvensis L 4 Specimen kindly lent by M. A. FAUVEL. 803. Meloe violaceus ? : between right eye and the base of the right antenna arise two supernumerary antennas from the head. Of these one has 3 joints and the other has one. KHAATZ, Deut. ent. Zt., 1877, xxi. p. 57, PI. i.fig. 22.

The following example is mentioned here, though its nature is quite obscure. In it there is a suggestion that parts of two extra antennae are present, but the extra parts seem to be peripheral to the parts which they repeat. As my stay in Rouen was short I was not able to give as much time to this specimen as I should have wished '. 804. Melolontha vulgaris £ : left antenna abnormal. This case differs wholly from any other that I know of. I can only describe it in a most tentative way. The appearance when the lamellae were cleaned and separated was as shewn in Fig. 179. Joints 1 — 8 are fairly normal, but peripheral to this place there were

FIG. 179. Left antenna of Melolontha nil<i«ri*, No. 804. The numbers are set in tentative suggestion of the possible nature of the parts. (In Rouen Mus.) 1 This antenna was when I saw it covered with mould and dirt. In washing it I accidentally detached it from the head, but I mounted it again carefully with the specimen. a nmiilMT iif lamella •. ^>\ne like normal lamellae, others quite i insular. A- far afi I c.,iild make out, the divisions were as -hewn in tinfigure. and I have affixed numbers to the several l>:iris in illu-trat t«\\ <>\' their possible nature. The appearance suggests that thi-iv is an irregular repetition of a pair of clubs peripheral to tinnormal antenna, but I can form no opinion as to tinmorphology of the parts. Originally described by .Mm \>i i;i;vs. Col. anorm., 18(SO, p. 12, fiy. [Description and figure altogether misleading.]

805. Bembidium striatum (Carabidw): left maxillary palp arises l.y a tirst joint enlarged towards its apex, bearing ////•"' sepa r,-ite terminal joints bead of one. < )i these joints one stands apart on a small process of the first joint, Inn the other two are placed close together, on either side ot the apex of the Hrst joint, and diverge from each other at about a ri^ht an^-le. JACQUELIN-DUVAL, Ann. Soc. Ent. France, 1850, Ser. 2, vin. p. ."i:;:;, I'/, it, ,\\i.

806. Helops sulcipennis (I let.): supernumerary, partially double apical joint arises from the 2nd joint of right maxillary palp. Tt is >et on /// r'n/lit niK/Iix to the plane of the normal palpus. JAYXK, H. F., Trams. Amer. Ent. Soc., isso. vin. p. Kil,///- H. S()7. Euprepia purpurea (Aretiida-): a specimen in which the right wings and ;uitriina were male and the left wings and antenna female, is declared to have !„,--, -,,-eil ,;// ,.1-trit ]><iir «/' ji,ilj,i. [No sutVicient desi'ription of this extraordinary occurrence is given ; an.l as the repetition of the palpi is only incidentally mentioned, it may he doubted whether a full examination was made.] FKEYKK, C. F., Beitr. ur Schmetterlingskunde, isi"). I'd/, v, p. 1'27, Tab. 4~)S,ji/i. 4.

THE facts as to Secondary Symmetries in Crustacea are so similar to those already detailed in Insects that, were it not for their value as confirmation of the principles indicated, it would be scarcely necessary to describe them at large. Some few of the cases have besides a special interest, as in them may be seen rudimentary or bud-like structures apparently presenting the lowest condition of paired parts in Secondary Symmetry. Precisely as in Insects there are a number of cases (including those last mentioned) where it would at first sight be supposed that the extra parts are single, but on inspection most of them prove double. Nevertheless there remain some few where this cannot be shewn, and strange -as it may seem, these must be admitted to be genuine examples of duplicity of limbs. Of them a special account will be given in another chapter.