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

confluent though each had a separate cesophageal opening and a separate arch from the aorta. In the drawing, for which I am also indebted to Professor Weldon, the cesophageal openings are not shewn. *140. Bdellostoma. In this genus the number of branchial sacs is variable, different numbers being found in different species and individual variation also occurring. The generic name Heptatrema was originally given by DUM£RIL from the presence of seven gill-sacs. In 1834 JOH. MULLER, finding that this character is not constant proposed the name Bdellostoma. Of three Cape specimens examined by him one had seven gill-sacs on each side, one had six on each side, and one had six on the right side and seven on the left. To these he gave the names B. heptatrema, B. hexatrema and B. Jteterotrema respectively (Abh. k. Ak. Wiss. Berlin, 1834, pp. 66, 67 and 79, Taf. vil.). Further observation has shewn that the number of gill-sacs in the Cape Bdellostoma is liable to individual variation, some specimens having six while others have seven. The name B. cirrhatum (GuXTHER, Cat. Brit. Mus., vm. 1870, p. 511) includes these and the New Zealand specimens. As to the relative frequency of specimens with six or seven pairs or with an asymmetrical arrangement I have no information. A collection lately brought from the Cape by Sedgvvick includes one specimen with six pairs and several with seven pairs.

HI. B. polytrema : single specimen from Chili, badly preserved hut apparently having fourteen pairs of gill-openings. GUNTHER I.C., p. .'.12. /;. bischoffii: single specimen, 10 gill-openings on each side. ibid. Ammocoetes : having eight branchial openings on each side instead of seven, the normal number. The shape of the mouth of this specimen also .•diiiormal, heing described as somewhat square. [No satis- t'.u-tory description.] EDWARD, THOMAS. Zuoloyist, xvi., p. 6097.

1 [-2. In connexioD with individual Variation in the number of gillsaca in Myxinoids it should be borne in mind that in Petromyzon tin -re are normally seven pairs of gill-sacs. The case of the Notidanidie may also be mentioned in this connexion. Among Selachians tin- Nut idanidse are peculiar in having a number of gillslits iitluT than live, and of them Hexanchus has six pairs, while Though tinevidence of this subject is well known and has often been collected, it may be convenient to give here some .detract of the farts in so far as the phenomena of Variation are illustrated by tln-m. Since cervical fistula; have been believed to result frci 1 1 tin- | H •! sistence of the embryonic branchial clefts, they may properly IK- considered in relation to the general question of Variation in the numher of gill-slits, while the development of external appendages pei'hap- Anally homologous with the external ears, directly concerns the subject of Meristic Variation. Man. The .subject has been studied by many observers, especially by ASI-III:I;SII\-, and by HEUsiNGER3, who brought together and abstracted 4<i cases, being all that had been described in Man up to isiii-. <;. Kis< IIKR* gives a full list of the literature of the subject up to |N7(>, with an analysis of Go cases. A further paper liy ! 1 1:1 siNci.i; contains a general account of these structures as 'he\ are toiind in Man and in the domestic animals. Additional casi 8j i.>;j, ther with a general discussion of the subject, especially in relation to tistuL-eon t he external ears, were given by Sir JAMES

' Balanoglossus. In li\v sjiccios with which I am acquainted, the number 'f ^'ill-liars ami >lits varies in prnportion to the size of the l><uly, and as it is not unlik'-ly that tln-c animals coniinn.' to LMI>\V tlirnu^lmut life, it is probable that the miiiilirr i if ln-aii'-hia is always increasing by formation <>t new ^ill-slits at the posterior end d" ill.' branchial re^imi. The same is probably true of Amphioxiis. \ -• .11 toast, I>,' ii<tii!i.< f.,/// ,•,,!!./, H/li'. Hnlin. lsH'2. H Irch.f. I'.it/i. .ln>it. a. I'ln/.t.. 1st; i, \\ix.

PACiET1 in 1878. Lastly, the whole evidence as to cervical fistulse and the structures associated with them has been fully collected up to 1889 and tabulated by KOSTANECKI and MIELECKI-, who also discuss in detail the relations of these abnormalities to the facts of development. The following account is taken from these sources. For figures the reader is referred to the original memoirs. *14o. Cervical fistulas are generally known as orifices placed in the region of the neck, leading into a sinus of greater or less extent, varying in size from a mere pit to a 'duct some inches in length. In the greater number of cases the sinus ends blindly, but in about a third of recorded cases (K. and M.) it passes inwards to open into the pharynx, forming thus a communication between the pharyngeal cavity and the exterior. Such passages are spoken of as complete cervical fistulas, those which have an external but no internal opening being external incomplete fistulas. Besides these there are cases of diverticula from the pharynx or oesophagus which do not reach the exterior, and these are known as internal incomplete fistulas.

Cervical fistula? are more commonly present on one side only, but in a good many cases they have occurred on both sides. According to Fischer they are more common on the right side than on the left. The following statistics are given by him. 65 persons had 79 fistulas : .')! unilateral, 1-4 bilatei'al: 20 complete, 53 without an opening to the pharynx : of the unilateral cases 33 were on the right and 13 on the left : 34 in males, 30 in females. There was evidence of heredity in 21 cases.

The external opening is very small and may either be on the surface of the skin or elevated on a minute papilla. Sometimes it is covei'ed by a small flap of skin as with a valve, in other cases it is placed as a fissure between two lips. The positions in. which the external openings of cervical fistulas are found are very variable, but in the great majority of cases the opening is close to the middle line in the neighbourhood of the sterno-clavicular articulation, generally from a few lines to an inch above it, on either the inner or the outer border of the sterno-cleido-mastoid muscle. In rarer cases the external opening is placed at the level of the middle of the cricoid cartilage, and is sometimes just behind the angle of the jaw. These positions are not however at all precisely maintained, but vary a good deal in different cases. When the external opening is in the higher situation and the fistula is complete, a sound may then be passed into the pharynx, but when the external opening is low, the duct when present passes upwards covered by skin only, in a straight line so far as the upper limit of the larynx, at which point it turns at a sharp angle upwards and inwards. For this reason it is not possible in such cases to follow the course to the pharynx by means of a sound, but in some of them the presence of an internal opening has been proved by the injection of fluids having colour or taste. The position of the internal openings is also variable, and from the nature of the case has been accurately

2 VON KOSTANECKI untl VON MIELECKI, Arch. f. path. Anat. u. Phys., cxx. and cxxi. determined in comparatively few instances. In a case dissected by NI.I BOFEB1 thenwas a fistula on each side, the external opening >.f tin- rL'ht was .', in. from the middle line and 7 lines above the cla\ide. that of the left was 3 — 4 lines higher and further from the miildlr line. The right internal opening was on the posterior border of tinpharyngo-palatine muscle, behind the cornu of the hyoid near the t-.ii.-il, tlif left internal opening being rather higher than the right. Internal openings of such fistulahave also been seen on the edge of tli.- arcuphai yngo-palatinus, also in the neighbourhood of the root of tintongue. SKIM;I.J gives a ease in which there were two fistula1, the -•lie .,11 the right side in the upper position, and the other in the middle line at al.out the same level, but whether either of these communicated uith the pharynx could not be made out. The twin-brother of the -aimint'ant had a single minute fistula.

The duets of cervical h'stuhe are usually of greater calibre than the external openings but they are rarely wider than a fine quill. The walls antough and the lining epithelium, is sometimes flat and sometimeeiliat.-d. The degree to which the walls are sensitive differs in di He rent eases. The external opening is described in several instances as having a reddish colour. In three cases of the presence of branchial fistulain female patients, it is recorded that the external openings became intlaiued during the menstrual periods.

I'r-.in the point of view of the naturalist the chief interest of eer\ ical fistulaarises in connexion with the question of their morphology. Since the time of Ascherson the view has been commonly accepted that these structures arise by persistence of embryonic gillclefts, and some of the recent writers3 on the subject have gone so far as t.. apportion the various forms of cervical fistula? among the several gill .-lefts from the first to the fourth, according to the situations of the external openings, giving diagrams shewing the regions occupied by each. As KosTANKCKi and .M i Ki.KCKi point out, this apportionment is quite arbitrary; for in the development of the neck the external in- \ aginations f-,r all the clefts behind the hyoid arch become included in the sinus cervicalis of Rabl (sinus pnecervicalis of His), which is e\ent ually dosed by the growth of the opercular process from the hyoid arch. The external opening of a cervical fistula may thus re]. resent a part of the sinus cervicalis still left open, but it cannot on the ground of its position be referred to any gill-cleft in particular. Sii.-h reference could only be properly made on the ground of the position of the internal opening and the course of the duct in relation to structures \\hose relation to the visceral clefts is known. Moreover owing to the \\ay in which the .'5rd and 4th clefts are shifted in\\ards by the formation of the sinus cervicalis, Kostanecki and Mielecki consider that they are practically excluded. The same authors after an analysis of the cases in which the position of the internal opening has been properly ascertained, come to the conclusion that in all these it falls within the region of the 2nd visceral sac

(hyo-branchial). Besides they point out that the evidence in the few cases in which the course of the duct has been traced, shewed that it passed between the external and internal carotids. In their judgment, therefore, cervical fistula? are all to be referred to the second (hyo-branchial) cleft. Next it is to be remembered that according to many observers (especially His) there is at no period a complete connexion between the outer gill-clefts and the evagination from the pharynx or branchial sacs, but the membrane separating these chambers is stated by tin-in never to be broken down. If this account is accepted, it is, as Kostanecki and jMielecki have said, necessary to suppose that in the case of any complete cervical fistula a communication between the exterior and the pharynx has arisen by some abnormal occurrence. This is illustrated by reference to the normal condition of the first or hyo-mandibular cleft. Here the auditory meatus represents an external incomplete fistula, and the Eustacbian tube an internal incomplete fistula, the two being separated by the tympanic membrane. In a single case given by VIRCIIOAV' a complete passage existed congenitally in this position, together with great abnormality in position and form of the external ear.

From the evidence it may thus on the whole be concluded that incomplete external fistula? result from imperfect closure of the sinus cervicalis, and that incomplete internal fistula? may arise by persistence of one of the branchial sacs, but it is doubtful whether many cases of the latter properly belong to the category of branchial fistula? at all. 144. Abnormal appendages attached to the neck have been described by several observers, and by those who have discussed the subject of cervical fistula? some account of these appendages is generally given. In the neighbourhood of the external ears, especially near the antitragus, such structures having the form of small warts or flaps of skin are not very uncommon. Their presence is generally associated with deformity of the external ear, and often with what are known as "aural fistula?2." In the region of the neck, supernumerary auricles

2 Aural fistulas are spoken of by many writers as being of the same nature as cervical or branchial fistula?. They are blind ducts or pits, opening on some part of the external ear and are nearly always associated with other abnormalities either in the form of the ear or defective hearing, &c. (Scmriiz, De fist, colli coni/en., Inaug. Diss., Halle, 1873 [not seen, W. B.]; URBANTSCHITSCH, Moiiatxch. f. Ohrfiih., 1877, transl. Edin. Med. Jour., xxm. 1878, p. 690.) They may be either unilateral, or on both sides of the body. SIR JAMES PAGET (Trans. Med. Cliir. Soc.. LXI., p. 41) described the occurrence of such fistulae in the ears of several members of a family, many of whom were affected with deafness. The supposed connexion of these fistulas with cervical fistulas was in this case suggested by the fact that several cases of actual cervical fistulas occurred in the same family, several of its members having both cervical and aural fistulas. From the evidence of the not infrequent association of the two kinds of malformation most writers (PAGET, URBANTSCHITSCH, &c.) consider that the aural fistulas must be biauchial in origin and may be taken to represent the first (hyo-mandibular) cleft.

KOSTANECKI and MIELECKI (I. c.), following His, point out that since in no case has an aural fistula ever been known to communicate with the auditory rneatus or tympanic cavity, this belief is unsupported; and in addition, that from the mode of development of the external ear from a number of tubercles, it is are nm.-li raivr. l.ut in several instances they have attained a consiilt-rablr .|i-v.-l..j.in.-iit. < »f tlii> class of variation the f..ll»u ini,' well-kn«)\vn case tinni"-t n-markaUf.

14-5 A healthy female iiifiint was brought to Ciuy's Hospital in 1851 on account of Og ;_'io',vthabout tinmiddle of the lateral cervical region-. The i i.'in..ved until February Is.-.s, when they were found to have i,,,,, i ... , situated over about the centre of the sterno-cleidojna-toid mil sell s. Co the 1 'h«-y resembled the tissue of the lobe of tinauricle, an. I tli- v contained within them a linn resist ins,' nucleus like the cartilage oi the

Tli.-v w. ic also I with peculiarly delicate, soft, downy hairs. lik.' tli. lohe ..f ih" ear. They W.T.- excised without difficulty. Each was supplied with a Mnall artrry. Tiny appeared to be intimately associated with the fibres of tli.- pliity-uia m ..t' dippinde, per than this structure, and to be entirely itical Mcti.'ii WH- made in the Lui^' axis of cadi growth; and the tissues of lli.' liilii< and ..t ill.' libn'-rai tilav'"1 <>f the auri.-lr wendearly distinguished. The -hap.' i >( tlic Qbro-oartilage resembled iimir i,\- less closely m parts, the outline of the pi.'per auiiele, ami its ti.v-m > were the same. BlBKETT, J., Tninx. Path. Soc. / and., a., I*.'.*, p. i ts. tig.'.

-ii!)iei.-ni to suppo-e that aural fistula' arise by the imperfect union of these tubercles. The fart, however, that th.-e various defects in development of the branchial appaiatu- and iN del i\ at i\ .•- ai v fre,|ueiitly a>^.oei:itet 1 t. ^'et her is well established. A- indicating the freipiency of a--nciat inn with disease of the ear, Urbautschitsch iinntioiis that in •Jiion aural caseB, 12 iiistauc.'- of aural fistulfe were seen. The -am.' author ^-ives a reiuai kalile ca-e of the occurrence of aural fistula on the riijlit "iilij in many memhei - ..f the same family with other ini]iortant particulars (/. c.). In Lancet, ]*"•*. n. ]'. :'.'.i'.i (cpioting KAUVIV), and in a jiaper by VIUCHOW HpiMtni^' \Vii.i. i i, .In-/;, ji.itli. .lii<it. I'lii/.*., lsi',1, x\x. p. •_'•_'-"•. ivteivnce is made to a Case "I" < V-I.I'.I.IIM, I'l-'ict. m.rtiix. (/«• .n/rc in<>ii.-.tri I/inn., Norimb., Itlsj, pp. 3(5

. i ciribing a child with "four ears." On referring to the original however il aj. p.ars that this was merely a double mon-ti r. having two incomplete heads, and thus hears no analogy with the present examples. Several cases analogous to the above, though differing in the extent of the development, are on record '. KOSTAXECKI and MIKMCCKI (I. c.), who give references to the literature of the subject, consider together with VIRCIIOW and others, that there is no doubt that these supernumerary auricles may properly be regarded as "heterotopic" partial repetitions of the external ears. According to a view which has been held by the majority of writers on the subject, and which is in part alternative to that given above, it is suggested that the cartilages contained in these appendages are in reality parts of one or other of the usually undeveloped branchial arches behind the hyoicl. As against this suggestion it is to be remembered that in the subsequent development of the neck these arches are pushed in far from the surface, whereas the cartilages in question are always superficial. The usual histology of these bodies is in favour of the view that they are repetitions of the ear-cartilages, but on the other hand a specimen of cervical auricle in Mus. Coll. Surg. (No. 373, c) contains not only cartilage but also a small bone of complex form. But whether or not any part of such cervical auricles truly represents any part of the gill-bars, it is clear that these external projections having the structure of the ear, considered from the point of view of Variation must be regarded as partial repetitions of the ears, and there is a considerable probability that they stand to the sinus cervicalis in a relation similar to that which the normal external ear bears to the hyo-mandibular cleft, being according to the terminology here proposed, examples of repetition by forward Homceosis.

In this connexion the question of correlation between supernumerary auricles of the neck and cervical tistuke is especially important. If it is true that such auricles are repetitions of the ears, it might, on the analogy of other cases of repetition, be expected that they would usually be found bounding the external openings of fistula^. As a matter of fact they have several times been found in such a position, but the connexion between these two variations is by no means a close one, for cervical tistuhe are not as a rale accompanied by cervical auricles, nor are cervical auricles generally associated with cervical fistula*, such collocation being on the whole exceptional. It should also be mentioned that in a few cases small cartilaginous or bony structures have been found imbedded in the neighbourhood of cervical fistulas, but that similar structures have also occurred independently of any fistula2.

In many domestic animals both cervical fistuLe and auricles are well known and have been described by Heusinger3 from whom the following account is chiefly taken. 146. Pig- Cervical auricles are not uncommon and have been referred 1 A figure is given by SUTTON, J. B., in III. Meil. Xews, 1889 (repeated in "Ecolutiaii niul Dixfaae," by the same author IH'IO, p. 83), representing a large supernumerary auricle on the right side of the neck of a girl. The structure is represented as helicoid in form, closely resembling the normal ear. It is unfortunate that no description of this specimen is given: in tbe absence of such •description tbis quite unprecedented case cannot be accepted without reserve.

- Dermoids of many kinds occurring in the cervical region of Man and other animals are by many writers considered to arise by modification of tissues occluded from the walls of the branchial clefts. to a- di-t miruisliiir.: particular local breeds. They are generally paired ictures. Tli" t'ollouiiig case is exceptional in the fact that the auricle uapre-.-nt mi the left only, and that it was associated with an op, -nil,.,' po--ibly of a cervical fistula. A pig having a single appendage about 7 cm. lung attached under tinlower jaw on the left, is described liy Ki DEfi DESLONGCHAMPS '. It contained a stalk of cartilage stated IM ha\e i .-.-ml. led the cartilage of the ear. To this on either side was attached a small muscle. Unfortunately the appendage had been cut

• •I!' dose t«. the .skin. A small opening (pertain) was present on the skin near the appendage, and from this opening a small brush or tuft of bristleprut i -tided. Fistul:e in the neck of swine are well known as giving rise to a disease called u-eissK Jiorate ill Germany (Fr. la Sole or poil piqu£) from the fact that certain white bristles are found at the opening of the • liict. In the popular fancy it is supposed that the bristles themselves bore the perforation, but according to ZUXDEL* they are congenital and iif-'n bilateral. Ib-usinger agrees with Ziindel in regarding such openings MV branchial li-tula-.

147. Sheep :in,l Goats. < Vrviml fistula? unknown, but appendages on the neck imon. The -h> i p nl' tlir \\ 'd-ter marshes are described3 as having the neck bare «i \\nol, and an a)i|i. HI MIU ••• as of a fur-collar. Above the collar and below the jiliarynx tln-y li:i\i- a pair of appendages about the size of an acorn. Such ap- p,-n'la;_'>-< ar> -aid tu lie u<>t uncommon in Merinos'1. Among the Kalmuck and Kirghiz sheep ami ;_'<>ats -ueli auricles are said by Pallas5 to be common. In many ign 1.1. . - i.r ;_•... u- these auricles seem to be a constant character. In position tln-y may \aiy fr-mi tin angle of the jaw to the middle of the neck. The length is ii-ually aln mt :', in. hut they are recorded as reaching 15 cm. Figures of goats b:i\ing such auricles are vi\oi by Scxxox8. The anatomy of one of tbese bodies is nli.il 1>\ < MM i;u \:, and it is mentioned that a plate of cartilage was found in the intt n.'i-. A -imilar rai ula-i \\a- 1'mind by STEWAUT* together with striped muscular tiln'i'. (iniiliaiix ^'i\«- a ca-t- of two she-goats on a farm, one having cervical a|'i» adagi a, the ntb.T having muie. I-'.ach gave birth to a pair of kids at the same time. Kadi pair \\;>- a mall- and female, and in the one the male only had the appi inla^e- and in the nther the female ,,nly. The characterof th.- father of these kidwere not kno\\ n.

]4S. Horse. <i-r\i.-al auricles unkno\\n. Fistula (in the position considered by Beuringe] i" indicate the first branchial cleft) are common and are recognized by their action in -"iling the hair near tin1 i \teinal opening. Ki .- Ai-ii ri.ATioN. 'I'hc evidence as a whole o-ocs to slic\v that st riict ares, soinet im.-s of lar^v si/r. ha\ in-- several essential f'eatuivs "f the exti-inal ear, tbat m.iy in lad b.- fairly spoken of as n-])etilioiis of the ear. may by Hoinu-otie Variation appeal' on theneck of Man and oilier animals: further, thai these repetitions have been \H ' ..... :cur at the openin-s ,,f ci-rvical lisiuhe, surest ing a parison \\nli i be relalion of the external ear i«- the li\ o-inandibi ilar el ei i. bin that such * relation to cervical fistulais exceptional.

SOME of the phenomena of Meristic Variation are well seen in the case of mammae1, and especially in the modes by which increase in the number of these organs takes place. The facts regarding these variations in Man have so often been collected that it is scarcely necessary to detail them again. For our present purposes it will be sufficient to give a recapitulation of the chief observations in so far as they illustrate the phenomena of Variation.

The most important collections of the evidence on this subject are those of PUECH*, LEICHTENSTERNS, and WILLIAMS*, from whose papers references to all cases recorded up to 1890 maybe obtained. Besides these, BRUCE 5 has given a valuable account of a considerable number of new cases together with measurements and statistical particulars. These accounts contain almost all that is known on the subject but additional reference will be made to original authorities in a few special cases.

In Man supernumerary mammae or nipples nearly always occur on the front of the trunk, being usually placed at points on two imaginary lines drawn from the normal nipples, converging in the direction of the pubes. These lines may thus be spoken of as the •• Mammary lines." It is with reference to supernumerary mamma? occurring on these lines that the subject of mammary variations is chiefly important to the study of Meristic Variation. In addition to these, however, there are a few well authenticated examples of mammse placed in parts of the body other than the mammary lines and of these some mention must be made hereafter.

1 It will be understood that facts as to variations consisting in absence of mammfe or nipples and other such changes do not conae within the scope of this volume, but belong rather to the province of Substantive Variation. 3 LEICHTENSTERN, I'irch. Arc h. f. path. Aunt. u. Pliy*., 1878, LXXIII. p. 222. This collection was apparently made independently from that of PUECH. In tli.- great majority of cases (over 00 per cent., LEICHTEN- si i. UN ' ) of mamma' placed on the mammary lines, the supernumerar\ structures are below the normal ones, being then as a rule internal to ih. in, while those found above the normal mammas are 1 ss common and are external to the normal mamma;. The di- Man- -araiiug the normal from the supernumerary mamma-

.ily indifferent cases, and most conditions have been -• . 11 intermediate between a stage in which the nipple is bifid, and that in which completely separate supernumerary mammae are presented. It is of consequence to observe that there appears to be mi case in \\hich a supernumerary mamma is so large as the normal mamma of the same individual. Tindegn-o to which supernumerary structures of this nature are d«-\el,,ped is very various. They may be fully formed mamma-

with nipples, in th«- female capable of function; while in oth. r ML; n.. <li-tnr_;iiished glandular tissue of mammary character in connexion with them, or the}- may be tumours of mammary character \\ith»ut nij)ples or even definite ducts. Between these several con'lit ionthere is no sharp distinction. It appeal's therefore that thenai«- fii'n rudimentary or imperfect conditions possi- 1.1.- : either supernumerary nipples without recognizable mammary glands, shading off into small warty elevations of uncertain character, and on the other hand redundant portions of mammary gland \\ithout nipples. Tinlatter may be partially connected with the normal mamma- or quite separate from them. All these states of imperfect inn annnn-h more common than the complete supernumerary mamma-.