The Development of the Skull of Scyllium (Scyliorhinus) canicula L. · 2006-05-23 · The...

The Development of the Skull ofScyllium (Scyliorhinus) canicula L.

By

G. R. de Beer, M.A., B.Sc.Fellow of M e r t o n College, J e n k i n s o n L e c t u r e r i n E m b r y o l o g y , a n d

D e m o n s t r a t o r i n Zoology in t h e U n i v e r s i t y of Oxford.

W i t h P l a t e s 3 2 - 7 a n d 27 Text-f igures .

C O N T E N T S .

I N T R O D U C T I O N . . . . . . . . . . 3 9 1

D E S C R I P T I O N o r S T A G E S . . . . . . . . 5 9 3

D I S C U S S I O N : T H E A U D I T O R Y C A P S U L E . . . . . 617

T H E R E L A T I O N S O F T H E J A W S T O T H E B R A I N - C A S E . . . 622

T H E P R O B L E M O F T H E A C R O C H O R D A L . . . . . . 637

S U M M A R Y . . . . . . . . . . . 6 4 0

L I S T O F L I T E R A T U R E C I T E D . . . . . . . 6 4 1

E X P L A N A T I O N O F L E T T E R I N G 6 0 1

E X P L A N A T I O N O F P L A T E S 6 4 5

INTRODUCTION.

IT is little short of remarkable that the skull of S c y l l i u mc a n i c u l a , one of the commonest types used for laboratorydissection, should not have been the object of more detailedand recent study than it has. Parker's (1878) famous paper wasexcellent, although it contained a few mistaken observations,as pointed out by Ridewood (1895 and 1897), and it remainsthe sole complete work on the development of the skull ofS c y l l i u m . Some valuable observations on certain features ofthe development are contained in papers by van Wijhe (1904)and by Goodrich (1918); the former being principally concernedwith S q u a l u s a c a n t h i a s , and the latter with the segmenta-tion of the head. On the other hand, the skull of S q u a l u sa c a n t h i a s has been well and thoroughly studied by Sewert-zoff (1899),-Wells (1917), van Wijhe (1922), and Mori (1924).It shows certain differences as compared with S c y l l i u m , and

592 G. E. DB BEER

this paper is presented as an attempt to fill in a gap in ourknowledge and to provide material for a detailed comparisonbetween the developments of two fairly closely related forms.In addition, a discussion of the results obtained will lead toa consideration of certain problems of general interest invertebrate craniogeny.

The material on which the present study is based consistedof some thirty embryos of S c y l l i u m c a n i c u l a preparedaccording to van Wijhe's method (1902 and 1922) for demon-strating cartilage. After staining and while being dehydrated,the embryos were dissected under the binocular microscopewith the help of very fine Swiss watch-maker's forceps, andmounted in glass-cells which allow of study under the com-pound microscope on all sides. In addition, the whole ofProfessor Goodrich's and the late Dr. Jenkinson's splendidcollection of sections was available for the purpose of checkingthe results obtained by a study of the whole preparations, andfor determining the relations of those structures which areinvisible by the van Wijhe technique. The figures which illustratethis paper were drawn under the camera lucida at a magnifica-tion of 25 diameters, and the author is indebted to his wife forassistance in preparing them for reproduction. A carefullyselected set of transverse sections is given in the form of text-figures, for the purpose of elucidating certain matters. The workwas carried out in the Department of Zoology and ComparativeAnatomy of the Oxford University Museum, and, as always,the author enjoyed the kind encouragement of Professor E. S.Goodrich, P.E.S., to whom he wishes to express his gratitude.

Before the work had progressed very iar, it had becomeobvious that embryos of a given length from Naples were notidentical with those of the same length from Plymouth. Onthe whole, the Naples material showed a tendency to delay inthe chondrification of the anterior and dorsal regions of the skull,as compared with Plymouth material of the same degree ofdevelopment. The difference is purely an embryological andnot a morphological one, for the structures in the Naplesembryos, when they do chondrify, exhibit relations identicalwith those shown by the Plymouth embryos. In addition to

SKULL OF SCYLLIUM 593

Table of material of Scyllium canicula used forthis work.

Stage.

123456789

10111213

Adult

Length.

mm.2425282929£3030343536363745—

Provenance.

Plymouth?

NaplesNaplesNaplesNaplesPlymouthPlymouthPlymouthPlymouthNaplesPlymouthNaplesPlymouth

Illustrated in

Fig. 1, PL 32Fig. 2, PI. 32Fig. 3, PL 32Fig. 4, PL 32Figs. 5, 6, PL 32Figs. 7, 8, PL 32Figs. 9, 10, PL 33, and Text-figs. 1, 2Figs. 11, 12, PL 33Figs. 13, 14, PL 34Figs, 15, 16, PL 34; fig. 17, PL 35Fig. 18, PL 35Figs. 19, 20, PL 36Figs. 21, 22, 23, 24, PL 37Figs. 26, 27, 28, PL 36, and 25, PL 37

the material enumerated in the table given above, severalNaples embryos were studied and drawn, but they are notreproduced in the paper, for they merely show imperfectionor retardation of chondrification. The matter, however, isinteresting, as revealing a variation in the time factor indevelopment within a species. Analogous variations with regardto S q u a l u s a c a n t h i a s were found by van Wijhe (1922);he observed that cartilage made its first appearance in embryos22 mm. long from Heligoland and Helder, but 32 mm. long fromBoston; further, embryos collected in summer from Heligolandwere more advanced in their development than embryos ofsimilar size collected in winter from Helder.

The author wishes to express his appreciation of the kindhelp given him in provision of material by the PlymouthLaboratory of the Marine Biological Association, by the StazioneZoologica at Naples, and by Professor J. P. Hill, F.R.S.

DESCRIPTION OF STAGES.

S t a g e 1 (24 mm., Plymouth, fig. 1, PI. 32).

In several embryos of this stage of development, it has beenpossible to see that the parachordal cartilages are the first

594 G. R. DE BEER

elements of the skull to become chondrified. They extend froma point level with the front wall of the auditory sac to a pointabove the first gill-slit, and are slightly, thicker behind than theyare in front. There is as yet no sign of the dorsal process or ofthe occipital arch. Goodrich (1918) has shown that the sheetof mesenchyme from which the parachordal arises is doubtlessformed from sclerotomes 4 and 3, and perhaps also 2 and 1,though no signs of segmentation are any longer visible. On theother hand, scleromeres are formed in segments 5, 6, and 7, ascondensations of mesenchyme. At later stages, the cartilagewhich arises from these condensations becomes fused with theparachordal plate forming its occipital region, while the con-dyles appear to be derived from the scleromere of segment 8.

S t a g e 2 (25 mm., fig. 2, PL 32).

At this stage a number of other structures have made theirappearance, and are present as very young cartilage. Theseare: the skeletal elements of the mandibular and hyoid arches(palato-quadrate and Meckel's cartilage; hyomandibula andceratohyal); the trabeculae, and the orbital cartilages. All thesestructures are independent. Meckel's cartilage appears topossess only one centre of chondrification, and not two (oneach side) as reported by van Wijhe (1922) in S q u a l u s . AsGoodrich showed, the angle made between the parachordal andthe trabecula is not as acute in S c y l l i u m as it is in S q u a l u s .

S t a g e 3 (28 mm., Naples, fig. 3, PL 32).The parachordal cartilages are, of course, situated immedi-

ately beneath and median to the auditory sacs, and at this stagethey can be seen to send a process outwards and forwards oneach side towards the under surface of the front of each sac.At the same time, the lateral edge of the parachordal is nolonger straight. Just behind the process described above, andwhich is really the rudiment of the anterior basicapsular com-missure, the edge of the parachordal forms the so-called laminabasiotica, and bears an uprising knob, which is the rudiment ofthe dorsal process. Behind this, the parachordal is narrowerthan in the region of the lamina basiotica.


S t a g e 4 (29 mm., Naples, fig. 4, PI. 32).The cartilage of the parachordal is now continuous with a

thin sheet of cartilage which underlies the front region of theauditory sac. In Scy I l i um, therefore, as Goodrich has alreadyshown (1918), the auditory capsule chondrifies from the first incontinuity with the parachordal, whereas in S q u a l u s vanWijhe showed that the auditory capsule had independentcentres of chondrification. Van Wijhe is of the opinion thatthe early and close association between the parachordals and theauditory capsule indicates that the origin of the brain-case wasassociated with the provision of a firm attachment for the earas an organ of balance. The conditions in S c y l l i u m thereforegive him even stronger support than do those of S q u a l u s .

It may be noticed that the cartilage cells of the parachordalsextend forwards a little way along the notochord sheath (withoutyet invading it), but the foremost ends of the parachordalsdiverge from one another and from the notochord, which extendsforwards freely between them.

This embryo is of interest for two further reasons. It showsthe earliest origin of the vertebral elements in the form of thebasidorsal cartilages, and they are entirely independent anddistinct from one another. This is important in view of the factthan van Wijhe (1922) has described the vertebral elements asarising from four c o n t i n u o u s bands of cartilage along thenotochord, the dorsal bands corresponding to the basidorsalsand the ventral bands to the basiventrals, in both S c y l l i u mand S q u a l u s . At later stages, he shows in S q u a l u s thatthe ventral and dorsal band fuse on each side into a single band,and that eventually the separate basidorsal and basiventralcartilages are separated off from one another, presumably, ashe thinks, as a result of the growth in length and diameter ofthe notochord and of its sheath. My studies of S c y l l i u mhave shown me nothing of this, and the basidorsal and basi-ventral cartilages are independent from the start, as theyappeared to me to be in H e t e r o d o n t u s (de Beer, 1924). Itis only at subsequent stages that adjacent basidorsals andbasiventrals become connected by a thin film of cartilage alongthe outer surface of the elastica externa. There appears here,

596 G. R. DE BEER

therefore, to be a conflict of evidence concerning the methodof origin of the vertebral column, which I am at a loss to under-stand. Attention may be directed to the fact that in the an-terior region of the vertebral column, the basidorsals appearbefore the basiventrals, as is also the case in Sa lmo (deBeer, 1927).

The other point of interest with regard to this embryo con-cerns the polar cartilage. As is now well known, van Wijhediscovered a polar cartilage in S q u a l u s (1904) between thetrabecula and parachordal, and the independent existence ofthis structure has been confirmed in other forms. In S q u a l u sthe polar cartilage first becomes attached to the hind end of thetrabecula of its own side, and then with the under surface ofthe front edge of the parachordal. After numerous fruitlessattempts to find the polar cartilages as independent structuresin S c y l l i u m (a quest in which Professor van Wijhe was alsounsuccessful, as he kindly informed the author) it would seemthat they are at last discovered in this embryo, not as indepen-dent chondrifications, but as nodules of cartilage attached tothe under surface of the front edge of the parachordals, freefrom the trabeculae, and sending a small sharp process towardsone another. As will be shown in connexion with later stages,these structures answer all the requirements of polar cartilages,but the new feature which they show is that they are moreclosely associated with the parachordals than with the trabe-culae. This is, after all, a difference of small importance, andin A mi a (Pehrson, 1922; de Beer, 1926) the polar cartilageswould seem to be in a condition intermediate between that ofS q u a l u s and of S c y l l i u m , for in Amia the polar cartilagearises in association with both the trabecula and parachordal.The condition in S c y l l i u m is, however, interesting assuggesting the possibility that the cartilage which appears toform the anterior region of the parachordal may in realityrepresent the polar cartilage (as, for instance, in Teleostei).

S t a g e 5 (29£ mm., Naples, figs. 5, 6, PI. 32).

This embryo agrees with the previous one in that the polarcartilages are visible, attached to the under surface of the front


end of the parachordals. The trabeculae and orbital cartilagesare still free. The parachordals now show a prominence at theirhind ends, forming the rudiment of the occipital arch. It mayalso be noticed for comparison with earlier stages that the hindend of the parachordals now lies over the second gill-slit. In thevertebral column, the neural arches of the first three vertebraehave appeared, as minute nodules of cartilage, separate andindependent from the basidorsals. As regards the visceral archskeleton, ceratobranchials and epibranchials are now presentin the first, second, and third branchial arches. They arise in theform of a U wrapped round the adductor muscles, and whenthe arms of the U meet, the characteristic foramina are formed.

S t a g e 6 (30 mm., Naples, figs. 7, 8, PI. 32).

The trabeculae are attached to the polar cartilages, so thatnow for the first time the floor of the brain-case is continuous.There is, however, no continuity between right and left sides,for the notochord completely separates the parachordals. Theorbital cartilage is now attached by the pila antotica to theupper surface of the front edge of the parachordal of its ownside. In S q u a l u s van Wijhe describes an independentchondrification for the cartilago antotica (pila antotica) as wellas for the cartilago supraorbitalis (orbital cartilage). InS c y l l i u m a separate origin for the pila antotica has not beenfound. The lateral surface of the auditory capsule is nowchondrified, but it has not been possible to make out the twocentres of chondrification discovered by van Wijhe in S q u a l u s .The anterior end of the auditory capsule is attached to theparachordal by means of the anterior basicapsular commissure,but its posterior end is free. Between the capsule and the lateraledge of the parachordal there is therefore a slit, which, in itsanterior region represents a basicapsular fenestra, such as ispresent in S a 1 m o. But owing to the absence of a posteriorbasicapsular commissure, this fenestra is confluent with thefissura metotica behind it. The dorsal process of the parachordalmarks the approximate line of separation between these twoportions of the slit. The glossopharyngeal nerve passes outbehind the dorsal process and beneath the auditory capsule, in

598 G. E. DE BEER

the fissura metotica. The occipital arch is pierced by a foramenfor the posterior cranial root of the hypoglossal nerve. VanWijhe (1904) imagined that this condition, which he alsoobserved, indicated that the occipital arch represented a singleneural arch, pierced by a ventral nerve-root foramen, as arethe neural arches of S q u a l u s , but not of S c y I l i u m .Goodrich (1918), on the other hand, has shown that it is muchmore probable that the occipital arch represents two arches,related to the seventh and sixth sclerotomes respectively, andenclosing a hypoglossal root between them. The conditions inScy I l i um lend strong support to Goodrich's view, for in somecases, the enclosure of the nerve-root in a foramen is incomplete,and it then lies in a groove between the large posterior occipitalarch and the smaller anterior arch. There are, therefore, definitetraces of metameric segmentation in the metotic region of theskull o f S c y l l i u m . In the visceral arch skeleton, the pharyngo-branchials have now appeared in the first to third branchialarches. Basiventral cartilages are now present in the. vertebralcolumn.

S t a g e 7 (30 mm., Plymouth, figs. 9, 10, PI. 33, and Text-figs.1,2).This is one of the most interesting and important of the early

stages of development of the skull of S c y l l i u m . Anteriorly,the trabeculae converge towards an independent median rostralcartilage, while, on each side, each trabecula bears a processdirected outwards and forwards; the ethmoid process ofSewertzoff (1899), or the lamina orbitonasalis of van Wijhe.In front of the rostral cartilage on each side the paired nasalcartilages have appeared, in the form of transverse bars eachbearing three backwardly projecting processes, the outermosttwo meeting round the external nostril and enclosing it in aring. In front of these nasal cartilages is another pair of inde-pendent cartilages, representing the rudiments of the front wallof the nasal capsule.

Although the polar cartilage is attached to trabecula and para-chordal, its position is clearly marked out by notches: one infront, representing the original separation between the polar


cartilage and the trabecula, and lodging the efferent pseudo-branchial artery; and one behind, between the polar cartilageand the antero-ventral edge of the pila antotica, and lodgingthe pituitary vein. The polar cartilages were seen at an earlierstage to have a process directed towards their fellow of theopposite side. In this embryo, these processes have becomeinterconnected by means of a transverse cartilage which willhere be called the postpituitary commissure. As the recon-struction given in Text-fig. 1 shows, this commissure lies behindthe point of entry of the internal carotid arteries and beneath thenotochord. It is therefore not an acrochordal, and it is notthe same as the cartilage which van Wijhe found in S q u a l u sand which lies in front of the point of entry of the carotids.The latter, or precarotid commissure, is present in later stagesof S c y l l i u m , as will be seen below.

The orbital cartilage has extended and lapped round thetrochlear nerve, which is therefore enclosed in a foramen.From the postero-dorsal corner of the orbital cartilage, a processis directed backwards towards the auditory capsule; this is therudiment of the taenia marginalis, which will eventually convertthe incisura prootica into the foramen prooticum. The auditorycapsule is now fairly well developed, and the three canals andthe septa separating them from the main cavity of the capsulecan be seen by transparency. Anteriorly, the capsule is, ofcourse, attached to the parachordal by the broad anterior basi-capsular commissure; posteriorly it is still free. The embryoreconstructed in Text-figs. 1 and 2 is slightly further developedthan that illustrated in figs. 9 and 10, PI. 33, and the posteriorview shows that a shelf of cartilage stretches out from theparachordal on each side, beneath the posterior portion of theauditory capsule. This cartilaginous shelf plays an importantpart in producing the apparent passage of the glossopharyngealnerve through the auditory capsule, for in this region the audi-tory capsule has no cartilaginous floor of its own, and the nerve,lying above the shelf, passes through what is really a tunnel ofwhich the upper wall has broken down. Should a name berequired for this shelf, lamina hypotica may be suggested, soas to show that it is morphologically ventral to the floor of the

f3 f

TEXT-FIGS. 1 and 2.

Reconstruction from serial sections of an embryo of Scy l l iumc a n i c u l a about 30 mm. long, seen from in front.

Reconstruction from serial sections of an embryo of Scy l l iumc a n i c u l a about 30 mm. long, seen from behind.


auditory capsule, and in order to distinguish it from that otherportion of the parachordal plate which van Wijhe has called the

EXPLANATION OF LETTERING.

a, articular facet for the hyomandibula; abe, anterior basicapsular com-missure ; ac, auditor}' capsule; an, abducens nerve; aom, arteria ophthalmicamagna; as, auditory sac; asc, anterior semicircular canal; ami, auditorynerve; 6a 1, branchial arch 1 (to 5); be, basieochlear fissure; bef, basicapsularfeneatra; bdl, basidorsal 1 (to 5); bh, basihyal; bp, basal process; br 1,branchial rays of first (to fifth) arch; bvl, basiventral 1 (to 7); bvc, basi-vestibular commissure; c, occipital condyle; cac, cavity of auditory capsule;cbl, ceratobranchial of first (to fifth) arch; eel, centrum of first vertebra;of, foramen for internal carotid artery; eh, ceratohyal; en, cartilage con-necting front and side walls of nasal capsule; csp, cartilage connecting palato-quadrate and lamina orbito-nasalis; des, ductus canalis semicircularisposterioris; del, ductus endolymphaticus; dm, dura mater; dp, dorsal processof parachordal; ds, dorsum sellae; e, eye; eb 1, epibranchial of first (to fifth)arch; ee, elastica externa; eha, efferent hyoidean artery; ei, elastica interna;en, nostril; epa, efferent pseudobranchial artery; exdl, dorsal extrabranchialof first (to fourth) arch; exv 1, ventral extrabranchial of first (to third)arch; /, foramen magnum; fa, foramen for auditory nerve; fb, basicranialfenestra; fel, foramen for ductus endolymphaticus; fep, foramen for efferentpseudobranchial artery; fg, foramen for glossopharyngeal nerve; fh, hypo-physial fenestra; fie, interorbital canal for pituitary vein; fll, foramen en-closing lateral-line canal; fit, foramina for twigs of superficial ophthalmicbranch of facial nerve; fin, fissura metotica; fn, facial nerve; fne, front wallof nasal capsule; fo, foramen for optic nerve; foa, foramen for orbital artery;foe, foramen for oculomotor nerve; fol, olfactory foramen; fon, orbito-nasalforamen; fp, palatine branch of facial nerve; fpr, foramen prooticum; fsf,foramen for superficial ophthalmic branch of facial nerve; fso, superficialophthalmic branch of facial nerve; fst, foramen for superficial ophthalmicbranch of trigeminal nerve; ft, foramen for trochlear nerve; fv, foramen forvein; g, gap between lamina hypotica and true floor of auditor}' capsule;gc, canal for passage of glossopharyngeal nerve; gn, glossopharyngeal nerve;gs 1, gill-slit 1 (to 5); h, hyomandibula; ha, hyoid arch; hb, hindbrain;hbr 1, hypobranchial of the first (to third) arch; hf, foramen (or foramina) forhypoglossal nerve; hn, hypoglossal nerve; hrd, dorsal group of hyal rays;hrv, ventral group of hyal rays; hy, hypophysis; i 1, interdorsal 1 (to 5);ic, internal carotid artery; iom, inferior oblique muscle; ip, incisura prootica;is, invaded sheath of notochord; I, lagena; Ib, limit between planum antor-bitale and preoptic root of orbital cartilage; Ibo, lamina baaiotica; Ic, labialcartilage; Iho, lamina hypotica; lie, lateral-line canal; Inc, side wall of nasalcapsule; Ion, lamina orbito-nasalis; he, lateral semicircular canal; ma,mandibular arch; maa, macula ampullaris anterioris; mal, macula ampullaris

NO. 296 R r

602 G. R. DE BEER

lamina basiotiea, and which forms the floor of the more anteriorpart of the capsule (1922, p. 281). The roof of the auditorycapsule is incomplete, and through a hole it is possible to seethe basicapsular fenestra. Eventually this fenestra will beobliterated when the lateral edge of the parachordal (laminabasiotiea) and the ventral edge of the wall of the auditorycapsule have joined.

The hypoglossal nerve on one side of the embryo illustratedin figs. 9 and 10, PI. 33, has become enclosed in a foramen at thebase of the occipital arch, but on the other side it still lies ina notch, between the anterior and posterior occipital arches, asdescribed by Goodrich.

lateralis; man, macula neglecta; map, macula ampullaris posterioris; mas,macula sacculi; Me, Meckel's cartilage; innc, median wall of nasal capsule;mm, macula recessus utricularis; ms, mandihular visceral cleft; n, notochord;na 1, neural arch of first (to third) vertebra; nc, nasal cartilage; nci, innerprocess of nasal cartilage; ncm, middle process of nasal cartilage; nco, outerprocess of nasal cartilage; nep, notch for efferent pseudobranchial artery;lion, notch for optic nerve; npv, notch for pituitary vein; na, nostril; nt,spinal cord; nv, notch for vena capitis lateralis; nve, notch for vein; oa,occipital arch; oa 1, 2, or 3, segmental occipital arches; oc, orbital cartilage;ofm, obliterated region of fissura metotica; on, oculomotor nerve; opn, opticnerve; ora, orbital artery; cw, orbital sinus; p, parachordal; pa, pila antotica;pb, pituitary body; pbc, posterior basicapsular commissure; pc, polar carti-lage ; pec, precarotid commissure; pch, perichordal commissure; pev, posteriorcanal vacuity; pev, plica encephali ventralis; pf, parietal fossa; pi, plariumantorbitale; pma, premandibular arch; pn, profundus nerve; ppb, peri-pharyngeal band; ppc, postpituitary commissure; pq, palato-quadrate;pr 1, pharyngobranchial of the first (to fifth) arch; prr, preoptic root of theorbital cartilage; psc, posterior semicircular canal; pv, pituitary vein; r,rostrum; rem, rectus externus muscle; rif, rectus inferior muscle; rim, rectusintermis muscle; rsm, rectus superior muscle; ru, recessus utricularis; s,saccule; sbo, subocular cartilage; sc, spiracular cartilage; si, lateral rostralprocess; am, median rostral process; sp, spiracular pouch or cleft; spo,supraorbital cartilage; sps, subpituitary space; srp, suprarostral process;sso, spiracular sense-organ; st, stomodaeum; stp, stomodaeal pouch; sv 2,second spinal ventral nerve-root; t, trabecula; tep, tectum posterius; tf,foramen for thyroid gland-stalk; th, rudiments of thymus gland; tm, taeniamarginalis; In, trigeminal nerve; tp, trabecular plate; ts, tectum synoticum;tso, superficial ophthalmic branch of trigeminal nerve; u, utricle; uc, unpairedcartilage representing hypobranchials of fourth and fifth arches; v, velum;•vcl, vena capitis lateralis; vn, vagus nerve; vs, ventral sac of hypophysis.

SKULL OP SCYLLIUM 603

At this stage the notochord sheath has begun to be invadedby the cartilage cells of the parachordals. The result of thisinvasion, which begins at the posterior end of the parachordals,is to establish cartilaginous connexion between the parachordalsof the two sides, above and beneath the notochord, formingwhat may be called a perichordal commissure.

In the vertebral column, it is noticeable that the neural archesof the first and second vertebrae are situated considerably in frontof their respective basidorsals, while the neural arches and basi-dorsals of the third and following vertebrae are in contact. At thisstage the interdorsal cartilages have appeared, as independentcartilages, between the neural arches and on the same level as them.

The palato-quadrate shows a thickening on its dorsal surfacenear the front end, which foreshadows the basal (orbital) pro-cess. Further back, the fifth branchial arch has now chondrified.

S t a g e 8 (34 mm., Plymouth, figs. 11, 12, PL 33).This stage follows on easily from the previous one. The

trabeculae have fused with the median rostral cartilage, andin so doing have enclosed the hypophysial fenestra from in front.In the hind region of this fenestra, the postpituitary commissurehas appeared as an independent transverse bar of cartilage,between the polar cartilages, beneath the notochord, and behindthe point of entry of the carotid arteries. Thus, while thepostpituitary commissure forms a posterior boundary to thehypophysial fenestra, it does not form the anterior boundary ofthe basicranial fenestra, for it is ventral to the level of thatfenestra. In other words, the postpituitary commissure is notthe same thing as the acrochordal.

The cartilages representing the rudiments of the front wallof the nasal capsule have now become attached to the rostrum,while in the nasal cartilages the outermost two prongs are nolonger fused behind the nostril. The lamina orbitonasalis isfurther developed than in previous stages, and the taeniamarginalis now connects the orbital cartilage with the auditorycapsule. The latter still shows a basicapsular fenestra, visiblethrough a hole in the roof. The hinder part of the auditorycapsule is now underlain by the lamina hypotica, extending

R r 2

604 G. R. DB BEER

sideways like a shelf from the hindmost part of the parachordals.The glossopharyngeal nerve emerges from a notch between thelamina hypotica and the auditory capsule.

There are now two hypoglossal nerve-roots enclosed in fora-mina on each side, and a horizontal section through this region

TEXT-FIG. 3.Horizontal section through the occipital region of an embryo of

Soy l l ium c a n i c u l a 32 mm. long (32 H, 9-1-4), showing themetameric segmentation of the occipital arches in the posteriorparachordal region.

(Text-fig. 3) is very instructive as to the manner in which thesenerve-roots are enclosed between what are really segmentallyrepeated occipital arches. The perichordal commissure hasextended farther forwards, and, in the visceral skeleton thedorsal extrabranchial cartilages have appeared.

The basal process on the palato-qiiadrate is a well-developed

hb ds PPC w/ dm

tp sps p'cc fH^ ICTEXT-FIGS. 4 and 5.

Transverse section (G. R. de B. 29-5-3-6) showing the precarotid com-missure, postpituitary commissure, and dorsum sellae, all distinct.

Median longitudinal section (J. H. J., young, 18-1-3) showing sub-pituitary space, trabecular plate, precarotid commissure, post-pituitary commissure, dorsum sellae, internal carotid arteries,pituitary vein.

606 G. R. DE BEER

knob, projecting upwards and outwards, beneath the laminaorbitonasalis. The relations of this region are of considerableinterest, for, on the one hand, Edgeworth has observed a car-tilaginous connexion between the palato-quadrate and thelamina orbitonasalis (antorbital process; 1925, p. 255), and onthe other, Sewertzoff and Disler (1924) have reported the exis-tence of a (mesenchymatous ?) separate pharyngomandibularelement (in S c y l l i u m , S q u a l u s , and Muste lus) , whichultimately becomes attached to the basal (orbital) process of thepalato-quadrate. I have observed the mesenchymatous massto which Sewertzoff' and Disler have applied the term pharyngo-mandibular, but its independence appears to be doubtful. Itseems to be a mesenchymatous connexion between the palato-quadrate on the one hand and the lamina orbitonasalis (antor-bital process) and the subocular cartilage on the other. Thelatter element has not yet appeared at this stage, but it will bedescribed below. At later stages I have also observed traces ofchondrification in this mesenchymatous connexion, thus con-firming Edgeworth's observations. If Sewertzoff and Disler'sview is correct, it is important in showing that the mandibulararch was once divided into four elements, like the branchialarches. The evidence from S c y l l i u m , however, does notseem to be sufficient to decide this matter.

S t a g e 9 (35 mm., Plymouth, figs. 13, 14, PI. 34).

The embryo illustrated in figs. 13 and 14, PI. 34, has beendissected so as to obtain a better view of the neurocranium byremoving the visceral arches. Cartilage is now spreading be-tween the rostrum, the trabeculae, and the base of the laminaeorbitonasales, to form a fairly extensive plate. This plate isnotched on each side at its anterior edge, for the exit of a vein.

The lamina orbitonasalis now shows features of considerableinterest. Extending upwards from it and from the trabeculaat its base is a plate of cartilage which is curved, so that itpresents a convex face anteriorly towards the nasal sac and a con-cave face posteriorly towards the orbit. This plate is perforated atabout its centre by a foramen through which a vein passes fromthe orbit into the future cavity of the nasal capsule. Now

TEXT-FIGS. 6-25.

Selected transverse sections through an embryo of Scy l l i umc a n i c u l a about 35 mm. long (G. R. de Beer, 35):

Fig. 6.—Section 3-1-1. Eig. 8.—Section 4-4-1.Eig. 7.—Section 4-1-2. Eig. 9.—Section 4-4-10.

608 G. K. DE BEER

Scyl l iu rn has lost its profundus nerve, but a comparison withH e t e r o d o n t u s shows that this foramen is the one throughwhich the profundus nerve typically leaves the orbit for thenasal capsule. In other words, this foramen has the relationof an orbitonasal fissure. Owing to the curvature of the platethrough which it is pierced, when seen in dorsal view, thisorbitonasal foramen gives the appearance of being a notch inits anterior face.

The question arises as to the nature of this cartilaginous plate,and, in particular, of that part of it which is situated medianto the orbitonasal foramen. In other vertebrates, the orbito-nasal foramen or fissure lies between the planum antorbitaleof the nasal capsule and the side wall of the skull in the formof the orbital cartilage. Now, applying these relations toS c y l l i u m , it would seem that that portion of the plate whichlies median to the orbitonasal foramen is to be regarded as apreoptic root of the orbital cartilage, springing from the trabe-cula; Avhile the cartilage lateral to the orbitonasal foramen isthe planum antorbitale, springing from the lamina orbitonasalisor ethmoid process. The orbital cartilage is therefore representedby two separate portions: a preoptic root developed in closeassociation with the planum antorbitale, and a large platedeveloped in association Avith the pila antotica. These twoportions will eventually fuse together on each side, and alreadyat this stage an irregular piece of cartilage is present near thefuture point of fusion. In connexion with the previous stagemention was made of the subocular cartilage. This structurearises as a band of cartilage below the lateral edge of the trabe-cula and parachordal. In S c y l l i u m , as in P r i s t i u r u saccording to Sewertzoff's (1899) descriptions, the subocularcartilage appears to chondrify from two centres. The anteriorcentre is close to the lamina orbitonasalis and is involved in theconnexion with the palato-quadrate; the posterior centre is inthe neighbourhood of the foramen prooticum. Here the sub-ocular cartilage wraps round the orbital artery and is attachedto the skull-wall in front of and behind it, with the result thatthe artery emerges through a foramen on to what will become theupper surface of the subocular shelf. Eventually the anterior

SKULL OF SCYLLIUM GOO

and posterior portions of the subocular cartilage join, forminga continuous subocular shelf, which becomes attached to thelateral edge of the trabecula and paracbordal between thelamina orbitonasalis and the auditory capsule. But, for a longtime, the subocular cartilage is mostly free, as the transversesections show.

The foramen prooticum has become subdivided into theopenings for the main branches of the trigeminal and facialnerves, and the separate openings for the superficial ophthalmicbranches of these nerves.

The carotid arteries now enter the cranial cavity througha foramen formed by the development of a precarotid commis-sure joined on each side on to the postpituitary commissure.In the vertebral column, the basidorsals and basiventrals ofcorresponding segments are becoming joined. The occipitalarches have come into contact with the posterior wall of theauditory capsule, thus enclosing the fissura metotica to forma foramen jugulare, through which the vagus nerve passes. Thisnerve is separated from the glossopharyngeal by the approxima-tion of the under surface of the auditory capsule to the uppersurface of the lamina hypotica. The notch for the glosso-pharyngeal is just behind the articular facet for the hyornandi-bula. The basicapsular fenestra has been obliterated.

S t a g e 10 (36 mm., Plymouth, figs. 15, 16, PI. 34; fig. 17,PI. 35).

At this stage the outlines of the nasal capsule are marked out,and there is now a thin strand of cartilage connecting therostrum with the planum antorbitale, forming the rudiment ofthe front and side walls of the nasal capsule. The nasal cartilagesare still free, on the under surface of the future capsule. Car-tilaginous connexion has almost been established between themain body of the orbital cartilage and the preoptic root. Inthe embryo illustrated in figs. 15 and 16, PI. 34, the positionsof the thin film of cartilage connecting the lamina orbitonasaliswith the basal process of the palato-quadrate, and of the anteriorportion of the subocular cartilage, are indicated with dottedlines. A new feature at this stage is the supraorbital cartilage,

810 G. R. DB BEER

which projects laterally from the dorsal edge of the orbital

Kg. 10.—Section 5-1-3.Kg. 11.—Section 5-1-8.

Fig. 12.—Section 5-2-7.Fig. 13.—Section 5-3-9.

cartilage and taenia marginalis, and is continued back on tothe lateral surface of the auditory capsule. It is in the latter


region that this cartilage is best developed, and it forms an archthrough which the lateral line canal passes.

The most interesting feature shown by this stage concernsthe region between the polar cartilages. As in previous stages,there is a postpituitary commissure and a precarotid commis-sure, which, between them, enclose the carotid arteries in theirforamen and separate them from the remainder of the hypo-physial fenestra. In addition to these two transverse cartila-ginous bars, there are now a pair of processes which projecttowards one another from the foremost points of the parachor-dals, dorsal to the polar cartilages, and in intimate relation tothe dura mater, in which the notochord itself is also enclosed.These processes, which eventually meet one another, form thetrue dorsum sellae or front edge of the parachordal plate,beyond which the notochord projects a little way into the plicaencephali ventralis. These relations are further shown inText-figs. 4 and 5.

Up till now, the front ends of the parachordals had beendivergent from one another. But now that they become con-nected by a transverse dorsum sellae, a true basicranial fenestrais enclosed. This fenestra has not previously been describedin Selachians. Its existence has been confirmed in severalpreparations of S c y l l i u m made by the van Wijhe method,and also in serial sections. It eventually becomes completelyobliterated by the extension of the parachordals. The anterioredge of the dorsum sellae also seems to grow forwards, so thatin later forms it definitely overhangs the pituitary fossa orsella turcica from behind. A consequence of the formation ofthe postpituitary commissure and of the dorsum sellae isthat the pituitary vein becomes enclosed in its cartilaginousinterorbital canal.

Fig. 17, PI. 35, is a drawing of a ventral view of the nasalcartilages and of the mandibular and hyoid arches. It may beremembered that Parker (1878) referred to the nasal cartilagesas labial cartilages: a view which it is difficult to support.Between the ventral ends of the ceratohyals the basihyalhas appeared, and it is pierced by a foramen through whichthe stalk of the thyroid gland passes. The existence of this

612 G. R. DE BEER

foramen strongly suggests that the basihyal is really of pairednature.

S t a g e 11 (36 mm., Naples, fig. 18, PL 35).

This embryo has been drawn from the left side and slightlyfrom behind, to show the A'arious canals in the auditory capsuleas seen by transparency, and to show the relations of the occipitalarch to the auditory capsule, and the various elements of thevertebral column. In particular, it is possible to see how theposterior canal of the auditory capsule (it can scarcely be calleda semicircular canal, for it occupies a complete circle) has bulgedbackwards over the lamina hypotica, so that the points of exitof the glossopharyngeal and vagus nerves are fairly widelyseparated.

S t a g e 12 (37 mm., Plymouth, figs. 19, 20, PI. 36).

In this embryo, continuity has been established between themain body of the orbital cartilage and its preoptic root. Thereis still, however, a vacuity between the preoptic root and themain portion of the planum antorbitale, dorsal to the orbito-nasal foramen, due to delayed chondrification. The suborbitaland supraorbital cartilages are better developed than inprevious stages, and the film of cartilage between the laminaorbitonasalis and the palato-quadrate is still present. Thesupraorbital cartilage encloses between itself and the dorsaledge of the orbital cartilage the twigs of the superficial ophthal-mic branch of the facial nerve, innervating the supraorbitallateral line canal. The existence of these foramina, and of thearch further back through which the lateral line canal itselfpasses, is evidence in favour of the view that the supraorbitalcartilage, like the suborbital cartilage, has an independentorigin, and becomes attached to the orbital cartilage. Thepostpituitary commissure shows a small variation in that it isstill unattached to the polar cartilages. The precarotid com-missure, and the processes from the foremost ends of the para-chordals which go to form the dorsum sellae, are present. Alabial cartilage has appeared in the upper jaw, by the side of and


close to the anterior portion of the palato-quadrate. Its partnerin the lower jaw does not appear until later.

TEXT-FIGS. 14-17.

Tig. 14.—Section 6-1-3. Fig. 16.—Section 6-2-1.rig. 15.—Section 6-1-8. Fig. 17.—Section 6-2-6.

S t a g e 1 3 (45 mm., Naples, figs. 21, 22, 23, 24, PI. 37).

This stage is the oldest studied in this paper, and it leads oneasily to the conditions in the adult. But, like all the Naplesmaterial, it is in some respects more retarded than smallerembryos from Plymouth; for instance, the basidorsals and

614 G. R. DE BEER

basiventrals are still separate, whereas they were fused in theirrespective segments in the embryo described in stage 9.

The nasal capsule is further developed, and a considerableportion of its front and side walls is noAV formed. A pair ofcartilages rise up from the rostrum on each side, forming themedian wall of each nasal capsule, and, running upwards andforwards, are attached to the dorsal part of the front wall ofthe capsule. From near their point of attachment to this wall,the lateral rostral processes project forwards. The originalrostral cartilage extends forwards as the median rostral process,and, dorsal to it, a shorter and more blunt process extendsforwards between the median walls of the nasal capsules (thesupra-rostral process). The nasal cartilage is unattached to thefront wall of the nasal capsule, and its outermost prong is nowdetached from it, and forms a small separate cartilaginous curvedstrut round the outer and hinder sides of the external nostril.

The postpituitary and precarotid commissures and the pro-cesses forming the dorsum sellae are as in previous stages, butthe remainder of the hypophysial fenestra is beginning to beobliterated by islets of cartilage which will contribute to formthe trabecular plate. The basicranial fenestra is also reducedby further extension of the parachordal cartilage. A roof isnow present in the form of a small quadrangular cartilage. Thetwo hind corners of this cartilage are continuous Avith the occipi-tal arches, thus forming a tectum posterius. The two frontcorners are almost continuous with processes which extendtowards them from the roofs of the auditory capsules. "Whenthese establish connexion, a tectum synoticum will have beenformed. On each side of the quadrangular cartilage, therefore,there is an unchondrified region which gives rise to the parietalfossa of the adult skull, and which lies immediately dorsal tothe vacuity in the median wall of the auditory capsule throughwhich the posterior (semi-)circular canal bulges.

The spiracular cartilage is now present as a thin transverseplate immediately in front of the spiracle. As Bidewood (1895)showed, this cartilage has nothing to do with the ligamentsuspending the palato-quadrate. This ligament was erroneouslytermed the prespiracular ligament (whereas it is in reality


postspiracular) by Parker (1878). Huxley (1876) showed thatthe spiracular cartilage of H e t e r o d o n t u s answered, as

PSC K 9n iho I iha rfm'p:sc 3n

tho

TEXT-FIGS. 18-21.

Fig. 18.—Section 6-3-1. Fig. 20.—Section 6-4-1.Fig. 19.—Section 6-3-5. Fig. 21.—Section 6-4-8.

regards its relations, to the otic process of the frog, with whichhe regarded it as homologous.

The visceral arch skeleton is now complete; consistingof palato-quadrate, Meckel's cartilage, hyomandibula, cerato-hyal, five pharyngobranchials (the last two being joined),five epibranchials, five ceratobranchials, three hypobranchials

616 G. R. DE BEEE

corresponding to the first three branchial arches, and a mediancartilage representing the fused hypobranchials of the last twoarches, and a median basihyal which is still perforated by aforamen for the remains of the thyroid stalk. Four dorsal andthree ventral extrabranchial cartilages are present, as well asthe hyal and branchial rays. The hyal rays are grouped intotwo lots, the more dorsal being attached to the hyomandibulaand the more ventral to the ceratohyal.

It is interesting to compare the median views of longitudinalsections of the skull at this stage with those of the adult(figs. 24 and 25, PI. 37). The embryo illustrated in fig. 24,PL 37, is slightly younger than that in figs. 21 to 23, PL 37. Ithas been dissected and cut longitudinally so as to show themedian surface of the auditory capsule. The median wall of thecapsule is not complete, but an important part in its formationis taken by the dorsal process of the parachordal, which separatesthe auditory from the glossopharyngeal nerves. It is alsopossible to see how the posterior wall of the capsule has becomeattached to the occipital arch, and overlies the lamina hypotica,thus obliterating the fissura metotica between the glossopharyn-geal and vagus nerves. In the adult, the median wall of theauditory capsule has been completed, except of course forthe auditory foramen, the endolymphatic foramen, and for avacuity just behind the latter through which the posterior canalof the membranous labyrinth bulges. This is the region of theparietal fossa, where the roof of the brain-case dips downbetween the tectum posterius and the tectum synoticum, andthere is a vacuity in the roof on each side. As may be seen fromthe transverse sections, the posterior canal vacuity in the wallof the auditory capsule communicates, not with the cranialcavity, but with the tissues outside it, for the vacuity is shutoff from the cranial cavity by the dura mater, which in thisregion is unchondrified. The glossopharyngeal foramen is notmorphologically a perforation of the median wall of the capsule,but a remnant of the fissura metotica, between the capsule andthe parachordal plate. The relations of the cartilages to thecanal for the glossopharyngeal nerve are also shown in figs. 26-8,PL 36, in which it may be seen how the lamina hypotica forms


as it were a false bottom to the hinder part of the auditory-capsule.

Pigs. 24 and 25, PL 87, also serve to show how the post-pituitary and precarotid commissures and the processes formingthe dorsum sellae combine to form the well-marked ridge ofthe adult, overhanging the pituitary fossa (sella turcica) frombehind, and enclosing the pituitary vein in the cartilaginousinterorbital canal.

The A u d i t o r y C a p s u l e .

The study of the auditory capsule of ScyI l ium throwslight on two problems of interest, concerning the median wallof the capsule and the relations to the capsule of the glosso-pharyngeal nerve.

Gegenbaur (1872, p. 49) described the pit in the roof of theskull at the point towards which the anterior and posteriorsemicircular canals converge as the parietal fossa. At thebottom of this pit he said that there were two openings on eachside leading into the auditory capsule, but gave no furtherdetails concerning them. Parker (1878, p. 208) refers simplyto the ' aqueducts of the vestibule', yet his fig. 4 on PL xxxviiiappears to show the correct relations, although unlabelled.Wells (1917) and Daniel (1922) refer to openings for perilym-phatic spaces, by which they mean the vacuities through whichthe posterior semicircular canals protrude from the capsules.Quite recently, Norris (1929) has alluded to these vacuities, andtraced the history of their study at the hands of Geoffroy (1780),Monro (1785), Scarpa (1789), and Weber (1820). Scarpa re-garded the vacuities as equivalents of the fenestra ovalis of theear of higher forms; Weber considered them as the fenestrarotunda.

At the outset, it is necessary to realize that the auditoryorgan of Selachians, as admirably set forth by Eetzius (1881),differs in certain remarkable respects from that of other verte-brates. In particular, the posterior semicircular canal is worthyof attention. Instead of its anterior limb being joined on to theposterior limb of the anterior semicircular canal to form a cruscommune, the posterior canal in Selachians practically forms

NO. 296 S S

618 G. E. DB BEER

a complete circle or ring, apposed to the postero-medial surfaceof the saccule and utricle, and with the cavity of which itcommunicates by means of the ductus canalis semicircularis

Fig. 22.—Section 6-4-8.Pig. 23.—Section 7-1-3.

Fig. 24.—Section 7-2-5.Fig. 25.—Section 7-4-4.

posterioris. These relations are shown in the text-figures ofsections which accompany this paper. The lateral portion ofthe posterior canal is lodged in the cartilaginous capsule, butits median portion protrudes through the vacuity in the median


wall of the capsule. This vacuity (the posterior canal vacuity),places the cavity of the auditory capsule in communicationwith the surrounding regions, but not with the cranial cavity.As may easily be seen in the sections, the part of the posteriorcanal which bulges out through the vacuity is excluded fromthe cranial cavity by the dura mater, which in this region is notreinforced by cartilage. When in the dry skull a seeker is passedthrough the parietal fossa into the cranial cavity, it does notrespect the boundaries of the true cranial wall. Morphologicallythe parietal fossa opens only into the auditory capsule. InScy I l i um the vacuity through which the posterior semi-circular canal protrudes appears to be occluded only by themembranous wall of the canal itself, and not by a definitemembrane. Headers are referred to Norris's paper (1929) foran account of the conditions in other forms.

The existence of the posterior canal vacuity brings it aboutthat the median wall of the auditory capsule is incomplete fora certain short distance, and in this region there is therefore nocontribution from the auditory capsule to the formation of theroof of the skull. Posteriorly to the vacuity the roof of the skullis formed by the tectum posterius; anterior to it is the tectumsynoticum. The parietal fossa owes its existence to the presenceof the posterior canal vacuity.

Wholly distinct from the posterior canal vacuity, yet like itopening into the parietal fossa, is the foramen for the ductusendolymphaticus. It lies in front of the posterior canal vacuity,the ductus endolymphaticus and the posterior canal beingseparated by a wall of cartilage. In H e t e r o d o n t u s Norris(1929) states that the posterior canal vacuity has probablybecome confluent with the endolymphatic foramen, a conclusionwith which I agree.

Gegenbaur (1872) realized that the glossopharyngeal nerveon its passage through the wall of the skull traversed the cavityof the vestibule of the auditory capsule. Since primitively thenerve must have passed behind the capsule, it is of interest tosee how the conditions which prevail in Selachians may beexplained. The problem is identical with that concerned withthe passage of the facial nerve through the anterior region of

S S 2

620 G. R. DB BEER

the auditory capsule in U r o d e l a (Goodrich, 1930), and is tobe explained in the same way. Briefly what has happened is thata lateral shelf from the parachordal cartilages has extendedbeneath the floor of the auditory capsule. Originally theglossopharyngeal nerve passed out between this shelf (laminahypotica) beneath, and the floor of the capsule above, in a spacewhich represents the fissura metotica, and behind the basi-vestibular commissure, represented in Selachians by the dorsalprocess of the parachordal plate. Then the floor of the auditorycapsule became reduced and fenestrated while the laminahypotica persisted, and formed a vicarious lower cartilaginousboundary to the capsule. In one place the lamina hypoticadoes not quite reach the true floor of the capsule, just in frontof it, with the result that there is a gap in the cartilage, as maybe seen in Text-fig. 19. Behind this point the lateral edge ofthe lamina hypotica is confluent for a short distance with thelower edge of the outer wall of the capsule, but histologicaldifferences in the cartilage mark the limit to which each extends.Farther back again, the lamina hypotica extends freely to theside beneath the capsule which in this region has its true floorchondrified again. The lamina hypotica here forms the lowerborder of the glossopharyngeal notch. Meanwhile the cavityof the capsule and that portion of the fissura metotica whichacts as a glossopharyngeal canal are thrown into one, and theboundary between them is indicated only by some remnantsof membrane. At the same time the capsule has bulged back-wards to accommodate the large posterior canal, and so ithappens that the glossopharyngeal nerve in Selachians seemsto penetrate the cavity of the auditory capsule itself, passing infront of the posterior canal, on the inner hinder side of thelagena, and to run for a short distance close to the posteriorbranch of the auditory nerve. A minute examination of sectionsreveals traces of the obliterated fissura metotica, through whichboth glossopharyngeal and vagus nerves typically pass.

The disentangling of the relations of the glossopharyngealnerve to the auditory capsule in Selachians is important inconnexion with the conditions in the Teleostomes. As is wellknown, in the latter animals the median wall of the auditory


capsule is much reduced. Now, in the young embryo of Aci-p e n s e r s t e l l a t u s , the fissura metotica is present, the vagusand glossopharyngeal nerves traversing it in typical manner.At a later stage, however, the fissura metotica is divided bya fusion of the lateral edge of the parachordal with the medianwall of the auditory capsule into an anterior glossopharyngealforamen and a posterior typical vagus foramen. A true basi-vestibular commissure is present, forming the junction betweenparachordal and capsule, anterior to the glossopharyngeal.The fusion alluded to in the previous sentence is really formedby the cartilage which gives rise to what Gaupp (1906) hascalled the posterior basicapsular commissure. This commissurein Sa lmo lies behind the glossopharyngeal nerve, and for sometime there is no true basivestibular commissure, with the resultthat the glossopharyngeal nerve is enclosed in the basicapsularfenestra. The basicapsular fenestra of Sa lmo therefore con-tains a portion of the fissura metotica, and is not strictlyequivalent to the basicapsular (or basicochlear) fissure of highervertebrates, which is situated wholly in front of the basi-vestibular commissure, contains therefore no part of the fissurametotica, and is not traversed by the glossopharyngeal nerve(Text-fig. 26).

The absence of a median cartilaginous wall to the auditorycapsule, and the inclusion of the glossopharyngeal nerve in thebasicapsular fenestra in S a l m o , are responsible for the passageof the nerve through the capsule in that animal, for the mem-branes indicative of the true boundaries of the capsule havemostly broken down. This condition would appear, however,not to be universal in the Teleostei, for in Gasterosteus, Swinner-ton (1902) states that the cartilage of the auditory capsule 'iscontinuous both with the lateral process of the occipital archand with the postero-lateral border of the mesotic region, thusforming a complete boundary around the exit for the ninth andtenth nerves'. Here, then, the glossopharyngeal nerve clearlydoes not traverse a basicapsular fenestra, but accompanies thevagus through the fissura metotica. Gasterosteus, therefore,lacks a posterior basicapsular commissure, and the continuityof the cartilage of the auditory capsule with the ' postero-lateral

622 G. B. DB BEER

border of the mesotic region' must represent a true basivesti-bular commissure.

THE EELATIONS OF THE JAWS TO THE BRAIN-CASE

As is well known, Huxley (1876) in his classical memoireffected a subdivision of the types of suspension of the jaws into

26

Diagram to illustrate the relations between the glossopharyngealnerve and the auditory capsule in: A, the Selachian (Scyllium);B, the Teleost (Salmo); and C, the Mammal (Lepus).

three categories: viz. amphistylic, hyostylic, and autostylic.As it is matter of great importance to ascertain exactly what hemeant, some extracts from his paper will be quoted here.

(1) 'In the aniphisfcylic skull the palato-quadrate cartilageis quite distinct from the rest of the skull; but it is wholly, oralmost wholly, suspended by its own ligaments, the hyomandi-bular being small and contributing little to its support' (e.g.N o t i d a n u s = Hep tanchus ) .

(2)' The palato-quadrate cartilage is no longer continuous withthe chondrocranium (though the bony elements of that arch(i.e. mandibular) may unite suturally with those of the skull, asin the Plectognathi), but is, at most, united with it by ligament.


Moreover, the dorsal element of the hyoidean arch, or thehyomandibular, usually attains a large size and becomes thechief apparatus of suspension of the hinder end of the palato-quadrate cartilage with the skull. Skulls formed upon this type,which is exemplified in perfection in Ganoidei, Teleostei, andordinary Plagiostomes, may therefore be termed h y o s t y l i c '

(3) ' The part of the palato-quadrate cartilage which is unitedwith the skull, between the exits of the fifth and second nerves,answers to the "pedicle of the suspensorium" of the amphibian,while its backward and upward continuation on to the perioticcartilage corresponds with the otic process. As in the Amphibiaand in the higher Vertebrata, the mandibular arch is thusattached directly to the skull by that part of its own substancewhich constitutes the suspensorium. It may thus be said tobe a u t o s t y l i c '

While there can be little ambiguity about the term hyostylic(although it needs to be analysed further), the same is not trueof the terms autostylic and amphistylic. As regards theautostylic condition, it may mean simply that the mandibulararch is not dependent on the hyoid arch for part or most ofits suspension from the brain-case, or it may mean that themandibular arch is fused by means of one or more of its ownprocesses on to the cartilage of the brain-case. It is true thatliving fish do not present one of these alternatives without theother, but as the former may be a primitive feature and thelatter is probably secondary, there is reason to suspect that inextinct forms these alternatives were not inseparably associated.It is not quite clear from Gadow's (1888) descriptions whichmeaning he attached to the term 'simple autostylic form'.

It would seem that Huxley himself laid most stress on thefusion of the jaws with the brain-case, for, having finished withthe description of the autostylic forms, he goes on to say thatin all other fishes 'the palato-quadrate cartilage is no longercontinuous with the chondrocranium', suggesting therefore thatcontinuity is an essential feature of his autostylic category.This view of the meaning of the term autostylic is also sharedby Smith Woodward (1898) and Gregory (1904).

For the sake of clearness, the term autostylic will be taken

624 G. E. DB BEER

to mean this alternative, viz. that condition in which the mandi-hular arch is fused to the brain-case in the adult form. In suchforms, the hyomandibula plays no part in the suspension, butthis would seem to be because the hyoid arch has been reducedfrom a previous large and suspensorial condition, and notbecause the hyoid arch has not yet acquired a suspensorialfunction.

A difficulty arises from the fact that the palato-quadratecartilage has no less than four processes, viz. ethmoid, basal,ascending, and otic, by means of which suspension may beeffected, and it becomes necessary to distinguish betweenthem. H e p t an c h u s is amphistylic and attached by the basaland otic processes; H e t e r o d o n t u s (Cestracion) is attachedonly by the basal process (and of course the hyomandibula);S a 1 m o is attached only by the ethmoid process (and hyo-mandibula).

Another question which Huxley's classification raises is due tothe fact that it is essentially a functional classification, since it isprimarily based, not on whether a particular structure, say theotic process, exists, but on whether such a structure actuallyserves in the suspension of the jaws. Now in some forms theotic process exists, but it is too small to effect a functionalsuspension, as in S c y m n u s and A m i a, or it may be detachedfrom the palato-quadrate cartilage to form the spiracularcartilage, as in S c y l l i u m . Clearly S c y m n u s , Amia , andS c y l l i u m cannot be placed in the same group as H e p t a n -chus on Huxley's classification, and yet the possession of anotic process (albeit diminutive) deserves a recognition of theaffinity expressed between these forms and H e p t a n c h u sby the possession of a homologous structure. It seems, therefore,that while Huxley's scheme may most certainly be retained inorder to indicate the type of suspension of the jaws, anotherclassification based purely on morphological considerations isnecessary to express the phylogeny of the suspension.

An attempt was made to revise the classification of the typesof jaw-suspension by Gregory (1904), according to the followingsystem.


" Type.

Ancestral

Chlamydoselache

ScylliumHeterodontus

Heptanchus

RajaAcipenserPolypterus

Salmo

Chimaera

Ceratodus

Amphibia

Palato-quadrate.

Little or no attach-ment

Loose articulation(basal)

No articulationClose articulation(basal)

Close articulation(basal-)-otic)

Quite freeQuite freeArticulation (eth-moid)

Articulation (eth-moid)

Close fusion

Fusion (basal, as-cending, otic)

Fusion (ethmoid,basal, ascending,otic)

Hyomandibida.

Not suspen-sorial

Suspensorial

SuspensorialSuspensorial

Suspensorial

SuspensorialSuspensorialSuspensorial

Suspensorial

Not suspen-sorial

Not suspen-sorial

Not suspen-sorial

Term.

Palaeostylic

Hyostylic

HyostylicHyostylic

Amphyostylio

EuhyostylicMethyostylicMethyostylic

Methyostylic

Holostylic

Autostylic

Autostylic

A more recent classification is that of Edgeworth (1925) whichruns as follows:

Scyll ium Hyostylic.S q u a 1 u s Amphistylic and streptostylic.Heptanchus Autostylic and streptostylic.Heterodontus Amphistylic.Acipenser Hyostylic.Polyodon Hyostylic.Polypterus Hyostylic.A m i a Amphistylic.Lep idos teus Amphistylic.Salmo Autostylic anteriorly and hyostylic pos-

teriorly.

Stannius (1856) coined the terms ' streptostylic' and ' monimo-stylic' to express the mobility of the quadrate in the bonyskull. Furbringer (1900 and 1904) applied these terms, as doesEdgeworth, to the ca r t i l ag inous skull, to express thefreedom or attachment of the palato-quadrate to the brain-case.

626 G. R. DE BEER

Fuchs (1915) has pointed out that this extension of the use ofthe terms carries with it an alteration in their meaning. Andsince Versluys (1912) has shown that in the bony skull strepto-styly and monimostyly only constitute special cases of a morecomprehensive phenomenon, i.e. 'kinetism' and ' akinetism',it seems inadvisable to use the same expression streptostylicto denote the freedom of the palato-quadrate in the (cartila-ginous) skull of the Selachian and the mobility of the quadratein the (bony) skull of the Eeptilian. The latter meaning haspriority.

But be this as it may, Fiirbringer regarded the mobility ofthe palato-quadrate as the primitive condition (rIt is so-to-speak a morphological necessity', 1904, p. 585), whereas Edge-worth thinks that the fixed palato-quadrate is the moreprimitive. Before the question of the suspension of the jaws inthe cartilaginous skull can be tackled, attention must be turnedto some points of general principle. These are concerned with(1) the value of embryo logical evidence on questions of phylo-geny, and (2) problems of phylogeny and nomenclature.

Edgeworth is impressed by the fact that the joint between thepalato-quadrate cartilage and the neurocranium in S q u a l u sand L e p i d o s t e u s in the region of the basal process is pre-ceded in development by cartilaginous continuity; the samebeing probably true of H e p t a n c h u s also. This leads him toconclude that ' Selachii, Batoidei, and Teleostomi are descendedfrom autostylic and monimostylic ancestors in which there wasa pterygo-quadrate united to the Chondrocraniuru at threepoints' (1925, p. 257). This conclusion presupposes, that condi-tions which are embryonic in development must representancestral adult conditions in phylogeny. Elsewhere (de Beer,1930 A) I have endeavoured to draw attention to the fact thatthis conclusion is not sound, whether by reference to empiricalevidence, or by logical deduction from other available evidenceconcerning embryology and evolution. It is no more legitimateto say that the adult ancestral vertebrates were autostylicbecause in early stages of development of some forms a transientcartilaginous continuity exists between the palato-quadrate andthe neurocranium, than it would be to assert that the ancestral


adult nervous system was a solid rod because it develops as asolid rod in P e t r o m y z o n , L e p i d o s t e u s , Teleostei, andLepidos i ren . 1

There is then no proof that the ancestral vertebrates had jawsfused with their brain-case, and the facts that the only suchforms are those which are most specialized in other respects, andthat the postmandibular visceral arches are not confluent withthe axial skeleton, point rather to the view that the splanchno-cranium and neurocranium were originally distinct.

Now turning to questions of phylogeny and nomenclature, it1 The same criticism may be made in respect of the conclusions which

Edgeworth has based on his admirable researches into the method of originof the masticatory muscles. That the primordium of the muscles of the mandi-bular arch in non-Dipnoan fish should be divided into a dorsal (Constrictori dorsalis) and a ventral portion (Adductor mandibulae), while in Dipnoi andAmphibia the primordium is not so divided (a point which, incidentally, iscontested by Luther, 1914), i.e. remains in the condition through which thenon-Dipnoan fish pass, is no evidence that the Dipnoi and Amphibia arephylogenetically more primitive than the non-Dipnoan fish. It cannot beargued that because Caducibranchiate Urode la pass through a stage atwhich Perennibranchiates remain, the latter are the more primitive. Indeed,the secondary nature of the retention by adult descendants of characterswhich were larval or embryonic in the ancestors, is becoming more and moreclear. This paper is not concerned with muscles, but this matter must bementioned, for Edgeworth seeks to substantiate his view that the ancestralvertebrates were autostylic by means of the conclusions which he has drawnfrom his investigations into the development of muscles. As the3r stand, theseconclusions are inadmissible, and they receive final refutation from his furtherattempt to make the evidence from the development of the extrinsic eye-muscles coincide with them. In Selachians the eye-muscles develop, as iswell known, from the first three somites, and they preserve their innervationby the ventral nerve-roots of these segments. In Teleostomes the develop-ment is more obscure, and Edgeworth believes that they develop from thofirst and second somites only, while in Dipnoi and Amphibia he derives themfrom the first somite alone. These results lead him to conclude (1926, p. 32)that 'the conditions in Teleostomi and Plagiostomi may he consideredsecondary and tertiary modifications of those in Dipnoi, U r o d e l a , andAnura ' . As this view is inconsistent with van Wijhe's theory, Edgeworth(1928, p. 46) rejects the latter. It is only because Edgeworth has not dissoci-ated the conceptions of primitive and secondary in plrylogeny from those ofearly and late in ontogeny that he has thus been led to read the series Selachii—Teleostomi-Amphibia in a direction which is the opposite of that to which allthe evidence of zoology and palaeontology point.

628 G. R. DE BEER

is clear that the latter is most instructive when based on theformer. Further, the establishment of a phylogenetic seriescannot be based satisfactorily on a consideration of only oneor two sets of structures, although a single set of structures maybe very useful in showing when a series is n o t phylogenetic.All the evidence of zoology shows that the Chondrichthyes areat a lower and more primitive level of evolutionary organizationthan the Osteichthyes, which, in turn, are at a lower level thanthe Tetrapoda. In order to obtain the evidence which the sus-pension of the jaws presents on this problem, it is necessary,therefore, to examine the different structures which may beconcerned with suspension in the various forms, to establishtheir homology or non-homology by the m o r p h o l o g i c a lcriterion of their geometrical relations to neighbouring struc-tures, and to compare the results thus obtained. For reasonsstated above, e m b r y o l o g i c a l evidence cannot be reliedupon, since it is not possible to a s s e r t that ontogeneticfeatures have phylogenetic significance.

It may now be taken as accepted that Huxley's view is correctin regarding the basal process of the palato-quadrate as theoriginal dorsal end of the mandibular arch. Swinnerton (1902)and Sushkin (1927) have stressed the fact that the basal processis very persistent among the various types of fish, and Goodrich(1930) has summarized the evidence from other quarters infavour of Huxley's view. For a long time it seemed doubtfulwhether the orbital process of the Selachian represented anethmoid or a basal process. But the existence of an ethmoidarticulation in front of the orbital process in S c y m n u s andC h l a m y d o s e l a c h e , together with the fact that the ethmoidprocess forms the most anterior extremity of the palato-quadratecartilage in Teleostomes and higher forms, renders it reasonablycertain that the orbital process of Selachians is a basal process.This process may be situated far forward as in S c y I l i u m , orfar back as in S c y m n u s , depending on the relative lengthsof the jaws and the brain-case.

Began (1923) and Norman (1926) believe that the basalarticulation of L e p i d o s t e u s is not primitive, but has beensecondarily acquired. But in view of the prevalence of this


articulation throughout the vertebrates, and of the fact that itsmorphological relations in L e p i d o s t e u s are identical withthose of other forms, its independent acquisition by L e p i -d o s t e u s is hard to believe.

Before considering the distribution of the basal and otherprocesses in the various groups, attention must be turned tothe question as to whether the anterior visceral arches, i.e. themandibular and hyoid arches, ever showed the division intofour elements characteristic of the more posterior (or branchial)arches. The palato-quadrate and hyomandibula seem to repre-sent the epal elements of their respective arches, just as Meckel'scartilage and the ceratohyal appear to represent the ceratalelements. If such be the case, then it may be supposed that thepharyngomandibular, pharyngohyal, hypomandibular, andhypohyal elements originally existed. On the other hand, ifthe anterior visceral arches never were divided into four piecesbut only into two (i.e. palato-quadrate and Meckel's cartilage—•hyomandibula and ceratohyal), as Gegenbaur (1872) believed,the above-mentioned elements could not have existed.

Now, paired cartilages which seem to be hypohyals are presentin Sa 1 mo and in L a c e r t a , and the so-called 'hyoid' of therabbit shows evidence of paired origin. Among Selachii, hypo-hyals have been described in L a e m a r g u s by White (1895),and in H e p t a n c h u s by Braus (1906) and Piirbringer (1903).A pharyngohyal is present in C a l l o r h y n c h u s (Schauinsland,1903), and Luther (1909) claims to have found it in S t e g o -s t o m a , M u s t e l u s , and G a l e u s . There seems, therefore,to be considerable evidence in favour of the view that the hyoidarch was originally similar to the more posterior arches in beingdivided into four elements. It is probable that the disappearanceof the pharyngohyal was associated with the acquisition by thehyoid arch of the function of suspending the jaws, and it isinteresting to note that in C a l l o r h y n c h u s , the only formin which the pharyngohyal is at all well developed, the hyoidarch has no suspensory function.

As regards the mandibular arch, a hypomandibular has beendescribed in H e p t a n c h u s and L a e m a r g u s by White(1895), and in H e x a n c h u s by Fiirbringer (1903) and Sewert-

630 G. R. DE BEER

zoff (1927). Jaekel (1927) has described a hypomandibular inAcanthodes, and van Wijhe (1922), who observed that Meckel'scartilage has two centres of chondrification in S q u a l u s (andalso in birds), draws attention to the possibility of interpretinghis discovery in this way.

A pharyngomandibular is claimed in C a l l o r h y n c h u s byDean (1906), and an element corresponding to it in S c a p h i r -h y n c h u s by Sewertzoff (1923) and in L a e m a r g u s bySewertzoff and Disler (1924). In 1923 Sewertzoff suggested thatthe pharyngomandibular might be represented by the orbital( = basal) process of the palato-quadrate. In the following year,however, he altered his opinion, since he claimed that a pharyn-gomandibular arises separate from the orbital process in Mus-t e l u s , S c y l l i u m , and S q u a l u s , and later becomes fusedon to the median side of the process. Bugajew (1929) has re-ported cartilages which he regards as pharyngomandibulars inthree species of A c i p e n s e r . But the curious thing aboutthese elements is that they do not appear to be constant.A c i p e n s e r g u l d e n s t a d t i has two such nodules on eachside situated in the ligament which stretches between the basi-trabecular process and the palato-quadrate; A c i p e n s e rs t e l l a t u s , on the other hand, may have 4, 5, or 6, or 1 smalland 1 large nodule.

It is clear that the case for the pharyngomandibular cannotbe regarded as satisfactory, and the matter is made more difficultby Allis's (1923 A) claim, based on morphological considerations,that the pharyngomandibular is to be found in the polarcartilage of van Wijhe. In the present state of knowledge itis not possible to arrive at any definite conclusion regardingthe original condition of the mandibular arch. The basalprocess, regarded as the original dorsal end of the arch, maytherefore represent the end of the pharyngomandibular or ofthe epimandibular. Fortunately, this question is not of greatimportance for the following discussion.

The basal process is present in S c y l l i u m , H e p t a n c h u s ,S e y m n u s , S q u a l u s , H e t e r o d o n t u s , and other Squa-loids, but it has been lost in the Batoids; it was present inAcanthodii and Pleuracanthodii (Goodrich, 1909); there was

SKULL OF SCYLLIUM G31

a basal articulation in Osteolepidoti (Watson, 1926), in Coela-canthini (Stensi0, 1922), and in Palaeoniscoidei (Stensi0, 1921;and Watson, 1928); remnants of a former basal articulationare found in the form of a basitrabecular process in A c i p e n s e r(de Beer, 1925; and Sewertzoff, 1928); a basal articulation ispresent in L e p i d o s t e u s (Parker, 1882; Veit, 1907 and 1911;and de Beer, 1926); remnants of a basal process are present inAmi a (Pehrson, 1922; de Beer, 1926) and in Sa lmo (de Beer,1927); the basal process fuses with the neurocranium in Dipnoiand some Amphibia.

The otic process is present in H e p t a n c h u s , fossil Hetero-donti (Smith Woodward, 1898), Acanthodii and Pleuracan-thodii (Goodrich, 1909), Holocephali; a non-articulating oticprocess was present in Osteolepidoti (Watson, 1926) palaeoni-scoidei (Stensi0, 1921); vestiges of an otic process are visiblein P o l y p t e r u s (Sewertzoff, 1926A), L e p i d o s t e u s (Parker,1882), Amia (de Beer, 1926), and Sa lmo (de Beer, 1927);the otic process is fused with the auditory capsule in Dipnoiand some Amphibia.

The ethmoid process articulates with the neurocranium inScymnxis (Bugajew, 1930) and in Chlamydoselache (Allis,1923), and in all Teleostomes except Acipenseroidei; the ethmoidprocess is fused with the neurocranium in some Amphibia.

The ascending process is foreshadowed in Osteolepidoti(Watson, 1926), and it is fused with the neurocranium in Dipnoiand Amphibia.

The hyomandibula is suspensorial in all known fish exceptHolocephali and Dipnoi, and in no other forms.

It would seem that the above-mentioned facts may best beset into order as follows.

The most primitive suspending element is the basal process,or original dorsal end of the mandibular arch.

It is possible that in the earliest forms there was also a pre-mandibular arch, and the dorsal ends of these arches may beregarded as being loosely attached to the axial skeleton, inthe form of the notochord-sheath. For such an arrangement,Gregory's term P a l a e o s t y l i c may be used, and the Cyclo-stome condition as exemplified by Petromyzon may be regarded

632 G. R. DE BEER

as having been derived from it by fusion of the dorsal ends ofthe arches with the cartilage of the neurocranium, thus pro-viding a firm framework for the very specialized method offeeding by means of a rasping tongue which these forms possess.Should a term be required to denote this condition, P a r a u t o -s t y 1 i c may perhaps be proposed, suggesting that the fusion ofthe arches with the brain-case is quite independent of the trueautostyly of Dipnoi and Amphibia.

The next problem is to decide which of the two methods ofsuspension, by means of the otic process or of the hyomandibula,was evolved first. Fossils give little assistance here, for theearliest forms (Pleuracanthus, Acanthodes, Cladoselache, andHeterodonti) were amphistylic, and possessed both otic andhyomandibular suspensions (in addition to the basal process).An indication may, however, be obtained from a considerationof the Holocephah. These fish possess an otic process, but thehyomandibula is not only not suspensorial, the hyoid arch seemsto have retained its primitive dorsal element, the pharyngohyal.This so-called pharyngohyal might in reality be the firstpharyngobranchial, thus shifting the pharyngobranchials backone place. But a careful examination of Hubrecht's (1877) andSchauinsland's (1903) descriptions renders this unlikely. Theexistence of a pharyngohyal means that the Holocephah weredescended, e i t h e r from forms with an otic process and a non-suspensorial hyomandibula, or from forms in which thehyomandibula was suspensorial and subsequently lost thatfunction and reacquired the pharyngohyal (for the presence ofa pharyngohyal is inconsistent with a suspensorial hyomandi-bula). As the latter of these alternatives is unlikely, the onlytenable opinion is that the forms from which the Holocephahdescended had otic (and basal) processes, and a non-suspensorialhyoid arch, still retaining the pharyngohyal. The spiracularslit presumably was large, extending below the joint betweenhyomandibula and ceratohyal. Such a form is not separatelyprovided for in Huxley's scheme; it was not autostylic in theadopted sense because the arches cannot have been fused withthe brain-case, and it was not amphistylic since there was nosupport from the hyomandibula. The simplest way in which


to denote such a type would be to call it B a s i o t o s t y l i c ,thus drawing attention to the elements by means of which thesuspension is effected.

Prom this type the condition found in the Holocephali wouldhave been derived simply by fusion of the arch with the brain-case. Gregory's term H o l o s t y l i c is convenient as expressingthe completeness of the fusion, and because of its resemblanceto Holocephali. Prom the Basiotostylic type, the A m p h i s t y l i c(' Basiotohyostylic') condition was derived simply by the partici-pation of the hyoid arch in the suspension, with the consequentloss of the pharyngohyal and reduction in size of the spiracle.Here belong the primitive Selachians, Pleuracanthodii, Acan-thodii, and Cladoselachii. By a subsequent loss of the contactbetween the otic process and the auditory capsule consequentupon a reduction in the size of the former, or its separationfrom the palato-quadrate as a spiracular cartilage, the Hyo-s t y l i c forms were derived, as exampled by most modernSqualoid Selachians, e.g. S q u a l u s or S c y l l i u m ('Basihyo-stylic'), and in exaggerated form by the Batoids: the mostspecialized among these fish (' Euhyostylic'). In some Selachianscontact may also take place between the ethmoid process (fore-most extremity of the palato-quadrate cartilage) and the brain-case, as in S c y m n u s and Chlamydoselache. This scheme,in regarding the Hyostylic forms as descended from Amphistylicancestors, agrees Avith what palaeontological evidence there is,but not with Regan (1906) who considers that the families ofSelachians without an otic process are more primitive than thosewhich possess one.

The condition in the primitive Osteichthyes must have beenas follows. The otic and basal processes must have been present,suspending the jaws, and the ethmoid process also, for it isalmost universal in these forms. The hyomandibula must havebeen suspensorial, since it is difficult to imagine this conditionto have been lost and reacquired. It follows, therefore, thatthese hypothetical primitive Osteichthyes were derived fromAmphistylic forms simply by establishing the ethmoid articu-lation. Two additional points should be noticed. One is thatin all Osteichthyes the hyomandibula articulates with the

NO. 296 T t

634 G. R. DE BEER

auditory capsule dorsal to the vena capitis lateralis instead ofventral to it as in the Selachians. This condition was reachedby means of the formation of a jugular canal covered over bya cartilaginous bridge (lateral commissure, de Beer, 1926) acrosswhich the head of the hyomandibula was able to move up toits more dorsal position (Stensio, 1925). The other point is thatthe ascending process must have made its appearance at aboutthis time, for not only is it found in the Dipnoi but it is also fore-shadowed in the most primitive Teleostomes: the Osteolepidoti(Watson, 1926). The primitive Osteichthyes may therefore betermed P a r a m p h i s t y l i c to emphasize the fact that theirsuspension differs from that of their Amphystylic ancestors inthe relations of their hyomandibula. These forms must beregarded as the common ancestors of Osteolepidoti, Palaeonis-cids, and other Teleostomes, Dipnoi, and Tetrapoda (Watson,1925 and 1926). The Dipnoi lost the suspensorial function ofthe hyomandibula, and effected fusion between the basal,ascending, and otic processes and the brain-case, resulting inthe A u t o s t y l i c type. The Amphibia are similar, with theaddition of a fused ethmoid process in many forms. But sincein most Amniota the palato-quadrate is not fused with the brain-case, the question arises as to whether the primitive Tetrapodawere not Paramphistylic, in which case the Autostylic conditionof Dipnoi and modern Amphibia might have been independentlyacquired. This is the opinion of Luther (1914), and it agreeswith Watson's (1926) demonstration that living Dipnoi andAmphibia have become greatly specialized along convergentindependent lines.

The Teleostomes diverged from the Paramphistylic typesimply by losing the contact between the otic process and theauditory capsule, and preserving the ethmoid, basal, andhyomandibular suspensions. Such a condition, found in Osteo-lepidoti, Palaeoniscoidei, and L e p i d o s t e u s , may be termedT e l e o s t y l i c ('Ethmobasihyostylic), characteristic of Teleo-stomes. A further stage in the reduction of the suspensions isshown by P o l y p t e r u s , Amia , and the Teleostei, in whichthe contact between the basal process and the brain-case isalso lost ('Ethmohyostylic').


There remain the Acipenseroidei to be considered. Theseforms have a suspension effected solely by the hyomandibula,

HoLostyLicHoLocephaLi

ParautostyLicCycLostome

AmphistyLicPrimitiveSeLachii

ParamphistyLicPrimitive

Osteichthues

EthmohuostuiicModernTeLeostomi

Amniota2 7

TEXT-FIG. 27.Diagram to illustrate the probable phylogeny of the types of

suspension of the jaws.

and Sewertzoff (1926 B and 1928) is inclined to derive themdirectly from the Hyostylic Selachians. But not only have

T t 2

636 G. R. DE BEER

Traquair (1887), Smith Woodward (1898), Stensi© (1921), andWatson (1925) shown that the palaeontological evidence pointsto the Acipenseroidei having been derived from Palaeoniscoidei(and therefore from Paramphistylic forms), but the hyomandi-bula of Acipenseroidei articulates with the auditory capsuledorsal to the jugular canal (as in all Teleostomes), and therebydiffers from the condition in the Hyostylic Selachians. Gregory'sterm M e t h y o s t y l i c may therefore be appropriately usedto denote the fact that the suspension in Acipenseroidei iseffected by means of the hyomandibula while drawing attentionto the fact that this condition is not the same as the Hyostylyof Selachians.

Expressed in tabular form, this scheme is as follows:

Type. Suspension effected by. Term.

Ancestral

Petromyzon

PrimitiveChondrichthyes

HolooephaliPrimitiveSelachii

Squaloidei

BatoideiPrimitive

Osteichthyes

Dipnoi andAmphibia

PrimitiveTeleostomes

Teloostei

Acipenseroidei

Dorsal end of arch = basalprocess, attachment loose

Dorsal end of arch = basalprocess, fused

Basal and otic processes, at-tachment loose

Basal and otic processes, fusedBasal and otic processes andhyomandibula, attachmentloose, hyomandibula ventralto jugular vein

Basal process and hyomandi-bula, attachment loose

HyomandibulaEthmoid, basal, and otic pro-cesses, and hyomandibuladorsal to jugular vein, as-cending process small, at-tachment loose

Ethmoid, ascending, basal,and otic processes, fused

Ethmoid, and basal processesand hyomandibula, otic pro-cess reduced,attachment loose

Ethmoid process and hyo-mandibula basal and oticprocesses reduced, attach-ment loose

Hyomandibula(dorsal to jugu-lar vein) attachment loose

Palaeostylic

Parautostylic

Basiotostylic

HolostylicAmphistylic(Basiotohyostylic)

Hyostylic(Basihyostylic)

Hyostylic(Euhyostylic)Paramphistylic

Autostylic

Teleostylic(Ethmobasihyostylic)

Teleostylic(Ethmohyostylic)

Methyostylic


THE PROBLEM OF THE ACEOCHORDAL.

The skulls of nearly all vertebrates are characterized by thepossession of a pituitary fossa which is bounded posteriorlyby a more or less pronounced ridge known as the dorsum sellae.In the development of the various forms this ridge has beencalled the crista sellaris, acrochordal, and prootic bridge, andwhere it is well developed it projects up into the plica encephaliventralis. Behind the dorsum sellae it is very common to finda vacuity between it and the parachordal cartilages. Thisvacuity, which is traversed by the notochord at least in earlystages, is known as the fenestra basicranialis. In those formswhich possess a pila antotiea the dorsum sellae is typically adirect median continuation of the line of the pila antotiea. In1926 I treated the acrochordal, crista sellaris, and prootic bridgeas homologous structures, and the fenestra basicranialis ashomologous right through the vertebrate series.

This view, with which Allis was at one time in agreement, isnow opposed by him in a recent paper (1928). Allis distinguishestwo separate transverse cartilaginous structures at the frontof the parachordal plate. One of these he calls the commissuraacrochordalis, which he regards as situated in the region of thepremandibular segment, and in front of the true parachordals.The other is his conimissura transversa, which is the anterioredge of the parachordals proper, and which he regards as situatedin the mandibular segment. Further, Allis concludes that thevacuity in the basal plate commonly known as the fenestrabasicranialis may be of two kinds: (i) a fenestra prootica median's,between the commissura acrochordalis and the commissuratransversa and according to him to be found in S q u a l u s ,Ac ipense r , P o l y p t e r u s , Amia , and L e p i d o s t e u s ;(ii) a fenestra mesotica medialis, behind the commissuratransversa, and according to him to be found in S a 1 m o,A m i u r u s , S y n g n a t h u s , and L a c e r t a .

Allis admits that both these fenestrae are not described in anyone fish, and, indeed, it would seem that little evidence short ofthis could be really adequate to establish the truth of his con-tention. In the first place, attention must be called to what

G38 G. R. DB BEER

is surely a slip on Allis's part on p. 131 of his paper, where hesays that 'the fenestra basicranialis is . . . simply an openingbetween the trabeculo-polar bars which leads directly into theprimary extracranial subpituitary space'. What Allis is heredescribing is the fenestra hypohyseos, and the term fenestrabasicranialis must be restricted to the vacuity or vacuities inthe basal plate (i.e. the fenestrae prootica medialis and mesoticamedialis, if they are not identical).

Allis's view is open to attack along more than one line. Hewould regard the crista sellaris of L a c e r t a as not homologouswith the acrochordal of S q u a l u s , and yet the similarity inthe relations of these structures to the notochord and to thepilae antoticae, as shown by van Wijhe (1922), Gaupp (1900),and myself (1930 B), is such as to make it difficult to doubt theirhomology. Then again, Allis would deny the homology betweenthe fenestra situated behind the so-called prootic bridgeof A mi a and L e p i d o s t e u s on the one hand and that ofS a 1 m o on the other. But his reasons for doing so would seemto be insufficient. There may be slight topographical variationsbetween the fenestrae in question in these animals, which arehard to control owing to the absence of a pila antotica in S a 1 m oand A mi a and its (perhaps doubtful) presence in L e p i d o -s t e u s , with which to check the position of the prootic bridge.But their morphological similarity argues strongly in favourof their homology. It would indeed be odd if, in forms as closelyallied as are A m i a and S a 1 m o, structures, as similar as thebasicranial fenestrae in these two forms are, were of differentnature.

The relations to the pilae antoticae of the acrochordal ofSelachians and birds, and of the crista sellaris of reptiles, areidentical with those of the dorsum sellae of mammals (man,rabbit) and of the acrochordal cartilage of amphibians andDipnoi. The pilae antoticae in the mammals are of courserepresented by the posterior clinoid processes. The conclusionis, therefore, that the transverse ridge forming the dorsumsellae is homologous throughout the Gnathostome series, be itcalled acrochordal, crista sellaris, dorsum ephippii, or prooticbridge. Consequently, the vacuity in the basal or parachordal


plate immediately behind the dorsum sellae, the fenestra basi-cranialis, must also be homologous throughout.

There are considerable differences in the manner in which thedorsum sellae arises in the different groups of vertebrates, butthey cannot invalidate the conclusions arrived at above. InC e r a t o d u s and Amphibians the dorsum sellae is betterdenned as the hind border of the hypophysial fenestra, becausethe pituitary fossa is not well marked, and it is simply theanterior edge of the parachordal plate, and appears early. InA c i p e n s e r the so-called dorsum ephippii is likewise formedfrom the anterior edge of the parachordals according to Sewert-zoff (1928), and from my observations it would seem to extendfarther up the plica encephali ventralis at later stages ofdevelopment. In birds the acrochordal cartilage is the firstelement of the chondrocranium to appear, as Sonies (1907)showed and I can confirm. Among mammals the dorsumsellae arises in the rabbit at a late stage according to Voit(1909), as a transverse bar of cartilage connecting the twoposterior clinoid processes, which themselves represent the pilaeantoticae. In L e p i d o s t e u s , Amia (de Beer, 1926), andSa lmo (de Beer, 1927) the prootic bridge arises late as amedian plate of cartilage between the anterior ends of theparachordals.

In L a c e r t a (de Beer, 1930B) the crista sellaris arises latefrom paired cartilaginous extensions towards the middle linefrom the foremost points of the parachordals. In Selachiansthe dorsum sellae arises late and has a complicated origin. InS c y l l i u m , as described in this paper, there is formed a post-pituitary commissure and a precarotid commissure, between thepolar cartilages, beneath the notochord, and in the hind regionof the hypophysial fenestra. Later, the dorsum sellae, in theform of paired processes directed towards the notochord fromthe foremost points of the parachordals, forms the anterioredge to the parachordal plate. The conditions in S q u a l u sas described by van Wijhe (1922) are probably, though not quite,identical. He describes a cartilage which has the relations ofthat which is here called the precarotid commissure, and laterthere are formed paired processes which he says come from the

640 G. R. DE BEER

polar cartilages, and therefore correspond with what has herebeen called the postpituitary commissure.

The basicranial fenestra has now been observed in all themain groups of the Gnathostomes except the Dipnoi and theA n u r a . In U r o d e l a it appears to arise by absorption ofthe cartilage of the parachordals, but in all the remainder thebasicranial fenestra is an area of delayed chondrification.

There remains the question as to whether the dorsum sellaeis to be regarded as a part of the true cranial floor. There is nodoubt that this question must be answered with regard to theposition of the true wall of the brain-case as indicated bythe position of the dura mater, and not with regard to the carti-lage of a dried skull. Therefore, although the dorsum sellae doesproject up into the cavity of such a dry skull, since it is formedin intimate association with the dura mater, Allis must be rightin claiming the cartilage of the dorsum sellae as part of the trueskull-floor, and I wrong in denying it (1926).

SUMMARY.

1. The development of the skull of S c y l l i u m c a n i c u l ahas been studied from the first appearance of cartilage, throughthirteen stages, up to the point at which the main features of theadult skull have been acquired.

2. The parachordals are the first elements to chondrify, andevidence is presented confirming Goodrich's observations con-cerning the visible traces of metameric segmentation of themetotic region of the paraohordal.

3. The auditory capsule chondrifies from the first in con-tinuity with the parachordal, to which it is attached by theanterior basicapsular commissure.

4. The polar cartilages have not been found separate, butthey appear as nodules of cartilage attached to the undersurface of the anterior ends of the parachordals.

5. The orbital cartilage becomes attached to the parachordalby means of the pila antotica, and to the trabecula at the baseof the lamina orbitonasalis by means of the preoptic root.

6. The hind wall of the pituitary fossa is formed in a complexmanner, from a postpituitary commissure between the polar


cartilages, and a pair of inwardly directed processes from theforemost ends of the parachordals forming the dorsum sellae.There is also a precarotid commissure, enclosing the carotidarteries in a foramen between itself and the postpituitarycommissure.

7. The basicranial fenestra has been demonstrated.8. Arguments are given for rejecting Allis's view that the

so-called basicranial fenestrae throughout the craniates are nothomologous.

9. Attention is called to the vacuity in the median wall of theauditory capsule through which the posterior canal bulges, andto the fact that this vacuity is not to be confused with the fora-men endolymphaticum.

10. The relations of the glossopharyngeal nerve are described,and it is shown that its apparent passage through the cavityof the auditory capsule is to be ascribed to the fact that thelamina hypotica of the parachordal acts as a false floor to theauditory capsule, the true floor of which is in this region un-chondrified.

11. The problem of the relations of the jaws to the brain-caseis reviewed in the light of recent investigations, and a reasonedclassification is attempted.

12. It is noticed that chondriflcation is delayed in embryonicmaterial collected from Naples, as compared with material ofsimilar size and degree of development obtained from Plymouth.

L I S T OF LITERATURE CITED.

Allis, E. P. (1923).—"The cranial anatomy of Chlamydoselachus anguineus",'Acta Zool.', 4.

(1928).—"Concerning the pituitary fossa, the myodome and thetrigemino-facialis chamber in recent gnathostome fishes ", ' Journ. Anat.', 63.

Beer, G. R. de (1924).—"Contributions to the study of the development ofthe head in Heterodontus", 'Quart. Journ. Micr. Sci.', 68.

(1925).—"Contributions to the development of the skull in sturgeons",ibid., 69.

(1926).—"Studies on the vertebrate head. II. The orbito-temporalregion of the skull", ibid., 70-

(1927).—"The early development of the chondrocranium of Salmofario", ibid., 71.

(1930A).—'Embryology and Evolution.' Oxford.

642 G. B. DE BEER

Beer, G. R. de (1930B).—"The early development of the chondrocranium ofthe lizard", 'Quart. Journ. Micr. Sci.', 73.

Braus, H. (1906).—"Ueber den embryonalen Kiemenapparat von Heptan-chus", 'Anat. Anz.', 29.

Bugajew, J. (1929).—" tjber das Pharyngomandibular der Knorpelganoiden ",ibid., 67.

(1930).—"t)ber den Bau des Oberkieferapparates bei den Acipen-seroidei und den niederen Haifischen", ibid., 68.

Daniel, J. F. (1922).—'The Elasmobranch Fishes.' Univ. California Press.Dean, B. (1906).—'Chimaeroid fishes.' Carnegie Inst. Washington.Edgeworth, F. H. (1925).—"On the Autostylism of Dipnoi and Amphibia",

'Journ. Anat.', 59.(1926).—'The morphology of the eye-muscles in lower vertebrates.'

Programme British Association, Oxford meeting.(.1928).—"The development of some of the cranial muscles of Ganoid

fishes", 'Phil. Trans. Roy. Soc. B.', 217.Fuchs, H. (1915).—"Uber den Bau und Entwicklung des Schadels der

Chelone imbricata", 'Voeltzkow's Reise', 5.Furbringer, K. (1903).—"Beitrage zur Kenntnis des Visceralskelets der

Selachier", 'Morph. Jahrb.', 31.Furbringer, M. (1900).—"Zur vergleichenden Anatomie des Brustschulter-

apparates", 'Jena Zeit. Naturwiss.', 34.(1904).—"Zur Frage der Abstammung der Saugetiere", 'Jena.

Denkschr.' 11. (Festschr. f. Haeckel.)Gadow, H. (1888).—"On the modification of the first and second Visceral

arches, with especial reference to the homologies of the auditory ossicles",'Phil. Trans. Roy. Soc. B.', 179.

Gaupp, E. (1900).—"Das Chondrocranium von Lacerta agilis", 'Anat.Hefte', 15.

(1906).—"Die Entwicklung des Kopfskelettes", 'Hertwig's Handbuchder Vergl. und Exp. Entwick. der Wirb.', 3.

Gegenbaur, K. (1872).—'Untersuchungen zur vergleichenden Anatomic derWirbelthiere. III. Das Kopfskelet der Selachier.' Leipzig.

Goodrich, E. S. (1909).—" Vertebrata craniata. Cyclostomes and Fishes", in'A Treatise on Zoology'. London.

• (1918).—" On the development of the segments of the head in Scyllium ",'Quart. Journ. Micr. Sci.', 63.

(1930).—'Studies on the structure and development of vertebrates.'London.

Gregory, W. K. (1904).—"The relations of the anterior visceral arches tothe chondrocranium", 'Biol. Bull.', 7.

Hubrecht, A. W. (1877).—"Beitrag zur Kenntnis des Kopfskeletes derHolocephalen", 'Niederl. Arch. f. Zool.', 3.

Huxley, T. H. (1876).—"Contributions to Morphology. I. Ichthyopsida.On Ceratodus forsteri", 'Proc. Zool. Soc'

SKULL OF SCYLLITJM 648

Jaekel, O. (1927).—"Der Kopf der Wirbeltiere", 'Ergeb. der Anat. undEntwick.', 27.

Luther, A. (1909).—"Beitrage zur Kenntnis von Muskulatur und Skelett desKopfes des Haies Stegostoma tigrinum und der Holocephalen ", ' Acta Soc.Scicnt. Fennicae', 27.

. (1914).—"tJber die vom N. trigeminus versorgte Muskulatur derAmphibien", ibid., 44.

Mori, 0. (1924).—"tJber die Entwicklung des Schadelskelettes des DornhaiesAcanthias vulgaris", 'Zeit. f. Anat. und Entwick.', 73.

Norman, J. R. (1926).—"The development of the chondrooranium of theeel", 'Phil. Trans. Roy. Soo. B.', 214.

Norris, H. W. (1929).—"The Parietal fossa and related structures in theplagiostome fishes", ' Journ. Morph. and Physiol.', 48.

Parker, W. K. (1878).—"On the structure and development of the skull insharks and skates", "Trans. Zool. Soo.', 10.

(1882).—"On the development of the skull in Lepidosteus osseus",'Phil. Trans. Roy. Soc.'

Pehrson, T. (1922).—"Some points in the cranial development of Teleosto-mian fishes", 'Acta Zool.', 3.

Regan, C. T. (1906).—"A classification of the Selachian fishes", 'Proc. Zool.Soc'

(1923).—"The skeleton of Lepidosteus", ibid.Retzius, G. (1881).—'Das Gehororgan der Wirbeltiere.' Stockholm.Ridewood, W. G. (1895).—"On the spiracle and associated structures in

elasmobranch fishes", 'Anat. Anz.', 11.(1897).—"Note on the Extrabranchial Cartilages of Elasmobranch

Fishes", ibid., 13.Schauinsland, H. (1903).—"Beitrage zur Entwickelungsgeschichte und

Anatomie der Wirbeltiere", 'Zoologioa'.Sewertzoff, A. N. (1899).—"Die Entwickelung des Selachierschadels",

'Festschr. f. K. v. Kuppfer.'(1923).—"DieMorphologie des Visceralapparates der Elasmobranchier",

'Anat. Anz.', 56.Sewertzoff, A. N., and Dialer (1924).—"Das Pharyngomandibular bei den

Selachiern", ibid., 58.Sewertzoff, A. N. (1926A).—"Der Ursprung der Quadrupeda", 'Palaeont.

Zeitschr.', 7.(1926B).—"Studies on the bony skull of fishes", 'Quart. Journ. Micr.

Sci.', 70.(1927).—"Etudes sur les vertebres inferieurs. Structure de l'appareil

visceral des Elasmobranches", 'Pubb. Staz. Zool. Napoli', 8.(1928).—"The head skeleton and muscles of Acipenser ruthenus",

'Acta Zool.', 9.Smith Woodward, A. (1898).—'Outlines of vertebrate palaeontology.'

Cambridge.

644 G. R. DE BEER

Sorties, F. (1907).—"Ueber die Entwickelung des Chondrocraniums und derknorpeligen Wirbelsaule bei den Vogeln", 'Petrus Camper', 4.

Stannius, H. (1856).—'Handbuch der Anatomie der Wirbeltiere.'Stensi0, E. A. (1921).—'Triassic fishes from Spitzbergen. I.' Vienna.

(1922).—"tJber zwei Coelacanthiden aus dem Oberdevon von Wildun-gcn", 'Palaeont. Zeit.', 4.

(1925).—"On the head of the Macropetalichthyids", 'Field Museum ofNat. Hist. Chicago', 4.

Sushkin, P. P. (1927).—"On the modifications of the mandibular arches andtheir relations to the brain-case in the early Tetrapoda", 'Palaeont.Zeitschr.', 8.

Swinnerton, H. H. (1902).—"A contribution to the morphology of the teleo-stean head skeleton", 'Quart. Journ. Micr. Sci.', 45.

Traquair, R. H. (1887).—"Notes on Chondrosteus acipenseroides", 'Geol.Mag.', 4.

Veit, 0. (1907).—"tlber einige Besonderheiten am Primordialcranium vonLepidosteus oaseus", 'Anat. Hefte', 33.

• (1911).—"Die Entwickelung des Primordialcranium von Lepidosteusosseus", ibid., 44.

Versluys, J. (1912).—"Das Streptostylie-Problem", 'Zool. Jahrb.' suppl.,15,2.

Voit, M. (1909).—"Das Primordialkranium des Kaninchens", 'Anat. Hefte',38.

Watson, D. M. S. (1925).—"The structure of certain Palaeoniscids and therelationship of that group with other bony fish", 'Proc. Zool. Soc'

(1926).—"The evolution and origin of the Amphibia", 'Phil. Trans.Koy. Soc. B.', 214.

(1928).—"On some points in the structure of Palaeoniscid and alliedfish", 'Proc. Zool. Soc'

Wells, G. A. (1917).—"The skull of Acanthias vulgaris", 'Journ. Morph.', 28.White, P. J. (1895).—"The existence of skeletal elements between the

mandibular and hyoid arches in Hexanchus and Laemargus", 'Anat.Anz.', 9.

van Wijhe, J. W. (1902).—"A new method for demonstrating cartilaginousmikroskeletons", 'Proc. Kon. Akad. van Wetensch. te Amsterdam.'

(1904).—"Uber die Entwicklung des Kopfskeletts bei Selachiern",'Comptes Rendus 6e Congr. Internat. Zool. Berne.'

(1922).—"Friihe Entwicklungsstadien des Kopf- und Rumpfskelettsvon Acanthias vulgaris", 'Bijdr. tot de Dierk.', 22.


EXPLANATION OP PLATES 32 TO 37.All figures are of Scyllium canicula.

PLATE 32.

Kg. 1.—Left side view of stage 1 (24 mm., Plymouth).Fig. 2.—Left side view of stage 2 (25 mm.).Fig. 3.—Dorsal view of stage 3 (28 mm., Naples).Fig. 4.—Dorsal view of stage 4 (29 mm., Naples).Fig. 5.—Dorsal view, andFig. 6.—Left side view of stage 5 (29 J mm., Naples).Fig. 7.—Dorsal view, andFig. 8.—Left side view of stage 6 (30 mm., Naples).

PLATE 33.Fig. 9.—Dorsal view, andFig. 10.—Left side view of stage 7 (30 mm., Plymouth).Fig. 11.—Dorsal view, andFig. 12.—Left side view of stage 8 (34 mm., Plymouth).

PLATE 34.Fig. 13.—Dorsal view, andFig. 14.—Left side view of stage 9 (35 mm., Plymouth).Fig. 15.—Dorsal view, andFig. 16.—Left side view of stage 10 (36 mm., Plymouth).

PLATE 35.

Fig. 17.—Ventral view of nasal capsule and jaws of stage 10.Fig. 18.—View from left side and behind of stage 11 (36 mm., Naples).

PLATE 36.Fig. 19.—Dorsal view, andFig. 20.—Left side view of stage 12 (37 mm., Plymouth)Fig. 26.—Anterior face of left posterior portion of auditory capsule of

adult, seen looking back from in front.Fig. 27.—Anterior face of left posterior portion of auditory capsule of adult,

cut surface farther anterior to that of fig. 26, seen looking back from in front.Fig. 28.—Posterior view of left auditory capsule of adult.

PLATE 37.

Fig. 21.—Dorsal view, andFig. 22.—Left side view of stage 13 (45 mm., Naples).Fig. 23.—Ventral view of nasal capsule and visceral-arch skeleton of

stage 13.Fig. 24.—Median view of the left side of stage 13.Fig. 25.—Median view of the left side of adult.

Quart. Journ. Micr. Sci. Vol. 74, N. 8., PI 32

p<j f> Me

G. 11. iC- C. G. ilc B. del.

Quart. Journ. Micr. Sci. Vol. 74, N. S., PI. 33

G. n. •£.• a. a. lit B.

Quart. Journ. Micr. Set. Vol. 74, N. S., PI 34

,-s

G. R. Jb C. G. (le B. del.

Quart. Journ. liter. Sci. Vol. 74, N. S., PI 35

pi r t Me

: !f*l! \

Imm

17

na I ij

18

b'p i p'cj Me

ij. It. <b v. (r. as IS. del.

Quart. Journ. Micr. Sci. Vol. 74, N. S., PI, 36

a. R. cfc C. G. ch B. del.

Lnc tp f/,

/ ' . : fbn •

fri

pi

exdt

fu h

non oc fft fsf ff>r

bdl

R. & C. G. de B. del.

Quart. Journ. Micr. Sci. Vol. 74, JV. S., PI. 37

nco pi ton fh

oa psc Lsc af fpr ! H f/c ;qc L mn

Date post:	03-Jul-2020
Category:	Documents
Upload:	others
View:	1 times
Download:	0 times

The Development of the Skull of Scyllium (Scyliorhinus) canicula L. · 2006-05-23 · The...

Documents